**Part 2**

**Silicon Carbide: Electronic Devices and Applications** 

334 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Yamada, K. & Kamiya, N. (March 1999). High temperature mechanical properties of Si3N4-

1543-1550, ISSN 0955-2219

MoSi2 and Si3N4-SiC composites with network structures of second phases. *Materials Science and Engineering A*, Vol. 261, No. 1-2, pp. 270 - 277, ISSN 0921-5093 Yeomans, J. (2008). Ductile particle ceramic matrix composites--Scientific curiosities or

engineering materials?. *Journal of the European Ceramic Society*, Vol. 28, No. 7, pp.

**14** 

*India* 

**SiC Devices on Different Polytypes:** 

*Radio Physics and Electronics, University of Calcutta, West Bengal,* 

*Centre for Millimeter-Wave Semiconductor Devices and Systems (CMSDS), Institute of* 

Imaging, broadband communication and high-resolution spectroscopic applications in the mid- and far-infrared regions have underscored the importance of developing reliable solidstate sources operating in the frequency range from 0.3 Terahertz to 10.0 Terahertz (1000 to 30 µm wavelength). Recent studies suggest that Terahertz (THz) interactions can enable a variety of new applications on the wide range of solids, liquids, gases, including polymers and biological materials such as proteins and tissues. Compared to microwave and MMwave, far-infrared or THz frequency range has significant reduction in the antenna sizes and greater communication bandwidth. Commercial applications comprise thermal imaging, remote chemical sensing, molecular spectroscopy, medical diagnosis and surveillance. Military applications comprise night vision, rifle sight enhancement, missile tracking, space based surveillance and target recognition. Despite the technical advantages, the major challenge today in THz technology is the development of a portable high-power THz source. During the past few years, significant efforts were devoted to search of reliable

Recently, several solid-state physics research group, the world over, are focusing their research attention in developing semiconductor devices those can generate THz oscillations. A promising concept for THz sources utilizing plasma waves in a gated 2D electron gas (2DEG) was proposed in the early 90-ties. Thereafter, recent experimental observations and theoretical studies have revealed that resonant detection and coherent emission of THz radiation can be effectively induced by excitation of plasma oscillations in the electron channel of Field Effect Transistors (FET). Another promising THz source is the Quantum Cascade Laser (QCL). QCL were first demonstrated in 1994 based on a series of coupled quantum wells constructed using MBE. Although in the mid-infrared region (5< λ < 10 µm) these devices have been in development for more than ten years, it is only recently that the first THz laser has been reported at 4.4 THz. These lasers are made from 1,500 alternating layers of GaAs and AlGaAs and have produced 2.0 mW of peak power (20.0 nW average power). Advances in output power and operating wavelength continue at a rapid pace. Low Temperature Grown (LTG) GaAs photo-mixer can provide up to around 2µW of output power at the frequency of 1.0 THz and their operation frequency can be as high as 5 THz. Of the several available terahertz source technologies, those based on the difference frequency technique are very promising, as they can produce a relatively high power terahertz beam

**1. Introduction** 

semiconductor sources at the THz regime.

**Prospects and Challenges** 

Moumita Mukherjee

### **SiC Devices on Different Polytypes: Prospects and Challenges**

#### Moumita Mukherjee

*Centre for Millimeter-Wave Semiconductor Devices and Systems (CMSDS), Institute of Radio Physics and Electronics, University of Calcutta, West Bengal, India* 

#### **1. Introduction**

Imaging, broadband communication and high-resolution spectroscopic applications in the mid- and far-infrared regions have underscored the importance of developing reliable solidstate sources operating in the frequency range from 0.3 Terahertz to 10.0 Terahertz (1000 to 30 µm wavelength). Recent studies suggest that Terahertz (THz) interactions can enable a variety of new applications on the wide range of solids, liquids, gases, including polymers and biological materials such as proteins and tissues. Compared to microwave and MMwave, far-infrared or THz frequency range has significant reduction in the antenna sizes and greater communication bandwidth. Commercial applications comprise thermal imaging, remote chemical sensing, molecular spectroscopy, medical diagnosis and surveillance. Military applications comprise night vision, rifle sight enhancement, missile tracking, space based surveillance and target recognition. Despite the technical advantages, the major challenge today in THz technology is the development of a portable high-power THz source. During the past few years, significant efforts were devoted to search of reliable semiconductor sources at the THz regime.

Recently, several solid-state physics research group, the world over, are focusing their research attention in developing semiconductor devices those can generate THz oscillations. A promising concept for THz sources utilizing plasma waves in a gated 2D electron gas (2DEG) was proposed in the early 90-ties. Thereafter, recent experimental observations and theoretical studies have revealed that resonant detection and coherent emission of THz radiation can be effectively induced by excitation of plasma oscillations in the electron channel of Field Effect Transistors (FET). Another promising THz source is the Quantum Cascade Laser (QCL). QCL were first demonstrated in 1994 based on a series of coupled quantum wells constructed using MBE. Although in the mid-infrared region (5< λ < 10 µm) these devices have been in development for more than ten years, it is only recently that the first THz laser has been reported at 4.4 THz. These lasers are made from 1,500 alternating layers of GaAs and AlGaAs and have produced 2.0 mW of peak power (20.0 nW average power). Advances in output power and operating wavelength continue at a rapid pace. Low Temperature Grown (LTG) GaAs photo-mixer can provide up to around 2µW of output power at the frequency of 1.0 THz and their operation frequency can be as high as 5 THz. Of the several available terahertz source technologies, those based on the difference frequency technique are very promising, as they can produce a relatively high power terahertz beam

SiC Devices on Different Polytypes: Prospects and Challenges 339

inherent in the build–up of the avalanche current, coupled with the phase delay developed as the carriers traverse the depletion layer. The word IMPATT stands for "impact ionization avalanche transit time". IMPATT diodes employ impact–ionization and transit–time properties of semiconductor structures to produce negative resistance at microwaves and millimeter waves frequencies. The negative resistance arises from two delays which cause the current to lag behind the voltage. One is the 'avalanche delay' caused by finite buildup time of the avalanche current; the other is the 'transit–time delay' caused by the finite time required by the carriers to cross the "drift" region. When these two delays add up to half– cycle time, the diode electronic resistance is negative at the corresponding frequency. IMPATT devices have emerged as most powerful solid-state devices for generation of high CW and pulsed power at millimeter wave frequencies. These devices also provide high oscillator output power with high DC to RF conversion efficiency in Silicon Monolithic

In a practical mm-wave IMPATT oscillator the diode is embedded in a circuit which is resonant at a frequency within the negative–resistance band of the device. The oscillation is initiated by random noise fluctuations, which grows in a negative–resistance medium at the resonant frequency of the circuit. In practice the device has to be mounted either in a coaxial

In 1954, Schokley first studied the microwave negative resistance from the transit time delay of an electron bunch in a forward biased p-n junction diode. Afterwards, in 1958, W. T. Read showed that the finite delay between an applied RF voltage and the external current is due to the generation of carriers in a reverse biased p+-n i n+ diode under avalanche breakdown and the subsequent drift of carriers through the depletion layer. This would lead to a negative resistance of the device at microwave frequencies. In 1965, Johnston et al experimentally observed microwave oscillation from a simple p+nn+ device. At the same time Lee et al also reported oscillation at microwave frequency from a Read diode. Small signal analysis of avalanche diodes of general doping profiles was carried out by T. Misawa, who showed that the negative resistance occurs in any reverse biased p-n junction diode of

In 1968 Lee et al first reported resonant-cap mounted IMPATT oscillator together with a simplified equivalent circuit. Their report showed CW power of 100 mW at 50 GHz with an overall efficiency of 2 percent. In the same year Misawa [9] reported millimeter-wave oscillation from silicon avalanche diode with abrupt junction mounted in a resonant-cap waveguide circuit. His report shows a CW output power in the range of 23 to 150 mW for the 50 to 84 GHz frequency range with a maximum efficiency of 3 percent. In 1970, Misawa and Kenyon reported mechanical tuning characteristics of resonant-cap IMPATT oscillators at millimeter-wave frequencies. Since then , fast advances in semiconductor technology , rapid advances have been made towards further development of various IMPATT diode structures as well as IMPATT based oscillators and amplifiers to meet the to-days high power requirement in higher frequency band. In recent years lot of interest has been created regarding IMPATT diodes and oscillators based on SiC as a semiconducting material. Considerable advancement regarding device physics, device fabrication and the optimum circuit design for IMPATT oscillator and amplifiers has push the frequency range to mm and sub mm regions which has resulted in the emergence of IMPATTs the most powerful solid state devices for generation of microwave and mm

Millimeter Wave Integrated Circuits (SIMMWIC).

arbitrary doping profile.

wave power.

line or in a section of wave-guide or in a micro strip circuit.

over the frequency from 100 GHz to 3.5 THz, which is tunable. Recently THz output power level exceeding 10 mW (occasionally 100 mW) at the frequency of around 1.0 THz has been demonstrated with a special type of electro-optically tunable compact terahertz source.

It is clear from the brief review that the commercially available recent THz sources are complex and bulky. It will be more useful if THz frequency oscillation can be generated from a small sized single solid state source. Among all the two terminal solid state sources, IMPATT diodes have already been established as the most efficient semiconductor sources that can generate highest MM-wave power. Conventional Si and GaAs based IMPATT diodes were found to be reliable but they are limited by power and operating frequencies due to the limitation imposed by their inherent material parameters. WBG semiconductor such as Silicon Carbide (SiC) has received remarkable attention during the last decade as a promising device material for high-temperature, high-frequency and high- power device applications due to its high thermal conductivity, high saturation velocity of charge carriers and high critical field for breakdown. SiC exhibits higher value of thermal conductivity (3-10 times), critical electric field (5-10 times) and saturated carrier velocity (~ 2 times) compared to the conventional semiconductor materials such as Si and GaAs. For a better comparison of the possible high-power, high-frequency performances of these materials, some commonly knew FOMs (Figure of Merit). Taking *Keyes'* and *Johnson's* FOM for Si as unity, the *Keyes'* and *Johnson's* FOM for GaAs are 0.45 and 7.1, respectively, while those for SiC are 5.1 and 278. From the FOMs for high-frequency and high-temperature operation, SiC appears to be superior to both Si and GaAs.

SiC crystallizes in numerous polytypes. The three most common polytypes are the cubic phase, 3C and the hexagonal phases, 4H, and 6H-SiC. The cubic structure, referred to as β-SiC, is expected to have the highest saturation drift velocity. However, the energy bandgap of the 3C phase is significantly smaller than either the 4H or 6H phases, implying a lower breakdown voltage. In addition to this, β-SiC is difficult to grow in a mono-crystalline form due to its meta-stability resulting in a solid-state transformation into an alpha (α)-structure. Due to difficulty in the growth of β-SiC, most of the efforts for producing bulk monocrystalline growth have concentrated on the more easily prepared α-polytypes, referred to as 4H-SiC and 6H-SiC. Thus due to the availability and quality of reproducible single-crystal wafers in these polytypes, 4H- and 6H-SiC-based electronic devices presently exhibit the most promise. The energy band gap of >3.0 eV in hexagonal (4H and 6H) SiC enables the devices based on such materials, to support peak internal electric field (Ec) about ten times higher than Si and GaAs. Higher Ec increases the breakdown voltage, an essential criterion for generation of high output power in a device. Higher Ec also permits incorporation of higher doping level in the depletion layer of the device, which in turn, reduces the width of the active region. Thus the device layers can be made very thin. The transit times of carriers become very small in a thin layered semiconductor if the carrier drift velocities are high. The intrinsic material parameters of hexagonal SiC are thus favorable for the realization of highpower devices.

#### **2. History of IMPATT devices**

A device possesses negative resistance when the A.C. current lags the voltage by a phase angle between 90° and 270°. The negative resistance in an avalanche diode occurs as a result of 180° phase difference between the A.C. current and voltage in a p–n junction reverse– biased to avalanche breakdown. The phase difference is produced by the time delay

over the frequency from 100 GHz to 3.5 THz, which is tunable. Recently THz output power level exceeding 10 mW (occasionally 100 mW) at the frequency of around 1.0 THz has been demonstrated with a special type of electro-optically tunable compact terahertz source. It is clear from the brief review that the commercially available recent THz sources are complex and bulky. It will be more useful if THz frequency oscillation can be generated from a small sized single solid state source. Among all the two terminal solid state sources, IMPATT diodes have already been established as the most efficient semiconductor sources that can generate highest MM-wave power. Conventional Si and GaAs based IMPATT diodes were found to be reliable but they are limited by power and operating frequencies due to the limitation imposed by their inherent material parameters. WBG semiconductor such as Silicon Carbide (SiC) has received remarkable attention during the last decade as a promising device material for high-temperature, high-frequency and high- power device applications due to its high thermal conductivity, high saturation velocity of charge carriers and high critical field for breakdown. SiC exhibits higher value of thermal conductivity (3-10 times), critical electric field (5-10 times) and saturated carrier velocity (~ 2 times) compared to the conventional semiconductor materials such as Si and GaAs. For a better comparison of the possible high-power, high-frequency performances of these materials, some commonly knew FOMs (Figure of Merit). Taking *Keyes'* and *Johnson's* FOM for Si as unity, the *Keyes'* and *Johnson's* FOM for GaAs are 0.45 and 7.1, respectively, while those for SiC are 5.1 and 278. From the FOMs for high-frequency and high-temperature operation, SiC

SiC crystallizes in numerous polytypes. The three most common polytypes are the cubic phase, 3C and the hexagonal phases, 4H, and 6H-SiC. The cubic structure, referred to as β-SiC, is expected to have the highest saturation drift velocity. However, the energy bandgap of the 3C phase is significantly smaller than either the 4H or 6H phases, implying a lower breakdown voltage. In addition to this, β-SiC is difficult to grow in a mono-crystalline form due to its meta-stability resulting in a solid-state transformation into an alpha (α)-structure. Due to difficulty in the growth of β-SiC, most of the efforts for producing bulk monocrystalline growth have concentrated on the more easily prepared α-polytypes, referred to as 4H-SiC and 6H-SiC. Thus due to the availability and quality of reproducible single-crystal wafers in these polytypes, 4H- and 6H-SiC-based electronic devices presently exhibit the most promise. The energy band gap of >3.0 eV in hexagonal (4H and 6H) SiC enables the devices based on such materials, to support peak internal electric field (Ec) about ten times higher than Si and GaAs. Higher Ec increases the breakdown voltage, an essential criterion for generation of high output power in a device. Higher Ec also permits incorporation of higher doping level in the depletion layer of the device, which in turn, reduces the width of the active region. Thus the device layers can be made very thin. The transit times of carriers become very small in a thin layered semiconductor if the carrier drift velocities are high. The intrinsic material parameters of hexagonal SiC are thus favorable for the realization of high-

A device possesses negative resistance when the A.C. current lags the voltage by a phase angle between 90° and 270°. The negative resistance in an avalanche diode occurs as a result of 180° phase difference between the A.C. current and voltage in a p–n junction reverse– biased to avalanche breakdown. The phase difference is produced by the time delay

appears to be superior to both Si and GaAs.

power devices.

**2. History of IMPATT devices** 

inherent in the build–up of the avalanche current, coupled with the phase delay developed as the carriers traverse the depletion layer. The word IMPATT stands for "impact ionization avalanche transit time". IMPATT diodes employ impact–ionization and transit–time properties of semiconductor structures to produce negative resistance at microwaves and millimeter waves frequencies. The negative resistance arises from two delays which cause the current to lag behind the voltage. One is the 'avalanche delay' caused by finite buildup time of the avalanche current; the other is the 'transit–time delay' caused by the finite time required by the carriers to cross the "drift" region. When these two delays add up to half– cycle time, the diode electronic resistance is negative at the corresponding frequency. IMPATT devices have emerged as most powerful solid-state devices for generation of high CW and pulsed power at millimeter wave frequencies. These devices also provide high oscillator output power with high DC to RF conversion efficiency in Silicon Monolithic Millimeter Wave Integrated Circuits (SIMMWIC).

In a practical mm-wave IMPATT oscillator the diode is embedded in a circuit which is resonant at a frequency within the negative–resistance band of the device. The oscillation is initiated by random noise fluctuations, which grows in a negative–resistance medium at the resonant frequency of the circuit. In practice the device has to be mounted either in a coaxial line or in a section of wave-guide or in a micro strip circuit.

In 1954, Schokley first studied the microwave negative resistance from the transit time delay of an electron bunch in a forward biased p-n junction diode. Afterwards, in 1958, W. T. Read showed that the finite delay between an applied RF voltage and the external current is due to the generation of carriers in a reverse biased p+-n i n+ diode under avalanche breakdown and the subsequent drift of carriers through the depletion layer. This would lead to a negative resistance of the device at microwave frequencies. In 1965, Johnston et al experimentally observed microwave oscillation from a simple p+nn+ device. At the same time Lee et al also reported oscillation at microwave frequency from a Read diode. Small signal analysis of avalanche diodes of general doping profiles was carried out by T. Misawa, who showed that the negative resistance occurs in any reverse biased p-n junction diode of arbitrary doping profile.

In 1968 Lee et al first reported resonant-cap mounted IMPATT oscillator together with a simplified equivalent circuit. Their report showed CW power of 100 mW at 50 GHz with an overall efficiency of 2 percent. In the same year Misawa [9] reported millimeter-wave oscillation from silicon avalanche diode with abrupt junction mounted in a resonant-cap waveguide circuit. His report shows a CW output power in the range of 23 to 150 mW for the 50 to 84 GHz frequency range with a maximum efficiency of 3 percent. In 1970, Misawa and Kenyon reported mechanical tuning characteristics of resonant-cap IMPATT oscillators at millimeter-wave frequencies. Since then , fast advances in semiconductor technology , rapid advances have been made towards further development of various IMPATT diode structures as well as IMPATT based oscillators and amplifiers to meet the to-days high power requirement in higher frequency band. In recent years lot of interest has been created regarding IMPATT diodes and oscillators based on SiC as a semiconducting material. Considerable advancement regarding device physics, device fabrication and the optimum circuit design for IMPATT oscillator and amplifiers has push the frequency range to mm and sub mm regions which has resulted in the emergence of IMPATTs the most powerful solid state devices for generation of microwave and mm wave power.

SiC Devices on Different Polytypes: Prospects and Challenges 341

two face centered cubic sub lattices of the crystal vibrate in the opposite directions. Excitation of the optical phonons are possible when the electrons gain a minimum energy

frequency for optical mode of vibration and h is the Planck's constant. The values of *op*

GaAs and Silicon are obtained from neutron scattering experiments [18-21] and are of the order of 0.035 eV and 0.063 eV respectively. At a high field ( >107 V.m-1 ), the average

energy to the lattice via optical phonon is created and reaches a scattering limited

0.5 8 tanh( 2 ) [ ] 3 \* *op op*

 ε

*m*

[(1 )(1 )] *s op i i i*

Where lop and li is the mean free path for optical phonon collision and ionizing collision

collision , q is the charge of electron and E is the electric field. This theoretical investigation shows that the drift velocity for electrons in Si passes through a maximum before attaining saturation. Danda and Nicolet give an expression which fits well with the experimental results for Si for field dependence of carrier drift velocity in the following

( ) [1 exp( )] *<sup>o</sup> <sup>s</sup>*

Where v(E) is the carrier drift velocity at field E. The values of low field mobility (μo) of

At very high electric field (> 107 V.m-1) electrons (minority carriers) gain energy at a faster rate than they can lose through the emission of optical phonon in a reverse biased p-n junction. As

*<sup>E</sup> vE v*

*s*

*v* μ

 <sup>=</sup> + +

*l l qE* ε

ε

π

Drift velocity of carriers in Si at different field has been accurately determined by a number of workers using the time of flight technique and the space charge resistance technique [23- 26]. The time of flight technique provides direct measurement of the drift velocity of both majority and minority carriers accurately. In 1967, Duh and Moll measured the carrier drift velocity in Si at high electric field ( > 107 V.m-1 ) and it shows that a slow increase of vd in the field range ( 2.6 - 4.35 )x107 V.m-1. At a high field ( 2x107 V.m-1 ) impact ionization becomes an important scattering mechanism in addition to optical phonon scattering. At such high electric fields the energy gained by the electrons from the electric field is lost mostly in ionizing collisions that results an electron-hole (e-h) pair. According to Roy

*KT*

0.5

2 *op <sup>o</sup> h*

π<sup>=</sup> , where

 ω

ε

<sup>=</sup> (4)

ω

and thereby a transfer of

(5)

is the threshold energy for ionizing

=−− (6)

*<sup>o</sup>* is the angular

εfor

ε

equal to optical phonon energy or Ramam energy ( )

carrier kinetic energy exceeds the optical phonon energy *op*

average drift velocity independent of the electric field and is given by,

ε

*d*

and Ghosh, at the ionizing fields the drift velocity v (E) is expressed by

( )

carriers can be obtained from the slopes of v-E curves at low field.

respectively, vs is the saturated drift velocity, *<sup>i</sup>*

form:

**2.2 Impact ionization** 

*<sup>v</sup> v E <sup>l</sup>*

*v*

To understand the operation and performance of IMPATT devices and oscillators knowledge of the basic IMPATT phenomena is required and briefly discussed in the following section. To build an IMPATT oscillator the device has to be mounted in a suitable microwave/mm-wave circuits. The performance of the oscillator strongly depends not only on the device but also on the circuit in which the device is embedded .The various microwave/mm-wave circuits that are being widely used to construct IMPATT oscillators have been reviewed briefly.

A brief review of the fundamental physical processes involved in IMPATT action followed by a review of the various IMPATT structures and oscillators will be presented in this section. The factors, which determine the avalanche delay and the transit time delay for high frequency operation of IMPATT will also be briefly discussed.

#### **2.1 High field properties of charge carriers in IMPATT devices**

The different scattering interactions between the charge carriers and the lattice lead to the emission of both acoustic and optical phonons which give rise to the saturation of carrier's drift velocity in semiconductors which is one of the fundamental physical phenomenon involved in IMPATT action.

Drift velocity of charge carriers has been observed to be linear upto the electric field 104 V.m-1 and it reaches a scattering limited value independent of the electric field when the field is very high (>106 V.m-1 ). At low values of electric field (E), which the principal scattering phenomenon is acoustic phonon, the drift velocity (vd ) of charge carriers in semiconductor varies as :

$$
\omega\_d = \mu\_o E \tag{1}
$$

Where μ*<sup>o</sup>* is the low field mobility and can be expressed as,

$$
\mu\_o = \frac{q < v^2 \tau\_o >}{m^\* < v^2 >} \tag{2}
$$

Where q is the charge of the electron, m\* is the effective mass of the carrier, *<sup>o</sup>* τ is the relaxation time and v is the carrier velocity. The brackets in the above expression indicate Maxwellian average.

In the low field region (Ohmic region), the rate of energy through acoustic phonon collision is small and the scattering is isotropic. Assuming energy distribution to be Maxwellian at lattice temperature T and a constant mean free path, the low field mobility is given by,

$$
\mu\_o = \frac{4ql\_a}{\Re(2\pi m^\*KT)^{0.5}}\tag{3}
$$

Where la is the mean free path for acoustic phonon collision and K is the Boltzman constant. At high electric field ( >106 V.m-1 ) high-energy electrons ( hot electrons ) interact more strongly with lattice and there is a departure from the linear dependence of drift velocity with the electric field. The thermal equilibrium is lost because the rate of energy gained from the field is more that the amount lost to the crystal lattice through low energy acoustic phonon collision. At this high field, emission of optical phonons is a dominant phenomenon, which are quanta of high frequency thermal vibrations of the lattice in which

To understand the operation and performance of IMPATT devices and oscillators knowledge of the basic IMPATT phenomena is required and briefly discussed in the following section. To build an IMPATT oscillator the device has to be mounted in a suitable microwave/mm-wave circuits. The performance of the oscillator strongly depends not only on the device but also on the circuit in which the device is embedded .The various microwave/mm-wave circuits that are being widely used to construct IMPATT oscillators

A brief review of the fundamental physical processes involved in IMPATT action followed by a review of the various IMPATT structures and oscillators will be presented in this section. The factors, which determine the avalanche delay and the transit time delay for high

The different scattering interactions between the charge carriers and the lattice lead to the emission of both acoustic and optical phonons which give rise to the saturation of carrier's drift velocity in semiconductors which is one of the fundamental physical phenomenon

Drift velocity of charge carriers has been observed to be linear upto the electric field 104 V.m-1 and it reaches a scattering limited value independent of the electric field when the field is very high (>106 V.m-1 ). At low values of electric field (E), which the principal scattering phenomenon is acoustic phonon, the drift velocity (vd ) of charge carriers in

> *d o v E* = μ

2 <sup>2</sup> \* *<sup>o</sup> <sup>o</sup> q v m v* τ

< > <sup>=</sup> < >

relaxation time and v is the carrier velocity. The brackets in the above expression indicate

In the low field region (Ohmic region), the rate of energy through acoustic phonon collision is small and the scattering is isotropic. Assuming energy distribution to be Maxwellian at lattice temperature T and a constant mean free path, the low field mobility is given by,

> 4 3(2 \* ) *<sup>a</sup> <sup>o</sup>*

π

Where la is the mean free path for acoustic phonon collision and K is the Boltzman constant. At high electric field ( >106 V.m-1 ) high-energy electrons ( hot electrons ) interact more strongly with lattice and there is a departure from the linear dependence of drift velocity with the electric field. The thermal equilibrium is lost because the rate of energy gained from the field is more that the amount lost to the crystal lattice through low energy acoustic phonon collision. At this high field, emission of optical phonons is a dominant phenomenon, which are quanta of high frequency thermal vibrations of the lattice in which

*ql m KT*

0.5

<sup>=</sup> (3)

(1)

(2)

τ

is the

frequency operation of IMPATT will also be briefly discussed.

**2.1 High field properties of charge carriers in IMPATT devices** 

*<sup>o</sup>* is the low field mobility and can be expressed as,

μ

μ

Where q is the charge of the electron, m\* is the effective mass of the carrier, *<sup>o</sup>*

have been reviewed briefly.

involved in IMPATT action.

semiconductor varies as :

Where

μ

Maxwellian average.

two face centered cubic sub lattices of the crystal vibrate in the opposite directions. Excitation of the optical phonons are possible when the electrons gain a minimum energy equal to optical phonon energy or Ramam energy ( ) 2 *op <sup>o</sup> h* ε ω π <sup>=</sup> , where ω*<sup>o</sup>* is the angular frequency for optical mode of vibration and h is the Planck's constant. The values of *op* ε for GaAs and Silicon are obtained from neutron scattering experiments [18-21] and are of the order of 0.035 eV and 0.063 eV respectively. At a high field ( >107 V.m-1 ), the average carrier kinetic energy exceeds the optical phonon energy *op* ε and thereby a transfer of energy to the lattice via optical phonon is created and reaches a scattering limited average drift velocity independent of the electric field and is given by,

$$w\_d = \left[\frac{8\varepsilon\_{op}\tanh(\varepsilon\_{op}2KT)}{3\pi m^\*}\right]^{0.5} \tag{4}$$

Drift velocity of carriers in Si at different field has been accurately determined by a number of workers using the time of flight technique and the space charge resistance technique [23- 26]. The time of flight technique provides direct measurement of the drift velocity of both majority and minority carriers accurately. In 1967, Duh and Moll measured the carrier drift velocity in Si at high electric field ( > 107 V.m-1 ) and it shows that a slow increase of vd in the field range ( 2.6 - 4.35 )x107 V.m-1. At a high field ( 2x107 V.m-1 ) impact ionization becomes an important scattering mechanism in addition to optical phonon scattering. At such high electric fields the energy gained by the electrons from the electric field is lost mostly in ionizing collisions that results an electron-hole (e-h) pair. According to Roy and Ghosh, at the ionizing fields the drift velocity v (E) is expressed by

$$w(E) = \frac{\upsilon\_s}{\left[ (1 + \frac{l\_{op}}{l\_i})(1 + \frac{\mathcal{E}\_i}{l\_i q E}) \right]^{0.5}} \tag{5}$$

Where lop and li is the mean free path for optical phonon collision and ionizing collision respectively, vs is the saturated drift velocity, *<sup>i</sup>* ε is the threshold energy for ionizing collision , q is the charge of electron and E is the electric field. This theoretical investigation shows that the drift velocity for electrons in Si passes through a maximum before attaining saturation. Danda and Nicolet give an expression which fits well with the experimental results for Si for field dependence of carrier drift velocity in the following form:

$$v(E) = v\_s[1 - \exp(-\frac{\mu\_o E}{v\_s})] \tag{6}$$

Where v(E) is the carrier drift velocity at field E. The values of low field mobility (μo) of carriers can be obtained from the slopes of v-E curves at low field.

#### **2.2 Impact ionization**

At very high electric field (> 107 V.m-1) electrons (minority carriers) gain energy at a faster rate than they can lose through the emission of optical phonon in a reverse biased p-n junction. As

SiC Devices on Different Polytypes: Prospects and Challenges 343

Where Js is the initial reverse saturation current and M is called the multiplication factor. The current multiplication factors Mn and Mp for electrons and holes are given by J/Jns, and J/Jps respectively. A small amount of reverse saturation current (Jns, Jps) multiplied by very high multiplication factor grows to a very high current and this phenomenon is known as

*a x*

*o N E dx* = = α

Considering the carrier multiplication process initiated by both electrons and holes at the two edges of the depletion layer and unequal ionization rates of charge carriers, Lee et al derived the generalised breakdown condition of the p-n junction. In Fig. 1(a), Jps and Jns are the saturation currents for holes and electrons entering the depletion layer of a reverse biased p-n junction at x = -x1 and x = x2 respectively. The increase of electron and hole

*n p nn pp*

*<sup>n</sup> n n p p <sup>J</sup> J J*

 α

*n n p p <sup>J</sup> J J*

 α  αδ

 αδ

α

α

Since the diffusion current is very small compared to the drift current, then the hole drift current *p p J* = *qpv* and the electron drift current *n n J qnv* = , where *vp* and *vn* are the saturated drift velocities for holes and electrons, n and p are the carrier density for electrons and holes.

*<sup>J</sup> J J*

Using the boundary conditions ( 0) *n ns Jx J* = = and ( ) *<sup>n</sup> ps Jx W J J* = =− and using the

*dx* the above equation reduces to

*W W x*

*o o o*

/ 1 exp{ ( ) } [1 exp{ ( ) } ] *W W x*

*o o o* −+ − − = − − − *k k*

*s ps ps n p n np*

*JJ J* −+ − − =− − − αα

 α

/ exp{ ( ) } [1 exp{ ( ) } ]

α

*n p n np*

 α

*dx J dx dx*

αα

() 1

(10)

δ

= + (12)

=− − (13)

=− − = (14)

 αα

 αα

*dx dx* (15)

*J J Jx Jx* =− = + (11)

*x* may be written as,

avalanche breakdown. At breakdown M and J tends to infinity and then,

current at x is equal to the charges generated per second in distance

 δ

Therefore the continuity equations for electrons and holes can be written as,

*x* ∂

*p*

*x* ∂

∂

Thus the total drift current density ( ) *<sup>n</sup> <sup>p</sup> JJJ* = + is independent of x.

*x* ∂

∂

( ) *<sup>n</sup> <sup>n</sup> <sup>p</sup> n n*

*n p*

α α*dx M*

Eliminating *pJ* from equation (11) one obtains,

*x*

*o* − − α α

integrating factor exp{ ( ) }

∂

δ

a result, it collides with bound electron in the valence band and excites them into the conduction band, creating an e-h pair and the phenomenon is termed as **impact ionization**. Important parameters for impact ionization are the **ionization threshold energy** Et (i.e. minimum energy required to cause an ionizing collision) and the **ionization rate** α (i.e. average number of ionizing collisions by the carrier in traversing unit distance in the direction of electric field). From energy conservation principle, *Et* should be equal to the band gap energy *(Eg)*. The values of *Eg* for Si and GaAs at room temperature are 1.10eV and 1.43eV respectively. If both energy and momentum conservation are taken into account, the threshold energy *Et* should be equal to 1.5*Eg* for parabolic band structure having the same effective masses for the carrier. If the electron energy exceeds *Et*, emission of optical phonons or ionizing collision may produce an e-h pair. The probability of either types of collision depends on the mean free path for optical collision (*lop*) and on the mean free path for ionizing collision (*li*). The relative probability of second collision being an ionizing collision is *lop li.* The ionization rate (α ) is a function of *lop, li, Et* and *Eg*. In 1954, Wolff first assumed that the probability of ionizing collision is much greater than the optical phonon collision and is

valid at high electric field. However, Shockley derive an expression at low field, such that electrons acquire ionization threshold energy Et and then produces an ionizing collision in the first attempt without suffering a single optical phonon collision which is given by,

$$\alpha = \frac{qE}{r\mathcal{E}\_{op}} \exp(-\frac{E\_t}{qEl\_r}) \tag{7}$$

The most important theoretical study of field-dependence of ionization rate was carried out by G. A. Baraff, by solving Boltzman transport collision equation in terms of a space and energy dependent collision density, considering the acoustic phonon, optical phonon and ionizing collision. The values of ionization rate 'α' can be obtained from universal Baraffs plot for any semiconductor for which the parameters *lop*, E t and ε*op* are known.

#### **2.3 Avalanche breakdown**

Under typical doping profile and reverse bias condition of a p-n junction diode, the total voltage drop occurs across a very thin space charge depletion region. Thermally generated electrons and holes (minority carriers) in the p and n regions diffusing towards the n and p edges of the depletion layer results a small reverse saturation current in the reverse bias condition. A single minority carrier experiences there a very high electric field and creates eh pair by impact ionization. These generated electron and hole produces additional e-h pair as they further drift toward n and p sides.

If a single electron yield N number of e-h pairs while drifting across the avalanche region of length xa then,

$$N = \int\_{\sigma}^{\chi} \alpha(E) d\chi \tag{8}$$

Equal ionization rates for electrons and holes generate N2 number of e-h pairs and the process continues and this is known as avalanche multiplication. Therefore, the total current after avalanche multiplication becomes,

$$J = J\_s(1 + N + N^2 + \dots \dots \dots) = \frac{J\_s}{1 - N} = M I\_s \tag{9}$$

a result, it collides with bound electron in the valence band and excites them into the conduction band, creating an e-h pair and the phenomenon is termed as **impact ionization**. Important parameters for impact ionization are the **ionization threshold energy** Et (i.e.

average number of ionizing collisions by the carrier in traversing unit distance in the direction of electric field). From energy conservation principle, *Et* should be equal to the band gap energy *(Eg)*. The values of *Eg* for Si and GaAs at room temperature are 1.10eV and 1.43eV respectively. If both energy and momentum conservation are taken into account, the threshold energy *Et* should be equal to 1.5*Eg* for parabolic band structure having the same effective masses for the carrier. If the electron energy exceeds *Et*, emission of optical phonons or ionizing collision may produce an e-h pair. The probability of either types of collision depends on the mean free path for optical collision (*lop*) and on the mean free path for ionizing collision (*li*). The relative probability of second collision being an ionizing collision is *lop li.*

probability of ionizing collision is much greater than the optical phonon collision and is valid at high electric field. However, Shockley derive an expression at low field, such that electrons acquire ionization threshold energy Et and then produces an ionizing collision in the first attempt without suffering a single optical phonon collision which is given by,

The most important theoretical study of field-dependence of ionization rate was carried out by G. A. Baraff, by solving Boltzman transport collision equation in terms of a space and energy dependent collision density, considering the acoustic phonon, optical phonon and ionizing collision. The values of ionization rate 'α' can be obtained from universal Baraffs

Under typical doping profile and reverse bias condition of a p-n junction diode, the total voltage drop occurs across a very thin space charge depletion region. Thermally generated electrons and holes (minority carriers) in the p and n regions diffusing towards the n and p edges of the depletion layer results a small reverse saturation current in the reverse bias condition. A single minority carrier experiences there a very high electric field and creates eh pair by impact ionization. These generated electron and hole produces additional e-h pair

If a single electron yield N number of e-h pairs while drifting across the avalanche region of

*a x*

*o N E dx* = α

<sup>2</sup> (1 ..........) <sup>1</sup> *<sup>s</sup> s s <sup>J</sup> JJ NN MJ <sup>N</sup>*

Equal ionization rates for electrons and holes generate N2 number of e-h pairs and the process continues and this is known as avalanche multiplication. Therefore, the total

( )

α

plot for any semiconductor for which the parameters *lop*, E t and

ε

exp( ) *<sup>t</sup> op r qE E r qEl*

) is a function of *lop, li, Et* and *Eg*. In 1954, Wolff first assumed that the

= − (7)

ε

*op* are known.

(8)

= ++ + = = − (9)

α(i.e.

minimum energy required to cause an ionizing collision) and the **ionization rate**

The ionization rate (

**2.3 Avalanche breakdown** 

length xa then,

as they further drift toward n and p sides.

current after avalanche multiplication becomes,

α

Where Js is the initial reverse saturation current and M is called the multiplication factor. The current multiplication factors Mn and Mp for electrons and holes are given by J/Jns, and J/Jps respectively. A small amount of reverse saturation current (Jns, Jps) multiplied by very high multiplication factor grows to a very high current and this phenomenon is known as avalanche breakdown. At breakdown M and J tends to infinity and then,

$$N = \int\_{\rho}^{\chi\_g} \alpha(E) dx = 1 \tag{10}$$

Considering the carrier multiplication process initiated by both electrons and holes at the two edges of the depletion layer and unequal ionization rates of charge carriers, Lee et al derived the generalised breakdown condition of the p-n junction. In Fig. 1(a), Jps and Jns are the saturation currents for holes and electrons entering the depletion layer of a reverse biased p-n junction at x = -x1 and x = x2 respectively. The increase of electron and hole current at x is equal to the charges generated per second in distance δ*x* may be written as,

$$
\delta \delta \mathbf{J}\_n = -\delta \mathbf{J}\_p = \alpha\_n \mathbf{J}\_n \delta \mathbf{x} + \alpha\_p \mathbf{J}\_p \delta \mathbf{x} \tag{11}
$$

Therefore the continuity equations for electrons and holes can be written as,

$$\frac{\partial \mathbf{J}\_n}{\partial \mathbf{x}} = \alpha\_n \mathbf{J}\_n + \alpha\_p \mathbf{J}\_p \tag{12}$$

$$\frac{\partial \mathbf{J}\_p}{\partial \mathbf{x}} = -\alpha\_n \mathbf{J}\_n - \alpha\_p \mathbf{J}\_p \tag{13}$$

Since the diffusion current is very small compared to the drift current, then the hole drift current *p p J* = *qpv* and the electron drift current *n n J qnv* = , where *vp* and *vn* are the saturated drift velocities for holes and electrons, n and p are the carrier density for electrons and holes. Thus the total drift current density ( ) *<sup>n</sup> <sup>p</sup> JJJ* = + is independent of x.

Eliminating *pJ* from equation (11) one obtains,

*o*

$$\frac{\partial \mathbf{J}\_n}{\partial \mathbf{x}} = -(\alpha\_n - \alpha\_p)\mathbf{J}\_n = \alpha\_n \mathbf{J} \tag{14}$$

Using the boundary conditions ( 0) *n ns Jx J* = = and ( ) *<sup>n</sup> ps Jx W J J* = =− and using the integrating factor exp{ ( ) } *x n p* − − α α*dx* the above equation reduces to

$$I\_s - I\_{ps} + I\_{ps} \exp\{-\int\_{\alpha}^{W} (\alpha\_n - \alpha\_p) d\mathbf{x}\} = I[1 - \int\_{\alpha}^{W} \alpha\_n \exp\{-\int\_{\alpha}^{\cdot} (\alpha\_n - \alpha\_p) d\mathbf{x}\prime\prime\} d\mathbf{x}]$$

$$1 - k + k \exp\{-\int\_{\alpha}^{W} (\alpha\_n - \alpha\_p) d\mathbf{x}\prime\} = M[1 - \int\_{\alpha}^{W} \alpha\_n \exp\{-\int\_{\alpha}^{\cdot} (\alpha\_n - \alpha\_p) d\mathbf{x}\prime\prime\prime\} d\mathbf{x}\prime\prime] \tag{15}$$

SiC Devices on Different Polytypes: Prospects and Challenges 345

/ [1 exp{ ( ) }]/[1 exp{ ( ) } ] *W W x*

*o o o*

*n np*

 αα

1 1 / [1 exp{ ( ) } ] *W x*

In a similar way using the boundary condition, ( ) *<sup>p</sup> ns Jx o J J* = =− and ( ) *p ps Jx W J* = = we

exp{ ( ) } <sup>1</sup> [1 exp{ ( ) } ]

*o o*

 αα

α

*n np*

*<sup>k</sup> dx dx*

 αα*dx dx dx*

*ps*

 αα*dx dx dx*

− − − = (18)

*dx dx*

− − <sup>−</sup> (17)

 αα

*<sup>J</sup> <sup>M</sup> J* =

*n p W x <sup>o</sup> p np*

1 1 / <sup>1</sup> exp{ ( ) } *W x*

<sup>−</sup> −= − − +

*o o*

/ exp{ ( ) }. exp{ ( ) } *W Wx*

 α

When avalanche breakdown occurs i.e. M tends to infinity for a mixture of electron and hole

<sup>1</sup> / exp{ ( ) } *W x*

*n np*

/ exp{ ( ) }. exp{ ( ) } 1

*np p np*

In case of pure electron or hole injection i.e. in which multiplication is initiated purely by electrons ( 0, 0 *ps k J* = = ) or purely by holes ( 0, 0 *ns k J* = = ), the breakdown condition

 αα

<sup>−</sup> −− +

*np p np*

*dx*

Multiplying equation (15 ) by (1-k) and equation (16) by k and adding one obtain,

α

*o oo*

*<sup>J</sup>* <sup>=</sup> and *<sup>p</sup>*

*o o*

 α

*W Wx*

*o oo*

ξ

αα

*<sup>J</sup> <sup>M</sup>*

*k* α

ξ

αα

*ns*

= −−

*o o*

*n p n np*

 α

*dx dx dx*

 αα

/

*dx dx*

(16)

*dx dx*

=− − − (2.4.8)

Where M is the multiplication factor and J *<sup>s</sup>* is the total reverse saturation current. Thus

*M* = −+ − − − − − *k k* αα

α

Where , *ps*

ξ

get,

*s <sup>J</sup> <sup>k</sup>*

*<sup>J</sup>* <sup>=</sup> ,

Where 1 exp{ ( ) }

*M*

*k*

ξ

*k*

ξ

Now, *M* may be written as,

injection, one obtain,

reduces to

2 (1 ) *M <sup>n</sup> <sup>p</sup>* =− + *k M kM* where *<sup>n</sup>*

=−+ − − *k k*

*s <sup>J</sup> <sup>M</sup>*

*M*

 α

*W*

*o*

ξ

*n p*

*W*

*M*

α α

− −

ξ

α*dx*

*<sup>J</sup>* <sup>=</sup> and *n ns ps JJ J* = +

Fig. 1. (a) Avalanche multiplication; (b) carrier current profile and (c) Electric field profile in the depletion region of a reverse biased p-n junction.

Where , *ps s <sup>J</sup> <sup>k</sup> <sup>J</sup>* <sup>=</sup> , *s <sup>J</sup> <sup>M</sup> <sup>J</sup>* <sup>=</sup> and *n ns ps JJ J* = +

344 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Fig. 1. (a) Avalanche multiplication; (b) carrier current profile and (c) Electric field profile in

the depletion region of a reverse biased p-n junction.

Where M is the multiplication factor and J *<sup>s</sup>* is the total reverse saturation current. Thus

$$M = \left[1 - k + k \exp\{-\int\_{0}^{V} (\alpha\_{n} - \alpha\_{p}) dx\}\right] / \left[1 - \int\_{0}^{V} \alpha\_{n} \exp\{-\int\_{0}^{x} (\alpha\_{n} - \alpha\_{p}) dx'\} dx\right]$$

$$\frac{1}{M} = \frac{1}{\xi} \left[1 - \int\_{0}^{V} \alpha\_{n} \exp\{-\int\_{0}^{x} (\alpha\_{n} - \alpha\_{p}) dx'\} dx\right] \tag{2.4.8}$$

Where 1 exp{ ( ) } *W n p o* ξ α =−+ − − *k k* α*dx*

In a similar way using the boundary condition, ( ) *<sup>p</sup> ns Jx o J J* = =− and ( ) *p ps Jx W J* = = we get,

$$\frac{1}{M} = \frac{\exp\{-\int\_{\cdot}^{V} (\alpha\_n - \alpha\_p) d\mathbf{x}\}}{\xi} [1 - \int\_{\cdot}^{V} \alpha\_p \exp\{\int\_{\cdot}^{\cdot} (\alpha\_n - \alpha\_p) d\mathbf{x}\} \, d\mathbf{x}] \tag{16}$$

Multiplying equation (15 ) by (1-k) and equation (16) by k and adding one obtain,

$$1 - \frac{1}{M} = \frac{1 - k}{\xi} \int\_{\boldsymbol{\varrho}}^{V} \alpha\_n \exp\{-\int\_{\boldsymbol{\varrho}}^{\boldsymbol{x}} (\alpha\_n - \alpha\_p) d\mathbf{x}'\} d\mathbf{x} + $$
 
$$ \frac{k}{\xi} \exp\{-\int\_{\boldsymbol{\varrho}}^{V} (\alpha\_n - \alpha\_p) d\mathbf{x}\} \int\_{\boldsymbol{\varrho}}^{V} \alpha\_p \exp\{\int\_{\boldsymbol{\varrho}}^{\boldsymbol{x}} (\alpha\_n - \alpha\_p) d\mathbf{x}'\} d\mathbf{x} \tag{17} $$

Now, *M* may be written as,

2 (1 ) *M <sup>n</sup> <sup>p</sup>* =− + *k M kM* where *<sup>n</sup> ns <sup>J</sup> <sup>M</sup> <sup>J</sup>* <sup>=</sup> and *<sup>p</sup> ps <sup>J</sup> <sup>M</sup> J* =

When avalanche breakdown occurs i.e. M tends to infinity for a mixture of electron and hole injection, one obtain,

$$
\frac{1-k}{\xi} \int\_{o}^{W} \alpha\_{n} \exp\{-\int\_{o}^{x} (\alpha\_{n} - \alpha\_{p}) d\mathbf{x}'\} d\mathbf{x} + 
$$

$$
\frac{k}{\xi} \exp\{\int\_{o}^{W} (\alpha\_{n} - \alpha\_{p}) d\mathbf{x}\} \int\_{o}^{W} \alpha\_{p} \exp\{-\int\_{o}^{x} (\alpha\_{n} - \alpha\_{p}) d\mathbf{x}'\} d\mathbf{x} = 1 \tag{18}
$$

In case of pure electron or hole injection i.e. in which multiplication is initiated purely by electrons ( 0, 0 *ps k J* = = ) or purely by holes ( 0, 0 *ns k J* = = ), the breakdown condition reduces to

SiC Devices on Different Polytypes: Prospects and Challenges 347

intrinsic material .The device consists essentially of two regions ; One is the narrow p-region at which avalanche multiplication occurs .This region is also called the **high field region or the avalanche region** .The other is the i (or ν ) region through which the generated holes must drift while moving towards the p+ contact .This region is also called the **intrinsic region or the drift region** .When the reverse biased voltage is well above the punch through or breakdown voltage, the space between the n+p junction and the i-p+ junction happens to be the space-charge region . Carriers (holes) moving in the high field near the n+-p junction acquire energy to knock valance electrons into the conduction band, thus producing holeelectron pairs. The rate of pair production, or avalanche multiplication, is a sensitive nonlinear function of the field. By proper doping, the field can be given a relatively sharp peak so that avalanche multiplication is confined to a very narrow region at the n+-p junction. The electrons move in to the n+- region and the holes drift through the space charge region to the p+- region with a constant velocity *v*d (called the saturated drift velocity).The transit time of a hole across the drift region (i.e. i-region of length L) is given

The phenomenon of negative resistance in Read diode can be understood with reference to Fig. 3. In actual practice, to form oscillator, the diode is mounted in a microwave/mm-wave resonant circuit. An a.c. voltage can be maintained at a given frequency in the circuit thus the total voltage across the diode is the sum of the d.c. and a.c. voltages, mathematically : VT (t)=V DC + V D sin ωt , and the form of this total diode voltage is shown in Fig. 3(a). This total voltage causes breakdown at the n+ - p junction during the positive half cycle of the a.c. voltage when VT is above the breakdown value, and the carrier current (i.e. the hole current in this case) Io(t) generated at the n+ p junction by the avalanche multiplication grows exponentially with time while the voltage is above the critical (i.e. breakdown) value. During the negative half cycle, when VT is below the breakdown voltage for the diode, the current Io(t) decays exponentially to a small steady state value. The carrier current I0(t) is the current at the junction only and is in the form of a pulse of very short duration as shown in Fig. 3(b). Therefore, the carrier current I0(t) reaches its maximum in the middle of the a.c. voltage or lags by 900 behind the said a.c. voltage. The direction of the electric field is such that the generated holes are injected into the space-charge region towards the negative terminal. An equal number of generated electrons move to the left, back into the n+ - contact to maintain space charge neutrality .As the injected holes traverse the drift space, they induce a current Ie (t) in the external circuit which is approximately a square wave as shown in fig.3(c).The current Ie (t) flows in the external circuit for a time τ during which the holes are moving across the space-charge region. Thus, on the average, the external current Ie (t) due to the moving holes is delayed by τ/2 or 900 relative to the pulsed carrier current I0 (t) generated at the n+-p junction .Because the carrier current I0 (t) is already delayed by 900 relative to the a.c. voltage, the external current Ie (t) is then delayed by as a total of 1800 relative to the applied a.c. voltage .In general, a device exhibits negative resistance at its terminals when the a.c. current flowing though it lags the a.c. voltage by a phase angle

The device structure, doping profile and electric field distribution of Misawa diode i.e. a p-in diode reverse biased to avalanche breakdown is shown in Fig. 4 (a-c). Misawa assumed an

by τ = L/ *v*d.

which lies between 900 and 2700 .

**2.5.2 Misawa diode** 

$$1 - \frac{1}{M\_n} = \int\_o^W \alpha\_n \exp\{-\int\_o^x (\alpha\_n - \alpha\_p) d\mathbf{x}'\} dx = 1\tag{2.4.12a}$$

$$1 - \frac{1}{M\_p} = \int\_o^W \alpha\_p \exp\{\int (\alpha\_n - \alpha\_p) d\mathbf{x}'\} d\mathbf{x} = 1\tag{19}$$

For α α *<sup>n</sup>* = *<sup>p</sup>* , the above equation reduces to () 1 *a x o N E dx* = = α.

Therefore, the above equation governs the avalanche breakdown, the static and dynamic properties of IMPATT diodes. The enhancement of mobile space charge density modify the electric field profile, breakdown voltage and the depletion layer width, because the ionization rates get modified at various points in the space charge layer.

#### **2.4 Basic IMPATT phenomena**

The operation of an IMPATT device is based on two basic physical mechanisms: one is the **avalanche multiplication** caused by **impact ionization** [4] and the other is the **finite transit time** required by the charge carriers to **cross the depletion layer** with saturated drift velocity. The avalanche process turns out to be an inductive process causing a phase delay between the applied r.f. voltage and the generated r.f. current. The transit-time process adds an extra phase delay. Thus the device exhibits a high-frequency negative resistance when the combined phase delay due to the avalanche process and the finite transit time of the drifting carriers lie between 90 0 and 270 0 .

The IMPATT phenomena were studied by Read and Misawa by considering two different devices models. Read considered an n+-p-i-p+ diode structure and assumed that the spatial extent of avalanche zone is negligibly small. But Misawa in his p-i-n avalanche diode structure assumed an extended avalanche zone. However, in practical IMPATT structure like SDR, DDR etc, the avalanche zone is neither too thin like Read diode nor too wide like Misawa diode but it is intermediate between the two having finite avalanche zone width. In section 2.5.1 mechanisms of IMPATT mode of operation has been discussed with reference to (a) Read and (b) Misawa diodes and in the nest section and a brief overview of various practical IMPATT diode structures will be presented.

Several authors including Read and Misawa have carried out analysis of microwave/mmwave properties for different IMPATT structures and have found that the active diode impedance when the device generates microwave/mm-wave can be represented by a high frequency negative resistance in series with a capacitance .The magnitude of the negative resistance being much smaller than the capacitive impedance, the device is mainly capacitive.

#### **2.5 Mechanism of IMPATT mode of operation in (i) read diode and (ii) misawa diode 2.5.1 Read diode**

A schematic diagram of Read diode structure n+-p-i-p+ along with its doping profile and electric field distribution at reverse biased to avalanche breakdown is shown in Fig. 2. In the Read structure the superscript plus sign denotes very high doping and the i or ν refers to

<sup>1</sup> / <sup>1</sup> exp{ ( ) } 1

*M*−= − − =

<sup>1</sup> / <sup>1</sup> exp{ ( ) } 1

*M*−= − =

*p np*

 αα

*a x*

*o N E dx* = = α.

Therefore, the above equation governs the avalanche breakdown, the static and dynamic properties of IMPATT diodes. The enhancement of mobile space charge density modify the electric field profile, breakdown voltage and the depletion layer width, because the

The operation of an IMPATT device is based on two basic physical mechanisms: one is the **avalanche multiplication** caused by **impact ionization** [4] and the other is the **finite transit time** required by the charge carriers to **cross the depletion layer** with saturated drift velocity. The avalanche process turns out to be an inductive process causing a phase delay between the applied r.f. voltage and the generated r.f. current. The transit-time process adds an extra phase delay. Thus the device exhibits a high-frequency negative resistance when the combined phase delay due to the avalanche process and the finite transit time of the

The IMPATT phenomena were studied by Read and Misawa by considering two different devices models. Read considered an n+-p-i-p+ diode structure and assumed that the spatial extent of avalanche zone is negligibly small. But Misawa in his p-i-n avalanche diode structure assumed an extended avalanche zone. However, in practical IMPATT structure like SDR, DDR etc, the avalanche zone is neither too thin like Read diode nor too wide like Misawa diode but it is intermediate between the two having finite avalanche zone width. In section 2.5.1 mechanisms of IMPATT mode of operation has been discussed with reference to (a) Read and (b) Misawa diodes and in the nest section and a brief overview of various practical IMPATT diode structures will be

Several authors including Read and Misawa have carried out analysis of microwave/mmwave properties for different IMPATT structures and have found that the active diode impedance when the device generates microwave/mm-wave can be represented by a high frequency negative resistance in series with a capacitance .The magnitude of the negative resistance being much smaller than the capacitive impedance, the device is mainly

**2.5 Mechanism of IMPATT mode of operation in (i) read diode and (ii) misawa diode** 

A schematic diagram of Read diode structure n+-p-i-p+ along with its doping profile and electric field distribution at reverse biased to avalanche breakdown is shown in Fig. 2. In the Read structure the superscript plus sign denotes very high doping and the i or ν refers to

*n np*

 αα *dx dx*

*dx dx*

(2.4.12a)

(19)

*W x*

*W x*

*n o o*

*p o o*

*<sup>n</sup>* = *<sup>p</sup>* , the above equation reduces to () 1

ionization rates get modified at various points in the space charge layer.

For α α

presented.

capacitive.

**2.5.1 Read diode** 

**2.4 Basic IMPATT phenomena** 

drifting carriers lie between 90 0 and 270 0 .

α

α

intrinsic material .The device consists essentially of two regions ; One is the narrow p-region at which avalanche multiplication occurs .This region is also called the **high field region or the avalanche region** .The other is the i (or ν ) region through which the generated holes must drift while moving towards the p+ contact .This region is also called the **intrinsic region or the drift region** .When the reverse biased voltage is well above the punch through or breakdown voltage, the space between the n+p junction and the i-p+ junction happens to be the space-charge region . Carriers (holes) moving in the high field near the n+-p junction acquire energy to knock valance electrons into the conduction band, thus producing holeelectron pairs. The rate of pair production, or avalanche multiplication, is a sensitive nonlinear function of the field. By proper doping, the field can be given a relatively sharp peak so that avalanche multiplication is confined to a very narrow region at the n+-p junction. The electrons move in to the n+- region and the holes drift through the space charge region to the p+- region with a constant velocity *v*d (called the saturated drift velocity).The transit time of a hole across the drift region (i.e. i-region of length L) is given by τ = L/ *v*d.

The phenomenon of negative resistance in Read diode can be understood with reference to Fig. 3. In actual practice, to form oscillator, the diode is mounted in a microwave/mm-wave resonant circuit. An a.c. voltage can be maintained at a given frequency in the circuit thus the total voltage across the diode is the sum of the d.c. and a.c. voltages, mathematically : VT (t)=V DC + V D sin ωt , and the form of this total diode voltage is shown in Fig. 3(a). This total voltage causes breakdown at the n+ - p junction during the positive half cycle of the a.c. voltage when VT is above the breakdown value, and the carrier current (i.e. the hole current in this case) Io(t) generated at the n+ p junction by the avalanche multiplication grows exponentially with time while the voltage is above the critical (i.e. breakdown) value. During the negative half cycle, when VT is below the breakdown voltage for the diode, the current Io(t) decays exponentially to a small steady state value. The carrier current I0(t) is the current at the junction only and is in the form of a pulse of very short duration as shown in Fig. 3(b). Therefore, the carrier current I0(t) reaches its maximum in the middle of the a.c. voltage or lags by 900 behind the said a.c. voltage. The direction of the electric field is such that the generated holes are injected into the space-charge region towards the negative terminal. An equal number of generated electrons move to the left, back into the n+ - contact to maintain space charge neutrality .As the injected holes traverse the drift space, they induce a current Ie (t) in the external circuit which is approximately a square wave as shown in fig.3(c).The current Ie (t) flows in the external circuit for a time τ during which the holes are moving across the space-charge region. Thus, on the average, the external current Ie (t) due to the moving holes is delayed by τ/2 or 900 relative to the pulsed carrier current I0 (t) generated at the n+-p junction .Because the carrier current I0 (t) is already delayed by 900 relative to the a.c. voltage, the external current Ie (t) is then delayed by as a total of 1800 relative to the applied a.c. voltage .In general, a device exhibits negative resistance at its terminals when the a.c. current flowing though it lags the a.c. voltage by a phase angle which lies between 900 and 2700 .

#### **2.5.2 Misawa diode**

The device structure, doping profile and electric field distribution of Misawa diode i.e. a p-in diode reverse biased to avalanche breakdown is shown in Fig. 4 (a-c). Misawa assumed an

SiC Devices on Different Polytypes: Prospects and Challenges 349

Fig. 3. Voltage and currents in Read diode, (a) Total diode voltage, (b) Carrier current generated at n+-p junction by avalanche multiplication and (c) Current induced in the

external circuit.

Fig. 2. (a) Read (n+-p-i-p+) structure (b) Doping profile and (c) Electric field distribution.

Fig. 2. (a) Read (n+-p-i-p+) structure (b) Doping profile and (c) Electric field distribution.

Fig. 3. Voltage and currents in Read diode, (a) Total diode voltage, (b) Carrier current generated at n+-p junction by avalanche multiplication and (c) Current induced in the external circuit.

SiC Devices on Different Polytypes: Prospects and Challenges 351

Fig. 4. (a) Misawa diode (p-i-n) structure, (b) Doping profile (c) Electric Field distribution (d) Electric field and generated electron density wave and (e) Phase diagram (current and field)

uniform avalanching i.e. the field remains high enough for avalanche multiplication of charge carriers to take place throughout the active region (i.e. i-region). The origin of negative resistance can be understood as follows. The electrons and holes enter the i-region from the p- and n- regions respectively Fig. 4 (a) and they move with saturated velocities and generates holes and electrons simultaneously in this region. Now let us consider a fluctuation in the electron density (ñ0) in the i-region which will be moving towards the xdirection with saturated drift velocity being carried by the electrons moving is the same direction Fig. 4(d) .The electric field wave ( E ) caused by this perturbation in the electron density lags the electron density wave by 900. Now the generation rate of electron-hole pairs is larger both when the electric field is stronger and when there are more carriers. Therefore, in this case of an electron density perturbation the generation rate peaks somewhere between the place where the field is strongest and the place where the density is largest. This means that the generation rate (G) leads the electron density wave by less than 900 .It is to be noted that since the d.c. field is in negative x-direction, the field becomes strongest at its negative peak. The increased generation rate gives rise to an excess electron density (∆ñ) which lags the rate by 900. Thus the resultant electron density (ñ) gives a current ( *j* n) which lags the field by more than 900 .The current due to hole density wave also lags the field by more than 900 .The situation is shown in Fig. 4 (e) .Thus the current generated in the device lags the field by more than 90 0 and hence the device exhibits negative resistance property.

#### **2.6 Practical IMPATT diode**

The Read and Misawa diodes that have been discussed are idealized IMPATT structures. But practical IMPATT diodes which have been fabricated and are in wide use are intermediate between the two in the sense that the avalanche zone is well defined having a finite width neither too thin nor too wide. The charge is injected from a well defined avalanche zone into the drift zone approximately 900 behind the r.f. voltage and then the injected charge travels with saturated drift velocity in the drift region. The practical Single Drift Region (SDR) and Double Drift Region (DDR) IMPATT which are now commonly used belong to this category. Diodes have been fabricated from various semiconductor materials i.e. Germanium, Silicon, Gallium Arsenide, Indium Phosphide and recently from Silicon Carbide.

#### **2.7 Impedance properties and equivalent circuit of IMPATT diode**

A knowledge of the nature of device impedance is quite important to understand the mechanism of microwave/mm-waves generation by IMPATT diodes. Also an insight into the impedance properties of the device helps one to choose the microwave/mm-wave circuit necessary to construct oscillators with these devices. Several authors including Read , Gilden and Hines have analyzed the small-signal impedance properties of Read type IMPATT diode. Misawa has calculated the small signal impedance of the p-i-n avalanche diode both analytically and numerically. Gummel and Scharfetter extended the small signal analysis of Gilden and Hines and obtained small-signal admittance plots for a more realistic Read structure. Also Scharfetter and Gummel carried out large-signal numerical solution for Si Read diode and obtained the diode admittance as a function of frequency and r.f. voltage amplitude.

uniform avalanching i.e. the field remains high enough for avalanche multiplication of charge carriers to take place throughout the active region (i.e. i-region). The origin of negative resistance can be understood as follows. The electrons and holes enter the i-region from the p- and n- regions respectively Fig. 4 (a) and they move with saturated velocities and generates holes and electrons simultaneously in this region. Now let us consider a fluctuation in the electron density (ñ0) in the i-region which will be moving towards the xdirection with saturated drift velocity being carried by the electrons moving is the same

density lags the electron density wave by 900. Now the generation rate of electron-hole pairs is larger both when the electric field is stronger and when there are more carriers. Therefore, in this case of an electron density perturbation the generation rate peaks somewhere between the place where the field is strongest and the place where the density is largest. This means that the generation rate (G) leads the electron density wave by less than 900 .It is to be noted that since the d.c. field is in negative x-direction, the field becomes strongest at its negative peak. The increased generation rate gives rise to an excess electron density (∆ñ) which lags the rate by 900. Thus the resultant electron density (ñ) gives a current ( *j*

which lags the field by more than 900 .The current due to hole density wave also lags the field by more than 900 .The situation is shown in Fig. 4 (e) .Thus the current generated in the device lags the field by more than 90 0 and hence the device exhibits negative resistance

The Read and Misawa diodes that have been discussed are idealized IMPATT structures. But practical IMPATT diodes which have been fabricated and are in wide use are intermediate between the two in the sense that the avalanche zone is well defined having a finite width neither too thin nor too wide. The charge is injected from a well defined avalanche zone into the drift zone approximately 900 behind the r.f. voltage and then the injected charge travels with saturated drift velocity in the drift region. The practical Single Drift Region (SDR) and Double Drift Region (DDR) IMPATT which are now commonly used belong to this category. Diodes have been fabricated from various semiconductor materials i.e. Germanium, Silicon, Gallium Arsenide, Indium Phosphide and recently from

A knowledge of the nature of device impedance is quite important to understand the mechanism of microwave/mm-waves generation by IMPATT diodes. Also an insight into the impedance properties of the device helps one to choose the microwave/mm-wave circuit necessary to construct oscillators with these devices. Several authors including Read , Gilden and Hines have analyzed the small-signal impedance properties of Read type IMPATT diode. Misawa has calculated the small signal impedance of the p-i-n avalanche diode both analytically and numerically. Gummel and Scharfetter extended the small signal analysis of Gilden and Hines and obtained small-signal admittance plots for a more realistic Read structure. Also Scharfetter and Gummel carried out large-signal numerical solution for Si Read diode and obtained the diode admittance as a function of frequency and r.f. voltage

**2.7 Impedance properties and equivalent circuit of IMPATT diode** 

E ) caused by this perturbation in the electron

n)

direction Fig. 4(d) .The electric field wave (

property.

Silicon Carbide.

amplitude.

**2.6 Practical IMPATT diode** 

Fig. 4. (a) Misawa diode (p-i-n) structure, (b) Doping profile (c) Electric Field distribution (d) Electric field and generated electron density wave and (e) Phase diagram (current and field)

SiC Devices on Different Polytypes: Prospects and Challenges 353

Fig. 5. (a) Model of Read diode with avalanche region, drift region and inactive region. (b)

Equivalent circuit of the avalanche region.

Gilden and Hines derived an expression for the diode terminal impedance in a Read type structure by assuming a thin avalanche zone where space charge and signal delay is negligible and a wide drift zone where no carriers are formed but where space-charge and transit time effects are significant. Denoting Za as avalanche zone impedance, Zd as drift zone impedance and Rs as the passive resistance of the inactive zone , Gilden and Hines obtained the following expression for the terminal impedance

$$\mathbf{Z} = R\_s + Z\_a + Z\_d = R\_s + \frac{1}{j a \mathcal{C}\_a} \left( 1 - \frac{a\_a^2}{a^2} \right)^{-1} + \frac{1}{a \mathcal{C}\_d} \left| \left( 1 - \frac{a^2}{a\_a^2} \right)^{-1} \left( \frac{1 - \text{Cov}\theta}{\theta} \right) \right| + \frac{1}{j a \mathcal{C}\_d} $$

$$+ \frac{1}{j a \mathcal{C}\_d} \left\{ 1 - \left( 1 - \frac{a^2}{a\_a^2} \right)^{-1} \left( \frac{\text{Sim}\theta}{\theta} \right) \right\} \tag{20}$$

where ,

*Ca* = *a A l* <sup>∈</sup> , avalanche zone capacitance *Cd* = *d A l* <sup>∈</sup> , drift zone capacitance θ = *<sup>d</sup> d l v* ω , transit angle in the drift zone 2 ω*a* = <sup>0</sup> <sup>2</sup> *<sup>d</sup>* α′*v J* ∈ , avalanche resonance frequency with *d*( ) *dE* α α′ = , the derivative of ionization rate w.r.t. electronic field E , J0 is the d.c.

current density . *l*a and *l*d the avalanche zone and drift zone lengths, *v*d is the saturated drift velocity , A is the cross - sectional area , and ∈ is the permittivity .

It can be observed from equation (20) that the avalanche zone is represented by an antiresonant circuit as shown in Fig. 5 .The drift zone consists of a resistive and a reactive part .The resistive part (i.e. the real part of Zd) is negative for all frequencies above ω<sup>a</sup> (except for nulls at θ = 2π x integer) and below this frequency it is positive .It is important to note that the avalanche zone does not contribute to the device negative resistance. For small transit angles i.e. for θ <π /4, equation (20) reduces to :-

$$Z = R\_s + \frac{l\_d}{\upsilon\_d \in A} \left( 1 - \frac{a\nu^2}{a\varrho\_a^2} \right)^{-1} + \frac{1}{j\alpha \mathcal{C}} \left( 1 - \frac{a\nu^2}{a\varrho\_a^2} \right)^{-1} \tag{21}$$

Where , *a d <sup>A</sup> <sup>C</sup> l l* <sup>∈</sup> <sup>=</sup> + .

A typical plot of equation (21) is shown in Fig.6. It is observed that the diode reactance, X , changes from inductive to capacitive at the avalanche resonance frequency and also the resistive part , R, changes sign at the same frequency and become negative above ωa .Thus the device possesses negative resistance for all frequencies above ωa and there the device reactance is capacitive.

Gilden and Hines derived an expression for the diode terminal impedance in a Read type structure by assuming a thin avalanche zone where space charge and signal delay is negligible and a wide drift zone where no carriers are formed but where space-charge and transit time effects are significant. Denoting Za as avalanche zone impedance, Zd as drift zone impedance and Rs as the passive resistance of the inactive zone , Gilden and Hines

<sup>1</sup> <sup>2</sup>

+ 1 ω*Cd*

 θ*Sin*

θ

<sup>1</sup> <sup>2</sup> 2 <sup>1</sup> <sup>1</sup> *a* ω

<sup>−</sup> <sup>−</sup> <sup>−</sup>

ω

 θ*Cos*

+

(20)

θ

<sup>2</sup> 1 ω*a* ω

<sup>2</sup> 1 1

<sup>−</sup> − 

<sup>1</sup> <sup>2</sup>

*a* ω

<sup>−</sup> − −

′ = , the derivative of ionization rate w.r.t. electronic field E , J0 is the d.c.

1 1 2 2 2 2

ω

− − −+ −

 ω

 ω

<sup>∈</sup> (21)

current density . *l*a and *l*d the avalanche zone and drift zone lengths, *v*d is the saturated drift

It can be observed from equation (20) that the avalanche zone is represented by an antiresonant circuit as shown in Fig. 5 .The drift zone consists of a resistive and a reactive part .The resistive part (i.e. the real part of Zd) is negative for all frequencies above ω<sup>a</sup> (except for nulls at θ = 2π x integer) and below this frequency it is positive .It is important to note that the avalanche zone does not contribute to the device negative resistance. For small

> <sup>1</sup> 1 1 *<sup>d</sup> d a a*

*v A j C* ω

ω

A typical plot of equation (21) is shown in Fig.6. It is observed that the diode reactance, X , changes from inductive to capacitive at the avalanche resonance frequency and also the resistive part , R, changes sign at the same frequency and become negative above ωa .Thus the device possesses negative resistance for all frequencies above ωa and there the device

ω

1 *<sup>a</sup> j*ω*C*

obtained the following expression for the terminal impedance

1 *<sup>d</sup> j C*ω

Z = *Rs* + *Za* + *Zd* = *Rs* +

<sup>∈</sup> , avalanche zone capacitance

, transit angle in the drift zone

, avalanche resonance frequency

velocity , A is the cross - sectional area , and ∈ is the permittivity .

*l*

transit angles i.e. for θ <π /4, equation (20) reduces to :-

Z = *Rs* +

<sup>∈</sup> , drift zone capacitance

+

where ,

*a A l*

*d A l*

*a* = <sup>0</sup> <sup>2</sup> *<sup>d</sup>* α′*v J* ∈

with *d*( )

α

Where ,

*dE* α

*a d <sup>A</sup> <sup>C</sup> l l* <sup>∈</sup> <sup>=</sup> + .

reactance is capacitive.

*Ca* =

*Cd* =

θ = *<sup>d</sup> d l v* ω

2 ω

Fig. 5. (a) Model of Read diode with avalanche region, drift region and inactive region. (b) Equivalent circuit of the avalanche region.

SiC Devices on Different Polytypes: Prospects and Challenges 355

interest, XD can be approximated with sufficient accuracy by the reactance of the junction (chip) capacitance at the breakdown voltage. Thirdly, the values of the negative resistance are generally small compared with the usual transmission line impedances. Further, it may be mentioned that the magnitude of negative resistance of IMPATT diode varies with signal

The properties of IMPATT diode chip have so far been discussed, but the chip by itself is difficult to handle and it is prone to mechanical damage and environmental contamination. So, it is necessary to provide the diode chip with a robust, hermetically sealed package. The package also provides the necessary heat- sink arrangements for dissipating the heat from the diode chip to the ambient .A common type of commercially available S4 package has been shown in Fig.2.6.3.(b) and also equivalent circuit of the packaged IMPATT diode having packaged parasitics . To a good approximation, the package can be described by two reactive elements a series inductance, Lp and a shunt capacitance Cp .The exact values of Lp

Fig. 7. (a) IMPATT diode chip r.f. equivalent circuit and (b) Cross-section of the chip in S4

package with equivalent circuit of the packaged IMPATT diode.

level and that it increases with increasing signal level.

and Cp varies from one package style to another.

Fig. 6. Typical Impedance variation with frequency for a Read diode.

Gummel and Scharfetter extended the small-signal analysis of Gilden and Hines to include diode in which the avalanche region is not necessarily narrow .They obtained small signal admittance plots for a Read diode and for more realistic diode structures in which the avalanche region occupies an appreciable fraction of the total depletion region. They found that the optimum performance is achieved when the avalanche region is one third of the drift region. Scharfetter and Gummel have developed large-signal numerical analysis for Read diodes with a realistic doping profile including the effect of microwave circuit in which the diode is placed.

The results of investigation on impedance properties of IMPATT diode by various workers can be summed up to obtain an idea of the r.f. equivalent circuit for the IMPATT diode chip. In general, the diode equivalent circuit will consist of a negative resistance -RD in series with a capacitive impedance XD (or alternatively by a negative conductance -GD in parallel with a capacitive susceptance BD). The r.f. equivalent circuit of IMPATT diode chip is shown in Fig. 7 (a). Some important observation regarding the nature and values of RD and XD is worth mentioning. First, the magnitude of RD is usually much smaller than XD. Consequently, the magnitude of the chip impedance is approximately equal to XD. Secondly, for most cases of

Fig. 6. Typical Impedance variation with frequency for a Read diode.

which the diode is placed.

Gummel and Scharfetter extended the small-signal analysis of Gilden and Hines to include diode in which the avalanche region is not necessarily narrow .They obtained small signal admittance plots for a Read diode and for more realistic diode structures in which the avalanche region occupies an appreciable fraction of the total depletion region. They found that the optimum performance is achieved when the avalanche region is one third of the drift region. Scharfetter and Gummel have developed large-signal numerical analysis for Read diodes with a realistic doping profile including the effect of microwave circuit in

The results of investigation on impedance properties of IMPATT diode by various workers can be summed up to obtain an idea of the r.f. equivalent circuit for the IMPATT diode chip. In general, the diode equivalent circuit will consist of a negative resistance -RD in series with a capacitive impedance XD (or alternatively by a negative conductance -GD in parallel with a capacitive susceptance BD). The r.f. equivalent circuit of IMPATT diode chip is shown in Fig. 7 (a). Some important observation regarding the nature and values of RD and XD is worth mentioning. First, the magnitude of RD is usually much smaller than XD. Consequently, the magnitude of the chip impedance is approximately equal to XD. Secondly, for most cases of interest, XD can be approximated with sufficient accuracy by the reactance of the junction (chip) capacitance at the breakdown voltage. Thirdly, the values of the negative resistance are generally small compared with the usual transmission line impedances. Further, it may be mentioned that the magnitude of negative resistance of IMPATT diode varies with signal level and that it increases with increasing signal level.

The properties of IMPATT diode chip have so far been discussed, but the chip by itself is difficult to handle and it is prone to mechanical damage and environmental contamination. So, it is necessary to provide the diode chip with a robust, hermetically sealed package. The package also provides the necessary heat- sink arrangements for dissipating the heat from the diode chip to the ambient .A common type of commercially available S4 package has been shown in Fig.2.6.3.(b) and also equivalent circuit of the packaged IMPATT diode having packaged parasitics . To a good approximation, the package can be described by two reactive elements a series inductance, Lp and a shunt capacitance Cp .The exact values of Lp and Cp varies from one package style to another.

Fig. 7. (a) IMPATT diode chip r.f. equivalent circuit and (b) Cross-section of the chip in S4 package with equivalent circuit of the packaged IMPATT diode.

SiC Devices on Different Polytypes: Prospects and Challenges 357

illumination configuration: (a) Top Mounted and (b) Flip chip. The corresponding diagrams

In order to assess the role of leakage currents in controlling the dynamic properties of IMPATT oscillators at THz frequencies, simulation experiments are carried out on the effect of electron current multiplication factor, Mn , (keeping hole current multiplication factor Mp very high ~ 106) and Mp (keeping Mn very high ~ 106) on (i) the small-signal admittance characteristics, (ii) the negative resistivity profiles, (iii) quality factor at peak frequencies (Qp), (iv) device negative resistance at peak frequencies (-ZRP) and (v) maximum power

DC simulation program is used to obtain the E(x) and P(x) profiles of flat profile SiC IMPATT diodes which are designed and optimized for operation at 0.3 THz regime. The optimized design parameters and corresponding bias current densities for each diode are shown in Table 1. In figures 10(a-c), plots of E(x) and P(x) profiles of DDR SiC based unilluminated and illuminated (TM and FC configuration) IMPATTs are presented. It is intersting to note that there are small changes in the electric field profile due to the lowering of Mn, corresponding to TM illumination configuration, while the variation is comparatively much prominent due to the lowering of MP, corrsponding to FC illumination configuration. Analysis of P(x) profiles, as shown in figure 10(a-c), reveals that avalanche centre, at which JP =Jn, moves towards the metallurgical junction from n-side with the lowering of both Mn and Mp. Similar to E(x) profiles, P(x) profiles of SiC based DDRs are also much sensitive to hole leakage current. Table 2 shows that the 4H-SiC IMPATT breaks down at 135.0 V , which is atleast 22.7% higher than its 6H-SiC and 3C-SiC counterpart. The variation of breakdown voltage (VB) of the designed SiC based IMPATTs, with the enhancement of photo-leakage current is shown in Figure 11. It is clear that compare to TM

configuration the diode breakdown voltages decreases more. Moreover, the breakdown voltage of the illuminated hexagonal SiC based devices reduces much than its illuminated

4H-SiC 6.5 6.5 250.0 250.0 3.4 6H-SiC 8.0 8.0 250.0 250.0 3.5 3C-SiC 8.0 8.0 250.0 250.0 3.7

4H-SiC based diode is found to be more efficient (14%) than 6H-SiC and 3C-SiC based diodes, under almost similar operating condition. Moreover the negative conductance of the 4H-SiC IMPATT is found to be ~ 55.0% and 7.0% higher than 6H-SiC and 3C-SiC based IMPATT. The higher value of diode breadown voltage and negative conductance in 4H-SiC based diode icreases the RF power level. It is clear from the table 2, that 4H-SiC based

Width of the n region (nm)

Width of the p region (nm)

Current density (109 A m-2)

Doping conc. ( p region) (10 23m-3)

Table 1. Design Parameters of SiC IMPATT Diodes at 0.3THz Frequency

are shown in Figure 9.

output of DDR SiC IMPATTs.

cubic SiC counterpart.

DDR diode type

**4. Observations from simulation experiment** 

illumination configuration, in case of FC illumination

Doping conc. (n region) (10 23 m-3)

#### **3. Simulation experiment**

Different polytypes of SiC are shown in Figure 8. At first, SiC diodes are designed and optimized through a generalized double iterative simulation technique used for the analysis of IMPATT action [1]. The fundamental device equations, i.e. the one-dimensional *Poisson's*  equation and the combined current continuity equations under steady-state conditions, have been numerically solved subject to appropriate boundary conditions, through an accurate and generalized double iterative computer algorithm, described elsewhere. Iteration over the value and location of field maximum are carried out until the boundary conditions of E(x) and P(x) = [JP(x) – Jn(x)]/J0 are satisfied at both the edges of diode active layer. The DC solution gives the electric field E(x) profile, normalized current density P(x) profile, the maximum electric field (Em), drift voltage drop (VD), breakdown voltage (VB) and avalanche zone width (xa). The breakdown voltage (VB) is calculated by integrating the spatial field profile over the total depletion layer width. The boundary conditions for current density profiles are fixed by assuming a high multiplication factor (Mn, p) ~ 106, since it is well known that, avalanche breakdown occurs in the diode junction when the electric field is large enough such that the charge multiplication factors (Mn, Mp) become infinite. The edges of the depletion layer are also determined accurately from the DC analysis.

The small-signal analysis of the IMPATT diode provides insight into the high-frequency performance of the diode. The range of frequencies exhibiting negative conductance of the diode can easily be computed by *Gummel-Blue* method [2]. From the DC field and current profiles, the spatially dependent ionization rates that appear in the *Gummel-Blue* equations are evaluated and fed as input data for the small-signal analysis. The edges of the depletion layer of the diode, which are fixed by the DC analysis, are taken as the starting and end points for the small-signal analysis. The spatial variation of high-frequency negative resistivity and reactivity in the depletion layer of the diode are obtained under small-signal conditions by solving two second order differential equations in R(x, ω) and X(x, ω). R(x, ω) and X(x, ω) are the real and imaginary part of diode impedance Z (x,ω), such that, Z (x,ω) = R(x, ω) + j X(x, ω). The total integrated diode negative resistance (ZR) and reactance (Z x) at a particular frequency (ω) and current density J0, are computed from numerical integration of the R(x) and X(x) profiles over the active space-charge layer.

At resonance, the reactance of the resonant cavity is mainly capacitive in nature. When the magnitude of negative conductance of the diode |-G| is equal to the load conductance GL, the condition of resonance is satisfied and as a result, power is absorbed in GL and at the same time oscillation starts to build up in the circuit. *Adlerstein et al.* developed a method for determining RS from the threshold condition of IMPATT oscillation [3]. In the present method, the author has determined the value of series resistance (RS) from the admittance characteristics using a realistic analysis of *Gummel-Blue* [2] and *Adlerstein et al* [3]. The author has considered the effect of RS in the realistic analysis of output power from the THz diodes. The basic mechanism of optical control of IMPATT diode is discussed earlier. In summary, the leakage current entering the depletion region of the reversed biased p-n junction of an un-illuminated IMPATT diode is only due to thermally generated electron-hole pairs and it is so small that the multiplication factors (M n , p) become very high. When optical radiation of suitable wavelength (photon energy hc/λ > Eg) is incident on the active layer of the device, the leakage current increases significantly due to photo-generation of charge carriers. The enhancement of the leakage current under optical illumination of the devices is manifested as the lowering of Mn,p. The photo-sensitivity of IMPATTs are studied under two

Different polytypes of SiC are shown in Figure 8. At first, SiC diodes are designed and optimized through a generalized double iterative simulation technique used for the analysis of IMPATT action [1]. The fundamental device equations, i.e. the one-dimensional *Poisson's*  equation and the combined current continuity equations under steady-state conditions, have been numerically solved subject to appropriate boundary conditions, through an accurate and generalized double iterative computer algorithm, described elsewhere. Iteration over the value and location of field maximum are carried out until the boundary conditions of E(x) and P(x) = [JP(x) – Jn(x)]/J0 are satisfied at both the edges of diode active layer. The DC solution gives the electric field E(x) profile, normalized current density P(x) profile, the maximum electric field (Em), drift voltage drop (VD), breakdown voltage (VB) and avalanche zone width (xa). The breakdown voltage (VB) is calculated by integrating the spatial field profile over the total depletion layer width. The boundary conditions for current density profiles are fixed by assuming a high multiplication factor (Mn, p) ~ 106, since it is well known that, avalanche breakdown occurs in the diode junction when the electric field is large enough such that the charge multiplication factors (Mn, Mp) become infinite. The

edges of the depletion layer are also determined accurately from the DC analysis.

the R(x) and X(x) profiles over the active space-charge layer.

The small-signal analysis of the IMPATT diode provides insight into the high-frequency performance of the diode. The range of frequencies exhibiting negative conductance of the diode can easily be computed by *Gummel-Blue* method [2]. From the DC field and current profiles, the spatially dependent ionization rates that appear in the *Gummel-Blue* equations are evaluated and fed as input data for the small-signal analysis. The edges of the depletion layer of the diode, which are fixed by the DC analysis, are taken as the starting and end points for the small-signal analysis. The spatial variation of high-frequency negative resistivity and reactivity in the depletion layer of the diode are obtained under small-signal conditions by solving two second order differential equations in R(x, ω) and X(x, ω). R(x, ω) and X(x, ω) are the real and imaginary part of diode impedance Z (x,ω), such that, Z (x,ω) = R(x, ω) + j X(x, ω). The total integrated diode negative resistance (ZR) and reactance (Z x) at a particular frequency (ω) and current density J0, are computed from numerical integration of

At resonance, the reactance of the resonant cavity is mainly capacitive in nature. When the magnitude of negative conductance of the diode |-G| is equal to the load conductance GL, the condition of resonance is satisfied and as a result, power is absorbed in GL and at the same time oscillation starts to build up in the circuit. *Adlerstein et al.* developed a method for determining RS from the threshold condition of IMPATT oscillation [3]. In the present method, the author has determined the value of series resistance (RS) from the admittance characteristics using a realistic analysis of *Gummel-Blue* [2] and *Adlerstein et al* [3]. The author has considered the effect of RS in the realistic analysis of output power from the THz diodes. The basic mechanism of optical control of IMPATT diode is discussed earlier. In summary, the leakage current entering the depletion region of the reversed biased p-n junction of an un-illuminated IMPATT diode is only due to thermally generated electron-hole pairs and it is so small that the multiplication factors (M n , p) become very high. When optical radiation of suitable wavelength (photon energy hc/λ > Eg) is incident on the active layer of the device, the leakage current increases significantly due to photo-generation of charge carriers. The enhancement of the leakage current under optical illumination of the devices is manifested as the lowering of Mn,p. The photo-sensitivity of IMPATTs are studied under two

**3. Simulation experiment** 

illumination configuration: (a) Top Mounted and (b) Flip chip. The corresponding diagrams are shown in Figure 9.

In order to assess the role of leakage currents in controlling the dynamic properties of IMPATT oscillators at THz frequencies, simulation experiments are carried out on the effect of electron current multiplication factor, Mn , (keeping hole current multiplication factor Mp very high ~ 106) and Mp (keeping Mn very high ~ 106) on (i) the small-signal admittance characteristics, (ii) the negative resistivity profiles, (iii) quality factor at peak frequencies (Qp), (iv) device negative resistance at peak frequencies (-ZRP) and (v) maximum power output of DDR SiC IMPATTs.

#### **4. Observations from simulation experiment**

DC simulation program is used to obtain the E(x) and P(x) profiles of flat profile SiC IMPATT diodes which are designed and optimized for operation at 0.3 THz regime. The optimized design parameters and corresponding bias current densities for each diode are shown in Table 1. In figures 10(a-c), plots of E(x) and P(x) profiles of DDR SiC based unilluminated and illuminated (TM and FC configuration) IMPATTs are presented. It is intersting to note that there are small changes in the electric field profile due to the lowering of Mn, corresponding to TM illumination configuration, while the variation is comparatively much prominent due to the lowering of MP, corrsponding to FC illumination configuration. Analysis of P(x) profiles, as shown in figure 10(a-c), reveals that avalanche centre, at which JP =Jn, moves towards the metallurgical junction from n-side with the lowering of both Mn and Mp. Similar to E(x) profiles, P(x) profiles of SiC based DDRs are also much sensitive to hole leakage current. Table 2 shows that the 4H-SiC IMPATT breaks down at 135.0 V , which is atleast 22.7% higher than its 6H-SiC and 3C-SiC counterpart. The variation of breakdown voltage (VB) of the designed SiC based IMPATTs, with the enhancement of photo-leakage current is shown in Figure 11. It is clear that compare to TM illumination configuration, in case of FC illumination

configuration the diode breakdown voltages decreases more. Moreover, the breakdown voltage of the illuminated hexagonal SiC based devices reduces much than its illuminated cubic SiC counterpart.


Table 1. Design Parameters of SiC IMPATT Diodes at 0.3THz Frequency

4H-SiC based diode is found to be more efficient (14%) than 6H-SiC and 3C-SiC based diodes, under almost similar operating condition. Moreover the negative conductance of the 4H-SiC IMPATT is found to be ~ 55.0% and 7.0% higher than 6H-SiC and 3C-SiC based IMPATT. The higher value of diode breadown voltage and negative conductance in 4H-SiC based diode icreases the RF power level. It is clear from the table 2, that 4H-SiC based

SiC Devices on Different Polytypes: Prospects and Challenges 359

lowering of Mn and Mp. The simulated results for the illuminated 4H-SiC IMPATT diode (Table 4) indicate that the values of │-GP│and PRF decrease nearly by the same percentage (~3.8.0 %) as Mn decreases from 106 to 25, keeping the value of MP constant at 106. On the other hand, for a similar variation of Mp from 106 to 10 (keeping the value of Mn constant at 106), reduces the values of │-GP│ and PRF by ~ 35.0 %. Similar trend is reflected in Figure 13. In the case of 6H-SiC and 3C-SiC IMPATT diodes, as the value of Mn decreases from 106 to 25 (corresponding to TM illumination configuration), │-GP│ and PRF decrease by 5.2 % and 2% , respectively (Table 4). Similarly for the FC illumination configuration of 6H-SiC and 3C-SiC IMPATTs, the decrease of MP from 106 to 25, causes a reduction in the values of │-GP│ and PRF by 18.2% and 21.6 %,

> Estimated load conductance (g) (106Sm-2)

Negative resistance (-ZR) (10-9Ωm2)

Series Resistance (RS) (10-9 Ωm2)

Susceptance (B) (106 Sm-2)

4H-SiC 153.0 150.0 130.0 3.34 1.02

6H-SiC 95.0 152.0 63.7 2.95 1.33

3C-SiC 150.0 153.0 125.0 3.26 1.1

It is evident from Table 4, that in the case of 4H-SiC based TM diode, a lowering of Mn from 106 to 25 causes the diode negative resistance (-ZRp) to decrease by 3.0%, while there is a corresponding lowering of -ZRp by 51.0 % in case of FC diode. Similarly for TM 6H-SiC and 3C-SiC diodes, the value of -ZRp reduces by 20.0% and 10.0 %, respectively, as Mn decreases from 106 to 25 (Table 4), whereas, for a similar variation of MP, │-ZRp│ reduces by 40.0 % and 25.0 % , respectively in 6H-SiC and 3C-SiC based diodes. The device quality factors are also found to degrade with the lowering of Mn and MP, in case

It is further evident from figure 13 that in α-SiC and β-SiC based DDR devices a lowering of Mp causes larger upward shift in frequency than corresponding lowering of Mn. The optimum frequency of oscillation (fP) shifts upwards by 15.0 GHz and 45.0GHz respectively for TM and FC IMPATTs based on 4H-SiC. On the other hand, in case of 6H-SiC and 3C-SiC based TM IMPATTs, fP shifts upward by 1.0 GHz and 2.0 GHz, respectively. Whereas under FC illumination configuration, the values of fP shifts upward by 3.0 GHz and 7.0 GHz

Thus the studies reveal that, effects of photo-illumination on the frequency up shift as well as on the modulation of the THz behavior of the SiC devices are found to be more pronounced in FC illumination configuration than that for TM illumination configuration under similar operating condition. These results show an identical trend as observed

Table 3. Series resistance of the designed diodes at THz regime (frequency = 0.3 THz)

respectively.

Diode type

of all the SiC based diodes.

respectively for 6H-SiC and 3C-SiC based diodes.

previously for MM-wave SiC devices [4].

Negative conductance (-G) (106 Sm-2)

IMPATT is capable of delivering a RF power density of 36.45x1010 Wm-2 , which is ~2.5 times and 1.6 times higher than 6H-SiC and 3C-SiC based IMPATTs, respectively. Furthermore, the device negative resistance (-ZRP) at peak frequencies of the 4H-SiC based device is atleast 45% higher than its counterparts. The higher value of negative resistance is an essential criterian for getting sustained oscillation from the THz devices. The quality factor of 4H-SiC based device is found to be best among all the designed diodes.


Table 2. DC and high-frequency properties of SiC IMPATT Diodes at around 0.3 THz Frequency

The values of RS for all the three designed diodes are estimated from Adlerstein's approach, as mentioned in earlier section and the results are shown in Table 3. It is depicted that among all the designed diodes, the magnitude of parasitic positive series resistance is least in case of 4H-SiC based devices. Moreover, Table 3 also shows that the ratioes of negative resistance : positive series resistance in 4H-SiC , 6H-SiC and 3C-SiC based diodes are 3.3: 2.21: 2.96. The much higher value of negative resistance than its positive series resistance in 4H-SiC based device indicates that IMPATT diode based on 4H-SiC material system will be an potential candidate for generating THz power. The effects of RS on the admittance charecteristics of the devices are shown in Fgure 12. Figure 12 indicates that parasitic series resistance degrades the admittance charecteristics of the designed SiC based DDRs. The degradation of admittance charecteristics due to the presence of RS is more serious in 6H-SiC based IMPATT.

Thus the present study definitely establishes that the prospects of 4H-SiC based IMPATT as a high power, efficient THz source is far better than its 6H-SiC and 3C-SiC counterparts.

The effects of optical illumination on the THz behavior of the designed diodes are shown in Table 4. The computed values of –GP, -ZRP, PRF, fP and –QP for different electron and hole current multiplication factors are shown in Table 4. Admittance plots of all the SiC DDR IMPATTs under optical illumination are shown in Figure 13. It is evident from the Figure as well as from Table 4 that the values of │-GP│ at the optimum frequencies decrease with the lowering of Mn and Mp. At the same time, the frequency ranges over which the devices exhibit negative conductance, shift towards higher frequencies with the

IMPATT is capable of delivering a RF power density of 36.45x1010 Wm-2 , which is ~2.5 times and 1.6 times higher than 6H-SiC and 3C-SiC based IMPATTs, respectively. Furthermore, the device negative resistance (-ZRP) at peak frequencies of the 4H-SiC based device is atleast 45% higher than its counterparts. The higher value of negative resistance is an essential criterian for getting sustained oscillation from the THz devices. The quality factor of 4H-SiC

(108 V m-1) 4.25 3.8 5.65

(-Gp)(106 S m-2) 162.0 102.0 152.0

(1010 Wm-2) 36.45 15.15 22.99

Breakdown voltage (VB) (V) 135.0 109.0 110.0 Efficiency (η) (%) 14.0 12.0 12.5 Peak frequency (fp) (THz) 0.325 0.35 0.353

Device quality factor (-Qp) 1.26 2.45 1.79

Table 2. DC and high-frequency properties of SiC IMPATT Diodes at around 0.3 THz

The values of RS for all the three designed diodes are estimated from Adlerstein's approach, as mentioned in earlier section and the results are shown in Table 3. It is depicted that among all the designed diodes, the magnitude of parasitic positive series resistance is least in case of 4H-SiC based devices. Moreover, Table 3 also shows that the ratioes of negative resistance : positive series resistance in 4H-SiC , 6H-SiC and 3C-SiC based diodes are 3.3: 2.21: 2.96. The much higher value of negative resistance than its positive series resistance in 4H-SiC based device indicates that IMPATT diode based on 4H-SiC material system will be an potential candidate for generating THz power. The effects of RS on the admittance charecteristics of the devices are shown in Fgure 12. Figure 12 indicates that parasitic series resistance degrades the admittance charecteristics of the designed SiC based DDRs. The degradation of admittance charecteristics due to the presence of RS is more serious in 6H-

Thus the present study definitely establishes that the prospects of 4H-SiC based IMPATT as a high power, efficient THz source is far better than its 6H-SiC and 3C-SiC counterparts. The effects of optical illumination on the THz behavior of the designed diodes are shown in Table 4. The computed values of –GP, -ZRP, PRF, fP and –QP for different electron and hole current multiplication factors are shown in Table 4. Admittance plots of all the SiC DDR IMPATTs under optical illumination are shown in Figure 13. It is evident from the Figure as well as from Table 4 that the values of │-GP│ at the optimum frequencies decrease with the lowering of Mn and Mp. At the same time, the frequency ranges over which the devices exhibit negative conductance, shift towards higher frequencies with the

DDR 6H-SiC DDR 3C-SiC DDR

2.35 1.30 1.60

based device is found to be best among all the designed diodes.

Diode parameters 4H-SiC

Peak electric field (Em)

Peak negative conductance

Device negative resistance at peak frequency (-ZRP) (10-9Ωm2)

RF output power density (PRF)

Frequency

SiC based IMPATT.

lowering of Mn and Mp. The simulated results for the illuminated 4H-SiC IMPATT diode (Table 4) indicate that the values of │-GP│and PRF decrease nearly by the same percentage (~3.8.0 %) as Mn decreases from 106 to 25, keeping the value of MP constant at 106. On the other hand, for a similar variation of Mp from 106 to 10 (keeping the value of Mn constant at 106), reduces the values of │-GP│ and PRF by ~ 35.0 %. Similar trend is reflected in Figure 13. In the case of 6H-SiC and 3C-SiC IMPATT diodes, as the value of Mn decreases from 106 to 25 (corresponding to TM illumination configuration), │-GP│ and PRF decrease by 5.2 % and 2% , respectively (Table 4). Similarly for the FC illumination configuration of 6H-SiC and 3C-SiC IMPATTs, the decrease of MP from 106 to 25, causes a reduction in the values of │-GP│ and PRF by 18.2% and 21.6 %, respectively.


Table 3. Series resistance of the designed diodes at THz regime (frequency = 0.3 THz)

It is evident from Table 4, that in the case of 4H-SiC based TM diode, a lowering of Mn from 106 to 25 causes the diode negative resistance (-ZRp) to decrease by 3.0%, while there is a corresponding lowering of -ZRp by 51.0 % in case of FC diode. Similarly for TM 6H-SiC and 3C-SiC diodes, the value of -ZRp reduces by 20.0% and 10.0 %, respectively, as Mn decreases from 106 to 25 (Table 4), whereas, for a similar variation of MP, │-ZRp│ reduces by 40.0 % and 25.0 % , respectively in 6H-SiC and 3C-SiC based diodes. The device quality factors are also found to degrade with the lowering of Mn and MP, in case of all the SiC based diodes.

It is further evident from figure 13 that in α-SiC and β-SiC based DDR devices a lowering of Mp causes larger upward shift in frequency than corresponding lowering of Mn. The optimum frequency of oscillation (fP) shifts upwards by 15.0 GHz and 45.0GHz respectively for TM and FC IMPATTs based on 4H-SiC. On the other hand, in case of 6H-SiC and 3C-SiC based TM IMPATTs, fP shifts upward by 1.0 GHz and 2.0 GHz, respectively. Whereas under FC illumination configuration, the values of fP shifts upward by 3.0 GHz and 7.0 GHz respectively for 6H-SiC and 3C-SiC based diodes.

Thus the studies reveal that, effects of photo-illumination on the frequency up shift as well as on the modulation of the THz behavior of the SiC devices are found to be more pronounced in FC illumination configuration than that for TM illumination configuration under similar operating condition. These results show an identical trend as observed previously for MM-wave SiC devices [4].

SiC Devices on Different Polytypes: Prospects and Challenges 361

their locations shift towards the nn++ and pp++ edges of the drift layer with the decrease of Mn or MP. The depression of the peaks and the shift of the R(x) profiles are less pronounced in TM diode structure (Figure 14) while the same are more pronounced in FC diode structure (Figure 14). The optical illumination studies on the three types of SiC DDRs thus reveals that 4H-SiC based IMPATT is comparatively more photo- sensitive than its

Fig. 8. The stacking sequence of double layers of the three most common SiC polytypes

counterparts.


Table 4. Optical illumination effects on SiC IMPATTs in the THz regime (frequency = 0.3 THz).

To study the microscopic properties of the devices the authors has computed the spatial distribution of negative resistivity (R(x)) in the depletion layer of the device, which would give an insight into the region of depletion layer that contribute to THz power. The computed R(x) profiles at the respective optimum frequencies fP of the DDRs are shown in figure 14. The computed R(x) profiles in each case is characterized by two negative resistivity peaks (Rmax) in the middle of each drift layer along with a central negative resistivity minimum (Rmin) located near the metallurgical junction. Furthermore, the magnitude of negative resistivity peaks produced by holes in the hole drift layer (Rmax,,p) is appreciably higher compared with the peak produced by electrons in the electron drift layer (Rmax, n). This may be due to the fact that in SiC, hole is more ionizing carrier than electrons [5]. The study also reveals that under normal operating conditions of the DDRs, the values of Rmin is appreciable, which indicates that the central avalanche zone of all the DDR SiC IMPATTs contributes appreciably to the THz power although to a lesser extent than the corresponding drift zone. Some qualitative estimate of the THz-power contributed by the avalanche region of the devices can be obtained from the contribution of the smallsignal negative resistance (-ZR) from the avalanche zone, which is the area of the R(x) profile bounded by avalanche zone. The dependence of magnitudes of the negative resistivity peaks in the two drift layers of 4H-SiC, 6H-SiC and 3C-SiC diodes can be explained by considering the relative magnitudes of the ionization rates of electrons and holes in the avalanche zone. It is observed that in case of all the diodes, under optical illumination, the magnitudes of the peaks of the negative resistivity profiles decrease and

**-Gp (108 Sm-2)** 

**-ZRp (10-9 Ωm2)** 

**PRF (1010 Wm-2)** 

**-QP** 

**fp (THz)** 

(unilluminated) 106 106 0.325 162.0 2.35 36.90 1.26 4H-SiC (TM) 50 106 0.33 158.0 2.30 35.99 1.32 4H-SiC (TM) 25 ,, 0.34 156.0 2.29 35.54 1.34 4H-SiC (FC) 106 50 0.35 135.0 1.48 30.75 2.0 4H-SiC (FC) 106 25 0.37 120.0 1.15 27.33 2.50

(unilluminated) 106 106 0.35 101.0 1.30 14.99 2.58 6H-SiC (TM) 50 106 0.353 99.0 1.12 14.70 2.84 6H-SiC (TM) 25 ,, 0.36 96.0 1.03 14.25 2.91 6H-SiC (FC) 106 50 0.362 90.0 0.84 13.36 3.50 6H-SiC (FC) 106 25 0.38 85.4 0.76 12.68 3.80

(unilluminated) 106 106 0.353 152.0 1.60 22.99 1.80 3C-SiC (TM) 50 106 0.353 150.0 1.50 22.68 1.90 3C-SiC (TM) 25 ,, 0.355 149.0 1.44 22.53 1.94 3C-SiC (FC) 106 50 0.356 138.6 1.44 20.96 2.0 3C-SiC (FC) 106 25 0.360 125.0 1.20 18.91 2.32

Table 4. Optical illumination effects on SiC IMPATTs in the THz regime (frequency = 0.3

To study the microscopic properties of the devices the authors has computed the spatial distribution of negative resistivity (R(x)) in the depletion layer of the device, which would give an insight into the region of depletion layer that contribute to THz power. The computed R(x) profiles at the respective optimum frequencies fP of the DDRs are shown in figure 14. The computed R(x) profiles in each case is characterized by two negative resistivity peaks (Rmax) in the middle of each drift layer along with a central negative resistivity minimum (Rmin) located near the metallurgical junction. Furthermore, the magnitude of negative resistivity peaks produced by holes in the hole drift layer (Rmax,,p) is appreciably higher compared with the peak produced by electrons in the electron drift layer (Rmax, n). This may be due to the fact that in SiC, hole is more ionizing carrier than electrons [5]. The study also reveals that under normal operating conditions of the DDRs, the values of Rmin is appreciable, which indicates that the central avalanche zone of all the DDR SiC IMPATTs contributes appreciably to the THz power although to a lesser extent than the corresponding drift zone. Some qualitative estimate of the THz-power contributed by the avalanche region of the devices can be obtained from the contribution of the smallsignal negative resistance (-ZR) from the avalanche zone, which is the area of the R(x) profile bounded by avalanche zone. The dependence of magnitudes of the negative resistivity peaks in the two drift layers of 4H-SiC, 6H-SiC and 3C-SiC diodes can be explained by considering the relative magnitudes of the ionization rates of electrons and holes in the avalanche zone. It is observed that in case of all the diodes, under optical illumination, the magnitudes of the peaks of the negative resistivity profiles decrease and

**DDR diode type Mn Mp** 

4H-SiC

6H-SiC

3C-SiC

THz).

their locations shift towards the nn++ and pp++ edges of the drift layer with the decrease of Mn or MP. The depression of the peaks and the shift of the R(x) profiles are less pronounced in TM diode structure (Figure 14) while the same are more pronounced in FC diode structure (Figure 14). The optical illumination studies on the three types of SiC DDRs thus reveals that 4H-SiC based IMPATT is comparatively more photo- sensitive than its counterparts.

Fig. 8. The stacking sequence of double layers of the three most common SiC polytypes

SiC Devices on Different Polytypes: Prospects and Challenges 363

Fig. 10. (b) Electric field and normalised current density profiles of 6H-SiC IMPATT at THz region. E1 and P1 are un-illuminated diodes and E2,3 and P2,3 are illuminated TM (2) and

Fig. 10. (c) Electric field and normalised current density profiles of 3C-SiC IMPATT at THz region. E1 and P1 are un-illuminated diodes and E2,3 and P2,3 are illuminated TM (2) and FC

FC (3) diodes.

(3) diodes.

Fig. 9. (a) Top-Mounted IMPATT and (b) Flip-Chip IMPATT

Fig. 10. (a) Electric field and normalised current density profiles of 4H-SiC IMPATT at THz region. E1 and P1 are un-illuminated diodes and E2,3 and P2,3 are illuminated TM (2) and FC (3) diodes.

Fig. 10. (a) Electric field and normalised current density profiles of 4H-SiC IMPATT at THz region. E1 and P1 are un-illuminated diodes and E2,3 and P2,3 are illuminated TM (2) and FC

Fig. 9. (a) Top-Mounted IMPATT and (b) Flip-Chip IMPATT

(3) diodes.

Fig. 10. (b) Electric field and normalised current density profiles of 6H-SiC IMPATT at THz region. E1 and P1 are un-illuminated diodes and E2,3 and P2,3 are illuminated TM (2) and FC (3) diodes.

Fig. 10. (c) Electric field and normalised current density profiles of 3C-SiC IMPATT at THz region. E1 and P1 are un-illuminated diodes and E2,3 and P2,3 are illuminated TM (2) and FC (3) diodes.

SiC Devices on Different Polytypes: Prospects and Challenges 365

Fig. 12. Admittance plots of SiC DDR IMPATTs in the Terahertz region, dotted lines are

incorporating the series resistance effects.

Fig. 11. Plots of breakdown voltage as a function of electron and hole current multiplication factors of SiC THz IMPATTs.

Fig. 11. Plots of breakdown voltage as a function of electron and hole current multiplication

factors of SiC THz IMPATTs.

Fig. 12. Admittance plots of SiC DDR IMPATTs in the Terahertz region, dotted lines are incorporating the series resistance effects.

SiC Devices on Different Polytypes: Prospects and Challenges 367

Fig. 14. Effect of radiation on negative resistivity plots of SiC DDR IMPATTs in the

Terahertz region.

Fig. 13. Effect of photo-illumination on FC and TM illumination configuration of SiC (3C, 4H and 6H type) based terahertz DDR IMPATT diodes at room temperature (300K).

Fig. 13. Effect of photo-illumination on FC and TM illumination configuration of SiC (3C, 4H

and 6H type) based terahertz DDR IMPATT diodes at room temperature (300K).

Fig. 14. Effect of radiation on negative resistivity plots of SiC DDR IMPATTs in the Terahertz region.

**15** 

 *Brazil* 

**Recent Developments on Silicon Carbide Thin** 

**Films for Piezoresistive Sensors Applications** 

The increasing demand for microelectromechanical systems (MEMS) as, for example, piezoresistive sensors with capabilities of operating at high temperatures, mainly for automotive, petrochemical and aerospace applications, has stimulated the research of alternative materials to silicon in the fabrication of these devices. It is known that the high temperature operating limit for silicon-based MEMS sensors is about 150ºC (Fraga,

Silicon carbide (SiC) has shown to be a good alternative to silicon in the development of MEMS sensors for harsh environments due to its excellent electrical characteristics as wide band-gap (3 eV), high breakdown field strength (10 times higher than Si) and low intrinsic carrier concentration which allow stable electronic properties under harsh environments (Cimalla et al., 2007; Wright & Horsfall, 2007; Rajab et al., 2006). In addition, SiC exhibits high elastic modulus at high temperatures which combined with the excellent electronic properties make it very attractive for piezoresistive sensors applications (Kulikovsky et al.,

Silicon carbide can be obtained in bulk or film forms. In recent years, great progress has been made in the field of the growth of SiC bulk. Currently there are 6H-SiC, 4H-SiC and 3C-SiC wafers commercially available. However, these wafers are still very expensive (Hobgood et al., 2004; Camassel & Juillaguet, 2007), so encouraging studies on crystalline and amorphous SiC films deposited on silicon or SOI (Silicon-On-Insulator) substrates using appropriate techniques. The use of SiC films besides being less expensive has another advantage which is the well known processing techniques for silicon micromachining. The challenge is to obtain SiC films with mechanical, electrical and piezoresistive properties as

Nowadays, some research groups have studied the synthesis and characterization of SiC films obtained by different techniques namely, plasma enhanced chemical vapour deposition (PECVD), molecular beam epitaxy (MBE), sputtering, among others, aiming MEMS sensors applications (Chaudhuri et al., 2000; Fissel et al., 1995; Rajagopalan et al.,

**1. Introduction** 

2009).

2008).

good as the bulk form.

2003; Lattemann et al., 2003).

Mariana Amorim Fraga1,2, Rodrigo Sávio Pessoa2,3,

*2Plasma and Processes Laboratory, Technological Institute of Aeronautics* 

Homero Santiago Maciel2 and Marcos Massi2

*1Institute for Advanced Studies* 

*3 IP&D, University of Vale do Paraiba* 

#### **5. Acknowledgment(s)**

The author wish to acknowledge Director, CMSDS, University of Calcutta for providing necessary support to do this work. The author is grateful to (late) Prof. S. K. Roy, Professor and founder Director, CMSDS, University of Calcutta and (late) Prof. Nilratan Mazumder, Professor, West Bengal University of Technology, India for providing their valuable suggestions during this work.

#### **6. References**


### **Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications**

Mariana Amorim Fraga1,2, Rodrigo Sávio Pessoa2,3, Homero Santiago Maciel2 and Marcos Massi2 *1Institute for Advanced Studies 2Plasma and Processes Laboratory, Technological Institute of Aeronautics 3 IP&D, University of Vale do Paraiba Brazil* 

#### **1. Introduction**

368 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

The author wish to acknowledge Director, CMSDS, University of Calcutta for providing necessary support to do this work. The author is grateful to (late) Prof. S. K. Roy, Professor and founder Director, CMSDS, University of Calcutta and (late) Prof. Nilratan Mazumder, Professor, West Bengal University of Technology, India for providing their valuable

[1] Mukherjee, M; Mazumder, N; Roy, S.K; and Goswami, K ; "GaN IMPATT diode: A

[2] Gummel, H. K; Blue, J.L ; 'A small signal theory of avalanche noise in IMPATT Diodes',

[3] Adlerstein, M. G, Holway, Chu, S. L. "Measurement of series resistance in IMPATT

[4] Mukherjee, M; Mazumder, N; Dasgupta, A; "Effect of Charge Bump on Series

[5] Electronic Archive: New Semiconductor Materials, Characteristics and Properties

Technology, (2007), vol. 22, pp. 1258-1267.

IEEE Trans. Electron Devices (1967), vol. 14, p. 569.

(Online) www.ioffe.ru/SVA/NSM/Semicond/SiC.

diodes", IEEE Trans. Electron Devices, vol. ED- 30, p. 179,.

photosensitive high power Terahertz source", Semiconductor Science and

Resistance and Ka-Band performance of 4H-SiC IMPATT Oscillator", Proceedings of "IEEE International symposium on Integrated circuits" (IEEE-ISIC 2007), September 26-28, (2007), Nanyang Technological University, and IEEE- Singapore.

**5. Acknowledgment(s)** 

suggestions during this work.

pp. 61-64

**6. References** 

The increasing demand for microelectromechanical systems (MEMS) as, for example, piezoresistive sensors with capabilities of operating at high temperatures, mainly for automotive, petrochemical and aerospace applications, has stimulated the research of alternative materials to silicon in the fabrication of these devices. It is known that the high temperature operating limit for silicon-based MEMS sensors is about 150ºC (Fraga, 2009).

Silicon carbide (SiC) has shown to be a good alternative to silicon in the development of MEMS sensors for harsh environments due to its excellent electrical characteristics as wide band-gap (3 eV), high breakdown field strength (10 times higher than Si) and low intrinsic carrier concentration which allow stable electronic properties under harsh environments (Cimalla et al., 2007; Wright & Horsfall, 2007; Rajab et al., 2006). In addition, SiC exhibits high elastic modulus at high temperatures which combined with the excellent electronic properties make it very attractive for piezoresistive sensors applications (Kulikovsky et al., 2008).

Silicon carbide can be obtained in bulk or film forms. In recent years, great progress has been made in the field of the growth of SiC bulk. Currently there are 6H-SiC, 4H-SiC and 3C-SiC wafers commercially available. However, these wafers are still very expensive (Hobgood et al., 2004; Camassel & Juillaguet, 2007), so encouraging studies on crystalline and amorphous SiC films deposited on silicon or SOI (Silicon-On-Insulator) substrates using appropriate techniques. The use of SiC films besides being less expensive has another advantage which is the well known processing techniques for silicon micromachining. The challenge is to obtain SiC films with mechanical, electrical and piezoresistive properties as good as the bulk form.

Nowadays, some research groups have studied the synthesis and characterization of SiC films obtained by different techniques namely, plasma enhanced chemical vapour deposition (PECVD), molecular beam epitaxy (MBE), sputtering, among others, aiming MEMS sensors applications (Chaudhuri et al., 2000; Fissel et al., 1995; Rajagopalan et al., 2003; Lattemann et al., 2003).

Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications 371

In 2002, Toriyama & Sugiyama performed a theoretical analysis on the piezoresistivity of β-SiC based on electron transfer and the mobility shift mechanism and in 2004 a detailed experimental study on piezoresistive properties of single crystalline, polycrystalline, and

In parallel to these studies on characterization of piezoresistive properties of the SiC polytypes, some SiC sensors have been developed. In the ´90s, Okoije et al. developed 6H-SiC pressure sensors for high temperature applications and Ziermann et al. reported a piezoresistive pressure sensor with n-type β-SiC thin-film piezoresistors on Silicon-on-Insulators (SOI) substrate. In 2003, Atwell et al. simulated, fabricated and tested bulk

The good performance exhibited by the sensors based on 6H-SiC bulk and on 3C-SiC film have motivated studies on the piezoresistive properties of amorphous SiC (a-SiC) films produced at low temperatures by techniques such as PECVD and magnetron sputtering

Table 1 presents the *GF* and *TCR* values of different SiC types and of some other materials commonly used in piezoresistive sensors. As can be observed, the p-type Si has the greater

film Nitrogen Crystalline -31.8 400

**(ppm/ºC)** 

**Material Form Dopant Structure GF \* TCR** 

p-type Si bulk Boron Crystalline 140 1082 n-type Si bulk Phosphorus Crystalline -133 1920 Ge thin film Boron Amorphous 10 3100 Polysilicon thin film Boron Polycristalline 34 100 a-SiC thin film Nitrogen Amorphous 49 36

6H-SiC bulk Nitrogen Crystalline 15 -240

Table 1. Comparison among the properties of some piezoresistive materials reported in

*ij*

ρ

ρ

The piezoresistive effect can also be defined as the tensor relationship between applied

*ij kl*

<sup>Δ</sup> <sup>=</sup> (3)

π σ

where ρ is the resistivity, π is the piezoresistive coefficient and σ is the mechanical stress. In the case of a material with cubic structure, the stress has six components σ1, σ2, and σ<sup>3</sup> (along the axes of the cube) and σ4, σ5, and σ6 (the shear stresses) as shown in Figure 1.

nanocrystalline n-type 3C-SiC was reported by Eickhoff et al.

micromachined 6H-SiC piezoresistive accelerometers.

(Fraga, 2010, 2011a; Fraga et al., 2011b, 2011c).

*GF* whereas the a-SiC film the smaller *TCR*.

literature (Fraga, 2011c; Shor, 1993; Okoije 1998a).

stress and change in resistivity (Johns, 2005):

3C-SiC thick

\* GF measured at room temperature

**2.2 Physical description** 

The purpose of this chapter is to present an overview of the deposition techniques of SiC films, summarizing the deposition conditions that affect the piezoresistive properties of these films, the influence of the temperature on their piezoresistive properties and comparing the performance of piezoresistive sensors based on SiC films with those based in other materials. Moreover, the chapter focus attention is on the development of pressure sensors and accelerometers based on SiC films with suited piezoresistive properties to substitute the silicon in the microfabrication of these sensors so as to extend their endurance under harsh environment.

#### **2. Piezoresistive effect in SiC**

#### **2.1 Brief overview**

Piezoresistivity is a physical property which has been widely used to convert a mechanical signal into an electrical one, in different device types such as pressure sensors, accelerometers, tactile sensors, strain gauges and flow sensors, among others.

The piezoresistive effect was discovered by Lord Kelvin in 1856. This property is quantified in terms of gauge factor (*GF*), which is defined as the fractional change in the resistance per unit strain (Window, 1992):

$$GF = \frac{\Delta R}{R} \frac{1}{\varepsilon} \tag{1}$$

where *R* is the nominal electrical resistance and ε the strain. *GF* is a dimensionless number that depends on the crystallographic orientation and is related to the elastic or Young's modulus of the material (*E*) by the following expression,

$$E = \frac{\sigma}{\varepsilon} \tag{2}$$

where σ is the mechanical stress. A positive *GF* indicates an increase in resistance with stress increases whereas the negative correspond a decrease.

Thus, from eq. (1) and (2), the piezoresistivity can be defined as the fractional change in the resistivity of a material when submitted to a mechanical stress. The change in resistance arises from two effects: the change in the dimension of the resistor and the change in the resistivity of the material itself.

The large piezoresistive effect in silicon and germanium was first observed by Smith in 1954. Since then, it has been noted that the piezoresistive effect in semiconductor materials is highly anisotropic and exhibits a dependence on the dopant type, dopant concentration and crystalline orientation. Furthermore, in 1956 Morin et al. demonstrated the temperature dependence of the piezoresistance of silicon and germanium.

In 1968, Rapatskaya et al. were the first to report the piezoresistive properties of n-type α-SiC (6H-SiC). In the 70's three papers on piezoresistance in SiC were published by Guk: two on the piezoresistive characterization and temperature depence of the 6H-SiC polytype and one on the piezoresistance of β-SiC (3C-SiC). In 1993, Shor et al. have extended this study on piezoresistive properties of β-SiC discussing the GF and the temperature coefficient of resistance (*TCR*) of this material for several doping levels. In 1997, Strass et al. investigated the influence of crystal quality on the piezoresistive effect in β-SiC. In 1998, Okojie et al. determined the longitudinal and transverse *GF* and the *TCR* of n- and p-type 6H-SiC.

In 2002, Toriyama & Sugiyama performed a theoretical analysis on the piezoresistivity of β-SiC based on electron transfer and the mobility shift mechanism and in 2004 a detailed experimental study on piezoresistive properties of single crystalline, polycrystalline, and nanocrystalline n-type 3C-SiC was reported by Eickhoff et al.

In parallel to these studies on characterization of piezoresistive properties of the SiC polytypes, some SiC sensors have been developed. In the ´90s, Okoije et al. developed 6H-SiC pressure sensors for high temperature applications and Ziermann et al. reported a piezoresistive pressure sensor with n-type β-SiC thin-film piezoresistors on Silicon-on-Insulators (SOI) substrate. In 2003, Atwell et al. simulated, fabricated and tested bulk micromachined 6H-SiC piezoresistive accelerometers.

The good performance exhibited by the sensors based on 6H-SiC bulk and on 3C-SiC film have motivated studies on the piezoresistive properties of amorphous SiC (a-SiC) films produced at low temperatures by techniques such as PECVD and magnetron sputtering (Fraga, 2010, 2011a; Fraga et al., 2011b, 2011c).

Table 1 presents the *GF* and *TCR* values of different SiC types and of some other materials commonly used in piezoresistive sensors. As can be observed, the p-type Si has the greater *GF* whereas the a-SiC film the smaller *TCR*.


\* GF measured at room temperature

370 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

The purpose of this chapter is to present an overview of the deposition techniques of SiC films, summarizing the deposition conditions that affect the piezoresistive properties of these films, the influence of the temperature on their piezoresistive properties and comparing the performance of piezoresistive sensors based on SiC films with those based in other materials. Moreover, the chapter focus attention is on the development of pressure sensors and accelerometers based on SiC films with suited piezoresistive properties to substitute the silicon in the microfabrication of these sensors so as to extend their endurance

Piezoresistivity is a physical property which has been widely used to convert a mechanical signal into an electrical one, in different device types such as pressure sensors,

The piezoresistive effect was discovered by Lord Kelvin in 1856. This property is quantified in terms of gauge factor (*GF*), which is defined as the fractional change in the resistance per

> *<sup>R</sup>* <sup>1</sup> *GF R* ε

where *R* is the nominal electrical resistance and ε the strain. *GF* is a dimensionless number that depends on the crystallographic orientation and is related to the elastic or Young's

> *E* σ

> > ε

where σ is the mechanical stress. A positive *GF* indicates an increase in resistance with stress

Thus, from eq. (1) and (2), the piezoresistivity can be defined as the fractional change in the resistivity of a material when submitted to a mechanical stress. The change in resistance arises from two effects: the change in the dimension of the resistor and the change in the

The large piezoresistive effect in silicon and germanium was first observed by Smith in 1954. Since then, it has been noted that the piezoresistive effect in semiconductor materials is highly anisotropic and exhibits a dependence on the dopant type, dopant concentration and crystalline orientation. Furthermore, in 1956 Morin et al. demonstrated the temperature

In 1968, Rapatskaya et al. were the first to report the piezoresistive properties of n-type α-SiC (6H-SiC). In the 70's three papers on piezoresistance in SiC were published by Guk: two on the piezoresistive characterization and temperature depence of the 6H-SiC polytype and one on the piezoresistance of β-SiC (3C-SiC). In 1993, Shor et al. have extended this study on piezoresistive properties of β-SiC discussing the GF and the temperature coefficient of resistance (*TCR*) of this material for several doping levels. In 1997, Strass et al. investigated the influence of crystal quality on the piezoresistive effect in β-SiC. In 1998, Okojie et al.

determined the longitudinal and transverse *GF* and the *TCR* of n- and p-type 6H-SiC.

<sup>Δ</sup> <sup>=</sup> (1)

<sup>=</sup> (2)

accelerometers, tactile sensors, strain gauges and flow sensors, among others.

modulus of the material (*E*) by the following expression,

increases whereas the negative correspond a decrease.

dependence of the piezoresistance of silicon and germanium.

under harsh environment.

unit strain (Window, 1992):

resistivity of the material itself.

**2.1 Brief overview** 

**2. Piezoresistive effect in SiC** 

Table 1. Comparison among the properties of some piezoresistive materials reported in literature (Fraga, 2011c; Shor, 1993; Okoije 1998a).

#### **2.2 Physical description**

The piezoresistive effect can also be defined as the tensor relationship between applied stress and change in resistivity (Johns, 2005):

$$\frac{\Delta\rho\_{ij}}{\rho} = \pi\_{ij}\sigma\_{kl} \tag{3}$$

where ρ is the resistivity, π is the piezoresistive coefficient and σ is the mechanical stress. In the case of a material with cubic structure, the stress has six components σ1, σ2, and σ<sup>3</sup> (along the axes of the cube) and σ4, σ5, and σ6 (the shear stresses) as shown in Figure 1.

Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications 373

where *L* is the length and *A* the cross-sectional area of the resistor. When the resistor is subjected to a longitudinal stress, the resistivity, cross-sectional area and length will be

> *R LA R LA* ρ

This resistor that changes its resistivity with an applied stress is called piezoresistor. As can

and dimensions change *<sup>L</sup>*

depends on the material type. In your experiments, Smith observed that for silicon the change in resistivity gives a larger contribution to the resistance changes than the change in

Considering that the components associated with dimension change can be written as a

*L L*

*AWH AWH*

In the above equations the fractional change in length is equal to the longitudinal strain

Considering also that the longitudinal and transverse strain are related through equation:

where ν is the Poisson´s ratio of the material, thus the equation (7) can be simplified to

ρ ε

Thus, the gauge factor can be related to resistivity, longitudinal strain and Poisson's ratio by

<sup>1</sup> ( ) 1 2

ρ

*l GF v* ρ

Δ

Another important parameter to evaluate the piezoresistive effect is the temperature, whose influence on strain measurement cannot be neglected. When the ambient temperature

ε ρ

*R*

Δ Δ

*R*

*t l* ε νε

( ) 1 2 *<sup>l</sup>*

*v*

ΔΔ Δ

whereas the change in area is the sum of change in width *<sup>W</sup>*

*l*

2 *<sup>t</sup>*

ε

ε

=++ (7)

<sup>Δ</sup> <sup>=</sup> (8)

= += (9)

*W* Δ

= (10)

=++ (11)

= ++ (12)

and height *<sup>H</sup>*

*H* Δ . It is

, can be influenced by two

. The dominant factor

*R* Δ

*L* Δ , *<sup>A</sup> A* Δ

Δ ΔΔΔ

ρ

be observed in equation (7), a fractional resistance change, *<sup>R</sup>*

ρ

 Δ 

ρ

changed as shown in the equation below:

dimensions of the resistor (Smith, 1954).

function of the strain, we have:

known that *wHt*

εεε= = .

the following equation (Allameh et al., 2006):

and

factors: resistivity change

Fig. 1. Schematic illustration of the stress components.

The six stress components and six resistivity components result in a matrix with 36 piezoresistive coefficients. For the cubic crystal structure of materials such as silicon or β-SiC, the matrix simplifies to only three piezoresistive coefficients (π11, π12, and π44) as shown in the following equation (Singh et al., 2002):

$$
\begin{bmatrix}
\Delta\boldsymbol{\rho}\_{I} \\
\Delta\boldsymbol{\rho}\_{2} \\
\boldsymbol{\rho}\_{3} \\
\boldsymbol{\Delta\boldsymbol{\rho}}\_{4} \\
\Delta\boldsymbol{\rho}\_{5} \\
\Delta\boldsymbol{\rho}\_{6}
\end{bmatrix} = \begin{bmatrix}
\pi\_{II} & \pi\_{12} & \pi\_{12} & 0 & 0 & 0 \\
\pi\_{12} & \pi\_{II} & \pi\_{12} & 0 & 0 & 0 \\
\pi\_{12} & \pi\_{12} & \pi\_{11} & 0 & 0 & 0 \\
0 & 0 & 0 & \pi\_{44} & 0 & 0 & \boldsymbol{\sigma}\_{4} \\
0 & 0 & 0 & 0 & \pi\_{44} & 0 & \boldsymbol{\sigma}\_{5} \\
0 & 0 & 0 & 0 & 0 & \pi\_{44} & \boldsymbol{\rho}\_{6}
\end{bmatrix} \tag{4}
$$

Although equation (4) models the piezoresistive effect in silicon in the direction of the crystal axes, customarily this effect is measured using only two coefficients: πl that relates the resistance change due to stress in the longitudinal direction and πt in the transverse direction. Therefore, the total resistivity change of a material can be simplified considering only changes under longitudinal and transverse stress components,

$$\frac{\Delta\rho}{\rho} = \pi\_l \sigma\_l + \pi\_t \sigma\_t \tag{5}$$

The piezoresistive effect can be better understood by the analysis of the behavior of a resistor when submitted to a mechanical stress. It is known that the electrical resistance of an unstressed resistor is given by,

$$R = \rho \frac{L}{A} \tag{6}$$

where *L* is the length and *A* the cross-sectional area of the resistor. When the resistor is subjected to a longitudinal stress, the resistivity, cross-sectional area and length will be changed as shown in the equation below:

$$\frac{\Delta R}{R} = \frac{\Delta \rho}{\rho} + \frac{\Delta L}{L} + \frac{\Delta A}{A} \tag{7}$$

This resistor that changes its resistivity with an applied stress is called piezoresistor. As can be observed in equation (7), a fractional resistance change, *<sup>R</sup> R* Δ , can be influenced by two

factors: resistivity change ρ ρ Δ and dimensions change *<sup>L</sup> L* Δ , *<sup>A</sup> A* Δ . The dominant factor

depends on the material type. In your experiments, Smith observed that for silicon the change in resistivity gives a larger contribution to the resistance changes than the change in dimensions of the resistor (Smith, 1954).

Considering that the components associated with dimension change can be written as a function of the strain, we have:

$$\frac{\Delta L}{L} = \mathcal{E}\_l \tag{8}$$

and

372 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

The six stress components and six resistivity components result in a matrix with 36 piezoresistive coefficients. For the cubic crystal structure of materials such as silicon or β-SiC, the matrix simplifies to only three piezoresistive coefficients (π11, π12, and π44) as shown

> *12 12 11 12 11 12 11 12 12*

π

ρ

Δ

ρ

π

π

*0 0 0 0 0 0 0 0 0 0 0 0 0 0 0*

Although equation (4) models the piezoresistive effect in silicon in the direction of the crystal axes, customarily this effect is measured using only two coefficients: πl that relates the resistance change due to stress in the longitudinal direction and πt in the transverse direction. Therefore, the total resistivity change of a material can be simplified considering

*ll tt*

πσ πσ

The piezoresistive effect can be better understood by the analysis of the behavior of a resistor when submitted to a mechanical stress. It is known that the electrical resistance of an

> *<sup>L</sup> <sup>R</sup> <sup>A</sup>* <sup>=</sup> ρ

π

π

π

 

 

σ

σ

σ

σ

σ

σ

 

*44*

π

*44*

π

*0 0 0 0 0 0 0 0 0*

*44*

π

(4)

(6)

= + (5)

Fig. 1. Schematic illustration of the stress components.

 

 

 

 

*1*

ρ

unstressed resistor is given by,

Δρ

Δρ

Δρ

Δρ

Δρ

Δρ

π

π

π

=

only changes under longitudinal and transverse stress components,

in the following equation (Singh et al., 2002):

$$\frac{\Delta A}{A} = \frac{\Delta W}{W} + \frac{\Delta H}{H} = 2\varepsilon\_t \tag{9}$$

In the above equations the fractional change in length is equal to the longitudinal strain whereas the change in area is the sum of change in width *<sup>W</sup> W* Δ and height *<sup>H</sup> H* Δ . It is known that *wHt* εεε= = .

Considering also that the longitudinal and transverse strain are related through equation:

$$
\mathfrak{e}\_t = \mathsf{V} \mathfrak{e}\_l \tag{10}
$$

where ν is the Poisson´s ratio of the material, thus the equation (7) can be simplified to

$$\frac{\Delta R}{R} = \frac{\Delta \rho}{\rho} + \varepsilon\_l (1 + 2v) \tag{11}$$

Thus, the gauge factor can be related to resistivity, longitudinal strain and Poisson's ratio by the following equation (Allameh et al., 2006):

$$GF = \frac{1}{\varepsilon\_{\parallel}} \frac{\Delta \rho}{\rho} + (1 + 2v) \tag{12}$$

Another important parameter to evaluate the piezoresistive effect is the temperature, whose influence on strain measurement cannot be neglected. When the ambient temperature

Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications 375

When subjected to a mechanical stress, the electrical resistance of the resistors change

Whereas the four resistors have the same nominal resistance value (*R*1=*R*2=*R*3=*R*4) and that under mechanical stress the resistances R2 and R3 increases their values in +∆*R*, the resistances *R*1 and *R*4 decreases their values in -∆*R*. Therefore, the equation (16) can be

2 2

*<sup>R</sup> <sup>V</sup> <sup>S</sup>*

*V R R R R R V R RR*

*out s*

Given this, the sensitivity of a piezoresistive pressure sensor is determined by

( ) ( ) ( )( ) 3 3 4 4 11 33 2 2 4 4

( )

1 1 *out s*

1 1 *out s*

*Rg V g*

As shown in the previous section, in recent years many researchers have been reported on the piezoresistive characterization of different SiC polytypes aiming the applicability of these materials in sensors. When comparing these studies, it is observed that for a same SiC polytype a dispersion of different values can be obtained for piezoresistive coefficient, *GF*

It is known that the SiC has about 200 polytypes with different physical properties. This is one of the difficulties in characterizing the piezoresistivity in SiC. Moreover, studies show that maximum value of *GF* for SiC at room temperature is between 30 at 49 while for the monocrystalline p-type Si is 140 (see Table 1). However, all studies published until now have demonstrated the potential of the 6H-SiC and 3C-SiC polytypes besides a-SiC for the development of piezoresistive sensors for high temperature application. Given this, it is important to evaluate when it is advantageous to use SiC in piezoresistive sensors and

Several studies show that the SiC has mechanical and chemical stability at high temperatures. Due to these characteristics the application of SiC sensors is always associated with harsh environments. In these environments, silicon has mechanical and chemical limitations. At temperature greater than 500ºC, silicon deforms plastically under small loads

*RP V P*

Whereas, for a piezoresistive accelerometer, the sensitivity is defined as the electrical output

*<sup>R</sup> <sup>V</sup> <sup>S</sup>*

**3. When and why to use SiC films in piezoresistive sensors?** 

<sup>Δ</sup> + Δ − Δ <sup>Δ</sup> =− = (17)

<sup>Δ</sup> <sup>Δ</sup> = = Δ Δ (18)

<sup>Δ</sup> <sup>Δ</sup> = = (19)

<sup>Δ</sup> + Δ + Δ = − +Δ + +Δ +Δ + +Δ (16)

leading to a variation of the output voltage, according to the following relationship

*V RR R R V RR RR R R R R*

*out s*

simplified to

where ∆*P* is change in pressure.

per unit of applied acceleration:

and *TCR* (Okoije, 2002).

where *g* is the acceleration of gravity.

whether is better to use SiC in bulk or thin film form.

This analysis should begin with the following question: Why SiC?

changes, the electrical resistance of the resistor changes *T R R* <sup>Δ</sup> Δ . This influence is measured

through temperature coefficient of resistance (*TCR*) and temperature coefficient of gauge factor (*TCGF*) that describe the parts per million change in resistance (or *GF*) for every one degree change in temperature. These coefficients can be determined by

$$TCR = \frac{1}{\Delta T} \frac{\Delta R}{R} = \frac{1}{\Delta T} \frac{R\_T - R\_0}{R\_T} \tag{13}$$

$$T\text{CGF} = \frac{GF\_T - GF\_0}{GF\_0} \frac{1}{\Delta T} \tag{14}$$

where ∆*T* is the change in temperature, *R*0 and *GF*0 are the electrical resistance and the gauge factor measured at room temperature or reference temperature (usually 25ºC), respectively; and *R*T and*GF*T are the electrical resistance and gauge factor measured at an operating temperature. In a first analysis, the sensitivity of a piezoresistive sensor is evaluated in terms of *GF*, *TCR* and *TCGF*, i.e., a sensor with good performance should exhibit high *GF* and low *TCR*. For this, there is great interest by the piezoresistive characterization of materials with low *TCR*. In their study, Shor et al. reported that to reduce the effect of changing in temperature on the performance of a sensor the *TCR* should be positive and preferably constant, the *TCGF* negative and *TCR TCGF* > (Shor et al., 1993).

In respect to the layout of a piezoresistive sensor, in general the most used configuration for the resistors is the Wheatstone bridge. In this configuration, four resistors are connected in loop as shown in Figure 2 and the output voltage is related to the input voltage according to the following equation:

$$\frac{V\_{out}}{V\_s} = \frac{V\_A - V\_B}{V\_s} = \frac{R\_3}{R\_1 + R\_3} - \frac{R\_4}{R\_2 + R\_4} \tag{15}$$

where *V*s is the supply voltage and *V*out is the output voltage.

Fig. 2. Wheatstone bridge configuration.

When subjected to a mechanical stress, the electrical resistance of the resistors change leading to a variation of the output voltage, according to the following relationship

$$\frac{\Delta V\_{out}}{V\_s} = \frac{R\_3 + \Delta R\_3}{\left(R\_1 + \Delta R\_1\right) + \left(R\_3 + \Delta R\_3\right)} - \frac{R\_4 + \Delta R\_4}{\left(R\_2 + \Delta R\_2\right) + \left(R\_4 + \Delta R\_4\right)}\tag{16}$$

Whereas the four resistors have the same nominal resistance value (*R*1=*R*2=*R*3=*R*4) and that under mechanical stress the resistances R2 and R3 increases their values in +∆*R*, the resistances *R*1 and *R*4 decreases their values in -∆*R*. Therefore, the equation (16) can be simplified to

$$\frac{\Delta V\_{out}}{V\_s} = \frac{R + \Delta R}{2R} - \frac{\left(R - \Delta R\right)}{2R} = \frac{\Delta R}{R} \tag{17}$$

Given this, the sensitivity of a piezoresistive pressure sensor is determined by

$$S = \frac{\Delta R}{R} \frac{\text{l}}{\Delta P} = \frac{\Delta V\_{out}}{V\_s} \frac{\text{l}}{\Delta P} \tag{18}$$

where ∆*P* is change in pressure.

374 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

through temperature coefficient of resistance (*TCR*) and temperature coefficient of gauge factor (*TCGF*) that describe the parts per million change in resistance (or *GF*) for every one

> *<sup>R</sup> R R TCR TR T R* <sup>Δ</sup> <sup>−</sup> = = Δ Δ

> > <sup>1</sup> *GF GF <sup>T</sup> TCGF*

where ∆*T* is the change in temperature, *R*0 and *GF*0 are the electrical resistance and the gauge factor measured at room temperature or reference temperature (usually 25ºC), respectively; and *R*T and*GF*T are the electrical resistance and gauge factor measured at an operating temperature. In a first analysis, the sensitivity of a piezoresistive sensor is evaluated in terms of *GF*, *TCR* and *TCGF*, i.e., a sensor with good performance should exhibit high *GF* and low *TCR*. For this, there is great interest by the piezoresistive characterization of materials with low *TCR*. In their study, Shor et al. reported that to reduce the effect of changing in temperature on the performance of a sensor the *TCR* should be positive and preferably constant, the *TCGF*

In respect to the layout of a piezoresistive sensor, in general the most used configuration for the resistors is the Wheatstone bridge. In this configuration, four resistors are connected in loop as shown in Figure 2 and the output voltage is related to the input voltage according to

> *V R V V R V V RR R R* <sup>−</sup> = =− + +

*out A B s s*

where *V*s is the supply voltage and *V*out is the output voltage.

3 4 13 24

<sup>0</sup> 1 1 *<sup>T</sup>*

0 0

*GF T* <sup>−</sup> <sup>=</sup> <sup>Δ</sup>

*T*

*T*

. This influence is measured

(13)

(14)

(15)

*R R* <sup>Δ</sup> Δ 

changes, the electrical resistance of the resistor changes

negative and *TCR TCGF* > (Shor et al., 1993).

Fig. 2. Wheatstone bridge configuration.

the following equation:

degree change in temperature. These coefficients can be determined by

Whereas, for a piezoresistive accelerometer, the sensitivity is defined as the electrical output per unit of applied acceleration:

$$S = \frac{\Delta R}{R} \frac{1}{\mathcal{g}} = \frac{\Delta V\_{out}}{V\_s} \frac{1}{\mathcal{g}} \tag{19}$$

where *g* is the acceleration of gravity.

#### **3. When and why to use SiC films in piezoresistive sensors?**

As shown in the previous section, in recent years many researchers have been reported on the piezoresistive characterization of different SiC polytypes aiming the applicability of these materials in sensors. When comparing these studies, it is observed that for a same SiC polytype a dispersion of different values can be obtained for piezoresistive coefficient, *GF* and *TCR* (Okoije, 2002).

It is known that the SiC has about 200 polytypes with different physical properties. This is one of the difficulties in characterizing the piezoresistivity in SiC. Moreover, studies show that maximum value of *GF* for SiC at room temperature is between 30 at 49 while for the monocrystalline p-type Si is 140 (see Table 1). However, all studies published until now have demonstrated the potential of the 6H-SiC and 3C-SiC polytypes besides a-SiC for the development of piezoresistive sensors for high temperature application. Given this, it is important to evaluate when it is advantageous to use SiC in piezoresistive sensors and whether is better to use SiC in bulk or thin film form.

This analysis should begin with the following question: Why SiC?

Several studies show that the SiC has mechanical and chemical stability at high temperatures. Due to these characteristics the application of SiC sensors is always associated with harsh environments. In these environments, silicon has mechanical and chemical limitations. At temperature greater than 500ºC, silicon deforms plastically under small loads

Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications 377

Crystalline SiC (c-SiC) thin films can be produced by techniques that use temperatures higher than 1000°C as chemical vapour deposition (CVD) (Chaudhuri et al., 2000), molecular beam epitaxy (MBE) (Fissel et al., 1995) and electron cyclotron resonance (ECR) (Mandracci et al., 2001). However, it is known that this high substrate temperature required for growing crystalline SiC onto Si substrate can degrade the quality of the SiC/Si interface leading to many defects in the grown films, which often prevents the film processing in conjunction with other microfabrication processes involved in a MEMS device fabrication. Conversely, there are attractive processes for the synthesis of thin films at low temperature as those based on plasma assisted techniques, such as plasma chemical vapour deposition (PECVD) and plasma sputtering, which operate at temperatures below 600°C (Rajagopalan et al., 2003; Lattemann et al., 2003). But SiC films obtained at low temperature processes are amorphous (a-SiC) or nanocrystallines (nc-SiC) and, thus, can exhibit properties somewhat different from those observed in crystalline films (Foti, 2001). Because of this, a process usually used to improve the

Among the techniques used to deposit SiC films, in this chapter only four of them will be described: CVD, PECVD, magnetron sputtering and co-sputtering. These techniques were chosen because have been used with success in the deposition of undoped and doped SiC films for MEMS sensors application. A common point among them is the ease to perform the "in situ" doping by the addition of dopant gas (N2, PH3 or B2H6) during the film

One of the most popular (laboratory) thin film deposition techniques nowadays are those based on chemical deposition processes such as chemical vapor deposition (CVD) and plasma enhanced chemical vapor deposition (PECVD) (Grill, 1994; Ohring, 2002; Bogaerts et

CVD or thermal CVD is the process of gas phase heating (by a hot filament, for example (Gracio et al., 2010)) in order for causing the decomposition of the gas, generating radical species that by diffusion can reach and be deposited on a suitably placed substrate. It differs from physical vapor deposition (PVD), which relies on material transfer from condensedphase evaporant or sputter target sources (see section 4.2.). A reaction chamber is used for this process, into which the reactant gases are introduced to decompose and react with the substrate to form the film. Figure 3a illustrates a schematic of the reactor and its main components. Basically, a typical CVD system consists of the following parts: 1) sources and feed lines of gases; 2) mass flow controllers for metering the gas inlet; 3) a reaction chamber for decomposition of precursor gases; 4) a system for heating up the gas phase and wafer on

Concerning the gas chemistry of CVD process for SiC film production, usually silane (SiH4) and light hydrocarbons gases are used, such as propane or ethylene, diluted in hydrogen as a carrier gas (Chowdhury et al., 2011). Moreover, the main CVD reactor types used are

As a modification to the CVD system, PECVD arose when plasma is used to perform the decomposition of the reactive gas source. By chemical reactions in the plasma (mainly electron impact ionization and dissociation), different kinds of ions and radicals are formed which diffuse toward the substrate where chemical surface reactions are promoted leading

crystallinity of the a-SiC films is the annealing (Rajab et al., 2006).

**4.1 Chemical deposition processes: CVD and PECVD techniques** 

which the film is to be deposited; and 5) temperature sensors.

atmospheric pressure CVD (APCVD) and low-pressure CVD (LPCVD).

deposition.

al., 2002).

(Pearson et al., 1957). In addition, the silicon does not support prolonged exposure to corrosive media. Another important factor that should be considered is that silicon pressure sensors using p-n junction piezoresistors have exhibited good performance at temperatures up to 175ºC and the SOI sensors at temperatures up to 500ºC.

Among the semiconductor materials with potential to substitute the silicon in harsh environments, SiC is the most appropriate candidate because its native oxide is SiO2 which makes SiC directly compatible with the Si technology. This signifies that a sensor based on SiC can be developed following the same steps used in silicon sensors.

On the other hand, the chemical stability that have qualified SiC for harsh environments, makes it difficult to etch the bulk and to integrate any process step with already established Si based processes. Furthermore, the high cost of SiC wafer also difficult the development of "all of SiC" sensors. Faced with these difficulties the use of SiC thin films is quite attractive because the film can be grown on large-area Si substrates and by the ease of using conventional Si bulk micromachining techniques (Fraga et al., 2011a).

The second question is: When to use piezoresistive sensors based on SiC?

As already mentioned in the beginning of this section, at room temperature the monocrystalline silicon has greater *GF* than the SiC, i.e. sensors based on silicon operating on this condition has superior sensitivity. This fact shows that the use of SiC is only justified for specific applications in four main types of harsh environments, namely:


#### **4. Brief description of the main techniques to deposit SiC films**

Several techniques for obtaining thin films and bulks of SiC have been developed. Some companies that manufacture crystalline silicon wafers also offer SiC bulk wafers up to 4 inches in diameter. However, SiC wafers have an average price fifteen times higher than Si wafers with the same dimensions (Hobgood et al., 2004; Camassel & Juillaguet, 2007). Besides the high cost, another problem of the use of SiC substrates is the difficult micromachining process and high density of defects (Wu et al., 2001). In this context, there is a crescent interest in deposition techniques of SiC films on Si or SOI (Silicon-On-Insulator) substrates. These films can be produced in crystalline and amorphous forms.

(Pearson et al., 1957). In addition, the silicon does not support prolonged exposure to corrosive media. Another important factor that should be considered is that silicon pressure sensors using p-n junction piezoresistors have exhibited good performance at temperatures

Among the semiconductor materials with potential to substitute the silicon in harsh environments, SiC is the most appropriate candidate because its native oxide is SiO2 which makes SiC directly compatible with the Si technology. This signifies that a sensor based on

On the other hand, the chemical stability that have qualified SiC for harsh environments, makes it difficult to etch the bulk and to integrate any process step with already established Si based processes. Furthermore, the high cost of SiC wafer also difficult the development of "all of SiC" sensors. Faced with these difficulties the use of SiC thin films is quite attractive because the film can be grown on large-area Si substrates and by the ease of using

As already mentioned in the beginning of this section, at room temperature the monocrystalline silicon has greater *GF* than the SiC, i.e. sensors based on silicon operating on this condition has superior sensitivity. This fact shows that the use of SiC is only justified

a. Mechanically aggressive that involve high loads as in oil and gas industry applications which require sensors to operate in pressure ranges up to 35,000 psi and at

b. Thermally aggressive that involve high temperatures as in combustion control in gas turbine engines, where the operating temperatures are around 600°C (Vandelli, 2008) and in pressure monitoring during deep well drilling and combustion in aeronautical and automobile engines that require sensors to operate at temperatures ranging

c. Chemically aggressive or corrosive environment as in biomedical and petrochemical applications where chemical attack by fluids is one of the modes of degradation of devices. The SiC sensors are a good choice for these applications because at room temperature, there is no known wet chemical that etches single-crystal SiC (George et

d. Aerospace environment where sensors should to maintain their functionality under high cumulative doses of radiation. Due to well known chemical inertness of the SiC, sensors based on this material have exhibited great potential for these applications.

Several techniques for obtaining thin films and bulks of SiC have been developed. Some companies that manufacture crystalline silicon wafers also offer SiC bulk wafers up to 4 inches in diameter. However, SiC wafers have an average price fifteen times higher than Si wafers with the same dimensions (Hobgood et al., 2004; Camassel & Juillaguet, 2007). Besides the high cost, another problem of the use of SiC substrates is the difficult micromachining process and high density of defects (Wu et al., 2001). In this context, there is a crescent interest in deposition techniques of SiC films on Si or SOI (Silicon-On-Insulator)

up to 175ºC and the SOI sensors at temperatures up to 500ºC.

SiC can be developed following the same steps used in silicon sensors.

conventional Si bulk micromachining techniques (Fraga et al., 2011a). The second question is: When to use piezoresistive sensors based on SiC?

for specific applications in four main types of harsh environments, namely:

**4. Brief description of the main techniques to deposit SiC films** 

substrates. These films can be produced in crystalline and amorphous forms.

temperatures up to 200°C (Vandelli, 2008);

al., 2006);

between 300 and 600ºC (Stanescu & Voican, 2007);

Crystalline SiC (c-SiC) thin films can be produced by techniques that use temperatures higher than 1000°C as chemical vapour deposition (CVD) (Chaudhuri et al., 2000), molecular beam epitaxy (MBE) (Fissel et al., 1995) and electron cyclotron resonance (ECR) (Mandracci et al., 2001). However, it is known that this high substrate temperature required for growing crystalline SiC onto Si substrate can degrade the quality of the SiC/Si interface leading to many defects in the grown films, which often prevents the film processing in conjunction with other microfabrication processes involved in a MEMS device fabrication. Conversely, there are attractive processes for the synthesis of thin films at low temperature as those based on plasma assisted techniques, such as plasma chemical vapour deposition (PECVD) and plasma sputtering, which operate at temperatures below 600°C (Rajagopalan et al., 2003; Lattemann et al., 2003). But SiC films obtained at low temperature processes are amorphous (a-SiC) or nanocrystallines (nc-SiC) and, thus, can exhibit properties somewhat different from those observed in crystalline films (Foti, 2001). Because of this, a process usually used to improve the crystallinity of the a-SiC films is the annealing (Rajab et al., 2006).

Among the techniques used to deposit SiC films, in this chapter only four of them will be described: CVD, PECVD, magnetron sputtering and co-sputtering. These techniques were chosen because have been used with success in the deposition of undoped and doped SiC films for MEMS sensors application. A common point among them is the ease to perform the "in situ" doping by the addition of dopant gas (N2, PH3 or B2H6) during the film deposition.

#### **4.1 Chemical deposition processes: CVD and PECVD techniques**

One of the most popular (laboratory) thin film deposition techniques nowadays are those based on chemical deposition processes such as chemical vapor deposition (CVD) and plasma enhanced chemical vapor deposition (PECVD) (Grill, 1994; Ohring, 2002; Bogaerts et al., 2002).

CVD or thermal CVD is the process of gas phase heating (by a hot filament, for example (Gracio et al., 2010)) in order for causing the decomposition of the gas, generating radical species that by diffusion can reach and be deposited on a suitably placed substrate. It differs from physical vapor deposition (PVD), which relies on material transfer from condensedphase evaporant or sputter target sources (see section 4.2.). A reaction chamber is used for this process, into which the reactant gases are introduced to decompose and react with the substrate to form the film. Figure 3a illustrates a schematic of the reactor and its main components. Basically, a typical CVD system consists of the following parts: 1) sources and feed lines of gases; 2) mass flow controllers for metering the gas inlet; 3) a reaction chamber for decomposition of precursor gases; 4) a system for heating up the gas phase and wafer on which the film is to be deposited; and 5) temperature sensors.

Concerning the gas chemistry of CVD process for SiC film production, usually silane (SiH4) and light hydrocarbons gases are used, such as propane or ethylene, diluted in hydrogen as a carrier gas (Chowdhury et al., 2011). Moreover, the main CVD reactor types used are atmospheric pressure CVD (APCVD) and low-pressure CVD (LPCVD).

As a modification to the CVD system, PECVD arose when plasma is used to perform the decomposition of the reactive gas source. By chemical reactions in the plasma (mainly electron impact ionization and dissociation), different kinds of ions and radicals are formed which diffuse toward the substrate where chemical surface reactions are promoted leading

Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications 379

Therefore, the smaller electrode acquires a larger bias voltage and becomes negative with respect to the larger electrode. The negative sheath voltage accelerates the positive ions towards the substrate which is mounted on this smaller electrode, allowing the substrate to

In order to maximize the ion to neutral ratio of the plasma, the plasma must be operated at the lowest possible pressure. Nevertheless, the ions are only about 10 percent of the filmforming flux even at pressures as low as 50 mTorr. Lower pressures cannot be used as the plasma wills no longer strike. A second disadvantage of this source is the energy spread in the ion energy distribution, prohibiting a controlled deposition. This energy spread is due to inelastic collisions as the ions are accelerated towards the substrate. The effect of this energy spread is to lower the mean ion energy to about 0.4 of the sheath voltage. Still, another disadvantage of the rf PECVD source is that it is not possible to have independent control over the ion energy and the ion current, as they both vary with the rf power. On the other hand, PECVD allows the deposition of uniform films over large areas, and PECVD systems

The most used precursor gases to deposit SiC films by PECVD are SiH4, as the silicon source, and methane (CH4), as carbon source. Finally, Figure 4 illustrates the deposition mechanism of chemical vapor deposition technique (Grill, 1994). Basically the mechanism occurs by the following steps: (i) a predefined mix of reactant gases and diluents inert gases are introduced at a specified flow rate into the reaction chamber; (ii) a heat source is applied in order to dissociate the reactant gases; (iii) the resulting radical species diffuse to the substrate; (iv) the reactants get adsorbed on the surface of the substrate; (v) the reactants undergo chemical reactions with the substrate to form the film; and (vi) the gaseous by-products of the reactions are desorbed and evacuated from the reaction

Fig. 4. Chemical vapor deposition mechanism. Adapted from (Doi, 2006).

become bombarded by energetic ions facilitating reactions with substrate surface.

can be easily scaled up (Neyts, 2006).

chamber.

to film growth. The major advantage compared to simple CVD is that PECVD can operate at much lower temperatures. Indeed, the electron temperature of 2–5 eV in PECVD is sufficient for dissociation, whereas in CVD the gas and surface reactions occur by thermal activation. Hence, some coatings, which are difficult to form by CVD due to melting problems, can be deposited more easily with PECVD (Bogaerts et al., 2002; Peng et al., 2011). Among the kinds of plasma sources that have been used for this application stand out the radiofrequency (rf) discharges (Bogaerts et al., 2002), pulsed discharges (Zhao et al., 2010) and microwave discharges (Gracio et al., 2010).

Basically, in PECVD the substrate is mounted on one of the electrodes in the same reactor where the species are created (see Figure 3b). Here, we focused the rf discharge because it is the configuration more used in research and industry. The rf PECVD reactor essentially consists of two electrodes of different areas, where the substrate is placed on the smaller electrode, to which the power is capacitively coupled. The rf power creates a plasma between the electrodes. Due to the higher mobility of the electrons than the ions, a sheath is created next to the electrodes containing an excess of ions. Hence, the sheath has a positive space charge, and the plasma creates a positive voltage with respect to the electrodes. The electrodes therefore acquire a dc self-bias equal to their peak rf voltage (self-bias electrode). The ratio of the dc self-bias voltages is inversely proportional to the ratio of the squared electrode areas, i.e., V1/V2 = (A1/A2)2 (Lieberman & Lichtenberg, 2005).

Fig. 3. Schematic diagram of CVD (a) and PECVD (b) systems.

to film growth. The major advantage compared to simple CVD is that PECVD can operate at much lower temperatures. Indeed, the electron temperature of 2–5 eV in PECVD is sufficient for dissociation, whereas in CVD the gas and surface reactions occur by thermal activation. Hence, some coatings, which are difficult to form by CVD due to melting problems, can be deposited more easily with PECVD (Bogaerts et al., 2002; Peng et al., 2011). Among the kinds of plasma sources that have been used for this application stand out the radiofrequency (rf) discharges (Bogaerts et al., 2002), pulsed discharges (Zhao et al., 2010)

Basically, in PECVD the substrate is mounted on one of the electrodes in the same reactor where the species are created (see Figure 3b). Here, we focused the rf discharge because it is the configuration more used in research and industry. The rf PECVD reactor essentially consists of two electrodes of different areas, where the substrate is placed on the smaller electrode, to which the power is capacitively coupled. The rf power creates a plasma between the electrodes. Due to the higher mobility of the electrons than the ions, a sheath is created next to the electrodes containing an excess of ions. Hence, the sheath has a positive space charge, and the plasma creates a positive voltage with respect to the electrodes. The electrodes therefore acquire a dc self-bias equal to their peak rf voltage (self-bias electrode). The ratio of the dc self-bias voltages is inversely proportional to the ratio of the squared

electrode areas, i.e., V1/V2 = (A1/A2)2 (Lieberman & Lichtenberg, 2005).

Fig. 3. Schematic diagram of CVD (a) and PECVD (b) systems.

and microwave discharges (Gracio et al., 2010).

Therefore, the smaller electrode acquires a larger bias voltage and becomes negative with respect to the larger electrode. The negative sheath voltage accelerates the positive ions towards the substrate which is mounted on this smaller electrode, allowing the substrate to become bombarded by energetic ions facilitating reactions with substrate surface.

In order to maximize the ion to neutral ratio of the plasma, the plasma must be operated at the lowest possible pressure. Nevertheless, the ions are only about 10 percent of the filmforming flux even at pressures as low as 50 mTorr. Lower pressures cannot be used as the plasma wills no longer strike. A second disadvantage of this source is the energy spread in the ion energy distribution, prohibiting a controlled deposition. This energy spread is due to inelastic collisions as the ions are accelerated towards the substrate. The effect of this energy spread is to lower the mean ion energy to about 0.4 of the sheath voltage. Still, another disadvantage of the rf PECVD source is that it is not possible to have independent control over the ion energy and the ion current, as they both vary with the rf power. On the other hand, PECVD allows the deposition of uniform films over large areas, and PECVD systems can be easily scaled up (Neyts, 2006).

The most used precursor gases to deposit SiC films by PECVD are SiH4, as the silicon source, and methane (CH4), as carbon source. Finally, Figure 4 illustrates the deposition mechanism of chemical vapor deposition technique (Grill, 1994). Basically the mechanism occurs by the following steps: (i) a predefined mix of reactant gases and diluents inert gases are introduced at a specified flow rate into the reaction chamber; (ii) a heat source is applied in order to dissociate the reactant gases; (iii) the resulting radical species diffuse to the substrate; (iv) the reactants get adsorbed on the surface of the substrate; (v) the reactants undergo chemical reactions with the substrate to form the film; and (vi) the gaseous by-products of the reactions are desorbed and evacuated from the reaction chamber.

Fig. 4. Chemical vapor deposition mechanism. Adapted from (Doi, 2006).

Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications 381

created. Figure 6 shows a schematic drawing of a conventional dc magnetron sputtering discharge. The trapping of the secondary electrons results in a higher probability of electron impact ionization and hence higher plasma density, increasing the sputtering flux and allowing operation at lower pressures, bellows 10 mTorr. Furthermore, the discharge voltage can be lowered into the range of 300-700 V. The main problem with the magnetron sputtering configuration is that the sputtering is confined to a small area of the target cathode governed by the magnetic field. The discharge appears in the form a high-density annulus of width w and radius R, as seen in Figure 6. Sputtering occurs in the corresponding track of the target. This

Fig. 6. Schematic drawing of a conventional dc magnetron sputtering discharge. Adapted

film compositions by applied different power on each target (Medeiros et al., 2011).

**5. Requirements of SiC films for piezoresistive sensors application** 

sputtering as appropriate techniques to deposit SiC films on SiO2/Si (Zanola, 2004).

Deposition of SiC films by the Magnetron Sputtering technique is performed generally using a SiC target in Ar atmosphere or a silicon target with precursor gases Ar plus CH4 (Stamate et al., 2008). The dual magnetron (or co-sputtering) method also has been used to deposit SiC films. In this technique, the films are produced by co-sputtering of carbon and silicon targets (see Figure 7) with Ar as precursor gas (Kikuchi et al., 2002; Kerdiles et al., 2002). The co-sputtering technique offers as main advantage to obtaining of SiC films with different electrical, structural and mechanical properties by the variation of C/Si ratio in the film deposited (Kikuchi et al., 2002). Using this technique, it is possible to obtain a range of SiC

In order to develop piezoresistive sensors with high performance based on SiC films is necessary to optimize the properties of the SiC thin-film piezoresistors to maximize their sensitivity with the minimum temperature-dependent resistance variation (Luchinin &

The first step for this optimization is the choice of the technique to deposit SiC films onto an insulator on Si substrates. Silicon dioxide (SiO2) is the most used insulator material for this purpose, but some studies have showed silicon nitride (Si3N4) or aluminum nitride (AlN) as alternative materials. In general, good results have been achieved with the SiO2, although this material has a coefficient of thermal expansion (*CTE*) significantly lower than the SiC, giving rise to thermal stresses at the SiC/SiO2 interface. Many studies have shown CVD, PECVD and

area, known as the race track, is created by the uneven ion density.

from (Bogaerts et al., 2002).

Korlyakov, 2009).

#### **4.2 Physical deposition processes: Magnetron sputtering and co-sputtering techniques**

The physical deposition process comprise the physical sputtering and reactive sputtering techniques. Basically, these techniques differ when a neutral gas (physical sputtering) is added together with a reactive gas (reactive sputtering). In physical sputtering, ions (and atoms) from the plasma bombard the target, and release atoms (or molecules) of the target material. Argon ions at 500–1000 V are usually used. The sputtered atoms diffuse through the plasma and arrive at the substrate, where they can be deposited (Bogaerts et. al., 2002). In reactive sputtering, use is made of a molecular gas (for example, N2 or O2). Beside the positive ions from the plasma that sputter bombard the target, the dissociation products from the reactive gas will also react with the target. Hence, the film deposited at the substrate will be a combination of sputtered target material and the reactive gas (Bogaerts et al., 2002; Berg, 2005; Lieberman & Lichtenberg, 2005). The sputter deposition process is schematically presented in Figure 5.

Fig. 5. Schematic of sputtering process.

Basically the steps of sputtering process are the following: (i) the neutral gas is ionized by a external power supply, producing a glow discharge or plasma; (ii) a source (the cathode, also called the target) is bombarded in high vacuum by gas ions due to the potential drop acceleration in the cathode sheath; (iii) atoms from the target are ejected by momentum transfer and diffuse through the vacuum chamber; (iv) atoms are deposited on the substrate to be coated and form a thin film.

Because sputter yields are of order unity for almost all target materials, a very wide variety of pure metals, alloys, and insulators can be deposited. Physical sputtering, especially of elemental targets, is a well understood process enabling sputtering systems for various applications to be relatively easily designed. Reasonable deposition rates with excellent film uniformity, good surface smoothness, and adhesion can be achieved over large areas (Lieberman & Lichtenberg, 2005).

Typically, the sputtering process can be accomplished using a planar configuration of electrodes and a dc power supply, where one electrode is biased negatively (cathode) and suffer the sputtering process. However, the sputtering yield is directly dependent on the gas pressure (best sputtering rates are in the range of mTorr) a fact that compromises the efficiency of planar geometry for this application: it is great for pressures above 100 mTorr. To solve this problem, it was developed the magnetron discharge where the plasma is magnetically enhanced by placing magnets behind the cathode target, i.e., a crossed electric and magnetic field configuration is

The physical deposition process comprise the physical sputtering and reactive sputtering techniques. Basically, these techniques differ when a neutral gas (physical sputtering) is added together with a reactive gas (reactive sputtering). In physical sputtering, ions (and atoms) from the plasma bombard the target, and release atoms (or molecules) of the target material. Argon ions at 500–1000 V are usually used. The sputtered atoms diffuse through the plasma and arrive at the substrate, where they can be deposited (Bogaerts et. al., 2002). In reactive sputtering, use is made of a molecular gas (for example, N2 or O2). Beside the positive ions from the plasma that sputter bombard the target, the dissociation products from the reactive gas will also react with the target. Hence, the film deposited at the substrate will be a combination of sputtered target material and the reactive gas (Bogaerts et al., 2002; Berg, 2005; Lieberman & Lichtenberg, 2005). The sputter deposition process is

Basically the steps of sputtering process are the following: (i) the neutral gas is ionized by a external power supply, producing a glow discharge or plasma; (ii) a source (the cathode, also called the target) is bombarded in high vacuum by gas ions due to the potential drop acceleration in the cathode sheath; (iii) atoms from the target are ejected by momentum transfer and diffuse through the vacuum chamber; (iv) atoms are deposited on the substrate

Because sputter yields are of order unity for almost all target materials, a very wide variety of pure metals, alloys, and insulators can be deposited. Physical sputtering, especially of elemental targets, is a well understood process enabling sputtering systems for various applications to be relatively easily designed. Reasonable deposition rates with excellent film uniformity, good surface smoothness, and adhesion can be achieved over large areas

Typically, the sputtering process can be accomplished using a planar configuration of electrodes and a dc power supply, where one electrode is biased negatively (cathode) and suffer the sputtering process. However, the sputtering yield is directly dependent on the gas pressure (best sputtering rates are in the range of mTorr) a fact that compromises the efficiency of planar geometry for this application: it is great for pressures above 100 mTorr. To solve this problem, it was developed the magnetron discharge where the plasma is magnetically enhanced by placing magnets behind the cathode target, i.e., a crossed electric and magnetic field configuration is

**4.2 Physical deposition processes: Magnetron sputtering and co-sputtering** 

**techniques** 

schematically presented in Figure 5.

Fig. 5. Schematic of sputtering process.

to be coated and form a thin film.

(Lieberman & Lichtenberg, 2005).

created. Figure 6 shows a schematic drawing of a conventional dc magnetron sputtering discharge. The trapping of the secondary electrons results in a higher probability of electron impact ionization and hence higher plasma density, increasing the sputtering flux and allowing operation at lower pressures, bellows 10 mTorr. Furthermore, the discharge voltage can be lowered into the range of 300-700 V. The main problem with the magnetron sputtering configuration is that the sputtering is confined to a small area of the target cathode governed by the magnetic field. The discharge appears in the form a high-density annulus of width w and radius R, as seen in Figure 6. Sputtering occurs in the corresponding track of the target. This area, known as the race track, is created by the uneven ion density.

Deposition of SiC films by the Magnetron Sputtering technique is performed generally using a SiC target in Ar atmosphere or a silicon target with precursor gases Ar plus CH4 (Stamate et al., 2008). The dual magnetron (or co-sputtering) method also has been used to deposit SiC films. In this technique, the films are produced by co-sputtering of carbon and silicon targets (see Figure 7) with Ar as precursor gas (Kikuchi et al., 2002; Kerdiles et al., 2002). The co-sputtering technique offers as main advantage to obtaining of SiC films with different electrical, structural and mechanical properties by the variation of C/Si ratio in the film deposited (Kikuchi et al., 2002). Using this technique, it is possible to obtain a range of SiC film compositions by applied different power on each target (Medeiros et al., 2011).

#### **5. Requirements of SiC films for piezoresistive sensors application**

In order to develop piezoresistive sensors with high performance based on SiC films is necessary to optimize the properties of the SiC thin-film piezoresistors to maximize their sensitivity with the minimum temperature-dependent resistance variation (Luchinin & Korlyakov, 2009).

The first step for this optimization is the choice of the technique to deposit SiC films onto an insulator on Si substrates. Silicon dioxide (SiO2) is the most used insulator material for this purpose, but some studies have showed silicon nitride (Si3N4) or aluminum nitride (AlN) as alternative materials. In general, good results have been achieved with the SiO2, although this material has a coefficient of thermal expansion (*CTE*) significantly lower than the SiC, giving rise to thermal stresses at the SiC/SiO2 interface. Many studies have shown CVD, PECVD and sputtering as appropriate techniques to deposit SiC films on SiO2/Si (Zanola, 2004).

Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications 383

The most used technique to determine the value of *GF* of a piezoresistor is the cantilever deflection method. In this method, the piezoresistor is glued near to the clamped end of a cantilever beam and on the free end of the beam different loads are applied. The value of *GF* is obtained by monitoring the resistance change when the resistor is subjected to different applied stress. Once determined the *GF*, the *TCR* and the *TCGF* are determined to evaluate

Table 2 summarizes the main requirements that SiC film should present to be successfully used in the development of piezoresistive sensors. As can be seen, the resistivity of the SiC thin film should be low (preferably of the order of mΩ.cm) because its thickness in general less than 1.0 μm. As the depth of the SiC thin-film piezoresistor is equals the thickness film,

Among the many silicon-based microsensors, piezoresistive pressure sensors are one of the widely used products of microelectromechanical system (MEMS) technology. This type of sensor has dominated the market in recent decades due to characteristics such as high sensitivity, high linearity, and an easy-to-retrieve signal through bridge circuit. The main applications of Si-based piezoresistive pressure sensors are in the biomedical, industrial and automotive fields. However, these sensors have a drawback that is the influence of the temperature on their performance. For some applications, this temperature effect can be

Given this, many studies have been performed aiming to reduce the temperature effects on the performance of the sensor through the use of piezoresistive sensing elements formed by wide bandgap semiconductor thin film as the SiC. The goal is to develop sensors as small as possible and enable to operate at high temperatures. For this, besides making the piezoresistors based on material with suitable properties for high temperature applications should also be used stable electrical contacts with excellent environmental stability. It is known that the metallization type also influences the performance of the devices at harsh environments. Studies show that for SiC sensors the best hightemperature contacts are metal as Au, Ni, Ti and W and binary compounds such as TiSi2

A typical SiC thin-film based piezoresistive pressure sensor consists of SiC thin-film piezoresistors, configured in Wheatstone bridge, on a diaphragm. The monocrystalline silicon is the material most used to form the diaphragm due its mechanical properties which make it an excellent material for elastic structural members of a sensor. In addition, the Si diaphragms can be easily fabricated by KOH anisotropic etching from the backside of a (100) silicon wafer using

it is necessary a low resistivity film to form low electrical resistance piezoresistors.

**Electrical and Mechanical Characteristics Requirement**  Elastic modulus The greater Residual stress The lower Resistivity The lower GF The greater TCR The lower TCGF The lower Table 2. Main requirements of SiC films for piezoresistive sensor applications.

**6. Examples of piezoresistive sensors based on SiC films** 

compensated by an external circuit, which adds substantial cost to the sensor.

and WiSi2 (Cocuzza, 2003).

the influence of the temperature (see details on topic 2).

Fig. 7. Schematic diagram of magnetron co-sputtering deposition technique.

After the film deposition, the residual stress must be investigated. SiC films obtained by CVD have low residual stress due to high temperatures involved in this process. However, films obtained by PECVD and sputtering exhibit a significant tensile or compressive residual stress that is dependent on various deposition parameters. To reduce this stress post-deposition thermal annealing is usually performed (Zorman, 2006).

The following step is used to determine the chemical, physical and structural properties of the as-deposited SiC film. For piezoresistive sensor applications, it is fundamental the knowledge of the orientation, elastic modulus, doping concentration and resistivity of the film. After determining these properties, the piezoresistive characterization of the film is started. First, a test structure must be developed. Generally, this structure consists of a SiC thin-film piezoresistor fabricated by photolithography, lift-off and etching processes as illustrated in Figure 8.

Fig. 8. Schematic flow diagram of the SiC thin-film resistor fabrication process.

Fig. 7. Schematic diagram of magnetron co-sputtering deposition technique.

post-deposition thermal annealing is usually performed (Zorman, 2006).

Fig. 8. Schematic flow diagram of the SiC thin-film resistor fabrication process.

illustrated in Figure 8.

After the film deposition, the residual stress must be investigated. SiC films obtained by CVD have low residual stress due to high temperatures involved in this process. However, films obtained by PECVD and sputtering exhibit a significant tensile or compressive residual stress that is dependent on various deposition parameters. To reduce this stress

The following step is used to determine the chemical, physical and structural properties of the as-deposited SiC film. For piezoresistive sensor applications, it is fundamental the knowledge of the orientation, elastic modulus, doping concentration and resistivity of the film. After determining these properties, the piezoresistive characterization of the film is started. First, a test structure must be developed. Generally, this structure consists of a SiC thin-film piezoresistor fabricated by photolithography, lift-off and etching processes as The most used technique to determine the value of *GF* of a piezoresistor is the cantilever deflection method. In this method, the piezoresistor is glued near to the clamped end of a cantilever beam and on the free end of the beam different loads are applied. The value of *GF* is obtained by monitoring the resistance change when the resistor is subjected to different applied stress. Once determined the *GF*, the *TCR* and the *TCGF* are determined to evaluate the influence of the temperature (see details on topic 2).

Table 2 summarizes the main requirements that SiC film should present to be successfully used in the development of piezoresistive sensors. As can be seen, the resistivity of the SiC thin film should be low (preferably of the order of mΩ.cm) because its thickness in general less than 1.0 μm. As the depth of the SiC thin-film piezoresistor is equals the thickness film, it is necessary a low resistivity film to form low electrical resistance piezoresistors.


Table 2. Main requirements of SiC films for piezoresistive sensor applications.

#### **6. Examples of piezoresistive sensors based on SiC films**

Among the many silicon-based microsensors, piezoresistive pressure sensors are one of the widely used products of microelectromechanical system (MEMS) technology. This type of sensor has dominated the market in recent decades due to characteristics such as high sensitivity, high linearity, and an easy-to-retrieve signal through bridge circuit. The main applications of Si-based piezoresistive pressure sensors are in the biomedical, industrial and automotive fields. However, these sensors have a drawback that is the influence of the temperature on their performance. For some applications, this temperature effect can be compensated by an external circuit, which adds substantial cost to the sensor.

Given this, many studies have been performed aiming to reduce the temperature effects on the performance of the sensor through the use of piezoresistive sensing elements formed by wide bandgap semiconductor thin film as the SiC. The goal is to develop sensors as small as possible and enable to operate at high temperatures. For this, besides making the piezoresistors based on material with suitable properties for high temperature applications should also be used stable electrical contacts with excellent environmental stability. It is known that the metallization type also influences the performance of the devices at harsh environments. Studies show that for SiC sensors the best hightemperature contacts are metal as Au, Ni, Ti and W and binary compounds such as TiSi2 and WiSi2 (Cocuzza, 2003).

A typical SiC thin-film based piezoresistive pressure sensor consists of SiC thin-film piezoresistors, configured in Wheatstone bridge, on a diaphragm. The monocrystalline silicon is the material most used to form the diaphragm due its mechanical properties which make it an excellent material for elastic structural members of a sensor. In addition, the Si diaphragms can be easily fabricated by KOH anisotropic etching from the backside of a (100) silicon wafer using

Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications 385

As mentioned earlier, the cost of the 6H-SiC is also elevated which has stimulated the researches on SiC thin-film piezoresistive accelerometer. The simplest model for this accelerometer is illustrated in Figure 10. This accelerometer consists of a SiC thin-film piezoresistor (or four piezoresistors configured in Wheatstone bridge) on a silicon cantilever beam which has a rigid silicon proof mass attached at its free end. The basic principle of this type of sensor is that the acceleration moves the proof mass so deflecting the cantilever which works as a spring. The mass shift produces a variation of the internal stress of the spring that can be sensed by the piezoresistor. The value of the acceleration can be inferred by the measurement of the magnitude of the stress. The main problem of this accelerometer is that all

its structure is built on silicon which can limit the performance at harsh environments.

Fig. 10. Schematic illustration of a SiC thin-film based piezoresistive accelerometer.

the SiO2 or Si3N4 film as etch mask. It is also necessary to grow SiO2 or Si3N4 on the front side of the wafer to perform the electrical insulation of the SiC thin-film piezoresistors from the substrate. Generally, the SiC thin-film piezoresistors are produced by RIE (reactive ion etching). Figure 9 illustrates two piezoresistive pressure sensors based on SiC films: one with six PECVD a-SiC thin-film piezoresistors, configured in Wheatstone bridge, on a SiO2/Si square diaphragm with Ti/Au metallization (Fraga et al., 2011b) and the other with phosphorusdoped APCVD polycrystalline 3C-SiC piezoresistors on Si3N4/3C-SiC diaphragm with Ni metallization (Wu et al., 2006).

Fig. 9. Schematic illustration of piezoresistive pressure sensors based on SiC films.

Another sensor type that has been developed based on SiC is the accelerometer. However, for now, the studies are still focused on piezoresistive accelerometers based on 6H-SiC bulk substrate (Atwell et al., 2003) or on SiC thin-film capacitive accelerometers (Rajaraman et al., 2011).

This occurs because the capacitive accelerometer is usually more sensitive than piezoresistive one and furthermore can be used in a wide range of temperature. On the other hand, the capacitive accelerometers have elevated cost and necessity of signal conditioning circuit (Koberstein, 2005). The motivation to develop piezoresistive accelerometers on 6H-SiC bulk is the possibility of obtaining superior performance at high temperature in comparison with capacitive accelerometer.

the SiO2 or Si3N4 film as etch mask. It is also necessary to grow SiO2 or Si3N4 on the front side of the wafer to perform the electrical insulation of the SiC thin-film piezoresistors from the substrate. Generally, the SiC thin-film piezoresistors are produced by RIE (reactive ion etching). Figure 9 illustrates two piezoresistive pressure sensors based on SiC films: one with six PECVD a-SiC thin-film piezoresistors, configured in Wheatstone bridge, on a SiO2/Si square diaphragm with Ti/Au metallization (Fraga et al., 2011b) and the other with phosphorusdoped APCVD polycrystalline 3C-SiC piezoresistors on Si3N4/3C-SiC diaphragm with Ni

Fig. 9. Schematic illustration of piezoresistive pressure sensors based on SiC films.

temperature in comparison with capacitive accelerometer.

Another sensor type that has been developed based on SiC is the accelerometer. However, for now, the studies are still focused on piezoresistive accelerometers based on 6H-SiC bulk substrate (Atwell et al., 2003) or on SiC thin-film capacitive accelerometers (Rajaraman et

This occurs because the capacitive accelerometer is usually more sensitive than piezoresistive one and furthermore can be used in a wide range of temperature. On the other hand, the capacitive accelerometers have elevated cost and necessity of signal conditioning circuit (Koberstein, 2005). The motivation to develop piezoresistive accelerometers on 6H-SiC bulk is the possibility of obtaining superior performance at high

metallization (Wu et al., 2006).

al., 2011).

As mentioned earlier, the cost of the 6H-SiC is also elevated which has stimulated the researches on SiC thin-film piezoresistive accelerometer. The simplest model for this accelerometer is illustrated in Figure 10. This accelerometer consists of a SiC thin-film piezoresistor (or four piezoresistors configured in Wheatstone bridge) on a silicon cantilever beam which has a rigid silicon proof mass attached at its free end. The basic principle of this type of sensor is that the acceleration moves the proof mass so deflecting the cantilever which works as a spring. The mass shift produces a variation of the internal stress of the spring that can be sensed by the piezoresistor. The value of the acceleration can be inferred by the measurement of the magnitude of the stress. The main problem of this accelerometer is that all its structure is built on silicon which can limit the performance at harsh environments.

Fig. 10. Schematic illustration of a SiC thin-film based piezoresistive accelerometer.

Recent Developments on Silicon Carbide Thin Films for Piezoresistive Sensors Applications 387

Fraga, M.A. (2009) *Desenvolvimento de sensores piezoresistivos de SiC visando aplicação em sistemas aeroespaciais*, PhD thesis, Instituto Tecnológico de Aeronáutica. Fraga, M.A.; Furlan H.; Massi, M.; Oliveira I.C. (2010a) *Microsyst Technol*., Vol. 16, pp.925-

Fraga, M.A.; Furlan, H.; Massi, M.; Oliveira, I.C.; Mateus, C.F.R.; Rasia, L.A. ; (2011b)

Fraga, M.A.; Furlan H.; Pessoa, R.S. (2011c) *Comparison among performance of strain sensors* 

George T.; Son K.A.; Powers R.A.; Del Castillo L.Y.; Okojie R. (2006) *Harsh environment* 

Gracio, J. J.; Fan, Q. H.; Madaleno, J. C. (2010) *J. Phys. D: Appl. Phys.*, Vol.43, 374017 (22pp). Grill, A. (1994) *Cold Plasmas in Materials Fabrication: From Fundamentals to Applications*, IEEE

Guk, G. N.; Usol´tseva, N.Y.; Shadrin, V.S.; Mundus-Tabakaev, A.F. (1974a) *Sov. Phys. Solid* 

Guk, G.N.; Lyubimskii, V.M.; Gofman, E.P.; Zinovév, V.B; Chalyi, E.A. (1974b) *Sov. Phys.* 

Guk, G. N. ; Usol´tseva, N.Y. ; Shadrin, V.S. ;Prokop´eva, N.P (1976) *Sov. Phys. Solid State,*

Hobgood, H. McD.; Brady, M. F.; Calus, M. R.; Jenny, J. R.; Leonard, R. T.; Malta, D. P.;

Kerdiles, S.; Rizk, R.; Gourbilleau, F.; Perez-Rodriguez, A.; Garrido, B.; Gonzalez-Varona, O.; Morante, J.R. (2002) *Materials Science and Engineering B*, Vol.69–70, pp.530–535. Kikuchi, N; Kusano, E.; Tanaka, T.; Kinbara, A.; Nanto, H. (2002) *Surface and Coatings* 

Koberstein, L. (2005) *Modelagem de um acelerômetro de estado sólido, Master dissertation in* 

Kulikovsky V. ; Vorlíček, V.; Boháč, P.; Stranyánek, M.; Čtvrtlík, R. ; Kurdyumov, A. ; Jastrabik, L. (2008) *Surface & Coatings Technology,* Vol.202, pp.1738–1745. Lattemann, M.; Nold, E.; Ulrich, S.; Leiste, H.; Holleck, H. (2003) *Surface and Coatings* 

Lieberman, M.A.; Lichtenberg, A.J. (2005) *Principles of Plasma Discharges and Materials* 

Luchinin, V.V.; Korlyakov, A.V. (2009), *Materials and elements of constructions for extreme micro- and nanoengineering, Proceedings of the EUROCON,* pp.1242-1245. Mandracci, P.; Chiodoni, A.; Cicero, G.; Ferrero, S.; Giorgis, F.; Pirri, C.F.; Barucca, G.; Musumeci, P.; Reitano, R. (2001) *Applied Surface Science* Vol.184, pp.43-49. Medeiros, H. S.; Pessoa, R.S.; Sagás, J.C.; Santos, L.V.; Fraga, M.A.; Maciel, H.S.; Sobrinho, A. S. S.;

*by Dual Magnetron Sputtering System*, In : IX SBPMat, Ouro Preto-MG, Brazil. Morin, F. J.; Geballe, T. H.; Herring, C. (1956) *Physical Review*, Vol.10, No.2, pp.525-539. Neyts, E. (2006) *Mathematical Simulation of the Deposition of Diamond-like carbon (DLC) Films*,

Massi, M. (2010) *Effect of Concentration of Carbon and Silicon in the SiC Thin Film Deposition* 

Müller, S. G.; Powell, A. R.; Tsvetkov, V. F.; Glass, R. C.; Carter, C. H. (2004) *Mater.* 

*Smart Sensors, Actuators and MEMS Conference*, Prague, April 2011.

*based on different semiconductor thin films, Proceedings of the SPIE Microtechnologies -* 

*microtechnologies for NASA and terrestrial applications, Proceedings of IEEE sensors*,

Foti, G. (2001) *Applied Surface Science* Vol.184, pp.20–26.

*Microsyst. Technol*., In press.

Irvine, November 2005.

*Solid State,* pp.104-105.

*Sci. Forum*, Vol.457-460, pp.3-8.

*Technology*, Vol.149, pp.76–81.

*Technology*, Vol.174-175, pp.365-369.

*Processing, 2nd edn*, New York: Wiley.

PhD thesis, Universiteit Antwerpen.

Press, 257p.

Fraga, M.A (2011a) *Materials Science Forum*, Vol. 679-680, pp.217-220.

*State,* Vol.8, pp. 406-407 as cited in (Okoije, 2002).

Vol. 10, pp.83-84 as cited in (Shor et al., 1993).

*Electrical Engineering*, University of São Paulo.

Johns, G.K. (2005) *Journal of Applied Engineering Mathematics*, Vol. 2, pp. 1-5.

930.

#### **7. Summary**

It is notable that in recent years significant advances have been made in the SiC thin film technology for piezoresistive sensors application. These advances include improvement of deposition techniques to optimize the electrical, mechanical and piezoresistive properties of crystalline and amorphous SiC films which have enabled the development of sensors appropriate for harsh environments with costs lower than those based on SiC bulk.

This chapter reviewed the concepts of piezoresistivity, presented a brief survey on the studies of piezoresistive properties of SiC films, described the main techniques that are being used to deposit SiC films for MEMS sensor applications, discussed when and why to use SiC and what are the requirements that SiC films must attain to be applied successfully in piezoresistive sensors. Futhermore, it was shown examples of SiC film based pressure sensors and accelerometers.

#### **8. Acknowledgments**

The authors acknowledge the financial support of Brazilian agencies: program PNPD-CAPES (process number 02765/09-8), CNPq (process number 152912/2010-0) and AEB. We also would like to thank the institutions that have provided their infrastructure for the experiments: Plasma and Processes Laboratory of the Technological Institute of Aeronautics, Microfabrication Laboratory of the Brazilian Synchrotron Light Laboratory (LMF-LNLS)*,*  Institute for Advanced Studies (IEAv), Center of Semiconductor Components (CCS-UNICAMP), Faculty of Technology of São Paulo (FATEC-SP) and Associate Laboratory of Sensors (LAS-INPE)*.*

#### **9. References**


Foti, G. (2001) *Applied Surface Science* Vol.184, pp.20–26.

386 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

It is notable that in recent years significant advances have been made in the SiC thin film technology for piezoresistive sensors application. These advances include improvement of deposition techniques to optimize the electrical, mechanical and piezoresistive properties of crystalline and amorphous SiC films which have enabled the development of sensors

This chapter reviewed the concepts of piezoresistivity, presented a brief survey on the studies of piezoresistive properties of SiC films, described the main techniques that are being used to deposit SiC films for MEMS sensor applications, discussed when and why to use SiC and what are the requirements that SiC films must attain to be applied successfully in piezoresistive sensors. Futhermore, it was shown examples of SiC film based pressure

The authors acknowledge the financial support of Brazilian agencies: program PNPD-CAPES (process number 02765/09-8), CNPq (process number 152912/2010-0) and AEB. We also would like to thank the institutions that have provided their infrastructure for the experiments: Plasma and Processes Laboratory of the Technological Institute of Aeronautics, Microfabrication Laboratory of the Brazilian Synchrotron Light Laboratory (LMF-LNLS)*,*  Institute for Advanced Studies (IEAv), Center of Semiconductor Components (CCS-UNICAMP), Faculty of Technology of São Paulo (FATEC-SP) and Associate Laboratory of

Allameh, S. M.; Soboyejo, W.O.; Srivatsan, T.S. (2006) *Silicon-Based Microelectromechanical* 

Atwell, A.R.; Okojie, R.S., Kornegay, K.T.; Roberson, S.L.; Beliveau, A. (2003) *Sensors and* 

Bogaerts A.; Neyts E.; Gijbels R.; Mullen J. (2003) *Spectrochimica* Acta Part B, Vol.57, pp.609–

Chaudhuri, J.; Ignatiev, K.; Edgar, J.H.; Xie, Z.Y.; Gao, Y.; Rek, Z. (2000) *Mater. Sci. Eng. B*,

Chowdhury, I.; Chandrasekhar, M.V.S.; Klein, P.B.; Caldwell, J.D.; Sudarshan, T. (2011)

Doi, I. (2006) *Técnicas de deposição: CVD, notas de aula da disciplina IE726 – Processos de Filmes* 

Eickhoff, M.; Möller, M.; Kroetz, G.; Stutzmann, M. (2004) *J. Appl. Phys*., Vol.96, pp. 2872-2877. Fissel, A.; Kaiser, U.; Ducke, E.; Schroter, B.; Richter, W. (1995) *J. Cryst. Growth*, Vol.154,

Cimalla, V.; Pezoldt, J.; Ambacher, O. (2007) J. *Phys. D: Appl. Phys.*, Vol.40, pp.6386-6434. Cocuzza, M. (2003) *Development of silicon and silicon carbide-based micro-electromechanical* 

Camassel, J.; Juillaguet, S. (2007) J. *Phys. D: Appl. Phys.*, Vol.40, pp.6264-6277.

*Systems (Si-MEMS)*, In: *Advanced Structural Materials: Properties, Design Optimization,* 

appropriate for harsh environments with costs lower than those based on SiC bulk.

**7. Summary** 

sensors and accelerometers.

**8. Acknowledgments** 

Sensors (LAS-INPE)*.*

658.

pp.72-80.

*and Applications*, pp.63-94.

Vol.76, pp.217-224.

*Actuators A: Physical*, Vol.104, pp.11-18. Berg, S.; Nyberg, T. (2005) *Thin Solid Films*, Vol.476, pp.215-230.

*Journal of Crystal Growth*, Vol.316, pp.60-66.

*systems*, PhD thesis, Polytechnic of Turin.

*Finos*, Universidade Estadual de Campinas.

**9. References** 


**0**

**16**

**Opto-Electronic Study of SiC Polytypes:**

**Simulation with Semi-Empirical**

<sup>1</sup>*Department of Physics and Astronomy, King Saud University,*

*and Department of physics, National Taiwan University, Taipei 106*

<sup>2</sup>*Université de Lyon, CNRS, Ecole Normale Supérieure de Lyon, Institut de Chimie de*

The recent growing scientific and technological interest on silicon carbide (SiC) arises from its peculiar physical properties, i.e., its mechanical, and chemical stability. Moreover, SiC is considered to be a promising material for electronic and optical devices. Microelectronic devices made of SiC can be used in high-power, high-speed, high temperature, high-frequency, and even hard-radiation application (1)-(4). The strong bonding between Si and C atoms in SiC makes this material very resistant to high temperature and radiation damage. In view of this extraordinary application potential a thorough knowledge of the structural and electronic properties of SiC is a matter of both ionic interest and technological importance. In addition to its traditional use as an abrasive (carborundum) there is currently much interest in materials made from SiC fibres, which compare well with their carbon fibre counterparts. Over a two hundred chemically stable semiconducting polytypes of SiC exist, they have a high bulk modulus and generally wide band gap. From such difference in stacking order it is possible to get almost 200 different crystal structures (1)-(10) of which the two extremes are the pure cubic polytype (with zinc blende structure) and the pure hexagonal one (with wurtzite structure). SiC is the most prominent of a family of close packed materials which exhibit a one dimensional polymorphism called polytypism. In addition, numerous hexagonal and rhombohedral structures (11)-(19) of SiC have been identified in addition to the common cubic form. In fact, SiC is one of the few compounds which form many stable and long-range ordered modifications, so-called polytypes (11)-(17). Previously, SiC has been subject to many theoretical studies. With this respect, a variety of structural, electronic and optical properties in SiC have been investigated by many theoretical groups (12)-(15) and the results can be related to the experimental works (7)-(10). In the last years, first-principle calculations have been applied to determine the ground-state properties of cubic and hexagonal polytypes of SiC (19)-(53). Based on previous theoretical works, the high-pressure behavior (18)-(33), and the effect of atomic relaxation on structural properties

**1. Introduction**

**Tight-Binding Approach**

Amel Laref1 and Slimane Laref2

*Riyadh 11451, Saudi Arabia*

<sup>1</sup>*Taiwan* <sup>2</sup>*France*

*Lyon, Laboratoire de Chimie, Lyon*


### **Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach**

Amel Laref1 and Slimane Laref2

<sup>1</sup>*Department of Physics and Astronomy, King Saud University, Riyadh 11451, Saudi Arabia and Department of physics, National Taiwan University, Taipei 106* <sup>2</sup>*Université de Lyon, CNRS, Ecole Normale Supérieure de Lyon, Institut de Chimie de Lyon, Laboratoire de Chimie, Lyon* <sup>1</sup>*Taiwan* <sup>2</sup>*France*

#### **1. Introduction**

388 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Okojie, R.S.; Ned, A.A.; Kurtz, A.D.; Carr, W.N. (1996) α(6H)-*SiC pressure sensors for high* 

Okojie, R.S.; Ned, A.A.; Kurtz, A.D.; Carr, W.N. (1998a) *IEEE Trans. Elec. Dev.*, Vol. 45,

Okojie, R.S.; Ned, A.A.; Kurtz, A.D. (1998b) *Sensors and Actuators A: Physical*, Vol.66,

Okojie, R.S. (2002) *Fabrication and characterization of single-crystal silicon carbide MEMS.* In: *MEMS Handbook*, Mohamed Gad-el-Hak, pp.20.1-20.31, CRC Press.

Ohring, M. (2002) *Material Science of Thin Films: Deposition & Structure*, San Diego, CA :

Rajab, S.M.; Oliveira, I.C.; Massi, M.; Maciel, H.S.; dos Santos Filho, S.G.; Mansano, R.D.

Rajagopalan, T.; Wang, X.; Lahlouh, B.; Ramkumar, C.; Dutta, P.; Gangopadhyay, S. (2003)

Rajaraman, V.; Pakula, L.S.; Yang, H.; French, P.J.; P. M. Sarro (2011) *Int J Adv Eng Sci Appl* 

Rapatskaya, I.V.; Rudashevskii, G.E.; Kasaganova, M.G.; Islitsin, M.I.; Reifman, M.B.; E.F. Fedotova, E.F. (1968) *Sov. Phys. Solid State*, Vol. 9, pp. 2833-2835 as cited in Okoije, 2002. Singh, R.; Ngo, L.L.; Seng H.S.; Mok, F.N.C. (2002) *A Silicon piezoresistive pressure sensor,* 

Shor, J.S.; Goldstein, D.; Kurtz, A.D. (1993) *IEEE Trans. Elec. Dev.*, Vol. 40, pp.1093-1099.

*Conference on Solid State Sensors and Actuators,* Vol. 2, pp.1439-1442.

Wright, N. G.; Horsfall, A. B. (2007) *J. Phys. D: Appl. Phys.*, Vol. 40, pp.6345-6354.

Toriyama, T.; Sugiyama S. (2002) *Appl. Phys. Lett.*, Vol. 81, pp.2797-2799.

Window, A.L. (1992) *Strain Gauge Technology*, Springer, London.

*Carbide Micro Electromechanical Systems*, pp. 18-45.

*Engineering B*, Vol.114-115, pp. 279-283.

Stamate, M.D.; Lazar, I.; Lazar, G. (2008) *Journal of Non-Crystalline Solids,* Vol.354, pp.61–64. Stanescu, C.D.; Voican, C. (2007) *Accelerated Stress Testing of SiC Pressure Tranduce, In:* 

Strass, J.; Eickhoff, M.; Kroetz, G. (1997) *The influence of crystal quality on the piezoresistive* 

Vandelli, N. (2008) *SiC MEMS Pressure Sensors for Harsh Environment Applications, MicroNano*

Wu, C.H.; Stefanescu, H.; Zorman, C. A.; Mehregany, M. (2001) *Fabrication and Testing of Single Crystalline 3C-SiC piezoresistive Pressure Sensors*, In: *Eurosensors XV.* Wu C.H.; Zorman C.A..; Mehregany M. (2006) *IEEE Sensors Journal*, Vol.6, pp. 316-324. Zanola,P.; Bontempi, E.; Ricciardi, C.; Barucca, G.; Depero, L.E. (2004) *Materials Science and* 

Zhao, D.; Mourey, D.A.; Jackson, T.N. (2010) *Journal of Electronic Materials*, Vol. 39, No. 5, pp.

Ziermann, R.; von Berg, J.; Reichert, W.; Obermeier, E.; Eickhoff, M.; Krotz, G. (1997) *A* 

*high temperature pressure sensor with β-SiC piezoresistors on SOI substrates, International Conference on Solid State Sensors and Actuators*,Chicago, June 1997. Zorman, C.A.; Fu, X.; Mehregany M. (2006) *Deposition techniques for SiC MEMS In: Silicon* 

*Proceedings of the First IEEE International Workshop on Electronic Design, Test and* 

*Proceedings of Fascicle of Management and Technological Engineering,* Vol. VI, pp.779-784.

*effect of β-SiC between RT and 450°C measured by using microstructures*, In: *International* 

Pearson, G.L.; Read Jr., W.T.; Feldman, W.L. (1957) *Acta Mettalurgica*, Vol. 5, pp.181-191.

pp.146-149.

pp.785-790.

pp.200-204.

Academic Press, 2º edition, 794 p.

*Math*, Vol. 2, pp. 28-34

*Applications*, pp.181-184.

News, pp.10-12.

554-558.

Smith, C.S. (1954) *Physics Review,* Vol. 94, pp. 42–49.

Peng, X.; Matthews, A.; Xue, S. (2011) J. *Mater. Sci.*, Vol.46, pp.1–37.

*Journal of Applied Physics*, Vol.94(8), pp.5252-5260.

(2006) *Thin Solid Films*, Vol.515, pp.170-175.

*temperature applications*, Micro Electro Mechanical Systems (MEMS '96) Proceedings,

The recent growing scientific and technological interest on silicon carbide (SiC) arises from its peculiar physical properties, i.e., its mechanical, and chemical stability. Moreover, SiC is considered to be a promising material for electronic and optical devices. Microelectronic devices made of SiC can be used in high-power, high-speed, high temperature, high-frequency, and even hard-radiation application (1)-(4). The strong bonding between Si and C atoms in SiC makes this material very resistant to high temperature and radiation damage. In view of this extraordinary application potential a thorough knowledge of the structural and electronic properties of SiC is a matter of both ionic interest and technological importance. In addition to its traditional use as an abrasive (carborundum) there is currently much interest in materials made from SiC fibres, which compare well with their carbon fibre counterparts. Over a two hundred chemically stable semiconducting polytypes of SiC exist, they have a high bulk modulus and generally wide band gap. From such difference in stacking order it is possible to get almost 200 different crystal structures (1)-(10) of which the two extremes are the pure cubic polytype (with zinc blende structure) and the pure hexagonal one (with wurtzite structure). SiC is the most prominent of a family of close packed materials which exhibit a one dimensional polymorphism called polytypism. In addition, numerous hexagonal and rhombohedral structures (11)-(19) of SiC have been identified in addition to the common cubic form. In fact, SiC is one of the few compounds which form many stable and long-range ordered modifications, so-called polytypes (11)-(17). Previously, SiC has been subject to many theoretical studies. With this respect, a variety of structural, electronic and optical properties in SiC have been investigated by many theoretical groups (12)-(15) and the results can be related to the experimental works (7)-(10). In the last years, first-principle calculations have been applied to determine the ground-state properties of cubic and hexagonal polytypes of SiC (19)-(53). Based on previous theoretical works, the high-pressure behavior (18)-(33), and the effect of atomic relaxation on structural properties

can be treated where the transferability of the hopping parameters is required. Section 4 deals with some of our recent results of the electronic and optical properties of SiC polytypes. To reach these, we analyzed this statement in terms of optoelectronic properties of SiC polytypes.

Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach 391

**2. A review of the large band-gap SiC based semiconductor device technology**

The recent surge of activity in wide band-gap semiconductors has arisen from the need for electronic devices capable of operation at high power levels, and high-radiation resistance, and separately, a need for optical materials, especially emitters, which are active in the blue and ultraviolet (UV) wavelengths (63)-(69). In this aspect, there has been renewed interest in SiC as one of the wide band gap compounds with great potential for the next generation of electronic devices operating at high temperature (61)-(68). This compound has been also used primarily in light-emitting diodes. SiC's intrinsic material properties as well as its existence in various polytypes have led to a revival of technological interest. Crystal growth of SiC polytypes has recently shown considerable progress, the expectation now being that the manufacturing of different electronic devices becomes feasible. The wide band-gap semiconductor SiC, with its excellent thermal conductivity, large breakdown fields, and resistance to chemical attack, will be the material of choice for these applications. Realized prototype power devices of SiC, like rectifier diodes, and junction field-effect transistors, show indeed encouraging performance results under extreme conditions (54)-(66). In the optical device arena, the ever increasing need for higher density optical storage and full color display technologies are driving researchers to develop wide band-gap semiconductor emitting technologies which are capable of shorter wavelength operation. Since the different energy gap values of SiC all happen to lie in the visible range of the spectrum, SiC is an interesting optical device material. Indeed, blue light emitting diodes were the first electronic SiC devices which found a good sale. Some SiC polytypes are in addition most promising as photodetective material sensitive to ultraviolet radiation. SiC is a good candidate for a short wave length diode laser. Prototype transistors have been fabricated from SiC, and the microwave and high temperature performance of SiC transistors have been studied. Devices like field effect transistors, bipolar storage capacitors, and ultraviolet detectors have been fabricated (57)-(64). SiC has a relatively high atomic bonding energy which is responsible for its mechanical strength and chemical stability at high temperatures. This material can without major difficulty, be crystallized in several polytypes, primarily due to similar geometric structures and atomic bonds (1)-(11). The different stacking of C-Si bilayers remarkably influences the properties of SiC. The most pronounced example concerns their electronic structure. Hence, a controlled epitaxial growth of different polytypes on each other would lead to high-quality heterostructures of chemical identical material with a locally adjustable band gap (7)-(14). Meanwhile, growth of heterocrystalline structures seems to be possible (4), but exhibits problems with the reproducibility and the crystal quality. Another possibility to create a combination of two polytypes is a solid-solid phase transition, which transforms one polytype into another one (6)-(8). However, polytypism also gives some advantages for constructing electronic devices, for example homo-material heterostructures. Quantum wells can be made by embedding a SiC polytype in another polytype with a wider gap(55)-(60). Among the SiC polytypes, 6H is most easily prepared and best studied, while the 3C and 4H polytypes are attracting more attention for their superior electronic properties. The very simple structure 2H is, in fact, very rarely produced by the employed growth techniques. Already, commercial applications have been done but most of the developments in industry

Finally in section 5 we summarize and conclude.

were also investigated (14)-(18). Some attempts towards the explanation of the existence of a large number of metastable SiC polytypes have been also undertaken (14)-(37). The electronic band structures of some SiC polytypes have been calculated by several groups (14)-(47). Further studies went deep into the optical properties of SiC polytypes (14)-(33). The optical and spectroscopic properties of SiC polymorphs have also been investigated by many groups both experimentally (7)-(14) and theoretically (19)-(25). Due to the problem of sample availability, most measurements were on 6H-SiC and 3C-SiC (54)-(57). Very recently, some measurements on 4H-SiC have also been reported (58)-(61). There are considerable variations in the measured optical properties mainly because the photon energy is limited to less than 6.6 eV using the popular ellipsometry technique. The use of vacuum-ultraviolet (VUV) spectroscopy can extend the energy range significantly and so far has only been carried out on 6H-SiC (57). Recent advances in crystal growth of SiC have allowed the study of the optical properties of different polytypes (54)-(60). In addition, tight-binding (TB) method has proven to be very useful for the study of both semiconductors and metallic systems, especially in systems which are too large to be studying via ab-initio techniques. This method is about 2 or 3 orders of magnitude faster than the ab initio formulations, and at the same time it describes with suitable accuracy the electronic structure of the systems. The computational efficiency of the TB method derives from the fact that the Hamiltonian can be parametrized. Furthermore, the electronic structure information can be easily extracted from the TB hamiltonian, which, in addition, also contains the effects of angular forces in a natural way. In order to use a more realistic method, we present a TB model with *sp*3*s*\* basis, representing exact curvatures of lowest conduction bands. The TB approach is standard and widely used for the electronic properties of a wide variety of materials. In the present contribution we overview our most recent results on the electronic structures and optical properties of SiC polytypes (62). Hence, the SiC polytypes can be considered as natural superlattices, in which the successive layers consist of Hexagonal SiC material of possibly different width. Our TB model can treat SiC polytypes as superlattices consisting hexagonal bulk-like blocks. We have investigated to which extent it is acceptable approximation for existing polytypes when various of nH-SiC crystal are used to present polytype superlattices. Indeed, this is an accurate approximation by building blocks consist of n-layers of nH-SiC. By representing in general the polytypes as superlattices, we have applied our recent TB model (62) that can treat the dimensions of the superlattice. Within this model we take for each sublayer linear combination of atomic orbitals of hexagonal SiC which are subsequently matched at the interfaces to similar combinations in the adjacent sublayers by using the boundary conditions. Polytypic superlattices, in comparison with heterostructure superlattices, have two important additional features, namely (i) the polytypes are perfectly lattice-matched superlattices and (ii) the polytypes have an energy band offset between adjacent layers equal to zero by definition. We can obtain with our TB model the band structures and particularly the energy band gaps of SiC polytypes and their wave functions. Our recent TB model (62) is very efficient when extended it to investigate the electronic properties of wurtzite (wz) superlattices in (0001) direction.

This chapter is organized as follows: Section 1 provides a review for the large band-gap SiC based semiconductor device technology. In the next section we present the different polytism of SiC. A fundamental concept of the TB theory for SiC polytypes is described in section 3. Our recent TB model is specifically applied to study the electronic and optical properties of SiC polytypes and it can be applied to nH-SiC wurtzite superlattices. The present approach is also suited for all wurtzite semiconductor superlattices and large complex unit cells which 2 Silicon carbide

were also investigated (14)-(18). Some attempts towards the explanation of the existence of a large number of metastable SiC polytypes have been also undertaken (14)-(37). The electronic band structures of some SiC polytypes have been calculated by several groups (14)-(47). Further studies went deep into the optical properties of SiC polytypes (14)-(33). The optical and spectroscopic properties of SiC polymorphs have also been investigated by many groups both experimentally (7)-(14) and theoretically (19)-(25). Due to the problem of sample availability, most measurements were on 6H-SiC and 3C-SiC (54)-(57). Very recently, some measurements on 4H-SiC have also been reported (58)-(61). There are considerable variations in the measured optical properties mainly because the photon energy is limited to less than 6.6 eV using the popular ellipsometry technique. The use of vacuum-ultraviolet (VUV) spectroscopy can extend the energy range significantly and so far has only been carried out on 6H-SiC (57). Recent advances in crystal growth of SiC have allowed the study of the optical properties of different polytypes (54)-(60). In addition, tight-binding (TB) method has proven to be very useful for the study of both semiconductors and metallic systems, especially in systems which are too large to be studying via ab-initio techniques. This method is about 2 or 3 orders of magnitude faster than the ab initio formulations, and at the same time it describes with suitable accuracy the electronic structure of the systems. The computational efficiency of the TB method derives from the fact that the Hamiltonian can be parametrized. Furthermore, the electronic structure information can be easily extracted from the TB hamiltonian, which, in addition, also contains the effects of angular forces in a natural way. In order to use a more realistic method, we present a TB model with *sp*3*s*\* basis, representing exact curvatures of lowest conduction bands. The TB approach is standard and widely used for the electronic properties of a wide variety of materials. In the present contribution we overview our most recent results on the electronic structures and optical properties of SiC polytypes (62). Hence, the SiC polytypes can be considered as natural superlattices, in which the successive layers consist of Hexagonal SiC material of possibly different width. Our TB model can treat SiC polytypes as superlattices consisting hexagonal bulk-like blocks. We have investigated to which extent it is acceptable approximation for existing polytypes when various of nH-SiC crystal are used to present polytype superlattices. Indeed, this is an accurate approximation by building blocks consist of n-layers of nH-SiC. By representing in general the polytypes as superlattices, we have applied our recent TB model (62) that can treat the dimensions of the superlattice. Within this model we take for each sublayer linear combination of atomic orbitals of hexagonal SiC which are subsequently matched at the interfaces to similar combinations in the adjacent sublayers by using the boundary conditions. Polytypic superlattices, in comparison with heterostructure superlattices, have two important additional features, namely (i) the polytypes are perfectly lattice-matched superlattices and (ii) the polytypes have an energy band offset between adjacent layers equal to zero by definition. We can obtain with our TB model the band structures and particularly the energy band gaps of SiC polytypes and their wave functions. Our recent TB model (62) is very efficient when extended it to investigate the electronic properties of wurtzite (wz) superlattices in (0001)

This chapter is organized as follows: Section 1 provides a review for the large band-gap SiC based semiconductor device technology. In the next section we present the different polytism of SiC. A fundamental concept of the TB theory for SiC polytypes is described in section 3. Our recent TB model is specifically applied to study the electronic and optical properties of SiC polytypes and it can be applied to nH-SiC wurtzite superlattices. The present approach is also suited for all wurtzite semiconductor superlattices and large complex unit cells which

direction.

can be treated where the transferability of the hopping parameters is required. Section 4 deals with some of our recent results of the electronic and optical properties of SiC polytypes. To reach these, we analyzed this statement in terms of optoelectronic properties of SiC polytypes. Finally in section 5 we summarize and conclude.

#### **2. A review of the large band-gap SiC based semiconductor device technology**

The recent surge of activity in wide band-gap semiconductors has arisen from the need for electronic devices capable of operation at high power levels, and high-radiation resistance, and separately, a need for optical materials, especially emitters, which are active in the blue and ultraviolet (UV) wavelengths (63)-(69). In this aspect, there has been renewed interest in SiC as one of the wide band gap compounds with great potential for the next generation of electronic devices operating at high temperature (61)-(68). This compound has been also used primarily in light-emitting diodes. SiC's intrinsic material properties as well as its existence in various polytypes have led to a revival of technological interest. Crystal growth of SiC polytypes has recently shown considerable progress, the expectation now being that the manufacturing of different electronic devices becomes feasible. The wide band-gap semiconductor SiC, with its excellent thermal conductivity, large breakdown fields, and resistance to chemical attack, will be the material of choice for these applications. Realized prototype power devices of SiC, like rectifier diodes, and junction field-effect transistors, show indeed encouraging performance results under extreme conditions (54)-(66). In the optical device arena, the ever increasing need for higher density optical storage and full color display technologies are driving researchers to develop wide band-gap semiconductor emitting technologies which are capable of shorter wavelength operation. Since the different energy gap values of SiC all happen to lie in the visible range of the spectrum, SiC is an interesting optical device material. Indeed, blue light emitting diodes were the first electronic SiC devices which found a good sale. Some SiC polytypes are in addition most promising as photodetective material sensitive to ultraviolet radiation. SiC is a good candidate for a short wave length diode laser. Prototype transistors have been fabricated from SiC, and the microwave and high temperature performance of SiC transistors have been studied. Devices like field effect transistors, bipolar storage capacitors, and ultraviolet detectors have been fabricated (57)-(64). SiC has a relatively high atomic bonding energy which is responsible for its mechanical strength and chemical stability at high temperatures. This material can without major difficulty, be crystallized in several polytypes, primarily due to similar geometric structures and atomic bonds (1)-(11). The different stacking of C-Si bilayers remarkably influences the properties of SiC. The most pronounced example concerns their electronic structure. Hence, a controlled epitaxial growth of different polytypes on each other would lead to high-quality heterostructures of chemical identical material with a locally adjustable band gap (7)-(14). Meanwhile, growth of heterocrystalline structures seems to be possible (4), but exhibits problems with the reproducibility and the crystal quality. Another possibility to create a combination of two polytypes is a solid-solid phase transition, which transforms one polytype into another one (6)-(8). However, polytypism also gives some advantages for constructing electronic devices, for example homo-material heterostructures. Quantum wells can be made by embedding a SiC polytype in another polytype with a wider gap(55)-(60). Among the SiC polytypes, 6H is most easily prepared and best studied, while the 3C and 4H polytypes are attracting more attention for their superior electronic properties. The very simple structure 2H is, in fact, very rarely produced by the employed growth techniques. Already, commercial applications have been done but most of the developments in industry

Fig. 2. Three-dimensional perspective views of the primitive hexagonal unit cells of the 3*C*-(zinc-blende), 2*H*-(wurtzite), 4*H*-, 6*H*-, and 8*H*-SiC polytypes. The stacking sequences ABC (3*C*), AB (2*H*), ABCB (4*H*), ABCACB (6*H*) and ABCAB ACB (8*H*) are also indicated.

(C) or hexagonal (H). With reference to figure 2, if the first Si-C layer is labelled A, the next layer that can be placed according to a closed packed structure will be placed either on B or C. The different polytypes are constructed by permutations of these three positions. In figure 2 the stacking sequence is shown for the most common polytypes, 3C, 2H, 4H, 6H, and 8H, which are very interesting for their technological applications. Three-dimensional perspective views of the primitive hexagonal unit cells of the 2H-, 3C-, 4H-, 6H-, and 8H-SiC polytypes. In the case of SiC, the basic units are tetrahedrons with a C(Si) atom at the center, surrounded by four Si(C) atoms covalently bonded: these units are periodically repeated in closed-packed hexagonal layers, whose stacking sequence gives rise to the different polytypes. Though being different in the long range order, the several polytypes show a similar local chemical environment for both the carbon and silicon species; in particular each Si(C) atom is situated above the center of a triangle of C(Si) atoms and underneath a C(Si) atom belonging to the next layer in a tetrahedral coordination. The SiC-polytypes consist of double silicon-carbon layers which are stacked on top of each other in the c-axis direction. A local arrangement of three consecutive double layers is called hexagonal, if it is like the arrangement of double layers in wurzite. It is called cubic, if the stacking arrangement is the same as for the zinc-blende structure. The basic structural elements is the SiC bilayer composed of one Si [0001] plane and the adjacent C[0001] plane. The SiC polytypes are differentiated by the stacking sequence of the tetrahedrally bonded SiC bilayers, such that the individual bond lengths and local atomic environments are nearly identical, while the overall symmetry of the crystal is determined by the stacking periodicity. Each SiC bilayer, while maintaining the tetrahedral bonding scheme of the crystal, can be situated in one of three possible positions with respect to the lattice. These are each arbitrarily assigned the notation A, B, or C. Depending on the stacking order, the bonding between Si and C atoms in adjacent bilayer planes is either of a zincblende (cubic) or wurtzite (hexagonal) nature. Zincblende bonds are rotated 60◦ with respect to nearest neighbors while hexagonal bonds are mirror images (Figure 2). Each type of bond provides a slightly altered atomic environment making some lattice sites inequivalent in polytypes with mixed bonding schemes and reducing the overall symmetry. These effects are important when considering the substitutional incorporation and electronic transport properties of

Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach 393

Fig. 1. (a) HRTEM image, displaying that the 3C/6H-SiC polytipic transformation takes place by three bilayers thin lamellae twinned along the (111) planes and bounded along the (112) planes. Image (b) presents a magnification of the area marked by the square in (a) (56).

and research laboratories focus on two hexagonal polytypes : the 6H and 4H-SiC varieties. The polytypism of SiC makes it non-trivial to grow single phase material, but it also offers some potential advantages if crystal growth methods can be developed sufficiently to capitalize on the possibility of polytype homo/heterojunctions (see figure 1).

#### **2.1 Polytypism in SiC**

SiC is a wide band gap semiconductor that can be synthesized in a variety of polytypes: polytypism, can be considered as a one dimensional variant of polymorphism (1)-(8). Indeed, while the term polymorphism generally refers to the possibility of an element or compound to crystallize in different structures, polytypes only differ for the stacking sequence of atomic layers along one crystalline direction. We include SiC in the group of polytypes because of its simplicity and the fact that its hexagonality is 100%. All various SiC-polytypes have the same stoichiometry and the same bonding configuration between next nearest neigbors. More than 200 polytypes of SiC exhibiting a wide range of properties have been reported (1). There are a lot of more complex polytypes in which the bonding arrangement (cubic vs. hexagonal) are repeated periodically. Due to that periodic repetition the SiC-polytypes are also called to be natural superlattices. However, only few of those polytypes are commonly found and those are relatively simple compared to the rest. The bandgaps differ widely among the polytypes ranging from 2.3 eV for 3C-SiC to 2.9 eV in 6H SiC to 3.3 eV for 2H SiC. In general, the greater the wurtzite component, the larger the bandgap.

A shorthand has been developed to catalogue the literally infinite number of possible polytype crystal structure. In this notation the number of layers in the stacking direction, before the sequence is repeated, is combined with the letter representing the Bravais lattice type: cubic 4 Silicon carbide

Fig. 1. (a) HRTEM image, displaying that the 3C/6H-SiC polytipic transformation takes place by three bilayers thin lamellae twinned along the (111) planes and bounded along the (112) planes. Image (b) presents a magnification of the area marked by the square in (a) (56).

the possibility of polytype homo/heterojunctions (see figure 1).

the wurtzite component, the larger the bandgap.

**2.1 Polytypism in SiC**

and research laboratories focus on two hexagonal polytypes : the 6H and 4H-SiC varieties. The polytypism of SiC makes it non-trivial to grow single phase material, but it also offers some potential advantages if crystal growth methods can be developed sufficiently to capitalize on

SiC is a wide band gap semiconductor that can be synthesized in a variety of polytypes: polytypism, can be considered as a one dimensional variant of polymorphism (1)-(8). Indeed, while the term polymorphism generally refers to the possibility of an element or compound to crystallize in different structures, polytypes only differ for the stacking sequence of atomic layers along one crystalline direction. We include SiC in the group of polytypes because of its simplicity and the fact that its hexagonality is 100%. All various SiC-polytypes have the same stoichiometry and the same bonding configuration between next nearest neigbors. More than 200 polytypes of SiC exhibiting a wide range of properties have been reported (1). There are a lot of more complex polytypes in which the bonding arrangement (cubic vs. hexagonal) are repeated periodically. Due to that periodic repetition the SiC-polytypes are also called to be natural superlattices. However, only few of those polytypes are commonly found and those are relatively simple compared to the rest. The bandgaps differ widely among the polytypes ranging from 2.3 eV for 3C-SiC to 2.9 eV in 6H SiC to 3.3 eV for 2H SiC. In general, the greater

A shorthand has been developed to catalogue the literally infinite number of possible polytype crystal structure. In this notation the number of layers in the stacking direction, before the sequence is repeated, is combined with the letter representing the Bravais lattice type: cubic

Fig. 2. Three-dimensional perspective views of the primitive hexagonal unit cells of the 3*C*-(zinc-blende), 2*H*-(wurtzite), 4*H*-, 6*H*-, and 8*H*-SiC polytypes. The stacking sequences ABC (3*C*), AB (2*H*), ABCB (4*H*), ABCACB (6*H*) and ABCAB ACB (8*H*) are also indicated.

(C) or hexagonal (H). With reference to figure 2, if the first Si-C layer is labelled A, the next layer that can be placed according to a closed packed structure will be placed either on B or C. The different polytypes are constructed by permutations of these three positions. In figure 2 the stacking sequence is shown for the most common polytypes, 3C, 2H, 4H, 6H, and 8H, which are very interesting for their technological applications. Three-dimensional perspective views of the primitive hexagonal unit cells of the 2H-, 3C-, 4H-, 6H-, and 8H-SiC polytypes. In the case of SiC, the basic units are tetrahedrons with a C(Si) atom at the center, surrounded by four Si(C) atoms covalently bonded: these units are periodically repeated in closed-packed hexagonal layers, whose stacking sequence gives rise to the different polytypes. Though being different in the long range order, the several polytypes show a similar local chemical environment for both the carbon and silicon species; in particular each Si(C) atom is situated above the center of a triangle of C(Si) atoms and underneath a C(Si) atom belonging to the next layer in a tetrahedral coordination. The SiC-polytypes consist of double silicon-carbon layers which are stacked on top of each other in the c-axis direction. A local arrangement of three consecutive double layers is called hexagonal, if it is like the arrangement of double layers in wurzite. It is called cubic, if the stacking arrangement is the same as for the zinc-blende structure. The basic structural elements is the SiC bilayer composed of one Si [0001] plane and the adjacent C[0001] plane. The SiC polytypes are differentiated by the stacking sequence of the tetrahedrally bonded SiC bilayers, such that the individual bond lengths and local atomic environments are nearly identical, while the overall symmetry of the crystal is determined by the stacking periodicity. Each SiC bilayer, while maintaining the tetrahedral bonding scheme of the crystal, can be situated in one of three possible positions with respect to the lattice. These are each arbitrarily assigned the notation A, B, or C. Depending on the stacking order, the bonding between Si and C atoms in adjacent bilayer planes is either of a zincblende (cubic) or wurtzite (hexagonal) nature. Zincblende bonds are rotated 60◦ with respect to nearest neighbors while hexagonal bonds are mirror images (Figure 2). Each type of bond provides a slightly altered atomic environment making some lattice sites inequivalent in polytypes with mixed bonding schemes and reducing the overall symmetry. These effects are important when considering the substitutional incorporation and electronic transport properties of

direction, where *c* and *a* are labelled cation and anion atoms. The *pH* (*p* = 2, 4, 6, 8,..) contains 2(2*n*) atoms in a unit cell at *Ri* with five orbitals each; |*αj* >, where *α* denotes the *s*, *x*(= *px*), *y*(= *py*), *z*(= *pz*) and *s*∗(=excited *s*) orbitals, and *j* represents the site index in a unit cell

Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach 395

For each wave vector **k** in the Brillouin zone (BZ), the Bloch functions can be constructed by

Here *ξ* is a quantum number that runs over the basis orbitals *s*, *s*\*, *px*, *py*, and *pz* on the different types of sites *α* in a unit cell. The *N* wave vectors **k** lie in the first BZ with the origin

*Cξα* (*k*, *λ*)|*ξ*,*rα*, *k* >

*λ* denotes the band index and *Cξα* (**k**, *λ*) is the eigen-wavefunction, which can be obtained by

− *E<sup>λ</sup>* (*k*) *δξξ*� *δαα*�

*Ha Hac H*<sup>+</sup>

*Hc H*0*ac*

Here, the blocks *Hc*(*a*), *Hac*, and *H*0*ac* denote intra-material interactions for *pH* (*p* = 2, 4, 6, 8,..),

�

, *H*0*ac* =

and every element represents a 5x5 matrix. The blocks *Hca* and *H*0*ca* are expressed as:

*Hac* = � *a ac ac*<sup>+</sup> *c*

*Ha Hac*

� *aa ac ca cc* �

*Hc H*0*ac* . .

�

1 2 3 .. *n* − 1 *n* 1 2 3 .. *n* (5)

*ik*.*Rl*+*ik*.*r<sup>α</sup>* <sup>|</sup>*ξ*,*rα*, *Rl* <sup>&</sup>gt; (2)

< *ξ*,*rα*, *k*|*k*, *λ* > |*ξ*,*rα*, *k* > (3)

< *ξ*,*rα*, *k*|*k*, *λ* >= 0 (4)

<sup>0</sup> *ca*

⎤ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎥ ⎦

*Ha Hac Hc*

(6)

<sup>√</sup>*<sup>N</sup>* ∑ *l e*

of the *lth* unit cell at **R***l*, and **r***<sup>α</sup>* represents the positions of the atoms in this unit cell.

which runs from 1 through 2(2*n*).

solving the Schrödinger equation.

∑ *ξ*,*α*�

> ⎡ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎢ ⎣

*H* =

*n* − 1 *n* 1 2 . *n* − 1 *n*

��*ξ*,*rα*, *<sup>k</sup>*|*H*|*ξ*�

*Hc H*0*ac Ha*

the linear combination of atomic orbitals |*ξ*,**r***α*, **R***<sup>l</sup>* > :

<sup>|</sup>*ξ*,*rα*, *<sup>k</sup>* <sup>&</sup>gt;<sup>=</sup> <sup>1</sup>

The electronic eigen-states of the *pH* (*p* = 2, 4, 6, 8,..) are expanded as :

*ξ*,*α*


= ∑ *ξ*,*α*

> ,*rα*� , *k* �

> > *Hc* . . . *Ha*

Therefore, we obtain the Hamiltonian matrix for *pH* (*p* = 2, 4, 6, 8,..).

SiC. If the stacking is ABCABC..., the purely cubic, i.e., a zinc-blende structure consisting of two interpenetrating face-centered (fcc) cubic lattices. Zincblende structure commonly abbreviated as 3C SiC (or *β*-SiC) is realized (Figure 2). 3C SiC is the only possible cubic polytype. The stacking direction of the basal planes perpendicular to the planes is in fact [111] direction of the cubic unit cell of 3C-SiC as indicated in the figure. The family of hexagonal polytypes is collectively referred to as alpha SiC. The purely wurtzite ABAB... stacking sequence is denoted as 2H SiC reflecting its two bilayer stacking periodicity and hexagonal symmetry. All of the other polytypes are mixtures of the fundamental zincblende and wurtzite bonds. Some common hexagonal polytypes with more complex stacking sequences are 4H-, 6H- and 8H- SiC (Figure 2). Since the SiC polytypes are mixtures of cubic and hexagonal stackings, a quantity defined as the hexagonality H representing the fraction of hexagonal stackings out of all the stackings (cubic + hexagonal) in a polytype is used frequently to describe how much the polytype is cubic-like or hexagonal-like in structural sense [5]. As it is obvious from the definition, the hexagonality of 2H-SiC is 100 % and that of 3C-SiC is 0 %. It is naturally expected that a polytype with a smaller H should be closer to 3C, i.e., more cubic-like than one with a larger H in other material properties as well as in structure, and this is generally true for most of the polytypes. 4*H*−, 8*H*−SiC are composed equally of cubic and hexagonal bonds, while 6*H*−SiC is two-thirds cubic. Despite the cubic elements, each has overall hexagonal symmetry. All these polytypes have higher periodicity (more Si-C bilayers) along the c-axis than 2H-SiC and they are in general called *α*-SiC together with 2H-SiC. 4Hand 6H-SiC are the most common polytypes, and single crystal wafers of these polytypes are currently available and hence all recent research for making commercial devices out of SiC are focused on these polytypes.

#### **3. Empirical tight-binding model for hexagonal and n-hexagonal systems: General formalism of the tight-binding model for (0001) wurtzite:**

The tight-binding approximation for band structure calculations uses atomic energy parameters and the expansion of the electron wave functions in terms of a linear combination of atomic orbitals (LCAO). In the LCAO method, the basic problem is to find the Hamiltonian matrix elements between the various basis states, as in the original paper of Slater and Koster (70); the matrix elements can be written for the basis functions *sp*<sup>3</sup> considering various possible interactions. In our recent calculations, a standard semi-empirical *sp*3*s*\* tight-binding method (71) has been employed and the matrix elements are parametrized in order to reproduce the principal features to know the band structures.

The general form of the Hamiltonian is (72).

$$H(k) = \sum\_{bb',l} \sum\_{a\beta} e^{ik.R^l\_{bb'}} E^{bb'}\_{a\beta} \left( R^l\_{bb'} \right) \tag{1}$$

where *l* labels the sublayers, *b* and *b*� refer to the atomic basis within a sublayer, and *α* and *β* are atomiclike orbitals. Given the *Ebb*� *αβ* 's (bulk band structure) and the **<sup>R</sup>***<sup>l</sup> bb*�'s (SL geometry), we can construct the Hamiltonian matrix and diagonalize it directly for the eigensolutions.

In our recent study, we have performed a TB method with an *sp*3*s*<sup>∗</sup> basis set (71). We used the nearest-neighbor TB parameters with a basis of five orbitals (*s*, *px*, *py*, *pz*, and *s*\*) per atom. We have derived a TB Hamiltonian *pH* (*p* = 2, 4, 6, 8,..) for different polytypes of SiC from the wz TB model. The label *pH* (*p* = 2, 4, 6, 8,..) is the hexagonality for different polytypes. Consider a TB Hamiltonian of two different alternating wz crystals labelled "*ca*" in (0001) 6 Silicon carbide

SiC. If the stacking is ABCABC..., the purely cubic, i.e., a zinc-blende structure consisting of two interpenetrating face-centered (fcc) cubic lattices. Zincblende structure commonly abbreviated as 3C SiC (or *β*-SiC) is realized (Figure 2). 3C SiC is the only possible cubic polytype. The stacking direction of the basal planes perpendicular to the planes is in fact [111] direction of the cubic unit cell of 3C-SiC as indicated in the figure. The family of hexagonal polytypes is collectively referred to as alpha SiC. The purely wurtzite ABAB... stacking sequence is denoted as 2H SiC reflecting its two bilayer stacking periodicity and hexagonal symmetry. All of the other polytypes are mixtures of the fundamental zincblende and wurtzite bonds. Some common hexagonal polytypes with more complex stacking sequences are 4H-, 6H- and 8H- SiC (Figure 2). Since the SiC polytypes are mixtures of cubic and hexagonal stackings, a quantity defined as the hexagonality H representing the fraction of hexagonal stackings out of all the stackings (cubic + hexagonal) in a polytype is used frequently to describe how much the polytype is cubic-like or hexagonal-like in structural sense [5]. As it is obvious from the definition, the hexagonality of 2H-SiC is 100 % and that of 3C-SiC is 0 %. It is naturally expected that a polytype with a smaller H should be closer to 3C, i.e., more cubic-like than one with a larger H in other material properties as well as in structure, and this is generally true for most of the polytypes. 4*H*−, 8*H*−SiC are composed equally of cubic and hexagonal bonds, while 6*H*−SiC is two-thirds cubic. Despite the cubic elements, each has overall hexagonal symmetry. All these polytypes have higher periodicity (more Si-C bilayers) along the c-axis than 2H-SiC and they are in general called *α*-SiC together with 2H-SiC. 4Hand 6H-SiC are the most common polytypes, and single crystal wafers of these polytypes are currently available and hence all recent research for making commercial devices out of SiC are

**3. Empirical tight-binding model for hexagonal and n-hexagonal systems: General**

The tight-binding approximation for band structure calculations uses atomic energy parameters and the expansion of the electron wave functions in terms of a linear combination of atomic orbitals (LCAO). In the LCAO method, the basic problem is to find the Hamiltonian matrix elements between the various basis states, as in the original paper of Slater and Koster (70); the matrix elements can be written for the basis functions *sp*<sup>3</sup> considering various possible interactions. In our recent calculations, a standard semi-empirical *sp*3*s*\* tight-binding method (71) has been employed and the matrix elements are parametrized in

where *l* labels the sublayers, *b* and *b*� refer to the atomic basis within a sublayer, and *α* and *β*

we can construct the Hamiltonian matrix and diagonalize it directly for the eigensolutions. In our recent study, we have performed a TB method with an *sp*3*s*<sup>∗</sup> basis set (71). We used the nearest-neighbor TB parameters with a basis of five orbitals (*s*, *px*, *py*, *pz*, and *s*\*) per atom. We have derived a TB Hamiltonian *pH* (*p* = 2, 4, 6, 8,..) for different polytypes of SiC from the wz TB model. The label *pH* (*p* = 2, 4, 6, 8,..) is the hexagonality for different polytypes. Consider a TB Hamiltonian of two different alternating wz crystals labelled "*ca*" in (0001)

*αβ* 's (bulk band structure) and the **<sup>R</sup>***<sup>l</sup>*

(1)

*bb*�'s (SL geometry),

**formalism of the tight-binding model for (0001) wurtzite:**

order to reproduce the principal features to know the band structures.

*H* (*k*) = ∑ *bb*� ,*l* ∑ *αβ e ik*.*R<sup>l</sup> bb*� *Ebb*� *αβ Rl bb*� 

The general form of the Hamiltonian is (72).

are atomiclike orbitals. Given the *Ebb*�

focused on these polytypes.

direction, where *c* and *a* are labelled cation and anion atoms. The *pH* (*p* = 2, 4, 6, 8,..) contains 2(2*n*) atoms in a unit cell at *Ri* with five orbitals each; |*αj* >, where *α* denotes the *s*, *x*(= *px*), *y*(= *py*), *z*(= *pz*) and *s*∗(=excited *s*) orbitals, and *j* represents the site index in a unit cell which runs from 1 through 2(2*n*).

For each wave vector **k** in the Brillouin zone (BZ), the Bloch functions can be constructed by the linear combination of atomic orbitals |*ξ*,**r***α*, **R***<sup>l</sup>* > :

$$|\xi\_{\prime}r\_{\alpha},k> = \frac{1}{\sqrt{N}}\sum\_{l}e^{ik\cdot R\_{l}+ik\cdot r\_{\alpha}}|\xi\_{\prime}r\_{\alpha},R\_{l}>\tag{2}$$

Here *ξ* is a quantum number that runs over the basis orbitals *s*, *s*\*, *px*, *py*, and *pz* on the different types of sites *α* in a unit cell. The *N* wave vectors **k** lie in the first BZ with the origin of the *lth* unit cell at **R***l*, and **r***<sup>α</sup>* represents the positions of the atoms in this unit cell. The electronic eigen-states of the *pH* (*p* = 2, 4, 6, 8,..) are expanded as :

$$\begin{split} |k\_{\prime}\lambda\rangle &= \sum\_{\tilde{\xi}, \mathfrak{a}} < \tilde{\xi}\_{\prime} r\_{\mathfrak{a}\prime} k |k\_{\prime}\lambda > |\tilde{\xi}\_{\prime} r\_{\mathfrak{a}\prime} k > \\ &= \sum\_{\tilde{\xi}, \mathfrak{a}} \mathsf{C}\_{\tilde{\xi}\mathfrak{a}} \left( k\_{\prime} \lambda \right) |\tilde{\xi}\_{\prime} r\_{\mathfrak{a}\prime} k > \end{split} \tag{3}$$

*λ* denotes the band index and *Cξα* (**k**, *λ*) is the eigen-wavefunction, which can be obtained by solving the Schrödinger equation.

$$\sum\_{\tilde{\xi}, \mathfrak{a}'} \left[ \left< \mathfrak{f}\_{\prime} r\_{\mathfrak{a} \prime} k | H | \mathfrak{f}^{\prime}, r\_{\mathfrak{a}'} k \right> - E\_{\lambda} \left( k \right) \delta\_{\tilde{\xi} \xi} \delta\_{\mathfrak{a} \mathbf{a}'} \right] < \mathfrak{f}\_{\prime} r\_{\mathfrak{a} \prime} k | k, \lambda > = 0 \tag{4}$$

Therefore, we obtain the Hamiltonian matrix for *pH* (*p* = 2, 4, 6, 8,..).

Here, the blocks *Hc*(*a*), *Hac*, and *H*0*ac* denote intra-material interactions for *pH* (*p* = 2, 4, 6, 8,..), and every element represents a 5x5 matrix. The blocks *Hca* and *H*0*ca* are expressed as:

$$H\_{\rm ac} = \begin{bmatrix} a & ac \\ ac^+ & c \end{bmatrix}, \; H\_{\rm 0ac} = \begin{bmatrix} aa \ ac \\ ca \ cc \end{bmatrix} \tag{6}$$

all polytypes at the centre of the BZ. The valence band maximum (VBM) is found to be at the center of the BZ at Γ point for all polytypes. The zero energy is used for all polytypes. In the case of 2H-SiC, the CBM is at the K point with two equivalent CBMs (73), (74), (75), (77), while 4HSiC has its CBM at the M point giving three equivalent CBMs (22), (25),(76), (78),(79) [Figure 4]. For 6H-SiC, the theoretical calculations predict the conduction band supplying the global CBM to be very flat along the ML line and the CBM resides at some place on the line, resulting in six equivalent CBMs (22), (25), (78),(79). This has been confirmed experimentally from the Raman scattering measurement by Colwell et al. (80). However, the exact location of the CBM and the detailed shape of conduction band affecting the determination of effective electron mass are not yet well-established, either experimentally or theoretically. There are similarities between the band structures of the hexagonal polytypes, both in the valence and the conduction bands, especially between 4H, 6H and 8H-SiC structures. A significant difference between 2H and the other three hexagonal polytyes is that in 2H-SiC the two lowest conduction bands have an intersection along MK line and that the lowest band at K point has a one-dimensional representation (in the single group representation). Both in 4H, 6H and 8H-SiC the two lowest conduction bands at K point are degenerate. The intersection in 2H-SiC makes it possible for the second lowest band at the M point to provide a global conduction band minimum at the K point with C3*<sup>v</sup>* symmetry whereas the minimum for 4H-SiC is at M (C2*v*) and for 6H-, and 8H-SiC along the ML line (also C2*<sup>v</sup>* symmetry), 44 % out from M towards L. The variation in band energy gaps is coming from the different locations of CBMs. This is related with the stacking and period of each polytype. Interestingly, it is predicted theoretically that the offsets of VBMs among different polytypes are quite small, at most 0.10-0.13 eV for the case of 2H and 3C (11),(14). In other words, the VBMs of all polytypes are similarly located in energy. This means that the considerable variation of band gap for

Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach 397

different polytypes is mainly due to the difference of CBM location.

in good agreement with the other results (73), (74), (75), (77).

Another interesting point to note in the conduction band structures of SiC polytypes is the location of second CBM. According to the calculation done by Persson et al. (26),(38), the second CBM of 3C-SiC is at the same symmetry point (X) as the first one with 2.92 eV higher in energy and this was confirmed experimentally from optical absorption measurements with slightly larger energy difference ( 3.1 eV) between the two minima (13). Persson et al. calculations also show that the three hexagonal polytypes (2H, 4H, 6H) have their second CBMs located at the M point and the energy difference between the first and second CBMs is 0.60 eV for 2H, 0.122 eV for 4H, and 1.16 eV for 6H respectively. The energy position of the second CBM in 4H-SiC has been probed experimentally by BEEM (56)-(58) and optical phonon spectra measurements (59)-(63), with measured energy that ranges 0.10-0.14 eV above the first CBM. The band gaps of several common polytypes of SiC have been measured carefully by Choyke et al. from the optical absorption or luminescence spectra of the polytypes (27). The measured band gaps range widely from 2.390 eV for 3C-SiC to 3.330 eV for 2H-SiC and lot of work has been done to understand all details of the corresponding variations. Those for 4Hand 6H-SiC which are in between the two extreme cases in structure are measured to be 3.265 eV and 3.023 eV respectively. So, from fig.4, it is clear that the valence and the conduction bands are well described. Moreover, our results are in good agreement with the experimental results (74). All energies are with reference to the top of the valence band. The results show that SiC is an indirect gap semiconductor. In addition, the calculated energy gaps of SiC are

Values of lowest indirect forbidden gaps (*Eg*) are listed in Table 1 in comparison with the available data in the literature and experimental results. Our TB model provides good results

Fig. 3. Brillouin zones of (a) cubic (b) hexagonal structures.

The diagonal elements *H*(*j* = *a*, and *c*) correspond to intra-site energies, and the others contain the nearest atomic interactions in the same layer (*Hij*) or between two neighbor layers (*H*0*ij*) perpendicular to the (0001) direction. The terms *a* and *c* are regarded as the anion and cation atoms of the SiC semiconductor. The intra-material elements in the Hamiltonian can be formed uniquely by using the corresponding bulk parameters. Our TB parameters (62) give the correct indirect and direct gap in comparison with Ref.(73) and are checked for their transferability to all considered structures by calculating the optoelectronic properties of different polytypes of SiC. This method reduces the size of the Hamiltonian matrix considerably compared with methods based on plane-wave basis and allows us to treat localized states. Our TB Hamiltonian can be generalized to the wz based SL's in (0001) direction with two different compounds and is efficient when extended it to investigate the electronic properties of wz SL's. Then, we present some of our recent results which we have obtained by our TB model for electronic and optical properties of SiC polytypes.

#### **4. Electronic and optical properties of polytypic SiC**

We start this section with some of our recent results for SiC polytypes in order to illustrate the electronic and optical properties of this system. With a TB scheme, the detailed calculations of electronic structure and optical properties of different polytypes of SiC are presented.

#### **4.1 Electronic band structures of 3C-, 2H-, 4H-, 6H-, and 8H-SiC:**

A very important aspect of the polytypism of SiC is the change in energy band structure, and how it does appear in the different polytypes. Having established the geometric structure for the polytypes, the electronic band structure was calculated along the symmetry directions (62). Figure 3 shows the BZs of cubic, and hexagonal polytypes with high symmetry points marked. The labeling of the symmetry points and the three symmetry lines out from the Γ point in the relevant hexagonal Bzs are shown in Figure 3.

The corresponding band structure of 3C-SiC is shown in figure 4. The conduction band minimum (CBM) for 3C-SiC is lying at the X point and the number of CBMs equals to three (2). The resulting TB band structures of SiC polytypes (2H, 4H, 6H, and 8H) are also represented in Figure 4 versus high-symmetry lines A-L-M-Γ-A-H-K-Γ. For all polytypes the gap is systematically identified as an indirect one. The valence band maximum is located for 8 Silicon carbide

The diagonal elements *H*(*j* = *a*, and *c*) correspond to intra-site energies, and the others contain the nearest atomic interactions in the same layer (*Hij*) or between two neighbor layers (*H*0*ij*) perpendicular to the (0001) direction. The terms *a* and *c* are regarded as the anion and cation atoms of the SiC semiconductor. The intra-material elements in the Hamiltonian can be formed uniquely by using the corresponding bulk parameters. Our TB parameters (62) give the correct indirect and direct gap in comparison with Ref.(73) and are checked for their transferability to all considered structures by calculating the optoelectronic properties of different polytypes of SiC. This method reduces the size of the Hamiltonian matrix considerably compared with methods based on plane-wave basis and allows us to treat localized states. Our TB Hamiltonian can be generalized to the wz based SL's in (0001) direction with two different compounds and is efficient when extended it to investigate the electronic properties of wz SL's. Then, we present some of our recent results which we have

obtained by our TB model for electronic and optical properties of SiC polytypes.

We start this section with some of our recent results for SiC polytypes in order to illustrate the electronic and optical properties of this system. With a TB scheme, the detailed calculations of

A very important aspect of the polytypism of SiC is the change in energy band structure, and how it does appear in the different polytypes. Having established the geometric structure for the polytypes, the electronic band structure was calculated along the symmetry directions (62). Figure 3 shows the BZs of cubic, and hexagonal polytypes with high symmetry points marked. The labeling of the symmetry points and the three symmetry lines out from the Γ

The corresponding band structure of 3C-SiC is shown in figure 4. The conduction band minimum (CBM) for 3C-SiC is lying at the X point and the number of CBMs equals to three (2). The resulting TB band structures of SiC polytypes (2H, 4H, 6H, and 8H) are also represented in Figure 4 versus high-symmetry lines A-L-M-Γ-A-H-K-Γ. For all polytypes the gap is systematically identified as an indirect one. The valence band maximum is located for

electronic structure and optical properties of different polytypes of SiC are presented.

Fig. 3. Brillouin zones of (a) cubic (b) hexagonal structures.

**4. Electronic and optical properties of polytypic SiC**

point in the relevant hexagonal Bzs are shown in Figure 3.

**4.1 Electronic band structures of 3C-, 2H-, 4H-, 6H-, and 8H-SiC:**

all polytypes at the centre of the BZ. The valence band maximum (VBM) is found to be at the center of the BZ at Γ point for all polytypes. The zero energy is used for all polytypes. In the case of 2H-SiC, the CBM is at the K point with two equivalent CBMs (73), (74), (75), (77), while 4HSiC has its CBM at the M point giving three equivalent CBMs (22), (25),(76), (78),(79) [Figure 4]. For 6H-SiC, the theoretical calculations predict the conduction band supplying the global CBM to be very flat along the ML line and the CBM resides at some place on the line, resulting in six equivalent CBMs (22), (25), (78),(79). This has been confirmed experimentally from the Raman scattering measurement by Colwell et al. (80). However, the exact location of the CBM and the detailed shape of conduction band affecting the determination of effective electron mass are not yet well-established, either experimentally or theoretically. There are similarities between the band structures of the hexagonal polytypes, both in the valence and the conduction bands, especially between 4H, 6H and 8H-SiC structures. A significant difference between 2H and the other three hexagonal polytyes is that in 2H-SiC the two lowest conduction bands have an intersection along MK line and that the lowest band at K point has a one-dimensional representation (in the single group representation). Both in 4H, 6H and 8H-SiC the two lowest conduction bands at K point are degenerate. The intersection in 2H-SiC makes it possible for the second lowest band at the M point to provide a global conduction band minimum at the K point with C3*<sup>v</sup>* symmetry whereas the minimum for 4H-SiC is at M (C2*v*) and for 6H-, and 8H-SiC along the ML line (also C2*<sup>v</sup>* symmetry), 44 % out from M towards L. The variation in band energy gaps is coming from the different locations of CBMs. This is related with the stacking and period of each polytype. Interestingly, it is predicted theoretically that the offsets of VBMs among different polytypes are quite small, at most 0.10-0.13 eV for the case of 2H and 3C (11),(14). In other words, the VBMs of all polytypes are similarly located in energy. This means that the considerable variation of band gap for different polytypes is mainly due to the difference of CBM location.

Another interesting point to note in the conduction band structures of SiC polytypes is the location of second CBM. According to the calculation done by Persson et al. (26),(38), the second CBM of 3C-SiC is at the same symmetry point (X) as the first one with 2.92 eV higher in energy and this was confirmed experimentally from optical absorption measurements with slightly larger energy difference ( 3.1 eV) between the two minima (13). Persson et al. calculations also show that the three hexagonal polytypes (2H, 4H, 6H) have their second CBMs located at the M point and the energy difference between the first and second CBMs is 0.60 eV for 2H, 0.122 eV for 4H, and 1.16 eV for 6H respectively. The energy position of the second CBM in 4H-SiC has been probed experimentally by BEEM (56)-(58) and optical phonon spectra measurements (59)-(63), with measured energy that ranges 0.10-0.14 eV above the first CBM. The band gaps of several common polytypes of SiC have been measured carefully by Choyke et al. from the optical absorption or luminescence spectra of the polytypes (27). The measured band gaps range widely from 2.390 eV for 3C-SiC to 3.330 eV for 2H-SiC and lot of work has been done to understand all details of the corresponding variations. Those for 4Hand 6H-SiC which are in between the two extreme cases in structure are measured to be 3.265 eV and 3.023 eV respectively. So, from fig.4, it is clear that the valence and the conduction bands are well described. Moreover, our results are in good agreement with the experimental results (74). All energies are with reference to the top of the valence band. The results show that SiC is an indirect gap semiconductor. In addition, the calculated energy gaps of SiC are in good agreement with the other results (73), (74), (75), (77).

Values of lowest indirect forbidden gaps (*Eg*) are listed in Table 1 in comparison with the available data in the literature and experimental results. Our TB model provides good results

This work Experiement GW LDA+U GGA/LDA EPM ETB NTB 3*C* 2.389 2.39*<sup>c</sup>* 2.38*<sup>b</sup>* 2.52*<sup>d</sup>* 1.5*e*, 1.27*<sup>f</sup>* 2.30*<sup>g</sup>* 2.47*<sup>h</sup>* 1.33*<sup>i</sup>*

Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach 399

*<sup>a</sup>* Experimental data (75) *<sup>b</sup>*GW calculations (73) *<sup>c</sup>* Experimental data(74) *<sup>d</sup>* LMTO calculations (78) *<sup>e</sup>*GGA calculations (79) *<sup>f</sup>* LDA calculations (25) *<sup>g</sup>* EPM calculations (22) *<sup>h</sup>* ETB values (76) *<sup>i</sup>* NTB values (77)

which agree with other calculations (22), (25), (73), (75), (76), (77), (78),(79) and experimental data (74). The general findings that all considered polytypes are indirect semiconductors are not surprising, including that the conduction-band minimum is located at X point in the zinc-blende structure or at M in the hexagonal BZ of 2H. Diamond and silicon show a similar behavior, there the conduction-band minima are situated on the ΓX line near X [110]. The *X* point in the fcc BZ represents the position of the minimum in the zinc-blende 3C-SiC. Two of these X points are folded onto M points of the hexagonal BZ of the corresponding 2H structure. The exact positions depend on the details of the calculations, the ratio c/a of the hexagonal lattice constants, as well as the atomic positions within the hexagonal unit cells. Moreover, the upper valence band has the lowest energy in X, so that the repulsive interaction between the lowest conduction band and the highest valence band should be small. In the wurtzite structure, the situation is changed. First of all, the zinc-blende X is folded onto 2/3LM in the hexagonal BZ of 2H. This point has a lower symmetry and the bonding and antibonding combinations of the C 2s orbital and a Si 3p orbital, of which the state mainly consists, can interact with more closer lying states. The minimum at K point, that has a similar orbital character as the states at the zinc blende W point, gives rise to the lowest empty band. The energetical distance of the valence and conduction bands in K point is remarkably reduced. The resulting stronger interaction pushes the conduction-band minimum away from the valence bands. States on the LM line near M point form the lowest conduction-band minimum. Surely, the minimum in the wurtzite structure 2H-SiC is located at the **k** point in the center of the BZ edge parallel to the *c* axis similarly to hexagonal diamond (79). We find the conduction band minima at M point for 4H and, respectively, at about 0.63LM for 6H and 8H. This result is somewhat surprising since the fcc X point should map onto 1/3LM for 4H and M for 6H and 8H. That means that the simplifying folding argument is not exactly valid going from one polytype to another one. The actual arrangement of atoms and bonds in the unit cells gives rise to changes in the band positions and dispersion. The exact minimum position is particularly sensitive to the details of the atomic structure since the lowest conduction band between L and M is rather flat. This flatness increases with the lowering of the LM distance in k space. Increasing the period of the superstructure along the optical axis (line Γ-A in the BZ) causes band folding, which can be seen for the Γ-A and K-H

2*H* 3.33 3.30*<sup>a</sup>* 3.31*<sup>b</sup>* 3.33*<sup>d</sup>* 2.10*<sup>f</sup>*

8*H* 2.86 2.86*<sup>a</sup>* 2.84*<sup>b</sup>*

Table 1. Values of indirect gap of SiC polytypes (62).

4*H* 3.20 3.19*<sup>a</sup>* 3.18*<sup>b</sup>* 3.16*<sup>d</sup>* 2.57*e*, 2.18*<sup>f</sup>* 3.20*<sup>g</sup>* 6*H* 2.86 2.85*<sup>a</sup>* 2.84*<sup>b</sup>* 2.90*<sup>d</sup>* 2.28*e*, 1.96*<sup>f</sup>* 2.99*<sup>g</sup>*

Fig. 4. Band structures for 3C-, 2H-, 4H-, 6H-, and 8H-SiC calculated by our *sp*3*s*\* TB model (62).


Table 1. Values of indirect gap of SiC polytypes (62).

10 Silicon carbide

Fig. 4. Band structures for 3C-, 2H-, 4H-, 6H-, and 8H-SiC calculated by our *sp*3*s*\* TB model

(62).

*<sup>a</sup>* Experimental data (75) *<sup>b</sup>*GW calculations (73) *<sup>c</sup>* Experimental data(74) *<sup>d</sup>* LMTO calculations (78) *<sup>e</sup>*GGA calculations (79) *<sup>f</sup>* LDA calculations (25) *<sup>g</sup>* EPM calculations (22) *<sup>h</sup>* ETB values (76) *<sup>i</sup>* NTB values (77)

which agree with other calculations (22), (25), (73), (75), (76), (77), (78),(79) and experimental data (74). The general findings that all considered polytypes are indirect semiconductors are not surprising, including that the conduction-band minimum is located at X point in the zinc-blende structure or at M in the hexagonal BZ of 2H. Diamond and silicon show a similar behavior, there the conduction-band minima are situated on the ΓX line near X [110]. The *X* point in the fcc BZ represents the position of the minimum in the zinc-blende 3C-SiC. Two of these X points are folded onto M points of the hexagonal BZ of the corresponding 2H structure. The exact positions depend on the details of the calculations, the ratio c/a of the hexagonal lattice constants, as well as the atomic positions within the hexagonal unit cells. Moreover, the upper valence band has the lowest energy in X, so that the repulsive interaction between the lowest conduction band and the highest valence band should be small. In the wurtzite structure, the situation is changed. First of all, the zinc-blende X is folded onto 2/3LM in the hexagonal BZ of 2H. This point has a lower symmetry and the bonding and antibonding combinations of the C 2s orbital and a Si 3p orbital, of which the state mainly consists, can interact with more closer lying states. The minimum at K point, that has a similar orbital character as the states at the zinc blende W point, gives rise to the lowest empty band. The energetical distance of the valence and conduction bands in K point is remarkably reduced. The resulting stronger interaction pushes the conduction-band minimum away from the valence bands. States on the LM line near M point form the lowest conduction-band minimum. Surely, the minimum in the wurtzite structure 2H-SiC is located at the **k** point in the center of the BZ edge parallel to the *c* axis similarly to hexagonal diamond (79). We find the conduction band minima at M point for 4H and, respectively, at about 0.63LM for 6H and 8H. This result is somewhat surprising since the fcc X point should map onto 1/3LM for 4H and M for 6H and 8H. That means that the simplifying folding argument is not exactly valid going from one polytype to another one. The actual arrangement of atoms and bonds in the unit cells gives rise to changes in the band positions and dispersion. The exact minimum position is particularly sensitive to the details of the atomic structure since the lowest conduction band between L and M is rather flat. This flatness increases with the lowering of the LM distance in k space. Increasing the period of the superstructure along the optical axis (line Γ-A in the BZ) causes band folding, which can be seen for the Γ-A and K-H

**Mass** 3*C* 2*H* 4*H* 6*H* 8*H <sup>m</sup>*� 0.69 0.40 0.60 0.65 0.67 0.667*<sup>c</sup>* 0.42*<sup>a</sup>* 0.58*<sup>a</sup>* 0.68*<sup>a</sup>* 0.70*<sup>b</sup>* 0.43*<sup>f</sup>* 0.53*<sup>b</sup>* 0.44*<sup>b</sup>* 0.449*<sup>d</sup>* 0.58*<sup>e</sup>* 0.77*<sup>e</sup>* 0.68*<sup>f</sup>* 0.57*<sup>f</sup>* 0.75*<sup>f</sup> <sup>m</sup>*⊥<sup>1</sup> 0.25 0.24 0.36 1.19 1.38 0.247*<sup>c</sup>* 0.22*<sup>a</sup>* 0.33*<sup>a</sup>* 1.25*<sup>a</sup>* 0.24*<sup>b</sup>* 0.26*<sup>f</sup>* 0.31*<sup>e</sup>* 1.14*<sup>b</sup>* 0.23*<sup>f</sup>* 0.31*<sup>f</sup>* 1.42*<sup>e</sup>*

Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach 401

*<sup>m</sup>*⊥<sup>2</sup> 0.25 0.40 0.21 0.10 0.15 0.247*<sup>c</sup>* 0.42*<sup>a</sup>* 0.29*<sup>a</sup>* 0.13*<sup>a</sup>* 0.24*<sup>b</sup>* 0.43*<sup>f</sup>* 0.19*<sup>b</sup>* 0.43*<sup>e</sup>* 0.23*<sup>f</sup>* 0.28*<sup>e</sup>* 0.24*<sup>f</sup>* 0.28*<sup>f</sup>*

Table 2. Effective masses of electrons in the conduction-band-minima. All values in units of

*<sup>b</sup>* EPM calculations (22). *<sup>c</sup>* Experimental values (82). *<sup>d</sup>* LMTO (GW) (83). *<sup>e</sup>* LMTO (19). *<sup>f</sup>* FPLAPW (26).

reported by several groups, and they agree quite well with the theoretically calculated values (19), (26), (83). For 6H-SiC, only the longitudinal effective mass along the c-axis has been measured (82), but due to the peculiar band shape along the direction there is still a large inconsistency between the measured value and the calculated ones, even among the values calculated theoretically by different groups (19), (26), (83). However, only the hole effective masses of 4H-SiC have been measured experimentally and reported by Son et al. (81).

We have applied the tetrahedron technique directly from the eigenvalues and the angular momentum character of each state. This is done by dividing up the Brillouin zone into 48 tetrahedron cube. The total density of states (TDOS) of 3C-SiC, corresponding to the band structure is given in figure 5 (62). The 3C-SiC have valence band density of states qualitatively similar to homopolar semiconductors, except for the gap which opens at X point. This gap is related to different potential for the cation and anion potentials. This "antisymetric" gap has been proposed as a measure of crystal ionicity. The lowest states contain a low-lying C s-derived band about 17 eV below the VBM. The lowest states of the VB from -17 to -13 eV has primarily s character and is localized on the anion. The large peak at -10 to -7 eV comes primarily from the onset of the second valence band at points X and L. The states of this band is primarily of cation s character, it changes rapidly to anion p-like at the top of valence band. From the Fig 5, it is apparent that there is a significant amount of Si p hybridization all the way up to the VBM. A comparison with the corresponding DOS curve of the experimental

the free-electron mass m0 (62). *<sup>a</sup>* Experimental values (81).

**4.3 Total density of states of 3C-, 2H-, 4H-, 6H-, and 8H-SiC**

1.83*<sup>f</sup>*

directions for the SiC polytypes shown in Figure 4. The overall features of the band structures agree well with previous calculations . Differences concern the magnitudes of the various band gaps, where the effect is related to the variations in the position of the conduction-band minima. An interesting problem concerns the preparation of heterostructures on the base of chemically identical, but structural inequivalent semiconductors, more strictly speaking of different polytypes. The key parameters of such structures are the band offsets at the interface.

#### **4.2 Effective masses of 3C-, 2H-, 4H-, 6H-, and 8H-SiC**

The effective electron masses for the different polytypes have been calculated and measured experimentally by different scientific groups (19), (22), (26), (81), (82), (83). The values vary depending on the experimental techniques or model used, especially for the hexagonal polytypes. Results for the lowest conduction band minima in *K*, *M* points, or at the *LM* line near *M* point are calculated (62). For electrons we give the full inverse effective-mass tensor along the principal axis determined by the c axis of the structure and the position of the minimum in **<sup>k</sup>** space. We consider the longitudinal masses m|| parallel to the connection line between the minimum position and more strictly speaking parallel *M*Γ(4*H*), *K*Γ (2*H*) and (*LM*)<sup>Γ</sup> (6*H*, 8*H*). The two transverse masses m⊥1, m⊥<sup>2</sup> are distinguished according to the anisotropy of the system. m<sup>⊥</sup> denotes the transverse mass parallel to the c axis. In the calculation of m⊥<sup>1</sup> we use the direction *ML*. For the estimation of the second transverse mass m⊥<sup>2</sup> of the hexagonal polytypes we replace the correct direction by the line *MK* in an approximate manner.

Our previously calculated values of the electron effective masses in three principal directions with the tight-binding method (62) are presented in table 2 in comparison with other theoretical and experimental data. All values of the electron effective masses agree with experimental values, when available, and for 3C-, 2H-, and 4H-SiC, our results agree with the majority of earlier calculations (22). We report in the same table our values of *m*∗ for 8*H*-SiC. There is no available experimental data for comparison. No clear trend with the hexagonality or the extent of the unit cell can be derived from table 2 for the electron masses. This is not astonishing since the conduction band minima appear at different **k** points in the BZ. Only in 4*H* case one observes the minimum at the same point *M*. A remarkable anisotropy of the electron effective mass tensor is found for 6*H* and 4*H*. In space directions (nearly) parallel to *M*Γ and *L*Γ heavy electrons appear whereas the mass for the electron motion in the plane perpendicular to c axis but parallel to the edge *MK* of the hexagonal BZ is small. This is a consequence of the flatness of the lowest conduction bands in the most space directions. The electron-mass anisotropy in the 2*H* polytypes at *M* or *K* is much smaller. The findings for the conduction band masses have consequences for the electron mobility, since this property is proportional to the inverse mass. We expect that at least for the mostly available 6*H*− and 4*H*-SiC polytype, the current directions should be carefully selected. Otherwise, too small electron mobilities result. 2H-SiC have more parabolic behavior around their minima, whereas in 4H- and 6H-SiC the interaction between the two close-lying bands at the M point will affect the parabolicity, especially for the flat curvatures in the c direction. The best agreement between theory and experiment seems to be for the 2H-SiC pure hexagonal polytype and the 3C-SiC cubic polytype. For the 8H-SiC polytype there are not yet any experimental results for the effective electron masses. Also there is only one experimental report for the longitude effective mass of 6H-SiC. For the hole effective masses there are few theoretical reports of the polytypes and even fewer are the experimental values. The effective electron masses of 3C- and 4H-SiC have been measured experimentally and 12 Silicon carbide

directions for the SiC polytypes shown in Figure 4. The overall features of the band structures agree well with previous calculations . Differences concern the magnitudes of the various band gaps, where the effect is related to the variations in the position of the conduction-band minima. An interesting problem concerns the preparation of heterostructures on the base of chemically identical, but structural inequivalent semiconductors, more strictly speaking of different polytypes. The key parameters of such structures are the band offsets at the interface.

The effective electron masses for the different polytypes have been calculated and measured experimentally by different scientific groups (19), (22), (26), (81), (82), (83). The values vary depending on the experimental techniques or model used, especially for the hexagonal polytypes. Results for the lowest conduction band minima in *K*, *M* points, or at the *LM* line near *M* point are calculated (62). For electrons we give the full inverse effective-mass tensor along the principal axis determined by the c axis of the structure and the position of the minimum in **<sup>k</sup>** space. We consider the longitudinal masses m|| parallel to the connection line between the minimum position and more strictly speaking parallel *M*Γ(4*H*), *K*Γ (2*H*) and (*LM*)<sup>Γ</sup> (6*H*, 8*H*). The two transverse masses m⊥1, m⊥<sup>2</sup> are distinguished according to the anisotropy of the system. m<sup>⊥</sup> denotes the transverse mass parallel to the c axis. In the calculation of m⊥<sup>1</sup> we use the direction *ML*. For the estimation of the second transverse mass m⊥<sup>2</sup> of the hexagonal polytypes we replace the correct direction by the line *MK* in an

Our previously calculated values of the electron effective masses in three principal directions with the tight-binding method (62) are presented in table 2 in comparison with other theoretical and experimental data. All values of the electron effective masses agree with experimental values, when available, and for 3C-, 2H-, and 4H-SiC, our results agree with the majority of earlier calculations (22). We report in the same table our values of *m*∗ for 8*H*-SiC. There is no available experimental data for comparison. No clear trend with the hexagonality or the extent of the unit cell can be derived from table 2 for the electron masses. This is not astonishing since the conduction band minima appear at different **k** points in the BZ. Only in 4*H* case one observes the minimum at the same point *M*. A remarkable anisotropy of the electron effective mass tensor is found for 6*H* and 4*H*. In space directions (nearly) parallel to *M*Γ and *L*Γ heavy electrons appear whereas the mass for the electron motion in the plane perpendicular to c axis but parallel to the edge *MK* of the hexagonal BZ is small. This is a consequence of the flatness of the lowest conduction bands in the most space directions. The electron-mass anisotropy in the 2*H* polytypes at *M* or *K* is much smaller. The findings for the conduction band masses have consequences for the electron mobility, since this property is proportional to the inverse mass. We expect that at least for the mostly available 6*H*− and 4*H*-SiC polytype, the current directions should be carefully selected. Otherwise, too small electron mobilities result. 2H-SiC have more parabolic behavior around their minima, whereas in 4H- and 6H-SiC the interaction between the two close-lying bands at the M point will affect the parabolicity, especially for the flat curvatures in the c direction. The best agreement between theory and experiment seems to be for the 2H-SiC pure hexagonal polytype and the 3C-SiC cubic polytype. For the 8H-SiC polytype there are not yet any experimental results for the effective electron masses. Also there is only one experimental report for the longitude effective mass of 6H-SiC. For the hole effective masses there are few theoretical reports of the polytypes and even fewer are the experimental values. The effective electron masses of 3C- and 4H-SiC have been measured experimentally and

**4.2 Effective masses of 3C-, 2H-, 4H-, 6H-, and 8H-SiC**

approximate manner.


Table 2. Effective masses of electrons in the conduction-band-minima. All values in units of the free-electron mass m0 (62). *<sup>a</sup>* Experimental values (81).

*<sup>b</sup>* EPM calculations (22). *<sup>c</sup>* Experimental values (82). *<sup>d</sup>* LMTO (GW) (83). *<sup>e</sup>* LMTO (19). *<sup>f</sup>* FPLAPW (26).

reported by several groups, and they agree quite well with the theoretically calculated values (19), (26), (83). For 6H-SiC, only the longitudinal effective mass along the c-axis has been measured (82), but due to the peculiar band shape along the direction there is still a large inconsistency between the measured value and the calculated ones, even among the values calculated theoretically by different groups (19), (26), (83). However, only the hole effective masses of 4H-SiC have been measured experimentally and reported by Son et al. (81).

#### **4.3 Total density of states of 3C-, 2H-, 4H-, 6H-, and 8H-SiC**

We have applied the tetrahedron technique directly from the eigenvalues and the angular momentum character of each state. This is done by dividing up the Brillouin zone into 48 tetrahedron cube. The total density of states (TDOS) of 3C-SiC, corresponding to the band structure is given in figure 5 (62). The 3C-SiC have valence band density of states qualitatively similar to homopolar semiconductors, except for the gap which opens at X point. This gap is related to different potential for the cation and anion potentials. This "antisymetric" gap has been proposed as a measure of crystal ionicity. The lowest states contain a low-lying C s-derived band about 17 eV below the VBM. The lowest states of the VB from -17 to -13 eV has primarily s character and is localized on the anion. The large peak at -10 to -7 eV comes primarily from the onset of the second valence band at points X and L. The states of this band is primarily of cation s character, it changes rapidly to anion p-like at the top of valence band. From the Fig 5, it is apparent that there is a significant amount of Si p hybridization all the way up to the VBM. A comparison with the corresponding DOS curve of the experimental results reveals excellent agreement for energies below 4 eV. The bandwidths and energies are in good agreement with photoemission results (74).

The DOS was determined by the tetrahedron integration over a mesh that was generated by six cuts in the Γ-M direction of the BZ and included 112, 84, 78 and 56 points in the irreducible part of 2H, 4H, 6H and 8H BZ, respectively. In Figure 5 total densities of states of SiC polytypes (62) are shown which can be used for the interpretation of photoemission spectra of SiC. The lowest valence band in 6H SiC between -19.0 and -13.0 eV is dominated by *s*-electrons of C atom. The maximum at 14.85 eV in the total DOS is dominated by the *s*-electron of Si. The upper part of the valence band of 6*H* SiC is dominated mainly by the *p*-electron of C and Si. The conduction band is mainly dominated by the *s*, and *p*-electrons of Si, whereas *p*-electrons of C are less dominant. There is a noticeable difference of the p-state occupation for different polytypes and for different sites in the same polytype. The band width of the valence band agrees with previous works represented in many literatures [(22), (25), (73), (75), (77), (78)], where 18.0 eV were obtained. Our value of the valence band width (≈19.0 eV) of 6*H* SiC is lower than in cubic SiC as expected (25), (73),(78),(79).

In the figure 5, one can see that the valence band, as expected, consists of two subbands. The energy width of these subbands and the total bandwidth are very similar for the four polytypes. In *α*-SiC polytypes the lower-lying subbands is in the range from about -19.5 to -13 eV and is dominated by the atomic Si 3s+3p states and the localized atomic C 2s states, whereas the higher subband also consists of Si 3p and 2p states. In the higher subband the Si 3s and C 2p states dominate at lower energies and the C 2p states dominate at higher energies. Even if it is not straightforward to compare photoemission spectra with the DOS, the clear peak at about -11.1 eV, arising from the atomic C 2p and Si 3s states, can probably be identified with the experimental value -10.5 eV (74). Also, the total band-width and the width of the higher subband seem to be in agreement with experimental results. The calculated width of the total band (higher subband) is about 8.5 eV for all four polytypes, whereas the experimental results for *α*-SiC polytypes are about 10.0 eV (74).

Since band structures accurate close to the band gap are desired, it is useful to examine the density of states in this region. As found experimentally (74) and theoretically (25), (73),(78),(79), the major differences between the density of states of the individual SiC polytypes calculated with our TB model band structure is in the conduction bands. The results are compared with results from density-functional theory (DFT) (75). Both results of 2H-, 4H-, 6H-, and 8H-SiC show not only an increasing band gap, but an increase in the steepness of the rise in the density of states at the conduction band edge with increasing hexagonality.

#### **4.4 Optical absorption of 3C-, 2H-, 4H-, 6H-, and 8H-SiC:**

Many optical properties, such as the dielectric function, the reflectivity, absorption, etc.., are related to the band structure of cristalline solids. Most of them can be derived from the dielectric function which is measured directly and reliably by spectroscopy ellipsometry. It is worth calculating the optical absorption for different polytypes of SiC. Theoretically, the spectra are seen to be dependent on quantities such as density of states and matrix elements coupling the initial to final state. In the case of absorption spectra for bulk semiconductors, the main structures are observed to be correlated with the inter-band critical points. It is very common to assume that the dipole matrix elements involved are constant throughout the Brillouin zone and to compare the spectra directly with joint density of states (48).

We can compute the matrix elements starting from an empirical Hamiltonian even if the full wave functions are not known. The Slater-Koster method is computationally very economical Fig. 5. Density of states for the 3C-, 2H-, 4H-, 6H-, and 8H-SiC polytypes (62).

Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach 403

14 Silicon carbide

results reveals excellent agreement for energies below 4 eV. The bandwidths and energies are

The DOS was determined by the tetrahedron integration over a mesh that was generated by six cuts in the Γ-M direction of the BZ and included 112, 84, 78 and 56 points in the irreducible part of 2H, 4H, 6H and 8H BZ, respectively. In Figure 5 total densities of states of SiC polytypes (62) are shown which can be used for the interpretation of photoemission spectra of SiC. The lowest valence band in 6H SiC between -19.0 and -13.0 eV is dominated by *s*-electrons of C atom. The maximum at 14.85 eV in the total DOS is dominated by the *s*-electron of Si. The upper part of the valence band of 6*H* SiC is dominated mainly by the *p*-electron of C and Si. The conduction band is mainly dominated by the *s*, and *p*-electrons of Si, whereas *p*-electrons of C are less dominant. There is a noticeable difference of the p-state occupation for different polytypes and for different sites in the same polytype. The band width of the valence band agrees with previous works represented in many literatures [(22), (25), (73), (75), (77), (78)], where 18.0 eV were obtained. Our value of the valence band width (≈19.0 eV) of 6*H* SiC is

In the figure 5, one can see that the valence band, as expected, consists of two subbands. The energy width of these subbands and the total bandwidth are very similar for the four polytypes. In *α*-SiC polytypes the lower-lying subbands is in the range from about -19.5 to -13 eV and is dominated by the atomic Si 3s+3p states and the localized atomic C 2s states, whereas the higher subband also consists of Si 3p and 2p states. In the higher subband the Si 3s and C 2p states dominate at lower energies and the C 2p states dominate at higher energies. Even if it is not straightforward to compare photoemission spectra with the DOS, the clear peak at about -11.1 eV, arising from the atomic C 2p and Si 3s states, can probably be identified with the experimental value -10.5 eV (74). Also, the total band-width and the width of the higher subband seem to be in agreement with experimental results. The calculated width of the total band (higher subband) is about 8.5 eV for all four polytypes, whereas the

Since band structures accurate close to the band gap are desired, it is useful to examine the density of states in this region. As found experimentally (74) and theoretically (25), (73),(78),(79), the major differences between the density of states of the individual SiC polytypes calculated with our TB model band structure is in the conduction bands. The results are compared with results from density-functional theory (DFT) (75). Both results of 2H-, 4H-, 6H-, and 8H-SiC show not only an increasing band gap, but an increase in the steepness of the rise in the density of states at the conduction band edge with increasing hexagonality.

Many optical properties, such as the dielectric function, the reflectivity, absorption, etc.., are related to the band structure of cristalline solids. Most of them can be derived from the dielectric function which is measured directly and reliably by spectroscopy ellipsometry. It is worth calculating the optical absorption for different polytypes of SiC. Theoretically, the spectra are seen to be dependent on quantities such as density of states and matrix elements coupling the initial to final state. In the case of absorption spectra for bulk semiconductors, the main structures are observed to be correlated with the inter-band critical points. It is very common to assume that the dipole matrix elements involved are constant throughout the Brillouin zone and to compare the spectra directly with joint density of states (48). We can compute the matrix elements starting from an empirical Hamiltonian even if the full wave functions are not known. The Slater-Koster method is computationally very economical

in good agreement with photoemission results (74).

lower than in cubic SiC as expected (25), (73),(78),(79).

experimental results for *α*-SiC polytypes are about 10.0 eV (74).

**4.4 Optical absorption of 3C-, 2H-, 4H-, 6H-, and 8H-SiC:**

Fig. 5. Density of states for the 3C-, 2H-, 4H-, 6H-, and 8H-SiC polytypes (62).

in obtaining the full-zone band structure of semiconductors, and our procedure for the optical matrix elements requires little additional computation beyond solving the eigenvalue problem for the energies.

#### **4.4.1 Applications of optical matrix elements:**

Optical-absorption spectra in semiconductors are normally dominated by transitions from the valence to the conduction bands. Then, it is possible to compute the joint density of states (JDOS) for SiC polytypes that is given by the below formula. The purpose is to see how our TB calculations are extended to optical properties.

$$J(E) = \Omega \sum\_{cv} \int\_{FBZ} \frac{d^3k}{\left(2\pi\right)^3} \delta\left(E\_{cv}\left(k\right) - \hbar\omega\right) \tag{7}$$

Fig. 6. Joint density of states for the 3C-, 2H-, 4H-, 6H-, and 8H-SiC polytypes (62).

This chapter reviewed the general aspects of the optical properties as well as the electronic structures of SiC indirect-band-gap semiconductor polytypes. We presented our recent results of the band structures, the total densities of states, and the optical absorption of different polytypes of SiC using the empirical tight-binding approximation. The set of TB parameters is transferable to all hexagonal 2H-, 4H-, 6H-, and 8H-SiC structures. We illustrated how the tight-binding formalism can be used to accurately compute the electronic states in semiconductor hexagonal polytypes. This approximation is known to yield a sufficiently accurate conduction band and to give its minimum position correctly. To this end we have presented our new developed tight-binding model which carefully reproduces ab initio calculations and experimental results of SiC polytypes. It is likely that our TB approach could be applied to SiC polytypes with even larger unit cells than 8*H* using the variation in band gap with hexagonality in cases where experimental band gaps are undetermined. SiC system exhibits rather interesting features which differ greatly from those other semiconductor compounds with respect to optical properties as well as in electronic structure. We refer to this aspect of superperiodicity that is the possibility of modifying the optical properties of the material. Furthermore, the interband optical matrix elements can be tailored. It has been shown that our recent results indicate that our TB calculations are suitable for describing optical properties in more complex polar semi-conductors. In addition, the optical properties

Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach 405

**5. Summary and conclusions**

where Ω is the real-space unit-cell volume.

where *Ecv* = *Ec* (*k*) − *Ev* (*k*) for the JDOS per component (48).

We have computed the JDOS for SiC polytypes (62), hence, the interband transitions in Eq.(7) are all of the valence-conduction type. The interest in the JDOS lies in the fact that the momentum matrix elements are assumed constant over the Brillouin zone. The band summations in Eq.(7) involve all states in the valence band and lowest states in the conduction band. The summations in Eq.(7) are over special points in the Brillouin zone. In our calculations, we took 32, 28, 24, 20 special **k**-points for 2H, 4H, 6H and 8H respectively in the Brillouin zone (21).

#### **4.4.2 Joint density of states of 3C-, 2H-, 4H-, 6H-, and 8H-SiC:**

Before discussing the effect of the optical transition matrix elements, we consider the JDOS (see figure 6). In order to get more information on the interband transition, we present our recent calculated joint density of states for SiC polytypes (62). We have determined the transition responsible for the major contributions to these structures. This was done by finding the energy of the desired peak or shoulder on the joint density of states graph and then examining the contribution to joint density of states at that energy from the constituent interband transitions. The fundamental gap is well understood and is attributed to Γ<sup>15</sup> → *X*1transition in 3C-SiC. We examine a large peak associated to Γ<sup>15</sup> → *L*<sup>1</sup> transition which occurs at 3.8 eV. The second major peak in JDOS, comes from the transition Γ<sup>15</sup> → Γ<sup>1</sup> which occurs at 5.2 eV. However, our band structure is satisfactory with respect to these transitions. The principal behavior of the joint density of states is very similar for the various 2H-, 4H-, 6H-, 8H- polytypes considered. The two peaks below the ionic gap exhibit a different behavior with the number *n* of SiC bilayers in the unit cell. Whereas that at higher energy around 7 eV is rather independent of the polytype, the low-energy peak around 8 eV is broadened with rising number *n*. We relate this fact to the folding effect parallel to the *c* axis. It causes an opposite variation of the band curvature along the *LM* and *HK* lines in the hexagonal BZ (25), (73), (78). The intensity of the two most pronounced peaks at 3.5 and 2.1 eV in the region of the upper valence bands monotonically follow the hexagonality of the structures (25), (75), (78),(79). Strong contributions to these peaks also arise from the *LM* line. The most drastic change in the conduction band region occurs near the onset of the density of states. Its steepness over several eV again follows the hexagonality of the polytype. The particular shape of the onset however depends on the number of bilayers and therefore on the folding effect as already has been pointed out by Lee et al. (84). The consequences can be clearly seen in the joint density of states. Their low-energy tails increase with decreasing hexagonality.

Fig. 6. Joint density of states for the 3C-, 2H-, 4H-, 6H-, and 8H-SiC polytypes (62).

#### **5. Summary and conclusions**

16 Silicon carbide

in obtaining the full-zone band structure of semiconductors, and our procedure for the optical matrix elements requires little additional computation beyond solving the eigenvalue problem

Optical-absorption spectra in semiconductors are normally dominated by transitions from the valence to the conduction bands. Then, it is possible to compute the joint density of states (JDOS) for SiC polytypes that is given by the below formula. The purpose is to see how our

> *d*3*k* (2*π*)

We have computed the JDOS for SiC polytypes (62), hence, the interband transitions in Eq.(7) are all of the valence-conduction type. The interest in the JDOS lies in the fact that the momentum matrix elements are assumed constant over the Brillouin zone. The band summations in Eq.(7) involve all states in the valence band and lowest states in the conduction band. The summations in Eq.(7) are over special points in the Brillouin zone. In our calculations, we took 32, 28, 24, 20 special **k**-points for 2H, 4H, 6H and 8H respectively in

Before discussing the effect of the optical transition matrix elements, we consider the JDOS (see figure 6). In order to get more information on the interband transition, we present our recent calculated joint density of states for SiC polytypes (62). We have determined the transition responsible for the major contributions to these structures. This was done by finding the energy of the desired peak or shoulder on the joint density of states graph and then examining the contribution to joint density of states at that energy from the constituent interband transitions. The fundamental gap is well understood and is attributed to Γ<sup>15</sup> → *X*1transition in 3C-SiC. We examine a large peak associated to Γ<sup>15</sup> → *L*<sup>1</sup> transition which occurs at 3.8 eV. The second major peak in JDOS, comes from the transition Γ<sup>15</sup> → Γ<sup>1</sup> which occurs at 5.2 eV. However, our band structure is satisfactory with respect to these transitions. The principal behavior of the joint density of states is very similar for the various 2H-, 4H-, 6H-, 8H- polytypes considered. The two peaks below the ionic gap exhibit a different behavior with the number *n* of SiC bilayers in the unit cell. Whereas that at higher energy around 7 eV is rather independent of the polytype, the low-energy peak around 8 eV is broadened with rising number *n*. We relate this fact to the folding effect parallel to the *c* axis. It causes an opposite variation of the band curvature along the *LM* and *HK* lines in the hexagonal BZ (25), (73), (78). The intensity of the two most pronounced peaks at 3.5 and 2.1 eV in the region of the upper valence bands monotonically follow the hexagonality of the structures (25), (75), (78),(79). Strong contributions to these peaks also arise from the *LM* line. The most drastic change in the conduction band region occurs near the onset of the density of states. Its steepness over several eV again follows the hexagonality of the polytype. The particular shape of the onset however depends on the number of bilayers and therefore on the folding effect as already has been pointed out by Lee et al. (84). The consequences can be clearly seen in the joint density

<sup>3</sup> *δ* (*Ecv* (*k*) − *h*¯ *ω*) (7)

for the energies.

the Brillouin zone (21).

**4.4.1 Applications of optical matrix elements:**

TB calculations are extended to optical properties.

where Ω is the real-space unit-cell volume.

*<sup>J</sup>* (*E*) <sup>=</sup> <sup>Ω</sup> <sup>∑</sup>*cv*

where *Ecv* = *Ec* (*k*) − *Ev* (*k*) for the JDOS per component (48).

**4.4.2 Joint density of states of 3C-, 2H-, 4H-, 6H-, and 8H-SiC:**

of states. Their low-energy tails increase with decreasing hexagonality.

 *FBZ*

> This chapter reviewed the general aspects of the optical properties as well as the electronic structures of SiC indirect-band-gap semiconductor polytypes. We presented our recent results of the band structures, the total densities of states, and the optical absorption of different polytypes of SiC using the empirical tight-binding approximation. The set of TB parameters is transferable to all hexagonal 2H-, 4H-, 6H-, and 8H-SiC structures. We illustrated how the tight-binding formalism can be used to accurately compute the electronic states in semiconductor hexagonal polytypes. This approximation is known to yield a sufficiently accurate conduction band and to give its minimum position correctly. To this end we have presented our new developed tight-binding model which carefully reproduces ab initio calculations and experimental results of SiC polytypes. It is likely that our TB approach could be applied to SiC polytypes with even larger unit cells than 8*H* using the variation in band gap with hexagonality in cases where experimental band gaps are undetermined. SiC system exhibits rather interesting features which differ greatly from those other semiconductor compounds with respect to optical properties as well as in electronic structure. We refer to this aspect of superperiodicity that is the possibility of modifying the optical properties of the material. Furthermore, the interband optical matrix elements can be tailored. It has been shown that our recent results indicate that our TB calculations are suitable for describing optical properties in more complex polar semi-conductors. In addition, the optical properties

[29] A. O. Konstantinov and H. Bleichner, Appl. Phys. Lett. 71, 3700 (1997).

[35] A. Galeckas, J. Linnros, and P. Pirouz, Appl. Phys. Lett. 81, 883 (2002).

[39] H.P. Iwata, U. Lindefelt, S. berg, P.R. Briddon, Phys. Rev. B, Vol. 65 (2001).

[43] J. Camassel, S. Juillaguet, J. Phys. D: Appl.Phys. 40, 6264 (2007)

[44] A. Chen, P. Srichaikul, Phys. Stat. Sol. B202, 81( 1997). [45] W.Y. Ching et al., Mater. Sci. And Eng. A 422, 147-156 (2006).

[50] G. Litovchenko, Phys. Stat. Sol. B **162**, 447 (1990).

Chaussende, Mat. Sci. For. 645-648 (2010) pp 63-66.

[52] Y. Ward, R. J. Young, R. A. Shatwell, J Mater Sci 42 (2007), p.5135-5141.

[41] H.P. Iwata, U. Lindefelt, S. berg, P.R. Briddon, Physica B, 340-342 p.165, (2003). [42] H.P. Iwata, U. Lindefelt, S. berg, P.R. Briddon, Microelectronics J. 34 371 (2003).

[46] A. Galeckas, J. Linnros, B. Breitholtz, H. Bleichner, J. Appl. Phys. 90 (2) 980 (2001). [47] H. Iwata, U. Lindefelt, S. Oberg, and P. R. Briddon, Phys. Rev. B 65, 033203 (2001); Phys.

[48] T. G. Pedersen, K. Pedersen, and T. B. Kriestensen, Phys. Rev. B **63**, 20110 (2001).

[49] D. Vogel, P. Krüger, and J. Pollmann, Phys. Rev. B **52**, R14316 (1995); B **54**, 5495 (1996); B

[51] G.L. Sun, I.G. Galben-Sandulache, T. Ouisse, J.M. Dedulle, M. Pons, R. Madar, D.

[53] D. Chaussende, F. Baillet, L. Charpentier, E. Pernot, M. Pons, R. Madar, J. Electrochem.

[54] D. Chaussende, F. Mercier, A. Boulle, F. Conchon, M. Soueidan, G. Ferro, A. Mantzari, A. Andreadou, E.K. Polychroniadis, C. Balloud, S. Juillaguet, J. Camassel, M. Pons, J. Cryst.

[55] S. Bai, R. P. Devaty, W. J. Choyke, U. Kaiser, G. Wagner, and M. F. MacMillan, Appl. Phys.

[56] G. Zoulis, J. Sun, M. Beshkova, R. Vasiliauskas, S. Juillaguet, H. Peyre, M. Syvjrvi,

[57] S. Ha, S. Skowronski, J.J. Sumakeris, M.J. Paisley, M.K. Kas Phys.Rev. Lett. Vol. 92, No.

R. Yakimova, J. Camassel, Mat. Sci. For. Vols. 645-648 (2010) pp 179-182.

and P. Skytt, Mater. Sci. Forum 353-356, 727 (2001).

Lett. 81, 2785 (2002).

J. Appl. Phys. 93, 1577 (2003).

Rev. B 68, 174108 (2003).

Rev. B 68, 245309 (2003).

Soc. Vol. 150, (2003), p. G653.

**55**, 12836 (1997).

Growth 310 (2008).

Lett. 83, 3171 (2003).

17, 175504 (2004).

(2004).

(2002).

Lett. 79, 3056 (2001); Mater. Sci. Forum 389-393, 451 (2002).

[30] H. Lendenmann, F. Dahlquist, N. Johansson, R. Sonderholm, P. A. Nilsson, J. P. Bergman,

Opto-Electronic Study of SiC Polytypes: Simulation with Semi-Empirical Tight-Binding Approach 407

[31] R. S. Okojie, M. Zhang, P. Pirouz, S. Tumakha, G. Jessen, and L. J. Brillson, Appl. Phys.

[32] L. J. Brillson, S. Tumakha, G. Jessen, R. S. Okojie, M. Zhang, and P. Pirouz, Appl. Phys.

[33] H. Iwata, U. Lindefelt, S. berg, and P. R. Briddon, Mater. Sci. Forum 389-393, 439 (2002);

[34] P. Pirouz, M. Zhang, J.-L. Demenet, and H. M. Hobgood, J. Appl. Phys. 93, 3279 (2003).

[36] A. T. Blumenau, C. J. Fall, R. Jones, S. berg, T. Frauenheim, and P. R. Briddon, Phys.

[37] A. T. Blumenau, R. Jones, S. berg, and T. Frauenheim, Mater. Sci. Forum 457- 460, 453

[38] P. O. A. Persson, L. Hultman, H. Jacobson, J. P. Bergman, E. Janzn, J. M. Molina-Aldareguia, W. J. Clegg, and T. Tuomi, Appl. Phys. Lett. 80, 4852 (2002).

[40] H.P. Iwata, U. Lindefelt, S. berg, P.R. Briddon, J. Phys. Condens. Matter Vol.14 p. 12733,

suggest that the hexagonal SiC polytypes composed of an indirect-band-gap semiconductors offer a great potential for application to optical devices.

#### **6. References**


[29] A. O. Konstantinov and H. Bleichner, Appl. Phys. Lett. 71, 3700 (1997).

18 Silicon carbide

suggest that the hexagonal SiC polytypes composed of an indirect-band-gap semiconductors

[1] A. P Verma and P. Krishna, Polymorphism and Polytypism in Crystals, (Wiley, New York,

[2] W. J. Choyke, in *The Physics and Chemistry of Carbides, Nitrides and Borides,* NATO ASI Ser.,

[3] K. Karch, G. Wellenhofer, P. Pavone, U. Rö*β*ler, and D. Strauch, *Structural and Electronic Properties of SiC polytypes*, in Proceeding of 22nd International Conference On the Physics

[4] R. F. Davis, Z. Sitar, B. E. Williams, H. S. Kong, H. J. Kim, J. W. Palmour, J. A. Edmond, J.

[6] Diamond, SiC and Nitride Wide Bandgap Semiconductors, edited by C. H. Carter. Jr., et

[7] P. Y. Yu, M. Cardona, Fundamentals of Semiconductors  Physics and Material

[8] Silicon Carbide, III-Nitrides and Related Materials, edited by G. Pensl, H. Morko, B.

[9] D. B. Holt, B. G. Yacobi, Extended Defects in Semiconductors. Electronic Properties,

[10] V. I. Gavrilenko, A. V. Postnikov, N. I. Klyui, and V. G. Litovchenko, Phys. Stat. Sol. B **162**,

[11] M. S. Miao, S. Limpijumnong, and W. R. L. Lambrecht, Appl. Phys. Lett. 79, 4360 (2001).

[16] K. Karch, G. Wellenhofer, P. Pavone, U. Rö*β*ler, and D. Strauch, Structural and Electronic Properties of SiC polytypes, in Proceeding of 22nd International Conference On the

[17] J. Bernholc, A; Antonelli, C. Wang, and R. F. Davis, in *Proceedings of the Fifteenth*

[18] W. R. L. Lambrecht, B. Segall, M. Methfessel, and M. van Schilfgaarde, Phys. Rev. B **44**,

[19] W. R. L. Lambrecht, S. Limpijumnong, S. N. Rashkeev, and B. Segall, Phys. Stat. Sol. (b)

[20] W. van Haeringen, P. A. Bobbert, and W. H. Backes, Phys. Stat. Sol. (b) 202, 63 (1997).

[13] C. Tivarus, J.P. Pelz, M.K. Hudait, S.A. Ringel, Appl. Phys. Lett. 87 182105 (2005). [14] K. J. Russell, Ian Appelbaum, and V. Narayanamurti, Phys. Rev. B 71, 121311(R) (2005).

Ser. E, Appl. Sci. Vol. **185**, R Freer, Ed. (Kluver, Dordrecht, 1990), p. 563.

Ryu, J. T. Glass, and C. H. Carter, Jr. Mater. Sci. Eng B **1**, 77 (1988). [5] N. W. Jepps and T. F. Page, Progr. Cryst. Growth Charact. 7, 259 (1983).

Monemar, and E. Janzn (Trans Tech publications, Switzerland, 1998).

Device Effects and Structures, Cambridge University Press (2007).

[12] Walter R. L. Lambrecht and M. S. Miao, Phys. Rev. B 73, 155312 (2006).

*International Conference on Defects in Semiconductors*, Budapest, 1998.

[15] W. R. L . Lambrecht, and B. Segall, Phys. Rev. B41, 2832 (1990).

[22] G. Pennington and N. Goldsman, Phys. Rev. B 64, 45104 (2001).

[25] P. Kckell, B. Wenzien, and F. Bechstedt, Phys. Rev. B 50, 10761 (1994).

[27] W. J. Choyke, D. R. Hamilton, and Lyle Patrick, Phys. Rev. 133, 1163 (1964). [28] Lyle Patrick, D. R. Hamilton, and W. J. Choyke, Phys. Rev. 143, 526 (1994).

[24] C. Person, and U. Lindefelt, J. Appl. Phys. 82, 5496 (1997).

[26] C. Persson and U. Lindefelt, J. Appl. Phys. 82, 5496(1997).

Physics of Semiconductors, Vancouver, 1994.

[21] B. Adolph et al., Phys. Rev. B **55**, 1422 (1997).

[23] A. Bauer, Phys. Rev. B 57, 2647 (1998).

offer a great potential for application to optical devices.

of Semiconductors, Vancouver, 1994.

properties, Spinger (2005).

447 (1990).

3685 (1991).

202, 5 (1997).

al. (Materials Research Society, Pittsburgh, 1994).

**6. References**

1966).


**17** 

*France* 

**Ceramic Materials** 

**Dielectrics for High Temperature SiC Device** 

The keys to successful high power electronic systems are located as much in the ability to build high temperature power devices and to package them with the appropriate materials, as in the aptitude to reduce and control switching and conduction power losses. Particularly, high temperature low loss operation allows an increase in the power rating of these devices. The recent development of wide band gap semiconductor devices should

Wide band gap semiconductor materials, especially the most mature silicon carbide (SiC), should allow the electronics operation at high junction temperatures (>200°C), high voltages (>10 kV) or in harsh thermal environment, with faster switching and lower power losses active devices than the silicon (Si) counterparts. Such SiC devices impose more severe electrical and thermal stresses to the surrounding insulating materials (polymeric passivation and encapsulation materials and ceramic substrates). Lots of improvements have already been built-up at the die level; however, superior device performance degrees

Among the power device packaging materials for a high temperature operation, typical organic passivation and encapsulation appear nowadays as the most sensitive to the thermal constraints (*Tmax*=250 °C). Moreover, even if ceramic materials present a high isothermal stability (up to 600°C) they are very sensitive to the large passive or active thermal cycling induced by the power devices or by severe environmental constraints during operation. Therefore, research on high temperature dielectric materials tries to identify new polymeric and ceramic materials electrically, thermally and mechanically suited for the packaging of SiC power devices and to determine their effective limits (properties and durability). In this chapter after a section on the high temperature applicative needs and the new thermal and electrical constraints imposed by SiC devices on the surrounding insulating materials, a complete review of the polymers and ceramics insulating materials which are reported to potentially answer to the packaging issues is carried out through a presentation of their different main physical properties and the sensitive aging parameters in link with microstructure. Among the polymeric materials, BPDA/PDA polyimide (PI), fluorinated parylene (PA-F), polyamide-imide (PAI), and silicone (PDMS) will be studied. On the other hand, mainly aluminium nitride (AlN) and silicon nitride (Si3N4) ceramics will be presented.

**1. Introduction** 

allow improving power electronic systems.

could be reached using higher performance insulation materials.

**Insulation: Review of New Polymeric and** 

Sombel Diaham, Marie-Laure Locatelli and Zarel Valdez-Nava

*University of Toulouse – UPS – INPT – LAPLACE – CNRS* 


### **Dielectrics for High Temperature SiC Device Insulation: Review of New Polymeric and Ceramic Materials**

Sombel Diaham, Marie-Laure Locatelli and Zarel Valdez-Nava *University of Toulouse – UPS – INPT – LAPLACE – CNRS France* 

#### **1. Introduction**

20 Silicon carbide

408 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

[58] B. J. Skromme, K. Palle, C. D. Poweleit, L. R. Bryant, W. M. Vetter, M. Dudley, K. Moore, and T. Gehoski, Mater. Sci. Forum 389-393, 455 (2002); B. J. Skromme, K. C. Palle, M. K. Mikhov, H. Meidia, S. Mahajan, X. R. Huang, W. M. Vetter, M. Dudley, K. Moore, S.

[59] J. Q. Liu, H. J. Chung, T. Kuhr, Q. Li, and M. Skowronski, Appl. Phys. Lett. 80, 2111

[60] S. Izumi, H. Tsuchida, T. Tawara, I. Kamata, and K. Izumi, Mater. Sci. Forum 483-485, 323 (2005); S. Izumi, H. Tsuchida, I. Kamata, and T. Tawara, Appl. Phys. Lett. 86, 202108

[61] H. Fujiwara, T. Kimoto, T. Tojo, and H. Matsunami, Appl. Phys. Lett. 87, 051912 (2005).

[63] I.G. Galben-Sandulache, G.L. Sun, J.M. Dedulle, T. Ouisse, R. Madar, M. Pons, D.

[64] L. Latu-Romain, D. Chaussende, C. Balloud, S. Juillaguet, L. Rapenne, E. Pernot, J. Camassel, M. Pons, R. Madar, Mat. Sci. Forum, vols. 527-529 (2006), pp. 99-102. [65] A. A. Lebedev, P. L. Abramov, E. V. Bogdanova, S. P. Lebedev, D. K. Nelson, G. A. Oganesyan, A. S. Tregubova, R Yakimova, Semicond. Sci. Technol. 23, 075004 (2008). [66] A. A. Lebedev, P. L. Abramov, N. V. Agrinskaya, V. I. Kozub, S. P. Lebedev, G. A. Oganesyan, A. S. Tregubova, D. V. Shamshur, and M. O. Skvortsova, J. Appl. Phys. 105,

Smith, and T. Gehoski, Mater. Res. Soc. Symp. Proc. 742, 181 (2003).

[62] A. Laref, S. Laref, Phys. Stat. Sol. B 245, N1, 89-100 (2008).

Chaussende, Mat. Sci. For. 645-648 (2010) pp 55-58.

[67] M. Skowronski and S. Ha, J. Appl. Phys. 99, 011101 (2006).

[72] Y. C. Chang and D. E. Aspnes, Phys. Rev. B **41**, 12002 (1990).

[70] J. C. Slater and G.F. Koster, Phys. Rev **94**, 1498 (1954).

by K. H. Hellwege (Spring-Verlag, Berlin, 1982).

[80] P.J. Colwell and M.V. Klein, Phys. Rev. B 6, 498 (1972). [81] N. T. Son et al., Appl. Phys. Lett **66**, 107 (1995).

[82] R. Kaplan, and R. J. Wagner, Solid State Com. **55**, 67 (1985).

[69] T. Quisse, J. App. Phys. **75**, 2092 (1994).

[68] Y. Ward, R. J. Young, R. A. Shatwell, J Mater Sci 42 (2007), p.5135-5141.

[76] G.Theodorou, G. Tsegas, and E. Kaxiras, J. Appl. Phys **85**, 2179 (1999).

[71] P. Vogl and H. P. Hjalmarson, and J. D. Dow, J. Phys. Chem Solids **44**, 365 (1983).

[73] B. Wenzien, P. Käckell, F. Bechstedt, and G. Cappellini, Phys. Rev. B **52**, 10897 (1995). [74] R. G. Humphreys, D. Bimberg, W. J. Choyke, Solid. State. Com . 39, 163 (1981).

[75] Landolt-Börnstein, *Numerical and Functional Relationships in Science and Technology*, edited

[77] N. Bernstein, H. J. Gotsis, D. A. Papaconstantopoulos, and M. J. Mehl, Phys. Rev. B**71**,

[78] G. Cubiotti, Yu. N. Kucherenko, and V. N. Antonov, J. Phys. Cond. Mat **9**, 165 (1997). [79] J. Kuriplach, M. Sob, G. Brauer, W. Anwand, E.-M. Nicht, P. G. Coleman, and , N. Wagner,

[83] M. Willatzen, M. Cardona, and N. E. Christensen, Phys. Rev. B **51**, 13150 (1995). [84] K, -H. Lee, C. H. Park, B. -H. Cheong, K. J. Chang, Solid State Commun. **92**, 869 (1994).

(2002).

(2005).

023706 (2009).

075203 (2005).

Phys. Rev. B **59**, 1948 (1999).

The keys to successful high power electronic systems are located as much in the ability to build high temperature power devices and to package them with the appropriate materials, as in the aptitude to reduce and control switching and conduction power losses. Particularly, high temperature low loss operation allows an increase in the power rating of these devices. The recent development of wide band gap semiconductor devices should allow improving power electronic systems.

Wide band gap semiconductor materials, especially the most mature silicon carbide (SiC), should allow the electronics operation at high junction temperatures (>200°C), high voltages (>10 kV) or in harsh thermal environment, with faster switching and lower power losses active devices than the silicon (Si) counterparts. Such SiC devices impose more severe electrical and thermal stresses to the surrounding insulating materials (polymeric passivation and encapsulation materials and ceramic substrates). Lots of improvements have already been built-up at the die level; however, superior device performance degrees could be reached using higher performance insulation materials.

Among the power device packaging materials for a high temperature operation, typical organic passivation and encapsulation appear nowadays as the most sensitive to the thermal constraints (*Tmax*=250 °C). Moreover, even if ceramic materials present a high isothermal stability (up to 600°C) they are very sensitive to the large passive or active thermal cycling induced by the power devices or by severe environmental constraints during operation. Therefore, research on high temperature dielectric materials tries to identify new polymeric and ceramic materials electrically, thermally and mechanically suited for the packaging of SiC power devices and to determine their effective limits (properties and durability). In this chapter after a section on the high temperature applicative needs and the new thermal and electrical constraints imposed by SiC devices on the surrounding insulating materials, a complete review of the polymers and ceramics insulating materials which are reported to potentially answer to the packaging issues is carried out through a presentation of their different main physical properties and the sensitive aging parameters in link with microstructure. Among the polymeric materials, BPDA/PDA polyimide (PI), fluorinated parylene (PA-F), polyamide-imide (PAI), and silicone (PDMS) will be studied. On the other hand, mainly aluminium nitride (AlN) and silicon nitride (Si3N4) ceramics will be presented.

Dielectrics for High Temperature

mechanical damages and chemical contamination.

top pads and the ceramic substrate metallization circuit.

**2.3 Specific constraints induced by SiC properties** 

reported for '3D' structures.

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 411

layers (gate dielectric, primary and secondary passivations, intermetallic insulator, …). In particular, the secondary passivation is the top final coating layer elaborated at the wafer level state, before sawing the dies. Contrary to the other existing dielectrics which are inorganic (most often SiO2 and Si3N4, from tens of nm to the order of 1 μm in thickness), the secondary passivation is usually a spin-coated polyimide film (from several μm to few tens of μm thick). Its role is the die protection against premature electrical breakdown,

In a multichip semiconductor power device, the die backside contacts require to be insulated from each other and from their common mechanical substrate. Double-side metallized ceramic substrates are mostly used in this case, instead of polymer based substrates suitable for low power and low voltage ratings. Such metallized ceramic substrates allow the electrical interconnection between the dies soldered on them and with the external circuit. Besides their mechanical and insulating functions, the ceramics ensure the thermal interface with the intermediary dissipating baseplate or the cooling system directly. For the die topside electrical connections, several techniques exist today apart from the conventional wire bonding, which have been developed in order to improve the packaging electrical performance, the cooling efficiency, and the '3D' system integration capability. In particular, 'sandwich' structures involve a second metallized insulating substrate (with polyimide (Liu, 1999) or ceramic as dielectric layer) for the chip top electrodes connecting. Either metal posts or bumps (Mermet-Guyennet, 2008) (preliminary brazed on the chip metal pads), or solder bumps (Dieckerhoff, 2006) (preliminary deposited as well), or direct bonding (Bai, 2004), have been used for the attachment between the chip

Finally, the empty space, existing above the assembly (as in the conventional wire-bonded structures or in the pressure-contacted structures) or present within the gap of the 'sandwich' structures, has to be filled with an insulating material. Its role is to avoid premature electric breakdown and partial discharges, and to protect all the system against humidity and contaminations. This encapsulation function is generally satisfied using silicone gels, which minimize mechanical strains on the assembly. More recently, the use of polymeric underfills, with a thermal expansion coefficient close to the soldered joint ones, is

The superior features of silicon carbide compared to silicon ones are recalled in Table 1, in order to introduce their potential impacts on the die surrounding materials conditions under operation. The high temperature ability of this wide energy band gap semiconductor principally arises from its much lower intrinsic carrier density *ni*, allowing the translating of the thermal runaway onset (induced by prohibitive leakage currents) above at least 700 °C instead of at maximum 200 °C for silicon, depending on the device blocking voltage ratings. Because no other SiC physical intrinsic mechanism is supposed to limit Tj, the upper *Tjmax* temperature limitation for SiC devices is more likely to be imposed by the high temperature performance and stability of all the die surrounding materials and their related interfaces and by the market need besides. Up to now, several high temperature SiC based circuits and devices have been reported, demonstrating short term operations up to 300 °C or 400 °C ambient temperatures (Mounce, 2006; Funaki, 2007). Connected to the thermal aspect, it should be added that high temperatures, and large thermal cycling magnitudes, mean more

#### **2. Needs, insulation problematic and constraints**

The "high temperature" range and the applicative needs are presented in the first part of this section. Silicon carbide arises today as the solution for above 200 °C operations on the semiconductor point of view. The roles and the types of dielectrics in the current semiconductor devices are described then. Insulating passivation, encapsulation and substrate, involving polymeric or ceramic materials, are the main insulating functions to be satisfied by the device packaging. Besides the high temperature requirement, the specific constraints on these materials and their assembly due to the use of SiC are presented at last.

#### **2.1 Needs for high temperature semiconductor devices**

Silicon being the most widely used semiconductor material for active devices active devices, the latter maximal operating junction temperature (*Tj*) limitation fixes the threshold for the "high temperature" denomination. Hence, operations or environments above 200 °C are qualified as "high temperature", 200 °C being the highest maximal operating temperature for available silicon devices. For a long time, the list of high temperature electronics markets has been given as follows: deep well logging (300 °C), geothermal research (400 °C), space exploration (500 °C), for which the common points are the high ambient temperature (*Ta*) of the environment (as indicated into brackets) and their 'niche' specificity. The self -heating of semiconductor devices under operation has been identified as a predictable limitation for the silicon based electronics development for a while as well. Today, the trends for higher integration, or more elevated power level, leading to *Tj* higher than 200 °C, increase the list of the high temperature device markets. In fact, a simple relation between the junction temperature and the power losses (*Pd*) dissipated through the device can be written as follows:

$$T\_j = R\_{th\_{ji}} P\_d + T\_a \tag{1}$$

where *Rthja* is the thermal path resistance between the device dissipating junction and the system ambient. The wider field of the energy conversion either for industry or transportation applications is concerned nowadays. Indeed, embedded integrated power electronics (with reduced or suppressed cooling requirements, meaning very high *Rthja* values) as well as static converters closer to (or inside) hot engine areas (which may correspond simultaneously to elevated *Ta* and *Pd*), are wanted. The aims are mass, volume, and cost savings and higher *Tj* devices are required.

The recent silicon carbide components emergence (Cooper Jr. & Agarwal, 2002), with promising operating temperatures well above 200 °C (Raynaud, 2010) in the future, represents a perspective of offer which will even encourage new demands. As a consequence, the research for high temperature operating dielectrics suitable for the semiconductor die assembly has become essential for the development of the full systems, as insulating materials are among the key points for its performance and reliability.

#### **2.2 Dielectrics for power device insulation**

To realize a discrete (single die) or hybrid (multiple dies) semiconductor device, multiple materials playing different roles are assembled, all of them constituting the device packaging. The semiconductor die itself is not a single material element, as it exhibits different metallized areas (ohmic contact, insulated gate contact, …), and different dielectric

The "high temperature" range and the applicative needs are presented in the first part of this section. Silicon carbide arises today as the solution for above 200 °C operations on the semiconductor point of view. The roles and the types of dielectrics in the current semiconductor devices are described then. Insulating passivation, encapsulation and substrate, involving polymeric or ceramic materials, are the main insulating functions to be satisfied by the device packaging. Besides the high temperature requirement, the specific constraints on these materials and their assembly due to the use of SiC are presented at last.

Silicon being the most widely used semiconductor material for active devices active devices, the latter maximal operating junction temperature (*Tj*) limitation fixes the threshold for the "high temperature" denomination. Hence, operations or environments above 200 °C are qualified as "high temperature", 200 °C being the highest maximal operating temperature for available silicon devices. For a long time, the list of high temperature electronics markets has been given as follows: deep well logging (300 °C), geothermal research (400 °C), space exploration (500 °C), for which the common points are the high ambient temperature (*Ta*) of the environment (as indicated into brackets) and their 'niche' specificity. The self -heating of semiconductor devices under operation has been identified as a predictable limitation for the silicon based electronics development for a while as well. Today, the trends for higher integration, or more elevated power level, leading to *Tj* higher than 200 °C, increase the list of the high temperature device markets. In fact, a simple relation between the junction temperature and the power losses (*Pd*) dissipated through the device can be written as

*ja*

where *Rthja* is the thermal path resistance between the device dissipating junction and the system ambient. The wider field of the energy conversion either for industry or transportation applications is concerned nowadays. Indeed, embedded integrated power electronics (with reduced or suppressed cooling requirements, meaning very high *Rthja* values) as well as static converters closer to (or inside) hot engine areas (which may correspond simultaneously to elevated *Ta* and *Pd*), are wanted. The aims are mass, volume,

The recent silicon carbide components emergence (Cooper Jr. & Agarwal, 2002), with promising operating temperatures well above 200 °C (Raynaud, 2010) in the future, represents a perspective of offer which will even encourage new demands. As a consequence, the research for high temperature operating dielectrics suitable for the semiconductor die assembly has become essential for the development of the full systems,

To realize a discrete (single die) or hybrid (multiple dies) semiconductor device, multiple materials playing different roles are assembled, all of them constituting the device packaging. The semiconductor die itself is not a single material element, as it exhibits different metallized areas (ohmic contact, insulated gate contact, …), and different dielectric

as insulating materials are among the key points for its performance and reliability.

*T RP T j th d a* = + (1)

**2. Needs, insulation problematic and constraints** 

**2.1 Needs for high temperature semiconductor devices** 

and cost savings and higher *Tj* devices are required.

**2.2 Dielectrics for power device insulation** 

follows:

layers (gate dielectric, primary and secondary passivations, intermetallic insulator, …). In particular, the secondary passivation is the top final coating layer elaborated at the wafer level state, before sawing the dies. Contrary to the other existing dielectrics which are inorganic (most often SiO2 and Si3N4, from tens of nm to the order of 1 μm in thickness), the secondary passivation is usually a spin-coated polyimide film (from several μm to few tens of μm thick). Its role is the die protection against premature electrical breakdown, mechanical damages and chemical contamination.

In a multichip semiconductor power device, the die backside contacts require to be insulated from each other and from their common mechanical substrate. Double-side metallized ceramic substrates are mostly used in this case, instead of polymer based substrates suitable for low power and low voltage ratings. Such metallized ceramic substrates allow the electrical interconnection between the dies soldered on them and with the external circuit. Besides their mechanical and insulating functions, the ceramics ensure the thermal interface with the intermediary dissipating baseplate or the cooling system directly. For the die topside electrical connections, several techniques exist today apart from the conventional wire bonding, which have been developed in order to improve the packaging electrical performance, the cooling efficiency, and the '3D' system integration capability. In particular, 'sandwich' structures involve a second metallized insulating substrate (with polyimide (Liu, 1999) or ceramic as dielectric layer) for the chip top electrodes connecting. Either metal posts or bumps (Mermet-Guyennet, 2008) (preliminary brazed on the chip metal pads), or solder bumps (Dieckerhoff, 2006) (preliminary deposited as well), or direct bonding (Bai, 2004), have been used for the attachment between the chip top pads and the ceramic substrate metallization circuit.

Finally, the empty space, existing above the assembly (as in the conventional wire-bonded structures or in the pressure-contacted structures) or present within the gap of the 'sandwich' structures, has to be filled with an insulating material. Its role is to avoid premature electric breakdown and partial discharges, and to protect all the system against humidity and contaminations. This encapsulation function is generally satisfied using silicone gels, which minimize mechanical strains on the assembly. More recently, the use of polymeric underfills, with a thermal expansion coefficient close to the soldered joint ones, is reported for '3D' structures.

#### **2.3 Specific constraints induced by SiC properties**

The superior features of silicon carbide compared to silicon ones are recalled in Table 1, in order to introduce their potential impacts on the die surrounding materials conditions under operation. The high temperature ability of this wide energy band gap semiconductor principally arises from its much lower intrinsic carrier density *ni*, allowing the translating of the thermal runaway onset (induced by prohibitive leakage currents) above at least 700 °C instead of at maximum 200 °C for silicon, depending on the device blocking voltage ratings. Because no other SiC physical intrinsic mechanism is supposed to limit Tj, the upper *Tjmax* temperature limitation for SiC devices is more likely to be imposed by the high temperature performance and stability of all the die surrounding materials and their related interfaces and by the market need besides. Up to now, several high temperature SiC based circuits and devices have been reported, demonstrating short term operations up to 300 °C or 400 °C ambient temperatures (Mounce, 2006; Funaki, 2007). Connected to the thermal aspect, it should be added that high temperatures, and large thermal cycling magnitudes, mean more

Dielectrics for High Temperature

**3. Material choice criteria and main issues** 

for the device assembly insulating substrate are also discussed.

**3.1 Thermal stability and degradation of organic materials** 

0

20

40

Weight (%)

2Heating rate: 10 °C/min

60

*Td* 5%

80

100

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 413

As presented in the previous paragraph, the insulating passivation, encapsulation and substrate are the three main insulating functions to be satisfied by the device packaging, involving organic and ceramic materials. Besides their electrical role, the involved materials may play mechanical, and/or thermal, and/or chemical roles. The aim of this paragraph is to review the main limiting properties or the main influent constraints to be taken into account at high temperature, according to the dielectric nature or its role in the device. Used dielectrics or reported candidates, as materials for high temperature device packaging, are presented at the same time through the proposed result examples. In particular, biphenyltetracarboxilic dianhydride/*p*-phenylene diamine (BPDA/PDA) polyimide (PI), and flurorinated parylene (PA-F) are considered as interesting high temperature insulating surface coating. Limits of polydimethylsiloxane (PDMS) materials, currently used as volumic insulation for encapsulation purpose, are presented as well. The different ceramic/metal couples available

Thermal stability is a fundamental parameter for a long-term reliable high temperature operation of polymeric and other organic materials. It appears as the first stage in the material evaluation because it can ensure a stability of the other physical properties. Conventionally, the thermal stability is determined using thermal gravimetric analysis (TGA) either in oxidant or inert atmosphere. This consists in probing the mass loss of a material versus temperature under a controlled heating slope (dynamical TGA, DTGA) or time at a set temperature (isothermal TGA, ITGA). The degradation temperature (*Td*) is often defined as the 5%-mass loss onset in DTGA plots. Figure 1 shows a comparison of DTGA measurements of thermo-stable organic materials. According to the material structural chemistry, *Td* is more or less elevated. Thus, the thermal stability determined by the means of DTGA in nitrogen reports *Td* values of 606 °C, 455 °C, 537 °C, and 456 °C for BPDA/PDA

PI, PAI, PA-F and PDMS/silica materials, respectively (Diaham, 2009, 2011a, 2011b).

 **PI (BPDA/PDA) PAI PA-F PDMA/silica**

Fig. 1. Comparison of dynamical TGA of thermo-stable organic materials in nitrogen 2

300 400 500 600 700

0 200 400 600 800 1000

Temperature (°C)

severe thermo-mechanical stresses and fatigue on the device assembly parts, due to their different thermal expansion coefficients. Also, a higher *Ta* may lead to higher thermal conductivity requirement (for reduced Rthja), in order to preserve a sufficient power density level (and its related level of power losses dissipation) for the wanted system operation for a given *Tjmax* (according to relation (1)).


Table 1. Main 4H-SiC and Si semiconductor physical properties. 1

Beyond the high temperature operation ability and related constraints presented above, the high critical electric field EC is the other SiC specificity inducing major novel stresses to the die surrounding materials, in comparison to the silicon case. Here the insulating dielectrics are more specifically addressed with regard to this aspect. Because the one-order higher *EC* property allows faster and higher voltage devices with low conduction losses than the silicon one, SiC components are designed to operate with internal maximal electric fields at blocking state as close as possible to the SiC critical EC value. As a consequence, even for optimally designed junction termination structures for a given blocking voltage rating, electric field peak values as high as around 3 MV/cm exist near the semiconductor surface, at the device periphery (Locatelli, 2003). Moreover, smaller dimensions of the device are resulting from the higher *EC* ability of SiC, including shorter periphery protection extension. Higher average result values of the electrical field as well. The semiconductor surface passivation materials are concerned at first level by such electrical stress enhancement. Besides, the higher the blocking voltage rating, the more the encapsulating material (above the passivation coating) will be impacted too. Today, the record in terms of breakdown voltage for a single SiC component is 19 kV for a SF6 gas encapsulated diode demonstrator (Sugawara, 2001), and more than 50 kV might be achievable with SiC while 10 kV represent the Si device practical limit.

Last but not least, higher on-state current density, higher switching speed and smaller SiC dies (thanks to a combination of good *EC*, electron mobility μ*<sup>n</sup>*, and electron saturation velocity *vsat* properties), also represent new challenges, especially in terms of connecting materials and highly compact packaging structures. Specific constraints on the insulation elaboration techniques may result so.

 1Among the different SiC polytypes, 4H-SiC is the one used for the commercial power devices production

#### **3. Material choice criteria and main issues**

412 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

severe thermo-mechanical stresses and fatigue on the device assembly parts, due to their different thermal expansion coefficients. Also, a higher *Ta* may lead to higher thermal conductivity requirement (for reduced Rthja), in order to preserve a sufficient power density level (and its related level of power losses dissipation) for the wanted system operation for a

*Eg* @ 300 K (eV) 3.26 1.12

*EC* @ 300 K, for *Nd* = 1015 cm-3 (V/cm) 2.5x106 3x105

*<sup>n</sup>* @ 300 K, for *Nd* = 1015 cm-3 (cm2/V/s) 850 1,400 *vsat* @ 300 K (cm/s) 2.2x107 107

*th* @ 300 K (W/cm/K) 3.8 1

Beyond the high temperature operation ability and related constraints presented above, the high critical electric field EC is the other SiC specificity inducing major novel stresses to the die surrounding materials, in comparison to the silicon case. Here the insulating dielectrics are more specifically addressed with regard to this aspect. Because the one-order higher *EC* property allows faster and higher voltage devices with low conduction losses than the silicon one, SiC components are designed to operate with internal maximal electric fields at blocking state as close as possible to the SiC critical EC value. As a consequence, even for optimally designed junction termination structures for a given blocking voltage rating, electric field peak values as high as around 3 MV/cm exist near the semiconductor surface, at the device periphery (Locatelli, 2003). Moreover, smaller dimensions of the device are resulting from the higher *EC* ability of SiC, including shorter periphery protection extension. Higher average result values of the electrical field as well. The semiconductor surface passivation materials are concerned at first level by such electrical stress enhancement. Besides, the higher the blocking voltage rating, the more the encapsulating material (above the passivation coating) will be impacted too. Today, the record in terms of breakdown voltage for a single SiC component is 19 kV for a SF6 gas encapsulated diode demonstrator (Sugawara, 2001), and more than 50 kV might be achievable with SiC while 10 kV represent

Last but not least, higher on-state current density, higher switching speed and smaller SiC

velocity *vsat* properties), also represent new challenges, especially in terms of connecting materials and highly compact packaging structures. Specific constraints on the insulation

1Among the different SiC polytypes, 4H-SiC is the one used for the commercial power devices

μ

*<sup>n</sup>*, and electron saturation

Table 1. Main 4H-SiC and Si semiconductor physical properties. 1

dies (thanks to a combination of good *EC*, electron mobility

4H-SiC Si

1.2x1010 1014

6x10-8 2x103

given *Tjmax* (according to relation (1)).

*ni* @ 300 K (cm-3) *ni* @ 473 K (cm-3)

the Si device practical limit.

elaboration techniques may result so.

μ

λ

production

As presented in the previous paragraph, the insulating passivation, encapsulation and substrate are the three main insulating functions to be satisfied by the device packaging, involving organic and ceramic materials. Besides their electrical role, the involved materials may play mechanical, and/or thermal, and/or chemical roles. The aim of this paragraph is to review the main limiting properties or the main influent constraints to be taken into account at high temperature, according to the dielectric nature or its role in the device. Used dielectrics or reported candidates, as materials for high temperature device packaging, are presented at the same time through the proposed result examples. In particular, biphenyltetracarboxilic dianhydride/*p*-phenylene diamine (BPDA/PDA) polyimide (PI), and flurorinated parylene (PA-F) are considered as interesting high temperature insulating surface coating. Limits of polydimethylsiloxane (PDMS) materials, currently used as volumic insulation for encapsulation purpose, are presented as well. The different ceramic/metal couples available for the device assembly insulating substrate are also discussed.

#### **3.1 Thermal stability and degradation of organic materials**

Thermal stability is a fundamental parameter for a long-term reliable high temperature operation of polymeric and other organic materials. It appears as the first stage in the material evaluation because it can ensure a stability of the other physical properties. Conventionally, the thermal stability is determined using thermal gravimetric analysis (TGA) either in oxidant or inert atmosphere. This consists in probing the mass loss of a material versus temperature under a controlled heating slope (dynamical TGA, DTGA) or time at a set temperature (isothermal TGA, ITGA). The degradation temperature (*Td*) is often defined as the 5%-mass loss onset in DTGA plots. Figure 1 shows a comparison of DTGA measurements of thermo-stable organic materials. According to the material structural chemistry, *Td* is more or less elevated. Thus, the thermal stability determined by the means of DTGA in nitrogen reports *Td* values of 606 °C, 455 °C, 537 °C, and 456 °C for BPDA/PDA PI, PAI, PA-F and PDMS/silica materials, respectively (Diaham, 2009, 2011a, 2011b).

Fig. 1. Comparison of dynamical TGA of thermo-stable organic materials in nitrogen 2

<sup>2</sup>Heating rate: 10 °C/min

Dielectrics for High Temperature

**3.2 Thermal properties of ceramic materials** 

as aluminum nitride, for instance.

temperature uses.

technologies.

Dielectric breakdown

Thermal conductivity

Fracture toughness (MPa

substrates for SiC device insulation

Available substrate technologies for thick film metallization (metal)

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 415

All these illustrations lead to highlight that the thermal stability is a property difficult to quantify with accuracy. It depends strongly on various structural parameters (materials, …) and experimental conditions (type of measurements, atmosphere, temperature, …). However, it appears as an essential information for a first selection of materials for high

In a classical approach for power electronics, the substrates assure the mechanical link and the electrical insulation between the semiconductor die and the rest of the system. For high temperature applications, ceramic materials are a natural choice due to their thermal stability, and high thermal conductivity compared to polymer materials. Ceramic materials on their own present a high isothermal stability (up to 600°C) and seem to be self-sufficient in most cases to insulate electrically appropriately the semiconductor from the environment. However, the presence of an attached metal can be at the origin of several mechanical problems which will be treated in a later section. Furthermore, when high power densities are attained, heat extraction could need to be assisted by high-thermal conductivity ceramics

The choice of the appropriate insulating ceramic is related to a compromise of electrical properties, thermal characteristics and compatible technologies available to assemble the components. Table 2 presents the characteristics of some of the insulating ceramics that are commercially available to this date. Beryllium oxide (BeO) use is being more and more limited due to toxicity concerns, and is being replaced, when possible, by other ceramic

 Si3N4 AlN Al2O3 Dielectric constant 8-9 8-9 9-10 Loss factor 2x10-4 3x10-4 3x10-4 - 1x10-3 Resistivity (Ω m) > 1012 > 1012 > 1012

strength (kV/mm) 10-25 14-35 10-35

(W/m K) 40-90 120-180 20-30 Bending strength (MPa) 600-900 250-350 300-380 Young Module (GPa) 200-300 300-320 300-370

m1/2) 4-7 2-3 3-5 CTE (mm/m K) 2.7-4.5 4.2-7 7-9

Table 2. Main thermal, mechanical and electrical characteristics of candidate ceramic

Despite the availability of ceramic materials of very high thermal conductivity, as BeO or AlN, one must take into account the evolution of this property with temperature. Even in high thermal-conductivity ceramics, the phonon conduction path is disturbed as

AMB (Cu) DBC (Cu), AMB

(Al) DBC (Cu)

For polymers, the thermal stability is often related to the presence of benzene rings in the monomer structure. In the case of PI materials, it has been shown that the increase in the number of benzene rings contributes to an increase in the degradation temperature (Sroog, 1965). However, the degradation temperature can be also affected by the presence of low thermo-stable bonds in the macromolecular structure. As an example, even if BPDA/PDA and PMDA/ODA (Kapton-type) PI own the same number of benzene rings (i.e. three in the elementary monomer backbone), the absence of the C—O—C ether group in the case of BPDA/PDA PI allows increasing *Td* of 60 °C in nitrogen and 110 °C in air in comparison to *Td* of PMDA/ODA PI (see Figure 2). Indeed, this is due to the lower thermal stability of the ether bonds inducing earlier degradations than the rest of the structure (Sroog, 1965; Tsukiji, 1990).

Fig. 2. Dynamical TGA of different structural PI films 2

Although the degradation temperature obtained by DTGA appears as an important parameter for the evaluation of the thermal stability, it is not sufficient to valid that a polymer can endure high temperature during a very long time. In addition, some polymers can exhibit lower *Td* values while they display a more stable behavior during time. Therefore, short-term ITGA measurements are recommended in order to identify premature degradation processes. Figure 3 presents ITGA measurements of both BPDA/PDA PI and PA-F films in air. Whereas PA-F films own a lower dynamical *Td* value than PI films, they show a better stability under isothermal conditions. Hence, after 5,000 minutes at 350 °C in air atmosphere the weight loss of PA-F is only of 0.5 % compared to 2.4 % for BPDA/PDA PI.

Fig. 3. Comparison of the isothermal TGA of BPDA/PDA PI and PA-F films in air

All these illustrations lead to highlight that the thermal stability is a property difficult to quantify with accuracy. It depends strongly on various structural parameters (materials, …) and experimental conditions (type of measurements, atmosphere, temperature, …). However, it appears as an essential information for a first selection of materials for high temperature uses.

#### **3.2 Thermal properties of ceramic materials**

414 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

For polymers, the thermal stability is often related to the presence of benzene rings in the monomer structure. In the case of PI materials, it has been shown that the increase in the number of benzene rings contributes to an increase in the degradation temperature (Sroog, 1965). However, the degradation temperature can be also affected by the presence of low thermo-stable bonds in the macromolecular structure. As an example, even if BPDA/PDA and PMDA/ODA (Kapton-type) PI own the same number of benzene rings (i.e. three in the elementary monomer backbone), the absence of the C—O—C ether group in the case of BPDA/PDA PI allows increasing *Td* of 60 °C in nitrogen and 110 °C in air in comparison to *Td* of PMDA/ODA PI (see Figure 2). Indeed, this is due to the lower thermal stability of the ether bonds inducing earlier degradations than the rest of the structure (Sroog, 1965; Tsukiji, 1990).

N

\*

N O

O \* N

Although the degradation temperature obtained by DTGA appears as an important parameter for the evaluation of the thermal stability, it is not sufficient to valid that a polymer can endure high temperature during a very long time. In addition, some polymers can exhibit lower *Td* values while they display a more stable behavior during time. Therefore, short-term ITGA measurements are recommended in order to identify premature degradation processes. Figure 3 presents ITGA measurements of both BPDA/PDA PI and PA-F films in air. Whereas PA-F films own a lower dynamical *Td* value than PI films, they show a better stability under isothermal conditions. Hence, after 5,000 minutes at 350 °C in air atmosphere the weight loss

10-1 100 101 102 103 104

Time (minutes)

 350°C 400°C

Atmosphere: Air

Fig. 3. Comparison of the isothermal TGA of BPDA/PDA PI and PA-F films in air

**PI**

**PA-F**

**PA-F**

**PI**

O

O

N \*

O \*

n

n

BPDA/PDA

PMDA/ODA

O

O

O

O

0 200 400 600 800 1000

Temperature (°C)

Fig. 2. Dynamical TGA of different structural PI films 2

of PA-F is only of 0.5 % compared to 2.4 % for BPDA/PDA PI.

90

92

94

96

Isothermal weight (%)

98

100

PMDA/ODA in Air (4)

300 400 500 600

(4) (3) (2) (1)

BPDA/PDA in Air (2)

BPDA/PDA in N2 (1)

PMDA/ODA in N2 (3)

0

20

40

Weight (%)

60

80

100

In a classical approach for power electronics, the substrates assure the mechanical link and the electrical insulation between the semiconductor die and the rest of the system. For high temperature applications, ceramic materials are a natural choice due to their thermal stability, and high thermal conductivity compared to polymer materials. Ceramic materials on their own present a high isothermal stability (up to 600°C) and seem to be self-sufficient in most cases to insulate electrically appropriately the semiconductor from the environment. However, the presence of an attached metal can be at the origin of several mechanical problems which will be treated in a later section. Furthermore, when high power densities are attained, heat extraction could need to be assisted by high-thermal conductivity ceramics as aluminum nitride, for instance.

The choice of the appropriate insulating ceramic is related to a compromise of electrical properties, thermal characteristics and compatible technologies available to assemble the components. Table 2 presents the characteristics of some of the insulating ceramics that are commercially available to this date. Beryllium oxide (BeO) use is being more and more limited due to toxicity concerns, and is being replaced, when possible, by other ceramic technologies.


Table 2. Main thermal, mechanical and electrical characteristics of candidate ceramic substrates for SiC device insulation

Despite the availability of ceramic materials of very high thermal conductivity, as BeO or AlN, one must take into account the evolution of this property with temperature. Even in high thermal-conductivity ceramics, the phonon conduction path is disturbed as

Dielectrics for High Temperature

can be as low as 10-5 for very performing materials.

temperatures, it is observed that the magnitude of both

100 150 200 250 300

Temperature (°C)

100 150 200 250 300

100 Hz

10 Hz

1 H

z

0

.1 Hz

Temperature (°C)

(a, c) (from Diaham, 2010a) and alumina ceramic (b, d)

3

10-3

10-2

10-1

*tan* 

δ

100

101

102 (c)

6

ε*'*

(a) 0.1 Hz 1 Hz 10 Hz 100 Hz 1 kHz 10 kHz 100 kHz

the transition where the materials start to become semi-insulating (i.e.

conductivity (i.e. property completely controlled by the motion of charges).

1 Hz

0.1 Hz

> 100 kHz 100 Hz

9

10-3

10-2

10-1

*tan* 

δ

100

101

102 (d)

12

15

ε*'*

18

21 (b)

10 Hz

10 kHz

> 100 kHz

Fig. 5. Dielectric permittivity and loss factor versus temperature for BPDA/PDA PI films

1 kHz

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 417

materials, an acceptable upper limit for the loss factor can be situated around 10-2 while it

Figure 5 shows two examples of the high temperature dependence of the dielectric properties of good insulating dielectrics: (a, c) BPDA/PDA PI films and (b, d) Al2O3 ceramic. Typically, at low temperature (<100 °C), most of the thermo-stable dielectrics present a nonvariant relative permittivity and a loss factor below 10-2. On the contrary, for higher

all the more important as temperature is high and/or frequency is low. Such magnitudes cannot find explanations in simple dipolar polarization processes (Adamec, 1974). These huge values are mainly associated to interfacial polarization processes (i.e. either due to Maxwell-Wagner-Sillars (MWS) relaxation-type in heterogeneous specimen or electrode polarization) (Kremer & Schönhals, 2003). MWS relaxation and electrode polarization are involved by the drift of mobile charges across the materials towards bulk interfaces (different phases, impurities, …) or electrodes, respectively. Their occurrence corresponds to

Consequently, it appears as more judicious to investigate them in terms of electrical

ε*'* and *tan*

δ

ε*'*>>ε

100 150 200 250 300

Temperature (°C)

100 150 200 250 300

1 kHz

100 Hz

10 Hz

1 Hz

0.1 Hz

Temperature (°C)

exhibits a strong increase

*∞* and *tan*

δ>10-1).

1 kHz

100 Hz

10 Hz

1 Hz

0.1 Hz

temperature increases, so one must expect a decay of this property as temperature increases. Figure 4 shows the temperature dependence of the thermal conductivity of AlN and Al2O3 ceramic substrates (Chasserio, 2009). In the case of AlN, this value can decrease abruptly above 100 °C, attaining just over 100 W m-1 K-1 at 300 °C.

Fig. 4. Temperature dependence of the thermal conductivity for AlN and Al2O3 ceramic substrates (values taken from Chasserio, 2009)

#### **3.3 Electrical properties**

As the main function of dielectric materials in the environment of the power devices is to separate two different electrical potentials from one to the other, their electrical insulating properties are fundamental and must be accurately known versus temperature. Particularly, in the case of the insulation of high temperature SiC power devices and modules (above 200 °C), the electrical properties of the candidates need to be investigated in the same range.

#### **3.3.1 Dielectric permittivity and loss**

The low field dielectric properties are usually defined under the complex dielectric permittivity formalism (ε*\**), which is made up of the dielectric constant (real part) and the dielectric loss (imaginary part) (see eq. (2)). The ratio between the imaginary part and the real part corresponds to the dielectric loss factor (*tan*δ) (see eq. (3)):

$$
\varepsilon''(\alpha) = \varepsilon'(\alpha) - j\varepsilon''(\alpha) \tag{2}
$$

$$\tan \delta(\alpha) = \frac{\varepsilon \text{"(}\alpha\text{)}}{\varepsilon \text{"(}\alpha\text{)}} \tag{3}$$

where ε*'* and ε*''* represent respectively the real and imaginary parts of the complex dielectric permittivity, ωis the angular frequency and *j* = −1 .

The dielectric permittivity and loss result from polarization processes in the material bulk such as the orientation of dipole entities. This phenomenon is strongly dependent on the frequency of study. Moreover, the dipolar mobility being thermally activated, the polarization processes are also strongly temperature-dependent. For good insulating

temperature increases, so one must expect a decay of this property as temperature increases. Figure 4 shows the temperature dependence of the thermal conductivity of AlN and Al2O3 ceramic substrates (Chasserio, 2009). In the case of AlN, this value can decrease abruptly

> AlN Al2 O3

0 50 100 150 200 250 300 350 400

Fig. 4. Temperature dependence of the thermal conductivity for AlN and Al2O3 ceramic

As the main function of dielectric materials in the environment of the power devices is to separate two different electrical potentials from one to the other, their electrical insulating properties are fundamental and must be accurately known versus temperature. Particularly, in the case of the insulation of high temperature SiC power devices and modules (above 200 °C), the electrical properties of the candidates need to be investigated in the same range.

The low field dielectric properties are usually defined under the complex dielectric

dielectric loss (imaginary part) (see eq. (2)). The ratio between the imaginary part and the

''( ) tan ( ) '( ) ε ω

The dielectric permittivity and loss result from polarization processes in the material bulk such as the orientation of dipole entities. This phenomenon is strongly dependent on the frequency of study. Moreover, the dipolar mobility being thermally activated, the polarization processes are also strongly temperature-dependent. For good insulating

 εω

δ ω

\* εω

is the angular frequency and *j* = −1 .

δ

*''* represent respectively the real and imaginary parts of the complex dielectric

 ε ω

ε ω

*\**), which is made up of the dielectric constant (real part) and the

) (see eq. (3)):

( ) '( ) ''( ) = − *j* (2)

<sup>=</sup> (3)

Temperature (°C)

above 100 °C, attaining just over 100 W m-1 K-1 at 300 °C.

Thermal conductivity (W m-1 K-1

substrates (values taken from Chasserio, 2009)

**3.3.1 Dielectric permittivity and loss** 

ε

real part corresponds to the dielectric loss factor (*tan*

**3.3 Electrical properties** 

permittivity formalism (

where ε*'* and ε

permittivity,

ω

)

materials, an acceptable upper limit for the loss factor can be situated around 10-2 while it can be as low as 10-5 for very performing materials.

Figure 5 shows two examples of the high temperature dependence of the dielectric properties of good insulating dielectrics: (a, c) BPDA/PDA PI films and (b, d) Al2O3 ceramic. Typically, at low temperature (<100 °C), most of the thermo-stable dielectrics present a nonvariant relative permittivity and a loss factor below 10-2. On the contrary, for higher temperatures, it is observed that the magnitude of both ε*'* and *tan*δ exhibits a strong increase all the more important as temperature is high and/or frequency is low. Such magnitudes cannot find explanations in simple dipolar polarization processes (Adamec, 1974). These huge values are mainly associated to interfacial polarization processes (i.e. either due to Maxwell-Wagner-Sillars (MWS) relaxation-type in heterogeneous specimen or electrode polarization) (Kremer & Schönhals, 2003). MWS relaxation and electrode polarization are involved by the drift of mobile charges across the materials towards bulk interfaces (different phases, impurities, …) or electrodes, respectively. Their occurrence corresponds to the transition where the materials start to become semi-insulating (i.e. ε*'*>>ε*∞* and *tan*δ>10-1). Consequently, it appears as more judicious to investigate them in terms of electrical conductivity (i.e. property completely controlled by the motion of charges).

Fig. 5. Dielectric permittivity and loss factor versus temperature for BPDA/PDA PI films (a, c) (from Diaham, 2010a) and alumina ceramic (b, d)

Dielectrics for High Temperature

600 800 1000 1200 1400 1600 1800 2000

PAA precursor residues

Wave number (cm-1)

10 12 14 16 18 20 22 24 26 28

19.14°

1,5 1,6 1,7 1,8 1,9 2,0 2,1 10-17

1000 / T (K-1)

20.36°

Angle 2θ[°]

Intensity [counts/sec]

10-16 10-15 10-14 10-13 10-12 10-11 10-10

σ *DC* (

films

Ω-1 cm-1

)

1,5 1,6 1,7 1,8 1,9 2,0 2,1 10-14

(c) As-deposited

1000 / T (K-1)

Temperature (°C)

350 300 250 200

400

 1.5 μm 4 μm 8 μm

Intensity (a.u.)

400 (a)

10-13 10-12 10-11 10-10 10-9 10-8 10-7

σ *DC* (

Ω-1 cm-1

)

the σ

2011b).

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 419

controlling the charge motion across amorphous dielectrics. For high temperature operation, higher the *Tg*, wider is the temperature range of use. Figure 6c and 6d present respectively

> 1.5 μm 8.8 μm 20 μm

Annealed at 400 °C

350 300 250 200

Temperature (°C)

*DC* temperature dependence of PA-F before and after a 400 °C annealing and as a function of thickness. It is shown that both annealing and material thickness improve the electrical properties (DC conductivity decreases). Inlet plots show that the PA-F crystallinity and the crystallite size are increased either with a thermal treatment or increasing thickness. Consequently, when the volume of the crystalline phase is increased the motion of charges within the material becomes more difficult, thus reducing the DC conductivity (Diaham,

0 100 200 300 400 -0,75

<sup>1</sup> <sup>10</sup> <sup>100</sup> 4,6

Thickness ( μm)

1,5 1,6 1,7 1,8 1,9 2,0 2,1 10-17

1000 / T (K-1)

4,7 4,8 4,9 5,0 5,1

Crystallite size *D* (nm)

(d)

400

10-16 10-15 10-14 10-13 10-12 10-11 10-10

σ *DC* (

Fig. 6. Main parameters affecting the temperature dependence of the DC conductivity of various polymers: (a) thickness of BPDA/PDA PI films, (b) glass transition temperature in two different PAI films, (c) crystallization temperature for PA-F films, (d) thickness of PA-F

Ω-1 cm-1

)

Temperature (°C)

**PAI 2**

**PAI 1**

*Tg2*

Tg2

(b) **PAI 1**

Temperature (°C)

400 350 300 250

**PAI 2**

 1.4 μm 4.8 μm 49.4 μm

Tg1

1,5 1,6 1,7 1,8 1,9

1000 / T (K-1)

Temperature (°C)

350 300 250 200

*Tg1*


Heat flow (W/g)

10-13 10-12 10-11 10-10 10-9 10-8 10-7

σ *DC* (

Ω-1 cm-1

)

#### **3.3.2 Electrical conductivity**

Insulating materials are defined by a volume conductivity largely below 10-12 Ω-1 cm-1. The peculiar range of semi-insulating materials corresponds to the conductivity range between that of insulating ones and semiconductors (i.e. from 10-12 to 10-8 Ω-1 cm-1). When the conduction of mobile charges dominates the dielectric loss, compared to the dipolar processes, it is preferable to represent the loss in the formalism of the alternating conductivity (σ*AC*) as a function of frequency and temperature using eq. (4) (Kremer & Schönhals, 2003; Jonscher, 1983):

$$
\sigma\_{\rm AC}(f, T) = 2\pi f \varepsilon\_0 \varepsilon^{\prime\prime}(f, T) = \sigma\_{\rm DC}(T) + A(T)f^{\prime} \tag{4}
$$

where ε*0* is the vacuum permittivity, σ*DC* is the static volume conductivity, *A* is a temperature-dependent parameter and *s* is the exponent of the power law (0<*s*≤1).

In a large frequency range of study, the AC conductivity is made up of a high frequency linear contribution and an independent-frequency region at low frequency characterized by a static conductivity (σ*DC*) plateau. The DC conductivity is a temperature-dependent property following usually the Arrhenius-like behavior, described by eq. (5). Materials presenting a thermal transition in the investigated temperature range (e.g. glass transition region) follow the non-linear Vogel-Fulcher-Tamman (VFT) behavior given by eq. (6):

$$
\sigma\_{DC}(T) = \sigma\_{\infty} \exp\left[-\frac{E\_a}{k\_B T}\right] \tag{5}
$$

$$
\sigma\_{DC}(T) = \sigma\_{\infty} \exp\left[-\frac{DT\_0}{T - T\_0}\right] \tag{6}
$$

where σ*<sup>∞</sup>* is the conductivity at an infinite temperature, *Ea* is the activation energy, *kB* is the Boltzmann's constant, *D* is the material fragility and *T0* is the Vogel temperature.

DC conductivity is related to the structure and microstructure of the dielectric materials. Moreover, for a given material the dielectric properties are also strongly related to the way used to synthesize and process it. Hence, whereas it is difficult to predict a priori what will be the final DC conductivity from a theoretical point of view, it appears as impossible to estimate before what will be the impact of the material processing on this property. Consequently, it is fundamental to investigate, analyse and understand the origins of such variations of the DC conductivity in close relations with the material physico-chemical properties. Figure 6 presents the main parameters affecting the temperature dependence of the dc conductivity for various thermo-stable polymers. Figure 6a shows the variation of σ*DC* of 400 °C-cured BPDA/PDA PI films for different thicknesses from 1.5 µm to 20 µm. It is observable an increase in σ*DC* with increasing thickness. The inlet plot, showing the infrared spectra of the PI films, allows relating this evolution to the remaining presence after the material processing of PI precursor (polyamic acid, PAA) residues (Diaham, 2011a). These impurities are a source of ionic species increasing the electrical conduction. Figure 6b shows the temperature dependence of σ*DC* for two PAI films with different glass transition temperatures (*Tg*). The increase in *Tg* for PAI 2 (i.e. 335 °C against 280 °C for PAI 1 obtained by DSC in the inlet plot) allows shifting the onset of the σ*DC* increase towards higher temperature (Diaham, 2009). The glass transition is therefore an important parameter

Insulating materials are defined by a volume conductivity largely below 10-12 Ω-1 cm-1. The peculiar range of semi-insulating materials corresponds to the conductivity range between that of insulating ones and semiconductors (i.e. from 10-12 to 10-8 Ω-1 cm-1). When the conduction of mobile charges dominates the dielectric loss, compared to the dipolar processes, it is preferable to represent the loss in the formalism of the alternating

<sup>0</sup> ( , ) ''( , ) ( ) ( ) <sup>2</sup> *<sup>s</sup>*

In a large frequency range of study, the AC conductivity is made up of a high frequency linear contribution and an independent-frequency region at low frequency characterized by

property following usually the Arrhenius-like behavior, described by eq. (5). Materials presenting a thermal transition in the investigated temperature range (e.g. glass transition region) follow the non-linear Vogel-Fulcher-Tamman (VFT) behavior given by eq. (6):

( ) exp *<sup>a</sup> DC*

*DC*( ) exp *DT <sup>T</sup>*

DC conductivity is related to the structure and microstructure of the dielectric materials. Moreover, for a given material the dielectric properties are also strongly related to the way used to synthesize and process it. Hence, whereas it is difficult to predict a priori what will be the final DC conductivity from a theoretical point of view, it appears as impossible to estimate before what will be the impact of the material processing on this property. Consequently, it is fundamental to investigate, analyse and understand the origins of such variations of the DC conductivity in close relations with the material physico-chemical properties. Figure 6 presents the main parameters affecting the temperature dependence of the dc conductivity for various thermo-stable polymers. Figure 6a shows the variation of

of 400 °C-cured BPDA/PDA PI films for different thicknesses from 1.5 µm to 20 µm. It is

spectra of the PI films, allows relating this evolution to the remaining presence after the material processing of PI precursor (polyamic acid, PAA) residues (Diaham, 2011a). These impurities are a source of ionic species increasing the electrical conduction. Figure 6b shows

temperatures (*Tg*). The increase in *Tg* for PAI 2 (i.e. 335 °C against 280 °C for PAI 1 obtained

temperature (Diaham, 2009). The glass transition is therefore an important parameter

 σ∞

 σ∞

Boltzmann's constant, *D* is the material fragility and *T0* is the Vogel temperature.

*<sup>E</sup> <sup>T</sup>*

= −

= − <sup>−</sup>

*<sup>∞</sup>* is the conductivity at an infinite temperature, *Ea* is the activation energy, *kB* is the

 σ

*AC DC*

temperature-dependent parameter and *s* is the exponent of the power law (0<*s*≤1).

σ

 π εε

σ

σ

*AC*) as a function of frequency and temperature using eq. (4) (Kremer &

*f T* = =+ *f f T T AT f* (4)

*DC*) plateau. The DC conductivity is a temperature-dependent

*B*

*T T*

0 0

*DC* with increasing thickness. The inlet plot, showing the infrared

*DC* for two PAI films with different glass transition

σ

*k T*

*DC* is the static volume conductivity, *A* is a

(5)

(6)

*DC* increase towards higher

σ*DC*

**3.3.2 Electrical conductivity** 

σ

Schönhals, 2003; Jonscher, 1983):

σ

*0* is the vacuum permittivity,

σ

conductivity (

ε

a static conductivity (

observable an increase in

the temperature dependence of

σ

σ

by DSC in the inlet plot) allows shifting the onset of the

where

where σ controlling the charge motion across amorphous dielectrics. For high temperature operation, higher the *Tg*, wider is the temperature range of use. Figure 6c and 6d present respectively the σ*DC* temperature dependence of PA-F before and after a 400 °C annealing and as a function of thickness. It is shown that both annealing and material thickness improve the electrical properties (DC conductivity decreases). Inlet plots show that the PA-F crystallinity and the crystallite size are increased either with a thermal treatment or increasing thickness. Consequently, when the volume of the crystalline phase is increased the motion of charges within the material becomes more difficult, thus reducing the DC conductivity (Diaham, 2011b).

Fig. 6. Main parameters affecting the temperature dependence of the DC conductivity of various polymers: (a) thickness of BPDA/PDA PI films, (b) glass transition temperature in two different PAI films, (c) crystallization temperature for PA-F films, (d) thickness of PA-F films

Dielectrics for High Temperature

**3.3.3 Dielectric breakdown field** 

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 421

The dielectric strength is the capability of dielectrics to withstand high electric fields without failure. The dielectric breakdown field (*EBR*) is the upper limit of electric field that dielectrics can support under a voltage supply. Its value strongly depends on the electrode configuration (i.e. plane-plane or needle-plane electrodes). In homogeneous plane-plane

*BR*

*<sup>d</sup>* <sup>=</sup> (7)

<sup>−</sup> =− − (8)

is the scale parameter (i.e. the field

γis the

is the shape parameter quantifying the

4 5 6 7 8 9 10 11

*EBR* (MV/cm)

(b)

0.95

*F* **(%)**

95.7 63.2 27.1 9.5

Ø 0.3 mm β<sup>2</sup> =3,4 Ø 1.2 mm β<sup>2</sup> =7,1 Ø 2.4 mm β<sup>2</sup> =10,2

β γ

>>1 is related to a low scattering of the data) and

α

β

α

*BR <sup>V</sup> <sup>E</sup>*

( ) 1 exp *BR*

Even if the dielectric strength is an intrinsic parameter depending mainly on structural properties, it is the dielectric property the more sensitive to both experimental (electrode configuration, electrode surface, material thickness, voltage waveform, voltage ramp speed, ...) and environmental parameters (temperature, humidity, pressure, ...). If it is an important property to know, this appears as not self-sufficient for dimensioning electronic systems due to the extreme complexity of the electrical and thermal stresses induced by power devices and environmental severe stresses induced by applications. Consequently, the following section only gives the main experimental observable tendencies on the breakdown field of thermo-stable dielectrics. Recently, the influence of several parameters on the dielectric strength has been reported for BPDA/PDA PI and PA-F films (Diaham,

*<sup>E</sup> F E*

(a)

Fig. 9. Electrode diameter influence on the room temperature dielectric strength of

*F* **(%)**

95.7 63.2 27.1 9.5

0.95



log10(loge(1/1-

*F*))

0

1

*T*=25 °C

 Ø 0.3 mm Ø 1.2 mm Ø 2.4 mm

Ø 0.3 mm α=8,5 MV/cm, β<sup>1</sup>

Ø 1.2 mm α=8,64 MV/cm, β<sup>1</sup>

Ø 2.4 mm α=8,7MV/cm, β<sup>1</sup>

=16,2

=15,6

=16,5

Experimental breakdown values (EBR) exhibit a dispersion that requires statistical treatment in order to extract a mean value under the specific measurement conditions. Thus, the data

electrode configuration, the dielectric breakdown field is given by:

where *VBR* is the breakdown voltage and *d* is the dielectric thickness.

are usually analyzed using the Weibull distribution law (Weibull, 1951):

*BR*

β

where *F(EBR)* is the cumulative probability of failure,

γ=0).

0,1 1 10

(MV/cm)

*EBR*

BPDA/PDA PI (b) and PA-F (b) films

value for which 63.2% of the samples are failed),

width of the data distribution (i.e.

threshold parameter (often

2010b; Khazaka, 2011a).

*T*=25 °C

 Ø 0.3 mm Ø 0.5 mm Ø 1 mm Ø 2.5 mm Ø 5 mm



log10[loge(1/1-

*F*)]

0

1

Fig. 7. AC conductivity of various ceramics at (a) 300 °C, (b) 350 °C and (c) 400 °C

In the case of ceramic materials, it is difficult to detect the DC conductivity because of the presence of several interfacial relaxations (from internal or extrinsic origins) at low frequency. Moreover, the pure nature effect of the ceramic on the DC conductivity is difficult to be derived due to the strong additive influence on the synthesized materials. Figure 7 shows the frequency dependence of the AC conductivity of various ceramics at different temperatures. No evidence can be extracted on the substrate nature effect because all the substrates own different sintering processes (temperature, additive types and concentrations, …). However, these results let expect that most of the ceramics present relatively low DC conductivity less than 10-12 Ω-1 cm-1 at 400 °C such as some AlN or Si3N4 substrates.

Finally, Figure 8 presents the impact of the sintering process at 1800 °C (i.e. conventional thermal sintering and spark plasma sintering, SPS) on the AC conductivity of AlN ceramics with Y2O3 additives. The microstructure, density and the distribution of sintering additives impact the low frequency-dispersion of the dielectric properties. The SPS sintered AlN ceramic has lower AC conductivity values at high temperatures, even if the low-frequency plateau (i.e. DC conductivity) cannot be observed in the investigated frequency range.

Fig. 8. Sintering process influence on the AC conductivity of 1800 °C-sintered AlN: (a) conventional sintering process and (b) SPS sintering process. Bar length: 10 μm

#### **3.3.3 Dielectric breakdown field**

420 Silicon Carbide – Materials, Processing and Applications in Electronic Devices


In the case of ceramic materials, it is difficult to detect the DC conductivity because of the presence of several interfacial relaxations (from internal or extrinsic origins) at low frequency. Moreover, the pure nature effect of the ceramic on the DC conductivity is difficult to be derived due to the strong additive influence on the synthesized materials. Figure 7 shows the frequency dependence of the AC conductivity of various ceramics at different temperatures. No evidence can be extracted on the substrate nature effect because all the substrates own different sintering processes (temperature, additive types and concentrations, …). However, these results let expect that most of the ceramics present relatively low DC conductivity less than 10-12 Ω-1 cm-1 at 400 °C such as some AlN or Si3N4

Finally, Figure 8 presents the impact of the sintering process at 1800 °C (i.e. conventional thermal sintering and spark plasma sintering, SPS) on the AC conductivity of AlN ceramics with Y2O3 additives. The microstructure, density and the distribution of sintering additives impact the low frequency-dispersion of the dielectric properties. The SPS sintered AlN ceramic has lower AC conductivity values at high temperatures, even if the low-frequency plateau (i.e. DC conductivity) cannot be observed in the investigated

> 10-14 10-13 10-12 10-11 10-10 10-9 10-8 10-7 (b)

σ *AC* (

Fig. 8. Sintering process influence on the AC conductivity of 1800 °C-sintered AlN: (a) conventional sintering process and (b) SPS sintering process. Bar length: 10 μm

Ω-1 cm-1

)

Fig. 7. AC conductivity of various ceramics at (a) 300 °C, (b) 350 °C and (c) 400 °C

log10(Frequency) (Hz)

 AlN-A AlN-B Si3 N4 -A Si3 N4 -B Al2 O3

*T*=350°C

10-12 10-11 10-10 10-9 10-8 10-7 (c)


 25°C 110°C 150°C 210°C 250°C 310°C 350°C

log10(Frequency) (Hz)

σ *AC* (Ω-1 cm-1

)


log10(Frequency) (Hz)

 AlN-A AlN-B Si3 N4 -A

*T*=400°C


log10(Frequency) (Hz)

 AlN-A AlN-B Si3 N4 -A Si3 N4 -B Al2 O3

*T*=300°C


log10(Frequency) (Hz)

 25°C 110°C 150°C 210°C 250°C 310°C 350°C 10-12 10-11 10-10 10-9 10-8 10-7 (b)

σ *AC* (Ω-1 cm-1

)

10-12 10-11 10-10 10-9 10-8 10-7 (a)

substrates.

frequency range.

10-14 10-13 10-12 10-11 10-10 10-9 10-8 10-7 (a)

σ *AC* (

Ω-1 cm-1

)

σ *AC* (Ω-1 cm-1

)

The dielectric strength is the capability of dielectrics to withstand high electric fields without failure. The dielectric breakdown field (*EBR*) is the upper limit of electric field that dielectrics can support under a voltage supply. Its value strongly depends on the electrode configuration (i.e. plane-plane or needle-plane electrodes). In homogeneous plane-plane electrode configuration, the dielectric breakdown field is given by:

$$E\_{BR} = \frac{V\_{BR}}{d} \tag{7}$$

where *VBR* is the breakdown voltage and *d* is the dielectric thickness.

Experimental breakdown values (EBR) exhibit a dispersion that requires statistical treatment in order to extract a mean value under the specific measurement conditions. Thus, the data are usually analyzed using the Weibull distribution law (Weibull, 1951):

$$F\left(E\_{BR}\right) = 1 - \exp\left(-\frac{E\_{BR} - \gamma}{\alpha}\right)^{\beta} \tag{8}$$

where *F(EBR)* is the cumulative probability of failure, α is the scale parameter (i.e. the field value for which 63.2% of the samples are failed), β is the shape parameter quantifying the width of the data distribution (i.e. β>>1 is related to a low scattering of the data) and γ is the threshold parameter (often γ=0).

Even if the dielectric strength is an intrinsic parameter depending mainly on structural properties, it is the dielectric property the more sensitive to both experimental (electrode configuration, electrode surface, material thickness, voltage waveform, voltage ramp speed, ...) and environmental parameters (temperature, humidity, pressure, ...). If it is an important property to know, this appears as not self-sufficient for dimensioning electronic systems due to the extreme complexity of the electrical and thermal stresses induced by power devices and environmental severe stresses induced by applications. Consequently, the following section only gives the main experimental observable tendencies on the breakdown field of thermo-stable dielectrics. Recently, the influence of several parameters on the dielectric strength has been reported for BPDA/PDA PI and PA-F films (Diaham, 2010b; Khazaka, 2011a).

Fig. 9. Electrode diameter influence on the room temperature dielectric strength of BPDA/PDA PI (b) and PA-F (b) films

Dielectrics for High Temperature

2

0

**3.4.1 Thermal aging** 

dielectrics.

10

20

30

α (kV/mm)

40

50

α (MV/cm)

0 50 100 150 200 250 300 350 400

Al AlN1 AlN2 BN <sup>2</sup> O3 Si3

Ceramic nature

Temperature (°C)

*T*=450 °C (c)

 PI (BPDA/PDA) PAI PA-F (as-deposited)

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 423

4

0

10

20

30

α (kV/mm)

40

50

α (MV/cm)

1 10 100

Thickness (µm)

0 50 100 150 200 250 300 350 400 450

Temperature (°C)

 PI (BPDA/PDA) PA-F

 AlN process 1 AlN process 2 *T*=25 °C

(b)

(d)

(a)

N4

Fig. 10. Main parameters affecting the dielectric strength of various dielectrics: (a) temperature for PI, PAI and PA-F films, (b) thickness for PI and PA-F films at 25 °C, (c) temperature and ceramic nature for thick substrates (values taken from Chasserio, 2009), (d) two AlN substrates from different manufacturers (values taken from Chasserio, 2009)

For organic materials, the thermal aging appears among the more severe aging condition during long term service because temperature can carry out sufficient energy to break the structural bonds constituting the material skeleton. Although approximate models exist to predict accelerated aging under relatively smooth conditions, nowadays nobody can ensure their validity at very high temperatures near the limit of the polymer maximal operating temperature due to the absence of knowledge of the degradation mechanisms. Moreover, despite the importance of such a topic, there is a lack of studies in the literature dealing with long term thermal aging of polymers (Diaham, 2008; Khazaka, 2011b, Wayne Johnson, 2007; Zheng, 2007; Yao, 2010). It is indispensable to perform extremely long aging under such high temperature to validate high temperature reliability. In order to probe thermal-induced degradations, the dielectric breakdown strength is often appreciated because it gives information on the high field properties of

Figure 11 shows the dielectric strength evolution of BPDA/PDA PI films versus time for several aging temperatures in air. The figure compares also the dielectric strength evolution for films coated on different substrates. We can observe first that the life time

Figure 9 shows the electrode area influence on the cumulative probability versus *EBR* at room temperature for BPDA/PDA PI and PA-F films. For PI, it is possible to observe that the cumulative probability curve shifts towards lower breakdown fields with increasing the electrode diameter. The scale parameter α (*F*=63.2 %) decreases also with increasing the electrode diameter. In the same way, the shape parameter β (i.e. the slope of the fitting straight line) decreases with increasing the electrode diameter. These two simultaneous observations typically deal with an increase in the result scattering. They usually are characteristic of an increase in the probability to find defects or impurities in the material bulk leading to the failure of the insulating layer. In the case of PI films, this tendency is associated to the increase in the probability to find polyamic acid and solvent precursor residues in the film. Contrary to PI, PA-F exhibits an area independent dielectric strength behavior at high breakdown field. The fact that PA-F is a by-productless material could explain such a behavior. At low fields, an area dependence appears and is usually related to the presence of surfacic defects (i.e. stacking faults, pinholes, ...). Such studies allow often extrapolating dielectric strength for higher areas which can correspond to more practical cases.

Figure 10 presents the influence of the main other parameters on the dielectric breakdown field of dielectrics. The temperature dependence of the dielectric strength shows a general decrease in α with increasing temperature. For instance, thermo-stable polymers such as PI, PAI and PA-F films illustrate such a tendency (see Figure 10a) (Diaham, 2009, 2010b; Bechara, 2011). The thermal activation of the mobile charge transport and electromechanical constraints are usually brought to light to interpret the origin of the breakdown of polymers. Figure 10b shows the thickness dependence of the dielectric breakdown field of PI and PA-F films. It is usual to observe a general decrease in the breakdown field with increasing thickness for dielectric materials. Here also, this behavior can be explained by an increase in the probability to find defects in the dielectric layer. However, whatever the thickness investigated the dielectric strength remained high above 1 MV/cm.

As seen in the previous section, the processing parameters of ceramics have a great impact on dielectric properties evolution with temperature. When comparing AlN ceramic substrates from two different manufacturers, the differences in the processing conditions (i.e. organic binders, sintering additives, sintering temperature and dwell times) result in subtle differences in the final microstructures and crystallographic phase distributions, that modify considerably the dielectric strength evolution versus temperature (Chasserio, 2009). Figure 10c and 10d present the influence of the ceramic substrate nature and the impact of the sintering process of commercial AlN ceramics on the breakdown field. On one hand, AlN and Si3N4 ceramics appear as the materials owning the higher dielectric strength even at high temperature compared to Al2O3 and BN ceramics. However, for high temperature insulation applications cautions have to be taken, even in the choice of a same-type of ceramic. Indeed, from one supplier to another, breakdown field values can vary strongly in the high temperature range (see Figure 10d with two different commercial AlN).

#### **3.4 Aging and life time**

In power electronics applications, the high operating temperature (>200 °C) can result from either the ambient environment, the power dissipation, or a combination of both. Thus, after the first stage of initial material characterizations, it is necessary to follow the above properties during aging in harsh environment (temperature during time, thermal cycles, atmospheres, ...) in order to estimate the life time of dielectrics. In this section, the influences of the more usual aging conditions on the main sensitive parameters for each dielectric function in a power device assembly are presented.

Fig. 10. Main parameters affecting the dielectric strength of various dielectrics: (a) temperature for PI, PAI and PA-F films, (b) thickness for PI and PA-F films at 25 °C, (c) temperature and ceramic nature for thick substrates (values taken from Chasserio, 2009), (d) two AlN substrates from different manufacturers (values taken from Chasserio, 2009)

#### **3.4.1 Thermal aging**

422 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Figure 9 shows the electrode area influence on the cumulative probability versus *EBR* at room temperature for BPDA/PDA PI and PA-F films. For PI, it is possible to observe that the cumulative probability curve shifts towards lower breakdown fields with increasing the

line) decreases with increasing the electrode diameter. These two simultaneous observations typically deal with an increase in the result scattering. They usually are characteristic of an increase in the probability to find defects or impurities in the material bulk leading to the failure of the insulating layer. In the case of PI films, this tendency is associated to the increase in the probability to find polyamic acid and solvent precursor residues in the film. Contrary to PI, PA-F exhibits an area independent dielectric strength behavior at high breakdown field. The fact that PA-F is a by-productless material could explain such a behavior. At low fields, an area dependence appears and is usually related to the presence of surfacic defects (i.e. stacking faults, pinholes, ...). Such studies allow often extrapolating dielectric strength for higher areas

Figure 10 presents the influence of the main other parameters on the dielectric breakdown field of dielectrics. The temperature dependence of the dielectric strength shows a general

PAI and PA-F films illustrate such a tendency (see Figure 10a) (Diaham, 2009, 2010b; Bechara, 2011). The thermal activation of the mobile charge transport and electromechanical constraints are usually brought to light to interpret the origin of the breakdown of polymers. Figure 10b shows the thickness dependence of the dielectric breakdown field of PI and PA-F films. It is usual to observe a general decrease in the breakdown field with increasing thickness for dielectric materials. Here also, this behavior can be explained by an increase in the probability to find defects in the dielectric layer. However, whatever the thickness

As seen in the previous section, the processing parameters of ceramics have a great impact on dielectric properties evolution with temperature. When comparing AlN ceramic substrates from two different manufacturers, the differences in the processing conditions (i.e. organic binders, sintering additives, sintering temperature and dwell times) result in subtle differences in the final microstructures and crystallographic phase distributions, that modify considerably the dielectric strength evolution versus temperature (Chasserio, 2009). Figure 10c and 10d present the influence of the ceramic substrate nature and the impact of the sintering process of commercial AlN ceramics on the breakdown field. On one hand, AlN and Si3N4 ceramics appear as the materials owning the higher dielectric strength even at high temperature compared to Al2O3 and BN ceramics. However, for high temperature insulation applications cautions have to be taken, even in the choice of a same-type of ceramic. Indeed, from one supplier to another, breakdown field values can vary strongly in

the high temperature range (see Figure 10d with two different commercial AlN).

In power electronics applications, the high operating temperature (>200 °C) can result from either the ambient environment, the power dissipation, or a combination of both. Thus, after the first stage of initial material characterizations, it is necessary to follow the above properties during aging in harsh environment (temperature during time, thermal cycles, atmospheres, ...) in order to estimate the life time of dielectrics. In this section, the influences of the more usual aging conditions on the main sensitive parameters for each dielectric

investigated the dielectric strength remained high above 1 MV/cm.

with increasing temperature. For instance, thermo-stable polymers such as PI,

(*F*=63.2 %) decreases also with increasing the

(i.e. the slope of the fitting straight

β

α

electrode diameter. The scale parameter

which can correspond to more practical cases.

decrease in

α

**3.4 Aging and life time** 

function in a power device assembly are presented.

electrode diameter. In the same way, the shape parameter

For organic materials, the thermal aging appears among the more severe aging condition during long term service because temperature can carry out sufficient energy to break the structural bonds constituting the material skeleton. Although approximate models exist to predict accelerated aging under relatively smooth conditions, nowadays nobody can ensure their validity at very high temperatures near the limit of the polymer maximal operating temperature due to the absence of knowledge of the degradation mechanisms. Moreover, despite the importance of such a topic, there is a lack of studies in the literature dealing with long term thermal aging of polymers (Diaham, 2008; Khazaka, 2011b, Wayne Johnson, 2007; Zheng, 2007; Yao, 2010). It is indispensable to perform extremely long aging under such high temperature to validate high temperature reliability. In order to probe thermal-induced degradations, the dielectric breakdown strength is often appreciated because it gives information on the high field properties of dielectrics.

Figure 11 shows the dielectric strength evolution of BPDA/PDA PI films versus time for several aging temperatures in air. The figure compares also the dielectric strength evolution for films coated on different substrates. We can observe first that the life time

Dielectrics for High Temperature

applications.

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 425

Semicrystalline PA-F films (Parylene HT in commercial form) have been developed for their capability to support very high temperature during very long time even in oxidant atmosphere due to C—F bonds in the monomer structure. This relatively new material is supposedly stable for at least 1,000 hours at 350 °C in air atmosphere and 3 hours at 450 °C (see Figure 12) (Kumar, 2009). Nowadays only one study has been reported on the high temperature electrical properties of PA-F (Diaham, 2011b). This places PA-F among the potential suitable polymers for insulating coating in high temperature electronics

1 10 100 1000

Aging time in air (hours)

Fig. 12. Dielectric strength versus aging time of PA-F films for different temperatures of

Figure 13 shows the room temperature probed dielectric strength of silicone gel and silicone elastomer used for the device encapsulation function during thermal aging at 250 °C in air (Yao, 2010). The breakdown field falls down in the early first stage of aging, whatever the encapsulant materials, showing the difficulty nowadays to identify materials, for this important function of the packaging, able to operate at high temperature with reliability. Comparable results on other silicone-type materials have been reported recently by Zheng (Zheng, 2007). The penury of thick and soft materials appears as the main problem to increase the operating temperature of high voltage power devices above 250 °C, at least

One of the main problems in power electronic systems, besides the discrete materials performance is their heterogeneous mechanical properties. The thick insulating ceramics are especially under concern, more than the thinner passivation layers or the very soft encapsulating silicone gels classically used in power devices. In a first glance, SiC and the insulating ceramic substrates appear to have a similar CTE, but as stated earlier, metallic conductors that support the assemblies have much larger CTE, often 5 to 10 times larger. This makes the interface of ceramic and metal of the substrate component a susceptible point of failure. Thermal cycling amplifies this effect, as systems are exposed to wide

aging in air (values taken from Kumar, 2009). Measured at room temperature

without changing radically the architecture of power modules.

**3.4.2 Thermal cycling** 

temperature fluctuations over their lifetime.

Measured at 25 °C

0

1

α (MV/cm)

<sup>50</sup> 250 °C 350 °C 400 °C 450 °C

2

3

Air

depends on exposure temperatures. For instance, while the life time is more than 7,000 hours at 250 °C for PI coatings on stainless steel, it decreases strongly with increasing temperature: around 5,000 hours at 300 °C; 1,000 hours at 340 °C and 400 hours at 360 °C. This result underlines the thermal activation of the degradation. Secondly, the aging of PI coatings depends strongly on their substrate nature. Indeed, when the films are deposited on Si wafers, the life time of the material is strongly increased. For instance, the life time at 300 °C of films deposited on Si is superior than 5,000 hours while the same films deposited on metal substrates (stainless steel) is less than 5,000 hours. This can be interpreted by the the difference of the CTE between the PI films and the substrates. In the case of stainless steel substrates (CTE=17 ppm/°C), internal residual mechanical stresses are amounted in the BPDA/PDA PI layer (CTE=3-6 ppm/°C) which lead to premature degradation during thermal aging. The minimization of the CTE mismatch between the Si wafer (CTE=3 ppm/°C) and the BPDA/PDA PI film allows decreasing the mechanical stresses and so increasing the life time of the dielectric material. In the case of coatings on SiC wafers (for the component passivation function), similar results can be expected due to the compatible value of the SiC CTE value (3-5 ppm/°C) with the one of the BPDA/PDA PI.

Fig. 11. Dielectric strength versus aging time of BPDA/PDA PI films for different aging temperatures in air. Measured at 250 °C for aging at 250 °C and at 300 °C for higher aging temperatures. Stainless steel or silicon are used as film substrates.

depends on exposure temperatures. For instance, while the life time is more than 7,000 hours at 250 °C for PI coatings on stainless steel, it decreases strongly with increasing temperature: around 5,000 hours at 300 °C; 1,000 hours at 340 °C and 400 hours at 360 °C. This result underlines the thermal activation of the degradation. Secondly, the aging of PI coatings depends strongly on their substrate nature. Indeed, when the films are deposited on Si wafers, the life time of the material is strongly increased. For instance, the life time at 300 °C of films deposited on Si is superior than 5,000 hours while the same films deposited on metal substrates (stainless steel) is less than 5,000 hours. This can be interpreted by the the difference of the CTE between the PI films and the substrates. In the case of stainless steel substrates (CTE=17 ppm/°C), internal residual mechanical stresses are amounted in the BPDA/PDA PI layer (CTE=3-6 ppm/°C) which lead to premature degradation during thermal aging. The minimization of the CTE mismatch between the Si wafer (CTE=3 ppm/°C) and the BPDA/PDA PI film allows decreasing the mechanical stresses and so increasing the life time of the dielectric material. In the case of coatings on SiC wafers (for the component passivation function), similar results can be expected due to the compatible value of the SiC CTE value (3-5 ppm/°C) with the one of the

1 10 100 1000 10000

Aging time in air (hours)

Fig. 11. Dielectric strength versus aging time of BPDA/PDA PI films for different aging temperatures in air. Measured at 250 °C for aging at 250 °C and at 300 °C for higher aging

temperatures. Stainless steel or silicon are used as film substrates.

50 250 °C (steel) 300 °C (steel) 300 °C (Si) 340 °C (steel) 360 °C (steel)

0

1

2

α (MV/cm)

3

4

5

Air

BPDA/PDA PI.

Semicrystalline PA-F films (Parylene HT in commercial form) have been developed for their capability to support very high temperature during very long time even in oxidant atmosphere due to C—F bonds in the monomer structure. This relatively new material is supposedly stable for at least 1,000 hours at 350 °C in air atmosphere and 3 hours at 450 °C (see Figure 12) (Kumar, 2009). Nowadays only one study has been reported on the high temperature electrical properties of PA-F (Diaham, 2011b). This places PA-F among the potential suitable polymers for insulating coating in high temperature electronics applications.

Fig. 12. Dielectric strength versus aging time of PA-F films for different temperatures of aging in air (values taken from Kumar, 2009). Measured at room temperature

Figure 13 shows the room temperature probed dielectric strength of silicone gel and silicone elastomer used for the device encapsulation function during thermal aging at 250 °C in air (Yao, 2010). The breakdown field falls down in the early first stage of aging, whatever the encapsulant materials, showing the difficulty nowadays to identify materials, for this important function of the packaging, able to operate at high temperature with reliability. Comparable results on other silicone-type materials have been reported recently by Zheng (Zheng, 2007). The penury of thick and soft materials appears as the main problem to increase the operating temperature of high voltage power devices above 250 °C, at least without changing radically the architecture of power modules.

#### **3.4.2 Thermal cycling**

One of the main problems in power electronic systems, besides the discrete materials performance is their heterogeneous mechanical properties. The thick insulating ceramics are especially under concern, more than the thinner passivation layers or the very soft encapsulating silicone gels classically used in power devices. In a first glance, SiC and the insulating ceramic substrates appear to have a similar CTE, but as stated earlier, metallic conductors that support the assemblies have much larger CTE, often 5 to 10 times larger. This makes the interface of ceramic and metal of the substrate component a susceptible point of failure. Thermal cycling amplifies this effect, as systems are exposed to wide temperature fluctuations over their lifetime.

Dielectrics for High Temperature

**3.4.3 Atmosphere effects** 

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 427

Atmosphere nature acts as an important factor in the degradation of the polymers. In the case of BPDA/PDA PI films, Figure 15 shows the influence of the ambient atmosphere of aging on PI high temperature breakdown voltage. This result shows the increase in the life time when aging is performed into inert atmosphere. Oxygen atoms coming from air atmosphere lead to cut the PI monomer skeleton inducing a thermo-oxidative degradation processes (Khazaka, 2011b). In nitrogen atmosphere, the pure thermal degradation processes start at a further moment or temperature. This highlights the importance of using hermetic cases for power devices or to use oxygen barrier layers to protect PI films against oxygen. The effects of such

0 100 200 300 400 500 600 700 800 900

Aging time in air (hours)

Fig. 15. Influence of the ambient atmosphere on the life time of the mean breakdown voltage

On the contrary, the atmosphere nature seems to have low influence in the case of PA-F due to its low permeability to oxygen. This kind of materials could act as a good oxygen barrier coating over other oxygen sensitive polymers to protect them and increase their life time

This chapter summarizes recent worldwide research advances regarding reported insulating polymers and ceramics for high temperature power SiC devices and modules. A presentation of their main limiting physical properties regarding high temperature

Among polymeric materials, BPDA/PDA polyimide (PI) or fluorinated parylene (PA-F) are reported as interesting candidates for high temperature operation due to their highest and longest thermal stability. Moreover, they keep good dielectric properties even above 250 °C and even in oxidative atmosphere. PI film electrical properties are very sensitive to curing process while PA-F ones depend strongly on the crystallinity of the layer. However, even those materials may be not suitable for answering the highest temperature identified needs (above 400 °C) for long-term operation. Other polyamide-imide (PAI) and silicone elastomer (PDMS) materials, widely used up to now as thick insulating in electronic systems, exhibit a long-term operating limit below 250 °C. Today, it remains the issue of the existence of thick

*T*=360 °C

 Nitrogen Air

barriers (e.g. SiOx, SixNy) have been previously reported elsewhere (Khazaka, 2009).

applications, linked to microstructure analyses, is also presented.

Mean

of 4 μm-thick BPDA/PDA PI films at 360 °C

under high temperature conditions.

**4. Conclusion** 

*VBR* (Volts)

Fig. 13. Dielectric strength versus aging time of silicone and silicone elastomer materials aged at 250 °C in air (values taken from Yao, 2010), measured at room temperature.

In high temperature operating SiC devices, several technologies, such as Al2O3 direct-copperbond (DCB) and AlN DCB are limited in wide temperature cycling (Dupont, 2006a), as the ceramic is fractured by the mechanical stress that is imposed by the copper foil (see Figure 14a).

Fig. 14. (a) Failure on DCB substrate due to thermal cycling. Note the fracture across the ceramic material; the conchoidal fracture is initiated close to the copper foil edges. (b) Schema of metal work hardening of DCB ceramic substrate technologies during thermal cycling

The local mechanical stress is incremented by each cycling due to the work hardening of the copper foil, increasing its yield strength (see Figure 14b), this goes on until the maximum acceptable stress is attained at the ceramic, causing its failure (Dupont, 2006b).

For intermediate current levels, it is possible to diminish the metallization thickness to delay failure, or to make dimples applied to the edges of the copper foil (Dupont, 2006a). The availability of new substrate ceramics and available metallization types make possible to increase the reliability in increasing temperature cycling ranges. AlN can be brazed to aluminium, that has a higher CTE when compared to copper, but a lower recrystallization temperature (200-400 °C). This allows for a recrystallization after the work hardening imposed by the thermal cycling, keeping the constraints below the fracture limit of the AlN (Lei, 2009). On the other hand, Si3N4 has much higher fracture toughness, allowing it to resist the work hardening of copper across cycling (El Sawy & Fahmy, 1998). Si3N4 brazed to copper is claimed to last more than 5,000 cycles, ten times more than DCB technologies (Kyocera, 2009). Alternative approaches involve the use of low CTE metals as Kovar alloys (Lin & Yoon, 2005).

#### **3.4.3 Atmosphere effects**

426 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Measured at 25 °C Silicone elastomer 1

 Silicone Silicone elastomer 2

0 5 10 15 20 25 30

Aging time at 250 °C in air (days)

Fig. 13. Dielectric strength versus aging time of silicone and silicone elastomer materials aged at 250 °C in air (values taken from Yao, 2010), measured at room temperature.

Fig. 14. (a) Failure on DCB substrate due to thermal cycling. Note the fracture across the ceramic material; the conchoidal fracture is initiated close to the copper foil edges. (b) Schema of metal work hardening of DCB ceramic substrate technologies during thermal cycling

acceptable stress is attained at the ceramic, causing its failure (Dupont, 2006b).

The local mechanical stress is incremented by each cycling due to the work hardening of the copper foil, increasing its yield strength (see Figure 14b), this goes on until the maximum

For intermediate current levels, it is possible to diminish the metallization thickness to delay failure, or to make dimples applied to the edges of the copper foil (Dupont, 2006a). The availability of new substrate ceramics and available metallization types make possible to increase the reliability in increasing temperature cycling ranges. AlN can be brazed to aluminium, that has a higher CTE when compared to copper, but a lower recrystallization temperature (200-400 °C). This allows for a recrystallization after the work hardening imposed by the thermal cycling, keeping the constraints below the fracture limit of the AlN (Lei, 2009). On the other hand, Si3N4 has much higher fracture toughness, allowing it to resist the work hardening of copper across cycling (El Sawy & Fahmy, 1998). Si3N4 brazed to copper is claimed to last more than 5,000 cycles, ten times more than DCB technologies (Kyocera, 2009). Alternative approaches involve the use of low CTE metals as Kovar alloys

In high temperature operating SiC devices, several technologies, such as Al2O3 direct-copperbond (DCB) and AlN DCB are limited in wide temperature cycling (Dupont, 2006a), as the ceramic is fractured by the mechanical stress that is imposed by the copper foil (see Figure 14a).

0

(a) (b)

10

20

α (kV/mm)

(Lin & Yoon, 2005).

30

40

50

Atmosphere nature acts as an important factor in the degradation of the polymers. In the case of BPDA/PDA PI films, Figure 15 shows the influence of the ambient atmosphere of aging on PI high temperature breakdown voltage. This result shows the increase in the life time when aging is performed into inert atmosphere. Oxygen atoms coming from air atmosphere lead to cut the PI monomer skeleton inducing a thermo-oxidative degradation processes (Khazaka, 2011b). In nitrogen atmosphere, the pure thermal degradation processes start at a further moment or temperature. This highlights the importance of using hermetic cases for power devices or to use oxygen barrier layers to protect PI films against oxygen. The effects of such barriers (e.g. SiOx, SixNy) have been previously reported elsewhere (Khazaka, 2009).

Fig. 15. Influence of the ambient atmosphere on the life time of the mean breakdown voltage of 4 μm-thick BPDA/PDA PI films at 360 °C

On the contrary, the atmosphere nature seems to have low influence in the case of PA-F due to its low permeability to oxygen. This kind of materials could act as a good oxygen barrier coating over other oxygen sensitive polymers to protect them and increase their life time under high temperature conditions.

#### **4. Conclusion**

This chapter summarizes recent worldwide research advances regarding reported insulating polymers and ceramics for high temperature power SiC devices and modules. A presentation of their main limiting physical properties regarding high temperature applications, linked to microstructure analyses, is also presented.

Among polymeric materials, BPDA/PDA polyimide (PI) or fluorinated parylene (PA-F) are reported as interesting candidates for high temperature operation due to their highest and longest thermal stability. Moreover, they keep good dielectric properties even above 250 °C and even in oxidative atmosphere. PI film electrical properties are very sensitive to curing process while PA-F ones depend strongly on the crystallinity of the layer. However, even those materials may be not suitable for answering the highest temperature identified needs (above 400 °C) for long-term operation. Other polyamide-imide (PAI) and silicone elastomer (PDMS) materials, widely used up to now as thick insulating in electronic systems, exhibit a long-term operating limit below 250 °C. Today, it remains the issue of the existence of thick

Dielectrics for High Temperature

pp. 18-27

1851-1856

295-300

1321-1329

1999

19, 2011

Heidelberg, Germany

SiC Device Insulation: Review of New Polymeric and Ceramic Materials 429

Diaham, S.; Zelmat, S.; Locatelli, M.-L.; Dinculescu, S.; Lebey, T. & Decup, M. (2010b).

Diaham, S.; Locatelli, M.-L.; Lebey, T. & Malec, D. (2011a). Thermal Imidization

Diaham, S.; Bechara, M.; Locatelli, M.-L.; & Tenailleau, C. (2011b). Electrical Conductivity of

Dieckerhoff, S.; Guttowski, S. & Reichl, H. (2006). Performance Comparison of Advanced

Dupont, L (2006a), *Contribution à l'Etude de la Durée de Vie des Assemblages de Puissance dans* 

Funaki, T.; Balda, J. C.; Junghans, J.; Kashyap, A. S.; Mantooth, H.A.; Barlow, F.; Kimoto, T.

Liu, X.; Calata, J. N.; Wang, J. & Lu, G.Q. (1999). The Packaging of Integrated Power

Jonscher, A. K. (1983). *Dielectric Relaxation in Solids,* Chelsea Dielectrics Press, London, U.K. Khazaka, R.; Diaham, S.; Locatelli, M.-L. & Despax, B. (2009). Oxidative Thermal Aging

Khazaka, R.; Bechara, M.; Diaham, S.; & Locatelli, M.-L. (2011a). Parameters Affecting the

Khazaka, R.; Diaham, S.; Locatelli, M.-L.; Trupin, C. & Schlegel, B. (2011b). Thermal and

Kumar, R. (2009). Parylene HT: A High Temperature Vapor Phase Polymer for Electronics

*and Dielectric Phenomena (CEIDP)*, Cancun, Mexico, October 16-19, 2011 Kremer, F. & Schönhals, A. (2003). *Broadband Dielectric Spectroscopy,* Springer-Verlag, Berlin

*Applications*, Melbourne, Florida, USA, November 09-11, 2009

*Amplitude*, École Normale Supérieure, Cachan, Val-de-Marne, France Dupont, L.; Zoubir, K.; Lefebvre, S. & Bontemps, S. (2006b). Effects of metallization

Electronics, June 21-22, Paris, France, 2006

*Processing Technology*, Vol.77, pp. 266-272

Dielectric Breakdown of Polyimide Films: Area, Thickness and Temperature Dependence. *IEEE Transactions on Dielectrics and Electrical Insulation,* Vol.17, No.1,

Optimization of Polyimide Thin Films Using Fourier Transform Infrared Spectroscopy and Electrical Measurements. *Thin Solid Films,* Vol.519, No.6, pp.

Parylene F at High Temperature. *Journal of Electronic Materials,* Vol.40, No.3, pp.

Power Electronic Packages for Automotive Applications, Automotive Power

*des Environnements Haute Température et avec des Cycles Thermiques de Grande* 

thickness of ceramic substrates on the reliability of power assemblies under high temperature cycling. *Microelectronics Reliability*, Vol.46, No.9-11, pp. 1766-1771 El Sawy, A. & Fahmy, M. (1998), Brazing of Si3N4 Ceramic to Copper, *Journal of Materials* 

& Hikihara, T. (2007). Power Conversion with SiC Devices at Extremely High Ambient Temperatures, *IEEE Transaction on Power Electronics*, Vol.22, No.4, pp.

Electronics Modules Using Flip-Chip Technology, *Proceedings of the 17th Annual Power Electronics Seminar of Virginia Tech.*, Virginia, USA, September, pp. 361-367,

Effect on the Electrical Properties of BPDA/PDA Polyimide Films With and Without Oxygen Barrier Layers, *6th International Symposium on Polyimides and Other High Temperature/High Performance Polymers Synthesis, Characterization and* 

DC Breakdown Strength of Parylene F Thin Films, *Annual Report Conference on Electrical Insulation and Dielectric Phenomena (CEIDP)*, Cancun, Mexico, October 16-

Thermo-Oxidative Aging Effects on the Dielectric Properties of Thin Polyimide Films Coated on Metal Substrate, *Annual Report Conference on Electrical Insulation* 

Applications, *6th International Symposium on Polyimides and Other High* 

and soft insulating polymeric materials able to withstand high voltage even in the very high temperature range (>250 °C) during thousands of hours in order to answer the insulation of high temperature/high voltage SiC devices. Consequently, future research should concentrate towards this objective.

Regarding ceramics, the high thermal conductivity and the relatively invariant temperature dependence of the dielectric strength of aluminium nitride (AlN) and silicon nitride (Si3N4) place them as the more performing ceramic materials to realize metallized substrates for high temperature power electronic modules. However, the choice of their metallization nature and geometrical parameters is of first importance in order to improve the substrate life time. Thus, AlN/Al and Si3N4/Cu couples have already shown higher performances than classical DCB technologies, particularly in terms of thermal cycling resistance and could be good alternatives to answer the needs in high temperature and severe cycling substrate applications.

#### **5. Acknowledgment**

The authors would like to thank the 'Fondation Nationale de Recherche pour l'Aéronautique et l'Espace' (FNRAE) and the 'Direction Générale des Entreprises' (DGE) for the financial support of this work.

#### **6. References**


and soft insulating polymeric materials able to withstand high voltage even in the very high temperature range (>250 °C) during thousands of hours in order to answer the insulation of high temperature/high voltage SiC devices. Consequently, future research should

Regarding ceramics, the high thermal conductivity and the relatively invariant temperature dependence of the dielectric strength of aluminium nitride (AlN) and silicon nitride (Si3N4) place them as the more performing ceramic materials to realize metallized substrates for high temperature power electronic modules. However, the choice of their metallization nature and geometrical parameters is of first importance in order to improve the substrate life time. Thus, AlN/Al and Si3N4/Cu couples have already shown higher performances than classical DCB technologies, particularly in terms of thermal cycling resistance and could be good alternatives to answer the needs in high temperature and severe cycling

The authors would like to thank the 'Fondation Nationale de Recherche pour l'Aéronautique et l'Espace' (FNRAE) and the 'Direction Générale des Entreprises' (DGE) for

Adamec, V. (1974). Polarization Effect in an Epoxy Resin at Elevated Temperatures. *Polymer,* 

Bai, J. G.; Calata, J. N. & Lu, G. Q. (2004). Comparative Thermal and Thermo-Mechanical

Chasserio, N.; Guillemet-Fritsch, S.; Lebey, T.; & Dagdag, S. (2009). Ceramic Substrates for

Cooper Jr., J. A. & Agarwal, A. (2002). SiC Power Switching Devices - The Second Electronics Revolution, *Proceedings of IEEE*, Vol.90, No.6, p.956, June, 2002. Diaham, S.; Locatelli, M.-L. & Lebey, T. (2008). Improvement of Polyimide Electrical

Diaham, S.; Locatelli, M.-L.; Lebey, T. & Dinculescu, S. (2009). Dielectric and Thermal

Diaham, S.; Locatelli, M.-L.; Lebey, T. & Dinculescu, S. (2010a). Dielectric Measurements in

(CPES), pp. 147-153, Virginia Tech, Virginia, USA, April 18-20, 2004 Bechara, M.; Khazaka, R.; Diaham, S.; & Locatelli, M.-L. (2011). Crystallization Effect on the

*of Electrical Engineering,* accepted for publication

2549-5, Québec City, Canada, October 26-29, 2008

Virginia Beach, Virginia, USA, October 18-21, 2009

*Physical Journal Applied Physics,* Vol.49, pp. 10401

Analysis of Solder-Bump and Direct-Solder Bonded Power Devices Packages having Double Sided Cooling Capability, *Power Electronics Seminar Proceedings*

Dielectric Strength of Fluorinated Parylene at High Temperature. *European Journal* 

High-temperature Electronic Integration. *Journal of Electronic Materials,* Vol.38, No.1,

Properties during Short-Term of Thermal Aging, *Annual Report Conference on Electrical Insulation and Dielectric Phenomena (CEIDP)*, pp. 79-82, ISBN 978-1-4244-

Properties of Polyamide-imide (PAI) Films, *Annual Report Conference on Electrical Insulation and Dielectric Phenomena (CEIDP)*, pp. 482-485, ISBN 978-1-4244-4559-2,

Large Frequency and Temperature Ranges of an Aromatic Polymer. *The European* 

concentrate towards this objective.

substrate applications.

**5. Acknowledgment** 

**6. References** 

the financial support of this work.

pp. 164-174

Vol.15, No.8, pp. 496-498


**18** 

*Malaysia* 

**Water Jet Machining** 

*International Islamic University Malaysia* 

Ahsan Ali Khan and Mohammad Yeakub Ali

**Application of Silicon Carbide in Abrasive** 

Silicon carbide (SiC) is a compound consisting of silicon and carbon. It is also known as carborundum. SiC is used as an abrasive material after it was mass produced in 1893. The credit of mass production of SiC goes to Edward Goodrich Acheson. Now SiC is used not only as an abrasive, but it is also extensively used in making cutting tools, structural

AWJM is a well-established non-traditional machining technique used for cutting difficult-to machine materials. Nowadays, this process is being widely used for machining of hard materials like ceramics, ceramic composites, fiber-reinforced composites and titanium alloys where conventional machining fails to machine economically. The fact is that in AWJM no heat is developed and it has important implications where heat-affected zones are to be avoided. AWJM can cut everything what traditional machining can cut, as well as what traditional machining cannot cut such as too hard material (e.g. carbides), too soft material (e.g. rubber) and brittle material (e.g. glass, ceramics, etc.). The basic cutting tool used in water jet machining is highly pressurized water that is passed through a very small orifice, producing a very powerful tool that can cut almost any material. Depending on the materials, thickness of cut can range up to 25 mm and higher (Kalpakjian & Schmid, 2010). A water jet system consists of three components which are the water preparation system, pressure generation system and the cutting head and

As far as technology development is concerned, three types of water jet machining have been found and used. The first type is a typical water jet machining which was used in the middle of 18th century. The first attempt was in Russia in 1930s to cut hard rock using the pressurized water jet. The typical water jet machining used only water as the cutting tool which allows only cutting limited materials. The second type is AWJM as the improvement to the original water jet machining technique. Addition of abrasive to water enhances the capability of machining by many times. AWJM is an appropriate and cost effective technique for a number of uses and materials. Third type of AWJM includes cutting of difficult-to-machine materials, milling and 3-D-shaping, turning, piercing, drilling, polishing etc. These operations can be performed just by using plain water jet machining. However, due to special considerations such like the type of material or shape complexity of

the part to be produced, the addition of the abrasive material is required.

material, automotive parts, electrical systems, nuclear fuel parts, jewelries, etc.

**1. Introduction** 

motion system.

*Temperature/High Performance Polymers Synthesis, Characterization and Applications*, Melbourne, Florida, USA, November 09-11, 2009

Kyocera. Kyocera Si3N4 AMB Products Kyocera Si3N4 AMB Products (AMB ver.6.7). Retrieved March 1, 2009, from

http://www.leb.eei.uni-erlangen.de/winterakademie/2008/courses/course3\_ material.old/lifetimePower/course3\_material/lifetimePower/Reliability\_6.pdf


### **Application of Silicon Carbide in Abrasive Water Jet Machining**

Ahsan Ali Khan and Mohammad Yeakub Ali *International Islamic University Malaysia* 

*Malaysia* 

#### **1. Introduction**

430 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Kyocera. Kyocera Si3N4 AMB Products Kyocera Si3N4 AMB Products (AMB ver.6.7).

Lin, Z. & Yoon, R. J. (2005). An AlN-Based High Temperature Package for SiC Devices:

*Packaging Materials: Processes, Properties and Interfaces*, pp. 156-159, 2005 Locatelli, M. L.; Isoird, K.; Dinculescu, S.; Bley, V.; Lebey, T.; Planson, D.; Dutarde, E. &

http://www.leb.eei.uni-erlangen.de/winterakademie/2008/courses/course3\_ material.old/lifetimePower/course3\_material/lifetimePower/Reliability\_6.pdf Lei, T. G.; Calata, J. N.; Ngo, K. D. T. & Lu, G. (2009). Effects of Large-Temperature Cycling

Melbourne, Florida, USA, November 09-11, 2009

*Materials Reliability*, Vol.9, No.4, pp. 563-568

Retrieved March 1, 2009, from

30, 2001

*Insulation (IEEE)*, pp. 88-91, 1990

*Mechanics,* Vol.18, pp. 293-297

*Temperature/High Performance Polymers Synthesis, Characterization and Applications*,

Range on Direct Bond Aluminum Substrate. *IEEE Transactions on Device and* 

Materials and Processing. *Proceedings of the International Symposium on Advanced* 

Mermet-Guyennet, M. (2003). Study of Suitable Dielectric Materials Properties for High Electric Field and High Temperature Power Semiconductor Environment, *European Power Electronics Conference*, Toulouse, France, September 02-04, 2003 Mermet-Guyennet, M.; Castellazzi, A.; Lasserre, P. & Saiz, J. (2008). 3D Integration of Power

Semiconductor Devices Based on Surface Bump Technology, *5th International Conference on Integrated Power Electronic Systems* (CIPS), Nuremberg, Germany, 2008

Efficiency SiC Based Power Electronic Converters for Extreme Environments,

High Temperature Performances of Wide Bandgap Semiconductors for Vertical

(1965). Aromatic Polypyromellitimides From Aromatic Polyamic Acids. *Journal of* 

4H-SiC PiN Diodes with Low Power Losses. Proceedings of the 13th International Symposium on Power Semiconductor Devices & Ics (IEEE), Osaka, Japan, pp. 27-

Polyimide Films, *Conference Record of the International Symposium on Electrical* 

Polyimide Films: Area, Thickness and Temperature Dependence. *IEEE Transactions* 

Encapsulants for High-Voltage, High-Temperature Power Electronic Packaging,

Temperature Power Module, *40th International Symposium on Microelectronics* 

Mounce, S.; McPherson, B.; Schupbach, R. & Lostetter, A. (2006). Ultralightweight, High

*Proceedings of the IEEE Aerospace Conference*, New York, USA, pp. 1-19, 2006 Raynaud, C.; Tournier, D.; Morel, H. & Planson, D. (2010). Comparison of High Voltage and

Sroog, C. E.; Endrey, A. L.; Abramo, S. V.; Berr, C. E.; Edwards, W. M.; & Olivier, K. L.

Sugawara, Y.; Takayama, D.; Asano, K.; Singh, R.; Palmour, J. & Hayashi, T. (2001). 12-19 kV

Tsukiji, M.; Bitoh, W.; & Enomoto, J. (1990). Thermal Degradation and Endurance of

Wayne Johnson, R.; Wang, C.; Liu, Y. & Scofield, J. D. (2007). Dielectric Breakdown of

Weibull, W. (1951). A Statistical Distribution of Wide Applicability. *Journal of Applied* 

Yao, Y; Chen, Z.; Lu, G.-Q. ; Boroyevich, D. & Ngo, K. D. T. (2010). Characterization of

Zheng, Y. & Katsis, D. (2007). Investigation of Silicone-Based Encapsulants for a High

Power Devices. *Diamond and Related Materials*, Vol.19, No.1, pp. 1–6

*Polymer Science Part A: Polymer Chemistry,* Vol.3, pp. 1373-1390

*on Electronics Packaging Manufacturing,* Vol.30, No.3, pp. 182-193

*Electronic Components and Technology Conference (IEEE)*, 2010

*(IMAPS)*, San Jose, California, USA, November 11-15, 2007

Silicon carbide (SiC) is a compound consisting of silicon and carbon. It is also known as carborundum. SiC is used as an abrasive material after it was mass produced in 1893. The credit of mass production of SiC goes to Edward Goodrich Acheson. Now SiC is used not only as an abrasive, but it is also extensively used in making cutting tools, structural material, automotive parts, electrical systems, nuclear fuel parts, jewelries, etc.

AWJM is a well-established non-traditional machining technique used for cutting difficult-to machine materials. Nowadays, this process is being widely used for machining of hard materials like ceramics, ceramic composites, fiber-reinforced composites and titanium alloys where conventional machining fails to machine economically. The fact is that in AWJM no heat is developed and it has important implications where heat-affected zones are to be avoided. AWJM can cut everything what traditional machining can cut, as well as what traditional machining cannot cut such as too hard material (e.g. carbides), too soft material (e.g. rubber) and brittle material (e.g. glass, ceramics, etc.). The basic cutting tool used in water jet machining is highly pressurized water that is passed through a very small orifice, producing a very powerful tool that can cut almost any material. Depending on the materials, thickness of cut can range up to 25 mm and higher (Kalpakjian & Schmid, 2010). A water jet system consists of three components which are the water preparation system, pressure generation system and the cutting head and motion system.

As far as technology development is concerned, three types of water jet machining have been found and used. The first type is a typical water jet machining which was used in the middle of 18th century. The first attempt was in Russia in 1930s to cut hard rock using the pressurized water jet. The typical water jet machining used only water as the cutting tool which allows only cutting limited materials. The second type is AWJM as the improvement to the original water jet machining technique. Addition of abrasive to water enhances the capability of machining by many times. AWJM is an appropriate and cost effective technique for a number of uses and materials. Third type of AWJM includes cutting of difficult-to-machine materials, milling and 3-D-shaping, turning, piercing, drilling, polishing etc. These operations can be performed just by using plain water jet machining. However, due to special considerations such like the type of material or shape complexity of the part to be produced, the addition of the abrasive material is required.

Application of Silicon Carbide in Abrasive Water Jet Machining 433

1. Almost all types of materials can be machined by AWJM irrespective of their hardness, softness or brittleness. Almost all types of metals, plastics, fibrous materials, glass,

2. The surface machined by AWJM is smooth and usually they don't need any subsequent machining operation. Abrasives of very small size should be used to produce a smooth

3. AWJM is performed at room temperature. For that reason there is no problem of heat affected zone like other machining techniques. There is no structural change, no phase

4. The technique is environment friendly. Abrasives like SiC, garnet, alumina, silica sand, olivine together with water are environmental friendly. They don't emit any toxic vapor

5. A major problem in conventional machining like milling, drilling, etc. is burr forming. But AWJM doesn't produce any burr. Rather the technique is used for deburring.

In AWJM abrasives are added to water. The performance of AWJM to a great extend depends on the properties of abrasives. The geometry of cut is a key indicator of AWJM.

The main element of the abrasive water jet system is the abrasive jet. Water is pressurized up to 400 MPa and expelled through a nozzle to form a high-velocity jet. In AWJM abrasives are added to water using a specially shaped abrasive-jet nozzle from separate feed ports. As the momentum of water is transferred to the abrasives, their velocities increase rapidly. It results a focused, high-velocity stream of abrasives that exits the nozzle and performs the cutting action of the work surface. A schematic diagram of

Fig. 1. Abrasive water jet machining (Source: Kalpakjian & Schmid, 2010)

Normal water is filtered and passed to the intensifier. The intensifier acts as an amplifier as it converts the energy from the low-pressure hydraulic fluid into ultra-high pressure water. The hydraulic system provides fluid power to a reciprocating piston in the intensifier center section to amplify the water pressure. Using a control switch and a valve water is pressurized to the nozzle. Abrasive is added to water in the nozzle head (Fig 2) and the

transformation, no oxidation or no decarburization of the machined surface.

ceramics, rubbers, etc. can be machines by this technique.

surface.

or unpleasant odor.

**3. Elements of AWJM** 

AWJM is presented in Fig.1

**3.1 Water abrasive water jet machine** 

The use of the AWJM for machining or finishing purposes is based on the principle of erosion of the material upon which the jet is incident. The primary purpose of the abrasive material within the jet stream is to develop enough forces to erode the work material. However, the jet also accelerates the abrasive material to a high speed so that the kinetic energy of the abrasives is high enough to erode the work material. The secondary purpose of the water is to carry away both the abrasive material and the eroded material from the machining zone and clear the work area. AWJM gives a clean cut without any damage of the cut surface.

#### **2. Application of AWJM**

Generally, water jets are used for (Momber W. & Kovacevic, R., 1998)


In the area of manufacturing, the water jet-technique is used mainly for material cutting by plain water jets (e.g., plastics, thin metal sheets, textiles, foam, very hard materials like carbides, very soft materials like rubber, etc.). Sometimes burrs are formed due to machining of metals by conventional techniques. Those burrs can be removed by plain water jet machining. Some parts work under dynamic load and fatigue failure is the most common type of failure for those parts. Fatigue strength of those parts can be improves by peening the surface with a high pressure water jet. Fibrous materials like Kevlar cannot be machined by conventional machining techniques because of pullouts of the fibers. But AWJM can be employed to machine those materials without any pullout of fibers. AWJM can also be used for milling 3-D shapes. During abrasive water jet milling the surfaces not to be machined is masked before machining and only the areas to be machined are exposed to the jet head. Turning and grooving can also be performed on a lathe using an abrasive water jet. Piercing, drilling and trepanning are other cutting operations performed by AWJM. Water jet machining is a very common technique used to polish and improve work surface smoothness.

The performance of AWJM depends on some key factors. The hardness of the abrasive is an important factor. Harder the abrasive, faster and more efficient will be the machining process. Machining efficiency of abrasives also depends on their structure. Grain shape is another factor in evaluation of an abrasive material for abrasive water-jet process. Shape of abrasives is characterized by their relative proportions of length, width and thickness. During AWJM machining rate to a large extend depends on the size of the grains. Larger grains have higher kinetic energy and their cutting ability is also higher. But though the material removal rate of smaller grains is smaller, they are used for finishing works. Grainsize distribution and average grain size also play role in the performance of AWJM.


#### **3. Elements of AWJM**

432 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

The use of the AWJM for machining or finishing purposes is based on the principle of erosion of the material upon which the jet is incident. The primary purpose of the abrasive material within the jet stream is to develop enough forces to erode the work material. However, the jet also accelerates the abrasive material to a high speed so that the kinetic energy of the abrasives is high enough to erode the work material. The secondary purpose of the water is to carry away both the abrasive material and the eroded material from the machining zone and clear the work area. AWJM gives a clean cut without any damage of the

In the area of manufacturing, the water jet-technique is used mainly for material cutting by plain water jets (e.g., plastics, thin metal sheets, textiles, foam, very hard materials like carbides, very soft materials like rubber, etc.). Sometimes burrs are formed due to machining of metals by conventional techniques. Those burrs can be removed by plain water jet machining. Some parts work under dynamic load and fatigue failure is the most common type of failure for those parts. Fatigue strength of those parts can be improves by peening the surface with a high pressure water jet. Fibrous materials like Kevlar cannot be machined by conventional machining techniques because of pullouts of the fibers. But AWJM can be employed to machine those materials without any pullout of fibers. AWJM can also be used for milling 3-D shapes. During abrasive water jet milling the surfaces not to be machined is masked before machining and only the areas to be machined are exposed to the jet head. Turning and grooving can also be performed on a lathe using an abrasive water jet. Piercing, drilling and trepanning are other cutting operations performed by AWJM. Water jet machining is a very common technique used to polish and improve work surface

The performance of AWJM depends on some key factors. The hardness of the abrasive is an important factor. Harder the abrasive, faster and more efficient will be the machining process. Machining efficiency of abrasives also depends on their structure. Grain shape is another factor in evaluation of an abrasive material for abrasive water-jet process. Shape of abrasives is characterized by their relative proportions of length, width and thickness. During AWJM machining rate to a large extend depends on the size of the grains. Larger grains have higher kinetic energy and their cutting ability is also higher. But though the material removal rate of smaller grains is smaller, they are used for finishing works. Grain-

size distribution and average grain size also play role in the performance of AWJM.

AWJM has many advantages over other machining techniques. They are:

Generally, water jets are used for (Momber W. & Kovacevic, R., 1998)

cut surface.

smoothness.

**2. Application of AWJM** 

4. concrete hydrodemolition 5. rock fragmentation 6. solid stabilization 7. decontamination 8. demolition 9. metal recycling

10. manufacturing operations

3. paint, enamel and coating stripping

1. industrial cleaning 2. surface preparation

> In AWJM abrasives are added to water. The performance of AWJM to a great extend depends on the properties of abrasives. The geometry of cut is a key indicator of AWJM.

#### **3.1 Water abrasive water jet machine**

The main element of the abrasive water jet system is the abrasive jet. Water is pressurized up to 400 MPa and expelled through a nozzle to form a high-velocity jet. In AWJM abrasives are added to water using a specially shaped abrasive-jet nozzle from separate feed ports. As the momentum of water is transferred to the abrasives, their velocities increase rapidly. It results a focused, high-velocity stream of abrasives that exits the nozzle and performs the cutting action of the work surface. A schematic diagram of AWJM is presented in Fig.1

Fig. 1. Abrasive water jet machining (Source: Kalpakjian & Schmid, 2010)

Normal water is filtered and passed to the intensifier. The intensifier acts as an amplifier as it converts the energy from the low-pressure hydraulic fluid into ultra-high pressure water. The hydraulic system provides fluid power to a reciprocating piston in the intensifier center section to amplify the water pressure. Using a control switch and a valve water is pressurized to the nozzle. Abrasive is added to water in the nozzle head (Fig 2) and the

Application of Silicon Carbide in Abrasive Water Jet Machining 435

rapidly. It is worth mentioning that not all garnets are the same. There are wide variations in purity, hardness, sharpness, etc, that can also affect the cutting speed and operating cost. Garnet is a natural type of abrasive. Garnet has three basic structural components. They are Almandine (Fe3, Al2 (SiO4)3), Pyrope (Mg3Al2(SiO4)3) and Spessartite (Mn3Al2(SiO4)3). Garnet also contains impurities like SiO2, Al2O3, FeO, MnO, MgO, and CaO. The hardness of garnet abrasive particles of Almandine, Pyrope and Spessartite are 7-7.5 Mohs, 7.5 Mohs and 7-7.5 Mohs respectively. Aluminum Oxide (Al2O3) is another popular abrasive used in AWJM. It is also known as alumina. Its melting point is about 2,000°C and specific gravity is about 4.0. It is insoluble in water and organic liquids and slightly soluble in strong acids and alkalis. Alumina is available in two crystalline forms. Alpha alumina is composed of colorless hexagonal crystals. Gamma alumina is composed of minute colorless cubic crystals with specific gravity about 3.6 that are transformed to the alpha form at high temperatures. Alumina powder is formed by crushing crystalline Alumina. It is white when pure. Alumina is extremely tough and is wedge shaped. It is used for high-speed penetration in tough materials without excessive shedding or fracturing of the grains. It is mainly used for grinding high tensile strength materials like carbon steels, alloy steels, tough bronze and

hard woods. Other abrasives used in AWJM are olivine, slag, silica sand, etc.

For a divergent (V-shaped) slot b is larger than a as shown in Fig. 3.

The surface of cut is not vertical. It is characterized by a taper. Based on the width of cut at

T b a /2 <sup>R</sup> = − ( )

Hochheng & Chang, 1994 and Momber et al., 1996 investigated the top width of cut during AWJM of ceramics using magnesia and bauxite abrasives and found that the top width of cut decreases with the work feed. But the top width of cut increases with increase in

pressure and SOD. However, the slot may be of different shapes as shown in Fig. 4.

**3.3 Geometry of cut** 

Fig. 3. Usual geometry cut

the top and the bottom it is calculated as follows:

mixture comes out of the nozzle with a very high energy and pressure. In AWJM water is pressurized up to 55,000 pounds per square inch (psi) and then is forced to come out through a small orifice (round or square) at a speed of 2500 feet (762meters) per second, which is about two and half times the speed of sound. As water exits the nozzle at a high speed, the abrasive material is injected into the jet stream or sucked into the stream by a phenomenon known as the 'Venturi' effect. The main purpose of addition of abrasives is to enhance the jet length and improve the cutting ability of the jet. It was found by Chacko et al., 2003 that an addition of polymer to the water jet increases the jet penetration depth. There are two types of AWJM; the slurry and entrainment. The only thing that differentiates them is the way the abrasives are added to the water. In the slurry system, the abrasive is mixed with the water before the water being pressurized. The mix is then pressurized and passed to the end of the nozzle. This method causes extensive wear to the elements or parts of the water jet head due to the friction of the abrasives. In an entrainment system a pipe is connected to the water inlet. When the high-velocity pressurized water passes through the pipe, a vacuum is created causing the abrasive to be sucked into the water stream.

Fig. 2. Abrasive water jet cutting head*.*

#### **3.2 Abrasives**

SiC is known for its very high hardness and abrasion resistance. It is dark gray in color; its hardness and modulus of elasticity are 2,800 knoop (kg/mm2) and 476 GPa respectively. This material is produced according to specific technology to imitate the natural abrasive. It is heat resistant, and decomposes when heated to about 2,700°C. Very pure silicon carbide is white or colorless; crystals of it are used in semiconductors for high-temperature applications. Its small coefficient of expansion, which decreases with increasing temperature, high hardness and sharp crystal edges make it a very good abrasive. It is primarily used for grinding nonferrous materials such as brass, copper, bronze and aluminum. Other applications of SiC include grinding of glass, wood, rubber and plastics. Recently SiC is gaining popularity in as an excellent abrasive for AWJM.

But a survey shows that 90% of the AWJM is done using garnet (Mort, 1995). In industries 80 mesh garnet is a popular abrasive. It is possible to cut slightly faster rate with harder abrasives. However the harder abrasives also cause the mixing tube on the nozzle to wear rapidly. It is worth mentioning that not all garnets are the same. There are wide variations in purity, hardness, sharpness, etc, that can also affect the cutting speed and operating cost. Garnet is a natural type of abrasive. Garnet has three basic structural components. They are Almandine (Fe3, Al2 (SiO4)3), Pyrope (Mg3Al2(SiO4)3) and Spessartite (Mn3Al2(SiO4)3). Garnet also contains impurities like SiO2, Al2O3, FeO, MnO, MgO, and CaO. The hardness of garnet abrasive particles of Almandine, Pyrope and Spessartite are 7-7.5 Mohs, 7.5 Mohs and 7-7.5 Mohs respectively. Aluminum Oxide (Al2O3) is another popular abrasive used in AWJM. It is also known as alumina. Its melting point is about 2,000°C and specific gravity is about 4.0. It is insoluble in water and organic liquids and slightly soluble in strong acids and alkalis. Alumina is available in two crystalline forms. Alpha alumina is composed of colorless hexagonal crystals. Gamma alumina is composed of minute colorless cubic crystals with specific gravity about 3.6 that are transformed to the alpha form at high temperatures. Alumina powder is formed by crushing crystalline Alumina. It is white when pure. Alumina is extremely tough and is wedge shaped. It is used for high-speed penetration in tough materials without excessive shedding or fracturing of the grains. It is mainly used for grinding high tensile strength materials like carbon steels, alloy steels, tough bronze and hard woods. Other abrasives used in AWJM are olivine, slag, silica sand, etc.

#### **3.3 Geometry of cut**

434 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

mixture comes out of the nozzle with a very high energy and pressure. In AWJM water is pressurized up to 55,000 pounds per square inch (psi) and then is forced to come out through a small orifice (round or square) at a speed of 2500 feet (762meters) per second, which is about two and half times the speed of sound. As water exits the nozzle at a high speed, the abrasive material is injected into the jet stream or sucked into the stream by a phenomenon known as the 'Venturi' effect. The main purpose of addition of abrasives is to enhance the jet length and improve the cutting ability of the jet. It was found by Chacko et al., 2003 that an addition of polymer to the water jet increases the jet penetration depth. There are two types of AWJM; the slurry and entrainment. The only thing that differentiates them is the way the abrasives are added to the water. In the slurry system, the abrasive is mixed with the water before the water being pressurized. The mix is then pressurized and passed to the end of the nozzle. This method causes extensive wear to the elements or parts of the water jet head due to the friction of the abrasives. In an entrainment system a pipe is connected to the water inlet. When the high-velocity pressurized water passes through the

pipe, a vacuum is created causing the abrasive to be sucked into the water stream.

SiC is known for its very high hardness and abrasion resistance. It is dark gray in color; its hardness and modulus of elasticity are 2,800 knoop (kg/mm2) and 476 GPa respectively. This material is produced according to specific technology to imitate the natural abrasive. It is heat resistant, and decomposes when heated to about 2,700°C. Very pure silicon carbide is white or colorless; crystals of it are used in semiconductors for high-temperature applications. Its small coefficient of expansion, which decreases with increasing temperature, high hardness and sharp crystal edges make it a very good abrasive. It is primarily used for grinding nonferrous materials such as brass, copper, bronze and aluminum. Other applications of SiC include grinding of glass, wood, rubber and plastics.

But a survey shows that 90% of the AWJM is done using garnet (Mort, 1995). In industries 80 mesh garnet is a popular abrasive. It is possible to cut slightly faster rate with harder abrasives. However the harder abrasives also cause the mixing tube on the nozzle to wear

Recently SiC is gaining popularity in as an excellent abrasive for AWJM.

Fig. 2. Abrasive water jet cutting head*.*

**3.2 Abrasives** 

The surface of cut is not vertical. It is characterized by a taper. Based on the width of cut at the top and the bottom it is calculated as follows:

$$\mathbf{T\_{R}} = \begin{pmatrix} \mathbf{b} - \mathbf{a} \end{pmatrix} / \mathbf{2}$$

For a divergent (V-shaped) slot b is larger than a as shown in Fig. 3.

Fig. 3. Usual geometry cut

Hochheng & Chang, 1994 and Momber et al., 1996 investigated the top width of cut during AWJM of ceramics using magnesia and bauxite abrasives and found that the top width of cut decreases with the work feed. But the top width of cut increases with increase in pressure and SOD. However, the slot may be of different shapes as shown in Fig. 4.

Application of Silicon Carbide in Abrasive Water Jet Machining 437

Fig. 5. Surface cut by AWJM (Source: Momber & Kovacevic, 1998)

Fig. 6. Waviness of the cut surface (Source: Waterjet machining tolerances, 2011,

http://waterjets.org)

Fig. 4. Different shapes of the taper produced

V shaped taper is produced a result of the jet spending more time over an area to erode the top of the material more than the bottom. Also, "splash back" as the abrasives are bounced back from the material will tend to erode the sides. This is the most common type of taper. Reverse taper tends to happen during AWJM of soft materials where the material is rapidly eroded or when work feed rate is very slowly. Because as the jet stream expands farther away from the nozzle, it removes more material from the bottom than from the top. Barrel taper is produced where the middle is wider than the top or the bottom. Barrel taper is usually produced during machining of very thick materials. Rhomboid taper is actually normal V-shaped taper that has been tilted. It is produced when the nozzle is not perpendicular to the work surface.

Along the vertical surface the quality of the surface is not uniform. It can be divided into three zones (Fig. 5). At the top there is a small initial damaged zone height hIDZ. After that zone there is a smoother zone of height hSC. At the bottom of the surface there is a wavy surface of height hRC.

A typical waviness of the cut slot is shown in Fig. 6. Because the cutting tool is basically a beam of water, it acts as a "floppy tool". The jet lags between where it first enters the material and where it exits. Bottom of the jet lags behind the cutting head.

Barrel Taper Rhomboid or Trapezoidal Taper

V shaped taper is produced a result of the jet spending more time over an area to erode the top of the material more than the bottom. Also, "splash back" as the abrasives are bounced back from the material will tend to erode the sides. This is the most common type of taper. Reverse taper tends to happen during AWJM of soft materials where the material is rapidly eroded or when work feed rate is very slowly. Because as the jet stream expands farther away from the nozzle, it removes more material from the bottom than from the top. Barrel taper is produced where the middle is wider than the top or the bottom. Barrel taper is usually produced during machining of very thick materials. Rhomboid taper is actually normal V-shaped taper that has been tilted. It is produced

Along the vertical surface the quality of the surface is not uniform. It can be divided into three zones (Fig. 5). At the top there is a small initial damaged zone height hIDZ. After that zone there is a smoother zone of height hSC. At the bottom of the surface there is a wavy

A typical waviness of the cut slot is shown in Fig. 6. Because the cutting tool is basically a beam of water, it acts as a "floppy tool". The jet lags between where it first enters the

material and where it exits. Bottom of the jet lags behind the cutting head.

V-shaped Reverse Taper

Fig. 4. Different shapes of the taper produced

surface of height hRC.

when the nozzle is not perpendicular to the work surface.

Fig. 5. Surface cut by AWJM (Source: Momber & Kovacevic, 1998)

Fig. 6. Waviness of the cut surface (Source: Waterjet machining tolerances, 2011, http://waterjets.org)

Application of Silicon Carbide in Abrasive Water Jet Machining 439

Pressure: 35kbar

Pressure: 40kbar

Pressure: 45kbar

Pressure: 50kbar

Fig. 9. Surfaces of the machined insert carbide tools at constant flow rate (135g/min)

### **4. Machining of carbides by SiC**

#### **4.1 Influence of jet pressure on work surface roughness**

Experiments were conducted to investigate the influence of pressure on surface roughness. During the experiments the jet pressure were varied from 35 kbar to 50 kbar. Work feed rate, abrasive flow rate and depth of cut were kept contstant at 1.36 mm/min, 135 g/min and 3.18 mm respectively. The results show that increasing in pressure will decrease the roughness of work materials. In short surface finish becomes smoother and better. It is presented in Fig. 7. From Fig. 8 is evident that surface roughness drastically increases at a higher depth from the top surface. Fig. 9 shows the deterioration of surface smoothness along the depth of the cut surface.

Fig. 7. Effect of pressure on surface roughness

Fig. 8. Surface roughness at different depths from the top surface of carbide

Experiments were conducted to investigate the influence of pressure on surface roughness. During the experiments the jet pressure were varied from 35 kbar to 50 kbar. Work feed rate, abrasive flow rate and depth of cut were kept contstant at 1.36 mm/min, 135 g/min and 3.18 mm respectively. The results show that increasing in pressure will decrease the roughness of work materials. In short surface finish becomes smoother and better. It is presented in Fig. 7. From Fig. 8 is evident that surface roughness drastically increases at a higher depth from the top surface. Fig. 9 shows the deterioration of surface smoothness

> 35 40 45 50 pressure(kbar)

0.5 1.6 2.7 **depth(mm)**

Fig. 8. Surface roughness at different depths from the top surface of carbide

mid top bottom

> 35kbar 40bar 45kbar 50kbar

**4. Machining of carbides by SiC** 

along the depth of the cut surface.

0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000 8.000

0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000 8.000

**Ra**

Fig. 7. Effect of pressure on surface roughness

Ra

**4.1 Influence of jet pressure on work surface roughness** 

Pressure: 35kbar

Pressure: 40kbar

Pressure: 45kbar

Pressure: 50kbar

Fig. 9. Surfaces of the machined insert carbide tools at constant flow rate (135g/min)

Application of Silicon Carbide in Abrasive Water Jet Machining 441

Abrasive mass flow rate: 135 g/min

Abrasive mass flow rate: 145 g/min

Abrasive mass flow rate: 155 g/min

Abrasive mass flow rate: 165 g/min

Abrasive mass flow rate: 175 g/min

Fig. 12. Photographs of the machined carbides by varying the flow rate

#### **4.2 Influence of abrasive flow rate on surface roughness**

This section presents the results of the experiments conducted to investigate the influence of abrasive flow rate on surface finish. In these experiments abrasive flow rate was varied from 135 g/min to 175 g/min. Work feed rate and jet pressure were kept constant at 1.36 mm/min and 45 kbar respectively. It can be observed from Fig. 10 that abrasive flow rate play a vital role in water jet cutting. Surface roughness values are presented at the top, middle and at the bottom of the cut surface. The surface roughness decreases due to larger amount per minute of abrasives used. Means the surface of the cutting profile becomes smoother with higher flow rate. In Fig. 11 surface roughness at different depths from the top surface of carbide is presented*.* Increasing the flow rate will reduce surface roughness. Surfaces cut at different abrasive flow rate are presented in Fig. 12.

Fig. 10. Effect of abrasive mass flow rate on surface roughness during machining carbide.

Fig. 11. Surface roughness at different depths from the top surface of carbide

This section presents the results of the experiments conducted to investigate the influence of abrasive flow rate on surface finish. In these experiments abrasive flow rate was varied from 135 g/min to 175 g/min. Work feed rate and jet pressure were kept constant at 1.36 mm/min and 45 kbar respectively. It can be observed from Fig. 10 that abrasive flow rate play a vital role in water jet cutting. Surface roughness values are presented at the top, middle and at the bottom of the cut surface. The surface roughness decreases due to larger amount per minute of abrasives used. Means the surface of the cutting profile becomes smoother with higher flow rate. In Fig. 11 surface roughness at different depths from the top surface of carbide is presented*.* Increasing the flow rate will reduce surface roughness.

> 135 145 155 165 175 flowrate(g/min)

Fig. 10. Effect of abrasive mass flow rate on surface roughness during machining carbide.

0.5 1.6 2.7 depth(mm)

Fig. 11. Surface roughness at different depths from the top surface of carbide

mid top bottom

135g/min 145g/min 155g/min 165g/min 175g/min

**4.2 Influence of abrasive flow rate on surface roughness** 

Surfaces cut at different abrasive flow rate are presented in Fig. 12.

0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000

0.000 1.000 2.000 3.000 4.000 5.000 6.000 7.000

Ra

Ra

Abrasive mass flow rate: 135 g/min

Abrasive mass flow rate: 145 g/min

Abrasive mass flow rate: 155 g/min

Abrasive mass flow rate: 165 g/min

Abrasive mass flow rate: 175 g/min Fig. 12. Photographs of the machined carbides by varying the flow rate

Application of Silicon Carbide in Abrasive Water Jet Machining 443

For the analysis of the abrasive contamination, three zones with different depths namely Zone A, Zone B and Zone C have been analyzed. The zones of measuring abrasive

The relationship of the abrasive flow rate with abrasive contaminations at zones A, B and C are presented in Fig.14. In this figure zone 1, zone 2 and zone 3 stand for zones A, B and C respectively. In Fig. 14 Contamination has been presented for four different levels of pressures: 25 kpsi, 30 kpsi, 35kpsi and 40 kpsi with the changes of flow rate and feed rate. From these graphs it can be observed that the pattern of the graph is basically same for these sixteen samples. Most of the graphs show that the minimum amount of abrasive contamination is at Zone B. It is also observed that in most of the cases the maximum contamination is at zone C. As the abrasives move down they lose their kinetic energy and due to friction with the work surface they are embedded to it. This is also supported by the

Embedding of abrasives on the cut surface for a few experiments are shown Table 2.

**7**

**Graph of Contamination vs Zone for Experiment 6**

**01234 Zone** 

01234 **Zone**

**Graph of Contamination vs Zone for Experiment 8**

12 12

**5**

**6 6**

18

(a) Pressure 25 kpsi

 **Contamination**

**Contamination**

contamination are shown in Fig. 13.

Fig. 13. Measured zone of contamination

work of (Keyurkumar, 2004).

**Contamination**

**Contamination**

**6**

**Graph of Contamination vs Zone for Experiment 5**

**3**

**01234 Zone**

**11 10**

**7**

**01234 Zone**

**Graph of Contamination vs Zone for Experiment 7**

### **5. Contamination**

In AWJM material removal occurs through erosion and results from the interaction between an abrasive water jet and the work-piece materials. However, there are certain drawbacks in this technology. One of them is the contamination of the surfaces generated during the machining process by fractured abrasives. This contamination may generate serious problems with the further treatment of these surfaces, such as grinding, welding and/or coating. Other surface properties such as fatigue resistance will also influence negatively by surface contaminants.

Although particle embedment is a shortcoming of this technology, it can be minimized by controlling the machining parameters. However, it was reported that if the nozzle is oscillated during AWJM, then contamination reduces to a great extent (Chen et al., 2002 and Siores et al., 2006). The present study presents the influence of machining parameters on surface contamination of mild steel. Some investigations performed for the blasting of steel surfaces by air-driven solid particles found that the higher the blasting angle, the higher the contamination. Other than that, there is also a study to reduce the contamination by using the oscillation nozzle. The major aim of the present study is to investigate abrasive contamination on the surface of the mild steel during the AWJM. Furthermore, the parameters of the AWJM which are the feed rate, flow rate of the abrasive and the pressure were varied during the AWJ cutting process in order to study their level of contaminations with the changes of parameters. Then a quantitative microstructure analysis using Digital Camera Microscope was performed to investigate abrasive contamination at the cutting surfaces. In this microscope the number of embedded SiC abrasives can be counted when the image of the surface is captured. To investigate the contamination of the surfaces by abrasives, 16 experiments were conducted with variables shown in the Table 1.


Table 1. Experimental variables

In AWJM material removal occurs through erosion and results from the interaction between an abrasive water jet and the work-piece materials. However, there are certain drawbacks in this technology. One of them is the contamination of the surfaces generated during the machining process by fractured abrasives. This contamination may generate serious problems with the further treatment of these surfaces, such as grinding, welding and/or coating. Other surface properties such as fatigue resistance will also influence negatively by

Although particle embedment is a shortcoming of this technology, it can be minimized by controlling the machining parameters. However, it was reported that if the nozzle is oscillated during AWJM, then contamination reduces to a great extent (Chen et al., 2002 and Siores et al., 2006). The present study presents the influence of machining parameters on surface contamination of mild steel. Some investigations performed for the blasting of steel surfaces by air-driven solid particles found that the higher the blasting angle, the higher the contamination. Other than that, there is also a study to reduce the contamination by using the oscillation nozzle. The major aim of the present study is to investigate abrasive contamination on the surface of the mild steel during the AWJM. Furthermore, the parameters of the AWJM which are the feed rate, flow rate of the abrasive and the pressure were varied during the AWJ cutting process in order to study their level of contaminations with the changes of parameters. Then a quantitative microstructure analysis using Digital Camera Microscope was performed to investigate abrasive contamination at the cutting surfaces. In this microscope the number of embedded SiC abrasives can be counted when the image of the surface is captured. To investigate the contamination of the surfaces by

abrasives, 16 experiments were conducted with variables shown in the Table 1.

No. Pressure (kpsi) Flowrate (g/s) Feedrate (mm/min)

1 25 5 10 2 25 10 20 3 25 15 30 4 25 20 40 5 30 5 20 6 30 10 10 7 30 15 40 8 30 20 30 9 35 5 30 10 35 10 40 11 35 15 10 12 35 20 20 13 40 5 40 14 40 10 30 15 40 15 20 16 40 20 10

**5. Contamination** 

surface contaminants.

Table 1. Experimental variables

For the analysis of the abrasive contamination, three zones with different depths namely Zone A, Zone B and Zone C have been analyzed. The zones of measuring abrasive contamination are shown in Fig. 13.

Fig. 13. Measured zone of contamination

The relationship of the abrasive flow rate with abrasive contaminations at zones A, B and C are presented in Fig.14. In this figure zone 1, zone 2 and zone 3 stand for zones A, B and C respectively. In Fig. 14 Contamination has been presented for four different levels of pressures: 25 kpsi, 30 kpsi, 35kpsi and 40 kpsi with the changes of flow rate and feed rate. From these graphs it can be observed that the pattern of the graph is basically same for these sixteen samples. Most of the graphs show that the minimum amount of abrasive contamination is at Zone B. It is also observed that in most of the cases the maximum contamination is at zone C. As the abrasives move down they lose their kinetic energy and due to friction with the work surface they are embedded to it. This is also supported by the work of (Keyurkumar, 2004).

Embedding of abrasives on the cut surface for a few experiments are shown Table 2.

(a) Pressure 25 kpsi

Application of Silicon Carbide in Abrasive Water Jet Machining 445

**Graph of Contamination vs Zone for Experiment 10**

 **Zone**

**Graph of Contamination vs Zone for Experiment 12**

 **Zone** 

(d) Pressure 40 psi

**Contam ination**

**Contam ination**

**Graph of Contamination vs Zone for Experiment 9**

 **Zone** 

**Graph of Contamination vs Zone for Experiment 11**

 **Zone**

**Contam ination**

**Contam ination** Fig. 14. Contamination at different zones and at different pressures.

<sup>(</sup>c) Pressure 35 kpsi

**Contamination**

**Graph of Contamination vs Zone for Experiment 2**

 **Zone**

 **Zone**

**Graph of Contamination vs Zone for Experiment 14**

 **Zone** 

 **Zone**

**Graph of Contamination vs Zone for Experiment 16**

**<sup>5</sup>**

**Graph of Contamination vs Zone for Experiment 4**

**Graph of Contamination vs Zone for Experiment 1**

 **Zone**

 **Zone**

**Graph of Contamination vs Zone for Experiment 13**

 **Zone** 

**Graph of Contamination vs Zone for Experiment 15**

 **Zone**

**Contamination**

**Contamination**

**Graph of Contamination vs Zone for Experiment 3**

**<sup>15</sup>**

**Contamination**

**Contamination**

(b) Pressure 30 psi

**Contamination**

**Contamination**

(c) Pressure 35 kpsi

**Contamination** 

Fig. 14. Contamination at different zones and at different pressures.

Application of Silicon Carbide in Abrasive Water Jet Machining 447

In order to compare the capability of SiC with other abrasives, glass was taken as the work material. The main properties of glass are: hardness- 600 knoops, density- 2200 kg/m3, tensile strength- 70 MN/m2 and specific heat capacity- 750 J/kg oC. Three types of abrasives used in the present study were garnet, Al 2O3 and SiC. Their hardness is 1350 knoops, 2100 knoops and 2500 knoops respectively. Experiments were conducted on a water jet machine WJ 4080. The machine was equipped with a controller type 2100 CNC Control. The nozzle used for the abrasive water jet was made of carbide with the orifice diameter of 0.1 mm. The jet was perpendicular to the work surface. The abrasive water jet in cutting process is shown in Fig. 15. After the cutting process the top width and the bottom width of the slot was

Taper of cut was calculated according to the mathematical expression; TR = (b – a)/2, where TR, b and a are taper of cut, top width of cut and the bottom width of the cut respectively. Experimental investigations showed that during AWJM with different abrasives, the width of cut at the top of the slot was always greater than that at the bottom of the slots. It was explained by Wang et al., 1999 that as the abrasive particles move down the jet, they lose their kinetic energy and the relative strength zone of the jet is narrowed down. As a result, the width of cut at the bottom of the slot is smaller than that at the top. Influence of standoff distance (SOD) of the jet from the target material on the taper of cut during AWJM with different types of abrasives is illustrated in Fig. 16. It can be observed that the garnet abrasives produced the largest taper of cut followed by Al 2O3 and SiC abrasives. Among the three types of abrasives used, SiC is the hardest material and consequently it retains its cutting ability as it moves down. Therefore, the difference between the widths at the top and bottom of the slot is small and consequently, the taper angle is also smaller. On the other hand, garnet abrasives lose their sharpness and as a result the bottom width becomes much narrower than the top width. Fig. 16 also shows that for all kinds of abrasives, the taper of cut increases with SOD. This is due to the divergence shape of the jet. As SOD is increased,

**5.1 Comparison of SiC with other abrasives in AWJM** 

measured using an optical microscope Mitutiyo Hismet II.

**5.2 Effect of different cutting parameters on taper of cut** 

the jet focus area also increases resulting increase in the width of cut.

Fig. 15. Experimental set-up





Table 2. Embedding of the abrasives

**Experiment 5 Zone A Zone B Zone C** 

**Contaminations** 6 3 7

**Contamination** 5 6 6

**Experiment 13 Zone A Zone B Zone C** 

**Contamination** 17 10 18

**Contamination** 11 4 6

**Experiment 16 Zone A Zone B Zone C** 

**Experiment 6 Zone A Zone B Zone C** 

**Pressure :** 30

**Flow rate :** 5 g/s **Feed rate :** 20mm/min

**Pressure :** 30

**Flow rate :** 10

**Feed rate :** 10 mm/min

**Pressure :** 40

**Flow rate :** 5 g/s **Feed rate :** 40 mm/min

**Pressure :** 40

**Flow rate :** 20

**Feed rate :** 10 mm/min

Table 2. Embedding of the abrasives

kpsi

g/s

kpsi

kpsi

g/s

kpsi

#### **5.1 Comparison of SiC with other abrasives in AWJM**

In order to compare the capability of SiC with other abrasives, glass was taken as the work material. The main properties of glass are: hardness- 600 knoops, density- 2200 kg/m3, tensile strength- 70 MN/m2 and specific heat capacity- 750 J/kg oC. Three types of abrasives used in the present study were garnet, Al 2O3 and SiC. Their hardness is 1350 knoops, 2100 knoops and 2500 knoops respectively. Experiments were conducted on a water jet machine WJ 4080. The machine was equipped with a controller type 2100 CNC Control. The nozzle used for the abrasive water jet was made of carbide with the orifice diameter of 0.1 mm. The jet was perpendicular to the work surface. The abrasive water jet in cutting process is shown in Fig. 15. After the cutting process the top width and the bottom width of the slot was measured using an optical microscope Mitutiyo Hismet II.

Fig. 15. Experimental set-up

#### **5.2 Effect of different cutting parameters on taper of cut**

Taper of cut was calculated according to the mathematical expression; TR = (b – a)/2, where TR, b and a are taper of cut, top width of cut and the bottom width of the cut respectively. Experimental investigations showed that during AWJM with different abrasives, the width of cut at the top of the slot was always greater than that at the bottom of the slots. It was explained by Wang et al., 1999 that as the abrasive particles move down the jet, they lose their kinetic energy and the relative strength zone of the jet is narrowed down. As a result, the width of cut at the bottom of the slot is smaller than that at the top. Influence of standoff distance (SOD) of the jet from the target material on the taper of cut during AWJM with different types of abrasives is illustrated in Fig. 16. It can be observed that the garnet abrasives produced the largest taper of cut followed by Al 2O3 and SiC abrasives. Among the three types of abrasives used, SiC is the hardest material and consequently it retains its cutting ability as it moves down. Therefore, the difference between the widths at the top and bottom of the slot is small and consequently, the taper angle is also smaller. On the other hand, garnet abrasives lose their sharpness and as a result the bottom width becomes much narrower than the top width. Fig. 16 also shows that for all kinds of abrasives, the taper of cut increases with SOD. This is due to the divergence shape of the jet. As SOD is increased, the jet focus area also increases resulting increase in the width of cut.

Application of Silicon Carbide in Abrasive Water Jet Machining 449

entrance to the jet exit. As a result, taper of cut reduces with increase in jet pressure. Louis et al., 2003 indicates some other positive aspects of using higher pressure. He found that the depth of penetration of the jet increases and cutting efficiency improves with increase in pressure. On the other hand, abrasive flow rate can be reduced if the jet pressure is increased. However, taper of cut is smaller for SiC abrasives followed by Al2O3 and garnet. SiC abrasives being harder than Al2O3 and garnet abrasives retain their sharp edges both at the entrance and the exit of the jet and produce the smallest width of cut. On the other hand, garnet abrasives being comparatively softer lose the sharpness of their cutting edges when

**Influence of pressure on taper of cut**

Aluminum oxide SiC

garnet

Aluminum oxide SiC

garnet

0 20 40 60 **Jet pressure (ksi)**

0246 **SOD (mm)**

**Influence of SOD on average width of cut**

they are near the jet exit.

0

Fig. 18. Effect of pressure on taper of cut

0

Fig. 19. Effect of SOD on taper of cut

0.5

1

1.5

**average width od cut** 

**(mm)**

2

2.5

0.05

0.1

0.15

**Taper of cut**

0.2

0.25

Fig. 16. Influence of SOD on taper of cut

Fig. 17. Influence of feed rate on taper of cut

Fig. 17 shows the relationship between work feed rate and taper of cut during AWJM using different abrasive materials. For all types of abrasives the taper of cut shows an increasing trend with increase in work feed rate. With increase in work feed rate the machining zone is exposed to the jet for a shorter time. Cutting process is less effective at the jet exit that results an increase in taper of cut. Conner & Hashish, 2003 also found similar effect of feed rate on taper of cut during AWJM of aerospace materials using garnet abrasives. Garnet abrasives demonstrate a high taper of cut followed by SiC and Al 2O3.

Influence of pressure on taper of cut is illustrated in Fig. 18. Taper of cut decreases with increase in jet pressure for all the types of abrasives used. At a higher pressure the abrasives have higher energy and they retain their cutting ability as they move down from the jet

**Influence of SOD on taper of cut**

Aluminum oxide

SiC

garnet

Aluminum oxide SiC Garnet

0246 **SOD (mm)**

0 20 40 60 **work feed rate (mm/min)**

Fig. 17 shows the relationship between work feed rate and taper of cut during AWJM using different abrasive materials. For all types of abrasives the taper of cut shows an increasing trend with increase in work feed rate. With increase in work feed rate the machining zone is exposed to the jet for a shorter time. Cutting process is less effective at the jet exit that results an increase in taper of cut. Conner & Hashish, 2003 also found similar effect of feed rate on taper of cut during AWJM of aerospace materials using garnet abrasives. Garnet abrasives

Influence of pressure on taper of cut is illustrated in Fig. 18. Taper of cut decreases with increase in jet pressure for all the types of abrasives used. At a higher pressure the abrasives have higher energy and they retain their cutting ability as they move down from the jet

**Influence of work feed rate on taper of cut**

0

Fig. 16. Influence of SOD on taper of cut

0

Fig. 17. Influence of feed rate on taper of cut

demonstrate a high taper of cut followed by SiC and Al 2O3.

0.05

0.1

0.15

**taper of cut**

0.2

0.25

0.3

0.05

0.1

0.15

**Taper of cut** 

0.2

0.25

entrance to the jet exit. As a result, taper of cut reduces with increase in jet pressure. Louis et al., 2003 indicates some other positive aspects of using higher pressure. He found that the depth of penetration of the jet increases and cutting efficiency improves with increase in pressure. On the other hand, abrasive flow rate can be reduced if the jet pressure is increased. However, taper of cut is smaller for SiC abrasives followed by Al2O3 and garnet. SiC abrasives being harder than Al2O3 and garnet abrasives retain their sharp edges both at the entrance and the exit of the jet and produce the smallest width of cut. On the other hand, garnet abrasives being comparatively softer lose the sharpness of their cutting edges when they are near the jet exit.

Fig. 18. Effect of pressure on taper of cut

Fig. 19. Effect of SOD on taper of cut

Application of Silicon Carbide in Abrasive Water Jet Machining 451

work is exposed to the jet for a shorter period. The effect of pressure on average width of cut during AWJM is shown in Fig. 21. A higher pressure produces a jet of higher energy with capability of more effective cutting. From Fig. 19, Fig. 20 and Fig. 21 it was observed that in all the cases the average width of cut produced by SiC was higher than those produced by Al 2O3 and garnet abrasives. It can be concluded that hardness is a key property of abrasive

From the above discussions it can be concluded that during AWJM of carbides using SiC abrasives, machined surface roughness reduces if the jet pressure is increased. Surface smoothness deteriorates from the top of cut towards the exit of cut. The roughness of cut surface reduces with increase in abrasive flow rate since more abrasives are available per unit area of cut. The lower most zone of the cut surface is the most contaminated zone followed by the top most zone and the middle zone. Taper of cut increases with increase in SOD. Garnet abrasives produce a larger taper of cut followed by Al2O3 and SiC. This is due to higher hardness of SiC compared to Al2O3 and garnet. Taper of cut also increases with increase in work feed rate. But taper of cut reduces with increase in pressure. A higher pressure increases the kinetic energy of the abrasives and the divergence of the jet is reduced that causes a decrease in taper of cut. An increase in SOD increases the focus area of the jet and increases the average width of cut. But increase in feed rate reduces the average width of cut since the surface to be cut is exposed to the jet for a shorter time. A higher jet pressure increases the kinetic energy of the abrasive particles and enhances their cutting ability. As a result, increase in pressure causes increase in the average width of cut. SiC is harder than Al2O3 and garnet. As a result, its cutting ability is also higher than that of Al2O3 and garnet. Therefore, the average width of cut produced by SiC is higher than those

The authors of this work are indebted to the Research Management Center, International Islamic University Malaysia (IIUM) for its continuous help during the research work. The author is also grateful to Momber W. & Kovacevic, R. (1998), since some information has

Chen F., Patel K., Siores E. & Momber A. (2002). Minimizing particle contamination at

Chacko, V.; Gupta, A. & Summers, A. (2003). Comparative performance study of

abrasive water jet machined surfaces by a nozzle oscillation technique. *International Journal of Machine Tools & Manufacture*, Vol. 42, pp. 1385–1390, ISSN

polyacrylamide and xanthum polymer in abrasive slurry jet, *Proceedings of American Water Jet Conference,* Houston, Texas, USA [3] Hocheng, H. & and Chang, R. (1994). Material removal analysis in abrasive water jet cutting of ceramic plates. *Journal of Materials Processing Technology,* Vol. 40, pp. 287-304,

materials.

**6. Conclusions** 

produced by Al2O3 and garnet.

**7. Acknowledgement** 

been taken from their book.

0890-6955

ISSN 0924-0136

**8. References** 

#### **5.3 Effect of different parameters on average width of cut**

It has been established that though the abrasive water jet is a divergent one, the effective cutting zone of the jet is convergent, since the abrasives at the outer zone of the jet lose their kinetic energy. As a result, the width of cut at the jet entrance is always greater than the same at the jet exit. In Fig. 19 to Fig. 21 the average value of the widths of the jet entrance and jet exit has been taken as the width of cut. From Fig. 19 it is obvious that the average width of cut increases with increase in SOD which is due to the divergence shape of the jet. It was found that SiC produced the widest slot followed by Al2O3 and garnet. This is by virtue of higher hardness of SiC that enables more effective material removal.

Fig. 20. Effect of feed on width of cut

Fig. 21. Effect of pressure on width of cut

Influence of work feed rate on the average width of cut is illustrated in Fig. 20. Average width of cut decreases with increase in work feed rate since with the increase in feed rate the work is exposed to the jet for a shorter period. The effect of pressure on average width of cut during AWJM is shown in Fig. 21. A higher pressure produces a jet of higher energy with capability of more effective cutting. From Fig. 19, Fig. 20 and Fig. 21 it was observed that in all the cases the average width of cut produced by SiC was higher than those produced by Al 2O3 and garnet abrasives. It can be concluded that hardness is a key property of abrasive materials.

#### **6. Conclusions**

450 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

It has been established that though the abrasive water jet is a divergent one, the effective cutting zone of the jet is convergent, since the abrasives at the outer zone of the jet lose their kinetic energy. As a result, the width of cut at the jet entrance is always greater than the same at the jet exit. In Fig. 19 to Fig. 21 the average value of the widths of the jet entrance and jet exit has been taken as the width of cut. From Fig. 19 it is obvious that the average width of cut increases with increase in SOD which is due to the divergence shape of the jet. It was found that SiC produced the widest slot followed by Al2O3 and garnet. This is by

**Influence of feed rate on average width of cut**

Aluminum oxide SiC

Garnet

Aluminum oxide SiC garnet

0 20 40 60 **Work feed rate (mm/min)**

**Influence of pressure on average width of cut**

0 20 40 60 **jet pressure (ksi)**

Influence of work feed rate on the average width of cut is illustrated in Fig. 20. Average width of cut decreases with increase in work feed rate since with the increase in feed rate the

virtue of higher hardness of SiC that enables more effective material removal.

**5.3 Effect of different parameters on average width of cut** 

0

Fig. 20. Effect of feed on width of cut

0

Fig. 21. Effect of pressure on width of cut

0.5

1

1.5

**average width of cut (mm)**

2

2.5

0.5

1

1.5

**Avarage width of cut** 

**(mm)**

2

2.5

From the above discussions it can be concluded that during AWJM of carbides using SiC abrasives, machined surface roughness reduces if the jet pressure is increased. Surface smoothness deteriorates from the top of cut towards the exit of cut. The roughness of cut surface reduces with increase in abrasive flow rate since more abrasives are available per unit area of cut. The lower most zone of the cut surface is the most contaminated zone followed by the top most zone and the middle zone. Taper of cut increases with increase in SOD. Garnet abrasives produce a larger taper of cut followed by Al2O3 and SiC. This is due to higher hardness of SiC compared to Al2O3 and garnet. Taper of cut also increases with increase in work feed rate. But taper of cut reduces with increase in pressure. A higher pressure increases the kinetic energy of the abrasives and the divergence of the jet is reduced that causes a decrease in taper of cut. An increase in SOD increases the focus area of the jet and increases the average width of cut. But increase in feed rate reduces the average width of cut since the surface to be cut is exposed to the jet for a shorter time. A higher jet pressure increases the kinetic energy of the abrasive particles and enhances their cutting ability. As a result, increase in pressure causes increase in the average width of cut. SiC is harder than Al2O3 and garnet. As a result, its cutting ability is also higher than that of Al2O3 and garnet. Therefore, the average width of cut produced by SiC is higher than those produced by Al2O3 and garnet.

#### **7. Acknowledgement**

The authors of this work are indebted to the Research Management Center, International Islamic University Malaysia (IIUM) for its continuous help during the research work. The author is also grateful to Momber W. & Kovacevic, R. (1998), since some information has been taken from their book.

#### **8. References**


**19** 

*India* 

**Silicon Carbide Filled Polymer Composite for** 

**A Comparative Analysis of Experimental and** 

*1Department of Mechanical Engineering, National Institute of Technology, Rourkela, 2Department of Mechanical Engineering, National Institute of Technology, Hamirpur,* 

Polymer composites form important class of engineering materials and are commonly used in mechanical components. Because of their high strength-to-weight and stiffness-to-weight ratios, they are extensively used for a wide variety of structural applications as in aerospace, automotive, gear pumps handling industrial fluids, cams, power plants, bushes, bearing cages and chemical industries. Whereas, wear performance in nonlubricated condition is a key factor for the material selection and fabrication procedure (Hutchings, 1992). Glass fiber reinforced polymer composites traditionally show poor wear resistance due to the brittle nature of the fibers. Many researchers have been reported on the effect of fiber, filler and matrix materials so far in the literature regarding economical and functional benefits to both consumers and industrial manufacturers (Budinski, 1997; Chand et al., 2000; Tripathy and Furey, 1993). The addition of hard particulate ceramic fillers not only improves the wear performance of the particulate filled polymer composites but also reduce the cost of the composites. In order to obtain improve wear performances many researchers modified polymers using different fillers (Briscoe et al. 1974; Tanaka 1986; Bahadur et al, 1994;

Silicon carbide (SiC) is one such ceramic material that has great potential for overcoming the current inadequacies of abrasive products due to its inherent characteristic of being chemically inert and consequently resistant to improve mechanical and wear resistance material. It has an excellent abrasive nature and has been produced for grinding wheels and other for more than hundred years. Now-a-days the material has been developed into a high quality technical grade ceramic with very good mechanical properties. It is used in abrasives, ceramics, refractories, and other high-performance applications. Silicon carbide is composed of tetrahedra of carbon and silicon atoms with strong bonds in the crystal lattice. This produces a very strong material and not attacked by any acids or alkalis or molten salts

To this end, the present research work is undertaken to develop a new class of glass fiber based polymer composite filled with SiC particulate and study the effect of various

Bahadur and Tabor,1985; Kishore et al. 2000; Wang et al. 2003).

up to 800°C (Nordsletten et al. 1996).

**1. Introduction** 

**Erosive Environment Application:** 

Sandhyarani Biswas1, Amar Patnaik2 and Pradeep Kumar2

**FE Simulation Results** 


### **Silicon Carbide Filled Polymer Composite for Erosive Environment Application: A Comparative Analysis of Experimental and FE Simulation Results**

Sandhyarani Biswas1, Amar Patnaik2 and Pradeep Kumar2 *1Department of Mechanical Engineering, National Institute of Technology, Rourkela, 2Department of Mechanical Engineering, National Institute of Technology, Hamirpur, India* 

#### **1. Introduction**

452 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Conner, I & and Hashish, M. (2003). Abrasive water jet machining of aerospace structural

Kalpakjian, S. & Schmid, R. (2010). Manufacturing Engineering and Technology, Pentice

Keyurkumar, P. (2004). Quantitative Evaluation of Abrasive Contamination In Ductile

Louis, H.; Mohamed, M. & Pude, F. (2003). Cutting mechanism and cutting efficiency for

Momber, W.; Eusch, I & Kovacevic, R. (1996). Machining refractory ceramics with abrasive water jet. *Journal of Materials Science,* Vol. 31, pp. 6485-6493, ISSN 0022-2461 Momber W. & Kovacevic, R. (1998). *Principles of Abrasive Water Jet Machining,* Springer, ISBN

Mort, A. (1995). Results of abrasive water jet market survey, *Proceedings of 8th American Water* 

Siores, E.; Chen, L.; Lemma, E. & Wang, J. (2006). Optimizing the AWJ Cutting Process of

Wang, J. & Wong, K. (1999). A study of abrasive water jet cutting of metallic coated sheet

*Machine Tools & Manufacture.* Vol.42, pp. 781–789, ISSN 0890-6955

Ductile Materials Using Nozzle Oscillation Technique, *International Journal of* 

steel. *International Journal of Machine Tools and Manufacture,* Vol. 39, pp. 855-870,

*Jet Conference,* Vol. 1, pp. 259-289, Houston, Texas, USA

Waterjet machining tolerances, 2011, Available from http://waterjets.org

Texas, USA

Houston, Texas, USA

3540762396, London

ISSN 0890-6955

Hall, ISBN 978-981-06-8144-9, Singapore

Vol. 44, pp. 1125-1132, ISSN 0890-6955

sheet and thin plate materials. *Proceedings of American water Jet Conference*, Houston,

Material During Abrasive Water Jet Machining And Minimizing With A Nozzle Head Oscillation Technique. *International Journal of Machine Tools & Manufacture,* 

water pressures above 600 MPa. *Proceedings of American Water Jet Conference*,

Polymer composites form important class of engineering materials and are commonly used in mechanical components. Because of their high strength-to-weight and stiffness-to-weight ratios, they are extensively used for a wide variety of structural applications as in aerospace, automotive, gear pumps handling industrial fluids, cams, power plants, bushes, bearing cages and chemical industries. Whereas, wear performance in nonlubricated condition is a key factor for the material selection and fabrication procedure (Hutchings, 1992). Glass fiber reinforced polymer composites traditionally show poor wear resistance due to the brittle nature of the fibers. Many researchers have been reported on the effect of fiber, filler and matrix materials so far in the literature regarding economical and functional benefits to both consumers and industrial manufacturers (Budinski, 1997; Chand et al., 2000; Tripathy and Furey, 1993). The addition of hard particulate ceramic fillers not only improves the wear performance of the particulate filled polymer composites but also reduce the cost of the composites. In order to obtain improve wear performances many researchers modified polymers using different fillers (Briscoe et al. 1974; Tanaka 1986; Bahadur et al, 1994; Bahadur and Tabor,1985; Kishore et al. 2000; Wang et al. 2003).

Silicon carbide (SiC) is one such ceramic material that has great potential for overcoming the current inadequacies of abrasive products due to its inherent characteristic of being chemically inert and consequently resistant to improve mechanical and wear resistance material. It has an excellent abrasive nature and has been produced for grinding wheels and other for more than hundred years. Now-a-days the material has been developed into a high quality technical grade ceramic with very good mechanical properties. It is used in abrasives, ceramics, refractories, and other high-performance applications. Silicon carbide is composed of tetrahedra of carbon and silicon atoms with strong bonds in the crystal lattice. This produces a very strong material and not attacked by any acids or alkalis or molten salts up to 800°C (Nordsletten et al. 1996).

To this end, the present research work is undertaken to develop a new class of glass fiber based polymer composite filled with SiC particulate and study the effect of various

Silicon Carbide Filled Polymer Composite for Erosive Environment

aggregately in the proposed model.

angle)

Application: A Comparative Analysis of Experimental and FE Simulation Results 455

process instead of single particle (ElTobgy et al., 2005). In this study, 125 spherical shaped particles were used to ensure the accuracy of the proposed model. All the particles are striking the target area at random locations. There are 10 groups which contain 125 particles

(a) (b)

(c) (d)

Fig. 1. Schematic diagram of target composite material and nozzle (a: 30° impingement angle, b: 45° impingement angle, c: 60° impingement angle and d: 90° impingement

operational variables, material parameters and their interactive influences on erosive wear behavior of these composites. A nite element (FE) model (AUTO-DYN) of erosive wear is established for damage assessment and validated by a well designed set of experiments. The eroded surfaces of these composites are analyzed with scanning electron microscopy (SEM), and the erosion wear mechanisms of the composites are investigated.

#### **2. Experimental**

#### **2.1 Preparation of composites**

In this study, short E-glass fiber with 6mm length (Elastic modulus of 72.5 GPa and density of 2.59 gm/cc) is taken to prepare all the particulate filled (SiC) glass fiber reinforced polyester composites. The unsaturated isophthalic polyester resin (Elastic modulus 3.25GPa and density 1.35gm/cc) is manufactured by Ciba Geigy and locally supplied by Northern Polymers Ltd. New Delhi, India. The composite fabricated in two different parts. One part having different fiber loading with varying the fiber weight fraction from 10wt% to 50wt% at an increment of 10wt% and the second part, SiC filled short glass fiber reinforced polyester resin with three different percentages (0wt%, 10wt% and 20wt% of SiC). The mixture is poured into various moulds conforming to the requirements of various testing conditions and characterization standards. The entrapped air bubbles (if any) are removed carefully with a sliding roller and the mould is closed for curing at a temperature of 30°C for 24 h at a constant pressure of 10 kg/cm2.

#### **2.2 Air-jet erosion tester**

The solid particle erosion test rig as per ASTM G76 used in the present study consists of an air compressor, a particle feeder, an air particle mixing chamber and accelerating chamber. The equipment was designed to feed erodent particles into a high velocity air stream, which propelled the particles against the specimen surface (Strzepa et al., 1993; Routbort et al., 1981). The erodent particles entrained in a stream of compressed air and accelerated down to a 65mm long brass nozzle with 3mm inside diameter to impact on a specimen mounted on an angle xture. The velocity of the eroding particles is determined using rotating disc method (Ruff and Ives, 1975). The steady state erosion rate was determined by weighing the sample before and after the end of each test. While the impingement angles ranges from 30° to 90° and the test duration was 20min for each run. The erodent used for this test was river silica sand particle of three different sizes, i.e. 250, 350 and 450μm. The sample was cleaned with a blast of compressed air before each weighing to remove all loosely adhering debris. The mass loss from the target was measured with an analytical balance of ±0.01mg accuracy. The process is repeated every 10 minutes till the erosion rate attains a constant value called steady-state-erosion-rate. Finally, the worn surfaces of some selected samples are examined by scanning electron microscope JEOL JSM-6480LV.

#### **2.3 Finite element model**

In the present work, the erosive wear processes are modeled using an explicit dynamic code ANSYS/AUTO-DYN. The eight-node brick hexahedral elements with one integration point are used in the 3D simulation. The mesh is rened to a standard cubic element in order to calculate the erosion rate at the targeted area on the composites. It has been studied in literature that simulating a single particle was not sufficient to get valid results therefore subsequently considered three or more particles were needed to simulate the erosion

operational variables, material parameters and their interactive influences on erosive wear behavior of these composites. A nite element (FE) model (AUTO-DYN) of erosive wear is established for damage assessment and validated by a well designed set of experiments. The eroded surfaces of these composites are analyzed with scanning electron microscopy (SEM),

In this study, short E-glass fiber with 6mm length (Elastic modulus of 72.5 GPa and density of 2.59 gm/cc) is taken to prepare all the particulate filled (SiC) glass fiber reinforced polyester composites. The unsaturated isophthalic polyester resin (Elastic modulus 3.25GPa and density 1.35gm/cc) is manufactured by Ciba Geigy and locally supplied by Northern Polymers Ltd. New Delhi, India. The composite fabricated in two different parts. One part having different fiber loading with varying the fiber weight fraction from 10wt% to 50wt% at an increment of 10wt% and the second part, SiC filled short glass fiber reinforced polyester resin with three different percentages (0wt%, 10wt% and 20wt% of SiC). The mixture is poured into various moulds conforming to the requirements of various testing conditions and characterization standards. The entrapped air bubbles (if any) are removed carefully with a sliding roller and the mould is closed for curing at a temperature of 30°C for

The solid particle erosion test rig as per ASTM G76 used in the present study consists of an air compressor, a particle feeder, an air particle mixing chamber and accelerating chamber. The equipment was designed to feed erodent particles into a high velocity air stream, which propelled the particles against the specimen surface (Strzepa et al., 1993; Routbort et al., 1981). The erodent particles entrained in a stream of compressed air and accelerated down to a 65mm long brass nozzle with 3mm inside diameter to impact on a specimen mounted on an angle xture. The velocity of the eroding particles is determined using rotating disc method (Ruff and Ives, 1975). The steady state erosion rate was determined by weighing the sample before and after the end of each test. While the impingement angles ranges from 30° to 90° and the test duration was 20min for each run. The erodent used for this test was river silica sand particle of three different sizes, i.e. 250, 350 and 450μm. The sample was cleaned with a blast of compressed air before each weighing to remove all loosely adhering debris. The mass loss from the target was measured with an analytical balance of ±0.01mg accuracy. The process is repeated every 10 minutes till the erosion rate attains a constant value called steady-state-erosion-rate. Finally, the worn surfaces of some selected samples are examined

In the present work, the erosive wear processes are modeled using an explicit dynamic code ANSYS/AUTO-DYN. The eight-node brick hexahedral elements with one integration point are used in the 3D simulation. The mesh is rened to a standard cubic element in order to calculate the erosion rate at the targeted area on the composites. It has been studied in literature that simulating a single particle was not sufficient to get valid results therefore subsequently considered three or more particles were needed to simulate the erosion

and the erosion wear mechanisms of the composites are investigated.

**2. Experimental** 

**2.1 Preparation of composites** 

24 h at a constant pressure of 10 kg/cm2.

by scanning electron microscope JEOL JSM-6480LV.

**2.2 Air-jet erosion tester** 

**2.3 Finite element model** 

process instead of single particle (ElTobgy et al., 2005). In this study, 125 spherical shaped particles were used to ensure the accuracy of the proposed model. All the particles are striking the target area at random locations. There are 10 groups which contain 125 particles aggregately in the proposed model.

(a) (b)

Fig. 1. Schematic diagram of target composite material and nozzle (a: 30° impingement angle, b: 45° impingement angle, c: 60° impingement angle and d: 90° impingement angle)

Silicon Carbide Filled Polymer Composite for Erosive Environment

Table 1. Levels for various control factors

**3.1 Erosive wear of the composites** 

**3. Results and discussion** 

experimental technique.

**3.2 Steady state erosion rate** 

**3.2.1 Influence of impingement angle** 

erosion conditions such as the properties of target material.

Control factor Level

Application: A Comparative Analysis of Experimental and FE Simulation Results 457

A:Impact velocity 43 54 65 m/sec B: SiC content 0 10 20 % C:Impingement angle 30 60 90 Degree D:Stand-off distance 65 75 85 mm E:Erodent size 250 350 450 µm

The results have been organized and discussed in two sections. Firstly, the steady state erosion characteristics of the composites are determined for selected level of optimally controlled operating variables and compared the steady state results with the simple finite element simulation results to observe the variations in erosion rate with respective to impingement angle and the next, simulations results have been analyzed under Taguchi's

Solid-particle erosion is a complex wear phenomena influenced by a number of control factors such as impact velocity, angle of impingement, erodent particle size, stand-offdistance, materials properties, erodent particles geometry and environment temperature etc. Among these, impingement angle is the one of the most important parameter and widely studied parameter in the erosion study of materials (Hutchings, 1992; Tsuda et al., 2006). The erosion rate is measured of function of impingement angle, two types of material behavior generally observed in the target material i.e. ductile and brittle nature. The ductile nature of materials is characterized by maximum erosion rate at acute angle (15-30°) and for brittle behavior of materials, the maximum erosion rate is observed at normal impingement angle (90°). But as far as polymer matrix composites are concerned the composite materials show versatile in nature depending upon the fabrication procedure and type of reinforcing material. The reinforced composites show a semi-ductile behavior having the maximum erosion rate in the range of 45-60° (Hutchings, 1992), unlike the above two categories. This classification, however, is not absolute as the erosion of material has a strong dependence on

In the present study of SiC filled glass fiber-polyester composites, the erosion rate increases monotonically with the increase in impingement angle and reaches maximum at 45° impingement angle for particulate filled composites. However, for unfilled composite the maximum erosion rate is found to be at 60° impingement angle. This indicated that all the particulate filled and unfilled composites show semi-ductile erosion behaviour irrespective of filler content. Similarly, the finite element analysis simulated results are in good agreement with the experimental results as observed in Figure 2. As far as erosion resistance is concerned 20wt% SiC filled composites show better erosion resistance among other particulate filled and unfilled composites. Whereas, unfilled composites shows maximum erosion rate as compared with 10wt% and 20wt% SiC filled glass fiber reinforced polyester composites both in

experimental and finite element analysis simulated results as shown in Figure 2.

I II III Units

Each group has 12 particles which would impact the surface simultaneously and followed by another simultaneous particles group, and so on. According to the researchers, the distance between any two particles' centers in the same group is no less than 0.6r (r is the radius of the particles) to avoid the damage interaction (Woytowitz and Richman, 1999). The nite element model of the target material and simulated nozzle is shown in Figure 1. For the particles, the rotation degrees of freedom are constrained. Generally, the erosion rate (g/g) was used to characterize the erosion performance of the target materials.

#### **2.4 Taguchi experimental design**

Taguchi method is a statistical tool for the purpose of designing experimental procedure and mainly improving product quality. It uses the orthogonal array to set up the experiment for the advantages of less number and optimizes the process parameters by the analysis of signal-to-noise (SN) ratio. Taguchi method has become a powerful analysis tool for improving the experimental results to get high quality at low cost (Peace, 1993; Phadke, 1989). Therefore, a large number of factors are included so that non-significant variables can be identified at earliest opportunity**.** The impact of five such parameters are studied using L27 (313) orthogonal design. The operating conditions under which wear tests are carried out are given in Table 1. In conventional full factorial experiment design, it would require 35 = 273 runs to study five parameters each at three levels whereas, Taguchi's factorial experiment approach reduces it to only 27 runs offering a great advantage in terms of experimental time and cost. The experimental observations are further transformed into signal-to-noise (S/N) ratio. There are several S/N ratios available depending on the type of performance characteristics. The S/N ratio for minimum wear rate can be expressed as "lower is better" characteristic, which is calculated as logarithmic transformation of loss function as shown below (Peace, 1993).

Smaller is the better characteristic:

$$\frac{\mathbf{S}}{\mathbf{N}} = -10 \log \frac{1}{\mathbf{n}} \boldsymbol{\Sigma} \mathbf{Y}^2 \tag{1}$$

where, n the number of observations and y the observed experimental data.

The plan of the experiments is as follows: the first column is assigned to impact velocity (A), the second column to SiC content (B), the fifth column to impingement angle (C), the ninth column to stand-off distance (D) and the tenth column to erodent size (E), the third and fourth column are assigned to (A× B)1 and (A× B)2 respectively to estimate interaction between impact velocity (A) and SiC content (B), the sixth and seventh column are assigned to (B× C)1 and (B× C)2 respectively to estimate interaction between the SiC content (B) and impingement angle (C), the eight and eleventh column are assigned to (A × C)1 and (A× C)2 respectively to estimate interaction between the impact velocity (A) and impingement angle (C) and the remaining columns are used to estimate experimental errors. The output to be studied is erosion rate (Er) and the tests are repeated twice corresponding to 54 tests. Furthermore, a statistical analysis of variance (ANOVA) is performed to identify the process parameters that are statistically significant. With the S/N and ANOVA analyses, the optimal combination of the process parameters can be predicted to a useful level of accuracy. Finally, a confirmation experiment is conducted to verify the optimal process parameters obtained from the parameter design.


Table 1. Levels for various control factors

#### **3. Results and discussion**

456 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Each group has 12 particles which would impact the surface simultaneously and followed by another simultaneous particles group, and so on. According to the researchers, the distance between any two particles' centers in the same group is no less than 0.6r (r is the radius of the particles) to avoid the damage interaction (Woytowitz and Richman, 1999). The nite element model of the target material and simulated nozzle is shown in Figure 1. For the particles, the rotation degrees of freedom are constrained. Generally, the erosion rate

Taguchi method is a statistical tool for the purpose of designing experimental procedure and mainly improving product quality. It uses the orthogonal array to set up the experiment for the advantages of less number and optimizes the process parameters by the analysis of signal-to-noise (SN) ratio. Taguchi method has become a powerful analysis tool for improving the experimental results to get high quality at low cost (Peace, 1993; Phadke, 1989). Therefore, a large number of factors are included so that non-significant variables can be identified at earliest opportunity**.** The impact of five such parameters are studied using L27 (313) orthogonal design. The operating conditions under which wear tests are carried out are given in Table 1. In conventional full factorial experiment design, it would require 35 = 273 runs to study five parameters each at three levels whereas, Taguchi's factorial experiment approach reduces it to only 27 runs offering a great advantage in terms of experimental time and cost. The experimental observations are further transformed into signal-to-noise (S/N) ratio. There are several S/N ratios available depending on the type of performance characteristics. The S/N ratio for minimum wear rate can be expressed as "lower is better" characteristic, which is calculated as logarithmic transformation of loss

S 1 <sup>2</sup> 10log Y N n = − (1)

The plan of the experiments is as follows: the first column is assigned to impact velocity (A), the second column to SiC content (B), the fifth column to impingement angle (C), the ninth column to stand-off distance (D) and the tenth column to erodent size (E), the third and fourth column are assigned to (A× B)1 and (A× B)2 respectively to estimate interaction between impact velocity (A) and SiC content (B), the sixth and seventh column are assigned to (B× C)1 and (B× C)2 respectively to estimate interaction between the SiC content (B) and impingement angle (C), the eight and eleventh column are assigned to (A × C)1 and (A× C)2 respectively to estimate interaction between the impact velocity (A) and impingement angle (C) and the remaining columns are used to estimate experimental errors. The output to be studied is erosion rate (Er) and the tests are repeated twice corresponding to 54 tests. Furthermore, a statistical analysis of variance (ANOVA) is performed to identify the process parameters that are statistically significant. With the S/N and ANOVA analyses, the optimal combination of the process parameters can be predicted to a useful level of accuracy. Finally, a confirmation experiment is conducted to verify the optimal process parameters obtained

where, n the number of observations and y the observed experimental data.

(g/g) was used to characterize the erosion performance of the target materials.

**2.4 Taguchi experimental design** 

function as shown below (Peace, 1993). Smaller is the better characteristic:

from the parameter design.

#### **3.1 Erosive wear of the composites**

The results have been organized and discussed in two sections. Firstly, the steady state erosion characteristics of the composites are determined for selected level of optimally controlled operating variables and compared the steady state results with the simple finite element simulation results to observe the variations in erosion rate with respective to impingement angle and the next, simulations results have been analyzed under Taguchi's experimental technique.

#### **3.2 Steady state erosion rate**

#### **3.2.1 Influence of impingement angle**

Solid-particle erosion is a complex wear phenomena influenced by a number of control factors such as impact velocity, angle of impingement, erodent particle size, stand-offdistance, materials properties, erodent particles geometry and environment temperature etc. Among these, impingement angle is the one of the most important parameter and widely studied parameter in the erosion study of materials (Hutchings, 1992; Tsuda et al., 2006). The erosion rate is measured of function of impingement angle, two types of material behavior generally observed in the target material i.e. ductile and brittle nature. The ductile nature of materials is characterized by maximum erosion rate at acute angle (15-30°) and for brittle behavior of materials, the maximum erosion rate is observed at normal impingement angle (90°). But as far as polymer matrix composites are concerned the composite materials show versatile in nature depending upon the fabrication procedure and type of reinforcing material. The reinforced composites show a semi-ductile behavior having the maximum erosion rate in the range of 45-60° (Hutchings, 1992), unlike the above two categories. This classification, however, is not absolute as the erosion of material has a strong dependence on erosion conditions such as the properties of target material.

In the present study of SiC filled glass fiber-polyester composites, the erosion rate increases monotonically with the increase in impingement angle and reaches maximum at 45° impingement angle for particulate filled composites. However, for unfilled composite the maximum erosion rate is found to be at 60° impingement angle. This indicated that all the particulate filled and unfilled composites show semi-ductile erosion behaviour irrespective of filler content. Similarly, the finite element analysis simulated results are in good agreement with the experimental results as observed in Figure 2. As far as erosion resistance is concerned 20wt% SiC filled composites show better erosion resistance among other particulate filled and unfilled composites. Whereas, unfilled composites shows maximum erosion rate as compared with 10wt% and 20wt% SiC filled glass fiber reinforced polyester composites both in experimental and finite element analysis simulated results as shown in Figure 2.

Silicon Carbide Filled Polymer Composite for Erosive Environment

**3.3 Taguchi analysis and response optimization** 

SiC content (B) (%)

Expt. No.

Impact Velocity (A)(m/s)

Application: A Comparative Analysis of Experimental and FE Simulation Results 459

increase in erosion rate with increase in impact velocity can be attributed to increased penetration of particles on impact as a result of dissipation of greater amount of particle thermal energy to the target surface. This leads to more surface damage, enhanced sub-

The analysis is made using the computational software MINITAB 15. Table 2 shows the experimental design using L27 orthogonal array. The overall mean for the S/N ratio of erosion rate is found to be 61.92db for erosion rate is mentioned in the response table.

1 43 0 30 65 250 0.0003303 69.6224 2 43 0 60 75 350 0.0002466 72.1588 3 43 0 90 85 450 0.0001246 78.0908 4 43 10 30 75 350 0.0004458 67.0165 5 43 10 60 85 450 0.0002775 71.1347 6 43 10 90 65 250 0.0023721 52.4974 7 43 20 30 85 450 0.0006133 64.2461 8 43 20 60 65 250 0.0003333 69.5424 9 43 20 90 75 350 0.0006175 64.1873 10 54 0 30 75 450 0.0014625 56.6981 11 54 0 60 85 250 0.0028121 51.0194 12 54 0 90 65 350 0.0027000 51.3727 13 54 10 30 85 250 0.0000917 80.7558 14 54 10 60 65 350 0.0022625 52.9082 15 54 10 90 75 450 0.0027392 51.2476 16 54 20 30 65 350 0.0005450 65.2721 17 54 20 60 75 450 0.0001229 78.2078 18 54 20 90 85 250 0.0007804 62.1535 19 65 0 30 85 350 0.0024783 52.1171 20 65 0 60 65 450 0.0045143 46.9082 21 65 0 90 75 250 0.0031857 49.9359 22 65 10 30 65 450 0.0004354 67.2217 23 65 10 60 75 250 0.0009611 60.3442 24 65 10 90 85 350 0.0004091 67.7625 25 65 20 30 75 250 0.0002840 70.9336 26 65 20 60 85 350 0.0034362 49.2785 27 65 20 90 65 450 0.0035105 49.0927

Stand-off Distance (D)(mm)

Erodent size (E) (µm)

Erosion rate (Er) (g/g)

S/N Ratio (db)

critical crack growth etc. and consequently to the reduction in erosion resistance.

Impingement

angle (C) (Degree)

Table 2. Experimental design using L27 orthogonal array

(Impact velocity: 43 m/sec, stand-off distance: 75mm and erodent size: 450μm) Fig. 2. Inuence of impingement angle on erosion rates of composites

#### **3.2.2 Influence of impact velocity**

Similarly, the variation of erosion rate of unfilled and SiC filled composites with impact velocity is shown in Figure 3. Erosion trials are conducted at five different impact velocities.

(Impingement angle: 60°, stand-off distance: 75mm and erodent size: 450μm)

Fig. 3. Inuence of impact velocity on erosion rates of composites

It is seen, in the Figure 3 that for all the composite samples, the erosion rates gradually increases with the increase in impact velocity from 43m/sec to 65m/sec respectively. The increase in erosion rate with increase in impact velocity can be attributed to increased penetration of particles on impact as a result of dissipation of greater amount of particle thermal energy to the target surface. This leads to more surface damage, enhanced subcritical crack growth etc. and consequently to the reduction in erosion resistance.

#### **3.3 Taguchi analysis and response optimization**

458 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

 20wt% SiC (Expt.result) 20wt% SiC (FEM result) 10wt% SiC (Expt.result) 10wt% SiC (FEM result) 0 wt% SiC (Expt.result) 0 wt% SiC (FEM result)

30 45 60 75 90

**Impingement angle (Degree)**

Similarly, the variation of erosion rate of unfilled and SiC filled composites with impact velocity is shown in Figure 3. Erosion trials are conducted at five different impact velocities.

40 45 50 55 60 65

It is seen, in the Figure 3 that for all the composite samples, the erosion rates gradually increases with the increase in impact velocity from 43m/sec to 65m/sec respectively. The

**Impact velocity (m/sec)**

5.0x10-4 1.0x10-3 1.5x10-3 2.0x10-3 2.5x10-3 3.0x10-3 3.5x10-3 4.0x10-3

(Impact velocity: 43 m/sec, stand-off distance: 75mm and erodent size: 450μm) Fig. 2. Inuence of impingement angle on erosion rates of composites

> 0 wt% SiC 10wt% SiC 20wt% SiC

(Impingement angle: 60°, stand-off distance: 75mm and erodent size: 450μm) Fig. 3. Inuence of impact velocity on erosion rates of composites

**Erosion rate (g/g)**

**3.2.2 Influence of impact velocity** 

5.0x10-4

1.0x10-3

1.5x10-3

2.0x10-3

**Erosion rate (g/g)**

2.5x10-3

3.0x10-3

The analysis is made using the computational software MINITAB 15. Table 2 shows the experimental design using L27 orthogonal array. The overall mean for the S/N ratio of erosion rate is found to be 61.92db for erosion rate is mentioned in the response table.


Table 2. Experimental design using L27 orthogonal array

Silicon Carbide Filled Polymer Composite for Erosive Environment

75

70

65

60

**SN ratios**

55

50

70

65

60

**SN ratios**

55

50

Signal-to-noise: Smaller is better

Signal-to-noise: Smaller is better

Application: A Comparative Analysis of Experimental and FE Simulation Results 461

**Interaction Plot for SN ratios** Data Means

> 0 10 20

30 60 90

C

B

43 54 65

**A**

**Interaction Plot for SN ratios** Data Means

43 54 65

**A**

Fig. 5b. Interaction graph between factor A and factor C (A×C) for erosion rate

Fig. 5a. Interaction graph between factor A and factor B (A×B) for erosion rate

The Effect of control factors on erosion rate is shown in Figure 4. It is observed from response graph that the combination of factors settings are A1, B3, C1, D3 and E1 have been found to be the optimum factor level for the erosion rate is concerned on the basis of smaller-the-better characteristics. The corresponding interaction graphs are plotted in the Figures 5a-c.

Fig. 4. Effect of control factors on erosion rate

The Effect of control factors on erosion rate is shown in Figure 4. It is observed from response graph that the combination of factors settings are A1, B3, C1, D3 and E1 have been found to be the optimum factor level for the erosion rate is concerned on the basis of smaller-the-better characteristics. The corresponding interaction graphs are plotted in the

> **Main Effects Plot for SN ratios** Data Means

> > 0 10 20 30 60 90

B C

250 350 450

43 54 65

D E

A

65 75 85

Signal-to-noise: Smaller is better

Fig. 4. Effect of control factors on erosion rate

**Mean of SN ratios**

Figures 5a-c.

Fig. 5a. Interaction graph between factor A and factor B (A×B) for erosion rate

Fig. 5b. Interaction graph between factor A and factor C (A×C) for erosion rate

Silicon Carbide Filled Polymer Composite for Erosive Environment

Removal of matrix materials

Expose of short glass fibers

Application: A Comparative Analysis of Experimental and FE Simulation Results 463

(a) (b)

(c) (d)

Removal of matrix

materials

Fig. 5c. Interaction graph between factor B and factor C (B×C) for erosion rate

#### **3.4 Surface morphology**

The erosion wear mechanisms of SiC filled glass fiber reinforced polyester composites eroded surfaces are observed as per Taguchi experimental design by scanning electron microscopy (SEM). Figure 6a and 6b shows the SEM of eroded composite sample studied at lower impingement angle (see Table 2, Experiment 4). The random distribution of SiC fillers on the composite surface and removable of matrix material on the composite surface are clearly observed from the figures. Figure 6c shows the increase in erosion rate with higher impingement angle (90°) for 10wt% of SiC filled glass-polyester composites (see Table 2, Experiment 6). Figure 6d shows a hole formed after SiC particle was removed from the surface. The inside surface of the hole seemed very smooth and clear which indicated that SiC particles debonded from the matrix surface with the propagation of interfacial cracks due in part to the poor interfacial bond strength (see Table 2, Experiment 11). This is due to the increase in impact velocity from 43m/s to 54m/s and more energy to chip-off the target material.

Similarly Figure 6e and 6f show the fibers are protruded above the worn surface due to SiC particles are removed from the upper surface of the composites. However, there is a signicant difference between Figures 6e and 6f due to the change in impingement angle and change in impact velocity. Thus, after the removal of the matrix material, there could be a layer of glass fiber and SiC particulates bonded on the matrix material. This indicated the favorable effect of good interfacial bond strength on the wear performance for the composites which helped prolong the lifetime of the SiC particulates to bear erosion and protect the matrix material before it was removed away. It has been reported by few of researchers that the impact on brittle materials at an oblique angle produced radial cracks at an angle to the surface and they can contribute to only matrix material loss (Scattergood et al., 1981; Lawn, 1993). Radial cracks can also contribute to material removal when they drive through a relatively thin wall. In such a case, the material loss will occur without the formation of a lateral crack. Due to the above wear mechanism the larger erodent produce

**Interaction Plot for SN ratios** Data Means

> 30 60 90

C

72.5

70.0

67.5 65.0

62.5 60.0

**SN ratios**

**3.4 Surface morphology** 

57.5 55.0

Signal-to-noise: Smaller is better

54m/s and more energy to chip-off the target material.

0 10 20

**B**

The erosion wear mechanisms of SiC filled glass fiber reinforced polyester composites eroded surfaces are observed as per Taguchi experimental design by scanning electron microscopy (SEM). Figure 6a and 6b shows the SEM of eroded composite sample studied at lower impingement angle (see Table 2, Experiment 4). The random distribution of SiC fillers on the composite surface and removable of matrix material on the composite surface are clearly observed from the figures. Figure 6c shows the increase in erosion rate with higher impingement angle (90°) for 10wt% of SiC filled glass-polyester composites (see Table 2, Experiment 6). Figure 6d shows a hole formed after SiC particle was removed from the surface. The inside surface of the hole seemed very smooth and clear which indicated that SiC particles debonded from the matrix surface with the propagation of interfacial cracks due in part to the poor interfacial bond strength (see Table 2, Experiment 11). This is due to the increase in impact velocity from 43m/s to

Similarly Figure 6e and 6f show the fibers are protruded above the worn surface due to SiC particles are removed from the upper surface of the composites. However, there is a signicant difference between Figures 6e and 6f due to the change in impingement angle and change in impact velocity. Thus, after the removal of the matrix material, there could be a layer of glass fiber and SiC particulates bonded on the matrix material. This indicated the favorable effect of good interfacial bond strength on the wear performance for the composites which helped prolong the lifetime of the SiC particulates to bear erosion and protect the matrix material before it was removed away. It has been reported by few of researchers that the impact on brittle materials at an oblique angle produced radial cracks at an angle to the surface and they can contribute to only matrix material loss (Scattergood et al., 1981; Lawn, 1993). Radial cracks can also contribute to material removal when they drive through a relatively thin wall. In such a case, the material loss will occur without the formation of a lateral crack. Due to the above wear mechanism the larger erodent produce

Fig. 5c. Interaction graph between factor B and factor C (B×C) for erosion rate

Silicon Carbide Filled Polymer Composite for Erosive Environment

**3.5 ANOVA and the effects of factors** 

have less significant contribution on erosion rate.

Total 26 2713.18

Table 3. ANOVA table for erosion rate

Error 4 395.79 395.79 98.95

95%.

Application: A Comparative Analysis of Experimental and FE Simulation Results 465

deeper radial cracks on the eroded surfaces. The tendency for material loss to occur from radial cracking should increase with increase in erodent size (Lee et al., 2005; Milman et al., 1999). However, with the increasing content of the SiC particles in the composites from 10wt% to 20wt%, the wear rate of the composites increased gradually, reached a maximum and then declined gradually. With the further increase in impact velocity from 54m/sec to 65m/sec for 20wt% SiC filled short glass fiber reinforced polyester composites the material removal on the surface is more but the erosion resistance become more as compared with other particulate filled glass-polyester composites as shown in Figure 6g and 6h (see Table 2, Experiment 25). In the present study, the matrix material is removed from the composite surface due to continuous impact of erodent particles with sharp angles and high impact velocity, but the reinforcing glass fibers and SiC particulates are removed slowly then the

matrix material. This may be due to the inclusions of high hardness of SiC particles.

The results of the experimental trials were investigated using the ANOVA statistical analysis method. Table 3 shows the results of the ANOVA with the erosion rate (Er) for SiC filled short glass fiber reinforced polyester composites. The objective of ANOVA is to analyze the influence of impact velocity (A), SiC content (B), impingement angle (C), stand-off distance (D) and erodent size (E) on the total variance of the results. This analysis was undertaken for a level of significance of 5% that is for a level of confidence of

From Table 3, it is concluded that impact velocity (p = 0.191), impingement angle (p = 0.365), stand-off distance (p = 0.470) and SiC content (p = 0.539) have great influence on the erosion rate. The interactions of impact velocity and SiC content (p = 0.283) has most significant effect on erosion rate but the factor erodent size (p = 0.828), interaction between SiC content and impingement angle (p = 0.717) and impact velocity and impingement angle (p = 0.684)

Source DF Seq SS Adj SS Adj MS F P

A 2 510.00 510.00 255.00 2.58 0.191 B 2 143.49 143.49 71.75 0.73 0.539 C 2 258.94 258.94 129.47 1.31 0.365 D 2 181.24 181.24 90.62 0.92 0.470 E 2 39.22 39.22 19.61 0.20 0.828 A×B 4 733.23 733.23 183.31 1.85 0.283 B×C 4 213.85 213.85 53.46 0.54 0.717 A×C 4 237.42 237.42 59.35 0.60 0.684

deeper radial cracks on the eroded surfaces. The tendency for material loss to occur from radial cracking should increase with increase in erodent size (Lee et al., 2005; Milman et al., 1999). However, with the increasing content of the SiC particles in the composites from 10wt% to 20wt%, the wear rate of the composites increased gradually, reached a maximum and then declined gradually. With the further increase in impact velocity from 54m/sec to 65m/sec for 20wt% SiC filled short glass fiber reinforced polyester composites the material removal on the surface is more but the erosion resistance become more as compared with other particulate filled glass-polyester composites as shown in Figure 6g and 6h (see Table 2, Experiment 25). In the present study, the matrix material is removed from the composite surface due to continuous impact of erodent particles with sharp angles and high impact velocity, but the reinforcing glass fibers and SiC particulates are removed slowly then the matrix material. This may be due to the inclusions of high hardness of SiC particles.

#### **3.5 ANOVA and the effects of factors**

464 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

(e) (f)

(g) (h)

Fig. 6. SEM micrographs of the eroded glass fiber-polyester composites filled with SiC

The results of the experimental trials were investigated using the ANOVA statistical analysis method. Table 3 shows the results of the ANOVA with the erosion rate (Er) for SiC filled short glass fiber reinforced polyester composites. The objective of ANOVA is to analyze the influence of impact velocity (A), SiC content (B), impingement angle (C), stand-off distance (D) and erodent size (E) on the total variance of the results. This analysis was undertaken for a level of significance of 5% that is for a level of confidence of 95%.

From Table 3, it is concluded that impact velocity (p = 0.191), impingement angle (p = 0.365), stand-off distance (p = 0.470) and SiC content (p = 0.539) have great influence on the erosion rate. The interactions of impact velocity and SiC content (p = 0.283) has most significant effect on erosion rate but the factor erodent size (p = 0.828), interaction between SiC content and impingement angle (p = 0.717) and impact velocity and impingement angle (p = 0.684) have less significant contribution on erosion rate.


Table 3. ANOVA table for erosion rate

Silicon Carbide Filled Polymer Composite for Erosive Environment

**5. Acknowledgement** 

Washington DC.

**6. References** 

34.

A221.

pp. 20-27.

153-156.

17, No. 5, pp. 361-368.

UK.

pp. 303–309.

Application: A Comparative Analysis of Experimental and FE Simulation Results 467

5. In eroded samples observed in SEM shows mostly two types of wear mechanisms i.e. micro-cutting and micro-ploughing actions. As far as SiC filled glass polyester composite is concerned the matrix material is removed at faster rate from the composite surface due to continuous impact of erodent particles with sharp angles and high impact velocity, but the reinforcing glass fibers and SiC particulates are removed slowly then the matrix

The authors are grateful to the nancial supports of the research project Ref. No.

Bahadur, S. & Tabor, D. (1985). Role of fillers in friction and wear behaviour of HDPE In:

Bahadur, S.; Fu, Q.; & Gong, D. (1994). The effect of reinforcement and the synergism between CuS and carbon fiber on the wear of nylon, *Wear*, Vol. 178, pp. 123-130. Briscoe, B. J.; Pogosion, A. K.; & Tabor, D. (1974). The friction and wear of high Density

Budinski, K.G. (1997). Resistance of particle abrasion to selected plastics, *Wear*, Vol. 203–204,

Chand, N.; Naik, A.M.; & Neog i , S. (2000). Three-body abrasive s wear of short glass ber

ElTobgy, M. S.; Ng, E. & Elbestawi, M. A. (2005). Finite Element Modeling of Erosive Wear, *International Journal of Machine Tools and Manufacture*, Vol. 45, pp. 1337-1346. Hutchings, I. M. (1992). Ductile brittle transitions and wear maps for the erosion and

Kishore.; Sampathkumaran, P.; Seetharamu, S.; Vynatheya, S.; Murali, A.; Kumar, R. K.

Lawn, B. (1993). *Fracture in Brittle Solids*, 2nd ed., Cambridge University Press, Cam-bridge,

Lee, S. H.; Lee, Y. I.; Kim, Y.W.; Xie, R. J.; Mitomo, M. & Zhan, G. D. (2005). Mechanical

Milman, Y.V.; Chugunova, S. I.; Goncharova, I. V.; Chudoba, T.; Lojkowski, W. & Gooch, W.

Nordsletten, L.; Hogasen, A.; Konttinen, Y.; Santavirta, S.; Aspenberg, P.; Aasen, A. (1996).

*Polymer wear and its control*, Volume 287-268 (L.H. Lee (ed.) ACM symposium series,

polyethylene; the action of lead oxide and copper oxide fillers, *Wear*, Vol.27, pp. 19-

abrasion of brittle materials, *Journal of Physics D: Applied Physics*, Vol. 25, pp. A212-

(2000). SEM observations of the effect of velocity and load on the slide wear characteristics glass-fabric-epoxy composites with different fillers, *Wear,* Vol. 237,

Properties of Hot-Forged Silicon Carbide Ceramics, *Scripta Materialia*, Vol. 52, pp.

(1999). Temperature Dependence of Hardness in Silicon-Carbide Ceramics with Different Porosity, *International Journal of Refractory Metals and Hard Materials*, Vol.

Human monocytes stimulation by particles of hydroxyapatite, silicon carbide, and

material. This may be due to the inclusions of high hardness of SiC particles.

SR/FTP/ETA-49/08 by Department of Science and Technology, India.

polyester composite, *Wear*, Vol. 242, pp.38–46.

#### **3.6 Confirmation experiment**

To determine the optimal conditions and to compare the result with the predicted performance, it is necessary to perform a conrmation experiment. If the generated design fails to meet the predicted requirement, the process must be reiterated using a new system unit and finally the required criteria are satised. The conrmation experiment is performed by conducting a new series of test condition in combination of the significant factors and their respective interaction levels on erosion rate as reported in Table 3. The nal step is to predict and verify the improvement of the erosion resistance. The predictive value η1 using the optimal levels of the input parameters can be calculated as:

$$\overline{\mathbf{u}}\_{\overline{1}} = \overline{\mathbf{T}} + (\overline{\mathbf{A}}\_{2} - \overline{\mathbf{T}}) + (\overline{\mathbf{B}}\_{2} - \overline{\mathbf{T}}) + [(\overline{\mathbf{A}}\_{2}\overline{\mathbf{B}}\_{2} - \overline{\mathbf{T}}) - (\overline{\mathbf{A}}\_{2} - \overline{\mathbf{T}}) - (\overline{\mathbf{B}}\_{2} - \overline{\mathbf{T}})] + (\overline{\mathbf{C}}\_{3} - \overline{\mathbf{T}}) + (\overline{\mathbf{D}}\_{2} - \overline{\mathbf{T}}) + (\overline{\mathbf{E}}\_{3} - \overline{\mathbf{T}}) \tag{3}$$

<sup>η</sup>1 : Predicted average

T : Overall experimental average

A ,B ,C ,D and E 22 3 2 <sup>3</sup> Mean response for factors and interactions at designated levels. By combining like terms, the equation reduces to

$$
\overline{\mathbf{u}}\_{1} = \overline{\mathbf{A}}\_{2}\overline{\mathbf{B}}\_{2} + \overline{\mathbf{C}}\_{3} + \overline{\mathbf{D}}\_{2} + \overline{\mathbf{E}}\_{3} - 3\overline{\mathbf{T}} \tag{4}
$$

After solving the above predictive equation the erosion rate is found to be <sup>η</sup> 57.61dB <sup>1</sup> <sup>=</sup> and the experimental result is 56.33 dB. The resulting model seems to be capable of predicting wear rate to a reasonable level of accuracy. An error of 2.22% for the S/N ratio of wear rate is observed. This validates the development of the mathematical model for predicting the measures of performance based on knowledge of the input parameters.

#### **4. Conclusions**


5. In eroded samples observed in SEM shows mostly two types of wear mechanisms i.e. micro-cutting and micro-ploughing actions. As far as SiC filled glass polyester composite is concerned the matrix material is removed at faster rate from the composite surface due to continuous impact of erodent particles with sharp angles and high impact velocity, but the reinforcing glass fibers and SiC particulates are removed slowly then the matrix material. This may be due to the inclusions of high hardness of SiC particles.

#### **5. Acknowledgement**

The authors are grateful to the nancial supports of the research project Ref. No. SR/FTP/ETA-49/08 by Department of Science and Technology, India.

#### **6. References**

466 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

To determine the optimal conditions and to compare the result with the predicted performance, it is necessary to perform a conrmation experiment. If the generated design fails to meet the predicted requirement, the process must be reiterated using a new system unit and finally the required criteria are satised. The conrmation experiment is performed by conducting a new series of test condition in combination of the significant factors and their respective interaction levels on erosion rate as reported in Table 3. The nal step is to predict and verify the improvement of the erosion resistance. The predictive value η1 using

<sup>3</sup> <sup>η</sup> T (A T) (B T) [(A B T) (A T) (B T)] (C T) (D T ) ( ) 1 2 2 22 2 2 3 2 =+ − + − + − − − − − + − + − + − *E T* (3)

<sup>η</sup> A B C D E 3T 1 22 3 2 <sup>3</sup> = ++ +− (4)

A ,B ,C ,D and E 22 3 2 <sup>3</sup> Mean response for factors and interactions at designated levels.

After solving the above predictive equation the erosion rate is found to be <sup>η</sup> 57.61dB <sup>1</sup> <sup>=</sup> and the experimental result is 56.33 dB. The resulting model seems to be capable of predicting wear rate to a reasonable level of accuracy. An error of 2.22% for the S/N ratio of wear rate is observed. This validates the development of the mathematical model for predicting the measures of performance based on knowledge of the input parameters.

1. The present work successfully fabricated SiC filled short glass fiber reinforced polyester composites by simple hand-lay-up techniques and also showed that it was feasible to add SiC particles to the glass-polyester resin composites to improve their erosion resistance. 2. The particle filled composites show good tribological properties as compared with the unfilled glass fiber reinforced polyester composites. The erosion rate for 20wt% SiC filled composites shows superior erosion resistance as compared with the rest of the

3. The variation of erosion rate with impingement angles, the material loss is dictated mainly more at 60° impingement angles for unfilled composites and for filled composites the maximum erosion resistance show around 45° impingement angle both

4. On comparision with the experimental results, the FE model (ANSYS/AUTO-DYN) is much closer to the experimental results. The major advantages of simulated results are during experimental study it is very difficult to analysis the flow direction and particularly at low impingement angle most of the erodent particles are sliding on the target composite materials instead of reback of erodent particles from the composite surface. However, finite element simulated model can be easily implemented to measure the residual stress and the

depth of penetration which is difficult to determine by experimental method.

the optimal levels of the input parameters can be calculated as:

**3.6 Confirmation experiment** 

<sup>η</sup>1 : Predicted average

**4. Conclusions** 

T : Overall experimental average

filled and unfilled composites.

in experimental and finite element simulated results.

By combining like terms, the equation reduces to


**20** 

*Malaysia* 

Nor Zaihar Yahaya

*Universiti Teknologi Petronas* 

**Comparative Assessment of Si Schottky Diode** 

Semiconductors are important as they can serve as switching agents. Unlike semiconductor that work in linear mode such as in power amplifiers and linear regulators, large amount of energy is lost in the power circuit before the processed energy reaches the output. This applies for power conversion from source to load which requires high efficiency. Power will be dissipated in the form of heat once the system has low efficiency (Batarseh, 2004). A Silicon Schottky, (SiS) is a common diode used in power electronics circuits, whereas Silicon Carbide Schottky, (SiCS) is a diode that overall could perform the same operation but at a higher efficiency rate, for example in terms of

An ideal semiconductor device would perform within these criteria; possessing large breakdown voltage, low voltage drop during on-state, high switching speed and low power loss. To increase the performance of a semiconductor device, doping process will be experienced by the device, where the characteristic of the device will be altered by adding some impurity atoms to the pure semiconductor material. The material will then be recognized as extrinsic material of *n-*type and *p-*type. A predetermined number of impurity atoms will be added into the silicon or germanium based semiconductor. For silicon, the ntype is created by introducing impurity elements with five valence electrons (*pentavalent*), such as antimony, arsenic and phosphorus. The n*-*type semiconductor will have electrons as majority carriers due to one extra free electron to move within the newly formed n*-*type material. On the other hand, p*-*type material is formed by doping the silicon crystal with impurity atoms having three valence electrons such as boron, gallium and indium. A *p-*type semiconductor will have holes as majority carriers due to insufficient number of electrons to

A forward bias or "on" condition is established once the positive potential is applied to the *p-type* material and the negative potential to the *n-type* material. The application of forwardbias potential will "pressure" electrons in the *n-type* materials and holes in the *p-type* material to recombine with the ions near the boundary and reduce the width of depletion region. If an external potential of volts is applied across the *p-n* junction such that the positive terminal is connected to the *n-type* material and the negative terminal is connected to the *p-type* material, the number of uncovered positive ions in the depletion region of the n-type material will increase because there are large number of free electrons drawn to the

**1. Introduction** 

switching losses.

complete the covalent bonds.

**Family in DC-DC Converter** 

diamond: in vitro studies of new prosthesis coatings. *Biomaterials*, Vol. 17, pp.1521– 1527.


### **Comparative Assessment of Si Schottky Diode Family in DC-DC Converter**

Nor Zaihar Yahaya *Universiti Teknologi Petronas Malaysia* 

#### **1. Introduction**

468 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Patnaik, A.; Biswas, S.; Kaundal, R.; Satapathy, A. (2011). Damage Assessment of Short Glass

Routbort, J. L.; Gulden, M. E. & Marshall, E. (1981). Particle Size Distribution Effects on the Solid Particle Erosion of Brittle Materials, *Wear*, Vol. 71, pp. 363-373. Ruff, A.W.; Ives, L.K. (1975). Measurement of solid particle velocity in erosive wear, *Wear*,

Scattergood, R. O.; Routbort, J. L. & Turner, A. P. L. (1981). Velocity and Size Dependence of

Strzepa, P.; Zamirowski, E. J.; Kupperman, J. B.; Goretta, K. C. & Routbort, J. L. (1993).

Tanaka, K., (1986). Effect of various fillers on the friction and wear of PTFE-based

Tripathy, B. S.; Furey, M.J.; (1993). Tribological behaviour of unidirectional graphite-epoxy

Tsuda, K.; Kubouchi, M.; Sakai, T.; Saputra, A. H. & Mitomo, N. (2006). General Method for Predicting the Sand Erosion Rate of GFRP, *Wear*, Vol. 260, pp. 1045-1052. Wang, J.; Gu, M.; Songhao, Ge. S. (2003). The role of the influence of MoS2 on the

Woytowitz, P. J. & Richman, R. H. (1999). Modeling of Damage from Multiple Impacts by

Indentation, Erosion and Strength Degradation of Silicon-Alloyed Pyrolytic

composites, In: *Friction and Wear of Polymer composites,* Vol. 205, pp. 137-174,

tribological properties of carbon fiber reinforced Nylon 1010 composites, *Wear,* Vol.

Peace, G. S. (1993). *Taguchi methods: A hand-on approach*. Reading, MA: Addison-Wesley. Phadke, M. S. (1989). *Quality engineering using robust design*. Englewood Cliffs, NJ: Prentice-

open access publisher, ISBN 978-953-307-1100-7.

the Erosion Rates in Silicon, *Wear*, Vol. 67, pp. 227-232.

(Friedrich K editor), Elsevier, Amsterdam.

Spherical Particles, *Wear*, Vol. 233-235, pp. 120-133.

Carbon, *Journal of Materials Science*, Vol. 28, pp. 5917–5921.

and carbon-PEEK composites. *Wear*, Vol. 162–164, pp. 385–391.

1527.

Hall.

Vol. 35, pp. 195–199.

255, pp. 774-779.

diamond: in vitro studies of new prosthesis coatings. *Biomaterials*, Vol. 17, pp.1521–

Fiber Reinforced Polyester Composites: A comparative Study, *Composites*, In Tech

Semiconductors are important as they can serve as switching agents. Unlike semiconductor that work in linear mode such as in power amplifiers and linear regulators, large amount of energy is lost in the power circuit before the processed energy reaches the output. This applies for power conversion from source to load which requires high efficiency. Power will be dissipated in the form of heat once the system has low efficiency (Batarseh, 2004). A Silicon Schottky, (SiS) is a common diode used in power electronics circuits, whereas Silicon Carbide Schottky, (SiCS) is a diode that overall could perform the same operation but at a higher efficiency rate, for example in terms of switching losses.

An ideal semiconductor device would perform within these criteria; possessing large breakdown voltage, low voltage drop during on-state, high switching speed and low power loss. To increase the performance of a semiconductor device, doping process will be experienced by the device, where the characteristic of the device will be altered by adding some impurity atoms to the pure semiconductor material. The material will then be recognized as extrinsic material of *n-*type and *p-*type. A predetermined number of impurity atoms will be added into the silicon or germanium based semiconductor. For silicon, the ntype is created by introducing impurity elements with five valence electrons (*pentavalent*), such as antimony, arsenic and phosphorus. The n*-*type semiconductor will have electrons as majority carriers due to one extra free electron to move within the newly formed n*-*type material. On the other hand, p*-*type material is formed by doping the silicon crystal with impurity atoms having three valence electrons such as boron, gallium and indium. A *p-*type semiconductor will have holes as majority carriers due to insufficient number of electrons to complete the covalent bonds.

A forward bias or "on" condition is established once the positive potential is applied to the *p-type* material and the negative potential to the *n-type* material. The application of forwardbias potential will "pressure" electrons in the *n-type* materials and holes in the *p-type* material to recombine with the ions near the boundary and reduce the width of depletion region. If an external potential of volts is applied across the *p-n* junction such that the positive terminal is connected to the *n-type* material and the negative terminal is connected to the *p-type* material, the number of uncovered positive ions in the depletion region of the n-type material will increase because there are large number of free electrons drawn to the

Comparative Assessment of Si Schottky Diode Family in DC-DC Converter 471

From Fig. 1, the four Si atoms make a covalent bonding with a single Carbon, (C) atom in order to form a SiC. The C atom is located in the middle of the structure, and the distances between all atoms which marked C-Si are the same. The position of C and Si in periodic

The SiC possesses an increased tolerance to radiation damage, making it a preferred material for defense and aerospace applications. Due to high tolerance of temperature in SiC material (up to 650°C) (Silicon Carbide Electronics, 2006), it is used in various industries,

In terms of schottky's perspective, SiCS diode has higher critical field and higher barrier height than SiS diode. These two advantages give result in a reduced on-resistance and lower leakage current (Kearney et al., 1990). It has been demonstrated that the SiC material

table is shown in Fig. 2.

Fig. 2. Position of Carbon and Silicon in Periodic Table

such as aircraft, automotive, communications, power and spacecraft.

has the potential to improve power FET performance (Balaga, 1989).

Fig. 3. Energy band diagram of a semiconductor (Ozpincci & Tolbert, 2003)

The characteristic of SiC diode being a wide bandgap semiconductor results in more energy to excite the electron from its covalent bonding during turn-off compared to SiS diode. As shown in Fig. 3, the energy level is measured from the distance between the conduction band and the valence band of the semiconductor that is Eg = Ec - Ev. An insulator would have a larger bandgap that it would take a lot of energy for the electrons to move from the

positive potential of the applied voltage. The number of uncovered negative ions will also increase in the p-type material. Thus, the net effect is a widening of the depletion region and the diode is reverse-biased (Boylestad & Nashelsky, 1999).

#### **2. Silicon Schottky (SiS) diodes**

The Schottky diode or Schottky Barrier diode is a semiconductor device that is widely used as a mixer or detector diode. It is also used in power applications as a rectifier, because of its low forward voltage drop leading to lower level of power loss (Malvino, 1980).

Schottky diode is a unipolar device, where the current transport is mainly due to majority carriers. It does not rely on holes or electrons when they enter the opposite type of region as in the case of a conventional diode and therefore it gives better speed. This diode also has low turn-on voltage and high frequency capability with low capacitance (Mohammed et al., 2005).

#### **2.1 Silicon Carbide (SiC) material**

Silicon Carbide, (SiC) is a type of wide-bandgap semiconductor having advantages of fast recovery times. SiC diode is seen to have no change (or lesser) on switching loss when temperature increases, whereas SiS diode's behaviour changes in temperature (Ozpincci & Tolbert, 2003). This device has the potential to operate more efficiently by producing less heat and able to work at high temperature compared to Silicon, (Si) diodes. The cause of increasing temperature is the increase in electron's thermal energy which causes reduction of barrier height in the SiS diodes. Therefore, the power losses in SiS diode is due to the increase in its peak reverse recovery current (Chinthavali et al., 2004).

The SiC diode also comes in small size and lighter weight compared to normal schottky diode. It has a bandgap energy at most three times higher and due to this, it also gives better electrical breakdown strength about 10 times higher. This means that electronic devices in SiC can operate at a voltage of 5 to 20 times higher and with current density about 200 to 400 times higher (Wide Bandgap Semiconductor Devices, 2006).

The normal schottky diode has small forward voltage and the reverse breakdown voltage cannot be made too high (currently, approx. 100 to 200 volts). It is normally used for general rectification such as the rectification of power supplies for low voltage and high current application or for high frequency rectification with small reverse recovery time.

Fig. 1. The tetragonal bonding of a carbon atom with the four nearest silicon neighbours (IFM, 2006).

positive potential of the applied voltage. The number of uncovered negative ions will also increase in the p-type material. Thus, the net effect is a widening of the depletion region and

The Schottky diode or Schottky Barrier diode is a semiconductor device that is widely used as a mixer or detector diode. It is also used in power applications as a rectifier, because of its

Schottky diode is a unipolar device, where the current transport is mainly due to majority carriers. It does not rely on holes or electrons when they enter the opposite type of region as in the case of a conventional diode and therefore it gives better speed. This diode also has low turn-on voltage and high frequency capability with low capacitance (Mohammed

Silicon Carbide, (SiC) is a type of wide-bandgap semiconductor having advantages of fast recovery times. SiC diode is seen to have no change (or lesser) on switching loss when temperature increases, whereas SiS diode's behaviour changes in temperature (Ozpincci & Tolbert, 2003). This device has the potential to operate more efficiently by producing less heat and able to work at high temperature compared to Silicon, (Si) diodes. The cause of increasing temperature is the increase in electron's thermal energy which causes reduction of barrier height in the SiS diodes. Therefore, the power losses in SiS diode is due to the

The SiC diode also comes in small size and lighter weight compared to normal schottky diode. It has a bandgap energy at most three times higher and due to this, it also gives better electrical breakdown strength about 10 times higher. This means that electronic devices in SiC can operate at a voltage of 5 to 20 times higher and with current density about 200 to 400

The normal schottky diode has small forward voltage and the reverse breakdown voltage cannot be made too high (currently, approx. 100 to 200 volts). It is normally used for general rectification such as the rectification of power supplies for low voltage and high current

Fig. 1. The tetragonal bonding of a carbon atom with the four nearest silicon neighbours

application or for high frequency rectification with small reverse recovery time.

low forward voltage drop leading to lower level of power loss (Malvino, 1980).

increase in its peak reverse recovery current (Chinthavali et al., 2004).

times higher (Wide Bandgap Semiconductor Devices, 2006).

the diode is reverse-biased (Boylestad & Nashelsky, 1999).

**2. Silicon Schottky (SiS) diodes** 

**2.1 Silicon Carbide (SiC) material** 

et al., 2005).

(IFM, 2006).

From Fig. 1, the four Si atoms make a covalent bonding with a single Carbon, (C) atom in order to form a SiC. The C atom is located in the middle of the structure, and the distances between all atoms which marked C-Si are the same. The position of C and Si in periodic table is shown in Fig. 2.

Fig. 2. Position of Carbon and Silicon in Periodic Table

The SiC possesses an increased tolerance to radiation damage, making it a preferred material for defense and aerospace applications. Due to high tolerance of temperature in SiC material (up to 650°C) (Silicon Carbide Electronics, 2006), it is used in various industries, such as aircraft, automotive, communications, power and spacecraft.

In terms of schottky's perspective, SiCS diode has higher critical field and higher barrier height than SiS diode. These two advantages give result in a reduced on-resistance and lower leakage current (Kearney et al., 1990). It has been demonstrated that the SiC material has the potential to improve power FET performance (Balaga, 1989).

Fig. 3. Energy band diagram of a semiconductor (Ozpincci & Tolbert, 2003)

The characteristic of SiC diode being a wide bandgap semiconductor results in more energy to excite the electron from its covalent bonding during turn-off compared to SiS diode. As shown in Fig. 3, the energy level is measured from the distance between the conduction band and the valence band of the semiconductor that is Eg = Ec - Ev. An insulator would have a larger bandgap that it would take a lot of energy for the electrons to move from the

Comparative Assessment of Si Schottky Diode Family in DC-DC Converter 473

The trr is obtained by adding ta and tb. Whilst reverse recovery current is the rate of fall current multiplied with the time taken due to stored charge. The IRR is directly proportional

(b) Soft recovery (a) Abrupt recovery

IRR = √ (2 x QRR di/dt) (1)

A chopper circuit, better known as a dc-to-dc converter is used to obtain variable dc voltage from a constant voltage dc source. The SiS and SiCS diodes are characterized using this circuit. The diodes under test (D1\_SiC and D2\_Si) represent each of the diodes used in the

The values of Rg1 and Rg2 used in the simulation are 21 Ω, with temperature at 27 C and Vcc

The circuit shown in Fig. 5 is constructed by arranging the load resistor and load inductor in series whereas the diode under test is in parallel to the loads. The pulse voltage (Vpulse) is in series with the gate of MOSFET and a limiting resistor, Rg1 is placed in between the gate and

The DC source current from Vcc will provide current during turn-on of the switch (MOSFET). The turn-on and turn-off of the switch will be determined by Vpulse. *Vgg1* will provide pulse signal to the MOSFET (M1) and the signal will appear at *Vgs*. The pulse signal will then forward bias the gate-source junction of the MOSFET, using current that passes through *Rg1*, or known as *Ig* and eventually the voltage. As a result, the MOSFET is turned on. The drain current will increase slowly until the pulse signal drops to zero. The current

to di/dt. The formula for IRR is given by Eq. (1) (Power Electronic Circuits, 2006).

It can be seen that if the rate of fall current is high, the IRR will also be high.

Fig. 4. Diode reverse recovery

Major components used in the simulation are: M1 and M2: IRF520 – 9.2A/100V MOSFET

DUT (D1\_SiC): SDP06S60/INF – 6A/600V SiCS diode (D2\_Si): SB30-03F – 3A/30V SiS diode

**4. Methodology** 

Rload = 55 Ω Iload = 500 uH

simulation.

is 25 V.

Vpulse.

valence band to the conduction band whereas a conductor would have no forbidden band. The wider the bandgap of a semiconductor is, the more thermal energy is needed to excite the electrons to the valence band. Therefore a wide bandgap semiconductor could operate at higher temperature without affecting its electrical property.

#### **3. Diode characteristics**

The static property of a diode includes the I-V and reverse characteristics. The SiS diode would have a lower voltage drop than the SiCS diode. During turn-on, there is a high level injection of carrier for SiS diode that leads to a smaller amount of voltage to forward bias the diode. Due to smaller band-gap in SiS diode compared to SiCS, a higher voltage is required to forward bias the SiCS diode (Yahaya & Chew, 2004). However, SiCS can handle large reverse voltage before having an overshoot of leakage current compared to SiS diode.

The study also looks at the dynamic charateristics for both SiS and SiCS diode in terms of forward voltage drop, reverse recovery time and reverse recovery current which is given in Table 1 (Pierobon et al., 2002).


Table 1. Dynamic Characteristics Comparison

From Table 1, the dynamic characteristic of SiCS diode shows that the reverse recovery time is independent on temperature, in addition to having negligible reverse recovery current and low switching losses. On the other hand, SiS diode shows an increase in reverse recovery time, trr and reverse recovery current, IRR as temperature increases as well as high switching losses.

#### **3.1 Reverse recovery**

Reverse recovery is one measurable quantity that can be used to indicate the performance and hence the efficiency of the device. The reverse recovery in a diode occurs when a device conducts in forward bias long enough to establish steady state due to the presence of minority charge carriers. These charges must be removed prior to blocking in reverse direction (Ahmed, 1999).

Fig. 4 shows the characteristic of reverse recovery exhibited by a diode. The diode current conducts in reverse direction due to free carriers in the diode. The ta is the time which is based on the charge stored in the depletion region of the junction. It is seen that the charges are removed faster in abrupt recovery as shown in Fig. 4 (b) compared to soft recovery, Fig. 4 (a).

Fig. 4. Diode reverse recovery

The trr is obtained by adding ta and tb. Whilst reverse recovery current is the rate of fall current multiplied with the time taken due to stored charge. The IRR is directly proportional to di/dt. The formula for IRR is given by Eq. (1) (Power Electronic Circuits, 2006).

$$\mathbf{I\_{RR}} = \sqrt{\left(\mathbf{2} \times \mathbf{Q\_{RR}} \,\mathrm{di/dt}\right)}\tag{1}$$

It can be seen that if the rate of fall current is high, the IRR will also be high.

#### **4. Methodology**

472 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

valence band to the conduction band whereas a conductor would have no forbidden band. The wider the bandgap of a semiconductor is, the more thermal energy is needed to excite the electrons to the valence band. Therefore a wide bandgap semiconductor could operate at

The static property of a diode includes the I-V and reverse characteristics. The SiS diode would have a lower voltage drop than the SiCS diode. During turn-on, there is a high level injection of carrier for SiS diode that leads to a smaller amount of voltage to forward bias the diode. Due to smaller band-gap in SiS diode compared to SiCS, a higher voltage is required to forward bias the SiCS diode (Yahaya & Chew, 2004). However, SiCS can handle large reverse voltage before having an overshoot of leakage current compared to SiS diode. The study also looks at the dynamic charateristics for both SiS and SiCS diode in terms of forward voltage drop, reverse recovery time and reverse recovery current which is given in

**Characteristics SiC Schottky (SDP 04S60) Si Schottky (SB30-03F)** 

Current Negligible Temperature dependent

From Table 1, the dynamic characteristic of SiCS diode shows that the reverse recovery time is independent on temperature, in addition to having negligible reverse recovery current and low switching losses. On the other hand, SiS diode shows an increase in reverse recovery time, trr and reverse recovery current, IRR as temperature increases as well as high switching losses.

Reverse recovery is one measurable quantity that can be used to indicate the performance and hence the efficiency of the device. The reverse recovery in a diode occurs when a device conducts in forward bias long enough to establish steady state due to the presence of minority charge carriers. These charges must be removed prior to blocking in reverse

Fig. 4 shows the characteristic of reverse recovery exhibited by a diode. The diode current conducts in reverse direction due to free carriers in the diode. The ta is the time which is based on the charge stored in the depletion region of the junction. It is seen that the charges are removed faster in abrupt recovery as shown in Fig. 4 (b) compared to soft recovery, Fig. 4 (a).

change of temperature Temperature dependent

higher temperature without affecting its electrical property.

Reverse Recovery Time Does not change much in

Table 1. Dynamic Characteristics Comparison

Switching Losses Low High

(I) Rating 600V/4A 30V/3A

**3. Diode characteristics** 

Table 1 (Pierobon et al., 2002).

Reverse Recovery

Voltage (V) and Current

**3.1 Reverse recovery** 

direction (Ahmed, 1999).

A chopper circuit, better known as a dc-to-dc converter is used to obtain variable dc voltage from a constant voltage dc source. The SiS and SiCS diodes are characterized using this circuit. The diodes under test (D1\_SiC and D2\_Si) represent each of the diodes used in the simulation.

Major components used in the simulation are: M1 and M2: IRF520 – 9.2A/100V MOSFET DUT (D1\_SiC): SDP06S60/INF – 6A/600V SiCS diode (D2\_Si): SB30-03F – 3A/30V SiS diode Rload = 55 Ω Iload = 500 uH

The values of Rg1 and Rg2 used in the simulation are 21 Ω, with temperature at 27 C and Vcc is 25 V.

The circuit shown in Fig. 5 is constructed by arranging the load resistor and load inductor in series whereas the diode under test is in parallel to the loads. The pulse voltage (Vpulse) is in series with the gate of MOSFET and a limiting resistor, Rg1 is placed in between the gate and Vpulse.

The DC source current from Vcc will provide current during turn-on of the switch (MOSFET). The turn-on and turn-off of the switch will be determined by Vpulse. *Vgg1* will provide pulse signal to the MOSFET (M1) and the signal will appear at *Vgs*. The pulse signal will then forward bias the gate-source junction of the MOSFET, using current that passes through *Rg1*, or known as *Ig* and eventually the voltage. As a result, the MOSFET is turned on. The drain current will increase slowly until the pulse signal drops to zero. The current

Comparative Assessment of Si Schottky Diode Family in DC-DC Converter 475

Fig. 6 shows the Vpulse setting used in this work. The DC voltage provided by the Vpulse is set to 20 V and the same applies to V2. V1 is set to be 0 V. V1 and V2 are for maximum and minimum voltage of the pulse respectively. The rise and fall time of the pulse are both configured to be 30 ns and 20 ns. The frequency of the pulse is 40 kHz with 0.5 duty ratio. Therefore, the period

Fig. 7 shows the signal waveform from the Vpulse. The signal is the same for both circuits since the parameters used in both circuits are identical. Here, Vpulse for SiCS diode circuit is shown to represent Vpulse for both circuits. The correct signal shows square wave with pulse period of 25 µs and the maximum voltage is at 20 V whereas the minimum is 0 V. For the

In order to find the voltage across gate (g) and source (s) of the MOSFET, voltagedifferential marker is used. The marker will be placed at the gate and source according to

Fig. 8. Finding Vgs of Silicon Schottky and Silicon Carbide Schottky diode using voltage

(PER) is 25 µs and the pulse width (PW), 12.5 µs, representing the 0.5 duty ratio.

Fig. 7. Vgg1 (Vpulse) Signal

**4.1 Finding Vgs and Vds**

differential probe.

duty ratio of 0.5, half period is seen at 12.5 µs.

polarity and current flow. The illustration is shown in Fig. 8.

will stop flowing once *Ig* drops below the threshold value of the MOSFET. Since there is no current flowing through drain, the MOSFET is turned off.

In the loop containing *D1\_SiC*, *Rload* and *Iloa*d, during turn-on of MOSFET, *D1\_SiC* will be turned off due to no current flowing through *D1\_SiC*. The DC current from the DC source will flow through the resistor, *Rload* , inductor, *Iload*, drain of the MOSFET and then to the gate at the source of the MOSFET. When the current flows through *Iload*, it charges up the inductor.

*D1\_SiC* is turned on once MOSFET is turned off. This occurs when current stored in the *Iload* (inductor) starts to flow and goes through *D1\_SiC*. *D1\_SiC* will then be in forward biased until MOSFET is turned on again by *Vgg1* (Vpulse) signal. Just a few moment before *D1\_SiC* is turned off, the current will be forced to flow in reverse direction. This is when reverse recovery current appears, which is the interest of this work.

The cycle of the signal will repeat again by charging and discharging *Iload* in inductor to turn on and off of the MOSFET and *D1\_SiC*. The PSpice settings are shown in Fig. 6.

Fig. 6. Vpulse Setting

Fig. 6 shows the Vpulse setting used in this work. The DC voltage provided by the Vpulse is set to 20 V and the same applies to V2. V1 is set to be 0 V. V1 and V2 are for maximum and minimum voltage of the pulse respectively. The rise and fall time of the pulse are both configured to be 30 ns and 20 ns. The frequency of the pulse is 40 kHz with 0.5 duty ratio. Therefore, the period (PER) is 25 µs and the pulse width (PW), 12.5 µs, representing the 0.5 duty ratio.

Fig. 7. Vgg1 (Vpulse) Signal

474 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

will stop flowing once *Ig* drops below the threshold value of the MOSFET. Since there is no

In the loop containing *D1\_SiC*, *Rload* and *Iloa*d, during turn-on of MOSFET, *D1\_SiC* will be turned off due to no current flowing through *D1\_SiC*. The DC current from the DC source will flow through the resistor, *Rload* , inductor, *Iload*, drain of the MOSFET and then to the gate at the

The cycle of the signal will repeat again by charging and discharging *Iload* in inductor to turn

source of the MOSFET. When the current flows through *Iload*, it charges up the inductor. *D1\_SiC* is turned on once MOSFET is turned off. This occurs when current stored in the *Iload* (inductor) starts to flow and goes through *D1\_SiC*. *D1\_SiC* will then be in forward biased until MOSFET is turned on again by *Vgg1* (Vpulse) signal. Just a few moment before *D1\_SiC* is turned off, the current will be forced to flow in reverse direction. This is when reverse

on and off of the MOSFET and *D1\_SiC*. The PSpice settings are shown in Fig. 6.

current flowing through drain, the MOSFET is turned off.

recovery current appears, which is the interest of this work.

Fig. 5. Inductive Load Chopper Circuit

Fig. 6. Vpulse Setting

Fig. 7 shows the signal waveform from the Vpulse. The signal is the same for both circuits since the parameters used in both circuits are identical. Here, Vpulse for SiCS diode circuit is shown to represent Vpulse for both circuits. The correct signal shows square wave with pulse period of 25 µs and the maximum voltage is at 20 V whereas the minimum is 0 V. For the duty ratio of 0.5, half period is seen at 12.5 µs.

#### **4.1 Finding Vgs and Vds**

In order to find the voltage across gate (g) and source (s) of the MOSFET, voltagedifferential marker is used. The marker will be placed at the gate and source according to polarity and current flow. The illustration is shown in Fig. 8.

Fig. 8. Finding Vgs of Silicon Schottky and Silicon Carbide Schottky diode using voltage differential probe.

Comparative Assessment of Si Schottky Diode Family in DC-DC Converter 477

The PSpice software is already equipped with a function in calculating power loss. The conventional way is by using the equation P = IV, but by using PSpice, the power loss can be measured straightforward. The power loss function is within the 'add trace' function for

**4.4 Finding the effect of varying frequency to the reverse recovery loss in the diode**  The frequency of the inductive load chopper circuit used in this project is obtained from the Vpulse. Therefore, in order to vary the frequency, the period (PER) within the Vpulse setting will be adjusted according to formula *f=1/T*, where in this case T is the period (PER). It is also noted that after the period has been changed, the PW (pulse width) is changed to

The overshoot gate and drain voltages of MOSFET during turn-on in both circuits are measured based on SICS and SiS diodes as well as the load current. In addition, the respective turn-off reverse recovery overshoot current is also determined to observe the

Fig. 11 shows the voltage waveform of Vgs for SiCS and SiS diode. There is voltage overshoot seen during the turn-on of the MOSFET and in Fig. 12, the overshoot portion

SiCS

Fig. 11. Vgs of switch M1 and M2 applied at SiC Schottky Diode and Si Schottky Diode

SiS

**4.3 Finding diode turn-off power loss and MOSFET turn-on power loss** 

example W(M1) and W(M2).

configure 0.5 duty ratio setting.

**5. Results and discussion** 

**5.1 Results of Vgs and Vds** 

(circle) is enlarged.

Circuit respectively.

influence of carbide material property in the diode.

Again, similar method is applied to measure Vds. The voltage differential marker is placed at the drain and source of the MOSFET as shown in Fig. 9.

Fig. 9. Finding Vds of Silicon Schottky and Silicon Carbide Schottky diode using voltage differential probe.

The simulation is carried out one at a time starting with finding the voltage across gate and source, and then followed by finding the voltage across the drain and source. Any overshoots or ringing will be noticed and the results are measured.

#### **4.2 Finding reverse recovery current**

The next process in the simulation is to capture the reverse recovery current produced by SiCS and SiS diode. The current marker is now placed at the node terminal of the diode and then the circuit is simulated.

Fig. 10 shows the location where current marker/probe is placed on the circuit in order to measure SiCS and SiS currents.

Fig. 10. Current probe placed on the diode under test (DUT)

A diode current will be displayed and by using the 'zooming' tool, the reverse recovery currents of both diodes are measured and analyzed.

#### **4.3 Finding diode turn-off power loss and MOSFET turn-on power loss**

The PSpice software is already equipped with a function in calculating power loss. The conventional way is by using the equation P = IV, but by using PSpice, the power loss can be measured straightforward. The power loss function is within the 'add trace' function for example W(M1) and W(M2).

#### **4.4 Finding the effect of varying frequency to the reverse recovery loss in the diode**

The frequency of the inductive load chopper circuit used in this project is obtained from the Vpulse. Therefore, in order to vary the frequency, the period (PER) within the Vpulse setting will be adjusted according to formula *f=1/T*, where in this case T is the period (PER). It is also noted that after the period has been changed, the PW (pulse width) is changed to configure 0.5 duty ratio setting.

#### **5. Results and discussion**

476 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Again, similar method is applied to measure Vds. The voltage differential marker is placed at

Fig. 9. Finding Vds of Silicon Schottky and Silicon Carbide Schottky diode using voltage

overshoots or ringing will be noticed and the results are measured.

Fig. 10. Current probe placed on the diode under test (DUT)

currents of both diodes are measured and analyzed.

The simulation is carried out one at a time starting with finding the voltage across gate and source, and then followed by finding the voltage across the drain and source. Any

The next process in the simulation is to capture the reverse recovery current produced by SiCS and SiS diode. The current marker is now placed at the node terminal of the diode and

Fig. 10 shows the location where current marker/probe is placed on the circuit in order to

A diode current will be displayed and by using the 'zooming' tool, the reverse recovery

the drain and source of the MOSFET as shown in Fig. 9.

differential probe.

**4.2 Finding reverse recovery current** 

then the circuit is simulated.

measure SiCS and SiS currents.

The overshoot gate and drain voltages of MOSFET during turn-on in both circuits are measured based on SICS and SiS diodes as well as the load current. In addition, the respective turn-off reverse recovery overshoot current is also determined to observe the influence of carbide material property in the diode.

#### **5.1 Results of Vgs and Vds**

Fig. 11 shows the voltage waveform of Vgs for SiCS and SiS diode. There is voltage overshoot seen during the turn-on of the MOSFET and in Fig. 12, the overshoot portion (circle) is enlarged.

Fig. 11. Vgs of switch M1 and M2 applied at SiC Schottky Diode and Si Schottky Diode Circuit respectively.

Comparative Assessment of Si Schottky Diode Family in DC-DC Converter 479

From Fig. 14, it is noticed that the MOSFET's Vds overshoot is visible in SiCS diode circuit having value of 26.956 V. Whilst for SiS diode circuit, no overshoot is recorded at Vds**.**  However, the peak Vds value is 25.277 V for SiS better than SiCS. This is the drawback found

Fig. 15 indicates the load resistor's current, IRload for both circuits. From the figure, the maximum value of IRload in SiCS diode circuit is 230.766 mA, with a minimum of 45.078 mA. As for IRload in SiS diode circuit, the maximum is 232.897 mA and minimum of 54.207mA. The values obtained from Fig. 15 above will be used to calculate load output power of the

SiS

SiS

SiCS

Fig. 14. Vds overshoot of forward switch M1 (SiC circuit) and M2 (Si circuit)

SiS

SiCS

Fig. 15. IRload for Silicon Carbide Schottky and Silicon Schottky Circuit

SiCS

in the findings.

circuits.

As seen in Fig. 12, the overshoot voltage of MOSFET using SiS diode is higher than using SiCS diode with 6.0217 V, compared to MOSFET with SiCS of only 5.0484 V.

From this result, MOSFET turn-on power loss is also smaller in SiCS diode circuit compared to SiS circuit, due to low voltage ringing during turn-on, which will be later discussed in this chapter.

Fig. 13 shows the voltage across drain and source of the MOSFET for both circuits using SiCS and SiS diode. On the other hand, Fig. 14 shows the overshoot during turn-on as referred to the circle shown in Fig. 11.

Fig. 12. Vgs overshoot of M1 (SiCS circuit) and M2 (SiS circuit)

Fig. 13. Vds of switch M1 and M2 applied at SiC Schottky Diode and Si Schottky Diode Circuit respectively

As seen in Fig. 12, the overshoot voltage of MOSFET using SiS diode is higher than using

From this result, MOSFET turn-on power loss is also smaller in SiCS diode circuit compared to SiS circuit, due to low voltage ringing during turn-on, which will be later discussed in this

Fig. 13 shows the voltage across drain and source of the MOSFET for both circuits using SiCS and SiS diode. On the other hand, Fig. 14 shows the overshoot during turn-on as

SiCS diode with 6.0217 V, compared to MOSFET with SiCS of only 5.0484 V.

SiCS

SiS

Fig. 12. Vgs overshoot of M1 (SiCS circuit) and M2 (SiS circuit)

SiS

Fig. 13. Vds of switch M1 and M2 applied at SiC Schottky Diode and Si Schottky Diode

SiCS

chapter.

Circuit respectively

referred to the circle shown in Fig. 11.

From Fig. 14, it is noticed that the MOSFET's Vds overshoot is visible in SiCS diode circuit having value of 26.956 V. Whilst for SiS diode circuit, no overshoot is recorded at Vds**.**  However, the peak Vds value is 25.277 V for SiS better than SiCS. This is the drawback found in the findings.

Fig. 15 indicates the load resistor's current, IRload for both circuits. From the figure, the maximum value of IRload in SiCS diode circuit is 230.766 mA, with a minimum of 45.078 mA. As for IRload in SiS diode circuit, the maximum is 232.897 mA and minimum of 54.207mA. The values obtained from Fig. 15 above will be used to calculate load output power of the circuits.

Fig. 14. Vds overshoot of forward switch M1 (SiC circuit) and M2 (Si circuit)

Fig. 15. IRload for Silicon Carbide Schottky and Silicon Schottky Circuit

Comparative Assessment of Si Schottky Diode Family in DC-DC Converter 481

advantage of carbide is that the leakage current from anode to cathode is lower due to the fact that SiC structure of metal-semiconductor barrier is two times higher than Si and its smaller intrinsic carrier concentration (Scheick et al., 2004), (Libby et al., 2006). The IRR in SiCS diode is also smaller than SiS as SiC has no stored charges where a majority carrier device could operate without high-level minority carrier injection. Therefore, during the turn-off of the SiCS diode, most of the stored charges are removed (Bhatnagar & Baliga, 1993). The low switching losses of SiCS diode is due to high breakdown field of SiCS which results in reduced blocking layer thickness, in conjunction to the reduced charges (Chintivali

SiCS

SiS

Fig. 17. Diode Current, Id at Silicon Schottky and Silicon Carbide Schottky Diode

Fig. 18. Reverse Recovery Current of Silicon Schottky and Silicon Carbide Schottky Diode

SiS

SiCS

et al., 2005).

The load power for the circuits are obtained from calculation:


b. Silicon Schottky diode circuit:

```
IRload,avg = (IRload,max - IRload,min) / 2 
 = (232.297 mA – 54.207 mA) / 2 
 = 89.045 mA
```
With Rload value of 55 Ω, the output power (Pout) is obtained:

	- = (89.045 mA)2 x 55 Ω
	- = 436.096 mW

From the calculation, the output power, Pout generated by SiCS diode circuit is 474.100 mW and 436.096 mW for SiS diode circuit. The Pout of SiCS diode is higher by 8.016 %. This is because SiCS diode provides higher output current, thus higher efficiency.

Fig. 16. Source current, Is, Current across diode, Id and load current, IRload

Fig. 16 shows the flow of current to the load. This explanation is referred to current divider for diode current, Id = Is - IRload. The IRload of SiCS diode is obviously lower than SiCS due to lower IRload. Therefore, the SiS diode is proven to have larger power loss.

The carbide element in SiCS diode helps in increasing the output current and hence the output power of the circuit. This is due to the fact that SiC has lower reverse recovery current, IRR thus lower power losses at the diode during turn-off.

#### **5.2 Results of reverse recovery current**

From Fig. 17, it can be seen that there are negative overshoot during turn-off of the diode having IRR below 0A. In this simulation, the transient setting is set to be 100 µs.

Fig. 18 shows a significant difference of IRR overshoot between SiCS diode and SiS diode. It is observed that the IRR of SiS diode is -1.0245 A, whereas -91.015 mA for SiS diode. The

From the calculation, the output power, Pout generated by SiCS diode circuit is 474.100 mW and 436.096 mW for SiS diode circuit. The Pout of SiCS diode is higher by 8.016 %. This is

IRload

Fig. 16 shows the flow of current to the load. This explanation is referred to current divider for diode current, Id = Is - IRload. The IRload of SiCS diode is obviously lower than SiCS due to

The carbide element in SiCS diode helps in increasing the output current and hence the output power of the circuit. This is due to the fact that SiC has lower reverse recovery

From Fig. 17, it can be seen that there are negative overshoot during turn-off of the diode

Fig. 18 shows a significant difference of IRR overshoot between SiCS diode and SiS diode. It is observed that the IRR of SiS diode is -1.0245 A, whereas -91.015 mA for SiS diode. The

because SiCS diode provides higher output current, thus higher efficiency.

Id

Is

Fig. 16. Source current, Is, Current across diode, Id and load current, IRload

lower IRload. Therefore, the SiS diode is proven to have larger power loss.

having IRR below 0A. In this simulation, the transient setting is set to be 100 µs.

current, IRR thus lower power losses at the diode during turn-off.

**5.2 Results of reverse recovery current** 

The load power for the circuits are obtained from calculation:

With Rload value of 55 Ω, the output power (Pout) is obtained:

With Rload value of 55 Ω, the output power (Pout) is obtained:

a. Silicon Carbide Schottky diode circuit:

IRload,avg = (IRload,max - IRload,min) / 2 = (230.766 mA – 45.078 mA) / 2

b. Silicon Schottky diode circuit: IRload,avg = (IRload,max - IRload,min) / 2 = (232.297 mA – 54.207 mA) / 2

= 92.844 mA

= 89.045 mA

Pout = IRload,avg2 x RRload,load = (89.045 mA)2 x 55 Ω = 436.096 mW

Pout = IRload,avg2 x RRload,load = (92.844 mA)2 x 55 Ω = 474.100 mW

advantage of carbide is that the leakage current from anode to cathode is lower due to the fact that SiC structure of metal-semiconductor barrier is two times higher than Si and its smaller intrinsic carrier concentration (Scheick et al., 2004), (Libby et al., 2006). The IRR in SiCS diode is also smaller than SiS as SiC has no stored charges where a majority carrier device could operate without high-level minority carrier injection. Therefore, during the turn-off of the SiCS diode, most of the stored charges are removed (Bhatnagar & Baliga, 1993). The low switching losses of SiCS diode is due to high breakdown field of SiCS which results in reduced blocking layer thickness, in conjunction to the reduced charges (Chintivali et al., 2005).

Fig. 17. Diode Current, Id at Silicon Schottky and Silicon Carbide Schottky Diode

Fig. 18. Reverse Recovery Current of Silicon Schottky and Silicon Carbide Schottky Diode

Comparative Assessment of Si Schottky Diode Family in DC-DC Converter 483

Fig. 20 shows that MOSFET turn-on power loss in SiS diode circuit (20.619 W) is higher than in SiCS diode (790.777 mW). The higher power loss of MOSFET SiS diode indicates higher power loss produced by the diode during turn-off. The carbide material in SiCS diode is the main factor why such lower power loss is generated. From the results for Vgs of the MOSFET, it can be seen that lower current spike is observed in SiCS diode circuit during turn-on. With lower voltage ringing effect in SiCS diode, lower power loss will be produced by the MOSFET. It is found that, carbide material in SiCS diode has eventually given some

SiCS

SiS

influence in improving the circuit's performance.

Fig. 20. MOSFET turn-On Power Loss during DUT turn-Off

From Fig. 19, it can be seen that SiS diode has a turn-off loss of 3.0704 W larger than SiCS diode, 818.590 mW. With higher IRR, more power loss will be dissipated because more power is required for the diode to be fully turned off due to a larger stored charge.

Fig. 19. Turn Off Loss of Silicon Schottky and Silicon Carbide Schottky Diode

From Fig. 19, it can be seen that SiS diode has a turn-off loss of 3.0704 W larger than SiCS diode, 818.590 mW. With higher IRR, more power loss will be dissipated because more

SiS

SiCS

power is required for the diode to be fully turned off due to a larger stored charge.

Fig. 19. Turn Off Loss of Silicon Schottky and Silicon Carbide Schottky Diode

Fig. 20 shows that MOSFET turn-on power loss in SiS diode circuit (20.619 W) is higher than in SiCS diode (790.777 mW). The higher power loss of MOSFET SiS diode indicates higher power loss produced by the diode during turn-off. The carbide material in SiCS diode is the main factor why such lower power loss is generated. From the results for Vgs of the MOSFET, it can be seen that lower current spike is observed in SiCS diode circuit during turn-on. With lower voltage ringing effect in SiCS diode, lower power loss will be produced by the MOSFET. It is found that, carbide material in SiCS diode has eventually given some influence in improving the circuit's performance.

Fig. 20. MOSFET turn-On Power Loss during DUT turn-Off

Comparative Assessment of Si Schottky Diode Family in DC-DC Converter 485

higher reverse recovery current than silicon carbide schottky diode. Therefore, lesser power losses are generated in silicon carbide schottky diode with 91.12 % improvement. The results also confirmed that the ringing at the switch (MOSFET) has been reduced by 16.16 %. Eventually, the carbide element has helped in achieving higher output power by 8 %. The turn-off losses in diodes have also been reduced by 73.34 % using silicon carbide schottky diode as well as the MOSFET turn-on power losses which is reduced by 96.16 % mainly due

The authors wish to thank Universiti Teknologi PETRONAS for providing financial support

[1] Ahmed, A. (1999) *Power Electronics for Technology*, Purdue University-Calumet, Prentice

[2] Baliga, B. J. (1989) Power semiconductor device figure of merit for high-frequency applications, *IEEE Electron Device Letters*, Vol. 10, Iss. 10, pp. 455-457. [3] Batarseh, I. (2004), *Power Electronic Circuits*, University of Central Florida: John Wiley &

[4] Bhatnagar, M. & Baliga, B. J. (1993) Comparison of 6H-SiC, 3C-SiC, and Si for power devices, *IEEE Transactions on Electronics Devices*, Vol. 40, Iss. 3, pp. 645-655. [5] Boylestad, R. L. & Nashelsky, L. (1999) *Electronic Devices and Circuit Theory*, 7th Edition,

[6] Chintivali, M. S.; Ozpineci, B. & Tolbert, L. M. (2005) High-temperature and high-

[7] Chinthavali, M. S.; Ozpineci, B. & Tolbert, L. M. (2004) Temperature-dependent

[8] IFM, Materials Science Division Linköpings Universitet, Crystal Structure of Silicon

http://www.ifm.liu.se/matephys/AAnew/research/sicpart/kordina2.htm. [9] Kearney, M. J.; Kelly, M. J.; Condie, A. & Dale, I. (1990) Temperature Dependent Barrier

[10] Libby, R. L.; Ise, T. & Sison, L. (2006) Switching Characteristics of SiC Schottky Diodes

http://www.dilnet.upd.edu.ph/~irc/pubs/local/libby-switching.pdf. [11] Malvino, A. P. (1980) *Transistor Circuit Approximation*, 3rd Edition, McGraw-Hill, Inc. http://www.eng.uwi.tt/depts/elec/staff/rdefour/ee33d/s2\_rrchar.html [12] Mohammed, F.; Bain, M.F.; Ruddell, F.H.; Linton, D.; Gamble, H.S. & Fusco, V.F.,

[13] National Aeronautics and Space Administration, Silicon Carbide Electronics (2006)

Performance, *IEEE Electronic Letters*, Vol. 26, Iss. 10, pp. 671 – 672.

*IEEE Transactions*, Vol. 52, Iss. 7, pp. 1384 – 1391.

frequency performance evaluation of 4H-SiC unipolar power devices, *Applied Power Electronics Conference and Exposition 2005, Twentieth Annual IEEE*, Vol. 1, pp. 322-

characterization of SiC power electronic devices, *IEEE Power Electronics in* 

Heights In Bulk Unipolar Diodes Leading To Improved Temperature Stable

in a Buck DC-DC Converter, *Proc. Electronic and Communications Engineering Conf*,

(2005) A Novel Silicon Schottky Diode for NLTL Applications, *Electron Devices,* 

to the reduction in reverse recovery current.

Prentice Hall International, Inc.

*Transportation*, pp. 43-47.

Carbide (2006)

**7. Acknowledgment** 

to publish this work.

Hall.

Sons, Inc.

328.

**8. References** 


Table 2. Simulation Results

From Table 2, SiS diode has higher peak IRR of -1.0245 A compared to SiCS diode, - 91.015mA. As for turn-off loss of both diodes, it also shows that SiS diode generates more losses. This is also applied to MOSFET power loss during turn-on where there shows an improvement of 96.16 % when SiCS diode is used.

#### **5.3 The effect of varying frequency to the reverse recovery loss of the diode under test (DUT)**

From Fig. 21, it is obvious that SiCS diode circuit does not experience much difference in frequency variation. As for SiS diode, it shows an increase in power loss. However, it is also noted that once frequency is higher than 50 kHz, the power loss in SiS diode is maintained at around 3.6 W to 3.7 W. Nevertheless, SiCS diode has shown the ability in operating at higher switching frequency with minimal power loss.


Fig. 21. Graph of Power Loss vs Frequency of Silicon Schottky and Silicon Carbide Schottky Diode

#### **6. Conclusion**

This work is about the comparative study of silicon schottky and silicon carbide schottky diode using PSpice simulation. An inductive load chopper circuit is used in the simulation and the outputs in terms of reverse recovery, turn-off power losses of both diodes and turnon losses of the MOSFET are analyzed. It is proven that silicon schottky diode has produced higher reverse recovery current than silicon carbide schottky diode. Therefore, lesser power losses are generated in silicon carbide schottky diode with 91.12 % improvement. The results also confirmed that the ringing at the switch (MOSFET) has been reduced by 16.16 %. Eventually, the carbide element has helped in achieving higher output power by 8 %. The turn-off losses in diodes have also been reduced by 73.34 % using silicon carbide schottky diode as well as the MOSFET turn-on power losses which is reduced by 96.16 % mainly due to the reduction in reverse recovery current.

#### **7. Acknowledgment**

The authors wish to thank Universiti Teknologi PETRONAS for providing financial support to publish this work.

#### **8. References**

484 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Output Power, Pout 436.096mW 474.100mW 8.016%

DUT Turn-Off Loss 3.0704W 818.59mW 73.34%

Loss 20.619W 790.777mW 96.16%

From Table 2, SiS diode has higher peak IRR of -1.0245 A compared to SiCS diode, - 91.015mA. As for turn-off loss of both diodes, it also shows that SiS diode generates more losses. This is also applied to MOSFET power loss during turn-on where there shows an

**5.3 The effect of varying frequency to the reverse recovery loss of the diode under** 

From Fig. 21, it is obvious that SiCS diode circuit does not experience much difference in frequency variation. As for SiS diode, it shows an increase in power loss. However, it is also noted that once frequency is higher than 50 kHz, the power loss in SiS diode is maintained at around 3.6 W to 3.7 W. Nevertheless, SiCS diode has shown the ability in operating at

Fig. 21. Graph of Power Loss vs Frequency of Silicon Schottky and Silicon Carbide Schottky

This work is about the comparative study of silicon schottky and silicon carbide schottky diode using PSpice simulation. An inductive load chopper circuit is used in the simulation and the outputs in terms of reverse recovery, turn-off power losses of both diodes and turnon losses of the MOSFET are analyzed. It is proven that silicon schottky diode has produced

**Diode SiC Schottky Diode Percentage** 


**Improvement (%)** 

**Characteristics Si Schottky** 

improvement of 96.16 % when SiCS diode is used.

higher switching frequency with minimal power loss.

Peak Reverse Recovery Current, Irr

**test (DUT)** 

Diode

**6. Conclusion** 

MOSFET Turn-On

Table 2. Simulation Results


http://www.ifm.liu.se/matephys/AAnew/research/sicpart/kordina2.htm.


**21** 

*1Germany 2Russia* 

**Compilation on Synthesis, Characterization** 

P. Hoffmann1, N. Fainer2, M. Kosinova2, O. Baake1 and W. Ensinger1

During the last years the interest in silicon and boron carbonitrides developed remarkably. This interest is mainly based on the extraordinary properties, expected from theoretical considerations. In this time significant improvements were made in the synthesis of silicon carbonitride SiCxNy and boron carbonitride BCxNy films by both physical and chemical

In the Si–C–N and B-C-N ternary systems a set of phases is situated, namely diamond, SiC, β-Si3N4, c-BN, B4C, and β-C3N4, which have important practical applications. SiCxNy has drawn considerable interest due to its excellent new properties in comparison with the Si3N4 and SiC binary phases. The silicon carbonitride coatings are of importance because they can potentially be used in wear and corrosion protection, high-temperature oxidation resistance, as a good moisture barrier for high-temperature industrial as well as strategic applications. Their properties are low electrical conductivity, high hardness, a low friction coefficient, high photosensitivity in the UV region, and good field emission characteristics. All these characteristics have led to a rapid increase in research activities on the synthesis of SiCxNy compounds. In addition to these properties, low density and good thermal shock resistance are very important requirements for future aerospace and automobile parts applications to enhance the performance of the components. SiCxNy is also an important material in microand nano-electronics and sensor technologies due to its excellent mechanical and electrical properties. The material possesses good optical transmittance properties. This is very useful for membrane applications, where the support of such films is required (Fainer et al., 2007,

The structural similarity between the allotropic forms of carbon and boron nitride (hexagonal BN and graphite, cubic BN and diamond), and the fact that B-N pairs are isoelectronic to C-C pairs, was the basis for predictions of the existence of ternary BCxNy compounds with notable properties (Samsonov et al., 1962; Liu et al., 1989; Lambrecht & Segall, 1993; Zhang et al., 2004). This prediction has stimulated intensive research in the last 40 years towards the synthesis of ternary boron carbonitride. BCxNy compounds are interesting in both the cubic (c-BCN) and hexagonal (h-BCN) structure. On the one hand, the

2008; Mishra, 2009; Wrobel, et al., 2007, 2010; Kroke et al., 2000).

**1. Introduction** 

methods.

**and Properties of Silicon and Boron** 

*1Technische Universität Darmstadt, Materials Science 2Nikolaev Institute of Inorganic Chemistry, SB RAS* 

**Carbonitride Films** 

http://www.grc.nasa.gov/WWW/SiC/index.html.


### **Compilation on Synthesis, Characterization and Properties of Silicon and Boron Carbonitride Films**

P. Hoffmann1, N. Fainer2, M. Kosinova2, O. Baake1 and W. Ensinger1 *1Technische Universität Darmstadt, Materials Science 2Nikolaev Institute of Inorganic Chemistry, SB RAS 1Germany 2Russia* 

#### **1. Introduction**

486 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

[14] Ozpincci, B. & Tolbert, L. M. (2003) Characterization of SiC Schottky Diodes at Different Temperatures, *IEEE Power Electronics Letters*, Vol. 1, No. 2, pp. 54-57. [15] Ozpincci, B. & Tolbert, L. M. (2003) Comparison of Wide-Bandgap Semiconductos For Power Electronics Applications, Oak Ridge National Laboratory, Tennessee. [16] Pierobon, R.; Buso, S.; Citron, M.; Meneghesso, G.; Spiazzi, G. & Zanon, E. (2002)

[17] Power Electronic Circuits (2006) University of West Indies. http://www.eng.uwi.tt/depts/elec/staff/rdefour/ee33d/s1\_dvice.html. [18] Purdue University Nanoscale Center, Wide Bandgap Semiconductor Devices (2006) http://www.nanodevices.ecn.purdue.edu/widebandgap.html.

[19] Scheick, L.; Selva, L. & Becker, H. (2004) Displacement Damage-induced Catastrophic

[20] Yahaya, N. Z. & Chew, K. K. (2004) Comparative Study of The Switching Energy

Characterization of SiC Diodes for Power Applications, *IEEE Power Electronics* 

Second Breakdown in Silicon Carbide Schottky Power Diodes, *Nuclear Science IEEE* 

Losses Between Si PiN and SiC Schottky Diode, *National Power & Energy Conference*,

http://www.grc.nasa.gov/WWW/SiC/index.html.

*Specialists Conference*, Vol. 4, pp. 1673 – 1678.

*Transactions*, Vol. 51, Iss. 6, pp. 3193- 3200.

pp. 216-229.

During the last years the interest in silicon and boron carbonitrides developed remarkably. This interest is mainly based on the extraordinary properties, expected from theoretical considerations. In this time significant improvements were made in the synthesis of silicon carbonitride SiCxNy and boron carbonitride BCxNy films by both physical and chemical methods.

In the Si–C–N and B-C-N ternary systems a set of phases is situated, namely diamond, SiC, β-Si3N4, c-BN, B4C, and β-C3N4, which have important practical applications. SiCxNy has drawn considerable interest due to its excellent new properties in comparison with the Si3N4 and SiC binary phases. The silicon carbonitride coatings are of importance because they can potentially be used in wear and corrosion protection, high-temperature oxidation resistance, as a good moisture barrier for high-temperature industrial as well as strategic applications. Their properties are low electrical conductivity, high hardness, a low friction coefficient, high photosensitivity in the UV region, and good field emission characteristics. All these characteristics have led to a rapid increase in research activities on the synthesis of SiCxNy compounds. In addition to these properties, low density and good thermal shock resistance are very important requirements for future aerospace and automobile parts applications to enhance the performance of the components. SiCxNy is also an important material in microand nano-electronics and sensor technologies due to its excellent mechanical and electrical properties. The material possesses good optical transmittance properties. This is very useful for membrane applications, where the support of such films is required (Fainer et al., 2007, 2008; Mishra, 2009; Wrobel, et al., 2007, 2010; Kroke et al., 2000).

The structural similarity between the allotropic forms of carbon and boron nitride (hexagonal BN and graphite, cubic BN and diamond), and the fact that B-N pairs are isoelectronic to C-C pairs, was the basis for predictions of the existence of ternary BCxNy compounds with notable properties (Samsonov et al., 1962; Liu et al., 1989; Lambrecht & Segall, 1993; Zhang et al., 2004). This prediction has stimulated intensive research in the last 40 years towards the synthesis of ternary boron carbonitride. BCxNy compounds are interesting in both the cubic (c-BCN) and hexagonal (h-BCN) structure. On the one hand, the

Compilation on Synthesis, Characterization

**2.1.1.3 Radio frequency magnetron sputtering** 

from Si31C35N25O9 up to Si5C89N3O3.

**2.1.1.4 Reactive DC magnetron sputtering** 

light structure.

(Si36.9C30.4N32.7).

GPa.

temperatures.

and Properties of Silicon and Boron Carbonitride Films 489

SiCN films were deposited on Si(100) substrates by RF sputtering methods using SiC targets and N2 as reactant gas (Chen et al., 2009). A high substrate temperature is not favorable for the N2 incorporation into the SiCN films. The stoichiometry of these SiCN films was given as Si32.14C39.10N28.76, which is close to SiCN. The film grown at room temperature showed a

Amorphous SiC*x*N*y* films were prepared by RF magnetron reactive sputtering using sintered SiC targets and a mixture of Ar and N2 (99.999%) (Xiao et al., 2000; Li et al., 2009). The results revealed the formation of complex networks among the three elements Si, C and N, and the existence of different chemical bonds in the SiC*x*N*y* films such as C-N, C=N, C≡N, Si-C and Si-N. The stoichiometry of the as-deposited films was found to be close to SiCN

Nanostructured and amorphous SiCxNy films have been deposited by magnetron sputtering of SiC under reactive gas environment at 700-1000°C (Lin et al., 2002). Gas mixtures containing CH4 and N2 with various ratios were used for deposition. As the CH4/N2 ratio was increased, the SiCxNy films changed from mirror-like smooth films to column-like and ridge-like C-rich SiCxNy nanostructures. The chemical composition of these films varied

SiCN films have been produced by means of reactive magnetron sputtering of a Si target in an Ar/N2/C2H2 atmosphere (Hoche et al, 2008). Depending on their position in the Si– C–N phase diagram, the hardness of the films varies over a broad range, with maximum values at about 30 GPa, while Young's modulus remains in a narrow range around 200

The nano-composite SiCN thin films on silicon, glass and steel have been produced by magnetron sputtering at different substrate temperatures ranging from 100°C to 500°C at 400 W RF power from SiC targets in Ar/N2 atmosphere (Mishra et al, 2008; Mishra, 2009). The nanocomposite SiCN films were found to have nanocrystals of 2–15 nm of the β-C3N4 phase distributed in an amorphous matrix. The microhardness values of the films were found to vary between 25–47 GPa and was dependent on deposition and substrate

SiCN films were deposited on n-type Si(100) and glass substrates by RF reactive magnetron sputtering of a polycrystalline silicon target under mixed reactive gases of C2H2 and N2 (Peng et al., 2010). The SiCN films deposited at room temperature are amorphous, and the

Si–C–N films were deposited on p-type Si(100) substrates by DC magnetron co-sputtering of silicon and carbon in nitrogen–argon mixtures using a single sputter target with variable Si/C area ratios (Vlcek et al, 2002). The substrate temperature was adjusted at Ts=600°C by an ohmic heater and the RF-induced negative substrate bias voltage, Ub was 500V. With a rising Ar concentration in the gas mixture, the Si content in the films rapidly increases (from 19 to 34 at.% for a 40 at.% Si fraction in the erosion target area), while the C content decreases (from 34 to 19 at.%) at an almost constant N concentration (39–43 at.%). As a result, the N–Si and Si–N bonds dominate over the respective N–C and Si–O bonds,

C, Si and O compositions in the films are sensitive to the RF power, except N.

preferred in a pure N2 discharge, and the film hardness increases up to 40 GPa.

synthesis of c-BCN is aimed at the production of super-hard materials since properties between those of cubic boron nitride (c-BN) and diamond would be obtained (Kulisch, 2000; Solozhenko et al., 2001). On the other hand, h-BCN has potential applications in microelectronics (Kawaguchi, 1997), since it is expected to behave as semiconductor of varying band gap depending on the composition and atomic arrangement (Liu et al., 1989), or in the production of nanotubes (Yap, 2009).

#### **2. Methods of synthesis**

Considerable efforts in the synthesis of SiCxNy and BCxNy films have been made by a large variety of deposition methods (both physical and chemical techniques).

#### **2.1 Physical Vapour Deposition (PVD)**

#### **2.1.1 Silicon carbonitrides**

#### **2.1.1.1 Laser based methods**

CSi*x*N*y* thin films were grown on Si(100) substrates by pulsed laser deposition (PLD) assisted by a radio frequency (RF) nitrogen plasma source (Thärigen et al., 1999). Up to about 30 at% nitrogen and up to 20 at% silicon were found in the hard amorphous thin films (23 GPa).

SiCxNy films were grown on silicon substrates using the pulsed laser deposition (PLD) technique (Soto et. al., 1998; Boughaba et. al, 2002). A silicon carbide (SiC) target was ablated by the beam of a KrF excimer laser in a nitrogen (N2) background gas. Smooth, amorphous films were obtained for all the processing parameters. The highest values of hardness and Young´s modulus values were obtained in the low-pressure regime, in the range of 27–42 GPa and 206–305 GPa, respectively.

SiCxNy thin films have been deposited by ablation a sintered silicon carbide target in a controlled N2 atmosphere (Trusso et al., 2002). The N2 content was found to be dependent on the N2 partial pressure and did not exceed 7.5%. A slight increase of sp3 hybridized carbon bonds has been observed. The optical band gap Eg values were found to increase up to 2.4 eV starting from a value of 1.6 eV for a non-nitrogenated sample.

#### **2.1.1.2 Radio frequency reactive sputtering**

Nanocrystalline SiCxNy thin films were prepared by reactive co-sputtering of graphite and silicon on Si(111) substrates (Cao et al., 2001). The films grown with pure nitrogen gas are exclusively amorphous. Nanocrystallites of 400–490 nm in size were observed by atomic force microscopy (AFM) in films deposited with a mixture of N2+Ar.

Amorphous silicon carbide nitride thin films were synthesized on single crystal Si substrates by RF reactive sputtered silicon nitride target in a CH4 and Ar atmosphere (Peng et. al, 2001). The refractive index decreased with increasing target voltage.

SiCN films were deposited by RF reactive sputtering and annealed at 750°C in nitrogen atmosphere (Du et al., 2007). The as-deposited film did not show photoluminescence (PL), whereas strong PL peaks appeared at 358 nm, 451 nm, and 468 nm after annealing.

The a-SiCxNy thin films were deposited by reactive sputtering from SiC target and N2/Ar mixtures (Tomasella et al., 2008). For more than vol.30 % of nitrogen in the gas mixture, a N– saturated Si-C-N film was formed. All the structural variations led to an increase of the optical band gap from 1.75 to 2.35 eV.

synthesis of c-BCN is aimed at the production of super-hard materials since properties between those of cubic boron nitride (c-BN) and diamond would be obtained (Kulisch, 2000; Solozhenko et al., 2001). On the other hand, h-BCN has potential applications in microelectronics (Kawaguchi, 1997), since it is expected to behave as semiconductor of varying band gap depending on the composition and atomic arrangement (Liu et al., 1989),

Considerable efforts in the synthesis of SiCxNy and BCxNy films have been made by a large

CSi*x*N*y* thin films were grown on Si(100) substrates by pulsed laser deposition (PLD) assisted by a radio frequency (RF) nitrogen plasma source (Thärigen et al., 1999). Up to about 30 at% nitrogen and up to 20 at% silicon were found in the hard amorphous thin films

SiCxNy films were grown on silicon substrates using the pulsed laser deposition (PLD) technique (Soto et. al., 1998; Boughaba et. al, 2002). A silicon carbide (SiC) target was ablated by the beam of a KrF excimer laser in a nitrogen (N2) background gas. Smooth, amorphous films were obtained for all the processing parameters. The highest values of hardness and Young´s modulus values were obtained in the low-pressure regime, in the range of 27–42

SiCxNy thin films have been deposited by ablation a sintered silicon carbide target in a controlled N2 atmosphere (Trusso et al., 2002). The N2 content was found to be dependent on the N2 partial pressure and did not exceed 7.5%. A slight increase of sp3 hybridized carbon bonds has been observed. The optical band gap Eg values were found to increase up

Nanocrystalline SiCxNy thin films were prepared by reactive co-sputtering of graphite and silicon on Si(111) substrates (Cao et al., 2001). The films grown with pure nitrogen gas are exclusively amorphous. Nanocrystallites of 400–490 nm in size were observed by atomic

Amorphous silicon carbide nitride thin films were synthesized on single crystal Si substrates by RF reactive sputtered silicon nitride target in a CH4 and Ar atmosphere (Peng et. al,

SiCN films were deposited by RF reactive sputtering and annealed at 750°C in nitrogen atmosphere (Du et al., 2007). The as-deposited film did not show photoluminescence (PL),

The a-SiCxNy thin films were deposited by reactive sputtering from SiC target and N2/Ar mixtures (Tomasella et al., 2008). For more than vol.30 % of nitrogen in the gas mixture, a N– saturated Si-C-N film was formed. All the structural variations led to an increase of the

whereas strong PL peaks appeared at 358 nm, 451 nm, and 468 nm after annealing.

variety of deposition methods (both physical and chemical techniques).

to 2.4 eV starting from a value of 1.6 eV for a non-nitrogenated sample.

force microscopy (AFM) in films deposited with a mixture of N2+Ar.

2001). The refractive index decreased with increasing target voltage.

or in the production of nanotubes (Yap, 2009).

**2.1 Physical Vapour Deposition (PVD)** 

GPa and 206–305 GPa, respectively.

**2.1.1.2 Radio frequency reactive sputtering** 

optical band gap from 1.75 to 2.35 eV.

**2. Methods of synthesis** 

**2.1.1 Silicon carbonitrides 2.1.1.1 Laser based methods** 

(23 GPa).

SiCN films were deposited on Si(100) substrates by RF sputtering methods using SiC targets and N2 as reactant gas (Chen et al., 2009). A high substrate temperature is not favorable for the N2 incorporation into the SiCN films. The stoichiometry of these SiCN films was given as Si32.14C39.10N28.76, which is close to SiCN. The film grown at room temperature showed a light structure.

#### **2.1.1.3 Radio frequency magnetron sputtering**

Amorphous SiC*x*N*y* films were prepared by RF magnetron reactive sputtering using sintered SiC targets and a mixture of Ar and N2 (99.999%) (Xiao et al., 2000; Li et al., 2009). The results revealed the formation of complex networks among the three elements Si, C and N, and the existence of different chemical bonds in the SiC*x*N*y* films such as C-N, C=N, C≡N, Si-C and Si-N. The stoichiometry of the as-deposited films was found to be close to SiCN (Si36.9C30.4N32.7).

Nanostructured and amorphous SiCxNy films have been deposited by magnetron sputtering of SiC under reactive gas environment at 700-1000°C (Lin et al., 2002). Gas mixtures containing CH4 and N2 with various ratios were used for deposition. As the CH4/N2 ratio was increased, the SiCxNy films changed from mirror-like smooth films to column-like and ridge-like C-rich SiCxNy nanostructures. The chemical composition of these films varied from Si31C35N25O9 up to Si5C89N3O3.

SiCN films have been produced by means of reactive magnetron sputtering of a Si target in an Ar/N2/C2H2 atmosphere (Hoche et al, 2008). Depending on their position in the Si– C–N phase diagram, the hardness of the films varies over a broad range, with maximum values at about 30 GPa, while Young's modulus remains in a narrow range around 200 GPa.

The nano-composite SiCN thin films on silicon, glass and steel have been produced by magnetron sputtering at different substrate temperatures ranging from 100°C to 500°C at 400 W RF power from SiC targets in Ar/N2 atmosphere (Mishra et al, 2008; Mishra, 2009). The nanocomposite SiCN films were found to have nanocrystals of 2–15 nm of the β-C3N4 phase distributed in an amorphous matrix. The microhardness values of the films were found to vary between 25–47 GPa and was dependent on deposition and substrate temperatures.

SiCN films were deposited on n-type Si(100) and glass substrates by RF reactive magnetron sputtering of a polycrystalline silicon target under mixed reactive gases of C2H2 and N2 (Peng et al., 2010). The SiCN films deposited at room temperature are amorphous, and the C, Si and O compositions in the films are sensitive to the RF power, except N.

#### **2.1.1.4 Reactive DC magnetron sputtering**

Si–C–N films were deposited on p-type Si(100) substrates by DC magnetron co-sputtering of silicon and carbon in nitrogen–argon mixtures using a single sputter target with variable Si/C area ratios (Vlcek et al, 2002). The substrate temperature was adjusted at Ts=600°C by an ohmic heater and the RF-induced negative substrate bias voltage, Ub was 500V. With a rising Ar concentration in the gas mixture, the Si content in the films rapidly increases (from 19 to 34 at.% for a 40 at.% Si fraction in the erosion target area), while the C content decreases (from 34 to 19 at.%) at an almost constant N concentration (39–43 at.%). As a result, the N–Si and Si–N bonds dominate over the respective N–C and Si–O bonds, preferred in a pure N2 discharge, and the film hardness increases up to 40 GPa.

Compilation on Synthesis, Characterization

K, respectively.

**2.1.2.1 Laser based methods** 

and Properties of Silicon and Boron Carbonitride Films 491

B(s) + C(s) + N2(g) → BCN(s) (1)

 B(s) + C(s) + NH3(g) → BCN(s) + H2(g) (2) The BCN obtained according to reactions (1) and (2) was characterized by a somewhat larger unit cell parameter (0.6845 nm) than that of hexagonal boron nitride (0.6661 nm) or graphite (0.6708 nm). As the authors reported, the BCN powder was oxidized at 1073 K. This result indicates that this material did not contain carbon or boron carbide, because the interaction of these compounds with oxygen starts already at a temperature of 773 and 873

Using a disk combining together two semidisks, one of h-BN and one of graphite, as target, Perrone et al. deposited at room temperature polycrystalline films: a mixture of c-BCN and h-BCN by PLD in vacuum and amorphous h-BN in nitrogen gas ambient (Perrone, 1998; Dinescu, 1998). The targets used by Teodorescu et al. for film deposition were both a half C and half BN disk and a ¾ h-BN and ¼ C disk (Teodorescu et al., 1999). The influence of substrate temperature on composition and crystallinity of BCN films has been investigated. Films deposited on heated substrates are amorphous, while films produced at room temperature are polycrystalline. Wada et al. deposited BCN films from a hot-pressure BCN target consisting of graphite and h-BN powder in an 1:1 ratio (Wada et al., 2000). Later the same group (Yap et al., 2001) demonstrated that BCN films with the composition of BC2N can be obtained by RF plasma-assisted pulsed laser deposition (PLD) at 800°C on Si substrate, but these films were carbon doped BN compounds (BN:C). Furthermore, hybridized BCN films can be deposited on Ni substrate under similar synthesis conditions. Another laser-based technique was pulsed laser ablation of a sintered B4C target in the environment of a nitrogen plasma generated from ECR microwave discharge in nitrogen gas, with growing films being simultaneously bombarded by the low-energy nitrogen plasma stream (Ling, 2002; Pan, 2003). The prepared films are composed of boron, carbon, and nitrogen with an average atomic B/C/N ratio of 3:1:3.8. It was found that the assistance of the ECR nitrogen plasma facilitated nitrogen incorporation and film formation. Nitrogen ion beam generated by a Kaufman ion gun was applied to assist reactive PLD of BCN thin films from sintered B4C (Ying, 2007). It is demonstrated that with nitrogen ion beam assistance, BCN films with nitrogen content of more than 30 at.% can be synthesized. The bonding characteristics and crystalline structure of the films were also found to be influenced by the substrate temperature. With increasing substrate temperature to 600°C, the BCN films exhibit nanocrystalline nature. Recently, amorphous BCN films were

produced by laser ablation of B4C target in nitrogen atmosphere (Yang, 2010).

conditions exhibit a structure of polycrystalline BC2N (Yue et al., 2000).

Ternary boron carbonitride thin lms were prepared by RF reactive sputtering method from a hexagonal h-BN target in an Ar-CH4 atmosphere. The lms with different C contents were obtained by varying the CH4 partial pressure. The lms deposited under the optimum

BCN lms of diverse compositions have been deposited by magnetron sputtering, mainly from h-BN and graphite targets (Ulrich et al., 1998, 1999; Zhou et al., 2000; Lei et al., 2001; Yokomichi et al., 2002; Liu et al., 2005, 2006) or B4C target (Louza et al., 2000; Martinez et al., 2001; Bengy et

**2.1.2.2 Radio frequency reactive sputtering** 

**2.1.2.3 Radio frequency magnetron sputtering** 

#### **2.1.1.5 Ion Beam Sputtering Assisted Deposition (IBAD)**

SiCN films have been successfully synthesized at a temperature below 100°C from an adenine (C5N5H5)-silicon-mixed target sputtered by an Ar ion beam (Wu et al., 1999). The chemical composition of these films varied from Si24C60N13O3 up to Si32C34N 19O15. Only amorphous films for Si-rich SiCN were obtained, while the films with low Si incorporation and deposited at high Ar ion beam voltage contained nanocrystallites.

High-dose nitrogen ion implantation into SiC is a possible way to produce a-SiCxNy (Ishimaru et al., 2003; Suvorova et al., 2009). SiC crystal target was implanted by nitrogen ions at ambient temperature up to a fluence of 5×1017 N+/cm2, followed by thermal annealing at 1500°C for 30 min. a-SiCxNy possesses an intermediate bond length between Si–C and Si–N.

#### **2.1.1.6 Dual Ion Beam Sputtering (DIBS)**

SiCN films were deposited by dual ion beam sputtering (DIBS) of a SiC target in mixed Ar/N2 atmosphere at 100°C (Zhou et al., 2010). The results showed that the variations of surface roughness and hardness for the SiCN films with the assisting ion beam energy were in the range of 7–27 nm and 23–29 GPa, respectively.

#### **2.1.1.7 Combined High Power Pulse Magnetron Sputtering (HPPMS) - DC sputtering**

Amorphous SiCN coatings were synthesized by conventional DC and RF magnetron sputtering as well as with a combined sputtering process using one target in the DC mode and one target in the HPPMS mode (Hoche et al, 2010). The SiCN's Young's modulus of approximately 210 GPa makes SiCN coatings promising for the deposition onto steel. Structural differences can originate from the different carbon sources. By using acetylene a distinct amount of carbon ions can be achieved in the plasma.

#### **2.1.1.8 An arc enhanced magnetic sputtering hybrid system**

SiCN hard films have been synthesized on stainless steel substrates by an arc enhanced magnetic sputtering hybrid system using a Si target and graphite target in gases mixed of Ar and N2 (Ma et al., 2008). The microstructure of the SiCN films with a high silicon content are nanocomposites in which nano-sized crystalline C3N4 hard particles are embedded in the amorphous SiCN matrix. The hardness of the SiCN films is found to increase with increasing silicon contents, and the maximum hardness is 35 GPa. The SiCN hard films show a low friction coefficient of 0.2.

#### **2.1.1.9 Microwave Electron Cyclotron Resonance (ECR) plasma enhanced unbalance magnetron sputtering**

SiCN thin films were prepared by microwave ECR plasma enhanced unbalanced magnetron sputtering (Gao et al., 2007). The Si–C–N bonds increased from 17.14% to 23.56% while the graphite target voltage changed from 450V to 650V. The optical gap value progressively decreases from 2.65 to 1.95 eV as the carbon content changes from 19.7 at.% to 26.4 at.%. The maximum hardness of the thin films reaches 25 GPa.

#### **2.1.2 Boron carbonitrides**

The goal to synthesize boron carbonitride with the participation of the gas phase and to examine its structure and properties was put forward by Kosolapova et al. (Kosolapova et al., 1971). The product corresponding to BCN composition, as indicated by chemical analysis, was obtained by nitrogenization of a mixture of amorphous boron and carbon black in nitrogen or ammonia within the temperature range 2073 –2273K.

$$\rm{B}\_{(s)} + \rm{C}\_{(s)} + \rm{N}\_{2(g)} \rightarrow \rm{BCN}\_{(s)} \tag{1}$$

$$\text{B}\_{\text{(s)}} + \text{C}\_{\text{(s)}} + \text{NH-}\_{\text{(g)}} \rightarrow \text{BCN}\_{\text{(s)}} + \text{H}\_{\text{2(g)}} \tag{2}$$

The BCN obtained according to reactions (1) and (2) was characterized by a somewhat larger unit cell parameter (0.6845 nm) than that of hexagonal boron nitride (0.6661 nm) or graphite (0.6708 nm). As the authors reported, the BCN powder was oxidized at 1073 K. This result indicates that this material did not contain carbon or boron carbide, because the interaction of these compounds with oxygen starts already at a temperature of 773 and 873 K, respectively.

#### **2.1.2.1 Laser based methods**

490 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

SiCN films have been successfully synthesized at a temperature below 100°C from an adenine (C5N5H5)-silicon-mixed target sputtered by an Ar ion beam (Wu et al., 1999). The chemical composition of these films varied from Si24C60N13O3 up to Si32C34N 19O15. Only amorphous films for Si-rich SiCN were obtained, while the films with low Si incorporation

High-dose nitrogen ion implantation into SiC is a possible way to produce a-SiCxNy (Ishimaru et al., 2003; Suvorova et al., 2009). SiC crystal target was implanted by nitrogen ions at ambient temperature up to a fluence of 5×1017 N+/cm2, followed by thermal annealing at 1500°C for 30

SiCN films were deposited by dual ion beam sputtering (DIBS) of a SiC target in mixed Ar/N2 atmosphere at 100°C (Zhou et al., 2010). The results showed that the variations of surface roughness and hardness for the SiCN films with the assisting ion beam energy were

SiCN hard films have been synthesized on stainless steel substrates by an arc enhanced magnetic sputtering hybrid system using a Si target and graphite target in gases mixed of Ar and N2 (Ma et al., 2008). The microstructure of the SiCN films with a high silicon content are nanocomposites in which nano-sized crystalline C3N4 hard particles are embedded in the amorphous SiCN matrix. The hardness of the SiCN films is found to increase with increasing silicon contents, and the maximum hardness is 35 GPa. The SiCN hard films

**2.1.1.9 Microwave Electron Cyclotron Resonance (ECR) plasma enhanced unbalance** 

SiCN thin films were prepared by microwave ECR plasma enhanced unbalanced magnetron sputtering (Gao et al., 2007). The Si–C–N bonds increased from 17.14% to 23.56% while the graphite target voltage changed from 450V to 650V. The optical gap value progressively decreases from 2.65 to 1.95 eV as the carbon content changes from 19.7 at.% to 26.4 at.%. The

The goal to synthesize boron carbonitride with the participation of the gas phase and to examine its structure and properties was put forward by Kosolapova et al. (Kosolapova et al., 1971). The product corresponding to BCN composition, as indicated by chemical analysis, was obtained by nitrogenization of a mixture of amorphous boron and carbon

black in nitrogen or ammonia within the temperature range 2073 –2273K.

**2.1.1.7 Combined High Power Pulse Magnetron Sputtering (HPPMS) - DC sputtering**  Amorphous SiCN coatings were synthesized by conventional DC and RF magnetron sputtering as well as with a combined sputtering process using one target in the DC mode and one target in the HPPMS mode (Hoche et al, 2010). The SiCN's Young's modulus of approximately 210 GPa makes SiCN coatings promising for the deposition onto steel. Structural differences can originate from the different carbon sources. By using acetylene a

**2.1.1.5 Ion Beam Sputtering Assisted Deposition (IBAD)** 

**2.1.1.6 Dual Ion Beam Sputtering (DIBS)** 

show a low friction coefficient of 0.2.

maximum hardness of the thin films reaches 25 GPa.

**magnetron sputtering** 

**2.1.2 Boron carbonitrides** 

in the range of 7–27 nm and 23–29 GPa, respectively.

distinct amount of carbon ions can be achieved in the plasma. **2.1.1.8 An arc enhanced magnetic sputtering hybrid system** 

and deposited at high Ar ion beam voltage contained nanocrystallites.

min. a-SiCxNy possesses an intermediate bond length between Si–C and Si–N.

Using a disk combining together two semidisks, one of h-BN and one of graphite, as target, Perrone et al. deposited at room temperature polycrystalline films: a mixture of c-BCN and h-BCN by PLD in vacuum and amorphous h-BN in nitrogen gas ambient (Perrone, 1998; Dinescu, 1998). The targets used by Teodorescu et al. for film deposition were both a half C and half BN disk and a ¾ h-BN and ¼ C disk (Teodorescu et al., 1999). The influence of substrate temperature on composition and crystallinity of BCN films has been investigated. Films deposited on heated substrates are amorphous, while films produced at room temperature are polycrystalline. Wada et al. deposited BCN films from a hot-pressure BCN target consisting of graphite and h-BN powder in an 1:1 ratio (Wada et al., 2000). Later the same group (Yap et al., 2001) demonstrated that BCN films with the composition of BC2N can be obtained by RF plasma-assisted pulsed laser deposition (PLD) at 800°C on Si substrate, but these films were carbon doped BN compounds (BN:C). Furthermore, hybridized BCN films can be deposited on Ni substrate under similar synthesis conditions. Another laser-based technique was pulsed laser ablation of a sintered B4C target in the environment of a nitrogen plasma generated from ECR microwave discharge in nitrogen gas, with growing films being simultaneously bombarded by the low-energy nitrogen plasma stream (Ling, 2002; Pan, 2003). The prepared films are composed of boron, carbon, and nitrogen with an average atomic B/C/N ratio of 3:1:3.8. It was found that the assistance of the ECR nitrogen plasma facilitated nitrogen incorporation and film formation. Nitrogen ion beam generated by a Kaufman ion gun was applied to assist reactive PLD of BCN thin films from sintered B4C (Ying, 2007). It is demonstrated that with nitrogen ion beam assistance, BCN films with nitrogen content of more than 30 at.% can be synthesized. The bonding characteristics and crystalline structure of the films were also found to be influenced by the substrate temperature. With increasing substrate temperature to 600°C, the BCN films exhibit nanocrystalline nature. Recently, amorphous BCN films were produced by laser ablation of B4C target in nitrogen atmosphere (Yang, 2010).

#### **2.1.2.2 Radio frequency reactive sputtering**

Ternary boron carbonitride thin lms were prepared by RF reactive sputtering method from a hexagonal h-BN target in an Ar-CH4 atmosphere. The lms with different C contents were obtained by varying the CH4 partial pressure. The lms deposited under the optimum conditions exhibit a structure of polycrystalline BC2N (Yue et al., 2000).

#### **2.1.2.3 Radio frequency magnetron sputtering**

BCN lms of diverse compositions have been deposited by magnetron sputtering, mainly from h-BN and graphite targets (Ulrich et al., 1998, 1999; Zhou et al., 2000; Lei et al., 2001; Yokomichi et al., 2002; Liu et al., 2005, 2006) or B4C target (Louza et al., 2000; Martinez et al., 2001; Bengy et

Compilation on Synthesis, Characterization

**2.1.2.7 Ion beam implantation** 

**2.1.2.8 Electron-cyclotron-wave-resonance PACVD** 

**2.2 Chemical Vapor Deposition (CVD)** 

corresponding to the existence of Si(C4-nNn) units.

films were found to be sensitive to their N content.

**2.2.1 Silicon carbonitrides** 

**2.2.1.1 Thermal CVD** 

other applications.

2007).

BCxNy films.

and Properties of Silicon and Boron Carbonitride Films 493

targets. Ar+N2 gases were added to the deposition atmosphere under pressure of 0.1–0.3 Pa. The deposition parameters included the substrate bias, the flow rate and ratio of the reactive gases have been varied. The analytical results (FEGSEM, HRTEM and XRD, see section 4) showed that the films revealed an amorphous cauliflower-like columnar structure (Tsai,

BCN hybrid thin films were grown from ion beam plasma of borazine (B3N3H6) on highly oriented pyrolytic graphite substrate at room temperature, 600°C, and 850°C. The substrate temperature and ion fluence were shown to have significant effects on the coordination and

Nanocrystalline BCN thin lms were prepared on n-type Si(100) wafers using the electroncyclotron-wave-resonance plasma-assisted chemical vapor deposition, whereby the energy for precursor ions was adjusted between 70 and 180 eV. ECR plasma of nitrogen was asymmetrically RF biased to sputter the high-purity h-BN/graphite target (Cao, 2003).

Chemical vapor deposition (CVD) is one of the potential growing techniques of SiCxNy and

The Si-C-N deposits were obtained by CVD using the mixture of gaseous compounds such as SiCl4, NH3, H2, and C3H8 at very high temperatures from 1100 up to 1600°C (Hirai et al., 1981). The obtained amorphous deposits were mixtures of amorphous a-Si3N4, SiC and

The SiCxNy coatings were obtained by CVD at 1000–1200 °C using TMS–NH3–H2 (Bendeddouche et al., 1997). It was found that SiCxNy films are not simply a mixture of the phases SiC and Si3N4, and have a more complex relationship between the three elements,

Cubic crystalline Si1–x–yCxNy films have been grown using various carbon sources by rapidthermal CVD (Ting et al., 2002). The heat source was an ultraviolet halogen lamp with highenergy density. A mixture of carbon source, NH3, and SiH4 diluted in hydrogen was used as the source gas and introduced to the furnace. The different carbon sources are SiH3CH3, C2H4, and C3H8. The substrate's temperature was raised quickly from room temperature to 1000°C with a temperature raising rate in the range of 300–700°C/min. The Si1–x–yCxNy films grown with C3H8 gas possesses the most desirable characteristics for electronic devices and

a-SiCN:H films were successfully obtained through an in-house developed vapor-transport CVD technique in a N2 atmosphere (Awad et al, 2009). Polydimethylsilane (PDMS) was used as a precursor for both silicon and carbon, while NH3 was mixed with argon to ensure the in-situ nitrogenation of the films. The increase of the N fraction in the a-SiCN:H films resulted in an increase of the average surface roughness from 4 to 12 nm. The a-SiCN:H

pyrolytic C (up to 10 wt. %). The deposits surface had a pebble-like structure.

elemental binding states in BCN hybrid films (Uddin et al., 2005a, 2005b, 2006)

al., 2009; Nakao et al., 2010) or B and graphite targets (Byon et al., 2004; Kim et al., 2004; Zhuang et al., 2009). In most cases the films were amorphous. It has been concluded that various intermediate compounds were obtained under different experimental conditions. Ulrich et al. (Ulrich et al., 1998, 1999) still obtained BCN lms with C and BN phase separation. Liu et al. (Liu et al., 2005, 2006) also obtained the lms of atomic-level BCN compounds from h-BN and graphite targets under various experimental conditions. In addition to the synthesis of microscopic ternary BCN lms, the correlation between the chemical composition of lms and the choice of targets has also been discussed. Lousa et al. (Lousa et al., 2000) found that the atomic ratio of B/C in the lms kept almost constant as 4:1, similar to that of the target (B4C).

#### **2.1.2.4 Reactive DC magnetron sputtering**

Reactive DC magnetron sputtering technique has been investigated to grow BCxNy films. Thin films were synthesized by pulsed DC magnetron sputtering from BN + C (Martinez et al., 2002) or B4C (Johansson et al., 1996; Freire et al., 2001; Reigada et al., 2001; Chen et al., 2006) or B4C + C (Xu et al., 2006a, 2006b) targets in Ar/N2 atmosphere. Effects of target power, target pulse frequency, substrate bias and pulse frequency on surface roughness were studied. Linss et al. used a set of targets with different B/C ratios (B, B4C, BC, BC4, C) (Linss et al., 2004a, 2004b). Real ternary phases, presenting BCN bonds, were only found at low nitrogen contents; in boron-rich films. At higher nitrogen contents, the FTIR and XPS spectra were dominated by BN, CC/CN and C≡N bonds, suggesting a phase separation into BN and C/CNx phases.

#### **2.1.2.5 Ion Beam Assisted Deposition (IBAD)**

During the last 10 years ion beam assisted deposition is used for boron carbonitride film deposition. The lms were deposited by evaporating B4C or B targets to produce BCN lms. The assistance was performed with ions from the precursor gas nitrogen. IBAD has permitted to cover a wide range of compositions as a function of deposition parameters. Albella's group (Gago et al., 2000, 2001, 2002a, 2002b, 2002c) also reported that the c-BCN coatings had been synthesized successfully through evaporating B4C target and the simultaneous bombardment of the ions from the mixture gas Ar+N2+CH4. Subsequently, they paid much attention to studying the chemical composition and bonding of the BCN coatings (Caretti et al., 2003, 2004, 2007, 2010). The structure of the BCxN compounds grown by IBAD has shown to be quite sensitive to the C concentration (Caretti et al., 2010), as expected for compounds with supposedly different mechanical and electronic properties. The structure varies from a hexagonal laminar phase when x<1 to a fully amorphous compound for x≥4. For x=1, the compound consists of curved hexagonal planes in the form of a fullerene-like structure, being an intermediate structure in the process of amorphization due to C incorporation (Caretti et al., 2007, 2010).

Boron carbonitride (BCN) coatings were deposited on Si(100) wafers and Si3N4 disks by using IBAD from a boron carbide target. The BCN coatings were synthesized by the reaction between boron and carbon vapor as well as nitrogen ion simultaneously. The influence of deposition parameters such as ion acceleration voltage, ion acceleration current density and deposition ratio on the surface roughness and mechanical properties of the BCN coatings was investigated (Fei Zhou et al., 2006a, 2006b, 2006c).

#### **2.1.2.6 Cathodic arc plasma deposition**

Tsai et al. demonstrated that boron carbon nitride (BCN) thin films were deposited on Si (100) substrates by reactive cathodic arc evaporation from graphite and B4C composite targets. Ar+N2 gases were added to the deposition atmosphere under pressure of 0.1–0.3 Pa. The deposition parameters included the substrate bias, the flow rate and ratio of the reactive gases have been varied. The analytical results (FEGSEM, HRTEM and XRD, see section 4) showed that the films revealed an amorphous cauliflower-like columnar structure (Tsai, 2007).

#### **2.1.2.7 Ion beam implantation**

492 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

al., 2009; Nakao et al., 2010) or B and graphite targets (Byon et al., 2004; Kim et al., 2004; Zhuang et al., 2009). In most cases the films were amorphous. It has been concluded that various intermediate compounds were obtained under different experimental conditions. Ulrich et al. (Ulrich et al., 1998, 1999) still obtained BCN lms with C and BN phase separation. Liu et al. (Liu et al., 2005, 2006) also obtained the lms of atomic-level BCN compounds from h-BN and graphite targets under various experimental conditions. In addition to the synthesis of microscopic ternary BCN lms, the correlation between the chemical composition of lms and the choice of targets has also been discussed. Lousa et al. (Lousa et al., 2000) found that the atomic ratio of B/C in the lms kept almost constant as 4:1, similar to that of the target (B4C).

Reactive DC magnetron sputtering technique has been investigated to grow BCxNy films. Thin films were synthesized by pulsed DC magnetron sputtering from BN + C (Martinez et al., 2002) or B4C (Johansson et al., 1996; Freire et al., 2001; Reigada et al., 2001; Chen et al., 2006) or B4C + C (Xu et al., 2006a, 2006b) targets in Ar/N2 atmosphere. Effects of target power, target pulse frequency, substrate bias and pulse frequency on surface roughness were studied. Linss et al. used a set of targets with different B/C ratios (B, B4C, BC, BC4, C) (Linss et al., 2004a, 2004b). Real ternary phases, presenting BCN bonds, were only found at low nitrogen contents; in boron-rich films. At higher nitrogen contents, the FTIR and XPS spectra were dominated by BN, CC/CN and C≡N bonds, suggesting a phase separation into

During the last 10 years ion beam assisted deposition is used for boron carbonitride film deposition. The lms were deposited by evaporating B4C or B targets to produce BCN lms. The assistance was performed with ions from the precursor gas nitrogen. IBAD has permitted to cover a wide range of compositions as a function of deposition parameters. Albella's group (Gago et al., 2000, 2001, 2002a, 2002b, 2002c) also reported that the c-BCN coatings had been synthesized successfully through evaporating B4C target and the simultaneous bombardment of the ions from the mixture gas Ar+N2+CH4. Subsequently, they paid much attention to studying the chemical composition and bonding of the BCN coatings (Caretti et al., 2003, 2004, 2007, 2010). The structure of the BCxN compounds grown by IBAD has shown to be quite sensitive to the C concentration (Caretti et al., 2010), as expected for compounds with supposedly different mechanical and electronic properties. The structure varies from a hexagonal laminar phase when x<1 to a fully amorphous compound for x≥4. For x=1, the compound consists of curved hexagonal planes in the form of a fullerene-like structure, being an intermediate structure in the process of amorphization

Boron carbonitride (BCN) coatings were deposited on Si(100) wafers and Si3N4 disks by using IBAD from a boron carbide target. The BCN coatings were synthesized by the reaction between boron and carbon vapor as well as nitrogen ion simultaneously. The influence of deposition parameters such as ion acceleration voltage, ion acceleration current density and deposition ratio on the surface roughness and mechanical properties of the BCN coatings

Tsai et al. demonstrated that boron carbon nitride (BCN) thin films were deposited on Si (100) substrates by reactive cathodic arc evaporation from graphite and B4C composite

**2.1.2.4 Reactive DC magnetron sputtering** 

**2.1.2.5 Ion Beam Assisted Deposition (IBAD)** 

due to C incorporation (Caretti et al., 2007, 2010).

was investigated (Fei Zhou et al., 2006a, 2006b, 2006c).

**2.1.2.6 Cathodic arc plasma deposition** 

BN and C/CNx phases.

BCN hybrid thin films were grown from ion beam plasma of borazine (B3N3H6) on highly oriented pyrolytic graphite substrate at room temperature, 600°C, and 850°C. The substrate temperature and ion fluence were shown to have significant effects on the coordination and elemental binding states in BCN hybrid films (Uddin et al., 2005a, 2005b, 2006)

#### **2.1.2.8 Electron-cyclotron-wave-resonance PACVD**

Nanocrystalline BCN thin lms were prepared on n-type Si(100) wafers using the electroncyclotron-wave-resonance plasma-assisted chemical vapor deposition, whereby the energy for precursor ions was adjusted between 70 and 180 eV. ECR plasma of nitrogen was asymmetrically RF biased to sputter the high-purity h-BN/graphite target (Cao, 2003).

#### **2.2 Chemical Vapor Deposition (CVD)**

Chemical vapor deposition (CVD) is one of the potential growing techniques of SiCxNy and BCxNy films.

#### **2.2.1 Silicon carbonitrides**

#### **2.2.1.1 Thermal CVD**

The Si-C-N deposits were obtained by CVD using the mixture of gaseous compounds such as SiCl4, NH3, H2, and C3H8 at very high temperatures from 1100 up to 1600°C (Hirai et al., 1981). The obtained amorphous deposits were mixtures of amorphous a-Si3N4, SiC and pyrolytic C (up to 10 wt. %). The deposits surface had a pebble-like structure.

The SiCxNy coatings were obtained by CVD at 1000–1200 °C using TMS–NH3–H2 (Bendeddouche et al., 1997). It was found that SiCxNy films are not simply a mixture of the phases SiC and Si3N4, and have a more complex relationship between the three elements, corresponding to the existence of Si(C4-nNn) units.

Cubic crystalline Si1–x–yCxNy films have been grown using various carbon sources by rapidthermal CVD (Ting et al., 2002). The heat source was an ultraviolet halogen lamp with highenergy density. A mixture of carbon source, NH3, and SiH4 diluted in hydrogen was used as the source gas and introduced to the furnace. The different carbon sources are SiH3CH3, C2H4, and C3H8. The substrate's temperature was raised quickly from room temperature to 1000°C with a temperature raising rate in the range of 300–700°C/min. The Si1–x–yCxNy films grown with C3H8 gas possesses the most desirable characteristics for electronic devices and other applications.

a-SiCN:H films were successfully obtained through an in-house developed vapor-transport CVD technique in a N2 atmosphere (Awad et al, 2009). Polydimethylsilane (PDMS) was used as a precursor for both silicon and carbon, while NH3 was mixed with argon to ensure the in-situ nitrogenation of the films. The increase of the N fraction in the a-SiCN:H films resulted in an increase of the average surface roughness from 4 to 12 nm. The a-SiCN:H films were found to be sensitive to their N content.

Compilation on Synthesis, Characterization

synthesized in the heated gas mixture:

boron trichloride and methyl cyanide

temperature ranges:

chemical and phase composition of the obtained compounds.

showed that the film is a uniform material with grain size of about 10 nm.

lattice parameters a=2.5Å and c=3.4Å (Kouvetakis et al., 1989).

compounds of this ternary system remained poorly investigated.

C5B2N. The material was stable to heating up to 1973K.

materials:

and Properties of Silicon and Boron Carbonitride Films 495

compounds, as initial substances for obtaining boron carbonitride films. The first attempt of CVD production of boron carbonitride was reported by Badyan and co-authors (Badyan et al., 1972a) in which they used the CVD process with BCl3, CCl4, N2, and H2 as starting

 BCl3+ CCl4+ N2 + H2 → BCN + HCl (3) At the synthesis temperature 2223K, they obtained solid solution with the (BN)xC1-2x composition, which was confirmed by X-ray diffraction (XRD) data. The authors assumed that the obtained material is a solution with substitution at the atomic level, as a result of substitution of a pair of carbon atoms in the hexagonal graphite lattice by nitrogen and boron atoms. Experimentally determined density of the material was 2.26±0.02 g/cm3, which is close to the density of graphite (2.26 g/cm3) and h-BN (2.27 g/cm3). At a temperature above 2273 K, the obtained compound decomposed yielding boron carbide B4C, graphite and nitrogen. Unfortunately, these works contain only a few data on the

The BCN material was more thoroughly characterized for the first time by Kaner et al. (Kaner et al., 1987). In this paper, boron carbonitride with graphite-like structure was

 BCl3 + C2H2 + NH3 → BCN + HCl (4) In order to prevent the formation of h-BN, the authors recommend at first to mix BCl3 and C2H2 (they do not react at low temperature), and then add ammonia into the hot region of the reactor. Chemical composition of the products obtained at 673 and 973 K was B0.485C0.03N0.485 and B0.35C0.30N0.35, respectively. The X-ray photoelectron analysis demonstrated that this material is not a simple mixture of boron nitride and graphite. The B1s and N1s spectra indicate that boron is bound both to carbon and to nitrogen atoms, while nitrogen atoms are bound both to carbon and to boron. These compounds exhibited semiconductor properties at room temperature. Transmission electron microscopy (TEM)

Further investigations of the synthesis of boron carbonitride, involving the initial mixture of

 BCl3 + CH3CN → BC2N + HCl (5) at temperature above 1173 K resulted in obtaining the stoichiometric compound BC2N with

The synthesis of boron carbonitride by CVD from the gaseous mixture of boron trichloride, ammonia and acetylene at 973-1323 K resulted in obtaining BCN solid solution (Saugnas et al., 1992). Both amorphous and polycrystalline films were obtained; their composition was

Nevertheless, by the 90-ies the chemical and phase composition, and properties of the

The h-BN films containing small amount of carbon and hydrogen as impurities were synthesized by means of CVD. The formula ascribed to this compound was BN(C,H). The synthesis of the films was performed using different initial gas mixtures within different

#### **2.2.1.2 Hot-wire CVD method (HWCVD)**

SixNyCz:H films were produced by HWCVD, plasma assisted HWCVD (PA-HWCVD) and plasma enhanced (PECVD) using a gas mixture of SiH4, C2H4 and NH3 without hydrogen dilution (Ferreira et al., 2006). For the HWCVD process the filament temperature was kept at 1900°C while for the PECVD component an RF power of 130W was applied. HWCVD films have higher carbon incorporation. PA-HWCVD films are N rich. PECVD films contain C and N bonded preferentially in the hydroxyl groups and the main achieved bonds are those related to C–H, C–N and Si–CHx–Si.

a-SiCN:H thin films were deposited by HWCVD using SiH4, CH4, NH3 and H2 as precursors (Swain et al., 2008). Increasing the H2 flow rate in the precursor gas more carbon is introduced into the a-SiCN:H network resulting in a decrease of the silicon content in the films from 41 at.% to 28.8 at.% and sp2 carbon cluster increases when the H2 flow rate is increased from 0 to 20 sccm.

#### **2.2.1.3 Plasma Enhanced CVD (PECVD)**

SiOCH and SiNCH films were deposited using TMS, mixed with O2 or N2. (Latrasse et al, 2009*).* Plasmas of O2/TMS and N2/TMS gas mixtures can be sustained between 5 and 25 Pa.

SiCN cone arrays were synthesized on Si wafers using a microwave plasma CVD reactor with gas mixtures of CH4, SiH4, Ar, H2 and N2 as precursors (Cheng et al., 2006). The typical process temperature was 900°C. The SiCN cones have nanometer-sized tips and their roots vary from nanometers to micrometers. Field emission characteristic of SiCN cone arrays shows a low turn-on field with relatively high current density.

The amorphous SiCN films were grown on the Si(100) and fused silica substrates by microwave CVD using a mixture of SiH4, NH3, CH4 and H2 gases in various proportions (Chen et al., 2005). The stronger affinity of silicon to bond with nitrogen than to bond with carbon results in the complete absence of Si–C bonds in a-SiCN thin films.

SiCN coatings deposited on a Si substrate are produced by PECVD using methyltrichlorosilane (MTCS), N2, and H2 as starting materials (Ivashchenko et al, 2007). The coatings are nanostructured and represent β-C3N4 crystallites embedded into the amorphous a-SiCN matrix with a hardness of 25 GPa and an Young's modulus of above 200 GPa). SiCN thin films deposited by PACVD using TMS and NH3 have been investigated in order to determine their corrosion protective ability (Loir et al, 2007).

SiCN films were synthesized on Si wafer by microwave plasma CVD (MWCVD) with CH4 (99.9%), high-purity N2 (99.999%) as precursors, and additional Si column as sources (Cheng et al, 2004). When no hydrogen was introduced, the well-faceted crystals can be achieved at modest N2 flow rate. A higher temperature results in second nucleation on previous crystals, larger crystalline size, and perfect crystalline facet.

Large and well faceted hexagonal crystallites in SiCN films can grow on Si and Ti substrates under higher nitrogen gas flow in the gaseous mixture of CH4 and H2 in the normal process of diamond deposition using a microwave plasma chemical vapor deposition (MP-CVD) (Fu et al., 2001).

#### **2.2.2 Boron carbonitrides**

The processes of CVD, considered in the present review, can be divided into three groups: 1) use of boron trichloride, 2) use of boron hydride, and 3) use of complex boron-nitrogen

SixNyCz:H films were produced by HWCVD, plasma assisted HWCVD (PA-HWCVD) and plasma enhanced (PECVD) using a gas mixture of SiH4, C2H4 and NH3 without hydrogen dilution (Ferreira et al., 2006). For the HWCVD process the filament temperature was kept at 1900°C while for the PECVD component an RF power of 130W was applied. HWCVD films have higher carbon incorporation. PA-HWCVD films are N rich. PECVD films contain C and N bonded preferentially in the hydroxyl groups and the main achieved bonds are those

a-SiCN:H thin films were deposited by HWCVD using SiH4, CH4, NH3 and H2 as precursors (Swain et al., 2008). Increasing the H2 flow rate in the precursor gas more carbon is introduced into the a-SiCN:H network resulting in a decrease of the silicon content in the films from 41 at.% to 28.8 at.% and sp2 carbon cluster increases when the H2 flow rate is

SiOCH and SiNCH films were deposited using TMS, mixed with O2 or N2. (Latrasse et al, 2009*).* Plasmas of O2/TMS and N2/TMS gas mixtures can be sustained between 5 and 25

SiCN cone arrays were synthesized on Si wafers using a microwave plasma CVD reactor with gas mixtures of CH4, SiH4, Ar, H2 and N2 as precursors (Cheng et al., 2006). The typical process temperature was 900°C. The SiCN cones have nanometer-sized tips and their roots vary from nanometers to micrometers. Field emission characteristic of SiCN cone arrays

The amorphous SiCN films were grown on the Si(100) and fused silica substrates by microwave CVD using a mixture of SiH4, NH3, CH4 and H2 gases in various proportions (Chen et al., 2005). The stronger affinity of silicon to bond with nitrogen than to bond with

SiCN coatings deposited on a Si substrate are produced by PECVD using methyltrichlorosilane (MTCS), N2, and H2 as starting materials (Ivashchenko et al, 2007). The coatings are nanostructured and represent β-C3N4 crystallites embedded into the amorphous a-SiCN matrix with a hardness of 25 GPa and an Young's modulus of above 200 GPa). SiCN thin films deposited by PACVD using TMS and NH3 have been investigated in

SiCN films were synthesized on Si wafer by microwave plasma CVD (MWCVD) with CH4 (99.9%), high-purity N2 (99.999%) as precursors, and additional Si column as sources (Cheng et al, 2004). When no hydrogen was introduced, the well-faceted crystals can be achieved at modest N2 flow rate. A higher temperature results in second nucleation on previous

Large and well faceted hexagonal crystallites in SiCN films can grow on Si and Ti substrates under higher nitrogen gas flow in the gaseous mixture of CH4 and H2 in the normal process of diamond deposition using a microwave plasma chemical vapor deposition (MP-CVD) (Fu

The processes of CVD, considered in the present review, can be divided into three groups: 1) use of boron trichloride, 2) use of boron hydride, and 3) use of complex boron-nitrogen

**2.2.1.2 Hot-wire CVD method (HWCVD)** 

related to C–H, C–N and Si–CHx–Si.

**2.2.1.3 Plasma Enhanced CVD (PECVD)** 

shows a low turn-on field with relatively high current density.

carbon results in the complete absence of Si–C bonds in a-SiCN thin films.

order to determine their corrosion protective ability (Loir et al, 2007).

crystals, larger crystalline size, and perfect crystalline facet.

increased from 0 to 20 sccm.

Pa.

et al., 2001).

**2.2.2 Boron carbonitrides** 

compounds, as initial substances for obtaining boron carbonitride films. The first attempt of CVD production of boron carbonitride was reported by Badyan and co-authors (Badyan et al., 1972a) in which they used the CVD process with BCl3, CCl4, N2, and H2 as starting materials:

$$\text{BCl}\_3 + \text{CCl}\_4 + \text{N}\_2 + \text{H}\_2 \rightarrow \text{BCN} + \text{HCl} \tag{3}$$

At the synthesis temperature 2223K, they obtained solid solution with the (BN)xC1-2x composition, which was confirmed by X-ray diffraction (XRD) data. The authors assumed that the obtained material is a solution with substitution at the atomic level, as a result of substitution of a pair of carbon atoms in the hexagonal graphite lattice by nitrogen and boron atoms. Experimentally determined density of the material was 2.26±0.02 g/cm3, which is close to the density of graphite (2.26 g/cm3) and h-BN (2.27 g/cm3). At a temperature above 2273 K, the obtained compound decomposed yielding boron carbide B4C, graphite and nitrogen. Unfortunately, these works contain only a few data on the chemical and phase composition of the obtained compounds.

The BCN material was more thoroughly characterized for the first time by Kaner et al. (Kaner et al., 1987). In this paper, boron carbonitride with graphite-like structure was synthesized in the heated gas mixture:

$$\text{BCl}\_3 + \text{C}\_2\text{H}\_2 + \text{NH}\_3 \rightarrow \text{BCN} + \text{HCl} \tag{4}$$

In order to prevent the formation of h-BN, the authors recommend at first to mix BCl3 and C2H2 (they do not react at low temperature), and then add ammonia into the hot region of the reactor. Chemical composition of the products obtained at 673 and 973 K was B0.485C0.03N0.485 and B0.35C0.30N0.35, respectively. The X-ray photoelectron analysis demonstrated that this material is not a simple mixture of boron nitride and graphite. The B1s and N1s spectra indicate that boron is bound both to carbon and to nitrogen atoms, while nitrogen atoms are bound both to carbon and to boron. These compounds exhibited semiconductor properties at room temperature. Transmission electron microscopy (TEM) showed that the film is a uniform material with grain size of about 10 nm.

Further investigations of the synthesis of boron carbonitride, involving the initial mixture of boron trichloride and methyl cyanide

$$\text{BCl}\_3 + \text{CH}\_3\text{CN} \to \text{BC}\_2\text{N} + \text{HCl} \tag{5}$$

at temperature above 1173 K resulted in obtaining the stoichiometric compound BC2N with lattice parameters a=2.5Å and c=3.4Å (Kouvetakis et al., 1989).

The synthesis of boron carbonitride by CVD from the gaseous mixture of boron trichloride, ammonia and acetylene at 973-1323 K resulted in obtaining BCN solid solution (Saugnas et al., 1992). Both amorphous and polycrystalline films were obtained; their composition was C5B2N. The material was stable to heating up to 1973K.

Nevertheless, by the 90-ies the chemical and phase composition, and properties of the compounds of this ternary system remained poorly investigated.

The h-BN films containing small amount of carbon and hydrogen as impurities were synthesized by means of CVD. The formula ascribed to this compound was BN(C,H). The synthesis of the films was performed using different initial gas mixtures within different temperature ranges:

Compilation on Synthesis, Characterization

and Properties of Silicon and Boron Carbonitride Films 497

Amorphous BCN:H films were first prepared by Montasser et al. in 1984 by means of RF and microwave plasma-stimulated CVD using the initial gas mixture composed of diborane, ethane (or methane) and nitrogen (or argon) (Montasser et al., 1984, 1985, 1990). The synthesis of transparent stable films of hydrogenated boron carbonitride BxCyNz:H is described on substrates made of NaCl, Si and glass (at room temperature). Film deposition rate was 2-12 nm/min, refractive index 1.3-1.6. Correlation between micro-hardness and chemical composition of the film was established; in turn, it depends on synthesis conditions: total pressure in the reactor, concentration ratio of the initial compounds B2H6:CH4 (or C2H6), and plasma discharge power. The BxCyNz:H films exhibited very complicated IR spectra; the authors have specially stressed that it is impossible to make conclusions concerning types of

Amorphous BCxNy:H films were prepared in a capacitively coupled RF-PECVD reactor at deposition temperatures <200oC starting from B2H6+CH4+N2+H2 gas mixture (Dekempeneer et al., 1996, 1997). Films were deposited on Si, steel and glass substrates. By varying the partial pressure of the gases, the composition was varied in a wide area of the B-N-C triangle. The same initial gas mixture was used by Polo et al. (Polo et al., 1998, 1999). It

Both amorphous and polycrystalline coatings were deposited by microwave low pressure CVD (LPCVD) using a mixture of B2H6+CH4+NH3+H2 at substrate temperature in the range 800-1350°C. (Stanishevsky, 2010). Amorphous coatings were usually formed at lower substrate temperatures and were non-homogeneous across the coating thickness. Polycrystalline coatings were generally represented by both diamond and boron nitride phases. In one case, a polycrystalline coating with the composition of B2CN4 was fabricated. The turbostratic structure of BCxNy with various compositions was synthesized by biasassisted hot-filament CVD (HFCVD) (Yu et al., 1999a, 1999b; Wang, 1999) within the temperature range 873-1273 K from B2H6+CH4+N2+H2 mixture. Investigation of the films by means of XPS demonstrated that the three atoms B, C, N are chemically bound. Boron carbonitride is the main phase in all the deposited samples, though in some cases (at high temperature) this phase was co-deposited with boron carbide. The growth rate of BCN films decreased substantially with increased temperature. Chemical composition and morphology of the layers were also dependent on deposition temperature. The turbostratic BCxNy films

were also grown by HFCVD from mixture of B2H6+CH4+NH3+H2 (Xie et al., 1998).

Laser-assisted CVD was used for preparation of single-phase BCxNy layers at low temperature in a gas atmosphere containing B2H6+CH4+NH3, where the starting composition ratio could be varied in a large range. Layers exhibited turbostratic structure. Some planar structure, containing especially CB2N groups, were suggested for the "unit

BCN films were deposited from mixture B2H6+CH4+N2+H2 with electron beam excited plasma-chemical vapor deposition (EBEP-CVD) (Hasegawa et al., 2002, 2003). By controlling the flow rate ratios of the process gases, films with composition expressed as

The review highlights of the synthesis, processing and properties of non-oxide silicon-based bulk ceramics materials derived from silazanes and polysilazanes (Kroke et al., 2000).

chemical bonds in the material basing only on the IR spectroscopic data.

was found that the films had a less ordered structure.

cell" of CBN solid solutions (Morjan et al., 1999).

BxCyN, where x=0.9-4.7 and y=0.5-6.0 were obtained.

**3.1 Silicon carbonitrides** 

**3. Syntheses of layers by single-source precursors** 

$$\begin{aligned} \text{At } 873-1273 \text{K (mixed substate) (Kawaguchiet al., 1991):}\\ \text{BCl}\_3 + \text{NH}\_3 + \text{C}\_2\text{H}\_2 \rightarrow \text{BN(C, H) + 3HCl} \end{aligned} \tag{6}$$

$$\begin{aligned} \text{1 at } 1473 \text{ - 2273K (graphite substrate) (Kawaguchiet al., 1991);}\\ \text{BCl}\_3 + \text{NH}\_3 + \text{C}\_2\text{H}\_4 \rightarrow \text{BN(C, H) + 3HCl} \end{aligned} \tag{7}$$

$$\begin{aligned} \text{At } 1473-2273 \text{K (graphite substrate) (Yokoshima et al., 1990);}\\ \text{BCl}\_3 + \text{NH}\_3 + \text{CH}\_4 \rightarrow \text{BN(C, H) + 3HCl} \end{aligned} \tag{8}$$

In all these cases, BCl3 and hydrocarbons (C2H2, C2H4 or CH4) were mixed beforehand to avoid the formation of boron nitride; ammonia was admitted directly into the reaction region near the substrate. The X-ray diffraction patterns of the BN(C,H) film synthesized according to reaction (6) were recorded by means of powder diffraction (Kawaguchi et al., 1991); the patterns contain a very broad (001) reflex and several reflexes the positions of which are close to the positions of peaks in t-BN. Additionally, the films synthesized according to reactions (7) and (8) exhibited diffraction patterns with only one diffraction reflex (001), the position of which is close to the positions of reflexes in h-BN, t-BN or graphite. The (100) and (101) reflexes are very weak and broadened. This result indicates that the BN(C,H) films obtained by means of CVD at high temperature possess the structure similar to that of t-BN. A similar ternary compound BC0,43N0,29 with turbostratic structure was synthesized on graphite at Т=1650K from a mixture of boron trichloride, methane, ammonia, and hydrogen at reduced pressure (Bessmann et al., 1990).

Amorphous boron–carbon–nitrogen (a-BCN) lms have been fabricated by hot-wire CVD using BCl3, C2H2 and N2 or NH3 (Yokomichi et al., 2004).

The N concentration of the lms synthesized by using a N2 was below several at.%, and Cl atoms were incorporated to 3–5 at.%. The N concentration increased and the Cl concentration decreased by using NH3 gas. In the case of NH3 gas, the N concentration was nearly equal to the B concentration in most cases. The nearly identical concentrations in N and B resulted from high chemical reactivity between the BCl3 and NH3 gases, and the decrease in Cl concentration resulted from the removal as HCl due to NH3 gas. These results indicate that the combination of BCl3 and NH3 is suitable for fabrication of a-BCN lms by the CVD method.

BCN lms were deposited by PECVD from a mixture of BCl3+C2H4+N2+H2+Ar in an industrial-scale DC plasma CVD plant (Kurapov et al., 2003, 2005). It was shown that the power density at the substrate has a large effect on the structure evolution of the BCN thin lms. The authors suggest that with increasing power density the structure of the deposited lms changed from an orientation where the c-axis is parallel to the substrate surface to a more randomly oriented structure.

During the last 10 years a group from Osaka University, Japan, studies intensively the PECVD synthesis from BCl3+CH4+N2+H2 mixtures and the properties of BCxNy films. The BCxNy films produced at 650°C were polycrystalline (Aoki et al., 2007, 2008a, 2008b, 2009a, 2009b, 2009c, 2009d; Etou, et al., 2002; Kimura, et al., 2005, 2009; Mazumder et al., 2009; Nesládek et al., 2001; Okada et al., 2006; Shimada et al., 2006; Sugino & Hieda, 2000; Sugino et al., 2000, 2001, 2002, 2008, 2010; Sugiyama et al., 2003, 2002; Tai et al., 2003; Umeda et al., 2004;Watanabe et al., 2008; Yuki et al., 2004; Zhang et al., 2005).

at 873 – 1273K nickel substrate Kawaguchi et al., 1991 ;

at 1473 – 2273K graphite substrate Kawaguchi et al., 1991 ;

at 1473 – 2273K graphite substrate Yokoshima et al., 199 ;

In all these cases, BCl3 and hydrocarbons (C2H2, C2H4 or CH4) were mixed beforehand to avoid the formation of boron nitride; ammonia was admitted directly into the reaction region near the substrate. The X-ray diffraction patterns of the BN(C,H) film synthesized according to reaction (6) were recorded by means of powder diffraction (Kawaguchi et al., 1991); the patterns contain a very broad (001) reflex and several reflexes the positions of which are close to the positions of peaks in t-BN. Additionally, the films synthesized according to reactions (7) and (8) exhibited diffraction patterns with only one diffraction reflex (001), the position of which is close to the positions of reflexes in h-BN, t-BN or graphite. The (100) and (101) reflexes are very weak and broadened. This result indicates that the BN(C,H) films obtained by means of CVD at high temperature possess the structure similar to that of t-BN. A similar ternary compound BC0,43N0,29 with turbostratic structure was synthesized on graphite at Т=1650K from a mixture of boron trichloride, methane,

Amorphous boron–carbon–nitrogen (a-BCN) lms have been fabricated by hot-wire CVD

The N concentration of the lms synthesized by using a N2 was below several at.%, and Cl atoms were incorporated to 3–5 at.%. The N concentration increased and the Cl concentration decreased by using NH3 gas. In the case of NH3 gas, the N concentration was nearly equal to the B concentration in most cases. The nearly identical concentrations in N and B resulted from high chemical reactivity between the BCl3 and NH3 gases, and the decrease in Cl concentration resulted from the removal as HCl due to NH3 gas. These results indicate that the combination of BCl3 and NH3 is suitable for fabrication of a-BCN lms by

BCN lms were deposited by PECVD from a mixture of BCl3+C2H4+N2+H2+Ar in an industrial-scale DC plasma CVD plant (Kurapov et al., 2003, 2005). It was shown that the power density at the substrate has a large effect on the structure evolution of the BCN thin lms. The authors suggest that with increasing power density the structure of the deposited lms changed from an orientation where the c-axis is parallel to the substrate surface to a

During the last 10 years a group from Osaka University, Japan, studies intensively the PECVD synthesis from BCl3+CH4+N2+H2 mixtures and the properties of BCxNy films. The BCxNy films produced at 650°C were polycrystalline (Aoki et al., 2007, 2008a, 2008b, 2009a, 2009b, 2009c, 2009d; Etou, et al., 2002; Kimura, et al., 2005, 2009; Mazumder et al., 2009; Nesládek et al., 2001; Okada et al., 2006; Shimada et al., 2006; Sugino & Hieda, 2000; Sugino et al., 2000, 2001, 2002, 2008, 2010; Sugiyama et al., 2003, 2002; Tai et al., 2003; Umeda et al.,

( )

3 3 22 ( )

3 3 24 ( )

3 34

BCl NH C H BN C, H 3HCl ++ → +

BCl NH CH BN C, H 3HCl ++→ +

ammonia, and hydrogen at reduced pressure (Bessmann et al., 1990).

using BCl3, C2H2 and N2 or NH3 (Yokomichi et al., 2004).

2004;Watanabe et al., 2008; Yuki et al., 2004; Zhang et al., 2005).

the CVD method.

more randomly oriented structure.

( ) ( )

BCl NH C H BN C, H 3HCl ++ → + (6)

( ) ( )

( ) ( )

(7)

(8)

0

Amorphous BCN:H films were first prepared by Montasser et al. in 1984 by means of RF and microwave plasma-stimulated CVD using the initial gas mixture composed of diborane, ethane (or methane) and nitrogen (or argon) (Montasser et al., 1984, 1985, 1990). The synthesis of transparent stable films of hydrogenated boron carbonitride BxCyNz:H is described on substrates made of NaCl, Si and glass (at room temperature). Film deposition rate was 2-12 nm/min, refractive index 1.3-1.6. Correlation between micro-hardness and chemical composition of the film was established; in turn, it depends on synthesis conditions: total pressure in the reactor, concentration ratio of the initial compounds B2H6:CH4 (or C2H6), and plasma discharge power. The BxCyNz:H films exhibited very complicated IR spectra; the authors have specially stressed that it is impossible to make conclusions concerning types of chemical bonds in the material basing only on the IR spectroscopic data.

Amorphous BCxNy:H films were prepared in a capacitively coupled RF-PECVD reactor at deposition temperatures <200oC starting from B2H6+CH4+N2+H2 gas mixture (Dekempeneer et al., 1996, 1997). Films were deposited on Si, steel and glass substrates. By varying the partial pressure of the gases, the composition was varied in a wide area of the B-N-C triangle. The same initial gas mixture was used by Polo et al. (Polo et al., 1998, 1999). It was found that the films had a less ordered structure.

Both amorphous and polycrystalline coatings were deposited by microwave low pressure CVD (LPCVD) using a mixture of B2H6+CH4+NH3+H2 at substrate temperature in the range 800-1350°C. (Stanishevsky, 2010). Amorphous coatings were usually formed at lower substrate temperatures and were non-homogeneous across the coating thickness. Polycrystalline coatings were generally represented by both diamond and boron nitride phases. In one case, a polycrystalline coating with the composition of B2CN4 was fabricated.

The turbostratic structure of BCxNy with various compositions was synthesized by biasassisted hot-filament CVD (HFCVD) (Yu et al., 1999a, 1999b; Wang, 1999) within the temperature range 873-1273 K from B2H6+CH4+N2+H2 mixture. Investigation of the films by means of XPS demonstrated that the three atoms B, C, N are chemically bound. Boron carbonitride is the main phase in all the deposited samples, though in some cases (at high temperature) this phase was co-deposited with boron carbide. The growth rate of BCN films decreased substantially with increased temperature. Chemical composition and morphology of the layers were also dependent on deposition temperature. The turbostratic BCxNy films were also grown by HFCVD from mixture of B2H6+CH4+NH3+H2 (Xie et al., 1998).

Laser-assisted CVD was used for preparation of single-phase BCxNy layers at low temperature in a gas atmosphere containing B2H6+CH4+NH3, where the starting composition ratio could be varied in a large range. Layers exhibited turbostratic structure. Some planar structure, containing especially CB2N groups, were suggested for the "unit cell" of CBN solid solutions (Morjan et al., 1999).

BCN films were deposited from mixture B2H6+CH4+N2+H2 with electron beam excited plasma-chemical vapor deposition (EBEP-CVD) (Hasegawa et al., 2002, 2003). By controlling the flow rate ratios of the process gases, films with composition expressed as BxCyN, where x=0.9-4.7 and y=0.5-6.0 were obtained.

#### **3. Syntheses of layers by single-source precursors**

#### **3.1 Silicon carbonitrides**

The review highlights of the synthesis, processing and properties of non-oxide silicon-based bulk ceramics materials derived from silazanes and polysilazanes (Kroke et al., 2000).

Compilation on Synthesis, Characterization

materials for advanced technology.

carbonitride network formation.

temperature stability.

**3.1.6 Tris(dimethylamino)silane (TrDMAS)** 

**3.1.7 Bis(trimethylsilyl)carbodiimide (BTSC)** 

SiCN films is sensitive to the contribution of the Si-CH2-N links.

**3.1.5 Bis(dimethylamino)methylsilane (BDMAMS)** 

500°C).

1.87.

**3.1.4 Bis(dimethylamino)dimethylsilane (BDMADMS)** 

and Properties of Silicon and Boron Carbonitride Films 499

SiCN thin films for membrane application were deposited by PECVD from bis(dimethylamino)dimethylsilane (BDMADMS) (Kafrouni et al., 2010). Single gas permeation tests have been carried out and a helium permeability of about 10−7 mol m−<sup>2</sup> s−1 Pa−1 was obtained with an ideal selectivity of helium/nitrogen of about 20. Moreover these PECVD membranes also seem to be stable at higher temperature in air (up to

a-Si:C:N:H films were produced by RPCVD from dimethylaminodimethylsilane (Blaszczyk-Lezak et al., 2005, 2006). The films deposited at different substrate temperatures (30–400°C). Strong adhesion to a substrate, high hardness (H=28–35GPa), low friction coefficient (μ=0.04, against stainless steel), and strong resistance to wear (predicted from high "plasticity index" values H/E°=0.10–0.12) were found for these films suggest that these materials are promising coatings for improving tribological properties of engineering

Siliconnitride-like films were deposited at low temperatures using RF inductively coupled plasma fed with bis(dimethylamino)-dimethylsilane (BDMADMS) and argon (Ar) (Mundo et al., 2005). The results indicate that at high power input and low monomer-to-Ar ratio, low carbon and high nitrogen content films can be obtained, stable and with a refractive index of

The RPCVD with bis(dimethylamino)methylsilane precursor was used for the synthesis of Si:C:N films (Blaszczyk-Lezak et al., 2007). The increase of TS enhances crosslinking in the film via the formation of nitridic Si–N and carbidic Si–C bonds. On the basis of the structural data a hypothetical crosslinking reaction has been proposed, contributing to silicon

Amorphous SiCN films were fabricated by RPCVD using H2 and TrDMAS, (Me2N)3SiH, as a novel single-source precursor, being a carrier of Si-N and C-N units (Wrobel et al., 2010). The Arrhenius plot of the temperature dependence of the film density implies that for TS>200°C a thermally enhanced crosslinking process predominates, and the density reaches a high value of ρ≈3.0 g cm-3 at TS=350°C. SiCN films are morphologically homogeneous materials exhibiting very low surface roughness Rrms=0.3 nm. The photoluminescence of

Amorphous SiCN coatings were prepared on steel substrates by RF-PECVD from BTSC (Zhou et al., 2006; Probst et al., 2005; Stelzner et al., 2005). The results of the studies show that the coatings obtained on the RF-powered electrode (cathode) were black, thick (>20 µm) and hard (21–29GPa), while those grown on the grounded electrode (anode) were yellow, thin (<4 µm) and soft (~5GPa). The surfaces of all coatings were very smooth with a maximum rms roughness between 2 nm and 5 nm for an area of 5µm×5µm. Wear tests at 600°C showed that the coatings posses an excellent high-

At the present time, the alternative way of synthesis of silicon carbonitride films is through the use of low-toxicity siliconorganic compounds of various compositions and structures used as single source-precursors containing all the necessary elements Si, C, and N in one molecule. These compounds are of special interest because the molecular structure of the initial organosilicon compound affects the chemical and phase compositions plus the microstructure of deposited silicon carbonitride films.

#### **3.1.1 Hexamethyldisilazane (HMDSN)**

SiCN films were deposited by HWCVD method using HMDSN which is an organic liquid material (Izumi et al, 2006; Limmanee et al, 2008). It is found that the composition ratio of SiCN can be controlled by changing the flow rate of NH3. SiCN films can be deposited at the substrate temperature of 100°C. The dielectric constant can be controlled from 2.9 to 7 by changing the flow rate of NH3. The best efficiency of 13.75% for polycrystalline silicon solar cells using a-SiCN:H films was achieved at the temperature of 750°C.

SiCN films were obtained at a substrate temperature of 250°C by HWCVD using HMDSN (Nakayamada et.al, 2008). No SiCN film thickness was changed at all for 1 week in 10wt.% H2SO4. A high corrosion resistance was confirmed.

SiCN nanopowders with different chemical compositions and characteristics can be prepared by CO2 laser pyrolysis of organosilicon precursors (HMDSN or TMDSN, see section 3.1.9) or their mixture with silane (Dez et.al, 2002). A correlation is established between the synthesis conditions of powders and their chemical composition, morphology, structure and thermal stability.

a-SiCN thin films were deposited at 250-500°C using a microwave plasma assisted CVD process fed with a mixture of CH4, N2, Ar and hexamethyldisilazane (Bulou et al., 2010). The increase of the CH4 rate results in less organic, films of higher density and in an increase of the refractive index. The CH4 addition to the gaseous mixture leads to a value of the Si/N ratio of films very close to stoichiometric Si3N4.

Si:C:N:H thin films were deposited by PECVD using HMDSN as monomer and Ar as carrier gas (Vassallo et al., 2006). The films become more amorphous and inorganic at increasing RF plasma power. The wettability of the film has been studied and related to the chemical composition and to the morphology of the deposited layers.

SiCxNy films were synthesized with the composition varying in a wide range from those similar to silicon carbide to those similar to silicon nitride. HMDS was used by PECVD as single-source precursor in the mixtures with helium, nitrogen or ammonia in the wide range of temperatures from 100 up to 800°С and RF plasma powers from 15 up to 70 W (Fainer et.al., 1999, 2000, 2001a, 2001b, 2003, 2004, 2008).

#### **3.1.2 Ethylsilazane**

Thin films of amorphous Si-C-N were grown on Si(100) substrates by the pyrolysis of ethylsilazane in mixtures with H2 in the temperature range of 873-1073K. (Bae et al., 1992). It was shown that the refraction index of these films varied from 1.81 to 2.09, elastic recoil detection decreased from 21 to 8% in the range of temperatures from 873K to 1073K. The chemical composition of the films was determined to be Si43C7N48O2.

#### **3.1.3 Polysilazane**

Non-stoichiometric X-ray-amorphous Si3+xN4Cx+y was deposited during pyrolysis of polysilazane at 1440°С. (Schonfelder, 1993). The heating up to 1650°C results in formation of a mixture of the nanocomposites Si3N4/SiC or Si3N4/SiC/C.

At the present time, the alternative way of synthesis of silicon carbonitride films is through the use of low-toxicity siliconorganic compounds of various compositions and structures used as single source-precursors containing all the necessary elements Si, C, and N in one molecule. These compounds are of special interest because the molecular structure of the initial organosilicon compound affects the chemical and phase compositions plus the

SiCN films were deposited by HWCVD method using HMDSN which is an organic liquid material (Izumi et al, 2006; Limmanee et al, 2008). It is found that the composition ratio of SiCN can be controlled by changing the flow rate of NH3. SiCN films can be deposited at the substrate temperature of 100°C. The dielectric constant can be controlled from 2.9 to 7 by changing the flow rate of NH3. The best efficiency of 13.75% for polycrystalline silicon solar

SiCN films were obtained at a substrate temperature of 250°C by HWCVD using HMDSN (Nakayamada et.al, 2008). No SiCN film thickness was changed at all for 1 week in 10wt.%

SiCN nanopowders with different chemical compositions and characteristics can be prepared by CO2 laser pyrolysis of organosilicon precursors (HMDSN or TMDSN, see section 3.1.9) or their mixture with silane (Dez et.al, 2002). A correlation is established between the synthesis conditions of powders and their chemical composition, morphology,

a-SiCN thin films were deposited at 250-500°C using a microwave plasma assisted CVD process fed with a mixture of CH4, N2, Ar and hexamethyldisilazane (Bulou et al., 2010). The increase of the CH4 rate results in less organic, films of higher density and in an increase of the refractive index. The CH4 addition to the gaseous mixture leads to a value of the Si/N

Si:C:N:H thin films were deposited by PECVD using HMDSN as monomer and Ar as carrier gas (Vassallo et al., 2006). The films become more amorphous and inorganic at increasing RF plasma power. The wettability of the film has been studied and related to the chemical

SiCxNy films were synthesized with the composition varying in a wide range from those similar to silicon carbide to those similar to silicon nitride. HMDS was used by PECVD as single-source precursor in the mixtures with helium, nitrogen or ammonia in the wide range of temperatures from 100 up to 800°С and RF plasma powers from 15 up to 70 W (Fainer

Thin films of amorphous Si-C-N were grown on Si(100) substrates by the pyrolysis of ethylsilazane in mixtures with H2 in the temperature range of 873-1073K. (Bae et al., 1992). It was shown that the refraction index of these films varied from 1.81 to 2.09, elastic recoil detection decreased from 21 to 8% in the range of temperatures from 873K to 1073K. The

Non-stoichiometric X-ray-amorphous Si3+xN4Cx+y was deposited during pyrolysis of polysilazane at 1440°С. (Schonfelder, 1993). The heating up to 1650°C results in formation of

microstructure of deposited silicon carbonitride films.

H2SO4. A high corrosion resistance was confirmed.

ratio of films very close to stoichiometric Si3N4.

et.al., 1999, 2000, 2001a, 2001b, 2003, 2004, 2008).

composition and to the morphology of the deposited layers.

chemical composition of the films was determined to be Si43C7N48O2.

a mixture of the nanocomposites Si3N4/SiC or Si3N4/SiC/C.

cells using a-SiCN:H films was achieved at the temperature of 750°C.

**3.1.1 Hexamethyldisilazane (HMDSN)** 

structure and thermal stability.

**3.1.2 Ethylsilazane** 

**3.1.3 Polysilazane** 

#### **3.1.4 Bis(dimethylamino)dimethylsilane (BDMADMS)**

SiCN thin films for membrane application were deposited by PECVD from bis(dimethylamino)dimethylsilane (BDMADMS) (Kafrouni et al., 2010). Single gas permeation tests have been carried out and a helium permeability of about 10−7 mol m−<sup>2</sup> s−1 Pa−1 was obtained with an ideal selectivity of helium/nitrogen of about 20. Moreover these PECVD membranes also seem to be stable at higher temperature in air (up to 500°C).

a-Si:C:N:H films were produced by RPCVD from dimethylaminodimethylsilane (Blaszczyk-Lezak et al., 2005, 2006). The films deposited at different substrate temperatures (30–400°C). Strong adhesion to a substrate, high hardness (H=28–35GPa), low friction coefficient (μ=0.04, against stainless steel), and strong resistance to wear (predicted from high "plasticity index" values H/E°=0.10–0.12) were found for these films suggest that these materials are promising coatings for improving tribological properties of engineering materials for advanced technology.

Siliconnitride-like films were deposited at low temperatures using RF inductively coupled plasma fed with bis(dimethylamino)-dimethylsilane (BDMADMS) and argon (Ar) (Mundo et al., 2005). The results indicate that at high power input and low monomer-to-Ar ratio, low carbon and high nitrogen content films can be obtained, stable and with a refractive index of 1.87.

#### **3.1.5 Bis(dimethylamino)methylsilane (BDMAMS)**

The RPCVD with bis(dimethylamino)methylsilane precursor was used for the synthesis of Si:C:N films (Blaszczyk-Lezak et al., 2007). The increase of TS enhances crosslinking in the film via the formation of nitridic Si–N and carbidic Si–C bonds. On the basis of the structural data a hypothetical crosslinking reaction has been proposed, contributing to silicon carbonitride network formation.

#### **3.1.6 Tris(dimethylamino)silane (TrDMAS)**

Amorphous SiCN films were fabricated by RPCVD using H2 and TrDMAS, (Me2N)3SiH, as a novel single-source precursor, being a carrier of Si-N and C-N units (Wrobel et al., 2010). The Arrhenius plot of the temperature dependence of the film density implies that for TS>200°C a thermally enhanced crosslinking process predominates, and the density reaches a high value of ρ≈3.0 g cm-3 at TS=350°C. SiCN films are morphologically homogeneous materials exhibiting very low surface roughness Rrms=0.3 nm. The photoluminescence of SiCN films is sensitive to the contribution of the Si-CH2-N links.

#### **3.1.7 Bis(trimethylsilyl)carbodiimide (BTSC)**

Amorphous SiCN coatings were prepared on steel substrates by RF-PECVD from BTSC (Zhou et al., 2006; Probst et al., 2005; Stelzner et al., 2005). The results of the studies show that the coatings obtained on the RF-powered electrode (cathode) were black, thick (>20 µm) and hard (21–29GPa), while those grown on the grounded electrode (anode) were yellow, thin (<4 µm) and soft (~5GPa). The surfaces of all coatings were very smooth with a maximum rms roughness between 2 nm and 5 nm for an area of 5µm×5µm. Wear tests at 600°C showed that the coatings posses an excellent hightemperature stability.

Compilation on Synthesis, Characterization

different gaseous additives (ammonia, nitrogen, hydrogen).

used as boron, carbon, and nitrogen sources.

reactivity of boron with oxygen in the absence of N2.

**3.2.1 Trimethylamine borane (TMAB)** 

**3.2 Boron carbonitrides** 

and Properties of Silicon and Boron Carbonitride Films 501

During the recent years, special attention was paid to the introduction of volatile compounds – single-source precursors - containing all the necessary atoms (boron, carbon, and nitrogen) for the synthesis of boron carbonitrides. The use of complex organoelemental volatile compounds should be considered as an essential step forward. Since these compounds are incombustible and rather stable toward reactions in the natural atmosphere, their application in technology is preferable over chemically active boron trichloride and diborane. These substances have a well-defined ratio of B:C:N due to their stoichiometry, and they can be evaporated and easily handled due to their chemical and physical properties. With these compounds, one can obtain layers of different composition using

In CVD processes, molecular precursors such as dimethylamine borane (CH3)2HN⋅BH3 (DMAB), trimethylamine borane (CH3)3N⋅BH3 (TMAB), triethylamine borane (C2H5)3N⋅BH3 (TEAB), N,N',N''-trimethylborazine (CH3)3N3B3H3, N,N',N''-triethylborazine (C2H5)3N3B3H3, tris-(dimethylamino)borane B(N(CH3)3, (N-pyrrolidino)diethylborane C8H18BN, pyridine borane C5H5NBH3, and triazaborabicyclohexane BN3H2(CH2)6 have been

Kosinova et al. pioneered the use of trimethylamine borane complex (CH3)3N⋅BH3 in both RF PECVD (40.68 MHz) and LPCVD processes for BCN film deposition (Kosinova et al., 2001, 2003a, 2003b; Fainer et al., 2001). Boron carbonitride lms were grown by PECVD using TMAB and its mixtures with ammonia, hydrogen, or helium. The effects of the starting-mixture composition and substrate temperature on the chemical composition of the deposits were studied. The results indicate that the initial composition of the gas mixture, the nature of the activation gas, and substrate temperature play a key role in determining the deposition kinetics and the physicochemical properties of the deposits. Depending on

h-BCN films with a thickness of ≈4 μm were synthesized on Si(100) substrate by RF (13.56 MHz, 1kW) and microwave (2,45 GHz) PECVD using mixture of TMAB and H2 as precursors (Mannan et al., 2007). The temperature of deposition was 300 and 600°C for RF PECVD, and 840-850°C for MF PECVD processes. The films were amorphous with an inhomogeneous microstructure confirmed by XRD and FEG-SEM. XPS and FTIR suggested that the films were consisted of a variety of bonds between B, C and N atoms such as B-N, B-C and C-N. Oxygen was inevitably incorporated as a contaminant (13-15 at.%). The effects of the deposition conditions, including microwave power and carrier gas, on the film properties was studied by Kida (Kida, 2009). The deposition time varied between 0.5 h and 2 h with microwave powers of 200, 300, and 400W (2.45 GHz). Substrate temperatures depended on the microwave power applied and ranged between 700 and 900°C. N2 and a gas mixture of CH4 (10vol.%) and H2 (90vol.%) were used as carrier gases. The films deposited were found to have fibrous nanostructures consisting of nanosized fibers. For the films deposited under N2 flow, boron and nitrogen contents of the films increased as the microwave power increased, leading to the formation of B–N and C–N bonds, as confirmed by FTIR. Moreover, deposition at higher microwave power reduced the oxygen content in the films. However, for films deposited under CH4+H2 flow, B–O bond formation dominated (B30C15N4O51), owing to the high

these process parameters, one can obtain h-BN, h-BN + B4C, or h-BCxNy lms.

#### **3.1.8 Dimethyl(2,2-dimethylhydrazino)silane (DMDMHS) and dimethyl-***bis***dimethylhydrazino silane (DM-***bis***-DMHSN)**

SiCN films were synthesised by RPECVD using a novel single-source precursors dimethyl(2,2-dimethylhydrazino) silane (CH3)2HSiNHN(CH3)2, (DMDMHS) and dimethyl*bis*-dimethylhydrazino silane (CH3)2Si[NHN(CH3)2]2 (DM-*bis*-DMHSN), which are silyl derivatives of 1,1-dimethylhydrazine (Smirnova et al., 2003). The films were found to be predominantly amorphous with a number of crystallites embedded in an unstructured matrix. The crystalline phase can be indexed as a tetragonal cell with lattice parameters a=9.6 Å and c=6.4 Å. This novel material has an optical band gap varying within the energy range from 2.0 to 4.7 eV.

#### **3.1.9 Tetramethyldisilazane (TMDSN)**

Si:C:N films were produced by RPCVD from mixture of a 1,1,3,3-tetramethyldisilazane precursor with H2 (Blaszczyk-Lezak et al., 2006a, 2006b; Wrobel et al., 2007). An increase in TS leads to the elimination of the organic groups and subsequent crosslinking via the formation of Si-C and Si-N networks. In view of the relatively high hardness (16 GPa) and a low friction coefficient µ value (0.02-0.05 against stainless steel) found for the a-Si:C:N film deposited at TS=400°C, this material may be useful as a tribological coating for metals.

#### **3.1.10 Hexamethylcyclotrisilazane (HMCTS)**

Silicon nitride films were obtained in glow-discharge plasma from HMCTS in a mixture with nitrogen gas or ammonia at low temperatures (below 150°С) (Voronkov et al., 1981). Chemical composition was analyzed with IR-spectroscopy and demonstrated that in the films obtained at such conditions are present Si-N, C-C, Si-H (or Si-C≡N) and N-H bonds.

The silicon nitride films synthesized from HMCTS with a set of additional gases such as NH3, H2, and N2 by PECVD at temperatures (150-400°С) and plasma power of 5-50 W (Brooks& Hess, 1987, 1988). The films obtained from a gas mixture (HMCTS + NH3) and characterized by lesser than 4 at.% carbon and hydrogen content of about 25 at.% are close to the chemical composition of silicon nitride films. The Si-N bonds are dominant. The films obtained from the mixture (HMCTS + H2), contain significant amount of carbon (30-40 at.%) and 21 at.% of hydrogen. These films contain both Si-N and Si-C bonds.

SiCN films were obtained by RPECVD using HMCTS in a mixture with helium or nitrogen in the range of temperatures of 100-750°С and plasma powers of 15-50W (Fainer et al, 2009a, 2009b). The low temperature SiCxNyOz:H films are compounds with chemical bonds among the main elements Si, N, and C together with impurity elements such as hydrogen and oxygen. The empirical formula of the high-temperature films is represented by SiCxNy. The absensence of hydrogen in these films leads to good thermal stability and microhardness. These films exhibit an excellent transparency with a transmittance of ~92–95% in the spectral range λ=380–2500 nm.

#### **3.1.11 N-bromhexamethyldisilazane**

SiCN films were producted from the new volatile organosilicon compound Nbromhexamethyldisilazane (Smirnova et al, 2008). An increase in the refractive index from 1.5 to 2.2 and a decrease in the optical band gap width from 4.5 to 2.1 eV is observed as the chemical composition of the films changes in the temperature interval of 470–870K.

#### **3.2 Boron carbonitrides**

500 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

SiCN films were synthesised by RPECVD using a novel single-source precursors dimethyl(2,2-dimethylhydrazino) silane (CH3)2HSiNHN(CH3)2, (DMDMHS) and dimethyl*bis*-dimethylhydrazino silane (CH3)2Si[NHN(CH3)2]2 (DM-*bis*-DMHSN), which are silyl derivatives of 1,1-dimethylhydrazine (Smirnova et al., 2003). The films were found to be predominantly amorphous with a number of crystallites embedded in an unstructured matrix. The crystalline phase can be indexed as a tetragonal cell with lattice parameters a=9.6 Å and c=6.4 Å. This novel material has an optical band gap varying within the energy

Si:C:N films were produced by RPCVD from mixture of a 1,1,3,3-tetramethyldisilazane precursor with H2 (Blaszczyk-Lezak et al., 2006a, 2006b; Wrobel et al., 2007). An increase in TS leads to the elimination of the organic groups and subsequent crosslinking via the formation of Si-C and Si-N networks. In view of the relatively high hardness (16 GPa) and a low friction coefficient µ value (0.02-0.05 against stainless steel) found for the a-Si:C:N film deposited at TS=400°C, this material may be useful as a tribological coating for

Silicon nitride films were obtained in glow-discharge plasma from HMCTS in a mixture with nitrogen gas or ammonia at low temperatures (below 150°С) (Voronkov et al., 1981). Chemical composition was analyzed with IR-spectroscopy and demonstrated that in the films obtained at such conditions are present Si-N, C-C, Si-H (or Si-C≡N) and N-H bonds. The silicon nitride films synthesized from HMCTS with a set of additional gases such as NH3, H2, and N2 by PECVD at temperatures (150-400°С) and plasma power of 5-50 W (Brooks& Hess, 1987, 1988). The films obtained from a gas mixture (HMCTS + NH3) and characterized by lesser than 4 at.% carbon and hydrogen content of about 25 at.% are close to the chemical composition of silicon nitride films. The Si-N bonds are dominant. The films obtained from the mixture (HMCTS + H2), contain significant amount of carbon (30-40 at.%)

SiCN films were obtained by RPECVD using HMCTS in a mixture with helium or nitrogen in the range of temperatures of 100-750°С and plasma powers of 15-50W (Fainer et al, 2009a, 2009b). The low temperature SiCxNyOz:H films are compounds with chemical bonds among the main elements Si, N, and C together with impurity elements such as hydrogen and oxygen. The empirical formula of the high-temperature films is represented by SiCxNy. The absensence of hydrogen in these films leads to good thermal stability and microhardness. These films exhibit an excellent transparency with a transmittance of ~92–95% in the spectral

SiCN films were producted from the new volatile organosilicon compound Nbromhexamethyldisilazane (Smirnova et al, 2008). An increase in the refractive index from 1.5 to 2.2 and a decrease in the optical band gap width from 4.5 to 2.1 eV is observed as the

chemical composition of the films changes in the temperature interval of 470–870K.

and 21 at.% of hydrogen. These films contain both Si-N and Si-C bonds.

**3.1.8 Dimethyl(2,2-dimethylhydrazino)silane (DMDMHS) and dimethyl-***bis***-**

**dimethylhydrazino silane (DM-***bis***-DMHSN)** 

range from 2.0 to 4.7 eV.

range λ=380–2500 nm.

**3.1.11 N-bromhexamethyldisilazane** 

metals.

**3.1.9 Tetramethyldisilazane (TMDSN)** 

**3.1.10 Hexamethylcyclotrisilazane (HMCTS)** 

During the recent years, special attention was paid to the introduction of volatile compounds – single-source precursors - containing all the necessary atoms (boron, carbon, and nitrogen) for the synthesis of boron carbonitrides. The use of complex organoelemental volatile compounds should be considered as an essential step forward. Since these compounds are incombustible and rather stable toward reactions in the natural atmosphere, their application in technology is preferable over chemically active boron trichloride and diborane. These substances have a well-defined ratio of B:C:N due to their stoichiometry, and they can be evaporated and easily handled due to their chemical and physical properties. With these compounds, one can obtain layers of different composition using different gaseous additives (ammonia, nitrogen, hydrogen).

In CVD processes, molecular precursors such as dimethylamine borane (CH3)2HN⋅BH3 (DMAB), trimethylamine borane (CH3)3N⋅BH3 (TMAB), triethylamine borane (C2H5)3N⋅BH3 (TEAB), N,N',N''-trimethylborazine (CH3)3N3B3H3, N,N',N''-triethylborazine (C2H5)3N3B3H3, tris-(dimethylamino)borane B(N(CH3)3, (N-pyrrolidino)diethylborane C8H18BN, pyridine borane C5H5NBH3, and triazaborabicyclohexane BN3H2(CH2)6 have been used as boron, carbon, and nitrogen sources.

#### **3.2.1 Trimethylamine borane (TMAB)**

Kosinova et al. pioneered the use of trimethylamine borane complex (CH3)3N⋅BH3 in both RF PECVD (40.68 MHz) and LPCVD processes for BCN film deposition (Kosinova et al., 2001, 2003a, 2003b; Fainer et al., 2001). Boron carbonitride lms were grown by PECVD using TMAB and its mixtures with ammonia, hydrogen, or helium. The effects of the starting-mixture composition and substrate temperature on the chemical composition of the deposits were studied. The results indicate that the initial composition of the gas mixture, the nature of the activation gas, and substrate temperature play a key role in determining the deposition kinetics and the physicochemical properties of the deposits. Depending on these process parameters, one can obtain h-BN, h-BN + B4C, or h-BCxNy lms.

h-BCN films with a thickness of ≈4 μm were synthesized on Si(100) substrate by RF (13.56 MHz, 1kW) and microwave (2,45 GHz) PECVD using mixture of TMAB and H2 as precursors (Mannan et al., 2007). The temperature of deposition was 300 and 600°C for RF PECVD, and 840-850°C for MF PECVD processes. The films were amorphous with an inhomogeneous microstructure confirmed by XRD and FEG-SEM. XPS and FTIR suggested that the films were consisted of a variety of bonds between B, C and N atoms such as B-N, B-C and C-N. Oxygen was inevitably incorporated as a contaminant (13-15 at.%). The effects of the deposition conditions, including microwave power and carrier gas, on the film properties was studied by Kida (Kida, 2009). The deposition time varied between 0.5 h and 2 h with microwave powers of 200, 300, and 400W (2.45 GHz). Substrate temperatures depended on the microwave power applied and ranged between 700 and 900°C. N2 and a gas mixture of CH4 (10vol.%) and H2 (90vol.%) were used as carrier gases. The films deposited were found to have fibrous nanostructures consisting of nanosized fibers. For the films deposited under N2 flow, boron and nitrogen contents of the films increased as the microwave power increased, leading to the formation of B–N and C–N bonds, as confirmed by FTIR. Moreover, deposition at higher microwave power reduced the oxygen content in the films. However, for films deposited under CH4+H2 flow, B–O bond formation dominated (B30C15N4O51), owing to the high reactivity of boron with oxygen in the absence of N2.

Compilation on Synthesis, Characterization

contain high amounts of tantalum (BCN-Ta layers).

TDMAB and N2 at 350°C (Aoki et al., 2010a, 2010b).

**3.2.5 (N-pyrrolidino)-diethylborate (PEB)** 

**3.2.6 Borazine derivatives** 

**3.2.6.1 Trimethylborazine** 

the coating was 2.05 – 2.09 g/cm3.

temperature 650-750°C.

and Properties of Silicon and Boron Carbonitride Films 503

substrate temperatures lead to the formation of hemispheres (approximately 10 μm in diameter)

BCN-layers were deposited in a hot-filament supported reactor using TDMAB (Weissenbacher et al., 2002, 2006). These experiments were strongly influenced by the stability of the Ta-filament. At filament temperatures of 1500°C layer deposition on the surface of the filament takes place, at temperatures higher than 2000°C liquid phase formation led to filament breakdown in many cases. The morphology of the deposited BCN layers on hard metal substrates (WC-Co) depends on the deposition conditions and films

Methyl-BCN lms were deposited by plasma-assisted CVD (PACVD) using mixture of

Hexagonal boron carbonitride hybrid lms - sp2-BCN phase with h-BN-like conguration have been synthesized on Si(100) and on highly oriented pyrolytic graphite, respectively, by RF-PACVD using TDMAB (Mannan et al., 2009a, 2009b, 2010). The deposition was performed at different RF powers of 400–800W, at the working pressure of 2×10-1 Torr and

The influence of the B/C/N containing single-source precursors pyridine-borane (PB) and triazaborabicyclodecane (TBBD) on the chemical composition of boron carbonitride thin films was investigated in (Hegemann et al., 1997, 1999). The films are deposited via a PACVD process, activated by 13.56 MHz radio frequency (RF). N2, Ar and He serve as carrier gases. It becomes evident, that from a certain bias voltage, the self-bias in capacitively RF electrical discharges mainly influences the chemical composition of the BCN films independent of the kind of the used precursor. Films that were either deposited in He using a low power density or in N2 using a high power density showed comparable

BCN films were deposited on polycarbonate and silicon wafer by means of different RF PACVD (inductively coupled and capacitively coupled RF PACVD), by use of liquid organic compound (N-pyrrolidino)diethylborane (C8H18BN) as precursors. Deposition was carried out on at 95–120°C. A mixture of argon, hydrogen, and nitrogen was used as process gas. The layer shows a columnar structure. The composition of BCN lms deposited ranged

Amorphous semiconductor BCN films were produced by means of pyrolysis of borazine derivatives (tris(1,3,2-benzodiazaborolo)borazine) at 1073 K (Maya, 1988a, 1988b; Maya, & Harris, 1990). Quartz, titanium and silicon were used as substrates. Chemical analysis showed that the synthesized material had the composition BC5.2N1.8H1.9O0.45. The density of

Mixtures of *N,N',N''*-trimethylborazine = (CH3)3N3B3H3 (TMB) (the B:C:N ratio 1:1:1) and argon have been used to deposit BCN:H films by means of ECR PECVD processes at 100- 150°C (Weber et al., 1992, 1993). Amorphous BCN layers were deposited on a polycarbonate substrate by RF PACVD at low temperature (95-120°C) using a mixture of TMB and H2, N2, or Ar (Wöhle et al., 1999). The composition of the layers varied in a wide range. The boron

properties. Analysis of these films showed their chemical composition to be BC4N.

between BC7.3N0.8 and BC0.9N0.6 (Wöhle et al., 1999; Ahn et al., 2003).

on the surface, thus increasing the roughness of the BCN layers (Gammer et al., 2002).

#### **3.2.2 Dimethylamine borane (DMAB)**

Boron nitride films were obtained by means of PECVD of DMAB+(CH3)2NH⋅BH3 - in mixture with ammonia (Schmolla et al., 1983; Bath et al., 1989, 1991, 1994) or nitrogen (Baehr et al., 1997; Boudiombo et al., 1997; Abdellaoui et al., 1997).

Amorphous and poor crystalline phases in the B-C-N system were obtained by plasma chemical decomposition of DMAB in mixture with hydrogen and argon (Loeffler et al., 1997).

#### **3.2.3 Triethylamine-borane (TEAB)**

In the study of Levy et al., films consisting of B-N-C-H have been synthesized by LPCVD using the liquid precursor TEAB = (C2H5)3N⋅BH3 complex both with and without ammonia (Levy et al., 1995). In the absence of NH3, the growth rate dependency on temperature follows an Arrhenius behaviour with an apparent activation energy of 11 kcal/mole. The addition of NH3 has the effect of lowering the deposition temperature to 300°C and doubling the apparent activation energy. The deposits were found to be in all cases amorphous. A significant increase in carbon concentration was observed above 650°C due to the break up of the amine molecule. The addition of NH3 was used to reduce the carbon content in the films. The same result of the ammonia effect was also obtained by Kosinova et al. (Kosinova et al., 1999, 2001). The BCN films were deposited on Si(100), GaAs(100) and fused silica substrates using TEAB with and without ammonia by both LPCVD and RF-PECVD (40.68 MHz) methods.

With TEAB in the direct current glow discharge plasma process (GD-PECVD) the highest carbon concentrations (48–73 at.%) in BCN films are obtained without using an additional carbon source (Thamm et al., 2005). Elastic recoil detection analysis (ERDA) measurements yield information on the layer composition regarding the concentrations of the elements boron, carbon, nitrogen, and hydrogen. The hydrogen content in the produced BCN layers strongly depends on the substrate temperature and increases up to 35 at.%. Depth profiles show a homogeneous distribution of the elements B, C, N, and H over the entire layer thickness.

The paper of Mannan et al. presents the chemical bonding states and the local structures of oriented hexagonal BCN films with the grain size of around 100 nm synthesized by microwave PECVD (MW-PECVD) using mixture TEAB and CH4+H2 as the carrier gas (Mannan et al., 2008). The deposition was performed at different microwave power settings of 200–500W at working pressure of 5.0 Torr. The substrate temperature was measured to be 750 and 850°C, respectively. It was estimated a particle size of around 100-150 nm. The crystallinity was not good as the hexagonal structures appeared in a short-range order which could not be detected by XRD.

#### **3.2.4 Tris-(dimethylamino)borane (TDMAB)**

Homogeneous carbon boronitride coatings were produced with cold-wall CVD varying the temperature of the deposition substrate from 800°C up to 1400°C using tris-(dimethylamino) borane B[N(CH3)2]3 as a single-source molecular precursor. The deposition temperature has an influence on the growth rate as well as on the coating composition (C: 35–75at%; B: 12–40at%; N: 7–24at%). Below 700°C substrate temperature no deposition can be observed. At temperatures between 700°C and 800°C the layers grow very slowly and they are oriented parallel to the substrate´s surface. If temperatures are raised to 900°C the layer already seems to be under stress as it cracks into small pieces during cooling to room temperature. Higher

Boron nitride films were obtained by means of PECVD of DMAB+(CH3)2NH⋅BH3 - in mixture with ammonia (Schmolla et al., 1983; Bath et al., 1989, 1991, 1994) or nitrogen (Baehr

Amorphous and poor crystalline phases in the B-C-N system were obtained by plasma chemical decomposition of DMAB in mixture with hydrogen and argon (Loeffler et al., 1997).

In the study of Levy et al., films consisting of B-N-C-H have been synthesized by LPCVD using the liquid precursor TEAB = (C2H5)3N⋅BH3 complex both with and without ammonia (Levy et al., 1995). In the absence of NH3, the growth rate dependency on temperature follows an Arrhenius behaviour with an apparent activation energy of 11 kcal/mole. The addition of NH3 has the effect of lowering the deposition temperature to 300°C and doubling the apparent activation energy. The deposits were found to be in all cases amorphous. A significant increase in carbon concentration was observed above 650°C due to the break up of the amine molecule. The addition of NH3 was used to reduce the carbon content in the films. The same result of the ammonia effect was also obtained by Kosinova et al. (Kosinova et al., 1999, 2001). The BCN films were deposited on Si(100), GaAs(100) and fused silica substrates using TEAB with and without ammonia by both LPCVD and RF-

With TEAB in the direct current glow discharge plasma process (GD-PECVD) the highest carbon concentrations (48–73 at.%) in BCN films are obtained without using an additional carbon source (Thamm et al., 2005). Elastic recoil detection analysis (ERDA) measurements yield information on the layer composition regarding the concentrations of the elements boron, carbon, nitrogen, and hydrogen. The hydrogen content in the produced BCN layers strongly depends on the substrate temperature and increases up to 35 at.%. Depth profiles show a homogeneous distribution of the elements B, C, N, and H over the entire layer

The paper of Mannan et al. presents the chemical bonding states and the local structures of oriented hexagonal BCN films with the grain size of around 100 nm synthesized by microwave PECVD (MW-PECVD) using mixture TEAB and CH4+H2 as the carrier gas (Mannan et al., 2008). The deposition was performed at different microwave power settings of 200–500W at working pressure of 5.0 Torr. The substrate temperature was measured to be 750 and 850°C, respectively. It was estimated a particle size of around 100-150 nm. The crystallinity was not good as the hexagonal structures appeared in a short-range order

Homogeneous carbon boronitride coatings were produced with cold-wall CVD varying the temperature of the deposition substrate from 800°C up to 1400°C using tris-(dimethylamino) borane B[N(CH3)2]3 as a single-source molecular precursor. The deposition temperature has an influence on the growth rate as well as on the coating composition (C: 35–75at%; B: 12–40at%; N: 7–24at%). Below 700°C substrate temperature no deposition can be observed. At temperatures between 700°C and 800°C the layers grow very slowly and they are oriented parallel to the substrate´s surface. If temperatures are raised to 900°C the layer already seems to be under stress as it cracks into small pieces during cooling to room temperature. Higher

**3.2.2 Dimethylamine borane (DMAB)** 

**3.2.3 Triethylamine-borane (TEAB)** 

PECVD (40.68 MHz) methods.

which could not be detected by XRD.

**3.2.4 Tris-(dimethylamino)borane (TDMAB)** 

thickness.

et al., 1997; Boudiombo et al., 1997; Abdellaoui et al., 1997).

substrate temperatures lead to the formation of hemispheres (approximately 10 μm in diameter) on the surface, thus increasing the roughness of the BCN layers (Gammer et al., 2002).

BCN-layers were deposited in a hot-filament supported reactor using TDMAB (Weissenbacher et al., 2002, 2006). These experiments were strongly influenced by the stability of the Ta-filament. At filament temperatures of 1500°C layer deposition on the surface of the filament takes place, at temperatures higher than 2000°C liquid phase formation led to filament breakdown in many cases. The morphology of the deposited BCN layers on hard metal substrates (WC-Co) depends on the deposition conditions and films contain high amounts of tantalum (BCN-Ta layers).

Methyl-BCN lms were deposited by plasma-assisted CVD (PACVD) using mixture of TDMAB and N2 at 350°C (Aoki et al., 2010a, 2010b).

Hexagonal boron carbonitride hybrid lms - sp2-BCN phase with h-BN-like conguration have been synthesized on Si(100) and on highly oriented pyrolytic graphite, respectively, by RF-PACVD using TDMAB (Mannan et al., 2009a, 2009b, 2010). The deposition was performed at different RF powers of 400–800W, at the working pressure of 2×10-1 Torr and temperature 650-750°C.

The influence of the B/C/N containing single-source precursors pyridine-borane (PB) and triazaborabicyclodecane (TBBD) on the chemical composition of boron carbonitride thin films was investigated in (Hegemann et al., 1997, 1999). The films are deposited via a PACVD process, activated by 13.56 MHz radio frequency (RF). N2, Ar and He serve as carrier gases. It becomes evident, that from a certain bias voltage, the self-bias in capacitively RF electrical discharges mainly influences the chemical composition of the BCN films independent of the kind of the used precursor. Films that were either deposited in He using a low power density or in N2 using a high power density showed comparable properties. Analysis of these films showed their chemical composition to be BC4N.

#### **3.2.5 (N-pyrrolidino)-diethylborate (PEB)**

BCN films were deposited on polycarbonate and silicon wafer by means of different RF PACVD (inductively coupled and capacitively coupled RF PACVD), by use of liquid organic compound (N-pyrrolidino)diethylborane (C8H18BN) as precursors. Deposition was carried out on at 95–120°C. A mixture of argon, hydrogen, and nitrogen was used as process gas. The layer shows a columnar structure. The composition of BCN lms deposited ranged between BC7.3N0.8 and BC0.9N0.6 (Wöhle et al., 1999; Ahn et al., 2003).

#### **3.2.6 Borazine derivatives**

Amorphous semiconductor BCN films were produced by means of pyrolysis of borazine derivatives (tris(1,3,2-benzodiazaborolo)borazine) at 1073 K (Maya, 1988a, 1988b; Maya, & Harris, 1990). Quartz, titanium and silicon were used as substrates. Chemical analysis showed that the synthesized material had the composition BC5.2N1.8H1.9O0.45. The density of the coating was 2.05 – 2.09 g/cm3.

#### **3.2.6.1 Trimethylborazine**

Mixtures of *N,N',N''*-trimethylborazine = (CH3)3N3B3H3 (TMB) (the B:C:N ratio 1:1:1) and argon have been used to deposit BCN:H films by means of ECR PECVD processes at 100- 150°C (Weber et al., 1992, 1993). Amorphous BCN layers were deposited on a polycarbonate substrate by RF PACVD at low temperature (95-120°C) using a mixture of TMB and H2, N2, or Ar (Wöhle et al., 1999). The composition of the layers varied in a wide range. The boron

Compilation on Synthesis, Characterization

**4. Physical and chemical characterization** 

physical and/or chemical charcterization of the products.

**4.1 Ellipsometry** 

**4.2 Indentation** 

Wikipedia, and from other not-authorizised sources in the internet.

Kosinova et al., 1999; Hoffmann et al., 2007; Baake et al., 2010).

and Properties of Silicon and Boron Carbonitride Films 505

of hydrogen atmosphere compared to the experiments in argon atmosphere. This can be explained by etching reactions of the atomic hydrogen. Due to the etching reactions of the atomic hydrogen the IR-spectra also look different. Nevertheless, it was not possible to identify a cubic phase from the IR or XRD measurements (Weissenbacher et al., 2003).

Mainly for judging the synthesis success, the products have to be characterized for physical and chemical properties. The analytical procedures are quite different, belonging to the interest of the scientists. One part is interested in the physical parameters, e.g., as mechanical (e.g., dimension, hardness, roughness), optical (e.g., reflectivity, absorption coefficient), thermal (e.g., conductivity), and electrical and electronic (e.g., resistivity, conductivity, band gap) properties. An other part examines the chemical properties, e.g., the elemental composition, the phase state (e.g., crystallinity, amorphous state), the character of chemical bonds (e.g., participation of the elements, single/double/triple, energy, distances, ionic/covalent/metal/van der Waals), and the chemical reactivity (e.g., inertness).

The first citation of a carbonitride, to our knowledge, was given by a Soviet group (Dubovik & Struk, 1965). They referred to papers, published earlier (Samsonov et al., 1962; Chelepenkouv et al., 1964). These scientists synthesized the product by nitriding of boron carbide and were interested in an excellent thermal resistance. The test of the material as crucibles at 1600- 2000°C was successful, but a fundamental characterization was not performed. This situation changed a little later, when the elemental composition and structure of carbonitrides of different productions were determined (Kosolapova et al., 1971). Such information was derived from X-ray and electron diffraction measurements. Later on, nearly all papers dealing with the synthesis of silicon or boron carbonitrides contain a voluminous chapter with

The description of the methods is partly taken from text books, lecture notes, application notes of manufacturers of analytical instruments, from the internet encyclopaedia

Ellipsometry is an optical method in material science and surface physics (Fujiwara, 2007; Tompkins, 2006). It permits to determine the real and the imaginary part of the complex dielectrical refractory index and of the thickness of thin layers, as well. It can be used for various materials, as organic and inorganic substances (metals, semiconductors, insulators). Ellipsometry is working in a wide range: UV, visible part, and far in the IR (THz range) light. By means of this method the change of the polarization state at reflection or transmission is determined. This change is given by the relation of the reflection coefficients vertical *rs* and parallel *rp* to the incidence plane of the polarized light. In nearly all cases a model analysis is carried out, in which the optical constants are described by a model dielectric function. In all former papers of the authors the absorption coefficient, refractory index, and the thickness of the silicon and boron carbonitride layers are determined by ellipsometry (Fainer et al., 1997;

The hardness of the synthesized products is determined by usual methods, described in text books. The so called Mohs` hardness (mainly for minerals) is divided in a graduation

Obviously, there is a relation between the physical and the chemical parameters.

content of the lms ranged from 3at% to 42at%, the carbon content from 16at.% to 80at.%, and oxygen content from 2at.% to 10 at.%. BCN films with hexagonal turbostratic graphite like structure were deposited by both isothermal chemical vapour deposition (ITCVD) under atmospheric pressure and PECVD from gaseous mixtures of trimethylborazine, toluene and ammonia at 950°C (Stockel, 2002, 2003). Parallels between ITCVD and PECVD films emerged in the case of chemical composition and the correlation between carbon content and hardness values. Considerable differences exist with regard to the microstructure, especially the texture of the films. Moreover in ITCVD films the carbon is preferentially incorporated between the BN basal planes, whereas in PECVD films it is incorporated preferentially in between the BN basal planes as well. BCN coatings were deposited on polycarbonate by means of a capacitively or inductively coupled RF-PACVD (capacitive coupled plasma (CCP) vs. inductively coupled plasma (ICP)) using the elemental-organic compound trimethylborazine as precursor. The influence of the plasma parameters on the properties of films has been discussed (Ahn et al., 2003).

BCN:H lms on silicon substrates were deposited with two different PECVD techniques (Thamm et al., 2005, 2007). A microwave plasma with RF-bias enhancement (MW-PECVD) and a direct current glow discharge plasma system (GD-PECVD) was used with TMB and benzene as an additional carbon source. Argon and nitrogen were used as plasma gases. Substrate temperature, substrate bias and gas composition were varied. The hydrogen content in the produced BCN layers strongly depends on the substrate temperature and increases up to 35at.%. Depth proles show a homogeneous distribution of the elements B, C, N, and H over the entire layer thickness. Impurities such as oxygen or argon are detected in only small quantities (below 0.5at.%) and the concentration does not increase towards the surface. The hydrogen content mostly depends on the substrate temperature during the coating process. If the layers are deposited on 50°C substrate temperature (MW-PECVD) the hydrogen content increases up to 35at.%. If the temperature is increased up to 800°C, only 8at.% hydrogen are detected, independently of the plasma gas. The variation of the MW plasma power and RF-bias have no significant effects on the layer composition. Multicomponent films were grown by PECVD from N-trimethylborazine–nitrogen mixtures at temperatures from 373 to 973 K and varied RF power. According to XPS and IR spectroscopy results, the major component of the films is boron nitride. The films grown at temperatures below 673 K contain hydrogen. The higher temperature films contain carbon (Sulyaeva et al., 2007, 2009, 2010).

#### **3.2.6.2 Triethylborazine (TEB) and Tripropylborazine (TPB)**

Thermolysis of N-triethylborazine = (C2H3)N3B3H3 (TEB) and N-tripropylborazine = (C3H7)N3B3H3 (TPB) at 500°C produces homogeneous, amorphous boron carbonitride phases, whose compositions are dependent upon the borazine substituent, and whose structures are similar to that of icosahedral boron carbide B4C (Brydson et al., 2002). The deposition of BCN-layers by decomposing TEB was performed in a hot-filament CVD apparatus and for the gas activation a carburized Ta wire was used. Untreated, etched and diamond coated WC–Co hardmetal inserts were used as substrates. Substrate temperature was 890°C. The deposition parameters filament temperature (1800–2200°C), precursor flow rate and gas atmosphere were varied. The layer growth rates increased with the precursor flow rate and showed less influence of the substrate material. Layer morphology was of the pyrolytic type. When the layers became thick they tended to eliminate from the substrate. Due to the atomic hydrogen produced at the filament the deposition rate decreased in case

content of the lms ranged from 3at% to 42at%, the carbon content from 16at.% to 80at.%, and oxygen content from 2at.% to 10 at.%. BCN films with hexagonal turbostratic graphite like structure were deposited by both isothermal chemical vapour deposition (ITCVD) under atmospheric pressure and PECVD from gaseous mixtures of trimethylborazine, toluene and ammonia at 950°C (Stockel, 2002, 2003). Parallels between ITCVD and PECVD films emerged in the case of chemical composition and the correlation between carbon content and hardness values. Considerable differences exist with regard to the microstructure, especially the texture of the films. Moreover in ITCVD films the carbon is preferentially incorporated between the BN basal planes, whereas in PECVD films it is incorporated preferentially in between the BN basal planes as well. BCN coatings were deposited on polycarbonate by means of a capacitively or inductively coupled RF-PACVD (capacitive coupled plasma (CCP) vs. inductively coupled plasma (ICP)) using the elemental-organic compound trimethylborazine as precursor. The influence of the plasma

BCN:H lms on silicon substrates were deposited with two different PECVD techniques (Thamm et al., 2005, 2007). A microwave plasma with RF-bias enhancement (MW-PECVD) and a direct current glow discharge plasma system (GD-PECVD) was used with TMB and benzene as an additional carbon source. Argon and nitrogen were used as plasma gases. Substrate temperature, substrate bias and gas composition were varied. The hydrogen content in the produced BCN layers strongly depends on the substrate temperature and increases up to 35at.%. Depth proles show a homogeneous distribution of the elements B, C, N, and H over the entire layer thickness. Impurities such as oxygen or argon are detected in only small quantities (below 0.5at.%) and the concentration does not increase towards the surface. The hydrogen content mostly depends on the substrate temperature during the coating process. If the layers are deposited on 50°C substrate temperature (MW-PECVD) the hydrogen content increases up to 35at.%. If the temperature is increased up to 800°C, only 8at.% hydrogen are detected, independently of the plasma gas. The variation of the MW plasma power and RF-bias have no significant effects on the layer composition. Multicomponent films were grown by PECVD from N-trimethylborazine–nitrogen mixtures at temperatures from 373 to 973 K and varied RF power. According to XPS and IR spectroscopy results, the major component of the films is boron nitride. The films grown at temperatures below 673 K contain hydrogen. The higher temperature films contain carbon

Thermolysis of N-triethylborazine = (C2H3)N3B3H3 (TEB) and N-tripropylborazine = (C3H7)N3B3H3 (TPB) at 500°C produces homogeneous, amorphous boron carbonitride phases, whose compositions are dependent upon the borazine substituent, and whose structures are similar to that of icosahedral boron carbide B4C (Brydson et al., 2002). The deposition of BCN-layers by decomposing TEB was performed in a hot-filament CVD apparatus and for the gas activation a carburized Ta wire was used. Untreated, etched and diamond coated WC–Co hardmetal inserts were used as substrates. Substrate temperature was 890°C. The deposition parameters filament temperature (1800–2200°C), precursor flow rate and gas atmosphere were varied. The layer growth rates increased with the precursor flow rate and showed less influence of the substrate material. Layer morphology was of the pyrolytic type. When the layers became thick they tended to eliminate from the substrate. Due to the atomic hydrogen produced at the filament the deposition rate decreased in case

parameters on the properties of films has been discussed (Ahn et al., 2003).

(Sulyaeva et al., 2007, 2009, 2010).

**3.2.6.2 Triethylborazine (TEB) and Tripropylborazine (TPB)** 

of hydrogen atmosphere compared to the experiments in argon atmosphere. This can be explained by etching reactions of the atomic hydrogen. Due to the etching reactions of the atomic hydrogen the IR-spectra also look different. Nevertheless, it was not possible to identify a cubic phase from the IR or XRD measurements (Weissenbacher et al., 2003).

### **4. Physical and chemical characterization**

Mainly for judging the synthesis success, the products have to be characterized for physical and chemical properties. The analytical procedures are quite different, belonging to the interest of the scientists. One part is interested in the physical parameters, e.g., as mechanical (e.g., dimension, hardness, roughness), optical (e.g., reflectivity, absorption coefficient), thermal (e.g., conductivity), and electrical and electronic (e.g., resistivity, conductivity, band gap) properties. An other part examines the chemical properties, e.g., the elemental composition, the phase state (e.g., crystallinity, amorphous state), the character of chemical bonds (e.g., participation of the elements, single/double/triple, energy, distances, ionic/covalent/metal/van der Waals), and the chemical reactivity (e.g., inertness). Obviously, there is a relation between the physical and the chemical parameters.

The first citation of a carbonitride, to our knowledge, was given by a Soviet group (Dubovik & Struk, 1965). They referred to papers, published earlier (Samsonov et al., 1962; Chelepenkouv et al., 1964). These scientists synthesized the product by nitriding of boron carbide and were interested in an excellent thermal resistance. The test of the material as crucibles at 1600- 2000°C was successful, but a fundamental characterization was not performed. This situation changed a little later, when the elemental composition and structure of carbonitrides of different productions were determined (Kosolapova et al., 1971). Such information was derived from X-ray and electron diffraction measurements. Later on, nearly all papers dealing with the synthesis of silicon or boron carbonitrides contain a voluminous chapter with physical and/or chemical charcterization of the products.

The description of the methods is partly taken from text books, lecture notes, application notes of manufacturers of analytical instruments, from the internet encyclopaedia Wikipedia, and from other not-authorizised sources in the internet.

#### **4.1 Ellipsometry**

Ellipsometry is an optical method in material science and surface physics (Fujiwara, 2007; Tompkins, 2006). It permits to determine the real and the imaginary part of the complex dielectrical refractory index and of the thickness of thin layers, as well. It can be used for various materials, as organic and inorganic substances (metals, semiconductors, insulators). Ellipsometry is working in a wide range: UV, visible part, and far in the IR (THz range) light. By means of this method the change of the polarization state at reflection or transmission is determined. This change is given by the relation of the reflection coefficients vertical *rs* and parallel *rp* to the incidence plane of the polarized light. In nearly all cases a model analysis is carried out, in which the optical constants are described by a model dielectric function. In all former papers of the authors the absorption coefficient, refractory index, and the thickness of the silicon and boron carbonitride layers are determined by ellipsometry (Fainer et al., 1997; Kosinova et al., 1999; Hoffmann et al., 2007; Baake et al., 2010).

#### **4.2 Indentation**

The hardness of the synthesized products is determined by usual methods, described in text books. The so called Mohs` hardness (mainly for minerals) is divided in a graduation

Compilation on Synthesis, Characterization

**4.4 Atomic Force Microscopy (AFM)** 

Sugimoto et al., 2007).

**4.5 X-ray (XRD), electron, and neutron diffraction** 

the atomic structure of new materials.

2d .

and Properties of Silicon and Boron Carbonitride Films 507

 Fig. 1. Typical SEM images of surfaces of PECVD SiCxNy film (left) and BCxNy (right).

The atomic force microscopy (AFM) or scanning force microscopy (SFM) provides a three-dimensional profile of the sample surface (in comparison to SEM, which works two-dimensional). It is of very high-resolution on the order of fractions of a nanometer. The AFM consists of a cantilever with a sharp tip (silicon or silicon nitride) to scan the sample surface. When the probe is brought to the proximity of the specimen mechanical contact force, van der Waals forces, capillary forces, chemical bonding, electrostatic and magnetic forces, Casimir forces, and/or solvation forces influence the deflection of the cantilever. The deflection is, typically, measured by a laser spot reflected from the top of the cantilever. In best cases, individual surface atoms can be identified. The advantages of AFM are in comparison to SEM: The receiving of a three-dimensional image, the sample viewing without any treatments (coatings), working perfectly in ambient air, giving true atomic resolution in ultra-high vacuum (and in liquid environment). The disadvantages are: Single scan image size in the micrometer range (height: 10-20 µm, scanning area: 150x150 µm2), and a limitation in the scanning speed (Giessibl, 2003;

The diffraction methods are used to determine the arrangement of atoms within a crystal. That directly means that the method is not suitable for amorphous materials. Beneath the structure and phases, a quantitative determination of the elemental composition can be derived. Therefore, X-ray diffraction was first used for the characterization of carbonitrides (Kosolapova et al., 1971). X-ray crystallography is called the chief method for characterizing

A regular array of scatterers (crystal) produces a regular array of spherical waves. In most directions these waves cancel one another out through destructive interference, they add

where d = spacing between diffraction planes, Θ=incident angle, n=any integer, and λ=wavelength of the beam. That is successful because the wavelength λ is typically of the

sin Θ = n . λ, (8)

constructively in a few specific directions, determined by Bragg´s law (8)

scale of 10 parts. These parts are defined by the materials which scratch the surface of a sample. The indentation, and mainly the nano-indentation is a local (positional) highresolution method for the characterization of thin solid samples. The method is based on the measurement of force and path during an elastic/plastic contact of a hard tester (indenter) with the sample. The advantages of this procedure are: Extremely highresolution lateral and in the depth, no optical measurement at indentation is necessary, simultaneous determination of the hardness and of the Young´s modulus, generation and measurement of defined scratches. In most cases, a diamond tip is pressed into the sample with increasing force. Then the force is deceased to zero. First, an elastic deformation is observed, followed by a plastic deformation at higher force. Only the elastic deformation is traced back during the unloading. Several methods are given to evaluate the data. In most cases, the contact stiffness is determined at the beginning of the unloading (Bauer-Marschallinger et al., 2007).

#### **4.3 Scanning electron microscopy (SEM)**

In this type of an electron microscope the sample is scanned by a high-energy beam of electrons. Utilizable information are mainly derived from signals of secondary electrons (SE), of back-scattered electrons (BSE), and of characteristic X-rays. The secondary electrons can produce very high-resolved images of the sample surface, showing details less than 5 nm in size. Because the intensity of the BSE signal is strongly related to the atomic number (Z) of the specimen, BSE images provide information about the lateral distribution of different elements. In case of relaxation of the excited atoms in the sample, characteristic X-rays are emitted. These are used to determine the elemental composition and to measure the abundance of the elements in the sample. The information depth is related to the primary electron energy (0.5 keV – 40 keV), to the mean atomic number and the density of the sample, and to the energy of the emitted X-rays (Moseley´s law gives the relation between energy and atomic number). The electrons typically are thermoionically emitted from a tungsten (W) or a lanthanum hexaboride (LaB6) cathode. For conventional imaging in SEM, the specimen must be electrically conductive. Nonconductive specimens, e.g. ceramics, tend to charge and, therefore, have to be coated by an ultrathin conducting layer (e.g., gold, carbon). Back scattered electron imaging, quantitative X-ray analysis, and X-ray mapping of materials (e.g., metals, ceramics, geological) require that the surfaces have to be ground and polished to an ultra smooth surface. Special high resolution coating techniques are used for high magnification imaging of inorganic thin films. As an example, SEM images of carbonitride layers are given in Fig. 1.

The most common imaging mode collects the low-energy (< 50 eV) secondary electrons (SE), which originate from a thin (few nanometers) surface layer of the sample. These electrons are biased to about + 2 kV and detected by scintillator-photomultiplier systems. The highenergetic back-scattered or reflected electrons are usually measured by detectors of the scintillation or semiconductor types. The X-rays are detected by wavelength-(WDXRS) or energy-dispersive (EDXRS) systems. Expended systems contain several WDXRS detectors (arranged vertically and horizontally) and one EDXRS detector – so called electron micro probe analyzer (EMPA). More detailed information is given in the text books and in special papers. As introduction the monography of Goldstein et al. is suggested (Goldstein et al., 1981).

scale of 10 parts. These parts are defined by the materials which scratch the surface of a sample. The indentation, and mainly the nano-indentation is a local (positional) highresolution method for the characterization of thin solid samples. The method is based on the measurement of force and path during an elastic/plastic contact of a hard tester (indenter) with the sample. The advantages of this procedure are: Extremely highresolution lateral and in the depth, no optical measurement at indentation is necessary, simultaneous determination of the hardness and of the Young´s modulus, generation and measurement of defined scratches. In most cases, a diamond tip is pressed into the sample with increasing force. Then the force is deceased to zero. First, an elastic deformation is observed, followed by a plastic deformation at higher force. Only the elastic deformation is traced back during the unloading. Several methods are given to evaluate the data. In most cases, the contact stiffness is determined at the beginning of the unloading (Bauer-

In this type of an electron microscope the sample is scanned by a high-energy beam of electrons. Utilizable information are mainly derived from signals of secondary electrons (SE), of back-scattered electrons (BSE), and of characteristic X-rays. The secondary electrons can produce very high-resolved images of the sample surface, showing details less than 5 nm in size. Because the intensity of the BSE signal is strongly related to the atomic number (Z) of the specimen, BSE images provide information about the lateral distribution of different elements. In case of relaxation of the excited atoms in the sample, characteristic X-rays are emitted. These are used to determine the elemental composition and to measure the abundance of the elements in the sample. The information depth is related to the primary electron energy (0.5 keV – 40 keV), to the mean atomic number and the density of the sample, and to the energy of the emitted X-rays (Moseley´s law gives the relation between energy and atomic number). The electrons typically are thermoionically emitted from a tungsten (W) or a lanthanum hexaboride (LaB6) cathode. For conventional imaging in SEM, the specimen must be electrically conductive. Nonconductive specimens, e.g. ceramics, tend to charge and, therefore, have to be coated by an ultrathin conducting layer (e.g., gold, carbon). Back scattered electron imaging, quantitative X-ray analysis, and X-ray mapping of materials (e.g., metals, ceramics, geological) require that the surfaces have to be ground and polished to an ultra smooth surface. Special high resolution coating techniques are used for high magnification imaging of inorganic thin films. As an example, SEM images of carbonitride layers are

The most common imaging mode collects the low-energy (< 50 eV) secondary electrons (SE), which originate from a thin (few nanometers) surface layer of the sample. These electrons are biased to about + 2 kV and detected by scintillator-photomultiplier systems. The highenergetic back-scattered or reflected electrons are usually measured by detectors of the scintillation or semiconductor types. The X-rays are detected by wavelength-(WDXRS) or energy-dispersive (EDXRS) systems. Expended systems contain several WDXRS detectors (arranged vertically and horizontally) and one EDXRS detector – so called electron micro probe analyzer (EMPA). More detailed information is given in the text books and in special papers. As introduction the monography of Goldstein et al. is suggested (Goldstein et al.,

Marschallinger et al., 2007).

given in Fig. 1.

1981).

**4.3 Scanning electron microscopy (SEM)** 

Fig. 1. Typical SEM images of surfaces of PECVD SiCxNy film (left) and BCxNy (right).

#### **4.4 Atomic Force Microscopy (AFM)**

The atomic force microscopy (AFM) or scanning force microscopy (SFM) provides a three-dimensional profile of the sample surface (in comparison to SEM, which works two-dimensional). It is of very high-resolution on the order of fractions of a nanometer. The AFM consists of a cantilever with a sharp tip (silicon or silicon nitride) to scan the sample surface. When the probe is brought to the proximity of the specimen mechanical contact force, van der Waals forces, capillary forces, chemical bonding, electrostatic and magnetic forces, Casimir forces, and/or solvation forces influence the deflection of the cantilever. The deflection is, typically, measured by a laser spot reflected from the top of the cantilever. In best cases, individual surface atoms can be identified. The advantages of AFM are in comparison to SEM: The receiving of a three-dimensional image, the sample viewing without any treatments (coatings), working perfectly in ambient air, giving true atomic resolution in ultra-high vacuum (and in liquid environment). The disadvantages are: Single scan image size in the micrometer range (height: 10-20 µm, scanning area: 150x150 µm2), and a limitation in the scanning speed (Giessibl, 2003; Sugimoto et al., 2007).

#### **4.5 X-ray (XRD), electron, and neutron diffraction**

The diffraction methods are used to determine the arrangement of atoms within a crystal. That directly means that the method is not suitable for amorphous materials. Beneath the structure and phases, a quantitative determination of the elemental composition can be derived. Therefore, X-ray diffraction was first used for the characterization of carbonitrides (Kosolapova et al., 1971). X-ray crystallography is called the chief method for characterizing the atomic structure of new materials.

A regular array of scatterers (crystal) produces a regular array of spherical waves. In most directions these waves cancel one another out through destructive interference, they add constructively in a few specific directions, determined by Bragg´s law (8)

$$\mathbf{2d} \cdot \sin \Theta = \mathbf{n} \cdot \lambda \,\tag{8}$$

where d = spacing between diffraction planes, Θ=incident angle, n=any integer, and λ=wavelength of the beam. That is successful because the wavelength λ is typically of the

Compilation on Synthesis, Characterization

and their sizes are relatively large.

**4.6 X-ray Reflectometry (XRR)** 

**4.8 Infrared Spectroscopy (IR)** 

and Properties of Silicon and Boron Carbonitride Films 509

complex structures of (nanocrystalline) intermetallic compounds and zeolites. Electron diffraction is used in solid state physics and chemistry in devices as transmission electron

For structural determination of light elements neutron diffraction is the favoured method (Ibberson & David, 2002). For application of monochromatic, intense beams nuclear reactors and spallation neutron sources are used. Neutrons being uncharged scatter much more readily from the atomic nuclei than from the electrons. As mentioned, neutron scattering is very useful for observing the positions of light elements, especially hydrogen, which is essentially invisible in the X-ray diffraction. The samples are mostly exposed as powders

X-ray reflectometry is a highly accurate method for the determination of layer thickness (Als-Nielsen, 2001). The reflectance of the sample is measured as a function of the grazing incidence angle of X-rays. Due to interference effects of the radiation from each layer, oscillations are observed in the reflectance curve. The oscillation period is mainly determined by the layer thickness, the oscillation amplitude depends on the densities and on the surface roughness. The reflectance curve can be fitted according to the Fresnel equations. XRR can provide information on the thickness, roughness and density of thin films on a substrate. Using

Nuclear magnetic resonance is one of the most important methods for the determination of the molecular structure of mostly organic and organometallic, rarely of inorganic compounds (Keeler 2005). All stable isotopes that contain an odd number of protons and/or neutrons have an intrinsic magnetic moment. The most commonly studied nuclei are 1H, 13C, 19F, and 31P. The liquid or solid sample is placed between very narrow positioned poles of a huge magnet. The magnet produces a field of extremely high stability and homogeneity. This field is varied by a so called sweep coil. In the measurements the applied magnetic field is reduced (absorbed) by induction effects, yielding an "effective field strength". This effect is also called "shielding" and varies with the electron distribution of atoms. The change of a resonance line due to binding effects is called the chemical shift. The spectra are recorded in relation to the proton signal of tetramethylsilane = Si(CH3)4 (TMS). An example of NMR being used in the determination of a structure is that of buckminsterfullerene (C60). As this structure contains no hydrogen, 13C NMR has to be used (longer acquisition time since 13C is not the common isotope of carbon). However, the spectrum was obtained and was found to

synchrotron radiation layers can be distinguished with similar optical constants.

exhibit a single peak, confirming the unusual structure of C60 (Taylor et al., 1990).

The infrared spectroscopy deals with the electromagnetic spectrum of the near- (14000-4000 cm-1; 0.8-2.5 µm; to excite overtone or harmonic vibrations) , mid- (4000-400 cm-1; 2.5-25 µm; to study fundamental vibrations and associated rotational-vibrational structures) and far-infrared (400-10 cm-1; 25-1000 µm; used for rotational spectroscopy) regions; this light has a longer wavelength (lower frequency) than visible light. IR spectroscopy has been applied successfully for characterization in organic and inorganic compounds. Also it was successfully utilized in the field of semiconductor microelectronics (Lau 1999). Molecules absorb specific frequencies

**4.7 Nuclear Magnetic Resonance spectroscopy (NMR)** 

microscopy (TEM, see 4.13) or scanning electron microscopy (SEM, see 4.3).

same order of magnitude as the spacing d between planes of the crystal (1-100 Ǻ). The initial studies revealed the typical radii of atoms, and confirmed many theoretical models of chemical bonding, such as the tetrahedral bonding of carbon in the diamond structure (Bragg & Bragg, 1913). In material sciences, many complicated systems (e.g., fullerenes) were analyzed using single-crystal methods. The Cambridge Structural Database contains over 500,000 structures; over 99 % of them were determined by X-ray diffraction.

In laboratories X-ray diffraction devices contain X-ray tubes (X-ray generators) as sources. Electrons are thermally emitted from a metal and extracted through a strong electric potential (e.g., 50 kV) and directed to a metal plate (mostly Cu). Thereby, bremsstrahlung and some characteristic lines are emitted. Usually, the X-ray tube has a stationary anode (2 kW). For application of an intensive beam a rotating anode (14 kW) is used. The brightest and most useful X-ray source is a synchrotron. These systems allow a better resolution, and they make it convenient to tune the wavelength (energy) of the radiation. An XRD diffraction diagram recorded at a synchrotron beam is given in Fig. 2.

Fig. 2. The typical XRD-SR pattern of nanocomposite SiCxNy film shows that the peaks' position is close to standard α-Si3N4 phase. There are some unknown peaks at small diffraction angles indicated by "?", which do not correspond to peaks positions of other known phases of the Si-C-N system.

The intensities of the reflections are recorded by photographic film, area detector or by a charge-coupled device (CCD) image sensor. To collect the total information, the crystal must be rotated step-by-step through 180° (Θ/2Θ-geometry). As an additional method grazing incidence XRD is used.

Electrons interact with positively charged atomic nuclei and with the surrounding electrons, whereas X-rays interact with valence electrons only, and neutrons scatter by atomic nuclei. Whereas X-rays interact relatively weakly with the electrons, for some applications electron beams are used to examine relatively thin crystals (> 100 nm). The strong interaction of charged electrons with matter by Coulomb forces (about 1000 times stronger than for Xrays) allows determination of the atomic structure of extremely small volumes. The field of application of electron diffraction ranges from bio molecules over organic thin films to the complex structures of (nanocrystalline) intermetallic compounds and zeolites. Electron diffraction is used in solid state physics and chemistry in devices as transmission electron microscopy (TEM, see 4.13) or scanning electron microscopy (SEM, see 4.3).

For structural determination of light elements neutron diffraction is the favoured method (Ibberson & David, 2002). For application of monochromatic, intense beams nuclear reactors and spallation neutron sources are used. Neutrons being uncharged scatter much more readily from the atomic nuclei than from the electrons. As mentioned, neutron scattering is very useful for observing the positions of light elements, especially hydrogen, which is essentially invisible in the X-ray diffraction. The samples are mostly exposed as powders and their sizes are relatively large.

#### **4.6 X-ray Reflectometry (XRR)**

508 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

same order of magnitude as the spacing d between planes of the crystal (1-100 Ǻ). The initial studies revealed the typical radii of atoms, and confirmed many theoretical models of chemical bonding, such as the tetrahedral bonding of carbon in the diamond structure (Bragg & Bragg, 1913). In material sciences, many complicated systems (e.g., fullerenes) were analyzed using single-crystal methods. The Cambridge Structural Database contains

In laboratories X-ray diffraction devices contain X-ray tubes (X-ray generators) as sources. Electrons are thermally emitted from a metal and extracted through a strong electric potential (e.g., 50 kV) and directed to a metal plate (mostly Cu). Thereby, bremsstrahlung and some characteristic lines are emitted. Usually, the X-ray tube has a stationary anode (2 kW). For application of an intensive beam a rotating anode (14 kW) is used. The brightest and most useful X-ray source is a synchrotron. These systems allow a better resolution, and they make it convenient to tune the wavelength (energy) of the radiation. An XRD

Fig. 2. The typical XRD-SR pattern of nanocomposite SiCxNy film shows that the peaks' position is close to standard α-Si3N4 phase. There are some unknown peaks at small diffraction angles indicated by "?", which do not correspond to peaks positions of other

The intensities of the reflections are recorded by photographic film, area detector or by a charge-coupled device (CCD) image sensor. To collect the total information, the crystal must be rotated step-by-step through 180° (Θ/2Θ-geometry). As an additional method grazing

Electrons interact with positively charged atomic nuclei and with the surrounding electrons, whereas X-rays interact with valence electrons only, and neutrons scatter by atomic nuclei. Whereas X-rays interact relatively weakly with the electrons, for some applications electron beams are used to examine relatively thin crystals (> 100 nm). The strong interaction of charged electrons with matter by Coulomb forces (about 1000 times stronger than for Xrays) allows determination of the atomic structure of extremely small volumes. The field of application of electron diffraction ranges from bio molecules over organic thin films to the

known phases of the Si-C-N system.

incidence XRD is used.

over 500,000 structures; over 99 % of them were determined by X-ray diffraction.

diffraction diagram recorded at a synchrotron beam is given in Fig. 2.

X-ray reflectometry is a highly accurate method for the determination of layer thickness (Als-Nielsen, 2001). The reflectance of the sample is measured as a function of the grazing incidence angle of X-rays. Due to interference effects of the radiation from each layer, oscillations are observed in the reflectance curve. The oscillation period is mainly determined by the layer thickness, the oscillation amplitude depends on the densities and on the surface roughness. The reflectance curve can be fitted according to the Fresnel equations. XRR can provide information on the thickness, roughness and density of thin films on a substrate. Using synchrotron radiation layers can be distinguished with similar optical constants.

#### **4.7 Nuclear Magnetic Resonance spectroscopy (NMR)**

Nuclear magnetic resonance is one of the most important methods for the determination of the molecular structure of mostly organic and organometallic, rarely of inorganic compounds (Keeler 2005). All stable isotopes that contain an odd number of protons and/or neutrons have an intrinsic magnetic moment. The most commonly studied nuclei are 1H, 13C, 19F, and 31P. The liquid or solid sample is placed between very narrow positioned poles of a huge magnet. The magnet produces a field of extremely high stability and homogeneity. This field is varied by a so called sweep coil. In the measurements the applied magnetic field is reduced (absorbed) by induction effects, yielding an "effective field strength". This effect is also called "shielding" and varies with the electron distribution of atoms. The change of a resonance line due to binding effects is called the chemical shift. The spectra are recorded in relation to the proton signal of tetramethylsilane = Si(CH3)4 (TMS). An example of NMR being used in the determination of a structure is that of buckminsterfullerene (C60). As this structure contains no hydrogen, 13C NMR has to be used (longer acquisition time since 13C is not the common isotope of carbon). However, the spectrum was obtained and was found to exhibit a single peak, confirming the unusual structure of C60 (Taylor et al., 1990).

#### **4.8 Infrared Spectroscopy (IR)**

The infrared spectroscopy deals with the electromagnetic spectrum of the near- (14000-4000 cm-1; 0.8-2.5 µm; to excite overtone or harmonic vibrations) , mid- (4000-400 cm-1; 2.5-25 µm; to study fundamental vibrations and associated rotational-vibrational structures) and far-infrared (400-10 cm-1; 25-1000 µm; used for rotational spectroscopy) regions; this light has a longer wavelength (lower frequency) than visible light. IR spectroscopy has been applied successfully for characterization in organic and inorganic compounds. Also it was successfully utilized in the field of semiconductor microelectronics (Lau 1999). Molecules absorb specific frequencies

Compilation on Synthesis, Characterization

(right).

and Properties of Silicon and Boron Carbonitride Films 511

Fig. 4. Typical Raman spectra of PECVD SiCxNy film grown at 373-673 K (left) and 873-973 K

The Raman spectra provide a fingerprint by which the molecule can be identified. In solid state physics Raman spectroscopy characterize materials and find the crystallographic orientation of a sample. Raman spectra of a carbonitride samples are shown in Fig. 4. In nanotechnology, a Raman microscope can be used to analyze nanowires to better understand the composition of the structures. Variations of Raman spectroscopy have been developed. The usual purpose is to enhance the sensitivity (e.g., surface-enhanced Raman), to improve the statial resolution (Raman microscopy), or to acquire very specific

For excitation of atoms to emit X-rays, radiation of higher energy is required. For X-ray fluorescence (XRF) the emission of X-ray tubes, of radionuclides, or of synchrotron radiation is applied (Beckhoff et al.(Eds.), 2006; Van Grieken & Markowicz (Eds.), 2002). For the excitation of X-rays in SEM (see 4.3) electrons are used. The principle structure of a XRF device to determine the elemental composition is quite simple: source – sample – detector. The main properties of XRF systems are: simple element identification by Moseley´s law; to be applied for non-destructive analysis of solid samples (compact or powders); simple conditions for the elemental region 20<Z>92; for light elements vacuum conditions, windowless detectors and special sources are needed; the detection limit is in the range 10-4-10-7 g/g is related to the device, the source, the detector, the element, and the composition of the sample. The information depth depends on the incident energy, the composition of the sample, and the energy of the excited X-radiation – it varies from the µm-range to some cm-range. For a quantitative determination, calibration procedures using standard materials of a composition similar to the unknown sample or an evaluation using fundamental parameter are used. The principle of XRF is the excitation of electrons to a higher energy level (their relaxation leads to the emission of X-rays). These X-rays are divided in K-, L-, and M series, depending on the ground level of the electrons. Various cathodes (e.g., Cr, Pd, W) of the tubes or various radionuclides (55Fe, 109Cd, 241Am) are used for a high efficient excitation. The incident X-ray beam usually is directed in an angle of 45° on the sample. The very intensive synchrotron radiation has various advantages: Continuous spectrum from about 1 eV up to many 10 keV, and tunable for optimal excitation (resonance), highly collimated beam. Semiconductor diodes are applied for energy-dispersive detection (EDX). An EDX spectrum of the light elements in a

information (resonance Raman) (Lombardi & Birke, 2008).

**4.10 Wavelength and/or energy dispersive X-ray spectroscopy (WDX/EDX)** 

that are characteristic of their structures; these absorptions are at resonant frequencies. A molecule to be "IR active" must be associated with changes in the permanent dipole. Molecules with the number N atoms have many vibrational modes: linear have 3N-5 and nonlinear molecules have 3N-6 vibrational degrees of freedom (e.g., H2O has 3 degrees or modes). Stretching, bending, rocking, wagging and twisting vibrations were distinguished. Gaseous samples are measured in cells with a long path length of 5-10 cm. Liquid samples are placed between two plates of salt (e.g., NaCl, KBr, CaF). Solid samples are crushed with e.g. Nujol (oily agent) and smeared as a thin film onto salt plates. Sometimes microtomy cuts can be used. The techniques´ arrangement consists of a light source (e.g., Nernst glower, Globar = SiC rod, mercury vapour lamp), a monochromator (former: Prisms, today: Gratings) or interferometer (mainly for FTIR technique), the sample position, and the detection arrangement (thermocouple, Golay detector, PbS cell, PbSe cell, or other semiconductors). Thus absorption spectra will be recorded as function of the frequency. Today the Fouriertransform-technique is used mostly. It permits to analyze a large amount of data when the total spectrum is recorded simultaneously (Fellgett´s advantage or multiplex advantage). By this way both speed and signal-to-noise ratio are improved (White 1990). Fig. 3 exhibits the FTIR spectrum of a carbonitride sample.

Fig. 3. FTIR spectra of PECVD SiCxNy film grown at 973 K: total view and deconvolution of the main adsorption band into components.

#### **4.9 Raman spectroscopy**

Raman spectroscopy permits like IR spectroscopy to study vibrational and rotational modes of molecules (Gardiner, 1989). Whereas IR spectroscopy is mostly working in the absorption mode, the Raman scatter spectra are recorded in 90° geometry. Therefore both methods are complementary. Raman spectroscopy relies on inelastic (Raman) scattering of monochromatic (laser) light in the visible, near IR and near UV range, resulting in the energy of the laser photons being shifted down (Stokes lines) and up (Anti-Stokes lines). The light from the sample is collected with a lens, sent through a monochromator (holographic gratings, multiple dispersion stages, Czerny-Turner arrangements), and directed to a detector (photomultiplier, CCD camera).

that are characteristic of their structures; these absorptions are at resonant frequencies. A molecule to be "IR active" must be associated with changes in the permanent dipole. Molecules with the number N atoms have many vibrational modes: linear have 3N-5 and nonlinear molecules have 3N-6 vibrational degrees of freedom (e.g., H2O has 3 degrees or modes). Stretching, bending, rocking, wagging and twisting vibrations were distinguished. Gaseous samples are measured in cells with a long path length of 5-10 cm. Liquid samples are placed between two plates of salt (e.g., NaCl, KBr, CaF). Solid samples are crushed with e.g. Nujol (oily agent) and smeared as a thin film onto salt plates. Sometimes microtomy cuts can be used. The techniques´ arrangement consists of a light source (e.g., Nernst glower, Globar = SiC rod, mercury vapour lamp), a monochromator (former: Prisms, today: Gratings) or interferometer (mainly for FTIR technique), the sample position, and the detection arrangement (thermocouple, Golay detector, PbS cell, PbSe cell, or other semiconductors). Thus absorption spectra will be recorded as function of the frequency. Today the Fouriertransform-technique is used mostly. It permits to analyze a large amount of data when the total spectrum is recorded simultaneously (Fellgett´s advantage or multiplex advantage). By this way both speed and signal-to-noise ratio are improved (White 1990). Fig. 3 exhibits the

Fig. 3. FTIR spectra of PECVD SiCxNy film grown at 973 K: total view and deconvolution of

Raman spectroscopy permits like IR spectroscopy to study vibrational and rotational modes of molecules (Gardiner, 1989). Whereas IR spectroscopy is mostly working in the absorption mode, the Raman scatter spectra are recorded in 90° geometry. Therefore both methods are complementary. Raman spectroscopy relies on inelastic (Raman) scattering of monochromatic (laser) light in the visible, near IR and near UV range, resulting in the energy of the laser photons being shifted down (Stokes lines) and up (Anti-Stokes lines). The light from the sample is collected with a lens, sent through a monochromator (holographic gratings, multiple dispersion stages, Czerny-Turner arrangements), and directed to a

FTIR spectrum of a carbonitride sample.

the main adsorption band into components.

detector (photomultiplier, CCD camera).

**4.9 Raman spectroscopy** 

Fig. 4. Typical Raman spectra of PECVD SiCxNy film grown at 373-673 K (left) and 873-973 K (right).

The Raman spectra provide a fingerprint by which the molecule can be identified. In solid state physics Raman spectroscopy characterize materials and find the crystallographic orientation of a sample. Raman spectra of a carbonitride samples are shown in Fig. 4. In nanotechnology, a Raman microscope can be used to analyze nanowires to better understand the composition of the structures. Variations of Raman spectroscopy have been developed. The usual purpose is to enhance the sensitivity (e.g., surface-enhanced Raman), to improve the statial resolution (Raman microscopy), or to acquire very specific information (resonance Raman) (Lombardi & Birke, 2008).

#### **4.10 Wavelength and/or energy dispersive X-ray spectroscopy (WDX/EDX)**

For excitation of atoms to emit X-rays, radiation of higher energy is required. For X-ray fluorescence (XRF) the emission of X-ray tubes, of radionuclides, or of synchrotron radiation is applied (Beckhoff et al.(Eds.), 2006; Van Grieken & Markowicz (Eds.), 2002). For the excitation of X-rays in SEM (see 4.3) electrons are used. The principle structure of a XRF device to determine the elemental composition is quite simple: source – sample – detector. The main properties of XRF systems are: simple element identification by Moseley´s law; to be applied for non-destructive analysis of solid samples (compact or powders); simple conditions for the elemental region 20<Z>92; for light elements vacuum conditions, windowless detectors and special sources are needed; the detection limit is in the range 10-4-10-7 g/g is related to the device, the source, the detector, the element, and the composition of the sample. The information depth depends on the incident energy, the composition of the sample, and the energy of the excited X-radiation – it varies from the µm-range to some cm-range. For a quantitative determination, calibration procedures using standard materials of a composition similar to the unknown sample or an evaluation using fundamental parameter are used. The principle of XRF is the excitation of electrons to a higher energy level (their relaxation leads to the emission of X-rays). These X-rays are divided in K-, L-, and M series, depending on the ground level of the electrons. Various cathodes (e.g., Cr, Pd, W) of the tubes or various radionuclides (55Fe, 109Cd, 241Am) are used for a high efficient excitation. The incident X-ray beam usually is directed in an angle of 45° on the sample. The very intensive synchrotron radiation has various advantages: Continuous spectrum from about 1 eV up to many 10 keV, and tunable for optimal excitation (resonance), highly collimated beam. Semiconductor diodes are applied for energy-dispersive detection (EDX). An EDX spectrum of the light elements in a

Compilation on Synthesis, Characterization

Fig. 6a. NEXAFS spectra of Si3N4, SiC, and SiCxNy.

of these orbitals (Hemraj-Benny et al., 2006).

Fig. 6b. NEXAFS spectra of BN, B4C, and BCxNy.

**4.12 X-ray Photoelectron Spectroscopy (XPS)** 

and Properties of Silicon and Boron Carbonitride Films 513

In a total-reflection geometry (see Figs. 6a,b) at the sample, the information depth using photon detection decreases to about 5 nm (Baake et al., 2010). As an example, detailed spectral resonances at the carbon K-edge yield information about the bonding environment of this atom, such as functionalized species and chemisorbed impurities. The lower π\* resonance can provide insights into bond hybridization, while the σ\* resonance is a measure of the intramolecular bond length. Since the light from the synchrotron source used is linearly polarized, the intensity of the π\* and σ\* transitions will be sensitive to the orientation

The great power of NEXAFS derives from its elemental specificity. Additionally, NEXAFS can be used to determine: presence of defects and amorphous content in carbon nanotubes, varying degrees of bond hybridization in mixed sp2/sp3-bonded carbon materials, degree of vertical alignment in nanotube samples, nature of oxygen-containing functional groups on nanotube surfaces (Hemraj-Benny et al., 2006). Applying grazing-incidence XRF-NEXAFS it will be possible to build up a profile of chemical bonding in multilayered samples (Pagels et al., 2010).

X-ray photoelectron spectroscopy (XPS) is a technique that measures the elemental composition, empirical formula, chemical state and the electronic state of the elements in the

silicon carbonitride sample is given in Fig. 5. In that case the spectra are recorded simultaneously. For wavelength-dispersive (WDX) detection a system with a diffraction monocrystal (or monolayer), collimators, and a combined proportional/scintillation counter are used. In that case the spectra are recorded sequentially. This system exhibits a better energetic resolution. For a quantitative analysis, calibration measurements are inserted in the procedure using reference materials with a composition similar to the sample to be analyzed. In all cases a validation is necessary: Either by measurement of well-known standard reference materials (SRM) or by comparison with the result of an independent analytical method.

Fig. 5. EDS spectrum of PECVD SiCxNy grown at 973 K using a (HMDS+He+NH3) mixture.

A special arrangement is used in the so called "total-reflection XRF (TXRF)". The incident beam is directed onto a flat (polished) sample at an angle of 1-2°. In case of total reflection the penetration depth is in the nm-range. Therefore, this method can be used for surface analysis. As sample carrier quartz, plexiglass, glassy carbon, or boron nitride are applied. The absolute lower limit of detection (LLD) varies between 5 and 100 pg, that corresponds to a relative LLD of about 0.1-2 ng/g. The field of application is very broad: microsamples (droplets, particles), surface analysis, thin films from high vacuum techniques, semiconductor handling, evaporation and sputter procedures, and laser mirror production (Hein et al., 1992). A powerful technique can be build up by arranging a TXRF unit at a synchrotron beam line (Beckhoff et al., 2007).

#### **4.11 Near-edge X-ray Absorption Fine Structure spectroscopy (NEXAFS)**

NEXAFS, also known as "X-ray Absorption Near Edge Structure" (XANES), is an absorption spectroscopy method. Electrons from the core level are excited by photons to partially filled and empty states and the consequent emission of photoelectrons is measured (Hemraj-Benny et al., 2006). The resulting core hole is filled either via an Auger process or by capture of an electron from another shell followed by emission of a fluorescent photon. The absorption intensity is detected as a function of the exciting photons´ energy. NEXAFS measurements can detect specific bonds in molecules as well as the angular dependence of the specific orbitals involved (Stöhr 1992). The fluorescent photons originate from the the top 200 nm of the film, whereas the Auger electrons arise from the top 10 nm.

silicon carbonitride sample is given in Fig. 5. In that case the spectra are recorded simultaneously. For wavelength-dispersive (WDX) detection a system with a diffraction monocrystal (or monolayer), collimators, and a combined proportional/scintillation counter are used. In that case the spectra are recorded sequentially. This system exhibits a better energetic resolution. For a quantitative analysis, calibration measurements are inserted in the procedure using reference materials with a composition similar to the sample to be analyzed. In all cases a validation is necessary: Either by measurement of well-known standard reference materials (SRM) or by comparison with the result of an independent analytical method.

Fig. 5. EDS spectrum of PECVD SiCxNy grown at 973 K using a (HMDS+He+NH3) mixture. A special arrangement is used in the so called "total-reflection XRF (TXRF)". The incident beam is directed onto a flat (polished) sample at an angle of 1-2°. In case of total reflection the penetration depth is in the nm-range. Therefore, this method can be used for surface analysis. As sample carrier quartz, plexiglass, glassy carbon, or boron nitride are applied. The absolute lower limit of detection (LLD) varies between 5 and 100 pg, that corresponds to a relative LLD of about 0.1-2 ng/g. The field of application is very broad: microsamples (droplets, particles), surface analysis, thin films from high vacuum techniques, semiconductor handling, evaporation and sputter procedures, and laser mirror production (Hein et al., 1992). A powerful technique can be build up by arranging a TXRF unit at a

**4.11 Near-edge X-ray Absorption Fine Structure spectroscopy (NEXAFS)** 

top 200 nm of the film, whereas the Auger electrons arise from the top 10 nm.

NEXAFS, also known as "X-ray Absorption Near Edge Structure" (XANES), is an absorption spectroscopy method. Electrons from the core level are excited by photons to partially filled and empty states and the consequent emission of photoelectrons is measured (Hemraj-Benny et al., 2006). The resulting core hole is filled either via an Auger process or by capture of an electron from another shell followed by emission of a fluorescent photon. The absorption intensity is detected as a function of the exciting photons´ energy. NEXAFS measurements can detect specific bonds in molecules as well as the angular dependence of the specific orbitals involved (Stöhr 1992). The fluorescent photons originate from the the

synchrotron beam line (Beckhoff et al., 2007).

Fig. 6a. NEXAFS spectra of Si3N4, SiC, and SiCxNy.

In a total-reflection geometry (see Figs. 6a,b) at the sample, the information depth using photon detection decreases to about 5 nm (Baake et al., 2010). As an example, detailed spectral resonances at the carbon K-edge yield information about the bonding environment of this atom, such as functionalized species and chemisorbed impurities. The lower π\* resonance can provide insights into bond hybridization, while the σ\* resonance is a measure of the intramolecular bond length. Since the light from the synchrotron source used is linearly polarized, the intensity of the π\* and σ\* transitions will be sensitive to the orientation of these orbitals (Hemraj-Benny et al., 2006).

Fig. 6b. NEXAFS spectra of BN, B4C, and BCxNy.

The great power of NEXAFS derives from its elemental specificity. Additionally, NEXAFS can be used to determine: presence of defects and amorphous content in carbon nanotubes, varying degrees of bond hybridization in mixed sp2/sp3-bonded carbon materials, degree of vertical alignment in nanotube samples, nature of oxygen-containing functional groups on nanotube surfaces (Hemraj-Benny et al., 2006). Applying grazing-incidence XRF-NEXAFS it will be possible to build up a profile of chemical bonding in multilayered samples (Pagels et al., 2010).

#### **4.12 X-ray Photoelectron Spectroscopy (XPS)**

X-ray photoelectron spectroscopy (XPS) is a technique that measures the elemental composition, empirical formula, chemical state and the electronic state of the elements in the

Compilation on Synthesis, Characterization

**4.13 Transmission Electron Microscopy (TEM)** 

preparation (Baram & Kaplan, 2008).

where h.

energy").

O-

and Properties of Silicon and Boron Carbonitride Films 515

h . ν = Ekin + EBV(k), (9)

energy. The determination of the different binding energies of an element in a sample is the most important power of XPS. It is stated by the "chemical shift" in comparison to a pure substance. For fixing the energy resolution over the total measuring region the electrons are limited to a constant velocity before their entrance into the analyzer ("pass

In transmission electron microscopy (TEM) an electron beam is transmitted through an ultra thin sample. An image is formed from the interaction (e.g., absorption, diffraction) of the electrons with the specimen. The electrons are guided through an expanded electron optical column. The imaging device is a fluorescent screen, a photographic film, or a CCD camera (Fultz & Howe, 2007; Rose, 2008). The analytical power of a TEM is described by the resolution properties: By reduction of spherical aberrations a magnification of 50 million times (resolution: 0.5 Ǻ=50 pm) is reached. The ability to determine the position of atoms has made the high-resolution TEM (HRTEM) an indispensable tool for nanotechnology research, including heterogeneous catalysis and the development of semiconductor devices for electronics and photonics (O´Keefe & Allard, 2004). High quality samples will have a thickness of only a few tens of nanometers. Preparation of TEM specimen is specific to the material under analysis. Some of the methods for preparing such samples are: Tissue sectioning by a microtome, sample staining, mechanical milling, chemical etching, and ion etching (sputtering). Recently, focussed ion beams (FIB) have been used for sample

For measurement of the fine structure of absorption edges to determine chemical differences in nano structures, electron energy loss spectroscopy (EELS) can be used. This method is a

The advantages of secondary ion mass spectrometry (SIMS) can shortly be described as: Detection limit in the range of parts per million (ppm) or below, all elements can be measured (H-U), full isotopic analysis, atomic and molecular detection, rapid data acqisition, and three dimensional imaging capability (depth profiling) (Goldsmith et al., 1999). SIMS is based on the impact of primary ions (0.5-20 keV) on the sample surface, resulting in the sputtering of positive and negative secondary ions (atomic and molecular), electrons, and neutral species. SIMS instruments are build up by a primary ion source (e.g.,

, O2+, Cs+), a sample manipulation system, a secondary ion extraction system, magnetic and electric fields mass spectrometer (double focussing) (also quadrupole and time of flight devices are applied), and several kinds of detectors (Faraday cup, electron multiplier, microchannel plate). As an example, a SIMS profile is given in Fig. 8 of a layered sample

As positive ions are only a small fraction of the total sputtered material, a method called "secondary neutrals mass spectrometry (SNMS)" is in use. The transformation of raw

spectral or image intensities into meaningful concentrations is still challenging.

supplement to NEXAFS and XPS (mainly for nano sized samples).

**4.14 Secondary Ion Mass Spectrometry (SIMS)** 

with the substrate Si(100) and a BCN layer on a Cu layer.

ν = irradiation energy, Ekin=energy of the emitting electron, and EBV(k)=binding

sample (Briggs & Seah 1990). The spectra are obtained by irradiation of a sample with a monochromatic X-ray beam while simultaneously measuring the kinetic energy and number of electrons that escape from a surface layer of 1 to 10 nm. XPS is a surface analysis technique. It is also known as "Electron Spectroscopy for Chemical Analysis (ESCA)". XPS detects all elements except hydrogen and helium.

Fig. 7b. XPS spectra of BN, B4C, and BCxNy.

A typical spectrum is a plot of the intensity of the electrons detected versus their binding energy of (see Fig.7a,b for several silicon and boron compounds). The peaks in such a spectrum correspond to the electron configuration of the compounds in the sample. The number of detected electrons in a characteristic peak is directly related to the amount of the element in the area irradiated. To generate atomic percentage values the signal intensity must be corrected by the "relative sensitivity factor (RSF)".

Monochromatic Al Kα X-rays with an energy of 1486.7 eV are used for excitation. As electron analyzer a "Concentric Hemispherical Analyzer (CHA)" is applied with a channeltron.

Under optimum conditions, the quantitative accuracy of the atomic percent values is 90-95% for major peaks and 60-80% of the true value for weaker signals (10-20% of the strongest signal). The binding energy will be determined via the equation (9)

sample (Briggs & Seah 1990). The spectra are obtained by irradiation of a sample with a monochromatic X-ray beam while simultaneously measuring the kinetic energy and number of electrons that escape from a surface layer of 1 to 10 nm. XPS is a surface analysis technique. It is also known as "Electron Spectroscopy for Chemical Analysis (ESCA)". XPS

A typical spectrum is a plot of the intensity of the electrons detected versus their binding energy of (see Fig.7a,b for several silicon and boron compounds). The peaks in such a spectrum correspond to the electron configuration of the compounds in the sample. The number of detected electrons in a characteristic peak is directly related to the amount of the element in the area irradiated. To generate atomic percentage values the signal intensity

Monochromatic Al Kα X-rays with an energy of 1486.7 eV are used for excitation. As electron analyzer a "Concentric Hemispherical Analyzer (CHA)" is applied with a channeltron. Under optimum conditions, the quantitative accuracy of the atomic percent values is 90-95% for major peaks and 60-80% of the true value for weaker signals (10-20% of the strongest

detects all elements except hydrogen and helium.

Fig. 7a. XPS spectra of Si3N4, SiC, and SiCxNy.

Fig. 7b. XPS spectra of BN, B4C, and BCxNy.

must be corrected by the "relative sensitivity factor (RSF)".

signal). The binding energy will be determined via the equation (9)

$$
\mathbf{h} \cdot \mathbf{v} = \mathbf{E}\_{\text{kin}} + \mathbf{E}\_{\text{B}} \mathbf{v}(\mathbf{k}),\tag{9}
$$

where h. ν = irradiation energy, Ekin=energy of the emitting electron, and EBV(k)=binding energy. The determination of the different binding energies of an element in a sample is the most important power of XPS. It is stated by the "chemical shift" in comparison to a pure substance. For fixing the energy resolution over the total measuring region the electrons are limited to a constant velocity before their entrance into the analyzer ("pass energy").

#### **4.13 Transmission Electron Microscopy (TEM)**

In transmission electron microscopy (TEM) an electron beam is transmitted through an ultra thin sample. An image is formed from the interaction (e.g., absorption, diffraction) of the electrons with the specimen. The electrons are guided through an expanded electron optical column. The imaging device is a fluorescent screen, a photographic film, or a CCD camera (Fultz & Howe, 2007; Rose, 2008). The analytical power of a TEM is described by the resolution properties: By reduction of spherical aberrations a magnification of 50 million times (resolution: 0.5 Ǻ=50 pm) is reached. The ability to determine the position of atoms has made the high-resolution TEM (HRTEM) an indispensable tool for nanotechnology research, including heterogeneous catalysis and the development of semiconductor devices for electronics and photonics (O´Keefe & Allard, 2004). High quality samples will have a thickness of only a few tens of nanometers. Preparation of TEM specimen is specific to the material under analysis. Some of the methods for preparing such samples are: Tissue sectioning by a microtome, sample staining, mechanical milling, chemical etching, and ion etching (sputtering). Recently, focussed ion beams (FIB) have been used for sample preparation (Baram & Kaplan, 2008).

For measurement of the fine structure of absorption edges to determine chemical differences in nano structures, electron energy loss spectroscopy (EELS) can be used. This method is a supplement to NEXAFS and XPS (mainly for nano sized samples).

#### **4.14 Secondary Ion Mass Spectrometry (SIMS)**

The advantages of secondary ion mass spectrometry (SIMS) can shortly be described as: Detection limit in the range of parts per million (ppm) or below, all elements can be measured (H-U), full isotopic analysis, atomic and molecular detection, rapid data acqisition, and three dimensional imaging capability (depth profiling) (Goldsmith et al., 1999). SIMS is based on the impact of primary ions (0.5-20 keV) on the sample surface, resulting in the sputtering of positive and negative secondary ions (atomic and molecular), electrons, and neutral species. SIMS instruments are build up by a primary ion source (e.g., O- , O2+, Cs+), a sample manipulation system, a secondary ion extraction system, magnetic and electric fields mass spectrometer (double focussing) (also quadrupole and time of flight devices are applied), and several kinds of detectors (Faraday cup, electron multiplier, microchannel plate). As an example, a SIMS profile is given in Fig. 8 of a layered sample with the substrate Si(100) and a BCN layer on a Cu layer.

As positive ions are only a small fraction of the total sputtered material, a method called "secondary neutrals mass spectrometry (SNMS)" is in use. The transformation of raw spectral or image intensities into meaningful concentrations is still challenging.

Compilation on Synthesis, Characterization

**5.1 Silicon carbonitride compounds** 

improve existing ones.

bonds.

**5. Properties of carbonitride compounds** 

having higher hardness than the one of diamond.

and Properties of Silicon and Boron Carbonitride Films 517

Currently, strict conditions of modern technologies and aggressive working environment dictate higher requirements for construction materials quality. Two approaches are implemented to create new advanced materials: Synthesize radically new materials, or

In the last twenty years, researchers from different countries are studying the possibility to synthesize a new class of multifunctional materials based on the ternary compound silicon carbonitride SiCN. Varying the elemental composition of silicon carbonitrides, that is, synthesis of any set of compounds, corresponding to the ternary phase diagram of Si-C-N from silicon and carbon nitrides to silicon carbide, diamond, and their mixtures, can obtain

It is assumed that these materials may possess the unique properties combining the best ones of the compounds mentioned, such as high mechanical strength and hardness, high thermal resistance, and chemical inertness. Silicon carbide SiC is studied as a promising high-temperature semiconductor material. It is known that silicon nitride Si3N4 is one of the key materials of modern electronics and a basic component of the ceramic composites. In recent years, there have been active attempts to synthesize carbon nitride C3N4 as a material

According to the literature, in those years several researchers have attempted to obtain silicon nitride films, not only with the use of ammonolysis of monosilane widely applied at that time, but also to develop many alternative ways of synthesis, in particular, with the use of organosilicon compounds. In the beginning of the 80-ies of the last century scientists from the Irkutsk Institute of Chemistry, specialized in synthesis of organosilicon compounds, used them as single-source precursors to obtain silicon nitride films. Hence, silicon nitride films were obtained in glow-discharge plasma from HMCTS in mixtures with N2 or NH3 at low temperatures (below 150°С) (Voronkov et al., 1981). There Si-N, C-C, Si-H (or Si-C≡N) and N-H chemical bonds were determined in the films obtained at such conditions. Later silicon nitride films were deposited by PECVD using a mixture of HMCTS and a wider set of additional gases such as NH3, H2, and N2, and higher temperatures up to 400°С and plasma power (5-50 W) (Brooks & Hess, 1987, 1988). The set of characterization methods has been expanded. We can assume that so called silicon nitride films in reality consist of silicon carbonitride, whereas the films obtained from the mixture HMCTS+H2 have significant amounts of carbon (30-40at.%) and 21at.% of hydrogen and contain both Si-N and Si-C

Lateron, the films were obtained by plasma enhanced chemical decomposition using HMCTS in the mixture with helium or nitrogen in the temperature range of 100-750°С and plasma powers of 15-50 W (Fainer et al., 2009a, 2009b). Physical and chemical as well as functional properties of these films were studied by FTIR, Raman spectroscopy, XPS, EDXRS, XRD using synchrotron radiation, SEM, AFM, nanoindentation, ellipsometry, spectrophotometry, and electrophysical methods. The evaluation of the results obtained by spectroscopic methods showed that the low temperature SiCxNy films are compounds in which chemical bonding are present among Si, N, and C and with impurity elements, such as hydrogen and oxygen. Thus, a formula SiCxNyOz:H is more correct. Electrophysical and mechanical characteristics, and other physicochemical properties have allowed new consideration of these SiCxNyOz:H films as perspective interlayer dielectric films in

new materials with desired physical and chemical properties in a wide range.

Fig. 8. SIMS profile of a layered system BCN/Cu/Si.

#### **4.15 Rutherford back scattering (RBS)**

Rutherford back scattering (RBS) is a method applied in material science for the determination of the composition, the structure and of the depth profile in a sample (Oura et al., 2003). A beam of high energy (1-3 MeV) ions is directed on a sample. The ions partly backscattered at nuclei (the scattering at electrons leads to some extend to a decrease of the resolution) are detected. The energy of these backscattered ions is a function of the mass of the atoms (and of the scatter angle), at which the collision take place. An RBS instrument consists of an ion source (linear particle accelerator or an alpha particle source) and an energy sensitive detector (silicon surface barrier detector). In practice, the compositional depth profile can be determined from an intensity-energy measurement. The elements are characterized by the peak position in the spectrum and the depth can be derived from the width and shifted position of these peaks. Crystal structures (channeling) and surface information can also be evaluated from the spectra.

#### **4.16 Elastic Recoil Detection Analysis (ERDA)**

Elastic recoil detection analysis is a nuclear technique in materials science to obtain elemental concentration depth profiles in thin films. An energetic ion beam is directed at the sample to be depth profiled. As in RBS an elastic nuclear interaction with the atoms of the sample is observed. The energy of the incident ions (some MeV) is enough to recoil the atoms which are detected with a suitable detector. The advantage in ERDA is that all atoms of the sample can be recoiled if a heavy incident beam is used. For example, a 200 MeV Au beam is used with an ionization detector. In the right recoil angle the scattered incident beam ions do not reach the detector. ERDA is often used with a relatively low energy 4He beam (2 MeV) for depth profiling of hydrogen.

### **5. Properties of carbonitride compounds**

#### **5.1 Silicon carbonitride compounds**

516 Silicon Carbide – Materials, Processing and Applications in Electronic Devices

Rutherford back scattering (RBS) is a method applied in material science for the determination of the composition, the structure and of the depth profile in a sample (Oura et al., 2003). A beam of high energy (1-3 MeV) ions is directed on a sample. The ions partly backscattered at nuclei (the scattering at electrons leads to some extend to a decrease of the resolution) are detected. The energy of these backscattered ions is a function of the mass of the atoms (and of the scatter angle), at which the collision take place. An RBS instrument consists of an ion source (linear particle accelerator or an alpha particle source) and an energy sensitive detector (silicon surface barrier detector). In practice, the compositional depth profile can be determined from an intensity-energy measurement. The elements are characterized by the peak position in the spectrum and the depth can be derived from the width and shifted position of these peaks. Crystal structures (channeling) and surface

Elastic recoil detection analysis is a nuclear technique in materials science to obtain elemental concentration depth profiles in thin films. An energetic ion beam is directed at the sample to be depth profiled. As in RBS an elastic nuclear interaction with the atoms of the sample is observed. The energy of the incident ions (some MeV) is enough to recoil the atoms which are detected with a suitable detector. The advantage in ERDA is that all atoms of the sample can be recoiled if a heavy incident beam is used. For example, a 200 MeV Au beam is used with an ionization detector. In the right recoil angle the scattered incident beam ions do not reach the detector. ERDA is often used with a relatively low energy 4He

Fig. 8. SIMS profile of a layered system BCN/Cu/Si.

information can also be evaluated from the spectra.

**4.16 Elastic Recoil Detection Analysis (ERDA)** 

beam (2 MeV) for depth profiling of hydrogen.

**4.15 Rutherford back scattering (RBS)** 

Currently, strict conditions of modern technologies and aggressive working environment dictate higher requirements for construction materials quality. Two approaches are implemented to create new advanced materials: Synthesize radically new materials, or improve existing ones.

In the last twenty years, researchers from different countries are studying the possibility to synthesize a new class of multifunctional materials based on the ternary compound silicon carbonitride SiCN. Varying the elemental composition of silicon carbonitrides, that is, synthesis of any set of compounds, corresponding to the ternary phase diagram of Si-C-N from silicon and carbon nitrides to silicon carbide, diamond, and their mixtures, can obtain new materials with desired physical and chemical properties in a wide range.

It is assumed that these materials may possess the unique properties combining the best ones of the compounds mentioned, such as high mechanical strength and hardness, high thermal resistance, and chemical inertness. Silicon carbide SiC is studied as a promising high-temperature semiconductor material. It is known that silicon nitride Si3N4 is one of the key materials of modern electronics and a basic component of the ceramic composites. In recent years, there have been active attempts to synthesize carbon nitride C3N4 as a material having higher hardness than the one of diamond.

According to the literature, in those years several researchers have attempted to obtain silicon nitride films, not only with the use of ammonolysis of monosilane widely applied at that time, but also to develop many alternative ways of synthesis, in particular, with the use of organosilicon compounds. In the beginning of the 80-ies of the last century scientists from the Irkutsk Institute of Chemistry, specialized in synthesis of organosilicon compounds, used them as single-source precursors to obtain silicon nitride films. Hence, silicon nitride films were obtained in glow-discharge plasma from HMCTS in mixtures with N2 or NH3 at low temperatures (below 150°С) (Voronkov et al., 1981). There Si-N, C-C, Si-H (or Si-C≡N) and N-H chemical bonds were determined in the films obtained at such conditions. Later silicon nitride films were deposited by PECVD using a mixture of HMCTS and a wider set of additional gases such as NH3, H2, and N2, and higher temperatures up to 400°С and plasma power (5-50 W) (Brooks & Hess, 1987, 1988). The set of characterization methods has been expanded. We can assume that so called silicon nitride films in reality consist of silicon carbonitride, whereas the films obtained from the mixture HMCTS+H2 have significant amounts of carbon (30-40at.%) and 21at.% of hydrogen and contain both Si-N and Si-C bonds.

Lateron, the films were obtained by plasma enhanced chemical decomposition using HMCTS in the mixture with helium or nitrogen in the temperature range of 100-750°С and plasma powers of 15-50 W (Fainer et al., 2009a, 2009b). Physical and chemical as well as functional properties of these films were studied by FTIR, Raman spectroscopy, XPS, EDXRS, XRD using synchrotron radiation, SEM, AFM, nanoindentation, ellipsometry, spectrophotometry, and electrophysical methods. The evaluation of the results obtained by spectroscopic methods showed that the low temperature SiCxNy films are compounds in which chemical bonding are present among Si, N, and C and with impurity elements, such as hydrogen and oxygen. Thus, a formula SiCxNyOz:H is more correct. Electrophysical and mechanical characteristics, and other physicochemical properties have allowed new consideration of these SiCxNyOz:H films as perspective interlayer dielectric films in

Compilation on Synthesis, Characterization

relationships were determined.

properties were determined.

GPa, that exceeds hardness of both Si3N4 and SiC.

for light emitting diodes (LED) in blue and UV spectrum areas.

and NMR.

and Properties of Silicon and Boron Carbonitride Films 519

around silicon have been confirmed by TEM/EELS analyses. Also links were observed between the three elements: Silicon, nitrogen and carbon, which was confirmed by FTIR,

Remote microwave hydrogen plasma CVD (RP-CVD) was used with BDMADMS as precursor for the synthesis of silicon carbonitride (Si:C:N) films (Blaszczyk-Lezak et al., 2007). The Si:C:N films were characterized by XPS and FTIR, as well as by AFM. The increase of TS enhances crosslinking in the film via the formation of nitridic Si–N and carbidic Si–C bonds. On the basis of the structural data a hypothetical crosslinking reaction

Si:C:N films were produced by RPCVD from a 1,1,3,3-TMDSN precursor and at a substrate temperature in the range of 30–400°C (Wrobel et al., 2007). The effects of the substrate temperature on the rate and yield of the RP-CVD process and chemical structure (examined by FTIR) of the resulting films were investigated. The Si:C:N film properties were characterized in terms of density, hardness (2.5-16 GPa), Young´s modulus (43-187 GPa), and friction coefficient (0.02-0.05). With the IR structural data, reasonable structure–property

Physical, optical, and mechanical properties were investigated of amorphous hydrogenated silicon carbonitride (a-Si:C:N:H) films produced by the remote PECVD from (dimethylamino)dimethylsilane in relation to their chemical composition and structure (Blaszczyk-Lezak et al., 2006). The films deposited at different substrate temperatures (30–400°C) were characterized in terms of their density (1.95-2.27 g/cm3), refractive index (1.8-2.07), adhesion to a substrate, hardness (24-35 GPa), Young´s modulus (150-198 GPa), friction coefficient (0.036-0.084), and resistance to wear predicted from the "plasticity index" values H/E°=0.10–0.12. The correlations between the film compositional parameters, expressed by the atomic concentration ratios N/Si and C/Si, as well as structural parameters described by the relative integrated intensities of the absorption IR bands from the Si–N, Si–C, and C–N bonds, and the XPS Si2p band from the Si–C bonds (controlled by substrate temperature) were investigated. On the basis of the results of these studies, reasonable compositional and structural dependencies of film

In his review Badzian proposed stable and solid phases in the ternary system Si-N-C as silicon carbonitride (Badzian, 2002). Silicon carbonitride films obtained at 1000-1200°С from mixture of tetramethylsilane, ammonia and hydrogen are characterized by a hardness of 38

Crystalline films of silicon carbonitride were obtained by MW-PECVD using H2, CH4, N2, and SiH4 mixture (Chen et al., 1998). The ternary compound (CSi)xNy exhibits a hexagonal structure and consists of a network wherein Si and C are substitutional elements. While the N content of the compound is in the range 35–40 at.%, the fraction of Si varies and can be as low as 10 at.%. The preliminary lattice parameters *a* and *c* are 5.4 and 6.7 Å, respectively. Photoluminescence of silicon carbonitride films has been studied as well. The direct band gap of crystalline (CSi)xNy is 3.8 eV at room temperature. The measurements of optical properties have shown that SiCN is a perspective wide-band material with energies suitable

Si–C–N films were deposited on p-type Si(100) substrates by DC magnetron co-sputtering of silicon and carbon using a single sputter target with variable Si/C area ratios in nitrogen– argon mixtures (Vlcek et al., 2002). As a result, the N–Si and Si–N bonds dominate over the

contributing to silicon carbonitride network formation have been proposed.

microelectronics devices of novel generation. The empirical formula of the high-temperature films is represented as SiCxNy. It was established that the films are nanocomposite materials consisting of an amorphous part and nanocrystals with a size of 1-60 nm having lattice parameters close to those of the standard phase α-Si3N4. According to the Raman spectroscopic data, the films synthesized at a high temperature (up to 1023K) contain an insignificant number of graphite nanocrystals. The films synthesized from the mixture of HMCTS and helium or nitrogen exhibit an excellent transparency with a transmittance of 92– 95% in the spectral range λ=380–2500 nm.

Thus, the increase number of research techniques and improving their accuracy revealed that the films obtained from one and the same single-source precursor HMCTS are silicon carbonitrides. SiCxNy films are nanocomposite materials consisted of an amorphous part and distributed nanocrystals having lattice parameters close to those of the standard phase α-Si3N4. The films grown at above 973K contain inclusions of free graphite nanocrystals with a size of about 1 nm.

The compilation of publications, especially the earlier ones shows that among the authors involved in the synthesis of silicon carbonitride, no assumptions exist about what is meant by the term "carbonitride". Typically, the researchers saw it as a material having in its structure the elements of Si, C, and N. In this case, it may be a mixture of individual compounds as Si3N4, C, and SiC, and/or ternary SiCxNy compounds of variable composition.

What is silicon carbonitride, what its possible structure, let us consider some examples. In one of the first publications Si-C-N deposits were obtained by CVD using mixtures of gaseous compounds such as SiCl4, NH3, H2, and C3H8 and very high temperatures from 1100 up to 1600°C (Hirai & Goto, 1981). The obtained amorphous deposits were mixtures of amorphous a-Si3N4, SiC, and pyrolytic C (up to 10 weight %). The deposits surface had a pebble-like structure.

Thin films of amorphous silicon nitride and silicon carbonitride were grown on Si(100) substrates by pyrolysis of ethylsilazane [CH2CH3SiHNH] in mixtures with ammonia or hydrogen in the temperature range of 873-1073K (Bae et al., 1992). The films were studied by AES, RBS, and nuclear reaction analysis. It was shown, that the refraction index varied from 1.81 to 2.09. The hydrogen content was determined by ERDA to decrease from 21 to 8±1% in silicon carbonitride with increasing deposition temperature (873-1073K). According to AES the chemical composition of the films was determined as Si43C7 N48 O2. The silicon carbonitride films contained the bonds Si-C-N and Si-H.

Non-stoichiometric X-ray-amorphous Si3+xN4Cx+y was deposited during pyrolysis of polysilazane at 1440°С (Schonfelder et al., 1993). The heating up to 1650°C results in formation of a mixture of nanocomposites Si3N4/SiC or Si3N4/SiC/C.

SiCxNy coatings were obtained by CVD at 1000–1200°C using TMS–NH3–H2 (Bendeddouche et al., 1997). These coatings were analyzed by XPS, Raman spectrometry, FTIR, TEM/EELS and 29Si magic-angle spinning NMR (29Si MAS-NMR). The main bonds are Si–C, Si–N, and C–C in these films. It was demonstrated that silicon carbonitride coatings obtained at high temperatures are nonhydrogenated. To clarify the chemical environment of silicon atoms by carbon and nitrogen atoms the SiKL2,3L2,3 line shapes were analyzed. It was shown that these peaks are decomposed into components corresponding to an intermediate position between the tetrahedra Si(C)4 and Si(N)4, i.e., silicon carbonitride films are not simply a mixture of phases of SiC and Si3N4, and have a more complex relationship between the three elements, corresponding to the existence of Si(C4-n Nn) units. Mixed coordination shells

microelectronics devices of novel generation. The empirical formula of the high-temperature films is represented as SiCxNy. It was established that the films are nanocomposite materials consisting of an amorphous part and nanocrystals with a size of 1-60 nm having lattice parameters close to those of the standard phase α-Si3N4. According to the Raman spectroscopic data, the films synthesized at a high temperature (up to 1023K) contain an insignificant number of graphite nanocrystals. The films synthesized from the mixture of HMCTS and helium or nitrogen exhibit an excellent transparency with a transmittance of 92–

Thus, the increase number of research techniques and improving their accuracy revealed that the films obtained from one and the same single-source precursor HMCTS are silicon carbonitrides. SiCxNy films are nanocomposite materials consisted of an amorphous part and distributed nanocrystals having lattice parameters close to those of the standard phase α-Si3N4. The films grown at above 973K contain inclusions of free graphite nanocrystals with

The compilation of publications, especially the earlier ones shows that among the authors involved in the synthesis of silicon carbonitride, no assumptions exist about what is meant by the term "carbonitride". Typically, the researchers saw it as a material having in its structure the elements of Si, C, and N. In this case, it may be a mixture of individual compounds as Si3N4, C, and SiC, and/or ternary SiCxNy compounds of variable

What is silicon carbonitride, what its possible structure, let us consider some examples. In one of the first publications Si-C-N deposits were obtained by CVD using mixtures of gaseous compounds such as SiCl4, NH3, H2, and C3H8 and very high temperatures from 1100 up to 1600°C (Hirai & Goto, 1981). The obtained amorphous deposits were mixtures of amorphous a-Si3N4, SiC, and pyrolytic C (up to 10 weight %). The deposits surface had a

Thin films of amorphous silicon nitride and silicon carbonitride were grown on Si(100) substrates by pyrolysis of ethylsilazane [CH2CH3SiHNH] in mixtures with ammonia or hydrogen in the temperature range of 873-1073K (Bae et al., 1992). The films were studied by AES, RBS, and nuclear reaction analysis. It was shown, that the refraction index varied from 1.81 to 2.09. The hydrogen content was determined by ERDA to decrease from 21 to 8±1% in silicon carbonitride with increasing deposition temperature (873-1073K). According to AES the chemical composition of the films was determined as Si43C7 N48 O2. The silicon

Non-stoichiometric X-ray-amorphous Si3+xN4Cx+y was deposited during pyrolysis of polysilazane at 1440°С (Schonfelder et al., 1993). The heating up to 1650°C results in

SiCxNy coatings were obtained by CVD at 1000–1200°C using TMS–NH3–H2 (Bendeddouche et al., 1997). These coatings were analyzed by XPS, Raman spectrometry, FTIR, TEM/EELS and 29Si magic-angle spinning NMR (29Si MAS-NMR). The main bonds are Si–C, Si–N, and C–C in these films. It was demonstrated that silicon carbonitride coatings obtained at high temperatures are nonhydrogenated. To clarify the chemical environment of silicon atoms by carbon and nitrogen atoms the SiKL2,3L2,3 line shapes were analyzed. It was shown that these peaks are decomposed into components corresponding to an intermediate position between the tetrahedra Si(C)4 and Si(N)4, i.e., silicon carbonitride films are not simply a mixture of phases of SiC and Si3N4, and have a more complex relationship between the three elements, corresponding to the existence of Si(C4-n Nn) units. Mixed coordination shells

95% in the spectral range λ=380–2500 nm.

a size of about 1 nm.

pebble-like structure.

carbonitride films contained the bonds Si-C-N and Si-H.

formation of a mixture of nanocomposites Si3N4/SiC or Si3N4/SiC/C.

composition.

around silicon have been confirmed by TEM/EELS analyses. Also links were observed between the three elements: Silicon, nitrogen and carbon, which was confirmed by FTIR, and NMR.

Remote microwave hydrogen plasma CVD (RP-CVD) was used with BDMADMS as precursor for the synthesis of silicon carbonitride (Si:C:N) films (Blaszczyk-Lezak et al., 2007). The Si:C:N films were characterized by XPS and FTIR, as well as by AFM. The increase of TS enhances crosslinking in the film via the formation of nitridic Si–N and carbidic Si–C bonds. On the basis of the structural data a hypothetical crosslinking reaction contributing to silicon carbonitride network formation have been proposed.

Si:C:N films were produced by RPCVD from a 1,1,3,3-TMDSN precursor and at a substrate temperature in the range of 30–400°C (Wrobel et al., 2007). The effects of the substrate temperature on the rate and yield of the RP-CVD process and chemical structure (examined by FTIR) of the resulting films were investigated. The Si:C:N film properties were characterized in terms of density, hardness (2.5-16 GPa), Young´s modulus (43-187 GPa), and friction coefficient (0.02-0.05). With the IR structural data, reasonable structure–property relationships were determined.

Physical, optical, and mechanical properties were investigated of amorphous hydrogenated silicon carbonitride (a-Si:C:N:H) films produced by the remote PECVD from (dimethylamino)dimethylsilane in relation to their chemical composition and structure (Blaszczyk-Lezak et al., 2006). The films deposited at different substrate temperatures (30–400°C) were characterized in terms of their density (1.95-2.27 g/cm3), refractive index (1.8-2.07), adhesion to a substrate, hardness (24-35 GPa), Young´s modulus (150-198 GPa), friction coefficient (0.036-0.084), and resistance to wear predicted from the "plasticity index" values H/E°=0.10–0.12. The correlations between the film compositional parameters, expressed by the atomic concentration ratios N/Si and C/Si, as well as structural parameters described by the relative integrated intensities of the absorption IR bands from the Si–N, Si–C, and C–N bonds, and the XPS Si2p band from the Si–C bonds (controlled by substrate temperature) were investigated. On the basis of the results of these studies, reasonable compositional and structural dependencies of film properties were determined.

In his review Badzian proposed stable and solid phases in the ternary system Si-N-C as silicon carbonitride (Badzian, 2002). Silicon carbonitride films obtained at 1000-1200°С from mixture of tetramethylsilane, ammonia and hydrogen are characterized by a hardness of 38 GPa, that exceeds hardness of both Si3N4 and SiC.

Crystalline films of silicon carbonitride were obtained by MW-PECVD using H2, CH4, N2, and SiH4 mixture (Chen et al., 1998). The ternary compound (CSi)xNy exhibits a hexagonal structure and consists of a network wherein Si and C are substitutional elements. While the N content of the compound is in the range 35–40 at.%, the fraction of Si varies and can be as low as 10 at.%. The preliminary lattice parameters *a* and *c* are 5.4 and 6.7 Å, respectively. Photoluminescence of silicon carbonitride films has been studied as well. The direct band gap of crystalline (CSi)xNy is 3.8 eV at room temperature. The measurements of optical properties have shown that SiCN is a perspective wide-band material with energies suitable for light emitting diodes (LED) in blue and UV spectrum areas.

Si–C–N films were deposited on p-type Si(100) substrates by DC magnetron co-sputtering of silicon and carbon using a single sputter target with variable Si/C area ratios in nitrogen– argon mixtures (Vlcek et al., 2002). As a result, the N–Si and Si–N bonds dominate over the

Compilation on Synthesis, Characterization

≈55°.

206–305 GPa, respectively.

and Properties of Silicon and Boron Carbonitride Films 521

crucibles (Fainer et al., 2008). SiCxNy coatings were grown on fused silica substrates from hexamethyldisilazane with helium or ammonia in the temperature interval of 873-1073 K. Change of surface morphology, elemental composition and wetting angles were studied after the interaction of the surface of SiCxNy layers with the silicon melt at 1423 K by SEM, EDX and sessile drop measurements. The drop measurements after interaction of liquid Si (≈1450°C) with the surface of SiC4N sample determined a wetting angle of ≈90° that implying a poor wetting. The lack of etching figures on the SiCxNy surface proved, that no chemical reaction starts of Si melt with the SiCxNy coating. In case of silicon carbonitride with larger concentration of nitrogen (Si2C3N2) wetting angle was obtained as ≈60° close to that one of Si melt on Si3N4 of

Silicon carbonitride (SiCxNy) films were grown on silicon substrates using the PLD technique (Boughaba et al., 2002). A SiC target was ablated by the beam of a KrF excimer laser in a N2 background gas. The morphology, structure, composition, as well as the optical and mechanical properties of the coatings were investigated as functions of the N2 pressure (1– 30 mTorr) and substrate temperature (250–650°C). Smooth, amorphous films were obtained for all the processing parameters. The hardness, Young´s modulus of the films were found to be a function of the growth regime; the highest values of the hardness and Young´s modulus values were obtained in the low-pressure regime, in the range of 27–42 GPa and

A visible-blind ultraviolet (UV) photodetector (PD) with metal-semiconductor-metal (MSM) structure has been developed on a cubic-crystalline SiCN film (Chang et al., 2003). The cubic-crystalline SiCN film was deposited on Si substrate with rapid thermal CVD (at 1150°C) using SiH4, C3H8, NH3, and H2 mixture. The optoelectronical performances of the SiCN-MSMPD have been examined by the measurement of photo and dark currents and the current ratio under various operating temperatures. The current ratio for 254 nm UV light of the detector is about 6.5 at room temperature and 2.3 at 200°C, respectively. The results are better than for the counterpart SiC of 5.4 at room temperature, and less than 2 for above 100 °C, thus offering potential applications for low-cost and high-temperature UV detection. The internal stress, optical gap, and chemical inertness were examined of amorphous silicon-nitride films incorporating carbon prepared by RF magnetron sputtering (Yasui et al., 1989). The carbon composition of the films was less than 15 at.%. The optical band gap was barely affected by the carbon addition. The internal stress was compressive in all films and increased up to 7.3×108 N/cm2 in a-SiN:H films proportional to the nitrogen content, and decreased to less than half in carbon-free films. The buffered HF etch rate increased to greater than 1 μm/min in proportion to the nitrogen content in SiN:H films. The etch rate

In several papers thin films of silicon carbonitride are described with compositions varying in the wide range from similar to silicon carbide to similar to silicon nitride. These were synthesized by PECVD using HMDS as single-source precursor in the mixtures with helium, nitrogen or ammonia in the wide range of temperatures from 100 up to 800°С and RF plasma powers from 15 up to 70 W (Fainer et al., 1999, 2000, 2001a, 2001b, 2003, 2004, 2008). The nondestructive method XRD-SR was developed to determine phase composition and crystallinity of the obtained films composed of lightweight elements (Si, N, C) using the facilities of the station "Anomalous Scattering" (International Siberian Center for Synchrotron and Terahertz Radiation, Budker Institute of Nuclear Physics, SB RAS, Novosibirsk, Russia). The application of SR-XRD and high-resolution electron microscopy

decreased by about one order of magnitude with the addition of carbon.

N–C and Si–O bonds (XPS), preferred in a pure nitrogen discharge, and the film hardness increases up to 40 GPa.

SiCN coatings were deposited on silicon substrates (350°C) by PECVD using mixtures of methyltrichlorosilane (MTCS), nitrogen, and hydrogen (Ivashchenko et al., 2007). The coatings were characterized by AFM, XRD, and FTIR. Their mechanical properties are determined with nanoindentation. The abrasion wear resistance is examined using a ballon-plane (calowear) test and adhesion to the base was tested using a scratch test. The XRD measurement indicates that the coatings are nanostructured and represent β-C3N4 crystallites embedded into an amorphous a-SiCN matrix. The coatings deposited at a higher nitrogen flow rate are amorphous. β-C3N4 crystallites embedded into the amorphous a-SiCN matrix promote an increase in hardness (25 GPa) and Young's modulus (above 200 GPa) of SiCN coatings.

Tribological tests have revealed that the friction coefficients of the coatings containing nitrogen are two to three times smaller than those based on SiC and deposited on a silicon substrate. The ball-on-plane tests show that the nanostructured coatings also exhibit the highest abrasive wear resistance. These findings demonstrate that the SiCN films deposited using MTCS show good mechanical and tribological properties and can be used as wearresistant coatings.

SiCN hard films have been synthesized on stainless steel substrates by an arc enhanced magnetic sputtering hybrid system using a silicon target and graphite target in mixed gases of Ar and N2 (Ma et al., 2008). The XRD results indicate that basically the SiCN films are amorphous. However, the HRTEM results confirm that the microstructure of the SiCN films with a high silicon content are nanocomposites in which nano-sized crystalline C3N4 hard particles are embedded in the amorphous SiCN matrix. The hardness of the SiCN films is found to increase with increasing silicon content, and the maximum hardness is 35 GPa. The SiCN hard films show a surprising low friction coefficient of 0.2 when the silicon content is relatively low.

SiCN films have been produced by means of reactive magnetron sputtering of a silicon target in an argon/nitrogen/acetylene atmosphere (Hoche et al., 2008). The mechanical, chemical, and structural properties have been thoroughly investigated by means of indentation hardness testing, pin on disk wear testing in reciprocating sliding motion, glow discharge optical emission spectroscopy (GDOES), FTIR, Raman spectroscopy, XPS. The main aim of this investigation was to establish the relationship between deposition conditions, resulting mechanical, chemical, structural, and the respective wear properties. Analogous to their position in the Si–C–N phase diagram, the hardness of the films varies over a broad range, with maximum values of around 30 GPa, while Young's modulus remains in a narrow range around 200 GPa. XPS spectra showed the main component to be Si–C, but Si–N and to a minore extent C–C bonds were also detected. Further, IR spectra suggested the presence of the carbodiimide group. Raman spectra show a varying ratio of sp3 to sp2 carbon, depending on deposition condition. The hardest films were found along the SiC–Si3N4 tie line. In dry sliding their brittleness coupled with a high friction coefficient led to premature coating failure. Carbon rich films have a very low friction coefficient leading to good wear behaviour in dry conditions, but their ability to withstand high Hertzian pressures is reduced. The low friction coefficient of is attributed to more graphitic structures of the free carbon in the films.

To decrease the level of contamination of silicon melts during the Czochralski process the novel protective layer of silicon carbonitride was proposed for the inner surface of quartz

N–C and Si–O bonds (XPS), preferred in a pure nitrogen discharge, and the film hardness

SiCN coatings were deposited on silicon substrates (350°C) by PECVD using mixtures of methyltrichlorosilane (MTCS), nitrogen, and hydrogen (Ivashchenko et al., 2007). The coatings were characterized by AFM, XRD, and FTIR. Their mechanical properties are determined with nanoindentation. The abrasion wear resistance is examined using a ballon-plane (calowear) test and adhesion to the base was tested using a scratch test. The XRD measurement indicates that the coatings are nanostructured and represent β-C3N4 crystallites embedded into an amorphous a-SiCN matrix. The coatings deposited at a higher nitrogen flow rate are amorphous. β-C3N4 crystallites embedded into the amorphous a-SiCN matrix promote an increase in hardness (25 GPa) and Young's modulus (above 200 GPa) of

Tribological tests have revealed that the friction coefficients of the coatings containing nitrogen are two to three times smaller than those based on SiC and deposited on a silicon substrate. The ball-on-plane tests show that the nanostructured coatings also exhibit the highest abrasive wear resistance. These findings demonstrate that the SiCN films deposited using MTCS show good mechanical and tribological properties and can be used as wear-

SiCN hard films have been synthesized on stainless steel substrates by an arc enhanced magnetic sputtering hybrid system using a silicon target and graphite target in mixed gases of Ar and N2 (Ma et al., 2008). The XRD results indicate that basically the SiCN films are amorphous. However, the HRTEM results confirm that the microstructure of the SiCN films with a high silicon content are nanocomposites in which nano-sized crystalline C3N4 hard particles are embedded in the amorphous SiCN matrix. The hardness of the SiCN films is found to increase with increasing silicon content, and the maximum hardness is 35 GPa. The SiCN hard films show a surprising low friction coefficient of 0.2 when the silicon content is

SiCN films have been produced by means of reactive magnetron sputtering of a silicon target in an argon/nitrogen/acetylene atmosphere (Hoche et al., 2008). The mechanical, chemical, and structural properties have been thoroughly investigated by means of indentation hardness testing, pin on disk wear testing in reciprocating sliding motion, glow discharge optical emission spectroscopy (GDOES), FTIR, Raman spectroscopy, XPS. The main aim of this investigation was to establish the relationship between deposition conditions, resulting mechanical, chemical, structural, and the respective wear properties. Analogous to their position in the Si–C–N phase diagram, the hardness of the films varies over a broad range, with maximum values of around 30 GPa, while Young's modulus remains in a narrow range around 200 GPa. XPS spectra showed the main component to be Si–C, but Si–N and to a minore extent C–C bonds were also detected. Further, IR spectra suggested the presence of the carbodiimide group. Raman spectra show a varying ratio of sp3 to sp2 carbon, depending on deposition condition. The hardest films were found along the SiC–Si3N4 tie line. In dry sliding their brittleness coupled with a high friction coefficient led to premature coating failure. Carbon rich films have a very low friction coefficient leading to good wear behaviour in dry conditions, but their ability to withstand high Hertzian pressures is reduced. The low friction coefficient of is attributed to more graphitic

To decrease the level of contamination of silicon melts during the Czochralski process the novel protective layer of silicon carbonitride was proposed for the inner surface of quartz

increases up to 40 GPa.

SiCN coatings.

resistant coatings.

relatively low.

structures of the free carbon in the films.

crucibles (Fainer et al., 2008). SiCxNy coatings were grown on fused silica substrates from hexamethyldisilazane with helium or ammonia in the temperature interval of 873-1073 K. Change of surface morphology, elemental composition and wetting angles were studied after the interaction of the surface of SiCxNy layers with the silicon melt at 1423 K by SEM, EDX and sessile drop measurements. The drop measurements after interaction of liquid Si (≈1450°C) with the surface of SiC4N sample determined a wetting angle of ≈90° that implying a poor wetting. The lack of etching figures on the SiCxNy surface proved, that no chemical reaction starts of Si melt with the SiCxNy coating. In case of silicon carbonitride with larger concentration of nitrogen (Si2C3N2) wetting angle was obtained as ≈60° close to that one of Si melt on Si3N4 of ≈55°.

Silicon carbonitride (SiCxNy) films were grown on silicon substrates using the PLD technique (Boughaba et al., 2002). A SiC target was ablated by the beam of a KrF excimer laser in a N2 background gas. The morphology, structure, composition, as well as the optical and mechanical properties of the coatings were investigated as functions of the N2 pressure (1– 30 mTorr) and substrate temperature (250–650°C). Smooth, amorphous films were obtained for all the processing parameters. The hardness, Young´s modulus of the films were found to be a function of the growth regime; the highest values of the hardness and Young´s modulus values were obtained in the low-pressure regime, in the range of 27–42 GPa and 206–305 GPa, respectively.

A visible-blind ultraviolet (UV) photodetector (PD) with metal-semiconductor-metal (MSM) structure has been developed on a cubic-crystalline SiCN film (Chang et al., 2003). The cubic-crystalline SiCN film was deposited on Si substrate with rapid thermal CVD (at 1150°C) using SiH4, C3H8, NH3, and H2 mixture. The optoelectronical performances of the SiCN-MSMPD have been examined by the measurement of photo and dark currents and the current ratio under various operating temperatures. The current ratio for 254 nm UV light of the detector is about 6.5 at room temperature and 2.3 at 200°C, respectively. The results are better than for the counterpart SiC of 5.4 at room temperature, and less than 2 for above 100 °C, thus offering potential applications for low-cost and high-temperature UV detection.

The internal stress, optical gap, and chemical inertness were examined of amorphous silicon-nitride films incorporating carbon prepared by RF magnetron sputtering (Yasui et al., 1989). The carbon composition of the films was less than 15 at.%. The optical band gap was barely affected by the carbon addition. The internal stress was compressive in all films and increased up to 7.3×108 N/cm2 in a-SiN:H films proportional to the nitrogen content, and decreased to less than half in carbon-free films. The buffered HF etch rate increased to greater than 1 μm/min in proportion to the nitrogen content in SiN:H films. The etch rate decreased by about one order of magnitude with the addition of carbon.

In several papers thin films of silicon carbonitride are described with compositions varying in the wide range from similar to silicon carbide to similar to silicon nitride. These were synthesized by PECVD using HMDS as single-source precursor in the mixtures with helium, nitrogen or ammonia in the wide range of temperatures from 100 up to 800°С and RF plasma powers from 15 up to 70 W (Fainer et al., 1999, 2000, 2001a, 2001b, 2003, 2004, 2008). The nondestructive method XRD-SR was developed to determine phase composition and crystallinity of the obtained films composed of lightweight elements (Si, N, C) using the facilities of the station "Anomalous Scattering" (International Siberian Center for Synchrotron and Terahertz Radiation, Budker Institute of Nuclear Physics, SB RAS, Novosibirsk, Russia). The application of SR-XRD and high-resolution electron microscopy

Compilation on Synthesis, Characterization

materials for CVD.

and Properties of Silicon and Boron Carbonitride Films 523

arrangements and the electronic structures of three models of BC2N were studied. A correlation was found between the structural symmetries and the conducting properties. Two structures were found to have semiconducting gaps and one to be metallic. This behaviour is similar to the relation of graphite to BN. This paper initiated a world wide activity in synthesizing of BCxNy by various methods and characterizing the products by an increasing number of analytical methods. Beneath the interest for the chemical structure, the elemental composition, the speciation (chemical bonding), and the relation between

About 10 years later a review on BCN materials was published (Kawaguchi 1997). The chemical bond energies are given as B-N: 4.00 eV, C-C: 3.71 eV, N-C: 2.83 eV, and N-N: 2.11 eV. Furthermore, the product is described by a possible replacement of nitrogen by carbon in h-BN. The conductivity of BC2N was found to be variable over several orders of magnitude at room temperature related to the synthesis conditions. The conductivity of BC3N was 10 times lower than that of carbon plates, and slightly larger than that of BC2N the increase at temperatures between 25 and 700°C shows, that BC3N is stated to be a semiconductor. Additionally, photoluminescence and cathodoluminescence were observed for BN(C,H) films, intercalation chemistry is discussed, and an application of intercalated Li into B/C/N is proposed for Li battery systems. Mainly, for the future it is desirable to receive large-crystalline B/C/N materials, e.g., by a selection of appropriate starting

In the same year BCN samples were prepared by nitridation of B4C (Kurmaev et al. 1997). For characterization X-ray emission, XRD, Raman, and TEM-EELS were used. New signals were found (no B4C, no graphite, no h-BN), which confirmed the structural model in which

BCN films were deposited by RF magnetron sputtering from h-BN and graphite targets in an Ar-N2 gas mixture (Zhou et al. 2000). A large variety of analytical methods was used: XPS, Auger, FTIR, Raman, XRD, and nanoindentation. B-N, B-C, and C-N bonds were identified. No phase separation between h-BN and graphite was observed. Amorphous BC2N films with an atomically smooth surface were obtained. As mechanical and tribological parameters were measured: Hardness in the range 10-30 GPa, microfriction coefficient was 0.11 under a load of

In the following years a number of papers was published by a Spanish group. Their method of production was the IBAD technique. Therein B4C was evaporated with concurrent N2+ bombardment (Gago et al., 2001a, 2001b, 2002a, 2002b). Various methods were used to identify the character of the products: NEXAFS, FTIR, Raman, HRTEM, and time-of-flight-ERDA. The results can be summarized as follows: c-BCN and h-BCN (B50C10N40, solubility of C in h-BN about 15%) were identified, and the transition from amorphous BxC to h-BNlike structures was observed. As physical parameters a hardness of 35 GPa, a Young´s

Fullerene-like B-C-N products were synthesized by dual cathode sputtering (Hellgren et al., 2004). By means of RBS, SEM, HRTEM, and nanoindentation a fullerene-like microstructure

The incorporation of carbon into the crystal structure of h-BN was stated first by S.C. Ray

In these years, a systematic examination of BCN products can be observed from the literature. For chemical bonding determination mainly XPS and NEXAFS (also FTIR) are

boron nitride monolayers are in random intercalation with the graphite ones.

modulus, a friction coefficient of 0.05, and thermal stability were measured.

1000 µN, and the Young´s modulus was within 100-200 GPa.

was determined and an elastic response was observed.

(Ray et al., 2004) using XRD and NEXAFS examinations.

chemical situation and physical properties were investigated, up to now.

with selective area electron diffraction (HRTEM-SAED) yielded to the result that silicon carbonitride films contain nanocrystals close to α-Si3N4, distributed in amorphous matrix of the film, i.e. the films are nanocomposite. The spectroscopic results (FTIR, XPS, EDX, AES, Raman) clarified that silicon carbonitride is a ternary compound, in which complex chemical bonds between all three elements – silicon, carbon and nitrogen with impurity of oxygen and inclusion of nanocrystalline graphite - are formed. The formation of mixed Si(C4-*n*N*n*) units could be proposed in the films. Apparently, the formation of nanocrystals With a phase composition close to the standard α-Si3N4 and the presence of silicon atoms surrounded by nitrogen and carbon atoms, suggests that some places in the crystal lattice occupied by silicon atoms may be substituted by isovalent carbon atoms. The formation of a substitutional solid solution is in fact possible. The films possess high transparency in the spectral region of 270–3500 nm and a large variation of band gap from 2.0 to 5.3 eV. Hydrogenated silicon oxycarbonitrides are perspective low-k dielectrics in the silicon technology of new generation. Presence of complex chemical bonds between three elements and nanocrystals in the films allowed obtaining films with higher hardness of above 30 GPa as compared with mixture phases such as α-Si3N4, SiC or C.

#### **5.2 Boron carbonitride compounds**

In the last 20 years the publications dealing with BCN are countless. They are dealing with the production, as described in the paragraphs 2 and 3. Additionally, the methods of characterization of BCN compounds to determine the elemental composition, the crystal structure, the chemical bonding, and several physical properties are abundant. All over the world (e.g., China, France, Germany, Japan, Korea, Spain, Russia, United States, and others) research and commercial materials science institutes were and are engaged in this field. The importance of BCN compounds is shown by the recent edition of a monography (Yap, 2009). Obviously, it is not possible to touch all the activities and to comment them. The selection we have made is therefore somewhat subjective and somewhat accidental.

The first activities on boron carbonitride dealt with high-melting substances, mainly to be applied in space technique. For these specimen neither physical nor chemical characterization is described in the relevant papers (Samsonov et al., 1962; Chepelenkouv et al., 1964). Nearly 10 years later, another group (Kosolapova et al., 1971) using XRD measurements characterized the products from elemental composition data as BCN. The structure of this boron carbonitride is based on BN with a somewhat increased period c of the crystal lattice. The black powder with a particle (branched) size of the order of 1 µm showed a density of 2.13 g/cm3 (determined by pycnometry). As secondary constituents or as impurities boron carbide B4C and graphite C were identified.

In the first (to our knowledge) experimental paper on BCN from the United States (Kaner et al., 1987) another group dealing with BCN is cited (Badzian, 1972). In the paper of Kaner et al. outstanding analytical methods as XRD and XPS were applied for the characterization of the product, not being a mixture of BN+C but a specific new chemical compound BxCyNz with a ratio of boron and nitrogen approximately 1:1 and an increasing fraction of C with increasing temperature at synthesis. This new compound shows a room temperature conductivity σ = 6x10-4 S/cm (whereas BN is an insulator), a thermal band gap of 0.2 eV, and is intercalated by strong reducing and oxidizing agents.

Referring to the papers of Badyan et al. and Kaner et al. a calculation examination of the BCN compounds was performed by Liu et al. (Liu et al., 1989). The possible atomic

with selective area electron diffraction (HRTEM-SAED) yielded to the result that silicon carbonitride films contain nanocrystals close to α-Si3N4, distributed in amorphous matrix of the film, i.e. the films are nanocomposite. The spectroscopic results (FTIR, XPS, EDX, AES, Raman) clarified that silicon carbonitride is a ternary compound, in which complex chemical bonds between all three elements – silicon, carbon and nitrogen with impurity of oxygen and inclusion of nanocrystalline graphite - are formed. The formation of mixed Si(C4-*n*N*n*) units could be proposed in the films. Apparently, the formation of nanocrystals With a phase composition close to the standard α-Si3N4 and the presence of silicon atoms surrounded by nitrogen and carbon atoms, suggests that some places in the crystal lattice occupied by silicon atoms may be substituted by isovalent carbon atoms. The formation of a substitutional solid solution is in fact possible. The films possess high transparency in the spectral region of 270–3500 nm and a large variation of band gap from 2.0 to 5.3 eV. Hydrogenated silicon oxycarbonitrides are perspective low-k dielectrics in the silicon technology of new generation. Presence of complex chemical bonds between three elements and nanocrystals in the films allowed obtaining films with higher hardness of above 30 GPa

In the last 20 years the publications dealing with BCN are countless. They are dealing with the production, as described in the paragraphs 2 and 3. Additionally, the methods of characterization of BCN compounds to determine the elemental composition, the crystal structure, the chemical bonding, and several physical properties are abundant. All over the world (e.g., China, France, Germany, Japan, Korea, Spain, Russia, United States, and others) research and commercial materials science institutes were and are engaged in this field. The importance of BCN compounds is shown by the recent edition of a monography (Yap, 2009). Obviously, it is not possible to touch all the activities and to comment them. The selection

The first activities on boron carbonitride dealt with high-melting substances, mainly to be applied in space technique. For these specimen neither physical nor chemical characterization is described in the relevant papers (Samsonov et al., 1962; Chepelenkouv et al., 1964). Nearly 10 years later, another group (Kosolapova et al., 1971) using XRD measurements characterized the products from elemental composition data as BCN. The structure of this boron carbonitride is based on BN with a somewhat increased period c of the crystal lattice. The black powder with a particle (branched) size of the order of 1 µm showed a density of 2.13 g/cm3 (determined by pycnometry). As secondary constituents or

In the first (to our knowledge) experimental paper on BCN from the United States (Kaner et al., 1987) another group dealing with BCN is cited (Badzian, 1972). In the paper of Kaner et al. outstanding analytical methods as XRD and XPS were applied for the characterization of the product, not being a mixture of BN+C but a specific new chemical compound BxCyNz with a ratio of boron and nitrogen approximately 1:1 and an increasing fraction of C with increasing temperature at synthesis. This new compound shows a room temperature conductivity σ = 6x10-4 S/cm (whereas BN is an insulator), a thermal band gap of 0.2 eV,

Referring to the papers of Badyan et al. and Kaner et al. a calculation examination of the BCN compounds was performed by Liu et al. (Liu et al., 1989). The possible atomic

we have made is therefore somewhat subjective and somewhat accidental.

as impurities boron carbide B4C and graphite C were identified.

and is intercalated by strong reducing and oxidizing agents.

as compared with mixture phases such as α-Si3N4, SiC or C.

**5.2 Boron carbonitride compounds** 

arrangements and the electronic structures of three models of BC2N were studied. A correlation was found between the structural symmetries and the conducting properties. Two structures were found to have semiconducting gaps and one to be metallic. This behaviour is similar to the relation of graphite to BN. This paper initiated a world wide activity in synthesizing of BCxNy by various methods and characterizing the products by an increasing number of analytical methods. Beneath the interest for the chemical structure, the elemental composition, the speciation (chemical bonding), and the relation between chemical situation and physical properties were investigated, up to now.

About 10 years later a review on BCN materials was published (Kawaguchi 1997). The chemical bond energies are given as B-N: 4.00 eV, C-C: 3.71 eV, N-C: 2.83 eV, and N-N: 2.11 eV. Furthermore, the product is described by a possible replacement of nitrogen by carbon in h-BN. The conductivity of BC2N was found to be variable over several orders of magnitude at room temperature related to the synthesis conditions. The conductivity of BC3N was 10 times lower than that of carbon plates, and slightly larger than that of BC2N the increase at temperatures between 25 and 700°C shows, that BC3N is stated to be a semiconductor. Additionally, photoluminescence and cathodoluminescence were observed for BN(C,H) films, intercalation chemistry is discussed, and an application of intercalated Li into B/C/N is proposed for Li battery systems. Mainly, for the future it is desirable to receive large-crystalline B/C/N materials, e.g., by a selection of appropriate starting materials for CVD.

In the same year BCN samples were prepared by nitridation of B4C (Kurmaev et al. 1997). For characterization X-ray emission, XRD, Raman, and TEM-EELS were used. New signals were found (no B4C, no graphite, no h-BN), which confirmed the structural model in which boron nitride monolayers are in random intercalation with the graphite ones.

BCN films were deposited by RF magnetron sputtering from h-BN and graphite targets in an Ar-N2 gas mixture (Zhou et al. 2000). A large variety of analytical methods was used: XPS, Auger, FTIR, Raman, XRD, and nanoindentation. B-N, B-C, and C-N bonds were identified. No phase separation between h-BN and graphite was observed. Amorphous BC2N films with an atomically smooth surface were obtained. As mechanical and tribological parameters were measured: Hardness in the range 10-30 GPa, microfriction coefficient was 0.11 under a load of 1000 µN, and the Young´s modulus was within 100-200 GPa.

In the following years a number of papers was published by a Spanish group. Their method of production was the IBAD technique. Therein B4C was evaporated with concurrent N2 + bombardment (Gago et al., 2001a, 2001b, 2002a, 2002b). Various methods were used to identify the character of the products: NEXAFS, FTIR, Raman, HRTEM, and time-of-flight-ERDA. The results can be summarized as follows: c-BCN and h-BCN (B50C10N40, solubility of C in h-BN about 15%) were identified, and the transition from amorphous BxC to h-BNlike structures was observed. As physical parameters a hardness of 35 GPa, a Young´s modulus, a friction coefficient of 0.05, and thermal stability were measured.

Fullerene-like B-C-N products were synthesized by dual cathode sputtering (Hellgren et al., 2004). By means of RBS, SEM, HRTEM, and nanoindentation a fullerene-like microstructure was determined and an elastic response was observed.

The incorporation of carbon into the crystal structure of h-BN was stated first by S.C. Ray (Ray et al., 2004) using XRD and NEXAFS examinations.

In these years, a systematic examination of BCN products can be observed from the literature. For chemical bonding determination mainly XPS and NEXAFS (also FTIR) are

Compilation on Synthesis, Characterization

(produced with NH3).

solution (Caretti et al., 2010).

(Mannan et al., 2011).

NNIOa.

**7. References** 

**6. Acknowledgements** 

and Properties of Silicon and Boron Carbonitride Films 525

Various single source precursors (TMAB, TEAB, TMB) were introduced in a PECVD system. XPS, NEXAFS, and SEM/EDX were used for chemical identification. As results are determined h-BCN with stoichiometric formulas B2C3N (produced without NH3) or B2CN3

Thick (20-70 nm) amorphous BxCyNz films were produced by DMAB ((CH3)2HN:BH3) in a CVD procedure (Wu et al., 2010). XPS and SIMS were used for the determination of the elemental composition. The stoichiometry factors varied drastically: 0.46 ≤ x ≤ 0.68; 0.07 ≤ y

As can be derived from a large number of papers, the synthesized compounds are h-BCN in which carbon is replacing to some extent nitrogen in the hexagonal boron nitride structure. An extended TEM examination enlarge the knowledge in this field (Caretti et al., 2010). For low carbon content the h-BN is preserved in boron carbonitride compounds. By increasing the carbon content towards BCN stoichiometry (1<x>2) the hexagonal stacking sequence tends into a fullerene-like structure. Increasing the carbon content to the composition BC4N, the sample exhibit an amorphous structure. Surprisingly, the authors call their compounds "solid solutions", although in various papers the chemical bonds B-C, B-N, and C-N were determined, yielding a defined chemical, completely hybridized compound and not a

Only a few papers announced the production of c-BCN (e.g., Gago et al., 2001a). The yield of this material (in IBAD), proposed to be as hard as diamond, was related to the optimization of the deposition temperature, the Ar content in the gas mixture, to the assisting current density, and to the ion energy. Although, the identification of c-BCN is still not proved

The authors acknowledge the financial support granted by the Deutsche Forschungsgemeinschaft (DFG) for the research projects "nanolayer speciation" (EN 207/22- 1) and "chemical and physical characterization of nanolayers" (EN 207/22-2). The authors of the Russian Federation thank RFBR for the grant 07-03-91555-NNIOa and 10-03-91332-

Abdellaoui, A., Bath, A., Bouchikhi, B., & Baehr, O. (1997). Structure and optical properties

Ahn, H., Alberts, L., Kim, Y.-M., Yu J., & Rie, K.T. (2003). BCN coatings at low temperature

Ahn, H., Klimek, K.S., & Rie K.-T. (2003). BCN coatings by RF PACVD at low temperature.

Aoki, H., Shima, H., Kimura, C., & Sugino, T. (2007). Characterization of boron carbon

*Surf. Coat. Technol.*, Vol. 174 –175, pp. 1225–1228, ISSN 0257-8972

of boron nitride thin films prepared by PECVD. *Mater. Sci. Engin*. *B,* Vol. B47, No. 3,

using PACVD: capacitive vs. inductive plasma coupling. *Surf. Coat. Technol.*, Vol.

nitride film for humidity sensor. *Diamond Related MaterialsDiamond Relat. Mater*,

Als-Nielsen, J. (2001). *Elements of Modern X-Ray Physics.* Wiley, New York

169 –170, (June 2003), pp. 251–253, ISSN 0257-8972

Vol. 16, No. 4-7, pp. 1300-1303, ISSN 0925-9635

pp.257-262, ISSN 0921-5107

≤ 0.43; 0.01 ≤ z ≤ 0.26. The results on thick BCN films are encouraging.

used, and the hardness is measured by nanoindentation. Caretti et al. described an experimental reliable change of carbon in BCxN yielding hexagonal structure (Caretti et al., 2004). They describe a hardness of 17 GPa, a Young´s modulus of 170 GPa, and friction and wear experiments. An increase of the carbon flux is followed by an increase of carbon in the product (increase of the sp3 fraction) that improves the mechanical properties. Morant et al. and Zhou et al. produced samples with a hardness of 33 GPa, determined the roughness, and established excellent friction properties (Morant et al., 2005; Zhou et al., 2006). The chemical properties were determined by XPS with an identification of B-N, B-C, and C-N bonds. The highest value for the hardness of 40 GPa were published in 2005 (Kosinova et al, 2005).

One of first papers dealing with the production of BCN compounds by using a large molecule as precursor is authored by Uddin et al. (Uddin et al., 2005). The product was identified as graphite-like BCN with B-C, B-N, and B-C-N hybrids.

Beneath the usual characterisation of BCN compounds by XPS and FTIR, the chemical behaviour (solubility) in acidic, neutral, and alkaline solutions was examined (Byon et al., 2006). In HCl no anodic dissolution was observed, in NaOH the dissolution depends on the potential and is increasing with increasing pH.

The group from Osaka, Japan, synthesized polycrystalline BCN by PECVD (Tai et al., 2003). Various properties of the films were investigated in the last years: e.g., electrical and optical characteristics (Yuki et al., 2004), influence of UV radiation on dielectric constant (Zhang et al., 2005), adaptation as humidity sensor (Aoki et al., 2007), acid and alkaline wet influence on quality of LSI devices (Watanabe et al., 2008), modification of the tunneling controlled field emission (Sugino et al., 2010).

BCN compounds were synthesized by DC reactive sputtering of B4C target in a gas mixture of N2 and Ar (Xu et al., 2006). The composition of the product depends on the N2/Ar ratio. By nanoindentation the surface morphology and roughness were examined.

A method of BCN production by PECVD with TMB (+benzene) is described by Thamm et al. (Thamm et al., 2007). The main result is: The structure and the mechanical properties are in strong dependence on the substrate temperature.

An amorphous product was synthesized with corrosion protection properties better than B4C and CNx (Chen et al. 2006) for commercial application. This is attributed to the smoother morphology of BxCyNz films. The hardness was determined to be 20±3 GPa, and the Young´s modulus to 210±30 GPa.

BCN compounds were produced by ball milling of h-BN, graphite and polypropylene (Torres et al., 2007). SEM, XRD, FTIR, and NEXAFS examinations yielded compositions as BCN, BC2N, BC4N, BCNH2, a-BCN, and a-BC4N. The particles are nearly spherical in shape (60 nm), whereas the crystallites have a size of about 1 nm. Tribological studies were performed on a-BC4N films with a thickness of 2 µm (Caretti et al., 2007). Nanoindentation shows a hardness of 18 GPa and a Young´s modulus of 170 GPa, whereas the wear examinations yielded in a constant rate of 2x10-7 mm3/Nm and a coefficient of friction of 0.2.

h-BCN was synthesized in a PECVD with triethylamine borane (TEAB) or with tris- (dimethylamine) borane (TDEAB) as single source precursors (Mannan et al., 2008, 2009). The chemical characterization by FTIR, XPS and NEXAFS showed B-N, B-C, C-N, and B-C-N bonds. A h-BCN (or sp2-BCN) was produced with a microhardness of 4 GPa (nanoindentation).

used, and the hardness is measured by nanoindentation. Caretti et al. described an experimental reliable change of carbon in BCxN yielding hexagonal structure (Caretti et al., 2004). They describe a hardness of 17 GPa, a Young´s modulus of 170 GPa, and friction and wear experiments. An increase of the carbon flux is followed by an increase of carbon in the product (increase of the sp3 fraction) that improves the mechanical properties. Morant et al. and Zhou et al. produced samples with a hardness of 33 GPa, determined the roughness, and established excellent friction properties (Morant et al., 2005; Zhou et al., 2006). The chemical properties were determined by XPS with an identification of B-N, B-C, and C-N bonds. The highest value for the hardness of 40 GPa were published in 2005 (Kosinova et al,

One of first papers dealing with the production of BCN compounds by using a large molecule as precursor is authored by Uddin et al. (Uddin et al., 2005). The product was

Beneath the usual characterisation of BCN compounds by XPS and FTIR, the chemical behaviour (solubility) in acidic, neutral, and alkaline solutions was examined (Byon et al., 2006). In HCl no anodic dissolution was observed, in NaOH the dissolution depends on the

The group from Osaka, Japan, synthesized polycrystalline BCN by PECVD (Tai et al., 2003). Various properties of the films were investigated in the last years: e.g., electrical and optical characteristics (Yuki et al., 2004), influence of UV radiation on dielectric constant (Zhang et al., 2005), adaptation as humidity sensor (Aoki et al., 2007), acid and alkaline wet influence on quality of LSI devices (Watanabe et al., 2008), modification of the tunneling controlled

BCN compounds were synthesized by DC reactive sputtering of B4C target in a gas mixture of N2 and Ar (Xu et al., 2006). The composition of the product depends on the N2/Ar ratio.

A method of BCN production by PECVD with TMB (+benzene) is described by Thamm et al. (Thamm et al., 2007). The main result is: The structure and the mechanical properties are

An amorphous product was synthesized with corrosion protection properties better than B4C and CNx (Chen et al. 2006) for commercial application. This is attributed to the smoother morphology of BxCyNz films. The hardness was determined to be 20±3 GPa, and

BCN compounds were produced by ball milling of h-BN, graphite and polypropylene (Torres et al., 2007). SEM, XRD, FTIR, and NEXAFS examinations yielded compositions as BCN, BC2N, BC4N, BCNH2, a-BCN, and a-BC4N. The particles are nearly spherical in shape (60 nm), whereas the crystallites have a size of about 1 nm. Tribological studies were performed on a-BC4N films with a thickness of 2 µm (Caretti et al., 2007). Nanoindentation shows a hardness of 18 GPa and a Young´s modulus of 170 GPa, whereas the wear examinations yielded in a constant rate of 2x10-7 mm3/Nm and a

h-BCN was synthesized in a PECVD with triethylamine borane (TEAB) or with tris- (dimethylamine) borane (TDEAB) as single source precursors (Mannan et al., 2008, 2009). The chemical characterization by FTIR, XPS and NEXAFS showed B-N, B-C, C-N, and B-C-N bonds. A h-BCN (or sp2-BCN) was produced with a microhardness of 4 GPa

By nanoindentation the surface morphology and roughness were examined.

identified as graphite-like BCN with B-C, B-N, and B-C-N hybrids.

potential and is increasing with increasing pH.

in strong dependence on the substrate temperature.

field emission (Sugino et al., 2010).

the Young´s modulus to 210±30 GPa.

coefficient of friction of 0.2.

(nanoindentation).

2005).

Various single source precursors (TMAB, TEAB, TMB) were introduced in a PECVD system. XPS, NEXAFS, and SEM/EDX were used for chemical identification. As results are determined h-BCN with stoichiometric formulas B2C3N (produced without NH3) or B2CN3 (produced with NH3).

Thick (20-70 nm) amorphous BxCyNz films were produced by DMAB ((CH3)2HN:BH3) in a CVD procedure (Wu et al., 2010). XPS and SIMS were used for the determination of the elemental composition. The stoichiometry factors varied drastically: 0.46 ≤ x ≤ 0.68; 0.07 ≤ y ≤ 0.43; 0.01 ≤ z ≤ 0.26. The results on thick BCN films are encouraging.

As can be derived from a large number of papers, the synthesized compounds are h-BCN in which carbon is replacing to some extent nitrogen in the hexagonal boron nitride structure. An extended TEM examination enlarge the knowledge in this field (Caretti et al., 2010). For low carbon content the h-BN is preserved in boron carbonitride compounds. By increasing the carbon content towards BCN stoichiometry (1<x>2) the hexagonal stacking sequence tends into a fullerene-like structure. Increasing the carbon content to the composition BC4N, the sample exhibit an amorphous structure. Surprisingly, the authors call their compounds "solid solutions", although in various papers the chemical bonds B-C, B-N, and C-N were determined, yielding a defined chemical, completely hybridized compound and not a solution (Caretti et al., 2010).

Only a few papers announced the production of c-BCN (e.g., Gago et al., 2001a). The yield of this material (in IBAD), proposed to be as hard as diamond, was related to the optimization of the deposition temperature, the Ar content in the gas mixture, to the assisting current density, and to the ion energy. Although, the identification of c-BCN is still not proved (Mannan et al., 2011).

#### **6. Acknowledgements**

The authors acknowledge the financial support granted by the Deutsche Forschungsgemeinschaft (DFG) for the research projects "nanolayer speciation" (EN 207/22- 1) and "chemical and physical characterization of nanolayers" (EN 207/22-2). The authors of the Russian Federation thank RFBR for the grant 07-03-91555-NNIOa and 10-03-91332- NNIOa.

### **7. References**

Als-Nielsen, J. (2001). *Elements of Modern X-Ray Physics.* Wiley, New York


Compilation on Synthesis, Characterization

ISSN 0897-4756

0003-2700

0021-8979

and Properties of Silicon and Boron Carbonitride Films 527

Baehr, O., Thevenin, P., Bath, A., Koukab, A., Losson, E., & Lepley, B. (1997). Preparation of

Bath A., van der Put P.J., Lepley B. (1989). Study of boron nitride gate insulators grown by

Bath, A., van der Put, P.J., Becht, J.G.M., Schooman, J., & Lepley, B. (1991). Plasma enhanced

Bath, A., Baehr, O., Barrada, M., Lepley, B., van der Put, P.J., & Schoonman, J. (1994).

Bauer-Marschallinger, J., Reitinger, B., Pfaff, H. (2007). Nanoindentation und

Beckhoff, B., Kanngießer, B., Langhoff, N., Wedell, R., & Wolff, H. (Eds.), (2006). *Practical X-*

Beckhoff, B., Fliegauf, R., Kolbe, M., Müller, M., Weser, J., & Ulm, G., (2007). Reference-Free

Bendeddouche, A., Berjoan, R., Beche, E., Merle-Mejean, T., Schamm, S., Serin, V., Taillades,

Bengu, E., Genisel, M. F., Gulseren, O. & Ovali, R. (2009). Theoretical and spectroscopic

Besmann, T. M. (1990). Chemical Vapor Deposition in the Boron-Carbon-Nitrogen System. *J.Amer. Ceram. Soc.*, Vol.73, Is. 8, April 1990, pp.2498–2501, ISSN 0002-7820 Blaszczyk-Lezak, I., Wrobel, A. M., Kivitorma, M. P. M., & Vayrynen, I. J. (2005). Silicon

Blaszczyk-Lezak, I., Wrobel, A.M., Aoki, T., Nakanishi, Y., Kucinska, I., & Tracz. A. (2006).

Blaszczyk-Lezak, I., Wrobel, A.M., & Bielinski, D.M. (2006). Remote nitrogen microwave

*Ray Fluorescence Analysis.* Springer Verlag, ISBN-10 3-540-28503-9

applications. *Mater. Sci. Engin*. *B*, Vol.B46, pp.101-104, ISSN 0921-5107 Baram, M., Kaplan, W.D., (2008). Quantitative HRTEM analysis of FIB prepared specimens*.* 

*J. Microscopy,* Vol.232, No.3, pp. 395-405, ISSN 0022-2720

InP. *J. Appl. Phys.,* Vol.70, No.8, pp.4366-4370, ISSN 0021-8979

*Sci.*, V.39, pp.135-140, ISSN 0169-4332

*Films*, Vol.241, pp.278-281, ISSN 0040-6090

*Industrielles Symposium Mechatronik.* Linz, Austria

Vol. 518, No. 5, pp. 1459-1464, ISSN 0040-6090

(January 2005), pp. 44-52, ISSN 0948-1907

(February 2006), pp. 24–34, ISSN 0040-6090

(February 2006), pp. 35–41, ISSN 0040-6090

Ethylsilazane in Ammonia or Hydrogen. *Chem. Mater.*, Vol.4, No.2, pp.478-483,

boron nitride thin films by microwave plasma enhanced CVD, for semiconductor

low temperature plasma enhanced chemical vapor deposition on InP. *Appl. Surf.* 

chemical vapor deposition and characterization of boron nitride gate insulators on

Plasma enhanced chemical vapour deposition of boron nitride onto InP. *Thin Solid* 

Oberflächenwellen (Laser-ultraschall) zur Charakterisierung dünner Schichten.

Total Reflection X-ray Fluorescence Analysis of Semiconductor Surfaces with Synchrotron Radiation*. Analytical Chemistry,* Vol.79, No.20, pp.7873-7382, ISSN

G., Pradel, A., & Hillel, R. (1997). Structural characterization of amorphous SiCxNy chemical vapor deposited coatings*. J.Appl.Phys.,* Vol.81, No.9, pp.6147-6154, ISSN

investigations on the structure and bonding in B–C–N thin films. *Thin Solid Films*,

Carbonitride Films Produced by Remote Hydrogen Microwave Plasma CVD Using a (Dimethylamino)dimethylsilane Precursor. *Chem. Vap. Deposition,* Vol. 11, No. 1,

Remote nitrogen microwave plasma chemical vapor deposition from a tetramethyldisilazane precursor. 1. Growth mechanism, structure, and surface morphology of silicon carbonitride films*. Thin Solid Films*, Vol. 497, No.1-2,

plasma chemical vapor deposition from a tetramethyldisilazane precursor. 2. Properties of deposited silicon carbonitride films. *Thin Solid Films,* Vol. 497, No. 1-2,


Aoki, H., Tokuyama, S., Sasada, T., Watanabe, D., Mazumder, M. K., Kimura, C., & Sugino,

Aoki, H., Masuzumi, T., Hara, M., Watanabe, D., Kimura, C., & Sugino, T. (2009). Boron

Aoki, H., Masuzumi, T., Watanabe, D., Mazumder, M.K., Sota, H., Kimura, C., & Sugino, T.

Aoki, H., Tokuyama, S., Masuzumi, T., Hara, M., Mazumder, M. K., Watanabe, D., Kimura,

Aoki, H., Hara, M., Masuzumi, T., Ahmed, F., Kimura, C., & Sugino, T. (2010). Dry etching

Awad, Y., El Khakani, M.A., Aktik, C., Mouine, J., Camiré, N., Lessard, M., Scarlete, M., Al-

Baake, O., Hoffmann, P.S., Kosinova, M.L., Klein, A., Pollakowski, B., Beckhoff, B., Fainer,

Badyan, A., Nemiski, T., Appenheimer, S., & Olkusnik, E. (1972). Crystal structure in the

Badzian, A. (2002). Stability of Silicon Carbonitride Phases*. J. Am. Ceram. Soc.*, Vol.85, pp.16–

Badzian, A.R., Niemyski, T., Appenheimer, S., & Olkusnik,E. (1972). Graphite-Boron Nitride

Bae, Y. W., Du, H., Gallois, B., Gonsalvest, K. E., & Wilkens, B. J. (1992). Structure and

*Chemistry,* Vol.398, No.2, (June 2010), pp.1077-1084, ISSN 1618-2642

*polumetallah,* pp. 362-366, "Nauka I technika", Minsk (in Russian)

Vol.3,pp.747-753, Salt Lake City, UT, 1972

0169-4332

0169-4332

ISSN 0925-9635

ISSN 0925-9635

0257-8972

20

Hsinchu, April 2009

T. (2008). Dry etching properties of boron carbon nitride (BCN) films using carbon fluoride gas. *Diamond Relat. Mater,* Vol. 17, No. 7-10, pp.1800-1804, ISSN 0925-9635 Aoki, H., Watanabe, D., Moriyama, R., Mazumder, M. K., Komatsu, N., Kimura, C., &

Sugino, T. (2008). Influence of moisture on BCN (low-*K*) film for interconnection reliability. *Diamond Relat. Mater*, Vol. 17, No. 4-5, pp. 628-632, ISSN 0925-9635 Aoki, H., Hara, M., Masuzumi, T., Mazumder, M. K., Ooi, N., Watanabe, D., Kimura, C., &

Sugino, T. (2009). Influence of Cu electroplating solutions on boron carbon nitride (BCN) film. *Appl. Surf. Sci.*, Vol. 255, No. 6, (October 2008), pp. 3719-3722, ISSN

carbon nitride film containing hydrogen for 2nm node low-k interconnection, *VLSI Technology, Systems, and Applications (VLSI-TSA) 2009 International Symsosium*,

(2009). Influence of oxygen plasma treatment on boron carbon nitride film composition. *Appl. Surf. Sci.,* Vol. 255, No. 6, October 2008, pp. 3635-3638, ISSN

C., & Sugino, T. (2009). Synthesis of methyl boron nitride film as low-*k* insulating film for LSI interconnection. *Diamond Relat. Mater*, Vol. 18, No. 5-8, pp. 1048-1051,

properties of methyl-BCN lm with C4F8 gas for Cu/low-k interconnection. *Diamond Relat. Mater,* Vol. 19, No.5-6, (January 2010), pp. 507–509, ISSN 0925-9635 Aoki, H., Masuzumi, T., Hara, M., Lu, Z., Watanebe, D., Kimura, C., & Sugino, T. (2010).

Inuence of the methyl group on the dielectric constant of boron carbon nitride lms containing it. *Diamond Relat. Mater*, Vol. 19, No.12, (June 2010), pp. 1437–1440,

Abadleh, H.A., & Smirani, R. (2009). Structural and mechanical properties of amorphous silicon carbonitride films prepared by vapor-transport chemical vapor deposition. *Surf. Coat. Technol..* Vol. 204, No.4, (August 2009), pp.539–545, ISSN

N.I., Trunova, V.A., & Ensinger, W. (2010). Analytical characterization of BCxNy films generated by LPCVD with triethylamine borane. *Analytical & Bioanalytical* 

system boron – carbon – nitrogen, In: *Himicheskaya svyaz v poluprovodnikah i* 

solid solution by chemical vapor deposition. *Proc. Third Inter.Conf.CVD.* 

Chemistry of Silicon Nitride and Silicon Carbonitride Thin Films Deposited from

Ethylsilazane in Ammonia or Hydrogen. *Chem. Mater.*, Vol.4, No.2, pp.478-483, ISSN 0897-4756


Compilation on Synthesis, Characterization

ISSN 0925-9635

pp.1532-1537, ISSN 0925-9635

No.9, pp.565-567, ISSN 0741-3106

2005), pp.500–503, ISSN 0254-0584

6, pp.1589-1594, ISSN 0022-2461

ISSN 0040-6090

and Properties of Silicon and Boron Carbonitride Films 529

Caretti, I.; ., Jimenéz, I.; ., Gago, R.; ., Cáceres, D.; ., Abendroth, B.; ., & Albella, J.M., (2004).

Caretti I., Albella J.M., Jiménez I. (2007). Tribological study of amorphous BC4N coatings.

Caretti, I., Torres, R., Gago, R., Landa-Cánovas, A.R., & Jiménez, I. (2010). Effect of Carbon

Chang, W.-R., Fang, Y.-K., Ting, Sh.-F., Tsair, Y.-Sh., Chang, Ch.-N., Lin, Ch.-Yu, & Chen,

Charitidis, C., Panayiotatos, Y., & Logothetidis, S. (2003). A quantitative study of the nano-

Chen, C.W., Huang, C.C., Lin, Y.Y., Chen, L.C. & Chen, K.H. (2005). The affinity of Si–N and

Chen, Y.; Chung, Y.-W.; & Li, S.-Y., (2006). Boron carbide and boron carbonitride thin films

Chen, Z., Lin, H., Zhou, J., Ma, Z., & Xie, E. (2009). IR studies of SiCN films deposited by RF sputtering method*. J. Alloys and Compounds,* Vol.487, pp. 531–536, ISSN 0925-8388 Cheng, W., Jiang, J., Zhang, Y., Zhu, H., & Shen, D. (2004). Effect of the deposition

Chepelenkouv, Y.V., Em, A.P., Borodulin, P,.Y., & Momot, L.B. (1964). Thermal stability of

Dekempeneer, E.H.A., Meneve, J., Kuypers, S., & Smeets, J. (1996). Tribological properties of

Dekempeneer, E., Wagner, V., Gibson, P., Meneve, J., Kuypers, S., van Izendoorn, L., Smeets,

coatings. *Diamond Film & Technol*. Vol.7, No.3, pp. 181-193., ISSN 0917-4540 Derre, A., Filipozzi, L., Bouyer, F. & Marchand, A. (1994) Parametric study of the chemical

Dez, R., Teґneґgal1, F., Reynaud, C., Mayne, M., Armand, X. & Herlin-Boime, N. (2002).

No.6, (February 2010), pp. 1949–1951, ISSN 0897-4756

*Relat. Mater.,* Vol. 12, No. 3-7, pp. 1088-1092, ISSN 0925-9635

Vol. 72, No.19, (March 1998), pp.2463-2465, ISSN 0003-6951

refractory crucibles based on boron nitride. No.6, p.253.

Vol.200, No.12-13 (February 2005), pp.4072-4077, ISSN 0257-8972

Tribological properties of ternary BCN films with controlled composition of bonding structure. *Diamond Relat. Mater,Diamond Relat. Mater.*, Vol.13, No.4-8,

*Diamond Relat. Mater, Diamond Relat. Mater*., Vol. 16, No.1, (March 2001), pp. 63–73,

Incorporation on the Microstructure of BCxN (x=0.25, 1, and 4) Ternary Solid Solutions Studied by Transmission Electron Microscopy. *Chem. Mater*., Vol. 22,

Sh.-F. (2003). The Hetero-Epitaxial SiCN/Si MSM Photodetector for High-Temperature Deep-UV Detecting Applications. *IEEE Electron Device Letters,* Vol. 24,

scratch behavior of boron and carbon nitride films. *Diamond Relat. MaterDiamond* 

Si–C bonding in amorphous silicon carbon nitride (a-SiCN) thin film. *Diamond Relat. MaterDiamond Relat. Mater.,* Vol.14, No.3-7, pp.1126– 1130, ISSN 0925-9635 Chen, L. C., Chen, C. K., Wei, S. L., Bhusari, D. M., Chen, K. H. & Huang, Y. S. (1998).

Crystalline silicon carbon nitride: A wide band gap semiconductor. *Appl. Phys. Let.,*

as protective coatings in ultra-high density hard disc drives. *Surf. Coat. Technol*.,

conditions on the morphology and bonding structure of SiCN films*. Materials Chemistry and Physics*, Vol. 85, No.2-3, (January 2004), pp.370–376, ISSN 0254-0584 Cheng, W., Lin F., Shi W., Ma X., Shen D., & Zhang, Y. (2006). Synthesis and field emission

property of SiCN cone arrays. *Materials Chemistry and Physics*, Vol.98, (September

r.f. PACVD amorphous B-N-C coatings. *Thin Solid Films*, Vol. 281-282, pp. 331-333,

J., Geurts, J., & Caudano, R. (1997) Tribology and microstructure of PACVD B-N-C

vapour deposition of carbon-boron-nitrogen compounds. *J. Mater. Sci.*, Vol. 29, No.

Laser synthesis of silicon carbonitride nanopowders; structure and thermal


Blaszczyk-Lezak, I., Wrobel, A.M., & Bielinski, D.M. (2006). Remote hydrogen microwave

Blaszczyk-Lezak, I., Wrobel, A.M., Kivitorma, M.P.M, Vayrynen, I.J., & Tracz, A. (2007).

Boudiombo, J., Baehr, O., Boudrioua, A., Thevenin, P., Loulergue, J.C., & Bath, A. (1997).

Boughaba, S., Sproule, G.I., McCaffrey, J.P., Islam, M., & Graham, M.J. (2002). Synthesis of

Bragg, W.H., Bragg, W.L., (1913). The structure of the diamond*. Nature,* Vol.91, No.2283,

Brooks, T.A., Hess, D.W. (1987). Plasma-enhanced chemical vapor deposition of silicon

Brooks, T.A., Hess, D.W. (1988). Deposition chemistry and structure of plasma-deposited

Brydson, R., Daniels, H., Fox, M.A., Greatrex R., & Workman, C. (2002). Synthesis of a new

Bulou, S., Le Brizoual, L., Miska, P., de Poucques, L., Hugon, R., & Belmahi, M. (2010). a-

Byon, E., Son, M., Hara, N. & Sugimoto, K. (2004). Corrosion behavior of boron-carbon-

Byon, E.; ., Son, M.; ., Lee, K.-H.; ., Nam, K.-S.; ., Hara, N.; ., Sugimoto, K., (2006).

Cao, Z. X. (2001). Nanocrystalline silicon carbonitride thin films prepared by plasma beamassisted deposition*. Thin Solid Films*, Vol. 401, pp.94-101, ISSN 0040-6090 Cao, Z. X., Oechsner, H. (2003). Optical and mechanical characteristics of nanocrystalline

Caretti, I., Jimenez, I., & Albella J.M. (2003). BCN films with controlled composition obtained

*Films*., Vol.153, No. 1-3, (March 1987), pp.521-529, ISSN 0040-6090

Vol.64, No. 2, (January 1988), pp. 841–849, ISSN 0021-8979

borazines. *Chem. Commun*., No.7, (February 2002), pp. 718–719

*Sci.*, Vol. 253, No. 17, (June 2007), pp. 7211-7218, ISSN 0169-4332

2006), pp. 1650–1658, ISSN 0925-9635

Vol. 402, pp.99–110, ISSN 0040-6090

Briggs, D.,; Seah, M.P., (1990). *Practical Surface Analysis.* Vol.1, Wiley

*Science and Engineering*., Vol. 12, pp.012002-012005

*J. Appl. Phys*. Vol. 93, No. 2, (October 2002), pp. 1186-1189

ISSN 0921-5107

p.557, ISSN 0028-0836.

197-200, ISSN 0040-6090

4686

plasma chemical vapor deposition of silicon carbonitride films from a (dimethylamino)dimethylsilane precursor: Compositional and structural dependencies of film properties*. Diamond Relat. Mater.*, Vol. 15, No. 10, (October

Silicon carbonitride by remote microwave plasma CVD from organosilicon precursor: Growth mechanism and structure of resulting Si:C:N films. *Appl. Surf.* 

Modes of propagating light waves in thin films of boron nitride deposited by plasma enhanced chemical vapor deposition. *Mater. Sci. Engin. B,* Vol.B46, pp.96-98,

amorphous silicon carbonitride films by pulsed laser deposition*. Thin Solid Films*,

nitride from 1,1,3,3,5,5 – hexamethylcyclotrosilazane and ammonia. *Thin Solid* 

silicon nitride films from 1,1,3,35,5 – hexamethylcyclotrosilazane. *J. Appl. Phys*.,

boron carbonitride with a B4C-like structure from the thermolysis of N-alkylated

SiCxNy thin films deposited by a microwave plasma assisted CVD process using a CH4/N2/Ar/HMDSN mixture: methane rate effect*. IOP Conf. Series: Materials* 

nitride films grown by magnetron sputtering. *Thin Solid Films*, Vol. 447-448, pp.

Electrochemical properties of boron-carbon-nitride films formed by magnetron sputtering*. Electrochim.Acta,* Vol.51, No.13 (August 2005), pp.2662-2668, ISSN: 0013-

boron carbonitride lms synthesized by plasma-assisted physical vapor deposition.

by the interaction between molecular beams of B and C with nitrogen ion beams. *Diamond Relat. Mater,Diamond Relat. Mater*. Vol. 12, pp. 1079–1083, ISSN 0925-9635


Compilation on Synthesis, Characterization

ISSN 0022-3697

ISSN 1087-6596

pp.1711-1721

New Delhi, India, November, 2008

pp.1361–1366, ISSN 0022-3093

pp.1397-1401, ISSN 0734-2101

1–12, ISSN 1862-6300

1435, ISSN 0022-3727

0925-9635

Inc., ISBN 0-470-01608-6

Springer, ISBN 3540738851

1995), pp. 1747-1757; ISSN 0008-6223

and Properties of Silicon and Boron Carbonitride Films 531

Fainer, N.I., Kosinova, M.L., Rumyantsev, Yu.M., Maximovskii, E.A., & Kuznetsov, F.A.

Fainer, N.I., Bakovets, V.V., Kosinova, M.L., Rumayntsev, Yu.M., Maximovski, E.A., &

Fainer, N.I., Golubenko, A.N., Rumyantsev, Yu.M., & Maximovskii, E.A. (2009). Use of

Fainer, N.I., Rumyantsev, Yu.M., Kesler, V.G., Maximovski, E.A. & Kuznetsov, F.A. (2009).

Filipozzi, L., Derré, A., Conard, J., Piraux, L., & Marchand, A. (1995). Local order and

Filipozzi, L., Piraux, L., Marchand, A., Derré, A., Adouard, A., & Kinany-Alaoui, M. (1997).

Fitzpatrick, P.R., & Ekerdt, J.G., (2008). Film continuity and interface bonding of thin boron

Freire Jr., F.L., Reigada, D.C., & Prioli, R. (2001). Boron Carbide and Boron–Carbon Nitride

Fu, Yo., Sun*,* C. Q, Du, H., & Yan, B. (2001). Crystalline carbonitride forms harder than the

Fujiwara, H. (2007). *Spectroscopic Ellipsometry: Principles and Applications.* John Wiley & Sons

Fultz, B. & Howe, J. (2007). *Transmission Electron Microscopy and Diffractometry of Materials.* 

Gago, R., Jimenez, I., & Albella J.M. (2000). Boron]carbon]nitrogen compounds grown by ion beam assisted evaporation. *Thin Solid Films* , Vol. 373, pp. 277-281, ISSN 0040-6090 Gago, R., Jimenéz, I., Sajavaara, T., Rauhala, E., & Albella, J.M., (2001). X-Ray absorption

(2008). Thin silicon carbonitride films are perspective low-k materials. *Journal of Physics and Chemistry of Solids*, Vol. 69, No. 2-3, (February-March 2008), pp. 661-668,

Kuznetsov, F.A. (2008). SiCN coating as the protecting layer of quartz crucibles used for silicon crystallization. *Abstracts of 14-th APAM General Assembly and conference "State of materials research and new trends in material science"*, pp. 84-85,

Hexamethylcyclotrisilazane for Preparation of Transparent Films of Complex Compositions*. Glass Physics and Chemistry,* Vol.35, No.3, (August 2008), pp. 274–283,

Synthesis and thermal stability of nanocomposite SiCxNy:H films from cycle siliconorganic precursor. *ECS Trans*., Vol. 25, No. 8, pp.921-926, ISSN 1938-5862 Ferreira, E., Fortunato, I., Vilarinho, P., Viana, A.S., Ramos, A.R., Alves, E., & Martins, R.,

(2006). Hydrogenated silicon carbon nitride films obtained by HWCVD, PA-HWCVD and PECVD techniques. *Journal of Non-Crystalline Solids*, Vol.352, No.9-20,

electrical properties of boron carbonitride films. *Carbon*, Vol. 33, IsNo. 12, (August

Unusual behavior of the magnetoresistance of boron carbonitride films at low temperature. *Journal of Materials Research,* Vol. 12 , IsNo. 07, (December 1996),

carbonitride films on Ge(100) and Si(100). *J. Vac. Sci. Technol.A,* Vol.26, No.6,

Films Deposited by DC-Magnetron Sputtering: Structural Characterization and Nanotribological Properties. *Physica status solidi (a),* Vol. 187, No. 1, (July 2001), pp

hexagonal Si-carbonitride crystallite. *J. Phys. D: Appl. Phys.*, Vol.34, No.9, pp.1430–

studies of cubic boron-carbon-nitrogen films grown by ion beam assisted evaporation. *Diamond and Related. Materials,* Vol.10, No.3-7, pp.1165-1169, ISSN

stability*. J. European Ceramic Society*, Vol. 22, No.16, (February 2002), pp.2969–2979, ISSN 0955-2219


Di Mundo, R., d'Agostino, R., Fracassi, F., & Palumbo, F. (2005). A Novel Organosilicon

*Appl. Surf. Sci., Appl. Surf. Sci.*, Vol. 127-129, pp. 692-696, ISSN 0169-4332 Du*,* X.-W., Fu, Y., Sun, J., Yao, P., & Cui. L. (2007). Intensive light emission from SiCN films

Dubovik, T.V. & Struk, L.I. (1965). Electric insulation tube of boron carbonitride for

Etou, Y., Tai, T., Sugiyama, T., & Sugino, T. (2002). Characterization of boron carbon nitride

Fainer, N.I., Rumyantsev, Yu.M., Kosinova, M.L., Yurjev, G.S., Maximovskii, E.A., &

Fainer, N.I., Kosinova, M.L., Rumyantsev, Yu.M. & Kuznetsov, F.A. (1999). RPECVD thin

Fainer, N.I., Kosinova, M.L., Yurjev, G.S., Rumyantsev, Yu.M. & Asanov, I.P. (2000). The

Fainer, N.I., Kosinova, M.L., & Rumyantsev, Y.M. (2001). Thin films of silicon and boron

Fainer, N.I., Maximovskii, E.A., Kosinova, M.L., & Rumyantsev, Yu.M. (2001). The study of

Fainer, N.I., Maximovskii, E.A., Rumyantsev, Yu.M., Kosinova, M.L. & Kuznetsov, F.A.

films. *NIMA.* Vol. 470, No. 1, (September 2001), pp.193-197, ISSN 0168-9002 Fainer, N.I., Rumyantsev, Yu.M., Golubenko, A.N., Kosinova, M.L., & Kuznetsov, F.A.

Vol. 45, Supplement 1, (January 2004 ), pp. S65-S70, ISSN 0022-4766

Fainer, N., Rumyantsev, Y., Kosinova, M., Maximovski, E., Kesler, V. (2007). Low-k

ISSN 0955-2219

pp.107-110

9002

988, ISSN 0925-9635

614-617, ISSN 0169-4332

Pr8-775, ISSN 1155-4339

XLV, No. 3., pp. 101-108, ISSN ISSN 0373-0247

(April 2007), pp. 9269–9274, ISSN 0257-8972

378-381, pp.493-498, ISSN 0255-5476

pp.456–460, ISSN 0254-0584

stability*. J. European Ceramic Society*, Vol. 22, No.16, (February 2002), pp.2969–2979,

Source for Low Temperature Plasma Deposition of Silicon Nitride-like Thin Films. *Plasma Process. Polym.,* Vol. 2, No. 8 (June, 2005), pp. 612–617, ISSN 1612-8850 Dinescu, M., Perrone, A., Caricato, A.P., Mirenghi, L., Gerardi, C., Ghica, C., & Frunza, L.

(1998). Boron carbon nitride films deposited by sequential pulses laser deposition.

by reactive RF magnetron sputtering. *Materials Chemistry and Physics*. Vol.103,

protection of metal thermocouples. *Translated from Poroshkovaya Metallurgia,* No.5,

films with a low dielectric constant. *Diamond Relat. Mater*, Vol. 11, No. 3-6, pp. 985-

Kuznetsov F.A. (1997). The investigation of properties of silicon nitride films obtained by RPECVD from hexamethyldisilazane. *Appl. Surf. Sci.*, Vol. 113-l 14, pp.

silicon carbonitride films using hexamethyldisilazane*. J. Phys. IV,* V. 9, pp.Pr8-769 -

structure study of thin boron and silicon carbonitride films by diffraction of synchrotron radiation. *NIMA*., Vol. 448, No. 1-2, (June 2000), pp.294-298, ISSN 0168-

carbonitrides: synthesis, composition and structure research. *Russ. Chem. J.*, Vol.

thermal annealing influence on structure and phase composition of silicon carbonitride by diffraction of synchrotron radiation*. Materials Science Forum*, Vol.

(2001). Study of structure and phase composition of nanocrystal silicon carbonitride

(2003). Synthesis of nanocrystalline silicon carbonitride films by remote plasma enhanced chemical vapor deposition using the mixture of hexamethyldisilazane with helium and ammonia*. J. Crys. Growth*., Vol. 248, pp.175-179, ISSN 0022-0248 Fainer, N.I., Kosinova, M.L., Rumyantsev, Yu.M., & Kuznetsov, F.A. (2004). Physical and

chemical properties of silicon carbonitride nanocrystalline films*. J. Struct. Chem.*,

dielectrics on base of silicon carbon nitride films. *Surf. Coat. Technol.,* Vol. 201,


Compilation on Synthesis, Characterization

84, ISSN: 0168-9002

ISSN: 0925-9635.

ISSN 0168-583X

10.1039/B601939K

1996), p.3100-3108, ISSN 0734-2101

1196–1203, ISSN 0169-4332

scientific publications, ISBN 0-19-850091-2

*Ceramics*, Vol. 46, No.11-12, pp.543-549, ISSN 1068-1302

and Properties of Silicon and Boron Carbonitride Films 533

Hemraj-Benny, T., Banerjee, S., Sambasivan, S., Balasubramanian, M., Fischer, D.A., Eres, G.,

*Surf. Coat. Technol.*, Vol.202, No.22-23, pp.5567–5571, ISSN 0257-8972 Hoche H., Pusch C., Riedel R., Claudia F. & Klein A. (2010). Properties of SiCN coatings for

*Surf. Coat. Technol.*, Vol.205, Suppl.1, pp.S21–S27, ISSN 0257-8972

Puretzky, A.A., Geohegan, D.B., Lowndes, D.H., Han, W., Misevich, J.A., & Wong, S.S. (2006). Near-Edge X-ray Absorption Fine Structure Spectroscopy as a Tool for Investigating Nanomaterials*. Small,* Vol.2, No.1, pp.26-35, online ISSN: 1613-6829 Hirai, T., & Goto, T. (1981). Preparation of amorphous Si3N4,-C plate by chemical vapour deposition. *Journal of Materials Science,.* Vol. 16, No.1, pp.17-23, ISSN: 0022-2461 Hoche, H., Allebrandt, D., Bruns, M., Riedel, R. & Fasel, C. (2008). Relationship of chemical

and structural properties with the tribological behavior of sputtered SiCN films.

high temperature applications — Comparison of RF-, DC- and HPPMS-sputtering*.*

Pollakowski, B.; Trunova, V.; Ulm, G.; Weser, J., (2007). Chemical bonding in carbonitride nanolayers. *Nuclear Instruments and Methods A,* Vol.575, No.1-2, pp.78-

structure and properties of hard Si-B-C-N films prepared by reactive magnetron sputtering. *Diamond and Rel. ated Materials*., Vol.16, No.1 (March 2006), pp.29-36,

*determination from powder diffraction data. IUCr monographs on crystallography*, Oxford

silicon carbonitride formed by high-dose nitrogen ion implantation into silicon carbide*. Nuclear Instruments and Methods in Physics Research B*, Vol.206, pp.994–998,

M. V., Karpets, M. V., Stegnii, A. I., & Grishnova L. A. (2007). Hard plasma chemical coatings based on silicon carbon nitride. *Powder Metallurgy and Metal* 

ceramics. *J.Mater.Chem*., Vol.16, No.27, pp.2792-2799,. ISSN 0959-9428DOI:

temperature deposition of cubic BN:C films by unbalanced direct current magnetron sputtering of a B4C target. *J. Vac. Sci. Technol. A,* Vol.14, No.6, (August

silicon carbonitride films by plasma enhanced chemical vapor deposition (PECVD) using bis(dimethylamino)dimethylsilane (BDMADMS), as membrane for a small molecule gas separation. *Appl. Surf. Sci.*., Vol. 257, No. 4, (December 2010), pp.

carbon-nitrogen materials of graphite-like structure. *Materials Research Bulletin,*

Hoffmann, P.; Baake, O.; Beckhoff, B.; Ensinger, W.; Fainer, N.; Klein, A.; Kosinova, M.;

Houška, J.; Vlček, J.; Potocký, Š.; Peřina, V., (2007). Influence of substrate bias voltage on

Ibberson, R.M.; ., David, W.I.F., (2002). Neutron powder diffraction*.* In: *Structure* 

Ishimaru, M., Naito, M., Hirotsu, Y., & Sickafus, K. E. (2003). Amorphous structures of

Ivashchenko, L. A., Ivashchenko, V. I., Porada, O. K., Dub, S. M., Skrinskii, P. L., Ushakov,

Izumi, A., Oda, K. (2006). Deposition of SiCN films using organic liquid materials by HWCVD method*. Thin Solid Films*., Vol.501, No.1-2, pp.195–197, ISSN 0040-6090 Jäschke, T.; Jansen, M., (2006). A new borazine-type single source precursor for Si/B/C/N

Johansson, M. P., Ivanov, I., Hultman, L., Münger, E. P., & Schütze, A. (1996). Low-

Kafrouni, W., Rouessac, V., Julbe, A., & Durand. , J. (2010). Synthesis and characterization of

Kaner, R. B., Kouvetakis, J., Warble, C. E., Sattler, M. L., and & Bartlett, N. (1987). Boron-

Vol. 22, IsNo. 3, (March 1987), pp. 399-404, ISSN: 0025-5408


Gago, R., Jimenéz, I., Albella, J.M., & Terminello, L.J., (2001). Identification of ternary boron-

Gago, R., Jiménez, I., Agulló-Rueda, F., Albella, J.M., Czigány, Zs., & Hultman, L. (2002).

Gago, R.; Jimenéz, I.; Garcia, I.; Albella, J.M., (2002). Growth and characterisation of boron-

Gago, R., Jimenez, I., Kreissig, U., & Albella, J.M. (2002). X-Ray absorption study of the

Gao, P., Xu, J., Piao, Y., Ding, W., Wang, D., Deng, X., & Dong, C. (2007). Deposition of

Gardiner, D.J. (1989). *Practical Raman spectroscopy.* Springer-Verlag, ISBN 978-0387502540 Giessibl, F.J. (2003). Advances in atomic force microscopy. *Reviews of Modern Physics. ,* 

Biomedical Applications of Microprobe Analyses. Academic Press

Goldsmith, J.G.; ., Lazof, D.B.; ., & Linton, R.W. (1999). *Secondary ion mass spectrometry.* 

Goldstein, G.I.; Newbury, D.E.; Echlin, P.; Joy, D.C.; Fiori, C.; & Lifshin, E. (1981). *Scanning* 

Hasegawa, T., Yamamoto, K. & Kakudate, Y. (2002). Synthesis of B-C-N thin films by

Hasegawa, T., Yamamoto, K. & Kakudate, Y. (2003). Influence of raw gases on B-C-N films

Hegemann, D., Riedel, R., Drebler, W., Oehr, C., Schindler, B., & Brunner H. (1997). Boron

Hegemanna, D., Riedel, R., & Oehr C. (1999). Influence of single-source precursors on

Hein, M., Hoffmann, P., Lieser, K.H., & Ortner, H., (1992). Application of X-ray fluorescence

Hellgren, N., Berlind, T., Gueorguiev, G.K., Johansson, M.P., Stafström, S., & Hultman, L.

*electron microscopy and x-ray microanalysis.* Plenum Press, New York, ISBN

electron beam excited plasma CVD. *Diamond and Relat.ed Mater.ials*, Vol. 11, No.3-6,

prepared by electron beam excited plasma CVD. *Diamond Relat.ed Mater.ials*, Vol.

Carbonitride Thin Films by PACVD of Single-Source Precursors. *Chem. Vap.* 

PACVD-derived boron carbonitride thin films. *Thin Solid Films*, Vol. 339, (August

analysis with total-reflection (TXRF) in material science. *Fresenius J.Anal.Chem.,* 

(2004). Fullerene-like B-C-N thin films: a computational and experimental study. *Mater. Sci. Engin. B.,* Vol.113, No.3, (August 2004), pp.242-247, ISSN: 0921-5107

*Diamond and Relat.ed Mater*.ials, Vol. 11, pp. 1295–1299, ISSN 0925-9635 Gammer, K., Kolber, T., Piplits, K., Nowikow, K., Tang, X., Haubner, R., & Hutter, H. (2002).

*Lett.,* Vol.78, (April 2001), pp.3430-3432, ISSN 0003-6951

2002), pp. 5177-5182, ISSN 0021-8979

98–102, ISSN 0040-6090

0257-8972

030640768X

Vol.64, No.3-4, pp.199-204, ISSN 0042-207X

Vol.75, No.3, pp.949-983, ISSN: 0034-6861

12, No.10-11, pp. 1945-1048, ISSN 0925-9635

Vol.343, No.9-10, pp.760-764, ISSN 0937-0633

*Deposition*, Vol. 3, No. 5, pp.257-262, ISSN 1521-3862

pp. 1290-1294, ISSN 0925-9635

1999), pp. 154-159, ISSN 0040-6090

carbon-nitrogen hexagonal phases by x-ray absorption spectroscopy. *Appl. Phys.* 

Transition from amorphous boron carbide to hexagonal boron carbon nitride thin lms induced by nitrogen ion assistance. *J. Appl. Phys*., Vol. 92, No. 9, (August

carbon-nitrogen coatings obtained by ion beam assisted evaporation*. Vacuum,* 

bonding structure of BCN compounds enriched in carbon by CH4 ion assistance.

Characterization of carbon-boron nitride-coatings prepared with chemical vapor deposition by auger electron spectroscopy, electron probe microanalysis and secondary ion mass spectrometry. *Thin Solid Films*, Vol. 406, (December 2001), pp.

silicon carbon nitride thin films by microwave ECR plasma enhanced unbalance magnetron sputtering. *Surf. Coat. Technol..,* Vol. 201, No.9-11, pp.5298–5301, ISSN


Compilation on Synthesis, Characterization

No. 4, (May 2002), pp. 366–373, ISSN 0020-1685

2003-08, pp. 708-715, ISSN 1938-5862

2005-09, pp.1082-1087, ISSN 1938-5862

Vol. 34, IsNo. 1-3, pp. 1-7, ISSN: 0379-6779

No 1, pp. 22-26, ISSN 1068-1302

2000), pp.97-199, ISSN 0927-796X

2005), pp.613-623, ISSN: 0272-4324.

No.16, (February 2009), pp.2343–2349, ISSN 0257-8972

207X.

9810223528

and Properties of Silicon and Boron Carbonitride Films 535

Kosinova, M.L., Fainer, N.I., Rumyantsev, Yu.M., Maximovski, E.A., Kuznetsov, F.A.,

Kosinova, M.L.,.Fainer, N.I., Sulayeva, V.S., Rumyantsev, Yu.M., Kuznetsov, F.A.,

Kosolapova, T.Ya., Makarenko, G.N., Serebryakova, T.I., Prilutskii, E.V., Khorpyakov,

Kouvetakis, J., Sasaki, T., Shen, C., Hagiwara, R., Lerner, M., Krishnan, K. M. & Bartlett, N.

Kroke, E., Li, Y.-L., Konetschny, C., Lecomte, E., Fasel, C. , & Riedel, R. (2000). Silazane

Kulisch, W. (2000). Deposition of Diamond-Like Superhard Materials in the B/C/N System. *Phys. stat. sol*. *(a)*, Vol. 177, No.1, (December 1999), pp. 63-79, ISSN 1862-6300 Kurapov, D., Neuschütz, D., Cremer, R., Pedersen, T., Wuttig, M., Dietrich, D., Marx, G., &

Kurapov, D. and Schneider, J. M. (2005). Structure, Composition, Mechanical, and

Kurmaev, E.Z.; ., Ezhov, A.V.; ., Shamin, S.N.; ., Cherkashenko, V.M.; ., Andreev, Y.G.; ., &

Lau, W.S., (1999). *Infrared characterization for microelectronics*. World Scientific. ISBN

Lei, M. K., Li, Q., Zhou, Z. F., Bello, I., Lee, C. S.& Lee, S. T. (2001). Characterization and

*Films*, Vol. 389, No. 1-2, (march March 2001), pp. 194-199, ISSN 0040-6090 Levy, R.A., Mastromatteo, E., Grow, J.M., Paturi, V., & Kuo, W.P. (1995). Low pressure

*Mater. Res*. Vol. 10, No. 2, (October 1994), pp. 320-327, ISSN 0884-2914

Chemical Vapor Deposition from Trimethylamineborane. *Inorganic Mater*., Vol. 39,

Terauchi, M., Shibata, K., & Satoh, F. (2003). Growth of homogeneous and gradient BCxNy films by PECVD using trimethylamino borane complex. *ECS Proc.,* Vol.

Maximovski, E.A., Cao, Z.X., Terauchi, M., Shibata, K., & Satoh, F. (2005). Synthesis, nanoindentation and AFM studies of CVD boron nitride films». *ECS Proc*., Vol.

O.T., & Chernysheva, O.I. (1971). Nature of boron carbonitride I. Conditions of preparation of boron carbonitride. *Powder Metallurgy and Metal Ceramics*, Vol. 10,

(1989). Novel aspects of graphite intercalation by fluorine and fluorides and new B/C, C/N and B/C/N materials based on the graphite network. *Synthetic Metals,*

derived ceramics and related materials. *Mater. Sci. Eng*., Vol. R26, No.4-6, (January

Schneider, J. M. (2002). Synthesis and mechanical properties of BCN coatings deposited by PECVD. *Vacuum*, Vol. 68, IsNo. 4, July 2002, pp.335-339, ISSN: 0042-

Tribological Properties of BCN Films Deposited by Plasma Enhanced Chemical Vapor Deposition. *Plasma Chemistry and Plasma Processing,* Vol. 25, No. 6, (aApril

Lundström, T., (1997). X-ray Emission Spectra and Structural Models of BCN Materials. *J.Alloys and Compounds,* Vol.248, No.1-2, pp.86-89, ISSN: 0925-8388 Lambrecht, W.L., & Segall, B. (1993). Anomalous band-gap behavior and phase stability of *c*-BN–diamond alloys. *Phys. Rev*. *B*, Vol. B 47, No. 15, pp. 9289, ISSN 1098-0121 Latrasse, L., Lacoste, A., Sánchez-López, J.C., Bès, A., Rayar, M. & Pelletier, J. (2009). High

deposition rates of uniform films in tetramethylsilane-based plasmas generated by elementary microwave sources in matrix configuration*. Surf. Coat. Technol.*, Vol.203,

optical investigation of BCN film deposited by RF magnetron sputtering. *Thin Solid* 

chemical vapor deposition of B-C-N-H films from triethylamine borane complex. *J.* 


Kawaguchi, M., Nozaki, K., Kita, Y., & Doi, M. (1991). Photoluminescence characteristics of

Kawaguchi, M., & Bartlett, N. (1995). Syntheses, Structure, and Intercalation Chemistry of

Kawaguchi, M., Kawashima, T., Nakajima, T. (1996). Syntheses and Structures of New

Kawaguchi, M. (1997). B/C/N Materials Based on the Graphite Network. *Adv.Mater*., Vol.9,

Kawaguchi, M., Sugiyama, A. (1997). Syntheses and characterization of graphite-like

Kawaguchi, M., & Wakukawa, Y. (1999). Synthesis of Graphite-like Material of Composition

Kawaguchi, M., Wakukawa, Y., & Kawano, T. (2001). Preparation and electronic state of

Kim, D. H., Byon, E., Lee, S., Kim, J.-K. & Ruh, H. (2004). Characterization of ternary boron

Kimura, C., Okada, K., Funakawa, Sh., Sakata, Sh., & Sugino, T. (2005). Electron affinity and

Kimura, C., Sota, H., Aoki, H., & Sugino, T. (2009). Atomic bonds in boron carbon nitride

Kosinova, M.L., Fainer, N.I., Rumyantsev, Yu.M., Golubenko, A.N., & Kuznetsov, F.A.

Kosinova, M.L., Rumyantsev, Yu.M., Fainer, N.I., Maximovski, E.A., & Kuznetsov, F.A.

Kosinova, M. L., Rumyantsev, Yu. M., Golubenko, A. N., Fainer, N.I., Ayupov, B.M.,

Keeler, J.; (2005). *Understanding NMR Spectroscopy.* John Wiley & Sons. ISBN 0470017864 Kida, T., Shigezumi, K., Mannan, Md. A., Akiyama, M., Baba, Y., & Nagano, M. (2009).

pp.3926-3930, ISSN 0022-2461

Marcel Dekker, Inc., New York-Basel-Hong Kong.

8, No. 6, pp.1197–1201, ISSN 0897-4756

No.2, (1997) pp. 88-93, ISSN: 1341-1683

No. 8, pp.615-625, ISSN 1521-4095

pp. 259-263, ISSN: 0379-6779

pp. 1143–1146, ISSN 0042-207X

447-448, pp. 192-196, ISSN 0040-6090

pp. Pr3-987- Pr3-994, ISSN 1155-4339

Vol. 14, No. 3-7, pp. 719-723, ISSN: 0925-9635

*Relat. Mater,* Vol. 18, No. 2-3, pp. 478-481, ISSN: 0925-9635

*France*,. Vol. 9, pp. Pr8-915- Pr8-921, ISSN 1155-4339

ISSN: 0008-6223

BN(C,H) prepared by chemical vapour deposition. *J. Mater. Sci*., Vol.26, No.14,

B/C/N Materials Based on the Graphite Network, In: *Fluorine-Carbon and Fluoride-Carbon Materials*. *Chemistry, Physics, and Applications*, Nakajima,T., pp.187-238,

Graphite-like Materials of Composition BCN(H) and BC3N(H). *Chem. Mater.*, Vol.

material of composition BC6N2(H). *Materials Science Research International*, Vol. 3,

BC6N by CVD method at high temperature. *Carbon*, Vol. 37, No. 1, pp.147-163,

graphite-like layered material BC6N. *Synthetic Metals*, Vol. 125, IsNo. 2, July 2001,

Synthesis of boron carbonitride (BCN) lms by plasma-enhanced chemical vapor deposition using trimethylamine borane as a molecular precursor. *Vacuum,* Vol. 83,

carbon nitride films synthesized by RF magnetron sputtering. *Thin Solid Films,* Vol.

field emission characteristics of boron carbon nitride film. *Diamond Relat. Mater*,

films synthesized by remote plasma-assisted chemical vapor deposition. *Diamond* 

(1999). LPCVD boron carbonitride films from triethylamine borane. *J. Phys. IV* 

(2001). The structure study of boron carbonitride lms obtained by use of trimethylamine borane complex. *NIMA*, Vol. 470, pp. 253–257, ISSN 0168-9002 Kosinova, M.L.,. Fainer, N.I,. Sulayeva, V.S,. Rumyantsev, Yu.M,. Terauchi, M.,. Shibata, K,.

Satoh, F., Tanaka, M., & Kuznetsov, F.A. (2001). Structure and composition investigation of RPECVD SiCN and LPCVD BCN films. *J. Phys. IV France*,. Vol. 11,

Dolgovesova, I.P., Kolesov, B.A., Kaichev, V. V., & Kuznetsov, F.A. (2003). Chemical Composition of Boron Carbonitride Films Grown by Plasma-Enhanced Chemical Vapor Deposition from Trimethylamineborane. *Inorganic Mater*., Vol. 39, No. 4, (May 2002), pp. 366–373, ISSN 0020-1685


Compilation on Synthesis, Characterization

ISSN 0257-8972

ISSN: 1387-1811.

3697

1618-2642

and Properties of Silicon and Boron Carbonitride Films 537

Majoulet, O.; ., Alauzun, J.G.; ., Gottardo, L.; ., Gervais, C.; ., Schuster, M.E.; ., Bernard, S.; .,

Mannan, Md.A., Nagano, M., Shigezumi, K., Kida, T., Hirao, N., & Baba, Y. (2008).

Mannan, Md.A.; ., Noguchi, H.; ., Kida, T.; ., Nagano, M.; ., Hirao, M.; ., Baba, Y., (2008).

Mannan, Md.A.; ., Nagano, M.; ., Kida, T.; ., Hirao, N.; ., & Baba, Y.; (2009). Characterization

Mannan, Md.A., Kida, T., Noguchi, H., Nagano, M, Shimoyama, I., Hirao, N., Baba, Y.

Mannan, Md.A., Noguchi, H., Kida T., Nagano, M, Hirao, N., Baba, Y. (2010). Growth and

Mannan, Md.A., Baba, Y., Kida, T., Nagano, M., Shimoyama, I., Hirao, N., & Noguchi, H.,

Martínez, E., Lousa, A. & Esteve, J. (2001). Micromechanical and microtribological properties

Maya L. (1988). Aminoborane polymers as precursors of C-N-B ceramic materials. *J. Am.* 

Maya L. (1988). Semiconducting amorphous film containing carbon, nitrogen and boron. *J.* 

Maya, L. & Harris, L.A., (1990). Pyrolitic deposition of carbon films containing nitrogen and/or boron. *J. Am. Ceram*. *Soc.*, Vol. 73, No.7, (February 1990), pp. 1912-191 Mazumder, M. K., Aoki, H., Masuzumi, T., Hara, M., Watanabe, D., Kimura, C., Fukagawa,

*Electrochem. Soc.,* Vol.135, No.5, pp.1278-1281, ISSN: 0013-4651

*Diamond Relat. Mater,* Vol. 10, No. 9-10, pp. 1892-1896, ISSN: 0925-9635 Martínez, C., Kyrsta, S., Cremer, R. & Neuschütz, D. (2002). Influence of the composition of

Vol. 518, No.15, (November 2009), pp. 4163–4169, ISSN 0040-6090

*Amer. J. Appl. Sci*., Vol. 5, No.6, pp. 736-741, ISSN 1546-9239

Vol.11, No.3, (May 2009), pp.100-105, ISSN: 1369-8001

No.1364, (January 2009), pp.503-507, ISSN 1882-0743

No.6, (September 2010), pp.1780-1786, ISSN: 0040-6090.

*Ceram. Soc.,* Vol.71, No. 12, pp.1104-1109

419-422, ISSN: 0925-9635

sputtering hybrid system*. Surf. Coat. Technol.*, Vol.202, No.22-23, pp.5379–5382,

& Miele, P., (2011). Ordered mesoporous silicoboron carbonitride ceramics from boron-modified polysilazanes: Polymer synthesis, processing and properties*. Microporous and Mesoporous Mater*., Vol.140, No.1-3, (September 2010), pp.40-50,

Characterization of Boron Carbonitride (BCN) Thin Films Deposited by Radiofrequency and Microwave Plasma Enhanced Chemical Vapor Deposition.

Chemical bonding states and local structures of the oriented hexagonal BCN films synthesized by microwave CVD*. Materials Science in Semiconductor Processing,*

of BCN films synthesized by radiofrequency plasma enhanced chemical vapor deposition*. J.Phys.Chem.Solids,* Vol.70, No.1, (August 2008), pp.20-25, ISSN 0022-

(2009). Atomic arrangement, composition and prientattion of hexagonal BCN films synthesized by radiofrequency plasma enhaced CVD*. J. Ceram. Soc. Jap.*, Vol.117,

characterization of stoichiometric BCN lms on highly oriented pyrolytic graphite by radiofrequency plasma enhanced chemical vapor deposition. *Thin Solid Films*,

(2011). Orientation of B-C-N hybrid films deposited on Ni(111) and polycrystalline Ti substrates explored by X-ray absorption spectroscopy*. Thin Solid Films,* Vol.519,

of BCN thin films near the B4C composition deposited by r.f. magnetron sputtering.

BCN films deposited by reactive magnetron sputtering on their properties. *Analytical and Bioanalytical Chemistry*, Vol. 374, No. 4, (July 2002), pp. 709-711, ISSN

M., Umeda, M., Kusuhara, M. & Sugino, T. (2009). Properties of boron carbon nitride (BCN) film after plasma ashing *Diamond Relat. Mater,* Vol. 18, No. 2-3, pp.


Li, W.L., Yang, J.L., Zhao, Y. & Fei, W.D. (2009). Effect of assistant RF plasma on structure

Limmanee, A., Otsubo, M., Sugiura, T., Sato, T., Miyajima, Sh., Yamada, A., & Konagai, M.

Lin, H.Y., Chen, Y.C., Lin, C.Y., Tong, Y.P., Hwa, L.G., Chen, K.H., & Chen. L.C. (2002). Field

Ling, H.,; Wu, J. D.,; Sun, J.;, Shi, W.,; Ying Z. F. & Li, F. M. (2002). Electron cyclotron

Linss, V., Rodil, S. E., Reinke, P., Garnier, M. G., Oelhafen, P., Kreissig, U. & Richter, F.

Linss, V., Barzola-Quiquia, J., Häussler, P. & Richter, F. (2004). Structural properties of d.c.

Lippmann, G., (Coord.), (2006-2010). *Progress Summary 2009.* http://www.google.de/search?client=firefox-a&rls=org.mozilla%3Ad Liu A.Y., Wentzcovitch R.A., & Cohen M.L. (1989). Atomic arrangement and electronic structure of BC2N. *Phys. Rev*. *B*, Vol. B 39, No. 3, pp. 1760-1765, ISSN 1098-0121 Liu, L., Wang, Y., Feng, K., Li, Y., Li, W., Zhao, Ch. & Zhao, Yo. (2006). Preparation of boron

Loeffler, J., Konyashin, I., Bill, J., Uhlig, H., & Aldinger, F. (1997). PACVD nano-crystalline

Loir, A.-S., Pech, D., Steyer, P., Gachon, Y., Heґau, C., & Saґnchez-Loґpez, J. C., (2007).

Lombardi, J.R. & Birke, R.L., (2008). An Unified Approach to Surface-Enhanced Raman

Lousa, A., Esteve, J., Muhl, S. & Martínez, E. (2000). BCN thin films near the B4C

Ma, R., & Bando, Y., (2003). Pyrolytic grown B-C-N and BN nanotubes*. Science & Technology* 

Ma, S., Xu, B., Wu, G., Wang, Y., Ma, F., Ma, D., Xu, K. & Bell, T. (2008). Microstructure and

Vol. 252, No. 12, (June 2005), pp. 4185-4189, ISSN 0169-4332

*Relat.ed Mater*.ials, Vol.6, No.5-7, pp.608-611, ISSN 0925-9635

*Polym.*, Vol. 4, pp.173–179, ISSN 1612-8850

*Mater*, Vol. 9, No. 3-6, pp. 502-505, ISSN: 0925-9635

*of Advanced Materials,* Vol.4, No.5, pp.403-407, ISSN: 1468-6996

target*. J.Alloys Comp.*, Vol.482, pp.317–319, ISSN 0925-8388

No.5, pp.652–655, ISSN 0040-6090

pp.85–91, ISSN 0040-6090

2004), pp.66-75, ISSN 0040-6090

0925-9635.

0040-6090

1932-7447

and properties of SiCN thin films prepared by RF magnetron sputtering of SiC

(2008). Effect of thermal annealing on the properties of a-SiCN:H films by hot wire chemical vapor deposition using hexamethyldisilazane. *Thin Solid Films,* Vol. 516,

emission of nanostructured amorphous SiCN films deposited by reactive magnetron sputtering of SiC in CH4/N2 atmosphere*. Thin Solid Films*, Vol. 416,

resonance plasma-assisted pulsed laser deposition of boron carbon nitride films. *Diamond and Related. Mater.ials,* Vol. 11, No. 9 (February 2002), pp. 1623-1628. , ISSN

(2004). Bonding characteristics of DC magnetron sputtered B–C–N thin films investigated by Fourier-transformed infrared spectroscopy and X-ray photoelectron spectroscopy. *Thin Solid Films,* Vol. 467, No. 1-2, (March 2004), pp. 76-87, ISSN

magnetron sputtered B–C–N thin films and correlation to their mechanical properties: a new empirical formula. *Thin Solid Films*, Vol. 467, No. 1-2, (March

carbon nitride thin films by radio frequency magnetron sputtering. *Appl. Surf. Sci.,*

B-C-N thin films obtained by use of an organoboron precursor*. Diamond and* 

Study of the Si Chemical Bonding and the Semiconductive Behavior of the SiCN Coatings and their Correlation with Anti-Corrosion Properties*. Plasma Process.* 

Spectroscopy*. Journal of Physical Chemistry C*, Vol.112, No.14, pp.5605-5617, ISSN

composition deposited by radio frequency magnetron sputtering. *Diamond Relat.* 

mechanical properties of SiCN hard films deposited by an arc enhanced magnetic

sputtering hybrid system*. Surf. Coat. Technol.*, Vol.202, No.22-23, pp.5379–5382, ISSN 0257-8972


Compilation on Synthesis, Characterization

ISSN 0169-4332

ISSN 0169-4332

379, ISSN 0925-9635

ISSN: 0257-8972

base of boron nitride.

pp.931-939, ISSN 0038-1101

0169-4332

(April 2003), pp. 298-305. , ISSN 0169-4332

and Properties of Silicon and Boron Carbonitride Films 539

Pan, W. J.; ., Sun, J.; ., Ling, H.; ., Xu, N.; ., Ying Z. F., & Wu, J. D. (2003). Preparation of thin

Peng, X., Song, L., Meng, J., Zhang, Y., & Hu. X., (2001). Preparation of silicon carbide nitride

Peng, Y., Zhou, J., Zhang, Z., Zhao, B., & Tan, X. (2010). Influence of radiofrequency power

Perrone, A.; ., Caricato, A. P.; ., Luches, A.; ., Dinescu, M.; ., Ghica, C.; ., Sandu, V., and &

Polo, M.C., Martínez, E., Esteve, J. & Andújar, J,L. (1998). Preparation of B-C-N thin films by

Polo, M.C., Martínez, E., Esteve, J., & Andújar, J,.L. (1999). Micromechanical properties of

*MaterDiamond Relat. Mater*., Vol. 8, No.2-5, pp. 423-427, ISSN 0925-9635 Popov, C., Saito, K., Ivanov, B., Koga, Y., Fujiwara, Sh. and Shanov, V. (1998) Chemical

*Solid Films,* Vol. 312, IsNo. 1-2, (August 1997), pp. 99-105, ISSN: 0040-6090 Popov, C., Saito, K., Yamamoto, K., Ouchi, A., Nakamura, T., Ohana, Yo., & Koga, Yo. (1998)

*Journal of Materials Science*, Vol. 33, No. 5, pp.1281-1286, ISSN: 0022-2461 Probst, D., Hoche, H., Zhou, Y., Hauser, R., Stelzner, T., Scheerer, H., Broszeit, E., Berger, C.,

Ray, S.C.; ., Tsai, H.M.; ., Chiou, J.W.; ., Jan, J.C.; ., Kumar, K.; ., Pong, W.F.; ., Chien, F.Z.; .,

Reigada, D.C. & Freire, F.L. Jr. (2001). Nitrogen incorporation into boron carbide films

Rose, H.H.; (2008). Optics of high-performance electron microscopes*. Technology of Advanced* 

Samsonov, G.V., Semenov, Yu.N., & Borodulin, P.Ya. (1962). Refractory materials on the

Saugnas, F., Teyssandier, F., & Marchand, A. (1992). Characterization of C-B-N solid

Schmolla W., Hartnagel H.L. (1983). Amorphous BN films produced in a double-plasma

*Materials,* Vol.9, No.1, (March 2008), pp.014107, ISSN: 1468-6996

*Soc.,* Vol.75, No.1, pp.161-169, ISSN 1551-2916

200, No. 1-4, (1 October 2005), pp. 355– 359, ISSN 0257-8972

films of carbon-based compounds. *Appl. Surf. Sci.Appl. Surf. Sci.,* Vol. 218, No. 1-4,

thin films by sputtering of silicon nitride target. *Appl. Surf. Sci.*, Vol.173, pp.313-317,

on compositional, structural and optical properties of amorphous silicon carbonitride films*. Appl. Surf. Sci.*, Vol.256, No.7, (September 2009), pp.2189–2192,

Andrei, A. (1998). Boron carbonitride films deposited by pulsed laser ablation. *Appl. Surf. Sci.Appl. Surf. Sci.,* Vol. 133, No. 4, (February 1998), pp. 239-242. , ISSN

r.f. plasma assisted CVD. *Diamond and Relat.ed Mater.ials*, Vol. 7, No.2-5, pp. 376-

BN and B-C-N cotings obtained by r.f. plasma-assisted CVD. *Diamond Relat.* 

vapour deposition of BC2N films and their laser-induced etching with SF6. *Thin* 

Synthesis of nitrogen-rich B–C–N materials from melamine and boron trichloride.

Riedel, R., Stafast, H., & Koke, E. (2005). Development of PE-CVD Si/C/N:H films for tribological and corrosive complex-load conditions*. Surf. Coat. Technol..*, Vol.

Tsai, M.-H.; ., Chattopadhyay, S.; ., Chen, L.C.; ., Chien, S.C.; ., Lee, M.T.; ., Lin, S.T.; ., & Chen, K.H., (2004). X-ray absorption studies of boron-carbon-nitrogen (BxCyNz) ternary alloys*. Diamond Rel.Materials,* Vol.13, No.4-8, pp.1553-1557, ISSN: 0925-9635

deposited by dc-magnetron sputtering: film microstructure and tribological properties. *Surface and & Coatings Technology,* Vol. 142-144, (July 2001), pp.889-893,

solutions deposited from a gaseous phase between 900o and 1050oC. *J Am. Ceram.* 

reactor for semiconductor application. *Solid State Electronics,* Vol.26, No.10,


Mishra, S.K., Bhattacharyya, A.S. (2008). Effect of substrate temperature on the adhesion

Montasser, K., Hattori, S., & Morita, S. (1984). Transparent B-C-N thin films formed by

Montasser, K., Hattori, S., & Morita, S. (1985). Characterization of hard transparent B-C-N-H

Montasser, R., Morita, S., & Hattori, S. (1991). A promising boron-carbon-nitrogen thin film.

Moore, A. W., Strong, S. L., Doll, G. L.,. Dresselhaus, M. S, Spain, I. L., Bowers, C. W., Issi, J.

Morant, C.; ., Prieto, P.; ., Bareño, J.; ., Sanz, J.M.; ., & Elizalde, E., (2006). Hard BCxNythin

Morjan, I., Conde, O., Oliveira, M., & Vasiliu, F. (1999). Structural characterization of CxByNz

Nakao, S., Sonoda, T., Tsugawa, K., Choi, J. & Kato, T. (2009). Effects of nitrogen gas ratio on

sputtering. *Vacuum,* Vol. 84, No. 5, (June 2009), pp. 642-647, ISSN 0042-207X Nakayamada, T., Matsuo, K., Hayashi, Y., Izumi, A. & Kadotani, Y. (2008). Evaluation of

materials. *Thin Solid Films.,*Vol. 516, No.5, pp.656–658, ISSN 0040-6090 Nesládek, M., Vaněček, M., Meykens, K., Haenen, K., Manca, J., De Schepper, L., Pace, E.,

*Metrology for Nanotechnology, Gaithersburg, MD (2004)* 

*Science: An Introduction.* Springer-Verlag, ISBN 3-540-00545-5

*Mat. Sci. Forum.*.,. Vol.54-55. , pp.295-312, ISSN 0255-5476

*Materials Letters,* Vol. 62, No.3, (May 2007), pp.398–402, ISSN: 0167-577X Mishra, S. K. (2009). Nano and Nanocomposite Superhard Coatings of Silicon Carbonitride

6, pp.345–354, ISSN 1546-542X

*Appl. Phys.*, Vol.58, No.8, pp.3185-3189

ISSN 0040-6090

pp.5109-5118

0925-9635

0168-583X.

211, ISSN: 0040-6090

pp. 95-105, ISSN 0040-6090

properties of magnetron sputtered nano-composite Si–C–N hard thin films*.* 

and Titanium Diboride by Magnetron Sputtering. *Int. J. Appl. Ceram. Technol.*, Vol.

plasma chemical vapour deposition. *Thin Solid Films*, Vol.117, No. 4, pp.311-317,

thin films formed by plasma chemical vapor deposition at room temperature. *J.* 

P. and & Piraux, L. (1989) Properties and characterization of codeposited boron nitride and carbon materials. *J. Appl. Phys.*, Vol.65, IsNo. 12, February 1989,

films grown by dual ion beam sputtering*. Thin Solid Films,* Vol.515, No1, pp.207-

(x=0.1 to x=0.2) layers obtained by laser-driven synthesis. *Thin Solid Films*, Vol. 340,

composition and microstructure of BCN films prepared by RF magnetron

corrosion resistance of SiCN films deposited by HWCVD using organic liquid

Pini, A., Verona Rinati, G., Kimura, C., Etou, Y., & Sugino, T. (2001). Study of UV and Subgap Photocurrent Response in Diamond and BCN Thin Films for Detector Applications. *Physica status solidi (a),* Vol. 185, No. 1, pp. 107–113, ISSN: 1862-6300 O´Keefe, M.A.; Allard, L.F., (2004). Sub-Ǻngstrom Electron Microscopy for Sub-Ǻngstrom

Nano-Metrology. *National Nanotechnology Initiative Workshop on Instrumentation and* 

treated with plasma. *Diamond Relat. Mater*, Vol. 15, No. 4-8, pp. 1000-1003, ISSN:

B.; ., Kanngießer, B.; ., Beckhoff, B., (2010). GIXRF-NEXAFS investigations on buried ZnO/Si interfaces: A first insight in changes of chemical states due to annealing of the soecimen. *Nucl.Instr.Meth.B,* Vol.268, No.3-4, pp.370-373, ISSN:

Okada, K., Kimura, C., & Sugino, T. (2006). Surface properties of boron carbon nitride films

Oura, K.; ., Lifshits, V.G.; ., Saranin, A.A.; ., Zotov, A.V.; ., Katayama, M., (2003). *Surface* 

Pagels, M.; ., Reinhardt, F.; ., Pollakowski, B.; ., Roczen, M.; ., Becker, C.; ., Lips, K.; ., Rech,


Compilation on Synthesis, Characterization

ISSN 0257-8972

pp. 487–494, ISSN 0020-1685

pp.5319–5322, ISSN 0169-4332

1117-1121, ISSN: 0925-9635

4936.

Vol. 19, No. 5-6, pp. 545-547, ISSN: 0925-9635

Vol. 12, No. 3-7, pp. 1113-1116, ISSN: 0925-9635

and Properties of Silicon and Boron Carbonitride Films 541

Sugino, T., Kikuchi, J., Kawai, S., Kimura, C., & Aoki, H. (2010). Improved tunneling

Sugiyama, T., Tai, T. & Sugino, T. (2002). Effect of annealing on dielectric constant of boron

Sugiyama, T., Tai, T., Okamoto, A., Yoshitake, M., & Sugino, T. (2003). Influence of the

Sulyaeva, V.S., Rumyantsev, Yu.M., Kosinova, M.L., Golubenko, A.N., Fainer, N.I., &

Sulyaeva, V.S., Kosinova, M.L., Golubenko, A.N., Rumyantsev, Yu.M., Fainer, N.I., &

Sulyaeva, V. S., Kosinova, M. L., Rumyantsev, Yu. M., Golubenko, A. N., Fainer, N. I.,.

Suvorova, A. A., Nunney, T., & Suvorov, A. V. (2009). Fabrication of Si–C–N compounds in

Swain, B. P., Hwang, N. M. (2008). Study of structural and electronic environments of

Tai, T., Sugiyama, T., & Sugino, T. (2003). Growth and characterization of boron carbon

Taylor, R.; ., Hare, J.P.; ., Abdul-Sada, A.K.; ., & Kroto, H.W., (1990). Isolation, separation

Teodorescu, V.S.; ., Luches, A.; ., Dinu, R.; ., Zocco, A.; ., Ciobanu, M.F.; ., Martino, M.; .,

Thamm, T., Körner, K.-U., Bohne, W., Strub, E., Röhrich, J., Stöckel, S., & Marx G. (2005).

Thamm; T., Wett, D., Bohne, W., Strub, E., Röhrich, J., Szargan, R., Marx, G., & Goedel,

*Sci*., Vol. 252, No.1, (March 2005), pp. 223–226, ISSN 0169-4332

*Appl. Phys. Lett.*, Vol.80, No. 22, April 2002, pp.4214-1 - –4214-4

*Transaction*, Vol.25, No. 8, part 1. , pp. 343-348, ISSN 1938-5862

*Research B*, Vol. 267, No. 8-9, pp. 1294–1298, ISSN 0168-583X

*Proces.*, Vol. 69, No. 7, pp.S667-S670. , ISSN 0947-8396

controlled field emission of Boron carbon nitride nanofilm. *Diamond Relat. Mater,* 

carbon nitride films synthesized by plasma-assisted chemical vapor deposition.

annealing process on properties of Cu/BCN/p-Si structure. *Diamond Relat. Mater,*

Kuznetsov, F.A. (2007). Plasma enhanced chemical vapour deposition of BCxNy films prepared from N-trimethylborazine: Modelling, synthesis and characterization. *Surf. Coat. Technol.,* Vol.201, No.22-23, April 2007, pp.9009–9014,

Kuznetsov, F.A. (2009). PECVD BCxNy films from mixtures of N-trimethylborazine with hydrogen and ammonia: modelling, synthesis and characterization*. ECS* 

Alferova, N. I., Ayupov, B. M., Gevko, P. N., Kesler, V. G., Kolesov, B. A., Maksimovskii, E. A., Myakishev, K. G., Yushina, I. V., & Kuznetsov, F. A. (2010). Properties of BCxNy Films Grown by Plasma Enhanced Chemical Vapor Deposition from N-Trimethylborazine–Nitrogen Mixtures. *Inorganic Materials*, Vol. 46, No. 5,

silicon carbide by ion implantation. *Nuclear Instruments and Methods in Physics* 

hydrogenated amorphous silicon carbonitride (a-SiCN:H) films deposited by hot wire chemical vapor deposition*. Appl. Surf. Sci.*, Vol. 254, No.17, (February 2008),

nitride films with low dielectric constant. *Diamond Relat. Mater,* Vol. 12, No. 3-7, pp.

and characterization of the fullerenes C60 and C70: the third form of carbon*. Journal of the Chemical Society, Chemical Communications* Vol.20, pp.1423-1425. , ISSN: 0022-

Sandu, V. and & Dinescu, M. (1999). Influence of the substrate temperature on BCN films deposited by sequential pulsed laser deposition. *App. Phys. A: Mater. Sci.* &

Characterization of PECVD boron carbonitride layers. *Appl. Surf. Sci., Appl. Surf.* 

W.A. (2007). Investigations on PECVD boron carbonitride layers by means of


Schonfelder, H., Aldinger, E., Riedel, R. (1993). Silicon carbonitrides - A novel class of materials, *J. de Physique IV,* Vol. 3, No. C7, pp.1293-1298, ISSN 1155-4339 Shimada, Y., Chikamatsu, K., Kimura, C., Aoki, H., & Sugino, T. (2006). Effect of plasma

Smirnova T.P., Badalian, A.M., Yakovkina, L.V., Kaichev, V.V., Bukhtiyarov, V.I., Shmakov,

Smirnova, T. P., Badalian, A. M., Borisov, V. O., Bakhturova, L. F., Kaichev, V. V.,

Solozhenko, V.L., Andrault, D., Mezouar, M., & Rubie, D.C. (2001). Synthesis of superhard cubic BC2N. *Appl. Phys. Lett*., Vol. 78, No. 10, (November 2000), pp. 1385-1387. Soto, G., Samano, E. C., Machorro, R., & Cota, L. (1998). Growth of SiC and SiCxNy films

Stanishevsky, A., Li, H., Badzian, A., Badzian, T., Drawl, W., Khriachtchev, L. & McDaniel,

Stelzner, Th., Arold, T. M., Falk, F., Stafast, H., Probst, D. & Hoche, H. (2005). Single source

Stockel, S., Weise, K., Dietrich, D., Thamm, T., Braun, M., Cremer, R., Neuschutz, D., & G.

Stöckel, S., Weise, K., Thamm, T., Körner, K.-U., Dietrich, D.,& Marx· G. (2003). Comparative

Sugimoto, Y.; ., Pou, P.; ., Abe, M.; ., Jelinek, P.; ., Perez, R.; ., Morita, S.; ., Custance, O.;

Sugino, T. & Hieda, H. (2000). Field emission characteristics of boron carbon nitride films

Sugino, T., Tai, T. & Etou, Y. (2001). Synthesis of boron nitride film with low dielectric

Sugino, T., Kimura, C. & Aoki, H. (2008). Surface properties and field emission of boron nitride. *Diamond Relat. Mater*, Vol. 17, No. 7-10, pp. 1764-1769, ISSN: 0925-9635

*Diamond Relat. Mater,* Vol. 10, No. 3-7, pp. 1375-1379, ISSN: 0925-9635 Sugino, T., Etou, Y., Tai, T., & Mori, H. (2002). Dielectric constant of boron carbon nitride

No. 3, pp. 1459-1463, ISSN: 0169-4332

Vol.16, pp.1311-1315, ISSN 0734-2101

ISSN 0257-8972

465–471, ISSN 0040-6090

884–890, ISSN 1618-2642

Stöhr, J., (1992). *NEXAFS Spectroscopy.* Springer, Berlin

microscopy. *Nature,* Vol.446, No.7131, pp.64-67.

Vol. 9, No. 3-6, pp. 1233-1237, ISSN: 0925-9635

Vol. 80, No. 4, (November 2001), p. 649-652

Vol. 429, No. 1-2, (April), pp.144–151, ISSN 0040-6090

*Solid Films*, Vol. 398-399, pp. 270-274, ISSN 0040-6090

44, No. 12, (November 2007), pp. 1312–1318, ISSN 0020-1685

treatment on interface property of BCN/GaN structure. *Appl. Surf. Sci.,* Vol. 253,

A.N., Asanov, I.P., Rachlin, V.I., & Fomina, A.N. (2003). SiCN alloys obtained by remote plasma chemical vapour deposition from novel precursors. *Thin Solid Films*.,

Podgorbunskaya, T. A., & Rakhlin V. I. (2008). Plasma-Chemical Deposition of SiCN Films from Volatile N-Bromhexamethyldisilazane. *Inorganic Materials.,* Vol.

by pulsed laser ablation of SiC in Ar and N2 environments. *J. Vac. Sci. Technol. A*,

E. (2001). B-C-N coatings prepared by microwave chemical vapor deposition. *Thin* 

precursors for plasma-enhanced CVD of SiCN films, investigated by mass spectrometry. *Surf. Coat. Technol.*., Vol. 200, No. 1-4, (October 2005), pp. 372– 376,

Marx. (2002). Influence of composition and structure on the mechanical properties of BCN coatings deposited by thermal CVD. *Thin Solid Films*. , Vol. 420 –421, pp.

investigations of structure and properties of BCN coatings deposited by thermal and plasma-enhanced CVD. *Anal. Bioanal. Chem*. Vol. 375, (November 2002), pp.

(2007). Chemical identification of individual surface atoms by atomic force

synthesized by plasma-assisted chemical vapor deposition. *Diamond Relat. Mater,*

constant for its application to silicon ultralarge scale integrated semiconductors.

films synthesized by plasma-assisted chemical-vapor deposition. *Appl. Phys. Lett.,*


Compilation on Synthesis, Characterization

Dekker, New York

2005), pp. 7993-8000, ISSN 0169-4332

ISSN 0002-3264 (in Russian)

4, May 1995, pp.2880-2883

*pages*)

1398

ISSN: 0040-6090

and Properties of Silicon and Boron Carbonitride Films 543

Van Grieken, R.; ., Markowicz, A. (Eds.), (2002). *Handbook of X-ray Spectrometry.* Marcel

Vassallo, E., Cremona, A., Ghezzi, F., Dellera, F., Laguardia, L., Ambrosone, G., & Coscia, U.

Vlcek*,* J., Kormunda, M., Czek, J., Perina, V., & Zemek, J. (2002). Influence of nitrogen–

Voronkov, M.G., Sulimin, A.D., Yachmenev, V.V., Mirskov, R.G., Kokin, V.N., & Chernova,

Wada, Y.; Yap, Y. K.; Yoshimura, M.; Mori, Y. & Sasaki, T. (2000). The control of B---N and

Wang E.G. (1999). A new development in covalently bonded carbon nitride and related materials. *Adv. Mater*., Vol.11. , No.13, pp.1129-1133, ISSN 1521-4095 Wang, Z.-C.; ., Kamphowe, T.W.; ., Katz, S.; ., Peng, J.; ., Seifert, H.J.; ., Bill, J.; ., & Aldinger,

Watanabe, D., Aoki, H., Moriyama, R., Mazumder, M. K., Kimura, C. & Sugino, T. (2008).

Watanabe, M. O., Itoh, S., Mizushima, K., and & Sasaki, T. (1995). Electrical properties of

Watanabe, M. O., Itoh, S., Mizushima, K., and & Sasaki, T. (1996). Bonding characterization

Watanabe, M.O., Sasaki, T., Itoh, S. and & Mizushima, K. (1996) Structural and electrical

Weber, A., Bringmann, U., Klages, C.P., Taube, K., Döllein, G., Meyer, H., & Weidenbach, G.

Weber, A., Bringmann, U., Nikulski, R., & Klages, C.-P. (1993). Growth of cubic boron

Weissenbacher, R., Haubner, R., Aigner, K., & Lux B. (2002). Interactions of Ta-filaments

Weissenbacher, R., Haubner, R. (2003). HF-CVD of B, C, N containing layers on various

No. 2-4, (March 1993), pp. 201-206, ISSN 0925-9635

No.2, (October 2001), pp. 191–197, ISSN 0925-9635

*Mater.ials*, Vol. 12, No.3-7, pp. 884–890, ISSN 0925-9635

*and & Coatings Technology*, Vol.160, pp.74–81, ISSN 0257-8972

*Relat.ed Materials*, Vol. 9, No. 3-6, pp. 620-624. , ISSN 0925-9635

*J.Mater.Sc.Letters,* Vol.19, No.19, pp.1701-1704, ISSN: 0261-8028.

*Diamond Relat. Mater*, Vol. 17, No. 4-5, pp. 669-672, ISSN: 0925-9635

(2006). Structural and optical properties of amorphous hydrogenated silicon carbonitride films produced by PECVD*. Appl. Surf. Sci..* Vol. 252, No. 22, (8 October

argon gas mixtures on reactive magnetron sputtering of hard Si–C–N films. *Surface* 

V.G. (1981). Production of silicon nitride films from hexamethylcyclotrisilazane in high-frequency glow discharge*. Dokl. AN USSR. ,* Vol. 259, No.5, pp.1130-1132,

B---C bonds in BCN films synthesized using pulsed laser deposition. *Diamond and* 

F., (2000). Effects of polymer thermolysis on composition, structure and hightemperature stability of amorphous silicoboron carbonitride ceramics.

Characterization of BCN film after wet process for interconnection integration.

BC2N thin films prepared by chemical vapor deposition. *J. Appl. Phys.,* Vol.78, No.

of BC2N thin films. *Appl. Phys. Lett.,* Vol.68, IsNo. 21, (March 1996), pp.2962-2964 (*3* 

characterization of BC2N thin films. *Thin Solid Films,* Vol. 281-282, pp. 334-336,

(1992). Plasma deposition of BN, BCN:H and Me-BCN:H films using Ntrimethylborazine (Me = Ti, Nb). *J. de Phys. III*, Vol. 2, No.8, (April 1992), pp. 1391-

nitride and boron-carbon-nitrogen coatings using N-trimethylborazine in an electron cyclotron resonance plasma process. *Diamond and Related. Mater.ials,* Vol. 2,

during hot-filament CVD of BCN-layers. *Diamond and Relat.ed Materials,* Vol. 11,

substrates by the decomposition of Triethylborazine. *Diamond and Related.* 

ERDA, XPS and nano-indentation measurements. *Microchim. Acta*., Vol. 156, No.1-2, (April 2006), pp. 53–56, ISSN: 0026-3672


Thärigen, T., Lippold, G., Riede, V., Lorenz, M., Koivusaari, K. J., Lorenz, D., Mosch, S.,

Tomasella, E., Rebib, F., Dubois, M., Cellier, J., & Jacquet, M. (2008). Structural and optical

Tompkins, H.G., (2006). *A User´s Guide to Ellipsometry.* Dover Publications Inc., Mineola,

Torres, R.; ., Caretti, I.; ., Gago, R.; ., Martin, Z.; ., Jimenéz, I., (2007). Bonding structure of

Trusso, S., Barreca, F., & Neri, F. (2002). Bonding configurations and optical band gap for

Tsai, P.-C. (2007). The deposition and characterization of BCN films by cathodic arc plasma

Uddin, Md.N., Shimoyama, I., Baba, Y., Sekiguchi, T., Nath, K.G., & Nagano, M. (2005). B-C-

Uddin, Md.N., Shimoyama, I., Baba, Y., Sekiguchi, T., Nath, K.G. & Nagano, M. (2006).

Uddin, Md.N., Shimoyama, I., Baba, Y., Sekiguchi, T., & Nagano, M. (2006). X-ray

Ulrich, S., Ehrhardt, H., Theel, T., Schwan, J., Westermeyr, S., Scheib, M., Becker, P.,

Ulrich, S., Kratzsch, A., Leiste, H., Scheib, M., Schloßmacher, P., Holleck, H., Binder, J.,

Umeda, Sh., Yuki, T., Sugiyama, T., & Sugino, T. (2004). Boron carbon nitride film with low

*Relat. Mater,* Vol. 13, No. 4-8, pp. 1135-1138, ISSN: 0925-9635

*Series,* Vol.100, (February 2008), pp.082045. ISSN 1742-6596

*Appl. Phys*., Vol. 92, No.5, pp.2485-2489, ISSN 0021-8979

*Sci*. Vol.241, (October 2004), pp. 246-249, ISSN 0169-4332

Vol.99, (February 2006), pp. 084902-1 – 084902-5

1997), pp. 839-844, ISSN: 0925-9635

(April 2006), pp. 53–56, ISSN: 0026-3672

1346, ISSN 0361-5235

ISBN 0-486-45028-7

pp.1450-1454, ISSN: 0925-9635

5108–5113, ISSN 0257-8972

497-502

ISSN: 0257-8972

ERDA, XPS and nano-indentation measurements. *Microchim. Acta*., Vol. 156, No.1-2,

Grau, P., Hesse, R., Streubel, P., & Szargan. R. (1999). Hard amorphous CSixNy thin films deposited by RF nitrogen plasma assisted pulsed laser ablation of mixed graphite/Si3N4-targets. *Thin Solid Films*, Vol. 348, pp.103-113, ISSN 0040-6090 Ting, S.-F., Fang, Y.-K.,. Hsien, W.-T., Tsair, Y.-S., Chang, C.-N., Lin, C.-S., Hsieh, M.-C.,

Chiang, H.-C., & Ho, J.-J. , (2002). Heteroepitaxial Silicon-Carbide Nitride Films with Different Carbon Sources on Silicon Substrates Prepared by Rapid-Thermal Chemical-Vapor Deposition. *Journal of Electronic materials,* Vol. 31, No.12, pp.1341-

properties studies of sputtered a-SiCN thin films. *Journal of Physics: Conference* 

BCN nanopowders prepared by ball milling*. Diamond Rel. Materials,* Vol.16, No.4-7,

nitrogenated amorphous silicon carbide films prepared by pulsed laser ablation*. J.* 

evaporation. *Surf. Coat. Technol.Surf. Coat. Technol.,* Vol. 201, (August 2007), pp.

N hybrid synthesis by high-temperature ion implantation. *Appl. Surf. Sci.,Appl. Surf.* 

Synthesis and characterization of oriented graphitelike B-C-N hybride. *J.Appl. Phys*.,

photoelectron spectroscopic observation on B-C-N hybrids synthesized by ion beam deposition of borazine. *J.Vac.Sci.Technol. A*, Vol.23, No.3, (February 2005), pp.

Oechsner, H., Dollinger, G. & Bergmaier, A. (1998). Phase separation in magnetron sputtered superhard BCN thin films. *Diamond Relat. Mater*, Vol. 7, No. 6, (October

Schild, D., Westermeyer, S., Becker, P., & Oechsner, H. (1999). Variation of carbon concentration, ion energy, and ion current density of magnetron-sputtered boron carbonitride films. *Surface and & Coatings Technology,* Vol. 116-119, pp. 742-750.

dielectric constant as passivation film for high speed electronic devices. *Diamond* 


Compilation on Synthesis, Characterization

295, ISSN

0042-207X

ISSN 0040-6090

4419-0085-2, London, New York.

1579-1585, ISSN: 0925-9635

Vol.10, No.3-7, pp.1137-1141, ISSN: 0925-9635

*of Ceramics*, pp.116-119, Fukuoka, Japan, 1990

98, No. 3, June 2005, pp. 036105-1 - 036105-4

pp.210-217, ISSN: 0040-6090.

*J. Mater. Res*. Vol.14, No.3, Ppp.1137-1141, ISSN 0884-2914

and Properties of Silicon and Boron Carbonitride Films 545

Yap, Y. K.; ., Wada, Y.; ., Yamaoka, M.; ., Yoshimura, M.; ., Mori, Y., & Sasaki, T. (2001). Bond

Yap, Y.K. (Ed.). (2009). *B-C-N Nanotubes and Related Nanostructures*, Springer, ISBN 978-1-

Yap, Y.K., Wada, Y., Yamaoka, M., Yoshimura, M., Mori, Y., & Sasaki, T., (2001). Bond

Yasui, K., Fukuda, K., Katoh, H., & Kaneda, Sh. (1989). The influence of carbon addition on

Ying, Z.F.; ., Yu, D.; ., Ling, H.; ., Xu, N.; ., Lu, Y.F.; ., Sun J., & Wu, J.D. (2007). Synthesis of

Yokomichi, H., Funakawa, T., & Masuda, A. (2002). Preparation of boron–carbon–nitrogen

Yokomichi, H., Futakuchi, T., Yasuoka, M., & Kishimoto, N. (2004). Fabrication of

Yu J., & Wang E.G (1999). Turbostratic boron carbonitride film and its field-emitting behavior. *Appl. Phys. Lett*., Vol.74, No.20, pp.2948-2950, ISSN 0003-6951 Yu J., Wang E.G., & Xu G. (1999). Synthesis and characterization of B-С-N on molybdenum.

Yue, J., Cheng, W., Zhang, X., He, D., & Chen, G. (2000). Ternary BCN thin lms deposited

Yuki, T., Umeda, Sh., & Sugino, T. (2004). Electrical and optical characteristics of boron

*Diamond Relat. Mater,* Vol. 13, Issues No.4-8, pp. 1130-1134, ISSN 0925-9635 Zhang, Y., Sun, H., & Chen, C.F. (2004). Superhard Cubic BC2N Compared to Diamond. *Phys. Rev. Lett*., Vol. 93, No. 19, pp.195504-1 – 195504-4, ISSN 0031-9007 Zhang, Zh., Kimura, C., & Sugino, T. (2005). Wavelength-dependent effect of ultraviolet

Zhou, F., Adachi, K., & Kato, K., (2006). Influence of deposition parameters on surface

*Crystalline Solids*, Vol. 338-340, June 2004, pp. 509-512, ISSN 0022-3093 Yokoshima, S., Goto, T., & Hirai, T. (1990). Dielectric properties of BN-C thick films

*of Non-Crystalline Solids,* Vol. 111, No.1-3, pp.173-177, ISSN 0022-3093 Yasui. H., Hirose, Y., Awazu, K., & Iwaki, M. (2000). The properties of BCN lms formed by

No. 3-7, (December 2001), pp. 1137-1141. , ISSN 0925-9635

modification of BCN films on Ni substrate. *Diamond and Relat.ed Mater.ials*, Vol. 10,

modification of BCN films on Ni substrate. *Diamond Relat. MaterDiamond Rel.Mater*.,

the internal stress and chemical inertness of amorphous silicon-nitride films. *Journal* 

ion beam assisted deposition. *Colloids and Surfaces B: Biointerfaces*, Vol. 19, pp. 291–

BCN thin films by nitrogen ion beam assisted pulsed laser deposition from a B4C target. *Diamond Relat. MaterDiamond Relat. Mater.,* Vol. 16, No. 8 (January 2007), pp.

thin films by magnetron sputtering. *Vacuum*, Vol. 66, No. 3-4, pp. 245-249, ISSN

amorphous boron–carbon–nitrogen films by hot-wire CVD. *Journal of Non-*

prepared by chemical vapor deposition. *Proceed. 28th Symposium on the Basic Science* 

by reactive sputtering. *Thin Solid Films,* Vol. 375, (November 1999), pp. 247-250,

carbon nitride films synthesized by plasma-assisted chemical vapor deposition.

irradiation on dielectric constant of boron carbon nitride films *J. Appl. Phys.,* Vol.

roughness and mechanical properties of boron carbon nitride coatings synthesized by ion beam assisted deposition. *Thin Solid Films,* Vol.497, No. 1-2, (October 2005),


Weissenbacher, R., Haubner, R. (2006). Deposition of B, C, N coatings on WC-Co substrate –

White, R., (1990). *Chromatography/Fourier transform infrared spectroscopy and its applications.* 

Wöhle, J., Ahn, H., & Rie, K.-T., (1999). BCN coatings on polymer substrates by plasma CVD

Wrobel, A.M., Blaszczyk-Lezak, I., & Walkiewicz-Pietrzykowska, A. (2007). Silicon

*Polymer Science.,* Vol. 105, No. 1, (July 2007), pp.122–129, ISSN 0021-8995 Wrobel, A. M., Blaszczyk-Lezak, I., Uznanski, P., & Glebocki, B. (2010). Silicon Carbonitride

*Deposition*., Vol.16, No.7-9 ( September 2010), pp. 211–215, ISSN 0948-1907 Wu, F.; ., Valle, N.; ., Fitzpatrick, R.; ., Ekerdt, J.G.; ., Houssiau, L.; ., Migeon, H.-N., (2011).

using the MCsx+mode. *Surf.Interface Anal.,* Vol. 43, No. 1-2, pp. 669–672 Wu, J.-J., Wu, C.-T., Liao, Y. –C., Lu, T.-R., Chen, L.C., Chen, K.H., Hwa, L.-G., Kuo, C.-T., &

Xiao, X.-Ch., Li,Ya-W., Song, Li-X., Peng, X.-F., & Hu, X.-F. (2000). Structural analysis and

*Thin Solid Films*, Vol. 355-356, pp. 417-422, ISSN 0040-6090

No.5, pp. 1441–1446, ISSN:1533-4880

2009), pp. 1863-1867, ISSN 0257-8972

ISSN: 0022-3093

http://books.google.com/books?id=t2VSNnFoO3wC&pg=PA7

No.5, (November 2005), pp.374-379, ISSN 0263-4368

1171, ISSN 0257-8972

Analytical problems indicating c-BN formation. *Int. J. Ref. Met. Hard Mater*., Vol. 24,

Marcel Dekker, New York.

at low temperature. *Surf. Coat. Technol.Surf. Coat. Technol.,* Vol. 116-119, pp. 1166-

Carbonitride Thin-Film Coatings Fabricated by Remote Hydrogen–Nitrogen Microwave Plasma Chemical Vapor Deposition from a Single-Source Precursor: Growth Process, Structure, and Properties of the Coatings. *Journal of Applied* 

(SiCN) Films by Remote Hydrogen Microwave Plasma CVD from Tris(dimethylamino)silane as Novel Single-Source Precursor. *Chem. Vap.* 

Elaboration and quantitative investigation of BCN-type films by dynamic SIMS

Ling, K.-J. (1999). Deposition of silicon carbon nitride films by ion beam sputtering.

microstructural observation of SiCxNy films be reactive sputtering of SiC in N2 and Ar*. Appl. Surf. Sci.*, Vol. 156, No. 1-4, (February 2000), pp.155-160, ISSN 0169-4332 Xie, Z.Y., Edgar, J.H., McCormick, T.L, & Sidorov, M.V. (1998). The effects of the

simultaneous addition of diborane and ammonia on the hot-filament-assisted chemical vapor deposition of diamond II. Characterization of diamond and BCN film. *Diamond and Relat.ed Mater.ials*, Vol.7, No.9, pp.1357-1363, ISSN 0925-9635 Xu, S.; Ma, X.; Tang, G., (2006). A Study of Nano-Mechanical Properties and Nano-Scratch

Behaviour of Boron Carbonitride Films*. J.Nanoscience and Nanotechnology,* Vol.6,

Flow Ratios by DC Reactive Magnetron Sputtering Plasma Science, *IEEE* 

of nitrogen pressure on structure and optical properties of pulsed laser deposited BCN thin films. *Surf. Coat. Technol.Surf. Coat. Technol.,* Vol. 204, No. 11, (November

carbon-containing cubic boron nitride films by pulsed-DC magnetron sputtering

stabilities in various atmospheres of SiB0.5C1.5N0.5 nano-sized powders fabricated by mechanical alloying technique. *J.Non-Crystalline Solids,* Vol.356, No.6-8, pp.326-333,

Xu, Sh., ., Ma, X., & Su, M. (2006). Investigation of BCN Films Deposited at Various N2/Ar

*Transactions on Issue Date*, Vol. 34, No.4, pp. 1199 – 1203, ISSN 0093-3813. Yang, Q.; ., Wang, C.B.; ., Zhang, S.; ., Zhang, D.M.; ., Shen, Q., & Zhang, L.M. (2010). Effect

Yang, T.-S., Tsai, T.-H., Lee, Ch.-H., Cheng, Ch.-L. , & Wong, M.-Sh. (2001). Deposition of

Yang, Z.-H.; ., Jia D.-C.; ., Duan, X.-M.; ., Zhou, Y., (2010). Microstructure and thermal

*Thin Solid Films,* Vol. 398-399, pp. 285-290, ISSN 0040-6090


Zhou, F., Adachi, K., & Kato, K. (2006). Comparoson of tribological property of a-C, a-CNx

Zhou, F., Adachi, K., & Kato, K. (2006). Friction and wear behavior of BCN coatings sliding

Zhou, F., Yue, B., Wang, X., Wu, X. & Zhuge L. (2010). Surface roughness, mechanical

Zhou, Y., Probst, D., Thissen, A., Kroke, E., Riedel, R., Hauser, R., Hoche, H., Broszeit, E.,

Zhou, Z. F., Bello, I., Lei, M. K., Li, K. Y., Lee, C. S., & Lee, S. T. (2000). Synthesis and

Zhuang, Ch., Zhao, J., Jia, F., Guan, Ch., Wu, Zh., Bai, Y. & Jiang, X. (2009). Tuning bond

*Technol*, Vol. 200, (March 2005), pp. 4471-4478, ISSN 0257-8972

2005), pp. 301–310, ISSN 0043-1648

2005), pp.1325–1335, ISSN 0955-2219

pp. 713-717, ISSN 0257-8972

ISSN 0925-8388

0257-8972

and BCN coatings sliding against SiC balls in water*. Surf. Coat. Technol.Surf. Coat.* 

against ceramic and steel balls in various environments. *Wear*, Vol. 261, (October

properties and bonding structure of silicon carbon nitride films grown by dual ion beam sputtering*. J. Alloys Compounds,* Vol.492, No. 1-2, (March 2010), pp.269–276,

Kroll, P. & Stafast H. (2006). Hard silicon carbonitride films obtained by RF-plasmaenhanced chemical vapour deposition using the single-source precursor bis(trimethylsilyl)carbodiimide*. J.Europ. Ceram. Soc.*, Vol. 26, No. 8, (4 February

characterization of boron carbon nitride films by radio frequency magnetron sputtering. *Surf. Coat. Technol.Surf. Coat. Technol*, Vol. 128-129, pp. 334-340, ISSN

contents in B–C–N films via temperature and bias voltage within RF magnetron sputtering. *Surf. Coat. Technol.Surf. Coat. Technol*, Vol. 204, No 5, (September 2009),

### *Edited by Moumita Mukherjee*

Silicon Carbide (SiC) and its polytypes, used primarily for grinding and high temperature ceramics, have been a part of human civilization for a long time. The inherent ability of SiC devices to operate with higher efficiency and lower environmental footprint than silicon-based devices at high temperatures and under high voltages pushes SiC on the verge of becoming the material of choice for high power electronics and optoelectronics. What is more important, SiC is emerging to become a template for graphene fabrication, and a material for the next generation of sub-32nm semiconductor devices. It is thus increasingly clear that SiC electronic systems will dominate the new energy and transport technologies of the 21st century. In 21 chapters of the book, special emphasis has been placed on the "materials" aspects and developments thereof. To that end, about 70% of the book addresses the theory, crystal growth, defects, surface and interface properties, characterization, and processing issues pertaining to SiC. The remaining 30% of the book covers the electronic device aspects of this material. Overall, this book will be valuable as a reference for SiC researchers for a few years to come. This book prestigiously covers our current understanding of SiC as a semiconductor material in electronics. The primary target for the book includes students, researchers, material and chemical engineers, semiconductor manufacturers and professionals who are interested in silicon carbide and its continuing progression.

Silicon Carbide - Materials, Processing and Applications in Electronic Devices

Silicon Carbide

Materials, Processing and Applications

in Electronic Devices

*Edited by Moumita Mukherjee*

Photo by michal812 / iStock