# Bond order

when SiH4 is positively charged.

SiH4 –

SiH4 +

Si2H6 –

Si2H6 +

SiH6 –

SiH4 –

Si2H6 –

Si2H6 +

bond strength.

Moreover, there are many processes where ion or electron beams are used to enhance the kinetics at low temperature. For example, high quality films can be grown at low temperature by ion-beam-assisted deposition (IBAD) in the sputtering or evaporation processes. Also, there is a process called gas-mediated electron or ion-beam-induced deposition and etching, where beams of electrons and ions are used to modify a surface locally at micron and submicron dimensions or fabricating in three dimensions [42–44]. In all these processes, the enhanced kinetics is not clearly understood, vaguely explained by the bombarding energy of ions or electrons. However, it is highly probable that the enhanced kinetics should come from weakening of the bond strength by charge. Moreover, the effect of charge on weakening the bond strength and thereby enhancing the kinetics seems to be very general, even related with the catalytic effect and the enzyme activity in biology.

Zheng et al. [45] observed the superplastic deformation behavior of nanoscale amorphous silica nearroom temperature using in situ experiments inside a TEM with low beam intensities without obvious rise in sample temperature. They called the phenomenon electron-assisted superplasticity. After imaging for the positioning of the sample and the diamond flat punch, the beam was blocked with the condenser lens aperture, and the silica particle was com‐ pressed with the Hysitron Pico-indenter [46, 47]. The particle is plastically deformed because of the e-beam irradiation it had experienced during the imaging. After ∼40% compression [46], where the contact pressure is estimated to be 9.2 GPa, the beam was brought on the sample to image the particle. On unloading, the total plastic percent compression was 27%. On compressive loading again with the beam on, surprisingly, the plastic flow of the glass was continuous and smooth, with no sign of shear banding or cracking. Much easier flow was observed with the ensuing beam-on deformation in a second set of in situ experi‐ ments, resulting in a pancake shape. Although the contact area kept increasing, the forces required to deform were at levels considerably lower than those in the beam-off condition. This phenomenon of electron-assisted superplasticity might also be explained by weaken‐ ing of bond strength by charge.

### **3. Non classical crystallization in diamond deposition at low pressure**

The concept that thin films grow by the building blocks of CNPs was first suggested in the low pressure synthesis of diamonds by Hwang et al. [7]. In this section, it will be briefly reviewed how the TCN was developed as the growth mechanism of diamonds in its low pressure synthesis.

**Figure 4.** Thermodynamic and kinetic description of the metastable diamond formation. Reprinted with permission from [10]. Copyright 2010 Elsevier.

Low pressure deposition of diamond was reported by pioneering scientists Derjaguin and Fedoseev [48], Spitsyn et al. [49] in HWCVD and Matsumoto et al. [50] in PECVD. Although its process has been studied in depth [51–54], the underlying principle has not been clearly understood. The fundamental question is why diamond can be deposited at low pressure whereas graphite is more stable than diamond.

This question can be explained by **Figure 4** [10]. **Figure 4** describes the general thermodynamic and kinetic explanation of the case where the metastable phase can be formed more dominantly than the stable phase The Gibbs free energy of the stable phase is lower than that of the metastable phase. This difference in free energy determines the difference in the driving force of formation. In other words, the driving force for the formation of the stable phase from the unstable phase is higher than that of the metastable phase.

However, the kinetic barrier, such as a nucleation barrier, is a more important factor that determines the tendency of the formation since the formation of the stable and the metasta‐ ble phases is a kinetically parallel process. The formation of metastable diamond at low pressure might be explained by this concept. To check this possibility, the nucleation barrier between graphite and diamond should be compared. The Gibbs free energy of nucleation is composed of the driving force for precipitation and the surface energy of the nucleus. Isotropic surface energies of diamond [55] and graphite [56] are respectively 3.7 J/m2 and 3.1 J/m2 . Since the surface energy of graphite is smaller than that of diamond and the driving force for the precipitation of graphite from the gas phase is larger than that of diamond, the total Gibbs free energy of diamond is higher than that of graphite in all ranges of radius as shown in **Figure 5(a)**. Diamond is less stable than graphite in all ranges of radius in **Figure 5(a)**. Considering only **Figure 5(a)**, it appears that the formation of diamond at low pressure cannot be approached by the concept of the metastable phase formation of **Figure 4**. However, **Figure 5(a)** does not correctly compare the Gibbs free energy between diamond and graph‐ ite because the comparison should be made with respect to the number of atoms as shown in **Figure 5(b)** [57] instead of being made with respect to the radius. Although the surface energy of diamond is larger than that of graphite, the molar volume of diamond (3.41 cm3 / mole) is much smaller than that of graphite (5.405 cm3 /mole). Therefore, the surface energy multiplied by the molar area for diamond (8.38 × 10−4 J) is smaller than that for graphite (9.55 × 10−4 J).

**Figure 4.** Thermodynamic and kinetic description of the metastable diamond formation. Reprinted with permission

32 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Low pressure deposition of diamond was reported by pioneering scientists Derjaguin and Fedoseev [48], Spitsyn et al. [49] in HWCVD and Matsumoto et al. [50] in PECVD. Although its process has been studied in depth [51–54], the underlying principle has not been clearly understood. The fundamental question is why diamond can be deposited at low pressure

This question can be explained by **Figure 4** [10]. **Figure 4** describes the general thermodynamic and kinetic explanation of the case where the metastable phase can be formed more dominantly than the stable phase The Gibbs free energy of the stable phase is lower than that of the metastable phase. This difference in free energy determines the difference in the driving force of formation. In other words, the driving force for the formation of the stable phase from the

However, the kinetic barrier, such as a nucleation barrier, is a more important factor that determines the tendency of the formation since the formation of the stable and the metasta‐ ble phases is a kinetically parallel process. The formation of metastable diamond at low pressure might be explained by this concept. To check this possibility, the nucleation barrier between graphite and diamond should be compared. The Gibbs free energy of nucleation is composed of the driving force for precipitation and the surface energy of the nucleus. Isotropic surface energies of diamond [55] and graphite [56] are respectively 3.7 J/m2 and 3.1 J/m2

the surface energy of graphite is smaller than that of diamond and the driving force for the precipitation of graphite from the gas phase is larger than that of diamond, the total Gibbs free energy of diamond is higher than that of graphite in all ranges of radius as shown in **Figure 5(a)**. Diamond is less stable than graphite in all ranges of radius in **Figure 5(a)**. Considering only **Figure 5(a)**, it appears that the formation of diamond at low pressure cannot be approached by the concept of the metastable phase formation of **Figure 4**. However, **Figure 5(a)** does not correctly compare the Gibbs free energy between diamond and graph‐ ite because the comparison should be made with respect to the number of atoms as shown

. Since

from [10]. Copyright 2010 Elsevier.

whereas graphite is more stable than diamond.

unstable phase is higher than that of the metastable phase.

**Figure 5.** (a) Dependence of Gibbs free energy on the radius of diamond and graphite nanoparticle. (b) Dependence of Gibbs free energy on the number of carbon atoms for diamond and graphite. Reprinted with permission from [57]. Copyright 1997 Elsevier.

The number of atoms at which the Gibbs free energies of diamond and graphite intersect as shown in **Figure 5(b)** can be derived as,

$$n^\* = 36\pi \left( \frac{\sigma\_{du} \left( \Omega\_{du} \right)^{\frac{2}{3}} - \sigma\_{yu} \left( \Omega\_{yu} \right)^{\frac{2}{3}}}{\Delta \mu^{du \to yu}} \right)^3 \tag{2}$$

where *σ*dia and *σ*gra are respectively the surface energies of diamond and graphite, *Ω*dia and *Ω*gra respectively the atomic volumes of diamond and graphite, and Δ*μ*dia→gra the free energy difference per atom between diamond and graphite. The reported surface energies of diamond and graphite are 3.7 Jm−2 and 3.1 Jm−2, respectively. The chemical potential difference between

diamond and graphite, Δ*μ*dia→gra, at 927°C is calculated to be −1.2101×10−20 J/atom. From these values, *n*\* for (Eq. 2) is estimated to be 351. This means that for a nucleus containing less than 351 atoms, diamond is more stable than graphite. Although the nucleation curves in **Figure 5(b)**, which are evaluated from the reported surface energy data of some uncertain‐ ty, have uncertainties, they definitely tell that diamond can be more stable than graphite for a sufficiently small size. In **Figure 5(b)**, the nucleation barrier of diamond is slightly larger than that of graphite, indicating that graphite would nucleate more dominantly than diamond. However, it should be noted that the difference of the nucleation barrier between diamond and graphite is rather small. If the nucleation barrier of diamond was slightly smaller than that of graphite, diamond synthesis at low pressure would be much easier than now and gas activation such as hot filament and plasma would not be necessary. Oxygen-deficient burning of any carbon-containing materials such as coals, fuels, and woods might have produced diamonds instead of soot.

In reality, graphitic or amorphous carbon is formed without gas activation. It is known that gas activation is essential to the low-pressure synthesis of diamond. If we assume that the role of gas activation is to reduce the surface energy of diamond, the nucleation of diamond can be more dominant than that of graphite. For example, if the gas activation reduces the surface energy of diamond by 10%, *n*\* in (Eq. 2) becomes 1784 at 927°C, and the nucleation barrier of diamond becomes lower than that of graphite as shown in **Figure 5(b)**.

Hwang et al. [7, 58] suggested that the essential role of gas activation is to generate an abundant amount of electric charge rather than atomic hydrogen and that the negative charge stabilizes diamond over graphite. They further suggested that charged diamond nanoparticles are generated in the gas phase, becoming the building blocks of diamond films. They reached this conclusion from the fact that the well-established experimental observation of simultaneous diamond deposition and graphite etching violates the second law of thermodynamics if diamond deposition occurs by individual atoms.

However, in the diamond CVD community, the 'atomic hydrogen hypothesis', which was suggested by Spitsyn et al. [49], has been the most popular explanation for diamond growth at low pressure. According to the hypothesis, atomic hydrogen, which is produced by gas activation such as hot filament or plasma, etches graphite much faster than diamond and therefore low-pressure synthesis of metastable diamond is possible. This hypothesis is equivalent to saying that even though diamond is less stable than graphite, diamond can grow dominantly over graphite in the presence of atomic hydrogen.

However, this statement violates the second law of thermodynamics because if stable graphite etches away, less stable diamond must etch away. Although the atomic hydrogen hypothesis has a significant drawback of violating the second law of thermodynamics as to the irreversible transfer of carbon atoms [7], it is widely accepted in the diamond CVD community largely because simultaneous diamond deposition and graphite etching are experimentally observed. In other words, the experimental observation appears to violate the second law. This is a very interesting situation. Great attention should be paid to this seeming contradiction between the experimental observation and the second law because it is expected that if such a contradiction is solved, a big discovery may be made.

Let's make a rigorous thermodynamic analysis of the simultaneous diamond deposition and graphite etching. The criterion for the irreversible transfer of atoms between the system and the surrounding, which corresponds to an open system in thermodynamics, is the chemical potential. Chemical potential is defined as the partial derivative of the Gibbs free energy with respect to the number of atoms under constant temperature and pressure [59]. The total Gibbs free energy is decreased when atoms transfer from a phase with high chemical potential to a phase with low chemical potential.

diamond and graphite, Δ*μ*dia→gra, at 927°C is calculated to be −1.2101×10−20 J/atom. From these

34 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

351 atoms, diamond is more stable than graphite. Although the nucleation curves in **Figure 5(b)**, which are evaluated from the reported surface energy data of some uncertain‐ ty, have uncertainties, they definitely tell that diamond can be more stable than graphite for a sufficiently small size. In **Figure 5(b)**, the nucleation barrier of diamond is slightly larger than that of graphite, indicating that graphite would nucleate more dominantly than diamond. However, it should be noted that the difference of the nucleation barrier between diamond and graphite is rather small. If the nucleation barrier of diamond was slightly smaller than that of graphite, diamond synthesis at low pressure would be much easier than now and gas activation such as hot filament and plasma would not be necessary. Oxygen-deficient burning of any carbon-containing materials such as coals, fuels, and woods might have produced

In reality, graphitic or amorphous carbon is formed without gas activation. It is known that gas activation is essential to the low-pressure synthesis of diamond. If we assume that the role of gas activation is to reduce the surface energy of diamond, the nucleation of diamond can be more dominant than that of graphite. For example, if the gas activation reduces the surface

Hwang et al. [7, 58] suggested that the essential role of gas activation is to generate an abundant amount of electric charge rather than atomic hydrogen and that the negative charge stabilizes diamond over graphite. They further suggested that charged diamond nanoparticles are generated in the gas phase, becoming the building blocks of diamond films. They reached this conclusion from the fact that the well-established experimental observation of simultaneous diamond deposition and graphite etching violates the second law of thermodynamics if

However, in the diamond CVD community, the 'atomic hydrogen hypothesis', which was suggested by Spitsyn et al. [49], has been the most popular explanation for diamond growth at low pressure. According to the hypothesis, atomic hydrogen, which is produced by gas activation such as hot filament or plasma, etches graphite much faster than diamond and therefore low-pressure synthesis of metastable diamond is possible. This hypothesis is equivalent to saying that even though diamond is less stable than graphite, diamond can grow

However, this statement violates the second law of thermodynamics because if stable graphite etches away, less stable diamond must etch away. Although the atomic hydrogen hypothesis has a significant drawback of violating the second law of thermodynamics as to the irreversible transfer of carbon atoms [7], it is widely accepted in the diamond CVD community largely because simultaneous diamond deposition and graphite etching are experimentally observed. In other words, the experimental observation appears to violate the second law. This is a very interesting situation. Great attention should be paid to this seeming contradiction between the experimental observation and the second law because it is expected that if such a contradiction

diamond becomes lower than that of graphite as shown in **Figure 5(b)**.

in (Eq. 2) becomes 1784 at 927°C, and the nucleation barrier of

for (Eq. 2) is estimated to be 351. This means that for a nucleus containing less than

values, *n*\*

diamonds instead of soot.

energy of diamond by 10%, *n*\*

diamond deposition occurs by individual atoms.

is solved, a big discovery may be made.

dominantly over graphite in the presence of atomic hydrogen.

Since graphite is more stable than diamond at the given temperature and pressure of the CVD process according to the phase diagram of carbon [47, 48], the chemical potential of carbon in diamond is higher than that in graphite, which can be written as,

$$
\mu\_{\text{gra}} > \mu\_{\text{c}}^{\text{gra}} \,, \tag{3}
$$

where the subscript *'C'* and the superscripts *'dia'* and *'gra'* mean carbon, diamond, and graphite, respectively. Since reversible etching and deposition are not driven by the difference of the chemical potential and do not produce a net flux, the reversible etching and deposition can occur simultaneously. On the other hand, irreversible etching or deposition is driven by the chemical potential difference and produces a net flux. The irreversible etching and deposition cannot occur simultaneously.

The experimental observation of simultaneous diamond deposition and graphite etching indicates that the chemical potential of carbon in diamond is lower than that in gas and that the chemical potential of carbon in graphite is higher than that in gas. The irreversible graphite etching can be written as

$$
\mu\_{\varepsilon}^{\text{ga}} > \mu\_{\varepsilon}^{\text{ga}},
\tag{4}
$$

where the superscript *'gas'* represents the gas phase. Similarly, the irreversible diamond deposition can be written as

$$
\mu\_{\text{c}}^{\text{gas}} > \mu\_{\text{c}}^{\text{dia}} \tag{5}
$$

In this paper, etching and deposition refer to the irreversible process and from now on, those words will be used without the adjective 'irreversible'. The simultaneous diamond deposition and graphite etching indicate that Eqs. (4) and (5) should be simultaneously satisfied and can be written as

$$
\mu\_{\rm c}^{\rm gra} > \mu\_{\rm c}^{\rm gas} > \mu\_{\rm c}^{\rm gra},
\tag{6}
$$

Therefore, the chemical potential of carbon in graphite is larger than that in diamond, which is identical to saying that diamond is more stable than graphite. This contradicts with Eq. (3). Therefore, it is quite evident that the simultaneous diamond deposition and graphite etching or the atomic hydrogen hypothesis violates the second law of thermodynamics.

The contradiction becomes more evident if chemical potentials of carbon for different phases are represented in diagrams as shown in **Figure 6**. **Figure 6** compares chemical potentials of carbon for graphite, diamond, and gas based on the carbon phase diagram (**Figure 6(a)**) and the atomic hydrogen hypothesis or the experimental observation of simultaneous diamond deposition and graphite etching (**Figure 6(b)**).

In the interpretation of this paradoxical observation by the atomic hydrogen hypothesis, it is implicitly assumed that the unbalanced etching rate between graphite and diamond can change the stability between graphite and diamond. This hypothesis neglects the fact that thermodynamic stability is not affected by kinetics. It should be noted that the thermodynamic stability determined by the Gibbs free energy is a state function, which depends only on the initial and final states, not on the kinetic path.

**Figure 6.** (a) Chemical potentials of carbon for different phases and the direction of carbon atoms indicated by the car‐ bon phase diagram and (b) those indicated by simultaneous diamond deposition and graphite etching based on the assumption that the building block should be an individual atom. Reprinted with permission from [10]. Copyright 2010 Elsevier.

Since the second law of thermodynamics cannot be violated, there must be some logical error in interpreting the experimental observation of simultaneous diamond deposition and graphite etching. What would be the logical error? One thing that should be noted is that etching of diamond and graphite occurs under the condition where the concentration of CH4 is high enough to deposit both phases. In order to understand this situation, let's consider the phase diagram of the C-H system as shown in **Figure 7** [6].

Therefore, the chemical potential of carbon in graphite is larger than that in diamond, which is identical to saying that diamond is more stable than graphite. This contradicts with Eq. (3). Therefore, it is quite evident that the simultaneous diamond deposition and graphite etching

The contradiction becomes more evident if chemical potentials of carbon for different phases are represented in diagrams as shown in **Figure 6**. **Figure 6** compares chemical potentials of carbon for graphite, diamond, and gas based on the carbon phase diagram (**Figure 6(a)**) and the atomic hydrogen hypothesis or the experimental observation of simultaneous diamond

In the interpretation of this paradoxical observation by the atomic hydrogen hypothesis, it is implicitly assumed that the unbalanced etching rate between graphite and diamond can change the stability between graphite and diamond. This hypothesis neglects the fact that thermodynamic stability is not affected by kinetics. It should be noted that the thermodynamic stability determined by the Gibbs free energy is a state function, which depends only on the

**Figure 6.** (a) Chemical potentials of carbon for different phases and the direction of carbon atoms indicated by the car‐ bon phase diagram and (b) those indicated by simultaneous diamond deposition and graphite etching based on the assumption that the building block should be an individual atom. Reprinted with permission from [10]. Copyright

Since the second law of thermodynamics cannot be violated, there must be some logical error in interpreting the experimental observation of simultaneous diamond deposition and graphite etching. What would be the logical error? One thing that should be noted is that etching of diamond and graphite occurs under the condition where the concentration of CH4 is high enough to deposit both phases. In order to understand this situation, let's consider the

phase diagram of the C-H system as shown in **Figure 7** [6].

or the atomic hydrogen hypothesis violates the second law of thermodynamics.

36 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

deposition and graphite etching (**Figure 6(b)**).

initial and final states, not on the kinetic path.

2010 Elsevier.

**Figure 7.** Phase diagram of the C–H system under 2700 Pa. The graphite and diamond lines indicate the maximum solubility of carbon in the gas phase. Reprinted with permission from [6]. Copyright 1996 Elsevier.

The vertical dashed line is for the composition of 1% CH4-99% H2, which is typically used in the gas-activated diamond CVD process. At the substrate temperature around 1200 K (927 °C), the vertical line is inside the two-phase region of gas + solid, which means that both diamond and graphite have a driving force for precipitation. This prediction is opposite to the experi‐ mental observation of graphite etching. This disagreement between the phase diagram and the experimental observation comes from the implicit assumption that the gas-phase nuclea‐ tion does not take place. If gas-phase nucleation takes place, the equilibrium carbon content in the gas phase follows the solubility line in **Figure 7**. If diamond nucleates in the gas phase, the carbon content in the gas phase would be the solubility line for diamond. If graphite nucleates in the gas phase, the carbon content in the gas phase would be the solubility line for graphite. Carbon solubility in graphite is lower than that in diamond because graphite is more stable than diamond. This solubility tends to be minimum at ∼1300 K and increases with decreasing temperature; it has a retrograde solubility. If diamond or graphite nucleates in the gas phase, carbon in the gas phase would be depleted. When this carbon-depleted gas approaches the substrate, at the temperature of which equilibrium carbon solubility is much higher, etching occurs for both diamond and graphite.

Therefore, the experimental observation of graphite etching indicates that the gas-phase nucleation of diamond or graphite occurred. Besides, in order to explain the experimental observation of simultaneous diamond deposition and graphite etching without violating the second law of thermodynamics, the nuclei should have a diamond structure and at the same time those diamond nuclei should be the building blocks of the growing diamond. According to this scenario, both diamond and graphite are etched away by the atomic unit but diamond deposition occurs by the building block of gas phase-nucleated nanoparticles. This scenario is free from the thermodynamic paradox. Therefore, the puzzling phenomenon of simultaneous diamond deposition and graphite etching provides strong evidence that diamond deposition should occur by the building blocks of gas phase-nucleated nanoparticles.

To check this scenario, the gas-phase nucleation in the diamond CVD system must be exper‐ imentally confirmed. To confirm the existence of CNPs generated in the gas phase, Jeon et al. [11] attached an energy analyzer combined with a Wien filter to the diamond hot filament CVD (HFCVD) reactor. Since the processing pressure in diamond HFCVD is higher than the required pressure for Wien filter measurements, differential pumping was done through an orifice (1.2 mmØ) and a skimmer (2 mmØ) between the measuring chamber and the CVD reactor. By such a system, they could measure the size distribution of negatively charged carbon nanoparticles in the gas phase shown in **Figure 8** [11]. The processing condition of measurement was a gas mixture of 1.5% CH4−98.5% H2, a reactor pressure of 800 Pa, and a wire temperature of 2100°C. The peak occurred at ∼3000 atomic mass units, which corresponds to ∼250 carbon atoms. Based on the negative current measured on the Faraday cup, they estimated the number density of negatively charged nuclei as ∼10<sup>6</sup> mm−3.

**Figure 9** shows the paradox-free direction of carbon flux before and after gas-phase nucleation [10]. Before gas-phase nucleation, the gas phase is supersaturated and the driving force is for the deposition of both diamond and graphite (**Figure 9(a)**). On the other hand, after gas-phase nucleation, the gas phase is undersaturated and the driving force is for etching of both diamond and graphite (**Figure 9(b)**). After gas-phase nucleation, however, the gas phase is undersatu‐ rated and has the driving force for etching of both diamond and graphite (**Figure 9(b)**). Since both diamond and graphite are etched away by the atomic unit under such conditions, the atomic contribution to diamond growth is negative. Nevertheless, diamond crystals grow under such conditions. This means that the growing diamond is contributed 100% by the building blocks of gas-phase nucleated diamond nanoparticles, whereas the diamond film

**Figure 8.** Mass distribution of negatively charged carbon nanoparticles generated in the HWCVD reactor. Reprinted with permission from [11]. Copyright 2000 Elsevier.

deposited on the substrate is being etched atomically. In order for diamond films to grow by the building blocks of CNPs, it must be assumed that CNPs should undergo self-assembly and epitaxial recrystallization. And in order to undergo epitaxial recrystallization, CNPs should be liquid-like. This liquid-like property would come from the weakening of bond strength by charge as explained in the previous section.

diamond deposition and graphite etching provides strong evidence that diamond deposition

To check this scenario, the gas-phase nucleation in the diamond CVD system must be exper‐ imentally confirmed. To confirm the existence of CNPs generated in the gas phase, Jeon et al. [11] attached an energy analyzer combined with a Wien filter to the diamond hot filament CVD (HFCVD) reactor. Since the processing pressure in diamond HFCVD is higher than the required pressure for Wien filter measurements, differential pumping was done through an orifice (1.2 mmØ) and a skimmer (2 mmØ) between the measuring chamber and the CVD reactor. By such a system, they could measure the size distribution of negatively charged carbon nanoparticles in the gas phase shown in **Figure 8** [11]. The processing condition of measurement was a gas mixture of 1.5% CH4−98.5% H2, a reactor pressure of 800 Pa, and a wire temperature of 2100°C. The peak occurred at ∼3000 atomic mass units, which corresponds to ∼250 carbon atoms. Based on the negative current measured on the Faraday cup, they

**Figure 9** shows the paradox-free direction of carbon flux before and after gas-phase nucleation [10]. Before gas-phase nucleation, the gas phase is supersaturated and the driving force is for the deposition of both diamond and graphite (**Figure 9(a)**). On the other hand, after gas-phase nucleation, the gas phase is undersaturated and the driving force is for etching of both diamond and graphite (**Figure 9(b)**). After gas-phase nucleation, however, the gas phase is undersatu‐ rated and has the driving force for etching of both diamond and graphite (**Figure 9(b)**). Since both diamond and graphite are etched away by the atomic unit under such conditions, the atomic contribution to diamond growth is negative. Nevertheless, diamond crystals grow under such conditions. This means that the growing diamond is contributed 100% by the building blocks of gas-phase nucleated diamond nanoparticles, whereas the diamond film

**Figure 8.** Mass distribution of negatively charged carbon nanoparticles generated in the HWCVD reactor. Reprinted

with permission from [11]. Copyright 2000 Elsevier.

mm−3.

should occur by the building blocks of gas phase-nucleated nanoparticles.

38 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

estimated the number density of negatively charged nuclei as ∼10<sup>6</sup>

**Figure 9.** The chemical potentials diagram of carbon in the diamond, graphite, and gas phases before and after gasphase nucleation. Reprinted with permission from [10]. Copyright 2010 Elsevier.

On the other hand, there is another puzzling phenomenon in the diamond CVD process: under the same deposition condition where diamond films grow on a silicon substrate, highly porous skeletal graphitic soot particles grow on an iron substrate and carbon nanotubes grow on a palladium substrate as shown in **Figure 10** [7, 61]. These results imply that the deposition mechanisms of diamond, soot, and carbon nanotubes should be closely related to one another.

From the previous analysis based on **Figures 4**–**9**, we know that diamond nucleates in the gas phase. Besides, **Figure 8** confirms that these nuclei are negatively charged. If this knowledge is combined with **Figure 10**, negatively charged diamond nuclei deposit as a diamond film, soot, and carbon nanotubes respectively on silicon, iron and palladium substrates. Since etching takes place atomically, the building blocks of the diamond film, soot, and carbon nanotubes must be negatively charged nanoparticles. In relation with this effect of the substrate, Huh et al. [61] studied the correlation between the deposition behavior and the charge transfer rate of the substrate and found out that there exists an almost perfect correlation that the substrate with a low charge transfer rate produces diamond whereas the substrate with a high charge transfer rate produces soot or carbon nanotubes.

According to Hwang et al. [7], when negatively charged diamond nuclei approach the silicon substrate which has a low charge transfer rate, they maintain a diamond phase and undergo self-assembly like nano-colloids and epitaxial recrystallization, producing the diamond film as shown in **Figure 10(a)**. However, when they approach the iron substrate, which has a high charge transfer rate, they lose charge to the iron substrate just before landing and transform to graphite. And the neutral graphite nanoparticles make random Brownian aggregations and produce porous skeletal soot on the iron substrate as shown in **Figure 10(b)**. In this scenario, the stability of diamond comes from charge. In other words, when carbon nanoparticles are charged, they have a diamond structure; otherwise, they have a graphite structure. In order to confirm the effect of charge on the stability of diamond, Hwang et al. [7] compared the deposition behavior between floating and grounded iron substrates after 2 hours of deposition. On the grounded iron substrate, soot particles continued to grow. On the floating iron substrate, however, diamond crystals were grown on the initially formed soot particles.

TEM observation of the soot particles by Huh et al. [61] revealed that iron nanoparticles are trapped inside each soot particle. In other words, each soot particle encloses an iron nanopar‐ ticle three dimensionally like an onion. If iron nanoparticles are not enclosed three dimen‐ sionally by carbon, it might be possible that carbon nanotubes grow. From the fact that carbon nanotubes are evolved on the palladium substrate as shown in **Figure 10(c)**, an insight can be drawn as to the growth mechanism of carbon nanotubes [62]. One possible driving force for the growth of one-dimensional carbon would be the reduction of the electrostatic energy. This possibility is supported by the bias effect on the growth of carbon nanotubes reported by Lee and Hwang [63]. They found out that carbon nanotubes grow only when the substrate is positively biased in the hot filament CVD process using a high methane concentration of 20% CH4−80% H2. This possibility is further supported by the experimental observation that onedimensional carbon such as carbon nanotubes or carbon fibers grow when the catalytic metal nanoparticles are placed on the insulating substrate such as glass, but do not grow when they are placed on the conducting substrate [64].

**Figure 10.** (a) Diamond, (b) soot, and (c) carbon nanotubes grown respectively on Si, Fe, and Pd substrates. Reprinted with permission from [7, 61]. Copyright 1996 Elsevier.

### **4. Deposition behavior and analysis of non-classical crystallization in the CVD process**

self-assembly like nano-colloids and epitaxial recrystallization, producing the diamond film as shown in **Figure 10(a)**. However, when they approach the iron substrate, which has a high charge transfer rate, they lose charge to the iron substrate just before landing and transform to graphite. And the neutral graphite nanoparticles make random Brownian aggregations and produce porous skeletal soot on the iron substrate as shown in **Figure 10(b)**. In this scenario, the stability of diamond comes from charge. In other words, when carbon nanoparticles are charged, they have a diamond structure; otherwise, they have a graphite structure. In order to confirm the effect of charge on the stability of diamond, Hwang et al. [7] compared the deposition behavior between floating and grounded iron substrates after 2 hours of deposition. On the grounded iron substrate, soot particles continued to grow. On the floating iron substrate, however, diamond crystals were grown on the initially formed soot particles.

40 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

TEM observation of the soot particles by Huh et al. [61] revealed that iron nanoparticles are trapped inside each soot particle. In other words, each soot particle encloses an iron nanopar‐ ticle three dimensionally like an onion. If iron nanoparticles are not enclosed three dimen‐ sionally by carbon, it might be possible that carbon nanotubes grow. From the fact that carbon nanotubes are evolved on the palladium substrate as shown in **Figure 10(c)**, an insight can be drawn as to the growth mechanism of carbon nanotubes [62]. One possible driving force for the growth of one-dimensional carbon would be the reduction of the electrostatic energy. This possibility is supported by the bias effect on the growth of carbon nanotubes reported by Lee and Hwang [63]. They found out that carbon nanotubes grow only when the substrate is positively biased in the hot filament CVD process using a high methane concentration of 20% CH4−80% H2. This possibility is further supported by the experimental observation that onedimensional carbon such as carbon nanotubes or carbon fibers grow when the catalytic metal nanoparticles are placed on the insulating substrate such as glass, but do not grow when they

**Figure 10.** (a) Diamond, (b) soot, and (c) carbon nanotubes grown respectively on Si, Fe, and Pd substrates. Reprinted

are placed on the conducting substrate [64].

with permission from [7, 61]. Copyright 1996 Elsevier.

### **4.1. Comparison of deposition behavior between floating and grounded substrates in the silicon CVD process**

The generation of CNPs in the gas phase was experimentally confirmed using a differential mobility analyzer (DMA) in many CVD systems [16, 17, 65]. Hwang et al. [9, 10] suggested that these CNPs are the building blocks of thin films and nanostructures in most CVD processes. Experimental confirmation of CNPs in the gas phase is not sufficient to say that they are the building blocks. However, the Si-Cl-H system has retrograde solubility like the C-H system [66]. Besides, in the CVD of the Si-Cl-H system, a paradoxical phenomenon of simul‐ taneous deposition and etching of silicon was reported [67]. This phenomenon also violates the second law of thermodynamics if the silicon film grows by individual atoms. In order to avoid the violation, the charged silicon nuclei formed in the gas phase should be the building blocks of growing silicon films [9, 10].

**Figure 11.** FESEM images: (a) low-magnification and (b) high-magnification images of a floating silicon substrate and (c) low-magnification and (d) high-magnification images of a grounded silicon substrate at a N2 flow rate of 500 sccm. Reprinted with permission from [64]. Copyright 2014 Elsevier.

In order to show CNPs to be building blocks of thin films or nanostructures in a silicon CVD process, Youn et al. [64] compared the deposition behavior between floating and grounded silicon substrates. Charge build-up would be maximized on the floating substrate and minimized on the grounded substrate. The result is shown in **Figure 11**.

**Figure 11(a)** and **(b)** shows respectively low- and high-magnification FESEM images of the surface microstructure of the floating substrate after deposition for 2 hours. The substrate temperature is 900°C and the gas flow rates are 5 sccm 10% SiH4−90% He, 50 sccm H2, and 500 sccm N2. **Figure 11(c)** and **(d)** shows respectively low- and high-magnification FESEM images of the surface microstructure of the grounded substrate with the deposition conditions being the same as those of **Figure 11(a)** and **(b)**. Silicon nanowires grew extensively on the floating substrate as shown in **Figure 11(a)** and **(b)**, whereas no silicon nanowire grew but only silicon nanoparticles were deposited on the grounded substrate as shown in **Figure 11(c)** and **(d)**. Diameters of the nanowires in **Figure 11(b)** are about 10–30 nm. In **Figure 11(b)**, without catalytic metals or seeds of silicon oxide, silicon nanowires were grown on the substrate. Neither vapor-liquid-solid (VLS) [68] nor oxide-assisted growth (OAG) [69] mechanism can explain this deposition behavior.

Floating and grounding of the substrate are the only differences in the processing condition between **Figure 11(b)** and **(d)**. Therefore, charge build-up would be the only difference between **Figure 11(b)** and **(d)**. Considering this, the growth of silicon nanowires would be attributed to the charge build-up. In other words, the electrostatic interaction between the CNPs and the substrate is responsible for the growth of the silicon nanowires.

**Figure 12.** FESEM images for (a) plane view and (b) cross section of films deposited on a floating silicon substrate and (c) plane view and (d) cross section of films deposited on a grounded silicon substrate at a N2 flow rate of 1000 sccm. Reprinted with permission from [64]. Copyright 2014 Elsevier.

When the N2 flow rate was increased from 500 sccm to 1000 sccm, films instead of nano‐ wires were grown as shown in **Figure 12**. **Figure 12(a)** and **(b)** shows respectively FESEM images of the surface and the cross section of a film deposited on a floating silicon sub‐ strate. **Figure 12(c)** and **(d)** shows those deposited on a grounded substrate. The deposition condition was the same as that of **Figure 11** except the gas flow rate of N2. The film on the grounded substrate in **Figure 12(c)** and **(d)** was much more porous than that on the float‐ ing substrate in **Figure 12(a)** and **(b)**. Therefore, quite different microstructures are evolved between the floating and grounded substrates, indicating that the dense film in **Figure 12(a)** and **(b)** resulted from the electrostatic interaction between the CNPs and the growing surface. The film thicknesses of **Figure 12(b)** and **(d)** were respectively ∼220 and ∼190 nm, indicat‐ ing that the growth rate of the film on the floating substrate was higher than that on the grounded substrate. Besides, the film in **Figure 12(b)** is much denser than that in **Fig‐ ure 12(d)**. Therefore, the mass increase on the floating substrate in **Figure 12(b)** is much higher than that on the grounded substrate in **Figure 12(d)**. This study also indicates that the deposition behavior of charged silicon nanoparticles differs dramatically between floating and grounded substrates due to the difference in the accumulation of charge between floating and grounded substrates. These results imply that charge plays a critical role in the evolu‐ tion of dense films and smooth nanowires.

temperature is 900°C and the gas flow rates are 5 sccm 10% SiH4−90% He, 50 sccm H2, and 500 sccm N2. **Figure 11(c)** and **(d)** shows respectively low- and high-magnification FESEM images of the surface microstructure of the grounded substrate with the deposition conditions being the same as those of **Figure 11(a)** and **(b)**. Silicon nanowires grew extensively on the floating substrate as shown in **Figure 11(a)** and **(b)**, whereas no silicon nanowire grew but only silicon nanoparticles were deposited on the grounded substrate as shown in **Figure 11(c)** and **(d)**. Diameters of the nanowires in **Figure 11(b)** are about 10–30 nm. In **Figure 11(b)**, without catalytic metals or seeds of silicon oxide, silicon nanowires were grown on the substrate. Neither vapor-liquid-solid (VLS) [68] nor oxide-assisted growth (OAG) [69] mechanism can

42 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Floating and grounding of the substrate are the only differences in the processing condition between **Figure 11(b)** and **(d)**. Therefore, charge build-up would be the only difference between **Figure 11(b)** and **(d)**. Considering this, the growth of silicon nanowires would be attributed to the charge build-up. In other words, the electrostatic interaction between the

**Figure 12.** FESEM images for (a) plane view and (b) cross section of films deposited on a floating silicon substrate and (c) plane view and (d) cross section of films deposited on a grounded silicon substrate at a N2 flow rate of 1000 sccm.

When the N2 flow rate was increased from 500 sccm to 1000 sccm, films instead of nano‐ wires were grown as shown in **Figure 12**. **Figure 12(a)** and **(b)** shows respectively FESEM images of the surface and the cross section of a film deposited on a floating silicon sub‐ strate. **Figure 12(c)** and **(d)** shows those deposited on a grounded substrate. The deposition condition was the same as that of **Figure 11** except the gas flow rate of N2. The film on the grounded substrate in **Figure 12(c)** and **(d)** was much more porous than that on the float‐

Reprinted with permission from [64]. Copyright 2014 Elsevier.

CNPs and the substrate is responsible for the growth of the silicon nanowires.

explain this deposition behavior.

### **4.2. Comparison of deposition behavior between conducting and non-conducting surfaces**

The concept of non-classical crystallization during PECVD was suggested by Cabarrocas et al. [23, 24, 70]. They reported that silicon nanoparticles formed in the gas phase during the PECVD process are incorporated into a polymorphous structure. They also suggested that high-quality films can be deposited at a high growth rate by incorporating those nanoparticles. In order to confirm whether nanoparticles, which are generated in the gas phase, should be charged or not, Cabarrocas et al. [70] examined the deposition behavior on 5×5 cm<sup>2</sup> glass substrates partially coated with Cr. They also compared the deposition behavior between the glass substrates with and without hydrogen plasma pretreatment as shown in **Figure 13**.

In the case of the glass substrate without any plasma treatment, no film was deposited on the uncoated part as shown in the right of **Figure 13**. In both cases of left and right in **Figure 13**, film deposition took place only on the Cr-coated part, which is a kind of selective deposi‐ tion. On the other hand, for the glass substrate exposed to hydrogen plasma, the entire surface of the substrate was deposited as shown in the left of **Figure 13**. These results strongly support their hypothesis that the building blocks of the deposition are charged.

**Figure 13.** Image of two deposition behavior on partly Cr-coated glass substrates with a hydrogen plasma pretreat‐ ment (left) and without plasma pretreatment (right). Reprinted with permission from [70]. Copyright 2008 Elsevier.

### **4.3. Effect of flow rate on the deposition behavior in the silicon CVD process**

The effect of various parameters in the CVD process should be newly understood based on the new paradigm of non-classical crystallization by CNPs. In this section, the effect of a carrier gas flow rate on the deposition behavior will be examined. Youn et al. [71] varied the flow rate of nitrogen carrier gas during atmospheric pressure silicon CVD and observed the micro‐ structure evolution of films or nanostructures as shown in **Figure 14**.

**Figure 14(a)–(d)** shows the FESEM images of the surface microstructure of films or nanostruc‐ tures synthesized on a quartz substrate respectively at nitrogen flow rates of 300, 500, 700, and 1000 sccm. At 300 sccm (**Figure 14(a)**), nanowires were grown with nanoparticles attached to nanowires. At 500 sccm (**Figure 14(b)**), nanowires were grown extensively. At 700 sccm (**Figure 14(c)**), a film was grown at first and then a porous structure evolved on the initial film. At 1000 sccm (**Figure 14(d)**), a relatively dense film was grown.

**Figure 14.** FESEM images of silicon films and nanostructures. The various microstructures of silicon evolved on a quartz substrate at N2 gas flow rates of (a) 300, (b) 500, (c) 700, and (d) 1000 sccm. Reprinted with permission from [71]. Copyright 2013 Elsevier.

These results indicate that the carrier gas flow rate greatly affects microstructure evolution. To examine the effect of the carrier gas flow rate on the generation of CNPs in the gas phase, insitu measurements of the CNPs were made using a DMA-FCE (Faraday cup electrometer) system under the same conditions as those of **Figure 14**. **Figure 15(a)** and **(b)** shows respectively the size distribution of negatively and positively charged silicon nanoparticles at various carrier flow rates. As the flow rate decreased, the number concentration of negative and positive CNPs decreased but the peak of the size distribution increased. In the case of positive CNPs in **Figure 15(b)**, the number concentration of ∼40 nm nanoparticles at 1000 sccm of N2 is roughly 100 times larger than that at 500 sccm. When the carrier flow rate is lower than 500 sccm, nanoparticles smaller than 20 nm were not detected.

**4.3. Effect of flow rate on the deposition behavior in the silicon CVD process**

44 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

structure evolution of films or nanostructures as shown in **Figure 14**.

At 1000 sccm (**Figure 14(d)**), a relatively dense film was grown.

Copyright 2013 Elsevier.

The effect of various parameters in the CVD process should be newly understood based on the new paradigm of non-classical crystallization by CNPs. In this section, the effect of a carrier gas flow rate on the deposition behavior will be examined. Youn et al. [71] varied the flow rate of nitrogen carrier gas during atmospheric pressure silicon CVD and observed the micro‐

**Figure 14(a)–(d)** shows the FESEM images of the surface microstructure of films or nanostruc‐ tures synthesized on a quartz substrate respectively at nitrogen flow rates of 300, 500, 700, and 1000 sccm. At 300 sccm (**Figure 14(a)**), nanowires were grown with nanoparticles attached to nanowires. At 500 sccm (**Figure 14(b)**), nanowires were grown extensively. At 700 sccm (**Figure 14(c)**), a film was grown at first and then a porous structure evolved on the initial film.

**Figure 14.** FESEM images of silicon films and nanostructures. The various microstructures of silicon evolved on a quartz substrate at N2 gas flow rates of (a) 300, (b) 500, (c) 700, and (d) 1000 sccm. Reprinted with permission from [71].

These results indicate that the carrier gas flow rate greatly affects microstructure evolution. To examine the effect of the carrier gas flow rate on the generation of CNPs in the gas phase, insitu measurements of the CNPs were made using a DMA-FCE (Faraday cup electrometer) system under the same conditions as those of **Figure 14**. **Figure 15(a)** and **(b)** shows respectively the size distribution of negatively and positively charged silicon nanoparticles at various carrier flow rates. As the flow rate decreased, the number concentration of negative and positive CNPs decreased but the peak of the size distribution increased. In the case of positive CNPs in **Figure 15(b)**, the number concentration of ∼40 nm nanoparticles at 1000 sccm of N2

**Figure 14** shows that the amount of deposition on the substrate in terms of mass is decreased with decreasing flow rate of the carrier gas. This aspect would be related with **Figure 15**, which shows that the number concentration of CNPs generated in the gas phase is decreased with decreasing flow rate. The study indicated that the microstructure evolution of films, nano‐ wires, and nanoparticles is closely related to the number concentration of CNPs, which is affected by the carrier gas flow rate.

**Figure 15.** The number concentration and size distribution of (a) negative and (b) positive CNPs at various N2 gas flow rates of 300, 500, 700, and 1000 sccm. Reprinted with permission from [71]. Copyright 2013 Elsevier.

Then why do the amount of the deposition and the number concentration of CNPs decrease with decreasing flow rate? What is found is that silicon tends to deposit on the tube wall near the inlet in the temperature zone of 600–800 °C, which is higher than the decomposition temperature of SiH4. It is further observed that the amount of silicon deposited on the tube wall increases with decreasing flow rate of the carrier gas. This means that the loss of silicon on the tube wall before the hot zone at the center of the reactor is responsible for the decrease in amount of deposit on the substrate and the number concentration of charged gas-phase nuclei with decreasing flow rate.

With decreasing flow rate of the carrier gas, the mean residence time of the gas would be increased in the reactor although the total amount was decreased due to the loss on the tube wall before the hot zone. The slight increase of the particle diameter of maximum size with decreasing flow rate as shown in **Figure 15** might come from the increased mean residence time. Therefore, the deposition on the tube wall depends on the flow rate, which can be understood by the drag force imposed on CNPs by impinging molecules. When nanoparticles are the building blocks of crystal growth, the drag force should be considered in understanding the deposition behavior. Even under the condition where CNPs are generated in abundance, they might not be able to deposit because of the drag force if the flow rate in the direction parallel to the substrate is high enough. In relation to the drag force, Park et al. [72] studied the effect of the bias on the deposition behavior of ZnO nanostructures as follows.

### **4.4. Effect of bias on the deposition behavior of ZnO nanostructures in the CVD process**

It was often found that even under the condition where CNPs are generated abundantly in the gas phase, the deposition of films or nanostructures does not occur on the substrate in specific locations of the CVD. Since non-deposition of films or nanostructures under the generation condition of CNPs would be caused by the drag force, the drag force is an important factor to be considered in the deposition of CNPs. Even when the deposition by CNPs was inhibited by the drag force, the existence of CNPs may be revealed by applying the electrical bias to the substrate. If the bias overcomes the drag force, CNPs that could not deposit on a substrate should be attracted toward the substrate and the deposition occurs. Park et al. [72] confirmed this possibility by examining the bias effect on the deposition behavior in the CVD reactor where charged ZnO nanoparticles are generated in the gas phase.

**Figure 16.** Size distribution of charged ZnO nanoparticles that were generated in the gas phase during the synthesis of ZnO nanostructures. Reprinted with permission from [72]. Copyright 2015 Elsevier.

**Figure 16** shows the size distribution of ZnO CNPs measured by the DMA-FCE system. This result indicates that the number of CNPs in gas phase of the CVD system is large enough. However, they do not land on the substrate placed in 450 °C zone, which is rather near the outlet, being away from the center of the reactor. Then, why don't they land on the substrate placed in 450 °C zone as shown in **Figure 17(a)**?

One possible answer would be the drag force. The drag force is influenced by the relative velocity of particles in the medium [73]. Considering gas flow dynamics in the reactor, the hot gas that comes from the heated reactor zone would move upward over the colder gas in the unheated zone because of the density difference arising from the temperature gradient. This gas stream would exert a drag force on the nanoparticles that are generated in the hot zone so that they would move upward over the gas of 450 °C zone. To test this possibility, the electric bias was applied to the substrate holder.

parallel to the substrate is high enough. In relation to the drag force, Park et al. [72] studied

**4.4. Effect of bias on the deposition behavior of ZnO nanostructures in the CVD process**

It was often found that even under the condition where CNPs are generated abundantly in the gas phase, the deposition of films or nanostructures does not occur on the substrate in specific locations of the CVD. Since non-deposition of films or nanostructures under the generation condition of CNPs would be caused by the drag force, the drag force is an important factor to be considered in the deposition of CNPs. Even when the deposition by CNPs was inhibited by the drag force, the existence of CNPs may be revealed by applying the electrical bias to the substrate. If the bias overcomes the drag force, CNPs that could not deposit on a substrate should be attracted toward the substrate and the deposition occurs. Park et al. [72] confirmed this possibility by examining the bias effect on the deposition behavior in the CVD reactor

**Figure 16.** Size distribution of charged ZnO nanoparticles that were generated in the gas phase during the synthesis of

**Figure 16** shows the size distribution of ZnO CNPs measured by the DMA-FCE system. This result indicates that the number of CNPs in gas phase of the CVD system is large enough. However, they do not land on the substrate placed in 450 °C zone, which is rather near the outlet, being away from the center of the reactor. Then, why don't they land on the substrate

ZnO nanostructures. Reprinted with permission from [72]. Copyright 2015 Elsevier.

placed in 450 °C zone as shown in **Figure 17(a)**?

the effect of the bias on the deposition behavior of ZnO nanostructures as follows.

46 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

where charged ZnO nanoparticles are generated in the gas phase.

The result is shown in **Figure 17**. As previously reported, charged tetrapod ZnOs as well as charged ZnO nanoparticles are generated in the gas phase during the synthesis of ZnO nanostructures [74]. This fact helps us understand the results of **Figure 17**. When +100 V was applied to the substrate holder of stainless steel, ZnO nanoparticles of ∼10 nm were deposited not in a little amount on the silicon substrate as shown in **Figure 17(b)**. When the bias was increased to +300 V, slightly larger ZnO nanoparticles of ∼40 nm were deposited in an appreciable amount as shown in **Figure 17(c)**. When the bias was increased even further to +600 V, tetrapod ZnOs were deposited as shown in **Figure 17(d)**. SEM observations at higher magnification revealed that tetrapod-ZnOs were deposited together with smaller ZnO nanoparticles. **Figure 17(b)–(d)** shows that as the bias voltage increased, the size of particles that could be deposited increased. This bias experiment shows that the charged ZnO nano‐ particles are under the influence of the drag force and explains why CNPs measured in **Figure 16** did not land on the substrate in **Figure 17(a)**. A drag force exerting on particles by the gas flow is expressed as

$$F\_D = 1 / \, 2C\_D A \rho \mathbf{v}^2 \tag{7}$$

where *FD* = drag force, *CD* = drag coefficient, *A* = cross sectional area perpendicular to the flow, *ρ* = density of the medium, *ν* = velocity of the body relative to the medium. This equation reveals that particles with a large area have a larger drag force and thereby larger resistance to move toward the substrate than those with a smaller area. Therefore, in order to deposit large particles, a high bias voltage is needed because of the high drag force arising from the upward gas stream caused by the temperature gradient. Non-deposition in **Figure 17(a)** can be understood on the basis of the drag force: under zero bias voltage, ZnO particles generated in the gas phase would just flow out with the carrier gas because the drag force inhibited the particles from moving toward the substrate. Also, under the bias of +100 V in **Figure 17(b)**, small particles of ∼10 nm could be deposited, but large particles would flow out with the carrier gas. Because the drag force for small charged particles is weak, only a small electric force is needed to overcome the drag force of the small particles, which results in the deposition of small charged particles on the substrate. Under the bias of +300 V in **Figure 17(c)**, slightly larger particles of ∼40 nm could be deposited with much larger particles being flowed out with the carrier gas. Under the bias of +600 V in **Figure 17(d)**, however, even larger tetrapod particles with a leg length of ∼600 nm could be deposited on a substrate, indicating that the electric force is high enough to overcome the drag force. The gravity of these nanosized particles appeared to be much smaller than the electric force and could be ignored. These results indicated that the drag force becomes an important factor in deposition when nano‐ particles are formed in the gas phase.

**Figure 17.** Deposition behavior of charged ZnO particles under the positive biases of (a) 0 V, (b) +100 V, (c) +300 V, and (d) +600 V. Reprinted with permission from [72]. Copyright 2015 Elsevier.

### **4.5. Effect of alternating bias on the deposition behavior of silicon in the CVD process**

In the previous section, it was shown that even under the condition where CNPs are generated in abundance, they might not contribute to deposition because of the drag force. In this case, applying the DC bias would be effective to deposit those CNPs in the gas phase when CNPs are somewhat conducting as shown in **Figure 17**. When the CNPs are insulating, however, the positive bias would attract the negative CNPs and then negative charge would build up on the surface. This built-up charge would diminish the effect of the positive bias and then the positive bias would not be pronounced any more. In this case, the alternating bias would be effective because it prevents charge build up on the growing surface.

Youn at el. [75] studied the effect of alternating bias in silicon deposition on stainless substrate holder during thermal CVD. The bias frequency and voltage were varied. **Figure 18** shows the FESEM images of the surface morphology of silicon films deposited on the quartz substrates for 2 hours at reactor temperature of 900°C with gas flow rates of 5 sccm 10% SiH4−90% He, 50 sccm H2, and 1000 sccm N2. **Figure 18(a)–(d)** is for the zero bias and the bias voltage of ±100 V at frequencies of 0.2, 0.5, and 5 Hz, respectively. **Figure 18(a)** shows the microstructure of the film which was deposited under the condition of the zero bias. The film has some roughness on the surface. This surface microstructure of the film is similar to that of films deposited by PECVD [76] and HWCVD [66]. **Figure 18(b)**, the microstructure of film which was deposited at a frequency of 0.2 Hz, shows the surface is covered by flake-like structures. At 0.5 Hz, the surface became smoother than that of 0.2 Hz and a flake-like structure disappeared as shown in **Figure 18(c)**. The surface of film which was deposited at 5 Hz had a very porous structure as shown in **Figure 18(d)**.

results indicated that the drag force becomes an important factor in deposition when nano‐

48 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 17.** Deposition behavior of charged ZnO particles under the positive biases of (a) 0 V, (b) +100 V, (c) +300 V, and

In the previous section, it was shown that even under the condition where CNPs are generated in abundance, they might not contribute to deposition because of the drag force. In this case, applying the DC bias would be effective to deposit those CNPs in the gas phase when CNPs are somewhat conducting as shown in **Figure 17**. When the CNPs are insulating, however, the positive bias would attract the negative CNPs and then negative charge would build up on the surface. This built-up charge would diminish the effect of the positive bias and then the positive bias would not be pronounced any more. In this case, the alternating bias would be

Youn at el. [75] studied the effect of alternating bias in silicon deposition on stainless substrate holder during thermal CVD. The bias frequency and voltage were varied. **Figure 18** shows the FESEM images of the surface morphology of silicon films deposited on the quartz substrates

**4.5. Effect of alternating bias on the deposition behavior of silicon in the CVD process**

(d) +600 V. Reprinted with permission from [72]. Copyright 2015 Elsevier.

effective because it prevents charge build up on the growing surface.

particles are formed in the gas phase.

**Figure 18.** FESEM image of silicon deposited during 2 hour on a quartz substrate at (a) zero bias, (b) AC bias voltage of ±100 V with frequencies 0.2 Hz, (c) AC bias voltage of ±100 V with frequencies 0.5 Hz, and (d) AC bias voltage of ±100 V with frequencies 5 Hz. Reprinted with permission from [75]. Copyright 2012 Elsevier.

The effect of the bias voltage was also examined. The bias was applied to the stainless substrate holder at the biases of ±50 V, ±100 V, ±150 V, and ±200 V with a frequency of 1 Hz. **Figure 19** shows the FESEM images of the surface morphology of silicon films deposited under the same processing condition as that of **Figure 18**. Under the bias of ±50 V, the film microstructure tended to have a flake-like structure as shown in **Figure 19(a)**. When the bias was increased to ±100 V, the microstructure tended to have round nodules as shown in **Figure 19(b)**. When the bias was increased to ±150 V, the microstructure became porous as shown in **Figure 19(c)**. On applying ±200 V bias, the nanoparticles of ∼100 nm were aggregated as a chain-like structure with their surface covered with numerous nano-rods as shown in **Figure 19(d)**. The nanopar‐

**Figure 19.** FESEM images of silicon films deposited on a quartz substrate at AC bias voltages of (a) ±50 V, (b) ±100 V, (c) ±150 V, and (d) ±200 V with the frequency of 1 Hz applied to the substrate holder. Reprinted with permission from [75]. Copyright 2012 Elsevier.

ticles of ∼50 nm were deposited on the surface, being smaller than those deposited at ±150 V. The microstructure deposited at ±200 V was much more porous than that deposited at ±150 V.

These results imply that controlling the AC bias frequency or the magnitude of bias voltages may produce various microstructures. Besides, the AC bias technique can provide a promising method to produce highly porous microstructures, which are favorable for low dielectric materials.

Various microstructures can be synthesized by changing the bias frequency and the magnitude of bias voltages during deposition. By this technique, films of a layered structure with each layer of different microstructures can be synthesized. For example, the first layer is made dense, the second layer being porous, and the third layer being dense again. These results show clearly that the bias, either AC or DC, can be a new processing parameter in the CVD process where CNPs are generated.

### **4.6. Effect of size of CNPs on CVD process**

The size of CNPs is also a critical factor in thin film growth by CVD. Therefore, informa‐ tion about the size distribution of CNPs is important in controlling the microstructure evolution of films. The size of nanoparticles increases normally as the concentration ofreactant source gases increases. **Figure 20** shows the measured mass distributions of negative CNPs for four different methane concentrations during HWCVD [60]. During the mass distribu‐ tion measurements, diamonds were deposited in situ on a Mo substrate placed near the orifice

with a substrate temperature of 750 °C. The diamond films deposited at 1% and 1.5% CH4, where nanoparticles of 200–300 carbon atoms were dominant as shown in **Figure 20**, showed good crystalline quality as shown in **Figure 21(a)**. For methane concentrations of 3% and 5%, where the nanoparticles in the gas phase contained more than 1000 carbon atoms and the size distribution became much broader in **Figure 20**, ball-like diamonds were deposited as shown in **Figure 21(b)**. This structure is often called a cauliflower structure because it looks similar to the vegetable cauliflower.

ticles of ∼50 nm were deposited on the surface, being smaller than those deposited at ±150 V. The microstructure deposited at ±200 V was much more porous than that deposited at ±150 V. These results imply that controlling the AC bias frequency or the magnitude of bias voltages may produce various microstructures. Besides, the AC bias technique can provide a promising method to produce highly porous microstructures, which are favorable for low dielectric

**Figure 19.** FESEM images of silicon films deposited on a quartz substrate at AC bias voltages of (a) ±50 V, (b) ±100 V, (c) ±150 V, and (d) ±200 V with the frequency of 1 Hz applied to the substrate holder. Reprinted with permission from

50 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Various microstructures can be synthesized by changing the bias frequency and the magnitude of bias voltages during deposition. By this technique, films of a layered structure with each layer of different microstructures can be synthesized. For example, the first layer is made dense, the second layer being porous, and the third layer being dense again. These results show clearly that the bias, either AC or DC, can be a new processing parameter in the CVD process

The size of CNPs is also a critical factor in thin film growth by CVD. Therefore, informa‐ tion about the size distribution of CNPs is important in controlling the microstructure evolution of films. The size of nanoparticles increases normally as the concentration ofreactant source gases increases. **Figure 20** shows the measured mass distributions of negative CNPs for four different methane concentrations during HWCVD [60]. During the mass distribu‐ tion measurements, diamonds were deposited in situ on a Mo substrate placed near the orifice

materials.

where CNPs are generated.

[75]. Copyright 2012 Elsevier.

**4.6. Effect of size of CNPs on CVD process**

**Figure 20.** Measured mass distributions of negatively charged carbon clusters extracted from the hot filament reactor using gas mixtures of 1% CH4-99% H2, 1.5% CH4-98.5% H2, 3% CH4-97% H2, and 5% CH4-95% H2. Reprinted with per‐ mission from [60]. Copyright 2001 Elsevier.

To explain such an effect of size of CNPs on the microstructure evolution, weakening of bond strength by charge should be considered. In a soft charging condition such as thermal and hot wire CVD, which is in contrast with the hard charging condition in plasma CVD, most nanoparticles are expected to be singly charged. In this case, the charge effect on weakening of the bond strength would be diminished as the size increases. In other words, the smaller CNPs would be more liquid-like than the larger ones. Therefore, large CNPs generated under the high methane concentration in **Figure 20** would be less liquid-like so that CNPs frequently fail to undergo epitaxial recrystallization, producing ball-like or cauliflower diamond struc‐ tures as shown in **Figure 21(b)**. On the other hand, small CNPs generated under the low methane concentration in **Figure 20** would be more liquid-like so that CNPs mostly undergo epitaxial recrystallization, producing high-quality diamonds with well-defined facets as shown in **Figure 21(a)**.

To confirm the effect of nanoparticle size on the deposition behavior, Yoshida and his collea‐ gues [77–81] made extensive studies on the epitaxial growth of films with a building block of nanoparticles by the method called thermal plasma flash evaporation. They could deposit high-quality epitaxial YBa2Cu3O7−xfilms with a growth rate as high as 16 nm/s by this technique. They also could estimate the size of the nanoparticles to be about 0.3–10 nm using a microtrench fabricated on a Si wafer [79]. They observed that small 1–2 nm nanoparticles made epitaxial spiral growth, medium size 3 nm nanoparticles produced epitaxial 2-dimensional particles, and large nanoparticles over 3 nm produced non-epitaxial island grains by scanning tunneling microscopy (STM) [80].

**Figure 21.** SEM images of diamond films deposited *in situ* during the measurement of the mass distribution of CNPs at 2100°C wire temperature and 800 Pa reactor pressure: (a) 1% CH4-99%nH2, (b) 3% CH4-97%H2. Reprinted with permis‐ sion from [60]. Copyright 2001 Elsevier.

**Figure 22.** Size distribution of positively (a) and negatively (b) charged silicon nanoparticles measured by PBMS at a wire temperature of 1800°C with various process pressures during HWCVD. Reprinted with permission from [82]. Copyright 2014 Elsevier.

The reactor pressure also affects the size of CNPs. Yoo et al. [82] investigated the size distri‐ bution of CNPs generated in the gas phase during deposition of Si films by HWCVD through particle beam mass spectrometry (PBMS).

**Figure 22** shows the size distribution of CNPs generated in the gas phase measured by PBMS in the HWCVD process with a 4% SiH4−96% H2 gas mixture and a wire temperature of 1800°C at 1.0, 1.5, and 2.0 Torr. The peak of the number concentration occurs at 9–12 nm at which the number concentrations of both positive and negative CNPs are 106 –107 /cm3 . These nanopar‐ ticles were observed by TEM after being captured from the gas phase on a TEM grid membrane. As the pressure is increased from 0.3 to 2 Torr, not only the size and the number of captured nanoparticles are decreased, but also the rate of deposition is increased. An increase in the distance at which nanoparticles were captured from the hot wires under 1.5 Torr also reduced the size and number of nanoparticles.

To confirm the effect of nanoparticle size on the deposition behavior, Yoshida and his collea‐ gues [77–81] made extensive studies on the epitaxial growth of films with a building block of nanoparticles by the method called thermal plasma flash evaporation. They could deposit high-quality epitaxial YBa2Cu3O7−xfilms with a growth rate as high as 16 nm/s by this technique. They also could estimate the size of the nanoparticles to be about 0.3–10 nm using a microtrench fabricated on a Si wafer [79]. They observed that small 1–2 nm nanoparticles made epitaxial spiral growth, medium size 3 nm nanoparticles produced epitaxial 2-dimensional particles, and large nanoparticles over 3 nm produced non-epitaxial island grains by scanning

52 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 21.** SEM images of diamond films deposited *in situ* during the measurement of the mass distribution of CNPs at 2100°C wire temperature and 800 Pa reactor pressure: (a) 1% CH4-99%nH2, (b) 3% CH4-97%H2. Reprinted with permis‐

**Figure 22.** Size distribution of positively (a) and negatively (b) charged silicon nanoparticles measured by PBMS at a wire temperature of 1800°C with various process pressures during HWCVD. Reprinted with permission from [82].

The reactor pressure also affects the size of CNPs. Yoo et al. [82] investigated the size distri‐ bution of CNPs generated in the gas phase during deposition of Si films by HWCVD through

tunneling microscopy (STM) [80].

sion from [60]. Copyright 2001 Elsevier.

Copyright 2014 Elsevier.

particle beam mass spectrometry (PBMS).

**Figure 23.** TEM images of initial silicon nanoparticles deposited for 20 sec after delay times of (a) 0 min and (b) 5 min and (c) and (d) showing respectively the HR images of (a) and (b). Reprinted with permission from [83]. Copyright 2012 Elsevier.

Chung et al. [83] show that the size of silicon nanoparticles generated in the gas phase during HWCVD tended to increase with increasing processing time and the size tended to be saturated above 15 min. This means that the steady state in terms of the size of CNPs is not reached in the reactor until 15 min. Therefore, in the initial stage of deposition before 15 min, small CNPs would deposit whereas in the later stage after 15 min, large CNPs would deposit. Therefore, if beginning of the deposition is delayed using the shutter above the substrate, the size of CNPs to be deposited can be controlled. Based on this observation, Chung et al. [83] used the concept of the delay time in deposition to control the microstructure. How the size of CNPs changes with the delay time can be examined by capturing CNPs on a TEM grid membrane with different delay times.

For example, **Figure 23(a)** and **(b)** shows the silicon nanoparticles captured on a Cu TEM grid membrane for 20 sec with delay times of 0 and 5 min, respectively. **Figure 23(c)** and **(d)** is the magnified view showing the lattice image of **Figure 23(a)** and **(b)**, respectively. The sizes of crystalline silicon nanoparticles in **Figure 23(a)** and **(b)** are ∼10 and ∼28 nm, respectively. The lattice image in **Figure 23(c)** shows a (220) plane, and that of **Figure 23(d)** consists of (220) and (111) planes, indicating that the silicon particle is a bicrystal.

**Figure 24.** Resistivity of silicon films with the initial silicon layer deposited for 300 sec after delay times of 0, 3, 5, and 10 min. Reprinted with permission from [83]. Copyright 2012 Elsevier.

In order to study the effect of a different initial layer, the initial layer was deposited on the glass substrate for 20 sec after delay times of 0, 3, 5, and 10 min using the shutter. Then, these four samples were further deposited for 300 sec under the same processing conditions as that of **Figure 23**, and the resistivity of the deposited silicon films was measured. **Figure 24** shows how the resistivity of the silicon films varies with delays times. The thickness of all the films was ∼170 nm. The resistivity of a silicon film deposited without a delay time (or a delay time of 0 min) is 0.32 Ω-cm. It should be noted that such low resistivity is due to HCl addition; silicon films deposited without HCl by PECVD or HWCVD have resistivity typically in the range of 5–50 Ω-cm [84, 85]. The resistivities of the silicon films prepared with delay times of 3, 5, and 10 min are 0.21, 0.13, and 0.16 Ω-cm, respectively. The resistivity tended to be saturated after a delay time of 5 min. Silicon films with a delay time show consistently lower resistivity than those without a delay time. The lowest resistivity of 0.13 Ω -cm of the film with a delay time of 5 min is 2.4 times lower than that without a delay time.

This resistivity change with delay times can be understood considering the property of CNPs. Without a delay time, the size of CNPs would be minimal and more liquid-like. These liquidlike CNPs, which would be accommodated to the substrate of an amorphous glass structure, would change to an amorphous structure, resulting in high resistivity. In contrast, with a delay time of 10 min, the size of CNPs would be maximal and less liquid-like. These CNPs would retain their crystalline silicon structure instead of changing to an amorphous structure, resulting in low resistivity.

### **4.7. Low-temperature deposition of crystalline silicon**

with the delay time can be examined by capturing CNPs on a TEM grid membrane with

54 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

For example, **Figure 23(a)** and **(b)** shows the silicon nanoparticles captured on a Cu TEM grid membrane for 20 sec with delay times of 0 and 5 min, respectively. **Figure 23(c)** and **(d)** is the magnified view showing the lattice image of **Figure 23(a)** and **(b)**, respectively. The sizes of crystalline silicon nanoparticles in **Figure 23(a)** and **(b)** are ∼10 and ∼28 nm, respectively. The lattice image in **Figure 23(c)** shows a (220) plane, and that of **Figure 23(d)** consists of (220) and

**Figure 24.** Resistivity of silicon films with the initial silicon layer deposited for 300 sec after delay times of 0, 3, 5, and

In order to study the effect of a different initial layer, the initial layer was deposited on the glass substrate for 20 sec after delay times of 0, 3, 5, and 10 min using the shutter. Then, these four samples were further deposited for 300 sec under the same processing conditions as that of **Figure 23**, and the resistivity of the deposited silicon films was measured. **Figure 24** shows how the resistivity of the silicon films varies with delays times. The thickness of all the films was ∼170 nm. The resistivity of a silicon film deposited without a delay time (or a delay time of 0 min) is 0.32 Ω-cm. It should be noted that such low resistivity is due to HCl addition; silicon films deposited without HCl by PECVD or HWCVD have resistivity typically in the range of 5–50 Ω-cm [84, 85]. The resistivities of the silicon films prepared with delay times of 3, 5, and 10 min are 0.21, 0.13, and 0.16 Ω-cm, respectively. The resistivity tended to be saturated after a delay time of 5 min. Silicon films with a delay time show consistently lower resistivity than those without a delay time. The lowest resistivity of 0.13 Ω -cm of the film with a delay

This resistivity change with delay times can be understood considering the property of CNPs. Without a delay time, the size of CNPs would be minimal and more liquid-like. These liquid-

(111) planes, indicating that the silicon particle is a bicrystal.

10 min. Reprinted with permission from [83]. Copyright 2012 Elsevier.

time of 5 min is 2.4 times lower than that without a delay time.

different delay times.

There is a strong need for the low temperature synthesis of crystalline silicon films for electronics such as displays and solar cells [86–89]. The most popular method to synthesize microcrystalline films is to deposit amorphous phase silicon on a glass substrate than to crystallize it by excimer laser, thermal annealing, or metal-induced crystallization [90–92]. Since these methods increase the production cost, it is necessary to deposit the crystalline silicon film directly on a glass substrate.

From the view point of classical crystallization, silicon diffusivity below 600 °C is so low that amorphous silicon is expected on a glass substrate below 600 °C. Direct deposition of crystal‐ line silicon at temperatures below 600 °C is not possible. Nonetheless, the direct deposition of microcrystalline silicon films on a glass substrate below 400 °C has been reported using HWCVD system [93–95] or PECVD system [96]. These results cannot be approached by the concept of classical crystallization. However, this low temperature deposition of crystalline silicon can be explained by non-classical crystallization. If crystalline silicon nanoparticles were formed in the gas phase in a high-temperature region near the hot wire and then incorporated into the films at low temperature, crystalline silicon could be deposited on a substrate at temperature below 600 °C.

If this explanation is correct, crystalline silicon nanoparticles should exist in the gas phase, which can be confirmed by capturing them on a TEM grid membrane. For this experimental confirmation, Lee et al. [84] designed the HWCVD, which consists of three chambers by two stage orifices. Each chamber was maintained in different pressures by two-stage differential pumping. The first chamber is the HWCVD reactor with the working pressure maintained at 27 Pa under a flowing gas mixture of SiH4 and H2, and the working pressure of the third chamber being maintained at 1.3 × 10−3 Pa during the process. The CNPs were generated in the first chamber and extracted through the first and the second orifices into the third chamber.

These CNPs were captured on the TEM grid located in the third chamber at room temperature. The wire temperature was 1560°C and the concentration of SiH4 in the gas mixture of SiH4- H2 was varied as 20%, 15%, 10%, and 5%. The TEM image of the crystalline silicon CNPs captured on the TEM grid membrane is shown in **Figure 25** [84]. The CNPs which were captured on the TEM grid membrane had the crystallinity and the lattice fringes of CNPs were clearly revealed. The size of the crystalline nanoparticles decreased with decreasing SiH4 concentration with 7–8 nm for 20% SiH4 (**Figure 25(a)**), 6–7 nm for 15% SiH4 (**Figure 25(b)**), 5– 6 nm for 10% SiH4 (**Figure 25(c)**), and 4–5 nm for 5% SiH4 (**Figure 25(d)**).

**Figure 25.** TEM images of silicon nanoparticles which were captured on the amorphous carbon membrane at room temperature by HWCVD with (a) 20%, (b) 15%, (c) 10% and (d) 5% flow rate of SiH4 at a wire temperature of 1560°C. Reprinted with permission from [84]. Copyright 2008 Elsevier.

**Figure 26.** Raman spectra of the silicon films deposited at various HCl flow rates. The dotted line is for a single crystal silicon wafer as a reference. Reprinted with permission from [97]. Copyright 2009 Elsevier.

Classical crystallization cannot explain such formation of crystalline silicon nanoparticles at room temperature because the atomic diffusion rate is negligible. **Figure 25** strongly indicates that CNPs were formed in the gas phase of the first chamber during HWCVD and landed on the TEM grid in the third chamber.

There is an appreciable amount of amorphous silicon in the TEM images. This amorphous phase is expected to have been formed in the low-temperature region far from the hot wire. Therefore, to minimize the amorphous phase, low-temperature precipitation or nucleation of silicon should be inhibited. To inhibit the precipitation of silicon at low temperature, Chung et al. [97, 98] suggested HCl addition to SiH4. Further, they confirmed that the crystallinity of silicon increased with increasing HCl addition as revealed in the Raman spectra of **Figure 26**. Since the Si-Cl-H system has retrograde solubility [99], if HCl is added to the Si-H system, the solubility of silicon in the gas phase increases with decreasing temperature, which means the equilibrium amount of silicon decreases with decreasing temperature.

This means that when gas-phase nucleation occurs, the driving force near the substrate temperature becomes for etching as in the case of the C-H system for diamond CVD. Therefore, HCl addition inhibits the precipitation of amorphous silicon at low temperature. Then, the crystalline silicon nanoparticles, which are precipitated at high temperature, will exclusively contribute to deposition without additional precipitation of amorphous silicon at low tem‐ perature, improving the crystallinity of the film.

**Figure 25.** TEM images of silicon nanoparticles which were captured on the amorphous carbon membrane at room temperature by HWCVD with (a) 20%, (b) 15%, (c) 10% and (d) 5% flow rate of SiH4 at a wire temperature of 1560°C.

56 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 26.** Raman spectra of the silicon films deposited at various HCl flow rates. The dotted line is for a single crystal

Classical crystallization cannot explain such formation of crystalline silicon nanoparticles at room temperature because the atomic diffusion rate is negligible. **Figure 25** strongly indicates that CNPs were formed in the gas phase of the first chamber during HWCVD and landed on

silicon wafer as a reference. Reprinted with permission from [97]. Copyright 2009 Elsevier.

Reprinted with permission from [84]. Copyright 2008 Elsevier.

the TEM grid in the third chamber.

**Figure 27.** TEM images of the *n*-type crystalline silicon film with HCl (a) a bright-field image, and (b) image of the selective area diffraction pattern with (c) and (d) being high-resolution images on the regions indicated in (a). Reprint‐ ed with permission from [98]. Copyright 2011 Elsevier.

It is known that an amorphous silicon incubation layer, which is as thick as hundreds of nanometers, is formed on the glass substrate before crystalline silicon starts to form when the low temperature deposition of crystalline silicon is done by HWCVD or PECVD. The formation of such an amorphous incubation layer is a serious problem for thin-film transistor (TFT) applications because the thickness of the TFT should be typically less than ∼200 nm.

Chung et al. [97, 98] could successfully minimize the formation of amorphous incubation layer by adding HCl as shown in **Figure 27**.

**Figure 27** shows the cross section HRTEM images of the crystalline silicon film deposited at the flow rate ratios of [HCl]/[SiH4] = 7.5 and [PH3]/[SiH4]=0.028. The whole deposition area of **Figure 27(a)** consists of crystalline silicon. More specifically, the high-resolution image at the middle of the film in **Figure 27(c)**reveals that the grain boundary is free from any amorphous phase between the two grains showing (111) and (220) planes. This aspect is highly in contrast with the previous result, where HCl was not added, that the crystalline silicon nanoparticles were embedded in the amorphous silicon phase. **Figure 27** showed that the amorphous incubation layer could be reduced markedly or almost removed by HCl addition. This is an example showing how new understanding of thin film growth by non-classical crystallization can be applied.

**Figure 28.** Growth rate and electron mobility from time-resolved microwave conductivity measurements of the total pressure and the distance of inter-electrode. Reprinted with permission from [70]. Copyright 2008 Elsevier.

As to the possibility of incorporation of particles generated in the gas phase during PECVD, Cabarrocas et al. [23, 24, 70, 96] suggested that the nanoparticles generated in the gas phase can be incorporated into films. Cabarrocas et al. [70] reported that the incorporation of crystalline nuclei generated in the gas phase increased the film deposition rate and electron mobility of films. This result is shown in **Figure 28**, where the x-axis is the product of the total pressure and inter-electrode distance. Both film growth rate and electron mobility are maxi‐ mum when the process pressure is ∼ 6 Torr.

They could deposit fully crystallized films, consisting of 68% small grain materials and 28% monocrystalline silicon with a 4% fraction of voids at substrate temperature of 200 °C, RF power at 20 W and gas flow rates at 1, 5, and 36 sccm respectively for SiF4, H2, and Ar. Considering the substrate temperature of 200 °C, the deposition of microcrystalline films also cannot be explained by classical crystallization. Therefore, the high crystallinity of films also would be attributed to the self-assembly of CNPs formed in the gas phase.

However, the temperature of the plasma is too low to form the crystalline silicon phase in the gas phase. This question would be explained by the role of charges produced in the plasma CVD, by which not only the kinetics of the precursor decomposition is enhanced, but also the deposition of crystalline films is made possible at such low temperatures. All these enhanced kinetics can be explained if the charge weakens the bond strength of molecules and nanopar‐ ticles.

### **5. Conclusion**

Chung et al. [97, 98] could successfully minimize the formation of amorphous incubation layer

58 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 27** shows the cross section HRTEM images of the crystalline silicon film deposited at the flow rate ratios of [HCl]/[SiH4] = 7.5 and [PH3]/[SiH4]=0.028. The whole deposition area of **Figure 27(a)** consists of crystalline silicon. More specifically, the high-resolution image at the middle of the film in **Figure 27(c)**reveals that the grain boundary is free from any amorphous phase between the two grains showing (111) and (220) planes. This aspect is highly in contrast with the previous result, where HCl was not added, that the crystalline silicon nanoparticles were embedded in the amorphous silicon phase. **Figure 27** showed that the amorphous incubation layer could be reduced markedly or almost removed by HCl addition. This is an example showing how new understanding of thin film growth by non-classical crystallization

**Figure 28.** Growth rate and electron mobility from time-resolved microwave conductivity measurements of the total

As to the possibility of incorporation of particles generated in the gas phase during PECVD, Cabarrocas et al. [23, 24, 70, 96] suggested that the nanoparticles generated in the gas phase can be incorporated into films. Cabarrocas et al. [70] reported that the incorporation of crystalline nuclei generated in the gas phase increased the film deposition rate and electron mobility of films. This result is shown in **Figure 28**, where the x-axis is the product of the total pressure and inter-electrode distance. Both film growth rate and electron mobility are maxi‐

They could deposit fully crystallized films, consisting of 68% small grain materials and 28% monocrystalline silicon with a 4% fraction of voids at substrate temperature of 200 °C, RF power at 20 W and gas flow rates at 1, 5, and 36 sccm respectively for SiF4, H2, and Ar. Considering the substrate temperature of 200 °C, the deposition of microcrystalline films also cannot be explained by classical crystallization. Therefore, the high crystallinity of films also

would be attributed to the self-assembly of CNPs formed in the gas phase.

pressure and the distance of inter-electrode. Reprinted with permission from [70]. Copyright 2008 Elsevier.

mum when the process pressure is ∼ 6 Torr.

by adding HCl as shown in **Figure 27**.

can be applied.

Non-classical crystallization, where the growth units of deposition in CVD process are CNPs, was revealed by the rigorous thermodynamic analysis of the paradoxical experimental observation of simultaneous diamond deposition and graphite etching. This mechanism was further supported by the puzzling phenomenon of diamond deposition on a silicon substrate and porous skeletal graphitic soot deposition on an iron substrate. This new paradigm of crystal growth was also experimentally confirmed in many CVD processes, which means that non-classical crystallization is a general growth mechanism of thin films and nanostructures in CVD processes. Lots of fundamental and application studies should be made in this new paradigm of crystal growth.

### **Acknowledgements**

This work was supported by the Global Frontier R&D Program (2013-073298) on Center for Hybrid Interface Materials (HIM) funded by the Ministry of Science, ICT & Future Planning and National Research Foundation of Korea(NRF) grant funded by the Ministry of Science, ICT & Future Planning(MSIP) (NO. NRF-2015R1A5A1037627).

### **Author details**

Jae-soo Jung1 and Nong-moon Hwang2\*

\*Address all correspondence to: nmhwang@snu.ac.kr

1 Department of Material Science and Engineering, Seoul National University, Gwanak-gu, Seoul, Republic of Korea

2 Department of Material Science and Engineering, Seoul National University and Research Institute of Advanced Materials (RIAM), Seoul National University, Gwanak-gu, Seoul, Republic of Korea

### **References**


model. J. Korean Ceram. Soc. 2001;38:218. DOI: http://www.dbpia.co.kr/Journal/ ArticleDetail/NODE00737510

[14] Kim J Y, Kim D Y, Hwang N M. Spontaneous generation of negatively charged clusters and their deposition as crystalline films during hot-wire silicon chemical vapor deposition. Pure Appl. Chem. 2006;78:1715. DOI: 10.1351/pac200678091715

**References**

[1] Boufendi L, Plain A, Blondeau J P, Bouchoule A. Laure C Toogood M, Measurements of particle size kinetics from nanometer to micrometer scale in a low-pressure argonsilane radio-frequency discharge. Appl. Phys. Lett. 1992;60:169. DOI: 10.1063/1.106981

[2] Howling A A, Sansonnens L, Dorier J L, Hollenstein C. Negative hydrogenated silicon ion clusters as particle precursors in RF silane plasma deposition experiments J. Phys.

[3] Veprek S, Ambacher O, Rieger W, Schopper K, Veprek-Heijman M G J. Clusters in a silane glow discharge: mechanism of their formation and how to avoid them. Mater.

[4] Garscadden A, Ganguly B N, Haaland P D, Williams J. Overview of growth and behaviour of clusters and particles in plasmas. Plasma Sources Sci. Technol. 1994;3:239.

[5] Stoffels E, Stoffels W W, Kroesen G M W, Hoog F J D. Dust formation and charging in an Ar/SiH4 radio-frequency discharge. J. Vac. Sci. Technol. A. 1996;14:556. DOI:

[6] Hwang N M, Yoon D Y. Thermodynamic approach to the paradox of diamond formation with simultaneous graphite etching in the low pressure synthesis of dia‐

[7] Hwang N M, Hahn J H, Yoon D Y. Charged cluster model in the low pressure synthesis of diamond. J. Cryst. Growth. 1996;162:55. DOI: 10.1016/0022-0248(95)00943-4

[8] Hwang N M, Hahn J H, Yoon D Y. Chemical potential of carbon in the low pressure synthesis of diamond. J. Cryst. Growth. 1996;160:87. DOI: 10.1016/0022-0248(95)00548-

[9] Hwang N M, Kim D Y. Charged clusters in thin film growth. Int. Mater. Rev.

[10] Hwang N M, Lee D K. Charged nanoparticles in thin film and nanostructure growth by chemical vapour deposition. J. Phys. D: Appl. Phys. 2010;43:483001. Review Paper.

[11] Jeon J D, Park C J, Kim D Y, Hwang N M. Experimental confirmation of charged carbon clusters in the hot filament diamond reactor. J. Cryst. Growth. 2001;213:79. DOI:

[12] Ahn H S, Park H M, Kim D Y, Hwang N M. Observation of carbon clusters of a few nanometers in the oxyacetylene diamond CVD process. J. Cryst. Growth. 2002;234:399.

[13] Jeon I D, Gueroudji L, Kim D Y, Hwang N M. Temperature dependence of the depo‐ sition behavior of yttria-stabilized zirconia CVD films: approach by charged cluster

mond. J. Cryst. Growth. 1996;160:98. DOI: 10.1016/0022-0248(95)00549-8

D: Appl. Phys. 1993;26:1003. DOI: 10.1088/0022-3727/26/6/019

60 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Res. Soc. Symp. Proc. 1993;297:13. DOI: 10.1557/PROC-297-13

DOI: 10.1088/0963-0252/3/3/001

2004;49:171. DOI: 10.1179/095066004225021891

DOI: 10.1088/0022-3727/43/48/483001

DOI: 10.1016/S0022-0248(01)01748-1

10.1016/S0022-0248(00)00358-4

10.1116/1.580144

X


[39] Zheng H, Smith R K, Jun Y W, Kisielowski C, Dahmen U. Observation of single colloidal platinum nanocrystal growth trajectories. Science. 2009;324(5932):1309–1312. DOI: 10.1126/science.1172104

[26] Nunomura S, Kita M, Koga K, Shiratani M, Watanabe Y. Fabrication of nanoparticle composite porous films having ultralow dielectric constant. Japan. J. Appl. Phys.

62 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

[27] Glasner A, Kenat J. The crystallization of KCl from aqueous solutions in the presence of lead ions. I. A calorimetric study. J. Cryst. Growth. 1968;2:119–127. DOI:

[28] Glasner A, Skurnik S. A new mechanism for the crystallization and growth of ionic crystals, with special reference to KCl in the presence of Pb2+ ions. Israel J. Chem.

[29] Glasner A, Tassa M. The thermal effects of nucleation and crystallization of KBr and KCl solutions. II. The heat of nucleation and the supersaturated solution. Israel J. Chem.

[30] Glasner A, Tassa M. The thermal effects of nucleation and crystallization of KBr and KCL solutions. III. The heat of crystallization and the co-precipitation of lead ions. Israel

[31] Sunagawa I. Morphology of minerals 'Morphology of Crystals'. Tokyo: Terra Sci. 26

[32] Sunagawa I. Growth and morphology of diamond crystals under stable and metastable conditions. J. Cryst. Growth. 1990;99:1156–1161. DOI: 10.1016/S0022-0248(08)80100-5

[33] Botsaris G D, Reid R C. Comments on the letter by Glasner and Skurnik entitled 'Growth of potassium chloride crystals from aqueous solutions. I. the effect of lead chloride'. J.

[34] Meldrum F C, Cölfen H. Controlling mineral morphologies and structures in biological and synthetic systems. Source of the Document Chemical Reviews. 2008;108:4332–4432.

[35] Niederberger M, Cölfen H. Oriented attachment and mesocrystals: non-classical crystallization mechanisms based on nanoparticle assembly. Phys Chem Chem Phys.

[36] Cölfen H, Antonietti M. Mesocrystals: inorganic superstructures made by highly parallel crystallization and controlled alignment. Angew. Chem. – Int. Ed. 2005;44(35):

[37] Liao H G, Cui L, Whitelam S, Zheng H. Real-time imaging of Pt3Fe nanorod growth in solution. Science. 2012;336(6084):1011–1014. DOI: 10.1126/science.1219185

[38] Yuk J M, Park J, Ercius P, Kim K, Hellebusch D J. Crommie M F, Lee J Y, Zettl. Highresolution EM of colloidal nanocrystal growth using graphene liquid cells A. Science.

2005;44:50–52. DOI: 10.1143/JJAP.44.L1509

1968;6:69–72. DOI: 10.1002/ijch.196800011

1974;12:799–816. DOI: 10.1002/ijch.197400069

J. Chem. 1974;12:817–826. DOI: 10.1002/ijch.197400070

Chem. Phys. 1967;47:3689. DOI: 10.1063/1.1712456

2006;8:3271–3287. DOI: 10.1039/B604589H

5576–5591. DOI: 10.1002/anie.200500496

2012;336(6077):61–64. DOI: 10.1126/science.1217654

10.1016/0022-0248(68)90073-0

1987:1;511–587.

DOI: 10.1021/cr8002856


vapor deposition. Sol. Energy Mater. Sol. Cells. 2011;95(1):211–214. DOI: 10.1016/ j.solmat.2010.04.021

[67] Kumomi H, Yonehara T, Nishigaki Y, Sato N. Selective nucleation based epitaxy (sentaxy): investigation of initial nucleation stages. Appl. Surf. Sci. 1990;41–42:638–642. DOI: 10.1016/0169-4332(89)90135-9

[53] Kamo M, Sato Y, Matsumoto S, Setaka N. Diamond synthesis from gas phase in microwave plasma. J. Cryst. Growth. 1983;62(3):642–644. DOI:

64 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

[54] Haubner R, Lux B. Diamond growth by hot-filament chemical vapor deposition: state of the art. Diamond Relat. Mater. 1993;2(9):1277–1294. DOI:

[56] Fedosayev D V, Deryagin B V, Varasavskaja I G. The growth of diamond and graphite from the gas phase. Surf. Coat. Technol. 1989;38(1–2):131–248. DOI:

[57] Choi K, Kang S J L, Jang H M, Hwang N M. Nucleation behavior in the presence of charge in the CVD diamond process. J. Cryst. Growth. 1997;172(3–4):416–425. DOI:

[58] Hwang N M, Bahng G W, Yoon D N. Thermodynamics and kinetics for nucleation of diamond in the chemical vapor deposition process. Diamond Relat. Mater. 1992;1(2–

[59] Gaskell D R. Introduction to the Thermodynamics of Materials. Washington, DC:

[60] Jeon J D, Park C J, Kim D Y, Hwang N M. Effect of methane concentration on size of charged clusters in the hot filament diamond CVD process. J. Cryst. Growth. 2001;223:6.

[61] Huh J M, Yoon D Y, Kim D Y, Hwang N M. Effect of substrate materials in the lowpressure synthesis of diamond: approach by theory of charged clusters. Z. Metallk.

[62] Yang Q, Tang Y, Yang S L, Li Y S, Hirose A. Simultaneous growth of diamond thin films and carbon nanotubes at temperatures ⩽550 °C. Carbon. 2008;46(4):589–595. DOI:

[63] Lee J I, Hwang N M. Generation of negative-charge carriers in the gas phase and their contribution to the growth of carbon nanotubes during hot-filament chemical vapor

[64] Youn W K, Lee S S, Lee J Y, Kim C S, Hwang N M, Iijima S. Comparison of the deposition behavior of charged silicon nanoparticles between floating and grounded substrates.

[65] Kim C S, Youn W K, Hwang N M. Generation of charged nanoparticles and their deposition during the synthesis of silicon thin films by chemical vapor deposition. J.

[66] Chung Y B, Lee D K, Lim J S, Hwang N M. Reduction of amorphous incubation layer by HCl addition during deposition of microcrystalline silicon by hot-wire chemical

deposition. Carbon. 2008;46(12):1588–1592. DOI: 10.1016/j.carbon.2008.07.006

J. Phys. Chem. C. 2014;118(22):11946–11953. DOI: 10.1021/jp5001144

Appl. Phys. 2010;108(1):014313. DOI: 10.1063/1.3452352

[55] Belokurov B V. Isotropic surface energies of diamond. Zh. Fiz. Khim. 1960;34:440

10.1016/0022-0248(83)90411-6

10.1016/0925-9635(93)90008-P

10.1016/0257-8972(89)90129-1

10.1016/S0022-0248(96)00759-2

DOI: 10.1016/S0022-0248(00)00358-4

10.1016/j.carbon.2008.01.005

2005;96(3):225–232. DOI: 10.3139/146.101024

4):191–194. DOI: 10.1016/0925-9635(92)90023-H

Taylor and Francis; 1995. DOI: 10.1016/j.msea.2008.06.016


[91] Jang J, Oh J Y, Kim S K, Choi Y J, Yoon S Y, Kim C O. Electric-field-enhanced crystal‐ lization of amorphous silicon. Nature. 1998;395:481–483. DOI: 10.1038/26711

[79] Terashima K, Yamaguchi N, Hattori T, Takamura Y, Yoshida T. High rate deposition of thick epitaxial films by thermal plasma flash evaporation. Pure Appl. Chem.

66 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

[80] Takamura Y, Hayasaki K, Terashima K, Yoshida. High-rate deposition of YBa2Cu3O7– x films by hot cluster epitaxy. J. Appl. Phys. 1998;84:5084. DOI: 10.1063/1.368757 [81] Yamaguchi N, Sasajima Y, Terashima K, Yoshida T. Molecular dynamics study of cluster deposition in thermal plasma flash evaporation. Thin Solid Films. 1999;345:34–

[82] Yoo S W, Hong J S, Lee S S, Kim C S, Kim T, Hwang N M. Nonclassical crystallization in low-temperature deposition of crystalline silicon by hot-wire chemical vapor deposition. Cryst. Growth Des. 2014;14(12):6239–6247. DOI: 10.1021/cg5008582 [83] Chung Y B, Lee S H, Bae S H, Park H K, Jung J S, Hwang N M. Effect of the initial structure on the electrical property of crystalline silicon films deposited on glass by hotwire chemical vapor deposition. J. Nanosci. Nanotechnol. 2012;12:5947–5951. DOI:

[84] Lee S S, Ko M S, Kim C S, Hwang N M. Gas phase nucleation of crystalline silicon and their role in low-temperature deposition of microcrystalline films during hot-wire chemical vapor deposition. J. Cryst. Growth. 2008;310;3659–3662. DOI: 10.1016/

[85] Krishnan A T, Bae S, Fonash S J. Low temperature microcrystalline silicon thin film resistors on glass substrates. Solid State Electron. 2000;44:1163–1168. DOI: 10.1016/

[86] Persidis E, Baur H, Pieralisi F, Schalberger P, Fruehauf N. Area laser crystallized LTPS TFTs with implanted contacts for active matrix OLED displays. Solid State Electron.

[87] Shah A, Torres P, Tscharner R, Wyrsch N, Keppner H. Photovoltaic technology: the case for thin-film solar cells. Science. 1999;285:692–698. DOI: 10.1126/science.

[88] Uchikoga, Ibaraki N. Low temperature poly-Si TFT-LCD by excimer laser anneal. Thin

[89] Kahn H, Tayebi N, Ballarini R, Mullen R, Heuer A. Sens. Fracture toughness of Polysilicon MEMS devices. Actuators A – Phys. 2000;82:274–280. DOI: 10.1016/

[90] Shimoda T, Matsuki Y, Furusawa M, Aoki T, Yudasaka I, Tanaka H, Iwasawa H, Wang D, Miyasaka M, Takeuchi Y. Solution-processed silicon films and transistors. Nature.

Solid Films. 2001;383:19–24. DOI: 10.1016/S0040-6090(00)01644-8

1998;70:1193–1197. DOI: 10.1351/pac199870061193

37. DPI:10.1016/S0040-6090(99)00074-7

10.1166/jnn.2012.6415

j.jcrysgro.2008.05.009

S0038-1101(00)00057-5

S0924-4247(99)00366-0

285.5428.692

2008;52:455–461. DOI: 10.1016/j.sse.2007.10.014

2006;440:783–786. DOI: 10.1038/nature04613


### **MOCVD Grown HgCdTe Heterostructures**

Pawel Madejczyk, Waldemar Gawron, Artur Keblowski and Adam Piotrowski

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/62952

### **Abstract**

HgCdTe heterostructures are widely applied for IR (infrared) detector constructing. Donor‐ and acceptor‐doping researches in (100) and (111) oriented HgCdTe layers grown by MOCVD have been studied. Fully doped HgCdTe heterostructures with acceptor concentration range between 1014 and 5 × 10<sup>17</sup> cm-3 and donor concentration range between 1014 and 1 × 10<sup>18</sup> cm-3 and without post‐grown annealing have been reported. The electrical and chemical characterizations of HgCdTe structures grown at 360°C on GaAs sub‐ strates using DIPTe have been described. Infrared photodiodes with different composi‐ tion x were constructing on the basis of obtained heterostructures enabling signal detection of any wavelength from 1 µm to above 20 µm covering SWIR, MWIR and LWIR spectral ranges. Presented experimental results show that MOCVD technology enables to grow HgCdTe structures dedicated for HOT devices.

**Keywords:** HgCdTe, MOCVD, Doping, Heterostructure, infrared detector

### **1. Introduction**

Mercury cadmium telluride Hg1-xCdxTe or MCT is a pseudobinary alloy whose development was commenced by Lawson and coworkers [1]. HgCdTe ternary compound is an excellent material for infrared detectors construction. Its position is conditioned by following character‐ istics:


**•** high optical coefficients enabling high quantum efficiency.

Above‐mentioned properties are simply consequence of the energy band structure of this zinc‐ blende semiconductor. The special advantages of HgCdTe are ability to obtain wide range of carrier concentrations, low dielectric constant, and high mobility of current carriers. The very small change of lattice constant with composition enables to grow high‐quality layered and graded gap structures [2]. Thus, HgCdTe can be used for detectors operating at various modes [photodiode, photoconductor, or metal‐insulator‐semiconductor (MIS) detector].

This chapter reviews work from literature and some unpublished work from the authors' own laboratory that has been carried out into determining suitable extrinsic dopant species and sources for use in MOCVD growth of MCT heterostructures. The authors especially would like to present MOCVD technology with wide range of composition and donor/acceptor doping and without post‐grown annealing as an excellent tool for (HOT) HgCdTe infrared photodetectors construction.

### **2. MOCVD grown MCT**

MOCVD growth of MCT was firstly demonstrated in 1981 by Irvine and Mullin [3]. The reason for the delayed start in the growth of the mercury alloys in comparison to III–V compounds growth has been linked to the high mercury vapour pressures which needed some unconven‐ tional modifications to the MOCVD systems. MOCVD growth of MCT is determined by high vapour pressures of mercury that are necessarily to keep equilibrium conditions over the growing layer. This results from the instability of HgTe bonds in comparison with CdTe bonds and requires much lower growth temperatures than are typical for more stable compounds [4].

### **2.1. Thermodynamic considerations**

The mercury chalcogenides are characterized by weak bonds of mercury what causes a higher equilibrium vapour pressure. If the evaporating species are Hg and Te2, then the solid–vapour equilibrium is given by the following expression:

$$K = p\_{\rm Hg} \sqrt{p\_{\rm re}} \tag{2.1}$$

where pHg is the equilibrium vapour pressure of mercury, pTe<sup>2</sup> is the equilibrium vapour pressure of Te2, and K is an equilibrium constant which is a function of temperature. Eq. (2.1) implies that the equilibrium pressures of the component elements are linked and that there is a range of pressures over which the solid remains in equilibrium as a single phase. From the pressure–temperature phase diagram for HgTe presented in **Figure 1**, we can conclude that at MOCVD growth temperatures, the pressure can vary by three orders of magnitude and remain in equilibrium with a single phase of HgTe. However, it is clear from Eq. (2.1) that Te<sup>2</sup> partial pressure will be varying across this phase field in the opposite sense to the Hg partial pressure.

**Figure 1.** Mercury pressure–temperature phase diagram for HgTe. The p/n transition area indicates the transition from p‐type to n‐type for HgTe rich MCT annealed under equilibrium conditions [5].

The upper‐phase boundary corresponds to the saturation of Hg over liquid mercury, and therefore, a horizontal line can be drawn from the equilibrium (pHg, T) to the intersection with the Hg saturated boundary in order to determine the equilibrium mercury source temperature. An example of a tie line is shown in **Figure 1** for the growth on Te rich‐phase boundary at 350°C. If the HgTe layer is to be grown on Hg‐rich‐phase boundary, then the source and substrate temperature will be similar, which means an isothermal reaction cell. This is unacceptable for MCT MOCVD because Cd and Te organometallics will pyrolyse on the reactor cell wall before they reach the substrate.

The delivery of the vapour pressure of the Hg to the substrate for growth at 350°C must be greater than the minimum equilibrium pressure of 2 × 10-3 bar. In the author's system, the Hg source is an elemental mercury. The Hg partial pressure is about 30 mbar when the Hg zone temperature is maintained at 220°C during the IMP growth process.

### **2.2. MOCVD Growth approaches: IMP and DAG**

**•** high optical coefficients enabling high quantum efficiency.

photodetectors construction.

**2. MOCVD grown MCT**

**2.1. Thermodynamic considerations**

to the Hg partial pressure.

equilibrium is given by the following expression:

Above‐mentioned properties are simply consequence of the energy band structure of this zinc‐ blende semiconductor. The special advantages of HgCdTe are ability to obtain wide range of carrier concentrations, low dielectric constant, and high mobility of current carriers. The very small change of lattice constant with composition enables to grow high‐quality layered and graded gap structures [2]. Thus, HgCdTe can be used for detectors operating at various modes

This chapter reviews work from literature and some unpublished work from the authors' own laboratory that has been carried out into determining suitable extrinsic dopant species and sources for use in MOCVD growth of MCT heterostructures. The authors especially would like to present MOCVD technology with wide range of composition and donor/acceptor doping and without post‐grown annealing as an excellent tool for (HOT) HgCdTe infrared

MOCVD growth of MCT was firstly demonstrated in 1981 by Irvine and Mullin [3]. The reason for the delayed start in the growth of the mercury alloys in comparison to III–V compounds growth has been linked to the high mercury vapour pressures which needed some unconven‐ tional modifications to the MOCVD systems. MOCVD growth of MCT is determined by high vapour pressures of mercury that are necessarily to keep equilibrium conditions over the growing layer. This results from the instability of HgTe bonds in comparison with CdTe bonds and requires much lower growth temperatures than are typical for more stable compounds [4].

The mercury chalcogenides are characterized by weak bonds of mercury what causes a higher equilibrium vapour pressure. If the evaporating species are Hg and Te2, then the solid–vapour

where pHg is the equilibrium vapour pressure of mercury, pTe<sup>2</sup> is the equilibrium vapour pressure of Te2, and K is an equilibrium constant which is a function of temperature. Eq. (2.1) implies that the equilibrium pressures of the component elements are linked and that there is a range of pressures over which the solid remains in equilibrium as a single phase. From the pressure–temperature phase diagram for HgTe presented in **Figure 1**, we can conclude that at MOCVD growth temperatures, the pressure can vary by three orders of magnitude and remain in equilibrium with a single phase of HgTe. However, it is clear from Eq. (2.1) that Te<sup>2</sup> partial pressure will be varying across this phase field in the opposite sense

*Hg Te*<sup>2</sup> *Kp p* = (2.1)

[photodiode, photoconductor, or metal‐insulator‐semiconductor (MIS) detector].

70 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

There are two techniques for MOCVD growth of HgCdTe: the first based on direct growth of the ternary alloy—direct alloy growth (DAG) and the second based on interdiffused creation of the HgCdTe following initial deposition of alternating thin layers of CdTe and HgTe with total period thicknesses lower than 150 nm—interdiffused multilayer process (IMP). The latter technique takes advantage of the rapid interdiffusion rates (D~10-11–10-13 cm2 /s) of the cations in HgTe and CdTe at the typical growth temperatures in the range of 350–380°C. In IMP, the

**Figure 2.** Composition profiles versus thickness modelled from laser reflectance data for a portion of an MOCVD‐IMP growth run for x = 0.23 HgCdTe at different times (reproduced from reference [6]).

Hg1-xCdxTe final x composition is determined by the relative thickness of the HgTe and CdTe cycles which are easily monitored by controlling the times of their growth [4].

The sequence of four calculated compositional profiles for an IMP structure with 93.6 nm of HgTe and 27.6 nm of CdTe was presented on **Figure 2**. The cation interdiffusion process in IMP MOCVD growth technique of HgCdTe was demonstrated by monitoring the laser reflectance signal and modelling the IMP process. The HgCdTe was grown at 380°C. The first profile (the solid line) was taken at t = 300 s. It shows that formation of pure HgTe is inhibited by the fast diffusion of Cd atoms. The preceding CdTe layer is narrowing by interdiffusion but maintains composition x = 1 in the centre. The second period taken 35 s later is at the commence of the next CdTe growth phase, by which time the preceding CdTe layer has a maximum Cd composition of 0.6. The diffusion of Hg in CdTe is much slower than the diffusion of Cd in HgTe, and thus, the CdTe phase remains on the binary composition in the contrary to the HgTe phase. In the third time period taken 50 s after the experiment beginning, the x composition of CdTe layer has decreased below 0.3 and at after t = 180 s, entire structure is nearly completely homogeneous. The IMP growth technique is relatively ease in implementation, and its decisive advantage depends on the ability to control the Cd/Te ratio during the CdTe cycle of the growth. The dopants can be preferentially directed to the intended lattice site by controlling the Cd/Te ratio. Good uniformity in alloy composition and film thickness has been demon‐ strated by both techniques [7].

### **2.3. Growth mechanisms**

A variety of different models have been used to describe the growth mechanisms of narrow bandgap II–VI semiconductors, and no clear mechanism has yet emerged from this research [4]. The haemolytic fission of alkyl radicals is a stepwise process where there is a different bond energy for the first and second radicals. In the case of DMCd, the bond energy is 193 kJ mole-1 for the first methyl radical and 88 kJ mole-1 for the second. The decomposition can simply be described by the following stepwise reactions where the first step is rate limiting:

$$(CH\_3)\_2Cd \to CH\_3Cd^\* + CH\_3 \tag{2.2}$$

$$CH\_3Cd^\* \rightarrow CH\_3^\* + Cd \tag{2.3}$$

More comprehensive study concerning HgCdTe growth mechanisms, you can find in refer‐ ences [4, 8].

### **2.4. Substrate type and orientation**

Hg1-xCdxTe final x composition is determined by the relative thickness of the HgTe and CdTe

**Figure 2.** Composition profiles versus thickness modelled from laser reflectance data for a portion of an MOCVD‐IMP

The sequence of four calculated compositional profiles for an IMP structure with 93.6 nm of HgTe and 27.6 nm of CdTe was presented on **Figure 2**. The cation interdiffusion process in IMP MOCVD growth technique of HgCdTe was demonstrated by monitoring the laser reflectance signal and modelling the IMP process. The HgCdTe was grown at 380°C. The first profile (the solid line) was taken at t = 300 s. It shows that formation of pure HgTe is inhibited by the fast diffusion of Cd atoms. The preceding CdTe layer is narrowing by interdiffusion but maintains composition x = 1 in the centre. The second period taken 35 s later is at the commence of the next CdTe growth phase, by which time the preceding CdTe layer has a maximum Cd composition of 0.6. The diffusion of Hg in CdTe is much slower than the diffusion of Cd in HgTe, and thus, the CdTe phase remains on the binary composition in the contrary to the HgTe phase. In the third time period taken 50 s after the experiment beginning, the x composition of CdTe layer has decreased below 0.3 and at after t = 180 s, entire structure is nearly completely homogeneous. The IMP growth technique is relatively ease in implementation, and its decisive advantage depends on the ability to control the Cd/Te ratio during the CdTe cycle of the growth. The dopants can be preferentially directed to the intended lattice site by controlling the Cd/Te ratio. Good uniformity in alloy composition and film thickness has been demon‐

A variety of different models have been used to describe the growth mechanisms of narrow bandgap II–VI semiconductors, and no clear mechanism has yet emerged from this research [4]. The haemolytic fission of alkyl radicals is a stepwise process where there is a different bond energy for the first and second radicals. In the case of DMCd, the bond energy is 193 kJ mole-1 for the first methyl radical and 88 kJ mole-1 for the second. The decomposition can simply be

described by the following stepwise reactions where the first step is rate limiting:

cycles which are easily monitored by controlling the times of their growth [4].

72 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

growth run for x = 0.23 HgCdTe at different times (reproduced from reference [6]).

strated by both techniques [7].

**2.3. Growth mechanisms**

The proper selection of substrate material and orientation has been a significant field for research in narrow gap II–VI semiconductors because it was recognized to be a limiting factor in the quality of the epilayers. Generally, we can divide substrates onto two categories: the first category is lattice matched II–VI substrates: CdZnTe and CdSeTe, and the second category is lattice mismatched substrates: GaAs, Si and sapphire (Al2O3). The CdZnTe ternary is expensive material, and it suffers severe segregation causing non‐uniformity in alloy composition, and CdSeTe substrates have high impurity concentrations arising from the selenium source. Much of the MOCVD growth has concentrated on orientations close to the (100), normally with some misorientation to reduce the size of macro‐defects, otherwise known as hillocks or pyramids. In a detailed analysis of the frequency and shapes of defects on different misorientations, it was concluded by Snyder et al. [9] that the optimum orientation was (100) 3°–4° towards the (111)B face. The most extensively used alternative substrate has been GaAs with a few microns of CdTe buffer layer reducing the 14% lattice mismatch between CdTe and GaAs. Most of the MOCVD growth onto silicon substrates has used a GaAs buffer layer to step the change in lattice parameter between silicon and CdTe. Substrate orientation as well as the HgCdTe orientation has been extensively reported to have fundamental influence on crystallographic defects, surface morphology, residual background concentration, and both donor‐ and acceptor‐doping efficiency. **Figure 3** presents atomic arrangements of the (100) and (h,11)B surfaces on the (011) plane of HgCdTe structure. The HgCdTe crystallographic orientation influence on doping efficiency will be studied in next paragraphs.

**Figure 3.** Atomic arrangements of the (100) and (h,11)B surfaces on the (011) plane of HgCdTe structure [10].

### **3. Growth of HgCdTe in Aix 200 system**

HgCdTe epitaxial growth was carried out in the horizontal reactor of AIX‐200 Aixtron MOCVD system (**Figure 4**). It consists of a horizontal rectangular aperture liner enclosed in an outer circular quartz tube. The system is designed to operate in the laminar flow regime with process pressures from 50 to 1000 mbar using a butterfly valve for pressure control. Reactor pressure of 500 mbar was used for all successful growth runs. Hydrogen is used as a carrier gas. Dimethylcadmium (DMCd) and diisopropyltelluride (DIPTe) are used as precursors. Ethyl‐ iodide (EtI) is used as a donor dopant source and TDMAAs (or AsH<sup>3</sup> in our previous research‐ es) as an acceptor dopant source. DMCd and EtI are delivered through the one channel, while DIPTe and TDMAAs are delivered through the lower channel over elemental mercury bath.

**Figure 4.** The scheme of the gas supply installation and the reactor in AIX 200 MOCVD system adapted to HgCdTe growth.

The Aixtron's gas foil rotation technique has been applied for better composition uniformity. There are two temperature zones in the reactor: the Mercury source zone and the growth zone with graphite susceptor controlled by external infrared heaters. High‐temperature anneal was used before each growth run for reactor cleaning. Gas delivery system is additionally equipped with ultrasonic precursor concentration monitors—Piezocon and reflectometer—EpiEye. The usage of piezocons contributed to a better repeatability of the growth processes. Adaptation of EpiEye reflectometer allows for *in situ* monitoring. Typically, a 3–4‐µm thick CdTe layer is used as a buffer layer reducing stress caused by crystal lattice mismatch between GaAs substrate and HgCdTe epitaxial layer structure. The buffer plays also a role of Ga diffusion barrier. The interdiffused multilayer process (IMP) technique was applied for the HgCdTe deposition [6]. HgCdTe was grown at 350°C with mercury source kept at 210°C. The II/VI mole ratio was kept in the range from 1.5 to 5 during CdTe cycles of the IMP process. An acceptor and donor doping has been examined over the wide range of compositions and doping levels of 5 × 1014–5 × 1017 cm-3 have been obtained. Obtained HgCdTe heterostructures have been not annealed neither during the growth process (*in situ*) nor after the growth (*ex situ*) [11].

The set of metalorganic precursors gas supply suitable for Te and Cd delivery was presented in **Figure 5**. In our system, such a set supplies the reactor with DMCd and DIPTe. The H2 carrier gas controlled by MFC (mass flow controller) is introduced into the bubbler and taking metalorganic precursor flows to the reactor. The bubbler is placed in the thermostatic bath. The liquid precursor temperature is one of the parameters determining precursor partial pressure in the reactor.

**3. Growth of HgCdTe in Aix 200 system**

growth.

HgCdTe epitaxial growth was carried out in the horizontal reactor of AIX‐200 Aixtron MOCVD system (**Figure 4**). It consists of a horizontal rectangular aperture liner enclosed in an outer circular quartz tube. The system is designed to operate in the laminar flow regime with process pressures from 50 to 1000 mbar using a butterfly valve for pressure control. Reactor pressure of 500 mbar was used for all successful growth runs. Hydrogen is used as a carrier gas. Dimethylcadmium (DMCd) and diisopropyltelluride (DIPTe) are used as precursors. Ethyl‐ iodide (EtI) is used as a donor dopant source and TDMAAs (or AsH<sup>3</sup> in our previous research‐ es) as an acceptor dopant source. DMCd and EtI are delivered through the one channel, while DIPTe and TDMAAs are delivered through the lower channel over elemental mercury bath.

74 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 4.** The scheme of the gas supply installation and the reactor in AIX 200 MOCVD system adapted to HgCdTe

The Aixtron's gas foil rotation technique has been applied for better composition uniformity. There are two temperature zones in the reactor: the Mercury source zone and the growth zone with graphite susceptor controlled by external infrared heaters. High‐temperature anneal was used before each growth run for reactor cleaning. Gas delivery system is additionally equipped with ultrasonic precursor concentration monitors—Piezocon and reflectometer—EpiEye. The usage of piezocons contributed to a better repeatability of the growth processes. Adaptation of EpiEye reflectometer allows for *in situ* monitoring. Typically, a 3–4‐µm thick CdTe layer is used as a buffer layer reducing stress caused by crystal lattice mismatch between GaAs substrate and HgCdTe epitaxial layer structure. The buffer plays also a role of Ga diffusion barrier. The interdiffused multilayer process (IMP) technique was applied for the HgCdTe deposition [6]. HgCdTe was grown at 350°C with mercury source kept at 210°C. The II/VI mole ratio was kept in the range from 1.5 to 5 during CdTe cycles of the IMP process. An acceptor and donor doping has been examined over the wide range of compositions and doping levels of 5 × 1014–5 × 1017 cm-3 have been obtained. Obtained HgCdTe heterostructures have been not annealed neither during the growth process (*in situ*) nor after the growth (*ex situ*) [11].

**Figure 5.** The set of metalorganic precursors gas supply suitable for Te and Cd delivery.

The partial pressures of cadmium and tellurium metalorganic compounds were calculated using following formulas:

$$P\_p\left[mbar\right] = P\_k\left[mbar\right]\frac{\mathcal{Q}\_{M0}\left[\frac{mol}{min}\right]}{\mathcal{Q}\_{Total}\left[\frac{mol}{min}\right]}\tag{3.1}$$

where *QTotal* is the total gas flow through the reactor

$$\left[\mathcal{Q}\_{\rm MO}\left[\frac{mol}{min}\right] = \frac{\mathcal{Q}\_{\rm S}}{22.4} \cdot \frac{p\left(\nu\right)}{p\_c - p\left(\nu\right)}\tag{3.2}$$

*pc* is the pressure in the reactor, *p(v)* is vapour pressure of the metalorganic in the bubbler, *QS* is the MFC source flow.

For metalorganic compounds, their vapour pressure is calculated using following formula:

$$\mathcal{Q}\_{\rm MO} \left[ \frac{mol}{min} \right] = \frac{\mathcal{Q}\_{\rm S}}{22.4} \cdot \frac{p(\nu)}{p\_c - p(\nu)} \tag{3.3}$$

where T is bubbler temperature in Kelvins, and A and B are material constants shown in **Table 1**.


**Table 1.** Material constant examples of the metalorganic sources.

In the author's system, the Hg source is an elemental mercury. The Hg partial pressure can be read from **Figure 1** and is about 30 mbar when the Hg zone temperature is maintained at 220°C during the IMP growth process.

The examples of surface morphology of obtained (111) and (100) oriented HgCdTe layers are presented in **Figures 6** and **7**, respectively. Both layers were not doped intentionally and are about 10 µm thick. The pictures were obtained using optical microscope with Nomarski contrast with ×1000 magnification. Both layers were grown on (100) GaAs substrate with 3° misoriented towards the (111)B face. The (111) CdTe buffer growth on (100) GaAs was obtained using Te flush during nucleation. The (100) CdTe buffer growth on (100) GaAs was obtained using Cd flush during nucleation. Both obtained layer are characterized by high uniformity in the composition and the thickness as well due to the Aixtron's gas foil rotation technique. It is clearly seen that the smoothness of (100) HgCdTe is much better in comparison with (111) HgCdTe what is attributable to microtwins on (111) orientation. The surface roughness coefficient Rq is equal 70 and 6 nm for (111) and (100) orientation, respectively. Rq coefficient was measured on Veeco optical profiler type Wyko NT 1100.

**Figure 6.** The example of surface morphology of obtained (111) oriented HgCdTe layer.

**Figure 7.** The example of surface morphology of obtained (100) oriented HgCdTe layer.

### **3.1. Undoped electrical properties**

where *QTotal* is the total gas flow through the reactor

is the MFC source flow.

**Table 1**.

*MO*

*MO*

**Material constant Metalorganic sources**

**Table 1.** Material constant examples of the metalorganic sources.

was measured on Veeco optical profiler type Wyko NT 1100.

during the IMP growth process.

( ) ( ) . 22.4 *S*

( ) ( ) . 22.4 *S*

*c*

where T is bubbler temperature in Kelvins, and A and B are material constants shown in

A 1850 2309 1715 2109 B 7764 8288 7877 8280

In the author's system, the Hg source is an elemental mercury. The Hg partial pressure can be read from **Figure 1** and is about 30 mbar when the Hg zone temperature is maintained at 220°C

The examples of surface morphology of obtained (111) and (100) oriented HgCdTe layers are presented in **Figures 6** and **7**, respectively. Both layers were not doped intentionally and are about 10 µm thick. The pictures were obtained using optical microscope with Nomarski contrast with ×1000 magnification. Both layers were grown on (100) GaAs substrate with 3° misoriented towards the (111)B face. The (111) CdTe buffer growth on (100) GaAs was obtained using Te flush during nucleation. The (100) CdTe buffer growth on (100) GaAs was obtained using Cd flush during nucleation. Both obtained layer are characterized by high uniformity in the composition and the thickness as well due to the Aixtron's gas foil rotation technique. It is clearly seen that the smoothness of (100) HgCdTe is much better in comparison with (111) HgCdTe what is attributable to microtwins on (111) orientation. The surface roughness coefficient Rq is equal 70 and 6 nm for (111) and (100) orientation, respectively. Rq coefficient

ë û - (3.2)

ë û - (3.3)

Cd Te EtI Zn

*c*

*pc* is the pressure in the reactor, *p(v)* is vapour pressure of the metalorganic in the bubbler, *QS*

For metalorganic compounds, their vapour pressure is calculated using following formula:

*mol Q p v <sup>Q</sup> min p p v*

é ù <sup>=</sup> ê ú

*mol Q p v <sup>Q</sup> min p p v*

76 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

é ù <sup>=</sup> ê ú

Indispensable in understanding the electrical properties of doped layers is the ability in interpretation of undoped layers results. The electrical properties of undoped MOCVD‐grown MCT are dependent on the different parameters such as contamination from alkyls or substrates [3, 12], crystallographic defects such as microtwins on (111) orientation, quench cooling conditions and native defects like mercury vacancies and others [13]. Although such defects do not affect interpretation of highly doped layers like contact layers, they may confuse interpretation of low‐doped layers like absorbing layers in photodiodes structure. **Figure 8** presents residual background concentration of undoped HgCdTe layers versus consecutive number of growth processes carried out within 3 years in author's lab.

**Figure 8.** Residual background concentrations of undoped HgCdTe layers versus the numbers of growth processes.

The residual background concentration is a matter of huge concern of engineers form each semiconductor laboratory because it is the serious test of the equipment and applied materials purity as well as the result of the efforts of the staff. The mean residual concentration main‐ tained in our laboratory is about 4 × 10<sup>15</sup> cm-3 for (111) orientation and 4.5 × 10<sup>14</sup> cm-3 for (100) orientation. An order of magnitude higher background concentration for (111) concentration is attributable to the microtwins presence. In the **Figure 8**, we can observe single points with positive (hole) concentrations above level of 10<sup>16</sup> cm-3 for (100) orientations. These happened during experiments with lowering the mercury zone temperature in order to reduce the hillock density. Then, the mercury vacancies were created. The lowest background donor concentra‐ tion obtained in our system is about 2 × 10<sup>14</sup> cm-3 what is similar with the best results reported from other labs. The matter of residual donor concentration is particularly important during p‐type doping at the low level.

### **4. Doping study**

substrates [3, 12], crystallographic defects such as microtwins on (111) orientation, quench cooling conditions and native defects like mercury vacancies and others [13]. Although such defects do not affect interpretation of highly doped layers like contact layers, they may confuse interpretation of low‐doped layers like absorbing layers in photodiodes structure. **Figure 8** presents residual background concentration of undoped HgCdTe layers versus consecutive

78 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 8.** Residual background concentrations of undoped HgCdTe layers versus the numbers of growth processes.

p‐type doping at the low level.

The residual background concentration is a matter of huge concern of engineers form each semiconductor laboratory because it is the serious test of the equipment and applied materials purity as well as the result of the efforts of the staff. The mean residual concentration main‐ tained in our laboratory is about 4 × 10<sup>15</sup> cm-3 for (111) orientation and 4.5 × 10<sup>14</sup> cm-3 for (100) orientation. An order of magnitude higher background concentration for (111) concentration is attributable to the microtwins presence. In the **Figure 8**, we can observe single points with positive (hole) concentrations above level of 10<sup>16</sup> cm-3 for (100) orientations. These happened during experiments with lowering the mercury zone temperature in order to reduce the hillock density. Then, the mercury vacancies were created. The lowest background donor concentra‐ tion obtained in our system is about 2 × 10<sup>14</sup> cm-3 what is similar with the best results reported from other labs. The matter of residual donor concentration is particularly important during

number of growth processes carried out within 3 years in author's lab.

Improved performance IR photodiodes which require the growth of doped heterostructures have been theoretically modelled [14, 15]. Such structures require control of the dopant concentrations and sharp transitions between either dopant or different concentrations within a structure combined with bandgap changes induced by varying the alloy composition. To realize these doped junction structures, the dopants chosen need to have low diffusion coefficients at the growth temperature. Another considerable issue during doped heterostruc‐ tures deposition is the control of any dopant memory effects. Some dopant sources may demonstrate 'run‐to‐run' memory effects in which the dopant is found to contaminate the run subsequent to the one in which the dopant was used. The dopant may induce memory effects within a single run which can limit the abruptness of the dopant transition. Then, the order of the layers within the structure can influence the sharpness of the electrical junction. Vast majority of papers concerning the influence of substrate orientation on impurity incorporation have pointed the complexity of the phenomena in play and have only given qualitative interpretations.

### **4.1. Donor‐doping control**

The selection for donor dopant in MCT is between the group III elements (Al, Ga, In) on metal sites and the group VII elements (Cl, Br I) on the Te sites. Despite some concerns over the diffusion coefficient of the group III species, these have been the most widely studied with most work concentrating on indium as the slowest diffuser of the group III elements [16, 17]. Although the reactive nature of the halogen elements, iodine doping has been studied at PRL [18] for the first time. The real aim of donor doping for the most device applications is to obtain control of the donor concentrations at ≈10<sup>15</sup> cm-3 for absorbing layers and at ≈10<sup>17</sup> cm-3 for n+ contact layers.

### *4.1.1. Indium‐ and aluminium‐doping studies*

Trimethyl indium (TMIn) has been studied most widely as a doping source [13]. Despite its low vapour pressure, this source yields very high dopant incorporation. In incorporated preferentially into HgTe (20–50 times higher than in CdTe). This effect was use to attempt to lower the doping range by only injecting the TMIn during the CdTe cycles. However, the controllable doping range with a bubbler temperature of 2°C was still only from 2 × 10<sup>17</sup> to 3  × 10<sup>18</sup> cm-3. Lower bubbler temperatures (-10°C) have been applied to obtain minimum concentration of 5 × 10<sup>16</sup> cm**-**<sup>3</sup> in MCT grown at 370°C using DAG. 'Effuser' mode bubbler operation with TMIn was investigated to dope at low level, but results were not promising. Although this did enable layers with ND**-**A = 2 × 10<sup>16</sup> cm-3 to be grown, the rate of increase in the donor concentration with bubbler temperature was very abrupt, making reproducible control very difficult. Similar rapid doping transitions were observed using a lower vapour pressure source [ethyldimethyl indium (EDMIn)] in bubbler mode meant that control of the donor concentration was only possible down to ≈1017 cm**-**<sup>3</sup> [19]. This threshold effect was assigned to alkyl adsorption effects which were dependent on reactor design. Adsorption in the gas lines and manifold can cause 'run‐to‐run' memory effects extended over several growth runs. The electrical properties of as‐grown In‐doped layers, which contain significant Hg vacancy concentrations, suggest that only 30% of the In is electrically active and that the In introduces additional Hg vacancies causing autocompensation in these layers [13].

### *4.1.2. Iodine‐doping studies*

The group VII elements, substituting on Te lattice sites, were expected to be slower diffusers. Preliminary researches of iodine doping in MOCVD were carried out with elemental iodine I2. It has a vapour pressure of 0.3 Torr at 25°C. Electrically active donor incorporation was observed at low—1015 cm-3 doping levels with high mobilities. The proper control with abrupt doping profiles was not demonstrated, however, because I<sup>2</sup> reacts with DMCd. Improved doping efficiency has been expected when alkyl iodides were applied. In DAG deposition, iodine doping from isopropyl iodide has been reported to levels as low as 5 × 1015 cm-3. However, isopropyl iodide introduced a memory effect. The initial choice for a MO iodide source was ethyliodide (EtI) due to its commercial availability. EtI is a highly effective precursor for doping without any memory effects [20]. Controlled iodine doping has been achieved in the range of 3 × 1014 –2 × 1018 cm-3 with 100% electrical activation following a standard Hg‐rich stoichiometric anneal at 235°C. However, some works present similar donor‐ doping results without annealing. A double‐dilution bubbler configuration is presented in **Figure 9**, and it has been used to obtain controlled low‐level doping.

**Figure 9.** The double‐dilution bubbler configuration for EtI doping.

The special doping structures contribute to reduce the number of growth experiments required to assess a doping source. Multilevel 'staircase' structures quickly yielded dopant source calibration curves and growth of a highly doped layer between two low‐doped regions in a 'sandwich' structure yielded information on diffusion and reactor memory effects. **Figure 10** shows a 'staircase' (111) HgCdTe structure obtained using EtI which was assessed by SIMS and differential 77 K Hall profiling. Presented SIMS profile is shown from the growth per‐ spective, while vast majority of such curves published in the literature are referenced from the layer surface. The errors in differential Hall concentrations become larger when the etch step size is around 2 µm. Consequently, it is very difficult to align the step positions with the grown layer interfaces. The correlation obtained between the chemical and electrical results implied the iodine incorporated on the correct Te lattice site; and that within the accuracy of the techniques, the activation efficiency is high. However, when we look at the **Figure 10**, the EtI‐ doping efficiency (the adjacency between chemical and electrical concentration) is decreasing with increasing dose level. This was corroborated by further layers, and **Figure 11** shows the resultant EtI calibration curve showing the linear control of both iodine and 77 K Hall con‐ centration as a function of the injected EtI concentration in the reactor. For the comparison calibration curve for (100), HgCdTe layers is shown. We can see above one order of magnitude higher EtI‐doping efficiency: incorporation and activation for (111)HgCdTe than for (100). The iodine should be located in the tellurium sublattice sites to act as a donor. The Te sides on the (111)B surface provide a more stable adsorption site with three bonds from the underlying group II atoms. On the contrary, the (100) surface provides a weaker adsorption site with two bonds (**Figure 3**).

and manifold can cause 'run‐to‐run' memory effects extended over several growth runs. The electrical properties of as‐grown In‐doped layers, which contain significant Hg vacancy concentrations, suggest that only 30% of the In is electrically active and that the In introduces

80 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The group VII elements, substituting on Te lattice sites, were expected to be slower diffusers. Preliminary researches of iodine doping in MOCVD were carried out with elemental iodine I2. It has a vapour pressure of 0.3 Torr at 25°C. Electrically active donor incorporation was observed at low—1015 cm-3 doping levels with high mobilities. The proper control with abrupt doping profiles was not demonstrated, however, because I<sup>2</sup> reacts with DMCd. Improved doping efficiency has been expected when alkyl iodides were applied. In DAG deposition, iodine doping from isopropyl iodide has been reported to levels as low as 5 × 1015 cm-3. However, isopropyl iodide introduced a memory effect. The initial choice for a MO iodide source was ethyliodide (EtI) due to its commercial availability. EtI is a highly effective precursor for doping without any memory effects [20]. Controlled iodine doping has been achieved in the range of 3 × 1014 –2 × 1018 cm-3 with 100% electrical activation following a standard Hg‐rich stoichiometric anneal at 235°C. However, some works present similar donor‐ doping results without annealing. A double‐dilution bubbler configuration is presented in

additional Hg vacancies causing autocompensation in these layers [13].

**Figure 9**, and it has been used to obtain controlled low‐level doping.

**Figure 9.** The double‐dilution bubbler configuration for EtI doping.

*4.1.2. Iodine‐doping studies*

**Figure 10.** SIMS and differential excess donor concentration profiles of a 'staircase' (111) HgCdTe structure, doped us‐ ing EtI [21].

**Figure 11.** Calibration curve obtained with EtI for (111) and (100) oriented HgCdTe layers.

### **4.2. Acceptor doping**

Acceptor doping should be provided by applying elements from the group IA (Li, Na, K), IB (Cu, Ag, Au) or group VB (P, As, Sb, Bi). The group I elements should be positioned on the metal lattice sites in order to provide theirs activity as the acceptors. The elements from group V need to be located on the tellurium sublattice sites. The group V elements are slower‐ diffusing species in HgCdTe and are versatile dopant capable of producing a variety of abrupt, doped junctions. If the group V elements were located on metal sites then they are likely to reveal donor behaviour. The most widely used element for acceptor doping is arsenic [22–28], although researches involving phosphorus and antimony have also been reported [13].

**Figure 12** presents incorporation and activation of As from TDMAAs and AsH3 versus precursor partial pressure for (111)HgCdTe. The chemical As concentration was determined by SIMS measurements. There is also comparison with 77K NA-D (100)HgCdTe doped from TDMAAs. The As chemical concentration increases roughly proportionally with As precursor partial pressure up to 10 ppm. The levels of As incorporation from AsH3 and TDMAAs into growing (111)HgCdTe samples are equal considering the measurement uncertainty. Arsenic precursors were introduced to the reactor during CdTe cycles of IMP growth process. The arsenic atoms appearance accelerates DMCd pyrolysis in the reactor that causes high shift in x composition. Chemical concentration of arsenic atoms measured by SIMS does not comply with acceptor electrical concentration determined by 77‐K Hall measurements so we do not observe full arsenic activation. Also, for the same dose of arsenic in the reactor, the holes concentration in the (100)HgCdTe is about one order of magnitude higher than in (111)HgCdTe. The arsenic incorporation considerably depends on crystallographic orienta‐ tion. This dependence can be explained by considering that the two Cd and As atoms are readily active in the incorporation process. The created Cd–As species can break into on the surface if proper conditions are realized. In this respect, the (100) plane offers the best config‐ uration because this surface has two double dangling bonds available for both Cd and Te (or substituting As) surface atoms (**Figure 3**). This mechanism can thus account for the highest doping observed with (100) substrates. Going away from (100) plane towards (111)A or (111)B leads to reduced As incorporation, and this corresponds to the fact that the relative fraction of (100) terraces decreases at the expense of (111) steps.

**Figure 12.** Incorporation and activation of As from TDMAAs and AsH3 versus partial pressure of precursors.

### **5. Heterostructures and devices**

**Figure 11.** Calibration curve obtained with EtI for (111) and (100) oriented HgCdTe layers.

82 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Acceptor doping should be provided by applying elements from the group IA (Li, Na, K), IB (Cu, Ag, Au) or group VB (P, As, Sb, Bi). The group I elements should be positioned on the metal lattice sites in order to provide theirs activity as the acceptors. The elements from group V need to be located on the tellurium sublattice sites. The group V elements are slower‐ diffusing species in HgCdTe and are versatile dopant capable of producing a variety of abrupt, doped junctions. If the group V elements were located on metal sites then they are likely to reveal donor behaviour. The most widely used element for acceptor doping is arsenic [22–28], although researches involving phosphorus and antimony have also been reported [13].

**Figure 12** presents incorporation and activation of As from TDMAAs and AsH3 versus precursor partial pressure for (111)HgCdTe. The chemical As concentration was determined by SIMS measurements. There is also comparison with 77K NA-D (100)HgCdTe doped from TDMAAs. The As chemical concentration increases roughly proportionally with As precursor partial pressure up to 10 ppm. The levels of As incorporation from AsH3 and TDMAAs into growing (111)HgCdTe samples are equal considering the measurement uncertainty. Arsenic precursors were introduced to the reactor during CdTe cycles of IMP growth process. The arsenic atoms appearance accelerates DMCd pyrolysis in the reactor that causes high shift in x composition. Chemical concentration of arsenic atoms measured by SIMS does not comply with acceptor electrical concentration determined by 77‐K Hall measurements so we do not observe full arsenic activation. Also, for the same dose of arsenic in the reactor, the holes concentration in the (100)HgCdTe is about one order of magnitude higher than in

**4.2. Acceptor doping**

After achieving device‐quality MCT layers with doping control over the ranges required, the growth of doped junctions can be studied. When both donor and acceptor doping is required, for example in P+ /π/N<sup>+</sup> photodiodes, we prefer MCT with (111) crystallographic orientation because of limits with high‐concentration doping control in (100)MCT. The capital letters P and N in device description means acceptor‐ and donor‐doped layers with wider gap material. The upper index '+' denote high doping. If the particular device requires p‐type doping only then is constructed on the basis of (100)HgCdTe. As an example, the MWIR photoresistors are presented in the Section 5.3. Cryogenic cooling of detectors has always been the disadvantage of sensitive infrared (IR) systems. Conventional HgCdTe IR photodetectors should be cooled down below ambient temperature to reduce noise and leakage currents resulting from thermal generation processes. Presented in previous section progress in MOCVD technology let construct IR detectors operating without cryogenic cooling. Uncooled—operating in room temperature or thermoelectrically cooled devices—operating typically at 230 K is named as higher operating temperature (HOT) devices. The examples of IR HOT detectors operating in different IR spectral ranges are shown in the following sections.

### **5.1. LWIR photodiodes**

The dark current in LWIR HgCdTe photodiodes is primary determined by Auger generation processes at elevated temperatures. The low‐doped absorber layer becomes intrinsic, and the carrier concentration is higher than the doping level. The device structures with a combination of exclusion and extraction junctions in N+ /π/P<sup>+</sup> configurations have demonstrated suppres‐ sion of Auger mechanisms by reducing the absorber carrier density below thermal equilibrium in a reverse bias condition. Classical N+ /π/P<sup>+</sup> structure has been expanded with graded interface layers denoted as 'G'. Graded doping and composition x layers represent the real structure which profile is shaped by interdiffusion processes during Hg1-xCdxTe growth at 350°C. Thus, N<sup>+</sup> /G/π/G/P<sup>+</sup> /G/n+ HgCdTe photodiode structure has been obtained as it is shown in **Figure 13**. In order to improve electrical contact properties of metallization to P+ layer, the structure was upgraded with p+ /n+ tunnelling junction on the top.

**Figure 13.** The LWIR N+ /G/π/G/P<sup>+</sup> /G/n+ HgCdTe photodiode diagram. *x* is the alloy composition, *NA*—the acceptor concentration, and *ND*—the donor concentration [29].

The cladding N+ ‐layer should be thicker than minority carrier diffusion length and doped at possible highest level providing reduction of series resistance. Analogously parameters of cladding P+ ‐layer are defined. The thickness of active π‐type region (absorbing layer) should be shorter than minority carrier diffusion length. Generally, the thickness of absorber layer is a compromise between requirements of high absorption efficiency and low thermal genera‐ tion. In our experiments, the thickness of absorber layer is varied from 3 to 6 µm. The acceptor doping of the absorber should be at possible low level just to overcompensate the donor background concentration of the material.

of sensitive infrared (IR) systems. Conventional HgCdTe IR photodetectors should be cooled down below ambient temperature to reduce noise and leakage currents resulting from thermal generation processes. Presented in previous section progress in MOCVD technology let construct IR detectors operating without cryogenic cooling. Uncooled—operating in room temperature or thermoelectrically cooled devices—operating typically at 230 K is named as higher operating temperature (HOT) devices. The examples of IR HOT detectors operating in

84 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The dark current in LWIR HgCdTe photodiodes is primary determined by Auger generation processes at elevated temperatures. The low‐doped absorber layer becomes intrinsic, and the carrier concentration is higher than the doping level. The device structures with a combination

/π/P<sup>+</sup>

/π/P<sup>+</sup>

in **Figure 13**. In order to improve electrical contact properties of metallization to P+

sion of Auger mechanisms by reducing the absorber carrier density below thermal equilibrium

interface layers denoted as 'G'. Graded doping and composition x layers represent the real structure which profile is shaped by interdiffusion processes during Hg1-xCdxTe growth at

tunnelling junction on the top.

configurations have demonstrated suppres‐

structure has been expanded with graded

layer, the

HgCdTe photodiode structure has been obtained as it is shown

HgCdTe photodiode diagram. *x* is the alloy composition, *NA*—the acceptor

different IR spectral ranges are shown in the following sections.

**5.1. LWIR photodiodes**

350°C. Thus, N<sup>+</sup>

**Figure 13.** The LWIR N+

/G/π/G/P<sup>+</sup>

concentration, and *ND*—the donor concentration [29].

/G/n+

of exclusion and extraction junctions in N+

in a reverse bias condition. Classical N+

/G/π/G/P<sup>+</sup>

structure was upgraded with p+

/G/n+

/n+

The SIMS profiles of N<sup>+</sup> /G/π/G/P<sup>+</sup> /G/n<sup>+</sup> heterostructure measured by CAMECA IMS 6F using positive and negative Cs ions are presented in **Figures 14** and **15**, respectively. The *x*‐compo‐ sition represented by the solid green lines has been calculated taking into account the measured SIMS points for Cd, Hg and Te elements. There is quite good convergence between projected and measured values in the absorber region. There are discrepancies between positive and negative ion measurements which result from incorrect SIMS ions calibration; about half of the order of magnitude differences between positive and negative ions for the arsenic‐doping profile. The diffusion processes during the epitaxial growth cause gradient profile in the interface layers adjacent to the absorber region.

**Figure 14.** SIMS profile through N<sup>+</sup> /G/π/G/P<sup>+</sup> /G/n<sup>+</sup> (111)HgCdTe heterostructure measured by using positive caesium ions.

**Figure 15.** SIMS profile through N+ /G/π/G/P<sup>+</sup> /G/n+ (111)HgCdTe structure measured by using negative caesium ions.

The gradient slope of arsenic line between P<sup>+</sup> and absorber π‐region seen in **Figures 14** and **15** indicates arsenic diffusion from P<sup>+</sup> ‐region to absorber region what is dangerous especially in thin, 3‐µm thick, absorber layers. The iodine profile (the solid pink line with triangles) seen in **Figure 15** indicates expected donor‐doping levels: 1 × 10<sup>19</sup> cm-3 for n+ ‐layer and 1 × 10<sup>18</sup> cm-3 for N+ ‐layer with EI doses 90 and 19 ppms, respectively. Unwanted iodine presence in the absorber region is at the level of 1 × 10<sup>15</sup> cm-3, what is close to the SIMS detection limit for this element.

The Hg0.81Cd0.19Te N+ /G/π/G/P<sup>+</sup> /G/n+ photodiode's current–voltage characteristic is presented in **Figure 16** for device with active region doped with 7 ppms of TDMAAs dose during the active layer deposition. Measurements were taken at 300 K (without any cooling) using the Keithley 2400 sourcemeter. The electrical area of devices is 8.1 × 10-9 m<sup>2</sup> . The solid line denotes theoretically calculated results applying the *APSYS* simulation platform. The negative differential resistance between -150and -250 mV is attributable to the suppression of Auger processes due to the exclusion and extraction phenomena. The dark current is determined by both band‐to‐band (BTB) and trap‐assisted (TAT) tunnelling mechanisms which evidences the quality of the material, for example the point defects like mercury vacancies. More compre‐ hensive studies concerning fitting procedure have been reported in references [14, 15].

**Figure 16.** Modelled and measured characteristics of dark current density versus bias voltage for LWIR N+ /G/π/G/P + /G/n+ HgCdTe photodiode at room temperature [29].

The spectral characteristics have been measured at zero and reverse bias polarization using the Perkin Elmer FT‐IR Spectrometer type Spectrum 2000. **Figure 17** presents the current responsivity versus wavelength for long wavelength Hg0.81Cd0.19Te detector measured at 300  K. TDMAAs dose of 3.6 ppms was applied during the growth of the absorbing region. The values of the reverse biases are typically chosen considering current–voltage analysis and are typically beyond the threshold voltage *UT* (designated in **Figure 16**). We can increase the current responsivity up to 50 times by the use of the reverse bias and thus causing Auger generation process suppression.

**Figure 15.** SIMS profile through N+

The Hg0.81Cd0.19Te N+

N+

indicates arsenic diffusion from P<sup>+</sup>

/G/π/G/P<sup>+</sup>

**Figure 15** indicates expected donor‐doping levels: 1 × 10<sup>19</sup> cm-3 for n+

Keithley 2400 sourcemeter. The electrical area of devices is 8.1 × 10-9 m<sup>2</sup>

/G/π/G/P<sup>+</sup>

/G/n+

86 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The gradient slope of arsenic line between P<sup>+</sup> and absorber π‐region seen in **Figures 14** and **15**

thin, 3‐µm thick, absorber layers. The iodine profile (the solid pink line with triangles) seen in

in **Figure 16** for device with active region doped with 7 ppms of TDMAAs dose during the active layer deposition. Measurements were taken at 300 K (without any cooling) using the

theoretically calculated results applying the *APSYS* simulation platform. The negative differential resistance between -150and -250 mV is attributable to the suppression of Auger processes due to the exclusion and extraction phenomena. The dark current is determined by both band‐to‐band (BTB) and trap‐assisted (TAT) tunnelling mechanisms which evidences the quality of the material, for example the point defects like mercury vacancies. More compre‐ hensive studies concerning fitting procedure have been reported in references [14, 15].

‐layer with EI doses 90 and 19 ppms, respectively. Unwanted iodine presence in the absorber region is at the level of 1 × 10<sup>15</sup> cm-3, what is close to the SIMS detection limit for this element.

(111)HgCdTe structure measured by using negative caesium ions.

‐region to absorber region what is dangerous especially in

/G/n+ photodiode's current–voltage characteristic is presented

‐layer and 1 × 10<sup>18</sup> cm-3 for

. The solid line denotes

**Figure 17.** Current responsivity versus wavelength for LWIR HgCdTe photodiodes measured at room temperature [29].

### **5.2. MWIR photoresistors**

The voltage responsivity versus wavelength for (100) and (111) HgCdTe MWIR photoresistors is presented in **Figure 18**. Photoresistors constructed from (100)HgCdTe have higher respon‐ sivity than these constructed from (111)HgCdTe. As it was discussed in the Section 4.2, there is higher acceptor‐doping efficiency in (100)HgCdTe in comparison with (111)HgCdTe. However, presented values of the voltage responsivities are not fully comparable, because of the differences in Hg1-xCdxTe compositions in presented devices. The absorber composition in (100)HgCdTe was x = 0.337 whereas in (111)HgCdTe was x = 0.331. The cut‐off wavelength λCO of measured photoresistors moves to longer waves with decreasing temperature because the energy gap is narrower.

**Figure 18.** The voltage responsivity versus wavelength for (100) and (111) HgCdTe MWIR photoresistors.

### **6. Summary**

In spite of other competitive technologies like InAs/GaSb superlattices, MCT material takes leading position in infrared detector industry at the time of writing this chapter. Donor‐ and acceptor‐doping researches in (100) and (111) oriented HgCdTe layers grown by MOCVD have been studied. Fully doped HgCdTe heterostructures with acceptor concentration range between 1014 and 5 × 10<sup>17</sup> cm-3 and donor concentration range between 1014 and 1 × 10<sup>18</sup> cm-3 and without post‐grown annealing have been reported. The electrical and chemical characteriza‐ tions of HgCdTe structures grown at 360°C on GaAs substrates using DIPTe have been described. Infrared photodiodes with different composition x were constructing on the basis of obtained heterostructures enabling signal detection of any wavelength from 1 µm to above 20 µm covering SWIR, MWIR and LWIR spectral ranges. Presented experimental results show that MOCVD technology enables to grow HgCdTe structures dedicated for HOT devices.

### **Acknowledgements**

**5.2. MWIR photoresistors**

energy gap is narrower.

**6. Summary**

The voltage responsivity versus wavelength for (100) and (111) HgCdTe MWIR photoresistors is presented in **Figure 18**. Photoresistors constructed from (100)HgCdTe have higher respon‐ sivity than these constructed from (111)HgCdTe. As it was discussed in the Section 4.2, there is higher acceptor‐doping efficiency in (100)HgCdTe in comparison with (111)HgCdTe. However, presented values of the voltage responsivities are not fully comparable, because of the differences in Hg1-xCdxTe compositions in presented devices. The absorber composition in (100)HgCdTe was x = 0.337 whereas in (111)HgCdTe was x = 0.331. The cut‐off wavelength λCO of measured photoresistors moves to longer waves with decreasing temperature because the

88 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 18.** The voltage responsivity versus wavelength for (100) and (111) HgCdTe MWIR photoresistors.

In spite of other competitive technologies like InAs/GaSb superlattices, MCT material takes leading position in infrared detector industry at the time of writing this chapter. Donor‐ and acceptor‐doping researches in (100) and (111) oriented HgCdTe layers grown by MOCVD have been studied. Fully doped HgCdTe heterostructures with acceptor concentration range between 1014 and 5 × 10<sup>17</sup> cm-3 and donor concentration range between 1014 and 1 × 10<sup>18</sup> cm-3 and without post‐grown annealing have been reported. The electrical and chemical characteriza‐ tions of HgCdTe structures grown at 360°C on GaAs substrates using DIPTe have been

This study was supported by Polish National Science Centre as the Research Project No. UMO‐ 2013/08/A/ST5/00773 and Polish National Centre for Research and Development as the Project No. TANGO1/2665576/NCBR/2015

### **Author details**

Pawel Madejczyk1\*, Waldemar Gawron1 , Artur Keblowski2 and Adam Piotrowski2

\*Address all correspondence to: pmadejczyk@wat.edu.pl

1 Military University of Technology, Warsaw, Poland

2 VIGO System S.A., Ozarow Mazowiecki, Poland

### **References**


[17] Irvine SJC, Bajaj J, Bubulac LO, Lin WP, Gedridge RW, Higa KT. A new n‐type doping precursor for MOCVD‐IMP growth of detector quality MCT. Journal of Electronic Materials. 1993;22(8):859–864. doi:10.1007/BF02817498

[6] Svoronos SA, Woo WW, Irvine SJC, Sankur HO, Bajaj J. Model of the interdiffused multilayer process. Journal of Electronic Materials. 1996;25:1561–1571. doi:10.1007/

90 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

[7] Mitra P, Case FC, Reine MB. Progress in MOVPE of HgCdTe for advanced infrared detectors. Journal of Electronic Materials. 1998;27:510–520. doi:10.1007/s11664‐998‐

[8] Bell W, Stevenson J, Cole‐Hamilton DJ, Halis JE. Evidence for a surface‐bound free

[9] Snyder DW, Mahajan S, Ko EI, Sides PJ. Effect of substrate misorientation on surface morphology of homoepitaxial CdTe films grown by organometallic vapor phase

[10] Svob L, Cheye I, Lusson A, Ballutaud D, Rommeluere JF, Marfaing Y. Crystallographic orientation dependence of As incorporation in MOVPE‐grown CdTe and correspond‐ ing acceptor electrical state activation. Journal of Crystal Growth. 1998;184/185:459–

[11] Madejczyk P, Gawron W, Piotrowski A, Klos K, Rutkowski J, Rogalski A. Improvement in performance of high‐operating temperature HgCdTe photodiodes. Infrared Physics

[12] Madejczyk P, Gawron W, Piotrowski A, Klos K, Rutkowski J, Rogalski A. Influence of TDMAAs acceptor precursor on performance improvement of HgCdTe photodiodes.

[13] Maxey CD, GaleIG, Clegg JB, Whiffin. Doping studies in MOVPE‐grown HgCdTe. Semiconductor Science and Technology. 1993;8:S183–S196. doi:10.1088/0268‐1242/8/1S/

[14] Martyniuk P, Gawron W, Madejczyk P, Rogalski A, Piotrowski J. Modelling of HgCdTe LWIR detector for high operations temperature conditions. Metrology and Measure‐

[15] Martyniuk P, Gawron W, Pawluczyk J, Keblowski A, Madejczyk P, Rogalski A. Dark current suppression in HOT LWIR HgCdTe heterostructures operating in non‐ equilibrium mode. Journal of Infrared and Millimeter Waves. 2015;34(4):385–390. [16] Gough JS, Houlton MR, Irvine SJC, Shaw N, Young ML, Astles MG. The growth and properties of In‐doped metalorganic vapor phase epitaxy interdiffused multilayer process (HgCd)Te. Journal of Vacuum Science and Technology. 1991;B9(3):1687–1690.

and Technology. 2011;54:310–315. doi:10.1016/j.infrared.2010.12.036

epitaxy. Applied Physics Letters. 1991;58(8):848–850. doi:10.1063/1.104509

mercury and/or Me2Cd under MOVPE conditions obtained from deuterium‐labelled precursors. Polyhedron. 1994;13:1253–1265. doi:10.1016/S0277‐5387(00)80259‐4

Pr2Te in the presence or absence of

radical mechanism during the decomposition of i

464. doi:10.1016/S0022‐0248(98)80096‐1

ment Systems. 2013;XX(2):159–170.

doi:10.1116/1.585400

Acta Physica Polonica A. 2010;118(6):1199–1204.

BF02655400

0007‐5

042


iodine doping. Journal of Electronic Materials. 1995;24(9):1077–1085. doi:10.1007/ BF02653056

[29] Madejczyk P, Gawron W, Martyniuk P, Keblowski A, Piotrowski A, Pawluczyk J, Pusz W, Kowalewski A, Piotriwski J, Rogalski A. MOCVD grown HgCdTe device structure for ambient temperature LWIR detectors. Semiconductor Science and Technology. 2013;28:105017–105023. doi:10.1088/0268‐1242/28/10/105017

### **Hot Filament Chemical Vapor Deposition: Enabling the Scalable Synthesis of Bilayer Graphene and Other Carbon Materials**

Frank Mendoza, Tej B. Limbu, Brad R. Weiner and Gerardo Morell

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/63921

### **Abstract**

iodine doping. Journal of Electronic Materials. 1995;24(9):1077–1085. doi:10.1007/

[29] Madejczyk P, Gawron W, Martyniuk P, Keblowski A, Piotrowski A, Pawluczyk J, Pusz W, Kowalewski A, Piotriwski J, Rogalski A. MOCVD grown HgCdTe device structure for ambient temperature LWIR detectors. Semiconductor Science and Technology.

2013;28:105017–105023. doi:10.1088/0268‐1242/28/10/105017

92 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

BF02653056

The hot filament chemical vapor deposition (HFCVD) technique is limited only by the size of the reactor and lends itself to be incorporated into continuous roll-to-roll industrial fabrication processes. We discuss the HFCVD reactor design and the interplay between the reactor parameters, such as filament and substrate temperatures, filamentto-substrate distance, and total pressure. Special attention is given to the large-area synthesis of bilayer graphene on copper, which is successfully grown by HFCVD with transmittance greater than 90% in the visible region and no gaps. We also discuss the HFCVD synthesis of carbon nanotubes, microcrystalline diamond, and nanocrystal‐ line diamond.

**Keywords:** hot filament chemical vapor deposition, bilayer graphene, diamond, carbon nanotubes, 2D material

### **1. Introduction**

Graphene can be obtained by a variety of techniques, including mechanical and chemical exfoliation of graphite, and chemical vapor deposition (CVD) methods, such as hot wall and hot filament CVD. Each technique has its own advantages and disadvantages that have to be weighed according to the intended application. The mechanical exfoliation of graphite produces high-quality monolayer graphene flakes [1, 2], whereas chemical exfoliation produces gra‐ phene flakes with significant structural defects [3]. Although the individual micron-scale

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

graphene flakes can be used to make miniature test devices that provide useful insight into the physical properties and potential applications of graphene [3], the reality is that graphene powders obtained by exfoliation methods are not useful for the mass production of reliable and reproducible electronic devices. However, hot-wall CVD appears to be the most straightfor‐ ward method to produce monolayer graphene films with low defect density, typically on Cu substrates [4]. Nonetheless, the graphene films obtained by this technique are polycrystalline and discontinuous [5]. **Table 1** compares the physical properties of graphene obtained by different methods.


**Table 1.** Properties of graphene films obtained by different methods.

An alternative approach to advance the potential applications of graphene is to focus the attention on bilayer graphene, which has a zero band gap like monolayer graphene. Further‐ more, the band gap of bilayer graphene has been shown to be tunable, and its domain walls have shown ballistic electron-conducting channels [6]. These exciting physical properties expand the potential applications of bilayer graphene beyond those of monolayer graphenes.

Mendoza et al. [7] reported that large-area bilayer graphene can be grown by hot filament chemical vapor deposition (HFCVD) with transmittance greater than 90% in the visible region and no gaps. The size of the graphene wafers that can be obtained by HFCVD is only limited by the size of the reactor and lends itself to be incorporated into a continuous roll-to-roll fabrication process, thus lowering the cost per unit area and enabling the integration of bilayer graphene in large-scale industrial production lines. The HFCVD method has the potential to enable large-area applications of graphene, such as transparent electrodes for flat panel displays and solar cells. In this chapter, we review the HFCVD state of the art for the synthesis of bilayer graphene and provide insight into how to tune the synthesis process to obtain optimum bilayer graphene films. We also discuss the HFCVD synthesis of diamond and carbon nanotubes.

### **2. The hot filament chemical vapor deposition reactor**

graphene flakes can be used to make miniature test devices that provide useful insight into the physical properties and potential applications of graphene [3], the reality is that graphene powders obtained by exfoliation methods are not useful for the mass production of reliable and reproducible electronic devices. However, hot-wall CVD appears to be the most straightfor‐ ward method to produce monolayer graphene films with low defect density, typically on Cu substrates [4]. Nonetheless, the graphene films obtained by this technique are polycrystalline and discontinuous [5]. **Table 1** compares the physical properties of graphene obtained by

94 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Sample size (mm) Charge carrier mobility**

**(cm2 V−1s−1)**

50 [5] 50 [5] 15,000 [14] 150–300 [15]

An alternative approach to advance the potential applications of graphene is to focus the attention on bilayer graphene, which has a zero band gap like monolayer graphene. Further‐ more, the band gap of bilayer graphene has been shown to be tunable, and its domain walls have shown ballistic electron-conducting channels [6]. These exciting physical properties expand the potential applications of bilayer graphene beyond those of monolayer graphenes.

Mendoza et al. [7] reported that large-area bilayer graphene can be grown by hot filament chemical vapor deposition (HFCVD) with transmittance greater than 90% in the visible region and no gaps. The size of the graphene wafers that can be obtained by HFCVD is only limited by the size of the reactor and lends itself to be incorporated into a continuous roll-to-roll fabrication process, thus lowering the cost per unit area and enabling the integration of bilayer

~1 [5] ~1 [8] (2L) >104

overlapping flakes [3]

overlapping flakes [3]

CVD [9] 0.5 [5] ~5 [10] 5000 [9] 900 [11]

≤0.1 [3] Large area of

~100 [3] Large area of

**Table 1.** Properties of graphene films obtained by different methods.

**(room temperature)**

100 (for a layer of overlapping flakes) [5]

200–300 (for a layer of overlapping flakes) [8]

30 [7] ~2500 [This work] ~800 [This work]

 (at low temperature) [9] **Sheet resistance (Ω/sqr) / 2L**

30 kΩ Drop-cast film [3]

8–150 kΩ Multilayer Langmuir-Blodgett

102 [9]

film [3]

~650 [16]

different methods.

Mechanical exfoliation [1, 2]

Chemical exfoliation [3]

Chemical exfoliation through graphene oxide [3]

Epitaxial growth (SiC) [12, 13]

HFCVD [7] ≤0.1 [7]

0.1–0.2 [16]

**Method Crystallite size (μm)**

The hot filament chemical vapor deposition (HFCVD) technique relies on a heated coiled wire to decompose the precursor reactants present in the gas mixture and deposit a film on the substrate surface kept near the filament at lower temperatures. HFCVD was introduced in 1979 for the synthesis of amorphous silicon films from silane gas at low substrate temperatures and high deposition rate [17]. For the synthesis of carbon materials (e.g., diamond, carbon nanotubes, graphene), a gas mixture consisting of a hydrocarbon (e.g., methane, acetylene) diluted in hydrogen is decomposed utilizing a refractory metal filament (such as tungsten, tantalum, or rhenium) that is resistively heated and maintained to 1800–2300°C, as measured with a dual wavelength pyrometer. Heterogeneous reactions at the hot filament surface and the emissive properties of the incandescent filament [18–21] initiate the decomposition of the hydrocarbon gases and the subsequent cascade of chemical vapor reactions. The filament temperature plays an essential role and influences the type and quality of the carbon materials that are grown. For example, at filament temperatures below 1800°C, little or no diamond film is obtained. Higher filament temperatures up to 2300°C lead to higher growth rates and higher quality polycrystalline diamonds. However, the higher the filament temperature, the faster the filaments become carburized. The metal carbides are brittle, resulting in relatively short

**Figure 1.** Internal details of the HFCVD setup for large-area graphene film deposition.

filament lifetimes and unsteady growth conditions. For stability and reproducibility, the filament material of choice is rhenium, which carburizes very slowly and lasts many cycles before requiring replacement.

For a 5-cm wafer, the HFCVD reactor typically consists of a 20-liter six-way cross stainless steel chamber. The substrate holder sits on the top of the substrate heater at the center of the chamber. The filament assembly is kept 5–10 mm above the substrate holder. **Figure 1** shows an actual picture of the main internal components in an HFCVD system. The filament wire diameter, length, and geometry must be optimized according to the intended film size, as well as the filament temperature and gas pressure required by the specific chemical vapor deposi‐ tion reaction. It is necessary to assess the filament material's resistivity as a function of temperature and pressure. For example, the empirical resistivity of rhenium is approximately 2 × 10−6 Ω⋅m at 2000°C and 3 × 103 Pa. The filament is held fixed by two parallel electrodes made of a refractory metal, typically molybdenum, to avoid evaporation of the electrode material onto the growing film. Coil-shaped filaments are often used because they enhance the total surface area available for initiating the heterogeneous decomposition reactions, but they tend to bend downward over time during film growth causing unsteady growth condi‐ tions. A filament arrangement consisting of multiple parallel stretched filaments of 0.25–0.50 mm diameter, as shown in **Figure 1**, is usually a better choice because it enhances the uni‐ formity and total area of the resulting films while also providing considerable surface area for the heterogeneous decomposition reactions. However, the multiple filament arrangement requires very high-current power supplies.

The substrate temperature required for the synthesis of a particular material is determined by the residence time and mobility of reactant molecules and radicals that adsorb at the substrate surface. In the HFCVD reactor, the substrate temperature is intrinsically affected by the filament temperature and power, filament-to-substrate distance, and total pressure. Therefore, active substrate temperature control is needed in most cases in order to be able to optimize the quality of the films grown. While the filament temperature is dictated by the energy required to break the precursor gas molecules to initiate the chemical vapor reactions, the filament-tosubstrate distance and total pressure are determined by the mean free path of critical radicals that form at or near the filament and must travel to the substrate surface before they get scavenged through other chemical reactions that do not lead to film deposition.

Besides determining the required filament and substrate temperatures, filament-to-substrate distance, and total pressure, the choice of substrate material and gas composition are also critical parameters to obtain the intended film type and quality. For example, under the same conditions, diamond can grow on molybdenum but not on quartz. The right gas-phase chemical environment and substrate temperature are conditions necessary but not sufficient to grow the intended film. The heterogeneous reactions that take place at the substrate surface are critically important in any chemical vapor deposition process. Therefore, the appropriate substrate must be provided to adsorb the reactive chemical species from the gas phase and catalyze the film synthesis reactions. After the right substrate is provided, the relative abun‐ dance of reactive chemical species will determine the film quality and growth rate and is primarily determined by gas composition. Priming or preparation of the substrate surface is usually needed because it is covered with all sorts of adsorbates from the ambient environment; seeding may also be needed in some cases. The priming may be done before insertion in the HFCVD chamber or during the initial HFCVD process. For example, diamond growth on silicon must be preceded by diamond seeding prior to insertion in the HFCVD and by the removal of the native silicon oxide layer, which can be accomplished by attack of the hydrogen radicals in the HFCVD [Solid State Communications 116, 217, 2000]. In this case, there is an induction period during which the intended film does not grow until the substrate surface is ready to adsorb the reactive species. In the case of graphene grown on copper foil, the copper oxide layer is first removed by exposure to hydrogen at 1000°C.

### **3. Reports on HFCVD graphene**

filament lifetimes and unsteady growth conditions. For stability and reproducibility, the filament material of choice is rhenium, which carburizes very slowly and lasts many cycles

96 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

For a 5-cm wafer, the HFCVD reactor typically consists of a 20-liter six-way cross stainless steel chamber. The substrate holder sits on the top of the substrate heater at the center of the chamber. The filament assembly is kept 5–10 mm above the substrate holder. **Figure 1** shows an actual picture of the main internal components in an HFCVD system. The filament wire diameter, length, and geometry must be optimized according to the intended film size, as well as the filament temperature and gas pressure required by the specific chemical vapor deposi‐ tion reaction. It is necessary to assess the filament material's resistivity as a function of temperature and pressure. For example, the empirical resistivity of rhenium is approximately

made of a refractory metal, typically molybdenum, to avoid evaporation of the electrode material onto the growing film. Coil-shaped filaments are often used because they enhance the total surface area available for initiating the heterogeneous decomposition reactions, but they tend to bend downward over time during film growth causing unsteady growth condi‐ tions. A filament arrangement consisting of multiple parallel stretched filaments of 0.25–0.50 mm diameter, as shown in **Figure 1**, is usually a better choice because it enhances the uni‐ formity and total area of the resulting films while also providing considerable surface area for the heterogeneous decomposition reactions. However, the multiple filament arrangement

The substrate temperature required for the synthesis of a particular material is determined by the residence time and mobility of reactant molecules and radicals that adsorb at the substrate surface. In the HFCVD reactor, the substrate temperature is intrinsically affected by the filament temperature and power, filament-to-substrate distance, and total pressure. Therefore, active substrate temperature control is needed in most cases in order to be able to optimize the quality of the films grown. While the filament temperature is dictated by the energy required to break the precursor gas molecules to initiate the chemical vapor reactions, the filament-tosubstrate distance and total pressure are determined by the mean free path of critical radicals that form at or near the filament and must travel to the substrate surface before they get

Besides determining the required filament and substrate temperatures, filament-to-substrate distance, and total pressure, the choice of substrate material and gas composition are also critical parameters to obtain the intended film type and quality. For example, under the same conditions, diamond can grow on molybdenum but not on quartz. The right gas-phase chemical environment and substrate temperature are conditions necessary but not sufficient to grow the intended film. The heterogeneous reactions that take place at the substrate surface are critically important in any chemical vapor deposition process. Therefore, the appropriate substrate must be provided to adsorb the reactive chemical species from the gas phase and catalyze the film synthesis reactions. After the right substrate is provided, the relative abun‐ dance of reactive chemical species will determine the film quality and growth rate and is primarily determined by gas composition. Priming or preparation of the substrate surface is

scavenged through other chemical reactions that do not lead to film deposition.

Pa. The filament is held fixed by two parallel electrodes

before requiring replacement.

2 × 10−6 Ω⋅m at 2000°C and 3 × 103

requires very high-current power supplies.

Singh et al. [16] reported the growth of large sp2 domain size (117–279 nm) single, bilayer, and multilayer graphene films on thick Ni foils using HFCVD at a relatively low substrate temperature of 700°C and low pressure (<200 Pa) at relatively short deposition time (10–30 min). The strategy used in this work is based on the mechanism by which hydrocarbon molecules decompose at the metal surface and diffuse into the bulk. Carbon diffuses into the bulk of nickel while it is kept hot and precipitates on the nickel surface forming nonuniform and discontinuous graphene layers during the fast cooling phase. This work presents the HFCVD as the advantage to dissociate at high temperature on filaments (tungsten filament at 1800–2200°C) source gases while temperature on heater is relatively low (700°C) compared with other CVD process obtaining high-quality films. Accordingly, with the different charac‐ terization (i.e., Raman spectra and SAED-TEM) presented in his report, the graphene samples (1–6 layers) obtained on nickel shows good crystallinity and also low sheet resistance values (630 Ω/sqr).

**Figure 2.** (a) Schematic diagram of the HFCVD used, (b) Raman spectra of the graphene obtained, (c) picture of the graphene transferred on SiO2 substrate, and (d) Raman mapping of a selected area of the hydrogenated graphene. (Im‐ age adapted with permission from Kataria et al. [22]).

Using Cu instead of Ni and higher substrate temperature, Kataria et al. [22] reported the growth of relatively large discontinuous patches (~100 µm2 ) of high-quality monolayer graphene at 1000°C by HFCVD, see **Figure 2a** the schematic diagram about HFCVD used. This report is summarized in **Figure 2**. Through **Figure 2b**, we can see that the graphene obtained is monolayer, which is corroborated by Raman area map in **Figure 2c**. The Raman spectra in these samples showed an additional band termed 2D' peak (2946 cm−1) in combination with D (~1353 cm−1) and G (1622 cm−1) bands, indicating hydrogenated graphene. Kataria's group analyzed the possibility to remove hydrogen from the graphene samples by applying temperatures up to 600°C in inert atmosphere. These procedures and Raman spectroscopy examination confirm the desorption of hydrogen from the graphene samples.

Behura et al. [23] reported the synthesis of nonuniform large-domain graphene films on Cu substrates inside a HFCVD reactor [24, 25]. From the report, it is not clear that they actually used the filament during the synthesis process. The material obtained shows high number of defects in the film. This is observable through the large intensity of D-band accordingly with the Raman spectra. Also, the intensity ratio between the G and 2D bands is ~0.7, and the FWHM of the 2D band is ~55 cm−1 indicating the growth of few-layer graphene. This report is sum‐ marized in **Figure 3**.

**Figure 3.** (a) SEM micrograph of graphene films on Cu surface with optical photograph shown in the inset, (b) Raman spectrum of graphene film on Cu surface. (Image adapted with permission from Behura et al. [23]).

### **4. Obtaining uniform bilayer graphene by HFCVD**

The general process to grow graphene by HFCVD is represented in **Figure 4**. The first step is the removal of the copper oxide layer from the copper substrate by exposure to hydrogen at 1000°C. We carry out this step at 4.7 × 103 Pa for 30 min. The second step is the actual growth of graphene by adding methane to the gas mixture and heating the filament to 1800°C.

Hot Filament Chemical Vapor Deposition: Enabling the Scalable Synthesis of Bilayer Graphene and Other Carbon Materials 99

**Figure 4.** Performance of the temperature during the cleaning process and deposition time.

Using Cu instead of Ni and higher substrate temperature, Kataria et al. [22] reported the growth

98 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

1000°C by HFCVD, see **Figure 2a** the schematic diagram about HFCVD used. This report is summarized in **Figure 2**. Through **Figure 2b**, we can see that the graphene obtained is monolayer, which is corroborated by Raman area map in **Figure 2c**. The Raman spectra in these samples showed an additional band termed 2D' peak (2946 cm−1) in combination with D (~1353 cm−1) and G (1622 cm−1) bands, indicating hydrogenated graphene. Kataria's group analyzed the possibility to remove hydrogen from the graphene samples by applying temperatures up to 600°C in inert atmosphere. These procedures and Raman spectroscopy examination confirm

Behura et al. [23] reported the synthesis of nonuniform large-domain graphene films on Cu substrates inside a HFCVD reactor [24, 25]. From the report, it is not clear that they actually used the filament during the synthesis process. The material obtained shows high number of defects in the film. This is observable through the large intensity of D-band accordingly with the Raman spectra. Also, the intensity ratio between the G and 2D bands is ~0.7, and the FWHM of the 2D band is ~55 cm−1 indicating the growth of few-layer graphene. This report is sum‐

**Figure 3.** (a) SEM micrograph of graphene films on Cu surface with optical photograph shown in the inset, (b) Raman

The general process to grow graphene by HFCVD is represented in **Figure 4**. The first step is the removal of the copper oxide layer from the copper substrate by exposure to hydrogen at

of graphene by adding methane to the gas mixture and heating the filament to 1800°C.

Pa for 30 min. The second step is the actual growth

spectrum of graphene film on Cu surface. (Image adapted with permission from Behura et al. [23]).

**4. Obtaining uniform bilayer graphene by HFCVD**

1000°C. We carry out this step at 4.7 × 103

) of high-quality monolayer graphene at

of relatively large discontinuous patches (~100 µm2

the desorption of hydrogen from the graphene samples.

marized in **Figure 3**.

Mendoza et al. [7] performed a parametric study that included the variation of methane concentration, total pressure, and substrate temperature. The filament-to-substrate distance and filament temperature were kept at 5 mm and 1800°C, respectively. The methane concen‐ tration was varied in the 1–50% range in 5% steps, while keeping the pressure at 4.7 × 103 Pa and the substrate temperature at 850°C. The ratio of the intensity of the 2D band to that of the G band is used to determine the number of graphene layers, while the ratio of the intensity of the D band to that of the G is used to determine the degree of disorder in the trigonal carbon network.

Graphene growth took place only in the 15–25% methane concentration range. The pressure was then changed in the 2.7–4.7 × 103 Pa range, and the quality of the graphene films deterio‐ rated as indicated by the D band. Increasing the pressure to 5.3 × 103 Pa yielded amorphous carbon. The substrate temperature was then lowered from 850 to 750°C in steps of 50°C, resulting in lower graphene quality as indicated by the D band. At temperatures higher than 850°C, the graphene quality improves and monolayer is achieved, but it grows in discontin‐ uous patches. In HFCVD, there seems to be a fundamental trade-off between continuousdiscontinuous coverage and bilayer-monolayer graphene [26].

Based on the above-described results, we extended the deposition time from 30 to 100 min for the HFCVD graphene growth in the temperature range from 750 to 1000°C, while keeping the gas mixture of 15% methane in hydrogen at 4.7 × 103 Pa. As shown in **Figure 5**, multilayer graphene is obtained below 850°C, and turbostratic bilayer graphene is obtained between 850 and 1000°C. The most uniform bilayer graphene is obtained at 950°C; the graphene film grown at 1000°C is discontinuous.

**Figure 5.** Graphene film evolution as a function of substrate temperature.

We then carried out graphene synthesis experiments as a function of time, keeping the substrate temperature at 950°C and the other parameters constant as well. **Figure 6** shows the Raman spectra of the graphene films as a function of deposition time. The optimum bilayer graphene quality and continuous surface is obtained for the 100-min deposition time. As the

**Figure 6.** Effect of growth time on bilayer graphene synthesis at 950°C.

deposition time is decreased in 5-min interval down to 20 min, the film continuity is compromised, and gaps remain open on the substrate surface as seen under the optical microscope. The graphene that is grown for less than 10 min shows a very large D band, indicating very small crystallite sizes (<10 nm). Hence, it appears that 100 min allows enough time for the graphene crystallites to grow, defects to heal, and gaps to close. Above 100 min, multilayer graphene starts to dominate the film composition.

**Figure 7.** Comparison of sheet resistance values reported for graphene.

**Figure 5.** Graphene film evolution as a function of substrate temperature.

**Figure 6.** Effect of growth time on bilayer graphene synthesis at 950°C.

We then carried out graphene synthesis experiments as a function of time, keeping the substrate temperature at 950°C and the other parameters constant as well. **Figure 6** shows the Raman spectra of the graphene films as a function of deposition time. The optimum bilayer graphene quality and continuous surface is obtained for the 100-min deposition time. As the

100 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 8.** (a) Graphene on copper foil, (b) graphene transferred onto PET, and (c) graphene transfer process from Cu to SiO2/Si.

The bilayer graphene grown by HFCVD should also be compared with other graphene reports in terms of transmittance and sheet resistance (Ω/sqr) [27–31]. **Figure 7** shows the sheet resistance values for HFCVD and hot wall or thermal chemical vapor deposition (TCVD) graphene with visible transmittance in the 90–95% range. The HFCVD bilayer graphene films grown at 850°C show an average sheet resistance in the order of 3 × 103 Ω/, whereas those grown at 950°C show an average sheet resistance value in the order of 8 × 102 Ω/sqr. The latter value is close to values reported for monolayer graphene grown by TCVD techniques, as seen in **Figure 8**. Moreover, the indium tin oxide (ITO) transparent electrode that is deposited on glass for flat panel systems has sheet resistance values in the order of 1.5 × 103 Ω/sqr [32].

### **5. Graphene transfer**

For transparent electrode applications, the bilayer graphene grown on copper substrates (**Figure 8a**) must be transferred onto the transparent substrate of choice (e.g., PET, **Figure 8b**). The transfer process is another critical step that can preserve or damage the integrity of the graphene films. First, a polymer support is attached to the graphene film, such as polymethyl methacrylate (PMMA), polydimethylsiloxane (PDMS), polyethylene terephthalate (PET). The polymer support is designed to hold the graphene film when the underlying Cu foil is etched away in iron chloride (FeCl3) at 70°C, as shown in **Figure 8c**. Once the Cu is removed, the graphene on polymer support is thoroughly cleaned in HCl to remove residual FeCl3 and Fe3+. Then, the graphene on polymer support is rinsed with deionized water several times and let to dry. The graphene is then ready to be transferred onto any substrate according to the intended application. In the example shown in **Figure 8**, the HFCVD bilayer graphene film was transferred onto a SiO2/Si wafer by pressing them firmly together and slowly peeling off the polymer support. The integrity of resulting graphene film depends greatly on carefully carrying out the transfer process. The electrical properties of graphene are particularly sensitive to remnant FeCl3 impurities and cracks produced in the transfer process.

### **6. HFCVD diamond**

Microcrystalline and nanocrystalline diamond films are grown by HFCVD on Si, Mo, and other carbide-forming substrates. The substrates need to be seeded by polishing the surface with <0.1-µm synthetic diamond powder. Before the clean reactive gas mixture is introduced into the HFCVD chamber, it is evacuated to 10−3 Pa or lower. The gas flow of methane and hydrogen is controlled to obtain fixed methane concentrations of 0.3 and 2.0% for microcrystalline and nanocrystalline diamond, respectively. The gas mixture enters at a rate of 100 sccm and is activated by a heated Re filament positioned at 8 mm above the substrate. The choice of Re for the filament material has the advantage that it does not react with carbon and therefore is not consumed during the diamond reaction. The filament is resistively heated in the range of 2300– 2500°C, as measured by an optical pyrometer. The total gas pressure of the chamber is kept at approximately 2.7 ± 0.1 Pa. The substrate temperature can be varied in the 600–900°C range. The growth rate varies in the 0.1–0.5 µm/h range depending on the substrate temperature [33]. The SEM images in **Figure 9** illustrate the differences in morphology and grain size between microcrystalline and nanocrystalline diamond.

**Figure 9.** Micrographs of (a) nanocrystalline and (b) microcrystalline diamond.

### **7. HFCVD carbon nanotubes**

The bilayer graphene grown by HFCVD should also be compared with other graphene reports in terms of transmittance and sheet resistance (Ω/sqr) [27–31]. **Figure 7** shows the sheet resistance values for HFCVD and hot wall or thermal chemical vapor deposition (TCVD) graphene with visible transmittance in the 90–95% range. The HFCVD bilayer graphene films grown at 850°C show an average sheet resistance in the order of 3 × 103 Ω/, whereas those

value is close to values reported for monolayer graphene grown by TCVD techniques, as seen in **Figure 8**. Moreover, the indium tin oxide (ITO) transparent electrode that is deposited on

For transparent electrode applications, the bilayer graphene grown on copper substrates (**Figure 8a**) must be transferred onto the transparent substrate of choice (e.g., PET, **Figure 8b**). The transfer process is another critical step that can preserve or damage the integrity of the graphene films. First, a polymer support is attached to the graphene film, such as polymethyl methacrylate (PMMA), polydimethylsiloxane (PDMS), polyethylene terephthalate (PET). The polymer support is designed to hold the graphene film when the underlying Cu foil is etched away in iron chloride (FeCl3) at 70°C, as shown in **Figure 8c**. Once the Cu is removed, the graphene on polymer support is thoroughly cleaned in HCl to remove residual FeCl3 and Fe3+. Then, the graphene on polymer support is rinsed with deionized water several times and let to dry. The graphene is then ready to be transferred onto any substrate according to the intended application. In the example shown in **Figure 8**, the HFCVD bilayer graphene film was transferred onto a SiO2/Si wafer by pressing them firmly together and slowly peeling off the polymer support. The integrity of resulting graphene film depends greatly on carefully carrying out the transfer process. The electrical properties of graphene are particularly

sensitive to remnant FeCl3 impurities and cracks produced in the transfer process.

Microcrystalline and nanocrystalline diamond films are grown by HFCVD on Si, Mo, and other carbide-forming substrates. The substrates need to be seeded by polishing the surface with <0.1-µm synthetic diamond powder. Before the clean reactive gas mixture is introduced into the HFCVD chamber, it is evacuated to 10−3 Pa or lower. The gas flow of methane and hydrogen is controlled to obtain fixed methane concentrations of 0.3 and 2.0% for microcrystalline and nanocrystalline diamond, respectively. The gas mixture enters at a rate of 100 sccm and is activated by a heated Re filament positioned at 8 mm above the substrate. The choice of Re for the filament material has the advantage that it does not react with carbon and therefore is not consumed during the diamond reaction. The filament is resistively heated in the range of 2300– 2500°C, as measured by an optical pyrometer. The total gas pressure of the chamber is kept at approximately 2.7 ± 0.1 Pa. The substrate temperature can be varied in the 600–900°C range.

Ω/sqr. The latter

Ω/sqr [32].

grown at 950°C show an average sheet resistance value in the order of 8 × 102

102 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

glass for flat panel systems has sheet resistance values in the order of 1.5 × 103

**5. Graphene transfer**

**6. HFCVD diamond**

Carbon nanotubes (CNTs) are grown by HFCVD on Cu at a substrate temperature of 900°C using a concentration of 2.0% methane in hydrogen. The growth parameters are similar to those required to grow nanocrystalline diamond; the crucial difference is the choice of substrate material, which should act as catalyst for the growth of CNTs. Before the clean reactive gas mixture is introduced into the HFCVD chamber, it is evacuated to 10−3 Pa or lower. The gas mixture enters at a rate of 100 sccm and is activated by a heated Re filament positioned at 8 mm above the substrate. The filament is resistively heated in the range of 2300–2500 °C, as measured by an optical pyrometer. The total gas pressure of the chamber is kept at approx‐ imately 2.7 ± 0.1 Pa. A thick layer of carbon nanotubes is readily obtained in about 15 min [34, 35].

The SEM images (**Figure 10a** and **b**) show that the films are composed of entangled clusters of nanotubes or nanofiber structures with the catalyst material present in the form of spherical tips. A more detailed by TEM (**Figure 10c**) shows that the fibers are bamboo-like carbon nanotubes (BCNTs) with diameters ranging from 50–100 nm and variable lengths. They contain nanocavities that are stacked one over the other with closed walls consisting of a number of graphene layers.

**Figure 10.** Scanning electron microscopy of (a) HFCVD carbon nanotubes, (b) Bamboo-like carbon nanotubes (BCNTs), and (c) transmission electron microscopy of BCNTs.

### **8. Summary**

In this chapter, we discussed the HFCVD reactor design and the interplay between the reactor parameters, such as filament and substrate temperatures, filament-to-substrate distance, and total pressure. Special attention was given to the large-area synthesis of bilayer graphene on copper, but we also discussed the HFCVD synthesis of microcrystalline diamond, nanocrys‐ talline diamond, and carbon nanotubes. Large-area bilayer graphene grown by HFCVD has transmittance greater than 90% in the visible region and no gaps. The size of the graphene wafers that can be obtained by HFCVD is only limited by the size of the reactor and lends itself to be incorporated into a continuous roll-to-roll fabrication process, thus lowering the cost per unit area and enabling the integration of bilayer graphene in large-scale industrial production lines. The HFCVD method has the potential to enable large-area applications of graphene, such as transparent electrodes for flat panel displays and solar cells.

### **Acknowledgements**

This work is funded by the Institute for Functional Nanomaterials (NSF Grant No. 1002410) and in part by PR NASA EPSCoR (NNX15AK43A).

### **Author details**

Frank Mendoza1,2\*, Tej B. Limbu1,2, Brad R. Weiner1,3 and Gerardo Morell1,2

\*Address all correspondence to: fran.mendoza@upr.edu

1 Institute for Functional Nanomaterials, University of Puerto Rico, San Juan, USA


### **References**

**Figure 10.** Scanning electron microscopy of (a) HFCVD carbon nanotubes, (b) Bamboo-like carbon nanotubes (BCNTs),

104 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

In this chapter, we discussed the HFCVD reactor design and the interplay between the reactor parameters, such as filament and substrate temperatures, filament-to-substrate distance, and total pressure. Special attention was given to the large-area synthesis of bilayer graphene on copper, but we also discussed the HFCVD synthesis of microcrystalline diamond, nanocrys‐ talline diamond, and carbon nanotubes. Large-area bilayer graphene grown by HFCVD has transmittance greater than 90% in the visible region and no gaps. The size of the graphene wafers that can be obtained by HFCVD is only limited by the size of the reactor and lends itself to be incorporated into a continuous roll-to-roll fabrication process, thus lowering the cost per unit area and enabling the integration of bilayer graphene in large-scale industrial production lines. The HFCVD method has the potential to enable large-area applications of graphene, such

This work is funded by the Institute for Functional Nanomaterials (NSF Grant No. 1002410)

and (c) transmission electron microscopy of BCNTs.

as transparent electrodes for flat panel displays and solar cells.

Frank Mendoza1,2\*, Tej B. Limbu1,2, Brad R. Weiner1,3 and Gerardo Morell1,2

1 Institute for Functional Nanomaterials, University of Puerto Rico, San Juan, USA

2 Department of Physics, University of Puerto Rico at Rio Piedras, San Juan, USA

3 Department of Chemistry, University of Puerto Rico at Rio Piedras, San Juan, USA

and in part by PR NASA EPSCoR (NNX15AK43A).

\*Address all correspondence to: fran.mendoza@upr.edu

**8. Summary**

**Acknowledgements**

**Author details**


[28] Buijnsters JG, Shankar P, van Enckevort WJP, Schermer JJ, and Ter Meulen JJ. The Adhesion of Hot-Filament CVD Diamond Films on AISI Type 316 Austenitic Stainless Steel. Diamond Relat. Mater. 2004;13:848–857.

[14] Wu X, Hu Y, Ruan M, Madiomanana NK, Hankinson J, Sprinkle M, Berger C, and Heer WA. Half integer quantum Hall effect in high mobility single layer epitaxial graphene.

[15] Krupka J and Strupinski W. Measurements of the sheet resistance and conductivity of

[16] Singh M, Jha HS, and Agarwal P. Growth of Large Sp2 domain size single and multilayer graphene films at low substrate temperature using hot filament chemical vapor

[17] Wiesmann H, Ghosh AK, McMahon T, and Strongin M. J. a-Si:H produced by high-

[18] Dupuie JL and Gulari E. Hot Filament enhanced chemical vapor deposition of AIN thin

[19] Jackson MJ, editor. Microfabrication and nanofacturing. Microfabrication and Nano‐ manofacturing. 1st ed. Boca Raton, FL: International Standard Book. 9780824724313;

[20] Wei QP, Ashfold MNR, Mankelevich YA, Yu ZM, Liu PZ, and Ma L. Diamond growth on WC-Co substrates by hot filament chemical vapor deposition: Effect of filament-

[21] Ali M, Ürgen M, and Atta MA. Tantalum carbide films synthesized by hot filament chemical vapor deposition technique. Surf. Coat. Technol. 2012;206:2833–2838.

[22] Kataria S, Patsha A, Dhara S, Tyagi AK, and Barshilia HC. J. Raman, Raman imaging on high-quality graphene grown by hot-filament chemical vapor deposition. Spectrosc.

[23] Behura SK, Mukhopadhyay I, Jani O, Yang Q, and Hirose A. Synthesis and character‐ ization of graphene films by hot filament chemical vapor deposition. Proceedings of

[24] Yang Q., Xiao C., Chen W., Singh AK, Asai T., and Hirose A, Growth mechanism and orientation control of well-aligned carbon nanotubes. Diam. Relat. Mater. 2003;12:1482–

[25] Behura SK, Nayak S, Yang Q, Hirose A, Jani O. J. Chemical Vapor Deposited Few-Layer Graphene as an Electron Emitter. Nanosci. and Nanotechnol. 2016;16:287–295

[26] Hawaldar R, Merino P, Correia MR, Bdikin I, Grácio J, Méndez J, Martín-Gago J-A, and Singh MK. Large-area high-throughput synthesis of monolayer graphene sheet by Hot

[27] Chen S, Cai W, Piner RD, Suk JW, Wu Y, Ren Y, Kang J, and Ruoff R. Synthesis and Characterization of Large-Area Graphene and Graphite Films on Commercial Cu-Ni

Filament Thermal Chemical Vapor Deposition. Sci. Rep. 2012;2:682.

Alloy Foils. Nano. Lett. 2011;11:3519–3525.

substrate separation. Diamond Relat. Mater. 2011;20:641–650.

the 24th CANCAM Saskatoon; Saskatchewan, Canada. 2013.

thin epitaxial graphene and SiC films. Appl. Phys. Lett. 2010;96:082101.

106 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

temperature thermal decomposition of silane. Appl. Phys. 1979;50:3752.

Appl. Phys. Lett. 2009;95:223108.

deposition. Mat. Lett. 2014;126:249–252.

films. Appl. Phys. Lett. 1991;59(5):549–551.

2010. 401 p.

2012;43:1864–1867.

1487.


### *In Situ* **Observation of Chemical Vapour Deposition Using Langasite Crystal Microbalance**

Hitoshi Habuka

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/62389

### **Abstract**

A method of *in situ* observation using langasite crystal microbalance (LCM) is descri‐ bed for chemical vapour deposition (CVD). First, the frequency behaviour of the LCM is expressed using the equation having the optimized coefficients in a wide range of gasphase conditions for the CVD. Next, by the LCM frequency behaviour, the existence of surface chemical reactions in a CVD reactor is determined. Additionally, the LCM can determine the lowest temperature for initiating the film deposition. In the last part, the temperature change related to the film formation process is described.

**Keywords:** Chemical vapour deposition, *In situ* observation, Langasite crystal micro‐ balance, Surface reaction, Gas properties

### **1. Introduction**

Chemical vapour deposition (CVD) is currently a fascinating technology for producing thin films in various advanced industries [1]. The CVD process is a complicated one having transport phenomena linked with the gas-phase and surface chemical reactions. For clarifying and designing the CVD process, the phenomena in the CVD reactor should be understood and optimized. For this purpose, the computational fluid dynamics (CFD) has been significantly advanced [2] and is actually very useful. In contrast, an experimental approach is still not easy [3], because the sensors seriously suffer from thermal, mechanical and chemical damage by the reactive and high-temperature ambient.

The *in situ* monitoring technique by means of the langasite crystal microbalance (LCM) [4–6] can be an appropriate experimental solution, because the LCM can sensitively detect various

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

changes in the CVD reactor, such as heat, flow and film deposition [5, 6]. The information obtained by the LCM should be used for understanding the CVD phenomena. Here, the LCM frequency nonlinearly influenced by various parameters should be clarified. Additionally, a practical process for the measurement and analysis should be developed. In this chapter, the process to *in situ* observe the CVD is thus explained using the LCM.

In Section 1, the frequency behaviour of the LCM is discussed using ambient gas mixtures at atmospheric pressure and at various temperatures. In order to express the LCM frequency decrease with the increasing concentrations of various gases in ambient hydrogen, the LCM frequency difference between the gas mixture and the carrier gas is practically expressed by optimizing the coefficients accounting for the gas properties. In Section 2, a method for determining the existence of surface chemical reactions is explained. The parameter, *C*(*T*), of the equation is used for expressing the LCM frequency change in a hydrogen–monomethyl‐ silane system. In Section 3, the lowest temperature for initiating the film deposition is evaluated using trichlorosilane gas and boron trichloride gas. In Section 4, the way to obtain the tem‐ perature change related to the film formation in a trichlorosilane–hydrogen system is descri‐ bed. First, the time constants for the LCM frequency change due to the surface and gas-phase temperature change are determined. Next, the continuous LCM frequency decrease is assigned to the weight increase by the film formation in a steady state. Based on an evaluation of the difference in the LCM frequency between those with and without the heat change related to the film formation, the surface temperature decrease caused by changing the precursor concentration is obtained.

### **2. Langasite crystal microbalance frequency behaviour over wide gas-phase conditions for chemical vapour deposition**

In this section, a relationship between the LCM frequency and the gas properties is expressed [7] as a practical equation applicable to the CVD conditions using several gases at various temperatures and gas concentrations.

### **2.1. Experimental procedure**

**Figure 1** shows the horizontal cold wall CVD reactor containing the langasite (La3Ga5SiO14) crystal microbalance (LCM) [4, 8]. This reactor consists of a gas supply system, a rectangularshaped quartz chamber and five infrared lamps. Hydrogen, nitrogen, trichlorosilane, mono‐ methylsilane and boron trichloride gases are used. The carrier gas is hydrogen. The LCM (Halloran Electronics, Tokyo, Japan) having a fundamental frequency of 10 MHz is placed 5 mm above the silicon wafer surface. The LCM frequency decreases with the increasing temperature [5, 6]. The silicon wafer and the LCM are simultaneously heated by infrared light from halogen lamps through the quartz chamber.

A typical process is shown in **Figure 2**. First, the LCM is heated to 160–600°C in hydrogen at atmospheric pressure. After waiting until the LCM frequency becomes stable, various gases are introduced at atmospheric pressure into the reactor chamber. The total gas flow rate is adjusted to 1 slm.

**Figure 1.** Chemical vapour deposition reactor containing a langasite crystal microbalance.

**Figure 2.** Process for measuring the frequency of the langasite crystal microbalance under various conditions.

### **2.2. LCM frequency and fluid property**

changes in the CVD reactor, such as heat, flow and film deposition [5, 6]. The information obtained by the LCM should be used for understanding the CVD phenomena. Here, the LCM frequency nonlinearly influenced by various parameters should be clarified. Additionally, a practical process for the measurement and analysis should be developed. In this chapter, the

110 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

In Section 1, the frequency behaviour of the LCM is discussed using ambient gas mixtures at atmospheric pressure and at various temperatures. In order to express the LCM frequency decrease with the increasing concentrations of various gases in ambient hydrogen, the LCM frequency difference between the gas mixture and the carrier gas is practically expressed by optimizing the coefficients accounting for the gas properties. In Section 2, a method for determining the existence of surface chemical reactions is explained. The parameter, *C*(*T*), of the equation is used for expressing the LCM frequency change in a hydrogen–monomethyl‐ silane system. In Section 3, the lowest temperature for initiating the film deposition is evaluated using trichlorosilane gas and boron trichloride gas. In Section 4, the way to obtain the tem‐ perature change related to the film formation in a trichlorosilane–hydrogen system is descri‐ bed. First, the time constants for the LCM frequency change due to the surface and gas-phase temperature change are determined. Next, the continuous LCM frequency decrease is assigned to the weight increase by the film formation in a steady state. Based on an evaluation of the difference in the LCM frequency between those with and without the heat change related to the film formation, the surface temperature decrease caused by changing the precursor

**2. Langasite crystal microbalance frequency behaviour over wide gas-phase**

In this section, a relationship between the LCM frequency and the gas properties is expressed [7] as a practical equation applicable to the CVD conditions using several gases at various

**Figure 1** shows the horizontal cold wall CVD reactor containing the langasite (La3Ga5SiO14) crystal microbalance (LCM) [4, 8]. This reactor consists of a gas supply system, a rectangularshaped quartz chamber and five infrared lamps. Hydrogen, nitrogen, trichlorosilane, mono‐ methylsilane and boron trichloride gases are used. The carrier gas is hydrogen. The LCM (Halloran Electronics, Tokyo, Japan) having a fundamental frequency of 10 MHz is placed 5 mm above the silicon wafer surface. The LCM frequency decreases with the increasing temperature [5, 6]. The silicon wafer and the LCM are simultaneously heated by infrared light

A typical process is shown in **Figure 2**. First, the LCM is heated to 160–600°C in hydrogen at atmospheric pressure. After waiting until the LCM frequency becomes stable, various gases

process to *in situ* observe the CVD is thus explained using the LCM.

concentration is obtained.

**conditions for chemical vapour deposition**

from halogen lamps through the quartz chamber.

temperatures and gas concentrations.

**2.1. Experimental procedure**

The relationship between the LCM frequency and the gas properties is described, accounting for the various relationships shown in **Figure 3**. Following a previous paper [9], the LCM frequency change, Δ*f*Gas, at a fixed temperature, *T* (K), is expressed using the product of the gas density, *ρ*gas, and gas viscosity, *η*Gas as follows:

$$
\Delta \! \! f\_{\rm Gas} = -C(T) \rho\_{\rm Ga}^{\vee} \eta\_{\rm Ga}^{\natural}, \tag{1}
$$

$$C(T) = \left(\frac{f\_c^{\times}}{\pi \rho^{\times}\_c \mu^{\times}\_c}\right) \tag{2}$$

where *f*c is the resonant frequency of the fundamental mode of the langasite crystal. *ρ*c and *μ*<sup>c</sup> are the density and shear modulus, respectively, of the langasite crystal. Based on Kanazawa et al. [9], the *x*, *y* and *z* values are 1.5, 0.5 and 0.5, respectively. In a vacuum, the Δ*f* value becomes zero.

The LCM frequency change from the vacuum condition to the hydrogen and to the gas mixtures are shown in **Figure 3** and are expressed by Eqs. (3) and (4).

$$
\Delta \mathcal{J}\_{\mathcal{U}\_2} = -C\left(T\right) \rho\_{\mathcal{U}\_2}^{\vee} \eta\_{\mathcal{U}\_2}^{\sharp} \quad \text{(in hydrogen)}.\tag{3}
$$

$$
\Delta \mathcal{J}\_{\text{Mix}} = -C \left( T \right) \rho\_{\text{Mix}}^{\text{y}} \eta\_{\text{Mix}}^{\text{z}} \quad \text{(in gas mixture)}.\tag{4}
$$

Although the LCM frequency in the vacuum, as a initial value, should be determined, experiment by experiment, the LCM frequency measurement in the vacuum is often not easy in the atmospheric and low-pressure CVD system. Thus, the accurate measurement of the LCM frequency difference from the vacuum condition, such as *<sup>Δ</sup> <sup>f</sup>* H2 and *Δf*Mix, is difficult.

**Figure 3.** Frequency of langasite crystal microbalance for hydrogen and gas mixtures.

In contrast, the LCM frequency change at the CVD condition from the carrier gas ambient, such as hydrogen ambient, is easily measured by simple operation. The LCM frequency change from that in the hydrogen (100%), Δ*f*Meas, is expressed by Eq. (5).

$$
\Delta \mathbf{f}\_{\text{Mens}} = \Delta \mathbf{f}\_{\text{Mix}} - \Delta \mathbf{f}\_{\text{H}\_2} \,. \tag{5}
$$

Following this, the Δ*f*Meas value can be accurately determined, while the mass change of the LCM occurs by means of the multiple use of the LCM.

By evaluating the ratio of the measured LCM frequency in the gas mixture to that in hydrogen, Δ*f*Meas is shown to have a relationship with the properties as described by Eq. (6).

$$
\Lambda \mathbf{f}\_{\text{Mess}} = \frac{\Lambda \mathbf{f}\_{\text{u}\_2}}{\rho\_{\text{u}\_2}^{\text{y}} \boldsymbol{\eta}\_{\text{u}\_2}^{\text{z}}} \, \rho\_{\text{Mir}}^{\text{y}} \boldsymbol{\eta}\_{\text{Mir}}^{\text{z}} - \Lambda \boldsymbol{f}\_{\text{H}\_2} \,. \tag{6}
$$

The *<sup>Δ</sup> <sup>f</sup>* H2 value can be sufficiently smaller than the first term in Eq. (6).

$$
\ln \Delta f\_{\text{Mes}} = y \ln \rho\_{\text{Mix}} + z \ln \eta\_{\text{Mix}} + A,\tag{7}
$$

where *A* is a constant. Using Eq. (7) accounting for various LCM frequencies and the fluid properties, the *y* and *z* values can be optimized.

In Eq. (7), the gas density and the gas viscosity are obtained following the ideal gas law and the Chapman–Enskog equation [10], respectively. The viscosity of the gas mixture is calculated following Pollard and Newman [11].

### **2.3. Influenced of gas density and viscosity**

Gas () , *Gas Gas y z* D =- *f CT* r h

*C C*

where *f*c is the resonant frequency of the fundamental mode of the langasite crystal. *ρ*c and *μ*<sup>c</sup> are the density and shear modulus, respectively, of the langasite crystal. Based on Kanazawa et al. [9], the *x*, *y* and *z* values are 1.5, 0.5 and 0.5, respectively. In a vacuum, the Δ*f* value becomes

The LCM frequency change from the vacuum condition to the hydrogen and to the gas

( ) H2 H H 2 2 ( gen in hydr )o . *y z* D =- *f CT* r h

Mix ( ) Mix Mix ( re). in gas mixtu *y z* D =- *f CT* r h

Although the LCM frequency in the vacuum, as a initial value, should be determined, experiment by experiment, the LCM frequency measurement in the vacuum is often not easy in the atmospheric and low-pressure CVD system. Thus, the accurate measurement of the LCM

pr m

*x C y z*

æ ö = ç ÷ ç ÷ è ø

( )

mixtures are shown in **Figure 3** and are expressed by Eqs. (3) and (4).

frequency difference from the vacuum condition, such as *<sup>Δ</sup> <sup>f</sup>* H2

**Figure 3.** Frequency of langasite crystal microbalance for hydrogen and gas mixtures.

zero.

*<sup>f</sup> C T*

112 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

(1)

(3)

(4)

and *Δf*Mix, is difficult.

(2)

**Figure 4** shows the Δ*f*Meas values obtained at 160, 400, 500 and 600°C using various gas mixtures, such as hydrogen–nitrogen, hydrogen–trichlorosilane, hydrogen–monomethylsilane. In this figure, the Δ*f*Meas value is plotted as a function of *ρ*<sup>y</sup>*μ*<sup>z</sup> assuming that both the y and z values are 0.5 following a previous study [9]. As shown in this figure, the Δ*f*Meas value follows a single trend and does not depend on the gas species. However, the behaviour shown in this figure is not linear.

Next, using Eq. (7), the *y* and *z* values are obtained. Using the measurement shown in **Figure 4**, both the *y* and *z* values are determined to be 1.3. As shown in **Figure 5**, the Δ*f*Meas value has a linear relationship with *ρ*1.3*μ*1.3. Similar to **Figure 4**, the LCM frequency change in **Figure 5** does not depend on the gas species.

Using Eqs. (1) and (2), the *C*(*T*) values at 160, 400, 500 and 600°C are obtained as shown in **Figure 6**. This figure shows that the *C*(*T*) value is expressed following the Arrhenius-type behaviour.

$$C\left(T\right) = \dots \times 10^{11} e^{\left(-2100/T\right)}\tag{8}$$

Following Eqs. (1), (2), (5) and (8) using the *y* and *z* values of 1.3, the Δ*f*Meas value is calculated and plotted versus the measurement from 160 to 600°C, as shown in **Figure 7**. This figure shows that the calculation could sufficiently reproduce the measurement.

**Figure 4.** The Δ*f*Meas value obtained for the gas mixture containing nitrogen, trichlorosilane and monomethylsilane in the hydrogen carrier gas at 160–600°C, plotted as a function of *ρ*0.5*μ*0.5.

**Figure 5.** The Δ*f*Meas value obtained for the gas mixture containing nitrogen, trichlorosilane and monomethylsilane in the hydrogen carrier gas at 160–600°C, plotted as a function of *ρ*1.3*μ*1.3.

*In Situ* Observation of Chemical Vapour Deposition Using Langasite Crystal Microbalance http://dx.doi.org/10.5772/62389 115

**Figure 6.** *C*(*T*) values changing with 1/*T*.

( ) 11 ( 2100/ ) 5.4 10 *<sup>T</sup> CT e* - = ´ (8)

Following Eqs. (1), (2), (5) and (8) using the *y* and *z* values of 1.3, the Δ*f*Meas value is calculated and plotted versus the measurement from 160 to 600°C, as shown in **Figure 7**. This figure shows

114 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 4.** The Δ*f*Meas value obtained for the gas mixture containing nitrogen, trichlorosilane and monomethylsilane in

**Figure 5.** The Δ*f*Meas value obtained for the gas mixture containing nitrogen, trichlorosilane and monomethylsilane in

that the calculation could sufficiently reproduce the measurement.

the hydrogen carrier gas at 160–600°C, plotted as a function of *ρ*0.5*μ*0.5.

the hydrogen carrier gas at 160–600°C, plotted as a function of *ρ*1.3*μ*1.3.

**Figure 7.** Correlation of the Δ*f*Meas values between the measurement and the calculation by Eqs. (1), (2), (5) and (8) using *y* and *z* of 1.3 at the temperatures of 160–600°C.

### **3. Method for determining chemical vapour deposition occurrence**

In this section, the method for determining the surface chemical reaction occurrence [12] is explained. The silicon carbide film formation from monomethylsilane gas is discussed.

### **3.1. LCM frequency**

The LCM frequency differences from ambient hydrogen in monomethylsilane–hydrogen system are shown in **Figure 8**. Immediately after opening the gas valve at 0 s, a high concen‐ tration of monomethylsilane gas, remained at 100% between the gas valve and the mass flow controller, is introduced into the reactor. During the high concentration gas passing through the reactor, the LCM frequency significantly decreases. Thereafter, the diluted monomethyl‐ silane gas reaches the LCM in the reactor. Thus, the fluctuation of the LCM frequency becomes quite small.

As shown in **Figure 8**, the LCM frequency difference measured at 300°C is very stable after 20 s. This indicates that the gas temperature and the fluid properties, such as the gas density and gas viscosity, are in a steady state. This simultaneously indicates that there is no chemical reaction at 300°C, due to no thermal change caused by no reaction heat. The behaviour at 400 and 500°C is similar to that at 300°C. The trend in the LCM frequency difference at 300, 400 and 500°C is the same and parallel to each other. Thus, the MMS-H2 system below 500°C is concluded to undergo no chemical reaction.

Although the LCM frequency difference at 550°C seems to be relatively stable, it very gradually increases after 20 s. Similar to this, the LCM frequency difference at 600°C also slightly increases with a larger gradient than that at 550°C. This gradual increase in the LCM frequency difference indicates that any transient change related to a chemical reaction, such as temperature, gas density and gas viscosity, occurred and continued in the reactor during the introduction of the monomethylsilane gas.

As shown in **Figure 8**, the LCM frequency increases with the decreasing temperature [13]. Additionally, the surface chemical reaction for the silicon carbide (SiC) film deposition from the monomethylsilane gas is endothermic [13]. The frequency decrease due to the weight

**Figure 8.** The LCM frequency change immediately after the introduction of the monomethylsilane gas to the hydrogen ambient atmosphere.

increase by the film deposition is overcompensated by the frequency increase due to the temperature decrease by the endothermic reaction. Thus, the LCM frequency continues to increase till reaching a steady state. From **Figure 8**, the chemical reaction occurring at the LCM surface is considered to continue after 200 s.

The temperature change should be enhanced by the increasing reaction rate, due to the greater reaction heat. In order to clearly show this trend, the LCM frequency gradient at various temperatures for the monomethylsilane–hydrogen system is evaluated, as shown in **Figure 9**. The LCM frequency gradient in the low-temperature range between 300–500°C is <0 Hz/s. This value is recognized to show the state with no chemicalreaction.In contrast, the LCM frequency gradient increases at 550°C from a value <0 to the positive value of 2 Hz/s. It further increases to a value >4 Hz/s at 600°C. Because the LCM frequency gradient increases with the increasing temperature, the surface chemical reaction at the LCM surface is initiated in the temperature range between 500 and 600°C.

**Figure 9.** LCM frequency gradient immediately after introducing the MMS gas at various temperatures.

### **3.2.** *C***(***T***) parameter**

**3.1. LCM frequency**

quite small.

monomethylsilane gas.

ambient atmosphere.

concluded to undergo no chemical reaction.

The LCM frequency differences from ambient hydrogen in monomethylsilane–hydrogen system are shown in **Figure 8**. Immediately after opening the gas valve at 0 s, a high concen‐ tration of monomethylsilane gas, remained at 100% between the gas valve and the mass flow controller, is introduced into the reactor. During the high concentration gas passing through the reactor, the LCM frequency significantly decreases. Thereafter, the diluted monomethyl‐ silane gas reaches the LCM in the reactor. Thus, the fluctuation of the LCM frequency becomes

116 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

As shown in **Figure 8**, the LCM frequency difference measured at 300°C is very stable after 20 s. This indicates that the gas temperature and the fluid properties, such as the gas density and gas viscosity, are in a steady state. This simultaneously indicates that there is no chemical reaction at 300°C, due to no thermal change caused by no reaction heat. The behaviour at 400 and 500°C is similar to that at 300°C. The trend in the LCM frequency difference at 300, 400 and 500°C is the same and parallel to each other. Thus, the MMS-H2 system below 500°C is

Although the LCM frequency difference at 550°C seems to be relatively stable, it very gradually increases after 20 s. Similar to this, the LCM frequency difference at 600°C also slightly increases with a larger gradient than that at 550°C. This gradual increase in the LCM frequency difference indicates that any transient change related to a chemical reaction, such as temperature, gas density and gas viscosity, occurred and continued in the reactor during the introduction of the

As shown in **Figure 8**, the LCM frequency increases with the decreasing temperature [13]. Additionally, the surface chemical reaction for the silicon carbide (SiC) film deposition from the monomethylsilane gas is endothermic [13]. The frequency decrease due to the weight

**Figure 8.** The LCM frequency change immediately after the introduction of the monomethylsilane gas to the hydrogen

**Figure 10** shows the LCM frequency change with the increase in *ρ*Mix 1.3*η*Mix 1.3 of the gas mixtures of monomethylsilane, nitrogen and hydrogen at 400, 500 and 600°C. In this figure, the frequency difference of the nitrogen–hydrogen system (white circles) follows Eq. (1) (*x* and *y* = 1.3) and shows a linear relationship. The monomethylsilane–hydrogen system (triangles) at 400 and 500°C coincides with the behaviour of the nitrogen–hydrogen system. In contrast, the mono‐ methylsilane–hydrogen system showed a slightly higher gradient in the low *ρ*1.3*η*1.3range. Thus, the *C*(*T*) value of Eq. (1) is expected to indicate the occurrence of a surface chemical reaction.

The *C*(*T*) values obtained using Eq. (1) are plotted versus 1/*T*. as shown in **Figure 11**. The *C*(*T*) values for the nitrogen–hydrogen system have a linear relationship following Arrhenius-type equation at temperatures from 300 to 600°C. In contrast, the monomethylsilane–hydrogen system shows an increase in the *C*(*T*) value at the higher temperatures.

In order to clearly recognize the difference, the *C*(*T*) value difference of the monomethylsilane– hydrogen system from the nitrogen–hydrogen system are shown in **Figure 12**. The *C*(*T*) value difference at temperatures lower than 500°C is around zero. Thus, the monomethylsilane gas at <500°C behaves the same as that of nitrogen. The monomethylsilane–hydrogen system shows an increased *C*(*T*) value to >1 × 1010 at temperatures between 500 and 550°C. Because these values are obtained separate from the fluid properties, the *C*(*T*) value change is under‐ stood as the surface chemical reaction for the CVD occurring at the higher temperatures. The obtained temperature is consistent with that from our previous study [6].

**Figure 10.** LCM frequency difference for monomethylsilane gas (dark triangle) and nitrogen gas (circle) from ambient hydrogen.

**Figure 11.** *C*(*T*) values of monomethylsilane–hydrogen system (triangles) and nitrogen–hydrogen system (circles).

**Figure 12.** *C*(*T*) value difference of MMS–H2 system from N2–H2 system.

### **4.** *In situ* **observation of chemical vapour deposition using SiHCl3 and BCl3 gases**

The film deposition behaviour using multiple precursors is explained [14]. The boron-doped silicon film is formed using the trichlorosilane (SiHCl3, TCS) gas and the boron trichloride (BCl3) gas for the film deposition and boron doping, respectively.

### **4.1. Chemical reaction occurrence**

The *C*(*T*) values obtained using Eq. (1) are plotted versus 1/*T*. as shown in **Figure 11**. The *C*(*T*) values for the nitrogen–hydrogen system have a linear relationship following Arrhenius-type equation at temperatures from 300 to 600°C. In contrast, the monomethylsilane–hydrogen

118 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

In order to clearly recognize the difference, the *C*(*T*) value difference of the monomethylsilane– hydrogen system from the nitrogen–hydrogen system are shown in **Figure 12**. The *C*(*T*) value difference at temperatures lower than 500°C is around zero. Thus, the monomethylsilane gas at <500°C behaves the same as that of nitrogen. The monomethylsilane–hydrogen system shows an increased *C*(*T*) value to >1 × 1010 at temperatures between 500 and 550°C. Because these values are obtained separate from the fluid properties, the *C*(*T*) value change is under‐ stood as the surface chemical reaction for the CVD occurring at the higher temperatures. The

**Figure 10.** LCM frequency difference for monomethylsilane gas (dark triangle) and nitrogen gas (circle) from ambient

**Figure 11.** *C*(*T*) values of monomethylsilane–hydrogen system (triangles) and nitrogen–hydrogen system (circles).

hydrogen.

system shows an increase in the *C*(*T*) value at the higher temperatures.

obtained temperature is consistent with that from our previous study [6].

The chemical reaction behaviour of boron trichloride gas is measured, as shown in **Figure 13**, at various boron trichloride concentrations and temperatures. At 400°C, the LCM frequency quickly decreases corresponding to the change in the fluid properties by an increase in the boron trichloride concentration. The LCM frequency is kept constant at each boron trichloride concentration. Thus, the boron trichloride gas does not have any chemical reaction and film deposition at this temperature. At 470°C, the LCM frequency sometimes shows flat and other times decrease. The film deposition occurrence is not obvious. In contrast, at 570°C, the LCM frequency decrease is clear at each boron trichloride concentration, as the typical behaviour showing the film deposition.

Similarly, the silicon film deposition from trichlorosilane gas is evaluated. Because the LCM frequency decrease occurs between 570 and 600°C, the silicon film deposition is determined to occur in this temperature range.

**Figure 13.** LCM frequency change at various boron trichloride concentrations and at 400, 470 and 570°C.

Next, the LCM frequency behaviour during the boron-doped silicon film deposition is observed at various temperatures using the trichlorosilane and boron trichloride gases, as shown in **Figure 14**. The LCM frequency shows a gradual increase and decrease in a short cycle at 330–500°C. In contrast, at 530 and 570°C, the LCM frequency continuously decreases.

**Figure 15** shows the LCM frequency gradient in the temperature range between 330 and 570°C. At temperatures lower than 470°C, the frequency gradient values are near 0 Hz s−1. At the temperatures higher than 500°C, the frequency gradient decreases to less than −3 Hz s−1. This behaviour indicates the occurrence of a continuous film deposition. Because the LCM fre‐ quency decreases for a long period, the film deposition is determined to occur in the temper‐ ature range between 470 and 530°C, specifically higher than 530°C.

**Figure 14.** LCM frequency change at various trichlorosilane and boron trichloride concentrations at 330–570°C.

**Figure 15.** LCM frequency gradient in the temperature range between 330 and 570°C. The trichlorosilane and boron trichloride gases were used for the film formation.

### **4.2. Growth rate**

**Figure 13.** LCM frequency change at various boron trichloride concentrations and at 400, 470 and 570°C.

120 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

ature range between 470 and 530°C, specifically higher than 530°C.

Next, the LCM frequency behaviour during the boron-doped silicon film deposition is observed at various temperatures using the trichlorosilane and boron trichloride gases, as shown in **Figure 14**. The LCM frequency shows a gradual increase and decrease in a short cycle at 330–500°C. In contrast, at 530 and 570°C, the LCM frequency continuously decreases.

**Figure 15** shows the LCM frequency gradient in the temperature range between 330 and 570°C. At temperatures lower than 470°C, the frequency gradient values are near 0 Hz s−1. At the temperatures higher than 500°C, the frequency gradient decreases to less than −3 Hz s−1. This behaviour indicates the occurrence of a continuous film deposition. Because the LCM fre‐ quency decreases for a long period, the film deposition is determined to occur in the temper‐

**Figure 14.** LCM frequency change at various trichlorosilane and boron trichloride concentrations at 330–570°C.

The film growth rate is shown in **Figure 16**, obtained following that the frequency decrease of 1 Hz corresponds to the weight increase of 6 ng cm−2 [9, 13, 15]. Here, the molecular weight and the density of the boron-doped silicon film are tentatively assumed to be an average of the silicon and boron.

The boron film growth rate at 570°C is near 1.5 × 10−9 mol cm−2 s−1 at the boron trichloride gas concentrations from 1 to 4%. Additionally, the silicon growth rate shows no obvious trend versus the trichlorosilane gas concentration from 1 to 4%. Similarly, the growth rate of the boron-doped silicon film formed from both the trichlorosilane and boron trichloride gases has no obvious trend, being about 1 × 10−9 mol cm−2 s−1. This growth rate behaviour is consistent with the saturation in the low-temperature silicon epitaxial growth process [16].

The change in the growth rate with the increasing temperature is shown in **Figure 17** as the Arrhenius plot. The boron growth rate is near 1 × 10−9 mol cm−2 s−1 and 1.5 × 10−9 mol cm−2 s−1 at 470 and 570°C, respectively. The silicon growth rate from trichlorosilane gas is about 5 × 10−10 mol cm−2 s−1 at temperatures lower than 530°C. It is too low value for determining the film deposition occurrence. Because the growth rate increases to nearly 10−9 mol cm−2 s−1 at 600°C, the silicon film deposition occurs at temperatures higher than 570°C.

At temperatures lower than 530°C, the film growth rate is about 5 × 10−10 mol cm−2 s−1. Because this value is similar to that of silicon, the film deposition is negligible. In contrast, the growth rate increases at temperatures higher than 530°C. It reaches 1 × 10−9 mol cm−2 s−1 at 570°C. The boron-doped silicon film is expected to be obtained at 570°C.

**Figure 16.** Growth rate at various trichlorosilane and boron trichloride gas concentrations. Silicon and boron film was grown using the combinations of trichlorosilane and boron trichloride gases of 2.4 and 1.0, 1.8 and 1.8, 1.2 and 2.8, and 0.6 and 4.0%.

**Figure 17.** Film growth rates changes of silicon, boron and boron-doped silicon versus 1/*T*.

#### **4.3. Film formation on substrate**

The boron-doped silicon film is formed at 570°C using the trichlorosilane gas and boron trichloride gas at 5 and 3%, respectively. The depth profile of the boron concentration is evaluated by secondary ion mass spectrometry (SIMS), as shown in **Figure 18**. While the boron concentration in the substrate is about 5 × 1016 atoms cm−3, it increases to that higher than 1020 atoms cm−3 near the film surface. Because the obtained film thickness is about 100 nm, the film growth rate is about 6–7 nm min−1. This value is comparable to 4 nm min−1 estimated from **Figures 16** and **17**. Thus, the growth rate obtained by the LCM is consistent with that by the film growth on the substrate.

**Figure 18.** Depth profile of boron concentration in the obtained silicon film.

### **4.4. Surface process**

**Figure 16.** Growth rate at various trichlorosilane and boron trichloride gas concentrations. Silicon and boron film was grown using the combinations of trichlorosilane and boron trichloride gases of 2.4 and 1.0, 1.8 and 1.8, 1.2 and 2.8, and

122 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 17.** Film growth rates changes of silicon, boron and boron-doped silicon versus 1/*T*.

The boron-doped silicon film is formed at 570°C using the trichlorosilane gas and boron trichloride gas at 5 and 3%, respectively. The depth profile of the boron concentration is evaluated by secondary ion mass spectrometry (SIMS), as shown in **Figure 18**. While the boron concentration in the substrate is about 5 × 1016 atoms cm−3, it increases to that higher than 1020 atoms cm−3 near the film surface. Because the obtained film thickness is about 100 nm, the film growth rate is about 6–7 nm min−1. This value is comparable to 4 nm min−1 estimated from

**4.3. Film formation on substrate**

0.6 and 4.0%.

The behaviours of film growth and doping are explained using Eqs. (9)–(12). The symbol '\*' indicates the chemisorbed state. Trichlorosilane is chemisorbed at the surface to produce \*SiCl2 and hydrogen chloride gas; \*SiCl2 is decomposed by hydrogen to form silicon at the surface [15]. Similarly, \*BCl2 is considered to be formed at the surface; it reacts with hydrogen to produce boron.

$$\text{SiHCl}\_3 \rightarrow \text{\*SiCl}\_2 + \text{HCl}\uparrow\tag{9}$$

$$\rm{^{4}SiCl}\_{2} + H\_{2} \rightarrow Si + 2HCl \uparrow \tag{10}$$

$$\text{BCl}\_3 + \frac{1}{2}\text{H}\_2 \rightarrow \text{\*} \text{BCl}\_2 + \text{HCl} \uparrow\tag{11}$$

$$\rm{^\*BCl}\_2 + H\_2 \rightarrow B + 2HCl \uparrow \tag{12}$$

For both gases, the intermediate species are chlorides which terminate the surface. The film growth rate at low temperatures is governed by the rates of the intermediate species adsorption and the chlorine removal. Thus, the growth rate of the boron-doped silicon is influenced by the slower process, that is by the silicon film growth rate. Because the LCM detects such details of the various behaviours, it can work as a sensitive monitor for studying the film growth behaviour.

### **5. Temperature change related to film formation process**

The surface temperature is one of the most important parameters for the CVD. In this section, the LCM is used in order to reveal and clarify the temperature behaviour related to the film formation [17]. For this purpose, the silicon film formation in a trichlorosilane–hydrogen system is used as one of the most popular systems.

### **5.1. LCM frequency behaviour**

The LCM frequency decreases corresponding to the weight increase on the LCM surface, following the Sauerbrey equation [15].

$$
\Delta f = -\frac{\Delta m \, f\_0^2}{A \left(\mu\_\text{c} \, \rho\_\text{c}\right)^{0.5}},
\tag{13}
$$

where *Λf* is the measured resonant frequency decrease, *f*0 is the intrinsic crystal frequency, *Δm* is the elastic mass change, *A* is the electrode area, *ρ*<sup>c</sup> is the density of the crystal, and *μc* is the shear modulus of the crystal.

Additionally, the LCM frequency depends on the fluid properties, as described by equation (14) [7, 8, 18].

$$
\Lambda f \propto \left(\rho \eta\right)^{\circ},
\tag{14}
$$

where *ρ* and *η* are the density and the viscosity, respectively, of the gas mixture. The *η* value is obtained from the literature [11]. The *x* value is 1.3 [7] in this section.

These parameters produce various changes in the LCM frequency during the CVD process. The change in the LCM frequency related to the film deposition by the trichlorosilane gas is classified by Parameters (*i*)–(*viii*) schematically shown in **Figure 19**.

Parameter (*i*) in **Figure 19** is the pressure and concentration of the gas remaining in the gas system. The LCM frequency very quickly and slightly increases due to the pressure increase in the reactor when opening the trichlorosilane gas valve. Immediately after this, the LCM frequency decreases and soon recovers by the trichlorosilane gas, which is remained at a high concentration in the gas line, reaching and passing the reactor. Parameter (*ii*) is the (*ρη*) 1.3 value of the gas mixture. Corresponding to this, the LCM frequency decreases from the initial frequency. Parameter (*iii*) is the heat capacity and the heat conduction. The increase in the heat capacity gradually causes the gas-phase temperature to decrease, resulting in an increase in the LCM frequency. Parameter (*iv*) is the reaction heat. The silicon film formation is an endothermic reaction [19]. This decreases the temperature and thus increases the LCM frequency. With the addition of the trichlorosilane gas, the film formation quickly begins and continues till termination of the trichlorosilane gas supply. Parameter (*v*) is the weight increase due to the film deposition. After a sufficient time for reaching a steady state, the weight increase by the film formation on the LCM appears as a continuous and linear decrease in the LCM frequency. Parameter (*vi*) is the (*ρη*)1.3 value. The decrease in this value immediately causes the LCM frequency to shift to a high value when the trichlorosilane gas supply is terminated. Parameter (*vii*) is the heat capacity and the heat conduction. Due to the heat capacity decrease because of the lack of trichlorosilane gas, the gas-phase temperature gradually increases and causes a decrease in the LCM frequency. Parameter (*viii*) is the weight increase due to the film deposition. After a sufficient period for achieving a steady state, the LCM frequency shift from that before the trichlorosilane gas supply corresponds to the increased weight of the film formed on the LCM.

the slower process, that is by the silicon film growth rate. Because the LCM detects such details of the various behaviours, it can work as a sensitive monitor for studying the film growth

124 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The surface temperature is one of the most important parameters for the CVD. In this section, the LCM is used in order to reveal and clarify the temperature behaviour related to the film formation [17]. For this purpose, the silicon film formation in a trichlorosilane–hydrogen

The LCM frequency decreases corresponding to the weight increase on the LCM surface,

( )

where *Λf* is the measured resonant frequency decrease, *f*0 is the intrinsic crystal frequency, *Δm* is the elastic mass change, *A* is the electrode area, *ρ*<sup>c</sup> is the density of the crystal, and *μc* is the

Additionally, the LCM frequency depends on the fluid properties, as described by equation

( ) , *x*

where *ρ* and *η* are the density and the viscosity, respectively, of the gas mixture. The *η* value

These parameters produce various changes in the LCM frequency during the CVD process. The change in the LCM frequency related to the film deposition by the trichlorosilane gas is

Parameter (*i*) in **Figure 19** is the pressure and concentration of the gas remaining in the gas system. The LCM frequency very quickly and slightly increases due to the pressure increase in the reactor when opening the trichlorosilane gas valve. Immediately after this, the LCM frequency decreases and soon recovers by the trichlorosilane gas, which is remained at a high

of the gas mixture. Corresponding to this, the LCM frequency decreases from the initial frequency. Parameter (*iii*) is the heat capacity and the heat conduction. The increase in the heat

concentration in the gas line, reaching and passing the reactor. Parameter (*ii*) is the (*ρη*)

rh

L µ*f*

is obtained from the literature [11]. The *x* value is 1.3 [7] in this section.

classified by Parameters (*i*)–(*viii*) schematically shown in **Figure 19**.

*c*

*m f <sup>f</sup> A* m r

2 0 0.5 c

,

<sup>D</sup> L =- (13)

(14)

1.3 value

**5. Temperature change related to film formation process**

system is used as one of the most popular systems.

**5.1. LCM frequency behaviour**

shear modulus of the crystal.

(14) [7, 8, 18].

following the Sauerbrey equation [15].

behaviour.

**Figure 19.** Schematic of changes in the frequency and the temperature of the LCM influenced by various parameters. Parameter (*i*): pressure and concentration of gas remaining in the gas system, parameter (*ii*): (*ρη*) 1.3, parameter (*iii*): heat capacity, parameter (*iv*): reaction heat, parameter (*v*): weight increase due to the film deposition, parameter (*vi*): (*ρη*)1.3, parameter (*vii*): heat capacity and parameter (*viii*): weight increase due to the film deposition.

If the temperature change due to the trichlorosilane gas did not exist, the temperature during the early stage of the film formation behaves like the thick dotted line, as shown in **Figure 19**. Thus, the LCM frequency difference between the solid line and the thick dotted line is considered to be a function of the temperature shift.

### **5.2. Reaction heat and heat transport**

By introducing the trichlorosilane gas into the reactor, the silicon film formation occurs along with the endothermic reaction heat and change of the physical properties of the gas mixture, as shown in **Figure 20**. Thus, the multiple thermal processes change the surface temperature. Here, the influence of each parameter is evaluated, following Steps 1, 2 and 3, as shown in **Figure 21**, taking into account the time constant for heat transport.

During Step1, the time constant for the surface temperature shift is evaluated without introducing a precursor in the ambient hydrogen. The quick lamp power decrease is assumed to show a significantly quick surface temperature decrease similar to that by the endothermic surface chemical reaction, as shown in **Figure 20**. After this, the gas-phase temperature of the near-surface region gradually decreases. These two processes are expected to have different time constants, such as very short and slightly long.

During Step 2, the influence of the change in the heat capacity and the heat conduction of the gas mixture are explained. The time constant for the gas-phase temperature shift induced by the precursor introduction, as shown in **Figure 20**, is evaluated at sufficiently low temperatures at which no chemical reaction occurs. The time constant for this process is expected to be longer than those for Step 1, because the gas-phase temperature shift occurs in the entire region of the reactor and not limited to the region near the surface.

During Step 3, the temperature change related to the film deposition is explained, accounting for the time constants obtained in Steps 1 and 2. During Step 3, the trichlorosilane concentration is stepwise changed from 0 to 3%. After the period corresponding to the time constants obtained in Steps 1 and 2, the LCM frequency behaviour expresses the film formation in a steady state. By extrapolation, the LCM frequency immediately after changing the precursor concentration is used for evaluating the temperature shift related to the film formation.

**Figure 20.** Parameters influencing the LCM temperature.


**Figure 21.** Three steps for evaluating temperature change during film deposition

### **5.3. Temperature and frequency**

**5.2. Reaction heat and heat transport**

By introducing the trichlorosilane gas into the reactor, the silicon film formation occurs along with the endothermic reaction heat and change of the physical properties of the gas mixture, as shown in **Figure 20**. Thus, the multiple thermal processes change the surface temperature. Here, the influence of each parameter is evaluated, following Steps 1, 2 and 3, as shown in

126 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

During Step1, the time constant for the surface temperature shift is evaluated without introducing a precursor in the ambient hydrogen. The quick lamp power decrease is assumed to show a significantly quick surface temperature decrease similar to that by the endothermic surface chemical reaction, as shown in **Figure 20**. After this, the gas-phase temperature of the near-surface region gradually decreases. These two processes are expected to have different

During Step 2, the influence of the change in the heat capacity and the heat conduction of the gas mixture are explained. The time constant for the gas-phase temperature shift induced by the precursor introduction, as shown in **Figure 20**, is evaluated at sufficiently low temperatures at which no chemical reaction occurs. The time constant for this process is expected to be longer than those for Step 1, because the gas-phase temperature shift occurs in the entire region of

During Step 3, the temperature change related to the film deposition is explained, accounting for the time constants obtained in Steps 1 and 2. During Step 3, the trichlorosilane concentration is stepwise changed from 0 to 3%. After the period corresponding to the time constants obtained in Steps 1 and 2, the LCM frequency behaviour expresses the film formation in a steady state. By extrapolation, the LCM frequency immediately after changing the precursor concentration is used for evaluating the temperature shift related to the film formation.

**Figure 21**, taking into account the time constant for heat transport.

time constants, such as very short and slightly long.

the reactor and not limited to the region near the surface.

**Figure 20.** Parameters influencing the LCM temperature.

The entire frequency dependence on the temperature is shown in **Figure 22**. This shows the frequency difference at various temperatures from that at room temperature. With the increasing temperature, the LCM frequency decreases, as shown in **Figure 22a**. At the higher temperatures, the LCM frequency more rapidly decreases than that at the lower temperatures. The temperature gradient is shown in **Figure 22b**. The gradient decreases with the increasing temperature. The frequency change due to the temperature change is about −400 Hz/K in the temperature range between 500 and 650°C.

**Figure 22.** LCM frequency changing with the temperature. (a) frequency change from that at room temperature and (b) frequency gradient (Hz/K) at various temperatures.

#### **5.4. Heat at surface**

In order to evaluate the LCM frequency behaviour caused by the quick surface temperature decrease, such as that by the reaction heat, the LCM frequency influenced by the stepwise lamp power shift is evaluated, as shown in **Figure 23**. The lamp power corresponding to about 1 K quickly decreases at 450 and 660°C.

**Figure 23.** LCM frequency change caused by the stepwise lamp power shift corresponding to about 1 K at (a) 450°C and (b) 660°C.

**Figure 23a** and **23b** shows the normalized LCM frequency change caused by the stepwise lamp power shift at 450 and 660°C, respectively. In both figures, the LCM frequency very quickly increases immediately after changing the lamp power. It then moderately increases and reaches the steady state.

These behaviours are expressed as a function of time, *t* (s), assuming that the fast and slow relaxation processes.

$$\Delta f = \left[ 230 \left( 1 - e^{\left( -\frac{t}{8} \right)} \right) + 90 \left( 1 - e^{\left( -\frac{t}{34} \right)} \right) \right]. \tag{15}$$

$$
\Delta \mathcal{J} = \left[ 370 \left( 1 - e^{\left( -\frac{l}{5} \right)} \right) + 80 \left( 1 - e^{\left( -\frac{l}{47} \right)} \right) \right]. \tag{16}
$$

In **Figure 23**, Eqs. (15) and (16) are indicated by the dotted lines. The quick temperature change, the first term, is considered to directly follow the lamp power decrease, that is the decrease in heat at the LCM surface. The slow temperature change, the second term, is due to the con‐ duction heat transport between the surface and the gas phase very near the surface. The influence of the reaction heat and heat conduction on the LCM frequency is expected to appear within 10 and 50 s, respectively, after introducing the trichlorosilane gas.

### **5.5. Heat transport around surface in gas phase**

power shift is evaluated, as shown in **Figure 23**. The lamp power corresponding to about 1 K

128 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 23.** LCM frequency change caused by the stepwise lamp power shift corresponding to about 1 K at (a) 450°C

**Figure 23a** and **23b** shows the normalized LCM frequency change caused by the stepwise lamp power shift at 450 and 660°C, respectively. In both figures, the LCM frequency very quickly increases immediately after changing the lamp power. It then moderately increases and

These behaviours are expressed as a function of time, *t* (s), assuming that the fast and slow

<sup>8</sup> <sup>34</sup> 230 1 90 1 . æ ö æ ö ç ÷ - -ç ÷ è ø è ø é ù æ öæ ö D= - + - ê ú ç ÷ç ÷ ë û è øè ø

<sup>5</sup> <sup>47</sup> 370 1 80 1 . æ ö æ ö ç ÷ - -ç ÷ è ø è ø é ù æ öæ ö D= - + - ê ú ç ÷ç ÷ ë û è øè ø

In **Figure 23**, Eqs. (15) and (16) are indicated by the dotted lines. The quick temperature change, the first term, is considered to directly follow the lamp power decrease, that is the decrease in heat at the LCM surface. The slow temperature change, the second term, is due to the con‐ duction heat transport between the surface and the gas phase very near the surface. The influence of the reaction heat and heat conduction on the LCM frequency is expected to appear

within 10 and 50 s, respectively, after introducing the trichlorosilane gas.

*t t*

*t t*

*fe e* (15)

*fe e* (16)

quickly decreases at 450 and 660°C.

and (b) 660°C.

reaches the steady state.

relaxation processes.

The LCM frequency changes due to the thermal properties, such as the heat capacity and the heat conductivity, are evaluated at the low temperatures, which do not cause the film depo‐ sition [14]. **Figure 24** shows that the LCM frequency change is due to the stepwise concentration change of hydrogen gas and trichlorosilane gas at 380°C. At conditions CA–CE, the hydrogen concentration and the trichlorosilane concentration decrease and increase, respectively.

At the beginning of condition CA, the LCM frequency quickly drops to less than −3000 Hz and recovers to −1000 Hz. The LCM frequency shift from 0 Hz to −1000 Hz corresponds to the increase in the gas density and the gas viscosity. Next, it gradually recovers to near −400 Hz. Till 500 s, the LCM frequency reaches the steady state. The gradual recovery of the LCM frequency is a result of the temperature decrease mainly due to the increase in the heat capacity of the gas mixture. Additionally, the temperature decrease at the surface is moderated by the heat balance with the gas phase *via* the heat conduction.

These heat transports overlaps and gives the gradual recovery of the LCM frequency. The LCM frequency at 380°C and at condition CD is expressed, as shown in **Figure 25**, using the following equation.

$$\Delta f = 164 \left( 1 - e^{\left( -\frac{l}{120} \right)} \right) . \tag{17}$$

From the temperature shift width shown in **Figure 25**, the temperature decrease by the trichlorosilane gas concentration change at 380°C is about 1 K. The time constant of these process is about 120 s. The time constant in Eq. (17) is longer than that of the second term in Eqs. (15) and (16). The time constant is assumed to have a similar value at the higher temper‐ atures, such as 600–700°C.

**Figure 24.** LCM frequency change caused by the stepwise concentration change in the hydrogen gas and trichlorosi‐ lane gas at 380°C.

**Figure 25.** LCM frequency change at 380°C caused by the gas concentrations of hydrogen and trichlorosilane from con‐ dition CC to CD in **Figure 24**.

#### **5.6. Film formation**

The silicon film is formed at 640°C along with measuring the LCM frequency, as shown in **Figures 26** and **27**. The hydrogen gas concentration decreases from 100 to 97%, while the trichlorosilane gas concentration increases from 0 to 3%. Because the silicon film growth rate is saturated at this temperature [16], the reaction heat remains the same for conditions CA–CE.

**Figure 26.** LCM frequency change caused by the stepwise change in the concentrations of hydrogen and trichlorosi‐ lane.

At condition CA, the trichlorosilane gas is added to the hydrogen gas. The LCM frequency shows a significant drop and a quick recovery within several seconds. After the quick recovery, the LCM frequency gradually increases. After the peak appearance, the LCM frequency gradually decreases accompanying the fluctuation due to a temperature fluctuation. Condi‐ tions CC, CD and CE show a similar LCM frequency behaviour without a significant drop, unlike that at the beginning of condition CA.

Taking into account the time constant corresponding to the various heat processes, the LCM frequency behaviour is evaluated, as shown in **Figure 27**. This shows the LCM behaviour at condition CD, as an example. Immediately after increasing the trichlorosilane concentration from 1.8 to 2.4%, the LCM frequency quickly decreases to about −550 Hz due to the increase in the gas density and the gas viscosity. Next, it shows a broad bottom for about 20 s. The LCM frequency gradually increases from 50 to 180 s. After showing a peak, the LCM frequency begins to decrease. This decrease is due to the film formation during the steady state.

Next, the LCM frequency gradient is evaluated. The maximum and minimum values are 1.6 Hz/s between Points A and B and 0.77 Hz/s between Points A and C, respectively. By this operation, the LCM frequency immediately after changing the trichlorosilane concentration is 245 to 410 Hz, as shown in **Figure 27**.

In addition, the flat bottom of the LCM frequency to 20 s after changing the trichlorosilane concentration may be caused by the balance among the changes in the gas density, the gas viscosity, the heat capacity and the heat conductivity. In order to obtain the possible minimum frequency value, the increasing trend in the LCM frequency between 20 and 180 s is extrapo‐ lated to that near several seconds. By this estimation, the frequency of about 50 Hz may be lower than that at the bottom. By adding these values, the LCM frequency change due to the temperature change by changing the trichlorosilane concentration from 1.8 to 2.4% is between 295 and 460 Hz. The surface temperature shift caused by changing the precursor concentration is considered to be about one degree.

**Figure 25.** LCM frequency change at 380°C caused by the gas concentrations of hydrogen and trichlorosilane from con‐

130 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The silicon film is formed at 640°C along with measuring the LCM frequency, as shown in **Figures 26** and **27**. The hydrogen gas concentration decreases from 100 to 97%, while the trichlorosilane gas concentration increases from 0 to 3%. Because the silicon film growth rate is saturated at this temperature [16], the reaction heat remains the same for conditions CA–CE.

**Figure 26.** LCM frequency change caused by the stepwise change in the concentrations of hydrogen and trichlorosi‐

At condition CA, the trichlorosilane gas is added to the hydrogen gas. The LCM frequency shows a significant drop and a quick recovery within several seconds. After the quick recovery, the LCM frequency gradually increases. After the peak appearance, the LCM frequency gradually decreases accompanying the fluctuation due to a temperature fluctuation. Condi‐ tions CC, CD and CE show a similar LCM frequency behaviour without a significant drop, unlike

dition CC to CD in **Figure 24**.

**5.6. Film formation**

lane.

that at the beginning of condition CA.

**Figure 27.** Evaluation of frequency change caused by the stepwise change from condition CC to CD in **Figure 26**.

### **Acknowledgements**

The author would like to thank Ms. Misako Matsui, Ms. Ayumi Saito and Mr. Kento Miyazaki of Yokohama National University for their intensive work. The author would like to thank Mr. Nobuyoshi Enomoto and Mr. Hitoshi Ueno of Halloran Electronics Co., Ltd., for his technical support. A part of this study was supported by JSPS KAKENHI Grant No. 25420772.

### **Author details**

Hitoshi Habuka\*

Address all correspondence to: habuka1@ynu.ac.jp

Department of Chemical Energy Engineering, Yokohama National University, Yokohama, Japan

### **References**


[8] H. Fritze, "High-temperature piezoelectric crystals and devices", J. Electroceram., 26 (2011) 122–161. doi:10.1007/s10832-011-9639-6

**Acknowledgements**

**Author details**

Hitoshi Habuka\*

Japan

**References**

2008).

Address all correspondence to: habuka1@ynu.ac.jp

The author would like to thank Ms. Misako Matsui, Ms. Ayumi Saito and Mr. Kento Miyazaki of Yokohama National University for their intensive work. The author would like to thank Mr. Nobuyoshi Enomoto and Mr. Hitoshi Ueno of Halloran Electronics Co., Ltd., for his technical

support. A part of this study was supported by JSPS KAKENHI Grant No. 25420772.

132 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Department of Chemical Energy Engineering, Yokohama National University, Yokohama,

[1] S. M. Sze, Semiconductor devices: physics and technology (John-Wiley, Hoboken, NJ,

[2] A. Bouteville, "Numerical simulation applied to chemical vapor deposition process, rapid thermal CVD and spray CVD", J. Optoelectron. Adv. Mater., 7 (2005) 599–606.

[3] U. Koehler, L. Andersohn and B. Dahlheimeer, "Time-resolved observation of CVD-

[4] M. Schulz, J. Sauerwald, H. Seh, H. Fritze and H. L. Tuller, "Defect chemistry based design of monolithic langasite structures for high temperature sensors", Solid State

[5] H. Habuka and Y. Tanaka, "In-situ monitoring of chemical vapor deposition from trichlorosilane gas and monomethylsilane gas using langasite crystal microbalance",

[6] H. Habuka and Y. Tanaka, "Langasite crystal microbalance frequency behavior over wide gas phase conditions for chemical vapor deposition", J. Surf. Eng. Mater. Adv.

[7] H. Habuka and M. Matsui, "Langasite crystal microbalance frequency behavior over wide gas phase conditions for chemical vapor deposition", Surf. Coat. Technol., 230

ECS J. Solid State Sci. Technol., 1 (2012) 62–65. doi:10.4236/jsemat.2013.31A009

growth of silicon on Si (111) with STM", Appl. Phys. A, 57 (1993) 491–497.

Ion., 184 (2011) 78–82. doi:10.1016/j.ssi.2010.08.009

Technol., 3 (2013) 61–66. doi:10.1016/j.surfcoat.2013.06.052

(2013) 312–315. doi:10.1016/j.surfcoat.2013.06.052


**CVD: Applications in Optical, Solar Cell and Solid State Devices**

### **Low‐Temperature PureB CVD Technology for CMOS Compatible Photodetectors**

Vahid Mohammadi and Stoyan Nihtianov

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/63344

### **Abstract**

In this chapter, a new technology for low‐temperature (LT, 400°C) boron deposition is developed, which provides a smooth, uniform, closed LT boron layer. This technolo‐ gy is successfully employed to create near‐ideal LT PureB (pure boron) diodes with low, deep junction‐like saturation currents, allowing full integration of LT PureB photodio‐ des with electronic interface circuits and other sensors on a single chip. In this way, smart‐sensor systems or even charge‐coupled device (CCD) or complementary metal oxide semiconductor (CMOS) ultraviolet (UV) imagers can be realised.

**Keywords:** low‐temperature boron deposition, ultrashallow p<sup>+</sup> n junction photodiode, chemical vapour deposition, UV photodetector, CMOS imager

### **1. Introduction**

Over the last few years, we have witnessed an increase in the demand for both high‐perform‐ ance ultraviolet and low‐energy electron detectors. These detectors are used in high‐tech applications such as optical lithography and electron microscopy, medical imaging, protein analysis and DNA sequencing, forensic analysis, disinfection and decontamination, space observation, etc.

To meet these demands, a new silicon detector technology has been proposed by TU Delft called "PureB" (pure boron) technology [1]. Conventional PureB technology is employed to fabricate high‐performance photodiodes for vacuum/extreme/deep‐UV (VUV/EUV/DUV) light, low‐energy electrons down to 200 eV, and X‐ray drift detectors [2–5]. A comparison of

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PureB technology to state‐of‐the‐art Si‐based UV and electron detectors is presented in **Tables 1** and **2**, respectively.


a SXUV: Si‐based n‐on‐p junction photodiodes with nitride metal silicide front window.

b Data are based on as‐deposited PureB photodetectors without any post‐processing thermal annealing.

c Value from measurement at PTB [2, 3].

dStability proven only in the VUV range [6].

e No more information is given in the reference.

f Δ: reduction in responsivity based on the data reported in the mentioned references.

**Table 1.** Performance overview of representative commercially available Si‐based UV detectors compared to the PureB UV photodetector.


a Electron signal gain, *G*ph, is defined as *I*ph/*I*beam, where *I*ph is the output current of the photodiode and *I*beam is the current of the incident electron beam, assuming a negligible dark current.

**Table 2.** Performance overview of representative commercially available, research‐reported Si‐based electron detectors compared to the PureB electron detector.

However, due to the relatively high temperature (HT) of boron deposition (500–700°C), it is hard to integrate this technology fully into a standard integrated circuit/ complementary metal oxide semiconductor (IC/CMOS) process flow.

In this chapter, a newly developed technology for low‐temperature (LT, 400°C) boron chemical vapour deposition (CVD) is presented, providing a smooth, uniform, closed LT boron layer. This technology is successfully employed to create near‐ideal LT PureB diodes with low, deep‐ junction‐like saturation currents. The low‐temperature deposition at 400°C makes it possible to integrate the LT PureB photodiodes fully with electronic interface circuits and other sensors on a single chip. In this way, smart‐sensor systems or even charge‐coupled device (CCD) or CMOS ultraviolet (UV) imagers can be realised.

What this chapter demonstrates is near ideal1 , highly sensitive, radiation‐hard, low leakage current diode characteristics. Moreover, since the boron deposition is conformal and highly selective to Si, PureB technology is shown to be an attractive candidate for creating junctions on silicon nanowires and advanced CMOS transistors including a source/drain in p‐type FinFETs. The impressive properties that PureB devices provide have resulted in this technol‐ ogy becoming commercially available very quickly for ASML EUV lithography tools and scanning electron microscopy (SEM) systems.

### **2. An analytical kinetic model for boron CVD**

In references [14] and [15], the pattern dependency and the loading effect of CVD boron deposition are identified as sources of non‐uniformity of the boron layer. Since investigating these two effects requires a thickness‐monitoring technique, an end‐of‐line resistance meas‐ urement is introduced as a non‐destructive, accurate means of monitoring the boron layer uniformity with fine resolution [16].

**Figure 1.** Temperature distribution profile simulated by commercial FLUENT© software for an ASM Epsilon One CVD reactor when the total pressure is ATM. The hydrogen gas is considered to be the main gas flowing over the susceptor, in order to simplify the simulation. The susceptor is heated up by an assembly of lamps to the deposition temperature (here 700°C) [19].

PureB technology to state‐of‐the‐art Si‐based UV and electron detectors is presented in **Tables 1**

138 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**IRD SXUV***<sup>a</sup>*

Junction type Schottky n‐on‐p p‐on‐n p‐on‐n

References [6, 7] [6, 8–10] [9, 10] [6, 8] [6] [2, 3]

**Table 1.** Performance overview of representative commercially available Si‐based UV detectors compared to the PureB

**Detector Commercial BSE Commercial vCD Delta doped e2V CCD 97 PureB BSE**

Electron signal gain, *G*ph, is defined as *I*ph/*I*beam, where *I*ph is the output current of the photodiode and *I*beam is the current

**Table 2.** Performance overview of representative commercially available, research‐reported Si‐based electron detectors

Electron signal gain at 500 eV ∼18 ∼52 ∼47 102 Electron signal gain at 200 eV Unknown Unknown ∼13 34 Stability Δ ∼ 5% Δ ∼ 5% Unknown Δ ∼ 4% References [11, 12] [11, 12] [13] [4]

at 1 KeV ∼46 ∼126 ∼112 213

Data are based on as‐deposited PureB photodetectors without any post‐processing thermal annealing.

SXUV: Si‐based n‐on‐p junction photodiodes with nitride metal silicide front window.

Δ: reduction in responsivity based on the data reported in the mentioned references.

**IRD UVG**

∼0.2*<sup>c</sup>* ∼0.265 ∼0.23 Unknown Unknown 0.267

Unknown*<sup>d</sup>* Not stable Good*<sup>e</sup>* Unknown Unknown Δ ∼ 3%*<sup>f</sup>*

∼0.03 ∼0.1 ∼0.01 ∼0.137 ∼0.1 0.102

∼0.02 ∼0.1 <0.01 Unknown Unknown 0.123

≤0.02 Unknown <0.01 <0.05 Unknown 0.116

Δ ∼ 2% Not stable Unknown Δ ∼ 10% Not stable Δ ∼ 4%

**Hamamatsu (S5226)**

**PureB***<sup>b</sup>*

and **2**, respectively.

EUV sensitivity (A/W) at 13.5 nm

Stability under

VUV sensitivity (A/W) at 193 nm

VUV sensitivity (A/W)at 157 nm

VUV sensitivity (A/W)at 121 nm

Stability under DUV/VUV

UV photodetector.

Electron signal gain*<sup>a</sup>*

Value from measurement at PTB [2, 3]. dStability proven only in the VUV range [6].

No more information is given in the reference.

of the incident electron beam, assuming a negligible dark current.

compared to the PureB electron detector.

a

b

c

e

f

a

EUV

**Detector ETH**

**PtSi‐nSi**

**IRD AXUV**

<sup>1</sup> Diode ideality factor, n≈1

In this section, an analytical model is established to describe the deposition kinetics and the deposition chamber characteristics that determine the deposition rate (DR) over the wafer. This pre‐deposition prediction tool can be used to improve the set‐up and control of the final deposition [17, 18]. It is also very useful for transferring recipes from one reactor to the other.

Some considerations such as the diffusion behaviour of the reactant (here BH3 species) through the stationary boundary layer over the wafer with the mechanism of the diffusion, the gas‐ phase processes and the related surface reactions are taken into account by this model. To develop the model, the actual parabolic gas velocity and temperature‐gradient profiles in the reactor are considered as well, both of which have been calculated theoretically and also simulated with FLUENT® software as shown in **Figure 1**.

The starting point for developing a mathematical model for the chemical and physical behaviour of the CVD process is the expression for the temperature and velocity profile for a fully developed flow. This can be calculated by solving the equation for energy conservation, the continuity equation and the equation for the motion of the carrier gas:

$$\mathbf{c}\_{\mathcal{S}} \, \rho\_{\mathcal{S}} \mu\_{\boldsymbol{x}} \frac{\partial T}{\partial \boldsymbol{x}} = \frac{\partial}{\partial \mathbf{x}} \left( \mathbf{k}\_{\boldsymbol{0}} T^{\boldsymbol{\beta}} \frac{\partial T}{\partial \mathbf{x}} \right) = \mathbf{0} \tag{1}$$

$$\frac{\partial}{\partial \mathbf{y}} \left( \begin{array}{c} \ \_0T \end{array} \stackrel{\mathcal{D}u\_x}{\partial \mathbf{y}} \right) = \text{constant} \tag{2}$$

where *ux*(*y*), *x*, *y*, µ0*T<sup>γ</sup>*, and k0*Tβ* are, respectively, the gas velocity profile, axial position in the direction of the gas flow, lateral position perpendicular to the direction of the gas flow, temperature‐dependent thermal conductivity and dynamic viscosity of the carrier gas.

**Figure 2.** Schematic illustration of a classical boundary layer and reactor conditions over the susceptor.

To develop an analytical model, first we consider a system with a constant temperature in which the active component rapidly decomposes at the susceptor, *y* = 0, for all axial flow positions *x* ≥ 0 (**Figure 2**). Therefore, the concentration of the reactant species at *y* = 0 is zero across the entire decomposition zone. The transport of material towards the susceptor in the *y*‐direction goes entirely via gas‐phase diffusion (i.e. laminar flow through the stationary boundary layer over the susceptor).

The equation for mass conservation for this case is as follows:

$$u(\mathbf{y})\frac{\partial C(\mathbf{x},\mathbf{y})}{\partial \mathbf{x}} = D \frac{\partial^2 C(\mathbf{x},\mathbf{y})}{\partial \mathbf{y}^2} \tag{3}$$

where *C* and *D* are the concentration profiles and the gas‐phase diffusion coefficient of the active component in the carrier gas, respectively, and *u*(*y*) is an expression for the parabolic velocity profile found by solving Eqs. (1) and (2) in the reactor chamber, which is given by

$$
\mu\left(\mathbf{y}\right) = 4\mu\_o \left(\frac{\mathbf{y}}{h} - \frac{\mathbf{y}^2}{h^2}\right) \tag{4}
$$

Equation (4) is difficult to solve for a parabolic velocity profile. Therefore, the problem is first solved for a constant flow velocity. Thereafter, the influence of a parabolic flow profile on the obtained results is evaluated.

By following the derivation procedure described in more detail in reference [18], the concen‐ tration profiles at *y* = 0 can be calculated as

$$\overline{C}\left(\mathbf{x}\right) = 0.692 C\_0 \exp\left[\frac{-2.52D}{h^2 u\_0} \mathbf{x}\right] \tag{5}$$

This is the average concentration profile of the reactants over the susceptor with a linear velocity profile. This profile will be used to develop the final deposition rate model for boron layer deposition.

Next, the deposition rate of the boron layer deposited in a CVD system by using B2H6 over a bare, non‐rotating silicon wafer can be calculated as [18]:

$$\mathrm{DR}\_{\mathrm{BaR}}^{\mathrm{B}}\left(\mathbf{x}\right) = 0.264 \eta \eta \beta\_1 \beta\_2 P\_{\mathrm{BH}\_1} \times \exp\left(-\frac{2.52D}{h^2 u\_0} \mathbf{x}\right) \tag{6}$$

Where

In this section, an analytical model is established to describe the deposition kinetics and the deposition chamber characteristics that determine the deposition rate (DR) over the wafer. This pre‐deposition prediction tool can be used to improve the set‐up and control of the final deposition [17, 18]. It is also very useful for transferring recipes from one reactor to the other. Some considerations such as the diffusion behaviour of the reactant (here BH3 species) through the stationary boundary layer over the wafer with the mechanism of the diffusion, the gas‐ phase processes and the related surface reactions are taken into account by this model. To develop the model, the actual parabolic gas velocity and temperature‐gradient profiles in the reactor are considered as well, both of which have been calculated theoretically and also

140 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The starting point for developing a mathematical model for the chemical and physical behaviour of the CVD process is the expression for the temperature and velocity profile for a fully developed flow. This can be calculated by solving the equation for energy conservation,

> <sup>0</sup> k 0 *g gx T T cu T xx x*

> > μ constant <sup>0</sup>

where *ux*(*y*), *x*, *y*, µ0*T<sup>γ</sup>*, and k0*Tβ* are, respectively, the gas velocity profile, axial position in the direction of the gas flow, lateral position perpendicular to the direction of the gas flow, temperature‐dependent thermal conductivity and dynamic viscosity of the carrier gas.

*<sup>x</sup> <sup>u</sup> <sup>T</sup> y y* g¶ ¶ æ ö

**Figure 2.** Schematic illustration of a classical boundary layer and reactor conditions over the susceptor.

boundary layer over the susceptor).

To develop an analytical model, first we consider a system with a constant temperature in which the active component rapidly decomposes at the susceptor, *y* = 0, for all axial flow positions *x* ≥ 0 (**Figure 2**). Therefore, the concentration of the reactant species at *y* = 0 is zero across the entire decomposition zone. The transport of material towards the susceptor in the *y*‐direction goes entirely via gas‐phase diffusion (i.e. laminar flow through the stationary

b

¶¶ ¶ æ ö = = ç ÷ ¶¶ ¶ è ø (1)

ç ÷ <sup>=</sup> ¶ ¶ è ø (2)

simulated with FLUENT® software as shown in **Figure 1**.

the continuity equation and the equation for the motion of the carrier gas:

r

$$\beta\_1 = \frac{1}{N\_0} \frac{\left(1 - \theta\_{\text{H}(\text{B})}\right)}{\left(m\_{\text{BH}\_3} \text{k}T\right)^{\frac{1}{2}}}.$$

and

**Figure 3.** Boron deposition rate extracted either from the model or experimentally as a function of (a) axial position, *x*, (b) main gas flow over the susceptor and (c) diborane partial pressure. In the figure, P1, P2 and P3 are the diborane partial pressure applied, that is, 3.39, 2.55 and 1.7 mtorr, respectively. F1, F2 and F3 are the gas flow applied, that is, 20, 15 and 10 slm, respectively. All experiments were performed at atmospheric pressure [18].

In **Figure 3**, several experimental results are compared to model predictions. It should be noted that a parameter fitting was performed for the ASM Epsilon One. The boron deposition rates are extracted as a function of (a) the axial position, *x*, for different gas flow and diborane partial pressure conditions; (b) gas flow over the susceptor; and (c) input diborane partial pressure including curves for two different axial positions. By only adjusting the reactor/process parameters, this model was also successfully transferred from the ASM Epsilon One to the Epsilon 2000 reactor, which has completely different reactor conditions. The experimental results and model predictions for the Epsilon 2000 are shown in **Figure 3a**. It is worth noting that this model is capable of predicting the deposition rate on any two‐dimensional (2D) uniformly or non‐uniformly patterned wafer such as those used for advanced device fabrica‐ tion. The very small (less than 5%) deviation of the experimental results and model prediction is plausibly related to the lateral diffusion of the diborane molecules, which becomes more evident at lower gas flow and diborane partial pressure.

It can be seen that the data calculated on the basis of this model fit well with the experimental results and have been very useful not only in the development of uniform boron layers with little pattern dependence but also in the transferring of recipes from one reactor to the other.

### **3. Low‐temperature boron deposition at 400°C**

BH on B 3 3 BH on B

*T T*

<sup>2</sup> 1 exp k k *E E*

142 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

 æ öæ ö - - - - = +- ç ÷ç ÷ è øè ø

**Figure 3.** Boron deposition rate extracted either from the model or experimentally as a function of (a) axial position, *x*, (b) main gas flow over the susceptor and (c) diborane partial pressure. In the figure, P1, P2 and P3 are the diborane partial pressure applied, that is, 3.39, 2.55 and 1.7 mtorr, respectively. F1, F2 and F3 are the gas flow applied, that is, 20,

In **Figure 3**, several experimental results are compared to model predictions. It should be noted that a parameter fitting was performed for the ASM Epsilon One. The boron deposition rates are extracted as a function of (a) the axial position, *x*, for different gas flow and diborane partial pressure conditions; (b) gas flow over the susceptor; and (c) input diborane partial pressure including curves for two different axial positions. By only adjusting the reactor/process parameters, this model was also successfully transferred from the ASM Epsilon One to the

15 and 10 slm, respectively. All experiments were performed at atmospheric pressure [18].

b

In this section, a newly developed technology for low‐temperature (400°C) boron deposition is discussed. The temperature dependency of the kinetics of the boron deposition on patterned Si/SiO2 surfaces in the temperature range of 700°C–400°C is examined. The recipe‐to‐deposit pure boron layer is explained. Also, selectivity issues that arise when the boron deposition temperature is reduced from 700°C to 400°C are analysed. Lastly, some provisions are recommended for minimising the undesirable boron deposition on oxide.

### **3.1. Temperature dependency of boron deposition**

To develop a new technology for low‐temperature (400°C) boron deposition on silicon, the temperature dependency of the deposition has to be studied to understand the kinetics of the boron deposition on patterned Si/SiO2 surfaces. This study will be described in this section.


a H\_Si/H\_B and are the silicon/boron atoms with H‐terminated dangling bonds or with free dangling bonds, respectively.

**Table 3.** Chemical reactions of the CVD boron deposition.

The kinetics of boron CVD is investigated in references [17] and [18]. The lateral gas‐phase diffusion length of boron atoms over silicon/boron surfaces during the CVD of pure boron layers is investigated in reference [20]. The chemical reactions contributing to the pure boron CVD are listed in **Table 3**.

For high‐temperature deposition (700°C), as described in reference [21], there would be enough energy to facilitate a reaction [R2]. This reaction releases hydrogen from the Si and/or B surface sites as H2 and leaves Si and/or B sites free for deposition. In this case, the deposition starts very fast, that is, after a few seconds of the surface being exposed to diborane gas under these conditions, and a monolayer of boron appears and covers the surface [1]. **Figure 4** shows a schematic of the HT deposition mechanisms for the first few monolayers of boron layers when a Si surface is exposed to diborane. In the high‐temperature deposition, the temperature‐ related reactions [R1]–[R4] proceed, leading to the adsorption of boron atoms, which are deposited and/or suspended, as well as migration of the boron atoms along the surface. This leads to a smooth, uniform, closed boron layer [17]. Triggering the intermediate reactions also ensures very smooth layers with a roughness that can be around 2 angstroms, as measured by HRTEM imaging and atomic‐force microscopy (AFM). This is illustrated by the analysis results shown in **Figure 5** for a 2‐nm‐thick layer.

**Figure 4.** Schematic of the deposition of the first few monolayers of boron layers for HT, 700°C, deposition when an Si surface is exposed to B2H6 [22].

However, the experiments show that by lowering the boron deposition temperature from 700 to 400°C, the deposition rate drops considerably from 0.4 to less than 0.01 nm/min for depo‐ sition in an H2 environment, with a laminar gas flow of 20 slm. In this case, even after a long deposition time at 400°C there is almost no measurable layer formed. Moreover, if such a layer is deposited, it is not closed enough to make it suitable for device applications. As a result, the performance of these devices is not comparable to the ones made by conventional 700°C boron deposition.

layers is investigated in reference [20]. The chemical reactions contributing to the pure boron

144 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

For high‐temperature deposition (700°C), as described in reference [21], there would be enough energy to facilitate a reaction [R2]. This reaction releases hydrogen from the Si and/or B surface sites as H2 and leaves Si and/or B sites free for deposition. In this case, the deposition starts very fast, that is, after a few seconds of the surface being exposed to diborane gas under these conditions, and a monolayer of boron appears and covers the surface [1]. **Figure 4** shows a schematic of the HT deposition mechanisms for the first few monolayers of boron layers when a Si surface is exposed to diborane. In the high‐temperature deposition, the temperature‐ related reactions [R1]–[R4] proceed, leading to the adsorption of boron atoms, which are deposited and/or suspended, as well as migration of the boron atoms along the surface. This leads to a smooth, uniform, closed boron layer [17]. Triggering the intermediate reactions also ensures very smooth layers with a roughness that can be around 2 angstroms, as measured by HRTEM imaging and atomic‐force microscopy (AFM). This is illustrated by the analysis results

**Figure 4.** Schematic of the deposition of the first few monolayers of boron layers for HT, 700°C, deposition when an Si

However, the experiments show that by lowering the boron deposition temperature from 700 to 400°C, the deposition rate drops considerably from 0.4 to less than 0.01 nm/min for depo‐ sition in an H2 environment, with a laminar gas flow of 20 slm. In this case, even after a long deposition time at 400°C there is almost no measurable layer formed. Moreover, if such a layer is deposited, it is not closed enough to make it suitable for device applications. As a result, the performance of these devices is not comparable to the ones made by conventional 700°C boron

CVD are listed in **Table 3**.

shown in **Figure 5** for a 2‐nm‐thick layer.

surface is exposed to B2H6 [22].

deposition.

**Figure 5.** HRTEM image (left) and atomic‐force microscopy (AFM) measurement (right) of a 6‐min HT‐deposited bor‐ on layer surface. The AFM image was taken in a 500×500 nm<sup>2</sup> scanning area. This HT layer is smooth and uniform at a thickness of about 2 nm with a root‐mean‐square (rms) roughness value of 0.204 nm [21].

Several factors are responsible for this negative effect. Firstly, as shown with the previously developed analytical kinetic model given in Eq. (6) [18], the deposition rate varies with the deposition temperature *T*; that is, by lowering the *T*, the DR will drop significantly. Secondly, at such a low deposition temperature, the gas‐phase diffusion length is very short. Lastly, by lowering the deposition temperature, the intermediate reactions that produce adsorption at 700°C are no longer effective.

For example, due to the very low amount of energy available, reactions [R2]–[R4] are not promoted in their forward direction. In addition, the presence of hydrogen in the reactor as a carrier gas can suppress the deposition reaction of [R1] because most of the surface sites are occupied by hydrogen that will not readily desorb at 400°C. Consequently, the first monolayer coverage takes more time at this temperature than at 700°C. Also, the inhibited removal of H from the Si surface at low temperatures plays an important role in increasing the roughness. This can be seen in **Figure 6** where the schematics of the deposition of the first few monolayers of boron are shown at LT, 400°C. The HRTEM image and AFM measurement of the LT‐ deposited boron layer at 400°C with the newly developed recipe discussed in the next section is shown in **Figure 7**. A layer roughness of around 6–8 angstroms was found for a 5‐nm‐thick layer.

By switching to N<sup>2</sup> as a carrier gas, reactions [R1–[R3] are triggered to proceed in the forward direction, while reaction [R4] is suppressed, which is the necessary intermediate reaction for providing a smooth, uniform, closed boron layer. Thus, only lowering the temperature in the conventional 700°C PureB recipe to 400°C does not guarantee a sufficient layer deposition as a promising capping layer. Therefore, for boron deposition at 400°C, a compromise between the layer deposition rate and layer uniformity and smoothness is necessary, which demands the development of a new recipe, to be discussed in detail in the next section.

**Figure 6.** Schematic of the deposition of the first few monolayers of boron layers for (a) HT, 700°C, and (b) LT, 400°C, deposition when an Si surface is exposed to B2H6 [22].

**Figure 7.** High‐resolution TEM image (bottom) and AFM measurement (top) of a boron layer deposited at 400°C for 16 min. In both cases, the thickness of the LT boron layer was extracted to about 5 nm, with a surface rms roughness val‐ ue of around 4–6 angstroms. The pictures on the left and right show a close‐up of the LT boron layer surface roughness and the interface with the Si substrate, respectively [22].

The final point to discuss about the temperature dependency of boron deposition is the investigation conducted of the deposition at the intermediate temperatures 600°C, 500°C and 450°C. By lowering the deposition temperature, the gas‐phase diffusion coefficient is lowered and the deposition rate based on Eq. (6) is reduced. In addition, the lower temperatures mean that intermediate reactions [R2]–[R4] are triggered less in their forward direction, resulting in a higher H‐coverage of the surface. Moreover, the gas‐phase diffusion lengths of boron atoms on both the Si and oxide surfaces also decrease. Altogether, this causes the deposited layer to become increasingly rough when shifting from 600°C to 500°C and 450°C. The properties of layers deposited on bare Si at temperatures between 700°C and 400°C are given in **Table 4**. What can be clearly seen is the increase in the roughness of the layers as the deposition temperature decreases.

For deposition at the lower temperatures, 450°C and 400°C, the ellipsometry measurement gave a roughness that is much greater than the average layer thickness, which may not be relevant. These parameters become very dependent on the exact deposition conditions because of both the absence of the other intermediate reactions ([R2]–[R4]) and the fact that the gas‐ phase diffusion lengths of the boron atoms on both the Si and oxide surfaces decrease.

Next, we will discuss the process and recipe requirements needed to make the boron deposi‐ tion possible at 400°C.


a The recipes only differ in time and temperature. Their structure is the same in all cases, even for 400°C, and the depositions were done only in an H2 ambient environment.

**Table 4.** Thickness and roughness of deposited layers extracted from ellipsometry measurements, as a function of deposition temperature and time.

### **3.2. LT boron deposition at 400°C**

**Figure 6.** Schematic of the deposition of the first few monolayers of boron layers for (a) HT, 700°C, and (b) LT, 400°C,

146 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 7.** High‐resolution TEM image (bottom) and AFM measurement (top) of a boron layer deposited at 400°C for 16 min. In both cases, the thickness of the LT boron layer was extracted to about 5 nm, with a surface rms roughness val‐ ue of around 4–6 angstroms. The pictures on the left and right show a close‐up of the LT boron layer surface roughness

The final point to discuss about the temperature dependency of boron deposition is the investigation conducted of the deposition at the intermediate temperatures 600°C, 500°C and 450°C. By lowering the deposition temperature, the gas‐phase diffusion coefficient is lowered and the deposition rate based on Eq. (6) is reduced. In addition, the lower temperatures mean that intermediate reactions [R2]–[R4] are triggered less in their forward direction, resulting in a higher H‐coverage of the surface. Moreover, the gas‐phase diffusion lengths of boron atoms on both the Si and oxide surfaces also decrease. Altogether, this causes the deposited layer to become increasingly rough when shifting from 600°C to 500°C and 450°C. The properties of

deposition when an Si surface is exposed to B2H6 [22].

and the interface with the Si substrate, respectively [22].

As discussed above, the presence of hydrogen gas can suppress the forward direction of reaction [R2]. This is necessary to release the hydrogen from the surface, to decrease the H‐ surface coverage, and to suppress reactions [R1] and [R3]. Therefore, the deposition rate is limited by reaction [R2]. The release of hydrogen is important as it facilitates the migration of deposited boron atoms along boron/silicon surfaces via the intermediate reactions [R3] and [R4], leading to a smooth, uniform, closed boron layer deposition. However, in the case of deposition in N2, unlike that of H2, reactions [R1]–[R3] are not suppressed, causing hydrogen atoms to be released more easily from the surface and reducing the H‐surface coverage. This leads to a higher deposition rate. However, due to the lower mobility of boron atoms in this environment, an increase in boron layer roughness was also expected. Therefore, for boron deposition at 400°C it is necessary to start the deposition in a nitrogen environment to facilitate the first monolayer deposition, and then switch to a hydrogen environment to make the surface smoother and to maintain the boron coverage over the entire silicon opening. This procedure can be repeated a few times with different cycle lengths, as shown schematically in **Figure 8**, to achieve a smooth, uniform, closed boron layer. Lowering the chamber pressure can also facilitate the release of hydrogen from the surface and decrease the H‐surface coverage. Thus, deposition at a lower ambient pressure is preferable.

**Figure 8.** Schematic illustration of carrier gases switching sequences in the LT PureB recipe [22].

The HRTEM image in **Figure 9** shows an LT boron layer deposited at 400°C at a chamber pressure of 95 torr, with the newly developed recipes applied in four‐ and six‐cycle sequences of carrier gas switching. As can be seen in these images, with more switching sequences a smoother boron layer can be created.

**Figure 9.** HRTEM image of an LT boron layer deposited after (left) the six‐ (right) and four‐cycle recipe sequences. The deposition chamber pressure was 95 torr [22].

Besides a new recipe, some additional treatments were also needed prior to the wafers being loaded into the reactor. The first treatment was a standard ex situ‐cleaning procedure prior to deposition (i.e. 10 min: HNO3 100% at 25°C; 5 min: DI water; 10 min: HNO3 70% at 110°C; and 5 min: DI water), immediately followed by a 4‐min HF (0.55%) dip to remove native oxide and H‐passivate the surface to prevent native oxide formation. This was followed by Marangoni drying, which is an effective substitute for spin‐rinse drying, to prevent the formation of wet spots. These wet spots readily form on mixed hydrophilic‐hydrophobic surfaces during spin‐ rinse drying, resulting in particle contamination from residues left behind after evaporation [23]. Unlike the 700°C deposition, for the 400°C deposition there was no in situ baking step [1]. Thus, the wafers were directly loaded into the pre‐prepared reactor at the 400°C deposition temperature. The lack of the in situ bake made the HF dip with Marangoni drying a very crucial step and meant that the wafers had to be loaded immediately into the N2‐purged load lock of the reactor to prevent any native oxide formation.

### **3.3. Issues with lowering the deposition temperature of the boron layers**

**Figure 8.** Schematic illustration of carrier gases switching sequences in the LT PureB recipe [22].

148 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

smoother boron layer can be created.

deposition chamber pressure was 95 torr [22].

The HRTEM image in **Figure 9** shows an LT boron layer deposited at 400°C at a chamber pressure of 95 torr, with the newly developed recipes applied in four‐ and six‐cycle sequences of carrier gas switching. As can be seen in these images, with more switching sequences a

**Figure 9.** HRTEM image of an LT boron layer deposited after (left) the six‐ (right) and four‐cycle recipe sequences. The

Besides a new recipe, some additional treatments were also needed prior to the wafers being loaded into the reactor. The first treatment was a standard ex situ‐cleaning procedure prior to deposition (i.e. 10 min: HNO3 100% at 25°C; 5 min: DI water; 10 min: HNO3 70% at 110°C; and 5 min: DI water), immediately followed by a 4‐min HF (0.55%) dip to remove native oxide and H‐passivate the surface to prevent native oxide formation. This was followed by Marangoni drying, which is an effective substitute for spin‐rinse drying, to prevent the formation of wet spots. These wet spots readily form on mixed hydrophilic‐hydrophobic surfaces during spin‐ rinse drying, resulting in particle contamination from residues left behind after evaporation During an HT (700°C) deposition, there is sufficient energy; therefore, the boron atoms that arrive above the oxide‐covered surfaces are very mobile and can move around and diffuse laterally to reach the Si surface and become available for the layer deposition. However, during an LT (400°C) deposition, due to the very low amount of available energy, all those boron atoms are very limited in terms of movement and therefore they remain wherever they are. This means that the concentration of the boron atoms over oxide‐covered surfaces can increase significantly so that the probability of desorption from the oxide‐free sites and/or defects becomes much lower than in the 700°C case. In fact, a significant amount of boron was observed to be deposited on the oxide. This was also promoted by the longer deposition time needed to achieve a reliably closed layer because of the low deposition rate and extra thickness needed to compensate for the large amount of roughness, as can be seen in **Figure 10**.

**Figure 10.** HRTEM images of flat/bevelled oxide surfaces of a small 40×40 µm<sup>2</sup> Si window with (a) LT and (b) HT boron layers [21, 22].

These parasitic boron depositions on the flat/bevelled oxide are unwanted and can cause problems in the subsequent processing steps. For example, the adhesion of plasma‐enhanced chemical vapour deposition (PECVD) layers of TEOS‐SiO2 was seen to degrade. This is illustrated in the SEM images of **Figure 11**. Similar adhesion issues were also witnessed in the case of physical vapour deposited (PVD) Al deposition on these surfaces. Boron depositions at 500°C can be performed without any parasitic deposition over most oxide surfaces, even at micron‐sized windows. However, still for some oxide qualities this kind of unwanted parasitic deposition was observed. An example is given in **Figure 12b** and **c**, where two different dyes with the same layout are shown with and without adhesion issues for subsequent PECVD TEOS and PVD Al layer depositions.

**Figure 11.** Examples of poor PECVD TEOS‐SiO2 adhesion as a result of parasitic boron deposition during LT deposi‐ tion of boron on flat/bevelled oxide surfaces near micron‐sized Si windows where the LOR is high [21].

**Figure 12.** Examples of poor adhesion and layer delamination for the same layout. Die (a) shows PECVD TEOS depos‐ ited over 400°C boron. PVD Al was deposited after PECVD TEOS deposition over 500°C boron for two different dies on the same wafer. Die (b) displays the adhesion problems. However, die (c) does not display the same adhesion prob‐ lems as die (b). The insets show the zoom‐in for different structures with micron‐sized openings [21].

### **3.4. Precautions to minimise the parasitic boron deposition at low temperatures**

To minimise the undesirable parasitic boron deposition on the oxide, there are a few precau‐ tions that can be taken into account. Firstly, the chances of parasitic boron deposition can be reduced by using the higher quality oxide, which has a lower density of oxide surface‐free sites and/or defects. Secondly, the global and local oxide coverage ratios can be optimised in the layout of the oxide mask to reduce the unnecessary boron atom accumulation over the oxide areas. Likewise, the deposition and process parameters, such as diborane partial pressure, gas flow and total deposition pressure, can be optimised to minimise the loading effects [15, 24]. These will otherwise also increase the accumulation of boron atoms over certain oxide areas, thus increasing the probability of parasitic boron deposition there. Also, we have found that this poor adhesion could depend on the ex/in situ treatments before the PECVD TEOS‐SiO<sup>2</sup> or PVD Al layers are deposited. Furthermore, the adhesion of these extra layers after boron deposition is also influenced by the exact conditions on the surface. For example, a large amount of hydrogen resides on the surface just after boron deposition and is in general known to reduce adhesion. This H coverage can be reduced by carrying out the baking steps in air. An investigation of these extra treatments would have to be included in the process develop‐ ment when working with LT boron deposition.

By taking the aforementioned precautions into account, the LT boron deposition was success‐ fully performed to create near‐ideal LT PureB photodiodes with low, deep‐junction‐like saturation currents.

### **4. LT PureB UV photodiode fabrication**

**Figure 11.** Examples of poor PECVD TEOS‐SiO2 adhesion as a result of parasitic boron deposition during LT deposi‐

**Figure 12.** Examples of poor adhesion and layer delamination for the same layout. Die (a) shows PECVD TEOS depos‐ ited over 400°C boron. PVD Al was deposited after PECVD TEOS deposition over 500°C boron for two different dies on the same wafer. Die (b) displays the adhesion problems. However, die (c) does not display the same adhesion prob‐

To minimise the undesirable parasitic boron deposition on the oxide, there are a few precau‐ tions that can be taken into account. Firstly, the chances of parasitic boron deposition can be reduced by using the higher quality oxide, which has a lower density of oxide surface‐free sites and/or defects. Secondly, the global and local oxide coverage ratios can be optimised in the layout of the oxide mask to reduce the unnecessary boron atom accumulation over the oxide areas. Likewise, the deposition and process parameters, such as diborane partial pressure, gas flow and total deposition pressure, can be optimised to minimise the loading effects [15, 24]. These will otherwise also increase the accumulation of boron atoms over certain oxide areas, thus increasing the probability of parasitic boron deposition there. Also, we have found that

lems as die (b). The insets show the zoom‐in for different structures with micron‐sized openings [21].

**3.4. Precautions to minimise the parasitic boron deposition at low temperatures**

tion of boron on flat/bevelled oxide surfaces near micron‐sized Si windows where the LOR is high [21].

150 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The post‐metal LT boron deposition was successfully performed to create p<sup>+</sup> n photodiodes with nm‐thin, boron‐only beam entrance windows and near‐theoretical UV sensitivity with negligible optical and electrical degradation [25]. **Figure 13** shows the processing steps of the

**Figure 13.** Processing steps of the fabricated photodiodes with boron‐only beam entrance windows, for boron deposi‐ tion both before (a) and after (b) metallisation [25].

fabricated photodiodes with boron‐only beam entrance windows, for boron deposition both before and after metallisation.

**Figure 14.** Measured I‐V characteristics of both pre‐metal HT and post‐metal LT PureB photodiodes with an active area of 9.6×9.6 mm<sup>2</sup> . Schematics of the fabricated photodiodes are shown in **Figure 13**. The boron layer thickness was meas‐ ured by ellipsometry to be 3.2 and 4.5 nm for the HT and LT photodiodes, respectively [25].

**Figure 15.** Measured VUV spectral responsivity as a function of wavelength for HT and LT PureB photodiodes (area: 9.6×9.6 mm<sup>2</sup> ) before and after high‐level irradiation and after a final cleaning procedure [22].

The I‐V characteristics measured for both pre‐metal HT and post‐metal LT PureB photodiodes with an active area of 9.6×9.6 mm<sup>2</sup> are shown in **Figure 14**. The boron layer thickness was measured by ellipsometry to be 3.2 and 4.5 nm for the HT and LT photodiodes, respectively. As can be seen, low deep‐junction‐like saturation currents and near‐ideal diode characteristics can be provided by LT boron deposition.

fabricated photodiodes with boron‐only beam entrance windows, for boron deposition both

152 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 14.** Measured I‐V characteristics of both pre‐metal HT and post‐metal LT PureB photodiodes with an active area

**Figure 15.** Measured VUV spectral responsivity as a function of wavelength for HT and LT PureB photodiodes (area:

) before and after high‐level irradiation and after a final cleaning procedure [22].

ured by ellipsometry to be 3.2 and 4.5 nm for the HT and LT photodiodes, respectively [25].

. Schematics of the fabricated photodiodes are shown in **Figure 13**. The boron layer thickness was meas‐

before and after metallisation.

of 9.6×9.6 mm<sup>2</sup>

9.6×9.6 mm<sup>2</sup>

**Figure 15** shows the measured responsivity in the VUV spectral range (120–400 nm) for both pre‐metal and post‐metal LT PureB photodiodes. As can be seen in this figure, the response for pre‐metal photodiodes is slightly higher than that for post‐metal ones. This could be explained by a small thinning of the boron layer due to the extra processing necessary to open the light entrance window (see **Figure 13**). The HT PureB device is also shown in **Figure 15** as a reference. As will be commented on subsequently, the rough LT boron surface proves to be much less chemically resilient than the smooth HT boron layers.

**Figure 16.** Measured responsivity across the surface in one direction of a 9.6×9.6 mm<sup>2</sup> LT PureB photodiode at wave‐ lengths of 193 and 121 nm, before and after high‐dose exposure, and after cleaning. The high‐dose exposure was per‐ formed with a circular beam spot of ø1 mm with a radiant exposure of 37 J/cm<sup>2</sup> [22].

The optical stability of the post‐metal LT PureB photodetector was measured by exposing the centre of the 10×10 mm<sup>2</sup> dies to high‐dose VUV irradiation. The high‐dose exposure was performed with a circular beam spot with a diameter of 1 mm, to a radiant exposure of 37 J/ cm<sup>2</sup> . The LT PureB photodiodes were found to be highly stable. To illustrate, the resulting responsivity at the centre of the post‐metal LT PureB photodiode is included in **Figure 16**.

A slight dip in responsivity is visible after the 121‐nm exposure. This is the result of the build‐ up of carbon contamination during exposure. To remove this layer, an ozone‐cleaning procedure was performed. For HT devices, the cleaning procedure successfully removes the contamination, and the original high responsivity is regained [26]. By contrast, this treatment cannot effectively remove the carbon contamination of the LT device, probably due to the rougher surface structure as shown in **Figure 13**. As can be seen, the rougher surface makes the carbon contamination removal more difficult than the smooth surface structures. Addi‐ tionally, this rougher surface structure makes the silicon surface below the boron layer more susceptible to oxidation [27]. The higher responsivity after standard ozone‐cleaning suggests that the boron layer is thinned considerably, as can be seen in **Figures 15** and **16**. On closer look at these figures, a reduced thinning can be clearly seen where the carbon contamination was present for both measurement wavelengths (i.e. in the middle of the active area). There‐ fore, although the as‐deposited LT boron layer is optically robust, it is more vulnerable to carbon contamination due to its rougher surface compared to HT boron layer, as can be seen in **Figure 9**.

### **5. Summary**

In this chapter, an analytical model is developed to describe the deposition kinetics and deposition chamber characteristics that determine the deposition rate over the wafer, which can be used as a pre‐deposition prediction tool to improve the set‐up and control of the final deposition. The model has been very useful both in the development of uniform boron layers with minimal pattern dependence and in the transferring of recipes from one reactor to the other.

A new technology for low‐temperature (400°C) boron deposition is presented, which provides a smooth, uniform, closed boron layer. The temperature dependency of the boron deposition on patterned Si/SiO2 surfaces in the temperature range of 400–700°C is investigated. Some selectivity issues that arise when the boron deposition temperature is reduced from 700°C to 400°C are discussed. Some provisions are recommended to minimise the undesirable boron deposition on the oxide. For boron deposition at 400°C, it is necessary to start the deposition in a nitrogen environment to facilitate the first monolayer deposition, and then switch to a hydrogen environment to make the surface smoother and to maintain the boron coverage over the entire silicon opening.

The LT PureB technology is then successfully employed to create near‐ideal LT PureB photodiodes with nm‐thin, boron‐only beam entrance windows and a near‐theoretical sensitivity for irradiation with either VUV/DUV/EUV light down to a wavelength of 10 nm. Very low dark current of only 15 pA at -1 V bias voltage was measured for post‐metal LT PureB photodiodes with an anode area of 9.6×9.6 mm2 , which we relate to the fact that no post‐PureB processing was performed.

It is believed that at temperatures of 400°C and lower, no silicon doping is possible during the boron deposition. Yet, the LT boron layers provide a structure with excellent p‐n junction‐like I‐V characteristics. Based on this observation, we make the assumption that during the damage‐free LT boron deposition, which is most likely a result of a chemical reaction, a kind of effective "hole" layer is built between the boron and silicon. This interface "hole" layer creates an electric field that repels the injected minority carriers (electrons) away from the interface.

### **Author details**

cannot effectively remove the carbon contamination of the LT device, probably due to the rougher surface structure as shown in **Figure 13**. As can be seen, the rougher surface makes the carbon contamination removal more difficult than the smooth surface structures. Addi‐ tionally, this rougher surface structure makes the silicon surface below the boron layer more susceptible to oxidation [27]. The higher responsivity after standard ozone‐cleaning suggests that the boron layer is thinned considerably, as can be seen in **Figures 15** and **16**. On closer look at these figures, a reduced thinning can be clearly seen where the carbon contamination was present for both measurement wavelengths (i.e. in the middle of the active area). There‐ fore, although the as‐deposited LT boron layer is optically robust, it is more vulnerable to carbon contamination due to its rougher surface compared to HT boron layer, as can be seen

154 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

In this chapter, an analytical model is developed to describe the deposition kinetics and deposition chamber characteristics that determine the deposition rate over the wafer, which can be used as a pre‐deposition prediction tool to improve the set‐up and control of the final deposition. The model has been very useful both in the development of uniform boron layers with minimal pattern dependence and in the transferring of recipes from one reactor to the

A new technology for low‐temperature (400°C) boron deposition is presented, which provides a smooth, uniform, closed boron layer. The temperature dependency of the boron deposition on patterned Si/SiO2 surfaces in the temperature range of 400–700°C is investigated. Some selectivity issues that arise when the boron deposition temperature is reduced from 700°C to 400°C are discussed. Some provisions are recommended to minimise the undesirable boron deposition on the oxide. For boron deposition at 400°C, it is necessary to start the deposition in a nitrogen environment to facilitate the first monolayer deposition, and then switch to a hydrogen environment to make the surface smoother and to maintain the boron coverage over

The LT PureB technology is then successfully employed to create near‐ideal LT PureB photodiodes with nm‐thin, boron‐only beam entrance windows and a near‐theoretical sensitivity for irradiation with either VUV/DUV/EUV light down to a wavelength of 10 nm. Very low dark current of only 15 pA at -1 V bias voltage was measured for post‐metal LT PureB

It is believed that at temperatures of 400°C and lower, no silicon doping is possible during the boron deposition. Yet, the LT boron layers provide a structure with excellent p‐n junction‐like I‐V characteristics. Based on this observation, we make the assumption that during the damage‐free LT boron deposition, which is most likely a result of a chemical reaction, a kind of effective "hole" layer is built between the boron and silicon. This interface "hole" layer

, which we relate to the fact that no post‐PureB

in **Figure 9**.

**5. Summary**

the entire silicon opening.

processing was performed.

photodiodes with an anode area of 9.6×9.6 mm2

other.

Vahid Mohammadi\* and Stoyan Nihtianov

\*Address all correspondence to: V.Mohammadi@tudelft.nl

Department of Microelectronics, Delft University of Technology, Delft, The Netherlands

### **References**


[21] Mohammadi V., Golshani N., Mok K.R.C., De Boer W.B., Derakhshandeh J., Nanver L.K. Temperature dependency of the kinetics of PureB CVD deposition over patterned Si/SiO2 surfaces. Microelectronic Engineering. 2014;125:45–50. DOI: http://dx.doi.org/ 10.1016/j.mee.2014.03.015

[8] Canfield L.R., Vest R.E., Korde R., Schmidtke H., Desor R. Absolute silicon photodiodes

[9] Scholze F., Brandt G., Mueller P., Meyer B., Scholz F., Tümmler J., et al. High‐accuracy

[10] Scholze F., Klein R., Muller R. Characterization of detectors for extreme UV radiation.

[11] FEI. High‐performance Microscopy Workflow Solutions [Internet]. [Updated: 2016].

[12] Roussel L.Y., Stokes D.J., Gestmann I., Darus M., Young R.J. Extreme high resolution scanning electron microscopy (XHR SEM) and beyond. In: SPIE 7378, Scanning

[13] Jacquot B.C., Hoenk M.E., Jones T.J., Cunningham T.J., Nikzad S. Direct detection of 100–5000 eV electrons with delta‐doped silicon CMOS and electron‐multiplying CCD

[14] Mohammadi V., De Boer W., Scholtes T.L., Nanver L.K. Pattern dependency of pure‐ boron‐layer chemical‐vapor depositions. ECS Transactions. 2012;45:39–48. DOI: http://

[15] Mohammadi V., De Boer W., Scholtes T.L.M., Nanver L.K. Local‐loading effects for pure‐boron‐layer chemical‐vapor deposition. ECS Transactions. 2013;50:333–341. DOI:

[16] Mohammadi V., Van de Kruijs R.W.E., Rao P.R., Sturm J.M., Nihtianov S. Influence of the surface oxide content of a boron capping layer on UV photodetector performance. In: 9th International Conference on Sensing Technology; 8–10 Dec.; Auckland, New

[17] Mohammadi V., De Boer W.B., Nanver L.K. Temperature dependence of chemical‐ vapor deposition of pure boron layers from diborane. Applied Physics Letters.

[18] Mohammadi V., De Boer W.B., Nanver L.K. An analytical kinetic model for chemical‐ vapor deposition of PureB layers from diborane. Journal of Applied Physics.

[19] Mohammadi V., Mohammadi S., Ramesh S., Nihtianov S. Numerical gas flow and heat transfer simulation in the ASM Epsilon 2000 CVD reactor for pure boron deposition. In: 23rd International Scientific Conference Electronics‐ET; 11–13 Sept.; Sozopol,

[20] Mohammadi V., Nihtianov S. Lateral gas phase diffusion length of boron atoms over Si/B surfaces during CVD of pure boron layers. AIP Advances. 2016;6:025103.1–

2012;101:111906.1–111906.4. DOI: http://dx.doi.org/10.1063/1.4752109

2012;112:113501.1–113501.10. DOI: http://dx.doi.org/10.1063/1.4767328

025103.8. DOI: http://dx.doi.org/10.1063/1.4941702

detector calibration for EUV metrology at PTB. SPIE. 2002;4688:680–689.

for 160 nm to 254 nm photons. Metrologia. 1998;35:329–334.

156 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Microscopy; 4 May; Monterey, CA. 2009. p. 73780W–73780W‐9.

imagers. IEEE Transactions on Electron Devices. 2012;59:1988–1992.

Metrologia. 2006;43:S6–S10.

dx.doi.org/10.1149/1.3700937

Zealand. 2015.

Bulgaria. 2014. p. 28–31.

http://dx.doi.org/10.1149/05004.0333ecst

Available at: http://www.fei.com/


### **Silicon-Rich Oxide Obtained by Low-Pressure Chemical Vapor Deposition to Develop Silicon Light Sources**

J. Alarcón-Salazar, R. López-Estopier,

E. Quiroga-González, A. Morales-Sánchez,

J. Pedraza-Chávez, I. E. Zaldívar-Huerta and

M. Aceves-Mijares

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/63012

### **Abstract**

Off stoichiometric silicon oxide, also known as silicon-rich oxide (SRO), is a lightemitting material that is compatible with silicon technology; therefore, it is a good candidate to be used as a light source in all-silicon optoelectronic circuits. The SRO obtained by low-pressure chemical vapor deposition (LPCVD) has shown the best luminescent properties compared to other techniques. In spite of LPCVD being a simple technique, it is not a simple task to obtain SRO with exact silicon excess in a reliable and repetitive way. In this work, the expertise obtained in our group to obtain SRO by LPCVD with precise variation is presented. Also, the characteristics of this SRO obtained in our group are revised and discussed. It is demonstrated that LPCVD is an excellent technique to obtain single layers and multilayers of nanometric single layers with good characteristics.

**Keywords:** SRO, LPCVD, photoluminescence, electroluminescence, multilayer

### **1. Introduction**

Chemical vapor deposition (CVD) is a versatile and economical technique used to deposit different materials. In the microelectronics industry, it has found a main place and it is a standard process. Currently, many efforts are being done to produce optoelectronic circuits using the mature technology of integrated circuits. A major drawback to integrate a whole

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

silicon circuit that manages both electronic and optical signals is that silicon does not emit light efficiently. There are serious restrictions to integrate a light source in such a system [1]. Basically, two approaches have been under study to solve the problem of the light source: One of them uses a reverse-biased *pn* junction [2], the other one uses light-emitting materials that are compatible with silicon [3, 4]. In spite of the fact that both approaches have shown that it is possible to integrate a complete optoelectronic system based on Si, there is still a wide field of possibilities to improve the efficiency of the light emitters [2, 5], and in consequence, of the whole system.

Off stoichiometric silicon oxide (with empirical formula SiO*x*), also known as silicon-rich oxide (SRO), is a good material to be used as a light source in all-silicon optoelectronics circuit [6]. It is obtained by different techniques, including plasma-enhanced and low-pressure chemical vapor deposition (PECVD and LPCVD) methods. The silicon excess in SRO obtained by LPCVD can be easily controlled during the deposition by the ratio of the reactive gases, in our case silane and nitrous oxide:

$$R\_0 = \frac{P\_{N\_2O}}{P\_{SH\_4}}\tag{1}$$

Dong et al. showed that for *R*0 = 10, 20, and 30, the corresponding silicon excess is about 12, 7 and 5 at %, respectively; that is, SiO*x* with *x* = 1.13, 1.50, and 1.63 [7]. However, it is worthy to mention that SiO*x* is a multiphase material composed of SiO2, elemental silicon, and SiO*y*. Thus, SiO*x* is an empirical formula to denote SRO (which may contain large Si excess), and SiO*y* is an oxide with stoichiometry deviating a bit from SiO2. Oxidation states obtained by X-ray photoelectron spectroscopy (XPS) of LPCVD-SRO for *R*0 =10, 20, and 30 are shown in [4]. Depending on the silicon excess, SRO has different characteristics: For low silicon excess, the density of SiO*y* compounds dominates, producing intense luminescence. However, for high silicon excess the density of elemental silicon increases, what reduces the emission, but increases the conductivity of the films.

In order to have intense luminescent SiO*x*, high-temperature treatments are required. For SRO obtained using LPCVD, 1100°C in nitrogen is the most favorable temperature. Additionally, the emission also depends on the *R*0. The emission increases as *R*0 increases, being *R*0 = 30 the top one; for *R*0 higher than 30, the emission reduces again. On the other hand, the conductivity increases as the *R*0 reduces, This compromise is important when electroluminescent devices are the goal. Therefore, structures that combine layers of high conductivity with layer of high emission properties are under study to obtain efficient light sources compatible with silicon.

SRO obtained using LPCVD is perhaps the most luminescent compared with SRO obtained by other methods [6]; however, in this technique, it is difficult to control the silicon excess with some precision, and to have films with controllable properties required of personal with expertise on this type of technique.

In this paper, details of our LPCVD deposition processes to obtain SRO single layers (SLs) and multilayers (MLs) with different *R*0, including details of our LPCVD system, are described. Also, the optical and structural characteristics of our SRO layers and multilayers are reviewed. Electroluminescence (EL), cathodoluminescence (CL), and photoluminescence (PL) results will be presented and discussed.

### **2. Our system**

silicon circuit that manages both electronic and optical signals is that silicon does not emit light efficiently. There are serious restrictions to integrate a light source in such a system [1]. Basically, two approaches have been under study to solve the problem of the light source: One of them uses a reverse-biased *pn* junction [2], the other one uses light-emitting materials that are compatible with silicon [3, 4]. In spite of the fact that both approaches have shown that it is possible to integrate a complete optoelectronic system based on Si, there is still a wide field of possibilities to improve the efficiency of the light emitters [2, 5], and in consequence, of the

160 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Off stoichiometric silicon oxide (with empirical formula SiO*x*), also known as silicon-rich oxide (SRO), is a good material to be used as a light source in all-silicon optoelectronics circuit [6]. It is obtained by different techniques, including plasma-enhanced and low-pressure chemical vapor deposition (PECVD and LPCVD) methods. The silicon excess in SRO obtained by LPCVD can be easily controlled during the deposition by the ratio of the reactive gases, in our

2

*SiH*

Dong et al. showed that for *R*0 = 10, 20, and 30, the corresponding silicon excess is about 12, 7 and 5 at %, respectively; that is, SiO*x* with *x* = 1.13, 1.50, and 1.63 [7]. However, it is worthy to mention that SiO*x* is a multiphase material composed of SiO2, elemental silicon, and SiO*y*. Thus, SiO*x* is an empirical formula to denote SRO (which may contain large Si excess), and SiO*y* is an oxide with stoichiometry deviating a bit from SiO2. Oxidation states obtained by X-ray photoelectron spectroscopy (XPS) of LPCVD-SRO for *R*0 =10, 20, and 30 are shown in [4]. Depending on the silicon excess, SRO has different characteristics: For low silicon excess, the density of SiO*y* compounds dominates, producing intense luminescence. However, for high silicon excess the density of elemental silicon increases, what reduces the emission, but

In order to have intense luminescent SiO*x*, high-temperature treatments are required. For SRO obtained using LPCVD, 1100°C in nitrogen is the most favorable temperature. Additionally, the emission also depends on the *R*0. The emission increases as *R*0 increases, being *R*0 = 30 the top one; for *R*0 higher than 30, the emission reduces again. On the other hand, the conductivity increases as the *R*0 reduces, This compromise is important when electroluminescent devices are the goal. Therefore, structures that combine layers of high conductivity with layer of high emission properties are under study to obtain efficient light sources compatible with silicon.

SRO obtained using LPCVD is perhaps the most luminescent compared with SRO obtained by other methods [6]; however, in this technique, it is difficult to control the silicon excess with some precision, and to have films with controllable properties required of personal with

*N O*

*P*

0

*R*

4

*<sup>P</sup>* <sup>=</sup> (1)

whole system.

case silane and nitrous oxide:

increases the conductivity of the films.

expertise on this type of technique.

We have two homemade LPCVD reactors, one for two- and another for four-inch wafers. Both reactors have the same layout, thus we will describe only one of them in a general way. Our laboratory is a teaching and research facility, therefore every day different materials have to be deposited and the equipment has to be very versatile. For this reason, we found that controlling it manually produces better results than using automatic parts, then the control of the gas flux using rotameter (ABB model 10A6131NB1B1X00) give us enough functionality. High throughput is not required and our main concern is to have good films with repetitive characteristics. Normally, polysilicon, silicon nitride, silicon oxynitride, and SROs are depos‐ ited in the reactor; however, we are not limited to only those materials. Perhaps, obtaining SROs with good characteristics is the most demanding, that is because small differences in silicon excess produce big changes in its characteristics. In the following paragraphs, we will concentrate on describing the details to obtain SRO in a controllable manner.

As shown in **Figure 1**, our LPCVD system is hot wall type that allows having a uniform temperature in the whole deposition chamber area. The heating element is a three-zone furnace, and a flat zone of ±2°C can be obtained. In the past, an analysis using multivariated experiment was carried out to study different parameters involved in the deposition process [8]. Based on that, we decide to maintain the wafer horizontally on a flat quartz wafer holder. The working temperature profile was chosen with an increasing slop to obtain lesser thickness variation, as shown in **Figure 2**. The increasing temperature compensates the changes of the boundary layer and produces a more uniform deposition through the flat wafer holder [9]. The deposition temperature allows to deposit SRO from *R*0 = 5 to *R*0 = 100.

**Figure 1.** Schematic of the LPCVD system. A three zones furnace and mechanical pump are used.

To obtain SRO, the reactive gases are N2O and SiH4 at 5%, the silane is diluted in N2. The high dilution of silane is a restriction of the system in order to increase the versatility of the reactor. Thereby, we have no possibilities to vary the chamber pressure varying a gas carrier. **Figure 3** shows a calibration graph of the pressure of silane and nitrous oxide as a function of the gas flux. The flux of the N2O is controlled by two rotameters as shown in the schematic of **Figure 1**. Double control of nitrous oxide allows for an efficient way to produce nanometric layers in multilayers with different *R*0, that is, layers of different silicon excess. For the two previous figures, the vacuum valve was 75% open and the base vacuum was at least 6 mTorr.

**Figure 2.** Working temperature profile to deposit SRO with different *R*0´s. The profile increases along the deposit area to compensate for the boundary layer.

**Figure 3.** Calibration graph of pressure as function of the gas flux of (a) silane and (b) nitrous oxide. In case of N2O, the flux is controlled by two rotameters.

The characteristics of the SRO depend strongly on the silicon excess; the flux ratio *R*0 is used to control such excess. In the LPCVD technique, it is easy controlling the *R*<sup>0</sup> by the ratio of partial pressure of each gas, defined in Eq. (1). In our case, the dilution of silane has to be taken into account. Therefore, Eq. (1) has to be written adding a multiplying factor (*F*) as

To obtain SRO, the reactive gases are N2O and SiH4 at 5%, the silane is diluted in N2. The high dilution of silane is a restriction of the system in order to increase the versatility of the reactor. Thereby, we have no possibilities to vary the chamber pressure varying a gas carrier. **Figure 3** shows a calibration graph of the pressure of silane and nitrous oxide as a function of the gas flux. The flux of the N2O is controlled by two rotameters as shown in the schematic of **Figure 1**. Double control of nitrous oxide allows for an efficient way to produce nanometric layers in multilayers with different *R*0, that is, layers of different silicon excess. For the two previous

**Figure 2.** Working temperature profile to deposit SRO with different *R*0´s. The profile increases along the deposit area

**Figure 3.** Calibration graph of pressure as function of the gas flux of (a) silane and (b) nitrous oxide. In case of N2O, the

to compensate for the boundary layer.

flux is controlled by two rotameters.

figures, the vacuum valve was 75% open and the base vacuum was at least 6 mTorr.

162 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

$$R\_0 = \frac{P \ast P\_{N\_2O}}{P\_{SH\_4}} \tag{2}$$

Considering that only 0.05 parts of gas corresponds to silane, *F* takes a value equal to 20. However, due to the natural variation of the gas provided by the supplier, when a new tank is used, a new calibration is carried out in such a way that the refractive index (*n*) and PL are kept within the known values. Then, normally, the multiplicative factor is different than 20. This procedure is also done periodically to assure the gases aging do not alter the characteristics of the SRO films.

In the everyday procedure to obtain always the same conditions, we fixed the *P*SiH4 and varied the *P*N2O as required by the *R*0; **Table 1** shows the set values used for different *R*0. To deposit single layers, the flux of silane is varied until the partial pressure is obtained, then the silane valve is closed and the nitrous oxide flux is set until the partial pressure is obtained. Next, both valves are shutoff and wafers are loaded. When the base vacuum is established, silane and nitrous oxide valves are simultaneously opened. The reactive gases mix and react producing variations in pump extraction and a reduction of the total pressure is observed. That is, the total pressure is different from that of the sum of the partial pressures. Also, a variation of the gases flux is observed. To obtain a repetitive process, it is necessary to adjust the fluxes to the values they had before the gases are mixed. This procedure assures to obtain repeatable characteristics. Also in **Table 1**, the refractive index, the rate of deposition, and the thickness variation from beginning to end of the wafer holder are presented.


**Table 1.** Deposit conditions in LPCVD system for different *R*0's, also the refractive index, the deposit rate and the thickness variation is presented.

There are many partial pressures combinations that fulfill Eq. (1). However, depending on each laboratory conditions, it is recommendable to set a linear relationship between the flux ratio and the partial pressure ratio of each *R*0. In our knowledge, it is very important to maintain a linear relationship of the ratios of pressure and flux (**Figure 4**) between the different *R*0's. Nevertheless, different deposit conditions such as pump valve aperture, system cleanliness, dilution of silane, and aging produce different relations. In our experiments, all the mentioned factors were taken into account varying the multiplicative factor (*F*) of Eq. (2). **Figure 4** shows different relationship varying the *F* due to different conditions of the system. In our system, using the vacuum valve open at 75% of the maximum aperture and purging with nitrogen the gas lines a couple of hour before the deposit are enough to maintain a linear relationship, and with it we obtain repetitive results.

**Figure 4.** Linear relationship between flux ratio and partial pressure ratio of each *R*0.

To deposit a multilayer with different *R*0's, the procedure is similar to that of single layers. However, in this case, the flux of N2O is set using the two flux controls. Using the gas that goes through one of them, one *R*0 is fixed (Ro1). The other *R*0 (Ro2) is fixed by using the second rotameter or both simultaneously, depending on the deposit conditions. The nitrous oxide valves are open during the time needed to obtain the Ro1 or Ro2 successively. The time required for each layer is established to have the nominal thickness, and it is calculated from the rate of deposit. If the layer being deposited is thin, then the time can be some seconds.

As it is expected in a low-pressure system, deposition in our system works under surface reaction kinetics limited [9, 10]. It implies that the deposit rate is low and good step coverage is obtained, and also by-products are trapped in the film, and that is why off stoichiometric silicon oxide is obtained. The LPCVD also has shown good step coverage, and in order to corroborate our step coverage results, the sticking factor and the gas arrival dates are estimated.

The sticking coefficient (*S*) is defined as

There are many partial pressures combinations that fulfill Eq. (1). However, depending on each laboratory conditions, it is recommendable to set a linear relationship between the flux ratio and the partial pressure ratio of each *R*0. In our knowledge, it is very important to maintain a linear relationship of the ratios of pressure and flux (**Figure 4**) between the different *R*0's. Nevertheless, different deposit conditions such as pump valve aperture, system cleanliness, dilution of silane, and aging produce different relations. In our experiments, all the mentioned factors were taken into account varying the multiplicative factor (*F*) of Eq. (2). **Figure 4** shows different relationship varying the *F* due to different conditions of the system. In our system, using the vacuum valve open at 75% of the maximum aperture and purging with nitrogen the gas lines a couple of hour before the deposit are enough to maintain a linear relationship, and

164 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

with it we obtain repetitive results.

**Figure 4.** Linear relationship between flux ratio and partial pressure ratio of each *R*0.

To deposit a multilayer with different *R*0's, the procedure is similar to that of single layers. However, in this case, the flux of N2O is set using the two flux controls. Using the gas that goes through one of them, one *R*0 is fixed (Ro1). The other *R*0 (Ro2) is fixed by using the second rotameter or both simultaneously, depending on the deposit conditions. The nitrous oxide valves are open during the time needed to obtain the Ro1 or Ro2 successively. The time required for each layer is established to have the nominal thickness, and it is calculated from the rate of deposit. If the layer being deposited is thin, then the time can be some seconds.

$$S = \frac{R\_{\rm r}}{A\_{\rm r}} \tag{3}$$

where *R*<sup>r</sup> is the reaction rate and *A*<sup>r</sup> is the arrival rate, respectively. The reaction rate is related with deposit rate; meanwhile the arrival rate is the velocity of the total flux of precursor gases that enters into the chamber. In our case, *A*r is estimated as

$$A\_r = \frac{F\_\text{T}}{A} \tag{4}$$

where *F*T is the sum of the fluxes of precursor gasses and A is the tubing area (our system uses ¼ inches diameter). Using Eqs. (3) and (4) and conditions from **Table 1**, the sticking coefficient is determined for each *R*0. **Table 2** shows results of *A*r, *R*r, and S for each *R*<sup>0</sup> from 5 to 50. The sticking coefficient has values in the order to 10−11. This value is rather too low; however, independent of the numerical value, it agrees with what is expected from a low-pressure system; see, for example, [9–11].


**Table 2.** Arrival and reaction rates and sticking coefficient for SRO-LPCVD deposited at 736°C.

### **3. Experimental procedure**

SRO films were deposited on <100> and low resistivity (5–10 Ω cm) silicon substrates by LPCVD at 736°C. The ratio between reactive gases nitrous oxide (N2O) and silane (SiH4) was varied to obtain films with different silicon excess. Single layers with *R*<sup>0</sup> values of 5, 10, 20, 25, 30, 35, and 50 were deposited, and will be labeled as SRO5, SRO10, SRO20, SRO25, SRO30, SRO35, and SRO50, respectively, for clarity. Also, two multilayer structures were fabricated. The multilayer is a stack of seven layers. One structure intercalates three SRO25 layers with four SRO5 layers, and the second one with four SRO10 layers. After deposition, all samples were thermally annealed at 1100°C for 3 h in nitrogen ambience to induce the silicon agglomeration.

Thickness and refractive index of all samples, including multilayer structure, were determined using a null ellipsometer Gaertner L117 with a laser He–Ne of 632.8 nm wavelength. The PL emission spectra were obtained with a Fluoromax-3 spectrometer; all the films were excited with UV radiation (300 nm) and the luminescence was measured from 370 to 1000 nm with a resolution of 1 nm. Optical filters were used in order to guarantee the wavelength of excitation beam. CL measurements were performed using a luminoscope equipment model ELM2-144, 0.3-mA current and 5 kV were used. The luminescence spectra (PL and CL) were measured at room temperature.

For electrical and electroluminescent studies, Metal-Insulator-Semiconductor (MIS) devices were fabricated, and we refer to them as light emitting capacitor (LEC). A ~250-nm-thick semitransparent n+ polycrystalline silicon (Poly) gate was deposited onto the SRO film surface by LPCVD. After a photolithography process step, square-shaped gates of 4-mm2 area were defined. The backside contact was formed with 1-µm thick aluminum layer by evaporation. Finally, the devices were thermally annealed at 480°C in forming gas.

A source meter Keithley model 2400 was used to obtain current versus voltage (I-V) curves. EL spectra were obtained by biasing the device with a constant DC voltage. The light emitted was collected with an optical fiber located facing the Poly gate and connected to the Fluoromax 3 spectrometer.

### **4. Composition of SRO by LPCVD**

SRO is a multiphase material composed of silicon oxides of different stoichiometry and Si nanocrystals (Si-ncs). In XPS spectra of this material, each Si 2p core level band is composed of bands originated in Si at different oxidation states (Si0 , Si1+, Si2+, Si3+, Si4+), which manifest themselves at different energies. The position of the peaks corresponding to Si0 and Si4+ (SiO2) is well known and is easily distinguishable [12], but the peaks related with silicon suboxides cannot be distinguished unequivocally in a complex spectrum composed of different Si oxide species; they have been usually studied at Si/SiO2 interfaces [13–16]. In this way, a quantitative analysis of such highly convoluted spectra is not straightforward. The material can be conveniently considered as composed of SiO2, elemental Si and SiO*y* with 0 < *y* < 2. **Figure 5** shows an XPS spectrum of SRO with *R*0 = 1 as example. The three phases can be clearly distinguished.

**Figure 5.** XPS spectrum of SRO with *R*0 =1 [17].

varied to obtain films with different silicon excess. Single layers with *R*<sup>0</sup> values of 5, 10, 20, 25, 30, 35, and 50 were deposited, and will be labeled as SRO5, SRO10, SRO20, SRO25, SRO30, SRO35, and SRO50, respectively, for clarity. Also, two multilayer structures were fabricated. The multilayer is a stack of seven layers. One structure intercalates three SRO25 layers with four SRO5 layers, and the second one with four SRO10 layers. After deposition, all samples were thermally annealed at 1100°C for 3 h in nitrogen ambience to induce the silicon agglomeration.

166 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Thickness and refractive index of all samples, including multilayer structure, were determined using a null ellipsometer Gaertner L117 with a laser He–Ne of 632.8 nm wavelength. The PL emission spectra were obtained with a Fluoromax-3 spectrometer; all the films were excited with UV radiation (300 nm) and the luminescence was measured from 370 to 1000 nm with a resolution of 1 nm. Optical filters were used in order to guarantee the wavelength of excitation beam. CL measurements were performed using a luminoscope equipment model ELM2-144, 0.3-mA current and 5 kV were used. The luminescence spectra (PL and CL) were measured at

For electrical and electroluminescent studies, Metal-Insulator-Semiconductor (MIS) devices were fabricated, and we refer to them as light emitting capacitor (LEC). A ~250-nm-thick

by LPCVD. After a photolithography process step, square-shaped gates of 4-mm2 area were defined. The backside contact was formed with 1-µm thick aluminum layer by evaporation.

A source meter Keithley model 2400 was used to obtain current versus voltage (I-V) curves. EL spectra were obtained by biasing the device with a constant DC voltage. The light emitted was collected with an optical fiber located facing the Poly gate and connected to the Fluoromax

SRO is a multiphase material composed of silicon oxides of different stoichiometry and Si nanocrystals (Si-ncs). In XPS spectra of this material, each Si 2p core level band is composed

is well known and is easily distinguishable [12], but the peaks related with silicon suboxides cannot be distinguished unequivocally in a complex spectrum composed of different Si oxide species; they have been usually studied at Si/SiO2 interfaces [13–16]. In this way, a quantitative analysis of such highly convoluted spectra is not straightforward. The material can be conveniently considered as composed of SiO2, elemental Si and SiO*y* with 0 < *y* < 2. **Figure 5** shows an XPS spectrum of SRO with *R*0 = 1 as example. The three phases can be clearly

themselves at different energies. The position of the peaks corresponding to Si0

Finally, the devices were thermally annealed at 480°C in forming gas.

polycrystalline silicon (Poly) gate was deposited onto the SRO film surface

, Si1+, Si2+, Si3+, Si4+), which manifest

and Si4+ (SiO2)

room temperature.

semitransparent n+

3 spectrometer.

distinguished.

**4. Composition of SRO by LPCVD**

of bands originated in Si at different oxidation states (Si0

**Table 3** presents the compositions of SRO with different *R*0's [17, 18]. As can be observed, the amount of elemental Si and the amount of SiO2 monotonically increase and decrease, respec‐ tively, when decreasing the *R*0. SiO*y* oscillates around 43%, for *R*0's higher than 3. The oscillating amount could be understood in the sense that SiO*y* is in reality a combination of different stoichiometries, which vary in proportion depending on the *R*0.


**Table 3.** Composition of SRO with different *R*0s.

From **Table 3**, it is possible to make a fit of the monotonically varying data (Si and SiO2). For the fit is considered that *R*<sup>0</sup> = 0 means 100% elemental Si, and that 0% elemental Si is obtained by *R*0 = 100 (no Si-ncs are observed from *R*0 = 30 [17]). From the fitting curves, it is possible to calculate the curve for SiO*y* as 100 - %Si - %SiO2. All calculated curves are shown in **Figure 6**.

**Figure 6.** Calculated curves for %Si, %SiO2, and %SiO*y*, obtained by fitting measured data.

For R0's below 8, the amount of SiO*<sup>y</sup>* is larger than the amount of SiO2. This is caused by the high Si excess. At larger R0's, the amount of SiO2 is larger than the one of SiO*y*, nevertheless, the amounts are almost constant (vary linearly with *R*0, with a small slope) at *R*0's over 20, when the amount of elemental Si is close to zero. Following this tendency, the amount of SiO*<sup>y</sup>* is not zero even at *R*<sup>0</sup> = 100 (the proportion is 44% SiOy by 56% SiO2). This result implies that the oxides obtained by LPCVD are in a large percentage non-stoichiometric, even at large *R*0's. This nature of the oxides may mean a large number of defects, many of them being luminescent, as will be made clear in Section 6.

For our CVD system, one can write the chemical reaction as

$$\rm SiH\_4 + (2+x)\*\/ N\_2O \to SiO\_x + 2H\_2O + (2+x)\*\/ N\_2 \tag{5}$$

where

$$\text{SiO}\_x = a \ast \text{Si} + b \ast \text{SiO}\_y + c \ast \text{SiO}\_2 \tag{6}$$

with *a, b*, and *c* being the atomic proportions of the different phases (*a* + *b* + *c* = 1).

It is worthy to mention that certain amount of nitrogen is incorporated in SiO*x* during the deposition, but it may be of maximum 1%. Lower amounts of nitrogen are presented in samples of smaller *R*0's [19]. These amounts do not change the material structurally, but may enhance its luminescence [20, 21].

It is also important to know the form how elemental Si is present in the samples. Through transmission electron microscopy (TEM) studies, it has been possible to evidence Si-ncs in samples with *R*0's below 20. By larger *R*0's, Si is in amorphous state or dispersed in the oxide matrix. **Figure 7** shows a plot of the sizes of the Si-ncs versus *R*<sup>0</sup> [17, 22, 23]. As can be observed, the Si-nc size decreases almost linearly with the increasing of *R*0, for *R*0's above 3. SRO with *R*<sup>0</sup> = 1 is closer to semi-insulating polysilicon (SIPOS), presenting much bigger grains.

**Figure 7.** Si-nC sizes depending on *R*0.

**Figure 6.** Calculated curves for %Si, %SiO2, and %SiO*y*, obtained by fitting measured data.

168 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

luminescent, as will be made clear in Section 6.

where

its luminescence [20, 21].

For our CVD system, one can write the chemical reaction as

with *a, b*, and *c* being the atomic proportions of the different phases (*a* + *b* + *c* = 1).

It is worthy to mention that certain amount of nitrogen is incorporated in SiO*x* during the deposition, but it may be of maximum 1%. Lower amounts of nitrogen are presented in samples of smaller *R*0's [19]. These amounts do not change the material structurally, but may enhance

It is also important to know the form how elemental Si is present in the samples. Through transmission electron microscopy (TEM) studies, it has been possible to evidence Si-ncs in samples with *R*0's below 20. By larger *R*0's, Si is in amorphous state or dispersed in the oxide matrix. **Figure 7** shows a plot of the sizes of the Si-ncs versus *R*<sup>0</sup> [17, 22, 23]. As can be observed,

For R0's below 8, the amount of SiO*<sup>y</sup>* is larger than the amount of SiO2. This is caused by the high Si excess. At larger R0's, the amount of SiO2 is larger than the one of SiO*y*, nevertheless, the amounts are almost constant (vary linearly with *R*0, with a small slope) at *R*0's over 20, when the amount of elemental Si is close to zero. Following this tendency, the amount of SiO*<sup>y</sup>* is not zero even at *R*<sup>0</sup> = 100 (the proportion is 44% SiOy by 56% SiO2). This result implies that the oxides obtained by LPCVD are in a large percentage non-stoichiometric, even at large *R*0's. This nature of the oxides may mean a large number of defects, many of them being

42 2 2 ++\* ® + ++\* (2 ) 2 (2 ) *<sup>x</sup> SiH x N O SiO H O x N* (5)

*x y* = \* +\* +\* <sup>2</sup> *SiO a Si b SiO c SiO* (6)

### **5. Electrical characteristics**

**Figure 8** shows the current density (*J*) as a function of the electric field (*E*), which it is defined as the ratio of applied voltage (*V*) and the thickness of the SRO film (*t*SRO). This J-E behavior corresponds to LEC with a single layer of SRO. All LECs are forwardly biased (accumulation mode) considering the substrate as reference.

**Figure 8.** J-E curve of SRO30 and SRO20-based LECs. SRO films thermally annealed at 1100°C.

As we can see, the presence of defects including the Si-nps, either crystalline or amorphous, and their density in the SRO films affect clearly the current transport when they are used in MIS devices. LEC with SRO30 films show a high current state (HCS) at low electric fields, and then after the applied voltage increases, the current is switched to a low conduction state (LCS), as shown in **Figure 8**. The switching from the HCS to LCS shown by SRO30-based devices was observed by our group before and for both forward and reverse bias [24–28]. That effect was related to the annihilation of conductive paths created by adjacent stable Si-nps and unstable silicon nanoclusters (Si-ncls) through structural changes and by the possible creation of defects (breaking off Si-Si bonds) [24, 25, 27]. Recent studies regarding the same electrical switching in SRO films was observed and related with a conductive filament [29–32]. The conductance switching behavior observed in that SRO films was explained also by structural changes through an electroforming process. In fact, the structural changes in the conductive filament was analyzed by in situ imaging TEM analysis, showing that the conductance switch is related with a crystallization and an amorphization process of Si-nps that creates the conductive filament [30]. These observations are in agreement with our asseverations about the conduc‐ tance switching observed in our SRO30-based LECs [24].

In the HCS regime, current jumps and drops, which are observed independently of the temperature of annealing, have been related to the creation and annihilation of the preferential conductive paths and with the appearing or disappearing of electroluminescent spots (EL dots) on the LEC surface [24, 25, 27, 28]. A clear correlation between current jumps/drops and EL dots appearing/disappearing was observed [27]. Once the current fluctuations disappear, through an electrical annealing, the current behavior stabilizes, as reported in [24, 28].

On the other hand, the electrical behavior of most of LECs with SRO20 films does not show current fluctuations. This effect has been related with the presence of well-separated and crystalline silicon nanoparticles (or Si-ncs) and mainly on the density of Si-nps [28]. The Si-nps density estimated from energy-filtered transmission electron microscopy (EFTEM) images of SRO20 films thermally annealed at 1100°C is ~2.46 × 1012 cm−2, about twice the Si-nps density in SRO30 with ~1.1 × 1012 cm−2 [27]. Therefore, a uniform network of conductive paths becomes possible as the Si-nps density increases, allowing a uniform charge flow through the whole capacitor area. Meanwhile, as the Si-nps density decreases (SRO30 films), the distance between them increases reducing the amount of available paths, with a resulting set of discrete and preferential conductive paths within the oxide.

Basically, there are four main mechanisms known to contribute in the carrier transport through a Si-rich oxide layer, including the direct tunneling, Fowler Nordheim tunneling (F-N), Poole-Frenkel (P-F) and the trap-assisted tunneling (TAT) [33–37]. It has been found that the TAT conduction mechanism predominates in our SRO30-based LECs, where the trap energy (*ϕ*<sup>t</sup> ) was estimated at about 1.99 eV [28]. This implies that traps, which are placed at ~2 eV below the conduction band, would be assisting the tunneling event. On the other hand, the P-F tunneling was found as the charge transport in the SRO20-based LECs. A relative permittivity (*ε*r) value of 9.16 was obtained from the P-F fit, closer to the relative permittivity of silicon (*ε*Si = 11.9), similar to other reports [36]. Relatively high permittivity values are a good indication of the large amount of silicon present as Si-nps within our SRO20 films. Moreover, it was found that the Si-nc size obtained through a relation between size and permittivity (obtained by P-F fit) is very close to that measured by high-resolution transmission electron microscopy (HRTEM) [28].

### **6. Electro-optical characteristics**

### **6.1. Single layer**

As we can see, the presence of defects including the Si-nps, either crystalline or amorphous, and their density in the SRO films affect clearly the current transport when they are used in MIS devices. LEC with SRO30 films show a high current state (HCS) at low electric fields, and then after the applied voltage increases, the current is switched to a low conduction state (LCS), as shown in **Figure 8**. The switching from the HCS to LCS shown by SRO30-based devices was observed by our group before and for both forward and reverse bias [24–28]. That effect was related to the annihilation of conductive paths created by adjacent stable Si-nps and unstable silicon nanoclusters (Si-ncls) through structural changes and by the possible creation of defects (breaking off Si-Si bonds) [24, 25, 27]. Recent studies regarding the same electrical switching in SRO films was observed and related with a conductive filament [29–32]. The conductance switching behavior observed in that SRO films was explained also by structural changes through an electroforming process. In fact, the structural changes in the conductive filament was analyzed by in situ imaging TEM analysis, showing that the conductance switch is related with a crystallization and an amorphization process of Si-nps that creates the conductive filament [30]. These observations are in agreement with our asseverations about the conduc‐

170 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

In the HCS regime, current jumps and drops, which are observed independently of the temperature of annealing, have been related to the creation and annihilation of the preferential conductive paths and with the appearing or disappearing of electroluminescent spots (EL dots) on the LEC surface [24, 25, 27, 28]. A clear correlation between current jumps/drops and EL dots appearing/disappearing was observed [27]. Once the current fluctuations disappear, through an electrical annealing, the current behavior stabilizes, as reported in [24, 28].

On the other hand, the electrical behavior of most of LECs with SRO20 films does not show current fluctuations. This effect has been related with the presence of well-separated and crystalline silicon nanoparticles (or Si-ncs) and mainly on the density of Si-nps [28]. The Si-nps density estimated from energy-filtered transmission electron microscopy (EFTEM) images of SRO20 films thermally annealed at 1100°C is ~2.46 × 1012 cm−2, about twice the Si-nps density in SRO30 with ~1.1 × 1012 cm−2 [27]. Therefore, a uniform network of conductive paths becomes possible as the Si-nps density increases, allowing a uniform charge flow through the whole capacitor area. Meanwhile, as the Si-nps density decreases (SRO30 films), the distance between them increases reducing the amount of available paths, with a resulting set of discrete and

Basically, there are four main mechanisms known to contribute in the carrier transport through a Si-rich oxide layer, including the direct tunneling, Fowler Nordheim tunneling (F-N), Poole-Frenkel (P-F) and the trap-assisted tunneling (TAT) [33–37]. It has been found that the TAT conduction mechanism predominates in our SRO30-based LECs, where the trap energy (*ϕ*<sup>t</sup>

was estimated at about 1.99 eV [28]. This implies that traps, which are placed at ~2 eV below the conduction band, would be assisting the tunneling event. On the other hand, the P-F tunneling was found as the charge transport in the SRO20-based LECs. A relative permittivity (*ε*r) value of 9.16 was obtained from the P-F fit, closer to the relative permittivity of silicon (*ε*Si = 11.9), similar to other reports [36]. Relatively high permittivity values are a good indication of the large amount of silicon present as Si-nps within our SRO20 films. Moreover, it was found

)

tance switching observed in our SRO30-based LECs [24].

preferential conductive paths within the oxide.

### *6.1.1. Photoluminescence*

PL spectra of annealed films from *R*<sup>0</sup> = 5 to *R*<sup>0</sup> = 50 are depicted in **Figure 9**. After annealing, all SRO samples present a main emission from ~600 to 850 nm and a negligible emission for some samples from ~380 to 500 nm. The emission intensity increases when the silicon excess decrease until a maximum of *R*0 = 30 and then the PL intensity decreases until it practically disappear.

**Figure 9.** Photoluminescence of annealed SRO films with *R*0 from 5 to 50.

As can be observed, the PL emission exhibits a shape dependence on the silicon excess, which could indicate different emission mechanism. Because of this, the multi-Gaussian deconvolu‐ tion of PL spectra was performed for some annealed samples, and the set of band positions have been determined (**Figure 10**). Each spectrum can be well fitted to a superposition of three Gaussian distributions: a main band (1) and two shoulders (2 and 3). Fit peaks are centered at (1) 710–730, (2) 780–790, and (3) 820 nm with FWHM of (1) 50–60, (2) 20–29, and (3) 18 nm, respectively.

**Figure 10.** PL spectra and fits from SRO films with different silicon excess. Symbols are experimental data, lines are the Gaussian fits and dash lines are distributions.

Peak position and intensity vary according to the silicon excess, as shown in **Figure 11**. There is a blue wavelength shift for all components when the silicon excess decreases (except for *R*<sup>0</sup> = 5), see **Figure 11(a)**. The main contribution of the luminescence is the peak 1 that increases rapidly as *R*0 increases until *R*0 = 30 and then decreases, as can be seen in **Figure 11(b)**. While peaks 2 and 3 slightly increase when *R*0 increases, due to this, samples with *R*<sup>0</sup> < 25 shows a shoulder in the near infrared region (NIR). Since there are different components that change with silicon excess, it can be assumed that PL emission is related with at least three different types of emission centers (or emission mechanism).

**Figure 11.** (a) Position and (b) intensity of the fit distributions for different silicon excess.

### *6.1.2. Cathodoluminescence*

have been determined (**Figure 10**). Each spectrum can be well fitted to a superposition of three Gaussian distributions: a main band (1) and two shoulders (2 and 3). Fit peaks are centered at (1) 710–730, (2) 780–790, and (3) 820 nm with FWHM of (1) 50–60, (2) 20–29, and (3) 18 nm,

172 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 10.** PL spectra and fits from SRO films with different silicon excess. Symbols are experimental data, lines are the

Peak position and intensity vary according to the silicon excess, as shown in **Figure 11**. There is a blue wavelength shift for all components when the silicon excess decreases (except for *R*<sup>0</sup> = 5), see **Figure 11(a)**. The main contribution of the luminescence is the peak 1 that increases rapidly as *R*0 increases until *R*0 = 30 and then decreases, as can be seen in **Figure 11(b)**. While peaks 2 and 3 slightly increase when *R*0 increases, due to this, samples with *R*<sup>0</sup> < 25 shows a shoulder in the near infrared region (NIR). Since there are different components that change with silicon excess, it can be assumed that PL emission is related with at least three different

respectively.

Gaussian fits and dash lines are distributions.

types of emission centers (or emission mechanism).

**Figure 11.** (a) Position and (b) intensity of the fit distributions for different silicon excess.

CL spectra from SRO films with different silicon excess are depicted in **Figure 12(a)**. The CL spectra of SRO with thermal treatment consist of a broad emission in the visible and NIR from ~400 to 850 nm. After annealing, intensity of the blue band at ~460 nm increases with increasing the *R*0. On the other hand, the intensity of the red-NIR CL band seems to have a maximum for *R*0 = 20.

**Figure 12.** (a) Cathodoluminescence spectra from SRO films with different silicon excess. (b) Gaussian fit of CL experi‐ mental spectrum, the best fit requires four distributions for *R*0 = 30 and six distributions for *R*0 = 20.

As CL emission has asymmetrical shape for all SRO samples, it can be assumed that CL emission is also due to different luminescent centers. Hence, multi-Gaussian deconvolution of CL spectra was also obtained, shown in **Figure 12(b)**. The best fit of CL spectra requires four and six components for *R*0 = 30 and 20, respectively. Peaks (or distributions) were obtained at about 460, 522, 643, and 714 nm for *R*0 = 30 and 447, 541, 645, 714, 780, and 823 nm for *R*0 = 20. Peaks obtained at 714, 780, and 823 nm in CL are centered in the same position than distribu‐ tions obtained from PL spectrum in *R*0 = 20. Furthermore, peak at 714 nm was obtained for Gaussian fit, in PL and CL in *R*0 = 30. Then, the red emission of the CL emission can be ascribed to the same PL emissive centers. PL distributions in higher wavelength are not observed in CL due to either destruction of the emissive centers or inefficient emission from low-energy emissive centers [38]. The latter one could occur increasingly because cathode-excited electrons acquire so high energy that they arrive to the higher emissive positions where they emit in the blue region (higher energy); however, almost none of the excited electrons reach that with lower energy; then, the red emission is not likely to occur in CL. Therefore, there could be several different kinds of emission traps located at different energy levels in SRO.

Depending on the emission wavelength, multiple luminescence centers have been reported in SiO2 films. Luminescent emission at 460 nm (2.7 eV), 520 nm (2.4 eV) and 650 nm (1.9 eV) nm are mainly related to defects such as oxygen deficiency-related centers (ODC) or oxygen vacancies [39–41], *E*'δ defect or peroxide radical [42] and non-bridging oxygen hole centers (NBOHC) [40, 41, 43], respectively. Since CL and PL measurements have shown luminescent peaks (or distributions) close to those wavelengths, such defects could be inside the SRO films.

### *6.1.3. Electroluminescence*

**Figure 13** shows the electroluminescence spectra from the SRO-based LECs. Blue electrolu‐ minescence is observed in the SRO30 film, as observed in **Figure 13(a)**. Nevertheless, this blue EL in whole area of LECs is obtained only after the current drop. The main EL peak remains at 468 nm even for different thermal annealing temperatures [28]. A long spectral shift, blue shift, of almost ~227 nm has been observed between the EL and PL band of the SRO30 films. Devices with SRO20 films emit a broad EL spectrum in the red region (713 nm), as observed in **Figure 13(b)**. An additional EL peak of low intensity is also observed at 468 nm. There exists also a blue shift of the EL spectra with respect to PL spectrum in SRO20 films. Nevertheless, both EL and PL spectra in SRO20 films appear in the red region, which could indicate that the same luminescent centers are involved. Images of the blue and red LECs are shown in the insets of the **Figure 13**. As we can see, intense EL is emitted in the whole area of the LEC devices.

**Figure 13.** EL spectra of (a) SRO30 and (b) SRO20 based LECs biased at different electric fields. Insets show images of each SRO-based LECs.

The spectral shift between PL and EL has been reported before and it has been explained according to three different mechanisms [35, 44, 45]. Our experimental results have suggested that the red EL observed in SRO20 films can be related with the combination of some surface defect on the Si-ncs, while the blue EL in SRO30 devices is consistent with the defect emission which could be intrinsically present or generated by electric field within the SRO matrix [28].

### **6.2. Multilayer**

### *6.2.1. Photoluminescence*

Multilayer structures were fabricated in order to improve the optical properties of the SRO films. Two samples were obtained, one of those is a combination of low silicon excess (*R*0 = 25) and high silicon excess (*R*0 = 5) and the second one is a combination of *R*0 = 25 and *R*0 = 10.

PL spectra of annealed multilayers and single layer are shown in **Figure 14(a)**. As can be seen the intensity emission is improved in the multilayer samples, where multilayer SRO10/SRO25 (ML-10/25) is the most intense. In order to obtain the components of every layer, the multi-Gaussian deconvolution of PL spectra was also obtained. **Figure 14(b)** and **(c)** shows the position and the intensity of the three peaks obtained from the Gaussian deconvolution. There is a blue-shift wavelength for the three peaks that can be due to the participation of high silicon excess (*R*0 = 5 or 10) on the PL. The intensity improvement could be due to the three components together, peak 1 is from *R*0 = 25 and the improvement of peaks 2 and 3 comes from high silicon excess (*R*0 = 5 or 10).

**Figure 14.** (a) Photoluminescence of annealed multilayer SRO films (*R*<sup>0</sup> = 10/25 and *R*<sup>0</sup> = 5/25) and a single layer of *R*0 = 25. (b) Position and (c) intensity of the fit distribution for single layer (*R*0 = 5, 10, and 25) and multilayers.

### *6.2.2. Electroluminescence*

*6.1.3. Electroluminescence*

each SRO-based LECs.

**6.2. Multilayer**

*6.2.1. Photoluminescence*

**Figure 13** shows the electroluminescence spectra from the SRO-based LECs. Blue electrolu‐ minescence is observed in the SRO30 film, as observed in **Figure 13(a)**. Nevertheless, this blue EL in whole area of LECs is obtained only after the current drop. The main EL peak remains at 468 nm even for different thermal annealing temperatures [28]. A long spectral shift, blue shift, of almost ~227 nm has been observed between the EL and PL band of the SRO30 films. Devices with SRO20 films emit a broad EL spectrum in the red region (713 nm), as observed in **Figure 13(b)**. An additional EL peak of low intensity is also observed at 468 nm. There exists also a blue shift of the EL spectra with respect to PL spectrum in SRO20 films. Nevertheless, both EL and PL spectra in SRO20 films appear in the red region, which could indicate that the same luminescent centers are involved. Images of the blue and red LECs are shown in the insets of the **Figure 13**. As we can see, intense EL is emitted in the whole area of the LEC devices.

174 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 13.** EL spectra of (a) SRO30 and (b) SRO20 based LECs biased at different electric fields. Insets show images of

The spectral shift between PL and EL has been reported before and it has been explained according to three different mechanisms [35, 44, 45]. Our experimental results have suggested that the red EL observed in SRO20 films can be related with the combination of some surface defect on the Si-ncs, while the blue EL in SRO30 devices is consistent with the defect emission which could be intrinsically present or generated by electric field within the SRO matrix [28].

Multilayer structures were fabricated in order to improve the optical properties of the SRO films. Two samples were obtained, one of those is a combination of low silicon excess (*R*0 = 25) and high silicon excess (*R*0 = 5) and the second one is a combination of *R*0 = 25 and *R*0 = 10.

**Figure 15** shows a scheme of the multilayered structure fabricated with its dimensional characteristics (left side), and a TEM image of the structure, which exposes the layers com‐ posing the SRO multilayer (right side). The goal of this structure is to improve the electrooptical properties of the LECs. In this multilayer, the luminescent properties of three layers with low silicon excess (SRO25) are combined with four conducting layers (SRO5).

**Figure 15.** Scheme of a light emitter capacitor with a multilayered SRO films (left side) and a TEM image of the SRO nanometric multilayer (right side).

The electroluminescence in this ML-SRO-based LEC is observed at forwardly bias considering the substrate as reference. A broad band with the main peak at about 600 nm is observed with *E* = 3.7 MV/cm, as we can see in **Figure 16**. As the voltage increases, the spectrum is divided in two. One peak at 680 nm and other at 450 nm, the blue peak increases to higher intensity than the red one. Apparently, SRO5 layers increase the conductivity across the structure and SRO25 layers produces the emission. The spectrum behavior shown in our samples have been observed in other reports [26].

However, the electric field needed to turn on the emission on an ML-SRO-based LEC is lower than an LEC of single layer (see **Figures 13** and **16**). This proves that the electro-optical properties of a ML-SRO-based LEC are improved, thereby the conductivity of the structure is increased by layers of high silicon excess, and luminescence response is conserved using layers of high *R*0.

**Figure 16.** EL spectra of a multilayered SRO5/SRO25-based LEC under different electric fields.

### **7. Conclusion**

In this chapter, details of a homemade hot-wall LPCVD system were presented. Also, impor‐ tant aspects of how to obtain SRO in a reliable and repetitive form were addressed. We show that in our system, it is possible to obtain single layers with variable silicon excess, and also good quality multilayered structures of nanometric layers. The structural, electrical, and luminescent characteristics of single- and multilayered structures were reviewed and dis‐ cussed.

### **Acknowledgements**

The electroluminescence in this ML-SRO-based LEC is observed at forwardly bias considering the substrate as reference. A broad band with the main peak at about 600 nm is observed with *E* = 3.7 MV/cm, as we can see in **Figure 16**. As the voltage increases, the spectrum is divided in two. One peak at 680 nm and other at 450 nm, the blue peak increases to higher intensity than the red one. Apparently, SRO5 layers increase the conductivity across the structure and SRO25 layers produces the emission. The spectrum behavior shown in our samples have been

176 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

However, the electric field needed to turn on the emission on an ML-SRO-based LEC is lower than an LEC of single layer (see **Figures 13** and **16**). This proves that the electro-optical properties of a ML-SRO-based LEC are improved, thereby the conductivity of the structure is increased by layers of high silicon excess, and luminescence response is conserved using layers

**Figure 16.** EL spectra of a multilayered SRO5/SRO25-based LEC under different electric fields.

In this chapter, details of a homemade hot-wall LPCVD system were presented. Also, impor‐ tant aspects of how to obtain SRO in a reliable and repetitive form were addressed. We show that in our system, it is possible to obtain single layers with variable silicon excess, and also good quality multilayered structures of nanometric layers. The structural, electrical, and luminescent characteristics of single- and multilayered structures were reviewed and dis‐

observed in other reports [26].

of high *R*0.

**7. Conclusion**

cussed.

Authors recognize the financial support of the CONACYT, particularly, J. Alarcón-Salazar for his grant with number 353251. Also, authors thank the microelectronic laboratory technicians Pablo Alarcon, Victor Aca, and Armando Hernández for their help during the fabrication process.

### **Author details**


\*Address all correspondence to: j.alarcon.sal@gmail.com

1 Electronics Department, National Institute of Astrophysics Optics and Electronics (INAOE), Luis Enrique Erro 1, Tonantzintla, Puebla, México

2 Professor CONACYT – IICO-UASLP, Karakorum, San Luis Potosí, México

3 Institute of Physics, Autonomous University of Puebla, México

4 Advanced Materials Research Center (CIMAV), Monterrey, Apodaca, Nuevo León, México

### **References**


the origin of light emission in silicon rich oxide obtained by low-pressure chemical vapor deposition. J. Nanomater. 2012;2012:890701. DOI:10.1155/2012/890701


high silicon content prepared by low pressure chemical vapor deposition. Thin Solid Films. 2011;519:8030–8036. DOI:10.1016/j.tsf.2011.06.020

[18] R. López-Estopier, M. Aceves-Mijares, and C. Falcony. Cathodo- and photo-lumines‐ cence of silicon rich oxide films obtained by LPCVD, cathodoluminescence. In: Naoki Yamamoto, editor. Chathodoluminescence. InTech; Shanghai China; 2012. DOI: 10.5772/34888

the origin of light emission in silicon rich oxide obtained by low-pressure chemical

[5] A. A. González-Fernández, J. Juvert, M. Aceves-Mijares, C. Domínguez. Monolithic integration of a silicon-based photonic transceiver in a CMOS process. IEEE Photonics

[6] A. Morales, J. Barreto, C. Domínguez, M. Riera, M. Aceves, and J. Carrillo. Comparative study between silicon-rich oxide films obtained by LPCVD and PECVD. Phys. E Low

[7] D. Dong, E. A. Irene, and D. R. Young. Preparation and some properties of chemically vapor-deposited Si-rich SiO and SigN, films. J. Electrochem. Soc. 1978;125(5):819–823.

[8] M. Aceves, L. A. Hernandez, and R. Murphy. Applying statistics to find the causes of variability of aluminum deposition: a case study. IEEE Trans. Semicond. Manuf.

[9] H. O. Pierson. Handbook of chemical vapor deposition (CVD): Principles, technology,

[10] Hwaiyu Geng. Chemical vapor deposition. In: Hwaiyu Geng, editor. Semiconductor manufacturing handbook. 1st ed. New York, USA: McGraw-Hill Handbooks; 2005.

[11] Y.-P. Zhao, J. T. Drotar, G.-C. Wang, and T.-M. Lu. Morphology transition during lowpressure chemical vapor deposition. Phys. Rev. Lett. 2001;87(13)1–4. DOI:10.1103/

[12] F. Verpoort, P. Persoon, L. Fiermans, G. Dedoncker, and L. Verdonck. SiO2/ Si(100)model support with AES and XPS in combination with MLCFA. J. Chem. Soc.,

[13] G. Hollinger and F. J. Himpsel. Probing the transition layer at the SiO2‐Si interface using core level photoemission. Appl. Phys. Lett. 1984;44(1):93–95. DOI:10.1063/1.94565 [14] F. J. Grundthaner, P. J. Grundthaner, R. P. Vasquez, B. F. Lewis, and J. Maserjian. Highresolution X-ray photoelectron spectroscopy as a probe of local atomic structure: Application to amorphous SiO2 and the Si-SiO2 interface. Phys. Rev. Lett. 1979;43(22):

[15] P. J. Grundthaner, M. H. Hecht, F. J. Grundthaner, and N. M. Johnson. The localization and crystallographic dependence of Si suboxide species at the SiO2/Si interface. J. Appl.

[16] K. Hirose, H. Nohira, K. Azuma, and T. Hattori. Photoelectron spectroscopy studies of SiO2/Si interfaces. Prog. Surf. Sci. 2007;82(1):3. DOI:10.1016/j.progsurf.2006.10.001 [17] E. Quiroga-González, W. Bensch, M. Aceves-Mijares, Z. Yu, R. López-Estopier, and K. Monfil-Leyva. On the photoluminescence of multilayer arrays of silicon rich oxide with

and applications. 2nd ed. New Jersey, USA: Noyes Publications; 1999.

Faraday Trans. 1997;93(19):3555. DOI:10.1039/A702279D

Dimens. Syst. Nanostruct. 2007;38(1–2):54–58. DOI:10.1016/j.physe.2006.12.056

vapor deposition. J. Nanomater. 2012;2012:890701. DOI:10.1155/2012/890701

178 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

J. 2016;8(1):7900213. DOI:10.1109/JPHOT.2015.2505144

DOI:10.1149/1.2131555

PhysRevLett.87.136102

1992;5(2):165–167. DOI:10.1109/66.136280

1683. DOI:10.1103/PhysRevLett.43.1683

Phys. 1987;61(2):629. DOI:10.1063/1.338215


luminescence in silicon nanoparticles based light emitting capacitors. Appl. Phys. Lett. 2011;99(17):1711102. DOI:10.1063/1.3655997


[39] M. Cervera, M. J. Hernández, P. Rodríguez, J. Piqueras, M. Avella, M. A. González, and J. Jimenez. Blue-cathodoluminescent layers synthesis by high-dose N+, C+ and B+ SiO2 implantation. J. Lumin. 2006;117(1):95–100. DOI:10.1016/j.jlumin.2005.03.015

luminescence in silicon nanoparticles based light emitting capacitors. Appl. Phys. Lett.

[28] L. Palacios Huerta, S. A. Cabañas Tay, J.-A. Luna López, M. Aceves, A. Coyopol, and A. Morales-Sánchez. Effect of the structure on luminescent characteristics of SRO-based light emitting capacitors. Nanotechnology. 2015;26(39):395202. DOI:

[29] J. Yao, Z. Sun, L. Zhong, D. Natelson, and J. M. Tour. Resistive switches and memories

[30] J. Yao, L. Zhong, D. Natelson, and J. M. Tour. In situ imaging of the conducting filament in a silicon oxide resistive switch. Sci. Rep. 2012;2(242):1-5. DOI:10.1038/srep00242 [31] A. Mehonic, A. Vrajitoarea, S. Cueff, S. Hudziak, H. Howe, C. Labbé, R. Rizk, M. Pepper, and A. J. Kenyon. Quantum conductance in silicon oxide resistive memory devices. Sci.

[32] A. Mehonic, S. Cueff, M. Wojdak, S. Hudziak, C. Labbé, R. Rizk, and A. Kenyon. Electrically tailored resistance switching in silicon oxide. Nanotechnology. 2012;23(45):

[33] B. H. Lai, C. H. Cheng, and G. R. Lin. Multicolor ITO/SiOx/P-Si/Al light emitting diodes with improved emission efficiency by small Si quantum dots. IEEE J. Quantum

[34] S. Cueff, S. Labbé, O. Jambois, Y. Berencén, A. J. Kenyon, B. Garrido, and R. Rizk. Structural factors impacting carrier transport and electroluminescence from Si nano‐ cluster-sensitized Er ions. Opt. Express. 2012;20(20):22490. DOI:10.1364/OE.20.022490

[35] B. H. Lai, C. H. Cheng, and G. R. Lin. Electroluminescent wavelength shift of Si-rich SiOx based blue and green MOSLEDs induced by O/Si composition Si-QD size

[36] J. M. Ramírez, Y. Berencén, L. López-Conesa, J. M. Rebled, F. Peiró, and B. Garrido. Carrier transport and electroluminescence efficiency of erbium-doped silicon nancrys‐

[37] R. Perera, A. Ikeda, R. Hattori, and Y. Kuroki. Trap assisted leakage current conduction in thin silicon oxynitride films grown by rapid thermal oxidation combined microwave excited plasma nitridation. Microelectron. Eng. 2003;65(4):357–370. DOI:10.1016/

[38] A. N. Trukhin., H. J. Fitting, T. Barfels, and V. Czarnowski. Cathodoluminescence and IR absorption of oxygen deficient silica – influence of hydrogen treatment. J. Non-Cryst.

Solids. 1999;260(1–2):132–140. DOI:10.1016/S0022-3093(99)00558-X

variations. Opt. Mater. Express. 2013;3(2):166. DOI:10.1364/OME.3.000166

tals superlattices. Appl. Phys. Lett. 2013;103:081102. DOI:10.1063/1.4818758

from silicon oxide. Nano Lett. 2010;10(10):4105. DOI:10.1021/nl102255r

180 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

2011;99(17):1711102. DOI:10.1063/1.3655997

Rep. 2013;3:2708. DOI:10.1038/srep02708

455201. DOI:10.1088/0957-4484/23/45/455201

S0167-9317(02)01025-0

Electron. 2011;47(5):698. DOI:10.1109/JQE.2011.2109699

10.1088/0957-4484/26/39/395202


### **High‐Density Plasma‐Enhanced Chemical Vapor Deposition of Si‐Based Materials for Solar Cell Applications**

H. P. Zhou, S. Xu and S. Q. Xiao

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/63529

### **Abstract**

High‐quality and low‐cost fabrication of Si‐based materials, in which many fundamen‐ tal and technology problems still remain, have attracted tremendous interests due to their wide applications in solar cell area. Low‐frequency inductively coupled plasma (LFICP) provides a new and competitive solution, thanks to its inherent advantages of high‐density plasma, low sheath potential, and low electron temperature, etc. The plasma characteristic‐dependent microstructures, optical and electronic properties of the LFICP CVD‐based hydrogenated amorphous/microcrystalline silicon and silicon oxides are systematically studied. Remote‐LFICP combing the high‐density plasma nature of ICP and mild ion bombardment on growing surface in remote plasma allows the deposition of high‐quality Si‐based materials providing excellent c‐Si surface passivation. The mechanism of surface passivation by LFICP CVD Si‐based materials, interaction between plasma species and growing surface are analyzed in terms of the plasma properties. These results pave the way for LFICP CVD utilization in Si‐based high‐efficiency and low‐cost solar cell fabrication.

**Keywords:** LFICP CVD, remote‐plasma, Si‐based materials, solar cells, passivation

### **1. Introduction**

Plasma‐enhanced chemical vapor deposition (PECVD) is an excellent alternative for deposit‐ ing thin films at lower temperature than those utilized in a conventional CVD process. Plasma generated by a glow discharge absorbs the electrical energy, and hence the neutral gases were dissociated into reactive radicals, ion, neutral atoms and molecules, and other highly excited

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

species through the electron impact instead of conventional thermal energy. These atomic and molecular fragments interact with a substrate and, depending on the nature of these interac‐ tions, either etching or deposition processes occur at the substrate. Therefore, the substrate can be maintained at a lower temperature keeping the reactions for the thin‐film deposition. These salient advantages of PECVD, i.e., low temperature nature,together with the high quality ofthe deposited films make PECVD be an important route for the fabrication of Si‐based materials, which have been the dominant material system for the present solar cell research and commer‐ cial market. PECVD‐based silicon nitride (SiNx:H) with the feedstock gases of SiH4 and NH3 act as the passivation and antireflection layer of the bulk Si solar cells. Fabrication of Si thin‐film solar cells consisting of the stack of p‐, i‐, and n‐type hydrogenated amorphous/microcrystal‐ line Si (a‐/µc‐Si:H) films also relies on PECVD technology. The interface quality and energy conversion efficiency of HIT (heterojunction with intrinsic thin layer) solar cell with the record efficiency of 25.6% [1] closely depends on the PECVD process because the symmetric intrinsic passivationlayersanddopedlayers(frontemitterandbacksurfacefieldlayers)arealsoprepared by PECVD technology [2].

Although PECVD‐associated Si‐based materials are dominant over other materials in present solar cell market, there are a variety of challenges due to the nature of the utilized PECVD method. The common PECVD uses the capacitively coupled plasma (CCP) between parallel plates driven by a 13.56 MHz RF power generator. The ionization degree of the neutral gas is at a low level of ∼10−6–10−4 because of the low electron density of ∼10<sup>9</sup> –1010/cm<sup>3</sup> in a CCP discharge mode. In principle, this means a low deposition rate of the targeted materials. As a consequence, the deposition rate of PECVD‐based a‐/µc‐Si:H is about 0.5 nm/s [3], which is not desirable for cost‐efficient, thin‐film solar cell productions. Furthermore, the intensive coupling of the plasma density and ion energy in CCP complicates the deposition of high‐ quality thin films and the achievement of high‐efficiency solar cells because strong ion bombardment would result in excessive defect formation in the films and on the growing surface/interface, and hence significant deterioration of the energy conversion efficiency of the solar cells [4]. These disadvantages of PECVD can be partially remedied by using very high‐ frequency (VHF) PECVD, which can effectively decouple the plasma density and electron density due to the increasing electron input power fraction of the total input power with frequency [5, 6] and enhance the deposition rate. However, there is not any report on the higher efficiency of the VHF‐PECVD‐based solar cells [7].

High‐density plasma represented by inductively‐coupled plasma (ICP) has increasingly attracted more interests due to its wide application areas such as microelectronics, display, and solar cells, etc. The prominent advantage of high density of ICP makes it very appropriate for the deposition of Si‐based materials (a‐/µc‐Si:H, silicon carbide, silicon oxide, and silicon nitride), which are widely used in solar cells. Although there are numerous reports on the ICP‐ based deposition of these materials, the as‐deposited Si‐based materials with low defect density and excellent interface features are few, not to mention the involved high‐efficiency solar cells. We have been making many efforts on low‐frequency (460 kHz) ICP (LFICP) excitation, characterization, and application in nano‐fabrication [8] and a wide range of thin films' synthesis [9, 10]. A discharging photograph of the LFICP facility and its schematic structure are shown in **Figure 1(a)** and **1(b)**, respectively. The power of the low‐frequency power generator is coupled into the reactive chamber sealed by a quartz window via a matching/tuning unit to sustain the reactive gases discharge.

**Figure 1.** Discharging photograph of the LFICP facility (a), and its schematic structure (b).

species through the electron impact instead of conventional thermal energy. These atomic and molecular fragments interact with a substrate and, depending on the nature of these interac‐ tions, either etching or deposition processes occur at the substrate. Therefore, the substrate can be maintained at a lower temperature keeping the reactions for the thin‐film deposition. These salient advantages of PECVD, i.e., low temperature nature,together with the high quality ofthe deposited films make PECVD be an important route for the fabrication of Si‐based materials, which have been the dominant material system for the present solar cell research and commer‐ cial market. PECVD‐based silicon nitride (SiNx:H) with the feedstock gases of SiH4 and NH3 act as the passivation and antireflection layer of the bulk Si solar cells. Fabrication of Si thin‐film solar cells consisting of the stack of p‐, i‐, and n‐type hydrogenated amorphous/microcrystal‐ line Si (a‐/µc‐Si:H) films also relies on PECVD technology. The interface quality and energy conversion efficiency of HIT (heterojunction with intrinsic thin layer) solar cell with the record efficiency of 25.6% [1] closely depends on the PECVD process because the symmetric intrinsic passivationlayersanddopedlayers(frontemitterandbacksurfacefieldlayers)arealsoprepared

184 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Although PECVD‐associated Si‐based materials are dominant over other materials in present solar cell market, there are a variety of challenges due to the nature of the utilized PECVD method. The common PECVD uses the capacitively coupled plasma (CCP) between parallel plates driven by a 13.56 MHz RF power generator. The ionization degree of the neutral gas is

discharge mode. In principle, this means a low deposition rate of the targeted materials. As a consequence, the deposition rate of PECVD‐based a‐/µc‐Si:H is about 0.5 nm/s [3], which is not desirable for cost‐efficient, thin‐film solar cell productions. Furthermore, the intensive coupling of the plasma density and ion energy in CCP complicates the deposition of high‐ quality thin films and the achievement of high‐efficiency solar cells because strong ion bombardment would result in excessive defect formation in the films and on the growing surface/interface, and hence significant deterioration of the energy conversion efficiency of the solar cells [4]. These disadvantages of PECVD can be partially remedied by using very high‐ frequency (VHF) PECVD, which can effectively decouple the plasma density and electron density due to the increasing electron input power fraction of the total input power with frequency [5, 6] and enhance the deposition rate. However, there is not any report on the higher

High‐density plasma represented by inductively‐coupled plasma (ICP) has increasingly attracted more interests due to its wide application areas such as microelectronics, display, and solar cells, etc. The prominent advantage of high density of ICP makes it very appropriate for the deposition of Si‐based materials (a‐/µc‐Si:H, silicon carbide, silicon oxide, and silicon nitride), which are widely used in solar cells. Although there are numerous reports on the ICP‐ based deposition of these materials, the as‐deposited Si‐based materials with low defect density and excellent interface features are few, not to mention the involved high‐efficiency solar cells. We have been making many efforts on low‐frequency (460 kHz) ICP (LFICP) excitation, characterization, and application in nano‐fabrication [8] and a wide range of thin films' synthesis [9, 10]. A discharging photograph of the LFICP facility and its schematic

–1010/cm<sup>3</sup>

in a CCP

at a low level of ∼10−6–10−4 because of the low electron density of ∼10<sup>9</sup>

efficiency of the VHF‐PECVD‐based solar cells [7].

by PECVD technology [2].

In comparison with the conventional CCP source, the LFICP source has several prominent merits as follows [11, 12]: (i) high densities of the plasma species; in particular, the electron number density *n*0 can reach 1013 cm−3 in the pressure range of a few pascal; (ii) low plasma sheath potentials of several or tens of volts, which are beneficial for the reduction in ion bombardment on the deposited films; (iii) low electron temperatures of a few electron volts in a broad range of discharge conditions; and (iv) excellent uniformity of the plasma parameters in the radical and axial directions. By using this novel plasma generator, Si‐based materials have been fabricated with a high‐deposition rate at low temperature. Hydrogenated, amor‐ phous silicon carbide [13] (a‐SiC*x*:H, presently used as transparent doped layers in thin film and heterojunction solar cells) with a wide carbon component *x* (0.09–0.71) has been attained by adjusting the feedstock gas ratio of CH4/SiH<sup>4</sup> dissociated by LFICP generator. Furthermore, low‐temperature (500 °C) deposition of nanocrystalline SiC (nanocrystal size of about 6 nm) has been developed by using LFICP [14]. Low‐temperature (100–200 °C) and high‐rate (>1 nm/ S) deposition of nc‐Si in the direct plasma region of LFICP have been realized [15]. In our previous works [16, 17], it was also found that the severe etching on silicon surface directly exposed to the high‐density hydrogen containing plasma drastically changes the morphology and microstructures of the surface. As a consequence, it is very easy to attain µc‐/nc‐Si:H with a high deposition rate. However, this high‐density ion region is not suitable for solar cell fabrications where the defect density in the silicon films and on the interface need to be elaborately minimized in terms of the plasma characteristics. Apparently, the severe ion bombardment in a direct ICP would introduce additional defects reducing the photovoltaic performance of the solar cells. Remote‐LFICP [18] is a feasible solution to avoid the ion bombardment damage on the surface through separating the plasma generation region and deposition region. High‐density plasma of hydrogen gas is generated and mainly confined in the region near the antenna, i.e., the plasma generation region. The energetic species transport into the growth region and react with the injected SiH4 neutral gas forming a variety of precursor radicals for the deposition of high‐quality thin film at low temperatures.

High‐quality Si‐based materials have been attained by means of remote‐LFICP:


The aim of this work is to systematically report the high‐density plasma‐enhanced chemical vapor deposition of device‐grade Si‐based materials by means of LFICP and its solar cell applications. The as‐deposited Si‐based materials will be studied in combination with the LFICP characteristics. Their soar cell applications mainly focus on the Si surface passivation using remote‐LFICP‐CVD‐based Si‐based materials. This article is organized as follows. Section 2 makes a brief introduction on a‐/µc‐Si:H and then focuses on the processing param‐ eters dependent microstructures and properties of the LFICP‐CVD intrinsic and doped a‐/µc‐ Si:H thin films. Section 3 focuses on the applications of LFICP‐CVD a‐/µc‐Si:H for c‐Si surface passivation. The influence of the plasma characteristics on the properties of LFICP‐based materials and growing surface properties are intensively investigated.

### **2. a-/μc-Si:H thin films**

### **2.1. a-Si:H**

Amorphous silicon (a‐Si) is a solid‐state material made of silicon atoms which are arranged in a short‐range order. With the absence of the constraints of periodicity in elemental amorphous silicon, it is difficult for each silicon atom to be linked up with four others. Compared with crystalline silicon (c‐Si), the average bond angles between nearest‐neighboring atoms are distorted. Some bonds are even broken and result in so‐called "dangling bonds". These two main defects of the lattice of a‐Si:H, i.e., bond distortion and dangling bonds, create large densities of allowed states within the forbidden band gap in such a way that the material cannot be effectively doped N‐ or P‐type and is not electronically useful. However, if the substance is hydrogenated, the hydrogen atoms saturate the dangling bonds and reduce the density of states within the forbidden gap. The bond distortion results in band tails near the valence and conduction bands. In these band tails, the electrons (or holes) are localized in space and do not directly participate in the electronic transport. The deep states near the middle of the band gap results from the non‐passivated dangling bonds. These defect centers are "amphoteric"—there are three charge states, i.e., +e, 0, and –e charges, leading to two levels (transitions between the 0/+ and −/0 charge states). The dangling bonds act as recombination centers for free electrons and holes.

### **2.2. μc‐Si:H**

High‐quality Si‐based materials have been attained by means of remote‐LFICP:

186 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

ity Si wafer (surface recombination of 70 cm/s obtained) [22].

36 cm/s for low resistivity n‐type Si wafer) [21].

materials and growing surface properties are intensively investigated.

**2. a-/μc-Si:H thin films**

**2.1. a-Si:H**

and holes.

**i.** As‐deposited a‐Si:H exhibits competitive surface passivation ability on crystalline silicon, thanks to the minimized inner and interface defect density [19];

**ii.** Silicon nitride with controlled bandgap due to the excellent dissociation ability of

**iii.** Remote‐LFICP‐based a‐SiOx:H provides effective surface passivation of low resistiv‐

The aim of this work is to systematically report the high‐density plasma‐enhanced chemical vapor deposition of device‐grade Si‐based materials by means of LFICP and its solar cell applications. The as‐deposited Si‐based materials will be studied in combination with the LFICP characteristics. Their soar cell applications mainly focus on the Si surface passivation using remote‐LFICP‐CVD‐based Si‐based materials. This article is organized as follows. Section 2 makes a brief introduction on a‐/µc‐Si:H and then focuses on the processing param‐ eters dependent microstructures and properties of the LFICP‐CVD intrinsic and doped a‐/µc‐ Si:H thin films. Section 3 focuses on the applications of LFICP‐CVD a‐/µc‐Si:H for c‐Si surface passivation. The influence of the plasma characteristics on the properties of LFICP‐based

Amorphous silicon (a‐Si) is a solid‐state material made of silicon atoms which are arranged in a short‐range order. With the absence of the constraints of periodicity in elemental amorphous silicon, it is difficult for each silicon atom to be linked up with four others. Compared with crystalline silicon (c‐Si), the average bond angles between nearest‐neighboring atoms are distorted. Some bonds are even broken and result in so‐called "dangling bonds". These two main defects of the lattice of a‐Si:H, i.e., bond distortion and dangling bonds, create large densities of allowed states within the forbidden band gap in such a way that the material cannot be effectively doped N‐ or P‐type and is not electronically useful. However, if the substance is hydrogenated, the hydrogen atoms saturate the dangling bonds and reduce the density of states within the forbidden gap. The bond distortion results in band tails near the valence and conduction bands. In these band tails, the electrons (or holes) are localized in space and do not directly participate in the electronic transport. The deep states near the middle of the band gap results from the non‐passivated dangling bonds. These defect centers are "amphoteric"—there are three charge states, i.e., +e, 0, and –e charges, leading to two levels (transitions between the 0/+ and −/0 charge states). The dangling bonds act as recombination centers for free electrons

LFICP facility was realized [20]. Meanwhile, the remote‐LFICP‐deposited silicon‐rich a‐SiNx:H displays excellent Si surface passivation (surface recombination velocity of µc‐Si:H is known to be a complex material consisting of crystalline and amorphous silicon phase plus grain boundaries [23]. **Figure 2(a)** shows a typical transmission electron microscopy (TEM) of µc‐Si:H. Apparent columnar grains are observed. The crystalline growth starts at nucleation centers near the interface between film and substrate. During the competitive growth between the nucleation centers, the diameter of the remaining crystalline increases leading to the cone‐shaped crystallites near the substrate surface [24]. **Figure 2(b)** displays the microstructure evolution from highly crystal µc‐Si:H to amorphous a‐Si:H. The highly crystal and amorphous structure are described in the left and right margin, respectively. From left to right, the column size decreases and more amorphous phase is introduced between the crystalline grains. This interrupts the columnar crystalline structure in the lateral direction.

Practically, µc‐Si:H is closely related to a‐Si:H, as the deposition technique for both materials is the same. However, the deposition parameters, such as discharge power density, working pressure, substrate temperature, etc., in the RF discharge deposition method, collectively determine the ultimate microstructures of the obtained silicon thin films. CCP‐PECVD is the most widely studied technique for the deposition of a‐/µc‐Si:H films. In this technique, hydrogen‐diluted silane is dissociated by electron impact from the RF field between two parallel plate electrodes to form SiHx radicals and atomic hydrogen. The generated species diffuse/drift to the growing surface. The species and their reaction products are either incorporated into the growing film or are re‐emitted from the surface into the gas phase. The radicals then undergo various secondary reactions during their transport to the substrate.

**Figure 2.** TEM micrograph of a µc‐Si:H film grown on silicon substrate by means of LFICP (a), and a schematic draw‐ ing to show the microstructure evolution from µc‐Si:H (left) to a‐Si:H (right) (Reproduced from Ref. [24]).

### **2.3. a‐/μc‐Si:H by LFICP CVD**

Although method of ICP has the advantages over the traditional CCP as demonstrated above, the reports on its applications in silicon thin‐film deposition are not abundant, not to mention the solar cell fabrication. For reference, some main progresses are summarized as follows. Li et al. [25] used the built‐in one‐turn coil inductance antenna driven by 13.56 MHz RF power to fabricate µc‐Si:H with tunable preferred orientation through controlling the hydrogen dilution. By the similar method with external conductance coil power by the same frequency generator, Shen et al. [26, 27] obtained low defect density ((3 x 1015/cm3 ) a‐Si:H films, from which a single‐junction a‐Si:H solar cell with energy conversion efficiency of 9.6% was successfully fabricated. The researchers from Hiroshima University realized the high‐rate deposition (up to 4 nm/s) of µc‐Si:H with a low‐level defect density in the year of 2001 [28, 29]. They systematically studied the films' optoelectronic properties such as the optical band gap, dark and photo conductivity, defect density, crystalline fraction ratio, etc. It seems that as‐deposited silicon films are device grade and fulfill the common requirements for solar cell fabrications. Unfortunately, they have not reported their subsequent achievements in their research. The reports on the application of ICP‐grown silicon thin films in heterojunction solar cells are not available. The only related work on the surface passivation by ICP‐grown a‐Si:H was reported in Ref. [30], where minority carrier lifetime of 53 µs was obtained on a P‐type silicon substrate (resistivity 1–20 Ωcm) with a 15‐nm‐thick a‐Si:H passivation layer.

For ICP reactor, there are two types of antenna configurations, namely cylindrical and planar. The present LFICP driven by a low frequency of 460 kHz power generator belongs to a planar coil configuration. As mentioned above, high‐density ion region in ICP is not suitable for a‐ Si:H thin‐film deposition for silicon surface passivation due to the severe ion bombardment. In order to decrease the ion bombardment, the distance *d* between the antenna and the substrate stage is increased by extending the deposition chamber height. In the a‐/µc‐Si:H deposition, two distance values of 33 cm or 53 cm were adopted for comparison purpose. The dependence of a‐/µc‐Si:H film microstructure and properties on the deposition parameters were studied. The n‐doped and p‐doped µc‐Si:H layer were obtained with the introduction of dopant gases of phosphine (PH3) and diborane (B2H6), respectively. The electrical properties of the doped films were analyzed.

### **2.4. Discharge power effect**

The a‐Si:H/µc‐Si:H thin films were deposited on the double‐side polished silicon and glass substrates cleaned by standard methods. The sample holder was heated to a temperature of 140 °C monitored by an internally equipped thermal couple. The flux of silane (H2) was kept constant at 5 sccm (20 sccm). The inductive RF discharge power was varied between 1.0–2.5 kW. The thicknesses of all the deposited thin films were measured directly from the cross‐ sectional scanning electron microscopy (SEM) image by using a JEOL JSM‐6700F filed emission scanning electron microscope. The deposition rate can be calculated from the thickness divided by deposition duration. FTIR measurements were performed on a Perkin‐Elmer FTIR 1725X spectrometer in the mid‐infrared range from 400 to 4000 cm−1 with an increment of 1 cm−1. The Cary 510 Bio spectrometer (300–1100nm) was used to measure the UV‐VIS transmission spectra.

### *2.4.1. Optical band gap and deposition rate*

The UV‐VIS transmission spectra (shown in **Figure 3(a)**) of the thin films deposited on glass at different RF powers were measured to extract the parameter of optical band gap *Eg*. Evident interference fringes due to multi‐layer interface reflections were observed in the spectra range longer than the absorption edge. The absorption edge occurs a redshift with deposition power, indicating shrinkage of the optical band gap. The transmission *T* was transformed into absorption *α* by the relation of *α*∞‐ln(*T*) using the thickness of the films directly read from the cross‐sectional SEM images. The optical band gap *Eg* of a‐/µc‐Si:H (Si an indirect band gap semiconductor) can be deduced from the Tauc equation

dilution. By the similar method with external conductance coil power by the same frequency

which a single‐junction a‐Si:H solar cell with energy conversion efficiency of 9.6% was successfully fabricated. The researchers from Hiroshima University realized the high‐rate deposition (up to 4 nm/s) of µc‐Si:H with a low‐level defect density in the year of 2001 [28, 29]. They systematically studied the films' optoelectronic properties such as the optical band gap, dark and photo conductivity, defect density, crystalline fraction ratio, etc. It seems that as‐deposited silicon films are device grade and fulfill the common requirements for solar cell fabrications. Unfortunately, they have not reported their subsequent achievements in their research. The reports on the application of ICP‐grown silicon thin films in heterojunction solar cells are not available. The only related work on the surface passivation by ICP‐grown a‐Si:H was reported in Ref. [30], where minority carrier lifetime of 53 µs was obtained on a P‐type

silicon substrate (resistivity 1–20 Ωcm) with a 15‐nm‐thick a‐Si:H passivation layer.

of the doped films were analyzed.

*2.4.1. Optical band gap and deposition rate*

**2.4. Discharge power effect**

spectra.

For ICP reactor, there are two types of antenna configurations, namely cylindrical and planar. The present LFICP driven by a low frequency of 460 kHz power generator belongs to a planar coil configuration. As mentioned above, high‐density ion region in ICP is not suitable for a‐ Si:H thin‐film deposition for silicon surface passivation due to the severe ion bombardment. In order to decrease the ion bombardment, the distance *d* between the antenna and the substrate stage is increased by extending the deposition chamber height. In the a‐/µc‐Si:H deposition, two distance values of 33 cm or 53 cm were adopted for comparison purpose. The dependence of a‐/µc‐Si:H film microstructure and properties on the deposition parameters were studied. The n‐doped and p‐doped µc‐Si:H layer were obtained with the introduction of dopant gases of phosphine (PH3) and diborane (B2H6), respectively. The electrical properties

The a‐Si:H/µc‐Si:H thin films were deposited on the double‐side polished silicon and glass substrates cleaned by standard methods. The sample holder was heated to a temperature of 140 °C monitored by an internally equipped thermal couple. The flux of silane (H2) was kept constant at 5 sccm (20 sccm). The inductive RF discharge power was varied between 1.0–2.5 kW. The thicknesses of all the deposited thin films were measured directly from the cross‐ sectional scanning electron microscopy (SEM) image by using a JEOL JSM‐6700F filed emission scanning electron microscope. The deposition rate can be calculated from the thickness divided by deposition duration. FTIR measurements were performed on a Perkin‐Elmer FTIR 1725X spectrometer in the mid‐infrared range from 400 to 4000 cm−1 with an increment of 1 cm−1. The Cary 510 Bio spectrometer (300–1100nm) was used to measure the UV‐VIS transmission

The UV‐VIS transmission spectra (shown in **Figure 3(a)**) of the thin films deposited on glass at different RF powers were measured to extract the parameter of optical band gap *Eg*. Evident interference fringes due to multi‐layer interface reflections were observed in the spectra range

) a‐Si:H films, from

generator, Shen et al. [26, 27] obtained low defect density ((3 x 1015/cm3

188 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

$$h\gamma\alpha(h\gamma) = B(h\gamma - E\_g)^2\tag{1}$$

where *B* is the edge width parameter related to the width of band tails in the film [31], and *hν* the photon energy. In general, *B* factor is also taken as a measure of the overall structural disorder and a higher value for *B* would indicate a lower degree of structure disorder [32]. In our calculation, the obtained value of *B* (>800 cm−1/2.eV−1/2) is larger than that of literatures, indicating less disorder and defects in the thin films.

The obtained *Eg* as a function of applied RF power together with the average deposition rate are both plotted in **Figure 3(b)**. A transition point of *Eg* was observed at power of 1.8 kW. Below 1.8 kW, *Eg* encounters a decrease from 2.30 eV for 1.0 kW to 2.10 eV for 1.8 kW. On the contrary, *Eg* increases from 2.10 eV for 1.8 kW to 2.14 eV for 2.5 kW. This changing trend of *Eg* is related to the crystallinity of the thin films and the hydrogen content, which will be discussed below based on the experiments of Raman scattering and FTIR. The increasing RF power results in the steady increase in the average deposition rate from 7 to 21 nm/min. These values obtained at antenna‐substrate distance *d* of 33 cm are about half of those corresponding to distance of 11 cm [33].

**Figure 3.** The UV transmission spectra of the thin films deposited at different RF powers on glass substrate (a), and deduced *Eg* (filled squares) and average deposition rate (filled circles) (b).

Another feature of the present films is the higher optical band gap *Eg* than that reported in Refs. [34–36] where *Eg* was obtained by the methods other than Tauc method (spectroscopic ellipsometry and surface photovoltage spectroscopy).

Although it was recognized that the Tauc method will lead to a larger value for *Eg*, it will give a correct variation trend of the value with respect to an independent parameter like RF power, substrate temperature, and so on. Herein, the values of *Eg* are above 2 eV, larger than the values (<1.9 eV) [33] obtained in earlier works performed on the direct‐plasma ICP system with other similar deposition parameters. This should be correlated to the increased hydrogen content (following FTIR results) in the thin films.

### *2.4.2. Raman scattering and crystallinity of the thin films*

The micro‐Raman scattering experiments were performed to estimate the crystal phase information in the thin films. The obtained spectra are displayed in **Figure 4(a)**. It was observed that the dominant peak gradually shifts from (480 to (520 cm−1 through a mixed phase state at the power of 1.8 kW. This change reflects phase evolution from completely amorphous to highly crystal with an applied RF power in the films. In fact, the crystallinity value can be estimated by using different methods described in Ref. [37]. In this work, the method based on Gaussian line fitting [38] is used, where *I*<sup>a</sup>

$$f\mathbf{c} = (I\_c + I\_i) / (I\_a + I\_c + I\_i) \tag{2}$$

**Figure 4.** RF power‐dependent Raman scattering spectra (a), and the Gaussian‐shaped curve decomposition of the Raman spectrum in the thin film deposited at RF powers of 1.8 kW. The inset of (b) shows the RF power‐dependent crystallinity of the thin films.

*I*i , and *I*c are the integrated area of each phase, i.e., the amorphous ((480 cm−1), intermediate ((510cm−1), and crystalline phase ((520 cm−1), respectively, in the Raman spectra. A typical deconvolution result is shown in **Figure 4(b)**, where the cyan, blue, and green curves (from left to right) denote the amorphous, intermediate, and crystal component, respectively. The experimental (open circle) and overall fitting result agrees well, indicating the validity of the deconvolution procedure. The calculated crystallinity as a function of RF power is included in the inset of **Figure 4(b)**. As expected, the crystallinity increases monotonously in the power range of 1.0–2.0 kW because of the production of denser and more energetic ion/radicals in the plasma and area near the surface of the substrate. Higher plasma density means more frequent impacts between the electron and the precursor gases, and thus a larger deposition rate. More energetic ion/radicals imply that the adsorbed radicals on the surface will have a greater diffusion length. This assists the adsorbed radicals to reach the most energetically favorable sites and gives rise to better crystallization of the synthesized thin films [39]. The crystallinity decreases from 67.0 to 61.6% when the RF power is further increased from 2.0 to 2.5 kW. This should be related to the rapid increase of deposition rate in the range of 2.0–2.5 kW shown in **Figure 3(b)**. One possible interpretation may be that: a higher (>2.0 kW) RF power excites more and more precursor ions like SiH, SiH2, and SH3 to improve the growth rate; on the other hand, the overly rapid growth minimizes the hydrogen etching effect on the existing weak and strained Si bonds. It is generally accepted that hydrogen etching is also an important mecha‐ nism to improve the crystallinity of Si thin films [40].

### *2.4.3. FTIR and hydrogen content in the thin films*

Although it was recognized that the Tauc method will lead to a larger value for *Eg*, it will give a correct variation trend of the value with respect to an independent parameter like RF power, substrate temperature, and so on. Herein, the values of *Eg* are above 2 eV, larger than the values (<1.9 eV) [33] obtained in earlier works performed on the direct‐plasma ICP system with other similar deposition parameters. This should be correlated to the increased hydrogen content

190 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The micro‐Raman scattering experiments were performed to estimate the crystal phase information in the thin films. The obtained spectra are displayed in **Figure 4(a)**. It was observed that the dominant peak gradually shifts from (480 to (520 cm−1 through a mixed phase state at the power of 1.8 kW. This change reflects phase evolution from completely amorphous to highly crystal with an applied RF power in the films. In fact, the crystallinity value can be estimated by using different methods described in Ref. [37]. In this work, the method based

**Figure 4.** RF power‐dependent Raman scattering spectra (a), and the Gaussian‐shaped curve decomposition of the Raman spectrum in the thin film deposited at RF powers of 1.8 kW. The inset of (b) shows the RF power‐dependent

, and *I*c are the integrated area of each phase, i.e., the amorphous ((480 cm−1), intermediate ((510cm−1), and crystalline phase ((520 cm−1), respectively, in the Raman spectra. A typical deconvolution result is shown in **Figure 4(b)**, where the cyan, blue, and green curves (from left to right) denote the amorphous, intermediate, and crystal component, respectively. The experimental (open circle) and overall fitting result agrees well, indicating the validity of the deconvolution procedure. The calculated crystallinity as a function of RF power is included in the inset of **Figure 4(b)**. As expected, the crystallinity increases monotonously in the power

( )/( ) *ci aci fc I I I I I* = + ++ (2)

(following FTIR results) in the thin films.

on Gaussian line fitting [38] is used, where *I*<sup>a</sup>

crystallinity of the thin films.

*I*i

*2.4.2. Raman scattering and crystallinity of the thin films*

The FTIR spectra of the thin films deposited with various RF power are shown in **Figure 5(a)**. Within a wavenumber range of 500–2300 cm−1, three Si‐H bond‐related bands were observed. The first one, located at around 640 cm−1 was unambiguously attributed to Si‐H rocking/ wagging mode. Its exact location gradually changed from 644 to 630 cm−1 with increasing RF power as displayed in **Figure 5(b)** (the filled circles). This is also correlated to the phase evolution from amorphous to crystal silicon [41]. The second one comprising of two sub‐peaks located at 845 and 890 cm−1 were identified as the bending mode of dihydrides SiH2 [41]. The third one at the region of (2100 cm−1 is recognized as the SiH*<sup>x</sup>* stretching mode. One can see that these hydrogen‐related signals become weak with increasing RF power, implying the decreasing bonded hydrogen content in the thin films. Actually, the bonded hydrogen content in the thin films can be estimated from the Si‐H bond rocking/wagging mode using the following equation:

$$C\_H = A\_o \int \frac{\alpha(o)}{o0} do = A\_o I\_o \tag{3}$$

where the oscillator strength *A*<sup>ω</sup> is of value of 1.6 × 1019 cm−2. The calculated hydrogen content at.% (defined as density ratio of hydrogen to that of hydrogen plus silicon) versus applied RF power is plotted as filled square in **Figure 5(b)**. The hydrogen content is in the range of 7.9–‐ 17.4% and exhibits a general decreasing trend with RF power. Meanwhile, much lower values (<8%) were obtained using the same method in the case of direct plasma (i.e., the film depo‐ sition region located in the plasma generation region) [15, 33]. A local fluctuation of hydrogen content was also observed at a power of 1.8 kW, where the thin film is of transition state from amorphous to crystal phase. Further increase in RF power from 2.0 to 2.5 kW leads to a little improvement of hydrogen content, which is consistent with the decreasing crystallinity in this power range shown in **Figure 4(b)**.

**Figure 5.** FTIR absorption spectra of the thin films deposited at different RF powers (a), and the RF power‐dependent hydrogen content (filled squares) and Si‐H wagging/rocking mode location (filled circles) (b).

#### **2.5. Hydrogen dilution effect**

Hydrogen in the deposited films originates from the diluent hydrogen gas and hydrogen atom bonded in silane molecules. From the point view of the deposition process, hydrogen facilitates the dissociation of silane through increasing the ion/electron impact, influences the surface energy of the growing surface via terminating the surface silicon atom, and etches the growing surface. Therefore, the hydrogen dilution should have a substantial effect on the microstruc‐ tures of the ultimate silicon thin films. In this set, the RF discharge power and substrate are fixed at 1.5 kW and 140 °C, respectively. The total flow rate of SiH4 and H2 is kept at 20 sccm. The respective flow rate of SiH4 and H<sup>2</sup> are changed with varying dilution ratio [H2]/[SiH4] = 0, 0.25, 0.5, 1, 2, 3, and 4. The working pressure in the deposition chamber is kept at 7.2 Pa.

**Figure 6** shows the average deposition rate and *Eg* deduced from the Tauc method described above. As expected, the deposition rate monotonically decreases with the decreasing silence concentration (i.e., increasing hydrogen dilution ratios). When the dilution ratio changes from 0 (pure silane) to 4, the deposition rate drops sharply from 42 nm/min to 10 nm/min. However, the case of *Eg* is more complicated with increasing dilution ratios. Within the range of ratio 0– 2, *Eg* increases from 2.17 to 2.32 eV. Further increase in dilution ratio results in decreasing *Eg*. At a dilution ratio of 4, *Eg* is 2.21 eV. It has been mentioned above that *Eg* of a‐/µc‐Si:H is related to the crystallinity and hydrogen content in the films. The corresponding Raman spectra shows that all the films deposited under present conditions are of amorphous structure without clear crystalline fractions. As pointed out above, the wagging/ rocking mode location in FTIR spectra is also related to the phase configuration in the films.

**Figure 6.** Dilution ratio‐dependent optical band gap and average deposition rate.

**Figure 5.** FTIR absorption spectra of the thin films deposited at different RF powers (a), and the RF power‐dependent

Hydrogen in the deposited films originates from the diluent hydrogen gas and hydrogen atom bonded in silane molecules. From the point view of the deposition process, hydrogen facilitates the dissociation of silane through increasing the ion/electron impact, influences the surface energy of the growing surface via terminating the surface silicon atom, and etches the growing surface. Therefore, the hydrogen dilution should have a substantial effect on the microstruc‐ tures of the ultimate silicon thin films. In this set, the RF discharge power and substrate are fixed at 1.5 kW and 140 °C, respectively. The total flow rate of SiH4 and H2 is kept at 20 sccm. The respective flow rate of SiH4 and H<sup>2</sup> are changed with varying dilution ratio [H2]/[SiH4] = 0, 0.25, 0.5, 1, 2, 3, and 4. The working pressure in the deposition chamber is kept at 7.2 Pa.

**Figure 6** shows the average deposition rate and *Eg* deduced from the Tauc method described above. As expected, the deposition rate monotonically decreases with the decreasing silence concentration (i.e., increasing hydrogen dilution ratios). When the dilution ratio changes from 0 (pure silane) to 4, the deposition rate drops sharply from 42 nm/min to 10 nm/min. However, the case of *Eg* is more complicated with increasing dilution ratios. Within the range of ratio 0– 2, *Eg* increases from 2.17 to 2.32 eV. Further increase in dilution ratio results in decreasing *Eg*. At a dilution ratio of 4, *Eg* is 2.21 eV. It has been mentioned above that *Eg* of a‐/µc‐Si:H is related to the crystallinity and hydrogen content in the films. The corresponding Raman spectra shows that all the films deposited under present conditions are of amorphous structure without clear crystalline fractions. As pointed out above, the wagging/ rocking mode location in FTIR spectra

hydrogen content (filled squares) and Si‐H wagging/rocking mode location (filled circles) (b).

192 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**2.5. Hydrogen dilution effect**

is also related to the phase configuration in the films.

**Figure 7** shows the extract waging/rocking vibration mode. With a small amount of hydrogen introduction (H2/SiH4 = 0.25), the hydrogen content CH increases from 26.3 to 27.5% and the wagging/rocking mode location shifts from 649 to 644 cm−1. CH drastically drops to 24.4% at a dilution ratio of 0.5. The values of CH and the location of wagging/rocking mode are compa‐ rably steady in the dilution range of 0.5–2. Beyond 2, CH decreases drastically to 12.2% and the mode location shifts to 631 cm−1 at a ratio of 4. The location shift to the small wavenumber side with increasing dilution ratios implies the phase evolution to the crystalline orientation.

**Figure 7.** Hydrogen content and the wagging/rocking mode location in the a‐Si:H films as functions of the dilution ration H2/SiH4.

In view of the above experimental results, hydrogen dilution in LFICP method is also a key parameter influencing the hydrogen content CH in the films. However, its effect on the crystallinity of the films is less pronounced than that of the discharge power.

### **2.6. Working pressure effect**

One of the inherent features of LFICP is low working pressure up to several tens of Pa. In this section, the working pressure is altered in the range (1–19Pa) enabling a stable discharge. The discharged power was kept at 2.5 kW. No external heating was applied to the substrate. The flow rate of SiH4 and H2 are 12.5 sccm. The Raman scattering experiments revealed the amorphous structure of these films. **Figure 8** displays the working pressure‐dependent *Eg* and average deposition rate. The deposition rate monotonically increases with working pressure from the 68 nm/min (1 Pa) to 142 nm/min (19 Pa). On the contrary, *Eg* becomes narrower with working pressure: 2.34 eV for 1 Pa, and 1.92 eV for 19 Pa.

**Figure 8.** The deposition rate and *Eg* as functions of working pressure.

**Figure 9(a)** shows the FTIR absorbance spectra of these films. The intensity of the waging/ rocking mode becomes weaker with increasing work pressure, indicating the effusion of the bonded hydrogen with increase in the working pressure. The detailed hydrogen content as a function of working pressure is shown in **Figure 9(b)**. The hydrogen content was reduced about 20% when the working pressure is increased from 1 to 19 Pa. The location of this mode experiences a slight shift to higher wavenumbers with working pressure, which implies an amorphous phase change. It is well known that an increase in the gas pressure enhances the collisions in the plasma and reduces the electron and ion energies, and hence decreases the crystallinity.

High‐Density Plasma‐Enhanced Chemical Vapor Deposition of Si‐Based Materials for Solar Cell Applications http://dx.doi.org/10.5772/63529 195

**Figure 9.** FTIR absorptance spectra of the a‐Si:H thin films deposited at different pressures (a), and deduced hydrogen content *CH* in the films (b).

It is worth noting that the stretching mode also decreases considerably with increasing working pressure. Assuming the stretching mode corresponds to SiH‐associated peak at 2000 cm−1 and the SiH2‐associated peak at 2100 cm−1, we will note the severe decrease of SiH2 fraction in this mode. This should be responsible for the reduced hydrogen content in the a‐Si:H films.

In summary, the working pressure substantially influences the optical band gap *Eg* and hydrogen content of the films. In present LFICP deposition, the high *Eg* values larger than 2 eV is mainly correlated to the high hydrogen content CH which corresponds to the ultra‐low working pressure.

### **2.7. Effect of substrate temperature**

In view of the above experimental results, hydrogen dilution in LFICP method is also a key parameter influencing the hydrogen content CH in the films. However, its effect on the

One of the inherent features of LFICP is low working pressure up to several tens of Pa. In this section, the working pressure is altered in the range (1–19Pa) enabling a stable discharge. The discharged power was kept at 2.5 kW. No external heating was applied to the substrate. The flow rate of SiH4 and H2 are 12.5 sccm. The Raman scattering experiments revealed the amorphous structure of these films. **Figure 8** displays the working pressure‐dependent *Eg* and average deposition rate. The deposition rate monotonically increases with working pressure from the 68 nm/min (1 Pa) to 142 nm/min (19 Pa). On the contrary, *Eg* becomes narrower with

**Figure 9(a)** shows the FTIR absorbance spectra of these films. The intensity of the waging/ rocking mode becomes weaker with increasing work pressure, indicating the effusion of the bonded hydrogen with increase in the working pressure. The detailed hydrogen content as a function of working pressure is shown in **Figure 9(b)**. The hydrogen content was reduced about 20% when the working pressure is increased from 1 to 19 Pa. The location of this mode experiences a slight shift to higher wavenumbers with working pressure, which implies an amorphous phase change. It is well known that an increase in the gas pressure enhances the collisions in the plasma and reduces the electron and ion energies, and hence decreases the

crystallinity of the films is less pronounced than that of the discharge power.

194 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

working pressure: 2.34 eV for 1 Pa, and 1.92 eV for 19 Pa.

**Figure 8.** The deposition rate and *Eg* as functions of working pressure.

crystallinity.

**2.6. Working pressure effect**

The substrate temperature substantially affects the plasma species diffusion on the surface and nucleation process, and hence the properties of the resulting a‐/µc‐Si:H thin films. In this set, the feed gas flow rate ratio is SiH4/H2 =5:20 (in sccm) and the discharge power is 2.0 kW. The substrate temperature is varied in the range of 27–150 °C. **Figure 10** shows the temperature‐

**Figure 10.** The substrate temperature‐dependent deposition rate and *Eg*.

dependent deposition rate and *Eg*. The increasing substrate temperature leads to the decrease in both of these two parameters. In the studied temperature range, deposition rate and *Eg* decreases by about 7.5 nm/min and 0.9 eV, respectively.

**Figure 11** shows the substrate temperature‐dependent Raman spectra. The obtained films are highly crystalline under the present deposition conditions. The elevated substrate temperature results in the enhanced intensity of the 520 cm−1 peak, and thus more order structure in the films.

**Figure 11.** Substrate temperature‐dependent Raman spectra.

The wagging/rocking mode in FTIR transmission spectra provides the bonded hydrogen numbers information which is displayed in **Figure 12**. The bonded hydrogen content CH is very sensitive to the substrate temperature, decreasing by about 21% with substrate temper‐ ature from room temperature to 150 °C.

**Figure 12.** Substrate temperature‐dependent hydrogen content in the films.

### **2.8. Effect of antenna‐substrate distance**

dependent deposition rate and *Eg*. The increasing substrate temperature leads to the decrease in both of these two parameters. In the studied temperature range, deposition rate and *Eg*

196 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 11** shows the substrate temperature‐dependent Raman spectra. The obtained films are highly crystalline under the present deposition conditions. The elevated substrate temperature results in the enhanced intensity of the 520 cm−1 peak, and thus more order structure in the

The wagging/rocking mode in FTIR transmission spectra provides the bonded hydrogen numbers information which is displayed in **Figure 12**. The bonded hydrogen content CH is very sensitive to the substrate temperature, decreasing by about 21% with substrate temper‐

decreases by about 7.5 nm/min and 0.9 eV, respectively.

**Figure 11.** Substrate temperature‐dependent Raman spectra.

**Figure 12.** Substrate temperature‐dependent hydrogen content in the films.

ature from room temperature to 150 °C.

films.

In the axial direction, the plasma properties should be dependent on the antenna‐substrate distance *d* between the RF antenna and the depositing substrate. Although the detailed plasma characterizations in different distances have not been conducted, the different plasma densi‐ ties, electron temperatures, and sheath potentials, etc. are expected. In principle, the neutral particles will become more and more important among species contributing to the film deposition due to the enhanced ion recombination with increasing distance *d*. It was recog‐ nized that the severe etching and ion bombardment produces additional defects on growing surface/substrate directly contacting with the high‐density plasma [16]. Therefore, the deposition region should avoid direct contact with the high‐density plasma during deposition. **Figure 13** shows the Raman spectra of the films deposited at different *d* values. The film is completely crystal in the case of direct plasma at *d* = 11 cm. In comparison, microcrystalline silicon is obtained when the distance is widened to 33 cm (namely, a semi‐remote plasma configuration). The film is completely amorphous at the distance of *d* = 53 cm (a remote plasma configuration). Meanwhile, the atom network evolution from crystalline to amorphous is accompanied with the deposition rate reduction shown in **Figure 14**. The drastic reduction of deposition rate with distance *d* reflects the decreasing number of precursors contributing to Si:H deposition.

**Figure 13.** Raman scattering spectra of the silicon thin films deposited at different antenna‐substrate distance *d* values. The other deposition parameters are as follows: discharge power 1.8 kW, substrate temperature 100 °C, and gas pres‐ sure of 7.2 Pa.

**Figure 14.** The antenna‐substrate distance‐dependent deposition rates.

**Figure 15** shows the evolution of FTIR absorbance spectra with respect to the distance *d*. All the H‐associated signals are enhanced with distance *d*. The hydrogen content CH deduced from the wagging/rocking mode is represented in the inset. CH increases sharply with *d* possibly due to the reduced plasma heating effect.

**Figure 15.** Comparison of FTIR absorbance spectra of the films deposited at different antenna‐substrate distance val‐ ues.

#### **2.9. Doped a‐/μc‐Si:H films by LFICP CVD**

In this section, the process parameters‐dependent microstructure and electrical properties of the n‐ and p‐doped a‐/µc‐Si:H films are studied to evaluate the doping ability of LFICP. Doping of a‐Si:H by PECVD was first demonstrated by Spear and Le Comber in Ref. [43]. The disor‐ dered structure of a‐Si:H was supposed to allow the incorporation of dopant into the amor‐ phous network, thus impeding the doping of the material. Subsequently, the underlying doping mechanism of a‐Si:H was further clarified by Street [44]. Nevertheless, the doping efficiency of a‐Si:H (<1%) as a disordered material is much lower than in crystalline silicon [45].

Here, the feed gas silane is pure without hydrogen pre‐dilution. The n‐type dopant resource gas phosphine (PH3) and p‐type dopant resource gas diborane (B2H6) both were H2‐diluted to a 15% concentration. In the following discussions, the gas consumptions are directly denoted as the nominal flow rate. It was found that the antenna‐substrate distance *d* also significantly affects the doping capability of LFICP. As an example, **Figure 16** shows the hole carrier concentration evolution with respect to the parameter of *d* under the same other parameters: SiH4:H2:B2H<sup>6</sup> = 5:20:0.4 (in sccm), discharge power 2.0 kW and low substrate temperatures of 150 or 100 °C. When *d* is increased from 11 to 53 cm, the hole concentration drops sharply spanning a wide range of 1019–1014 cm−3. The result at 100 °C is also included in the case of 33 cm. Recalling the Raman results at different *d* values (**Figure 13**), one will find that carrier density is inversely proportional to the crystallinity of the films. This is consistent with the fact that amorphous network is more difficult to dope. In the next section, it will be revealed that larger‐distance‐deposited a‐Si:H films provide more effective surface passivation effect on silicon. Hence, doping and passivation ability for LFICP‐grown a/‐µc‐Si seem to be a contra‐ diction.

**Figure 14.** The antenna‐substrate distance‐dependent deposition rates.

due to the reduced plasma heating effect.

**2.9. Doped a‐/μc‐Si:H films by LFICP CVD**

ues.

**Figure 15** shows the evolution of FTIR absorbance spectra with respect to the distance *d*. All the H‐associated signals are enhanced with distance *d*. The hydrogen content CH deduced from the wagging/rocking mode is represented in the inset. CH increases sharply with *d* possibly

198 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 15.** Comparison of FTIR absorbance spectra of the films deposited at different antenna‐substrate distance val‐

In this section, the process parameters‐dependent microstructure and electrical properties of the n‐ and p‐doped a‐/µc‐Si:H films are studied to evaluate the doping ability of LFICP. Doping of a‐Si:H by PECVD was first demonstrated by Spear and Le Comber in Ref. [43]. The disor‐

**Figure 16.** The hole carrier density as a function of antenna‐substrate distance *d* in the p‐doped films deposited at sub‐ strate temperature 150 °C (solid square) and 100 °C (open square).

### *2.9.1. p‐doping*

Different‐flux phosphine gas was introduced into the feed gases of SiH4:H<sup>2</sup> = 6:60 (in sccm) discharged by RF power of 3 kW with substrate temperature of 100 °C. **Figure 17(a)** shows the electrical parameters (from Hall effect measurement using the Ecopia HMS‐3000 system) of as‐deposited p‐doped a‐/µc‐Si:H films. The hole density *p* reaches its maximum value of 1.2 × 1019 cm3 at an intermediate B2H6 flux of 0.4 sccm. The ultimate conductivity *σ* displays the same trend with carrier density. The highest *σ* value is 0.58/Ωcm obtained at B2H6 flux of 0.4 sccm. The corresponding Raman scattering spectra of the films are shown in **Figure 17(b)**. A systematical shift from microcrystal to amorphous structure occurs with the gradual addition of dopant gas B2H6. The flux of 0.4 sccm is an optimum one, where the optimized amount of the dopant is introduced into the host µc‐Si:H structures.

**Figure 17.** B2H6 flux‐dependent electrical properties including the hole carrier density, mobility, resistivity, and con‐ ductivity (a), and corresponding Raman spectra of the a‐/µc‐Si:H films (b).

It is found that the doping is strongly dependent on the deposition temperature. Increasing substrate temperature improves the conductivity of the deposited films. **Figure 18(a)** shows the electrical properties of the doped a‐/µc‐Si:H films deposited in the substrate temperature range of 100–200 °C. The flux of B2H6 is kept at 0.4 sccm. The increased temperature facilitates the activation of the dopant in the film leading to the improvement in the hole density and degraded mobility. The resulting conductivity *σ* almost increases linearly with substrate temperature and *σ* reaches 2.1/Ωcm. **Figure 18(b)** shows the corresponding Raman scattering spectra. The peak gradually shifts from 505 to 510 cm−1 indicating the microstructure shift to the microcrystalline side. This ordered network resulted from the higher substrate temperature is beneficial to doping behavior.

**Figure 18.** Electrical properties of the doped films with B2H6 = 0.4 sccm at different substrate temperatures (a), and the corresponding Raman scattering spectra (b).

### *2.9.2. n‐doping*

*2.9.1. p‐doping*

1019 cm3

Different‐flux phosphine gas was introduced into the feed gases of SiH4:H<sup>2</sup> = 6:60 (in sccm) discharged by RF power of 3 kW with substrate temperature of 100 °C. **Figure 17(a)** shows the electrical parameters (from Hall effect measurement using the Ecopia HMS‐3000 system) of as‐deposited p‐doped a‐/µc‐Si:H films. The hole density *p* reaches its maximum value of 1.2 ×

200 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

trend with carrier density. The highest *σ* value is 0.58/Ωcm obtained at B2H6 flux of 0.4 sccm. The corresponding Raman scattering spectra of the films are shown in **Figure 17(b)**. A systematical shift from microcrystal to amorphous structure occurs with the gradual addition of dopant gas B2H6. The flux of 0.4 sccm is an optimum one, where the optimized amount of

**Figure 17.** B2H6 flux‐dependent electrical properties including the hole carrier density, mobility, resistivity, and con‐

It is found that the doping is strongly dependent on the deposition temperature. Increasing substrate temperature improves the conductivity of the deposited films. **Figure 18(a)** shows the electrical properties of the doped a‐/µc‐Si:H films deposited in the substrate temperature range of 100–200 °C. The flux of B2H6 is kept at 0.4 sccm. The increased temperature facilitates the activation of the dopant in the film leading to the improvement in the hole density and degraded mobility. The resulting conductivity *σ* almost increases linearly with substrate

the dopant is introduced into the host µc‐Si:H structures.

ductivity (a), and corresponding Raman spectra of the a‐/µc‐Si:H films (b).

at an intermediate B2H6 flux of 0.4 sccm. The ultimate conductivity *σ* displays the same

It seems that the n‐doping is more difficult than p‐doping in the case of higher antenna‐ substrate distance. Herein, the substrate temperature ranges from 100–200 °C, PH3 flux covers the range of 0.5–1.0 sccm, and the feed gas ratio SiH4:H2 is changed from 5:40 to 5:60 (in sccm). The discharge power is set at 2.75 kW. **Figure 19** shows the Raman scattering spectra of the a‐/ µc‐Si:H films n‐doped by different PH3 flux. The microstructure change with the PH3 intro‐ duction is not as monotonic as shown in **Figure 17** associated with the case of B2H6. **Table 1** presents the deposition parameters and associated electrical properties of the obtained n‐ doped a‐/µc‐Si films. At PH<sup>3</sup> flux of 0.5 and 0.62 sccm (100 °C), the electron density *n* is below 1014/cm3 . The former flux 0.5 sccm is not adequate for effective doping and the latter 0.62 sccm leads to an amorphous‐oriented network according to **Figure 19**. The optimum set is 0.75 sccm PH3 and 200 °C, where the highest conductivity is 3.45/Ωcm.

**Figure 19.** Raman scattering spectra of a‐/µc‐Si:H n‐doped at different PH3 fluxes.


**Table 1.** Deposition parameters and properties of n‐doped a‐/µc‐Si:H thin films.

**Figure 20** shows the n‐doped a‐/µc‐Si:H films deposited with different hydrogen dilution ratio. Up to the ratio of SiH4/H2 = 5:50, the film is a completely amorphous network impeding the

**Figure 20.** Raman scattering spectra of the n‐doped a‐/µc‐Si:H films deposited at different hydrogen dilution ratio.

doping behavior. The structure turns into microcrystalline at the ratio of SiH4/H2 = 5:60, where the better conductivity can be obtained in present experiments.

**Figure 21** shows a typical top view **(a)** and cross‐sectional **(b)** SEM image of a n‐doped µc‐Si:H film deposited at a temperature of 200 °C and dilution ratio of H2:SiH4 = 60:5. Nanostructured Si particles are uniformly distributed on the surface in **Figure 21(a)**. There are no any observ‐ able voids or apparent defects. The apparent column growth is observed in **Figure 21(b)**. The interface between the film and the Si substrate is clear and free of incubation layer.

**Figure 21.** Top view (a) and cross‐sectional SEM image of the n‐doped µc‐Si:H.

**Figure 19.** Raman scattering spectra of a‐/µc‐Si:H n‐doped at different PH3 fluxes.

**Table 1.** Deposition parameters and properties of n‐doped a‐/µc‐Si:H thin films.

**(kW) (sccm) (sccm) (sccm) (Pa) ( °C) (/cm3**

**Power SiH4 H2 PH3 Pressure** *T* **n** *μ ρ σ*

202 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

2.75 5 60 0.5 12 100 4.11E+14 17.85 859 0.001 2.75 5 60 0.75 12 100 4.95E+15 6.34 199 0.005 2.75 5 60 0.75 12 200 1.03E+20 0.21 0.29 3.45 2.75 5 60 1 12 200 7.56E+19 0.031 2.68 0.38

**Figure 20** shows the n‐doped a‐/µc‐Si:H films deposited with different hydrogen dilution ratio. Up to the ratio of SiH4/H2 = 5:50, the film is a completely amorphous network impeding the

**Figure 20.** Raman scattering spectra of the n‐doped a‐/µc‐Si:H films deposited at different hydrogen dilution ratio.

**) (cm2**

**/VS) (Ωcm) (/Ωcm)**

### **3. c‐Si surface passivation by LFICP CVD a‐/μc‐Si:H**

In Sections 1–2, the intrinsic and doped a‐/µc‐Si:H films based on LFICP‐CVD method are studied in terms of the plasma parameters. The optical band gap, bond configuration, micro‐ structure, and hydrogen content in the films are dependent on the deposition parameters of RF power, gas ratio, working pressure, and substrate temperature. In particular, the properties of the LFICP‐grown a‐/µc‐Si:H films are sensitive to the antenna‐substrate distance *d*. A high *d* value is inclined to a‐Si:H over µc‐Si:H growth due to avoiding direct immersion in high‐ density plasma. However, increasing *d* makes it difficult to dope the a‐/µc‐Si:H films. Improv‐ ing substrate temperature and hydrogen dilution can effectively compensate the doping obstacle from increased *d*. The conductivity of 2.10 and 3.45/Ωcm are achieved in p‐ and n‐ doped µc‐Si:H at a low substrate temperature of 200 °C.

In this section, the crystalline silicon surface passivation by LFICP‐CVD a‐/µc‐Si:H is investi‐ gated. In terms of the antenna‐substrate distance *d*, a‐/µc‐Si:H films display different passiva‐ tion level evaluated by the carrier lifetime measurements. The passivation effect of the film, which is strongly dependent on a variety of processing conditions, e.g., the pre‐deposition treatment of c‐Si, the film deposition parameters, and post‐deposition annealing, is physically analyzed. The carrier recombination mechanisms will be firstly reviewed.

### **3.1. Recombination and passivation in Si‐based solar cells**

Usually, the loss in photovoltaic solar cell comprises surface light loss, the transmitted light loss, thermalization loss (thermal electron loss). In a practical heterojunction (HJ) silicon solar cell, the a‐/µc‐Si:H layer is of several nanometers thick, which marginally absorbs sunlight. The photogeneration of excess charge carrier pairs mainly occur in the bulk c‐Si. In order to efficiently collect the photogenerated carriers which can be extracted and thus contribute to the externally retrievable current, reducing the carrier recombination in silicon, especially on the surface, is particularly important.

The carrier recombination originating from bulk silicon primarily comprises two types, i.e., the intrinsic and extrinsic ones. The intrinsic recombination, which mainly includes radiative [46] and Auger recombination [47], occurs in a pure and defect‐free semiconductor, while the extrinsic recombination stems from contaminants and defects which act as recombination centers.

Total bulk recombination is the sum of the above three terms and is expressed as

$$U = U\_{Rad} + U\_{Aag} + U\_{SRH} = \frac{\Delta n}{\tau\_{Rad}} + \frac{\Delta n}{\tau\_{Aag}} + \frac{\Delta n}{\tau\_{SRH}} = \frac{\Delta n}{\tau\_b} \tag{4}$$

where *τSRH* denotes the contribution from the extrinsic recombination of Shock‐Read‐Hall effect [48, 49].

Surface recombination is a special case of SRH recombination due to the localized states appearing at the surface. Different from the bulk SRH centers, these states usually form a set of states across the band gap instead of a single energy level. Surface recombination is usually evaluated in terms of surface recombination velocity (SRV) instead of lifetime. However, the principles are same to bulk SRH recombination. It mainly originates from the surface dangling bond, i.e., the missing of crystalline network. The chemically and mechanically created defect can also result in surface recombination. In a simple condition with a constant bulk lifetime *τ*<sup>b</sup> and a small constant SRV (two surfaces are same to each other), the effective lifetime *τeff* is expressed as

$$\frac{1}{\tau\_{\text{eff}}} = \frac{1}{\tau\_b} + \frac{2S\_{\text{eff}}}{W},\tag{5}$$

where *W* is the thickness of the bulk silicon.

High carrier lifetime is a critical issue in achieving high‐energy conversion efficiency in solar cells. There are many techniques that have been developed over the past two decades with the objective of measuring minority carrier lifetime in Si [50]. Two techniques, namely the transient photoconductance decay (TPCD) and the more recent quasi‐steady‐state photoconductance method (QSSPC), are the basis of the WCT‐120 tester, which are widely used in this work to characterize the solar cell precursors.

**3.1. Recombination and passivation in Si‐based solar cells**

the surface, is particularly important.

centers.

[48, 49].

expressed as

Usually, the loss in photovoltaic solar cell comprises surface light loss, the transmitted light loss, thermalization loss (thermal electron loss). In a practical heterojunction (HJ) silicon solar cell, the a‐/µc‐Si:H layer is of several nanometers thick, which marginally absorbs sunlight. The photogeneration of excess charge carrier pairs mainly occur in the bulk c‐Si. In order to efficiently collect the photogenerated carriers which can be extracted and thus contribute to the externally retrievable current, reducing the carrier recombination in silicon, especially on

204 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The carrier recombination originating from bulk silicon primarily comprises two types, i.e., the intrinsic and extrinsic ones. The intrinsic recombination, which mainly includes radiative [46] and Auger recombination [47], occurs in a pure and defect‐free semiconductor, while the extrinsic recombination stems from contaminants and defects which act as recombination

*Rad Aug SRH b*

DD DD = + + =++ = (4)

 t

= + (5)

ttt

where *τSRH* denotes the contribution from the extrinsic recombination of Shock‐Read‐Hall effect

Surface recombination is a special case of SRH recombination due to the localized states appearing at the surface. Different from the bulk SRH centers, these states usually form a set of states across the band gap instead of a single energy level. Surface recombination is usually evaluated in terms of surface recombination velocity (SRV) instead of lifetime. However, the principles are same to bulk SRH recombination. It mainly originates from the surface dangling bond, i.e., the missing of crystalline network. The chemically and mechanically created defect can also result in surface recombination. In a simple condition with a constant bulk lifetime *τ*<sup>b</sup> and a small constant SRV (two surfaces are same to each other), the effective lifetime *τeff* is

1 1 <sup>2</sup> , *eff*

*S*

High carrier lifetime is a critical issue in achieving high‐energy conversion efficiency in solar cells. There are many techniques that have been developed over the past two decades with the objective of measuring minority carrier lifetime in Si [50]. Two techniques, namely the transient photoconductance decay (TPCD) and the more recent quasi‐steady‐state photoconductance

*eff b*

 t*W*

t

where *W* is the thickness of the bulk silicon.

Total bulk recombination is the sum of the above three terms and is expressed as

*nn nn UU U U*

*Rad Aug SRH*

Compared with the optical management by conventional methods like an antireflection layer, the minimization of electronic losses at the crystal silicon surface is a more delicate challenge for high‐efficiency solar cell [51]. Recombination losses at silicon interface or surface can be minimized by using two passivation schemes, namely, via chemical passivation or filed passivation. The former aims to reduce the interface defects that originate from the dangling bonds by H atom, a thin dielectric, or semiconductor film. The latter is based on the fact that the electron/hole density at the interface will be significantly reduced by formation of a built‐ in potential from the introduction of a passivation layer. Based on these two strategies, various Si‐based materials, such as hydrogenated amorphous silicon (a‐Si:H) [52], silicon nitride (SiN) [53], thermal oxygenated silicon dioxide (SiO2) [54], and amorphous silicon carbide (a‐SiCx) [55] have been investigated for passivation purposes. Among these materials, a‐Si:H is still the best candidate for silicon heterojunction solar cells [56, 57] due to its excellent passivation properties obtained at low deposition temperatures, and also simple processing without the introduction of any atoms other than silicon and hydrogen.

The a‐Si:H (including SiOx, SiNx and SiCx) passivation layers were mainly deposited by using conventional CCP‐PECVD or hot wire chemical vapor deposition (HWCVD) in literatures. A detailed review of c‐Si surface passivation can be found in Refs. [2] and [51], where the millisecond level lifetime and below tens cm/s surface recombination velocity prevails. Schmidt et al. [59] used a‐Si1−*x*N*x* deposited at a temperature 300 °C ≤ *T* ≤ 400 °C and obtained recombination velocity *S* ≤ 10 cm/s. Martin et al. [60] adopted a‐SiC*x*:H as a c‐Si surface passivation layer yielding *S* ≤ 30 cm/s. Dauwe et al. [61] achieved a low recombination velocities *S* ≤ 10 cm/s by 200–250 °C deposited a‐Si:H. Descoeudres et al. [52] adopted a VHF‐PECVD (40.68 MHz) to deposit intrinsic silicon film for c‐Si surface passivation and obtained a lifetime of 5.9 ms in N‐type FZ Si wafer with a resistivity of 4.0 Ωcm. Furthermore, the authors proposed a new indicator of silane depletion fraction for good passivation effect. It was reported that the good interface passivation corresponds to the highly depleted silane plasma. Willem et al. [2] prepared the a‐Si:H film using three different CVD techniques, namely, PECVD, VHF‐ PECVD, and HWCVD for passivation purpose and found that all three deposition methods yield excellent surface passivation with milliseconds level lifetime values.

However, the ICP‐grown a‐Si:H for c‐Si surface passivation is rarely reported [30], not to mention that grown by LFICP.

### **3.2. c‐Si surface passivation by semi‐remote LFICP CVD a‐/μc‐Si:H**

It was found that a‐/µc‐Si films deposited at direct LFICP region (with an antenna‐substrate distance *d* = 11 cm) had an unsatisfied passivation effect for c‐Si due to the damaged interface by severe ion bombardment. Therefore, the distance must be lengthened in order to yield the high‐quality interface and good passivation. In this subsection, the adopted distance is *d* = 33 cm, where a semi‐remote plasma is obtained.

### *3.2.1. Passivation effect of a‐/μc‐Si:H deposited from hydrogen‐diluted silane*

In this subsection, a‐/µc‐Si:H films deposited from only hydrogen‐diluted silane at the antenna‐substrate distance of *d* = 33 cm is used to passivate c‐Si surface with different resis‐ tivity. First, the surface passivation effect of a‐/µc‐Si:H grown under different discharge powers has been investigated in the case of a higher substrate resistivity of 8 Ωcm. The effect of pre‐deposition hydrogen plasma treatment, post thermal annealing, and the correlation between the lifetime and the thickness of the incubation layer at the interface have been analyzed. At this antenna‐substrate distance, the resulting films possess fairly good passiva‐ tion even without any post‐deposition thermal annealing.

The a‐/µc‐Si:H thin films of thickness of 50 nm were deposited on the double‐side polished, p‐ type CZ Si (comparatively higher resistivity of 8 Ωcm, thickness of 600 µm, (100) orientation, and area of about 3 × 3 cm<sup>2</sup> ). The pre‐deposition hydrogen plasma treatment on the substrates was applied for various durations to investigate the effect of hydrogen plasma exposure on the surface recombination. The microstructures and properties of the thicker counterpart of the thin passivation layer were studied in Section 2.1. The effective minority carrier lifetime of the symmetrically passivated wafer was measured by using the quasi‐steady state photocon‐ ductance decay (QSSPCD) technique on the facility of WCT‐120.

### *3.2.1.1. Effect of pre‐deposition H2 plasma treatment on surface passivation*

Prior to the passivation layer deposition, the substrates were treated by H<sup>2</sup> plasma for different durations (0–90 s) to improve the interface properties. The final structure for the minority carrier lifetime measurement is shown in inset of **Figure 22**. All the specific minority carrier lifetimes were recorded at an excess carrier of 1015/cm<sup>3</sup> .

**Figure 22.** RF power‐dependent minority carrier lifetime (filled squares) and incubation layer thickness (open squares). The inset shows the passivation scheme for the lifetime measurement.

τeff values are plotted in **Figure 23** as a function of H2 plasma treatment time. τeff is 127 µs without H<sup>2</sup> plasma treatment. An intermediate treatment time (30–60 s) is beneficial to silicon surface passivation, which is evidenced by the improved lifetime value. However, overlong treatment (>60 s) is detrimental to the surface quality, resulting in a less lifetime value of 73 µs. The maximum value of carrier lifetime ∼230 µs was obtained at a time of 30 s.

**Figure 23.** The minority carrier lifetime of the sample exposed to H2 plasma for different time before the passivation layer deposition.

### *3.2.1.2. Influence of RF power on surface passivation*

*3.2.1. Passivation effect of a‐/μc‐Si:H deposited from hydrogen‐diluted silane*

206 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

tion even without any post‐deposition thermal annealing.

ductance decay (QSSPCD) technique on the facility of WCT‐120.

lifetimes were recorded at an excess carrier of 1015/cm<sup>3</sup>

The inset shows the passivation scheme for the lifetime measurement.

*3.2.1.1. Effect of pre‐deposition H2 plasma treatment on surface passivation*

and area of about 3 × 3 cm<sup>2</sup>

In this subsection, a‐/µc‐Si:H films deposited from only hydrogen‐diluted silane at the antenna‐substrate distance of *d* = 33 cm is used to passivate c‐Si surface with different resis‐ tivity. First, the surface passivation effect of a‐/µc‐Si:H grown under different discharge powers has been investigated in the case of a higher substrate resistivity of 8 Ωcm. The effect of pre‐deposition hydrogen plasma treatment, post thermal annealing, and the correlation between the lifetime and the thickness of the incubation layer at the interface have been analyzed. At this antenna‐substrate distance, the resulting films possess fairly good passiva‐

The a‐/µc‐Si:H thin films of thickness of 50 nm were deposited on the double‐side polished, p‐ type CZ Si (comparatively higher resistivity of 8 Ωcm, thickness of 600 µm, (100) orientation,

was applied for various durations to investigate the effect of hydrogen plasma exposure on the surface recombination. The microstructures and properties of the thicker counterpart of the thin passivation layer were studied in Section 2.1. The effective minority carrier lifetime of the symmetrically passivated wafer was measured by using the quasi‐steady state photocon‐

Prior to the passivation layer deposition, the substrates were treated by H<sup>2</sup> plasma for different durations (0–90 s) to improve the interface properties. The final structure for the minority carrier lifetime measurement is shown in inset of **Figure 22**. All the specific minority carrier

**Figure 22.** RF power‐dependent minority carrier lifetime (filled squares) and incubation layer thickness (open squares).

τeff values are plotted in **Figure 23** as a function of H2 plasma treatment time. τeff is 127 µs without H<sup>2</sup> plasma treatment. An intermediate treatment time (30–60 s) is beneficial to silicon surface passivation, which is evidenced by the improved lifetime value. However, overlong

.

). The pre‐deposition hydrogen plasma treatment on the substrates

The RF power‐dependent effective lifetime τeff is shown in **Figure 22** (filled squares). This curve can be separated into three stages: In the first stage (1.0–1.5 kW), τeff decreases from 67 to 42 µs; in the second stage (1.5–2.0 kW), τeff rapidly increases from 42 µs to its maximum of 234 µs; in the third stage (2.0–2.5 kW) τeff drops down to 129 µs. The observed pattern of τeff will be explained in terms of the properties of the thin films and the interface between the thin film and silicon substrate.

The passivation effect described by τeff is not only related to the deposited thin films itself but also the a‐Si:H/c‐Si interface quality. The low defect density in the thin films is presumably a consequence of hydrogen passivation. As mentioned above, the hydrogen content in these thin films is comparatively larger. However, in the current experiments, the maximum value of τeff is obtained at RF power of 2.0 kW, where the corresponding thin film contains the least hydrogen as shown in **Figure 5**. Therefore, the change of lifetime τeff should not simply be attributed to the change of hydrogen content with the RF power.

Another crucial mechanism to determine the passivation effect is the interface defect between the passivation layer and the substrate. It has been observed that during the deposition of a µc‐Si:H layer on top of a substrate, a 30–50 nm thick interlayer referred as incubation zone [63] is produced. Its crystallinity and thickness strongly depend on the deposition condition and the substrate material. In our experiments, the incubation layer was directly observed in the high‐magnification cross‐sectional SEM images of the deposited thin films on double‐side‐ polished Si substrate. **Figure 24** shows the high‐magnification cross‐sectional SEM images of the thin films deposited at the power of 1.2 (a), 1.5 (b), 1.8 (c) and 2.0 kW (d). The clear scratches and cracks come from the cross‐section preparation process. Except the sample deposited at 2.0 kW, evident incubation layers were observed between the crystal Si and the films. Fur‐ thermore, the thickness of the incubation layers can be determined from the SEM measurement tool by making measurement multiple times at different locations and the results as a function of RF power are plotted in **Figure 22** (the open squares). The error bars shown in **Figure 22** represent the standard deviations from the mean thickness values from the multiple metering. At a power of 1.5 kW, the thickness has a maximum of 80 ′ 7 nm (>the standard passivation layer thickness of 50 nm), while the lifetime has a minimum of 42 µs. No evident incubation layer was observed at the power of 2.0 kW, at which the greatest minority lifetime value was obtained. Linking the curve of incubation layer thickness with that of the lifetime, one will find that the incubation layer thickness changes with the RF power in an opposite way to the minority carrier lifetime τeff dose. This observation suggests the fact that the incubation layer on the interface between the passivation layer and the crystal Si will worsen the passivation effect of the a‐/µc‐Si:H thin films.

**Figure 24.** The cross‐sectional SEM images of the thin films deposited at RF power of 1.2 (a), 1.5 (b), 1.8 (c), and 2.0 k W (d). Evident incubation layers were observed at the interfaces.

The CVD deposition of Si thin films involves a complex combination of several processes including the arrival and removal of gas molecules or precursors at the substrate surface, the decomposition into reactive species, and the migration of these species on the surface where they can lead to nucleation and continued deposition [64]. The adsorbed reactive species are likely to come to rest when a position of minimum energy position is found. The minimum energy position can be a defect at the substrate surface, which will result in a new nucleus, or an existing nucleus [65]. At certain process conditions, the formation of nuclei will be delayed for some time known as the incubation time [64]. From the point view of the growth mecha‐ nism, the incubation layer (initial growth) will be rich of defects, which will lead to the intense recombination of the photon‐induced carriers and thus a decreasing lifetime value observed in **Figure 22**.

### *3.2.1.3. Influence of thermal annealing on surface passivation*

thermore, the thickness of the incubation layers can be determined from the SEM measurement tool by making measurement multiple times at different locations and the results as a function of RF power are plotted in **Figure 22** (the open squares). The error bars shown in **Figure 22** represent the standard deviations from the mean thickness values from the multiple metering. At a power of 1.5 kW, the thickness has a maximum of 80 ′ 7 nm (>the standard passivation layer thickness of 50 nm), while the lifetime has a minimum of 42 µs. No evident incubation layer was observed at the power of 2.0 kW, at which the greatest minority lifetime value was obtained. Linking the curve of incubation layer thickness with that of the lifetime, one will find that the incubation layer thickness changes with the RF power in an opposite way to the minority carrier lifetime τeff dose. This observation suggests the fact that the incubation layer on the interface between the passivation layer and the crystal Si will worsen the passivation

208 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 24.** The cross‐sectional SEM images of the thin films deposited at RF power of 1.2 (a), 1.5 (b), 1.8 (c), and 2.0 k W

The CVD deposition of Si thin films involves a complex combination of several processes including the arrival and removal of gas molecules or precursors at the substrate surface, the decomposition into reactive species, and the migration of these species on the surface where they can lead to nucleation and continued deposition [64]. The adsorbed reactive species are likely to come to rest when a position of minimum energy position is found. The minimum

(d). Evident incubation layers were observed at the interfaces.

effect of the a‐/µc‐Si:H thin films.

In order to investigate the influence of the post‐deposition annealing on the silicon surface, the samples were thermally annealed in vacuum and H2 flow atmosphere. For simplicity, one sample with an as‐deposited lifetime value of 196 µs was annealed just in vacuum in the temperature range of 230–500 °C. **Figure 25(a)** shows the evolution of lifetime τeff with annealing temperatures. The annealing duration at each temperature point is 30 min. One can see that the lifetime increases almost linearly in the range of 230–420 °C. Further increasing of temperature to 500 °C results in a significant decrease in lifetime value. The relatively low temperature (<500 °C) annealing in vacuum leads to the lifetime improvement by a factor of about 2.5. The corresponding FTIR transmission spectra of the sample before and after thermal annealing were measured in the wavenumber range of 550–950 cm−1 to clarify the H role in the surface passivation. The spectra of the as-deposited and the 420 and 500 °C annealed samples are shown in **Figure 25(b)**. The Si‐H wagging/rocking mode at about 630 cm−1 slightly changes with temperature increasing from 230 to 420 °C. In comparison, this mode considerably recedes when temperature is increased from 420 to 500 °C, which is indicative of lower hydrogen content after annealing. The mode around 850–900 cm−1 also becomes weak after annealing at a temperature of 420 and 500 °C.

**Figure 25.** The minority carrier lifetime of the sample annealed at different temperatures in vacuum (a), and FTIR transmission spectra (550–950 cm−1) of the as‐deposited sample and that annealed at 420 and 500 °C (b). Increase in the annealing temperature from 420 to 500 °C leads to the decreasing hydrogen content in the film.

The vacuum annealing results discussed above represent the important role of the hydrogen atom in the Si surface passivation. Low‐temperature (<420 °C) annealing enables the diffusion of hydrogen in the thin film towards the Si substrate, which effectively reduces the dangling bond density at the interface and enhances the minority carrier lifetime. At higher tempera‐ tures (>420 °C), hydrogen atom effuses leading to the re‐generation of the dangling bonds and thus a decrease in the lifetime value. Systematic annealing experiments were performed on the samples deposited at different RF powers, and the ultimate lifetime values (not shown here) presented the same trend with that of the as‐deposited samples shown in **Figure 22** (filled squares). These results show that the as‐deposited interface features, especially the incubation layer properties discussed above, are the crucial factors in determining the ultimate lifetime values in our case. Therefore, we will focus on the sample with the highest as‐deposited lifetime value. It is worth stressing that the ultimate lifetime value is dependent on the annealing circumstance. **Figure 26** exhibits the lifetime value evolution with annealing time at 420 °C in vacuum (filled squares), H2 flow (filled circles), and N2 flow. In the case of vacuum annealing, the lifetime value increases sharply in the 95 min and saturates at the value of 421 µs at about 185 min. On the contrary, the lifetime reaches the maximum of 524 µs in a short time of 11 min in the case of H2 flow. It seems that the presence of hydrogen in the annealing circumstance is essential to improve surface passivation by present a‐Si:H/µc‐Si:H thin films, despite the presence of rich hydrogen in the film as evidenced by the FTIR result. With N2 flow, lifetime reaches its maximum value of 425 µs in 5 min, followed by a drastic drop for time longer than 5 mins. The H2 preference is confirmed by multiple annealing experiments. It is well known that N2 is the commonly used protective gas for post‐deposition annealing of passivation layer a‐Si. But H2 circumstance annealing is indeed better than N2 in our experiments. The under‐ lying mechanism needs a further study to clarify.

**Figure 26.** The minority carrier lifetime as a function of annealing duration at 420 °C at an atmosphere of vacuum (fil‐ led squares), H2 flow (filled circles), and N2 flow (open circles).

### *3.2.1.4. Injection level‐dependent surface recombination*

The vacuum annealing results discussed above represent the important role of the hydrogen atom in the Si surface passivation. Low‐temperature (<420 °C) annealing enables the diffusion of hydrogen in the thin film towards the Si substrate, which effectively reduces the dangling bond density at the interface and enhances the minority carrier lifetime. At higher tempera‐ tures (>420 °C), hydrogen atom effuses leading to the re‐generation of the dangling bonds and thus a decrease in the lifetime value. Systematic annealing experiments were performed on the samples deposited at different RF powers, and the ultimate lifetime values (not shown here) presented the same trend with that of the as‐deposited samples shown in **Figure 22** (filled squares). These results show that the as‐deposited interface features, especially the incubation layer properties discussed above, are the crucial factors in determining the ultimate lifetime values in our case. Therefore, we will focus on the sample with the highest as‐deposited lifetime value. It is worth stressing that the ultimate lifetime value is dependent on the annealing circumstance. **Figure 26** exhibits the lifetime value evolution with annealing time at 420 °C in vacuum (filled squares), H2 flow (filled circles), and N2 flow. In the case of vacuum annealing, the lifetime value increases sharply in the 95 min and saturates at the value of 421 µs at about 185 min. On the contrary, the lifetime reaches the maximum of 524 µs in a short time of 11 min in the case of H2 flow. It seems that the presence of hydrogen in the annealing circumstance is essential to improve surface passivation by present a‐Si:H/µc‐Si:H thin films, despite the presence of rich hydrogen in the film as evidenced by the FTIR result. With N2 flow, lifetime reaches its maximum value of 425 µs in 5 min, followed by a drastic drop for time longer than 5 mins. The H2 preference is confirmed by multiple annealing experiments. It is well known that N2 is the commonly used protective gas for post‐deposition annealing of passivation layer a‐Si. But H2 circumstance annealing is indeed better than N2 in our experiments. The under‐

210 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 26.** The minority carrier lifetime as a function of annealing duration at 420 °C at an atmosphere of vacuum (fil‐

lying mechanism needs a further study to clarify.

led squares), H2 flow (filled circles), and N2 flow (open circles).

The effective minority carrier lifetime τeff is related to the effective surface recombination velocity *Seff* through Equation 5. By neglecting the bulk lifetime (τ*b*→∞, we can get the maximum value of the surface velocity from this equation. The injection level‐dependent *Seff* with and without annealing is plotted in **Figure 27**. Comparing the results of the as‐deposited and annealed, one can see that the low‐temperature annealing leads to the decreasing of the surface recombination, especially in the range of the low injection levels. The as‐deposited sample has a surface recombination of about 150 cm/s. The annealing at 230 °C for 11 min results in the decrease of *Seff* above the injection level of 1 x 15/cm3 and increase of *Seff* below the same level. This abnormal increase in *Seff* possibly stems from the excitation of the adsorbed atoms in the circumstance, which acts as the additional recombination centers to deteriorate passivation effect. Subsequent annealing at higher temperatures causes the continuous decreasing towards the injection level of 2.8 x 16/cm3 , where the Auger recombination plays a dominant role and the curves tend to converge. At the temperature of 420 °C, the lowest surface recombination velocity is 60 cm/s at the injection level of 1 x 15/cm3 . It should be emphasized that this *Seff* value is the upper limit because of the adopting of an infinity assignment for the bulk lifetime.

**Figure 27.** Injection level‐dependent surface recombination velocity of the as‐deposited sample and that annealed at different temperatures.

### *3.2.1.5. The high as‐deposited lifetime values in the case of semi‐remote LFICP*

The present as‐deposited lifetime value of 234 µs is much higher than that reported in literatures, where PECVD or HWCD methods were used to deposit the passivation a‐/µc‐Si‐ H layer. Jan‐Willem et al. [62] summarized the passivation effects of a‐Si:H deposited by various methods: 34 µs for PECVD, 43 µs for VHF PECVD, and 87 µs for HWCVD. In principle, passivation can be physically divided into two types, i.e., hydrogen‐related chemical passiva‐ tion and built‐in potential related field passivation. In the former, defect density can be remarkably lowered due to the reduced number of dangling bonds which is terminated by hydrogen atoms. With regard to the filed passivation, a‐/µc‐Si:H has a wider band gap than crystal silicon substrate, which results in a barrier on the interface preventing the recombina‐ tion of photogenerated carrier pairs through the surface/interface defects.

It has been shown that the pre‐deposition hydrogen plasma exposure significantly influences the surface passivation (see **Figure 23**). In the post‐deposition annealing, the interaction between hydrogen atom from the passivation layer and the c‐Si substrate is crucial to the surface passivation. Mitchell et al. [66] studied the thermal activation of the hydrogen‐related chemical passivation on silicon substrate and determined the thermal activation energy of *EA* through the following formula:

$$1\,\,\tau\_{\text{vac}} = A \exp(-E\_A \,/\, k\_B T) \tag{6}$$

where *τreac* is the reaction time between hydrogen atoms and silicon atoms, *EA* the activation energy of surface passivation, *kB* the Boltzmann constant, and *T* the temperature. According to Equation (6), a low *EA* value will point to a high reaction rate 1/*τreac* or a short reaction time *τreac*. In earlier works [33, 39], we analyzed the effect of high‐density hydrogen plasma on the crystallization mechanism of LFICP‐grown µc‐Si:H. Abundant atomic hydrogen resulting from the dissociation of SiH4 and H2 will result in high surface coverage by hydrogen, which reduces the activation energy *EA* and the reaction time *τreac*. In addition, high‐density excited hydrogen atoms will recombine with the precursor of Si‐H in the way described by the reaction H + Si‐H→Si‐ + H2 (with the dangling bond Si‐). This recombination process can produce effective local plasma heating on the growing surface, which will contribute to the hydrogen atom transport to the interface and bulk silicon and a considerable lifetime value even without any additional thermal annealing.

### *3.2.1.6. Incubation layer and its thickness control*

**Figure 28(a)** [**(b)**] shows a low‐ (high‐) magnification cross‐sectional TEM micrograph of LFICP‐grown µc‐Si:H on silicon substrate. In **Figure 28(a)**, a clear incubation layer (dark area) between the film and substrate is discriminated like in SEM images. **Figure 28(b)** exhibits more details of the interface. The grey area in the film corresponds to the microcrystalline Si fractions, which are conical conglomerates of microcrystals. As observed above, the incubation layer is detrimental for the improvement of lifetime and its thickness is highly dependent on the process parameters, e.g., the applied RF power in present experiments. Therefore, a deep understanding on the mechanism to control the formation of incubation layer is extremely important. It was recognized that the ion bombardment played a crucial role in the formation of incubation layer [67]. The ion beam works in two possible ways, namely (i) a surface effect: very small grains do not survive the high dose of ion irradiation, thereby reducing the nucleation rate, (ii) a bulk effect: the ion beam induces defect‐related grains growth. Accord‐ ingly, two strategies [67] to control the ion beam were proposed: the choice of the frequency of the power source generating the plasma, and the design of the electrode configuration to avoid its formation. In present antenna‐substrate distance *d* = 33 cm, most of the ions/electrons energy is around several eVs, below the threshold value of 16 eV for defect formation for Si. Therefore, the lengthened distance *d* makes it possible to minimize the effect of the ion bombardment to an acceptable level.

**Figure 28.** A typical cross‐sectional TEM micrograph of LFICP‐grown µc‐Si:H films on Si substrate at low (a) and high (b) magnification.

From the point of view of a growth mechanism, the key to diminish the initial growth of incubation layer seems to be the rapid nucleation without retard on the interface. In present experiment, this occurs at the power of 2.0 kW, near the transition point (1.8 kW) from a‐Si:H to µc‐Si:H. This is consistent with some authors' findings that the good passivation effect is obtained with microcrystalline‐oriented passivation layers, where the depletion fraction of SiH4 is comparatively higher than that of amorphous regimes [52, 68]. In the current study, although the silane depletion was not experimentally measured at certain RF power, the high depletion fraction at the power of 2.0 kW is expected. The detailed relation between the depletion fraction and the incubation time in the growth of a‐/µc‐Si:H needs to be further clarified.

### *3.2.1.7. Passivation of low resistivity c‐Si*

various methods: 34 µs for PECVD, 43 µs for VHF PECVD, and 87 µs for HWCVD. In principle, passivation can be physically divided into two types, i.e., hydrogen‐related chemical passiva‐ tion and built‐in potential related field passivation. In the former, defect density can be remarkably lowered due to the reduced number of dangling bonds which is terminated by hydrogen atoms. With regard to the filed passivation, a‐/µc‐Si:H has a wider band gap than crystal silicon substrate, which results in a barrier on the interface preventing the recombina‐

212 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

It has been shown that the pre‐deposition hydrogen plasma exposure significantly influences the surface passivation (see **Figure 23**). In the post‐deposition annealing, the interaction between hydrogen atom from the passivation layer and the c‐Si substrate is crucial to the surface passivation. Mitchell et al. [66] studied the thermal activation of the hydrogen‐related chemical passivation on silicon substrate and determined the thermal activation energy of *EA*

1 / exp(- / ) *reac A E kT A B*

where *τreac* is the reaction time between hydrogen atoms and silicon atoms, *EA* the activation energy of surface passivation, *kB* the Boltzmann constant, and *T* the temperature. According to Equation (6), a low *EA* value will point to a high reaction rate 1/*τreac* or a short reaction time *τreac*. In earlier works [33, 39], we analyzed the effect of high‐density hydrogen plasma on the crystallization mechanism of LFICP‐grown µc‐Si:H. Abundant atomic hydrogen resulting from the dissociation of SiH4 and H2 will result in high surface coverage by hydrogen, which reduces the activation energy *EA* and the reaction time *τreac*. In addition, high‐density excited hydrogen atoms will recombine with the precursor of Si‐H in the way described by the reaction H + Si‐H→Si‐ + H2 (with the dangling bond Si‐). This recombination process can produce effective local plasma heating on the growing surface, which will contribute to the hydrogen atom transport to the interface and bulk silicon and a considerable lifetime value even without

**Figure 28(a)** [**(b)**] shows a low‐ (high‐) magnification cross‐sectional TEM micrograph of LFICP‐grown µc‐Si:H on silicon substrate. In **Figure 28(a)**, a clear incubation layer (dark area) between the film and substrate is discriminated like in SEM images. **Figure 28(b)** exhibits more details of the interface. The grey area in the film corresponds to the microcrystalline Si fractions, which are conical conglomerates of microcrystals. As observed above, the incubation layer is detrimental for the improvement of lifetime and its thickness is highly dependent on the process parameters, e.g., the applied RF power in present experiments. Therefore, a deep understanding on the mechanism to control the formation of incubation layer is extremely important. It was recognized that the ion bombardment played a crucial role in the formation of incubation layer [67]. The ion beam works in two possible ways, namely (i) a surface effect: very small grains do not survive the high dose of ion irradiation, thereby reducing the nucleation rate, (ii) a bulk effect: the ion beam induces defect‐related grains growth. Accord‐

= (6)

tion of photogenerated carrier pairs through the surface/interface defects.

t

through the following formula:

any additional thermal annealing.

*3.2.1.6. Incubation layer and its thickness control*

It is more difficult to passivate a lower‐resistivity Si surface due to the defect associated with more recombination. However, the lower‐resistivity wafer is more meaningful in practical solar cell application. In this part, two types of wafers have been investigated: p‐type (100), 2 Ωcm, and n‐type (100), 3 Ωcm wafers. The process parameters such as discharge power, substrate temperature, dilution ratio, and post‐deposition annealing have been widely explored to improve their surface passivation. **Figure 29** shows the effective minority carrier lifetime τeff in the n‐type 3 Ωcm c‐Si passivated with a‐Si:H deposited at different substrate temperatures. τeff drastically drops from 164 µs to 25 µs with increasing substrate temperature from room temperature to 140 °C. This is possibly originated from the local epitaxial growth or reduced hydrogen content with increasing temperature.

**Figure 29.** As‐deposited lifetime values as a function of the substrate temperature. The other parameters are: SiH4:H2 = 12.5:12.5 (in sccm), discharge power 1.5 kW.

**Figure 30** displays τeff in both types c‐Si passivated by a‐Si:H deposited from hydrogen‐diluted silane with different dilution ratios (H2/SiH4). The discharge power and substrate temperature were kept at 2.0 kW and room temperature, respectively. The flow rate of SiH4 and H2 is 20:0, 12.5:12.5, 8.3:16.7, 5:20, corresponding to the hydrogen dilution ratio of 0, 1, 2, and 4, respec‐ tively. τeff reaches its maximum values at the dilution ratio of 1 in both p‐ and n‐type c‐Si. Based on the above analysis on passivation, the higher depletion of silane and reduced incubation layer are expected. **Figure 31** shows the discharge power‐dependent lifetime values, which reached its maximum at RF power of 2.5 kW. Above 2.5 kW, τeff begins to drop again.

**Figure 30.** As‐deposited lifetime values in both type c‐Si passivated by a‐Si:H grown from hydrogen‐diluted silane with different dilution ratios.

High‐Density Plasma‐Enhanced Chemical Vapor Deposition of Si‐Based Materials for Solar Cell Applications http://dx.doi.org/10.5772/63529 215

from room temperature to 140 °C. This is possibly originated from the local epitaxial growth

214 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 29.** As‐deposited lifetime values as a function of the substrate temperature. The other parameters are: SiH4:H2 =

**Figure 30** displays τeff in both types c‐Si passivated by a‐Si:H deposited from hydrogen‐diluted silane with different dilution ratios (H2/SiH4). The discharge power and substrate temperature were kept at 2.0 kW and room temperature, respectively. The flow rate of SiH4 and H2 is 20:0, 12.5:12.5, 8.3:16.7, 5:20, corresponding to the hydrogen dilution ratio of 0, 1, 2, and 4, respec‐ tively. τeff reaches its maximum values at the dilution ratio of 1 in both p‐ and n‐type c‐Si. Based on the above analysis on passivation, the higher depletion of silane and reduced incubation layer are expected. **Figure 31** shows the discharge power‐dependent lifetime values, which

reached its maximum at RF power of 2.5 kW. Above 2.5 kW, τeff begins to drop again.

**Figure 30.** As‐deposited lifetime values in both type c‐Si passivated by a‐Si:H grown from hydrogen‐diluted silane

or reduced hydrogen content with increasing temperature.

12.5:12.5 (in sccm), discharge power 1.5 kW.

with different dilution ratios.

**Figure 31.** As‐deposited lifetime values in N‐type 3 Ωcm c‐Si passivated by a‐Si:H grown under different RF powers.

The optimized lifetime results of these two types of low‐resistivity c‐Si have been shown in **Figure 32**. At the specific excess carrier density of 1015/cm3 , the lifetime value is 305 and 150 µs for the n‐type 3 Ωcm and p‐type 2 Ωcm c‐Si, respectively. It should be noted that these results are as‐deposited values without post‐deposition annealing. Unfortunately, post‐ deposition annealing cannot improve the lifetime values anymore in the present low‐resistivity c‐Si substrates. The hydrogen chemical annealing mentioned above has already raised the lifetime values to its maximum during the deposition process.

**Figure 32.** Injection level‐dependent minority carrier lifetime in two types of low‐resistivity c‐Si wafers: p‐type, 2 Ωcm and n‐type, 3 Ωcm.

Although the as‐deposited values without post‐deposition long time annealing are much bigger than that reported in references, the present lifetime values are still lower than the ms level results in the literatures. The plasma density in the growth region has already been reduced a lot by increasing the antenna‐substrate distance *d* to 33 cm. The semi‐remote LFICP grown a‐/µc‐Si provides a very high as‐deposited lifetime value in c‐Si (low and high resis‐ tivity). In turn, these high as‐deposited lifetime values reflect the still high‐density plasma in the present growth area, which provides effective chemical annealing during the growth process. On the other hand, the high‐density plasma makes the growth surface exposed to a pronounced ion bombardment, which will damage the interface between the substrate and the passivation layer. In terms of process parameters, the high‐discharge power points to the high silane depletion. Meanwhile, the ion bombardment on the substrate is inevitably enhanced under the increasing discharge power. In this sense, the key to improve the passi‐ vation effect lies in how to balance the silane depletion and ion bombardment on the interface.

### *3.2.2. Introduction of CO2 into the reactive gas*

a‐Si:H is easy for (partially) epitaxial growth on c‐Si even at low deposition temperatures, which will severely reduce the carrier lifetime values and thus the photovoltaic performance of HJ solar cell [56]. Incorporation of O into the amorphous silicon network can effectively suppress the epitaxial growth [69]. In the aspect of optics, the inherent blue light loss in a‐Si:H layer also restricts the solar cell optimization. As a consequence, wide‐banded Si‐based materials become a promising candidate. Thermally grown SiO<sup>2</sup> provides a state‐of‐art level of surface passivation with a reduced surface recombination velocity as low as 2 cm/s [54]. However, the thermally grown SiO2 suffers from the inherent disadvantages such as extremely high process temperature ((1000 °C) and low deposition rate. As a consequence, non‐stoichio‐ metric silicon oxide (SiO*x*) based on non‐thermal deposition methods have been proposed as an alternative to SiO2. Apart from conventional PECVD and ESR, VHF‐PECVD (13.56–110 MHz) has also been used to grow SiO*x*:H for c‐Si passivation [70, 71]. Mueller et al. [71] investigated the RF frequency effect on the passivation behavior for n‐type FZ silicon wafer (resistivity (0.5 Ωcm) and obtained the minority carrier lifetime of 480 µs at the optimum frequency of 70 MHz. Hydrogen‐diluted SiH4 acts as the feedstock gas, whereas the oxygen source varies from oxygen gas [72] to various composites such us N2O [73] and CO2 [74].

In this section, CO<sup>2</sup> is introduced into the reactive gases of SiH4+H<sup>2</sup> and efficiently dissociated in the LFICP system. The discharge power is set at 1.8 kW, flow rate of both SiH4 and H2 are 12.5 sccm, and no external heating is applied to the substrate. Various flow rates of CO<sup>2</sup> are introduced into the deposition chamber. The microstructure evolution with the introduction of CO2 in the obtained films is investigated. Two types of c‐Si wafers: n‐type, 3 Ωcm and p‐ type, 2 Ωcm, are passivated by these films. The underlying passivation mechanism is analyzed.

**Figure 33** shows *Eg* and deposition rate as functions of CO2 flow rate. The incorporation of CO2 into the reactive gases results in the effective widening of band gap *Eg* and the pronounced decrease in the deposition rate. *Eg* increases by about 0.3 eV with CO<sup>2</sup> increased from 0 to 12.5 sccm. According to Watanabe [75], SiO:H is a two‐phase material with an island of SiO in a matrix of a‐Si:H. In their two‐phase model, the oxygen‐rich phase is effective in increasing *Eg* and the Si‐rich phase contributes to high conductivity. The widened *Eg* is beneficial to the light transmittance through the window layer when SiO*x*:H is used as the window layer. Actually, Sritharathikhun et al. has succeeded in fabricating SiO*x*:H HIT solar cell on n‐type [76] and p‐ type [77] silicon substrate and obtained the competitive photovoltaic performance efficiency of 17.9% for n‐type substrate, and 15.3% for p‐type substrate.

High‐Density Plasma‐Enhanced Chemical Vapor Deposition of Si‐Based Materials for Solar Cell Applications http://dx.doi.org/10.5772/63529 217

**Figure 33.** Deposition rate and optical band gap *Eg* evolution with CO2 flow rate.

process. On the other hand, the high‐density plasma makes the growth surface exposed to a pronounced ion bombardment, which will damage the interface between the substrate and the passivation layer. In terms of process parameters, the high‐discharge power points to the high silane depletion. Meanwhile, the ion bombardment on the substrate is inevitably enhanced under the increasing discharge power. In this sense, the key to improve the passi‐ vation effect lies in how to balance the silane depletion and ion bombardment on the interface.

216 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

a‐Si:H is easy for (partially) epitaxial growth on c‐Si even at low deposition temperatures, which will severely reduce the carrier lifetime values and thus the photovoltaic performance of HJ solar cell [56]. Incorporation of O into the amorphous silicon network can effectively suppress the epitaxial growth [69]. In the aspect of optics, the inherent blue light loss in a‐Si:H layer also restricts the solar cell optimization. As a consequence, wide‐banded Si‐based materials become a promising candidate. Thermally grown SiO<sup>2</sup> provides a state‐of‐art level of surface passivation with a reduced surface recombination velocity as low as 2 cm/s [54]. However, the thermally grown SiO2 suffers from the inherent disadvantages such as extremely high process temperature ((1000 °C) and low deposition rate. As a consequence, non‐stoichio‐ metric silicon oxide (SiO*x*) based on non‐thermal deposition methods have been proposed as an alternative to SiO2. Apart from conventional PECVD and ESR, VHF‐PECVD (13.56–110 MHz) has also been used to grow SiO*x*:H for c‐Si passivation [70, 71]. Mueller et al. [71] investigated the RF frequency effect on the passivation behavior for n‐type FZ silicon wafer (resistivity (0.5 Ωcm) and obtained the minority carrier lifetime of 480 µs at the optimum frequency of 70 MHz. Hydrogen‐diluted SiH4 acts as the feedstock gas, whereas the oxygen source varies from oxygen gas [72] to various composites such us N2O [73] and CO2 [74].

In this section, CO<sup>2</sup> is introduced into the reactive gases of SiH4+H<sup>2</sup> and efficiently dissociated in the LFICP system. The discharge power is set at 1.8 kW, flow rate of both SiH4 and H2 are 12.5 sccm, and no external heating is applied to the substrate. Various flow rates of CO<sup>2</sup> are introduced into the deposition chamber. The microstructure evolution with the introduction of CO2 in the obtained films is investigated. Two types of c‐Si wafers: n‐type, 3 Ωcm and p‐ type, 2 Ωcm, are passivated by these films. The underlying passivation mechanism is analyzed.

**Figure 33** shows *Eg* and deposition rate as functions of CO2 flow rate. The incorporation of CO2 into the reactive gases results in the effective widening of band gap *Eg* and the pronounced decrease in the deposition rate. *Eg* increases by about 0.3 eV with CO<sup>2</sup> increased from 0 to 12.5 sccm. According to Watanabe [75], SiO:H is a two‐phase material with an island of SiO in a matrix of a‐Si:H. In their two‐phase model, the oxygen‐rich phase is effective in increasing *Eg* and the Si‐rich phase contributes to high conductivity. The widened *Eg* is beneficial to the light transmittance through the window layer when SiO*x*:H is used as the window layer. Actually, Sritharathikhun et al. has succeeded in fabricating SiO*x*:H HIT solar cell on n‐type [76] and p‐ type [77] silicon substrate and obtained the competitive photovoltaic performance efficiency

of 17.9% for n‐type substrate, and 15.3% for p‐type substrate.

*3.2.2. Introduction of CO2 into the reactive gas*

**Figure 34** shows the FTIR absorbance spectra of the corresponding films. Besides the Si‐H waging/rocking signal ((640 cm−1) and Si‐H stretching signal ((2100 cm−1) usually observed in a‐Si:H, some oxygen and carbon‐associated modes appear. The Si‐C stretching mode, which couples with the doublet signal of (SiH2)*x* ((845–890 cm−1) in the case of low CO2 flow rate, grows stronger, and eventually dominates the latter. The Si‐O‐Si stretching mode located around 1100 cm−1 gradually shifts from 1107 to 1120 cm−1. The former and latter are assigned to the nature of SiO2 and SiO*x*, respectively [78]. At minor CO2 addition, the distinct peak at 1107 cm−1 is possibly due to the remaining oxygen in the deposition chamber. The consecutive increase of CO2 makes the structure change from SiO2 to SiO*x*. Another effect of CO2 addition is the shift of the Si‐H stretching signal, indicating the incorporation of O into the Si‐H bond. In literatures, carbon was regarded as being excluded because of the formation of volatile CO. However, the FTIR method tracks carbon in the form of Si‐C bond in our experiment. There‐ fore, the obtained films can be strictly denoted as SiC1−*x*O*x*:H.

**Figure 34.** FTIR absorptance spectra of the deposited films with varying CO2 flow rates.

About 60 nm thick SiC1−*<sup>x</sup>*O*x*:H films were deposited on both sides of the two types of c‐Si substrate and the surface passivation was examined by the method of steady‐state photocon‐ ductance decay method described above. **Figure 35** shows the effective lifetime values of these two substrates passivated by the SiC1−*<sup>x</sup>*O*x* :H films deposited with different flux of CO2. At first glance, τeff degrades with the increasing addition of CO2. This should be correlated to the increasing defect from the introduction of C and O into the network. However, incorporation of minor CO2 is beneficial to the surface passivation. As an example, lifetime τeff has a highest value at flux of 2.5 sccm (1.25 sccm) in n‐type (p‐type) c‐Si. The optimized injection level‐ dependent lifetime values before and after thermal annealing (275 °C, 30 min) have been presented in **Figure 36**. The obtained best results for n‐type 3 Ωcm and p‐type 2Ωcm are 313 and 166 µs at the specific carrier density of 1015/cm3 , respectively.

**Figure 35.** The as‐deposited lifetime values in two types of c‐Si substrates passivated by SiC1−*x*O*x*:H deposited with dif‐ ferent flux of CO2. The lifetime values are taken at the specific excess carrier density of 1015/cm3 .

**Figure 36.** Injection level‐dependent lifetime values in the substrate of n‐type 3 Ωcm (a) and p‐type 2 Ωcm (b) before and after the thermal annealing. The fluxes of CO2 are 2.5 and 1.25 sccm in (a) and (b), respectively.

In order to clarify the passivation mechanism for the LFICP‐grown SiC1−*x*O*x*:H, high‐frequency (1MHz) C‐V measurements were performed on a metal‐semiconductor‐insulator (MIS) structure shown in the inset of **Figure 37**. The insulator layer is SiC1−*x*O*x*:H material. The front contact is a small Al dot with a diameter about 0.5 mm and the rear contact is sheet Al. Al electrode was deposited by the method of magnetron sputtering. The amplitude of the AC signal was 10 mV. The work function difference of Al and silicon was kept at −0.85 V [79]. The recorded C‐V curve is shown in **Figure 37**, from which a fixed charge density of −2.7×1011/cm<sup>2</sup> was deduced. However, this level is not high enough to contribute to the field passivation, which is usually accompanied by a fixed charge density of the order of 1012/cm<sup>2</sup> [74]. Therefore, the passivation effect of the present SiC1−*x*O*x*:H mainly stems from the chemical passivation like in a‐Si:H case.

**Figure 37.** The C‐V curve of the metal‐semiconductor‐insulator (MIS) structure, from which the fixed charge in the in‐ sulator can be deduced.

### **3.3. c‐Si surface passivation by remote LFICP CVD a‐Si:H**

About 60 nm thick SiC1−*<sup>x</sup>*O*x*:H films were deposited on both sides of the two types of c‐Si substrate and the surface passivation was examined by the method of steady‐state photocon‐ ductance decay method described above. **Figure 35** shows the effective lifetime values of these two substrates passivated by the SiC1−*<sup>x</sup>*O*x* :H films deposited with different flux of CO2. At first glance, τeff degrades with the increasing addition of CO2. This should be correlated to the increasing defect from the introduction of C and O into the network. However, incorporation of minor CO2 is beneficial to the surface passivation. As an example, lifetime τeff has a highest value at flux of 2.5 sccm (1.25 sccm) in n‐type (p‐type) c‐Si. The optimized injection level‐ dependent lifetime values before and after thermal annealing (275 °C, 30 min) have been presented in **Figure 36**. The obtained best results for n‐type 3 Ωcm and p‐type 2Ωcm are 313

218 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 35.** The as‐deposited lifetime values in two types of c‐Si substrates passivated by SiC1−*x*O*x*:H deposited with dif‐

**Figure 36.** Injection level‐dependent lifetime values in the substrate of n‐type 3 Ωcm (a) and p‐type 2 Ωcm (b) before

and after the thermal annealing. The fluxes of CO2 are 2.5 and 1.25 sccm in (a) and (b), respectively.

ferent flux of CO2. The lifetime values are taken at the specific excess carrier density of 1015/cm3

, respectively.

.

and 166 µs at the specific carrier density of 1015/cm3

In subsection 3.2.2, reactive gas CO2 was introduced into SiH4+H2 to improve the c‐Si surface passivation. The introduction of CO2 can widen the optical band gap of the passivation layer, and simultaneously suppress the partial epitaxial growth of the passivation layer on c‐Si. The passivated sample exhibits high as‐deposited lifetime values, whereas the lifetime values cannot be enhanced effectively after post‐deposition annealing. It is believed that the frequent ion bombardment from the high‐density plasma should be responsible for these two effects: it conveys energy to the growing surface acting as plasma annealing, and possibly results in damages to the interface. The damaged interface will define the hydrogen diffusion to the interface when annealing and results in a limited enhancement in lifetime. One possible solution is to keep the substrate away from the plasma generation space. In this section, the antenna‐substrate distance *d* is further enlarged to *d* = 53 cm to form remote plasma, and thus further decrease the ion bombardment on the surface.

**Figure 38** shows the lifetime τeff curves before and after the thermal annealing in 275 °C for different durations. In **Figure 38(a)**, the as‐deposited τeff value is 3 µs, and τeff surges upon annealing. The similar behavior is observed in the case of the n‐type 4 Ωcm c‐Si. The saturation values of 295 and 716 µs are achieved in p‐type 1 Ωcm and p‐type 4 Ωcm c‐Si, respectively. The lifetime values τeff reaches their maximums at annealing duration of 45 min. The present annealing behavior is different from that in the case of antenna‐substrate distance *d* = 33 cm. Herein, τeff is improved by more than two orders of magnitude after thermal annealing. This fact supports the assumption of the ion bombardment effect accompanying with a lower distance *d* value.

**Figure 38.** Injection level‐dependent lifetime values in the p‐type 1 Ωcm (a) and n‐type 3 Ω cm (b) c‐Si, before and after the thermal annealing. The thermal annealing was conducted at 275 °C in N2 atmosphere for different durations.

In a narrow tunable working pressure range, we have compared the lifetime curve of the passivated samples under different pressures. **Figure 39** shows the comparison between the two discharge pressures of 0.2 and 1.0 Pa. The lifetime values are measured in the passivated c‐Si substrates (n‐type 4 Ωcm) before and after the thermal annealing treatment. The as‐ deposited lifetime values at 0.2 Pa are much greater than that at 1.0 Pa. Nevertheless, the vice verse is true after the thermal annealing. As mentioned above, a higher pressure means a smaller mean free path of the particles and a higher deposition rate. A faster growth can minimize the partial epitaxial growth of a‐Si:H on c‐Si surface. In addition, the lower‐pressure mode possibly still causes some surface defects which limit the beneficial effect of thermal annealing.

solution is to keep the substrate away from the plasma generation space. In this section, the antenna‐substrate distance *d* is further enlarged to *d* = 53 cm to form remote plasma, and thus

220 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 38** shows the lifetime τeff curves before and after the thermal annealing in 275 °C for different durations. In **Figure 38(a)**, the as‐deposited τeff value is 3 µs, and τeff surges upon annealing. The similar behavior is observed in the case of the n‐type 4 Ωcm c‐Si. The saturation values of 295 and 716 µs are achieved in p‐type 1 Ωcm and p‐type 4 Ωcm c‐Si, respectively. The lifetime values τeff reaches their maximums at annealing duration of 45 min. The present annealing behavior is different from that in the case of antenna‐substrate distance *d* = 33 cm. Herein, τeff is improved by more than two orders of magnitude after thermal annealing. This fact supports the assumption of the ion bombardment effect accompanying with a lower

**Figure 38.** Injection level‐dependent lifetime values in the p‐type 1 Ωcm (a) and n‐type 3 Ω cm (b) c‐Si, before and after the thermal annealing. The thermal annealing was conducted at 275 °C in N2 atmosphere for different durations.

In a narrow tunable working pressure range, we have compared the lifetime curve of the passivated samples under different pressures. **Figure 39** shows the comparison between the two discharge pressures of 0.2 and 1.0 Pa. The lifetime values are measured in the passivated c‐Si substrates (n‐type 4 Ωcm) before and after the thermal annealing treatment. The as‐ deposited lifetime values at 0.2 Pa are much greater than that at 1.0 Pa. Nevertheless, the vice verse is true after the thermal annealing. As mentioned above, a higher pressure means a smaller mean free path of the particles and a higher deposition rate. A faster growth can minimize the partial epitaxial growth of a‐Si:H on c‐Si surface. In addition, the lower‐pressure mode possibly still causes some surface defects which limit the beneficial effect of thermal

further decrease the ion bombardment on the surface.

distance *d* value.

annealing.

**Figure 39.** Injection level‐dependent lifetime curves of the n‐type 4 Ωcm c‐Si passivated by a‐Si:H deposited under dif‐ ferent working pressures.

**Figure 40** shows the lifetime curves in two types of c‐Si substrates passivated by a‐Si:H deposited under two temperature points of RT and 140 °C. For the p‐type 2 Ωcm c‐Si substrate, substrate temperature does not strongly influence the lifetime curve. In comparison, increasing temperature from RT to 140 °C deteriorates the lifetime value severely in n‐type 4 Ωcm c‐Si substrate. This difference is possibly from the different conductivity values which strongly influence temperature‐dependent hydrogen diffusion on the surface or at the interface. The detailed mechanism needs further study to clarify.

**Figure 40.** Injection level‐dependent lifetime curve of two types of c‐Si substrates passivated by a‐Si:H deposited under different substrate temperatures.

**Figure 41** shows the best lifetime results we obtained in three types of c‐Si substrates. At the excess carrier density of 1015/cm3 , the best lifetime values are 1.3 ms, 620 µs, and 290 µs for the n‐type 4 Ωcm, p‐type 2 Ωcm, and p‐type 1 Ωcm c‐Si substrates, respectively.

**Figure 41.** The best lifetime curves obtained in three types of c‐Si substrates passivated by a‐Si:H deposited in remote LFICP with an antenna‐substrate distance of 53 cm.

### **3.4. c‐Si passivation by stack configuration and implied** *Voc*

It is well known that the passivation effect is strongly dependent on the passivation layer thickness. In a practical HIT solar cell, the total thickness of deposited functional layer on one side is several tens of nanometers, less than the commonly used thickness of 50 nm in passi‐ vation studies. In addition, in a practical HIT solar cell, the passivation configuration is a stack of intrinsic and doped a‐Si:H layers instead of above only intrinsic layer. Therefore, it is important to study the stack layer passivation configuration for the HIT solar cell applications. The passivation effect associated open‐circuit voltage, namely, the implied *Voc*, can be also determined from the QSSPC method. Using this method, we can estimate *Voc* of the solar cell precursor (without the metallization procedure).

Intrinsic and doped a‐Si:H films were symmetrically deposited on the p‐type 1 Ωcm c‐Si substrates to form the passivation configurations of n/i/P‐c‐Si/i/n or p/i/P‐c‐Si/i/p. This stack passivation scheme combines the chemical passivation (by intrinsic a‐Si:H) and the field passivation (by doped a‐Si:H). As an example, **Figure 42** shows the field passivation by p‐a‐ Si:H in (a) and the combination of both passivation mechanisms in (b). Addition of a p‐doped layer introduces an additional electrical filed to hinder the minority diffusion to the surface. The barrier ФBSF exponentially reduces the concentration of minority at the interface and thus the surface recombination. In **Figure 42(b)**, the interface defect density is reduced by the chemical passivation of addition of an intrinsic layer. Then, the solar cell precursor with the structure of n/i/P/i/p was fabricated to estimate the lifetime values. In these structures, the intrinsic and doped layer thicknesses are about 5 and 15 nm, respectively, estimated from the average deposition rate at corresponding deposition conditions. **Figure 43** shows the lifetime curves in different passivation configurations. Lifetime value τeff in p/i/P/i/p is much lower than that in n/i/P/i/n. It is reported that B dopant is prone to react with Si atom to form complex, increasing the interface defect density [80].

**Figure 41** shows the best lifetime results we obtained in three types of c‐Si substrates. At the

**Figure 41.** The best lifetime curves obtained in three types of c‐Si substrates passivated by a‐Si:H deposited in remote

It is well known that the passivation effect is strongly dependent on the passivation layer thickness. In a practical HIT solar cell, the total thickness of deposited functional layer on one side is several tens of nanometers, less than the commonly used thickness of 50 nm in passi‐ vation studies. In addition, in a practical HIT solar cell, the passivation configuration is a stack of intrinsic and doped a‐Si:H layers instead of above only intrinsic layer. Therefore, it is important to study the stack layer passivation configuration for the HIT solar cell applications. The passivation effect associated open‐circuit voltage, namely, the implied *Voc*, can be also determined from the QSSPC method. Using this method, we can estimate *Voc* of the solar cell

Intrinsic and doped a‐Si:H films were symmetrically deposited on the p‐type 1 Ωcm c‐Si substrates to form the passivation configurations of n/i/P‐c‐Si/i/n or p/i/P‐c‐Si/i/p. This stack passivation scheme combines the chemical passivation (by intrinsic a‐Si:H) and the field passivation (by doped a‐Si:H). As an example, **Figure 42** shows the field passivation by p‐a‐ Si:H in (a) and the combination of both passivation mechanisms in (b). Addition of a p‐doped layer introduces an additional electrical filed to hinder the minority diffusion to the surface. The barrier ФBSF exponentially reduces the concentration of minority at the interface and thus the surface recombination. In **Figure 42(b)**, the interface defect density is reduced by the chemical passivation of addition of an intrinsic layer. Then, the solar cell precursor with the structure of n/i/P/i/p was fabricated to estimate the lifetime values. In these structures, the

n‐type 4 Ωcm, p‐type 2 Ωcm, and p‐type 1 Ωcm c‐Si substrates, respectively.

222 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

, the best lifetime values are 1.3 ms, 620 µs, and 290 µs for the

excess carrier density of 1015/cm3

LFICP with an antenna‐substrate distance of 53 cm.

precursor (without the metallization procedure).

**3.4. c‐Si passivation by stack configuration and implied** *Voc*

**Figure 42.** The filed passivation created by p‐doped a‐Si:H (a), and the combination of chemical surface passivation and field passivation in a stack of an intrinsic and p‐doped layer (b).

**Figure 43.** The injection level‐dependent lifetime values in the p‐type 1 Ωcm c‐Si passivated by symmetric and asym‐ metric structures consisting of intrinsic and doped a‐Si:H.

The relation between *Voc* and effective lifetime τeff can be expressed as [81]:

$$W\_{oc} = \frac{kT}{q} \ln \frac{2J\_{ph}\tau\_{eff}(N\_A + \Delta n)}{qn\_i^2W} \tag{7}$$

where *Jph* is the photocurrent density, *NA* the background dopant density, and *ni* the intrinsic carrier concentration in silicon. As expected, a high lifetime value corresponds to a high *Voc*. In our p‐type 1 Ωcm Si (thickness of 300 µm), the implied *Voc* is about 670 and 620 mV corre‐ sponding to the solar cell precursor (n/i/P/i/p) lifetime of 150 and 64 µs, respectively.

### **4. Conclusions**

As a low‐temperature plasma source, LFICP has the inherent advantages of high‐density plasma, low sheath potential, and low electron temperature, etc., which deserves an excellent CVD route. In LFICP CVD, the parameters of the RF power, hydrogen dilution, working pressure, and substrate temperature significantly influence the microstructures and properties of the films. The ion bombardment on the growing surface in LFICP CVD can be effectively controlled by the parameter of antenna‐substrate distance *d*, providing an additional way to tune the microstructures and properties of Si‐based materials.

The passivation effects of LFICP CVD Si‐based materials are strongly dependent on the plasma configuration. In a region not far away the plasma generation area (semi‐remote LFICP), the passivated c‐Si shows a high as‐deposited carrier lifetime due to the hydrogen‐related chemical annealing during the deposition process. However, the additional thermal annealing does not lead to a considerable increase in the lifetime value. The addition of CO<sup>2</sup> into the reactive gases substantially influences the microstructure of the obtained thin films and improves the interface quality. In the region far away from plasma generation region (with a high antenna‐ substrate distance, remote‐LFICP), a‐Si:H passivated c‐Si shows a low as‐deposited and a high post‐deposition annealing lifetime value. The highest lifetime value of 1.3 ms after thermal annealing is obtained in passivated 4 Ωcm c‐Si substrate. After being passivated by the device‐ like scheme of undoped/doped silicon film stack, p‐type substrate (1 Ωcm, 300 µm) shows an implied open voltage of 670 mV.

### **Acknowledgements**

This work was jointly supported by AcRF Tier1 (No. RP 6/13 XS), A\*STAR, Singapore, and the National Science Foundation of China (Grant No. 51302028 and 61404061).

### **Author details**

(7)

the intrinsic

The relation between *Voc* and effective lifetime τeff can be expressed as [81]:

224 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

*oc*

*V*

tune the microstructures and properties of Si‐based materials.

**4. Conclusions**

implied open voltage of 670 mV.

**Acknowledgements**

2 2( ) = ln *ph eff A*

*kT J Nn*

*q qn W* t+ D

sponding to the solar cell precursor (n/i/P/i/p) lifetime of 150 and 64 µs, respectively.

where *Jph* is the photocurrent density, *NA* the background dopant density, and *ni*

*i*

carrier concentration in silicon. As expected, a high lifetime value corresponds to a high *Voc*. In our p‐type 1 Ωcm Si (thickness of 300 µm), the implied *Voc* is about 670 and 620 mV corre‐

As a low‐temperature plasma source, LFICP has the inherent advantages of high‐density plasma, low sheath potential, and low electron temperature, etc., which deserves an excellent CVD route. In LFICP CVD, the parameters of the RF power, hydrogen dilution, working pressure, and substrate temperature significantly influence the microstructures and properties of the films. The ion bombardment on the growing surface in LFICP CVD can be effectively controlled by the parameter of antenna‐substrate distance *d*, providing an additional way to

The passivation effects of LFICP CVD Si‐based materials are strongly dependent on the plasma configuration. In a region not far away the plasma generation area (semi‐remote LFICP), the passivated c‐Si shows a high as‐deposited carrier lifetime due to the hydrogen‐related chemical annealing during the deposition process. However, the additional thermal annealing does not lead to a considerable increase in the lifetime value. The addition of CO<sup>2</sup> into the reactive gases substantially influences the microstructure of the obtained thin films and improves the interface quality. In the region far away from plasma generation region (with a high antenna‐ substrate distance, remote‐LFICP), a‐Si:H passivated c‐Si shows a low as‐deposited and a high post‐deposition annealing lifetime value. The highest lifetime value of 1.3 ms after thermal annealing is obtained in passivated 4 Ωcm c‐Si substrate. After being passivated by the device‐ like scheme of undoped/doped silicon film stack, p‐type substrate (1 Ωcm, 300 µm) shows an

This work was jointly supported by AcRF Tier1 (No. RP 6/13 XS), A\*STAR, Singapore, and the

National Science Foundation of China (Grant No. 51302028 and 61404061).

H. P. Zhou1,2\*, S. Xu2\* and S. Q. Xiao3

\*Address all correspondence to: haipzhou@uestc.edu.cn; shuyan.xu@nie.edu.sg

1 School of Energy Science and Engineering, University of Electronic Science and Technology of China, Chengdu, Sichuan, China

2 Plasma Sources and Application Center, NIE, and Institute of Advanced Studies, Nanyang Technological University, Singapore

3 Key Laboratory of Advanced Process Control for Light Industry (Ministry of Education), Department of Electronic Engineering, Jiangnan University, Wuxi, Jiangsu, China

### **References**


[19] Zhou HP, Wei DY, Xu S, Xiao SQ, Xu LX, Huang SY, Guo YN, Khan S, Xu M. Crystalline silicon surface passivation by intrinsic silicon thin films deposited by low‐frequency inductively coupled plasma. Journal of Applied Physics 2012; 112:013708. DOI: http:// dx.doi.org/10.1063/1.4733701.

[7] Rech B, Repmann T, van den Donker MN, Berginski M, Kilper T, Hüpkes J, Calnan S, Stiebig H, Wieder S. Challenges in microcrystalline silicon based solar cell technology.

[8] Xu S, Levchenko I., Huang SY, Ostrikov K. Self‐organized vertically aligned single‐ crystal silicon nanostructures with controlled shape and aspect ratio by reactive plasma

[9] Denysenko I. B., Xu S., Long J. D., Rutkevych P. P., Azarenkov N. A. and Ostrikov K. Inductively coupled Ar/CH4/H2 plasma for low‐temperature deposition of ordered carbon nanostructures. J. Appl. Phys. 2004; 95: 2713‐2724. DOI: 10.1063/1.1642762 [10] Xiao SQ, Xu S, Ostrikov K. Low‐temperature plasma processing for Si photovoltaics. Materials Science and Engineering R. 2014; 78:1‐29. DOI: 10.1016/j.mser.2014.01.002

[11] Xu S, Ostrikov K, Li Y, Tsakadze EL, Jones IR. Low‐frequency, high‐density, inductively coupled plasma sources: operation and applications. Phys. Plasmas. 2001; 8: 2549‐2557.

[12] Ostrikov K, Xu S, Xu MY. Power transfer and mode transitions in low‐frequency inductively coupled plasmas. Journal of Applied Physics. 2000; 88:2268‐2271. DOI:

[13] Cheng QJ, Xu S, Long JD, Ni ZH, Rider AE, Ostrikov K. High‐rate, low‐temperature synthesis of composition controlled hydrogenated amorphous silicon carbide films in low‐frequency inductively coupled plasmas. Journal of Physics D: Applied Physics.

[14] Cheng QJ, Xu S, Long JD, Huang SY, Guo J. Homogeneous nanocrystalline cubic silicon carbide films prepared by inductively coupled plasma chemical vapor deposition.

[15] Cheng QJ, Xu S, Ostrikov K. Rapid, low‐temperature synthesis of nc‐Si in high‐density, non‐equilibrium plasmas: enabling nanocrystallinity at low hydrogen dilution. Journal

[16] Zhou HP, Xu LX, Xu S, Huang SY, Wei DY, Xiao SQ, Yan WS, Xu M. On the conductivity type conversion of the p‐type silicon exposed to low frequency inductively coupled plasma of Ar+H2. Journal of Physics D: Applied Physics. 2010; 43:505402. DOI:

[17] Xu S, Huang SY, Levchenko I, Zhou HP, Wei DY, Xiao SQ, Xu LX, Yan WS, Ostrikov K. Highly efficient silicon nanoarray solar cells by a single‐step plasma‐based process.

[18] Xiao SQ, Xu S, Zhou HP, Wei DY, Huang SY, Xu LX, Sern CC, Guo YN, Khan S. Amorphous/crystalline silicon heterojunction solar cells via remote inductively coupled plasma processing. Applied Physics Letters. 2012; 100:233902. DOI: http://

Advanced Energy Materials, 2011; 1:373‐376. DOI: 10.1002/aenm.201100085

Nanotechnology. 2007; 18:465601. DOI: 10.1088/0957‐4484/18/46/465601

of Material Chemistry. 2009; 19:5134‐5140. DOI: 10.1039/b904227

2008; 41:055406. DOI:10.1088/0022‐3727/41/5/055406

etching. Applied Physics Letters, 2009; 95:111505. DOI: 10.1063/1.3232210

Thin Solid Films, 2006; 511:548‐555. DOI: 10.1016/j.tsf.2005.12.161

226 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

DOI: 10.1063/1.1343887

10.1088/0022‐3727/43/50/505402.

dx.doi.org/10.1063/1.4721642

10.1063/1.1289055


deposition and its application for surface passivation of p‐type crystalline silicon. Japan Journal of Applied Physics. 2009; 48:066509‐1‐066509‐4. DOI: 10.1143/JJAP.48.066509


[41] Kaxiras E, Joannopoulos JD. Hydrogenation of semiconductor surfaces: Si and Ge (111). Physical Review B.1988; 37:8842‐8848. DOI: http://dx.doi.org/10.1103/PhysRevB. 37.8842

deposition and its application for surface passivation of p‐type crystalline silicon. Japan Journal of Applied Physics. 2009; 48:066509‐1‐066509‐4. DOI: 10.1143/JJAP.48.066509

[31] Mott NF, Davis EA. Electronic Processes in Non‐Crystalline Solids. second ed.,

228 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

[32] Matsuda A, Yamaoka T, Wolff S, KoyamaM, Imanishi Y, Kataoka H, Matsuura H, Tanaka K. Preparation of highly photosensitive hydrogenated amorphous Si‐C alloys from a glow‐discharge plasma. Journal of Applied Physics. 1986; 60:4205‐4207. DOI:

[33] Zhou HP, Wei DY, Xu S, Xiao SQ, Xu LX, Huang SY, Guo YN, Yan WS, Xu M. Dilution effect of Ar/H2 on the microstructures and photovoltaic properties of nc‐Si:H deposited in low frequency inductively coupled plasma. Journal of Applied Physics. 2011;

[34] Saha JK, Bahardoust B, Leong K, Gougam AB, Kherani N, Zukotynski S. Spectroscopic ellipsometry studies on hydrogenated amorphous silicon thin films deposited using DC saddle field plasma enhanced chemical vapor deposition system. Thin Solid Films.

[35] Cavalcoli D, Rossi M, Cavallini A. Defect states in nc‐Si:H films investigated by surface photovoltage spectroscopy. Journal of Applied Physics. 2011; 109:053719‐1‐053719‐5.

[36] Bagolini L, Mattoni A, Colombo L. Electronic localization and optical absorption in embedded silicon nanograins. Applied Physics Letters. 2009; 94:053115‐1‐053115‐3.

[37] Smit C, van Swaaij RACM, Donker H, Petit AMHN, Kessels WMM, van de Sanden MCM. Determining the material structure of microcrystalline silicon from Raman spectra. Journal of Applied Physics. 2003; 94:3582‐3588. DOI: 10.1063/1.1596364 [38] Myong SY, Lim KS, Konagai M. Effect of hydrogen dilution on carrier transport in hydrogenated boron‐doped nanocrystalline silicon‐silicon carbide alloys. Applied

[39] Cheng QJ, Xu S, Ostrikov K. Temperature‐dependent properties of nc‐si thin films synthesized in low‐pressure, thermally nonequilibrium, high‐density inductively coupled plasmas. Journal of Physics and Chemistry C. 2009; 113:14759‐14764. DOI:

[40] Solomon I, Drévillon B, Shirai H, Layadi N. Plasma deposition of microcrystalline silicon: the selective etching model. 15th International Conference on Amorphous Semiconductors: Science and Technology; 6‐10 September 1993; Cambridge: Journal of

Clarendon, Oxford; 1979. 590 p.

110:023517. DOI: 10.1063/1.3605288

DOI: 10.1063/1.3553583

DOI: 10.1063/1.3078281

10.1021/jp9047083

2011; 519:2863‐2866. DOI: 10.1016/j.tsf.2010.12.074

Physics Letters. 2006; 88:103120. DOI: 10.1063/1.2177641

Non‐Crystalline Solids. 1993; 164–166: 989‐992.

10.1063/1.337528


[67] Rath J K. Low temperature polycrystalline silicon: a review on deposition, physical properties and solar cell applications. Solar Energy Materials Solar Cells. 2003; 76:437‐ 487. DOI: 10.1016/S0927‐0248(02)00258‐1

[55] Martín I, Vetter M, Orpella A, Voz C, Puigdollers J, Alcubilla R. Characterization of a‐ SiCx:H films for c‐Si surface passivation. Applied Physics Letters. 2002; 81:476‐479.

230 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

[56] Fujiwara H, Kondo M. Effects of a‐Si:H layer thicknesses on the performance of a‐Si:H/ c‐Si heterojunction solar cells. Journal of Applied Physics. 2007; 101:054516. DOI:

[57] Wang Q, Page MR, Iwaniczko E, Xu Y, Roybal L, Baure R, To B, Yuan HC, Duda A, Fasoon F, Yan YF., Levi D, Meier D, Branz HM, Wang TH. Efficient heterojunction solar cells on p‐type crystal silicon wafers. Applied Physics Letters. 2010; 96:013507. DOI:

[58] Lee JY, Glunz SW. Investigation of various surface passivation schemes for silicon solar cells. Solar Energy Matters & Solar Cells. 2006; 90:82‐92. DOI: 10.1016/j.solmat.

[59] Schmidt J, Kerr M, Cuevas A. Surface passivation of Si solar cells using plasma‐ enhanced chemical vapor deposited SiN films and thermal SiO2/thermal SiN Stacks. Semiconductor Science Technology. 2001; 19:164‐170. DOI: 10.1088/0268‐1242/16/3/308

[60] Martin I, Vetter M, Orpella A, Puidollers J, Cuevas A, Alcublla R. Surface passivation of p‐type crystalline Si by plasma enhanced chemical vapor deposited amorphous SiCx:H films. Applied Physics Letters. 2001; 79:2199‐2201. DOI: 10.1063/1.1404406

[61] Dauwe S, Schmid J, Hezel R. Very low surface recombination velocities on p‐ and n‐ type silicon wafers passivated with hydrogenated amorphous silicon films. Proceed‐ ings of the Photovoltaic Specialists Conference; 19–24 May 2002; New York: IEEE; 2002.

[62] Willem J, Schuttauf A, Van der Werf KHM, Kielen IM, Van Sark WGJHM, Rath JK, Schropp REI. High quality crystalline silicon surface passivation by combined intrinsic and n‐type hydrogenated amorphous silicon. Applied Physics Letters. 2011; 98:153514.

[63] Vetterl O, Finger F, Carius R, Hapke P, Houben L, Kluth O, Lambertz A, Mück A, Rech B, Wagner H. Intrinsic microcrystalline silicon: a new material for photovoltaics. Solar Energy Materials Solar Cells. 2000; 62:97‐108. DOI: 10.1016/S0927‐0248(99)00140‐3

[64] Rem JB, Holleman J, Verweij JF. Incubation time measurements in thin-film deposition. Journal of Electrochemical Society. 1997; 144:2101‐2106. DOI: 10.1149/1.1837748

[65] Tickle AC. Thin‐Film Transistors: A New Approach to Microelectronic. John Wiley &

[66] Mitchell J, Macdonald D, Cuevas A. Thermal activation energy for the passivation of the n‐type crystalline silicon surface by hydrogenated amorphous silicon. Applied

Physics Letters. 2009; 94:162102. DOI: 10.1063/1.3120765

DOI: http://dx.doi.org/10.1557/PROC‐715‐A24.5

10.1063/1.2559975

10.1063/1.3284650

2005.02.007

p. 1246‐1249.

DOI: 10.1063/1.3662404

Sons, New York; 1969. 144 p.


Journal of Applied Physics D: Applied Physics. 2008; 41:185107. DOI:10.1088/0022‐ 3727/41/18/185107


### **Applications of CVD to Produce Thin Films for Solid‐State Devices**

A.M. Torres‐Huerta, M.A. Domínguez‐Crespo and A.B. López‐Oyama

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/63964

### **Abstract**

Journal of Applied Physics D: Applied Physics. 2008; 41:185107. DOI:10.1088/0022‐

[78] Chen MS, Santra AK, Goodman DW. Structure of thin SiO2 films grown on Mo(112). Physics Review B. 2004; 69:1124‐1133. DOI: http://dx.doi.org/10.1103/PhysRevB.

[79] Sze SM. Semiconductor Devices: Physics and Technology, John Wiley & Sons; 2002. 576

[80] Wolf SD, Kondo M. Nature of doped a‐Si:H/c‐Si interface recombination. Journal of

[81] Sinton RA, Cuevas A. Contactless determination of current–voltage characteristics and minority‐carrier lifetimes in semiconductors from quasi‐steady‐state photoconduc‐ tance data. Journal of Applied Physics. 1996; 69:2510‐2512. DOI: S0003‐6951(96)02743‐

Applied Physics. 2009; 105:103707‐103707‐6. DOI: 10.1063/1.3129578

232 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

3727/41/18/185107

69.155404

p.

X

Thin films of Pt‐YSZ and Pd‐ZrO2 cermets by chemical vapor deposition (CVD) from metallorganic precursors (MOCVD) were evaluated as electrode in solid‐state devices. Morphology and structural characteristics were studied by X‐ray diffraction (XRD), scanning electronic microscopy, atomic force microscopy (AFM), and transmission electronic microscopy (TEM). Electrochemical performance was determined using Tafel and electrochemical impedance spectroscopy methods. Metallorganic precursors were used (metal‐acetylacetonates), and argon and oxygen were used as the carrier and reactive gases, respectively. The particle average size was less than 20 nm, with high and uniform particle dispersion according to TEM measurements.

**Keywords:** thin films, CVD, platinum metals, zirconia, cermets

### **1. Introduction**

Chemical vapor deposition (CVD) is a fast and economic method to obtain controlled depos‐ its of high quality; the main drawbacks are the possible incorporation of impurities due to non‐ exhausted remains and poor adhesion of the film in some cases [1]. Chemical vapor deposition (CVD) from metallorganic precursors (MOCVD) is a rapidly developing method for produc‐ ing films and coatings of ceramic materials for a variety of applications [2]. Among solid electrolyte cells, commonly used for numerous applications including gas sensors, gas pumps, solid oxide fuel cells, and electrochemically promoted catalysts, platinum electrode deposited on yttria‐stabilized zirconia (YSZ) is one of the most widely studied examples [3] because of

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

YSZ conducts at temperatures above 300°C via O2- diffusion and platinum has always been of particular interest due to its electrocatalytic properties [4]. The electrochemical performance of electrode layers depends largely on the microstructure of the three‐phase boundaries (TPBs) [5] in which the ionic conductor, the electronic conductor, and the oxygen gas are in contact, being the best pathway to reduce oxygen to oxygen ion at the surface [6].

Nanoparticles exhibit very interesting features at the nanometer scale due to the quantum confinement, where the changes induced to the microstructure allow them to increase the specific surface area and, in the case of electrodes, more TPBs are arisen. These characteristics are desirable for high electrochemical performance. The higher sintering temperatures are related to the formation of undesirable reaction products highly resistive at the electro/ electrolyte interface. By reaching the nanometer size, the sintering temperature can be low facilitating the ceramic processing, due to the more TPBs which as a geometrical parameter required for high electrochemical performance [5].

The appearance of the material at nanometric scale can be achieved by modification of the preparation methods despite its identical chemical composition. The material undergoes a phase transformation by changing temperature and pressure or by combining different materials. The induced changes onto the microstructure modify the electrode behavior [4]. One approach to increase the TPBs areas, and in consequence the yield of the electrodes, is to prepare a cermet material that spreads the solid electrolyte into the electrode [5, 7–10]. The development in thin‐film technology and the application of ionic‐electronic composites as electrodes have significantly reduced the ohmic loss caused by the electrolyte [5]. The rate and selectivity of catalytic reactions occurring on metal and metal‐oxide porous catalyst‐electrode films deposited on solid electrolytes can be reversibly affected in a very pronounced and controlled manner by polarizing of the catalyst‐electrode, which can exceed the corresponding rate of ion transport through the solid electrolyte by several orders of magnitude [11]. Noble metals are historically known to be extremely non‐reactive, but the possibility for obtaining coatings of platinum metals by means of CVD on the substrates made of different materials was demonstrated widely [12]. Metal, oxide, and metal‐oxide composite systems can be obtained by means of CVD using the process of thermal decomposition of a mixture of the volatile initial metal complexes with organic ligands (precursors) on a heated surface [12]. Additionally, chemical vapor deposition (CVD) is one of the most important synthesis processes for making nanoparticles of metals, semiconductors, and magnetic materials. In some cases, Pt is deposited as continuous films whereas in other cases, it is preferred in the form of a dispersion of particles with a high surface area. On the other hand, ceramics based on zirconium dioxide (zirconia, ZrO2) are promising materials for multifunctional applica‐ tions. In previous works [13–16], ZrO2, IrO2, Pt, and IrO2‐YSZ thin films have been synthesized by MOCVD process and, in the case of ZrO2, a thermodynamic study to explain the phase and the conditions that promote the formation of pure ZrO2 from a β‐diketonate precursor (Zr‐ acetylacetonate, Zr(acac)4) and oxygen (O2) as carrier/reactive gas was reported [17], and now, the structural, nanocrystalline nature and morphological characterizations of Pt‐YSZ compo‐ sites are completed and reported; furthermore, in order to determine the performance of these composites as electrodes in YSZ solid‐state devices, electrochemical tests were also carried out. Additionally, the synthesis of Pd‐ZrO2 cermets by MOCVD and their characterization are reported. X‐ray diffraction (XRD), atomic force microscopy (AFM), scanning electron micro‐ scopy (SEM), Raman spectroscopy, and electrochemical characterizations were conducted in order to determine the structure, composition, morphology, oxidation state, and electrocata‐ lytic properties of these materials. Metallorganic precursors were used (metal‐acetylaceto‐ nates), and argon and oxygen were used as the carrier and reactive gases, respectively.

### **2. Experimental procedure**

YSZ conducts at temperatures above 300°C via O2- diffusion and platinum has always been of particular interest due to its electrocatalytic properties [4]. The electrochemical performance of electrode layers depends largely on the microstructure of the three‐phase boundaries (TPBs) [5] in which the ionic conductor, the electronic conductor, and the oxygen gas are in contact, being

234 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Nanoparticles exhibit very interesting features at the nanometer scale due to the quantum confinement, where the changes induced to the microstructure allow them to increase the specific surface area and, in the case of electrodes, more TPBs are arisen. These characteristics are desirable for high electrochemical performance. The higher sintering temperatures are related to the formation of undesirable reaction products highly resistive at the electro/ electrolyte interface. By reaching the nanometer size, the sintering temperature can be low facilitating the ceramic processing, due to the more TPBs which as a geometrical parameter

The appearance of the material at nanometric scale can be achieved by modification of the preparation methods despite its identical chemical composition. The material undergoes a phase transformation by changing temperature and pressure or by combining different materials. The induced changes onto the microstructure modify the electrode behavior [4]. One approach to increase the TPBs areas, and in consequence the yield of the electrodes, is to prepare a cermet material that spreads the solid electrolyte into the electrode [5, 7–10]. The development in thin‐film technology and the application of ionic‐electronic composites as electrodes have significantly reduced the ohmic loss caused by the electrolyte [5]. The rate and selectivity of catalytic reactions occurring on metal and metal‐oxide porous catalyst‐electrode films deposited on solid electrolytes can be reversibly affected in a very pronounced and controlled manner by polarizing of the catalyst‐electrode, which can exceed the corresponding rate of ion transport through the solid electrolyte by several orders of magnitude [11]. Noble metals are historically known to be extremely non‐reactive, but the possibility for obtaining coatings of platinum metals by means of CVD on the substrates made of different materials was demonstrated widely [12]. Metal, oxide, and metal‐oxide composite systems can be obtained by means of CVD using the process of thermal decomposition of a mixture of the volatile initial metal complexes with organic ligands (precursors) on a heated surface [12]. Additionally, chemical vapor deposition (CVD) is one of the most important synthesis processes for making nanoparticles of metals, semiconductors, and magnetic materials. In some cases, Pt is deposited as continuous films whereas in other cases, it is preferred in the form of a dispersion of particles with a high surface area. On the other hand, ceramics based on zirconium dioxide (zirconia, ZrO2) are promising materials for multifunctional applica‐ tions. In previous works [13–16], ZrO2, IrO2, Pt, and IrO2‐YSZ thin films have been synthesized by MOCVD process and, in the case of ZrO2, a thermodynamic study to explain the phase and the conditions that promote the formation of pure ZrO2 from a β‐diketonate precursor (Zr‐ acetylacetonate, Zr(acac)4) and oxygen (O2) as carrier/reactive gas was reported [17], and now, the structural, nanocrystalline nature and morphological characterizations of Pt‐YSZ compo‐ sites are completed and reported; furthermore, in order to determine the performance of these composites as electrodes in YSZ solid‐state devices, electrochemical tests were also carried out.

the best pathway to reduce oxygen to oxygen ion at the surface [6].

required for high electrochemical performance [5].

The experimental setup has been described elsewhere [15]. Amorphous quartz and YSZ were used as substrates; the composite electrodes were produced by a horizontal hot wall MOCVD. The employed precursors were Zr(acac)4, Y(acac)3, Pt(acac)2, and Pd(acac)<sup>2</sup> [Sigma‐Aldrich, ≥97%]. Argon and oxygen gases were used as the carrier and reactive gases, respectively. Before the MOCVD process, the quartz substrates were cleaned with a soapy solution and thereafter rinsed with tap water, deionized water, and acetone. The optimal experimental conditions for producing nanocomposite electrodes were identified as follows: 450–600°C deposition temperature (*T*dep), 1.0 Torr total pressure (*P*tot), 190–220°C precursor temperature (*T*prec), 130  sccm O2, and 180 sccm Ar flow rates (FR). The structural analysis and the particle size deter‐ mination were carried out by X‐ray diffraction (XRD, Siemens D5000 and Bruker D8 Ad‐ vanced) using a monochromatic Cu‐Kα1 radiation at 35 kV, 25 mA and a scan rate of 0.02° min-1; the average crystallite size was calculated from the diffraction peak broadening using the Scherrer equation; and the morphology, texture, and surface roughness of the films were examined by scanning electron microscopy (SEM, JEOL JSM‐6300), transmission electron microscopy (TEM, JEOL‐2000 FX‐II), and atomic force microscopy by AFM, Nanosurf easyscan 2 AFM/STM in contact mode with a Si cantilever and nominal force of 20 nN, respectively.

Finally, the electrochemical tests were conducted to determine the performance of these films as electrodes. The method for this purpose was a three terminal one: the working electrode (WE, 1.7 cm<sup>2</sup> in area) on one side of the YSZ pellet and the reference and the counter electrodes on the opposite side (RE, 0.5 cm<sup>2</sup> in area, CE, 1.2 cm<sup>2</sup> in area, respectively). Impedance meas‐ urements were carried out over the frequency range of 10<sup>5</sup> –10-2 Hz using a Solartron 1255 frequency response analyzer from 300 to 800°C in air. The signal applied to the cell was generally 10 mV rms. Potentiodynamic polarization was obtained from -400 mV to +400 mV at a scan rate of 1.67 mV s-1.

### **3. Results and discussion**

**Figure 1a** and **b** shows XRD spectra of platinum thin films obtained at 500°C and 550°C, and quartz is shown as reference (**Figure 1c**). Pt crystallites in face‐centered cubic structure (ICDD 04‐0802) grow in (1 1 1) orientation at 500°C while at 550°C, is polycrystalline, and at both temperature, the films are purely metallic and shiny. The broadening of the reflections was associated with a small crystallite size, determined by Scherrer equation:

$$t = \frac{0.9\lambda}{B\cos\theta\_g} \tag{1}$$

**Figure 1.** Pt thin‐film diffractograms (a) 500°C, (b) 550°C and (c) amorphous quartz.

Here, *t* is the mean crystalline size, *λ* is the wavelength of the X‐ray, *B* is the full width of a peak at half of the maximum of the peak (FWHM) in the diffraction spectra (measured in radians), and *θ* is the Bragg angle.

Then, the mean Pt crystallite size was 48 nm. Pt‐ZrO2 composite was obtained at 600°C (**Figure 2a**); ZrO<sup>2</sup> signals are weak, while Pt reflections are better defined than that for ZrO2. It was reported before that ZrO<sup>2</sup> at these particular experimental conditions was not obtained below 600°C [16]; the Pt‐ZrO<sup>2</sup> composites had to be synthesized at this temperature, taking in account the experimental conditions for ZrO2, that is, only oxygen was used as the carrier and reactive gas. The next step was the deposition of yttria‐stabilized zirconia (YSZ), shown in **Figure 3b**. These films were transparent. The XRD spectrum displays the signals correspond‐ ing to (1 1 1), (2 0 0), and (3 1 1) planes for cubic YSZ (ICDD 30‐1468); the Zr/Y weight ratio to obtain YSZ was 0.87:0.13, according to Wang et al. [18]; the mean crystallite size was 15 nm. The appearance of YSZ cubic microstructure could be the result of these nanocrystalline sizes [19, 20]. The formation of cubic structure of YSZ film at low temperatures is somewhat alike with the formation of metastable tetragonal structure as a result of nanocrystalline size effect in the transformation of zirconium phase [18, 19, 21]. Finally, once the individual thin films were synthesized, the composite Pt‐YSZ was deposited by combining the different experi‐ mental conditions (**Figure 3c**). The signals of both Pt and YSZ are clearly observed. In this case, the mean Pt crystallite size was 21 nm, while the YSZ size remains the same, 15 nm. The decrement in Pt crystallite size could be attributable to the film microstructure, which is avoiding the agglomeration, sintering, and, consequently, the growth of Pt particles. The XRD spectrum for amorphous quartz is shown in **Figures 1c**, **2b**, and **3a** in order to identify the signals arising from thin films. Experimental deposition conditions are summarized in **Table 1**.

**Figure 2.** Pt‐ZrO2 diffractograms (a) 600°C and (b) amorphous quartz.

<sup>=</sup> <sup>q</sup>*<sup>B</sup> <sup>λ</sup> <sup>t</sup> B* 0.9

236 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 1.** Pt thin‐film diffractograms (a) 500°C, (b) 550°C and (c) amorphous quartz.

radians), and *θ* is the Bragg angle.

Here, *t* is the mean crystalline size, *λ* is the wavelength of the X‐ray, *B* is the full width of a peak at half of the maximum of the peak (FWHM) in the diffraction spectra (measured in

Then, the mean Pt crystallite size was 48 nm. Pt‐ZrO2 composite was obtained at 600°C (**Figure 2a**); ZrO<sup>2</sup> signals are weak, while Pt reflections are better defined than that for ZrO2. It was reported before that ZrO<sup>2</sup> at these particular experimental conditions was not obtained below 600°C [16]; the Pt‐ZrO<sup>2</sup> composites had to be synthesized at this temperature, taking in account the experimental conditions for ZrO2, that is, only oxygen was used as the carrier and reactive gas. The next step was the deposition of yttria‐stabilized zirconia (YSZ), shown in **Figure 3b**. These films were transparent. The XRD spectrum displays the signals correspond‐ ing to (1 1 1), (2 0 0), and (3 1 1) planes for cubic YSZ (ICDD 30‐1468); the Zr/Y weight ratio to obtain YSZ was 0.87:0.13, according to Wang et al. [18]; the mean crystallite size was 15 nm. The appearance of YSZ cubic microstructure could be the result of these nanocrystalline sizes [19, 20]. The formation of cubic structure of YSZ film at low temperatures is somewhat alike with the formation of metastable tetragonal structure as a result of nanocrystalline size effect in the transformation of zirconium phase [18, 19, 21]. Finally, once the individual thin films

cos (1)

**Figure 3.** Diffractograms of (a) amorphous quartz, (b) YSZ, and (c) Pt‐YSZ at 600°C.


**Table 1.** Experimental deposition conditions for Pt‐YSZ film.

A typical view of the composite Pt‐YSZ film is given in **Figure 4**. The film is porous as a result of its columnar growth (**Figure 4d**, cross‐sectional view). Thickness of the film was estimated to be about 1 µm with a growth rate of 50 nm/min.

**Figure 4.** SEM images of Pt‐YSZ thin films (a–c) surface and (d) cross‐section.

**Figure 5** depicts the impedance spectrum at 300–800°C for Pt‐YSZ nanocomposite electrode films prepared on YSZ pellets. According to the associated capacitance value (10-5 F), reactions of charge transfer at the electrode/electrolyte interface were assigned to the small semicircle at high frequency. It seems that a diffusion mechanism rise at the low‐frequency region because of the large semicircle with capacitance values between 10-4 and 10-3 F, attributed to the oxygen diffusion at the finite thickness in the film. The bulk and grain boundary responses are missing due to the frequency scale. According to West et al. [22], *C* = ∼10-7 F corresponds to electrode‐ electrolyte interphase and *C* = ∼10-4 F is associated with the oxygen electrochemical reaction into the TPB; furthermore, a depressed semicircle could indicate an associated impedance to the oxygen path through the electrode [22]. The Pt‐YSZ nanocomposite electrodes displayed lower charge‐transfer resistance than diffusion resistance even at higher temperatures. This suggests easy interfacial charge‐transfer reactions probably controlled by the diffusion of oxygen through the structure of the Pt‐YSZ electrode.

**Thin film** *T***prec (°C)** *T***dep (°C) Gas flow (cm3**

A typical view of the composite Pt‐YSZ film is given in **Figure 4**. The film is porous as a result of its columnar growth (**Figure 4d**, cross‐sectional view). Thickness of the film was estimated

**Figure 5** depicts the impedance spectrum at 300–800°C for Pt‐YSZ nanocomposite electrode films prepared on YSZ pellets. According to the associated capacitance value (10-5 F), reactions of charge transfer at the electrode/electrolyte interface were assigned to the small semicircle at high frequency. It seems that a diffusion mechanism rise at the low‐frequency region because of the large semicircle with capacitance values between 10-4 and 10-3 F, attributed to the oxygen diffusion at the finite thickness in the film. The bulk and grain boundary responses are missing due to the frequency scale. According to West et al. [22], *C* = ∼10-7 F corresponds to electrode‐ electrolyte interphase and *C* = ∼10-4 F is associated with the oxygen electrochemical reaction into the TPB; furthermore, a depressed semicircle could indicate an associated impedance to the oxygen path through the electrode [22]. The Pt‐YSZ nanocomposite electrodes displayed lower charge‐transfer resistance than diffusion resistance even at higher temperatures. This suggests easy interfacial charge‐transfer reactions probably controlled by the diffusion of

Pt 180 500 180 ZrO2 220 600 130 Pt‐ZrO<sup>2</sup> 220 550 130 Y2O3‐ZrO<sup>2</sup> 220 600 130 Pt‐YSZ 220 600 130

**Table 1.** Experimental deposition conditions for Pt‐YSZ film.

to be about 1 µm with a growth rate of 50 nm/min.

**Figure 4.** SEM images of Pt‐YSZ thin films (a–c) surface and (d) cross‐section.

oxygen through the structure of the Pt‐YSZ electrode.

**/min)**

**Figure 5.** Impedance spectrum at 300–800°C for Pt‐YSZ nanocomposite electrode films prepared on YSZ pellets.

From polarization plots (Tafel plots) can be observed that as temperature increases, the current increases through the electrode‐electrolyte interphase (**Figure 6**). In the anodic branch where higher current was obtained than that obtained in cathodic branch, except at 600 and 700°C, at 600°C, the current in the cathodic branch is slightly higher than in anodic branch; and at 700°C, both values of currents (anodic and cathodic branches) seem to be the same; and the cathodic response looks similar at these temperatures (600 and 700°C). At 800°C, the cathodic response is lower than that observed at 600 and 700°C, but the current increases as temperature increases in the anodic branch. This behavior could be attributed to a decrement of TPB provoked by particle agglomeration, maybe Pt or YSZ, or by a reaction between Pt and YSZ as Badwal and de Bruin suggested [23]. **Table 2** shows the interchange current and Tafel slopes from **Figure 6**. According to the Tafel slopes, one order of reaction was obtained. Besides YSZ, no oxidized phases were found in Pt‐YSZ electrodes after heating in air, which indicates a high thermal stability (**Figure 7**).

**Figure 6.** Tafel plots of Pt‐YSZ at 300–800°C.

**Figure 5.** Impedance spectrum at 300–800°C for Pt‐YSZ nanocomposite electrode films prepared on YSZ pellets.

240 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

From polarization plots (Tafel plots) can be observed that as temperature increases, the current increases through the electrode‐electrolyte interphase (**Figure 6**). In the anodic branch where higher current was obtained than that obtained in cathodic branch, except at 600 and 700°C, at 600°C, the current in the cathodic branch is slightly higher than in anodic branch; and at


**Table 2.** Interchange current (*I*o) and Tafel slopes (*m*), from Pt‐YSZ.

**Figure 7.** XRD pattern of Pt‐YSZ/YSZ after impedance and Tafel tests.

The synthesis of cermets Pd‐ZrO2 by MOCVD and their characterization were conducted in order to determine the structure, composition, morphology, oxidation state, and electrocata‐ lytic properties of these materials. Metallorganic precursors were used (metal‐acetylaceto‐ nates), and argon and oxygen were used as the carrier and reactive gases, respectively. It is expected that these materials could be applied in water purification. The composite thin films were prepared using a ratio of Pd and Zr metals of 10–90 wt%, respectively. The deposition temperature (*T*dep) was varied from 450, 500, and 550°C, and pressure (*P*) was controlled at 1.0 Torr. According to XRD patterns (**Figure 8**), palladium metallic reflections are difficult to observe and only in the films at 500°C, these signals are detected. Pd‐ZrO2 coatings on stainless

**Figure 8.** XRD patterns of Pd‐ZrO2 on glass substrate deposited at (a) 450°C, (b) 500°C, and (c) 550°C for 30 min and 80–20% volume Ar‐O2 and (d) of glass.

steel were deposited to characterizing their electrochemical behavior. It was reported that an increasing in Pd concentration favors the zirconia tetragonal phase as well as a higher crystal size [24]. The Pd‐ZrO<sup>2</sup> coatings deposited on glass substrates were porous, uniform and thin (**Figure 9**). Apparently, the films exhibit good adherence, but an adherence test is needed to verify this appreciation; composites obtained at 450°C show higher porosity. **Figure 10** shows the AFM images for Pd‐ZrO2 at different temperatures; it can be suggested a columnar growth, characteristic for this type of technique and thin film. The average roughness is present in **Table 3**. Pd‐ZrO2 coatings on stainless steel, at 450 and 500°C, were uniform, and it is observed at the domes, which could indicate a columnar growth (**Figure 11**). The coatings display a shift in the corrosion potential (**Figure 12**) from 64 mV to 273 mV as consequence of deposition temperature.

**Figure 9.** Scanning electron micrographs of Pd‐ZrO<sup>2</sup> on glass substrate deposited at (a) 450°C, (b) 500°C, and (c) 550°C for 30 min and 80–20% volume Ar‐O2.

**Figure 10.** AFM micrographs of Pd‐ZrO2 on glass substrate deposited at (a) 450°C, (b) 500°C, and (c) 550°C for 30 min and 80–20% volume Ar‐O2.


**Table 3.** Average roughness for Pd‐ZrO2 thin films.

**Figure 7.** XRD pattern of Pt‐YSZ/YSZ after impedance and Tafel tests.

80–20% volume Ar‐O2 and (d) of glass.

The synthesis of cermets Pd‐ZrO2 by MOCVD and their characterization were conducted in order to determine the structure, composition, morphology, oxidation state, and electrocata‐ lytic properties of these materials. Metallorganic precursors were used (metal‐acetylaceto‐ nates), and argon and oxygen were used as the carrier and reactive gases, respectively. It is expected that these materials could be applied in water purification. The composite thin films were prepared using a ratio of Pd and Zr metals of 10–90 wt%, respectively. The deposition temperature (*T*dep) was varied from 450, 500, and 550°C, and pressure (*P*) was controlled at 1.0 Torr. According to XRD patterns (**Figure 8**), palladium metallic reflections are difficult to observe and only in the films at 500°C, these signals are detected. Pd‐ZrO2 coatings on stainless

242 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 8.** XRD patterns of Pd‐ZrO2 on glass substrate deposited at (a) 450°C, (b) 500°C, and (c) 550°C for 30 min and

**Figure 11.** SEM micrographs of Pd‐ZrO2 on stainless steel at (a) 450°C and (b) 500°C.

**Figure 12.** Tafel plot for Pd‐ZrO2 coatings on stainless steel at (a) 450°C, (b) 500°C, and (c) 550°C.

### **4. Conclusions**

Metal, oxide, and metal‐oxide composite systems can be obtained by means of CVD using the process of thermal decomposition of a mixture of the volatile initial metal complexes with organic ligands (precursors) on a heated surface. The possibility for obtaining coatings of platinum metals by means of CVD on the substrates made of different materials was demon‐ strated widely. The platinum‐YSZ particle average size was less than 20 nm, with high and uniform particle dispersion according to TEM measurements.

### **Acknowledgements**

Authors are grateful for the financial support by CONACYT, SNI, and IPN in the projects SIP2016‐0541, and they thank J. Ferretiz‐Anguiano, J.R. Vargas‐García, and H. Dorantes for their technical support.

### **Author details**

**Figure 11.** SEM micrographs of Pd‐ZrO2 on stainless steel at (a) 450°C and (b) 500°C.

244 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 12.** Tafel plot for Pd‐ZrO2 coatings on stainless steel at (a) 450°C, (b) 500°C, and (c) 550°C.

A.M. Torres‐Huerta1\*, M.A. Domínguez‐Crespo1 and A.B. López‐Oyama1,2

\*Address all correspondence to: atohuer@hotmail.com and atorresh@ipn.mx

1 National Polytechnic Institute, CICATA‐Altamira, Altamira, Tamaulipas, Mexico

2 CONACYT, Cd. México, Mexico

### **References**


## **Plasma-Enhanced Chemical Vapor Deposition: Where we are and the Outlook for the Future**

Yasaman Hamedani, Prathyushakrishna Macha, Timothy J. Bunning, Rajesh R. Naik and Milana C. Vasudev

Additional information is available at the end of the chapter

http://dx.doi.org/10.5772/64654

### **Abstract**

[7] K. Sasaki, J.P. Wurth, R. Gschwen, K. Godickemeier, L.J. Gauckler. Journal of the

246 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

[10] M.J.L. Ostergard, C. Clausen, C. Bagger, M. Mogensen. Electrochimica Acta 40(12)

[11] N. Kotsionopoulos, S. Bebelis. Journal of Applied Electrochemistry 35 (2005) 1253–1264. [12] I.K. Igumenov, N.V. Gelfond, P.S. Galkin, N.B. Morozova, N.E. Fedotova, G.I. Zhar‐ kova, V.I. Shipachev, E.F. Reznikova, A.D. Ryabtsev, N.P. Kotsupalo, V.I. Titarenko,

[13] A.M. Torres‐Huerta, J.R. Vargas‐Garcia. Journal of Metastable Nanocrystalline Mate‐

[14] A.M. Torres‐Huerta, J.R. Vargas‐Garcia. Journal Metastable Nanocrystalline Materials

[15] A.M. Torres‐Huerta, J.R. Vargas‐Garcia, M.A. Dominguez‐Crespo. Solid State Ionics

[16] A.M. Torres‐Huerta, M.A. Domínguez‐Crespo, E. Ramírez‐Meneses, J.R. Vargas‐

[17] A.M. Torres‐Huerta, J.R. Vargas‐Garcia, M.A. Dominguez‐Crespo, J.A. Romero‐

[18] H.B. Wang, C.R. Xia, G.Y. Meng, D.K. Peng. Materials Letters 44 (2000) 23–28.

[22] A.R. West, D.C. Sinclair, N. Hirose. Journal of Electroceramics 1 (1997) 65.

[23] S.P.S. Badwal, H.J. de Bruin. Physica Status Solidi A 54 (1979) 261.

[19] Y. Jiang, J. Gao, M. Liu, Y. Wang, G. Meng. Solid State Ionics 177 (2007) 3405–3410.

[20] H.Z. Song, H.B. Wang, S.W. Zha, D.K. Peng, G.Y. Meng. Solid State Ionics 156 (2003)

[21] J.S. Lee, T. Matsubara, T. Sei, T. Tsuchiya. Journal of Materials Science 32 (1997) 5249.

[24] E.M. Holmgreen, M.M. Yung, U.S. Ozkan. Journal of Molecular Catalyst A: Chemical

García. Applied Surface Science 255 (2009) 4792–4795.

Serrano. Journal of Alloys and Compounds 483 (2009) 394–398.

Yu.P. Dikov, V.V. Distler, M.I. Buleev. Desalination 136 (2001) 273–280.

Electrochemical Society 143(2) (1996) 530.

(1995) 1971.

rials 20–21 (2004) 393–398.

24–25 (2005) 399–402.

178 (2007) 1608–1616.

249.

270 (2007) 101–111.

[8] T. Tsai, S.A. Barnett. Solid State Ionics 93 (1997) 207. [9] T. Kenjo, M. Nishiya. Solid State Ionics 57 (1992) 295.

> Chemical vapor deposition (CVD) is a technique for the fabrication of thin films of polymeric materials, which has successfully overcome some of the issues faced by wet chemical fabrication and other deposition methods. There are many hybrid techni‐ ques, which arise from CVD and are constantly evolving in order to modify the properties of the fabricated thin films. Amongst them, plasma enhanced chemical vapor deposition (PECVD) is a technique that can extend the applicability of the method for various precursors, reactive organic and inorganic materials as well as inert materials. Organic/inorganic monomers, which are used as precursors in the PECVD technique, undergo disintegration and radical polymerization while exposed to a high-energy plasma stream, followed by thin film deposition. In this chapter, we have provided a summary of the history, various characteristics as well as the main applications of PECVD. By demonstrating the advantages and disadvantages of PECVD, we have provided a comparison of this technique with other techniques. PECVD, like any other techniques, still suffers from some restrictions, such as selection of appropriate monomers, or suitable inlet instrument. However, the remarkable properties of this technique and variety of possible applications make it an area of interest for research‐ ers, and offers potential for many future developments.

**Keywords:** plasma, polymers, vapor deposition

© 2016 The Author(s). Licensee InTech. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

### **1. Introduction**

Chemical vapor deposition (CVD) is a multifaceted procedure which is currently used for several applications such as the fabrication of coatings, powders, fibers, and uniform compo‐ nents. Metals, composites of nonmetallic materials such as carbon, silicon, carbides, nitrides, oxides, and intermetallics can be deposited through CVD. Considering one of the main applications of CVD, which is the synthesis of thin films and coatings, this procedure has in general overcome some of the problems Facing the chemical synthesis of thin films, and simplified the process by having a single-step uniform fabrication. In the CVD technique, a precursor gas flows into a chamber, over the heated substrates to be coated, and deposition of thin films on the surface occurs due to the chemical reaction in vapor phase as shown in **Figure 1A**. This procedure could be defined as an atomistic process where the primary Species deposited are atoms or molecules or a combination of both. Other common deposition procedures include physical vapor deposition technique (PVD), which uses evaporation, sputtering, and other physical processes to produce Vapors of materials instead of chemical processes. The PVD technique, however, has some drawbacks, such as low deposition rate and low pressure requirements. Moreover, this process might require subsequent annealing, which could be a drawback. Although these two techniques can be used separately, howev‐ er, there are procedures that have utilized The benefits of the combination of CVD and PVD techniques, such as a modified PEVCD method.

**Figure 1.** (A) Schematic of chemical vapor deposition technique demonstrating the various components of the instru‐ ment. (B) Microwave PECVD unit (reprinted with permission from [1], © Elsevier).

There are many hybrid techniques, which arise from CVD and are constantly evolving to modify the properties of the fabricated thin films [2], namely initiated CVD (iCVD), which utilizes an initiator to start the polymerization process because the initiator thermally decom‐ poses to form free radicals and they are absorbed on the surface and therefore lead to start the radical polymerization, Oxidative CVD (oCVD), which can be considered as an iCVD process and utilizes an oxidant molecule as the initiator for starting the polymerization [3], metalloorganic CVD (MOCVD), which operates at temperatures lower than CVD (300–800°C) and is used for achieving epitaxial growth and deposition of semiconductor materials [4, 5], Where deposition occurs at low pressure (10 –1000 Pa) to achieve and homogeneity of the deposited films [6], atmospheric pressure CVD (APCVD) in which deposition occurs at atmospheric pressure, which leads to lower equipment cost, avoids use of vacuum systems and simplicity of utilization and process control parameters [7] and plasma-enhanced CVD (PECVD), which uses electrical energy for producing a plasma, and the produced plasma activates the reaction by transferring the energy of its species to the precursors and induces free radical formation followed by radical polymerization [2]. Among them, PECVD can extend applicability of the vapor deposition process to various precursors, including reactive organic, inorganic, and inert materials. In this chapter, we will briefly demonstrate the characteristics of the PECVD technique, its advantages and disadvantages over other techniques, an overview of studies utilizing PECVD for both organic and inorganic materials deposition, as well as the future direction and potential applications.

**1. Introduction**

techniques, such as a modified PEVCD method.

Chemical vapor deposition (CVD) is a multifaceted procedure which is currently used for several applications such as the fabrication of coatings, powders, fibers, and uniform compo‐ nents. Metals, composites of nonmetallic materials such as carbon, silicon, carbides, nitrides, oxides, and intermetallics can be deposited through CVD. Considering one of the main applications of CVD, which is the synthesis of thin films and coatings, this procedure has in general overcome some of the problems Facing the chemical synthesis of thin films, and simplified the process by having a single-step uniform fabrication. In the CVD technique, a precursor gas flows into a chamber, over the heated substrates to be coated, and deposition of thin films on the surface occurs due to the chemical reaction in vapor phase as shown in **Figure 1A**. This procedure could be defined as an atomistic process where the primary Species deposited are atoms or molecules or a combination of both. Other common deposition procedures include physical vapor deposition technique (PVD), which uses evaporation, sputtering, and other physical processes to produce Vapors of materials instead of chemical processes. The PVD technique, however, has some drawbacks, such as low deposition rate and low pressure requirements. Moreover, this process might require subsequent annealing, which could be a drawback. Although these two techniques can be used separately, howev‐ er, there are procedures that have utilized The benefits of the combination of CVD and PVD

248 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 1.** (A) Schematic of chemical vapor deposition technique demonstrating the various components of the instru‐

There are many hybrid techniques, which arise from CVD and are constantly evolving to modify the properties of the fabricated thin films [2], namely initiated CVD (iCVD), which utilizes an initiator to start the polymerization process because the initiator thermally decom‐ poses to form free radicals and they are absorbed on the surface and therefore lead to start the radical polymerization, Oxidative CVD (oCVD), which can be considered as an iCVD process and utilizes an oxidant molecule as the initiator for starting the polymerization [3], metalloorganic CVD (MOCVD), which operates at temperatures lower than CVD (300–800°C) and is

ment. (B) Microwave PECVD unit (reprinted with permission from [1], © Elsevier).


**Table 1.** Comparison of the characteristics, advantages and disadvantages of the CVD, PVD, and PECVD techniques.

The CVD process, its modifications, and PVD each have their own unique advantages and applications. CVD may successfully overcome issues such as line-of sight deposition, which PVD commonly faces. This can improve the ease of coating both nonuniform and straight surfaces. From an economical perspective, CVD also offer advantages due to the ease of synthesis of thick coatings. The process also allows flexibility during deposition such as

codeposition of different materials, inclusion of plasma, or initiators to improve reactivity and operation at atmospheric pressure. On the other hand, CVD in its unmodified form has disadvantages, such as the requirement of higher temperatures (over 600°C). The high temperature is a constraint both due to the energy required to heat the gas phase, which increases the cost of the process and the limitation of materials and substrates that can be used due to their instability at higher temperatures. The other disadvantage of the process is the utilization of chemical precursors with high vapor pressure, such as halides, metal-carbonyl precursors, and hybrid ones which may be an issue due to some of their associated toxici‐ ties, as well as the limitation of types of materials that may be used as precursors. Moreover, the by-products of the CVD process maybe toxic and their neutralization can make the process expensive. However, PECVD, which has seen a rapid development in the last few years, has eliminated these problems to a great extent as discussed in **Table 1**.

### **1.1. History of PECVD**

The initial experiments using plasma in the CVD process dates back to 1950s and 1960s when the decomposition of organic compounds in the presence of an electron beam was first observed [8–10]. In the electro-optical systems, when a surface was exposed to an electron beam, it was covered by a thin film. Upon observing this effect, researchers reasoned that the formation of the thin film was due to the interactions between the electron beam and organic vapor present in the vacuum system. They concluded that the mechanism of formation of the thin layer was the free radical polymerization of the organic molecules, upon exposure to the electron beam followed by adherence to the target surface [8–10]. After roughly a decade, Buck et al. suggested that the observed mechanism of film formation can be beneficial for applica‐ tions where thin insulating films were required [11]. Subsequently, relative electron bombard‐ ment of appropriate chemical compounds was used for the fabrication of polymeric or metallic thin films. Christy et al. in 1960 used a defocused, low energy electron beam to form solid films composed of silicon oil. The factors affecting the film growth rate, such as oil pressure, strength of electron beam current, and temperature were also investigated. He showed that the produced films had excellent electrical insulating properties [11]. In 1961, Baker et al. have Demonstrated the production of metallic films from decomposition of organometallic vapors by an electron beam [12]. Among the initial reports on the use of PECVD, there was one in 1962, in the Electronics Laboratory of General Electric Company for the utilization of the direct current glow discharge system for the production of thin films [13]. Laendle et al. reported the use of an organosilicon compound and its decomposition in a low-energy plasma to fabricate a silicon oxide film [10]. The decomposition reported was related to the collisions that occurred between the excited or ionized gas atoms with the organosilicon molecules, which in turn resulted in the formation of free radicals, followed by their attachment on the substrate and finally due to continuous bombardment, fabrication of stable Si-O-Si network and the resultant thin film. In 1964, Davern et al. reported the use of the same process for the fabrication of silicon oxide films at room temperature, but with the use of radio frequency (RF) plasma production source to reduce the undesired sputtering observed in the previous report. A continuous bombardment of the surface was used to prevent the deposition of decomposition products. In the following year, they also investigated the use of the silicon dioxide films as a diffusion

masking against other diffusants such as N-type semiconductors including GaAs [14]. In 1965, the deposition of silicon oxide thin film capacitor was demonstrated in an AC glow discharge plasma. The produced film was amorphous, coherent and had good adherence. These capacitors also demonstrated thermal stability, resistance to humidity, and breakdown strength characteristics [15]. In 1971, Reinberg developed the "Reinberg reactor," which involved the use of a parallel plate, capacitively coupled RF reactor for depositions at lower temperatures for semiconductor encapsulation and optical coatings [16, 17]. This reactor was capable of inducing radial, laminar flow of the reactant gases over the substrate, resulted in enhancing the uniformity of the desired coating [16–18]. During this time, the process was also referred to as the chemical ion-plating and was commonly used for the deposition of silicon nitride films for encapsulation, and metal carbides [16]. Subsequently, the "plasma-assisted" or plasma-enhanced CVD as it is commonly known as evolved over the past few decades, and some of the process parameters and applications are discussed further on.

### **2. Methods**

codeposition of different materials, inclusion of plasma, or initiators to improve reactivity and operation at atmospheric pressure. On the other hand, CVD in its unmodified form has disadvantages, such as the requirement of higher temperatures (over 600°C). The high temperature is a constraint both due to the energy required to heat the gas phase, which increases the cost of the process and the limitation of materials and substrates that can be used due to their instability at higher temperatures. The other disadvantage of the process is the utilization of chemical precursors with high vapor pressure, such as halides, metal-carbonyl precursors, and hybrid ones which may be an issue due to some of their associated toxici‐ ties, as well as the limitation of types of materials that may be used as precursors. Moreover, the by-products of the CVD process maybe toxic and their neutralization can make the process expensive. However, PECVD, which has seen a rapid development in the last few years, has

250 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

The initial experiments using plasma in the CVD process dates back to 1950s and 1960s when the decomposition of organic compounds in the presence of an electron beam was first observed [8–10]. In the electro-optical systems, when a surface was exposed to an electron beam, it was covered by a thin film. Upon observing this effect, researchers reasoned that the formation of the thin film was due to the interactions between the electron beam and organic vapor present in the vacuum system. They concluded that the mechanism of formation of the thin layer was the free radical polymerization of the organic molecules, upon exposure to the electron beam followed by adherence to the target surface [8–10]. After roughly a decade, Buck et al. suggested that the observed mechanism of film formation can be beneficial for applica‐ tions where thin insulating films were required [11]. Subsequently, relative electron bombard‐ ment of appropriate chemical compounds was used for the fabrication of polymeric or metallic thin films. Christy et al. in 1960 used a defocused, low energy electron beam to form solid films composed of silicon oil. The factors affecting the film growth rate, such as oil pressure, strength of electron beam current, and temperature were also investigated. He showed that the produced films had excellent electrical insulating properties [11]. In 1961, Baker et al. have Demonstrated the production of metallic films from decomposition of organometallic vapors by an electron beam [12]. Among the initial reports on the use of PECVD, there was one in 1962, in the Electronics Laboratory of General Electric Company for the utilization of the direct current glow discharge system for the production of thin films [13]. Laendle et al. reported the use of an organosilicon compound and its decomposition in a low-energy plasma to fabricate a silicon oxide film [10]. The decomposition reported was related to the collisions that occurred between the excited or ionized gas atoms with the organosilicon molecules, which in turn resulted in the formation of free radicals, followed by their attachment on the substrate and finally due to continuous bombardment, fabrication of stable Si-O-Si network and the resultant thin film. In 1964, Davern et al. reported the use of the same process for the fabrication of silicon oxide films at room temperature, but with the use of radio frequency (RF) plasma production source to reduce the undesired sputtering observed in the previous report. A continuous bombardment of the surface was used to prevent the deposition of decomposition products. In the following year, they also investigated the use of the silicon dioxide films as a diffusion

eliminated these problems to a great extent as discussed in **Table 1**.

**1.1. History of PECVD**

The CVD process can operate in two different reactor schemes: a closed reactor and an open one. In the closed reactor CVD, which is the most common one, the species are placed in a container and the container is closed; however, in the open reactor CVD, or flowing-gas CVD, the chemicals are being entered the system continuously. However, in both systems, there is a reactant supply section, in which the reactant, which may be in a gas, liquid, or solid phase, must be transported into the reactor. In the case of gaseous reactants, they can be transferred into the reactor through pressure controllers, however, in the case of liquid or solid reactants, the reactants need to be heated up to their evaporation temperature, which can sometimes be problematic to the process. Then the reactants can be transported by an inert carrier gas to the reactor [5].

Depending on the activation energy used in the CVD process, such as temperature, photon, or plasma it can be categorized into several groups, namely, thermal CVD, laser/photo CVD, and plasma CVD. Thermal CVD typically requires high temperature in the ranges of 800– 2000°C, which can be produced by methods such as hot plate heating, and radiant heating. In the case of laser or photo CVD procedures, photons are produced using either a high power laser or ultraviolet (UV) radiation. Laser CVD involves the use of a laser for the generation of a strong beam of photons for the activation of CVD and photo CVD where the reaction is activated by the use of UV radiations for the dissociation of chemical bonds. However, in plasma CVD, or plasma-enhanced CVD (PECVD), the reaction is activated by the use of inert gas plasma. The utilization of plasma decreases the deposition temperature to a great degree in comparison with other methods such as thermal CVD. Deposition can occur at room temperature and the thermal effects are avoided and this opens up the opportunity of coating polymers and other materials with low melting temperatures [19]. As previously mentioned, PECVD is a variant of the CVD process in which an inert gas plasma is used for deposition of thin films. Thus far, CVD has been mostly applied to inorganic materials. The plasma, which is used in the PECVD technique, allows the usage of a wide range of precursors, both organic and inorganic due to the improved reactivity of the precursors [20, 21]. Plasma is a partially or fully ionized gas and generally is a mixture of electrons, charged particles, and neutral atoms, and therefore, the plasma state has extremely high energy; however, the plasma has no net charge, i.e., neutral. The energy that is available in a plasma discharge is used for various applications, one of these applications is the deposition of thin films and coatings.

In the PECVD process, an external energy source is required for the ionization of atoms and molecules (creation of the plasma), a pressure reduction system (maintain the plasma state), and finally, the existence of a reaction chamber. One method for plasma production is the heating of the gas; however, ionization temperatures are extremely high, and this can be limitation of this method. The other way of plasma production is the utilization of electrical energy, at various frequency discharges which can be subcategorized into audio frequency (10 or 20 kHz), radio frequency (13.56 MHz), and microwave frequency (2.45 GHz), and this can result in various types of plasma being produced. However, when a plasma produced by any of these techniques is utilized, the deposition species undergo fragmentation which forces them to become submonomers or free radicals and ions. The plasma induces radical or plasma polymerization in the monomers. Neutral molecules will be ionized or excited when the electrons and ions in the plasma interact with them, so they will become chemically reactive. The collision of a charged ion with a neutral atom will result in improved chemical reactivity and production of implantation reactions of atoms, radical generation, and polymer-forming reactions or an etching reaction [19, 22].

**Figure 2.** (A) Various processes that polymer surfaces can undergo in the time of exposure to the plasma stream (re‐ printed with permission from [16], © Springer). (B) Schematic of various modifications on polymer surfaces during plasma treatments (reprinted with permission from [27], © 2014 Walter de Gruyter).

The monomers that are used in this procedure are mainly in the gas or liquid states, which can be easily evaporated; however, there are studies which use solid monomers as well. Utilization of solid monomers requires inclusion of sublimation apparatus by which the solid monomer can sublime for deposition and this capability allows the use a vast range of materials as monomers [20, 23–26]. When a gaseous or liquid precursor with high vapor pressure is introduced into the PECVD reaction chamber, dissociation and activation of the precursor occur and in the presence of the plasma, which allows the deposition to happen at much lower temperatures compared to CVD. When the plasma comes in contact with the surface of a polymer substrate, modification of the surface can occur in different ways, namely *etching*, where the plasma treatment leads to the removal of materials from the surface; *deposition*, where precursors in the plasma stream are deposited as a plasma polymerized thin layer on the surface; *cross-linking and functionalization*, which involves modifications of the plasma polymers on the surface as shown in **Figure 2** [27–29].

or fully ionized gas and generally is a mixture of electrons, charged particles, and neutral atoms, and therefore, the plasma state has extremely high energy; however, the plasma has no net charge, i.e., neutral. The energy that is available in a plasma discharge is used for various

In the PECVD process, an external energy source is required for the ionization of atoms and molecules (creation of the plasma), a pressure reduction system (maintain the plasma state), and finally, the existence of a reaction chamber. One method for plasma production is the heating of the gas; however, ionization temperatures are extremely high, and this can be limitation of this method. The other way of plasma production is the utilization of electrical energy, at various frequency discharges which can be subcategorized into audio frequency (10 or 20 kHz), radio frequency (13.56 MHz), and microwave frequency (2.45 GHz), and this can result in various types of plasma being produced. However, when a plasma produced by any of these techniques is utilized, the deposition species undergo fragmentation which forces them to become submonomers or free radicals and ions. The plasma induces radical or plasma polymerization in the monomers. Neutral molecules will be ionized or excited when the electrons and ions in the plasma interact with them, so they will become chemically reactive. The collision of a charged ion with a neutral atom will result in improved chemical reactivity and production of implantation reactions of atoms, radical generation, and polymer-forming

**Figure 2.** (A) Various processes that polymer surfaces can undergo in the time of exposure to the plasma stream (re‐ printed with permission from [16], © Springer). (B) Schematic of various modifications on polymer surfaces during

The monomers that are used in this procedure are mainly in the gas or liquid states, which can be easily evaporated; however, there are studies which use solid monomers as well. Utilization of solid monomers requires inclusion of sublimation apparatus by which the solid monomer can sublime for deposition and this capability allows the use a vast range of materials as monomers [20, 23–26]. When a gaseous or liquid precursor with high vapor pressure is introduced into the PECVD reaction chamber, dissociation and activation of the precursor occur and in the presence of the plasma, which allows the deposition to happen at much lower temperatures compared to CVD. When the plasma comes in contact with the surface of a polymer substrate, modification of the surface can occur in different ways, namely *etching*,

plasma treatments (reprinted with permission from [27], © 2014 Walter de Gruyter).

applications, one of these applications is the deposition of thin films and coatings.

252 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

reactions or an etching reaction [19, 22].

**Figure 3.** Schematic of structural differences among conventional polymers and plasma polymers (reprinted with per‐ mission from [30], © 2015 Elsevier).

During the PECVD process, the plasma polymerization of the precursors, which are induced in the plasma stream, is of the random radical recombination type. There is always a radical that initiates the process of polymerization, due to the radicalization of the precursor, which at the end of the process, results in the formation of polymer. And due to the presence of the radicals in each chain of the produced polymer, cross linking of the polymer also occurs; therefore, the final resulting plasma polymer coating has a cross-linked structure [29, 30, 32]. Due to the existence of a greater degree of cross-linked and branched structures, the overall structure of the plasma polymeric thin film is not similar to the conventional polymers, as shown in **Figure 3** [30, 33]. Another evidence for this difference is that the structure of subunits of the plasma polymers, which are referred to as "monomers" for conventional polymers, is not conserved during the process; therefore, they cannot be characterized by their repeating units [33–35]. However, the structure of the subunit can be reserved only under specific conditions, in which all the external and internal plasma parameters are optimized.

### **2.1. Variation of pressure for plasma production**

The internal plasma parameters that affect the final polymerized film, include homogeneity of discharge, distribution of various species in the plasma, and energy of the species, and the external plasma parameters include reactor geometry, applied voltage, frequency, total pressure, and flow rate [32, 36]. Some of the parameters can be modified to create different variants of this technique, namely temperature, length of the deposition, pressure, inert gas flow rate, method of plasma production, and the power of the source. The most important ones among them are pressure and method of plasma production. Based on the pressure during deposition, the PECVD process can be classified as low pressure (LP- PECVD) and atmospheric pressure (AP-PECVD), with different specific applications [34]. Both of these techniques are similar to PECVD in the main steps of the procedure such as passage, transportation and absorption of the gaseous species over the surface, and production of the reaction products at the end [6]. Initial PECVD studies were performed mostly at low pressure to avoid the plasma instability due the mean-free path of the plasma species. But the utilization of AP-PECVD, which means having the plasma at an ambient pressure, has become an area of interest in the recent years. AP-PECVD offers two main advantages, namely, avoiding the cost of expensive vacuum pumping systems and transfer chambers from air to vacuum. The ease of controlling the plasma parameters such as current, gas flow, power, and voltage during the AP-PECVD process also makes it attractive and the schematic for the process is shown in **Figure 4**. Moreover, it has also been observed that APCVD demonstrates superior adherence of the coating on the substrate due to plasma activation [29, 34, 37]. Also, in LP-PECVD, there can be limitation on the size of the substrate to be treated, which is dependent on the size of the vacuum chamber, but it is not the case in AP-PECVD [38]. LP-PECVD has some other disad‐ vantages too, such as lower deposition rates compared with AP-PECVD. For some of the LP-PECVD techniques, applying higher temperatures is required to bring about more uniformity with less defects for the produced layer, but this can also be considered as a disadvantage for the procedure [6]. Besides, it has been demonstrated that the chemical structure of the materials is more effectively retained in the AP-PECVD due to the lower energy of plasma in comparison with LP-PECVD [37, 39]. Treating materials such as textiles, scaffolds, membranes, and hydrogels is an area where the use of AP-PECVD over LP-PECVD has had a considerable effect [40, 41].

**Figure 4.** (A) Atmospheric RF discharge used for the deposition of single walled carbon nanotubes (reprinted with per‐ mission from [42], © Elsevier). (B) Schematic of APCVD apparatus used for the deposition of zinc oxide (reprinted with permission from [43], © 2007 Elsevier).

The primary advantage of the low pressure plasmas is the high purity of the coatings produced compared to the atmospheric pressure plasmas. When utilizing LP-PECVD, the ratio of mass transport velocity to the velocity of reaction on the surface will be different. When the pressure decreases during LP-PECVD, the diffusion of the gas decreases and therefore the velocity of mass transport will also decrease correspondingly and the produced layer is expected to be more uniform [6]. However, in the AP-PECVD, the ratio of mass transport velocity and the velocity of reaction on the surface is equal to one and the two velocities are of the same order of magnitude. Moreover, another issue facing with AP-PECVD is the problem of maintaining a glow discharge under the atmospheric pressure, which however, has been developed to some extent [38]. One of the major differences between the LP-PECVD and AP-PECVD is in the plasma excitation step. In LP-PECVD, the collision between the monomer with an electron is responsible for activating the monomers and the excitation mechanism can be denoted as follows (1):

$$\mathbf{^\cdot M\_i + e^\cdot \longrightarrow M\_j^\ast + M\_k^\ast + e^\cdot} \tag{1}$$

Although this process exists in the AP-PECVD as well, but at atmospheric pressure, there are other reactions in which unstable high-energy atoms or molecules from the plasma gas exist too, so the aforementioned collision may not be the main collision responsible for activating the monomers. And therefore, the energy transfer in APCVD is more efficacious (2).

$$\text{M}^\* + \text{A}\_2 \xrightarrow{\text{-}} \text{M} + \text{A}^+ + \text{A} + \text{e}^- \tag{2}$$

### **2.2. Methods of plasma production**

variants of this technique, namely temperature, length of the deposition, pressure, inert gas flow rate, method of plasma production, and the power of the source. The most important ones among them are pressure and method of plasma production. Based on the pressure during deposition, the PECVD process can be classified as low pressure (LP- PECVD) and atmospheric pressure (AP-PECVD), with different specific applications [34]. Both of these techniques are similar to PECVD in the main steps of the procedure such as passage, transportation and absorption of the gaseous species over the surface, and production of the reaction products at the end [6]. Initial PECVD studies were performed mostly at low pressure to avoid the plasma instability due the mean-free path of the plasma species. But the utilization of AP-PECVD, which means having the plasma at an ambient pressure, has become an area of interest in the recent years. AP-PECVD offers two main advantages, namely, avoiding the cost of expensive vacuum pumping systems and transfer chambers from air to vacuum. The ease of controlling the plasma parameters such as current, gas flow, power, and voltage during the AP-PECVD process also makes it attractive and the schematic for the process is shown in **Figure 4**. Moreover, it has also been observed that APCVD demonstrates superior adherence of the coating on the substrate due to plasma activation [29, 34, 37]. Also, in LP-PECVD, there can be limitation on the size of the substrate to be treated, which is dependent on the size of the vacuum chamber, but it is not the case in AP-PECVD [38]. LP-PECVD has some other disad‐ vantages too, such as lower deposition rates compared with AP-PECVD. For some of the LP-PECVD techniques, applying higher temperatures is required to bring about more uniformity with less defects for the produced layer, but this can also be considered as a disadvantage for the procedure [6]. Besides, it has been demonstrated that the chemical structure of the materials is more effectively retained in the AP-PECVD due to the lower energy of plasma in comparison with LP-PECVD [37, 39]. Treating materials such as textiles, scaffolds, membranes, and hydrogels is an area where the use of AP-PECVD over LP-PECVD has had a considerable effect

254 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 4.** (A) Atmospheric RF discharge used for the deposition of single walled carbon nanotubes (reprinted with per‐ mission from [42], © Elsevier). (B) Schematic of APCVD apparatus used for the deposition of zinc oxide (reprinted

[40, 41].

with permission from [43], © 2007 Elsevier).

The different methods of plasma production, include microwave frequency (MW) plasma, also known as glow-discharge, radio frequency (RF) plasma, and arc plasma. By utilizing two electrodes in the reaction chamber, the high frequency power is applied to the gas to be ionized. When the microwave frequency is used, plasma is produced by the application of a high frequency electric field of magnitude of 2.45 GHz at a low pressure to a carrier gas such as oxygen, argon, or nitrogen, and the schematic of this process is shown in **Figure 5A**. At such high frequencies, electrons due to their small masses can accelerate toward the higher energy levels; however, this is not the case for the heavier ions, therefore, the temperature of electrons increases, but for heavier ions, it will remain low, and so this is a nonisothermal plasma. In the microwave plasmas, the electrons with their high energy levels are responsible for collision with the gas molecules to dissociate them and make them reactive for further plasma poly‐ merization and one such reactor is shown in **Figure 1B** [1]. This technique has been used for deposition of conductive films from a metal target. There are also many reports on the utilization of lower frequency discharges such as RF [23, 24]. Usually, in the case of an RF source, a matching box is used alongside the RF generator to reduce the reflected power, as shown in **Figure 5B**. The power setting differs based on the material deposited. The RF source is mainly used for nonconductive materials. Unlike for the LP-PECVD case in which micro‐ wave frequency or a microbeam plasma generator is used, production of a stable plasma in an AP-PECVD utilizing an RF source does not require a dielectric material between the electrodes [44]. Arc plasmas typically operate at moderate frequencies of 1 MHz lower than the micro‐ wave plasma sources.

**Figure 5.** (A) Schematic of carbon nanotube growth procedure with microwave PECVD apparatus (reprinted with per‐ mission from [45], © InTech). (B) Schematic of the radio frequency PECVD apparatus for the production of carbon nanowalls (reprinted with permission from [46], © 2005 Elsevier). (C) Schematic of a dielectric barrier discharge system (reprinted with permission from [47], © 2003 Springer).

However, arc plasmas are extremely hot and require higher power, so they are not usually utilized and some of the efforts to avoiding this include limiting the discharge time, which is the time applied for the electrons to accelerate in the electric field, by applying pulsed voltages or using a dielectric material between the electrode to reduce the accumulation of the electrons, which is called dielectric barrier discharges (DBD), as shown in **Figure 5C** [19, 34]. Arc plasma CVD techniques are mainly used for deposition of diamonds where the process occurs at a higher pressures compared to MW PECVD, and therefore, the number of collisions increases, and, consequently, the temperature is higher. Due to high temperatures, total dissociation of hydrogen molecules occurs and proves to be advantageous over the MW PECVD and thermal CVD due to the decreased production of atomic hydrogen species [19]. Due of the high temperature of the arc plasma deposition, substrate cooling is necessary, but it is difficult process; therefore, the high temperature of this procedure counts as a disadvantage of it; however, high rate of deposition and thickness of the produced layer are some of the advan‐ tages. The plasma at different frequencies (microwave, radio frequency, and arc plasma) can be produced by various methods such as inductively coupled or capacitive-coupled parallel plates and/or electron cyclotron resonance (ECR). For example, an MW plasma can be produced by the latter method, which is the case when the frequency of the applied electric field becomes matched with the natural frequency of the electrons when they orbit the magnetic field. The RF frequency plasma can be generated in the capacitive-coupled parallel plate configuration plasma reactor in which one of the electrodes is electrically grounded while the RF power is applied to the other electrode as shown in **Figure 5A** [46]. By considering the temperature of the species present in the plasma stream, the process can be divided into two categories. One is the case in which all the plasma species: ions, electrons, and neutral species all exist at the same temperatures, which means that plasma is at the local thermal equilibrium (LTE), and the other is when this equilibrium does not exist and the species are at different temperatures (non-LTE), and electrons may be at a higher temperature in comparison with the other particles. In a LTE plasma state, the density of electrons is usually higher than the non-LTE state, and therefore, the mean free path of the electrons is lower, and, consequently, the number of collisions increases. In the case of AP-PECVD, the plasma is expected to be in the LTE state since the higher pressure also leads to increased number of collisions; therefore, the temperature of the molecules and ions increases [43]. This case is called an arc plasma, which operates at lower frequencies compared to microwave and higher frequencies compared to RF sources (1 MHz) [19, 34].

wave frequency or a microbeam plasma generator is used, production of a stable plasma in an AP-PECVD utilizing an RF source does not require a dielectric material between the electrodes [44]. Arc plasmas typically operate at moderate frequencies of 1 MHz lower than the micro‐

256 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 5.** (A) Schematic of carbon nanotube growth procedure with microwave PECVD apparatus (reprinted with per‐ mission from [45], © InTech). (B) Schematic of the radio frequency PECVD apparatus for the production of carbon nanowalls (reprinted with permission from [46], © 2005 Elsevier). (C) Schematic of a dielectric barrier discharge system

However, arc plasmas are extremely hot and require higher power, so they are not usually utilized and some of the efforts to avoiding this include limiting the discharge time, which is the time applied for the electrons to accelerate in the electric field, by applying pulsed voltages or using a dielectric material between the electrode to reduce the accumulation of the electrons, which is called dielectric barrier discharges (DBD), as shown in **Figure 5C** [19, 34]. Arc plasma CVD techniques are mainly used for deposition of diamonds where the process occurs at a higher pressures compared to MW PECVD, and therefore, the number of collisions increases, and, consequently, the temperature is higher. Due to high temperatures, total dissociation of hydrogen molecules occurs and proves to be advantageous over the MW PECVD and thermal CVD due to the decreased production of atomic hydrogen species [19]. Due of the high temperature of the arc plasma deposition, substrate cooling is necessary, but it is difficult process; therefore, the high temperature of this procedure counts as a disadvantage of it; however, high rate of deposition and thickness of the produced layer are some of the advan‐ tages. The plasma at different frequencies (microwave, radio frequency, and arc plasma) can be produced by various methods such as inductively coupled or capacitive-coupled parallel

wave plasma sources.

(reprinted with permission from [47], © 2003 Springer).

PECVD can be considered as a nonline of sight process, because the plasma stream can surround the substrate, and therefore, leads to uniform deposition and alleviates some of the issues associated with ionic or e-beam sputtering processes. It has been well-documented in the literature that the resulting chemistry of the thin films produced by PECVD is unique and cannot be obtained by common wet deposition techniques. These fabricated thin films also demonstrate high solvent and corrosion resistance along with thermal and chemical stability [24]. The process results in efficient and controlled usage of precursors and the generation of very little by-products. The process allows the utilization of unconventional precursors and can be used for deposition on surfaces with complicated geometries [48]. Moreover, because the interactions are occurring just at the surface of the substrates, mechanical properties of the substrates will not change and the film that is formed is completely uniform [49]. The plasma polymers fabricated by this technique have high modulus, due to the dense cross-linking in their polymer chains. There are some reports that show these produced polymer layers have a high permeability coefficient of oxygen and nitrogen which is mainly related to the deposi‐ tion mechanism as well as the polymerization mechanism of plasma polymers [22]. Besides, if the plasma polymers contain ionic groups, they will have a wide variety of applications in various fields which require electrical conductivity, antistatic esterification, and biocompati‐ bility.

A major difference between the films produced by PECVD and CVD is that, the ones produced through PECVD have higher content of hydrogen, which is due to the utilization of plasma in the deposition process [50]. The synthesized plasma polymers have some disadvantages as well, such as instability against aging and humidity which is due to free radicals that exist in the plasma polymers as well as the existence of polar side groups that can absorb water, and therefore may render them unsuitable for dielectric thin films [49, 51]. Some ways of over‐ coming this problem such as heat treatment of plasma polymers in a vacuum which will decrease the radical concentration in the deposited thin films has been reported [35, 49, 51]. Another disadvantage is that undesirable compressive and residual stresses in the deposited film can occur at lower frequencies, which could lead to subsequent cracking and cause problems in many applications, specifically metallurgy [35]. Residual stresses can also affect thick oxide films where fracture has been observed [52]. Techniques for controlling the residual stress, which ultimately resulted in limiting the thickness of the produced layer, such as rapid thermal analysis (RTA) have been utilized [53]. However, this treatment may have other impacts on the characteristics of the produced film too; therefore, the behavior of the PECVD oxide layer following the heat treatment should be analyzed [54]. Also, the ion bombardment during the process could damage some sensitive substrates. Moreover, the problem of line-ofsight does exist to some extent and therefore, the regions that are directly exposed to the plasma stream will be more affected than the other regions. One other major issue is that due to the presence of several species in the plasma stream (with varying mass and charge), the film growth can be influenced by different species. The mechanism of this growth depends on the chemistry of the depositing species and the conditions used for deposition. For example, in DLC film deposition, it has been shown that both ions and neutral species are a part of the mass of the produced film [55]. However, in the production of silane or polymeric films, the role of radicals, as well as the influence of plasma pressure, has been considered [56, 57]. Therefore, development in the understanding the roles of different plasma species on the film growth, will enhance the film production, as well as control of the process [58].

### **3. Applications**

The initial application of PECVD was in microelectronics, namely electrical insulation films, conductive films, semiconductors, diffusion masks, and capacitors. Over a decade after the establishment of PECVD process, researchers started to utilize this technique for the fabrication of integrated circuits (ICs) [15, 52, 59], solar cells [43, 60, 61], transistors [62], photovoltaic and photonic applications [48, 59, 63, 64] such as notch filters and antireflective coatings [27]. Deposition of inorganic materials by PECVD also has applications in food packaging, scratch resistance, and biomedical applications [2, 65, 66]. PECVD has been extensively used in the development of organic polymers and organometallic compounds. Plasma polymers are used to avoid corrosion in many dielectric and optical devices. The film properties can be easily controlled to produce homogenous organic thin films on large substrates (silicon and glass) with varying mechanical, optical, chemical, thermal, and electric properties. Due to its relatively moderate nature, this technique has also recently been used for various biological applications. Biological applications include immobilization of the biomolecules (e.g. enzymes and proteins); biometallization of gold, silver and platinum nanoparticles with the aid of DNA, amino acids or proteins; Degradation and molecular weight reduction of biopolymers such as chitosan in order to enhance their properties; Changing surface functionalities by addition of functional groups such as amine, carboxylic and silane groups; Fabrication of organic/ inorganic hybrid film for enhancing the dielectric properties. Such surface treatment of implants and biomaterials for applications in biosensors, device fabrication and anti-biofoul‐ ing coatings [20, 25, 40, 67].

### **3.1. Silicon-based applications**

coming this problem such as heat treatment of plasma polymers in a vacuum which will decrease the radical concentration in the deposited thin films has been reported [35, 49, 51]. Another disadvantage is that undesirable compressive and residual stresses in the deposited film can occur at lower frequencies, which could lead to subsequent cracking and cause problems in many applications, specifically metallurgy [35]. Residual stresses can also affect thick oxide films where fracture has been observed [52]. Techniques for controlling the residual stress, which ultimately resulted in limiting the thickness of the produced layer, such as rapid thermal analysis (RTA) have been utilized [53]. However, this treatment may have other impacts on the characteristics of the produced film too; therefore, the behavior of the PECVD oxide layer following the heat treatment should be analyzed [54]. Also, the ion bombardment during the process could damage some sensitive substrates. Moreover, the problem of line-ofsight does exist to some extent and therefore, the regions that are directly exposed to the plasma stream will be more affected than the other regions. One other major issue is that due to the presence of several species in the plasma stream (with varying mass and charge), the film growth can be influenced by different species. The mechanism of this growth depends on the chemistry of the depositing species and the conditions used for deposition. For example, in DLC film deposition, it has been shown that both ions and neutral species are a part of the mass of the produced film [55]. However, in the production of silane or polymeric films, the role of radicals, as well as the influence of plasma pressure, has been considered [56, 57]. Therefore, development in the understanding the roles of different plasma species on the film

258 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

growth, will enhance the film production, as well as control of the process [58].

The initial application of PECVD was in microelectronics, namely electrical insulation films, conductive films, semiconductors, diffusion masks, and capacitors. Over a decade after the establishment of PECVD process, researchers started to utilize this technique for the fabrication of integrated circuits (ICs) [15, 52, 59], solar cells [43, 60, 61], transistors [62], photovoltaic and photonic applications [48, 59, 63, 64] such as notch filters and antireflective coatings [27]. Deposition of inorganic materials by PECVD also has applications in food packaging, scratch resistance, and biomedical applications [2, 65, 66]. PECVD has been extensively used in the development of organic polymers and organometallic compounds. Plasma polymers are used to avoid corrosion in many dielectric and optical devices. The film properties can be easily controlled to produce homogenous organic thin films on large substrates (silicon and glass) with varying mechanical, optical, chemical, thermal, and electric properties. Due to its relatively moderate nature, this technique has also recently been used for various biological applications. Biological applications include immobilization of the biomolecules (e.g. enzymes and proteins); biometallization of gold, silver and platinum nanoparticles with the aid of DNA, amino acids or proteins; Degradation and molecular weight reduction of biopolymers such as chitosan in order to enhance their properties; Changing surface functionalities by addition of functional groups such as amine, carboxylic and silane groups; Fabrication of organic/ inorganic hybrid film for enhancing the dielectric properties. Such surface treatment of

**3. Applications**

PECVD is recognized as a unique method of organic thin film deposition and is commonly used for the deposition of inorganic and organic, doped, and undoped films. Silicon composite films is one area where thin films are deposited from various precursors such as silane (SiH4), siloxane, and silazane mixed with other gases in varying ratios [68, 69]. The two predominant forms of silicon which get deposited are hydrogenated amorphous (a-Si:H) and microcrystal‐ line (µc-Si:H) silicon [70–72]. The films deposited through PECVD, are used in semiconductor devices, solar cells, and optically active device applications due to their optical, mechanical, and electrical properties. Recently, PECVD has also been utilized for the processing of flexible and printable electronic devices, due to its high process efficiency, large-scale patternability, lower cost, and environmentally friendly nature [73, 74]. AP-PECVD at low gas temperatures due to absence of vacuum equipment, seems beneficial in the surface processing of flexible materials and has been demonstrated in the lower temperature deposition of SiO*x* films on polymer substrates that are used as diffusion barrier and stable hydrophilic surfaces. The applications of flexible semiconductors can be seen in wearable electronics and organic electronics [75, 76].

### *3.1.1. Semiconductor devices*

The physical properties of PECVD-deposited films are known to be superior since they are highly cross-linked, uniform, and resistant to thermal and chemical changes and have low rates of corrosion. Plasma polymers find extensive use in dielectric films and optical coating due to lower cost and higher efficiency [77, 78]. Plasma polymerized organic-inorganic hybrid thin films with ratio controlled precursors has been used to optimize dielectric properties. Also, the characteristics of the thin film were modified by the gases' mixing ratio, plasma power, and chamber pressure [35, 79]. Silicon oxynitride (SiON) and silicon nitride (SiN) are some of the commonly used precursors for the deposition of insulating thin films for MIM (metalinsulator-metal) capacitors and allow easy manipulation of film properties [78]. It was observed that the varying the ratios of silane (SiH4), nitrous oxide (N2O), and ammonia (NH3), lead to the deposition of thin films of varying compositions. Properties including deposition rate, refractive index, and extinction index of the film were observed to improve with the increase in SiH4/N2O (NH3) ratio. Although both SiN and SiON have their advantages, SiN films are used for semiconductor applications due to their higher capacitance density, breakdown voltage and particle performance (**Figure 6A**–**C**) [63, 73, 80]. Silicon carbide, which has a high band gap and chemical inertness, is another material commonly deposited using PECVD. SiC is used widely as a material for MEMS devices due to its stability in harsh and high temperature environments [36, 81, 82]. PECVD, being a metal-compatible deposition method has been used for the deposition of α-SiC at 450°C using methane and silane precursor gases. Low frequency RF (40 Hz) power and pressure of range 700−1000 Torr, along with changes in the precursor ratios were utilized for the production of films of varying electrical and mechanical properties. The results indicate the potential of the α-SiC films as structural components for devices that need stable residual film stress at high temperatures of operation [36]. Although the thin films of SiC have demonstrated promise in the development of high temperature withstanding MEMS devices, the increase in surface roughness in deposited thicker films remains an obstacle. Increasing the carbon content of the film in mixed frequency condition and decreasing the RF power was observed to decrease the surface roughness, making it a potential solution. Such methods pave the way for future studies on the influence of PECVD process parameters on the growth rate, morphology, and microstructure in the field of semiconductors.

**Figure 6.** (A) Schematic representation of graphene FET device with silicon nitride coating. (B and C) SEM images of silicon nitride and silicon oxide deposited on graphene (A, B and C: reprinted with permission from [80] © 2010 ACS). (D) 2D surface morphology of n-µc-Si: H layer used in single junction a-Si solar cells [83]. (D, a) SEM images of p-type Silicon nanowires using PECVD on a glass substrate (Sn catalyzed, 600°C). (b) Deposition of a conformal intrinsic and n-type a-Si:H after a temperature drop of 175°C using PECVD. (c) Schematic representation of a PIN radial junction. (Reprinted with permission [84], © 2011 Elsevier). (E) I–V characteristics of solar cells with n-a-Si: H layer and n-µc-Si:H/n-a-Si: H bilayer (F and G: reprinted with permission [83], © 2015 Solar Energy).

### *3.1.2. Solar cells*

Various deposition techniques such as PVD, CVD, oxidation, plating, and spin-coating have been used extensively for the deposition of silicon thin films with good photoelectric properties [85]. Currently, the photovoltaic market is dominated by crystalline silicon cells, which accounted for nearly 95% of world's photovoltaic cell and module production in 2004 [85, 86]. PECVD is widely used in the microelectronics and solar cell production, specifically for the deposition of thin films from a mixture of gas-phase species onto a solid substrate [87]. There has been a continuous effort to establish the process conditions to improve the film quality and the rate of deposition. PECVD at 60 MHz excitation frequency is now used extensively in large area thin film technology and industrial applications [83, 88, 89].

An increase in growth rate (going up from 1.5 to 4 nm s−1) was obtained by switching from a low pressure (2−4 Torr) conventional RF plasma (13.56 MHz) to a 60 MHz excited plasmaassisted CVD was observed by Matsuda et al. and by several other research groups [88, 89]. The crystallinity of the deposited silicon plays a vital role in the film properties. It has been observed that even high quality amorphous silicon solar cells have very low efficiency in trapping light when compared to crystalline counterparts, which has led to various studies to improve solar power conversion (**Figure 6F**) [83]. Several groups have worked on the trans‐ formation of the silicon thin film from amorphous to nanocrystalline in the presence of nonthermal SiH4/H2 plasmas. The interactions between the hydrogen atoms of the plasma and the solid silicon matrix determine the crystallinity and other properties of the thin films. The presence of H atoms in between the strained Si-Si bonds led to the disordered-to-ordered transitions [70, 71]. The deposition parameters such as chamber pressure (*P*r), gas mixture composition, and flow rates, RF power density (*P*w), and substrate temperature can be varied to obtain the desired film properties, including the crystallinity [69, 70]. Microdoping is another process used for the betterment of optical, electrical and structural properties of a microcrys‐ talline silicon film for solar cell applications. The effect of boron microdoping through PECVD at high temperature and very high frequency (VHF, 60 MHz) PECVD at 200°C with a varying diborane flow rate of 0.00–0.30 sccm has been studied [90, 91]. Improved dark current flow and conductivity changes up to 10 orders of magnitude were observed upon boron doping was observed in these studies and could subsequently be used to improve solar conversion efficiency.

Another area where PECVD has been used extensively is the deposition of a passivation layer, which involves a shielding outer-layer deposited as a microcoating. In microelectronics and photovoltaics, a passivating layer can reduce surface recombination, which is a significant cause of losses in solar cells. Hydrogen atoms play an important role in the termination and passivation of silicon dangling bonds [70, 92]. Hydrogenated amorphous and nanocrystalline silicon films produced by PECVD are now widely being used in electronic and optoelectronic devices [70, 89]. Also, nanostructures have been used extensively to improve efficiency of solar cells in the recent times. One example is the use of silicon nanowires to improve the lighttrapping in low-cost silicon photovoltaic cells (**Figure 6E**) [84, 93]. The growth of radial junctions of hydrogenated amorphous silicon over p-doped crystalline silicon nanowires on a glass substrate using PECVD (single pump) and Sn catalysts has demonstrated a considerable increase in the dark current and conversion efficiencies [84]. Silicon solar cells produced on flexible plastic substrates at an optimum deposition temperature of ~200°C, typically are cheap and robust [61, 94, 95].

### *3.1.3. Optically active films*

and mechanical properties. The results indicate the potential of the α-SiC films as structural components for devices that need stable residual film stress at high temperatures of operation [36]. Although the thin films of SiC have demonstrated promise in the development of high temperature withstanding MEMS devices, the increase in surface roughness in deposited thicker films remains an obstacle. Increasing the carbon content of the film in mixed frequency condition and decreasing the RF power was observed to decrease the surface roughness, making it a potential solution. Such methods pave the way for future studies on the influence of PECVD process parameters on the growth rate, morphology, and microstructure in the field

260 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 6.** (A) Schematic representation of graphene FET device with silicon nitride coating. (B and C) SEM images of silicon nitride and silicon oxide deposited on graphene (A, B and C: reprinted with permission from [80] © 2010 ACS). (D) 2D surface morphology of n-µc-Si: H layer used in single junction a-Si solar cells [83]. (D, a) SEM images of p-type Silicon nanowires using PECVD on a glass substrate (Sn catalyzed, 600°C). (b) Deposition of a conformal intrinsic and n-type a-Si:H after a temperature drop of 175°C using PECVD. (c) Schematic representation of a PIN radial junction. (Reprinted with permission [84], © 2011 Elsevier). (E) I–V characteristics of solar cells with n-a-Si: H layer and n-µc-

Various deposition techniques such as PVD, CVD, oxidation, plating, and spin-coating have been used extensively for the deposition of silicon thin films with good photoelectric properties [85]. Currently, the photovoltaic market is dominated by crystalline silicon cells, which accounted for nearly 95% of world's photovoltaic cell and module production in 2004 [85, 86]. PECVD is widely used in the microelectronics and solar cell production, specifically for the deposition of thin films from a mixture of gas-phase species onto a solid substrate [87]. There has been a continuous effort to establish the process conditions to improve the film quality and the rate of deposition. PECVD at 60 MHz excitation frequency is now used extensively in

An increase in growth rate (going up from 1.5 to 4 nm s−1) was obtained by switching from a low pressure (2−4 Torr) conventional RF plasma (13.56 MHz) to a 60 MHz excited plasma-

Si:H/n-a-Si: H bilayer (F and G: reprinted with permission [83], © 2015 Solar Energy).

large area thin film technology and industrial applications [83, 88, 89].

of semiconductors.

*3.1.2. Solar cells*

Thin films with dye molecules are used in laser cavities, optical filters, and optical gas sensors [96–98]. Such thin films are synthesized using wet and sublimation techniques. PECVD avoids the harsh (chemical and temperature) and multistep process of these fabrication methods and can be used to deposit optically active thin films with tailorable properties. The deposition involves a partial polymerization of dye molecules that are evaporated over a substrate while exposed to an arc plasma. Ultrathin highly planar films (~100 nm) with controlled gradation of color were synthesized using the PECVD process. Such films can be used in development of photonic materials and devices such as sensors, and wavelength couplers [99, 100]. Plasma polymerization of fluorine-containing materials yield many properties such as low dielectric constant and refractive index, low surface energy, low coefficient of friction, low permeability constant, and good biocompatibility [23, 101]. Fabrication of thin films using plasma homoand copolymerization of octafluorocyclobutane (OFCB, C4F8) and hexamethyldisiloxane (HMDSO, C6H18Si2O) using PECVD (RF at 13.56 MHz, power 20–45 W and pressure 0.001–1 Torr) were studied for their optical properties [23, 102]. Rare earth-doped optical materials is an area that has also garnered interest in the recent times, Pitt et al. report the synthesis of silicon-rich SiO2 along with erbium doping. PECVD was used for the synthesis of such films due to low processing temperatures, ability of controlling stoichiometry, and growth condi‐ tions for the film. Microclusters of Si were found to be embedded in the erbium doped SiO2 films demonstrating strong absorption peaks in the visible region along with the strong near IR fluorescence due to the erbium ions [97, 103].

### *3.1.4. Quantum dots*

Silicon nanostructures have been found to exhibit luminescent properties, leading to many studies to understand them [63, 104]. Si-rich silicon nitride films have been grown by low frequency (440 kHz) PECVD using SiH4 and NH3 as reactant gases at a plasma power and a chamber pressure of 1000 W and 1500 mTorr, respectively. The total gas flow being 800 sccm, the ratio of NH3/SiH4 flows was changed between 2 and 10, and the flow rates were changed at a temperature of 370°C to get a different Si concentration in films. These kinds of films can be used in photovoltaic devices [105, 106].

### **3.2. Carbon-based nanostructures**

Carbon-based nanostructures can be classified into 2D structures such as nanowalls, graphene, and 3D structures such as nanotubes and nanowires. Based on their morphological features and orientation, such nanostructures are used for different applications. For example, conical structures give higher mechanical and thermal stability when compared to narrow ones and the large surface area of nanowalls prove helpful in electrochemical devices and gas storage [107, 108]. The main drawback in the synthesis of these nanostructures is the absence of a controlled and deterministic method for depositing large-scale assemblies, which can be used in nanoelectronics, biological applications (probes), field emission displays, and radiation sources. PECVD can be used to address these difficulties and modifications of the deposition conditions produces different morphologies, location, and orientation-specific carbon nano‐ structures such as carbon nanotubes, diamond-like carbon (DLC) films deposited using microwave PECVD [108, 109]. However, the use of PECVD for the synthesis of DLC films is limited because of the issue of being unable to reach harder tetrahedral amorphous carbon state, due to the presence of hydrogen in the gaseous precursor [109].

Carbon nanotubes (CNTs) formed from graphene sheets have excellent electronic and mechanical properties. CNTs in the range of 1 nm in diameter and a few micrometers in length can be classified broadly as single-walled or multiwalled CNTs. Single-walled nanotubes (SWNTs) are known for their semiconductor properties and find applications in

high performance electronics. SWNTs have been grown by using RF PECVD (13.56 MHz, power ~75 W, low pressure) with methane as a precursor and metallic seed catalyst as shown in **Figure 5**. MW PECVD process with frequency of 2.45 GHz, pressure of 20 Torr, and power of 5 kW could also be used to produce SWNTs [45, 110]. Various parameters and factors involved in the process influence the preferential growth of the CNTs (**Figure 7**) [108, 110, 112]. Vertically oriented SWNTs have been synthesized, and this alignment is attributed to the strong electric field in a plasma sheath [44, 113]. Such CNTs can be used in various applications such as field emission devices and nanoscale electrochemical actuators [111, 113– 115]. The formation of iron silicide (FeSi) roots on the Si substrate was observed to be helping in the formation of a base growth mode, which led to well-aligned CNTs deposition (**Figure 7B** and **C**) [114]. Recently, Bo et al. have reported the catalyst-free PECVD growth of graphene. The precursor was a volatile natural extract from the tea tree plant (*Melaleuca altornifolia*) [116]. The deposition rate was high and produced high quality graphene films without the presence of a catalyst. The graphene films at the end of this sustainable bottom up process, when tested yielded a high water contact angle of 135°, proving its potential use as a highly hydrophobic coating. Also, due to its mem-resistive behavior, such graphene could be used in the construction of memory devices [116].

of photonic materials and devices such as sensors, and wavelength couplers [99, 100]. Plasma polymerization of fluorine-containing materials yield many properties such as low dielectric constant and refractive index, low surface energy, low coefficient of friction, low permeability constant, and good biocompatibility [23, 101]. Fabrication of thin films using plasma homoand copolymerization of octafluorocyclobutane (OFCB, C4F8) and hexamethyldisiloxane (HMDSO, C6H18Si2O) using PECVD (RF at 13.56 MHz, power 20–45 W and pressure 0.001–1 Torr) were studied for their optical properties [23, 102]. Rare earth-doped optical materials is an area that has also garnered interest in the recent times, Pitt et al. report the synthesis of silicon-rich SiO2 along with erbium doping. PECVD was used for the synthesis of such films due to low processing temperatures, ability of controlling stoichiometry, and growth condi‐ tions for the film. Microclusters of Si were found to be embedded in the erbium doped SiO2 films demonstrating strong absorption peaks in the visible region along with the strong near

262 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Silicon nanostructures have been found to exhibit luminescent properties, leading to many studies to understand them [63, 104]. Si-rich silicon nitride films have been grown by low frequency (440 kHz) PECVD using SiH4 and NH3 as reactant gases at a plasma power and a chamber pressure of 1000 W and 1500 mTorr, respectively. The total gas flow being 800 sccm, the ratio of NH3/SiH4 flows was changed between 2 and 10, and the flow rates were changed at a temperature of 370°C to get a different Si concentration in films. These kinds of films can

Carbon-based nanostructures can be classified into 2D structures such as nanowalls, graphene, and 3D structures such as nanotubes and nanowires. Based on their morphological features and orientation, such nanostructures are used for different applications. For example, conical structures give higher mechanical and thermal stability when compared to narrow ones and the large surface area of nanowalls prove helpful in electrochemical devices and gas storage [107, 108]. The main drawback in the synthesis of these nanostructures is the absence of a controlled and deterministic method for depositing large-scale assemblies, which can be used in nanoelectronics, biological applications (probes), field emission displays, and radiation sources. PECVD can be used to address these difficulties and modifications of the deposition conditions produces different morphologies, location, and orientation-specific carbon nano‐ structures such as carbon nanotubes, diamond-like carbon (DLC) films deposited using microwave PECVD [108, 109]. However, the use of PECVD for the synthesis of DLC films is limited because of the issue of being unable to reach harder tetrahedral amorphous carbon

Carbon nanotubes (CNTs) formed from graphene sheets have excellent electronic and mechanical properties. CNTs in the range of 1 nm in diameter and a few micrometers in length can be classified broadly as single-walled or multiwalled CNTs. Single-walled nanotubes (SWNTs) are known for their semiconductor properties and find applications in

state, due to the presence of hydrogen in the gaseous precursor [109].

IR fluorescence due to the erbium ions [97, 103].

be used in photovoltaic devices [105, 106].

**3.2. Carbon-based nanostructures**

*3.1.4. Quantum dots*

**Figure 7.** (A) SEM image of CNTs grown using microwave PECVD (4 nm Fe film as a catalyst). (B) TEM images of various possible shapes of catalyst at the CNT growth, i.e., not well attached to substrate and completely enclosed). (C) Elongation of catalyst particle in vertical direction moving from (A) and (B) (reprinted with permission from [113], © 2005 ACS). (D–G) CNTs deposited from different mixture ratios of H2/C2H2. The flow rates of H2/C2H2 were 140/2, 160/5, 140/5, and 180/10 sccm, respectively (reprinted with permission from [110], © 2015 Creative Commons Attribu‐ tion (CC BY)).

Carbon nanowalls (CNWs) have been deposited using acetylene as precursors with iron as the catalyst. The growth parameters such as temperature, pressure, and gas flow rate were varied to get various samples of CNWs. The samples which displayed best field emission properties were the ones deposited at higher temperature and pressure (650°C with 120 sccm H2 flow rate and a pressure of 300 Pa) [46, 112]. Graphene an allotrope of carbon is an example of 2D carbon nanostructure with extraordinary properties such as being 100 times stronger than steel and conducting heat and electricity with a nearly transparent sheet like morphology [117]. Graphene deposition for dielectrics is a challenge because of discontinuity in film formation. This can be compensated by the deposition of SiN by PECVD, which provides excellent coverage of graphene for preservation of their carrier mobility for application in FETs [80]. Their unique structural features, i.e., vertical orientation of the substrate without any agglom‐ eration, sharp and exposed edges, and the controllable intersheet connectivity provide numerous advantages. These nanosheets can be deposited using PECVD on a variety of substrates using gas, solid, or liquid precursors [117]. Perpendicular orientation of the graphene sheets demonstrates better properties in comparison with conventional graphene films and is utilized for energy storage, sensors, and environmental applications. There are several unanswered questions in the development of carbon-based nanostructures by PECVD. Some of the unknowns involve the chemical species, or ions that are responsible for nano‐ structure growth, and their influence on the adhesion to the substrate, the rate determining step in their growth, and the effects of pressure and process parameters on the growth. Understanding some of these processes will help in the manipulation of the parameters to acquire required shape and orientation of carbon nanostructures and can lead to efficient large area deposition for various applications [108].

#### **3.3. Titanium-based applications**

Titanium has many applications due to its strength and low density and is used as an alloy, for thin film coatings, as substrates and photocatalysts. Its applications are wide-range and can be predominantly seen in medical prostheses (dental and orthoimplants), antimicrobial coatings, and electronic devices. RF PECVD has been most commonly used in the production of titanium thin films [118–123]. As a photocatalyst, titanium and its oxide (TiO2) films have numerous applications such as antimicrobial activity, waste water treatment, and for biomed‐ ical implants. The antimicrobial activity of titanium-oxide-coated cotton textile substrates (**Figure 8A**–**C**) and glass substrates were tested against *E. coli*. The coatings were uniform and bacterial inhibition increased correspondingly with the increase in the refractive index of the film [118]. Titania powder has also been used for photocatalytic activities and this property was harnessed in the degradation of certain organic compounds, which could further be adapted for waste water treatment, but powders have lower efficiency due to crystalline defects. Although crystalline TiO2 is more efficient, its surface area is not as high as powder. Hence, studies have been conducted to improve the surface area while retaining the crystal‐ linity aspects of TiO2. An excitation frequency of 40 kHz was employed in a cold-wall-type PECVD (precursor: titanium tetra-isopropoxide (Ti (O-*i*-C3H7)4), and oxygen gas carrier flow rate: 50 sccm and reduced pressure conditions ~30 Pa) was used to produce crystalline TiO2 particles, which had a high surface area (**Figure 8E**) and good photocatalytic properties [120]. The combination of PECVD deposited TiO2 films with silver nanoparticles have been studied (**Figure 8D**) to characterize the films for their photolytic activity and antimicrobial property (silver is known for its antimicrobial activity). These films demonstrated good photochemical activity at nearly room temperature and could be used for heat-sensitive substrates [119]. PECVD can also be used for other material-based precursors and here are some examples and applications of such depositions: the mixtures of Ti (O-*i*-C3H7)4 and oxygen can be used to produce amorphous TiO2 thin films by PECVD. The dielectric properties of TiO2 films were put to use by assuming that a TiO2/SiO2 bilayer would behave as two capacitors in series. The rate of deposition was observed to decrease with oxygen concentration and RF power and whereas an increase with equivalence ratio. The film thickness was seen to influence the dielectric constant of the film and it was pretty consistent, even in the presence of an interfacial SiO2 layer [42].

**Figure 8.** (A and B) Low and high magnifications SEM images of cotton knitwear deposited made of titanium oxide with RF PECVD (100 W). (C) SEM image of a single cotton fiber and it has been stripped partially. The EDX results in the two windows are for coated and uncoated fiber (reprinted with permission from [118], © 2005 Elsevier). (D) AFM pictures of the TiO2/PECVD films made at (a) 300°C, (b) 40°C using PECVD (2 × 2 µm) (reprinted with permission from [119], © 2009 Springer). (E) TEM image of the TiO2 powder produced using PECVD following treatment (800°C for 1.5 h) (reprinted with permission from [120], © 1999 Springer).

### **3.4. Biological applications**

Carbon nanowalls (CNWs) have been deposited using acetylene as precursors with iron as the catalyst. The growth parameters such as temperature, pressure, and gas flow rate were varied to get various samples of CNWs. The samples which displayed best field emission properties were the ones deposited at higher temperature and pressure (650°C with 120 sccm H2 flow rate and a pressure of 300 Pa) [46, 112]. Graphene an allotrope of carbon is an example of 2D carbon nanostructure with extraordinary properties such as being 100 times stronger than steel and conducting heat and electricity with a nearly transparent sheet like morphology [117]. Graphene deposition for dielectrics is a challenge because of discontinuity in film formation. This can be compensated by the deposition of SiN by PECVD, which provides excellent coverage of graphene for preservation of their carrier mobility for application in FETs [80]. Their unique structural features, i.e., vertical orientation of the substrate without any agglom‐ eration, sharp and exposed edges, and the controllable intersheet connectivity provide numerous advantages. These nanosheets can be deposited using PECVD on a variety of substrates using gas, solid, or liquid precursors [117]. Perpendicular orientation of the graphene sheets demonstrates better properties in comparison with conventional graphene films and is utilized for energy storage, sensors, and environmental applications. There are several unanswered questions in the development of carbon-based nanostructures by PECVD. Some of the unknowns involve the chemical species, or ions that are responsible for nano‐ structure growth, and their influence on the adhesion to the substrate, the rate determining step in their growth, and the effects of pressure and process parameters on the growth. Understanding some of these processes will help in the manipulation of the parameters to acquire required shape and orientation of carbon nanostructures and can lead to efficient large

264 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Titanium has many applications due to its strength and low density and is used as an alloy, for thin film coatings, as substrates and photocatalysts. Its applications are wide-range and can be predominantly seen in medical prostheses (dental and orthoimplants), antimicrobial coatings, and electronic devices. RF PECVD has been most commonly used in the production of titanium thin films [118–123]. As a photocatalyst, titanium and its oxide (TiO2) films have numerous applications such as antimicrobial activity, waste water treatment, and for biomed‐ ical implants. The antimicrobial activity of titanium-oxide-coated cotton textile substrates (**Figure 8A**–**C**) and glass substrates were tested against *E. coli*. The coatings were uniform and bacterial inhibition increased correspondingly with the increase in the refractive index of the film [118]. Titania powder has also been used for photocatalytic activities and this property was harnessed in the degradation of certain organic compounds, which could further be adapted for waste water treatment, but powders have lower efficiency due to crystalline defects. Although crystalline TiO2 is more efficient, its surface area is not as high as powder. Hence, studies have been conducted to improve the surface area while retaining the crystal‐ linity aspects of TiO2. An excitation frequency of 40 kHz was employed in a cold-wall-type PECVD (precursor: titanium tetra-isopropoxide (Ti (O-*i*-C3H7)4), and oxygen gas carrier flow rate: 50 sccm and reduced pressure conditions ~30 Pa) was used to produce crystalline TiO2 particles, which had a high surface area (**Figure 8E**) and good photocatalytic properties [120].

area deposition for various applications [108].

**3.3. Titanium-based applications**

PECVD is a technique which operates at lower temperatures, suitable for the deposition of organic precursors, unlike other processes. Synthesized thin films are biocompatible, adherent, and flexible for applications such as implants [2, 65]. PECVD offers several advantages in case of deposition of ultrathin layers of macromolecular materials such as alcohols, carboxylic acids, acid chlorides, simple amines, anhydrides, and ethylene glycols [124–130]. The chemical reactivity of the deposited layer is generally based on the functional group density and various degrees of cross-link density. The chemical reactivity of these functional layers also determines shelf life of the resultant device. These plasma-polymerized thin films are utilized in various biological applications such as antimicrobial coatings, biomaterials, tissue engineering, implant coatings, and bioelectronics [2, 66, 67, 131].

### *3.4.1. Antimicrobial and antifouling*

Antimicrobial surfaces are in demand owing to increase in bacterial colonization issues in medical devices, implants, and healthcare products. The lifetime of implanted prosthesis is brought down drastically due to bacterial infections, especially in immunocompromised patients. The use of antimicrobial surfaces has gained importance lately, because they do not lead to resistance (common example *Staphylococcus aureus* which is multidrug resistant). The PECVD process was an effective way of designing these surfaces because it can be used efficiently to control the coating properties and content of nanoparticles embedded. Biocidal activity of silver nanoclusters embedded in a PECVD deposited film is believed to be a result of effective progressive release of Ag ions from the nanoparticles in to the media surrounding it [119, 131–133]. In a recent study, polydimethylsiloxane (PDMS) surfaces were coated with PFM (pentafluorophenyl methacrylate) using PECVD with a continuous RF power (15 W), and pulsed plasma polymerization (DC of 0.5) for 3–5 min to produce a flexible and highly reactive surface with ester groups [134]. The incubation of these surfaces in an amine sugar (glucosa‐ mine) produced a controlled reductive surface capable of reducing silver salts to form microand nanostructured silver coatings. Gilabert et al. have demonstrated the antibacterial activity of such coatings due to their hydrophobic behavior which eliminated bacterial adhesion. The initial rapid release followed by a sustained release proved effective against *P. aeruginosa* and *S. aureus* bacterial adhesion but had no cytotoxic effects on the mammalian cell line (COS7 cells) when tested (**Figure 9**) [132]. Silver coatings are also commonly used on implants to avoid bacterial infections and biofilm formation [134].

**Figure 9.** (A) Schematic representation of synthesis of highly hydrophobic silver nanoparticle surface and silver ion release (reprinted with permission from [132], © 2015 ACS). (B) Silver nanoparticles on silicon oxide layer, a product of APCVD plasma jet (SEM micrograph) (reprinted with permission from [131], © 2012 Wiley). (C) SEM images depicting the surface morphology of silver modified PDMS samples. (D) Confocal images showing the bacterial viability on non‐ coated (PDMS) and coated (silver) surfaces. (E) Antibacterial activity of coated (silver) and noncoated (PDMS) surfaces on two different strains (*P. aeruginosa* and *S. aureus*) and cell viability of COS-7 cell-line (C–E: reprinted with permis‐ sion from [132], © 2015 ACS).

### *3.4.2. Wettability and cell adhesion*

*3.4.1. Antimicrobial and antifouling*

bacterial infections and biofilm formation [134].

sion from [132], © 2015 ACS).

Antimicrobial surfaces are in demand owing to increase in bacterial colonization issues in medical devices, implants, and healthcare products. The lifetime of implanted prosthesis is brought down drastically due to bacterial infections, especially in immunocompromised patients. The use of antimicrobial surfaces has gained importance lately, because they do not lead to resistance (common example *Staphylococcus aureus* which is multidrug resistant). The PECVD process was an effective way of designing these surfaces because it can be used efficiently to control the coating properties and content of nanoparticles embedded. Biocidal activity of silver nanoclusters embedded in a PECVD deposited film is believed to be a result of effective progressive release of Ag ions from the nanoparticles in to the media surrounding it [119, 131–133]. In a recent study, polydimethylsiloxane (PDMS) surfaces were coated with PFM (pentafluorophenyl methacrylate) using PECVD with a continuous RF power (15 W), and pulsed plasma polymerization (DC of 0.5) for 3–5 min to produce a flexible and highly reactive surface with ester groups [134]. The incubation of these surfaces in an amine sugar (glucosa‐ mine) produced a controlled reductive surface capable of reducing silver salts to form microand nanostructured silver coatings. Gilabert et al. have demonstrated the antibacterial activity of such coatings due to their hydrophobic behavior which eliminated bacterial adhesion. The initial rapid release followed by a sustained release proved effective against *P. aeruginosa* and *S. aureus* bacterial adhesion but had no cytotoxic effects on the mammalian cell line (COS7 cells) when tested (**Figure 9**) [132]. Silver coatings are also commonly used on implants to avoid

266 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

**Figure 9.** (A) Schematic representation of synthesis of highly hydrophobic silver nanoparticle surface and silver ion release (reprinted with permission from [132], © 2015 ACS). (B) Silver nanoparticles on silicon oxide layer, a product of APCVD plasma jet (SEM micrograph) (reprinted with permission from [131], © 2012 Wiley). (C) SEM images depicting the surface morphology of silver modified PDMS samples. (D) Confocal images showing the bacterial viability on non‐ coated (PDMS) and coated (silver) surfaces. (E) Antibacterial activity of coated (silver) and noncoated (PDMS) surfaces on two different strains (*P. aeruginosa* and *S. aureus*) and cell viability of COS-7 cell-line (C–E: reprinted with permis‐ In biological systems, the hydrophobic and hydration forces play an important role in the mediation of solute, for instance, in the adsorption and adhesion of proteins and cells (**Figure 9B**) [135, 136]. Production of bioabsorbable materials is one main area where PECVD has been utilized [137–139]. The design and synthesis of various functional hybrid film systems for SiO*x* film coatings with good mechanical properties at high deposition rates were reported. These SiO*x* surfaces are known for their hydrophilic and smooth properties that can be modified using plasma treatment [139, 140]. The deposition of SiO*x* films on poly (lactic acid) (PLA) substrate was synthesized by Chaiwong et al. using octamethylcyclotetrasiloxane (OMCTS: Si4O4C8H20) as the precursor and oxygen as carrier gas. Oxygen plasma treatment after deposition enhanced the wettability of the surface leading to an increased BSA protein adsorption and cell adhesion (preosteoblasts, fibroblasts) [139].

### *3.4.3. Biosensors*

Disease diagnosis using biosensors requires functionalization of sensors' surfaces with specific antigens to detect required biomolecules. Typically, biosensors use functionally modified polymer surfaces to immobilize proteins. Use of chemical synthesis methods is time consum‐ ing, complicated, and quality of film is difficult to control during the process. PECVD, being a single-step process, provides a venue to overcome some of these obstacles. Multiple studies have used PECVD for the deposition of various end-groups such as amines, carboxylic, mercapto, epoxy, and polyethylene glycol to address the problem of nonspecific binding of single-stranded DNA or other proteins of interest. Parameters such as higher RF power lead to increase in thickness of the film due to increased rate of deposition [140, 141]. Monomers such as amino silanes, allylamine, and allyl organophosphates have been commonly used for the deposition of thin films with reactive functional groups [58, 127].

### *3.4.4. Bioimplants*

Hydroxyapatite (HA) is a major mineral component of natural bone that can readily integrate with human tissues within just a short period after implantation. Therefore, it is widely used in bone and dental implants. Although it has multiple advantages, it has poor mechanical strength and load bearing capacity. Hence, it is used predominantly as a coating on metallic implants [142–144]. Some studies have shown that the inclusion of HA along with titanium in a plasma sputtering process promotes the formation of calcium titanate, which largely contributes to the adhesive strength between the alloy and the film. The structure and morphology of the coatings were controlled using plasma parameters such as DC substrate bias. It was observed that at higher DC (negative) bias voltages, the calcium oxide (CaO), and calcium titanate (CaTiO3) phase concentrations in the bioceramic coating increased. This means that the change (increase) in DC bias in turn increased the ion flux (CaO+ ) leading to the increased presence of CaO and CaTiO3 phases, and hence the biocompatibility (positive biomimetic response). Plasma-assisted RF magnetron sputtering deposition is also an efficient way of coating TiAlV orthopedic alloys with HA or any calcium phosphate-based mineral (**Figure 10A**) [144].

**Figure 10.** (A) SEM image of RF plasma deposited (RF power = 500 W). Calcium phosphate bioceramic film (1.2 µm) on an alloy (Ti6Al4V) as a bioimplant coating (reprinted with permission from [144], © 2007 Wiley). (C) Activity of im‐ mobilized GO*x* in polyacetylene coating on plastic foil (AP-PECVD) after different times of deposition (reprinted with permission from [40], © 2011 Wiley). (B) SEM images of cell and PECVD deposited material interactions. Cell spread‐ ing based on surface roughness confirms the dependence of the cell behavior on surface topography (reprinted with permission from [145], © ACS).

### *3.4.5. Food preservation and bioactive coatings*

Spoilage is a major problem in the food industry and various studies are trying to address this issue. Quintieri et al. used PECVD to coat bovine lactoferrin (BLF) and lactoferricin B (LfcinB) against spoilage by *Pseudomonas* strains. In one study, BLF and LfcinB were evaluated for their antimicrobial activity, when immobilized on plasma deposited films of ethylene/acrylic acid (pdEthAA) [146]. RF plasma with different input power, two fragmentation levels in plasma feed, and different –COOH group densities were used during the coatings, to vary the amount of LfcinB and BLF immobilization on the films [146]. Heyse et al. have reported the use of APCVD to form bioactive coatings (using enzymes, proteins, and other organic precursors) [40]. Though the temperature range during deposition did not exceed 32°C, the radical chemistry due to the plasma had harsh effects on the enzymes. There was considerable effort to minimize the effect, by atomizing the enzyme solution and encompassing it in water shell ("nano shuttle" formation) which subsequently protected the enzyme from the harsh plasma conditions. These enzymes were further trapped in the growing polymer network of organic polymers acetylene or pyrrole giving rise to bioactive films. The films formed using enzymes such as glucose oxide (GO*x*) (**Figure 10C**) and lipase were successfully used in sensor and catalysis applications. They tested it on other sensitive and hyperthermophilic proteins such as allophycocyanin and Apase [40].

The PECVD technique, as discussed has been used for a broad range of applications from photovoltaics, biological, to semiconductors. Most recently, this technique has been used in several instances where barrier properties were required for the encapsulation of organic lightemitting devices (OLED) by silicon oxide (SiOx) and nickel oxide (NiOx) PECVD-deposited films, organic electronic fabrication, and anticorrosive coatings. Room temperature deposition of polymer-like carbon films (PLC) for gate dielectric applications in organic thin film tran‐ sistors has also reported [148]. Improvements in the encapsulation process of various materials utilizing PECVD can lead to potential advancements in the research area of PECVD-based encapsulation. [147]. Other areas such as anticorrosion coating of metallic substrates such as copper via plasma polymeric films of dicyclopentadiene (DCPD) films is an area of future research [149]. Tailoring the surface properties of organic materials to become suitable for biomedical applications is another area in which PECVD holds promise for future applications [150].

### **Acknowledgements**

The authors would like to acknowledge the Provost's office at the University of Dartmouth and AFOSR.

### **Author details**

**Figure 10.** (A) SEM image of RF plasma deposited (RF power = 500 W). Calcium phosphate bioceramic film (1.2 µm) on an alloy (Ti6Al4V) as a bioimplant coating (reprinted with permission from [144], © 2007 Wiley). (C) Activity of im‐ mobilized GO*x* in polyacetylene coating on plastic foil (AP-PECVD) after different times of deposition (reprinted with permission from [40], © 2011 Wiley). (B) SEM images of cell and PECVD deposited material interactions. Cell spread‐ ing based on surface roughness confirms the dependence of the cell behavior on surface topography (reprinted with

268 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

Spoilage is a major problem in the food industry and various studies are trying to address this issue. Quintieri et al. used PECVD to coat bovine lactoferrin (BLF) and lactoferricin B (LfcinB) against spoilage by *Pseudomonas* strains. In one study, BLF and LfcinB were evaluated for their antimicrobial activity, when immobilized on plasma deposited films of ethylene/acrylic acid (pdEthAA) [146]. RF plasma with different input power, two fragmentation levels in plasma feed, and different –COOH group densities were used during the coatings, to vary the amount of LfcinB and BLF immobilization on the films [146]. Heyse et al. have reported the use of APCVD to form bioactive coatings (using enzymes, proteins, and other organic precursors) [40]. Though the temperature range during deposition did not exceed 32°C, the radical chemistry due to the plasma had harsh effects on the enzymes. There was considerable effort to minimize the effect, by atomizing the enzyme solution and encompassing it in water shell ("nano shuttle" formation) which subsequently protected the enzyme from the harsh plasma conditions. These enzymes were further trapped in the growing polymer network of organic polymers acetylene or pyrrole giving rise to bioactive films. The films formed using enzymes such as glucose oxide (GO*x*) (**Figure 10C**) and lipase were successfully used in sensor and catalysis applications. They tested it on other sensitive and hyperthermophilic proteins such

The PECVD technique, as discussed has been used for a broad range of applications from photovoltaics, biological, to semiconductors. Most recently, this technique has been used in several instances where barrier properties were required for the encapsulation of organic lightemitting devices (OLED) by silicon oxide (SiOx) and nickel oxide (NiOx) PECVD-deposited films, organic electronic fabrication, and anticorrosive coatings. Room temperature deposition

permission from [145], © ACS).

*3.4.5. Food preservation and bioactive coatings*

as allophycocyanin and Apase [40].

Yasaman Hamedani1 , Prathyushakrishna Macha1 , Timothy J. Bunning2 , Rajesh R. Naik2 and Milana C. Vasudev1\*

\*Address all correspondence to: mvasudev@umassd.edu

1 Department of Bioengineering, University of Massachusetts at Dartmouth, Dartmouth, MA, USA

2 Air Force Research Laboratory, Wright Patterson AFB, Dayton, OH, USA

### **References**


[20] Vasudev MC, Koerner H, Singh KM, Partlow BP, Kaplan DL, Gazit E, et al. Vertically aligned peptide nanostructures using plasma-enhanced chemical vapor deposition. Biomacromolecules. 2014;15(2):533–40.

[4] Eichfeld SM, Hossain L, Lin Y-C, Piasecki AF, Kupp B, Birdwell AG, et al. Highly scalable, atomically thin WSe2 grown via metal–organic chemical vapor deposition.

[5] Pierson HO. Handbook of Chemical Vapor Deposition (CVD): Principles, Technology

[6] Stoffel A, Kovács A, Kronast W, Müller B. LPCVD against PECVD for micromechanical applications. Journal of Micromechanics and Microengineering. 1996;6(1):1.

[7] Li X-L, Ge J-P, Li Y-D. Atmospheric pressure chemical vapor deposition: an alternative route to large-scale MoS2 and WS2 inorganic fullerene-like nanostructures and

[8] Poole KM. Electrode contamination in electron optical systems. Proceedings of the

[9] Ennos AE. The source of electron-induced contamination in kinetic vacuum systems.

[10] Alt LL, Ing SW, Laendle KW. Low‐temperature deposition of silicon oxide films.

[11] Christy RW. Formation of thin polymer films by electron bombardment. Journal of

[12] Baker AG, Morris WC. Deposition of metallic films by electron impact decomposition of organometallic vapors. Review of Scientific Instruments. 1961;32(4):458.

[13] Christy RW. Conducting thin films formed by electron bombardment of substrate.

[14] Ing SW, Davern W. Use of low‐temperature deposited silicon dioxide films as diffusion masks in GaAs. Journal of the Electrochemical Society. 1964;111(1):120–2.

[15] Ing SW, Davern W. Glow discharge formation of silicon oxide and the deposition of silicon oxide thin film capacitors by glow discharge techniques. Journal of the Electro‐

[16] Wertheimer MR. Plasma processing and polymers: a personal perspective. Plasma

[17] Reinberg AR. Extended Abtracts. The Electrochemical Society Meeting, San Francicso,

[19] Pierson HO. Handbook of Chemical Vapor Deposition (CVD). 2nd Ed. Norwich, NY:

nanoflowers. Chemistry – A European Journal. 2004;10(23):6163–71.

and Applications. Park Ridge, NJ: Noyes Publication; 1992.

270 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

ACS Nano. 2015;9(2):2080–7.

Physical Society Section B. 1953;66(7):542.

Applied Physics. 1960;31(9):1680–3.

chemical Society. 1965;112(3):284–8.

CA, May 12–17, 1974.

Journal of Applied Physics. 1962;33(5):1884–8.

Chemistry and Plasma Processing. 2013;34(3):363–76.

[18] Reinberg AR. Inventor, Patent No. US3757733 A, 1973.

William Andrew Publishing; 1999. p. iii.

British Journal of Applied Physics. 1954;5(1):27–31.

Journal of the Electrochemical Society. 1963;110(5):465.


[49] Tibbitt JM, Bell AT, Shen M. Dielectric relaxations in plasma-polymerized hydrocar‐ bons and fluorocarbons. Journal of Macromolecular Science: Chemistry. 1976;A10:519– 33.

[36] Peri B, Borah B, Dash RK. Effect of RF power and gas flow ratio on the growth and morphology of the PECVD SiC thin film s for MEMS applications. Bulletin of Materials

272 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

[37] Dams R, Vangeneugden D, Vanderzande D. Plasma deposition of thiophene deriva‐ tives under atmospheric pressure. Chemical Vapor Deposition. 2006;12(12):719–27.

[38] Schutze A, Jeong JY, Babayan SE, Park J, Selwyn GS, Hicks RF. The atmosphericpressure plasma jet: a review and comparison to other plasma sources. IEEE Transac‐

[39] Groenewoud LMH, Engbers GHM, White R, Feijen J. On the iodine doping process of plasma polymerised thiophene layers. Synthetic Metals. 2001;125(3):429–40.

[40] Heyse P, Van Hoeck A, Roeffaers MB, Raffin JP, Steinbüchel A, Stöveken T, et al. Exploration of Atmospheric pressure plasma nanofilm technology for straightforward bio‐active coating deposition: enzymes, plasmas and polymers, an elegant synergy.

[41] Da Ponte G, Sardella E, Fanelli F, Paulussen S, Favia P. Atmospheric pressure plasma deposition of poly lactic acid-like coatings with embedded elastin. Plasma Processes

[42] Nozaki T, Ohnishi K, Okazaki K, Kortshagen U. Fabrication of vertically aligned singlewalled carbon nanotubes in atmospheric pressure non-thermal plasma CVD. Carbon.

[43] Barankin MD, Gonzalez Ii E, Ladwig AM, Hicks RF. Plasma-enhanced chemical vapor deposition of zinc oxide at atmospheric pressure and low temperature. Solar Energy

[44] Babayan SE, Jeong JY, Tu VJ, Park J, Selwyn GS, Hicks RF. Deposition of silicon dioxide films with an atmospheric-pressure plasma jet. Plasma Sources Science and Technolo‐

[45] Hiramatsu M, Hori M. Aligned growth of single-walled and double-walled carbon nanotube films by control of catalyst preparation. Japanese Journal of Applied Physics.

[46] Shiji K, Hiramatsu M, Enomoto A, Nakamura M, Amano H, Hori M. Vertical growth of carbon nanowalls using rf plasma-enhanced chemical vapor deposition. Diamond

[47] Kogelschatz U. Dielectric-barrier discharges: their history, discharge physics, and industrial applications. Plasma Chemistry and Plasma Processing. 2003;23(1):1–46.

[48] Jiang H, Johnson WE, Grant JT, Eyink K, Johnson EM, Tomlin DW, et al. Plasma polymerized multi-layered photonic films. Chemistry of Materials. 2003;15(1):340–7.

Science. 2015;38(4):1105–12.

tions on Plasma Science. 1998;26(6):1685–94.

Plasma Processes and Polymers. 2011;8(10):965–74.

Materials and Solar Cells. 2007;91(10):924–30.

and Related Materials. 2005;14(3–7):831–4.

and Polymers. 2014;11(4):345–52.

2007;45(2):364–74.

gy. 1998;7(3):286.

2011;46(2):12–16.


[77] Seo HJ, Nam S-H, Kim S, Boo J-H. Organic and organic–inorganic hybrid polymer thin films deposited by PECVD using TEOS and cyclohexene for ULSI interlayer-dielectric application. Applied Surface Science. 2015;354:134–8.

[63] Binetti S, Acciarri M, Bollani M, Fumagalli L, Von Känel H, Pizzini S. Nanocrystalline silicon films grown by low energy plasma enhanced chemical vapor deposition for

[64] O'Mara WC. Liquid Crystal Flat Panel Displays: Manufacturing Science & Technology.

[65] Förch R, Chifen AN, Bousquet A, Khor HL, Jungblut M, Chu LQ, et al. Recent and expected roles of plasma‐polymerized films for biomedical applications. Chemical

[66] Pryce Lewis HG, Edell DJ, Gleason KK. Pulsed-PECVD films from hexamethylcyclo‐ trisiloxane for use as insulating biomaterials. Chemistry of Materials. 2000;12(11):3488–

[67] Tang L, Wu Y, Timmons RB. Fibrinogen adsorption and host tissue responses to plasma functionalized surfaces. Journal of Biomedical Materials Research. 1998;42(1):156–63.

[68] Sherman S, Wagner S, Mucha J, Gottscho RA. Substrate effect on plasma‐enhanced chemical vapor deposited silicon nitride. Journal of the Electrochemical Society.

[69] Martins R, Aguas H, Cabrita A, Tonello P, Silva V, Ferreira I, et al. New nanostructured silicon films grown by PECVD technique under controlled powder formation condi‐

[70] Sriraman S, Agarwal S, Aydil ES, Maroudas D. Mechanism of hydrogen-induced

[71] Marra DC, Edelberg EA, Naone RL, Aydil ES. Silicon hydride composition of plasmadeposited hydrogenated amorphous and nanocrystalline silicon films and surfaces.

[72] Kessels W, Smets A, Marra D, Aydil E, Schram D, Van De Sanden M. On the growth

[73] Martin-Palma RJ, Gago R, Torres-Costa V, Fernandez-Hidalgo P, Kreissig U, Martinez Duart JM. Optical and compositional analysis of functional SiOxCy:H coatings on

[74] Service RF. Patterning electronics on the cheap. Science (Washington, DC).

[75] Kim KN, Lee SM, Mishra A, Yeom GY. Atmospheric pressure plasmas for surface modification of flexible and printed electronic devices: a review. Thin Solid Films.

[76] Fanelli F. Thin film deposition and surface modification with atmospheric pressure dielectric barrier discharges. Surface and Coatings Technology. 2010;205(5):1536–43.

crystallization of amorphous silicon. Nature. 2002;418(6893):62–5.

Journal of Vacuum Science & Technology. 1998:3199–210.

mechanism of a-Si: H. Thin Solid Films. 2001;383(1):154–60.

optoelectronic applications. Thin Solid Films. 2005;487(1):19–25.

274 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

New York: Chapman & Hall; 1993.

Vapor Deposition. 2007;13(6–7):280–94.

94.

1997;144(9):3198–204.

tions. Solar Energy. 2001;69:263–9.

polymers. Thin Solid Films. 2006:2493–6.

1997;278(5337):383–4.

2016;598:315–34.


[103] Kenyon A, Trwoga P, Federighi M, Pitt C. Optical properties of PECVD erbium-doped silicon-rich silica: evidence for energy transfer between silicon microclusters and erbium ions. Journal of Physics: Condensed Matter. 1994;6(21):L319.

[90] Juneja S, Sudhakar S, Gope J, Lodhi K, Sharma M. Highly conductive boron doped micro/nanocrystalline silicon thin films deposited by VHF-PECVD for solar cell

[91] Engelhardt J, Frey A, Mahlstaedt L, Gloger S, Hahn G, Terheiden B. Boron emitters from doped PECVD layers for n-type crystalline silicon solar cells with LCO. Energy

[92] Matsuda A. Formation kinetics and control of microcrystallite in µc-Si: H from glow

[93] Norris KJ, Garrett MP, Zhang J, Coleman E, Tompa GS, Kobayashi NP. Silicon nanowire networks for multi-stage thermoelectric modules. Energy Conversion and Manage‐

[94] Rath J, Brinza M, Liu Y, Borreman A, Schropp R. Fabrication of thin film silicon solar cells on plastic substrate by very high frequency PECVD. Solar Energy Materials and

[95] Street R. Large area electronics, applications and requirements. Physica Status Solidi

[96] Yang P, Wirnsberger G, Huang HC, Cordero SR, McGehee MD, Scott B, et al. Mirrorless lasing from mesostructured waveguides patterned by soft lithography. Science.

[97] Yang C-C, Lin Y-H, Lin G-R, editors. Nonstoichiometric Si1-xGex based tunable saturable absorber for mode-locked erbium-doped fiber laser. In CLEO. OSA Technical

[98] Loerke J, Marlow F. Laser emission from dye-doped mesoporous silica fibers.

[99] Burt MC, Dave BC. An optical temperature sensing system based on encapsulation of a dye molecule in organosilica sol–gels. Sensors and Actuators B: Chemical. 2005;107(2):

[100] Barranco A, Aparicio F, Yanguas‐Gil A, Groening P, Cotrino J, González‐Elipe AR. Optically active thin films deposited by plasma polymerization of dye molecules.

[101] Limb SJ, Lau KK, Edell DJ, Gleason EF, Gleason KK. Molecular design of fluorocarbon film architecture by pulsed plasma enhanced and pyrolytic chemical vapor deposition.

[102] Chen K-S, Yang M-R, Hsu ST. Fabrication and characterization of fluorine-containing films using plasma polymerization of octafluorotoluene. Materials Chemistry and

Digest [online]. Optical Society of America; 2015. Paper JW2A.70.

Chemical Vapor Deposition. 2007;13(6–7):319–25.

Plasmas and Polymers. 1999;4(1):21–32.

Physics. 1999;61(3):214–8.

applications. Journal of Alloys and Compounds. 2015;643:94–9.

276 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

discharge plasma. Journal of Non-Crystalline Solids. 1983;59:767–74.

Procedia. 2014;55:235–40.

ment. 2015;96:100–4.

Solar Cells. 2010;94(9):1534–41.

(a). 1998;166(2):695–705.

2000;287(5452):465–7.

2002;14(23):1745–9.

552–6.


[129] Schiller S, Hu J, Jenkins A, Timmons R, Sanchez-Estrada F, Knoll W, et al. Chemical structure and properties of plasma-polymerized maleic anhydride films. Chemistry of Materials. 2002;14(1):235–42.

[116] Bo Z, Mao S, Han ZJ, Cen K, Chen J, Ostrikov KK. Emerging energy and environmental applications of vertically-oriented graphenes. Chemical Society Reviews. 2015;44(8):

278 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

[117] Jacob MV, Rawat RS, Ouyang B, Bazaka K, Kumar DS, Taguchi D, et al. Catalyst-free plasma enhanced growth of graphene from sustainable sources. Nano Letters.

[118] Szymanowski H, Sobczyk A, Gazicki-Lipman M, Jakubowski W, Klimek L. Plasma enhanced CVD deposition of titanium oxide for biomedical applications. Surface and

[119] Hájková P, Špatenka P, Krumeich J, Exnar P, Kolouch A, Matoušek J, et al. Antibacterial effect of silver modified TiO2/PECVD films. The European Physical Journal D.

[120] Ayllon J, Figueras A, Garelik S, Spirkova L, Durand J, Cot L. Preparation of TiO2 powder using titanium tetraisopropoxide decomposition in a plasma enhanced chemical vapor deposition (PECVD) reactor. Journal of Materials Science Letters. 1999;18(16):1319–21.

[121] Williams L, Hess D. Structural properties of titanium dioxide films deposited in an rf glow discharge. Journal of Vacuum Science & Technology A. 1983;1(4):1810–9.

[122] Maeda M, Watanabe T. Visible light photocatalysis of nitrogen-doped titanium oxide films prepared by plasma-enhanced chemical vapor deposition. Journal of the Electro‐

[123] Rosler RS, Engle GM. Plasma‐enhanced CVD of titanium silicide. Journal of Vacuum

[124] Candan S, Beck AJ, O'Toole L, Short RD. Effects of "processing parameters" in plasma deposition: acrylic acid revisited. Journal of Vacuum Science & Technology A.

[125] Beyer D, Knoll W, Ringsdorf H, Wang JH, Timmons RB, Sluka P. Reduced protein adsorption on plastics via direct plasma deposition of triethylene glycol monoallyl

[126] Tarducci C, Schofield W, Badyal J, Brewer S, Willis C. Synthesis of cross-linked ethylene glycol dimethacrylate and cyclic methacrylic anhydride polymer structures by pulsed

[127] Choukourov A, Biederman H, Slavinska D, Hanley L, Grinevich A, Boldyryeva H, et al. Mechanistic studies of plasma polymerization of allylamine. The Journal of Physical

[128] van Os M, Menges B, Timmons R, Knoll W, Foerch R, editors. Variations in the film chemistry of pulsed rf plasma deposited allylamine. In 13th International Symposium

on Plasma Chemistry. Beijing, China: Peking University Press; 1997.

ether. Journal of Biomedical Materials Research. 1997;36(2):181–9.

plasma deposition. Macromolecules. 2002;35(23):8724–7.

2108–21.

2015;15(9):5702–8.

2009;54(2):189–93.

1998;16(3):1702–9.

Coatings Technology. 2005;200(1):1036–40.

chemical Society. 2006;153(3):C186–C9.

Science & Technology B. 1984;2(4):733–7.

Chemistry B. 2005;109(48):23086–95.


[141] Gandhiraman RP, Gubala V, O'Mahony CC, Cummins T, Raj J, Eltayeb A, et al. PECVD coatings for functionalization of point-of-care biosensor surfaces. Vacuum. 2012;86(5):

[142] De Groot K, Geesink R, Klein C, Serekian P. Plasma sprayed coatings of hydroxylapa‐

[143] Friedman RJ, Black J, Galante JO, Jacobs JJ, Skinner HB. Current concepts in orthopaedic biomaterials and implant fixation. Journal of Bone and Joint Surgery: American.

[144] Long J, Sim L, Xu S, Ostrikov K. Reactive plasma‐aided RF sputtering deposition of hydroxyapatite bio‐implant coatings. Chemical Vapor Deposition. 2007;13(6–7):299–

[145] Intranuovo F, Favia P, Sardella E, Ingrosso C, Nardulli M, d'Agostino R, et al. Osteo‐ blast-like cell behavior on plasma deposited micro/nanopatterned coatings. Biomacro‐

[146] Quintieri L, Pistillo B, Caputo L, Favia P, Baruzzi F. Bovine lactoferrin and lactoferricin on plasma-deposited coating against spoilage Pseudomonas spp. Innovative Food

[147] Yu D, Yang Y-Q, Chen Z, Tao Y, Liu Y-F. Recent progress on thin-film encapsulation technologies for organic electronic devices. Optics Communications. 2016;362:43–9.

[148] Feng L, Anguita JV, Tang W, Zhao J, Guo X, Silva SRP. Room temperature grown high quality polymer-like carbon gate dielectric for organic thin-film transistors. Advanced

[149] Jeong D-C, Javid A, Wen L, Choi EJ, Park SY, Kim YH, et al. Low-temperature plasma polymerization of dicyclopentadiene for anti-corrosion properties. Polymer.

[150] Agnès T-M, Satoshi I, Bertrand-David S, Grant AC, Takuya T, Hiroyuki F, et al. Review on thin-film transistor technology, its applications, and possible new applications to

biological cells. Japanese Journal of Applied Physics. 2016;55(4S):04EA8.

tite. Journal of Biomedical Materials Research. 1987;21(12):1375–81.

280 Chemical Vapor Deposition - Recent Advances and Applications in Optical, Solar Cells and Solid State Devices

547–55.

306.

1993;75(7):1086–109.

molecules. 2010;12(2):380–7.

Electronic Materials. 2016;2(3):1–6.

2016;92:133–9.

Science & Emerging Technologies. 2013;20:215–22.

### *Edited by Sudheer Neralla*

This book provides an overview of chemical vapor deposition (CVD) methods and recent advances in developing novel materials for application in various fields. CVD has now evolved into the most widely used technique for growth of thin films in electronics industry. Several books on CVD methods have emerged in the past, and thus the scope of this book goes beyond providing fundamentals of the CVD process. Some of the chapters included highlight current limitations in the CVD methods and offer alternatives in developing coatings through overcoming these limitations.

Photo by nikkytok / Can Stock

Chemical Vapor Deposition - Recent Advances and Applications in

Optical, Solar Cells and Solid State Devices

Chemical Vapor Deposition

Recent Advances and Applications in

Optical, Solar Cells and Solid State Devices

*Edited by Sudheer Neralla*