With Cement. *Geotechnical Engineering Journal*. Vol.29, No.1, pp.29-44. **Part 4**

## **Nanostructured Materials for Electronic Industry**

460 Scanning Electron Microscopy

Yin, J.H. & Lai, C.K. (1998). Strength and Stiffness of Hong Kong Marine Deposit Mixed

**23** 

*Saudi Arabia* 

A. Alyamani1 and O. M. Lemine2

*1National Nanotechnology Research Centre, KACST, Riyadh, 2Physics Department, College of Sciences, Imam University Riyadh,* 

λ

= = (1)

. This wavelength can be

**FE-SEM Characterization of Some Nanomaterial** 

In 1931 Max Knoll and Ernst Ruska at the university of Berlin built the first electron microscope that use accelerated electrons as a source instead of light source. However, the first scanning electron microscope (SEM) was built in 1938 due to the difficulties of scanning the electrons through the sample. Electron microscope is working exactly the same as the

Since the invention of the electron microscope, it became one of the most useful instruments that has an impact in understanding scientific phenomena in different fields, such as physics, nanotechnology, medicine, chemistry biology..etc. Electron microscope has the ability to resolve objects ranging from part of nano-metre to micro-metre compared to light

In the first part of the chapter, we will describe some of the basics of electronic microscope and its applications. The second part will be dedicated to the results obtained mainly by SEM.

The principal of electron microscope is the same as a light microscope but instead of using visible light it use very energetic electrons as a source. However, the resolution of the optical microscope is limited by its wavelength compared to accelerated electrons which have very

changed according to the applied high voltage. Hence, according to Rayleigh's criterion the

*h h p mv*

where *h* = 6x10-34 J s is the Planck constant, *m* and *v* are the mass and velocity of the electron respectively. Since the electron can reach nearly the velocity of light *c* then we can use the

λ

relativistic equations. In this case the electron mass is changing according to:

of an electron is related to the momentum *p=mv* of the electron by: [2]

optical microscope expects it use a focused accelerated electron beam [1].

microscope that has a magnificationin the range of 1000 and resolution of 200 nm.

short wavelength.This is what makes it possible to see very small features.

In electron microscopes, electrons have very small wavelength

**1. Introduction** 

**2. Electron microscopy** 

λ

wavelength

**2.1 Fundamental principles of electron microscopy** 

## **FE-SEM Characterization of Some Nanomaterial**

## A. Alyamani1 and O. M. Lemine2

*1National Nanotechnology Research Centre, KACST, Riyadh, 2Physics Department, College of Sciences, Imam University Riyadh, Saudi Arabia* 

## **1. Introduction**

In 1931 Max Knoll and Ernst Ruska at the university of Berlin built the first electron microscope that use accelerated electrons as a source instead of light source. However, the first scanning electron microscope (SEM) was built in 1938 due to the difficulties of scanning the electrons through the sample. Electron microscope is working exactly the same as the optical microscope expects it use a focused accelerated electron beam [1].

Since the invention of the electron microscope, it became one of the most useful instruments that has an impact in understanding scientific phenomena in different fields, such as physics, nanotechnology, medicine, chemistry biology..etc. Electron microscope has the ability to resolve objects ranging from part of nano-metre to micro-metre compared to light microscope that has a magnificationin the range of 1000 and resolution of 200 nm.

In the first part of the chapter, we will describe some of the basics of electronic microscope and its applications. The second part will be dedicated to the results obtained mainly by SEM.

## **2. Electron microscopy**

#### **2.1 Fundamental principles of electron microscopy**

The principal of electron microscope is the same as a light microscope but instead of using visible light it use very energetic electrons as a source. However, the resolution of the optical microscope is limited by its wavelength compared to accelerated electrons which have very short wavelength.This is what makes it possible to see very small features.

In electron microscopes, electrons have very small wavelength λ . This wavelength can be changed according to the applied high voltage. Hence, according to Rayleigh's criterion the wavelength λof an electron is related to the momentum *p=mv* of the electron by: [2]

$$
\mathcal{X} = \frac{h}{p} = \frac{h}{mv} \tag{1}
$$

where *h* = 6x10-34 J s is the Planck constant, *m* and *v* are the mass and velocity of the electron respectively. Since the electron can reach nearly the velocity of light *c* then we can use the relativistic equations. In this case the electron mass is changing according to:

FE-SEM Characterization of Some Nanomaterial 465

thin specimen. Another scattering mechanism called elastic scattered where electrons don't loss their energy these scattered electrons can be used to get information about orientation

When the atoms bombarded with incident electrons, electrons will released from these atoms and this will leave the atom in the excited state. In order for the atom to return to the ground state, it needs to release the excess energy Auger electrons, X-Rays,and cathodoluminescence are three ways of relaxation. The x-ray is used to identify the elements and their concentrations in the specimen by using a technique called Energy –dispersive Xray analysis (EDX) technique. Chemical analysis can be done by using Auger electrons.

There are two types of electron microscopes. Scanning Electron Microscopes (SEM), and Transmission Electron Microscope (TEM), these types of microscopes detect electrons that emitted from the surface of the sample. The accelerated voltage is ranging from 10kV to 40kV for the SEM. The thickness of the specimen in this case is not important. In addition, the samples to be tested have to be electrically conductive; otherwise they would be overcharged with electrons. However, they can be coated with a conductive layer of metal

In TEM the transmitted electrons are detected, and in this case the specimen thickness is important and typically should not exceed 150 nm. The accelerated voltage in this case

Since the electrons are easily scattered in air all electron microscopes should operate under a

Electron Gun which is used to provide and supply electrons with the required energy.There are different types of electron gun;the old type was a bent piece of Tungsten wire with 100 micro-metresin diameter. Higher performance electron emitters consist of either single

As a materials processing technique, laser ablation was utilized for the first time in the 1960's, after the first commercial ruby laser was invented [4]. Nevertheless, as a thin film growth method it did not attract much research interest until the late 1980´s [5], when it has been used for growing high temperature superconductor films. Since then, the development of the pulsed laser deposition (PLD) technique has been more rapid and the amount of research devoted to this topic has increased dramatically [6]. The growth and quality of the resulting film will generally depend on a number of fundamental parameters, including the choice of substrate, the substrate temperature and the absolute and relative kinetic energies

All types of electron microscopes are basically consist of three basic components:

crystals of lanthanum hexaboride (LaB6) or from field emission guns.

and/or arrival rates of the various constituents within the plume.

and arrangement of atoms.

**2.2.4 Other interactions** 

or carbon.

> 100kV.

high vacuum.

**3. Experimental** 

**3.1 Pulse Laser Deposition (PLD)** 

**2.3 Types of electron microscopes** 

$$
\Delta m = \frac{m\_e}{\sqrt{1 - \left(v/c\right)^2}}\tag{2}
$$

where *me* is the rest mass of the electron. The energy *eV* transmitted to an electron is giving by:

$$eV = \left(m - m\_2\right)c^2\tag{3}$$

By using equations 1,2 and 3 the electron wavelength can be written as function of accelerated voltage: [3]

$$\mathcal{A} = \sqrt{\frac{1.5}{V\left(1 + V \ast 10^{-6}\right)}} mm \tag{4}$$

for example an accelrated voltage of 10 kV will yield a wavelength of 0.0122 nm. The extremely small wavelengths make it possible to see atomic structures using accelerated electrons.

#### **2.2 Interaction of accelerated electrons with the specimen**

The electron beam interacts with the specimen reveal useful information about the sample including: its surface features, size and shape of the features, composition and crystalline structure. The interaction of the electron beam with the specimen can be in different ways:

#### **2.2.1 Secondary electrons**

If the incident electronscome close enough to the atom then these electrons will give some of their energy to the specimen electrons mainly in the K-shell. As a result, these electrons will change their path and will ionize the electrons in the specimen atoms. These ionized electrons that escape the atoms are called secondary electrons. These electrons will move to the surface of the specimen and undergoing to elastics and inelastic collision until reaching the surface. However, due to their low energy ~ 5*eV* only those electrons that are close to the surface (~ 10 *nm*) will escape the surface and then can be detected and can used for imaging the topography of the specimen.

#### **2.2.2 Backscattered electrons**

When the incident electrons hit an atom directly, then they will be reflected or backscattered. Different atomic type of atoms will result in a different rate of backscattered electrons and hence the contrast of the image will vary as the atomic number of the specimen change, usually atoms with higher atomic number will appear brighter than those have lower atomic number.

#### **2.2.3 Transmitted electrons**

If the incident electrons pass through the specimen without any interaction with their atoms, then these electrons called transmitted electrons, these electrons are used to get an image of thin specimen. Another scattering mechanism called elastic scattered where electrons don't loss their energy these scattered electrons can be used to get information about orientation and arrangement of atoms.

## **2.2.4 Other interactions**

464 Scanning Electron Microscopy

*me <sup>m</sup>*

=

by:

electrons.

accelerated voltage: [3]

**2.2.1 Secondary electrons** 

the topography of the specimen.

**2.2.2 Backscattered electrons** 

have lower atomic number.

**2.2.3 Transmitted electrons** 

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

( ) <sup>2</sup>

( ) <sup>6</sup> 1.5 1 10

for example an accelrated voltage of 10 kV will yield a wavelength of 0.0122 nm. The extremely small wavelengths make it possible to see atomic structures using accelerated

The electron beam interacts with the specimen reveal useful information about the sample including: its surface features, size and shape of the features, composition and crystalline structure. The interaction of the electron beam with the specimen can be in different ways:

If the incident electronscome close enough to the atom then these electrons will give some of their energy to the specimen electrons mainly in the K-shell. As a result, these electrons will change their path and will ionize the electrons in the specimen atoms. These ionized electrons that escape the atoms are called secondary electrons. These electrons will move to the surface of the specimen and undergoing to elastics and inelastic collision until reaching the surface. However, due to their low energy ~ 5*eV* only those electrons that are close to the surface (~ 10 *nm*) will escape the surface and then can be detected and can used for imaging

When the incident electrons hit an atom directly, then they will be reflected or backscattered. Different atomic type of atoms will result in a different rate of backscattered electrons and hence the contrast of the image will vary as the atomic number of the specimen change, usually atoms with higher atomic number will appear brighter than those

If the incident electrons pass through the specimen without any interaction with their atoms, then these electrons called transmitted electrons, these electrons are used to get an image of

*V V*

λ

**2.2 Interaction of accelerated electrons with the specimen** 

*nm*

−

where *me* is the rest mass of the electron. The energy *eV* transmitted to an electron is giving

By using equations 1,2 and 3 the electron wavelength can be written as function of

*v c*

(2)

<sup>2</sup> *eV m m c* = − (3)

<sup>−</sup> <sup>=</sup> + ∗ (4)

When the atoms bombarded with incident electrons, electrons will released from these atoms and this will leave the atom in the excited state. In order for the atom to return to the ground state, it needs to release the excess energy Auger electrons, X-Rays,and cathodoluminescence are three ways of relaxation. The x-ray is used to identify the elements and their concentrations in the specimen by using a technique called Energy –dispersive Xray analysis (EDX) technique. Chemical analysis can be done by using Auger electrons.

#### **2.3 Types of electron microscopes**

There are two types of electron microscopes. Scanning Electron Microscopes (SEM), and Transmission Electron Microscope (TEM), these types of microscopes detect electrons that emitted from the surface of the sample. The accelerated voltage is ranging from 10kV to 40kV for the SEM. The thickness of the specimen in this case is not important. In addition, the samples to be tested have to be electrically conductive; otherwise they would be overcharged with electrons. However, they can be coated with a conductive layer of metal or carbon.

In TEM the transmitted electrons are detected, and in this case the specimen thickness is important and typically should not exceed 150 nm. The accelerated voltage in this case > 100kV.

Since the electrons are easily scattered in air all electron microscopes should operate under a high vacuum.

All types of electron microscopes are basically consist of three basic components:

Electron Gun which is used to provide and supply electrons with the required energy.There are different types of electron gun;the old type was a bent piece of Tungsten wire with 100 micro-metresin diameter. Higher performance electron emitters consist of either single crystals of lanthanum hexaboride (LaB6) or from field emission guns.

## **3. Experimental**

### **3.1 Pulse Laser Deposition (PLD)**

As a materials processing technique, laser ablation was utilized for the first time in the 1960's, after the first commercial ruby laser was invented [4]. Nevertheless, as a thin film growth method it did not attract much research interest until the late 1980´s [5], when it has been used for growing high temperature superconductor films. Since then, the development of the pulsed laser deposition (PLD) technique has been more rapid and the amount of research devoted to this topic has increased dramatically [6]. The growth and quality of the resulting film will generally depend on a number of fundamental parameters, including the choice of substrate, the substrate temperature and the absolute and relative kinetic energies and/or arrival rates of the various constituents within the plume.

FE-SEM Characterization of Some Nanomaterial 467

The field emission scanning electron microscope (FE-SEM) is a type of electron microscope that images the sample surface by scanning it with a high-energy beam of electrons in a raster scan pattern. Electron emitters from field emission gun was used. These types of electron emitters can produce up to 1000x the emission of a tungsten filament. However, they required much higher vacuum conditions. After the electrons beam exit the electron gun, they then confined and focused into a thin focused, monochromatic beam using metal aperturesand magnetic lenses. Finally, Detectors of each type of electrons are placed in the

Particles morphology of our samples was investigated using Nova 200 NanoLab field

Fig. 2 shows FESEM micrographs of ZnO thin films grown on sapphire substrate by pulse laser deposition at growth temperature from 685 to 750 °C by using a ZnO powder target at high grade. The experimental parameters are summarized in table.1. It is seen that with the substrate temperature increasing the morphology of ZnO thin lms have a little difference.

The effect of the distance between target and substrate on the morphology was also studied. Fig. 3 shows the FESEM images of ZnO thin film with different distance between the target

> Oxygen pressure

> > LASER

ENERGY(mJ)

SUBSTRATE

**3.3 Filed Emission Scanning Electron Microscopy (FESEM)** 

microscopes that collect signals to produce an image of the specimen.

The thickness of lms decreases with the increase of substrate temperature.

and thin film. It is clear that the distance affect the morphology of the film.

Distance between

target and the film

750 510 37.5 mm 150 mTorr 350 **Sapphire**

700 1230 37.5 mm 150 mTorr 350 **Sapphire**

685 1115 37.5 mm 150 mTorr 350 **Sapphire**

400 - 10 mm 150 mTorr 350 **Sapphire**

400 - 23mm 150 mTorr 350 **Sapphire**

emission scanning electron microscope (FE-SEM).

**4.1 Thin film prepared by Pulse Laser Deposition (PLD)** 

**4. Results** 

TEMPERATURE(C)

THICKNESS (nm)

Table.1.Growth parameters of ZnO thin films

The PLD process is shown in figure 1:

Fig. 1. Schematic presentation of the pulsed laser deposition process a) Laser − target interaction, b) Plume expansion and c) Film deposition [6].

The growth and quality of the resulting thin film will generally depend on a number of fundamental parameters, including the choice of substrate, the substrate temperature, *TS*, distance between target-substrate, pressure and laser energy.

In our case the laser energy was 300 mJ and the time was fixed at 60 minutes. For the others parameters (substrate, substrate temperature, pressure), different values were used.

#### **3.2 Mechanical Alloying (MA)**

The ball milling constitutes new promising methods to produce nanosized particles [7,8]. It has many advantages, e.g low cost, simple operation.The ball-milling is generally used as a mechanical co-grinding of powders, Initially different in nature, up to the preparation of a new powder, homogeneous in composition. The milling is done in cylindrical containers called vials and containing balls. The nature of the milling tools can be as diverse as steel, agate, tungsten carbide… The vials are generally filled under an inert atmosphere to avoid side reactions, since the particles are fractured during the milling process and, therefore, new highly reactive surfaces can react with the surrounding gases [8].

Several terms are used to call this technique: "Mechanical Alloying" when there is a chemical reaction between different powders, "Mechanical Grinding" or "Mechanical Milling" when the only goal is to modify the texture and/or the structure of a material (no chemical reaction is involved in the process).

Two kinds of milling systems were used to prepare our nanopowders (Vibrant and planetary milling) and different milling parameters were considered (milling times, balls to powders mass ratio, size of balls and rotation speed).

## **3.3 Filed Emission Scanning Electron Microscopy (FESEM)**

The field emission scanning electron microscope (FE-SEM) is a type of electron microscope that images the sample surface by scanning it with a high-energy beam of electrons in a raster scan pattern. Electron emitters from field emission gun was used. These types of electron emitters can produce up to 1000x the emission of a tungsten filament. However, they required much higher vacuum conditions. After the electrons beam exit the electron gun, they then confined and focused into a thin focused, monochromatic beam using metal aperturesand magnetic lenses. Finally, Detectors of each type of electrons are placed in the microscopes that collect signals to produce an image of the specimen.

Particles morphology of our samples was investigated using Nova 200 NanoLab field emission scanning electron microscope (FE-SEM).

## **4. Results**

466 Scanning Electron Microscopy

Fig. 1. Schematic presentation of the pulsed laser deposition process a) Laser − target interaction, b) Plume expansion and c) Film deposition [6].

new highly reactive surfaces can react with the surrounding gases [8].

distance between target-substrate, pressure and laser energy.

**3.2 Mechanical Alloying (MA)** 

reaction is involved in the process).

powders mass ratio, size of balls and rotation speed).

The growth and quality of the resulting thin film will generally depend on a number of fundamental parameters, including the choice of substrate, the substrate temperature, *TS*,

In our case the laser energy was 300 mJ and the time was fixed at 60 minutes. For the others

The ball milling constitutes new promising methods to produce nanosized particles [7,8]. It has many advantages, e.g low cost, simple operation.The ball-milling is generally used as a mechanical co-grinding of powders, Initially different in nature, up to the preparation of a new powder, homogeneous in composition. The milling is done in cylindrical containers called vials and containing balls. The nature of the milling tools can be as diverse as steel, agate, tungsten carbide… The vials are generally filled under an inert atmosphere to avoid side reactions, since the particles are fractured during the milling process and, therefore,

Several terms are used to call this technique: "Mechanical Alloying" when there is a chemical reaction between different powders, "Mechanical Grinding" or "Mechanical Milling" when the only goal is to modify the texture and/or the structure of a material (no chemical

Two kinds of milling systems were used to prepare our nanopowders (Vibrant and planetary milling) and different milling parameters were considered (milling times, balls to

parameters (substrate, substrate temperature, pressure), different values were used.

The PLD process is shown in figure 1:

## **4.1 Thin film prepared by Pulse Laser Deposition (PLD)**

Fig. 2 shows FESEM micrographs of ZnO thin films grown on sapphire substrate by pulse laser deposition at growth temperature from 685 to 750 °C by using a ZnO powder target at high grade. The experimental parameters are summarized in table.1. It is seen that with the substrate temperature increasing the morphology of ZnO thin lms have a little difference. The thickness of lms decreases with the increase of substrate temperature.

The effect of the distance between target and substrate on the morphology was also studied. Fig. 3 shows the FESEM images of ZnO thin film with different distance between the target and thin film. It is clear that the distance affect the morphology of the film.


Table.1.Growth parameters of ZnO thin films

FE-SEM Characterization of Some Nanomaterial 469

The conditions for production of α-Fe2O3 nano-crystallines by dry milling was studied. [9,10] Commercial α-Fe2O3 powder was used as the starting material. The mechanical milling was carried out in a planetary ball mill Fritsch Pulverisstte 6. The powder was ground in vial with 200g of mixture 1:1 in weight of stainless steel balls (10 and 15 mm in diameter). Different milling times were considered (1, 6, 12, 24 and 48h) and the sample to balls weight ratio was fixed to 1:10. The milling intensity was 250 rpm. Fig.4 shows scanning electron micrographs before and after milling. It is clear that un-milled powder shows an in homogeneities regarding particle size distribution (Fig. 4a). After milling, a reduction of the particle size can be observed with relatively better homogeneity (Fig. 4b-d)). SEM images for increasing milling times reveal clearly that large particles are in fact agglomerates of much

a

c

d

2**µm**

b

2**µm**

**4.2.2 Nanocrystalline zinc ferrite (ZnFe2O4)** 

Fig. 4. FESEM images for the samples milled at different times : a) 0h, b) 12h, c) 24h and d) 48h.

Nanocrystalline zinc ferrite (ZnFe2O4) is synthesized by high-energy ball-milling from a powders mixture of zinc oxide (ZnO) and hematite (α-Fe2O3). [11] Commercially powders of hematite (α-Fe2O3) and zinc oxide (ZnO) are used with equal molar (1:1) and were introduced into a stainless steel vials with stainless steel balls (12 mm and 6 mm in diameter) in a high energy mill (SPEX 8000 mixer mill). Different milling times were

2**µm** 2**µm**

**4.2 Nanopwders obtained by mechanical alloying** 

**4.2.1 Hematite (α-Fe2O3 ) nanocrystallines** 

smaller particles.

Fig. 2. FESEM images of thin lm grown on sapphire at substrate temperature of: (a) 750 °C, (b) 700 °C and (c) 685°C.

Fig. 3. **FESEM** images of thin lm grown on sapphire at substrate temperature of 400°C: a) distance between target and thin film =10mm and b) distance between target and thin film =23 mm

Fig. 2. FESEM images of thin lm grown on sapphire at substrate temperature of: (a) 750 °C,

 Fig. 3. **FESEM** images of thin lm grown on sapphire at substrate temperature of 400°C: a) distance between target and thin film =10mm and b) distance between target and thin

(b) 700 °C and (c) 685°C.

film =23 mm

## **4.2 Nanopwders obtained by mechanical alloying**

## **4.2.1 Hematite (α-Fe2O3 ) nanocrystallines**

The conditions for production of α-Fe2O3 nano-crystallines by dry milling was studied. [9,10] Commercial α-Fe2O3 powder was used as the starting material. The mechanical milling was carried out in a planetary ball mill Fritsch Pulverisstte 6. The powder was ground in vial with 200g of mixture 1:1 in weight of stainless steel balls (10 and 15 mm in diameter). Different milling times were considered (1, 6, 12, 24 and 48h) and the sample to balls weight ratio was fixed to 1:10. The milling intensity was 250 rpm. Fig.4 shows scanning electron micrographs before and after milling. It is clear that un-milled powder shows an in homogeneities regarding particle size distribution (Fig. 4a). After milling, a reduction of the particle size can be observed with relatively better homogeneity (Fig. 4b-d)). SEM images for increasing milling times reveal clearly that large particles are in fact agglomerates of much smaller particles.

Fig. 4. FESEM images for the samples milled at different times : a) 0h, b) 12h, c) 24h and d) 48h.

## **4.2.2 Nanocrystalline zinc ferrite (ZnFe2O4)**

Nanocrystalline zinc ferrite (ZnFe2O4) is synthesized by high-energy ball-milling from a powders mixture of zinc oxide (ZnO) and hematite (α-Fe2O3). [11] Commercially powders of hematite (α-Fe2O3) and zinc oxide (ZnO) are used with equal molar (1:1) and were introduced into a stainless steel vials with stainless steel balls (12 mm and 6 mm in diameter) in a high energy mill (SPEX 8000 mixer mill). Different milling times were

FE-SEM Characterization of Some Nanomaterial 471

Fig. 6. SEM micrographs of ZnO powder for different milling time: a) as received; b) as received at high magnification; c) milled for 3h; d) milled for 3h high magnification; e)

In summary, it is clear that scanning electron microscopy gives tremendous information about the microstructure of nanomaterials including thin film and nano-powders. In addition to that the signals coming from the sample can be used to get information about the

[1] David B Williams and C Barry Carter, "Transmission Electron Microscopy", Springer

[2] Nouredine Zietteli, " Quantum mechanic, concept and applications", Wiley 2001.

milled for 5h; f) milled for 5h at high magnification.

composition of the materials and the structure.

**5. References** 

2009

considered (6, 12 and 24) and two values of the balls to powders mass ratio were used (10:1 and 20:1). SEM micrographs of the samples before and after milling are shown in Figure 5 It is clear that unmilled powder shows a different chap of powders due to zinc oxides and hematite powders (fig.5a, 5b). After milling, a reduction of the crystallite size can be observed fig.5c. High magnification images (fig.5d) reveal clearly the formation of a new nanocrystalline different from the started materials.

## **4.2.3 Zinc oxides Nanocrystalline (ZnO)**

The effects of milling times on the mechanically milled ZnO powder are also studied [12]. Commercially ZnO powders with average particle size of about 1 µm and 99.9% of purity, were introduced into a stainless steel vials with stainless steel balls (12mm and 6mm in diameter) in a SPEX 8000 mixer mill, then milled for different milling periods of time. The balls to powder mass ratio was fixed to 10:1. SEM micrographs of the samples before after milling are shown in Figure.6. It is clear that un-milled powder shows un-homogeneities regarding particle size distribution, where the average size varies in the range 150 – 800 nm (Fig. 6a, 6b). After milling, a reduction of the particle size can be observed with relatively better homogeneity (Fig. 6c, 6e). High magnification images (Fig. 6d, 6f) reveal clearly that large particles are in fact agglomerates of much smaller particles. The average particle size after milling is less than 100nm.

Fig. 5. FESEM micrographs of mixtures (zinc oxides +hematite) powders as received (a) as received at high magnification (b) milled for12h (c) milled for 12h high magnification (d)

considered (6, 12 and 24) and two values of the balls to powders mass ratio were used (10:1 and 20:1). SEM micrographs of the samples before and after milling are shown in Figure 5 It is clear that unmilled powder shows a different chap of powders due to zinc oxides and hematite powders (fig.5a, 5b). After milling, a reduction of the crystallite size can be observed fig.5c. High magnification images (fig.5d) reveal clearly the formation of a new

The effects of milling times on the mechanically milled ZnO powder are also studied [12]. Commercially ZnO powders with average particle size of about 1 µm and 99.9% of purity, were introduced into a stainless steel vials with stainless steel balls (12mm and 6mm in diameter) in a SPEX 8000 mixer mill, then milled for different milling periods of time. The balls to powder mass ratio was fixed to 10:1. SEM micrographs of the samples before after milling are shown in Figure.6. It is clear that un-milled powder shows un-homogeneities regarding particle size distribution, where the average size varies in the range 150 – 800 nm (Fig. 6a, 6b). After milling, a reduction of the particle size can be observed with relatively better homogeneity (Fig. 6c, 6e). High magnification images (Fig. 6d, 6f) reveal clearly that large particles are in fact agglomerates of much smaller particles. The average particle size

Fig. 5. FESEM micrographs of mixtures (zinc oxides +hematite) powders as received (a) as received at high magnification (b) milled for12h (c) milled for 12h high magnification (d)

nanocrystalline different from the started materials.

**4.2.3 Zinc oxides Nanocrystalline (ZnO)** 

after milling is less than 100nm.

Fig. 6. SEM micrographs of ZnO powder for different milling time: a) as received; b) as received at high magnification; c) milled for 3h; d) milled for 3h high magnification; e) milled for 5h; f) milled for 5h at high magnification.

In summary, it is clear that scanning electron microscopy gives tremendous information about the microstructure of nanomaterials including thin film and nano-powders. In addition to that the signals coming from the sample can be used to get information about the composition of the materials and the structure.

## **5. References**


**24** 

*Ghana* 

Osei-Wusu Achaw

**A Study of the Porosity of Activated Carbons** 

The earliest mention of the significance of porosity in the performance of activated carbons is generally attributed to the French chemist Antoine-Alexandre-Brutus Bussy who in a 1822 publication suggested that porosity was important to the adsorptive properties of activated carbons. Since then a lot of research has gone into elucidating the nature of porosity of activated carbons, its development and measurement. In particular, a great deal of research has been spent on understanding factors that affect the development of porosity and how to model the porosity in terms of these factors. Similarly, much effort has gone into identifying accurate methods and procedures for characterizing activated carbons in general and particularly its pore structure. The continued interest in these research is because of the continued use and importance of activated carbons in industry and an unrelenting pursuit to improve on its performance. Characterization of porosity is often done indirectly by measurement of secondary data from which the requisite pore parameters are estimated. But direct methods also exist for characterizing the pore structure of activated carbons. Methods such as optical microscopy and scanning electron microscopy (SEM), in view of their ability to directly view the micro-structure of activated carbons have demonstrated enormous potential for use in the study and characterization of activated carbons [ Manocha et al., 2010; Lazslo et al., 2009*;* Achaw & Afrane, 2008]. However, this latter approach has only been applied in a very limited capacity in the past. Rather, industry and researchers alike continue to rely on the indirect methods to determine and quantify porosity in activated carbons. The indirect methods calculate activated carbon characteristics from measurement of other parameters that are generally thought to relate to the properties of interest. Adsorption measurements and related mathematical models wherein information regarding the pore structure of an activated carbon is determined are the most commonly used amongst the indirect methods. Porosity measurements using this approach extracts such pore characteristics as pore volume, surface area, pore size distribution and average pore diameter based on mathematical models of the adsorption process, information on the adsorbate and an adsorption isotherm. Besides adsorption measurements, several other indirect methods also exist to estimate the pore characteristics of activated carbons. Among these are immersion calorimetry, small angle scattering of X-rays (SAXS), small angle scattering of neutrons (SANS), and mercury porosimetry[Rigby & Edler, 2002; Stoeckli et al,

**1. Introduction** 

2002; Daley et. al., 1996].

**Using the Scanning Electron Microscope** 

*Department of Chemical Engineering, Kumasi Polytechnic, Kumasi,* 


## **A Study of the Porosity of Activated Carbons Using the Scanning Electron Microscope**

Osei-Wusu Achaw *Department of Chemical Engineering, Kumasi Polytechnic, Kumasi, Ghana* 

## **1. Introduction**

472 Scanning Electron Microscopy

[3] Arthur Beiser, " Concept of modern physics", McGraw-Hill, Inc, 1995 H.M. Smith and

[4] D. Dijkkamp, T. Venkatesan, X.D. Wu, S.A. Shaheen, N. Jisrawi, Y.H. Min-Lee, W.L.

[5] D.B. Chrisey and G.K. Hubler (Eds.), "Pulsed Laser Deposition of Thin Films", Wiley,

[6] Raphaël Janot and Daniel Guérard, Progress in Materials Science, Volume 50, Issue 1,

[7] E. Petrovsky, M.D. Alcala, J.M. Criado, T. Grygar, A. Kapicka and J. Subrt, J*. Magn.* 

[8] O.M.Lemine., A.Alyamani, M. Sajieddine and M.Bououdina,, Journal of alloys and

[9] O. M. Lemine , R. Msalam, M. Sajieddine , S. Mufti, A. Alyemani , A. F. Salem, Kh. Ziq and M. Bououdina, International Journal of Nanoscience, Vol. 8, No. 3 (2009) 1–8. [10] O.M. Lemine , M. Bououdina, M. Sajieddine, A. M. Al-Saie, M. Shafi, A. Khatab, M. Al-

[11] O. M. Lemine, A.Alyemani and M.Bououdina, Int. J. Nanoparticles, Vol. 2, 2009

McLean and M. Croft, Appl. Phys. Lett. 51 (1987) 619-621.

hilali1 and M. Henini, Physica B 406 (2011) 1989–1994

A.F. Turner, Appl. Opt. 4 (1965) 147.

*Magn. Mater*. 210 (2000), p. 257.

compounds, 502 (2010), pp. 279-282

New York, 1994.

January 2005, 1-92

The earliest mention of the significance of porosity in the performance of activated carbons is generally attributed to the French chemist Antoine-Alexandre-Brutus Bussy who in a 1822 publication suggested that porosity was important to the adsorptive properties of activated carbons. Since then a lot of research has gone into elucidating the nature of porosity of activated carbons, its development and measurement. In particular, a great deal of research has been spent on understanding factors that affect the development of porosity and how to model the porosity in terms of these factors. Similarly, much effort has gone into identifying accurate methods and procedures for characterizing activated carbons in general and particularly its pore structure. The continued interest in these research is because of the continued use and importance of activated carbons in industry and an unrelenting pursuit to improve on its performance. Characterization of porosity is often done indirectly by measurement of secondary data from which the requisite pore parameters are estimated. But direct methods also exist for characterizing the pore structure of activated carbons. Methods such as optical microscopy and scanning electron microscopy (SEM), in view of their ability to directly view the micro-structure of activated carbons have demonstrated enormous potential for use in the study and characterization of activated carbons [ Manocha et al., 2010; Lazslo et al., 2009*;* Achaw & Afrane, 2008]. However, this latter approach has only been applied in a very limited capacity in the past. Rather, industry and researchers alike continue to rely on the indirect methods to determine and quantify porosity in activated carbons. The indirect methods calculate activated carbon characteristics from measurement of other parameters that are generally thought to relate to the properties of interest. Adsorption measurements and related mathematical models wherein information regarding the pore structure of an activated carbon is determined are the most commonly used amongst the indirect methods. Porosity measurements using this approach extracts such pore characteristics as pore volume, surface area, pore size distribution and average pore diameter based on mathematical models of the adsorption process, information on the adsorbate and an adsorption isotherm. Besides adsorption measurements, several other indirect methods also exist to estimate the pore characteristics of activated carbons. Among these are immersion calorimetry, small angle scattering of X-rays (SAXS), small angle scattering of neutrons (SANS), and mercury porosimetry[Rigby & Edler, 2002; Stoeckli et al, 2002; Daley et. al., 1996].

A Study of the Porosity of Activated Carbons Using the Scanning Electron Microscope 475

porosity in activated carbons is developed but an additional insight into how to control its development. The different pore sizes play unique roles during activated carbon application. Indeed, the classification of the pores in activated carbons into micropores, mesopores and macropores is based more on the varied behavior of admolecules in these pore regimes than on the actual sizes of the pores. Thus, more than the total pore volume or total surface area of the activated carbons, the fraction of the total pore volume or surface area due to the various sizes of pores is of utmost importance. Again, understanding this development is essential for the design of models to describe the performance of activated carbons and the prediction of activated carbon behavior. According to IUPAC nomenclature [Sing, et al., 1985], micropores are those pores with width less than 2 nm. The micropores play the key role of providing the bulk of the surface involved in adsorption, which is the basis of many applications of activated carbons. The mesopores are wider than the micropores and have pore widths in the range 2 nm to 50 nm. The mesopores also play a role in adsorption albeit on a reduced scale compared to the micropores. The role of the mesopores becomes more important during the adsorption of large molecules that cannot be accommodated in the micropores. Finally, there are the macropores which have much larger pore sizes and which play the important role of being the conduits through which access to the interior of the activated carbon and hence to the mesopores and micropores are achieved. They are generally considered as being part of the external surface of the activated

Pores in activated carbons are areas of zero electron density in the carbon matrix. These constitute volume elements distributed throughout the particle and posses varied sizes and shapes. The individual volume elements are connected with each other through open channels which are themselves also volume elements. The volume elements are now known to originate from several sources. First, there are those whose source can be traced directly to the primary pore structure of the precursor material. Another group of pores are created as a result of the imperfections that arise from the arrangements of the lamellar constituent molecules (LCM) which are the building blocks of activated carbons. The LCM are layers of sheets which are made up of interconnected aromatic rings. They are formed when the precursor materials are subjected to heat treatment at the appropriate temperature and conditions [Bryne & Marsh, 1995; Evans & Marsh, 1995]. The imperfect arrangements of the LCM creates space in-between parallel layers of the molecules. Volume elements are also created when parts of the LCM are reacted away during contact with agents used in the activation process. The volume elements arising as a result of LCM arrangements and reactions constitute microporosity, and to a lesser extent mesoporosity in the activated carbon. The microporosity confers on activated carbons the unique ability to adsorb large quantities of a diverse range of molecules which makes activated carbons so useful in

The LCM and the accompanying microporosity are formed when the original cellular structure of the precursor material undergoes molecular transformation and reconstitution. During heat treatment of the precursor material a number of physical and chemical processes occur that culminate in the final activated carbon pore structure. Among these, first, moisture and other volatile constituents of the precursor material escape leaving voids that may be later transformed or retain themselves in the final activated carbon product. Secondly, the macromolecules of the precursor material breakdown, lose mostly oxygen and

carbon. The macropores have size greater than 50 nm.

separation processes and other applications.

The weakness of the indirect methods, is that they are based on models that do not always match with observed behavior of activated carbons. Others like mercury porosimetry are based on very simplified descriptions of the pore structure of activated carbons that are greatly deviated from pores observed directly using direct methods. Not surprisingly, pore characteristics estimated based on two different such models or methods rarely agree [Rodriguez-Reinoso and Linares-Solano, 1989]. The weaknesses notwithstanding, the indirect methods have thus far served a useful purpose of providing a framework for assessing and comparing activated carbons. In particular, they have provided a useful vehicle for predicting and evaluating the performance of these materials for industrial and other applications. The drawbacks of these methods, however, have meant that more consistent and reliable methods continue to be searched to measure the characteristics of activated carbons. The direct methods represent a viable option in that regard. Direct methods allow the direct viewing of the topography of the activated carbon surfaces which makes possible improved description of activated carbon properties such as pore shapes and pore orientation. When coupled with other methods or instrumentations, such as computerized image analysis, it is possible to estimate the pore characteristics of activated carbons more accurately. Again, these methods make possible a visual follow up of the stages of activated carbon manufacture which in turn makes possible the tracking of the changes that a precursor material goes through in forming an activated carbon. It thus offers enormous possibilities of shedding light on the pore development processes than hitherto known [Achaw & Afrane, 2008]. Already, in areas such as materials engineering, biology and medical sciences, the SEM has been extensively used to study and characterize the microstructure of substances [Chira et al., 2009; Vaishali et al., 2008; Chung, et al., 2008; Kamran, 1997]. The purpose of this chapter is to discuss the potential use of the SEM in understanding porosity development in activated carbon and pore structure characterization using micrographs of coconut shells at different stages during the manufacture of coconut shell-based activated carbons.

#### **2. Tracking porosity development using the scanning electron microscope**

To better control porosity in activated carbons, it is essential that its development during its preparation be well understood. It is now generally known that porosity in activated carbons is derived from three main sources, namely, the inherent cellular structure of the precursor material, the conditions extant during the preparation of activated carbons and the composition of the precursor material [Heschel & Klose, 1995; Raveendran et al., 1995; Evans & Marsh, 1979;]. How these factors combine to produce an activated carbon of a given specification has been and continues to be a subject of intense research. This continued search is borne out of the need to find newer applications for activated carbons and an unending desire to improve on the performance of activated carbons in such operations like filtration, gas and metal adsorption and separation, gas storage, and finally in water purification. In all these applications the performance of an activated carbon depends as much on the total pore volume as it does on the pore size distribution, the prevalence of a certain pore size regime, and the surface chemistry of the carbon. For instance, during operations involving molecular sieve activated carbons, that pore size characteristics is required that permits the separation of two or any number of molecules of differing molecular sizes. Achieving this kind of performance demands a special design of the pore structure of activated carbons. This in turn demands not only an understanding of how

The weakness of the indirect methods, is that they are based on models that do not always match with observed behavior of activated carbons. Others like mercury porosimetry are based on very simplified descriptions of the pore structure of activated carbons that are greatly deviated from pores observed directly using direct methods. Not surprisingly, pore characteristics estimated based on two different such models or methods rarely agree [Rodriguez-Reinoso and Linares-Solano, 1989]. The weaknesses notwithstanding, the indirect methods have thus far served a useful purpose of providing a framework for assessing and comparing activated carbons. In particular, they have provided a useful vehicle for predicting and evaluating the performance of these materials for industrial and other applications. The drawbacks of these methods, however, have meant that more consistent and reliable methods continue to be searched to measure the characteristics of activated carbons. The direct methods represent a viable option in that regard. Direct methods allow the direct viewing of the topography of the activated carbon surfaces which makes possible improved description of activated carbon properties such as pore shapes and pore orientation. When coupled with other methods or instrumentations, such as computerized image analysis, it is possible to estimate the pore characteristics of activated carbons more accurately. Again, these methods make possible a visual follow up of the stages of activated carbon manufacture which in turn makes possible the tracking of the changes that a precursor material goes through in forming an activated carbon. It thus offers enormous possibilities of shedding light on the pore development processes than hitherto known [Achaw & Afrane, 2008]. Already, in areas such as materials engineering, biology and medical sciences, the SEM has been extensively used to study and characterize the microstructure of substances [Chira et al., 2009; Vaishali et al., 2008; Chung, et al., 2008; Kamran, 1997]. The purpose of this chapter is to discuss the potential use of the SEM in understanding porosity development in activated carbon and pore structure characterization using micrographs of coconut shells at different stages during the

**2. Tracking porosity development using the scanning electron microscope**  To better control porosity in activated carbons, it is essential that its development during its preparation be well understood. It is now generally known that porosity in activated carbons is derived from three main sources, namely, the inherent cellular structure of the precursor material, the conditions extant during the preparation of activated carbons and the composition of the precursor material [Heschel & Klose, 1995; Raveendran et al., 1995; Evans & Marsh, 1979;]. How these factors combine to produce an activated carbon of a given specification has been and continues to be a subject of intense research. This continued search is borne out of the need to find newer applications for activated carbons and an unending desire to improve on the performance of activated carbons in such operations like filtration, gas and metal adsorption and separation, gas storage, and finally in water purification. In all these applications the performance of an activated carbon depends as much on the total pore volume as it does on the pore size distribution, the prevalence of a certain pore size regime, and the surface chemistry of the carbon. For instance, during operations involving molecular sieve activated carbons, that pore size characteristics is required that permits the separation of two or any number of molecules of differing molecular sizes. Achieving this kind of performance demands a special design of the pore structure of activated carbons. This in turn demands not only an understanding of how

manufacture of coconut shell-based activated carbons.

porosity in activated carbons is developed but an additional insight into how to control its development. The different pore sizes play unique roles during activated carbon application. Indeed, the classification of the pores in activated carbons into micropores, mesopores and macropores is based more on the varied behavior of admolecules in these pore regimes than on the actual sizes of the pores. Thus, more than the total pore volume or total surface area of the activated carbons, the fraction of the total pore volume or surface area due to the various sizes of pores is of utmost importance. Again, understanding this development is essential for the design of models to describe the performance of activated carbons and the prediction of activated carbon behavior. According to IUPAC nomenclature [Sing, et al., 1985], micropores are those pores with width less than 2 nm. The micropores play the key role of providing the bulk of the surface involved in adsorption, which is the basis of many applications of activated carbons. The mesopores are wider than the micropores and have pore widths in the range 2 nm to 50 nm. The mesopores also play a role in adsorption albeit on a reduced scale compared to the micropores. The role of the mesopores becomes more important during the adsorption of large molecules that cannot be accommodated in the micropores. Finally, there are the macropores which have much larger pore sizes and which play the important role of being the conduits through which access to the interior of the activated carbon and hence to the mesopores and micropores are achieved. They are generally considered as being part of the external surface of the activated carbon. The macropores have size greater than 50 nm.

Pores in activated carbons are areas of zero electron density in the carbon matrix. These constitute volume elements distributed throughout the particle and posses varied sizes and shapes. The individual volume elements are connected with each other through open channels which are themselves also volume elements. The volume elements are now known to originate from several sources. First, there are those whose source can be traced directly to the primary pore structure of the precursor material. Another group of pores are created as a result of the imperfections that arise from the arrangements of the lamellar constituent molecules (LCM) which are the building blocks of activated carbons. The LCM are layers of sheets which are made up of interconnected aromatic rings. They are formed when the precursor materials are subjected to heat treatment at the appropriate temperature and conditions [Bryne & Marsh, 1995; Evans & Marsh, 1995]. The imperfect arrangements of the LCM creates space in-between parallel layers of the molecules. Volume elements are also created when parts of the LCM are reacted away during contact with agents used in the activation process. The volume elements arising as a result of LCM arrangements and reactions constitute microporosity, and to a lesser extent mesoporosity in the activated carbon. The microporosity confers on activated carbons the unique ability to adsorb large quantities of a diverse range of molecules which makes activated carbons so useful in separation processes and other applications.

The LCM and the accompanying microporosity are formed when the original cellular structure of the precursor material undergoes molecular transformation and reconstitution. During heat treatment of the precursor material a number of physical and chemical processes occur that culminate in the final activated carbon pore structure. Among these, first, moisture and other volatile constituents of the precursor material escape leaving voids that may be later transformed or retain themselves in the final activated carbon product. Secondly, the macromolecules of the precursor material breakdown, lose mostly oxygen and

A Study of the Porosity of Activated Carbons Using the Scanning Electron Microscope 477

Fig. 1. Micrograph of surface of transverse section of raw coconut shell.

Fig. 2. Micrograph of surface of transverse section of carbonized coconut shell

Source: *Achaw & Afrane, 2008*

Source: *Achaw & Afrane, 2008*

hydrogen and reconstitute into aromatic rings which become the building blocks of the LCM. The new constituent molecules form to enclose the vacancies left by the escaping elements and molecules. These vacant lots also constitute porosity in activated carbons. The transformations are initiated during the pyrolysis of the precursor material and are continued and enhanced during the subsequent activation stage. The loss of volatile matter from the precursor materials occurs at all temperatures but aromatization of the material occurs at temperatures in excess of 700oC. Another important process occurs during the activation process to create new pores or enhance existing pores formed during the pyrolysis step. During activation, activation agents react with the carbon skeleton to create new pores or enlarge existing ones. It is also at the activation stage that other phenomena that facilitate pore creation manifest. For instance, inherent mineral matter such as alkali metals in the precursor material catalyze the pore formation process leading to such phenomena as pitting, channeling and pore enlargement [Bryne & Marsh, 1995]. Pores are also developed as a result of thermal stress on the cellular structure of the precursor material. This stress leads to the development of cracks, crevices, slits, fissures, and all manner of openings in the matrix of the ultimate carbon material. The events leading to the formation of pores occur mainly at the micro and sub micro levels and most of the products of the process such as LCM and associated carbon rings are hardly, directly, observable even with the most powerful of electron microscopes available today. As such these processes have most remained in the realm of theoretical discourse. However, there are other manifestations of these transformations that with the appropriate tools are observable. Using the SEM it has been possible to view images of some of the phenomena that engender porosity development in activated carbons. The ability of the SEM to distinguish objects as small as 1 nm makes it ideal for tracking the transformations happening in the precursor material during activated carbon formation.. This SEM has however not been fully exploited yet for the study of porosity development in activated carbons safe for the pioneering work of Achaw & Afrane, 2008. Figures 1-3 below, show SEM micrographs of sections of coconut shell at different stages during the preparation of coconut shell-based activated carbons. The images reveal details about these materials that shed useful light on aspects of porosity development in activated carbons.

Micrographs in Figures 1 – 3 reveal details of activated porosity development that, previously, has only being a matter of theoretical discourse. Samples for the SEM micrographs used in this study were prepared by cutting sections of well dried coconut shell (raw coconut shells, carbonized shells or coconut shell-based activated carbon) and mounting a on specimen stub with the help of a conductive silver adhesive. The specimen surfaces were thereafter sputter coated with a thin film of silver, placed in the sample holder and viewed with a Ziess DSM 962 electron microscope. A look at Figures 1 and 2 suggest that the original cellular structure of the coconut shells as seen from the transverse section are largely maintained albeit in a modified form following carbonization. In Figure 1 the largely isolated cylindrical units, see positions labeled A and B, has walls made up of layers of thin sheets. In Figure 2, these units seen in Figure 1 have joined together at the walls into a singular solid matrix interspersed with pores. The sheets of the walls are no longer visible in Figure 2. The joining of the walls and the fusing together of the sheets of the walls suggests a profound transformation at the molecular level in the shell during the pyrolysis.

hydrogen and reconstitute into aromatic rings which become the building blocks of the LCM. The new constituent molecules form to enclose the vacancies left by the escaping elements and molecules. These vacant lots also constitute porosity in activated carbons. The transformations are initiated during the pyrolysis of the precursor material and are continued and enhanced during the subsequent activation stage. The loss of volatile matter from the precursor materials occurs at all temperatures but aromatization of the material occurs at temperatures in excess of 700oC. Another important process occurs during the activation process to create new pores or enhance existing pores formed during the pyrolysis step. During activation, activation agents react with the carbon skeleton to create new pores or enlarge existing ones. It is also at the activation stage that other phenomena that facilitate pore creation manifest. For instance, inherent mineral matter such as alkali metals in the precursor material catalyze the pore formation process leading to such phenomena as pitting, channeling and pore enlargement [Bryne & Marsh, 1995]. Pores are also developed as a result of thermal stress on the cellular structure of the precursor material. This stress leads to the development of cracks, crevices, slits, fissures, and all manner of openings in the matrix of the ultimate carbon material. The events leading to the formation of pores occur mainly at the micro and sub micro levels and most of the products of the process such as LCM and associated carbon rings are hardly, directly, observable even with the most powerful of electron microscopes available today. As such these processes have most remained in the realm of theoretical discourse. However, there are other manifestations of these transformations that with the appropriate tools are observable. Using the SEM it has been possible to view images of some of the phenomena that engender porosity development in activated carbons. The ability of the SEM to distinguish objects as small as 1 nm makes it ideal for tracking the transformations happening in the precursor material during activated carbon formation.. This SEM has however not been fully exploited yet for the study of porosity development in activated carbons safe for the pioneering work of Achaw & Afrane, 2008. Figures 1-3 below, show SEM micrographs of sections of coconut shell at different stages during the preparation of coconut shell-based activated carbons. The images reveal details about these materials that shed useful light on

Micrographs in Figures 1 – 3 reveal details of activated porosity development that, previously, has only being a matter of theoretical discourse. Samples for the SEM micrographs used in this study were prepared by cutting sections of well dried coconut shell (raw coconut shells, carbonized shells or coconut shell-based activated carbon) and mounting a on specimen stub with the help of a conductive silver adhesive. The specimen surfaces were thereafter sputter coated with a thin film of silver, placed in the sample holder and viewed with a Ziess DSM 962 electron microscope. A look at Figures 1 and 2 suggest that the original cellular structure of the coconut shells as seen from the transverse section are largely maintained albeit in a modified form following carbonization. In Figure 1 the largely isolated cylindrical units, see positions labeled A and B, has walls made up of layers of thin sheets. In Figure 2, these units seen in Figure 1 have joined together at the walls into a singular solid matrix interspersed with pores. The sheets of the walls are no longer visible in Figure 2. The joining of the walls and the fusing together of the sheets of the walls suggests a profound transformation at the molecular level in the shell during the pyrolysis.

aspects of porosity development in activated carbons.

Fig. 1. Micrograph of surface of transverse section of raw coconut shell. Source: *Achaw & Afrane, 2008*

Fig. 2. Micrograph of surface of transverse section of carbonized coconut shell Source: *Achaw & Afrane, 2008*

A Study of the Porosity of Activated Carbons Using the Scanning Electron Microscope 479

matrix due to temperature changes in the activation process. Such cracks contribute to the overall surface of the activated carbon and as such are important. Yet discourse on porosity

Characterization of activated carbons is driven by the need to have qualitative and quantitative information which serve as the basis for comparison and selection of activated carbons for specific applications. Such data are also useful for modeling the behavior and performance of activated carbons. Furthermore, characterization provides feedback for use in the design and preparation of activated carbons. The characteristics often measured are density, abrasion resistance, surface area, average pore size, pore size distribution, pore shape, pore volume, and the surface chemistry of the carbon. Whilst there are well established standard methods for measuring the density and abrasion resistance, scientists and industry are still grappling with what accurate methods to use for measuring the others. Most current methods estimate these parameters indirectly from measurements of secondary data on the activated carbons. As a result there are still concerns with the accuracy of values determined for these parameters. The most popular method for characterizing activated carbons is through the measurement of adsorption data and application of mathematical models that relate the adsorption data to such characteristics as pore volume of the adsorbent and the properties of the adsorptive. Other indirect methods, namely, mercury porosimetry, immersion calorimetry, small angle scattering of X-rays (SAXS), neutrons (SANS), high-resolution transmission electron microscopy are also sometimes used to determine the characteristics of activated carbons. Then there are direct methods that hold enormous potential for characterizing activated carbons but which use are rarely mentioned in activated carbon literature. These latter methods are mainly the microscopic methods which enable the observation of micro- and sub-micro features of activated carbons and hence the direct measurement of these features. These microscopic

development of activated carbons often ignore these cracks.

**3. Characterization of activated carbons** 

methods are optical microscopy and the SEM.

**3.1 Adsorption methods of characterizing activated carbons** 

the solid at a fixed temperature and can be expressed mathematically as

*P P V V bV*

*V* is the equilibrium adsorbed amount (mmolg-1) of the adsorbate per unit mass of the adsorbent at a pressure P. *Vm* is the amount of gas required for monolayer coverage of the

These methods almost invariably combine the adsorption isotherm of a given adsorbateadsorbent system and a theoretical or empirical model of the adsorption process to estimate the characteristics of activated carbons [Machnikowski, et al., 2010; Noor & Nawi, 2008; Lozano-Castello et al., 2004; Stoeckli et al., 2002; Stoeckli et al., 2001 Yuna et al., 2001; Rodriguez-Reinoso, 1989]. Most commercial sorption equipment estimate activated carbon data using in-built software based on one version or the other of this approach. Of historical importance is the Langmuir model [ Gregg & Sing, 1982] which was first developed in 1916 to describe adsorption behavior on solid adsorbents in general. The model relates the adsorption of molecules in a gaseous medium onto a solid surface to the gas pressure above

> 1 *m m*

= + (1)

Fig. 3. Micrograph of transverse section of coconut shell-based activated carbon Source: *Achaw & Afrane, 2008*

Notice that the cylindrical units in Figure 1 has deformed into all manner of shapes after the pyrolysis, see Figure 2, positions marked C and D. The transformation of the pores continues to the activation stage, see positions marked E and F in Figure 3, where what were originally cylindrical shape has now become partially flattened cylinder. The micrograph of the activated carbon further demonstrates transformation of the matrix of the shell as a result of continued heating during the activation. The narrowing of the pore widths suggest a kind of deformation where the matrix softens and walls of the pores give way and close in on each other. Notice further that as a result of this deformation, some of the walls have completely fallen in on each other resulting in a complete zipped up of the pores, see positions marked G and H. The transformations suggest that the carbon matrix passes through a plastic phase as a result of the thermal treatment. Another phenomena observable from these micrographs is the preponderance of foreign materials in the pores of the carbonized product, see positions marked I and J on Figure 2, and the almost lack of these materials in the activated carbon of Figure 3. This means that the activation process serves the additional purpose of cleaning foreign materials from the carbon besides the creation of pores. It is nonetheless noteworthy that even at the activated carbon stage some pores still remain blocked by foreign material, see position marked K in Figure 3. This last observation is an indication that the activation process was not complete. It is anticipated that these foreign materials would be completely cleared at the end of the activation process. Yet another important feature of these micrographs is the position marked L on Figure 3 which is a crack in the carbon matrix probably developed as a result of thermal stress on the carbon matrix due to temperature changes in the activation process. Such cracks contribute to the overall surface of the activated carbon and as such are important. Yet discourse on porosity development of activated carbons often ignore these cracks.

## **3. Characterization of activated carbons**

478 Scanning Electron Microscopy

Fig. 3. Micrograph of transverse section of coconut shell-based activated carbon

Notice that the cylindrical units in Figure 1 has deformed into all manner of shapes after the pyrolysis, see Figure 2, positions marked C and D. The transformation of the pores continues to the activation stage, see positions marked E and F in Figure 3, where what were originally cylindrical shape has now become partially flattened cylinder. The micrograph of the activated carbon further demonstrates transformation of the matrix of the shell as a result of continued heating during the activation. The narrowing of the pore widths suggest a kind of deformation where the matrix softens and walls of the pores give way and close in on each other. Notice further that as a result of this deformation, some of the walls have completely fallen in on each other resulting in a complete zipped up of the pores, see positions marked G and H. The transformations suggest that the carbon matrix passes through a plastic phase as a result of the thermal treatment. Another phenomena observable from these micrographs is the preponderance of foreign materials in the pores of the carbonized product, see positions marked I and J on Figure 2, and the almost lack of these materials in the activated carbon of Figure 3. This means that the activation process serves the additional purpose of cleaning foreign materials from the carbon besides the creation of pores. It is nonetheless noteworthy that even at the activated carbon stage some pores still remain blocked by foreign material, see position marked K in Figure 3. This last observation is an indication that the activation process was not complete. It is anticipated that these foreign materials would be completely cleared at the end of the activation process. Yet another important feature of these micrographs is the position marked L on Figure 3 which is a crack in the carbon matrix probably developed as a result of thermal stress on the carbon

Source: *Achaw & Afrane, 2008*

Characterization of activated carbons is driven by the need to have qualitative and quantitative information which serve as the basis for comparison and selection of activated carbons for specific applications. Such data are also useful for modeling the behavior and performance of activated carbons. Furthermore, characterization provides feedback for use in the design and preparation of activated carbons. The characteristics often measured are density, abrasion resistance, surface area, average pore size, pore size distribution, pore shape, pore volume, and the surface chemistry of the carbon. Whilst there are well established standard methods for measuring the density and abrasion resistance, scientists and industry are still grappling with what accurate methods to use for measuring the others. Most current methods estimate these parameters indirectly from measurements of secondary data on the activated carbons. As a result there are still concerns with the accuracy of values determined for these parameters. The most popular method for characterizing activated carbons is through the measurement of adsorption data and application of mathematical models that relate the adsorption data to such characteristics as pore volume of the adsorbent and the properties of the adsorptive. Other indirect methods, namely, mercury porosimetry, immersion calorimetry, small angle scattering of X-rays (SAXS), neutrons (SANS), high-resolution transmission electron microscopy are also sometimes used to determine the characteristics of activated carbons. Then there are direct methods that hold enormous potential for characterizing activated carbons but which use are rarely mentioned in activated carbon literature. These latter methods are mainly the microscopic methods which enable the observation of micro- and sub-micro features of activated carbons and hence the direct measurement of these features. These microscopic methods are optical microscopy and the SEM.

#### **3.1 Adsorption methods of characterizing activated carbons**

These methods almost invariably combine the adsorption isotherm of a given adsorbateadsorbent system and a theoretical or empirical model of the adsorption process to estimate the characteristics of activated carbons [Machnikowski, et al., 2010; Noor & Nawi, 2008; Lozano-Castello et al., 2004; Stoeckli et al., 2002; Stoeckli et al., 2001 Yuna et al., 2001; Rodriguez-Reinoso, 1989]. Most commercial sorption equipment estimate activated carbon data using in-built software based on one version or the other of this approach. Of historical importance is the Langmuir model [ Gregg & Sing, 1982] which was first developed in 1916 to describe adsorption behavior on solid adsorbents in general. The model relates the adsorption of molecules in a gaseous medium onto a solid surface to the gas pressure above the solid at a fixed temperature and can be expressed mathematically as

$$\frac{P}{V} = \frac{P}{V\_m} + \frac{1}{bV\_m} \tag{1}$$

*V* is the equilibrium adsorbed amount (mmolg-1) of the adsorbate per unit mass of the adsorbent at a pressure P. *Vm* is the amount of gas required for monolayer coverage of the

A Study of the Porosity of Activated Carbons Using the Scanning Electron Microscope 481

adsorption in micropores is characterized more by pore filling than by surface coverage. As such, the application of the BET method does not always yield the correct result for the

Probably more accurate among the adsorption methods for the determination of pore characteristics of activated carbons is the Dubnin-Raduskevitch (DR) equation and its improved and more versatile version, the Dubnin-Astakov (DA) equation [Dubnin & Raduskekevich, 1947; Dubnin, 1989; Carrasco-Merin et al., 1996; Gil, 1998]. The DR equation is premised on the assumption that adsorption in micropores occurs by pore filling rather than by physical adsorption on the surface of the micropores. The equation relates the volume of pores, *W,* filled by an adsorbate at a given temperature *T* and relative pressure

> <sup>2</sup> exp[ ( ) ] *<sup>o</sup> o*

β

*o*

*<sup>E</sup>* = − (5)

*<sup>p</sup>* <sup>=</sup> (6)

*L nm E* ( ) 10.8( 11.4) = − *<sup>O</sup>* (7)

*<sup>E</sup>* = − (8)

*<sup>A</sup> W W*

where *W*o is the total volume of the micropores, *Eo* is the characteristic energy*,* and *β* is the affinity coefficient. Both *Eo* and β are system dependent. The deferential molar work of

*<sup>p</sup> A RT*

For slit-shaped pores, a relationship exists between *Eo* and the average micropore width *L*

The DR equation has a narrow range of application as it corresponds to mostly Type I isotherms. The Dubnin-Astakov (DA) equation which is a modification of the DR equation and which is applicable to a wide range of microporous carbons is therefore preferred. The

*o*

**3.2 Characterization of activated carbons with the scanning electron microscope** 

In the areas of porosity development and characterization of activated carbons, a number of issues still remain unresolved. The current state of knowledge has not been able to address all observed behavior and performance of activated carbons. For instance, to what extent does thermal stresses on the carbon matrix during pyrolysis affect porosity development.

*<sup>A</sup> W W*

When n=2, equation (8) becomes equal to the DR equation. Values of n between 1 and 4 are observed for most carbon adsorbents, n> 2 for molecular sieve carbons or carbons with highly homogeneous and small micropores, n < 2 for strongly activated carbons and heterogeneous micropore carbons. For monodisperse carbons, n=3 and for strongly

exp[ ( ) ] *<sup>n</sup>*

β

*o*

surface area of activated carbons, especially if they are predominantly microporous.

*p/po* and other parameters of the adsorption system as

adsorption on the adsorbent, A, is further defined as

generalized form of the DA equation is

heterogeneous carbons n= 2 [Carrasco-Merin et al., 1996].

as

adsorbent (mmolg-1), and b is a constant whose value depends on the temperature. A linear plot of equation (1) allows Vm to be evaluated from the gradient and hence the adsorbents surface area from the relation

$$S = V\_m \text{L}\sigma \tag{2}$$

S is the total surface area of the adsorbent (m2g-1), L is Avogadro's number and σ is the projected surface area of the adsorbate molecule. S is the sum total of pore surfaces and external (non-pore) surface of the adsorbent. Equation (1) is based on the assumption that i) there is a mono-layer adsorption, ii) there are no adsorbate-adsorbate interactions on the adsorbent surface, iii) the adsorbent has a homogeneous surface, iv) all adsorption sites on the adsorbent are equivalent and, v) the adsorbing gas adsorbs into an immobile state.

The Langmuir's model has been found to be of limited applications for activated carbons. In particular, activated carbon surfaces are rarely homogeneous. Generally, the assumptions have been found not to be consistent with observations therefore the Langmuir model is rarely used to characterize activated carbons. Consequently, other relatively more accurate models of adsorption are often used. One such model is the Braunnauer, Emmett and Teller (BET) model [Sing et. al., 1985]. The BET method has a much wider application and is most often used to interpret adsorption isotherms obtained using Nitrogen at 77K as the adsorbate. The model is an improvement on the Langmuir model in that it can account for multilayer adsorption. It relates the adsorption pressure and the volume of the adsorbed adsorbate according to the equation

$$\frac{p}{V(p^o - p)} = \frac{1}{V\_{mc}} + \frac{c - 1}{V\_{mc}} \frac{p}{p^o} \tag{3}$$

where

$$c = \exp\left(\frac{\Delta H\_A - \Delta H\_L}{RT}\right) \tag{4}$$

In equation (3), *V*mc is the monolayer capacity of the adsorbent, po is the saturation vapour pressure of the adsorbate gas, p is the pressure of the gas, and c is a constant which is exponentially related to the heat of first layer adsorption. *∆HA* is the heat of adsorption, *∆HL* is the heat of liquefaction of the adsorption fluid, T is the temperature, and R is the gas constant. A linear plot of equation (3) allows *V*mc to be determined from the intercept and from which the surface area of the activated carbon can be estimated using equation (2). The BET method has found a number of applications in adsorption studies and is especially used in the determination of the surface areas of adsorbents including activated carbons. Even so, the BET equation is unable to account for adsorption in a number of instances. For activated carbons equation (3) is only linear at p/po < 0.1. This introduces error into the measurement of those pores for which adsorption is possible at pressures for which p/po > 0.1. Secondly, the calculation of the adsorbent surface area using equation (2) requires knowledge of the projected surface area of the adsorbate molecule, in this case Nitrogen. This in turn requires that the adsorbate molecules (Nitrogen) be in a close packed, monolayer coverage on the adsorbent. The use of the method implicitly assumes that the value estimated for Vmc is necessarily accurate and that σ is constant for the adsorbate under all conditions. Further,

adsorbent (mmolg-1), and b is a constant whose value depends on the temperature. A linear plot of equation (1) allows Vm to be evaluated from the gradient and hence the adsorbents

> *S VL* = *<sup>m</sup>* σ

S is the total surface area of the adsorbent (m2g-1), L is Avogadro's number and σ is the projected surface area of the adsorbate molecule. S is the sum total of pore surfaces and external (non-pore) surface of the adsorbent. Equation (1) is based on the assumption that i) there is a mono-layer adsorption, ii) there are no adsorbate-adsorbate interactions on the adsorbent surface, iii) the adsorbent has a homogeneous surface, iv) all adsorption sites on the adsorbent are equivalent and, v) the adsorbing gas adsorbs into an immobile state.

The Langmuir's model has been found to be of limited applications for activated carbons. In particular, activated carbon surfaces are rarely homogeneous. Generally, the assumptions have been found not to be consistent with observations therefore the Langmuir model is rarely used to characterize activated carbons. Consequently, other relatively more accurate models of adsorption are often used. One such model is the Braunnauer, Emmett and Teller (BET) model [Sing et. al., 1985]. The BET method has a much wider application and is most often used to interpret adsorption isotherms obtained using Nitrogen at 77K as the adsorbate. The model is an improvement on the Langmuir model in that it can account for multilayer adsorption. It relates the adsorption pressure and the volume of the adsorbed

1 1

<sup>−</sup> = + − (3)

Δ −Δ <sup>=</sup> (4)

( ) *o o mc mc p p c V pp p V V*

( ) exp( ) *H H A L <sup>c</sup> RT*

In equation (3), *V*mc is the monolayer capacity of the adsorbent, po is the saturation vapour pressure of the adsorbate gas, p is the pressure of the gas, and c is a constant which is exponentially related to the heat of first layer adsorption. *∆HA* is the heat of adsorption, *∆HL* is the heat of liquefaction of the adsorption fluid, T is the temperature, and R is the gas constant. A linear plot of equation (3) allows *V*mc to be determined from the intercept and from which the surface area of the activated carbon can be estimated using equation (2). The BET method has found a number of applications in adsorption studies and is especially used in the determination of the surface areas of adsorbents including activated carbons. Even so, the BET equation is unable to account for adsorption in a number of instances. For activated carbons equation (3) is only linear at p/po < 0.1. This introduces error into the measurement of those pores for which adsorption is possible at pressures for which p/po > 0.1. Secondly, the calculation of the adsorbent surface area using equation (2) requires knowledge of the projected surface area of the adsorbate molecule, in this case Nitrogen. This in turn requires that the adsorbate molecules (Nitrogen) be in a close packed, monolayer coverage on the adsorbent. The use of the method implicitly assumes that the value estimated for Vmc is necessarily accurate and that σ is constant for the adsorbate under all conditions. Further,

(2)

surface area from the relation

adsorbate according to the equation

where

adsorption in micropores is characterized more by pore filling than by surface coverage. As such, the application of the BET method does not always yield the correct result for the surface area of activated carbons, especially if they are predominantly microporous.

Probably more accurate among the adsorption methods for the determination of pore characteristics of activated carbons is the Dubnin-Raduskevitch (DR) equation and its improved and more versatile version, the Dubnin-Astakov (DA) equation [Dubnin & Raduskekevich, 1947; Dubnin, 1989; Carrasco-Merin et al., 1996; Gil, 1998]. The DR equation is premised on the assumption that adsorption in micropores occurs by pore filling rather than by physical adsorption on the surface of the micropores. The equation relates the volume of pores, *W,* filled by an adsorbate at a given temperature *T* and relative pressure *p/po* and other parameters of the adsorption system as

$$\mathcal{W} = \mathcal{W}\_o \exp\left[-(\frac{A}{\beta \mathcal{E}\_o})^2\right] \tag{5}$$

where *W*o is the total volume of the micropores, *Eo* is the characteristic energy*,* and *β* is the affinity coefficient. Both *Eo* and β are system dependent. The deferential molar work of adsorption on the adsorbent, A, is further defined as

$$A = RT \frac{p}{p^o} \tag{6}$$

For slit-shaped pores, a relationship exists between *Eo* and the average micropore width *L* as

$$
\overline{L}(nm) = 10.8(E\_O - 11.4) \tag{7}
$$

The DR equation has a narrow range of application as it corresponds to mostly Type I isotherms. The Dubnin-Astakov (DA) equation which is a modification of the DR equation and which is applicable to a wide range of microporous carbons is therefore preferred. The generalized form of the DA equation is

$$\mathcal{W} = \mathcal{W}\_o \exp\left[-(\frac{A}{\beta \mathcal{E}\_o})^n\right] \tag{8}$$

When n=2, equation (8) becomes equal to the DR equation. Values of n between 1 and 4 are observed for most carbon adsorbents, n> 2 for molecular sieve carbons or carbons with highly homogeneous and small micropores, n < 2 for strongly activated carbons and heterogeneous micropore carbons. For monodisperse carbons, n=3 and for strongly heterogeneous carbons n= 2 [Carrasco-Merin et al., 1996].

#### **3.2 Characterization of activated carbons with the scanning electron microscope**

In the areas of porosity development and characterization of activated carbons, a number of issues still remain unresolved. The current state of knowledge has not been able to address all observed behavior and performance of activated carbons. For instance, to what extent does thermal stresses on the carbon matrix during pyrolysis affect porosity development.

A Study of the Porosity of Activated Carbons Using the Scanning Electron Microscope 483

that, in principle, even micropores could be exposed for study by SEM micrographs. Again, due to the very narrow electron beam employed, SEM micrographs have a large depth of field that yields a pseudo three-dimensional appearance useful for understanding the surface structure of a sample. Figures 4-7 are examples of SEM micrographs that reveal

details of activated carbons that only the direct methods can show.

Fig. 4. Micrograph of outer surface of raw coconut shell

Fig. 5. Micrograph of outer surface of carbonized coconut shell.

Source: *Achaw & Afrane, 2008*

Source: *Achaw & Afrane, 2008*

How realistic is the often used slit-shaped pore model in describing activated carbons. Similarly, in characterizing activated carbons the often used methods such as adsorption measurement and mercury porosimetry all rely on secondary data and mathematical models to estimate pore characteristics. But these methods are fraught with a number of drawbacks. In these methods, only those pores can be characterized that the adsorbate molecules could have access to. Also the mathematical models of adsorption which are the basis of estimating pore characteristics are based on assumptions most of which do not match with observations. Adsorption measurements in particular have other drawbacks that affect the accuracy of parameters estimated. For instance, phenomena such as activated diffusion and gate effects introduce errors into adsorption based estimates of pore parameters. Equally, idealized pore models such as the cylindrical or slit-shape pore which are the basis of a number of methods for estimating average pore width of activated carbons are too simplistic in the face of the observed complex nature of porosity in activated carbons. Then again some of the pore parameters are not at all amenable to estimation by the indirect methods. Such parameters like pore shape, pore location and distribution, and pore orientation have all eluded estimation by the indirect methods. These parameters nonetheless have important consequences on modeling and prediction of performance characteristics of activated carbons and therefore are worth estimating or measuring.

Direct methods, particularly, microscopy offers an alternative approach to resolving most of the drawbacks of the indirect methods. Using microscopy, it is possible to observe the micro and submicro-features of activated carbons directly and therefore makes possible a proper qualitative and quantitative description of its characteristics [Ito & Aguiar, 2009; Daley et al., 1996; Tomlinson, et al., 1995; Hefter, J., 1987; Ball & McCartney, 1981]. There are two types of this method, namely, optical microscopy which has a resolution of about 1 µm and electron microscopy whose resolution is much greater and in the range of about 1.5 nm and which can achieve a magnification of about 2,000,000x. The SEM is one version of the electron microscopy which uses a beam of electrons to scan the surface of a specimen and makes possible the direct observation of its surface features at the micro and submicro levels. Due to the huge magnifications and impressive resolutions achievable with the SEM it has been used in many areas of science and industry, particularly in materials engineering, biological and medical sciences for the study and characterization of the micro-structure of substances. It use provides an avenue to resolve some of the yet unresolved issues in activated carbon porosity development and characterization. The SEM functions exactly as its optical counterparts except that it uses a focused beam of electrons instead of light to "image" a specimen and gain information about its structure and composition. The SEM can yield information about the topography (surface features of an object), morphology (shape and size of the particles making up the surface of an object), composition (the elements that the object is composed of and the relative amounts of these) and crystallographic information (how the atoms are arranged in the object). This ability makes the SEM a hugely useful instrument for the study of activated carbons. The topographic information attainable using the SEM allows that surface features such as pore characteristics, the description of which has been a major preoccupation of activated carbon chemists, to be studied and measured directly from SEM micrographs. Also SEM's ability to reveal compositional details of a specimen makes it a potent instrument for studying the surface chemistry of activated carbons. Other features of the SEM that make it a unique instrument for studying activated carbons include its ability to reveal details of a sample less than 1 nm in size. This means

How realistic is the often used slit-shaped pore model in describing activated carbons. Similarly, in characterizing activated carbons the often used methods such as adsorption measurement and mercury porosimetry all rely on secondary data and mathematical models to estimate pore characteristics. But these methods are fraught with a number of drawbacks. In these methods, only those pores can be characterized that the adsorbate molecules could have access to. Also the mathematical models of adsorption which are the basis of estimating pore characteristics are based on assumptions most of which do not match with observations. Adsorption measurements in particular have other drawbacks that affect the accuracy of parameters estimated. For instance, phenomena such as activated diffusion and gate effects introduce errors into adsorption based estimates of pore parameters. Equally, idealized pore models such as the cylindrical or slit-shape pore which are the basis of a number of methods for estimating average pore width of activated carbons are too simplistic in the face of the observed complex nature of porosity in activated carbons. Then again some of the pore parameters are not at all amenable to estimation by the indirect methods. Such parameters like pore shape, pore location and distribution, and pore orientation have all eluded estimation by the indirect methods. These parameters nonetheless have important consequences on modeling and prediction of performance

characteristics of activated carbons and therefore are worth estimating or measuring.

Direct methods, particularly, microscopy offers an alternative approach to resolving most of the drawbacks of the indirect methods. Using microscopy, it is possible to observe the micro and submicro-features of activated carbons directly and therefore makes possible a proper qualitative and quantitative description of its characteristics [Ito & Aguiar, 2009; Daley et al., 1996; Tomlinson, et al., 1995; Hefter, J., 1987; Ball & McCartney, 1981]. There are two types of this method, namely, optical microscopy which has a resolution of about 1 µm and electron microscopy whose resolution is much greater and in the range of about 1.5 nm and which can achieve a magnification of about 2,000,000x. The SEM is one version of the electron microscopy which uses a beam of electrons to scan the surface of a specimen and makes possible the direct observation of its surface features at the micro and submicro levels. Due to the huge magnifications and impressive resolutions achievable with the SEM it has been used in many areas of science and industry, particularly in materials engineering, biological and medical sciences for the study and characterization of the micro-structure of substances. It use provides an avenue to resolve some of the yet unresolved issues in activated carbon porosity development and characterization. The SEM functions exactly as its optical counterparts except that it uses a focused beam of electrons instead of light to "image" a specimen and gain information about its structure and composition. The SEM can yield information about the topography (surface features of an object), morphology (shape and size of the particles making up the surface of an object), composition (the elements that the object is composed of and the relative amounts of these) and crystallographic information (how the atoms are arranged in the object). This ability makes the SEM a hugely useful instrument for the study of activated carbons. The topographic information attainable using the SEM allows that surface features such as pore characteristics, the description of which has been a major preoccupation of activated carbon chemists, to be studied and measured directly from SEM micrographs. Also SEM's ability to reveal compositional details of a specimen makes it a potent instrument for studying the surface chemistry of activated carbons. Other features of the SEM that make it a unique instrument for studying activated carbons include its ability to reveal details of a sample less than 1 nm in size. This means that, in principle, even micropores could be exposed for study by SEM micrographs. Again, due to the very narrow electron beam employed, SEM micrographs have a large depth of field that yields a pseudo three-dimensional appearance useful for understanding the surface structure of a sample. Figures 4-7 are examples of SEM micrographs that reveal details of activated carbons that only the direct methods can show.

Fig. 4. Micrograph of outer surface of raw coconut shell Source: *Achaw & Afrane, 2008*

Fig. 5. Micrograph of outer surface of carbonized coconut shell. Source: *Achaw & Afrane, 2008*

A Study of the Porosity of Activated Carbons Using the Scanning Electron Microscope 485

is a network of cracks which are more the result of thermal stress on the carbon matrix. This is totally in contrast to what was observed of the micrograph in Figure 3 which is more likely the result of re-arrangement of molecules in the carbon matrix. The observed porosity in Figure 6 further show the difficulty of defining a generalized pore structure for activated carbons. It calls to question such often used concepts as average pore size, and pore size distribution. Further, it raises questions about models associated with pore structure such as the slit-shaped model which is the basis of a number of mathematical models of adsorption in activated carbons or the cylindrical models used by mercury porosimetry and other empirical methods for estimating pore characteristics. Figures 6 & 7 are micrographs of different sections of an activated carbon. The two images clearly demonstrate the extent of inhomogeneity of the surfaces of activated carbons. Equally noteworthy of the micrographs in Figures 4-6 is the fact that there is hardly any common trend linking the structures in these micrographs. Whilst, hardly, any pore is observable at all in Figure 4, the structure in Figure 5 is fuzzy and confusing, and hardly yielding to any definition at all. Finally, even though the structure in Figure 6 has some semblance of order, it also defies any exact definition. An important observation of these micrographs is that pore development in activated carbons is as a result of several phenomena. Particularly, it seems that thermal stress plays an important role in pore development in these materials than previously thought. The foregoing observations demonstrate the strength of SEM in studying activated

The main features of the SEM are an electron source which provides the electrons that interact with the material to be examined, an arrangement of metal apertures, magnetic lenses and scanning coils or deflectors plates that confines, focuses and turns the beam of electrons into a thin and focused monochromatic beam which is accelerated towards the sample and which irradiates the specimen in a raster fashion [Goldstein, J et al., 2003; Reimer, L, 1998]. The interaction of the electrons with the specimen initiates a number of reactions inside the sample which results in the generation of signals which are taken advantage of to gain information about the sample. The SEM imaging process involves four major steps. These include sample preparation, the specimen scanning process, image formation and image analysis. The kind of preparation required of the sample depends on whether it is electrically conducting or not. Electrically conducting samples, for instance metals, only require minimal sample preparation prior to mounting on a sample stub for scanning and imaging. Non-conductive specimens such as activated carbons, however, must first be made conducting before mounting for study. Otherwise, these tend to charge when scanned by the electron beam leading to scanning faults and other image artifacts. Nonconducting samples are therefore first sputter coated with an ultra-thin coating of an electrically-conducting material before imaging. Other reasons for coating the sample surface are to increase the signal and surface resolution, especially with samples of low atomic number. Some of the commonly used materials for coating samples include gold, graphite, platinum, chromium, tungsten, osmium, and indium. For biological materials it is possible to increase the conductivity without coating by impregnating them with osmium before imaging. It is also possible to image non-conducting specimen without coating by using the Environmental SEM(ESEM) or the field emission gun (FEG) SEM [Schatten & Pawley, 2008; Hardt, T. A. 1999;]. Samples for SEM study do not need to be as thin as it is

carbons.

**4. The scanning electron microscope** 

Fig. 6. Micrograph of outer surface of coconut shell-based activated carbon *Source: Achaw & Afrane, 2008*

Fig. 7. A Micrograph of transverse section of coconut shell-based activated carbon *Source: This study*

The micrographs in Figures 4-7 are all of coconut shells at different stages during the preparation of activated carbons. Figures 4-6 are micrographs of the surfaces of the outer sections of the shell. The surface features observed from these micrographs are significantly different from those seen from the corresponding micrographs of on Figures 1-3, indicating that the nature of pores seen of activated carbons depend on the sections used. Crucially, the nature of porosity observed in Figure 6 is totally different from any description of porosity previously described except in the work of Achaw and Afrane (2008). What is observed here is a network of cracks which are more the result of thermal stress on the carbon matrix. This is totally in contrast to what was observed of the micrograph in Figure 3 which is more likely the result of re-arrangement of molecules in the carbon matrix. The observed porosity in Figure 6 further show the difficulty of defining a generalized pore structure for activated carbons. It calls to question such often used concepts as average pore size, and pore size distribution. Further, it raises questions about models associated with pore structure such as the slit-shaped model which is the basis of a number of mathematical models of adsorption in activated carbons or the cylindrical models used by mercury porosimetry and other empirical methods for estimating pore characteristics. Figures 6 & 7 are micrographs of different sections of an activated carbon. The two images clearly demonstrate the extent of inhomogeneity of the surfaces of activated carbons. Equally noteworthy of the micrographs in Figures 4-6 is the fact that there is hardly any common trend linking the structures in these micrographs. Whilst, hardly, any pore is observable at all in Figure 4, the structure in Figure 5 is fuzzy and confusing, and hardly yielding to any definition at all. Finally, even though the structure in Figure 6 has some semblance of order, it also defies any exact definition. An important observation of these micrographs is that pore development in activated carbons is as a result of several phenomena. Particularly, it seems that thermal stress plays an important role in pore development in these materials than previously thought. The foregoing observations demonstrate the strength of SEM in studying activated carbons.

#### **4. The scanning electron microscope**

484 Scanning Electron Microscopy

Fig. 6. Micrograph of outer surface of coconut shell-based activated carbon

Fig. 7. A Micrograph of transverse section of coconut shell-based activated carbon

The micrographs in Figures 4-7 are all of coconut shells at different stages during the preparation of activated carbons. Figures 4-6 are micrographs of the surfaces of the outer sections of the shell. The surface features observed from these micrographs are significantly different from those seen from the corresponding micrographs of on Figures 1-3, indicating that the nature of pores seen of activated carbons depend on the sections used. Crucially, the nature of porosity observed in Figure 6 is totally different from any description of porosity previously described except in the work of Achaw and Afrane (2008). What is observed here

*Source: Achaw & Afrane, 2008*

*Source: This study*

The main features of the SEM are an electron source which provides the electrons that interact with the material to be examined, an arrangement of metal apertures, magnetic lenses and scanning coils or deflectors plates that confines, focuses and turns the beam of electrons into a thin and focused monochromatic beam which is accelerated towards the sample and which irradiates the specimen in a raster fashion [Goldstein, J et al., 2003; Reimer, L, 1998]. The interaction of the electrons with the specimen initiates a number of reactions inside the sample which results in the generation of signals which are taken advantage of to gain information about the sample. The SEM imaging process involves four major steps. These include sample preparation, the specimen scanning process, image formation and image analysis. The kind of preparation required of the sample depends on whether it is electrically conducting or not. Electrically conducting samples, for instance metals, only require minimal sample preparation prior to mounting on a sample stub for scanning and imaging. Non-conductive specimens such as activated carbons, however, must first be made conducting before mounting for study. Otherwise, these tend to charge when scanned by the electron beam leading to scanning faults and other image artifacts. Nonconducting samples are therefore first sputter coated with an ultra-thin coating of an electrically-conducting material before imaging. Other reasons for coating the sample surface are to increase the signal and surface resolution, especially with samples of low atomic number. Some of the commonly used materials for coating samples include gold, graphite, platinum, chromium, tungsten, osmium, and indium. For biological materials it is possible to increase the conductivity without coating by impregnating them with osmium before imaging. It is also possible to image non-conducting specimen without coating by using the Environmental SEM(ESEM) or the field emission gun (FEG) SEM [Schatten & Pawley, 2008; Hardt, T. A. 1999;]. Samples for SEM study do not need to be as thin as it is

A Study of the Porosity of Activated Carbons Using the Scanning Electron Microscope 487

of the specimen. In the study of activated carbons however, qualitative data though useful as for instance in understanding pore development phenomena, it is quantitative data on pore structure- pore sizes, pore shapes, pore surface area, and pore size distribution- that are most useful for characterization and modeling of performance and behavior. Getting quantitative data from 2D SEM images of activated carbons, however, poses a number of challenges. First, 2D SEM cannot determine three-dimensional porosity of the materials because the conventional SEM is unable to observe images of inner parts of a specimen. Then also there are the difficulties associated with getting precise descriptions or a representative pore structure in view of the otherwise complex, varied and numerous pores in the field of view of the microscope. To get quantitative data of a specimen from conventional SEM images, the practice is to convert the 2D images into 3D from which the requisite quantitative data can be measured or estimated often with the help of computerized image analysis software. A number of methods for getting 3D data from 2D images are available [Marinello et al., 2008; Spowart, 2006; Spowart et al. 2003; Alkemper and Voorhees, 2001; Lyroudia et al., 2000 ]. These include stereology, photogrammetry, photometric stereo, and the more useful serial sectioning method. An automated variation of the serial sectioning method called the focus ion beam-scanning electron microscopy (FIB-SEM) is increasingly being used in the areas of materials engineering and biological sciences to study the micro-features of substances. This last method appears to have enormous

potential for use in the study and characterization of activated carbons.

In serial sectioning, 2D images of a sample are collected after a series of successive layers of equal width have been removed from the sample. Afterwards the stack of 2D data files (2D SEM images) is combined and processed in such a way that the microstructural features that are within the 3D data stack can be classified. This is most efficiently done using computerized image processing software. Serial sectioning is made up of two basic steps that are iteratively repeated until completion of the experiment. In the first step a nominally flat surface of the sample is prepared using any of a variety of methods such as cutting, polishing, ablating, etching, or sputtering. These processes remove a constant depth of material from the specimen between sections. The second step is to collect 2D characterization data after each section has been prepared, for instance by imaging with an SEM. Finally, computer software programs are used to construct a 3D array of the characterization data that can be subsequently rendered as an image or analyzed for morphological or topographical features of the sample. Using this method any micro and submicro features of the sample that can be distinguished by the SEM can also be characterized. Thus the capability of the methodology in characterizing topographical and

morphological features of substances is limited only by the resolution of the SEM.

carbons.

Automated serial sectioning techniques which facilitates the methodology exist for material removal and imaging. An example is the FIB-SEM which combines ion beam sectioning with SEM imaging to generate tomographic data that are well suited to characterize microstructural features of a sample in 3D via serial sectioning [Desbois et al., 2009; Orloff et al. 2003]. While less common, the FIB-SEM method has demonstrated the ability to complete 3D volumetric reconstruction at a resolution of 10 nm or better in all three dimensions. The method has been widely used for studies in materials engineering and life sciences and holds tremendous potential for characterizing the porosity of activated

the case in optical microscopy or transmission electron microscopy (TEM). The specimen size is dictated primarily by the size of the sample chamber and must be rigidly mounted on the specimen stub. Again, in view of the applied vacuum, it is important that the sample be completely dry. Hard, dry materials such as activated carbons can be examined with little further treatment besides sputter coating.

The signals of interest in the characterization of activated carbons using the SEM are the secondary electrons whose detection and imaging gives information on the surface topography and hence on the pore structure of the activated carbons, the backscattered electrons and the X-ray radiation which give complementary information of the chemical composition of the sample surface. Whilst secondary electrons can be detected and imaged in a conventional SEM instrument, harnessing signals from the backscattered electrons and X-ray radiation requires complementary instrumentation. The secondary electrons are produced when an incident electron in the primary electron beam excites an electron in an atom of the sample and loses most of its energy in the process. The excited electron moves to the surface of the sample where it can escape if it possesses sufficient energy. In view of their low energy, only secondary electrons that are very near the surface (<10 nm) can exit the sample and be detected. The secondary electrons are detected and accelerated onto a photomultiplier from which an amplified electrical signal output is displayed as a twodimensional intensity distribution that can be viewed and photographed on an analogue video display or converted and displayed or stored as a digital image.

Backscattered electrons (BSE) are produced when the primary electron beam hits the sample and some of the electrons are reflected or scattered back out of the specimen. The production of backscattered electrons varies directly with the specimen's atomic number. High atomic number elements backscatter electrons more strongly than low atomic number elements, and thus appear brighter in an image. BSE are therefore used to detect contrast between areas on the sample surface with different chemical compositions. This provides opportunity for examination of the chemistry of the surface of activated carbons. Dedicated backscattered electron detectors, usually either scintillator or semiconductor types are positioned above the sample to detect the backscattered electrons.

As earlier intimated, inelastic scattering, places the atoms of the sample in an excited state. The tendency therefore is for an atom to return to its ground or unexcited state. To achieve this, the atom gives off the excess energy. This may result in the production of X-rays, cathodoluminescence and Auger electrons. The relaxation energy is the fingerprint of each element in the sample. Thus detection and analysis of the relaxation energies enables the identification of the specific elements in the surface of a sample. When an SEM is equipped with energy-dispersive X-ray spectroscopy (EDX) or wavelength dispersive X-ray spectroscopy (WDS), it is possible to get information on the elemental composition on the surface of the specimen. For an activated carbon, this method can be used to study the effect on porosity development of specific elements in the carbon and generally the chemistry of the activated carbon surface [Afrane & Achaw, 2008].

#### **4.1 3D images in scanning electron microscope**

In conventional (standard) SEM a pseudo-three dimensional (3D) view of the sample surface can be observed directly. The standard SEM image is, however, really a two-dimensional (2D) structure from which mostly qualitative data is possible regarding the microstructure

the case in optical microscopy or transmission electron microscopy (TEM). The specimen size is dictated primarily by the size of the sample chamber and must be rigidly mounted on the specimen stub. Again, in view of the applied vacuum, it is important that the sample be completely dry. Hard, dry materials such as activated carbons can be examined with little

The signals of interest in the characterization of activated carbons using the SEM are the secondary electrons whose detection and imaging gives information on the surface topography and hence on the pore structure of the activated carbons, the backscattered electrons and the X-ray radiation which give complementary information of the chemical composition of the sample surface. Whilst secondary electrons can be detected and imaged in a conventional SEM instrument, harnessing signals from the backscattered electrons and X-ray radiation requires complementary instrumentation. The secondary electrons are produced when an incident electron in the primary electron beam excites an electron in an atom of the sample and loses most of its energy in the process. The excited electron moves to the surface of the sample where it can escape if it possesses sufficient energy. In view of their low energy, only secondary electrons that are very near the surface (<10 nm) can exit the sample and be detected. The secondary electrons are detected and accelerated onto a photomultiplier from which an amplified electrical signal output is displayed as a twodimensional intensity distribution that can be viewed and photographed on an analogue

Backscattered electrons (BSE) are produced when the primary electron beam hits the sample and some of the electrons are reflected or scattered back out of the specimen. The production of backscattered electrons varies directly with the specimen's atomic number. High atomic number elements backscatter electrons more strongly than low atomic number elements, and thus appear brighter in an image. BSE are therefore used to detect contrast between areas on the sample surface with different chemical compositions. This provides opportunity for examination of the chemistry of the surface of activated carbons. Dedicated backscattered electron detectors, usually either scintillator or semiconductor types are

As earlier intimated, inelastic scattering, places the atoms of the sample in an excited state. The tendency therefore is for an atom to return to its ground or unexcited state. To achieve this, the atom gives off the excess energy. This may result in the production of X-rays, cathodoluminescence and Auger electrons. The relaxation energy is the fingerprint of each element in the sample. Thus detection and analysis of the relaxation energies enables the identification of the specific elements in the surface of a sample. When an SEM is equipped with energy-dispersive X-ray spectroscopy (EDX) or wavelength dispersive X-ray spectroscopy (WDS), it is possible to get information on the elemental composition on the surface of the specimen. For an activated carbon, this method can be used to study the effect on porosity development of specific elements in the carbon and generally the chemistry of

In conventional (standard) SEM a pseudo-three dimensional (3D) view of the sample surface can be observed directly. The standard SEM image is, however, really a two-dimensional (2D) structure from which mostly qualitative data is possible regarding the microstructure

video display or converted and displayed or stored as a digital image.

positioned above the sample to detect the backscattered electrons.

the activated carbon surface [Afrane & Achaw, 2008].

**4.1 3D images in scanning electron microscope** 

further treatment besides sputter coating.

of the specimen. In the study of activated carbons however, qualitative data though useful as for instance in understanding pore development phenomena, it is quantitative data on pore structure- pore sizes, pore shapes, pore surface area, and pore size distribution- that are most useful for characterization and modeling of performance and behavior. Getting quantitative data from 2D SEM images of activated carbons, however, poses a number of challenges. First, 2D SEM cannot determine three-dimensional porosity of the materials because the conventional SEM is unable to observe images of inner parts of a specimen. Then also there are the difficulties associated with getting precise descriptions or a representative pore structure in view of the otherwise complex, varied and numerous pores in the field of view of the microscope. To get quantitative data of a specimen from conventional SEM images, the practice is to convert the 2D images into 3D from which the requisite quantitative data can be measured or estimated often with the help of computerized image analysis software. A number of methods for getting 3D data from 2D images are available [Marinello et al., 2008; Spowart, 2006; Spowart et al. 2003; Alkemper and Voorhees, 2001; Lyroudia et al., 2000 ]. These include stereology, photogrammetry, photometric stereo, and the more useful serial sectioning method. An automated variation of the serial sectioning method called the focus ion beam-scanning electron microscopy (FIB-SEM) is increasingly being used in the areas of materials engineering and biological sciences to study the micro-features of substances. This last method appears to have enormous potential for use in the study and characterization of activated carbons.

In serial sectioning, 2D images of a sample are collected after a series of successive layers of equal width have been removed from the sample. Afterwards the stack of 2D data files (2D SEM images) is combined and processed in such a way that the microstructural features that are within the 3D data stack can be classified. This is most efficiently done using computerized image processing software. Serial sectioning is made up of two basic steps that are iteratively repeated until completion of the experiment. In the first step a nominally flat surface of the sample is prepared using any of a variety of methods such as cutting, polishing, ablating, etching, or sputtering. These processes remove a constant depth of material from the specimen between sections. The second step is to collect 2D characterization data after each section has been prepared, for instance by imaging with an SEM. Finally, computer software programs are used to construct a 3D array of the characterization data that can be subsequently rendered as an image or analyzed for morphological or topographical features of the sample. Using this method any micro and submicro features of the sample that can be distinguished by the SEM can also be characterized. Thus the capability of the methodology in characterizing topographical and morphological features of substances is limited only by the resolution of the SEM.

Automated serial sectioning techniques which facilitates the methodology exist for material removal and imaging. An example is the FIB-SEM which combines ion beam sectioning with SEM imaging to generate tomographic data that are well suited to characterize microstructural features of a sample in 3D via serial sectioning [Desbois et al., 2009; Orloff et al. 2003]. While less common, the FIB-SEM method has demonstrated the ability to complete 3D volumetric reconstruction at a resolution of 10 nm or better in all three dimensions. The method has been widely used for studies in materials engineering and life sciences and holds tremendous potential for characterizing the porosity of activated carbons.

A Study of the Porosity of Activated Carbons Using the Scanning Electron Microscope 489

Desbois, G, Urai, J. L., & Kukla, P. A. (2009). Morphology of the pore space in claystones –

*Discuss.,* 4, 1–19, 2009, www.electronic-earth-discuss.net/4/1/2009/ Dubnin, M. M. (1989). Fundamentals of the theory of adsorption in micropores of carbon

Academic /Plenum Publishers, New York, ISBN 978-0-306-47292-3

2nd Edition, London, 1991, ISBN 0123009561

*Scanning* Vol. 19, pp. 426 - 430, ISSN: 1932-8745

7923-5939-9

1236, ISSN: 0008-6223

1- 12, ISSN 0957-0233

Gregg, S. J. and Sing, K. S. W. (1991). *Adsorption, Surface Area and Porosity*. Academic Press,

Hardt, T. A. (1999). Environmental SEM and related applications. In Impact of electron and

Hefter, J. (1987). Morphological characterizations of materials using low voltage scanning

Heschel, W. & Klose, E. (1995). On suitability of agricultural by-products for the manufacture of granular activated carbon. *Fuel,* Vol. 74, No. 12 , ISSN: 0016-2361 Ito, L. X. & Aguiar, M. L. (2009). A study of the porosity of gas filtration cakes. *Brazilian Journalof Chemical Engineering.* Vol. 26, No. 02, pp. 307 – 315. ISSN 0104-6632 Joshi, V. C., Khan, I. A., &. Sharaf, M. H. M. (2008). Use of Scanning Electron Microscopy in

Kamran, M. N. (1997).Fracture analysis of concrete using scanning electron microscopy*.* 

Lozano-Castello, D., Cazorla-Amoros,D., Linares-Solano, A. (2004).Usefulness of CO2

Lyroudia, K, Pantelidou, O. Mikrogeorgis, G., Nikopoulos, N., & Pitas,L. (2000). Three-

Md Noor., A. A. B. & Nawi, A. B. M. (2008).. Textural characteristics of activated carbons

electron microscopy. *Scanning Microsc.* 1(l), 13-21, ISSN: 1932-8745

evidence from BIB/FIB ion beam sectioning and cryo-SEM observations. *eEarth* 

adsorbents: Characteristics of their adsorption properties and microporous structures. *Pure & Applied Chemistry,* Vol. 10, pp. 1841 – 1843, ISSN 0033-4545 Evans, M. & Marsh, H. (1979). Origins of microporosity, mesoporosity, and macroporosity

in carbons and graphites, in *Characterization of Porous Solids*, Unger, K. K., Rouguerol, J., Sing, K. S. W., Kral, H., (eds). London, SCI, ISBN 0 - 444 – 42953 – Farkas, L., Major, N., Mihalko, A., Abraham, J. & Kozar, Z. (2009). Analysis of active carbon catalysis. *Material Sciences and Engineering, Miskolc,* Volume 34/2 , pp 41-51 Gil, A. (1998). Analysis of the micropore structure of various microporous materials from nitrogen adsorption at 77K. *Adsorption,* Vol. 4, Numbers 3-4, pp 197-206 Goldstein, J., Newbury D. E., Joy, D. C, Lyman, C. E., Echlin, P., Lifshin, E., Sawyer, L., and

Michael, J. R. (2003). *Scanning Electron Microscopy and X-Ray Microanalysis*. Kluwer

scanning probe microscopy on materials research. (eds). Rickerby, D. G. Giovanni Valdrè, G., & Valdrè, U, Kluwer Academic Publishers, Netherlands, p. 407, ISBN 0-

the Authentication of Botanicals. *Pharmacopeial Forum. Vol. 34, No.4, ISBN 0363-4655*

adsorption at 273 K for the characterization of porous carbons .*Carbon 42*, pp. 1231–

dimensional reconstruction: A new method for the evaluation of apical microleakage. *Journal of Endodontics, Vol. 26, No. 1, pp. 36- 38*, ISSN: 0099-2399 Machnikowski, J., Kierzek, K., Lis, K., Machnikowska, H., and Czepirski, L. (2010).Tailoring

porosity development in monolithic adsorbents made of KOH-activated pitch coke and furfuryl alcohol binder for methane storage. *Energy Fuels ,* 24, 3410–3414, Marinello, F., Bariani, P., Savio, E., Horsewell, A., & De Chiffre, L. (2008). Critical factors in

SEM 3D stereo microscopy. *Measurement Science and. Technology.* Vol. 19, No. 6, pp.

prepared from oil palm shells activated with ZnCl2 and pyrolysis under nitrogen

## **5. Conclusion**

In spite of the tremendous progress in development and use of activated carbons a number of questions still remain that the conventional indirect methods of studying activated carbon are still grappling to answer. The mode of porosity development is one such area. Another area is the characterization of the porosity of activated carbons, where the existing mathematical models and methods have still not succeeded in finding accurate ways to estimate pore parameters. Recent developments in scanning electron microscopy, especially in the conversion of 2D SEM to 3D and computerized image analysis has opened avenues for improved study of porosity development and characterization of activated carbon. This potential of the SEM has not really been adequately explored for the study of activated carbons as not much work exits in the literature in that regard. However, judging from the enormous strides researchers in the areas of materials engineering and the biological sciences have made in using this methodology to identify micro and submicro-features of substances, it is anticipated that its adaptation for use in the study of activated carbons would facilitate the study of porosity development and pore characterization. The SEM micrographs shown in this work clearly demonstrate this point. The major limitation of the SEM is the level of resolution achievable with it currently. At 1.5 nm, this poses difficulty in characterizing most micropores in activated carbons. It is nonetheless hoped that continued advances in SEM instrumentation will overcome this difficulty and facilitate the use of the SEM in the study of activated carbons.

## **6. References**


In spite of the tremendous progress in development and use of activated carbons a number of questions still remain that the conventional indirect methods of studying activated carbon are still grappling to answer. The mode of porosity development is one such area. Another area is the characterization of the porosity of activated carbons, where the existing mathematical models and methods have still not succeeded in finding accurate ways to estimate pore parameters. Recent developments in scanning electron microscopy, especially in the conversion of 2D SEM to 3D and computerized image analysis has opened avenues for improved study of porosity development and characterization of activated carbon. This potential of the SEM has not really been adequately explored for the study of activated carbons as not much work exits in the literature in that regard. However, judging from the enormous strides researchers in the areas of materials engineering and the biological sciences have made in using this methodology to identify micro and submicro-features of substances, it is anticipated that its adaptation for use in the study of activated carbons would facilitate the study of porosity development and pore characterization. The SEM micrographs shown in this work clearly demonstrate this point. The major limitation of the SEM is the level of resolution achievable with it currently. At 1.5 nm, this poses difficulty in characterizing most micropores in activated carbons. It is nonetheless hoped that continued advances in SEM instrumentation will overcome this difficulty and facilitate the use of the

Achaw, O-W. & Afrane, G. (2008). The evolution of pore structure of coconut shells during

Afrane, G. & Achaw, O-W. (2008). Effect of the concentration of inherent mineral elements

Alkemper, J., Voorhees P. W. (2001). Quantitative serial sectioning analysis. *Journal of* 

Ball, M. D. & McCartney, D. G. (1981). The measurement of atomic number and composition in a SEM using backscattered detectors. *J. Microsc.* 124, 57-68, ISSN: 1365-2818 Bryne, J. F. & Marsh, H. (1995). Origins and structure of porosity, in John W. Patrick (ed),

Carrasco-Marin, F., Alvarez Merino, M. A., & Moreno-Castilla, C. (1996). Microporous

Chung, W., Sharifi, V. N., Swithenbank J, Osammor, O., Nolan, A. (2008).. Characterisation

Dalye, M. A., Tandon, D., Economy, J. abd Hippi, E. J. (1996). Elucidating the porous

*Technology,* Vol. 99, No. 14, pp. 6678 – 6682, ISSN 0960-8524

*Microscopy*, Vol. 201, No.3, pp. 388–394, *ISSN: 1365-2818* 

the preparation of cocnut shell-based activated carbons. *Microporous and Mesoporous* 

on the adsorption capacity of coconut shell-based activated carbons, *Bioresource* 

*Porosity in carbons: Characterization and applications*, Edward Arnold, London, ISBN 0

activated carbons from a bituminous coal. *Fuel,* Vol. 75, No. 8, pp. 966-970, ISSN:

of Airborne Particulate Matter in a City Environment. *Modern Applied Science,* Vol.2,

structure of activated carbon fibers using direct and indirect methods. *Carbon,* Vol.

**5. Conclusion** 

SEM in the study of activated carbons.

*Materials*, 112, pp. 284 – 290, ISSN 1387-1811

**6. References** 

340544732

0016-2361

No. 4, p.17, ISSN 1913 – 1844

34, No. 10, pp 1191-1200, ISSN: 0008-6223


**25** 

Laura Frisk

*Finland* 

**Study of Structure and Failure Mechanisms** 

The trends in the electronics industry have for several decades been for smaller size combined with greater functionality. One enabler for this trend has been development of new packaging solutions which has required the development of new materials and also new interconnections technologies. In the development of these technologies it has been essential to have effective tools to study the structure of the packages and their failure mechanisms. Due to the versatility of electronics packages concerning materials, structures, and functions a plethora of different methods have been used. These include for example electrical characterization technologies, x-ray, scanning electron microscopy (SEM), scanning acoustic microscopy (SAM), optical microscopy, differential scanning calorimetry (DCS), and thermomechanical analysis (TMA) (Chan et al., 2000; Jang et al., 2008; Yim &

This chapter concentrates on flip chip technology, which is one of the technologies developed to miniaturise an electronics package. In this technology a bare chip is attached directly onto a substrate without wiring needed to connect the chip. As the attachment is done with the active side of the chip towards the substrate, the chip is flipped before bonding. Hence the name flip chip. Flip chip technology has several advantages as it enables the production of small, very high density packages. In addition, it has good electrical performance due to the short interconnection path (Lau, 2000). This method can be used to attach a chip directly onto a substrate, but also as an attachment method in single-chip packages, such as a ball grid array (BGA) and a chip scale package (CSP), and in multi-chip

One problem with this technology is that the quality of the joints is relatively difficult to assess. Semiconductor chips used in this technology may have hundreds of contacts which form an area array below the chip. In order to study the joints a cross-sectioning is often needed. Although cross-sections give lots of valuable information, they are restricted to very small area of a package and thereby often several cross-sections are needed. Additionally, the information of the interconnections may need to be increased using other techniques. For example scanning acoustic microscopy may be used to study the amount of delamination in a package. The cross-sections may be studied by optical microscopy. However, for detailed

information scanning electron microscopy is the preferred method of analysis.

**1. Introduction** 

Paik, 2001).

modules.

**in ACA Interconnections Using SEM** 

*Tampere University of Technology, Department of Electronics,* 

and carbon dioxide. *Journal of Physical Science,* Vol. 19(2), 93–104, 2008, ISSN: 1675- 3402


## **Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM**

Laura Frisk

*Tampere University of Technology, Department of Electronics, Finland* 

## **1. Introduction**

490 Scanning Electron Microscopy

Mfanacho, S. M. , Hemang, P. & Manocha, L. M. (2010). Enhancement of microporosity

Orloff, J., Utlaut, M., Swanson, L., (2003). *High Resolution Focused Ion Beams: FIB and Its Applications*. Kluwer Academic/Plenum*,* New York, ISBN 0-306-47350-X Raveendran, K., Ganesh, A., Khilar, K. C. (1995). Influence of mineral matter on boimasspyrolysis characteristics. *Fuel* Vol. 74, No. 12, ISSN: 0016-2361 Reimer, L. (1998). *Scanning Electron Microscopy: Physics of Image Formation and Microanalysis*,

Rigby, S. P. & Edler, K. J. (2002). The influence of mercury contact angle, surface tension,

Rodriguez-Reinoso, F. & Linares-Solano, A. (1989). In 'Chemistry and Physics of Carbon', . P. A. Thrower (ed), Dekker, Vol. 21 , New York, p.1, ISBN 0824781139 Rodriguez-Reinoso, F., An overview of methods of the characterization of activated carbons. *Pure & Applied Chemistry,* Vol. 61, No. 11, pp 1859 – 1867, 1989, ISSN 0033-4545 Schatten, H & Pawley, J. (2008). *Biological low voltage field emission scanning electron microscopy. Springer Science + Business Media, New York,* e-ISBN 978-0-387-72972-5

Sing., K. S. W., Everett, D. H., Haul, R. A. W., Moscou, L., Pierotti, R. A., Rouquerol, J. & Siemieniewska. (1985). Reporting on physiosorption data for gas/solid systems with special

Spowart, J. E. (2006). Automated serial sectioning for 3-D analysis of microstructures. *Scripta* 

Spowart, J. E., Mullens H. M., Puchala, B. T. (2003) Collecting and analyzing

Stoeckli, F., Guillot, A., Slasli, A. M.,a, and Hugi-Cleary, D. (2002). Microporosity in carbon

Stoeckli, F., Guillot, A., Slasli, A. M. & Hugi-Cleary, D. (2002). The comparison of

Tomlinson, J. B., Freeman, J. J., Sing, K. S. W., and Theocaris, C. R. (1995). Rates of activation

Yuna, C. H., Parka, Y. H., Park, C. R.,. (2001). Effects of pre-carbonization on porosity

and retraction mechanism on the interpolation of mercury porosimetry data.

reference to determination of surface and porosity. *Pure & Applied Chemistry,* Vol

microstructures in three dimensions: A fully automated approach. *JOM Journal of the Minerals, Metals and Materials Society,* Vol. 55, No. 10, pp. 35–37, ISSN 1047-4838

experimental and calculated pore size distributions of activated carbons. *Carbon* 

and scanning electron microscopy of polyaryllamide-derived chars. Carbon, Vol.

development of activated carbons from rice straw, *Carbon 39,* pp.559–567, ISSN:

3402

pp. 106 – 109, ISSN 0975 – 2595

57, p 603, ISSN 0033-4545

Springer-Verlag, New York, ISBN 3-540-63976-*4*

*Materialia,* Vol. 55, No.1, pp. 5–10, ISSN: 1359-6462

blacks. *Carbon 40, i*ssue 2, pp. 211-215, ISSN: 0008-6223

Vol. 40, No. 3, pp. 383-388, ISSN: 0008-6223

33, No. 6, pp. 789-793, ISSN: 0008-6223

0008-6223

Journal of Colloid and Interface Science 250, pp. 175-190

and carbon dioxide. *Journal of Physical Science,* Vol. 19(2), 93–104, 2008, ISSN: 1675-

through physical activation. PRAJÑĀ - *Journal of Pure and Applied Sciences,* Vol. 18,

The trends in the electronics industry have for several decades been for smaller size combined with greater functionality. One enabler for this trend has been development of new packaging solutions which has required the development of new materials and also new interconnections technologies. In the development of these technologies it has been essential to have effective tools to study the structure of the packages and their failure mechanisms. Due to the versatility of electronics packages concerning materials, structures, and functions a plethora of different methods have been used. These include for example electrical characterization technologies, x-ray, scanning electron microscopy (SEM), scanning acoustic microscopy (SAM), optical microscopy, differential scanning calorimetry (DCS), and thermomechanical analysis (TMA) (Chan et al., 2000; Jang et al., 2008; Yim & Paik, 2001).

This chapter concentrates on flip chip technology, which is one of the technologies developed to miniaturise an electronics package. In this technology a bare chip is attached directly onto a substrate without wiring needed to connect the chip. As the attachment is done with the active side of the chip towards the substrate, the chip is flipped before bonding. Hence the name flip chip. Flip chip technology has several advantages as it enables the production of small, very high density packages. In addition, it has good electrical performance due to the short interconnection path (Lau, 2000). This method can be used to attach a chip directly onto a substrate, but also as an attachment method in single-chip packages, such as a ball grid array (BGA) and a chip scale package (CSP), and in multi-chip modules.

One problem with this technology is that the quality of the joints is relatively difficult to assess. Semiconductor chips used in this technology may have hundreds of contacts which form an area array below the chip. In order to study the joints a cross-sectioning is often needed. Although cross-sections give lots of valuable information, they are restricted to very small area of a package and thereby often several cross-sections are needed. Additionally, the information of the interconnections may need to be increased using other techniques. For example scanning acoustic microscopy may be used to study the amount of delamination in a package. The cross-sections may be studied by optical microscopy. However, for detailed information scanning electron microscopy is the preferred method of analysis.

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 493

Several materials can be used in ICAs. The most widely used ICAs in the electronics industry are silver-filled epoxies, which also provide a high level of thermal conductivity. The popularity of epoxies is due to their excellent properties as a conductive adhesive. They have good adhesive strength, thermal stability and dielectric properties. Furthermore, they have good retention of these properties under thermomechanical stresses and under demanding conditions such as high humidity. However, other thermoset resins, such as silicones, cyanate esters, and cyanoacrylates, can also be used. Another option is thermoplastic resins. (Licari, 2005) Silver is the most commonly used filler material (Lau, 2003). The popularity of silver is due to its excellent conductivity and chemical stability

ICAs have been used in the electronics industry mainly as die-attach adhesives. However, lately they have also been proposed as an alternative to solders in surface mount and flip chip applications. For use in flip chip applications ICAs need to be carefully applied only on those areas which need to be conductive. Additionally, spreading of the adhesive should be prevented during the bonding process. A separate underfilling step is needed to improve the reliability of the joints. (Lau, 2003; Li, 2006) A typical cross section of an ICA flip chip

In an anisotropic conductive adhesive (ACA) the concentration of conductive particles is below the percolation threshold (Lau, 1995; Licari, 2005) and the adhesive does not conduct before the interconnection is formed. Typically the number of particles is 0.5% - 5 % by volume (Licari, 2005) but depends largely on the size and shape of the conducting particles and on the application the ACA is used in (Watanabe, 2004). Normally the particles are randomly dispersed in the matrix, but adhesives having uniformly dispersed particles have also been

During the ACA attachment process the adhesive is placed between the mating contacts. The ACA interconnection is established by applying pressure and heat simultaneously to the interconnection. When the temperature is raised the adhesive matrix will transform into low viscosity fluid (Tan et al., 2004), which allows excess adhesive to flow from the joints and fill the spaces around the contacts forming a physical connection between the parts to be attached. The conductive particles are trapped between the contacts and deform forming an electrical connection. As a result, electrical conduction is restricted to the z-direction and the electrical insulation in x-y directions is maintained. During cooling residual stresses are formed as a result of contraction of the adhesive matrix. In addition, residual stresses form

developed (Ishibashi & Kimura, 1996; Jin et al., 1993; Sungwook & Chappell, 2010).

(Morris 2005). Moreover, its oxide is highly conductive.

Fig. 1. Schematic illustration of an ICA flip chip joint with underfill.

joint is presented in Figure 1.

**2.2 Anisotropic conductive adhesives** 

Currently mainstream flip chip technology is based on solder bumps. These can be produced using both traditional tin-lead and new lead-free solders. However, mounting environmental concern has increased interest in electrically conductive adhesives, as they are environmentally friendly. In addition to being lead free, they can be used with substrate materials which do not withstand soldering temperatures. Thus they can be used to solve the problem caused by the high reflow temperature needed by most lead-free solders (Li & Wong, 2006).

Compared to solders adhesive materials are more complex as they are polymeric materials containing conductive particles. They have several advantages which makes their use profitable. However, due to their complex structure quality of the interconnections made by these materials needs to be determined carefully to attain good reliability. There are two types of electrically conductive adhesives. In isotropic conductive adhesives (ICA) the concentration of the conductive particles is high and they conduct in all directions. On the other hand, in anisotropic conductive adhesives (ACA) the concentration of conductive particles is low and the adhesive conducts in z-direction only after the bonding process. This chapter will concentrate on ACA materials used in flip chip applications. This chapter will discuss specifically how SEM may be utilised to study the quality and failure mechanisms of ACA interconnections.

## **2. Polymeric interconnections for electronics**

Electrically conductive adhesives used in electronics consist of polymer binder and conductive particles. Polymer resins used in these adhesives are inherently insulators. To obtain an electrically conductive adhesive they must therefore be filled with electrically conductive fillers, such as metal particles. In the following properties and materials of two main types of electrically conductive adhesives are discussed. Isotropic conductive adhesives (ICA) have high concentration of the particles and they conduct in every direction. These materials may be used to replace solders. If the concentration of conducting particles is low, an anisotropic conductive adhesive (ACA) is formed. ACAs conduct electrically only in a vertical direction and thereby may be used in very high density applications.

#### **2.1 Isotropic conductive adhesives**

Isotropically conductive adhesives are formed by adding enough conductive filler to a polymer matrix to transform it from an insulator into a conductor. This transformation has been explained by a percolation theory. When the concentration of conductive filler is increased, the resistivity of the adhesive drops dramatically above a critical concentration, and this is called the percolation threshold. It is believed that at this concentration the conductive particles contact each other forming a three dimensional network, which enables the conductivity. After the percolation threshold the resistivity decreases only slightly with increased concentration of the conductive filler. (Lau et al., 2003) The mechanical interconnection of an ICA joint is provided by the polymer matrix. If too high a concentration of the conductive filler is used, it may impair this interconnection. Thus the amount of conductive filler needs to be large enough to ensure good conductivity without sacrificing the mechanical properties of the adhesive (Lu, 2006). A typical volume fraction of the conductive filler is approximately 25 to 30 percent (Licari, 2005).

Currently mainstream flip chip technology is based on solder bumps. These can be produced using both traditional tin-lead and new lead-free solders. However, mounting environmental concern has increased interest in electrically conductive adhesives, as they are environmentally friendly. In addition to being lead free, they can be used with substrate materials which do not withstand soldering temperatures. Thus they can be used to solve the problem caused by the high reflow temperature needed by most lead-free solders (Li &

Compared to solders adhesive materials are more complex as they are polymeric materials containing conductive particles. They have several advantages which makes their use profitable. However, due to their complex structure quality of the interconnections made by these materials needs to be determined carefully to attain good reliability. There are two types of electrically conductive adhesives. In isotropic conductive adhesives (ICA) the concentration of the conductive particles is high and they conduct in all directions. On the other hand, in anisotropic conductive adhesives (ACA) the concentration of conductive particles is low and the adhesive conducts in z-direction only after the bonding process. This chapter will concentrate on ACA materials used in flip chip applications. This chapter will discuss specifically how SEM may be utilised to study the quality and failure mechanisms of

Electrically conductive adhesives used in electronics consist of polymer binder and conductive particles. Polymer resins used in these adhesives are inherently insulators. To obtain an electrically conductive adhesive they must therefore be filled with electrically conductive fillers, such as metal particles. In the following properties and materials of two main types of electrically conductive adhesives are discussed. Isotropic conductive adhesives (ICA) have high concentration of the particles and they conduct in every direction. These materials may be used to replace solders. If the concentration of conducting particles is low, an anisotropic conductive adhesive (ACA) is formed. ACAs conduct electrically only in a vertical direction and thereby may be used in very high density

Isotropically conductive adhesives are formed by adding enough conductive filler to a polymer matrix to transform it from an insulator into a conductor. This transformation has been explained by a percolation theory. When the concentration of conductive filler is increased, the resistivity of the adhesive drops dramatically above a critical concentration, and this is called the percolation threshold. It is believed that at this concentration the conductive particles contact each other forming a three dimensional network, which enables the conductivity. After the percolation threshold the resistivity decreases only slightly with increased concentration of the conductive filler. (Lau et al., 2003) The mechanical interconnection of an ICA joint is provided by the polymer matrix. If too high a concentration of the conductive filler is used, it may impair this interconnection. Thus the amount of conductive filler needs to be large enough to ensure good conductivity without sacrificing the mechanical properties of the adhesive (Lu, 2006). A typical volume fraction of

the conductive filler is approximately 25 to 30 percent (Licari, 2005).

Wong, 2006).

ACA interconnections.

applications.

**2.1 Isotropic conductive adhesives** 

**2. Polymeric interconnections for electronics** 

Several materials can be used in ICAs. The most widely used ICAs in the electronics industry are silver-filled epoxies, which also provide a high level of thermal conductivity. The popularity of epoxies is due to their excellent properties as a conductive adhesive. They have good adhesive strength, thermal stability and dielectric properties. Furthermore, they have good retention of these properties under thermomechanical stresses and under demanding conditions such as high humidity. However, other thermoset resins, such as silicones, cyanate esters, and cyanoacrylates, can also be used. Another option is thermoplastic resins. (Licari, 2005) Silver is the most commonly used filler material (Lau, 2003). The popularity of silver is due to its excellent conductivity and chemical stability (Morris 2005). Moreover, its oxide is highly conductive.

ICAs have been used in the electronics industry mainly as die-attach adhesives. However, lately they have also been proposed as an alternative to solders in surface mount and flip chip applications. For use in flip chip applications ICAs need to be carefully applied only on those areas which need to be conductive. Additionally, spreading of the adhesive should be prevented during the bonding process. A separate underfilling step is needed to improve the reliability of the joints. (Lau, 2003; Li, 2006) A typical cross section of an ICA flip chip joint is presented in Figure 1.

Fig. 1. Schematic illustration of an ICA flip chip joint with underfill.

## **2.2 Anisotropic conductive adhesives**

In an anisotropic conductive adhesive (ACA) the concentration of conductive particles is below the percolation threshold (Lau, 1995; Licari, 2005) and the adhesive does not conduct before the interconnection is formed. Typically the number of particles is 0.5% - 5 % by volume (Licari, 2005) but depends largely on the size and shape of the conducting particles and on the application the ACA is used in (Watanabe, 2004). Normally the particles are randomly dispersed in the matrix, but adhesives having uniformly dispersed particles have also been developed (Ishibashi & Kimura, 1996; Jin et al., 1993; Sungwook & Chappell, 2010).

During the ACA attachment process the adhesive is placed between the mating contacts. The ACA interconnection is established by applying pressure and heat simultaneously to the interconnection. When the temperature is raised the adhesive matrix will transform into low viscosity fluid (Tan et al., 2004), which allows excess adhesive to flow from the joints and fill the spaces around the contacts forming a physical connection between the parts to be attached. The conductive particles are trapped between the contacts and deform forming an electrical connection. As a result, electrical conduction is restricted to the z-direction and the electrical insulation in x-y directions is maintained. During cooling residual stresses are formed as a result of contraction of the adhesive matrix. In addition, residual stresses form

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 495

on the manufacturer. Typically the conductive particles are approximately 3-10 um in size. Nowadays, the most common conductive particles are nickel, which may be gold plated, and metal plated polymer particles. However, other materials, such as carbon fibres and solder balls, have also been used (Asai et al., 1995). The polymer particles are made of polystyrene cross-linked with divinyl benzene (Asai et al., 1995) and the metal plating on them may be of nickel, silver, or gold (Liu, 1996). The polymer particles are pliant and during the bonding process they deform, thereby forming the connection to the contacts. The deformation of the rigid nickel particles during the bonding process is less than that of the soft particles and the contact area formed is smaller. However, if the bonded contacts are made of softer metal, such as gold or copper, these contacts deform during bonding increasing the contact area with the rigid particles (Divigalpitiya & Hogerton, 2004; Yim &

In the flip chip process the interconnection is formed between pads on the substrate and bumps on the chip. However, as the cost of bumping may be unattractive in certain applications, bumpless chips are also used. The most commonly used bump materials are gold plated nickel, and gold. The gold bumps can be manufactured using an electroplating process. Copper bumps formed by similar electroplating techniques have also been considered as an alternative to gold because of their lower cost (Lau, 2000, Lau, 2003). However, copper oxidizes and corrodes easily, which may cause problems if it is used without plating. The nickel bumps can be made using an electroless plating process. This process has high potential for cost reduction, as it enables metal deposition directly on the aluminium pads on the chips. Thus the costly equipment needed in the electroplating process for sputtering, photoresist imaging, and electroplating is eliminated. The gold bumps may also be processed using a modified wire bonder to form stud bumps. The advantage of this process is that bumps can be formed on single chips in addition to whole

ACAs can be used either as films (ACF) or as pastes (ACA or ACP). The type of the ACA affects the bonding process and the equipment needed. ACFs are typically supplied in a reel and a dedicated in-line bonding machine is needed for cutting, aligning, and tacking to achieve high assembly speed. On the other hand, ACPs can be applied either by printing or by dispensing using a syringe. Even though the ACF process requires special equipment, ACFs are often used as they offer advantages compared to ACPs. The ACF process consumes less material than the ACP process. Moreover, the ACP process may destroy the

In the ACF bonding process the adhesive film is first cut to the correct size to cover the bonding area. After this the adhesive is aligned to the substrate and pretacked using light pressure and low temperature to attach the ACF to the substrate. After pretacking the carrier film on the ACF is removed. Next a chip is picked by a flip chip bonder. Typically a special flip chip bonder capable of simultaneously applying pressure and heat is used. The bumps on the chip and the pads on the substrate are aligned. The chip is pressed onto the substrate and heat is applied to the chip and the polymer matrix is cured. A schematic illustration of the bonding process used is presented in Figure 3. In case ACP material is

randomness of the particle distribution leading to problems in process quality.

Paik, 1998; Frisk & Ristolainen, 2005; Frisk & Kokko, 2006).

wafers. (Lau, 1995)

**2.2.2 The ACA bonding process** 

when the adhesive shrinks during curing. However, it has been shown that the residual stresses formed during cooling dominate (Kwon & Paik, 2004). This contraction builds up a sufficient force to create a stable, low-resistance connection. A typical cross section of an ACA flip chip joint is shown in Figure 2.

Fig. 2. Schematic illustration of an ACA joint between a chip and a substrate.

ACA joints have several advantages compared to underfilled solder interconnections. As the ACA process is solderless, there is no lead or alpha emission (Zhong, 2005). Moreover, the process is fluxless and no cleaning is required (Zhong, 2005). Furthermore, the process temperature is lower than that needed in soldering (Yim & Paik, 1998), which enables the use of heat sensitive or non-solderable substrate materials (Uddin et al., 2004). As the polymer matrix protects the contacts from mechanical damage and no underfilling is required (Lai & Liu, 1996), the ACA process costs less due to fewer processing steps (Yim & Paik, 2001). The ACA joining also enables very high interconnection density. On the other hand, the ACA joint has higher contact resistance and lower current capability than that made with solder (Jim & Paik, 1998; Zhong, 2005). Since the ACA has no self-alignment capability, a special bonding machine is needed for accurate alignment. During the bonding process heat and pressure also need to be applied simultaneously.

## **2.2.1 Materials used in ACAs**

Both thermoplastic and thermoset materials and their mixtures have been used as an ACA matrix. Initially, the ACAs were made of thermoplastic materials, as they have better reworkability and pot life (Lau, 1995). However, their stability at high temperatures is not good and the thermoplastic material is not strong enough to hold the conducting particles in position, which increases the contact resistance of the joint (Asai et al., 1995; Kim et al., 2004). Thermoset adhesives were developed to overcome the problems with thermoplastic adhesives. Thermoset adhesives are stable at high temperatures and enable low joint resistance. Epoxies are commonly used as an ACA matrix due to their good properties. Epoxies have excellent adhesion to a variety of substrates, due to the highly polar hydroxyl and ether groups (Luo, 2002). In addition, they have high glass transition temperature (Tg) and favourable melt viscosity (Kim et al., 2004; Yim & Paik, 1999). Furthermore, epoxies give low contact resistance and by selecting suitable curing agent long self-life and fast-cure properties can be achieved. The epoxy resin forms a crosslinked structure during bonding with good mechanical properties. However, their reworkability is problematic, as they are not thermally reversible and do not dissolve in common organic solvents (Lau, 1995).

The electrical conduction in ACA is formed by the conductive particles. The size, concentration and material of the conductive particles depend on the application area and on the manufacturer. Typically the conductive particles are approximately 3-10 um in size. Nowadays, the most common conductive particles are nickel, which may be gold plated, and metal plated polymer particles. However, other materials, such as carbon fibres and solder balls, have also been used (Asai et al., 1995). The polymer particles are made of polystyrene cross-linked with divinyl benzene (Asai et al., 1995) and the metal plating on them may be of nickel, silver, or gold (Liu, 1996). The polymer particles are pliant and during the bonding process they deform, thereby forming the connection to the contacts. The deformation of the rigid nickel particles during the bonding process is less than that of the soft particles and the contact area formed is smaller. However, if the bonded contacts are made of softer metal, such as gold or copper, these contacts deform during bonding increasing the contact area with the rigid particles (Divigalpitiya & Hogerton, 2004; Yim & Paik, 1998; Frisk & Ristolainen, 2005; Frisk & Kokko, 2006).

In the flip chip process the interconnection is formed between pads on the substrate and bumps on the chip. However, as the cost of bumping may be unattractive in certain applications, bumpless chips are also used. The most commonly used bump materials are gold plated nickel, and gold. The gold bumps can be manufactured using an electroplating process. Copper bumps formed by similar electroplating techniques have also been considered as an alternative to gold because of their lower cost (Lau, 2000, Lau, 2003). However, copper oxidizes and corrodes easily, which may cause problems if it is used without plating. The nickel bumps can be made using an electroless plating process. This process has high potential for cost reduction, as it enables metal deposition directly on the aluminium pads on the chips. Thus the costly equipment needed in the electroplating process for sputtering, photoresist imaging, and electroplating is eliminated. The gold bumps may also be processed using a modified wire bonder to form stud bumps. The advantage of this process is that bumps can be formed on single chips in addition to whole wafers. (Lau, 1995)

### **2.2.2 The ACA bonding process**

494 Scanning Electron Microscopy

when the adhesive shrinks during curing. However, it has been shown that the residual stresses formed during cooling dominate (Kwon & Paik, 2004). This contraction builds up a sufficient force to create a stable, low-resistance connection. A typical cross section of an

ACA joints have several advantages compared to underfilled solder interconnections. As the ACA process is solderless, there is no lead or alpha emission (Zhong, 2005). Moreover, the process is fluxless and no cleaning is required (Zhong, 2005). Furthermore, the process temperature is lower than that needed in soldering (Yim & Paik, 1998), which enables the use of heat sensitive or non-solderable substrate materials (Uddin et al., 2004). As the polymer matrix protects the contacts from mechanical damage and no underfilling is required (Lai & Liu, 1996), the ACA process costs less due to fewer processing steps (Yim & Paik, 2001). The ACA joining also enables very high interconnection density. On the other hand, the ACA joint has higher contact resistance and lower current capability than that made with solder (Jim & Paik, 1998; Zhong, 2005). Since the ACA has no self-alignment capability, a special bonding machine is needed for accurate alignment. During the bonding

Both thermoplastic and thermoset materials and their mixtures have been used as an ACA matrix. Initially, the ACAs were made of thermoplastic materials, as they have better reworkability and pot life (Lau, 1995). However, their stability at high temperatures is not good and the thermoplastic material is not strong enough to hold the conducting particles in position, which increases the contact resistance of the joint (Asai et al., 1995; Kim et al., 2004). Thermoset adhesives were developed to overcome the problems with thermoplastic adhesives. Thermoset adhesives are stable at high temperatures and enable low joint resistance. Epoxies are commonly used as an ACA matrix due to their good properties. Epoxies have excellent adhesion to a variety of substrates, due to the highly polar hydroxyl and ether groups (Luo, 2002). In addition, they have high glass transition temperature (Tg) and favourable melt viscosity (Kim et al., 2004; Yim & Paik, 1999). Furthermore, epoxies give low contact resistance and by selecting suitable curing agent long self-life and fast-cure properties can be achieved. The epoxy resin forms a crosslinked structure during bonding with good mechanical properties. However, their reworkability is problematic, as they are

not thermally reversible and do not dissolve in common organic solvents (Lau, 1995).

The electrical conduction in ACA is formed by the conductive particles. The size, concentration and material of the conductive particles depend on the application area and

Fig. 2. Schematic illustration of an ACA joint between a chip and a substrate.

process heat and pressure also need to be applied simultaneously.

**2.2.1 Materials used in ACAs** 

ACA flip chip joint is shown in Figure 2.

ACAs can be used either as films (ACF) or as pastes (ACA or ACP). The type of the ACA affects the bonding process and the equipment needed. ACFs are typically supplied in a reel and a dedicated in-line bonding machine is needed for cutting, aligning, and tacking to achieve high assembly speed. On the other hand, ACPs can be applied either by printing or by dispensing using a syringe. Even though the ACF process requires special equipment, ACFs are often used as they offer advantages compared to ACPs. The ACF process consumes less material than the ACP process. Moreover, the ACP process may destroy the randomness of the particle distribution leading to problems in process quality.

In the ACF bonding process the adhesive film is first cut to the correct size to cover the bonding area. After this the adhesive is aligned to the substrate and pretacked using light pressure and low temperature to attach the ACF to the substrate. After pretacking the carrier film on the ACF is removed. Next a chip is picked by a flip chip bonder. Typically a special flip chip bonder capable of simultaneously applying pressure and heat is used. The bumps on the chip and the pads on the substrate are aligned. The chip is pressed onto the substrate and heat is applied to the chip and the polymer matrix is cured. A schematic illustration of the bonding process used is presented in Figure 3. In case ACP material is

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 497

of ACA interconnections are discussed. Special attention is paid to the information obtained

As mentioned above, both optical and scanning electron microscopy may be used to study the cross-sections of ACA interconnections. Although, studying the cross-sections is typically very effective it has some drawbacks. The number of particles in ACA interconnections is typically quite low. Thus, when a cross-section of an interconnection is studied, the probability of seeing particles in interconnections is small even if there are sufficient particles in the interconnections to ensure proper joining. From a cross-section only one side of a chip is seen. Therefore, to determine planarity and alignment issues several cross-sections are needed. Furthermore, making a cross-section of a sample destroys it thereby gravely restricting further analysis. Optical microscopy may give valuable information especially when planarity and alignment are concerned. However, typically it does not give very good detailed information. Especially, if interfaces of different materials are studied an optical microscope does not give reliable information. On the other hand, SEM is often a very powerful tool to determine the quality and structure of an ACA interconnection. However, SEM analysis benefits from information of other analysis methods if they are available such as, for example, electrical characterisation and scanning

For SEM analysis the quality of the cross-sections needs to be good and they need to be clean. Typically epoxy moulding is used followed by grinding and polishing. However, other materials are possible such as acrylics. Use of high temperature mould materials may cause problems depending on the structure studied and its materials. As the ACA structure has many different materials the analysis is often challenging and the parameters used for SEM need to be determined according to the samples studied. Additionally, the area of interest in the interconnection affects the parameters. Both thin gold and carbon layers may be used to make samples electrically conductive. However, gold gives better quality of

Bonding parameters are the key factors when the quality and the reliability of the ACA interconnections are considered. The ACA process makes the bonding parameters especially important as the complex mechanical, rheological, and chemical properties of the ACA materials need to be considered (Dou, 2006). The most important parameters in the ACA bonding process are time, temperature and pressure. However, other bonding parameters, such as application rates of pressure and temperature, also affect the quality and reliability of the joints (Ogunjimi et al., 1996; Whalley et al., 1997). Moreover, the parameters may

Finding optimum process parameters necessitates careful study. Quite often the quality of the bonding parameters cannot be determined on the basis of electrical or adhesion measurements only and cross-sections are needed for verification of the interconnection quality. For thermoset adhesives the bonding temperature together with the bonding time determine the degree of cure of the adhesive matrix. The higher the bonding temperature

analysis and is recommended if elemental analysis is not needed.

from the interconnections using SEM analysis.

**3.1 SEM analysis of ACA interconnections** 

acoustic microscopy.

**3.2 Bonding parameters** 

interact with each other.

used the steps a and b are replaced by deposition of the ACP. After this steps c and d are performed similarly to the ACF process.

Fig. 3. Schematic illustration of the ACF flip chip bonding process: a) placement of ACF on the substrate, b) prebonding, c) alignment of the chip and the substrate and d) final bonding.

## **3. Evaluation of the quality of ACA interconnections**

The quality of the ACA interconnections may be studied using several techniques. In general, a good quality ACA interconnection is characterized by low contact resistance and good mechanical properties. Therefore, electrical measurements may be done to assess the quality. However, often this is not possible due to the design of the semiconductor chip. Electrical measurements may show alignment and planarity problems as higher resistance values. However, this is not always the case, as it is possible that the electrical connection seems good even though there are problems in the joints causing reliability problems during use. The alignment of an interconnection may be studied through the substrate in the case of transparent substrates such as glass or thin polyimide film. However, often the most effective way to examine both alignment and the structure of the interconnection is to make a cross-section of the structure and study them using either optical or scanning electron microscope. SEM especially is often very a effective tool yielding a plethora of information of the joint, which is important for optimisation of the bonding process.

The mechanical properties of the ACA interconnections may be studied using adhesion testing, which will indicate how well the ACA material is attached to the substrate and the chip. In this technique both adhesion strength and failure mechanism during testing will give valuable information. In the following several different parameters affecting the quality of ACA interconnections are discussed. Special attention is paid to the information obtained from the interconnections using SEM analysis.

## **3.1 SEM analysis of ACA interconnections**

496 Scanning Electron Microscopy

used the steps a and b are replaced by deposition of the ACP. After this steps c and d are

Fig. 3. Schematic illustration of the ACF flip chip bonding process: a) placement of ACF on the substrate, b) prebonding, c) alignment of the chip and the substrate and d) final bonding.

The quality of the ACA interconnections may be studied using several techniques. In general, a good quality ACA interconnection is characterized by low contact resistance and good mechanical properties. Therefore, electrical measurements may be done to assess the quality. However, often this is not possible due to the design of the semiconductor chip. Electrical measurements may show alignment and planarity problems as higher resistance values. However, this is not always the case, as it is possible that the electrical connection seems good even though there are problems in the joints causing reliability problems during use. The alignment of an interconnection may be studied through the substrate in the case of transparent substrates such as glass or thin polyimide film. However, often the most effective way to examine both alignment and the structure of the interconnection is to make a cross-section of the structure and study them using either optical or scanning electron microscope. SEM especially is often very a effective tool yielding a plethora of information

The mechanical properties of the ACA interconnections may be studied using adhesion testing, which will indicate how well the ACA material is attached to the substrate and the chip. In this technique both adhesion strength and failure mechanism during testing will give valuable information. In the following several different parameters affecting the quality

**3. Evaluation of the quality of ACA interconnections** 

of the joint, which is important for optimisation of the bonding process.

performed similarly to the ACF process.

As mentioned above, both optical and scanning electron microscopy may be used to study the cross-sections of ACA interconnections. Although, studying the cross-sections is typically very effective it has some drawbacks. The number of particles in ACA interconnections is typically quite low. Thus, when a cross-section of an interconnection is studied, the probability of seeing particles in interconnections is small even if there are sufficient particles in the interconnections to ensure proper joining. From a cross-section only one side of a chip is seen. Therefore, to determine planarity and alignment issues several cross-sections are needed. Furthermore, making a cross-section of a sample destroys it thereby gravely restricting further analysis. Optical microscopy may give valuable information especially when planarity and alignment are concerned. However, typically it does not give very good detailed information. Especially, if interfaces of different materials are studied an optical microscope does not give reliable information. On the other hand, SEM is often a very powerful tool to determine the quality and structure of an ACA interconnection. However, SEM analysis benefits from information of other analysis methods if they are available such as, for example, electrical characterisation and scanning acoustic microscopy.

For SEM analysis the quality of the cross-sections needs to be good and they need to be clean. Typically epoxy moulding is used followed by grinding and polishing. However, other materials are possible such as acrylics. Use of high temperature mould materials may cause problems depending on the structure studied and its materials. As the ACA structure has many different materials the analysis is often challenging and the parameters used for SEM need to be determined according to the samples studied. Additionally, the area of interest in the interconnection affects the parameters. Both thin gold and carbon layers may be used to make samples electrically conductive. However, gold gives better quality of analysis and is recommended if elemental analysis is not needed.

#### **3.2 Bonding parameters**

Bonding parameters are the key factors when the quality and the reliability of the ACA interconnections are considered. The ACA process makes the bonding parameters especially important as the complex mechanical, rheological, and chemical properties of the ACA materials need to be considered (Dou, 2006). The most important parameters in the ACA bonding process are time, temperature and pressure. However, other bonding parameters, such as application rates of pressure and temperature, also affect the quality and reliability of the joints (Ogunjimi et al., 1996; Whalley et al., 1997). Moreover, the parameters may interact with each other.

Finding optimum process parameters necessitates careful study. Quite often the quality of the bonding parameters cannot be determined on the basis of electrical or adhesion measurements only and cross-sections are needed for verification of the interconnection quality. For thermoset adhesives the bonding temperature together with the bonding time determine the degree of cure of the adhesive matrix. The higher the bonding temperature

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 499

Fig. 5. a) and b) examples of marked deformation of gold plated polymer particles.

pads prevented the penetration of the particles into the pads.

the bump and the pad.

**3.3 Effect of the substrate material** 

different substrate materials are given.

If rigid particles are used, the bump and the pad material also have a strong effect on the bonding pressure. With soft bump materials, such as gold or copper, the particles will sink into the bump forming a strong bond. If high bonding pressure is used, the particles will sink completely into the bumps and direct contact between the pad and the bump will be formed. An example of direct contact is presented in Figure 6. The hard nickel-plating on the

Fig. 6. Penetration of the nickel particles into the copper bump and direct contact between

Poor substrate and chip quality may cause coplanarity problems. This may increase the contact resistance, as not all joints have adequate deformation of the particles. For the flip chip process to be usable with the organic substrates, the planarity of the substrate is very important. Planarity issues often need to be determined using cross-sections. In the following examples of SEM analysis used for quality studies of ACA interconnections with

used the higher the degree of cure of the adhesive matrix is within the same amount of time (Chan & Luk, 2002a; Rizvi et al., 2005; Tan et al., 2004; Wu et al., 1997) as the higher temperature accelerates the crosslinking reaction (Uddin et al., 2004). Similarly, longer bonding time increases the degree of cure. Both bonding time and temperature needs to be high enough to achieve adequate curing of the adhesive matrix as the mechanical and chemical properties of the ACA have been found to depend heavily on the degree of cure (Wu et al., 1997). Too low degree of cure is often seen as high contact resistance values and also as inadequate adhesive strength.

The bonding pressure determines the deformation of the conductive particles. If the bonding pressure is too low the conductive particles cannot make good contact with the bonded surfaces and the contact resistance will be high (Chan & Luk, 2002b; Lau, 1995). In addition, the reliability of this kind of joint is poor (Frisk & Ristolainen, 2005; Lai & Liu, 1997). Examples of only slightly deformed particles are shown in Figure 4. In this case the variation is assumed to be caused by too fast curing of the adhesive during the bonding process. This adhesive has been designed to cure very quickly. Consequently it may have started to cure before the adhesive had flowed properly, leaving the particles insufficiently deformed.

Fig. 4. An example of slightly deformed particles: a) gold coated polymer particles and b) nickel particles.

When the bonding pressure is increased, the contact resistance typically decreases sharply at first before evening out (Yim & Paik, 1998; Yin et al., 2003). This is caused by greater deformation of the conductive particles leading to a larger contact area between the particle and contacts (Kwon et al., 2006; Yin et al., 2003). Figure 5 shows examples of properly deformed particles. However, if the pressure is increased too much, the contact resistance may start to increase again (Chan & Luk, 2002b; Yim & Paik, 1998). If metal plated polymer particles are used, too high pressure may crush the particles (Wang et al., 1998) and lead to direct contact between the bump and the pad. The cracking of the metal plating may separate it from the polymer core and reduce the amount of conductive path between the pad and the bump (Yim & Paik, 1998). Another problem with too high bonding pressure is elastic stress formed in the chip or in the substrate (Frisk et al., 2010; Lai & Liu, 1996). In some cases, the high temperature used during the bonding process may soften the substrate material used and it will deform markedly during the bonding process.

used the higher the degree of cure of the adhesive matrix is within the same amount of time (Chan & Luk, 2002a; Rizvi et al., 2005; Tan et al., 2004; Wu et al., 1997) as the higher temperature accelerates the crosslinking reaction (Uddin et al., 2004). Similarly, longer bonding time increases the degree of cure. Both bonding time and temperature needs to be high enough to achieve adequate curing of the adhesive matrix as the mechanical and chemical properties of the ACA have been found to depend heavily on the degree of cure (Wu et al., 1997). Too low degree of cure is often seen as high contact resistance values and

The bonding pressure determines the deformation of the conductive particles. If the bonding pressure is too low the conductive particles cannot make good contact with the bonded surfaces and the contact resistance will be high (Chan & Luk, 2002b; Lau, 1995). In addition, the reliability of this kind of joint is poor (Frisk & Ristolainen, 2005; Lai & Liu, 1997). Examples of only slightly deformed particles are shown in Figure 4. In this case the variation is assumed to be caused by too fast curing of the adhesive during the bonding process. This adhesive has been designed to cure very quickly. Consequently it may have started to cure before the

Fig. 4. An example of slightly deformed particles: a) gold coated polymer particles and b)

material used and it will deform markedly during the bonding process.

When the bonding pressure is increased, the contact resistance typically decreases sharply at first before evening out (Yim & Paik, 1998; Yin et al., 2003). This is caused by greater deformation of the conductive particles leading to a larger contact area between the particle and contacts (Kwon et al., 2006; Yin et al., 2003). Figure 5 shows examples of properly deformed particles. However, if the pressure is increased too much, the contact resistance may start to increase again (Chan & Luk, 2002b; Yim & Paik, 1998). If metal plated polymer particles are used, too high pressure may crush the particles (Wang et al., 1998) and lead to direct contact between the bump and the pad. The cracking of the metal plating may separate it from the polymer core and reduce the amount of conductive path between the pad and the bump (Yim & Paik, 1998). Another problem with too high bonding pressure is elastic stress formed in the chip or in the substrate (Frisk et al., 2010; Lai & Liu, 1996). In some cases, the high temperature used during the bonding process may soften the substrate

adhesive had flowed properly, leaving the particles insufficiently deformed.

also as inadequate adhesive strength.

nickel particles.

Fig. 5. a) and b) examples of marked deformation of gold plated polymer particles.

If rigid particles are used, the bump and the pad material also have a strong effect on the bonding pressure. With soft bump materials, such as gold or copper, the particles will sink into the bump forming a strong bond. If high bonding pressure is used, the particles will sink completely into the bumps and direct contact between the pad and the bump will be formed. An example of direct contact is presented in Figure 6. The hard nickel-plating on the pads prevented the penetration of the particles into the pads.

Fig. 6. Penetration of the nickel particles into the copper bump and direct contact between the bump and the pad.

#### **3.3 Effect of the substrate material**

Poor substrate and chip quality may cause coplanarity problems. This may increase the contact resistance, as not all joints have adequate deformation of the particles. For the flip chip process to be usable with the organic substrates, the planarity of the substrate is very important. Planarity issues often need to be determined using cross-sections. In the following examples of SEM analysis used for quality studies of ACA interconnections with different substrate materials are given.

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 501

deformation of the particles to be identical in the joints and increases reliability compared to

Fig. 7. a) and b) micrographs showing deformation of a FR-4 substrate

Fig. 8. Schematic cross-section of a printed circuit board made with SBU process.

Deformation of the particles between these test boards was also studied with SEM. With the test substrate, which did not have the RCC, the particles sank into the copper bump. However, the fairly thick nickel plating on the pads prevented the particles from sinking into it and caused deformation of the particles, as can be seen in Figure 10 a). In addition, due to relatively high magnification of the SEM micrograph the gold layers on both conducting particles and pad can be easily seen. The thinner nickel plating on the RCC test board enabled the particles to sink into the pad, as can be seen in Figure 10 b). Moreover, as the RCC gives in more under the pads during bonding than the FR-4 substrate, the deformation of the particles was less on the substrates with the RCC. Deformation of the particles is important for the reliability of the joints. Sufficiently deformed particles form a strong atomic interaction between the pad and the bump creating increased stability of the joint (Lai & Liu, 1996). Using SEM the interface between the particles and the pad or the bump may be studied and problems such as thin layers of polymer matrix may be detected. In both Figures 10 a) and 10 b) a good contact of particles is seen to both the pad and the

the substrate without the RCC.

bump .

#### **3.3.1 Glass fibre reinforced substrates**

Glass fibre reinforced materials are commonly used as substrates in electronics. The most widely used material is FR-4, which is a grade designated by the National Electrical Manufacturers Association (NEMA), which determines that the material is flame resistant, is primarily epoxy based, and has woven glass fibre reinforcement (Coombs, 2000). As the resin material between different FR-4s may vary, the properties of FR-4 materials are not identical. For example typical Tg of FR-4 is between 130 and 140 °C or between 170 and 180 °C. The popularity of the FR-4 is based on its good properties, availability, and low cost. When rigid glass fibre reinforced substrates are used in ACA applications, coplanarity problems may arise due to the woven structure of the substrate (Frisk & Cumini, 2009). During the bonding process the high temperature may soften the resin, which leads to its deformation under the pads. This deformation has been found to depend on the orientation of the glass fibres in the substrate and affect the electrical conductivity and reliability of the joints (Liu et al., 1999).

Deformation of substrate in ACA interconnections was studied using cross-sections and SEM (Frisk & Kokko, 2006; Frisk & Cumini, 2009; Frisk et al., 2010). SEM has proven to be a very effective method for such studies as the different materials and their interfaces may be clearly seen. Figure 7 shows micrographs of a FR-4 substrates after an ACA bonding process. Deformation of the substrates may be seen between contacts. Deformation is especially considerable in the areas where the glass fibres were far from the surface. This varying deformation of the substrate causes pressure variation in the joints (Pinardi, 1998). The varying pressure is important as it may result in different deformation of the particles leading to variation in the contact resistance and also impairing the reliability of the joints.

Quite often in flip chip applications with ACA materials very high wiring densities are needed. These are difficult to achieve with the traditional FR-4 substrates shown in Figure 7 One possibility to meet these demands is to use sequential build-up (SBU) processes. However, in this process conductor and dielectric layers are formed one after another on a rigid core board, which may be an FR-4 glass reinforced laminate (Tagagi et al., 2003). The electrical connection between the core board and the build-up layers is formed using microvia technologies. An example of a substrate made with the SBU process is presented in Figure 8. The typical dielectric materials used in SBU build-up layers are resin-coated copper foil (RCC or RCF), thermally cured resin, and photo-imageable resin (Tagagi et al., 2003). The most widely used dielectric material in the SBU process is RCC. The RCC is formed by adding a layer of resin to a thin copper foil, which is laminated to the core board. A typical resin material is epoxy. The RCC has several advantages and is suitable for processing in standard printed circuit board processes.

As the RCC layer does not have glass fibres, it is more pliable than an FR-4 substrate. Its effect on the interconnection structures was studied using cross-sections and SEM (Frisk & Kokko, 2006). During the bonding process the depression of the copper pads into the RCC was found to be much stronger than the depression into the FR-4 substrate. In Figure 9 an example of the depression is presented for test samples with the RCC test substrate. As can be seen, the RCC has deformed markedly more during the bonding process than the pure FR-4 substrates shown in Figure 7. Furthermore, some deformation of the FR-4 substrate beneath the RCC has occurred. Although the deformation of the RCC is greater, it is almost identical under every pad leading to more uniform distribution of pressure. This causes the

Glass fibre reinforced materials are commonly used as substrates in electronics. The most widely used material is FR-4, which is a grade designated by the National Electrical Manufacturers Association (NEMA), which determines that the material is flame resistant, is primarily epoxy based, and has woven glass fibre reinforcement (Coombs, 2000). As the resin material between different FR-4s may vary, the properties of FR-4 materials are not identical. For example typical Tg of FR-4 is between 130 and 140 °C or between 170 and 180 °C. The popularity of the FR-4 is based on its good properties, availability, and low cost. When rigid glass fibre reinforced substrates are used in ACA applications, coplanarity problems may arise due to the woven structure of the substrate (Frisk & Cumini, 2009). During the bonding process the high temperature may soften the resin, which leads to its deformation under the pads. This deformation has been found to depend on the orientation of the glass fibres in the substrate and affect the electrical conductivity and reliability of the

Deformation of substrate in ACA interconnections was studied using cross-sections and SEM (Frisk & Kokko, 2006; Frisk & Cumini, 2009; Frisk et al., 2010). SEM has proven to be a very effective method for such studies as the different materials and their interfaces may be clearly seen. Figure 7 shows micrographs of a FR-4 substrates after an ACA bonding process. Deformation of the substrates may be seen between contacts. Deformation is especially considerable in the areas where the glass fibres were far from the surface. This varying deformation of the substrate causes pressure variation in the joints (Pinardi, 1998). The varying pressure is important as it may result in different deformation of the particles leading to variation in the contact resistance and also impairing the reliability of the joints. Quite often in flip chip applications with ACA materials very high wiring densities are needed. These are difficult to achieve with the traditional FR-4 substrates shown in Figure 7 One possibility to meet these demands is to use sequential build-up (SBU) processes. However, in this process conductor and dielectric layers are formed one after another on a rigid core board, which may be an FR-4 glass reinforced laminate (Tagagi et al., 2003). The electrical connection between the core board and the build-up layers is formed using microvia technologies. An example of a substrate made with the SBU process is presented in Figure 8. The typical dielectric materials used in SBU build-up layers are resin-coated copper foil (RCC or RCF), thermally cured resin, and photo-imageable resin (Tagagi et al., 2003). The most widely used dielectric material in the SBU process is RCC. The RCC is formed by adding a layer of resin to a thin copper foil, which is laminated to the core board. A typical resin material is epoxy. The RCC has several advantages and is suitable for

As the RCC layer does not have glass fibres, it is more pliable than an FR-4 substrate. Its effect on the interconnection structures was studied using cross-sections and SEM (Frisk & Kokko, 2006). During the bonding process the depression of the copper pads into the RCC was found to be much stronger than the depression into the FR-4 substrate. In Figure 9 an example of the depression is presented for test samples with the RCC test substrate. As can be seen, the RCC has deformed markedly more during the bonding process than the pure FR-4 substrates shown in Figure 7. Furthermore, some deformation of the FR-4 substrate beneath the RCC has occurred. Although the deformation of the RCC is greater, it is almost identical under every pad leading to more uniform distribution of pressure. This causes the

**3.3.1 Glass fibre reinforced substrates** 

processing in standard printed circuit board processes.

joints (Liu et al., 1999).

deformation of the particles to be identical in the joints and increases reliability compared to the substrate without the RCC.

Fig. 7. a) and b) micrographs showing deformation of a FR-4 substrate

Fig. 8. Schematic cross-section of a printed circuit board made with SBU process.

Deformation of the particles between these test boards was also studied with SEM. With the test substrate, which did not have the RCC, the particles sank into the copper bump. However, the fairly thick nickel plating on the pads prevented the particles from sinking into it and caused deformation of the particles, as can be seen in Figure 10 a). In addition, due to relatively high magnification of the SEM micrograph the gold layers on both conducting particles and pad can be easily seen. The thinner nickel plating on the RCC test board enabled the particles to sink into the pad, as can be seen in Figure 10 b). Moreover, as the RCC gives in more under the pads during bonding than the FR-4 substrate, the deformation of the particles was less on the substrates with the RCC. Deformation of the particles is important for the reliability of the joints. Sufficiently deformed particles form a strong atomic interaction between the pad and the bump creating increased stability of the joint (Lai & Liu, 1996). Using SEM the interface between the particles and the pad or the bump may be studied and problems such as thin layers of polymer matrix may be detected. In both Figures 10 a) and 10 b) a good contact of particles is seen to both the pad and the bump .

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 503

The pliability of the flexible substrates may cause some problems during the bonding process. Figure 11 shows SEM micrographs of ACA interconnections with flexible liquid crystal polymer (LCP) substrates with two different pressures. A marked deformation of the liquid crystal polymer film with high pressure may be seen. Such deformation may cause problems. On the other hand, it may also even out some planarity problems as deformation can absorb the height variations (Connell, 1997; Savolainen, 2004) and thereby increase the quality of the interconnection. However, in some cases deformation may cause cracking of the wiring and thereby lead in to reliability problems. With the lower pressure the pressure exerted to the particles may not be high enough and therefore cause reliability problems.

Fig. 11. An example of deformation of the LCP substrate when a) low bonding pressure and

With thin flexible substrates routing may also be critical to the distribution of the pressure, if the substrate has several conductive layers (Lai & Liu, 1996). If double sided flexible substrates are used they may have wiring on both sides of the bonding area. This may cause uneven distribution of the pressure and deformation of contact areas. Figure 12 shows an interconnection with a polyimide substrate having double sided wiring. As can be seen, the gold bump has markedly deformed during the bonding process because of the wiring on the other side of the substrate. A similar effect may be seen in the LCP substrates in Figure 11. However, as the solder resist on the LCP substrate evened out the effect of the wiring, there is clearly less deformation. Consequently, the design of flexible circuitry is very important for good quality interconnections. Furthermore, such quality problems are difficult to detect

Another problem in a substrate may be overetching of the pads. As flexible substrates are often used in application where very high density substrates are needed such as attachment of driver chips is display application, overetching may cause marked problems. It reduces the contact area and may cause alignment and planarity problems. Figure 13 shows a SEM micrograph of a interconnection with an overetched polyimide (PI) substrate. In the original substrate layout the pad width was designed to be slightly greater than the bump. However, due to overetching the size of the pad is clearly less than that of the bump. Such overetching may be seen when substrates are quality checked before use. However, cross-section is a

good way to evaluate the effect of overetching on an interconnect.

b) high bonding pressure was used.

without cross-sectioning.

Fig. 9. Micrograph presenting the immersion of the pads in the RCC, when high bonding pressure is used.

Fig. 10. a) Micrograph presenting the deformation of the rigid nickel particle, when a FR-4 substrate is used. b) Micrograph presenting the immersion of the rigid nickel particle in the pad and the bump, when the substrate with the RCC is used.

#### **3.3.2 Flexible substrates**

ACA materials are often used with flexible substrates, which are fabricated using pliable unreinforced polymeric materials. Flexible substrates have several advantages compared to fibre-reinforced substrates and lighter and thinner products can be produced using them. Flexible substrate may absorb stress, which may be important for the reliability of the interconnections, especially in flip chip applications. In addition, the thermal transfer through a thin substrate is more effective. Furthermore, very high density substrates are available with flexible substrates and this is often critical for ACA applications. On the other hand, the thinness of the substrate may cause problems in the stability of the construction. Moreover, the cost of the flexible substrates is higher than that of the rigid substrates. (Coombs, 2001)

Fig. 9. Micrograph presenting the immersion of the pads in the RCC, when high bonding

Fig. 10. a) Micrograph presenting the deformation of the rigid nickel particle, when a FR-4 substrate is used. b) Micrograph presenting the immersion of the rigid nickel particle in the

ACA materials are often used with flexible substrates, which are fabricated using pliable unreinforced polymeric materials. Flexible substrates have several advantages compared to fibre-reinforced substrates and lighter and thinner products can be produced using them. Flexible substrate may absorb stress, which may be important for the reliability of the interconnections, especially in flip chip applications. In addition, the thermal transfer through a thin substrate is more effective. Furthermore, very high density substrates are available with flexible substrates and this is often critical for ACA applications. On the other hand, the thinness of the substrate may cause problems in the stability of the construction. Moreover, the cost of the flexible substrates is higher than that of the rigid

pad and the bump, when the substrate with the RCC is used.

pressure is used.

**3.3.2 Flexible substrates** 

substrates. (Coombs, 2001)

The pliability of the flexible substrates may cause some problems during the bonding process. Figure 11 shows SEM micrographs of ACA interconnections with flexible liquid crystal polymer (LCP) substrates with two different pressures. A marked deformation of the liquid crystal polymer film with high pressure may be seen. Such deformation may cause problems. On the other hand, it may also even out some planarity problems as deformation can absorb the height variations (Connell, 1997; Savolainen, 2004) and thereby increase the quality of the interconnection. However, in some cases deformation may cause cracking of the wiring and thereby lead in to reliability problems. With the lower pressure the pressure exerted to the particles may not be high enough and therefore cause reliability problems.

Fig. 11. An example of deformation of the LCP substrate when a) low bonding pressure and b) high bonding pressure was used.

With thin flexible substrates routing may also be critical to the distribution of the pressure, if the substrate has several conductive layers (Lai & Liu, 1996). If double sided flexible substrates are used they may have wiring on both sides of the bonding area. This may cause uneven distribution of the pressure and deformation of contact areas. Figure 12 shows an interconnection with a polyimide substrate having double sided wiring. As can be seen, the gold bump has markedly deformed during the bonding process because of the wiring on the other side of the substrate. A similar effect may be seen in the LCP substrates in Figure 11. However, as the solder resist on the LCP substrate evened out the effect of the wiring, there is clearly less deformation. Consequently, the design of flexible circuitry is very important for good quality interconnections. Furthermore, such quality problems are difficult to detect without cross-sectioning.

Another problem in a substrate may be overetching of the pads. As flexible substrates are often used in application where very high density substrates are needed such as attachment of driver chips is display application, overetching may cause marked problems. It reduces the contact area and may cause alignment and planarity problems. Figure 13 shows a SEM micrograph of a interconnection with an overetched polyimide (PI) substrate. In the original substrate layout the pad width was designed to be slightly greater than the bump. However, due to overetching the size of the pad is clearly less than that of the bump. Such overetching may be seen when substrates are quality checked before use. However, cross-section is a good way to evaluate the effect of overetching on an interconnect.

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 505

levels or higher stress cycle frequency during testing compared to those under normal operational conditions of the product (Suhir, 2002). Depending on the condition failures may occur through several mechanisms. It is important that the testing conditions are determined so that the failure mechanisms during testing are similar to those occurring under normal conditions of use. The test conditions depend decisively on the application of the product. For example, the test conditions for products used in military and space

Several different accelerated life tests have been used to study the reliability of ACA joints; see for example (Frisk & Ristolainen, 2005; Frisk & Cumini, 2006; Jang et al., 2008; Kim et al., 2004; Lai & Liu, 1996; Saarinen et al., 2011). These include high temperature aging tests, temperature cycling tests, high temperature and high humidity tests, and humidity and temperature cycling tests. The reliability and failure mechanisms of ACA interconnections depend on several factors, including the materials and bonding parameters. The materials used in the pads, bumps, and conductive particles need to be compatible with each other. The substrate material may also have a marked influence on the reliability of the joints. As the properties of the joints are much influenced by the bonding process, it is important that optimum bonding parameters are determined and used in the bonding process. In the

following failure mechanisms and analysis of ACA interconnections are discussed.

has been reported to decrease the shear strain in the joints (Kwon et al., 2005).

reducing the delamination and increasing the reliability.

The amount of warpage depends on the difference between the coefficient of thermal expansions of chip and substrate, but also on their stiffness. When a rigid fibre reinforced substrate is used, the shear stress caused by CTE mismatches is localised between the pad and the bump as the of deformation the substrate is less than that of the adhesive matrix (Lai et al., 1998). On the other hand, deformation of an unreinforced flexible substrate occurs much more easily due to its low modulus, and some of the shear stress may be absorbed by the substrate (Connell et al., 1997). This decreases the shear stress in the interconnection

**4.1 Failures of ACA interconnections with rigid FR-4 substrates under thermal cycling**  One of the major problems in ACA assembly using organic substrates is the great difference between the coefficient of thermal expansions (CTE) of chip and substrate. The bonding process is done at elevated temperatures. When the package is cooled down the contraction of the silicon chip is very small due to its low coefficient of thermal expansion and high Young's modulus (Kwon et al., 2005). On the other hand, the contraction of the substrate is much greater. At low temperatures this causes stresses to form between the chip and the substrate and the ACA flip chip package to warp downwards. At temperatures below its Tg the adhesive matrix holds the chip and the substrate together enabling this warpage (Kwon & Paik, 2006). However, when the Tg of the adhesive is exceeded, its mechanical strength decreases and it cannot provide mechanical support between the substrate and the chip. The warpage is evened out and both chip and substrate may expand with their inherent CTEs (Kwon & Paik, 2006). The warpage of the ACA package is presented in Figure 14. If the ambient temperature fluctuates, as in a temperature cycling test, the flip chip package warps repeatedly. The warping reduces the shear stress in the joints and a greater degree of warp

applications are much more rigorous than those for consumer electronics.

Fig. 12. Effect of the tracks on joint deformation on the PI substrate when the tracks are on both sides of the substrate.

Fig. 13. An overetched pad on the PI substrate.

## **4. Failure mechanisms of the ACA flip chip joints**

In general, a good ACA joint is characterized by low contact resistance. A key issue for the ACA to function properly is the retention of this contact resistance during the operational life of the product. Failure of a product can be defined as its inability to perform its intended function [For an ACA flip chip joint this typically means too great an increase in contact resistance. There is no single specific definition of failure of ACA joint resistance, as this depends largely on the application.

To improve the reliability of interconnections the reasons for their failure and the failure mechanisms must be understood. During the design phase reliability is typically assessed using accelerated environmental testing. The aim of such testing is to predict the future performance of a product in a shorter period of time than the service life of the product. Accelerated life tests can also be used to detect failure mechanisms occurring in products under different conditions of use. The acceleration is accomplished by using elevated stress

Fig. 12. Effect of the tracks on joint deformation on the PI substrate when the tracks are on

In general, a good ACA joint is characterized by low contact resistance. A key issue for the ACA to function properly is the retention of this contact resistance during the operational life of the product. Failure of a product can be defined as its inability to perform its intended function [For an ACA flip chip joint this typically means too great an increase in contact resistance. There is no single specific definition of failure of ACA joint resistance, as this

To improve the reliability of interconnections the reasons for their failure and the failure mechanisms must be understood. During the design phase reliability is typically assessed using accelerated environmental testing. The aim of such testing is to predict the future performance of a product in a shorter period of time than the service life of the product. Accelerated life tests can also be used to detect failure mechanisms occurring in products under different conditions of use. The acceleration is accomplished by using elevated stress

both sides of the substrate.

Fig. 13. An overetched pad on the PI substrate.

depends largely on the application.

**4. Failure mechanisms of the ACA flip chip joints** 

levels or higher stress cycle frequency during testing compared to those under normal operational conditions of the product (Suhir, 2002). Depending on the condition failures may occur through several mechanisms. It is important that the testing conditions are determined so that the failure mechanisms during testing are similar to those occurring under normal conditions of use. The test conditions depend decisively on the application of the product. For example, the test conditions for products used in military and space applications are much more rigorous than those for consumer electronics.

Several different accelerated life tests have been used to study the reliability of ACA joints; see for example (Frisk & Ristolainen, 2005; Frisk & Cumini, 2006; Jang et al., 2008; Kim et al., 2004; Lai & Liu, 1996; Saarinen et al., 2011). These include high temperature aging tests, temperature cycling tests, high temperature and high humidity tests, and humidity and temperature cycling tests. The reliability and failure mechanisms of ACA interconnections depend on several factors, including the materials and bonding parameters. The materials used in the pads, bumps, and conductive particles need to be compatible with each other. The substrate material may also have a marked influence on the reliability of the joints. As the properties of the joints are much influenced by the bonding process, it is important that optimum bonding parameters are determined and used in the bonding process. In the following failure mechanisms and analysis of ACA interconnections are discussed.

#### **4.1 Failures of ACA interconnections with rigid FR-4 substrates under thermal cycling**

One of the major problems in ACA assembly using organic substrates is the great difference between the coefficient of thermal expansions (CTE) of chip and substrate. The bonding process is done at elevated temperatures. When the package is cooled down the contraction of the silicon chip is very small due to its low coefficient of thermal expansion and high Young's modulus (Kwon et al., 2005). On the other hand, the contraction of the substrate is much greater. At low temperatures this causes stresses to form between the chip and the substrate and the ACA flip chip package to warp downwards. At temperatures below its Tg the adhesive matrix holds the chip and the substrate together enabling this warpage (Kwon & Paik, 2006). However, when the Tg of the adhesive is exceeded, its mechanical strength decreases and it cannot provide mechanical support between the substrate and the chip. The warpage is evened out and both chip and substrate may expand with their inherent CTEs (Kwon & Paik, 2006). The warpage of the ACA package is presented in Figure 14. If the ambient temperature fluctuates, as in a temperature cycling test, the flip chip package warps repeatedly. The warping reduces the shear stress in the joints and a greater degree of warp has been reported to decrease the shear strain in the joints (Kwon et al., 2005).

The amount of warpage depends on the difference between the coefficient of thermal expansions of chip and substrate, but also on their stiffness. When a rigid fibre reinforced substrate is used, the shear stress caused by CTE mismatches is localised between the pad and the bump as the of deformation the substrate is less than that of the adhesive matrix (Lai et al., 1998). On the other hand, deformation of an unreinforced flexible substrate occurs much more easily due to its low modulus, and some of the shear stress may be absorbed by the substrate (Connell et al., 1997). This decreases the shear stress in the interconnection reducing the delamination and increasing the reliability.

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 507

particle can be clearly seen. Additionally, agglomeration of material under the particles can be seen indicating failure caused by this sliding. Figure 17 b) shows a similar situation for

Fig. 15. Examples of delamination in test samples with FR-4 substrates after thermal cycling a) clear delamination between ACA and pad, b) delamination between bump and pad

In addition to temperature changes, humidity has been found to have a major influence on the reliability of ACA interconnections. Under humid conditions the adhesive matrix may deform as it absorbs water. The adhesive matrix may also relax due to the increased temperature. The effect of water absorption depends on the structure of the adhesive and also on the duration of the exposure. The most common adhesive material is epoxy. It has been suggested that the water absorbed in the epoxy polymer has two states to which the water molecules can diffuse. The water may either fill the free volume in the polymer matrix or form hydrogen bonds with the epoxy polymer. If the water is hydrogen-bonded, it causes swelling of the adhesive matrix (Chiang & Fernandez-Garcia, 2002; Luo et al., 2002). This swelling due to humidity typically increases with temperature, but decreases sharply across

d) delamination varying between the ACA-pad interface and the ACA-bump interface.

**4.2 Failures of ACA interconnections with rigid FR-4 substrates under humidity** 

which continues to ACA-chip interface, c) less pronounced delamination, and

**testing** 

nickel particle and similar agglomeration of material.

ACA interconnections with rigid FR-4 substrate were exposed to thermal cycling between - 40 °C and +125 ºC (Frisk & Kokko, 2006). When the failed test samples were studied using SEM, delamination was found in several test samples after testing. This delamination is probably caused by shear stress between the pad and the bump formed due to differences in the coefficient of thermal expansion of the substrate and the chip. As shown in Figures 16 a), 16 b), and 16 c) with SEM the delamination is seen clearly and its place can be determined. In some cases delamination may be less pronounced and therefore harder to detect. In Figure 16 c) another example of delamination which is more difficult to detect is shown.

Another phenomenon seen in these samples was cracking of the substrate material. This is assumed to be caused by repeated warping of the substrate as the duration of testing was several thousand cycles. This type of cracking was typically found in areas where the glass fibres were far from the surface of the substrate. They started from the corners of the pads and often continued into the substrate until they reached the glass fibres. Examples of such cracking are presented in Figure 16. The formation of such cracks has probably facilitated delamination between pad and the bumb by providing sites for crack initiation, as the cracks often connected to the delamination between the pad and the bump. An example of this is presented in Figures 16 b) and 16 c).

It has been suggested that during thermal testing above the Tg of the ACA matrix sliding between the pad and the bump occurs (Kwon et al, 2006; Uddin et al., 2004). This may break the conductive particles and lead to failure. Furthermore, this phenomenon may cause fretting and thereby the formation of oxides on the conductive surfaces, which increases the resistance of the joints above an acceptable value. When ACA interconnections having nickel particles or gold coated polymer particles, were studied after thermal cycling clear indication of this phenomenon was seen. Figure 17 a) shows a SEM micrograph of polymer particles form a failed ACA interconnection. In the picture cracking of the gold layer in the

Fig. 14. Warpage of the ACA flip chip package during temperature cycling test.

this is presented in Figures 16 b) and 16 c).

ACA interconnections with rigid FR-4 substrate were exposed to thermal cycling between - 40 °C and +125 ºC (Frisk & Kokko, 2006). When the failed test samples were studied using SEM, delamination was found in several test samples after testing. This delamination is probably caused by shear stress between the pad and the bump formed due to differences in the coefficient of thermal expansion of the substrate and the chip. As shown in Figures 16 a), 16 b), and 16 c) with SEM the delamination is seen clearly and its place can be determined. In some cases delamination may be less pronounced and therefore harder to detect. In Figure 16 c) another example of delamination which is more difficult to detect is shown.

Another phenomenon seen in these samples was cracking of the substrate material. This is assumed to be caused by repeated warping of the substrate as the duration of testing was several thousand cycles. This type of cracking was typically found in areas where the glass fibres were far from the surface of the substrate. They started from the corners of the pads and often continued into the substrate until they reached the glass fibres. Examples of such cracking are presented in Figure 16. The formation of such cracks has probably facilitated delamination between pad and the bumb by providing sites for crack initiation, as the cracks often connected to the delamination between the pad and the bump. An example of

It has been suggested that during thermal testing above the Tg of the ACA matrix sliding between the pad and the bump occurs (Kwon et al, 2006; Uddin et al., 2004). This may break the conductive particles and lead to failure. Furthermore, this phenomenon may cause fretting and thereby the formation of oxides on the conductive surfaces, which increases the resistance of the joints above an acceptable value. When ACA interconnections having nickel particles or gold coated polymer particles, were studied after thermal cycling clear indication of this phenomenon was seen. Figure 17 a) shows a SEM micrograph of polymer particles form a failed ACA interconnection. In the picture cracking of the gold layer in the

particle can be clearly seen. Additionally, agglomeration of material under the particles can be seen indicating failure caused by this sliding. Figure 17 b) shows a similar situation for nickel particle and similar agglomeration of material.

Fig. 15. Examples of delamination in test samples with FR-4 substrates after thermal cycling a) clear delamination between ACA and pad, b) delamination between bump and pad which continues to ACA-chip interface, c) less pronounced delamination, and d) delamination varying between the ACA-pad interface and the ACA-bump interface.

#### **4.2 Failures of ACA interconnections with rigid FR-4 substrates under humidity testing**

In addition to temperature changes, humidity has been found to have a major influence on the reliability of ACA interconnections. Under humid conditions the adhesive matrix may deform as it absorbs water. The adhesive matrix may also relax due to the increased temperature. The effect of water absorption depends on the structure of the adhesive and also on the duration of the exposure. The most common adhesive material is epoxy. It has been suggested that the water absorbed in the epoxy polymer has two states to which the water molecules can diffuse. The water may either fill the free volume in the polymer matrix or form hydrogen bonds with the epoxy polymer. If the water is hydrogen-bonded, it causes swelling of the adhesive matrix (Chiang & Fernandez-Garcia, 2002; Luo et al., 2002). This swelling due to humidity typically increases with temperature, but decreases sharply across

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 509

moisture in the substrate also facilitates the moisture absorption of the adhesive matrix, and may accelerate the formation of moisture related failures. The effect of humidity at elevated temperatures on the ACA joints was studied using an 85°C/85RH test. Flexible polyimide was used as a substrate material. When studied after testing using SEM, every test sample with the PI substrate which showed an open interconnection after testing also showed delamination. An example of the delamination is presented in Figure 18. The moisture absorption of polyimide is marked and it is assumed to be the reason for the formation of

Fig. 17. Micrograph of a failed ACA interconnection after thermal cycling a) with polymer

Fig. 18. a) An example of delamination after constant humidity testing on the PI substrate.b) close view of delamination after constant humidity testing on the PI substrate and on

ACA technique was also studied with thinned silicon chips. When silicon chips are thinned below 100 um, they become pliant and can be used in solutions where they are bent. Consequently, they can be used in flexible electronics. Thinning also allows the chips to dissipate more heat, which is important when the densities of the packages increase.

delamination during testing.

particles and b) with nickel particles.

particle.

the Tg of the polymer (Wong et al., 2000). Swelling due to moisture differs significantly between the materials used in flip chip packages causing the formation of hygroscopic stresses in the structure (Mercado et al., 2003; Wong et al., 2000).

Fig. 16. Examples of cracks in FR-4 substrates: a) marked deformation and clear cracking of the epoxy matrix, b) cracking which continues between the pad and the bump, c) less pronounced cracking which continues between the pad and the bump, and d) cracking of thin FR-4 substrate

The swelling of the adhesive matrix may be marked (Mercado et al., 2003) and concurrent with thermal expansion may cause the conductive particles to lose contact. However, it has been reported that the absorbed water may also weaken the mechanical properties of the adhesive. Unlike the hydrogen-bonded water, the water filling the free volume in the epoxy polymer does not cause swelling, as it occupies a volume that already exists (Chiang & Fernandez-Garcia, 2002). However, it acts as a plasticizer affecting the mobility of the chains and increasing chain flexibility, which decreases the Tg of the polymer. As the water acts like a plasticizer it may also impair the mechanical strength of the adhesive.

The substrate material may also have a great influence on the reliability of the joints under humid conditions (Frisk & Cumini, 2006). If the substrate material absorbs water, it penetrates the interfaces more easily and may cause delamination. A large amount of

the Tg of the polymer (Wong et al., 2000). Swelling due to moisture differs significantly between the materials used in flip chip packages causing the formation of hygroscopic

Fig. 16. Examples of cracks in FR-4 substrates: a) marked deformation and clear cracking of the epoxy matrix, b) cracking which continues between the pad and the bump, c) less pronounced cracking which continues between the pad and the bump, and d) cracking of

The swelling of the adhesive matrix may be marked (Mercado et al., 2003) and concurrent with thermal expansion may cause the conductive particles to lose contact. However, it has been reported that the absorbed water may also weaken the mechanical properties of the adhesive. Unlike the hydrogen-bonded water, the water filling the free volume in the epoxy polymer does not cause swelling, as it occupies a volume that already exists (Chiang & Fernandez-Garcia, 2002). However, it acts as a plasticizer affecting the mobility of the chains and increasing chain flexibility, which decreases the Tg of the polymer. As the water acts like

The substrate material may also have a great influence on the reliability of the joints under humid conditions (Frisk & Cumini, 2006). If the substrate material absorbs water, it penetrates the interfaces more easily and may cause delamination. A large amount of

a plasticizer it may also impair the mechanical strength of the adhesive.

thin FR-4 substrate

stresses in the structure (Mercado et al., 2003; Wong et al., 2000).

moisture in the substrate also facilitates the moisture absorption of the adhesive matrix, and may accelerate the formation of moisture related failures. The effect of humidity at elevated temperatures on the ACA joints was studied using an 85°C/85RH test. Flexible polyimide was used as a substrate material. When studied after testing using SEM, every test sample with the PI substrate which showed an open interconnection after testing also showed delamination. An example of the delamination is presented in Figure 18. The moisture absorption of polyimide is marked and it is assumed to be the reason for the formation of delamination during testing.

Fig. 17. Micrograph of a failed ACA interconnection after thermal cycling a) with polymer particles and b) with nickel particles.

Fig. 18. a) An example of delamination after constant humidity testing on the PI substrate.b) close view of delamination after constant humidity testing on the PI substrate and on particle.

ACA technique was also studied with thinned silicon chips. When silicon chips are thinned below 100 um, they become pliant and can be used in solutions where they are bent. Consequently, they can be used in flexible electronics. Thinning also allows the chips to dissipate more heat, which is important when the densities of the packages increase.

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 511

identification (RFID) tags and attachment of driver chips in display applications. ACA interconnections often show the typical trend in electronics for very small size with high functionality. In general this means high density of contacts in a chip and often also a large number of contacts per chip. Such applications are often very challenging both for studies of the interconnection process and the quality and reliability of the interconnections. Making cross-sections of the interconnections has proven to be efficient way to obtain detailed information about the interconnections structures, their quality and failure mechanisms, and this has been used effectively with other characterisation methods such as scanning acoustic microscope, DSC, and x-ray. Due to the small features currently common in electronics applications SEM is often the preferred method for examinations compared to optical

In this chapter several SEM analyses of ACA interconnections were described. Many of these have been critical to both understanding the bonding process of these materials and also for the development of the ACA interconnection techniques and their reliability. Studying cross sections with SEM has been shown to be an effective way to analyse several of the failure mechanisms found in ACA interconnections. However, in general a good understanding of SEM analysis technique is needed for analysis due to the complexity of the ACA structure. ACA interconnection has many interfaces and the failure may occur on any of these or in the bulk materials of the structure. For effective analysis the critical parts need to be already understood when the analysis is made, and therefore, a good understanding of the technique and materials is needed. Lately the use of ACA technology has increased and it has been adopted on new areas for example high temperature electronics and sensor applications. In the future, this will increase the need for detailed knowledge of this interconnection technique. Therefore, it is critical that studies such as presented in this chapter are continued as they give

Making cross-sections of studied samples is favoured in ACA flip chip applications due to the difficulties of studying the interconnections below a component. This also applies to many other interconnections and packaging technologies and similar methods have been used successfully in other applications. For example in the research of lead free solders cross-sections are systematically used for failure analysis and studies related to the microstructure of the interconnections. Additionally, in other techniques, such as flex on board attachments for example, in which flexible substrate is attached to a rigid substrate, and small packaging solutions such as Chip Scale Package (CSP) cross-sectioning and SEM analysis is often needed to determine the structure. In the future, the size of interconnections in electronics will decrease and their number will increase. As a consequence, the need for techniques capable for the analysis of such structures will increase markedly. SEM has proven to be an extremely useful tool for analysing electronics structures as it is relatively

I would like to thank my colleagues Dr. Kati Kokko, M.Sc. Kirsi Saarinen, M.Sc. Janne Kiilunen, and M.Sc. Sanna Lahokallio, and, additionally, my former colleague Dr. Anne Cumini for their help in this research. Furthermore, I would like to thank the staff at the

vital information for both development and applicability of this technique.

fast, typically easily available, and capable for analysing small features.

Institute of Material Science at Tampere University of Technology.

microscopy.

**6. Acknowledgment** 

However, thinned chips have certain drawbacks. They are more fragile than thicker chips, which needs to be taken into account when thinned chips are handled. Special tools may be needed since the fragile edges of the thinned chips are easily broken during handling. During the thinning and dicing processes a considerable amount of stress may be induced in the chips. It has been found that the thinning of the chips changes the shear stress distribution in a package (Frisk et al., 2011). In the analysis 50 μm thick chips were used instead of approximately 500 μm thick chips used in other studies. Very strong delamination was seen in these test samples which indicates failure mechanisms which is different from the one seen with the thicker chips. Examples of this delamination are shown in Figures 19 a) and 19 b). Furthermore cracking of the thinned chips was seen (Figure 19 c) and 19 d)). However, most of this occurred during the bonding process already before testing.

Fig. 19. Micrographs for ACA interconnections with thinned chips: a) marked delamination after humidity testing, b) close view of delamination under particle, c)and d) cracks in the thin chips.

## **5. Conclusion**

Anisotropic conductive adhesives (ACA) are an interesting interconnection method for several applications. Due to the low cost of the ACA process and capability for high density they nowadays dominate many fields for instance attachment of chips in radiofrequency

However, thinned chips have certain drawbacks. They are more fragile than thicker chips, which needs to be taken into account when thinned chips are handled. Special tools may be needed since the fragile edges of the thinned chips are easily broken during handling. During the thinning and dicing processes a considerable amount of stress may be induced in the chips. It has been found that the thinning of the chips changes the shear stress distribution in a package (Frisk et al., 2011). In the analysis 50 μm thick chips were used instead of approximately 500 μm thick chips used in other studies. Very strong delamination was seen in these test samples which indicates failure mechanisms which is different from the one seen with the thicker chips. Examples of this delamination are shown in Figures 19 a) and 19 b). Furthermore cracking of the thinned chips was seen (Figure 19 c) and 19 d)).

However, most of this occurred during the bonding process already before testing.

Fig. 19. Micrographs for ACA interconnections with thinned chips: a) marked delamination after humidity testing, b) close view of delamination under particle, c)and d) cracks in the

Anisotropic conductive adhesives (ACA) are an interesting interconnection method for several applications. Due to the low cost of the ACA process and capability for high density they nowadays dominate many fields for instance attachment of chips in radiofrequency

thin chips.

**5. Conclusion** 

identification (RFID) tags and attachment of driver chips in display applications. ACA interconnections often show the typical trend in electronics for very small size with high functionality. In general this means high density of contacts in a chip and often also a large number of contacts per chip. Such applications are often very challenging both for studies of the interconnection process and the quality and reliability of the interconnections. Making cross-sections of the interconnections has proven to be efficient way to obtain detailed information about the interconnections structures, their quality and failure mechanisms, and this has been used effectively with other characterisation methods such as scanning acoustic microscope, DSC, and x-ray. Due to the small features currently common in electronics applications SEM is often the preferred method for examinations compared to optical microscopy.

In this chapter several SEM analyses of ACA interconnections were described. Many of these have been critical to both understanding the bonding process of these materials and also for the development of the ACA interconnection techniques and their reliability. Studying cross sections with SEM has been shown to be an effective way to analyse several of the failure mechanisms found in ACA interconnections. However, in general a good understanding of SEM analysis technique is needed for analysis due to the complexity of the ACA structure. ACA interconnection has many interfaces and the failure may occur on any of these or in the bulk materials of the structure. For effective analysis the critical parts need to be already understood when the analysis is made, and therefore, a good understanding of the technique and materials is needed. Lately the use of ACA technology has increased and it has been adopted on new areas for example high temperature electronics and sensor applications. In the future, this will increase the need for detailed knowledge of this interconnection technique. Therefore, it is critical that studies such as presented in this chapter are continued as they give vital information for both development and applicability of this technique.

Making cross-sections of studied samples is favoured in ACA flip chip applications due to the difficulties of studying the interconnections below a component. This also applies to many other interconnections and packaging technologies and similar methods have been used successfully in other applications. For example in the research of lead free solders cross-sections are systematically used for failure analysis and studies related to the microstructure of the interconnections. Additionally, in other techniques, such as flex on board attachments for example, in which flexible substrate is attached to a rigid substrate, and small packaging solutions such as Chip Scale Package (CSP) cross-sectioning and SEM analysis is often needed to determine the structure. In the future, the size of interconnections in electronics will decrease and their number will increase. As a consequence, the need for techniques capable for the analysis of such structures will increase markedly. SEM has proven to be an extremely useful tool for analysing electronics structures as it is relatively fast, typically easily available, and capable for analysing small features.

### **6. Acknowledgment**

I would like to thank my colleagues Dr. Kati Kokko, M.Sc. Kirsi Saarinen, M.Sc. Janne Kiilunen, and M.Sc. Sanna Lahokallio, and, additionally, my former colleague Dr. Anne Cumini for their help in this research. Furthermore, I would like to thank the staff at the Institute of Material Science at Tampere University of Technology.

Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 513

Ishibashi, K. and Kimura, J. (1996). A New Anisotropic Conductive Film with Arrayed

Jin, S., Tiefel, T.H, Li-Ilan C., and Dahringer, D.W. (1993). Anisotropically Conductive

Kim, J., Kwon, S., and Ihm, D. (2004). Reliability and Thermodynamic Studied of an

Kwon, W. and Paik, K. (2004). Contraction stress build-up of anisotropic conductive films

Kwon, W., Yim, M., Paik, K., Ham, S., and Lee, S. (2005). Thermal Cycling Reliability and

Kwon, W., Ham, S., and Paik, K. (2006). Deformation Mechanism and its Effect on Electrical

Kwon, W. and Paik, K. (2006). Experimental Analysis of Mechanical and Electrical

Lai, Z. & Liu J. (1996). Anisotropically conductive adhesive flip-chip bonding on rigid and

Lai, Z., Lai, R., Persson, K., and Liu, J. (1998). Effect of Bump Height on the Reliability of

*Electronics Manufacturing*, Vol. 8, No. 3-4, 1998, pp. 217-24. ISSN 0960-3131 Lau, J. (1995). *Flip Chip Technologies*, McGraw-Hill, ISBN 0070366098, New York, USA. Lau, J. (2000). *Low Cost Flip Chip Technologies for DCA, WLCSP, and PBGA Assemblies*,

Lau, J., Wong, C.P., Lee, N., and Lee, R. (2003*). Electronics Manufacturing: with Lead-Free,* 

*Packaging*, Vol. 127, No. 2., 2005, pp.86-90. ISSN 1043-7398

29, No. 3, 2006, pp. 528-34. ISSN 1521-3331

McGraw-Hill, ISBN 0071351418, New York.

*Manufacturing Technology*, Vol. 19, No. 4, 1996, pp.752–7. ISSN 2156-3950 Jang K-W., C-K. Chung, W-S. Lee, K-W.Paik. (2008). Material properties of anisotropic

ISSN 0026-2714

977. ISSN 0148-6411

ISSN 0924-0136

ISSN 0026-2714

66. ISSN 2156-3950

0071386246, New York.

4628

Conductive Particles. *IEEE Transactions on Components, Packaging, and* 

conductive films (ACFs) and their flip chip assembly reliability in NAND flash memory applications. *Microelectronics Reliability*, Vol. 48, No. 7, 2008, pp. 1052-61.

Polymer Films with a Uniform Dispersion of Particles. *IEEE Transactions on Components, Hybrids, and Manufacturing Technology*, Vol. 16, No. 8, 1993, pp.972 -

Anisotropic Conductive Adhesive Film (ACAF) Prepared from Epoxy/Rubber Resins. *Journal of Material Processing Technology*, Vol. 152, No. 3, 2004, pp.357-62.

(ACFs) for flip-chip interconnection: Effect of thermal and mechanical properties of ACFs. *Journal of Applied Polymer Science*, Vol. 93, No. 6, 2004, pp. 2634-41. ISSN 1097-

Delamination of Anisotropic Conductive Adhesives Flip Chip on Organic Substrates with Emphasis on the Thermal Deformation. *Journal of Electronic* 

Conductivity of ACF Flip Chip Package Under Thermal Cycling Condition: An experimental Study. *Microelectronics Reliability*, Vol. 46, No.2-4, 2006, pp.589-99.

Characteristics of Metal-coated Conductive Spheres for Anisotropic Conductive Adhesives. *IEEE Transactions on Components and Packaging Technologies*, Volume

flexible printed circuit substrate. *IEEE Transactions on Components, Packaging, and Manufacturing Technology*, *Part B: Advanced Packaging*, Vol. 19, No. 3, 1996, pp. 644-

ACA Flip Chip Joining with FR4 Rigid and Polyimide Flexible Substrate. *Journal of* 

*Halogen-Free, and Conductive-Adhesive Materials*. McGraw-Hill Professional, ISBN

#### **7. References**


Asai, S., Saruta, U., Tobita, M., Takano, M., and Miyashita, Y. (1995) Development of an

Chan, Y. C. and Luk, D. Y. (2002a) Effects of Bonding Parameters on the Reliability

Chan, Y. C. and Luk, D. Y. (2002b) Effects of Bonding Parameters on the Reliability

Chiang, M. and Fernandez-Garcia, M. (2002) Relation of swelling and Tg depression to the

Connell, G., Zenner, R., and Gerber, J. (1997). Conductive Adhesive Flip Chip Bonding for

*Technology Conference*, ISBN: 0-7803-3857-X, San Jose, CA, USA, May, 1997. Coombs, C. (2001). *Coombs' Printed Circuits Handbook*. McGraw-Hill Professional, 5th edition,

Divigalpitiya, R. and Hogerton, P. (2004). Contact Resistance of Anisotropic Conductive

Frisk, L. and Ristolainen, E. (2005) .Flip Chip Attachment on Flexible LCP-Substrate using an ACF. *Microelectronics Reliability*, Vol. 45, No. 3-4, 2005, pp. 583-8. ISSN 0026-2714 Frisk, L. and Kokko, K. (2006). The Effects of Chip and Substrate Thickness on the Reliability

Frisk L. and A. Seppälä A. (2006). Reliability of Flip Chip Joints on LCP and PI Substrate.

Frisk L. and Kokko K. (2007). Effect of RCC on the reliability of adhesive flip chip joints. *Journal of Electronic Packagin*g, Vol. 129 No. 3, 2007, pp. 260-5. ISSN 1043-7398 Frisk, L. and Cumini, A. "Effect of Substrate Material and Thickness on Reliability of ACA

Frisk L. and Cumini A. (2010). Reliability of ACF interconnections on FR-4 substrates. *IEEE* 

Frisk, L., Saarinen K., and Kokko K. (2011). Reliability of ACA joined thinned chips on rigid

*Packaging Conference, EMPC*, Brighton, UK, September 2011.

*Polymer Science*, Vol. 87, No. 9, 2002, pp. 1436 - 44. ISSN 1097-4628

*Applied Polymer Science*, Vol. 56, No. 7, 1995, pp. 769-77. ISSN 1097-4628. Chan, Y., Hung, K., Tang, C., and Wu, C. (2000) Degradation Mechanisms of Anisotropic

*Manufacturing*, ISBN 0780364600, Espoo, Finland, June, 2000.

*Reliability*, Vol. 42, No. 8, 2002, pp.1185-94. ISSN 0026-2714

Vol. 42, No. 8, 2002, pp.1195-1204. ISSN 0026-2714

ISBN 0071350160.

0911

2009 pp. 15-23.

147. ISSN 1521-3331

pp. 194-9. ISSN 1551-4897.

No. 4, 2006, pp. 28-37. ISSN 0954-0911

Anisotropic Conductive Adhesive Film (ACAF) from Epoxy Resins. *Journal of* 

Conductive Adhesive Joints for Flip Chip on Flex application*s*. *Proceedings of 4th IEEE Conference on Adhesive Joining and Coating Technology in Electronics* 

Performance of Anisotropic Conductive Adhesive Interconnects for Flip-chip-onflex Packages Assembly I. Different Bonding Temperature. *Microelectronics* 

Performance of Anisotropic Conductive Adhesive Interconnects for Flip-chip-onflex Packages Assembly II. Different Bonding Pressure. *Microelectronics Reliability*,

apparent free volume of a particle-filled, epoxy-based adhesive. *Journal of Applied* 

Bumped and Unbumped Die. *Proceedings of the 47th IEEE Electronic Components and* 

Adhesives. *Journal of Microelectronics and Electronic Packaging*, Vol. 1, No. 3, 2004,

of ACA Bonded Flip Chip Joints. *Soldering and Surface Mount Technology*, Vol. 18,

*Soldering and Surface Mount Technology*, Vol. 18, No. 4, 2006, pp. 12-20. ISSN 0954-

Bonded Flip Chip Joints", Soldering and Surface Mount Technology, Vol. 21, No. 3,

*Transactions on Components and Packaging Technologies*, Vol. 33, No. 1, 2010, pp. 138-

substrates under humid conditions. *Proceedings of European Microelectronics and* 

**7. References** 


Study of Structure and Failure Mechanisms in ACA Interconnections Using SEM 515

Suhir, E. (2002). Accelerated Life Testing (ALT) in Microelectronics and Photonics: Its role,

Takagi, K., Honma, H., and Sasabe, T. (2003). Development of Sequential Build-up

Tan, S. C., Chan, Y. C., Chiu, Y. W., and Tan, C. W. (2004). Thermal Stability Performance of

Uddin, M. A., Alam, M. O., Chan Y. C., and Chan, H. P. (2004). Adhesion Strength and

Wang, X. Wang, Y., Chen, G., Liu, J., and Lai, Z. (1998). Quantitative Estimate of the

Watanabe, I. Fujinawa, T. Arifuku, M. Fujii, M. Yasushi, G. (2004). Recent advances of

Whalley, D., Mannan, S., and Williams, D. (1997). Anisotropic Conducting Adhesives for

Wong, E. H., Chan, K. C., Rajoo, R., and Lin, T. B. (2000). The Mechanics and Impact of

Wu, S. X., Zhang, C. Yeh, C, Wille, S., and Wyatt, K. (1997). Cure Kinetics and Mechanical

Yim, M. J. and Paik, K. W. (1998). Design and understanding of anisotropic conductive films

Yim, M. J. and Paik, K. W. (1999). The Contact Resistance and Reliability of Anisotropically

Yim, M. J. and Paik, K. W. (2001). Effect of Nonconducting Filler Additions on ACA

*Technologies*, Vol. 21, No. 2, 1998, pp.226-34. ISSN 1521-3331

*Reliability*, Vol. 44, No. 3, 2004, pp. 495-503. ISSN 0026-2714

*Electronic Packaging*, Vol. 124, No. 3, 2002, pp. 281-91. ISSN 1043-7398 Sungwook, M. and Chappell, W. J. (2010). Novel Three-Dimensional Packaging Approaches

pp. 3815-23. ISSN 0018-9480

pp.505-14. ISSN 0026-2714

Georgia, USA, August 2004.

Vegas, NV, USA, May, 2000.

2, 1999, pp.166-73, ISSN 1521-3323.

0144-5154

May, 1997.

19, No. 5, 2003, pp. 27 – 56. ISSN 0883-7554

21, no. 2, 1998, pp.248 – 51. ISSN 1070-9886

Attributes, Challenges, Pitfalls, and Interaction with Qualification Tests*. Journal of* 

Using Magnetically Aligned Anisotropic Conductive Adhesive for Microwave Applications. *IEEE Transactions on Microwave Theory and Techniques*, vol. 58, No. 12,

Multilayer Printed Wiring Boards in Japan. *IEEE Electrical Insulation Magazine*, Vol.

Anisotropic Conductive Film at Different Bonding Temperatures. *Microelectronics* 

Contact Resistance of Flip Chip on Flex Packages – Effect of Curing Degree of Anisotropic Conductive Film. *Microelectronics Reliability*, Vol. 44, No. 3, 2004,

Characteristics of Conductive Particles in ACA by using Nano Indenter. *IEEE Transactions on Components and Packaging, and Manufacturing Technology* , Vol.

interconnection technologies using anisotropic conductive films in flat panel display applications. *Proceedings of the 9th IEEE International symposium Advanced Packaging Materials: Processes, Properties and Interfaces*, ISBN 0780384369, Atlanta,

Electronic Assembly. *Assembly Automation*, Vol. 17, No. 1, 1997, pp.66-74. ISSN

Hygroscopic Swelling of Polymeric Materials in Electronic Packaging. *Proceedings of the 50th IEEE Electronics Components and Technology Conference*, ISBN 0780359089, Las

Properties of Conductive Adhesive. *Proceedings of the 47th IEEE Electronics Components and Technology Conference*, ISBN 0-7803-3857-X, San Jose, CA, USA,

(ACF's) for LCD packaging. *IEEE Transactions on Components and Packaging* 

Conductive Film (ACF). *IEEE Transactions on Advanced Packaging*, Vol. 22, No.

Properties and the Reliability of ACA Flip Chip on Organic Substrates. *IEEE* 


Li, Y. and Wong, C.P. (2006). Recent Advances of Conductive Adhesives as a Lead-free

Applications. *Materials Science and Engineering R 51*, 2006, ISSN 0927-796X. Licari, J. J. and Swanson, D. W. (2005). *Adhesives Technology for Electronic Applications*, William Andrew Publishing, ISBN 0815515138, Norwich, New York. Liu, J., Tolvgard, A., Malmodin, J., and Lai, Z. (1999). A Reliable and Environmentally

Liu, J. (2001). ACA Bonding Technology for Low Cost Electronics Packaging Applications –

Lu, D. (2006). Overview of Recent Advances on Isotropic Conductive Adhesives. *Proceeding* 

*Failure Analysis (HDP)*, ISBN 9781424404889, Shanghai, China, June 2006. Luo, S., Leisen, J., and Wong, C. (2002). Study on Mobility of Water and Polymer Chain in

Mercado, Lei., white, J., Sarihan, V., and Lee, T. (2003). Failure Mechanism Study of

Ogunjimi, A. O., Mannan, S. H., Whalley, D. C., & Williams, D. J. (1996). Assembly of Planar

Pinardi, K., Liu, J., Haug, R., Treutler, C., and Willander, M. (1998). Deformation study of

Rizvi, M. J., Chan, Y. C., Bailey, C., Lu, H., and Sharif, A. (2005). The Effect of Curing on

Saarinen, K., Frisk, L. and Ukkonen, L. (2011). Effects of different combinations of

*Interfaces*, ISBN 0780384369, Atlanta, Georgia, USA, January, 2004.

*Technology*, Vol. 19, No 4, 1996, pp. 257-63. ISSN 2156-3950

Binghamton, ISBN 0780349342, NY, USA, September, 1998.

*Packaging Technologies*, Vol. 26, No. 3, 2003, pp. 509-16. ISSN 1521-3331 Morris, J., Lee, J., and, Liu, J. (2005). Isotropic Conductive Adhesive Interconnect Technology

Vol. 22, No. 2, 1999, pp. 186-90, ISSN 1521-3331.

Vol. 13, No. 3, 2001, pp. 39-57, ISSN 0954-0911.

1, 2002, pp.1-8. ISSN 1097-4628

Wrocaw, Poland, January 2006.

0911

Alternative in Electronic Packaging: Material, Processing, Reliability and

Friendly Packaging Technology - Flip Chip Joining Using Anisotropically Conductive Adhesive", *IEEE Transactions on Components and Packaging Technology*,

Current Status and Remaining Challenges. *Soldering & Surface Mount Technology*,

*of 6th Conference on High Density Microsystem Design and Packaging and Component* 

Epoxy and Its Influence on Adhesion. *Journal of Applied Polymer Science*, Vol. 85, No.

Anisotropic Conductive Film (ACF) Packages*. IEEE Transactions on Components and* 

in Electronics Packaging Applications. *Proceedings of the 5th International Conference on Polymers and Adhesives in Microelectronics and Photonics*, ISBN 0780395530,

Array Components Using Anisotropic Conducting Adhesives-A Benchmark Study: Part I-Experiment. *IEEE Transactions on Components, Packaging, and Manufacturing* 

the PCB during the flip chip assembly processing anisotropically conductive adhesive (ACA) as a bonding agent. Proceedings of 3rd IEEE Conference on Adhesive Joining and Coating Technology in Electronics Manufacturing,

the Performance of ACF Bonded Chip-on-flex Assemblies After Thermal Ageing. *Soldering and Surface Mount Technology*, 2005, Vol. 17, No. 2, pp.40-8. ISSN 0954-

environmental test on the reliability of UHF RFID tags. *Procedings of 17th European Microelectronics and Packaging Conference, EMPC*, Brighton, UK September, 2011. Savolainen, P. (2004). Display driver packaging: ACF reaching the limits?. *Proceedings of the* 

*9th International Symposium on Advanced Packaging Materials: Processes, Properties and* 


**26** 

*India* 

**Exploring the Superconductors** 

Shiva Kumar Singh1,2,\*, Devina Sharma1, M. Husain2,

H. Kishan1, Ranjan Kumar3 and V.P.S. Awana1

*2Department of Physics, Jamia Millia Islamia, New Delhi, 3Department of Physics, Panjab University, Chandigrah,* 

*National Physical Laboratory (CSIR), New Delhi,* 

*1Quantum Phenomena and Applications,* 

**with Scanning Electron Microscopy (SEM)** 

The characterization of materials supports their development and in particular of superconductors, for their technological applications. Scanning electron microscopy (SEM) is one of these characterization techniques, whose data is used to estimate the properties, determine the shortcomings and hence improve the material. The phenomenon of superconductivity initially develops within the grain and eventually crosses over the grain boundaries, leading to the bulk. Hence SEM can be a useful tool to probe the microstructure of the superconductors and the properties related to it. Along with this the Energydispersive Spectroscopy (EDS) can tell about the chemical composition of compounds. Grain size and its connectivity can be seen through SEM and can be correlated with the corresponding properties. The superconducting materials developed for practical applications are some of the complex materials used today. These materials have large number of potential variables such as their processing conditions, composition, structure etc., whose dependence on the superconducting properties have to be analyzed critically. The characterization techniques are the tools that help to reveal and explore both the macro and microstructure of materials. It is known that the larger grains (reduction in grain boundaries) lead to increased pinning type behavior with enhanced *Jc* [1]*.* In contrast Rosko *et al.* [2] reported that *Jc* is determined by weak links and grain size has little role on it. Also, Smith *et al.* [3] interpreted reduction of *Jc* and activation of weak link type behavior with increasing grain size for YBa2Cu3O7-<sup>δ</sup> (YBCO) polycrystalline samples in terms of microcracks in large grains. The superconducting parameters are broadly divided into two categories; first, the intrinsic parameters such as penetration depth (*λ*), which are intrinsic to the material and are not affected by, grain size. On the other hand, values such as shielding/Meissner fraction, the inter- and intra-grain critical current density and diamagnetic fraction depend upon particle size of bulk superconductors. Thus SEM can be

very important to probe and in understanding the superconducting phenomena.

**1. Introduction** 

 \*

Corresponding Author

*Transactions on Components and Packaging Technologies*, Vol. 24, No. 1, 2001, pp.24-32. ISSN 1521-3331


## **Exploring the Superconductors with Scanning Electron Microscopy (SEM)**

Shiva Kumar Singh1,2,\*, Devina Sharma1, M. Husain2,

H. Kishan1, Ranjan Kumar3 and V.P.S. Awana1

*2Department of Physics, Jamia Millia Islamia, New Delhi,* 

*3Department of Physics, Panjab University, Chandigrah,* 

*India* 

## **1. Introduction**

516 Scanning Electron Microscopy

Yin, C.Y., Alam, M.O., Chan, Y.C., Bailey, C. and Lu, H. (2003). The Effect of Reflow Process

Zhong, Z. W. (2005). Various Adhesives for Flip Chips. *Journal of Electronic Packaging*, Vol.

*Reliability*, Vol. 43, No. 4, 2003, pp. 625-33. ISSN 0026-2714

127, No. 1, 2005, pp. 29-32. ISSN 1043-7398

ISSN 1521-3331

*Transactions on Components and Packaging Technologies*, Vol. 24, No. 1, 2001, pp.24-32.

on the Contact Resistance and Reliability of Anisotropic Conductive Film Interconnection for Flip Chip on Flex Applications. *Journal of Microelectronics* 

> The characterization of materials supports their development and in particular of superconductors, for their technological applications. Scanning electron microscopy (SEM) is one of these characterization techniques, whose data is used to estimate the properties, determine the shortcomings and hence improve the material. The phenomenon of superconductivity initially develops within the grain and eventually crosses over the grain boundaries, leading to the bulk. Hence SEM can be a useful tool to probe the microstructure of the superconductors and the properties related to it. Along with this the Energydispersive Spectroscopy (EDS) can tell about the chemical composition of compounds. Grain size and its connectivity can be seen through SEM and can be correlated with the corresponding properties. The superconducting materials developed for practical applications are some of the complex materials used today. These materials have large number of potential variables such as their processing conditions, composition, structure etc., whose dependence on the superconducting properties have to be analyzed critically. The characterization techniques are the tools that help to reveal and explore both the macro and microstructure of materials. It is known that the larger grains (reduction in grain boundaries) lead to increased pinning type behavior with enhanced *Jc* [1]*.* In contrast Rosko *et al.* [2] reported that *Jc* is determined by weak links and grain size has little role on it. Also, Smith *et al.* [3] interpreted reduction of *Jc* and activation of weak link type behavior with increasing grain size for YBa2Cu3O7-<sup>δ</sup> (YBCO) polycrystalline samples in terms of microcracks in large grains. The superconducting parameters are broadly divided into two categories; first, the intrinsic parameters such as penetration depth (*λ*), which are intrinsic to the material and are not affected by, grain size. On the other hand, values such as shielding/Meissner fraction, the inter- and intra-grain critical current density and diamagnetic fraction depend upon particle size of bulk superconductors. Thus SEM can be very important to probe and in understanding the superconducting phenomena.

*<sup>1</sup>Quantum Phenomena and Applications,* 

*National Physical Laboratory (CSIR), New Delhi,* 

<sup>\*</sup> Corresponding Author

Exploring the Superconductors with Scanning Electron Microscopy (SEM) 519

organizations, have developed standard test methods describing how to characterize microstructures quantitatively. The methods for grain size measurement are described in great detail in the ASTM Standard, E112, "Standard Test Methods for Determining Average Grain Size" [5]. The information below will provide cursory explanation to the methods for determining grain size in *ASTM* E112. The microstructural

*n* = 2G-1 Where, n is the number of grains per square inch measured at a magnification of 100x. Grain size or the value of G is most commonly measured by (a) Planimetric method and by (b)

In the Planimetric method [Fig. 1 (a)] (developed by Zay Jeffries in 1916), a count is made of the number of grains completely within a circle of known area and half of the number of grains intersected by the circle to obtain NA. Then, NA is related to G. This method is slow when done manually because the grains must be marked when counted to obtain an

quantity known as the *ASTM* micro gain size number, G is defined as

accurate count. This method is described in the section 9 of *ASTM E112*.

ii. The numbers of grains are counted that are completely within the area (ninside). iii. The number of grains is counted that are partially within the area (nintercepted).

iv. The number of grains per sq. mm, NA, is calculated from NA = f {ninside + ½(nintercepted)} v. The multiplier f is calculated from (M2/circle area), where M is the linear magnification

vi. From NA, we can calculate the *ASTM* grain size number, G, using the following formula

The *ASTM* grain size can also be determined using the intercept method (developed by Emil Heyn in 1904) counting either the number of grains intercepted (N) or the number of grain boundaries intersected (P) with a test line. *ASTM* recommends using a grid with three concentric circles (as shown in the Fig. 1 (b)) with a 500mm total line length. The count of the number of grains intercepted by the circle is N. To calculate the number of interceptions per mm, NL, we divide N by the true length of the circle. The true length (LT) is obtained by dividing the circumference of the circle by the magnification, M. Hence, NL = N/LT interceptions per mm. To calculate the grain size, we first determine the mean linear intercept length, *l*, which is the reciprocal of NL (or of PL, the number of grain boundary

G = {-6.644 (log10 *l*) – 3.288},

*ASTM E112* provides table that relates grains/in2 @ 100x and grains/mm2 @ 1x to *ASTM* grain size number G. Since the two methods are sensing different geometric aspects of the

intersections per unit length). G is calculated from an equation from E 112-96:

The basic steps of the procedure are as follows:

i. A circle of known size is inscribed over the SEM image.

from E 112-96:G = {3.322 (log10 NA) – 2.954}

Intercept method.

**(a) Planimetric method** 

of the image.

where, *l* is in mm.

**(b) Linear intercept method** 

In SEM electron beam is scanned across a sample's surface. When the electrons strike the sample, a verity of signals arises and produces elemental composition of the sample. *SEM*  with *EDS* is a major tool for qualitative and quantitative analyses which is done by bombarding a finely focused electron beam (electron probe) on the specimen, and measuring the intensities of the characteristic X-ray emitted. The three signals in SEM are the secondary electrons, backscattered electrons and X-rays, provide the greatest amounts of information. Secondary electrons are emitted from the atoms occupying the top surface and produce interpretable image of the surface. The contrast in the image is determined by the sample morphology. Backscattered electrons are primary electrons which are "reflected" from atoms in the solid. The contrast in the image produced is determinate by the atomic numbers of the elements in the sample. Therefore the image shows the distribution of different chemical elements in the sample. Since these electrons are emitted from the depth of the sample, the resolution of the image is not as good as for secondary electrons.

This chapter deals with the ability of SEM in extracting the information from superconductors. Since, the microstructure and topology of the materials determine largely its properties in terms of its grains and their connectivity, the utility of SEM in studying the properties of various superconductors discovered till date will be reviewed. The limitations will also be discussed. How other characterization tools, can provide better information along with *SEM*, will be explored.

## **2. SEM: As a characterization tool for superconductors**

Some of the important aspects with which *SEM* deals with, is the grain size, morphology and alignment, structural defects, chemical composition. While trying to optimize the transport properties, grain to grain alignment within the superconductor has to be considered. It is important to analyze the grain alignment and enhance it in order to relate it to the improvement of transport properties. Defects play an important role in determining the properties of superconductors, especially those of HTSc. While macro defects such as porosity, cracks, secondary phases etc. may adversely affect the transport critical super currents, on the other hand microscopic defects such as addition of nanoparticles, dislocation etc., can prove to be beneficial. As the size of the defects is smaller than the coherence length of the superconductor, they may act as pinning centers, thereby enhancing the critical current. Chemical composition of the material within the superconducting grain and at the grain boundaries has significant effect over its properties. The compositions and the change in the compositions can be measured or inferred from a variety of techniques such as energy dispersive spectroscopy. Backscattered electron imaging (BEI) can also be used to infer chemical compositional variations. In all, the complexity in the superconducting materials requires continuous research into their fundamental properties and evolution of new improved materials. For all these evaluations, various characterization tools have to be relied upon among which SEM and allied techniques such as EDS and BEI play an important role.

#### **2.1 Determination of grain size from** *SEM* **micrographs**

Grain size determination is perhaps one of the most commonly performed microstructural measurements from *SEM* micrographs. Standards organizations, including American Society for Testing and Materials (ASTM) [4] and some other national and international organizations, have developed standard test methods describing how to characterize microstructures quantitatively. The methods for grain size measurement are described in great detail in the ASTM Standard, E112, "Standard Test Methods for Determining Average Grain Size" [5]. The information below will provide cursory explanation to the methods for determining grain size in *ASTM* E112. The microstructural quantity known as the *ASTM* micro gain size number, G is defined as

$$n = \mathfrak{Z}^{G1}$$

Where, n is the number of grains per square inch measured at a magnification of 100x. Grain size or the value of G is most commonly measured by (a) Planimetric method and by (b) Intercept method.

#### **(a) Planimetric method**

518 Scanning Electron Microscopy

In SEM electron beam is scanned across a sample's surface. When the electrons strike the sample, a verity of signals arises and produces elemental composition of the sample. *SEM*  with *EDS* is a major tool for qualitative and quantitative analyses which is done by bombarding a finely focused electron beam (electron probe) on the specimen, and measuring the intensities of the characteristic X-ray emitted. The three signals in SEM are the secondary electrons, backscattered electrons and X-rays, provide the greatest amounts of information. Secondary electrons are emitted from the atoms occupying the top surface and produce interpretable image of the surface. The contrast in the image is determined by the sample morphology. Backscattered electrons are primary electrons which are "reflected" from atoms in the solid. The contrast in the image produced is determinate by the atomic numbers of the elements in the sample. Therefore the image shows the distribution of different chemical elements in the sample. Since these electrons are emitted from the depth

of the sample, the resolution of the image is not as good as for secondary electrons.

**2. SEM: As a characterization tool for superconductors** 

**2.1 Determination of grain size from** *SEM* **micrographs** 

along with *SEM*, will be explored.

play an important role.

This chapter deals with the ability of SEM in extracting the information from superconductors. Since, the microstructure and topology of the materials determine largely its properties in terms of its grains and their connectivity, the utility of SEM in studying the properties of various superconductors discovered till date will be reviewed. The limitations will also be discussed. How other characterization tools, can provide better information

Some of the important aspects with which *SEM* deals with, is the grain size, morphology and alignment, structural defects, chemical composition. While trying to optimize the transport properties, grain to grain alignment within the superconductor has to be considered. It is important to analyze the grain alignment and enhance it in order to relate it to the improvement of transport properties. Defects play an important role in determining the properties of superconductors, especially those of HTSc. While macro defects such as porosity, cracks, secondary phases etc. may adversely affect the transport critical super currents, on the other hand microscopic defects such as addition of nanoparticles, dislocation etc., can prove to be beneficial. As the size of the defects is smaller than the coherence length of the superconductor, they may act as pinning centers, thereby enhancing the critical current. Chemical composition of the material within the superconducting grain and at the grain boundaries has significant effect over its properties. The compositions and the change in the compositions can be measured or inferred from a variety of techniques such as energy dispersive spectroscopy. Backscattered electron imaging (BEI) can also be used to infer chemical compositional variations. In all, the complexity in the superconducting materials requires continuous research into their fundamental properties and evolution of new improved materials. For all these evaluations, various characterization tools have to be relied upon among which SEM and allied techniques such as EDS and BEI

Grain size determination is perhaps one of the most commonly performed microstructural measurements from *SEM* micrographs. Standards organizations, including American Society for Testing and Materials (ASTM) [4] and some other national and international In the Planimetric method [Fig. 1 (a)] (developed by Zay Jeffries in 1916), a count is made of the number of grains completely within a circle of known area and half of the number of grains intersected by the circle to obtain NA. Then, NA is related to G. This method is slow when done manually because the grains must be marked when counted to obtain an accurate count. This method is described in the section 9 of *ASTM E112*.

The basic steps of the procedure are as follows:


#### **(b) Linear intercept method**

The *ASTM* grain size can also be determined using the intercept method (developed by Emil Heyn in 1904) counting either the number of grains intercepted (N) or the number of grain boundaries intersected (P) with a test line. *ASTM* recommends using a grid with three concentric circles (as shown in the Fig. 1 (b)) with a 500mm total line length. The count of the number of grains intercepted by the circle is N. To calculate the number of interceptions per mm, NL, we divide N by the true length of the circle. The true length (LT) is obtained by dividing the circumference of the circle by the magnification, M. Hence, NL = N/LT interceptions per mm. To calculate the grain size, we first determine the mean linear intercept length, *l*, which is the reciprocal of NL (or of PL, the number of grain boundary intersections per unit length). G is calculated from an equation from E 112-96:

$$\mathbf{G} = \{\text{-}6.644 \, (\log 10 \, l) - 3.288\}\_{l}$$

where, *l* is in mm.

*ASTM E112* provides table that relates grains/in2 @ 100x and grains/mm2 @ 1x to *ASTM* grain size number G. Since the two methods are sensing different geometric aspects of the

Exploring the Superconductors with Scanning Electron Microscopy (SEM) 521

detector) and the X-axis shows the energy level of those counts. The EDS software associates

In this section we will discuss the *SEM* study of superconductors of different families. For high temperature superconductors (HTSc), *SEM* has been widely used to explore the superconducting behavior. Grain size matters a lot in deciding the superconducting parameters of cuprate HTSc. Decrease in shielding current is observed with decrease in particle size [6]. Magnesium diboride (MgB2) represent an attractive alternative to low temperature superconductors. For most of the practical applications, high critical current density (*J*c) in the presence of a magnetic field along with high upper critical field (*H*C2) and high irreversibility field (*H*irr) are required. Moderate impurities and nano (*n*) materials are being used to improve these parameters [7-14]. In particular, significant flux pinning enhancement in MgB2 is observed with *n-SiC* addition [15]. In case of newly discovered pnictides superconductors, *SEM* is being used in an ingenious way. The spatially resolved electrical transport properties have been studied on the surface of optimally-doped superconducting Ba(Fe1-*x*Co*x*)2As2 single crystal by using a four-probe scanning tunneling

**3.1 Grain size and gain connectivity in cuprate high temperature superconductors** 

One of the characteristics of high temperature superconductors (HTSc) is their small coherence length which is comparable with the unit cell. The coherence length is a key parameter for the performance of superconductors for applications, since this determines the size of the normally conducting core of the flux lines [17]. In order to control the motion of flux lines one needs a microstructure with defects as small as the coherence length. The extremely small coherence length of HTSc, which is for YBCO only 2.7 nm at 77 K within the ab-plane, is the reason that defects such as grain boundaries, which are very beneficial in low-temperature superconductors because they act as pinning defects, serve as weak links and limit the critical current, especially in the presence of an external magnetic field. HTSc bulk material can therefore be considered as a matrix of superconducting grains embedded

The cuprate superconductors belong to the family of HTSc in which Cu-O chains and planes are responsible for the conduction of super currents. The size as well as the shape of the grains varies in different cuprate HTSc [18-20]. This variation in microstructure in turn leads to different superconducting behavior. Although the nature of the occurrence of superconductivity in cuprate HTSc is same, *Tc* varies from 38 K in LSCO to 110 K in BSCCO system. This wide range of *Tc* itself indicates that micro-structural parameters are of much importance. In this section various cuprate HTSc will be discussed in the chronological

Discovered by Bednorz and Muller in 1986 [21], La2-xSrxCuO4 was the trendsetter breakthrough in the history of superconductivity leading to the new era of High Tc superconductivity. It has *Tc* of 38 K which is beyond the BCS limit of 30 K. Although it has

order of their discovery, in terms of the use of SEM for their characterization.

the energy level of the X-rays with the elements and shell levels that generated them.

**3. Exploring superconductors using** *SEM*

microscopy [16]. Results will be discussed in following subtitles.

in a non superconducting material.

**A. La2-xSrxCuO4**

three-dimensional grain structure, they will not give exactly the same value, but they will be close, generally within the experimental limitations of the measurements. In practice, these measurements are repeated on a number of fields in order to obtain a good estimate of the grain size.

(a)

Fig. 1. **(a)** Planimetric method, **(b)** Linear Intercept method

### **2.2 Chemical composition using EDS**

One of the most outstanding features of the *SEM-EDS* is that it allows elemental analysis and observation from an ultra micro area to a wide area on the specimen surface without destroying the specimen. Qualitative and quantitative analysis (by *EDS*) by electron probe takes advantage of the emission of characteristic X-radiation by electron interactions in the valence shell of atoms. Backscattered electron images in the *SEM* display compositional contrast that results from different atomic number and their distribution of elements. *EDS* allows one to identify what those particular elements are and their relative proportions, for example their atomic percentage. Initial EDS analysis usually involves the generation of an X-ray spectrum from the entire scan area of the SEM. In the X ray spectra generated from the entire scan, Y-axis shows the counts (number of X-rays received and processed by the detector) and the X-axis shows the energy level of those counts. The EDS software associates the energy level of the X-rays with the elements and shell levels that generated them.

## **3. Exploring superconductors using** *SEM*

520 Scanning Electron Microscopy

three-dimensional grain structure, they will not give exactly the same value, but they will be close, generally within the experimental limitations of the measurements. In practice, these measurements are repeated on a number of fields in order to obtain a good estimate of the

(a)

(b)

One of the most outstanding features of the *SEM-EDS* is that it allows elemental analysis and observation from an ultra micro area to a wide area on the specimen surface without destroying the specimen. Qualitative and quantitative analysis (by *EDS*) by electron probe takes advantage of the emission of characteristic X-radiation by electron interactions in the valence shell of atoms. Backscattered electron images in the *SEM* display compositional contrast that results from different atomic number and their distribution of elements. *EDS* allows one to identify what those particular elements are and their relative proportions, for example their atomic percentage. Initial EDS analysis usually involves the generation of an X-ray spectrum from the entire scan area of the SEM. In the X ray spectra generated from the entire scan, Y-axis shows the counts (number of X-rays received and processed by the

Fig. 1. **(a)** Planimetric method, **(b)** Linear Intercept method

**2.2 Chemical composition using EDS** 

grain size.

In this section we will discuss the *SEM* study of superconductors of different families. For high temperature superconductors (HTSc), *SEM* has been widely used to explore the superconducting behavior. Grain size matters a lot in deciding the superconducting parameters of cuprate HTSc. Decrease in shielding current is observed with decrease in particle size [6]. Magnesium diboride (MgB2) represent an attractive alternative to low temperature superconductors. For most of the practical applications, high critical current density (*J*c) in the presence of a magnetic field along with high upper critical field (*H*C2) and high irreversibility field (*H*irr) are required. Moderate impurities and nano (*n*) materials are being used to improve these parameters [7-14]. In particular, significant flux pinning enhancement in MgB2 is observed with *n-SiC* addition [15]. In case of newly discovered pnictides superconductors, *SEM* is being used in an ingenious way. The spatially resolved electrical transport properties have been studied on the surface of optimally-doped superconducting Ba(Fe1-*x*Co*x*)2As2 single crystal by using a four-probe scanning tunneling microscopy [16]. Results will be discussed in following subtitles.

#### **3.1 Grain size and gain connectivity in cuprate high temperature superconductors**

One of the characteristics of high temperature superconductors (HTSc) is their small coherence length which is comparable with the unit cell. The coherence length is a key parameter for the performance of superconductors for applications, since this determines the size of the normally conducting core of the flux lines [17]. In order to control the motion of flux lines one needs a microstructure with defects as small as the coherence length. The extremely small coherence length of HTSc, which is for YBCO only 2.7 nm at 77 K within the ab-plane, is the reason that defects such as grain boundaries, which are very beneficial in low-temperature superconductors because they act as pinning defects, serve as weak links and limit the critical current, especially in the presence of an external magnetic field. HTSc bulk material can therefore be considered as a matrix of superconducting grains embedded in a non superconducting material.

The cuprate superconductors belong to the family of HTSc in which Cu-O chains and planes are responsible for the conduction of super currents. The size as well as the shape of the grains varies in different cuprate HTSc [18-20]. This variation in microstructure in turn leads to different superconducting behavior. Although the nature of the occurrence of superconductivity in cuprate HTSc is same, *Tc* varies from 38 K in LSCO to 110 K in BSCCO system. This wide range of *Tc* itself indicates that micro-structural parameters are of much importance. In this section various cuprate HTSc will be discussed in the chronological order of their discovery, in terms of the use of SEM for their characterization.

## **A. La2-xSrxCuO4**

Discovered by Bednorz and Muller in 1986 [21], La2-xSrxCuO4 was the trendsetter breakthrough in the history of superconductivity leading to the new era of High Tc superconductivity. It has *Tc* of 38 K which is beyond the BCS limit of 30 K. Although it has

Exploring the Superconductors with Scanning Electron Microscopy (SEM) 523

In another particle size controlled study of non-superconducting La1*.*96Sr0*.*04CuO4 was made with SEM and IR spectra by S. Zhou et al [24]. They observed that as the particle size reduces, the IR band at around 685 cm-1, corresponding to in-plane Cu–O asymmetrical stretching mode, shifts to higher frequency and the magnetization exhibits a large enhancement at low temperature. A visible spin-glass transition was found under a relatively weak external field in the sample with the largest particle sizes. Whereas the sample with the smallest particle sizes exhibits no visible spin-glass transition. They suggested that surface effects play a dominant

Cuprate superconductors are very sensitive to oxygen content. Depending on oxygen content, YBa2Cu3O7-δ crystallizes in two phases. Tetragonal *P4/mmm* with δ = 0.60 results in non-superconducting and antiferromagnetic YBCO whereas orthorhombic *pmmm* with δ = 0.05 phase leads to a superconducting YBCO with *Tc* 93 K [25]. Also, the intra-grain signal depends much on oxygen content of the composition. The structure of YBCO can be viewed as (Ba,Sr)O/CuO2/RE/CuO2/(Ba,Sr)O slabs interconnected through a sheet of Cu and O with variable composition of CuOx.. Charge transport and high temperature superconductivity is believed to reside in the CuO2 planes of all known HTSc cuprates, except that CuO1+δ chains have been reported to participate in the *b*-axis transport of YBa2Cu3O7−δ [26]. In YBa2Cu3O7−δ (CuBa2YCu2O7−δ, Cu-1212) there are two different Cu sites, namely Cu1 and Cu2. Cu1 resides in CuOx chains and Cu2 in superconducting CuO2 planes. Even at macroscopic level, any contravene in integral CuO2 stacks, affects superconductivity drastically [27-28]. The CuOx chain acts as a charge reservoir and provides the mobile

 To understand the physics of the superconducting nature, investigation of doping various elements at Cu1 site was carried by some of us [29]. And it was found that the YBCO structure is versatile and changes with doped elements at Cu1 site. Single phase samples of 1212 type with different MOx layers showed the great flexibility of these rocksalt layers and variable structure formation. With different M, as the oxidation state and ionic state changes, carrier concentration and structure changes as well. While, Nb-, Fe-, Ru- and Al-1212 possess tetragonal *P4/mmm* space group structure, the Ga-1212 and Co-1212 are

The SEM images [Fig. 3] suggest that with the doping of variety of elements at the Cu1 site in Y-123 structure, the morphology also changes suggesting change in the structure of the new compounds formed. Change in structure was authenticated by the Rietveld analysis also. In another report Nalin *et al.* [30] studied effect of Zn doping at Cu1 site. With ac susceptibility inter and intra granular changes are studied. In the Zn-doped samples, the inter-grain peak got reduced dramatically. In the *χII* plots of Zn doped samples the intergrain peak superposes with intra-grain peak and inter-grain peak depresses further. It was concluded that as (see *SEM* images [Fig. 4]) the average grain size is increasing with Zn doping. The increased grain size provides more area for the eddy currents loops to persist in

A combined study through *SEM* and *EDS* for the compounds Y1-xCaxBa1.9Nd0.1Cu3Oy (YCBNCO) with x ≤ 0.40 have been made [31]. Back-scattered electron SEM micrographs of samples with x = 0.10 and 0.30 have been taken. The *SEM* studies proved that the samples

the individual grains, thus systematic enhancing the intra-grain peak.

role in determining the magnetic properties as the particle size reduces.

**B. YBa2Cu3O7-<sup>δ</sup>** 

carriers to superconducting CuO2 planes.

crystallized in orthorhombic *Ima2* space group.

higher *Tc* than conventional superconductors, its critical parameters such as critical temperature, field and current density, which are applicable in practical applications, are weaker. Various attempts had been made earlier to enhance these parameters and to understand the physics behind it. The values such as shielding/Meissner fraction, the interand intra-grain critical current density and diamagnetization fraction depends upon particle size of bulk superconductors. One of the earliest reports on the effect of particle size on the physical properties of a superconductor was by Chiang et al. [22], who varied particle size from 1 to 10 µm and found significant changes in the superconducting and physical properties. Another fact that has well been established is that the critical current measurements done in HTSc have shown a much lower value for polycrystalline bulk samples [1] than single crystals of the same compound. This difference cannot be attributed alone to the intrinsic nature (anisotropy etc.) of the material. In fact, the quasi-insulating grain boundaries of HTSc play a detrimental role in limiting the critical current and other superconducting and magnetic properties [22-23]. In a recent report D. Sharma et al. [18] have investigated the influence of grain size (sintering temperature) on various superconducting parameters in La1*.*85Sr0*.*15CuO4. From SEM micrographs [see Fig. 2 (a), (b) and (c)] it is clear that with the increase in the sintering temperature, there is a considerable increase in the grain size. Increment in the grain size brings in double boom to the superconductivity as it reduces the number of insulating grain boundaries (weak links) as well as increases the effective superconducting volume fraction. Thus both the inter and intra critical current density which is limited by the weak links is expected to enhance with increase in the grain size. Qualitative picture given by the SEM micrographs corroborates with the quantitative results obtained from various transport and magnetic measurements as given in the table 1.

Fig. 2. SEM of La1*.*85Sr0*.*15CuO4 samples sintered at (**a**) 900 °C, (**b**) 1000 °C, and (**c**) 1050 °C [Ref. 18].


Table 1. Critical current density, diamagnetization fraction and percentage volume fraction not penetrated by magnetic flux calculated from magnetic measurement data for different grain sizes of La1*.*85Sr0*.*15CuO4 [Ref. 18].

In another particle size controlled study of non-superconducting La1*.*96Sr0*.*04CuO4 was made with SEM and IR spectra by S. Zhou et al [24]. They observed that as the particle size reduces, the IR band at around 685 cm-1, corresponding to in-plane Cu–O asymmetrical stretching mode, shifts to higher frequency and the magnetization exhibits a large enhancement at low temperature. A visible spin-glass transition was found under a relatively weak external field in the sample with the largest particle sizes. Whereas the sample with the smallest particle sizes exhibits no visible spin-glass transition. They suggested that surface effects play a dominant role in determining the magnetic properties as the particle size reduces.

#### **B. YBa2Cu3O7-<sup>δ</sup>**

522 Scanning Electron Microscopy

higher *Tc* than conventional superconductors, its critical parameters such as critical temperature, field and current density, which are applicable in practical applications, are weaker. Various attempts had been made earlier to enhance these parameters and to understand the physics behind it. The values such as shielding/Meissner fraction, the interand intra-grain critical current density and diamagnetization fraction depends upon particle size of bulk superconductors. One of the earliest reports on the effect of particle size on the physical properties of a superconductor was by Chiang et al. [22], who varied particle size from 1 to 10 µm and found significant changes in the superconducting and physical properties. Another fact that has well been established is that the critical current measurements done in HTSc have shown a much lower value for polycrystalline bulk samples [1] than single crystals of the same compound. This difference cannot be attributed alone to the intrinsic nature (anisotropy etc.) of the material. In fact, the quasi-insulating grain boundaries of HTSc play a detrimental role in limiting the critical current and other superconducting and magnetic properties [22-23]. In a recent report D. Sharma et al. [18] have investigated the influence of grain size (sintering temperature) on various superconducting parameters in La1*.*85Sr0*.*15CuO4. From SEM micrographs [see Fig. 2 (a), (b) and (c)] it is clear that with the increase in the sintering temperature, there is a considerable increase in the grain size. Increment in the grain size brings in double boom to the superconductivity as it reduces the number of insulating grain boundaries (weak links) as well as increases the effective superconducting volume fraction. Thus both the inter and intra critical current density which is limited by the weak links is expected to enhance with increase in the grain size. Qualitative picture given by the SEM micrographs corroborates with the quantitative results obtained from various transport

and magnetic measurements as given in the table 1.

(**c**) 1050 °C [Ref. 18].

grain sizes of La1*.*85Sr0*.*15CuO4 [Ref. 18].

Fig. 2. SEM of La1*.*85Sr0*.*15CuO4 samples sintered at (**a**) 900 °C, (**b**) 1000 °C, and

Table 1. Critical current density, diamagnetization fraction and percentage volume fraction not penetrated by magnetic flux calculated from magnetic measurement data for different

Cuprate superconductors are very sensitive to oxygen content. Depending on oxygen content, YBa2Cu3O7-δ crystallizes in two phases. Tetragonal *P4/mmm* with δ = 0.60 results in non-superconducting and antiferromagnetic YBCO whereas orthorhombic *pmmm* with δ = 0.05 phase leads to a superconducting YBCO with *Tc* 93 K [25]. Also, the intra-grain signal depends much on oxygen content of the composition. The structure of YBCO can be viewed as (Ba,Sr)O/CuO2/RE/CuO2/(Ba,Sr)O slabs interconnected through a sheet of Cu and O with variable composition of CuOx.. Charge transport and high temperature superconductivity is believed to reside in the CuO2 planes of all known HTSc cuprates, except that CuO1+δ chains have been reported to participate in the *b*-axis transport of YBa2Cu3O7−δ [26]. In YBa2Cu3O7−δ (CuBa2YCu2O7−δ, Cu-1212) there are two different Cu sites, namely Cu1 and Cu2. Cu1 resides in CuOx chains and Cu2 in superconducting CuO2 planes. Even at macroscopic level, any contravene in integral CuO2 stacks, affects superconductivity drastically [27-28]. The CuOx chain acts as a charge reservoir and provides the mobile carriers to superconducting CuO2 planes.

 To understand the physics of the superconducting nature, investigation of doping various elements at Cu1 site was carried by some of us [29]. And it was found that the YBCO structure is versatile and changes with doped elements at Cu1 site. Single phase samples of 1212 type with different MOx layers showed the great flexibility of these rocksalt layers and variable structure formation. With different M, as the oxidation state and ionic state changes, carrier concentration and structure changes as well. While, Nb-, Fe-, Ru- and Al-1212 possess tetragonal *P4/mmm* space group structure, the Ga-1212 and Co-1212 are crystallized in orthorhombic *Ima2* space group.

The SEM images [Fig. 3] suggest that with the doping of variety of elements at the Cu1 site in Y-123 structure, the morphology also changes suggesting change in the structure of the new compounds formed. Change in structure was authenticated by the Rietveld analysis also. In another report Nalin *et al.* [30] studied effect of Zn doping at Cu1 site. With ac susceptibility inter and intra granular changes are studied. In the Zn-doped samples, the inter-grain peak got reduced dramatically. In the *χII* plots of Zn doped samples the intergrain peak superposes with intra-grain peak and inter-grain peak depresses further. It was concluded that as (see *SEM* images [Fig. 4]) the average grain size is increasing with Zn doping. The increased grain size provides more area for the eddy currents loops to persist in the individual grains, thus systematic enhancing the intra-grain peak.

A combined study through *SEM* and *EDS* for the compounds Y1-xCaxBa1.9Nd0.1Cu3Oy (YCBNCO) with x ≤ 0.40 have been made [31]. Back-scattered electron SEM micrographs of samples with x = 0.10 and 0.30 have been taken. The *SEM* studies proved that the samples

Exploring the Superconductors with Scanning Electron Microscopy (SEM) 525

Fig. 4. SEM images of YBa2Cu3−*<sup>x</sup>*Zn*x*O7−*<sup>δ</sup>*, (**a**) *x* = 0.01, (**b**) *x* = 0.03, (**c**) *x* = 0.05 and (**d**) *x* = 0.10

Bismuth based superconducting cuprates (in short named as BSCCO) is an another family of cuprate HTSc which are expressed by a general formula of (Bi*,* Pb)2Sr2Ca*<sup>n</sup>*−1Cu*n*O*y.* For (*n* = 1, 2, 3), these are abbreviated as Bi2201, Bi2212 and Bi2223 phases, whose superconducting transition temperatures (*T*c) are around 20, 85 and 110 K respectively [32]. Though the mechanism of superconductivity in HTSc superconductors has been extensively studied, it is still unclear. As a result of substitution experiments it is well known, that in HTSc's, there is a strong relationship between carrier concentration and transition temperature. In addition, intergrain carrier transportation is also a key factor in deciding the sharpness of the transition. It is a well known fact that intergrain region behaves as a non-conducting region. Thus grain connectivity becomes more important for the sharpness of transition and other critical parameters. The grain growth and its shape varies in Bi2−*<sup>x</sup>*Pb*x*Sr2CaCu2O8 (see Fig. 5) with substitution of Pb at Bi site (0< x < 0.40) [19]. From the flake type grain shape in pristine samples to needle type grain shape in *x* = 0.40 composition is observed from SEM micrographs. Moreover, improvement in the packing fraction and hence the inter-granular

[Ref. 30].

**C. (Bi, Pb)2Sr2Can−1CunOy** 

with x ≤ 0.20 are homogeneous and stone like grains with typical size of several microns. On the other hand the samples with x ≥ 0.20 are inhomogeneous. The *SEM* micrographs show that the stone-like grains and the sponge-like grains co-exist in the surface of the samples. The *EDS* results show that the constituted elemental ratios in both regions are different and hence the superconducting properties.

(e) Eu/Nb-1212 (f) Eu/Ru-1212

Fig. 3. SEM pictures of the M-1212: (a) AlSr2YCu2O7+δ, (b) CoSr2YCu2O7, (c) FeSr2YCu2O7+δ, (d) GaSr2YCu2O7, (e) NbSr2EuCu2O7+δ and (f) RuSr2EuCu2O7+δ. [Ref. 29].

Fig. 4. SEM images of YBa2Cu3−*<sup>x</sup>*Zn*x*O7−*<sup>δ</sup>*, (**a**) *x* = 0.01, (**b**) *x* = 0.03, (**c**) *x* = 0.05 and (**d**) *x* = 0.10 [Ref. 30].

#### **C. (Bi, Pb)2Sr2Can−1CunOy**

524 Scanning Electron Microscopy

with x ≤ 0.20 are homogeneous and stone like grains with typical size of several microns. On the other hand the samples with x ≥ 0.20 are inhomogeneous. The *SEM* micrographs show that the stone-like grains and the sponge-like grains co-exist in the surface of the samples. The *EDS* results show that the constituted elemental ratios in both regions are different and

(a) Al-1212 (b) Co-1212

(c) Fe-1212 (d) Ga-1212

(e) Eu/Nb-1212 (f) Eu/Ru-1212

Fig. 3. SEM pictures of the M-1212: (a) AlSr2YCu2O7+δ, (b) CoSr2YCu2O7, (c) FeSr2YCu2O7+δ,

(d) GaSr2YCu2O7, (e) NbSr2EuCu2O7+δ and (f) RuSr2EuCu2O7+δ. [Ref. 29].

hence the superconducting properties.

Bismuth based superconducting cuprates (in short named as BSCCO) is an another family of cuprate HTSc which are expressed by a general formula of (Bi*,* Pb)2Sr2Ca*<sup>n</sup>*−1Cu*n*O*y.* For (*n* = 1, 2, 3), these are abbreviated as Bi2201, Bi2212 and Bi2223 phases, whose superconducting transition temperatures (*T*c) are around 20, 85 and 110 K respectively [32]. Though the mechanism of superconductivity in HTSc superconductors has been extensively studied, it is still unclear. As a result of substitution experiments it is well known, that in HTSc's, there is a strong relationship between carrier concentration and transition temperature. In addition, intergrain carrier transportation is also a key factor in deciding the sharpness of the transition. It is a well known fact that intergrain region behaves as a non-conducting region. Thus grain connectivity becomes more important for the sharpness of transition and other critical parameters. The grain growth and its shape varies in Bi2−*<sup>x</sup>*Pb*x*Sr2CaCu2O8 (see Fig. 5) with substitution of Pb at Bi site (0< x < 0.40) [19]. From the flake type grain shape in pristine samples to needle type grain shape in *x* = 0.40 composition is observed from SEM micrographs. Moreover, improvement in the packing fraction and hence the inter-granular

Exploring the Superconductors with Scanning Electron Microscopy (SEM) 527

Fig. 6*. J*c measurements for Bi2-*x*Pb*<sup>x</sup>* Sr2CaCu2O8+*<sup>δ</sup>* (*x* = 0 to 0.40) [Ref. 19].

Fig. 7. Temperature dependence of the transport critical current densities, between 65 and 77

Reprinted with permission from [Supercond. Sci. Technol. 22 (2009) 034012].

K and self-field, for the four samples.

connectivity was seen in the samples for x = 0.0 up to *x* = 0*.*16, which degrades with further increase in *x*. Decrease in the grain alignment with increase in Pb content has also been seen in SEM micrographs.

Fig. 5. SEM micrographs of Bi2-*x*Pb*<sup>x</sup>* Sr2CaCu2O8+*<sup>δ</sup>* (**a**) *x* = 0.04, (**b**) *x* = 0.06, (**c**) *x* = 0.08 and (**d**) *x* = 0.16 (**e**) *x* = 0.20 and (**f**) *x* = 0.40[Ref. 19].

connectivity was seen in the samples for x = 0.0 up to *x* = 0*.*16, which degrades with further increase in *x*. Decrease in the grain alignment with increase in Pb content has also been seen

Fig. 5. SEM micrographs of Bi2-*x*Pb*<sup>x</sup>* Sr2CaCu2O8+*<sup>δ</sup>* (**a**) *x* = 0.04, (**b**) *x* = 0.06, (**c**) *x* = 0.08 and (**d**)

*x* = 0.16 (**e**) *x* = 0.20 and (**f**) *x* = 0.40[Ref. 19].

in SEM micrographs.

Fig. 6*. J*c measurements for Bi2-*x*Pb*<sup>x</sup>* Sr2CaCu2O8+*<sup>δ</sup>* (*x* = 0 to 0.40) [Ref. 19].

Fig. 7. Temperature dependence of the transport critical current densities, between 65 and 77 K and self-field, for the four samples.

Reprinted with permission from [Supercond. Sci. Technol. 22 (2009) 034012].

Exploring the Superconductors with Scanning Electron Microscopy (SEM) 529

Although cuprate superconductors exhibit very high transition temperatures, their in-field performance [34] is compromised by their large anisotropy, the result of which is to restrict high bulk current densities. On the other hand in-field performance (higher *Jc*), leads diboride of magnesium to much better candidate for application purposes besides its lower *Tc* than cuprates. With the magneto-optical (MO) and polarized light (PL), SEM was used to assess the issue of inhomogeneous and granular behaviour in MgB2 [35]. It was speculated through SEM that the strongly shielding high-*J*c regions are microstructurally subdivided on a scale of 100 nm. Also, in a darker central area a fine mixture of MgB2 and a boron-rich phase was found through SEM [Fig. 9]. They concluded that the strongly shielding regions contain a large number of high-angle grain boundaries. Thus along with MO and PL, SEM suggests that MgB2 is more similar to a low-*T*c metallic superconductor than to a high-*T*c copper oxide superconductor [35]. A sol-Gel synthesis of MgB2 nanowires is reported Nath et.al [36]. SEM study reveals formation of a thick mesh of nanowires. The nanowires are found to be ca. 50–100 nm in diameter with very smooth surfaces having lengths up to at least 20 micrometer. It is observed that nanowires oriented vertically with respect to the electron beam. Thus SEM also revealed a hexagonal cross section for MgB2 nanowires which is consistent with a degree of crystallinity. The Crystallinity of MgB2 nanowires was also supported by their selected area electron diffraction (*SAED*) study on some individual nanowires. As MgB2 has better candidature for practical applications various dopants has been added to improve its performance [6-14]. Arpita et al. [7] noted that with n-SiC addition though *T*c decreases, but critical current density (*J*c) and flux pinning improved significantly. Presence of Mg2Si phase was also revealed through SEM and EDS [Fig. 10]. Dual reaction occurs with n-SiC addition first n-SiC reacts with Mg forming Mg2Si and then free C is incorporated into MgB2 at B site [37]. Thus both reactions help in the pinning of vortices which results in improved superconducting performance. Mg2Si and excess carbon can be embedded within MgB2 grains as nanoinclusions. They argued that due to the substitution of C at the B site the formation of a nanodomain structure takes place due to the variation of Mg-B spacing. These nanodomain defects, having the size of 2-3 nm, can also behave as effective pinning centers. So, highly dispersed nanoinclusions within the grains and the presence of nanodomain defects act as pinning centers and thus result in the

**3.2 Flux pinning in MgB2**

improved *J*c*(H)* behavior for the n-SiC doped samples.

**A. REFeAsO1-xFx (1111)** 

**3.3 Pnictides: Chemical composition and electrical transport** 

Iron pnictides are the latest entrant in family of high temperature superconductors [38]. Superconductivity originates in parent pnictides *RE*FeAsO with doping of F at O site. The reactive nature of *REs* towards oxygen results a very critical synthesis condition for these compounds. Though the compounds are being synthesized in inert/oxygen controlled atmosphere, it is very hard to acquire the desired composition. Thus it is better to analyze the chemical composition of the synthesized compound before going insight and describing the physical properties. Thus *SEM* with *EDS* can be very useful in invoking the composition (especially effective F concentration) of the arsenides. The *SEM* analysis of the parent and non superconducting SmFeAsO compound after metallographic preparation reveals very small amounts of unreacted phases (iron arsenides), which are completely dissolved after

The critical current density data (see Fig. 6) for the same samples shows an increase in the current density with Pb substitution in pristine sample till x=0.16 after which it decreases with further increase in x. The decrease in conductivity for samples having *x* > 0*.*16 has been explained on the basis of the effects arising from decrease of the grain alignment, increase of porosity and secondary phases.

Also, A. Sotelo et al. [20] have studied the Lead (Pb) and Silver (Ag) doping of Bi-2212 samples. It was found that Pb doping results in the decrease of the transport critical current density (see Fig. 7), *J*c*,*t (from 4*.*4 × 107 to 6 × 106 Am−2 at 65 K and self-field) as well as in the worsening of the mechanical properties, by about 35% compared to the undoped samples. In contrast, Ag doping results in the improvement of both the critical current density and mechanical strength.

Fig. 8. Longitudinal SEM images obtained on annealed polished samples. (a) S1; (b) S2 (Ag); (c) S3 (Pb); and (d) S4 (PbAg). Phases can be identified as Bi-2201 (white contrast), Bi-free phases (dark grey; CuO, Sr14Cu24O41, and SrCuO2), plumbate-like phases and Bi-2212 (grey contrast), Ag (light grey contrast).

Reprinted with permission from [Supercond. Sci. Technol. 22 (2009) 034012].

These described effects are related to the microstructural observations (see Fig. 8) as, Pb doping dramatically reduces the texture, while Ag doping improves it. Moreover, for samples with Ag addition, an intergrowth of Bi-2223 inside the Bi-2212 grains is observed, which would explain the improved superconducting properties of these samples. The stability of these superconductors has been studied through the corrosion process in a moisture atmosphere [33]. By means of optical and SEM observations, several morphologies of the alteration products have been observed.

## **3.2 Flux pinning in MgB2**

528 Scanning Electron Microscopy

The critical current density data (see Fig. 6) for the same samples shows an increase in the current density with Pb substitution in pristine sample till x=0.16 after which it decreases with further increase in x. The decrease in conductivity for samples having *x* > 0*.*16 has been explained on the basis of the effects arising from decrease of the grain alignment, increase of

Also, A. Sotelo et al. [20] have studied the Lead (Pb) and Silver (Ag) doping of Bi-2212 samples. It was found that Pb doping results in the decrease of the transport critical current density (see Fig. 7), *J*c*,*t (from 4*.*4 × 107 to 6 × 106 Am−2 at 65 K and self-field) as well as in the worsening of the mechanical properties, by about 35% compared to the undoped samples. In contrast, Ag doping results in the improvement of both the critical current

Fig. 8. Longitudinal SEM images obtained on annealed polished samples. (a) S1; (b) S2 (Ag); (c) S3 (Pb); and (d) S4 (PbAg). Phases can be identified as Bi-2201 (white contrast), Bi-free phases (dark grey; CuO, Sr14Cu24O41, and SrCuO2), plumbate-like phases and Bi-2212 (grey

These described effects are related to the microstructural observations (see Fig. 8) as, Pb doping dramatically reduces the texture, while Ag doping improves it. Moreover, for samples with Ag addition, an intergrowth of Bi-2223 inside the Bi-2212 grains is observed, which would explain the improved superconducting properties of these samples. The stability of these superconductors has been studied through the corrosion process in a moisture atmosphere [33]. By means of optical and SEM observations, several morphologies

Reprinted with permission from [Supercond. Sci. Technol. 22 (2009) 034012].

porosity and secondary phases.

density and mechanical strength.

contrast), Ag (light grey contrast).

of the alteration products have been observed.

Although cuprate superconductors exhibit very high transition temperatures, their in-field performance [34] is compromised by their large anisotropy, the result of which is to restrict high bulk current densities. On the other hand in-field performance (higher *Jc*), leads diboride of magnesium to much better candidate for application purposes besides its lower *Tc* than cuprates. With the magneto-optical (MO) and polarized light (PL), SEM was used to assess the issue of inhomogeneous and granular behaviour in MgB2 [35]. It was speculated through SEM that the strongly shielding high-*J*c regions are microstructurally subdivided on a scale of 100 nm. Also, in a darker central area a fine mixture of MgB2 and a boron-rich phase was found through SEM [Fig. 9]. They concluded that the strongly shielding regions contain a large number of high-angle grain boundaries. Thus along with MO and PL, SEM suggests that MgB2 is more similar to a low-*T*c metallic superconductor than to a high-*T*c copper oxide superconductor [35]. A sol-Gel synthesis of MgB2 nanowires is reported Nath et.al [36]. SEM study reveals formation of a thick mesh of nanowires. The nanowires are found to be ca. 50–100 nm in diameter with very smooth surfaces having lengths up to at least 20 micrometer. It is observed that nanowires oriented vertically with respect to the electron beam. Thus SEM also revealed a hexagonal cross section for MgB2 nanowires which is consistent with a degree of crystallinity. The Crystallinity of MgB2 nanowires was also supported by their selected area electron diffraction (*SAED*) study on some individual nanowires. As MgB2 has better candidature for practical applications various dopants has been added to improve its performance [6-14]. Arpita et al. [7] noted that with n-SiC addition though *T*c decreases, but critical current density (*J*c) and flux pinning improved significantly. Presence of Mg2Si phase was also revealed through SEM and EDS [Fig. 10]. Dual reaction occurs with n-SiC addition first n-SiC reacts with Mg forming Mg2Si and then free C is incorporated into MgB2 at B site [37]. Thus both reactions help in the pinning of vortices which results in improved superconducting performance. Mg2Si and excess carbon can be embedded within MgB2 grains as nanoinclusions. They argued that due to the substitution of C at the B site the formation of a nanodomain structure takes place due to the variation of Mg-B spacing. These nanodomain defects, having the size of 2-3 nm, can also behave as effective pinning centers. So, highly dispersed nanoinclusions within the grains and the presence of nanodomain defects act as pinning centers and thus result in the improved *J*c*(H)* behavior for the n-SiC doped samples.

#### **3.3 Pnictides: Chemical composition and electrical transport**

#### **A. REFeAsO1-xFx (1111)**

Iron pnictides are the latest entrant in family of high temperature superconductors [38]. Superconductivity originates in parent pnictides *RE*FeAsO with doping of F at O site. The reactive nature of *REs* towards oxygen results a very critical synthesis condition for these compounds. Though the compounds are being synthesized in inert/oxygen controlled atmosphere, it is very hard to acquire the desired composition. Thus it is better to analyze the chemical composition of the synthesized compound before going insight and describing the physical properties. Thus *SEM* with *EDS* can be very useful in invoking the composition (especially effective F concentration) of the arsenides. The *SEM* analysis of the parent and non superconducting SmFeAsO compound after metallographic preparation reveals very small amounts of unreacted phases (iron arsenides), which are completely dissolved after

Exploring the Superconductors with Scanning Electron Microscopy (SEM) 531

secondary phases. On the other hand another *SEM* image of the same sample after metallographic preparation reveals that crystals are aggregated within a matrix constituted of FeAs, which was also evidenced by their *EDS* analysis. However, their XPS study speculated that the formation of secondary phases, such as FeAs and SmOF. The discrepancy with EDS data, indicating only the presence of FeAs in the matrix, was explained by different analysis depth: up to a few micrometers for *EDS* and only a few nanometers for *XPS*. However, the aggregation of REFeAsO0.85 crystals in FeAs matrix is also observed in Back Scattered *SEM* study [Fig. 11] of Ketnami et al. [40]. The absence of significant transport currents in polycrystalline samples has raised the concern that there is a significant depression of the superconducting order parameter at grain boundaries (GB) [41-43]. Remnant magnetization and *MO* studies of polycrystalline NdFeAsO0.85 and SmFeAsO0.85 uncovered that intergrain and intra-grain current densities had different temperature dependences and differed by three orders of magnitude, leaving open the possibility of an intrinsic GB blocking effect [40]. Moreover, the BSE-SEM images revealed that even the best SmFeAsO0.85 bulk had nonsuperconducting Fe-As and *RE*2O3 occupying at least three quarters of the REFeAsO0.85 GBs, making the active current path certainly much smaller than the geometrical cross-section of the sample [40]. Further to reveal the active local current paths, combined low temperature laser scanning microscopy (LTLSM) and SEM studies had been made [44]. With the *SEM* images they are able to show significant micro-structural differences between various regions of the sample. It is revealed that insulating Sm2O3 has a small surface to volume ratio and is mostly located within SmFeAsO0.85 grains, so it has the smallest effect on current transport. On the other hand the dark gray Fe-As phase wets many GBs, thus interrupting grain to grain supercurrent paths, which are further degraded by extensive cracking, sometimes at GBs.

Fig. 10. (a), (b) SEM images of pure MgB2 and 10 wt%n-SiC added samples. Reprinted with permission from [Nanotechnology 19 (2008) 125708]

Fig. 9. Polarized light microscope and magneto-optical images of the same area of sample B are compared in (**a)** and (**b)**, respectively. Bright regions of (**b)** indicate areas where magnetic flux has penetrated the sample after a field of 120 mT was applied after cooling the sample in zero field to 11 K. Image (**c)** presents a magnified view using SEM backscattered electron imaging of the strongly superconducting region marked with an arrow in (**a)** and (**b)**. At higher resolution, image **(d)**, a secondary electron examination of the central region in (**c)**, reveals that the area marked by an arrow in (**a)** and (**b)** has, 100-nm, fine-scale structure [Reprinted by permission from Macmillan Publishers Ltd: NATURE 410 (2001) 186].

sintering. In general, sintering greatly increases the density of the samples, but favours the formation of Sm2O3 small particles. This feature reveals that at the sintering temperature the formation of Sm2O3 competes with the thermodynamic stability of the oxy-pnictide. S. Kaciulis et al. [39] studied SmFeAsO0.85F0.15 sample with *SEM*, *EDS* and *XPS*. *SEM* image after the fracture manifests the crystals appear clean at the surface, without any contamination of

Fig. 9. Polarized light microscope and magneto-optical images of the same area of sample B are compared in (**a)** and (**b)**, respectively. Bright regions of (**b)** indicate areas where magnetic flux has penetrated the sample after a field of 120 mT was applied after cooling the sample in zero field to 11 K. Image (**c)** presents a magnified view using SEM backscattered electron imaging of the strongly superconducting region marked with an arrow in (**a)** and (**b)**. At higher

resolution, image **(d)**, a secondary electron examination of the central region in (**c)**, reveals that

sintering. In general, sintering greatly increases the density of the samples, but favours the formation of Sm2O3 small particles. This feature reveals that at the sintering temperature the formation of Sm2O3 competes with the thermodynamic stability of the oxy-pnictide. S. Kaciulis et al. [39] studied SmFeAsO0.85F0.15 sample with *SEM*, *EDS* and *XPS*. *SEM* image after the fracture manifests the crystals appear clean at the surface, without any contamination of

the area marked by an arrow in (**a)** and (**b)** has, 100-nm, fine-scale structure

[Reprinted by permission from Macmillan Publishers Ltd: NATURE 410 (2001) 186].

secondary phases. On the other hand another *SEM* image of the same sample after metallographic preparation reveals that crystals are aggregated within a matrix constituted of FeAs, which was also evidenced by their *EDS* analysis. However, their XPS study speculated that the formation of secondary phases, such as FeAs and SmOF. The discrepancy with EDS data, indicating only the presence of FeAs in the matrix, was explained by different analysis depth: up to a few micrometers for *EDS* and only a few nanometers for *XPS*. However, the aggregation of REFeAsO0.85 crystals in FeAs matrix is also observed in Back Scattered *SEM* study [Fig. 11] of Ketnami et al. [40]. The absence of significant transport currents in polycrystalline samples has raised the concern that there is a significant depression of the superconducting order parameter at grain boundaries (GB) [41-43]. Remnant magnetization and *MO* studies of polycrystalline NdFeAsO0.85 and SmFeAsO0.85 uncovered that intergrain and intra-grain current densities had different temperature dependences and differed by three orders of magnitude, leaving open the possibility of an intrinsic GB blocking effect [40]. Moreover, the BSE-SEM images revealed that even the best SmFeAsO0.85 bulk had nonsuperconducting Fe-As and *RE*2O3 occupying at least three quarters of the REFeAsO0.85 GBs, making the active current path certainly much smaller than the geometrical cross-section of the sample [40]. Further to reveal the active local current paths, combined low temperature laser scanning microscopy (LTLSM) and SEM studies had been made [44]. With the *SEM* images they are able to show significant micro-structural differences between various regions of the sample. It is revealed that insulating Sm2O3 has a small surface to volume ratio and is mostly located within SmFeAsO0.85 grains, so it has the smallest effect on current transport. On the other hand the dark gray Fe-As phase wets many GBs, thus interrupting grain to grain supercurrent paths, which are further degraded by extensive cracking, sometimes at GBs.

Fig. 10. (a), (b) SEM images of pure MgB2 and 10 wt%n-SiC added samples. Reprinted with permission from [Nanotechnology 19 (2008) 125708]

Exploring the Superconductors with Scanning Electron Microscopy (SEM) 533

Also, magnetic contamination is detected through *SEM* in the case of NdFeAsO single crystals grown out of NaAs flux under ambient pressure [45]. It is observed that some crystals show a lambda anomaly in the specific heat curve at ~12 K while the same is absent in others. They examined the cleaved (001) surfaces with *SEM* which were showing lambda anomaly. They turned to look at the edges of the crystals carefully. A 10-*μ*m-thick layer was observed on the edges of the NdFeAsO crystals. With *EDS* it was found that some particles of TaAs, were surrounded by some very fine particles. After removal of that impurity the lambda anomaly disappeared. Thus with the help of *SEM* and *EDS* we can find out the

Soon after the discovery of 1111 family, other superconducting families based on FeAs layers (122, 111 and 11 structure) were reported, such as (Ba*,*K)Fe2As2 [46], LiFeAs [47], and FeSe [48]. Among all of these iron-based superconductors, the 122-type superconductors with a *T*c of 38 K have a lower synthesis temperature and are oxygen free in comparison to the 1111-type. In addition, its *T*c is much higher than those of the 111 and 11-type superconductors. SEM is used ingeniously to study the single crystal of Ba(Fe/Co)2As2 [16]. The topographic images of cleaved surface of optimally doped BaFe1.8Co0.2As2 with *uniform*  contrast have been observed in the secondary electron emission images acquired by SEM. They used secondary electron emission mode as it can register the contrast according to topography, chemical composition, and surface barrier (work function or ionization energy) of the sample [49] Small darker regions in the SEM image have been identified as marked by a rectangular box in Fig 12 (a). An SEM zoom-in image in the dark region reveals microscopic *domain* structures [Fig. 12]. They made resolved electrical transport measurements with use of SEM which have provided direct evidence of the coupling between superconductivity and local environment that is reflected by Co-concentration variation. In the uniform regions, the superconducting transition occurs at *TC* = 22.1 K for 10% fixed percentage of the normal-state resistance. In the domain regions, although the onset superconducting transition temperature is found very close to that of the uniform regions, *TC* varied over a broader range of 0.3-3.2 K. In addition, resistance of the domain regions above the transition onset temperature was noticed higher than that of the uniform

Like HTSc's improvement in *Jc* is observed with increase of grain size in 122 systems also [50]. Effect of sintering temperature on the microstructure and superconducting properties of Sr0*.*6K0*.*4Fe2As2 bulk samples was made. It was found that the annealing temperature had little influence on the critical temperature *T*c. However, the irreversibility field *Hirr* and *J*<sup>c</sup> were significantly affected by the sintering temperature. The *SEM* images reveal although samples had similar microstructure, the grain size increases monotonically as the sintering temperature rises. The grain size was less influenced by temperatures over 850 C. It was concluded the *J*c enhancement may result mainly from better grain connectivity due to the

Although SEM is very useful in finding grain size, their connectivity and then revealing various microscopic properties with physics behind that but there are some limiting

actual cause which leads particular nature of a material.

region, indicating higher defect density in the domain regions.

**B. Ba/Sr/K/Fe2As2 (122)** 

decrease of impurity phases.

**4. Limitations** 

Fig. 11. BSE image of the (a) Sm and (b) Nd sample at high magnification. Although some grain boundaries are well connected, others are clearly obstructed by the Fe-As phase (dark contrast), Sm2O3 or Nd2O3 (white contrast) and cracks.

Reprinted with permission from [Supercond. Sci. Technol. 22 (2009) 015010].

Also, magnetic contamination is detected through *SEM* in the case of NdFeAsO single crystals grown out of NaAs flux under ambient pressure [45]. It is observed that some crystals show a lambda anomaly in the specific heat curve at ~12 K while the same is absent in others. They examined the cleaved (001) surfaces with *SEM* which were showing lambda anomaly. They turned to look at the edges of the crystals carefully. A 10-*μ*m-thick layer was observed on the edges of the NdFeAsO crystals. With *EDS* it was found that some particles of TaAs, were surrounded by some very fine particles. After removal of that impurity the lambda anomaly disappeared. Thus with the help of *SEM* and *EDS* we can find out the actual cause which leads particular nature of a material.

#### **B. Ba/Sr/K/Fe2As2 (122)**

532 Scanning Electron Microscopy

Fig. 11. BSE image of the (a) Sm and (b) Nd sample at high magnification. Although some grain boundaries are well connected, others are clearly obstructed by the Fe-As phase (dark

Reprinted with permission from [Supercond. Sci. Technol. 22 (2009) 015010].

contrast), Sm2O3 or Nd2O3 (white contrast) and cracks.

Soon after the discovery of 1111 family, other superconducting families based on FeAs layers (122, 111 and 11 structure) were reported, such as (Ba*,*K)Fe2As2 [46], LiFeAs [47], and FeSe [48]. Among all of these iron-based superconductors, the 122-type superconductors with a *T*c of 38 K have a lower synthesis temperature and are oxygen free in comparison to the 1111-type. In addition, its *T*c is much higher than those of the 111 and 11-type superconductors. SEM is used ingeniously to study the single crystal of Ba(Fe/Co)2As2 [16]. The topographic images of cleaved surface of optimally doped BaFe1.8Co0.2As2 with *uniform*  contrast have been observed in the secondary electron emission images acquired by SEM. They used secondary electron emission mode as it can register the contrast according to topography, chemical composition, and surface barrier (work function or ionization energy) of the sample [49] Small darker regions in the SEM image have been identified as marked by a rectangular box in Fig 12 (a). An SEM zoom-in image in the dark region reveals microscopic *domain* structures [Fig. 12]. They made resolved electrical transport measurements with use of SEM which have provided direct evidence of the coupling between superconductivity and local environment that is reflected by Co-concentration variation. In the uniform regions, the superconducting transition occurs at *TC* = 22.1 K for 10% fixed percentage of the normal-state resistance. In the domain regions, although the onset superconducting transition temperature is found very close to that of the uniform regions, *TC* varied over a broader range of 0.3-3.2 K. In addition, resistance of the domain regions above the transition onset temperature was noticed higher than that of the uniform region, indicating higher defect density in the domain regions.

Like HTSc's improvement in *Jc* is observed with increase of grain size in 122 systems also [50]. Effect of sintering temperature on the microstructure and superconducting properties of Sr0*.*6K0*.*4Fe2As2 bulk samples was made. It was found that the annealing temperature had little influence on the critical temperature *T*c. However, the irreversibility field *Hirr* and *J*<sup>c</sup> were significantly affected by the sintering temperature. The *SEM* images reveal although samples had similar microstructure, the grain size increases monotonically as the sintering temperature rises. The grain size was less influenced by temperatures over 850 C. It was concluded the *J*c enhancement may result mainly from better grain connectivity due to the decrease of impurity phases.

## **4. Limitations**

Although SEM is very useful in finding grain size, their connectivity and then revealing various microscopic properties with physics behind that but there are some limiting

Exploring the Superconductors with Scanning Electron Microscopy (SEM) 535

HTSc. Some of the important aspects with which *SEM* deals with, is the grain size, morphology and alignment, structural defects, chemical composition. It has been used widely to explore from HTSc's, diborides to pnictides. With the time ways to use *SEM* and to extract information from superconductors got improved. Earlier for HTSc's it was used simply in finding grain size, grain connectivity and to figure out impurity regions. With these parameters superconducting behaviour was explained. Along with this in diborides it was also used to figure out the pinning centers and hence to enhance the applicable parameters. In Pnictides with spatially resolved electrical transport measurements it provided direct evidence of the coupling between superconductivity and local environment variation. We suppose in future SEM will be more plausible to understand so that

Author S. K. Singh would like to acknowledge *CSIR*, India for providing fellowships. We are very much thankful NPG [Macmillan Publishers Ltd: NATURE 410 (2001) 186], IOP Publishing Ltd.{(SUST [Supercond. Sci. Technol. 22 (2009) 015010, Supercond. Sci. Technol. 22 (2009) 034012] and Nanotechnology **19** (2008) 125708)}, and APS [Phys. Rev. B 80 (2009) 214518] for providing permissions for the reprints of the images. We want to acknowledge the authors of the Ref. No. [15], [16], [20], [35], [40] for giving their consent to re-use the

[2] A Rosko, Y. M Chiang, J. S Moodera and D. A Rudman, American Ceram. Socie. (1988) 308 [3] D. S smith, S. Suasmoro, C.Gault, F. Caillaud and A. Smith, Revue Phys. Appl. 25 (1990) 61

[6] T. Nagano, Y. Tomioka, Y. Nakayama, K. Kishio, K. Kitazawa, Phys. Rev. B 48 (1993) 9689

[15] Arpita Vajpayee, V P S Awana, G L Bhalla and H Kishan Nanotechnology 19 (2008) 125708 [16] T.-H. Kim, R. Jin, L. R. Walker, J. Y. Howe, M. H. Pan, J. F. Wendelken, J. R.

[17] Charles P. Poole, Jr., Handbook of superconductivity, Academic Press, California (2000) [18] D. Sharma, Ranjan Kumar, H. Kishan and V.P.S. Awana, J Supercond Nov Magn 24

Thompson, A. S. Sefat, M. A. McGuire, B. C. Sales, D. Mandrus, A. P. Li, Phys. Rev.

superconductivity is better understood and improved.

[1] J. W Ekin, Adv. Ceram. Matter. 2 (1987) 586

[5] http://www.astm.org/Standards/E112.htm

[7] S X Dou *et al.* Appl. Phys. Lett. 81 (2002) 3419

[13] Xiang *et al.* Physica C 386 (2003) 611

B 80 (2009) 214518

(2011) 205

[8] W. K. Yeoh *et al.* Supercond. Sci. Technol. 19 (2006) 596 [9] H. Yamada, *et al.* Supercond. Sci. Technol. 19 (2006) 175 [10] C H Cheng, *et al.* Supercond. Sci. Technol. 16 (2003) 1182 [11] Senkowicz *et al.* Appl. Phys. Lett. 86 (2005) 202502 [12] R H T Wilke, *et al.* Phys. Rev. Lett. 92 (2004) 217003

[14] A. Matsumoto, *et al.* Supercond. Sci. Technol. 16 (2003) 926

[19] J. Kumar *et al.* J Supercond Nov Magn 23(2010) 493 [20] A Sotelo *et al.* Supercond. Sci. Technol. 22 (2009) 034012

**6. Acknowledgements** 

images.

**7. References** 

[4] http://www.astm.org/

conditions for it. It is not a complete characterization in the sense that it needs extra characterization techniques such as TEM, PL, XPS and MO to support the results. The sensitivity of SEM is known to be relatively poor for lighter elements such as B, C and O. Thus through EDS results, the actual percentage ratio cannot be determined as very light elements boron and carbon are lesser sensitive in comparison to others elements [15]. In nano-TiO2 doped samples [51] almost similar micrographs for all samples was found irrespective of whether they are doped with n-TiO2 or not. In their study with HRTEM it was concluded that several black holes that appear in the image are presumably the n-TiO2.

Fig. 12. Topographic images of cleaved surfaces of BaFe1.8Co0.2As2 single crystal. (a) SEM image showing uniform contrast with some dark regions near the edge of crystal, as marked by a rectangular box. (b) Zoom-in SEM image showing domain structures. Marked regions by symbols and a dash line in (a) and (b) indicate where the transport measurements and composition probing are carried out.

Reprinted (Fig.) with permission from [T.-H. Kim, R. Jin, L. R. Walker, J. Y. Howe, M. H. Pan, J. F. Wendelken, J. R. Thompson, A. S. Sefat, M. A.McGuire, B. C. Sales, D. Mandrus, A. P. Li, Phys. Rev. B 80 (2009) 214518].

## **5. Summary**

Summarily we can see that *SEM* has been widely used to explore the superconducting behavior. Grain size matters a lot in deciding the superconducting parameters of cuprate HTSc. Some of the important aspects with which *SEM* deals with, is the grain size, morphology and alignment, structural defects, chemical composition. It has been used widely to explore from HTSc's, diborides to pnictides. With the time ways to use *SEM* and to extract information from superconductors got improved. Earlier for HTSc's it was used simply in finding grain size, grain connectivity and to figure out impurity regions. With these parameters superconducting behaviour was explained. Along with this in diborides it was also used to figure out the pinning centers and hence to enhance the applicable parameters. In Pnictides with spatially resolved electrical transport measurements it provided direct evidence of the coupling between superconductivity and local environment variation. We suppose in future SEM will be more plausible to understand so that superconductivity is better understood and improved.

## **6. Acknowledgements**

534 Scanning Electron Microscopy

conditions for it. It is not a complete characterization in the sense that it needs extra characterization techniques such as TEM, PL, XPS and MO to support the results. The sensitivity of SEM is known to be relatively poor for lighter elements such as B, C and O. Thus through EDS results, the actual percentage ratio cannot be determined as very light elements boron and carbon are lesser sensitive in comparison to others elements [15]. In nano-TiO2 doped samples [51] almost similar micrographs for all samples was found irrespective of whether they are doped with n-TiO2 or not. In their study with HRTEM it was concluded that several black holes that appear in the image are presumably the n-TiO2.

Fig. 12. Topographic images of cleaved surfaces of BaFe1.8Co0.2As2 single crystal. (a) SEM image showing uniform contrast with some dark regions near the edge of crystal, as marked by a rectangular box. (b) Zoom-in SEM image showing domain structures. Marked regions by symbols and a dash line in (a) and (b) indicate where the transport measurements and

Reprinted (Fig.) with permission from [T.-H. Kim, R. Jin, L. R. Walker, J. Y. Howe, M. H. Pan, J. F. Wendelken, J. R. Thompson, A. S. Sefat, M. A.McGuire, B. C. Sales, D. Mandrus, A.

Summarily we can see that *SEM* has been widely used to explore the superconducting behavior. Grain size matters a lot in deciding the superconducting parameters of cuprate

composition probing are carried out.

P. Li, Phys. Rev. B 80 (2009) 214518].

**5. Summary** 

Author S. K. Singh would like to acknowledge *CSIR*, India for providing fellowships. We are very much thankful NPG [Macmillan Publishers Ltd: NATURE 410 (2001) 186], IOP Publishing Ltd.{(SUST [Supercond. Sci. Technol. 22 (2009) 015010, Supercond. Sci. Technol. 22 (2009) 034012] and Nanotechnology **19** (2008) 125708)}, and APS [Phys. Rev. B 80 (2009) 214518] for providing permissions for the reprints of the images. We want to acknowledge the authors of the Ref. No. [15], [16], [20], [35], [40] for giving their consent to re-use the images.

## **7. References**


**27** 

*USA* 

**Morphological and Photovoltaic Studies** 

Mukul Dubey and Hongshan He\*

**of TiO2 NTs for High Efficiency Solar Cells** 

*Center for Advanced Photovoltaics, Department of Electrical Engineering &* 

Highly ordered nanostructures, especially TiO2 NTs, have attracted considerable research interest in recent years due to their diverse applications in photocatalysis, photonic crystals, sensors, batteries and photovoltaic devices. The photophysical, photochemical, electrical and surface properties of these nanostructured materials depend highly on their morphology because of the quantum size effect. Hence it is critical to study the effect of morphology of the ordered nanostructures for device applications. In this chapter we will only focus on the TiO2 NT morphology in context of their applications in dye-sensitized

DSC is an electrochemical device that converts sunlight to electricity. The major components of DSC are photoelectrode, counterelectrode and electrolyte sandwiched between them. The photoelectrode is a dye-coated wide band gap semiconductor, such as TiO2, on a transparent conducting oxide (TCO) glass substrate. Dye molecules absorb sunlight and the electrons in the ground state are excited to the excited state. The electrons in the excited states inject into the conduction band of TiO2. The injected electrons transports to the TCO electrode via diffusion through TiO2 NPs. The electrons then flow through the external circuit to the counterelectrode, which is usually a platinized TCO glass. The redox species in the electrolyte, usually iodide, take the electron from counterelectrode, and are reduced to triiodide, which further gets oxidized by providing its electron to the ground state of dye molecule for its regeneration. There are several factors that affect the efficiency of DSC such as absorption band of dye molecule, electron injection efficiency from dye to TiO2, redox potential of electrolyte and charge transport through TiO2. The morphology of TiO2 photoelectrode is one critical factor that plays a pivotal role in the conversion of sunlight to electricity in DSCs. Remarkable breakthrough in photoelectrode by changing the planar structure to randomly packed mesoporous structure of TiO2 NPs improved the efficiency from less than 1 % to 8% by Grätzel *et al.* The mesoporous structures are promising due to their high surface area for the adsorption of photosensitzer leading to the improved light absorption and hence high efficiency. The photoelectrode was further optimized by introducing a compact layer with small TiO2 NPs and a scattering layer with large TiO2 NP underneath and at the top

**1. Introduction** 

solar cells (DSCs).

\* Corresponding Author

*Computer Science South Dakota State University, Brookings, SD,* 


## **Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells**

Mukul Dubey and Hongshan He\*

*Center for Advanced Photovoltaics, Department of Electrical Engineering & Computer Science South Dakota State University, Brookings, SD, USA* 

## **1. Introduction**

536 Scanning Electron Microscopy

[22] Chiang, Y.-M., Rudman, D.A., Leung, D.K., Ikeda, J.A.S., Roshko, A., Fabes, B.D.,

[29] Shiva Kumar, Anjana Dogra, M. Husain, H. Kishan and V.P.S. Awana, J. Alloys and

[30] N.P Liyanawaduge, Shiva Kumar Singh, Anuj kumar, V.P.S Awana and H.Kishan, J

[31] X.S. Wu, W.S. Tan, Y.M. Xu, E.M. Zhang,J. Du, A. Hu, S.S. Jiang, J. Gao Physica C 398

[33] O. Monnereau, Z.C. Kang, E. Russ, I. Suliga, G. Vacquier, T. Badéche, C. Boulesteix, A.

[34] Yeshurun, Y., Malozemoff, A. P. & Shaulov, A. Magnetic relaxation in high-temperature

[38] Y. Kamihara, T. Watanabe, M. Hirano, H. Hosono, J. Am. Chem. Soc. 130 (2008) 3296

[40] F. Kametani, A. A. Polyanskii, A. Yamamoto, J. Jiang, E. E. Hellstrom, A. Gurevich, D.

[41] A. Yamamoto, J. Jiang, C. Tarantini, N. Craig, A. A. Polyanskii, F. Kametani,F. Hunte, J.

C. Sales, D. K. Christen, and D. Mandrus, Appl. Phys. Lett. 92 (2008) 252501 [42] B. Senatore, G. Wu, R. H. Liu, X. H. Chen, and R. Flukiger, Phys. Rev. B 78 (2008) 054514 [43] R. Prozorov, M. E. Tillman, E. D. Mun, and P. C. Canfield, New J. Phys.11 (2009) 035004 [44] F. Kametani, P. Li,1 D. Abraimov, A. A. Polyanskii, A. Yamamoto, J. Jiang, E. E.

[45] J.-Q. Yan, Q. Xing, B. Jensen, H. Xu, K. W. Dennis, R. W. McCallum, and T. A. Lograsso

[49] Ma Y W, Gao Z S, Wang L, Qi Y P, Wang D L and Zhang X P Chin. Phys. Lett. 26 (2009)

[50] Zhiyu Zhang, Yanpeng Qi, Lei Wang, Zhaoshun Gao, DongliangWang, Xianping

[51] H. Kishan, V.P.S. Awana, T.M. de Oliveira, Sher Alam, M. Saito, O.F. de Lima Physica C

Zhang and Yanwei Ma Supercond. Sci. Technol. 23 (2010) 065009

C. Larbalestier, Z. A. Ren, J. Yang, X. L. Dong, W. Lu, and Z. X. Zhao, Supercond.

Jaroszynski, E. E. Hellstrom, D. C. Larbalestier, R.Jin, A. S. Sefat, M. A. McGuire, B.

Hellstrom, A. Gurevich, D. C. Larbalestier, Z. A. Ren, J. Yang, X. L. Dong, W. Lu,

[21] J.C. Bednorz and K.A. Müller, Zeitschrift für Physik B. 64(2) (1986) 189

[24] Shiming Zhou, Jiyin Zhao, Songnan Chu, Lei Shi Physica C 451 (2007) 38.

[27] M. Karppinen, V.P.S. Awana, Y. Morita, H. Yamauchi, Physica B, 312 (2003) 62

[26] R. Gagnon, C. Lupien, and L. Taillefer, Phys. Rev. B 50 (1994) 3458

Sup. and Novel Magn doi: 10.1007/s10948-010-1063-7

Casalot Applied Superconductivity April (1995) 197

superconductors. Rev. Mod. Phys. 68 (1996) 911

[36] Manashi Nath and B. A. Parkinson Adv. Mater. 18 (2006) 1865

and Z. X. Zhao Appl. Phys. Lett. 95 (2009) 142502

Physica C 152 (1988) 77 [23] J.W. Ekin *et al.* J. Appl. Phys. 62, (1987) 4821

[28] P.R. Slater, C. Greaves: Physica C 180 (1991) 299

[32] J. M. Tarascon *et al.* Phys. Rev. B 38 (1988) 8885

[35] L. D. Larbalestier *et al.* NATURE 410 (2001) 186

[37] S X Dou *et al.* Phys. Rev. Lett. 98 (2007) 097002

Sci. Technol. 22 (2009) 015010.

Phys. Rev. B. 84 1250

037401

458 (2007) 1

[46] M. Rotter, *et al.* Phys. Rev. B 78 (2008) 020503

[47] X C Wang, *et al.* Solid State. Commun. 148 (2008) 538 [48] Hsu F C *et al.* Proc. Natl Acad. Sci. USA 105 (2008) 14262

[39] S. Kaciulis *et al.* Surf. Interface Anal. 42 (2010) 692.

[25] R.J. Cava, Science 247 (1990) 656

compd. 352 (2010) 493

(2003) 131

Highly ordered nanostructures, especially TiO2 NTs, have attracted considerable research interest in recent years due to their diverse applications in photocatalysis, photonic crystals, sensors, batteries and photovoltaic devices. The photophysical, photochemical, electrical and surface properties of these nanostructured materials depend highly on their morphology because of the quantum size effect. Hence it is critical to study the effect of morphology of the ordered nanostructures for device applications. In this chapter we will only focus on the TiO2 NT morphology in context of their applications in dye-sensitized solar cells (DSCs).

DSC is an electrochemical device that converts sunlight to electricity. The major components of DSC are photoelectrode, counterelectrode and electrolyte sandwiched between them. The photoelectrode is a dye-coated wide band gap semiconductor, such as TiO2, on a transparent conducting oxide (TCO) glass substrate. Dye molecules absorb sunlight and the electrons in the ground state are excited to the excited state. The electrons in the excited states inject into the conduction band of TiO2. The injected electrons transports to the TCO electrode via diffusion through TiO2 NPs. The electrons then flow through the external circuit to the counterelectrode, which is usually a platinized TCO glass. The redox species in the electrolyte, usually iodide, take the electron from counterelectrode, and are reduced to triiodide, which further gets oxidized by providing its electron to the ground state of dye molecule for its regeneration. There are several factors that affect the efficiency of DSC such as absorption band of dye molecule, electron injection efficiency from dye to TiO2, redox potential of electrolyte and charge transport through TiO2. The morphology of TiO2 photoelectrode is one critical factor that plays a pivotal role in the conversion of sunlight to electricity in DSCs. Remarkable breakthrough in photoelectrode by changing the planar structure to randomly packed mesoporous structure of TiO2 NPs improved the efficiency from less than 1 % to 8% by Grätzel *et al.* The mesoporous structures are promising due to their high surface area for the adsorption of photosensitzer leading to the improved light absorption and hence high efficiency. The photoelectrode was further optimized by introducing a compact layer with small TiO2 NPs and a scattering layer with large TiO2 NP underneath and at the top

<sup>\*</sup> Corresponding Author

Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells 539

grown on titanium substrate and no technique was known to either grow or transfer the NT

Fig. 2. **(a)** TiO2 NP based DSC with front illumination geometry from photo-electrode side; **(b)** TiO2 NT based DSC with back illumination geometry from counter-electrode side

Front illumination in TiO2 NT-based DSCs can be realized through several recently reported methods. The first method is the growth of NTs on glass substrate with sputtered Ti metal on top. The sputtering must be performed at high temperatures to prevent peeling after anodization. Grimes *et al* recently reported a new method for sputtering Ti on FTO glass at low temperature that produced TiO2 NTs with lengths up to 33 μm after anodization. A cell with a 17 μm NT array achieved a conversion efficiency of 6.9%. Two concerns emerge with this process: (1) the time- consuming nature of sputtering several tens of micrometer Ti may increase cost, and (2) the FTO layer on the glass could be damaged during anodization.

The second method is to remove the NT array from the Ti foil and attach it to FTO glass. In 2008, Jong Hyeok Park *et al* put anodized Ti foil in 0.1 M HCl aqueous solution for 1 hour, obtained an NT membrane, and attached it to FTO glass with the help of titanium isopropoxide. They achieved 7.6% efficiency with 8 μm NT arrays. Although the team claimed that NT membranes could be handled with tweezers, optical images in their publication suggested that these NT membranes were very fragile. In 2009, Qinwei Chen *et al* reported a re-anodization process that was followed by immersing the foil in 10% aqueous H2O2 solution for 24 hours and resulted in large sized NT membranes. The NTs were then attached to FTO glass with the help of a TiO2 NP paste to achieve a conversion efficiency of 5.5%. Long-time immersion in solution diminishes the attractiveness of this mild process.

He *et al* also developed a method that can lift off the NT arrays in less than four minutes. The yellow membrane could easily be transferred to other substrates without any fracturing. He *et al* also developed a unique low temperature method to tightly plant the NT membranes on FTO glass. The NTs were embedded inside the NP layer. The DSCs with these films exhibited 6.1% efficiency using N719 as dye. It was found that the geometry of NT orientation on the glass substrate also plays a significant role in determining the efficiency of DSC. The test tube geometry of NTs with one end open and other end closed

films on to the transparent conducting substrate.

of normal TiO2 NPs respectively. Both improved electrical and optical properties of photoelectrode and hence the device efficiency. With those structures and ruthenium bipyridine dyes, a respectable efficiency of 11.5% has been achieved rendering the DSCs as promising and cost-effective alternative to its otherwise expensive silicon technology.

Fig. 1. **(a)** Schematic representation of electron transport in TiO2 NPs based photoelctrode; **(b)** electron transport in TiO2 NT based photoelectrode

The electron collection efficiency is a critical factor governing the overall photo conversion efficiency of solar cell. Various investigations suggest that the random morphology of polycrystalline TiO2 NPs exhibits high defect density, which leads to the electron losses via recombination and the reduced electron collection efficiency. The presence of numerous defects, grain boundaries and surface states provides several trapping/detrapping and recombination sites in the electron transport pathway. The presence of defects reduces the electron mobility leading to increased recombination and hence reduced cell performance. In this regard anodic TiO2 NTs proposed by Grimes *et al* is considered as an excellent electron acceptor for DSC. Architecturally, these NTs are well aligned in regular array perpendicular to the substrate leading to rapid unidirectional electron transport with reduced recombination. A schematic for difference in dimensionality of electron transport between random nanocrystalline particle network and one-dimensional NT is shown in Figure 1. The electron from dye molecules migrate directly from top of the NT to the bottom for electron collection without migration in a three dimensional network. A close to 100% electron collection efficiency at the bottom of the nantotube was observed. In addition, NTs also have strong light scattering behavior which increases the optical path length in the film and improve the light absorption efficiency for high solar cell efficiency.

Despite being promising both electrically and optically, the highest energy conversion efficiency obtained from NT based DSCs is only ~ 7%, which is much lower than the conventional NP based DSC. One of the disadvantages identified was the back illumination geometry of devices due to the presence of non-transparent Ti metal underneath the TiO2 NT arrays. The TiO2 NT arrays are usually grown directly from a thin layer of Ti metal, which is difficult to remove. This requires photo illumination from the counterelectrode (a platinum coated transparent conducting electrode) side as shown in Figure 2. The back illumination leads to significant loss in the photon flux by reflection from the platinum and absorption in the electrolyte. It was difficult to realize front illumination since the NTs were

of normal TiO2 NPs respectively. Both improved electrical and optical properties of photoelectrode and hence the device efficiency. With those structures and ruthenium bipyridine dyes, a respectable efficiency of 11.5% has been achieved rendering the DSCs as

Fig. 1. **(a)** Schematic representation of electron transport in TiO2 NPs based photoelctrode;

The electron collection efficiency is a critical factor governing the overall photo conversion efficiency of solar cell. Various investigations suggest that the random morphology of polycrystalline TiO2 NPs exhibits high defect density, which leads to the electron losses via recombination and the reduced electron collection efficiency. The presence of numerous defects, grain boundaries and surface states provides several trapping/detrapping and recombination sites in the electron transport pathway. The presence of defects reduces the electron mobility leading to increased recombination and hence reduced cell performance. In this regard anodic TiO2 NTs proposed by Grimes *et al* is considered as an excellent electron acceptor for DSC. Architecturally, these NTs are well aligned in regular array perpendicular to the substrate leading to rapid unidirectional electron transport with reduced recombination. A schematic for difference in dimensionality of electron transport between random nanocrystalline particle network and one-dimensional NT is shown in Figure 1. The electron from dye molecules migrate directly from top of the NT to the bottom for electron collection without migration in a three dimensional network. A close to 100% electron collection efficiency at the bottom of the nantotube was observed. In addition, NTs also have strong light scattering behavior which increases the optical path length in the film

Despite being promising both electrically and optically, the highest energy conversion efficiency obtained from NT based DSCs is only ~ 7%, which is much lower than the conventional NP based DSC. One of the disadvantages identified was the back illumination geometry of devices due to the presence of non-transparent Ti metal underneath the TiO2 NT arrays. The TiO2 NT arrays are usually grown directly from a thin layer of Ti metal, which is difficult to remove. This requires photo illumination from the counterelectrode (a platinum coated transparent conducting electrode) side as shown in Figure 2. The back illumination leads to significant loss in the photon flux by reflection from the platinum and absorption in the electrolyte. It was difficult to realize front illumination since the NTs were

**(b)** electron transport in TiO2 NT based photoelectrode

and improve the light absorption efficiency for high solar cell efficiency.

promising and cost-effective alternative to its otherwise expensive silicon technology.

grown on titanium substrate and no technique was known to either grow or transfer the NT films on to the transparent conducting substrate.

Fig. 2. **(a)** TiO2 NP based DSC with front illumination geometry from photo-electrode side; **(b)** TiO2 NT based DSC with back illumination geometry from counter-electrode side

Front illumination in TiO2 NT-based DSCs can be realized through several recently reported methods. The first method is the growth of NTs on glass substrate with sputtered Ti metal on top. The sputtering must be performed at high temperatures to prevent peeling after anodization. Grimes *et al* recently reported a new method for sputtering Ti on FTO glass at low temperature that produced TiO2 NTs with lengths up to 33 μm after anodization. A cell with a 17 μm NT array achieved a conversion efficiency of 6.9%. Two concerns emerge with this process: (1) the time- consuming nature of sputtering several tens of micrometer Ti may increase cost, and (2) the FTO layer on the glass could be damaged during anodization.

The second method is to remove the NT array from the Ti foil and attach it to FTO glass. In 2008, Jong Hyeok Park *et al* put anodized Ti foil in 0.1 M HCl aqueous solution for 1 hour, obtained an NT membrane, and attached it to FTO glass with the help of titanium isopropoxide. They achieved 7.6% efficiency with 8 μm NT arrays. Although the team claimed that NT membranes could be handled with tweezers, optical images in their publication suggested that these NT membranes were very fragile. In 2009, Qinwei Chen *et al* reported a re-anodization process that was followed by immersing the foil in 10% aqueous H2O2 solution for 24 hours and resulted in large sized NT membranes. The NTs were then attached to FTO glass with the help of a TiO2 NP paste to achieve a conversion efficiency of 5.5%. Long-time immersion in solution diminishes the attractiveness of this mild process.

He *et al* also developed a method that can lift off the NT arrays in less than four minutes. The yellow membrane could easily be transferred to other substrates without any fracturing. He *et al* also developed a unique low temperature method to tightly plant the NT membranes on FTO glass. The NTs were embedded inside the NP layer. The DSCs with these films exhibited 6.1% efficiency using N719 as dye. It was found that the geometry of NT orientation on the glass substrate also plays a significant role in determining the efficiency of DSC. The test tube geometry of NTs with one end open and other end closed

Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells 541

bottom leading to vertical cavity formation (d to f). The formation of round shape at the bottom of tube is still a topic of debate. It is proposed that this is results of volume expansion of TiO2 compared to the space available from metal loss leading to high stress at the interface, high electric field distribution density at pore bottom and enhancement in

Fig. 4. **(a)** Titanium substrate in the ionic environment of electrolyte; **(b)** Formation of porous oxide layer on exposed surface of titanium right after field is switched on; **(c)** Initial random

pore growth by dissolution; **(d)** elongation of pore geometry after few minutes of anodization; **(e)** development of regular array of pore geometry in the field direction; **(f)** fully developed NT array. Red and Black dots represent the fluoride & hydroxide ions

acidity at the pore bottom due to the external electric field.

Fig. 3. Electrochemical anodization set up.

respectively.

provides freedom in choosing the configuration of the freestanding NT fixation on substrate with either closed end or the open end on to the substrate. This finding suggests that both optically and electrically open end of the NT on to the substrate is superior to the other orientation and hence can help significantly in improving DSC efficiency.

Another challenge for the effective use of NT for DSC application is how to grow highly ordered TiO2 NT arrays. Many researchers have reported that the NT tends to cluster together and form bundles which not only inhibits the infiltration of dye and electrolyte throughout the thickness of film but also increases recombination by incorporating disorder induced defects. It was reported that fine polishing of the titanium substrate prior to growth minimized the cluster formation. Several reports also indicated that the bundle and micro crack formation in the film was due to the capillary stress during the sample drying process. The supercritical CO2 oxide drying technique was introduced, which indeed reduced the formation of clusters; however, the complete understanding of cluster formation is still elusive and requires further study.

To summarize the morphology of TiO2 NT plays a critical role in dye-sensitized solar cell. Study of the effect of morphology of TiO2 NT on DSC performance is therefore worthy of pursuit for achieving high conversion efficiency of the DSCs. In the following sections, we will discuss the growth mechanism of TiO2 NTs and approaches for highly ordered TiO2 NT array of NTs for DSC applications. We will also discuss how the effect of orientation of the NT on the TCO glass affects the photovoltaic properties of DSC.

## **2. Growth mechanism of TiO2 NTs**

This section reviews the growth mechanism of TiO2 NTs by potentiostatic anodization technique in fluoride-containing electrolyte. The NT formation in acidic electrolyte containing F ion is generally agreed to occur via the field assisted formation and dissolution of oxidized titanium surface. It involves two critical steps that occur simultaneously: formation of TiO2 on the titanium surface and the dissolution of oxide. The process can be described by following two reactions:

$$\text{Ti} + 2\text{H}\_2\text{O} \rightarrow \text{TiO}\_2 + 4e^- + 4H^+ \text{....} + \text{....} + \text{....} \text{(1)} \text{(Oxidation)}$$

$$\text{TiO}\_2 + 6\text{F}^- + 4H^+ \rightarrow \text{[}\text{TiF}\_6\text{]}^{2-} + 2\text{H}\_2\text{O} \text{....} + \text{....} \text{(2)} \text{(Dissolution)}$$

In this two-electrode setup, titanium serves as anode and platinum as cathode. The electrolyte is composed of ethylene glycol, ammonium fluoride and water. A constant DC voltage is applied across the electrodes as shown in Figure 3. After some time, a layer of TiO2 NTs will form on the surface of Ti metal. Figure 4 shows schematically how the TiO2 NTs are formed. When pristine Ti is immersed into electrolyte solution, it is surrounded by various ionic species such as OH- and F- . (a) Once the DC voltage is applied these ionic species tends to oxidize the surface of titanium substrate (b) forming a thin barrier layer of TiO2 as depicted in the equation 1 of reaction mechanism. Simultaneously the process of dissolution of TiO2 layer in presence of F ion occurs leading to the formation of random pores during the initial stage of growth process (c). The F ions localize to the bottom of the pore i.e. at the oxide/metal interface which further undergoes oxidation and dissolution processes. Since the concentration of F- ion is more at the bottom of the pore due to the external electric field; the effective dissolution of TiO2 is more pronounced at the pore bottom leading to vertical cavity formation (d to f). The formation of round shape at the bottom of tube is still a topic of debate. It is proposed that this is results of volume expansion of TiO2 compared to the space available from metal loss leading to high stress at the interface, high electric field distribution density at pore bottom and enhancement in acidity at the pore bottom due to the external electric field.

Fig. 3. Electrochemical anodization set up.

540 Scanning Electron Microscopy

provides freedom in choosing the configuration of the freestanding NT fixation on substrate with either closed end or the open end on to the substrate. This finding suggests that both optically and electrically open end of the NT on to the substrate is superior to the other

Another challenge for the effective use of NT for DSC application is how to grow highly ordered TiO2 NT arrays. Many researchers have reported that the NT tends to cluster together and form bundles which not only inhibits the infiltration of dye and electrolyte throughout the thickness of film but also increases recombination by incorporating disorder induced defects. It was reported that fine polishing of the titanium substrate prior to growth minimized the cluster formation. Several reports also indicated that the bundle and micro crack formation in the film was due to the capillary stress during the sample drying process. The supercritical CO2 oxide drying technique was introduced, which indeed reduced the formation of clusters; however, the complete understanding of cluster formation is still

To summarize the morphology of TiO2 NT plays a critical role in dye-sensitized solar cell. Study of the effect of morphology of TiO2 NT on DSC performance is therefore worthy of pursuit for achieving high conversion efficiency of the DSCs. In the following sections, we will discuss the growth mechanism of TiO2 NTs and approaches for highly ordered TiO2 NT array of NTs for DSC applications. We will also discuss how the effect of orientation of the

This section reviews the growth mechanism of TiO2 NTs by potentiostatic anodization technique in fluoride-containing electrolyte. The NT formation in acidic electrolyte

of oxidized titanium surface. It involves two critical steps that occur simultaneously: formation of TiO2 on the titanium surface and the dissolution of oxide. The process can be

2

− +

*Ti H O TiO e H Oxidation TiO F H TiF H O Dissolution*

In this two-electrode setup, titanium serves as anode and platinum as cathode. The electrolyte is composed of ethylene glycol, ammonium fluoride and water. A constant DC voltage is applied across the electrodes as shown in Figure 3. After some time, a layer of TiO2 NTs will form on the surface of Ti metal. Figure 4 shows schematically how the TiO2 NTs are formed. When pristine Ti is immersed into electrolyte solution, it is surrounded by

species tends to oxidize the surface of titanium substrate (b) forming a thin barrier layer of TiO2 as depicted in the equation 1 of reaction mechanism. Simultaneously the process of

pores during the initial stage of growth process (c). The F- ions localize to the bottom of the pore i.e. at the oxide/metal interface which further undergoes oxidation and dissolution processes. Since the concentration of F- ion is more at the bottom of the pore due to the external electric field; the effective dissolution of TiO2 is more pronounced at the pore

2 4 4 ..................... 1 6 4 2 ........... 2

ion is generally agreed to occur via the field assisted formation and dissolution

[ ] ( )( )

( )( )

. (a) Once the DC voltage is applied these ionic

ion occurs leading to the formation of random

orientation and hence can help significantly in improving DSC efficiency.

NT on the TCO glass affects the photovoltaic properties of DSC.

2 2

+ → ++

2 6 2

and F-

− + −

++ → +

elusive and requires further study.

**2. Growth mechanism of TiO2 NTs** 

described by following two reactions:

various ionic species such as OH-

dissolution of TiO2 layer in presence of F-

containing F-

Fig. 4. **(a)** Titanium substrate in the ionic environment of electrolyte; **(b)** Formation of porous oxide layer on exposed surface of titanium right after field is switched on; **(c)** Initial random pore growth by dissolution; **(d)** elongation of pore geometry after few minutes of anodization; **(e)** development of regular array of pore geometry in the field direction; **(f)** fully developed NT array. Red and Black dots represent the fluoride & hydroxide ions respectively.

Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells 543

surface morphology for NTs grown on as purchased commercially pure titanium foil for 15 minutes. The fingerprint of substrate crack structures and submicron heterogeneously

(a) (b)

(c) (d)

(e) (f)

Fig. 6. **(a)** TiO2 NTs after 30 min anodization; **(b)** TiO2 NT with a whirlpool geometry at crack site; **(c)** TiO2 NT cluster formed at crack lines; **(d)** Collapsed TiO2 NTs at crack lines; **(e)** TiO2 bundles throughout sample and **(f)** TiO2 bundles under higher magnification

We further investigated the local morphology of NTs near the crack sites which is shown in higher magnification SEM image of Figure 6 (b). Whirlpool geometry of NT distribution at the crack site was observed, which shows the strong influence of substrate morphology on the initial growth of NTs. This effect was more pronounced in the NT under short anodization time. The clusters are formed near the crack lines of the substrate. Uniformed NTs are observed on the surface without any cracks. We also observed that the tubes over the edges of cracks tended to collapse on each other forming intercrossed tubes as shown in

distributed morphology near crack site were clearly observed on the NT film.

## **3. Effect of substrate morphology on growth of TiO2 NTs**

The formation of NTs largely depends on the type and concentration of ionic species present in the electrolyte as well as the extrinsic parameters such as anodization voltage, time and temperature. By controlling these factors, TiO2 NTs having different length, diameter, and wall thickness can be obtained. However, it should be noted that field assisted directional dissolution of the oxide layer formed on titanium foil is a crucial step towards the formation of NTs which so far have been shown to depend on many variables such as electrolyte composition, concentration, anodization voltage and time, but least importance was given to the effect of substrate morphology on the growth of NTs which is discussed in the next section. We found that the morphology of titanium substrate also plays a key role in the morphological order of the NT thus formed. This section highlights the effect of morphological features of titanium substrate on NT growth which is further connected with the microscopic morphology drawing outline for the plausible reasons for the clustering of NTs and cost effective way to deal with it.

#### **3.1 Effect of mechanical treatment of titanium substrate on TiO2 NT growth**

Commercial Ti foil with thickness ~ 250 μm is usually used for the growth of the TiO2 NTs arrays. Before the anodization the Ti foil is cleaned by detergent, ethanol, toluene, and deionized water sequentially to remove any impurities on the surface. There are several commercial providers for Ti foil with high purity; however, the surface morphology of these as-purchased Ti foils is quite different. It was found the as-purchased Ti foil has many crack sites distributed throughout the surface of the substrate. Figure 5 shows the typical SEM image of the surface of one sample from Sigma-Aldrich. Many cracks were observed on the surface. The size of the cracks ranges from several hundred nanometer to several micrometer. The presence of such cracks leads to the formation of vertical gaps on the substrate leading to the absence of material up till certain depth. In addition there are several submicron range heterogeneous morphologies present in the vicinity of crack sites which render high degree of roughness to the substrate. The existence of cracks on the Ti surface leads to high degree of non-uniformity in the morphology of NTs thus formed resulting in the cluster and bundle formation of NT. Figure 6 (a) shows the SEM image of

Fig. 5. **(a)** Cracks or vertical gaps present on the surface of as purchased commercially pure titanium substrate; **(b)** magnified image of crack showing the absence of material up till certain depth.

The formation of NTs largely depends on the type and concentration of ionic species present in the electrolyte as well as the extrinsic parameters such as anodization voltage, time and temperature. By controlling these factors, TiO2 NTs having different length, diameter, and wall thickness can be obtained. However, it should be noted that field assisted directional dissolution of the oxide layer formed on titanium foil is a crucial step towards the formation of NTs which so far have been shown to depend on many variables such as electrolyte composition, concentration, anodization voltage and time, but least importance was given to the effect of substrate morphology on the growth of NTs which is discussed in the next section. We found that the morphology of titanium substrate also plays a key role in the morphological order of the NT thus formed. This section highlights the effect of morphological features of titanium substrate on NT growth which is further connected with the microscopic morphology drawing outline for the plausible reasons for the clustering of

**3.1 Effect of mechanical treatment of titanium substrate on TiO2 NT growth** 

Commercial Ti foil with thickness ~ 250 μm is usually used for the growth of the TiO2 NTs arrays. Before the anodization the Ti foil is cleaned by detergent, ethanol, toluene, and deionized water sequentially to remove any impurities on the surface. There are several commercial providers for Ti foil with high purity; however, the surface morphology of these as-purchased Ti foils is quite different. It was found the as-purchased Ti foil has many crack sites distributed throughout the surface of the substrate. Figure 5 shows the typical SEM image of the surface of one sample from Sigma-Aldrich. Many cracks were observed on the surface. The size of the cracks ranges from several hundred nanometer to several micrometer. The presence of such cracks leads to the formation of vertical gaps on the substrate leading to the absence of material up till certain depth. In addition there are several submicron range heterogeneous morphologies present in the vicinity of crack sites which render high degree of roughness to the substrate. The existence of cracks on the Ti surface leads to high degree of non-uniformity in the morphology of NTs thus formed resulting in the cluster and bundle formation of NT. Figure 6 (a) shows the SEM image of

(a) (b)

Fig. 5. **(a)** Cracks or vertical gaps present on the surface of as purchased commercially pure titanium substrate; **(b)** magnified image of crack showing the absence of material up till

**3. Effect of substrate morphology on growth of TiO2 NTs** 

NTs and cost effective way to deal with it.

certain depth.

surface morphology for NTs grown on as purchased commercially pure titanium foil for 15 minutes. The fingerprint of substrate crack structures and submicron heterogeneously distributed morphology near crack site were clearly observed on the NT film.

Fig. 6. **(a)** TiO2 NTs after 30 min anodization; **(b)** TiO2 NT with a whirlpool geometry at crack site; **(c)** TiO2 NT cluster formed at crack lines; **(d)** Collapsed TiO2 NTs at crack lines; **(e)** TiO2 bundles throughout sample and **(f)** TiO2 bundles under higher magnification

We further investigated the local morphology of NTs near the crack sites which is shown in higher magnification SEM image of Figure 6 (b). Whirlpool geometry of NT distribution at the crack site was observed, which shows the strong influence of substrate morphology on the initial growth of NTs. This effect was more pronounced in the NT under short anodization time. The clusters are formed near the crack lines of the substrate. Uniformed NTs are observed on the surface without any cracks. We also observed that the tubes over the edges of cracks tended to collapse on each other forming intercrossed tubes as shown in

Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells 545

direction bends toward the z- direction. The initial bending followed by z growth of NTs was further confirmed in Figure 6 (f) where it can be observed from one of the pits that the initial pore formation on the walls of the pit is in all three directions. However as the NTs grew longer they start bending in one direction which latter completely follows one directional growth. Interestingly it can be seen that the initial bending ranging to several microns leads to the collapse of NTs on each other leading to the formation of clusters. Hence formation of highly ordered NTs can be severely influenced by the substrate

Fig. 7. **(a)** SEM image of polished Ti substrate; **(b)** bundle formation and non-uniformed TiO2 NT morphology; **(c)** side view showing different length of NTs and the bundle

Removing structural disorder from NTs was recently a key concern in the area of DSC. Some techniques including post growth ultrasonic treatment and supercritical CO2 drying of NT samples showed promise in removing of the structural disorder. These techniques are very useful if the disorder in NT morphology is induced through impurities in the electrolyte, viscosity of the electrolyte or during drying of NTs after growth. Their applications to remove substrate induced disorder are limited. We employed a chemical etching process to solve this problem. The Ti substrates were immersed in 0.75 M HF ranging from 1 to 15 minutes. The cracks present on the substrate were removed completely in 10 minutes of etching time. Figure 9 (a) shows the SEM image of titanium foil etched for 5 minutes in 0.75 M HF where the crack features could still be observed. Figure 9 (b) shows

formation; **(d)** unevenly packed TiO2 NTs

(a) (b)

(c) (d)

morphology.

Figure 6 (d). The collapsing of the NTs on each other can potentially lead to the cluster formation which can be seen from Figure 6 (e & f).

Based on the results of anodization on commercially purchased Titanium, it can be observed that smooth surface for anodization is very crucial to obtain highly ordered morphology of NTs. Han et al and Lee et al reported two step anodization processes to obtain ordered morphology of NTs. In their report first anodization was performed for shorter time followed by removal of the first NT layer. The surface of Ti after removal of first layer was very smooth leading to highly ordered morphology of NT formed in the second step. On the other hand Kang et al reported electropolishing technique in which Ti substrate was electropolished to render it a smooth surface followed by anodization to form ordered NT structure. Both electropolishing and two step anodization processes were found promising to obtained highly ordered NT array.

However, these processes involves complex two step processes which is time consuming and expensive. An alternative approach could be the mechanical polishing of the substrate to remove cracks. To this end we have tried to polish the Ti substrate using fine sand paper. However, our SEM results shows that even with very fine sand paper the micron size scratches are developed on the surface scratches are developed on the surface. It can be clearly seen that there It can be clearly seen that there were significant clumping and clustering of the NTs. Additionally at many other places NTs were found to be completely broken. Based on the results it can be inferred that even the fine mechanical polishing can form micron level roughness which cannot be used to grow highly ordered NTs.

### **3.2 Effect of chemical treatment of titanium substrate on TiO2 NT morphology**

In order to further verify the effect of local substrate morphology on NT growth, we etched the titanium substrate for 30 minutes in 0.75 M hydro fluoric acid (HF) introducing high degree of surface roughness to the substrate. Figure 8 (a) shows the morphology of rough surface of titanium after etching. TiO2 NTs were then grown on the etched substrate for 15 minutes. It was observed that the initial pore formation for NT growth takes the local geometry of the substrate as shown in Figure 8 (b). The local pore formation might largely depend on the direction of local electric field was further confirmed by the NT formation in the etched substrate. Figure 8 (c) shows the SEM image of a large pit formed on the substrate due to etching. The pit shown in the image can be visualized to have three different planes i.e. x-y, y-z and x-z. It is interesting to note that the pore formation can be seen on all these three planes with their cross-sections perpendicular to the respective plane clearly indicating that the initial pore formation does depend on the direction of local electric field at the breakdown site this further depends on the local morphology of the substrate as shown in Figure 8 (d). The dependence of NT growth associated with the local electric field distribution corresponding to the substrate morphology can be a profound reason for the bundle and cluster formation in NTs which was further confirmed from SEM results. Figure 8 (e) shows the SEM image of NT at one of the crack sites of the NT film grown on etched substrate. It can be clearly observed that the NTs at crack site grew in different direction. Considering x-y plane to be the plane of substrate and z as direction normal to the substrate which is the preferred direction of NT growth, it can be clearly seen that the cross-sectional plane of NTs are facing in two different directions, one parallel to x-y plane highlighted with red circle and other in z- direction highlighted with yellow circle. The NTs facing x-y

Figure 6 (d). The collapsing of the NTs on each other can potentially lead to the cluster

Based on the results of anodization on commercially purchased Titanium, it can be observed that smooth surface for anodization is very crucial to obtain highly ordered morphology of NTs. Han et al and Lee et al reported two step anodization processes to obtain ordered morphology of NTs. In their report first anodization was performed for shorter time followed by removal of the first NT layer. The surface of Ti after removal of first layer was very smooth leading to highly ordered morphology of NT formed in the second step. On the other hand Kang et al reported electropolishing technique in which Ti substrate was electropolished to render it a smooth surface followed by anodization to form ordered NT structure. Both electropolishing and two step anodization processes were found promising

However, these processes involves complex two step processes which is time consuming and expensive. An alternative approach could be the mechanical polishing of the substrate to remove cracks. To this end we have tried to polish the Ti substrate using fine sand paper. However, our SEM results shows that even with very fine sand paper the micron size scratches are developed on the surface scratches are developed on the surface. It can be clearly seen that there It can be clearly seen that there were significant clumping and clustering of the NTs. Additionally at many other places NTs were found to be completely broken. Based on the results it can be inferred that even the fine mechanical polishing can

form micron level roughness which cannot be used to grow highly ordered NTs.

**3.2 Effect of chemical treatment of titanium substrate on TiO2 NT morphology** 

In order to further verify the effect of local substrate morphology on NT growth, we etched the titanium substrate for 30 minutes in 0.75 M hydro fluoric acid (HF) introducing high degree of surface roughness to the substrate. Figure 8 (a) shows the morphology of rough surface of titanium after etching. TiO2 NTs were then grown on the etched substrate for 15 minutes. It was observed that the initial pore formation for NT growth takes the local geometry of the substrate as shown in Figure 8 (b). The local pore formation might largely depend on the direction of local electric field was further confirmed by the NT formation in the etched substrate. Figure 8 (c) shows the SEM image of a large pit formed on the substrate due to etching. The pit shown in the image can be visualized to have three different planes i.e. x-y, y-z and x-z. It is interesting to note that the pore formation can be seen on all these three planes with their cross-sections perpendicular to the respective plane clearly indicating that the initial pore formation does depend on the direction of local electric field at the breakdown site this further depends on the local morphology of the substrate as shown in Figure 8 (d). The dependence of NT growth associated with the local electric field distribution corresponding to the substrate morphology can be a profound reason for the bundle and cluster formation in NTs which was further confirmed from SEM results. Figure 8 (e) shows the SEM image of NT at one of the crack sites of the NT film grown on etched substrate. It can be clearly observed that the NTs at crack site grew in different direction. Considering x-y plane to be the plane of substrate and z as direction normal to the substrate which is the preferred direction of NT growth, it can be clearly seen that the cross-sectional plane of NTs are facing in two different directions, one parallel to x-y plane highlighted with red circle and other in z- direction highlighted with yellow circle. The NTs facing x-y

formation which can be seen from Figure 6 (e & f).

to obtained highly ordered NT array.

direction bends toward the z- direction. The initial bending followed by z growth of NTs was further confirmed in Figure 6 (f) where it can be observed from one of the pits that the initial pore formation on the walls of the pit is in all three directions. However as the NTs grew longer they start bending in one direction which latter completely follows one directional growth. Interestingly it can be seen that the initial bending ranging to several microns leads to the collapse of NTs on each other leading to the formation of clusters. Hence formation of highly ordered NTs can be severely influenced by the substrate morphology.

Fig. 7. **(a)** SEM image of polished Ti substrate; **(b)** bundle formation and non-uniformed TiO2 NT morphology; **(c)** side view showing different length of NTs and the bundle formation; **(d)** unevenly packed TiO2 NTs

Removing structural disorder from NTs was recently a key concern in the area of DSC. Some techniques including post growth ultrasonic treatment and supercritical CO2 drying of NT samples showed promise in removing of the structural disorder. These techniques are very useful if the disorder in NT morphology is induced through impurities in the electrolyte, viscosity of the electrolyte or during drying of NTs after growth. Their applications to remove substrate induced disorder are limited. We employed a chemical etching process to solve this problem. The Ti substrates were immersed in 0.75 M HF ranging from 1 to 15 minutes. The cracks present on the substrate were removed completely in 10 minutes of etching time. Figure 9 (a) shows the SEM image of titanium foil etched for 5 minutes in 0.75 M HF where the crack features could still be observed. Figure 9 (b) shows

Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells 547

neighboring pits, providing global order in the overall morphology of NTs. In order to verify our assumption we performed 30 minutes of anodization to grow shorter NTs on the titanium substrate etched for 10 minutes in 0.75 M HF. Figure 10 (a) shows the SEM image of NTs grown on etched substrate for 30 minutes anodization time. The image clearly shows that the NTs followed the local morphology of each pit taking the overall geometry of the substrate. In addition clustering or collapse of NTs was also not observed anywhere on the surface suggesting that overall order in the morphology can be achieved by this process. However, the method can find its applicability only when longer NTs can be successfully grown with long range order which is the essential need for solar cells. To investigate the morphology of longer NTs, we performed anodization of the etched substrate for 5 hrs

Fig. 9. SEM images of titanium substrate etched in 0.75 M HF under different etching time.

Interestingly the SEM image of Figure 10 (b) shows that the NTs even after 5hrs of anodization time followed highly ordered morphology without cluster formation anywhere on the substrate. It was also evidenced that the NTs retained the concave geometry of the substrate shown highlighted in yellow circle of Figure 10 (c). The overall morphology of the NTs were observed to be comprised of several small concave shaped honeycomb structure grouped together to form structured NT film which can be seen from SEM image of Figure 10 (d). Thus it can be seen that the morphology of the NTs significantly depends on both the morphology of the substrate and simple chemical pretreatment of the substrate can prove to be useful in

growing oriented NTs which might further help in improving the efficiency of DSC.

(a) (b)

(c) (d)

which can lead to the formation of ~ 20 µm long NTs.

(**a**) 5 minutes; (**b, d**) 10 minutes; (**c**) 15 minutes.

Fig. 8. SEM images of **(a)** etched Ti substrate; **(b)** TiO2 NTs grown on etched Ti; **(c)** large pit of TiO2 NTs on etched Ti; **(d)** TiO2 NT at a crack site on etched Ti; **(e)** TiO2 NT at the edge of one pit **(f)** non uniform local electric field distribution near the rough surface of titanium.

the cracks or vertical gaps completely disappeared after 10 minutes of etching but also introducing high degree of surface roughness induced on the substrate. Further etching the substrate for 15 minutes led to highly disordered coarse surface as can be seen in Figure 9 (c). A closer investigation of individual pits formed after 10 minutes etching of the substrate as shown in Figure 9 (d) revealed that these pits offer a very smooth concave shaped surface with average size of 5 – 10 µm. This observation suggested that highly oriented NTs can be grown over these smooth surfaces with short range of order on the surface of the substrate. Further concavity of the pit structure can lead to small bending in the NTs with crosssection plane facing towards the center of conic cross-section. The small bending of NTs can further help preventing the NTs to interact and collapse over the NTs formed in the

(e)

Fig. 8. SEM images of **(a)** etched Ti substrate; **(b)** TiO2 NTs grown on etched Ti; **(c)** large pit of TiO2 NTs on etched Ti; **(d)** TiO2 NT at a crack site on etched Ti; **(e)** TiO2 NT at the edge of one pit **(f)** non uniform local electric field distribution near the rough surface of titanium. the cracks or vertical gaps completely disappeared after 10 minutes of etching but also introducing high degree of surface roughness induced on the substrate. Further etching the substrate for 15 minutes led to highly disordered coarse surface as can be seen in Figure 9 (c). A closer investigation of individual pits formed after 10 minutes etching of the substrate as shown in Figure 9 (d) revealed that these pits offer a very smooth concave shaped surface with average size of 5 – 10 µm. This observation suggested that highly oriented NTs can be grown over these smooth surfaces with short range of order on the surface of the substrate. Further concavity of the pit structure can lead to small bending in the NTs with crosssection plane facing towards the center of conic cross-section. The small bending of NTs can further help preventing the NTs to interact and collapse over the NTs formed in the

(a) (b)

(c) (d)

(f)

neighboring pits, providing global order in the overall morphology of NTs. In order to verify our assumption we performed 30 minutes of anodization to grow shorter NTs on the titanium substrate etched for 10 minutes in 0.75 M HF. Figure 10 (a) shows the SEM image of NTs grown on etched substrate for 30 minutes anodization time. The image clearly shows that the NTs followed the local morphology of each pit taking the overall geometry of the substrate. In addition clustering or collapse of NTs was also not observed anywhere on the surface suggesting that overall order in the morphology can be achieved by this process. However, the method can find its applicability only when longer NTs can be successfully grown with long range order which is the essential need for solar cells. To investigate the morphology of longer NTs, we performed anodization of the etched substrate for 5 hrs which can lead to the formation of ~ 20 µm long NTs.

Fig. 9. SEM images of titanium substrate etched in 0.75 M HF under different etching time. (**a**) 5 minutes; (**b, d**) 10 minutes; (**c**) 15 minutes.

Interestingly the SEM image of Figure 10 (b) shows that the NTs even after 5hrs of anodization time followed highly ordered morphology without cluster formation anywhere on the substrate. It was also evidenced that the NTs retained the concave geometry of the substrate shown highlighted in yellow circle of Figure 10 (c). The overall morphology of the NTs were observed to be comprised of several small concave shaped honeycomb structure grouped together to form structured NT film which can be seen from SEM image of Figure 10 (d). Thus it can be seen that the morphology of the NTs significantly depends on both the morphology of the substrate and simple chemical pretreatment of the substrate can prove to be useful in growing oriented NTs which might further help in improving the efficiency of DSC.

Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells 549

In our work Freestanding NT films were obtained by preferential etching of the TiO2/Ti interface followed by its fixation on TCO with colloidal TiO2 paste as adhesive layer. The SEM image of freestanding NT film reveals that one end of the NT is open while other end is closed rendering it to be like a test tube structure. Figure 11 (a) shows the morphology of open end of NT while Figure 11 (b) shows the surface morphology of the closed end side of NT. The freestanding NTs can be used in two different orientations for fixation on TCO substrate; one with open end of NT facing the substrate while other with closed end of NT facing the substrate as shown in Figure 12. This section thus tends to highlight the effect of NT orientation on DSC performance. It was reported earlier that the closed end of NT facing the substrate might be helpful in improving the efficiency of DSC by serving as a barrier layer in between substrate and TiO2 active layer improving the charge transport by minimizing the substrate/TiO2 interface recombination analogous to the compact layer in

Fig. 11. **(a)** Top view of NT showing one end to be open; **(b)** bottom view of NT showing

Fig. 12. Simplified DSC structure with CED and OED orientation of NT on TCO

(a) (b)

NP based DSC.

other end to be closed

Fig. 10. SEM images of TiO2 NTs under 5 hrs anodization on etched Ti substrate. **(a)** anodized surface at higher magnification; **(b)** anodized surface at lower magnification; **(c)** highly ordered NT with local concavity shown highlighted in yellow circle; **(d)** several concave geometries highlighted in yellow circles

## **4. Effect of TiO2 NT morphology on PV performance of DSC**

The TiO2 NTs on the Ti substrate can be used directly for the fabrication of DSCs. The Ti metal will function as same as TCO layer in conventional Gratzel type DSCs. Due to the non-transparency of Ti metal to the sunlight, the cell has to be illuminated from counter electrode (back illumination). In 2007, Grimes *et al* reported 6.89 % conversion efficiency of this type of cell using ruthenium dye (N719) as light absorber, 20 μm long TiO2 NT arrays for dye adsorption, and iodide/tri-iodide as electrolyte. Several other groups who fabricated DSCs with this configuration achieved efficiencies ~3% under similar conditions. In 2009, Grätzel *et al* reported a 3.59% conversion efficiency of DSCs using ruthenium dye (N719) as a light absorber, 14 μm long TiO2 NT array for dye adsorption, and ionic liquid as electrolyte. He *et al* also achieved an efficiency of 3.45% with this configuration. Since TiO2 NT arrays are often attached on the Ti foil and difficult to lift off, the NT arrays with Ti foil were used directly for cell fabrication. Sunlight must come from the rear of the cell. The absorption and reflection of sunlight by electrolyte and Pt counterelectrode respectively lead to reduction in photon flux reaching the dyes. Various techniques were reported from 2008 – 2010 for the growth, liftoff and fixation of NTs on transparent conducting substrate but they either lacked reproducibility or was time consuming.

Fig. 10. SEM images of TiO2 NTs under 5 hrs anodization on etched Ti substrate.

**4. Effect of TiO2 NT morphology on PV performance of DSC** 

concave geometries highlighted in yellow circles

either lacked reproducibility or was time consuming.

**(a)** anodized surface at higher magnification; **(b)** anodized surface at lower magnification; **(c)** highly ordered NT with local concavity shown highlighted in yellow circle; **(d)** several

The TiO2 NTs on the Ti substrate can be used directly for the fabrication of DSCs. The Ti metal will function as same as TCO layer in conventional Gratzel type DSCs. Due to the non-transparency of Ti metal to the sunlight, the cell has to be illuminated from counter electrode (back illumination). In 2007, Grimes *et al* reported 6.89 % conversion efficiency of this type of cell using ruthenium dye (N719) as light absorber, 20 μm long TiO2 NT arrays for dye adsorption, and iodide/tri-iodide as electrolyte. Several other groups who fabricated DSCs with this configuration achieved efficiencies ~3% under similar conditions. In 2009, Grätzel *et al* reported a 3.59% conversion efficiency of DSCs using ruthenium dye (N719) as a light absorber, 14 μm long TiO2 NT array for dye adsorption, and ionic liquid as electrolyte. He *et al* also achieved an efficiency of 3.45% with this configuration. Since TiO2 NT arrays are often attached on the Ti foil and difficult to lift off, the NT arrays with Ti foil were used directly for cell fabrication. Sunlight must come from the rear of the cell. The absorption and reflection of sunlight by electrolyte and Pt counterelectrode respectively lead to reduction in photon flux reaching the dyes. Various techniques were reported from 2008 – 2010 for the growth, liftoff and fixation of NTs on transparent conducting substrate but they

(a) (b)

(c) (d)

In our work Freestanding NT films were obtained by preferential etching of the TiO2/Ti interface followed by its fixation on TCO with colloidal TiO2 paste as adhesive layer. The SEM image of freestanding NT film reveals that one end of the NT is open while other end is closed rendering it to be like a test tube structure. Figure 11 (a) shows the morphology of open end of NT while Figure 11 (b) shows the surface morphology of the closed end side of NT. The freestanding NTs can be used in two different orientations for fixation on TCO substrate; one with open end of NT facing the substrate while other with closed end of NT facing the substrate as shown in Figure 12. This section thus tends to highlight the effect of NT orientation on DSC performance. It was reported earlier that the closed end of NT facing the substrate might be helpful in improving the efficiency of DSC by serving as a barrier layer in between substrate and TiO2 active layer improving the charge transport by minimizing the substrate/TiO2 interface recombination analogous to the compact layer in NP based DSC.

Fig. 11. **(a)** Top view of NT showing one end to be open; **(b)** bottom view of NT showing other end to be closed

Fig. 12. Simplified DSC structure with CED and OED orientation of NT on TCO

Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells 551

Fig. 14. Shows the cross-sectional SEM image of TiO2 NTs on FTO glass (**a**) CED orientation; (**b**) CED NT/TiO2 NP interface; (**c**) OED orientation; (**d**) OED NT/ TiO2 NP interface

In order to investigate the reason for difference in the PV performance of two structures we performed the cross-sectional SEM imaging of CED and OED structures shown in Figure 14. The interface between colloidal TiO2 NP layer and the NT for CED structure (shown in Figure 14 (a & b)) can be seen to have gaps in between these two layers which suggest that the electron transfer between these two layers is not efficient leading to excessive slow down of the electrons at this interface increasing the recombination probability. We attribute the poor interface quality of this structure to the round shaped closed end of the NT which might have prevented the colloidal particles to partially penetrate into the tube leading to weak interface formation which upon high temperature sintering of the film might have introduced gaps at the interface. Interestingly this feature was not observed in the case of OED structure as can be seen from the cross-sectional image of Figure 14 (c & d). The NTs were found to have formed very good interface by embedding itself into the NP matrix leaving behind no gaps. It can be seen from the image that even after sintering at high

In order to investigate the reason for higher photocurrent in OED structure we performed dye loading measurements for two cells. The dye loading densities for cells with OED and CED structures were found to be ~ 7.16 x 10-6 mol g-1 and 3.58 x 10-6 mol g-1 respectively which indicates higher dye loading for OED compared to CED structure and hence higher photocurrent. In addition we also anticipate that the improved photocurrent can also be a result of higher confinement of light in the active layer of TiO2 due to the nano-dome structure of closed end being on top leading to the increase in optical path length and hence

temperature the interface retained its good morphology.

(a) (b)

(c) (d)

In contrast it was also reported that ~ 2 – 3 µm thick layer of TiO2 at the closed end of NT might serve as an insulating layer between TCO/TiO2 layers which can be detrimental for effective charge transport from active layer to the electrode. In order to investigate the effect of closed end layer on PV performance, we fabricated DSC with two different orientations i.e. closed end facing the substrate and open end facing the substrate, hereafter referred to as CED and OED respectively. The DSCs fabricated with these two structures have apparently shown a big difference in their PV performance as can be seen from the J-V characteristics shown in Figure 13 (a).

Fig. 13. **(a)** J-V curve under illumination for cells with OED and CED structures; **(b)** EQE curves for cells with OED and CED structures

It was found that the OED structure had higher efficiency of 6.58% as opposed to 4.17% efficiency of CED structure. It was found that cell with OED structure exhibited higher values of short circuit current density (JSC), open circuit voltage (VOC) and fill factor (FF) compared to CED structure. The J-V data for the photovoltaic performance of two cells is provided in Table 1.


Table 1. J-V data for cells with OED and CED orientation of NTs.

In order to further support our J-V data we performed the external quantum efficiency (EQE) measurements on two cells as shown in Figure 13 (b). The EQE data was found to be very consistent with our J-V data where OED structure have shown greater quantum efficiency compared to CED structure. The current densities calculated from the EQE measurements were found to be ~ 15 and 10 mA/cm2 for OED and CED structures respectively which were in close agreement with the J-V data. Overall the cell performance indicated the superiority of the OED over CED orientation.

In contrast it was also reported that ~ 2 – 3 µm thick layer of TiO2 at the closed end of NT might serve as an insulating layer between TCO/TiO2 layers which can be detrimental for effective charge transport from active layer to the electrode. In order to investigate the effect of closed end layer on PV performance, we fabricated DSC with two different orientations i.e. closed end facing the substrate and open end facing the substrate, hereafter referred to as CED and OED respectively. The DSCs fabricated with these two structures have apparently shown a big difference in their PV performance as can be seen from the J-V characteristics

Fig. 13. **(a)** J-V curve under illumination for cells with OED and CED structures; **(b)** EQE

**NT length (µm)** 

OED 3 22 14.75 666 67.05 6.58 CED 2.7 23 9.5 642 68.45 4.17

In order to further support our J-V data we performed the external quantum efficiency (EQE) measurements on two cells as shown in Figure 13 (b). The EQE data was found to be very consistent with our J-V data where OED structure have shown greater quantum efficiency compared to CED structure. The current densities calculated from the EQE measurements were found to be ~ 15 and 10 mA/cm2 for OED and CED structures respectively which were in close agreement with the J-V data. Overall the cell performance

**JSC (mA/cm2) VOC (mV) FF (%) η (%)** 

It was found that the OED structure had higher efficiency of 6.58% as opposed to 4.17% efficiency of CED structure. It was found that cell with OED structure exhibited higher values of short circuit current density (JSC), open circuit voltage (VOC) and fill factor (FF) compared to CED structure. The J-V data for the photovoltaic performance of two cells is

shown in Figure 13 (a).

provided in Table 1.

**Orientation of NT** 

curves for cells with OED and CED structures

**NP layer thickness (µm)** 

Table 1. J-V data for cells with OED and CED orientation of NTs.

indicated the superiority of the OED over CED orientation.

Fig. 14. Shows the cross-sectional SEM image of TiO2 NTs on FTO glass (**a**) CED orientation; (**b**) CED NT/TiO2 NP interface; (**c**) OED orientation; (**d**) OED NT/ TiO2 NP interface

In order to investigate the reason for difference in the PV performance of two structures we performed the cross-sectional SEM imaging of CED and OED structures shown in Figure 14. The interface between colloidal TiO2 NP layer and the NT for CED structure (shown in Figure 14 (a & b)) can be seen to have gaps in between these two layers which suggest that the electron transfer between these two layers is not efficient leading to excessive slow down of the electrons at this interface increasing the recombination probability. We attribute the poor interface quality of this structure to the round shaped closed end of the NT which might have prevented the colloidal particles to partially penetrate into the tube leading to weak interface formation which upon high temperature sintering of the film might have introduced gaps at the interface. Interestingly this feature was not observed in the case of OED structure as can be seen from the cross-sectional image of Figure 14 (c & d). The NTs were found to have formed very good interface by embedding itself into the NP matrix leaving behind no gaps. It can be seen from the image that even after sintering at high temperature the interface retained its good morphology.

In order to investigate the reason for higher photocurrent in OED structure we performed dye loading measurements for two cells. The dye loading densities for cells with OED and CED structures were found to be ~ 7.16 x 10-6 mol g-1 and 3.58 x 10-6 mol g-1 respectively which indicates higher dye loading for OED compared to CED structure and hence higher photocurrent. In addition we also anticipate that the improved photocurrent can also be a result of higher confinement of light in the active layer of TiO2 due to the nano-dome structure of closed end being on top leading to the increase in optical path length and hence

Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells 553

Bisquert, J. & Zaban, A (2003). The trap limited diffusivity of electrons in nanoporous

Burke, A.; Ito, S, Snaith, H., Bach, U., Kwiatkowski, J. & Grätzel, M (2008). The function of a

Cameron, P.J. & Peter, L.M. (2003). Characterization of titanium dioxide blocking layers in

Cameron, P.J. & Peter, L.M. (2005). How important is the back reaction of electrons via the

Cao, C.; Zhang, G., Song, X. & Sun, Z. (2011). Morphology and microstructure of as

Cao, F.; Oskam, G., Meyer, G.J. & Searson, P.C. (1996). Electron transport in porous

Cass, M. J.; Qiu, F.L., Walker, A.B., Fisher, A.C. & Peter, L.M. (2003). Influence of grain

Chen, Q. & Xu, D. (2009). Large-scale, noncurling, and free-standing crystallized TiO2 NT

Chen, Q.; Xu, D., Wu, Z. & Liu, Z. (2008). Free-standing TiO2 NT arrays made by anodic oxidation and ultrasonic splitting. (July, 2008). *Nanotechnology*, 19, 365708 Crawford, G.A. & Chawla, N. (2009). Tailoring TiO2 NT growth during anodic oxidation by crystallographic orientation of Ti. (February, 2009)., *Scripta Materialia*, 60, 874-877 Dor, S.; Grinis, L., Ruhle, S. & Zaban. A. (2009). Electrochemistry in mesoporous electrodes:

sensitized solar cells (January 2009). *J. Phys. Chem. C*, 113 (5), 2022-2027 Dubey, M. & He, H. (2009). Morphological studies of vertically aligned TiO2 NTs for high

Dubey, M.; Shrestha, M., Zhong, Y., Galipeau, D. & He, H. (2011). TiO2 NT membranes on

Ghicov, A. & Schmuki, P. (2009). Self-ordering electrochemistry: a review on growth and

Han, L.; Koide, N., Chiba, Y., Islam, A., Komiya, R., Fuke, N., Fukui, A. & Yamanaka, R.

He, H.; Dubey, M., Zhong, Y., Shrestha, M. & Sykes, A.G. (2011). 2-(1-Acetyl-2-oxopropyl)-

He, H.; Sykes, A.G., Dubey, M., Yan, X., Galipeau, D. & Ropp, M. (2008). *Proc. 33rd IEEE* 

efficiency solar cell. *Proc. 34th IEEE Photovolt. Conf.* (2009), 002130

internal resistance. *Appl. Phys. Letts.* (May, 2005), 86, 213501

synthesized anodic TiO2 NT arrays. *Nanoscale Res. Lett* (2011), 6, 64

*Appl. Phys. A.*, 77, 507-514

14394-14400

17021-17027

6314

*Chem. B.*, 109, 930-936

(March, 2008). *Nano Letts.* 8, 4, 976-981

(October, 2002). *J. Phys. Chem. B*, 107, 113-119

2011)., *Nanotechnology*, 22, 285201

*Eur. J. Inorg. Chem.* 25, 3731-3738

(April, 2009), 2791-2808

*Photovolt. Spec. Conf.* 

semiconductor networks permeated with a conductive phase (November, 2002).

TiO compact layer in dye-sensitized solar cells incorporating "planar" organic dyes

dye-sensitized nanocrystalline solar cells (October, 2003). *J. Phys. Chem. B.*, 107,

substrate in dye-sensitized nanocrystalline solar cells? (October, 2004). *J. Phys.* 

nanocrystalline TiO2 photoelectrochemical cells (October, 1996). *J. Phys. Chem*, 100,

morphology on electron transport in dye sensitized nanocrystalline solar cells

arrays for dye-sensitized solar cells. (March, 2009). *J. Phys. Chem. C.*, 113, 15, 6310-

Influence of nanoporosity on the chemical potential of the electrolyte in dye

transparent conducting glass for high efficiency dye-sensitized solar cells. (May,

functionality of TiO2 NTs and other self aligned MOx structures. *Chem. Commun.*

(2005). Improvement of efficiency of dye-sensitized solar cells by reduction of

5,10,15,20-tetraphenyl Porphyrin and its transition metal complexes. (July, 2011).

improved absorption. A schematic for light confinement effect for CED & OED structures are shown in Figure 15 (a & b) respectively. Overall it can be seen that orientation of the NTs for cell fabrication also plays a critical role in determining the efficiency of DSC.

Fig. 15. (**a**) Schematic of light propagation through NT photoelectrode on FTO with (**a**) CED structure; and (**b**) OED structure

## **5. Conclusions**

We found that morphology of NTs largely depends on the macro and microstructural topology of the substrate. Removal of substrate induced disorder in the morphology might be difficult by using simple ultrasonication or drying processes. A simple chemical pretreatment of substrate leads to substantial change in the morphology of grown NTs that can help in obtaining highly oriented and ordered TiO2 NT arrays. The chemical pretreatment technique can find potential utility for being simple, cost effective and less time consuming. In addition we also found that the orientation of the NTs was critical in determining the efficiency of DSC. Hence a meticulous choice of NT orientation along with surface texturing of substrate can significantly help in engineering NT morphology for its successful implementation as a promising material for solar cells as well as other optoelectronic device applications.

#### **6. References**


improved absorption. A schematic for light confinement effect for CED & OED structures are shown in Figure 15 (a & b) respectively. Overall it can be seen that orientation of the NTs

Fig. 15. (**a**) Schematic of light propagation through NT photoelectrode on FTO with (**a**) CED

We found that morphology of NTs largely depends on the macro and microstructural topology of the substrate. Removal of substrate induced disorder in the morphology might be difficult by using simple ultrasonication or drying processes. A simple chemical pretreatment of substrate leads to substantial change in the morphology of grown NTs that can help in obtaining highly oriented and ordered TiO2 NT arrays. The chemical pretreatment technique can find potential utility for being simple, cost effective and less time consuming. In addition we also found that the orientation of the NTs was critical in determining the efficiency of DSC. Hence a meticulous choice of NT orientation along with surface texturing of substrate can significantly help in engineering NT morphology for its successful implementation as a promising material for solar cells as well as other

Albu, S.P.; Ghicov, A., Macak, J.M., Hahn, R. & Schmuki, P. (2007). Self-organized , free-

Ali, G.; Chen, C., Yoo, S.H., Kum, J.M. & Cho, S.O. (2011). Fabrication of complete titania

Anta, J.A.; Casanueva, F. & Oskam, G. (2006). A numerical model for charge transport and

Bisquert, J. & Vikhrenko, V. S. (2004). Interpretation of the time constant measured by

solar cells. (January, 2004). *J. Phys. Chem. B.*, 108, 2313-2322

standing TiO2 NT membrane for flow through photocatalytic applications. (April,

nanoporous structures via electrochemical anodization of Ti. *Nanoscale Research* 

recombination in dye-sensitized solar cells. (March, 2006). *J. Phys. Chem. B.*, 110,

kinetic techniques in nanostructured semiconductor electrodes and dye-sensitized

structure; and (**b**) OED structure

optoelectronic device applications.

*Letts.* 2011, 6:332

5372-5378

2007). Nano Letts., 7, 5, 1286-1289

**5. Conclusions** 

**6. References** 

for cell fabrication also plays a critical role in determining the efficiency of DSC.


Morphological and Photovoltaic Studies of TiO2 NTs for High Efficiency Solar Cells 555

Ofir, A.; Grinis, L. & Zaban, A. (2008). Direct measurement of the recombination losses via

Oshaki, Y.; Masaki, N., Kitamura, T., Wada, Y., Okamoto, T., Sekino, T., Niihara, K. &

Prakasam, H. E.; Shankar, K., Paulose, M., Varghese, O.K. & Grimes, C.A. (2007). A new

Robertson, N. (2006). Optimizing dyes for dye-sensitized solar cells (2006). *Angew. Chem. Int.* 

Roy, P.; Kim, D., Lee, K., Spiecker, E. & Schmuki, P. (2010). TiO2 NTs and their application in

Ruan, C.; Paulose, M., Varghese, O. K., Mor, G. K. & Grimes, C.A. (2005). Fabrication of

Santiago, F.; Bisquert, J., Belmonte, G., Boschloo, G. & Hagfeldt, A. (2005). Influence of

Santiago, F.F.; Barea, E.M., Bisquert, J., Mor, G.K., Shankar, K. & Grimes, C.A. (2008). High

Santiago, F.F.; Belmonte, G.G., Bisquert, J., Zaban, A. & Salvador, P. (2002). Decoupling of

Santiago, F.F.; Bisquert, J., Belmonte, G.G., Boschloo, G. & Hagfeldt, A. (2005). Influence of

Sero, I. M.; Dittrich, T., Belaidi, A., Belmonte, G.G. & Bisquert, J. (2005). Observation of

Sero, I.M. & Bisquert, J. (2003). Fermi level of surface states in TiO2 NPs. (June, 2003). *Nano* 

Sero, I.M.; Dittrich, T., Belmonte, G.G. & Bisquert, J. (2006). Determination of spatial charge

Shiga, A.; Tsujiko, A., Ide, T., Yae, S. & Nakato, Y (1998). Nature of electrical junction at the

dye-sensitized solar cells. *Nanoscale*. (December, 2009), 2, 45-49

doping. (August, 2008). *J. Am. Chem. Soc.*, 130, 34, 11312-11316

characterization. (October, 2005). *Phys. Chem. Chem. Phys.*, 7, 4157-4163 Park, J. H.; Lee, T-W. & Kang, M.G. (2008). Growth, detachment and transfer of highly-

*Phys. Chem. C.*, 112, 2279-2783

*Commun.*, 2867-2869

*C.* , 111, 20, 7235-7241

*Chem. B.*, 109, 33, 15754-15759

*Ed.* 45, 2338-2345

117-131

*B.*, 106, 334-339

117-131

14932-14938

*Letts.* 3, 7, 945-949

2006). *J. App. Phys.* 100, 1

(May, 1998). *J. Phys. Chem. B.*, 102, 6049-6055

the transparent conductive substrate in dye sensitized solar cells. (January, 2008). *J.* 

Yanagida, S. (2005). Dye-sensitized TiO2 NT solar cells: fabrication and electronic

ordered TiO2 NT arrays: use in dye-sensitized solar cells. (May, 2008). *Chem.* 

benchmark for TiO2 NT array growth by anodization. (April, 2007). *J. Phys. Chem.* 

highly ordered TiO2 NT arrays using an organic electrolyte. (July, 2005)., *J. Phys.* 

electrolyte in transport and recombination in dye-sensitized solar cells studied by impedance spectroscopy (November, 2004). *Solar Energy Material & Solar Cells*, 87,

carrier density and capacitance in TiO NT arrays induced by electrochemical

transport, charge storage, and interfacial charge transfer in the nanocrystalline TiO2/electrolyte system by impedance methods. (December, 2001). *J. Phys. Chem.* 

electrolyte in transport and recombination in dye-sensitized solar cells studied by impedance spectroscopy (November, 2004). *Solar Energy Materials & Solar Cells*, 87,

diffusion and tunneling recombination of dye-photoinjected electrons in ultrathin TiO2 layers by surface photovoltage transients. (May, 2005). *J. Phys. Chem. B.* 109,

separation of diffusing electrons by transient photovoltage measurements. (August,

TiO2/substrate interface for particulate TiO2 film electrodes in aqueous electrolyte


In, S-I.; Hou, Y., Abrams, B.L., Vesborg, P.C.K. & Chorkendorff, I. (2010). Controlled

Jennings, J. R.; Ghicov, A., Peter, L.M., Schmuki, P. & Walker, A.B. (2008). Dye-sensitized

Kang, S.H.; Kim, H.S., Kim, J-Y. & Sung, Y-E. (2009). An investigation on electron behavior

Kang, S.H.; Kim, J-Y., Kim, H-S. & Sung, Y-E. (2008). Formation and mechanistic study of

Karthikeyan, C.S. & Thelakkat, M (2008). Key aspects of individual layers in solid-state dye-

Kontos, A.G.; Kontos, A.I., Tsoukleirs, D.S., Likodimos, V., Kunze, J., Schmuki, P. & Falaras,

Lee, K-M.; Suryanarayanan, V. & Ho, K-C (2006). The influence of surface morphology of

Li, S.; Zhang, G, Guo, D, Yu, L & Zhang, W. (2009). Anodization fabrication of highly ordered TiO2 nanotubes. (May, 2009), J. Phys. Chem. C, 113, 12759-12765 Liberator, M.; Burtone, L., Brown, T.M.., Reale, A., Carlo, A.D., Decker, F., Caramori, S. &

Lin, C.J.; Yu, W-Y. & Chien, S-H. (2010). Transparent-electrodes of ordered opened-ended

Macak, J.M.; Hildebrand , H., Marten-Jhans, U. & Schmuki, P. (2008). Mechanistic aspects

Mohammadpour, A. & Shankar, K. (2010) Anodic TiO2 NT arrays with optical wavelength-

Mor, G. K.; Varghese, O.K., Paulose, M. & Grimes, C.A. (2005). Transparent highly ordered

Nazeeruddin, M. K.; Angelis, F. D., Fantacci, S., Selloni, A., Viscardi, G., Liska, P., Ito, S.,

Nusbaumer, H.; Zakeeruddin, S.M., Moser, J-E. & Grätzel, M (2003). An alternative redox couple for the dye-sensitized solar cell system. *Chem. Eur. J.* 2003, 9, 3756-3763 O'Regan, B. & Grätzel, M (1991). A low-cost, high-efficiency solar cell based on dyesensitized colloidal TiO2 films. (October, 1991). *Nature*, 353, 737-740

sized apertures. *J. Mater. Chem.* (September, 2010), 20, 8474-8477

(November, 2005). *J. Am. Chem. Soc.*, 127, 48, 16835-16847

surface morphology. (December, 2008). *Nanotechnology*, 20, 045603

electrons. (September, 2008). *J. Am. Chem. Soc.*, 130, 40, 13364-13372

(August, 2009), *Nanotechnology*, 20, 355307

2007). *Inorganica Chimica Acta*, 361, 635-655

*Energy materials & Solar Cells*, 90, 2398-2404

*Engineering Chemistry*, 14, 52-59

157, (5), E69-E74

173113

1296

*Mater. Chem.*, 20, 1073-1077

*Electroanalytical Chemistry*., 621, 254-266

directional growth of TiO2 NTs. (March, 2010). *Journal of The Electrochemical Society*,

solar cells based on oriented TiO NT arrays: transport, trapping and transfer of

employing vertically-aligned TiO2 NT electrodes for dye-sensitized solar cells.

self-ordered TiO2 NTs on Ti substrate. (June, 2007). *Journal of Industrial and* 

sensitized solar cells and novel concepts to improve their performance (April,

P. (2009). Photo-induced effects on self organized TiO2 NT arrays: the influence of

TiO2 coating on the performance of dye-sensitized solar cells (May, 2006). *Solar* 

Bignozzi, C.A. (2009). On the effect of Al2O3 blocking layer on the performance of dye solar cells with cobalt based electrolytes. (April, 2009). *Appl. Phys. Lett.* 94,

TiO2-NT arrays for highly efficient dye-sensitized solar cells. (December, 2009). *J.* 

and growth of large diameter self-organized TiO2 NTs. (January, 2008). *Journal of* 

TiO2 NT arrays via anodization of titanium thin films. *Adv. Funct. Mater.*, 15, 1291-

Takeru, B. & Grätzel. M. (2005). Combined experimental and DFT-TDDFT computational study of photoelectrochemical cell ruthenium sensitizers.


**28** 

*South Korea* 

**Synthesis and Characterisation** 

**Film for Enamel Wire** 

Xiaokun Ma and Sun-Jae Kim\*

**of Silica/Polyamide-Imide Composite** 

*Institute/Faculty of Nanotechnology and Adv. Materials Engin., Sejong University #98 Gunja-dong, Gwangjin-gu, Seoul,* 

In the past decade, the demand for polyamide-imide (PAI) and other high-temperature resistant polymeric materials has grown steadily because of their outstanding mechanical properties and excellent thermal and oxidative stability (Zhong, 2002; Sun, 2006; Yanagishita, 2001; Babooram, 2008). PAI is well-known for its low thermal expansion coefficient and dielectric constant. In microelectronics, PAI has been widely used as an interdielectric material, and in the large-scale integrated circuit industry, as an electrical insulation for conventional appliances (Kawakami, 1996, 1998, 2003; Rupnowski, 2006; Wu, 2005). Compared with pure polyimide and polyamide, PAI exhibits better process ability and heat-resistant properties. The application of PAI as a wire-coating material with thermal-resistant properties has attracted increasing interest (Chen, 1997; Ranade, 2002; Ma, 2007). However, with the introduction of higher-surge voltage devices, an increasing number of insulation electric breakdown cases have been reported. Insulation electric breakdown must be prevented because it may lead to electrical component failure or may endanger the people handling the component. Thus, the development of an organic/inorganic composite insulating material is essential in designing insulation for continuous use (Alexandre, 2000; Hossein, 2007; David, 1995; Yang, 2006). Polymer composites have received much attention, as various properties of the original matrix polymer can be considerably improved by adding a limited percentage of inorganic filler

(Jiao, 1989; Rangsunvigit, 2008; Xu, 2007; Hwang, 2008; Kim, 2007; Rankin, 1998).

Mosher, 2006; Kim, 2006; Ahn, 2006; Stathatos, 2004).

Silica has been commonly used as an inorganic component because it is effective in enhancing the mechanical and thermal properties of polymers. Various studies on the preparation of polymer/silica composite films have been conducted (Butterworth, 1995;

The properties of hybrid composites are affected by many factors, such as particle size, size distribution, and filler content. In addition, the inorganic particle shape, surface structure,

**1. Introduction** 

 \*

Corresponding Author


## **Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire**

Xiaokun Ma and Sun-Jae Kim\*

*Institute/Faculty of Nanotechnology and Adv. Materials Engin., Sejong University #98 Gunja-dong, Gwangjin-gu, Seoul, South Korea* 

## **1. Introduction**

556 Scanning Electron Microscopy

Shin, Y. & Lee, S. (2008). Self-organized regular arrays of anodic TiO2 nanotubes.

Sun, L.; Zhang, S., Sun, X. & He, X. (2010). Effect of the geometry of the anodized titania NT

Tschirch, J.; Bahnemann, D., Wark, M. & Rathousky, J. (2008). A comparative study into the

Vanmaekelbergh, D. & de Jongh, P.E (1999). Driving force for electron transport in porous nanostructured photoelectrode (January, 1999). *J. Phys. Chem. B.*, 103, 5, 747-750 Varghese, O.K.; Paulose, M. & Grimes, C.A. (2010). Long vertically aligned titania NTs on

Wang, J. & Lin. Z. (2008). Freestanding NT arrays with ultrahigh aspect ratio via electrochemical anodization. (November, 2007). *Chem. Mater.* 20, 1257-1261 Wei, M.; Konishi, Y., Zhou, H., Yanagida, M., Sugihara, H. & Arakawa, H. (2006). Highly

Xu, T.; He, H., Wang, Q., Dubey, M., Galipeau, D. & Ropp, M. (2008). *Proc. 33rd IEEE* 

Yuan, L.; Xurui, X., Dongshe, Z., Puhui, X. & Baowen, Z (2003). Light scattering

Zhang, L. & Han, Y. (2010). Effect of nanostructured titanium on anodization growth of selforganized TiO2 nanotubes. (December, 2009), Nanotechnology., 21, 055602 Zhu, J.; Hsu, C-M., Yu, Z., Fan, S. & Cui, Y. (2010). Nanodome solar cells with efficient light management and self-cleaning. (November, 2009). *Nano Lett.* , 10, 1979-1984 Zhu, K.; Neale, N.R., Miedaner, A. & Frank, A. J. (2007). Enhanced charge collection

Zhu, K.; Vinzant, T.B., Neale, N.R. & Frank, A.J. (2007). Removing structural disorder from

array on the performance of dye-sensitized solar cells. *J. Nanosci. Nanotechnol*. , 10,

photocatalytic properties of thin mesoporous layers of TiO2 with controlled mesoporosity (August, 2007). *Journal of Photochemistry and Photobiology A: Chemistry*,

transparent conducting oxide for highly efficient solar cells. (August, 2009). *N.* 

efficient dye-sensitized solar cells composed of mesoporous titanium dioxide

characteristic of TiO2 nanocrystalline porous films (May, 2003). *Chinese Science* 

efficiencies and light scattering in dye-sensitized solar cells using oriented TiO2 NT

oriented TiO NT arrays: Reducing the dimensionality of transport and recombination in dye-sensitized solar cells. (November, 2007). *Nano Lett.*, 7, 12,

(September, 2008), Nano Letts. 8, 10, 3171-3173

(January, 2006). *J. Mater. Chem.*, 16, 1287-1293

arrays. (December, 2006), *Nano Letts.*, 7, 1, 69-74

1-10

194, 181-188

*Nano*.2009.226

*Photovolt. Spec. Conf.* 

*Bulletin*, 48, 9

3739-3746

In the past decade, the demand for polyamide-imide (PAI) and other high-temperature resistant polymeric materials has grown steadily because of their outstanding mechanical properties and excellent thermal and oxidative stability (Zhong, 2002; Sun, 2006; Yanagishita, 2001; Babooram, 2008). PAI is well-known for its low thermal expansion coefficient and dielectric constant. In microelectronics, PAI has been widely used as an interdielectric material, and in the large-scale integrated circuit industry, as an electrical insulation for conventional appliances (Kawakami, 1996, 1998, 2003; Rupnowski, 2006; Wu, 2005). Compared with pure polyimide and polyamide, PAI exhibits better process ability and heat-resistant properties. The application of PAI as a wire-coating material with thermal-resistant properties has attracted increasing interest (Chen, 1997; Ranade, 2002; Ma, 2007). However, with the introduction of higher-surge voltage devices, an increasing number of insulation electric breakdown cases have been reported. Insulation electric breakdown must be prevented because it may lead to electrical component failure or may endanger the people handling the component. Thus, the development of an organic/inorganic composite insulating material is essential in designing insulation for continuous use (Alexandre, 2000; Hossein, 2007; David, 1995; Yang, 2006). Polymer composites have received much attention, as various properties of the original matrix polymer can be considerably improved by adding a limited percentage of inorganic filler (Jiao, 1989; Rangsunvigit, 2008; Xu, 2007; Hwang, 2008; Kim, 2007; Rankin, 1998).

Silica has been commonly used as an inorganic component because it is effective in enhancing the mechanical and thermal properties of polymers. Various studies on the preparation of polymer/silica composite films have been conducted (Butterworth, 1995; Mosher, 2006; Kim, 2006; Ahn, 2006; Stathatos, 2004).

The properties of hybrid composites are affected by many factors, such as particle size, size distribution, and filler content. In addition, the inorganic particle shape, surface structure,

<sup>\*</sup> Corresponding Author

Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire 559

the self-synthesised spherical silica particles were dispersed in separate PAI polymer matrices. The correlation of the silica particle size with the amount of silica dispersed in the

The spherical silica particles were prepared according to the Stőber procedure, which allows the preparation of monodispersed silica particles with particle sizes in the nanometer to submicron range. A 500 ml three-necked flask equipped with a mechanical stirrer was filled with 45 ml ethanol, 2 ml NH3 H2O, and 1 ml deionised H2O. The spherical silica synthesis was initiated by the rapid addition of 2 ml tetra-ethoxy-silane (TEOS) to a stirred solution. After the mixture was vigorously stirred for 2 h, spherical silica submicron particles of approximately 300 nm in size, were obtained at room temperature. Then, the silica particles were centrifuged at 10000 rpm for 30 min. The resultant silica particles were washed with

CTAB can be directly added into the submicron silica solutions, and the CTAB amount can be increased from 0 to 3wt % at the optimal temperature of 60 °C. After the surface modification process, the modified silica particles were collected via centrifugation and then

In the experiment, 100 ml deionised H2O and 1 g commercial silica nanoparticles were added to a flask and the solution was adjusted to pH 8 by the addition of 0.1 M NaOH. The silica nanoparticles were modified at 65 °C with the addition of CTAB under constant stirring. The dispersal state of the silica nanoparticles in the PAI matrix was improved by increasing the amount of CTAB from 0 to 3.0 wt%. After the surface modification process, the modified nanosilica particles were collected via suction filtration and then dried at 90 °C

The nanosilica/PAI composite films were prepared via simple ultrasonic blending. Two grams of PAI powder were dissolved in 3 ml N,N-dimethyl form amide (DMF). The silica nanoparticles were added into the solution, and the amount of silica was increased from 2 to 10 wt%. The mixture was put under ultrasonic dispersion for approximately 3 h at room temperature. The mixture solution was then cast on a square glass plate (5 cm × 5 cm) using a bar coater. Bar coaters are primarily used in applying a variety of coatings or emulsions to a multitude of substrates. The bar coater used in the current experiment had a 0.2 mm diameter and was made of stainless steel wire, which resulted in more uniform silica/PAI composite films. The films were initially heated to remove the solvent in the vacuum oven. The temperature was controlled as follows: at 60 °C for 4 h, at 80 °C for 2 h, increased to 120 °C for 1 h, and finally, kept at 160 °C for 1 h. The experimental details of the silica/PAI

PAI under optimal experimental conditions was discussed.

**2.1.1 Synthesis of the spherical silica submicron particles** 

ethanol and distilled H2O, and then modified with CTAB.

**2.1.2 Surface modification of the silica particles** 

**2.1.3 Preparation of the silica/PAI composite film** 

composite film preparation are shown in Scheme 1.

dried at 60 °C.

for 6 h.

**2.1 Preparation of the silica/PAI composite film** 

and mechanical properties of a filler (stiffness and strength, among others) play important roles in inorganic/organic composite material synthesis. In particular, the bond strength between the inorganic particles and the polymer matrix, influenced by the dispersion aid type or coupling agent used, should be improved (Kusakabe, 1996; Fuchigami, 2008; Castellano, 2005; Alexandre, 2000; Wu, 2006; Zheng, 2007; Ohki, 2005).

Silica nanoparticles, as important inorganic materials, have emerged as an area of intense interest because of their special physical and chemical properties, such as their small size, strong surface energy, high scattered performance, and thermal resistance (Ouabbas, 2009; Lee, 2006; Bhagat, 2008; Oh, 2009; Xue, 2009). However, the applications of silica nanoparticles are largely limited because of their highly energetic hydrophilic surface, which causes the silica nanoparticles to easily agglomerate. Surface modification methods using different surfactant agents may resolve this limitation. Thus, the strong interface adhesion between the organic matrix and the silica nanoparticles is the key to the application of silica nanoparticles as fillers.

Jadav et al. successfully synthesised a silica/polyamide nanocomposite film via interfacial polymerisation using two types of silica nanoparticles of 16 and 3 nm in size (Jadav, 2009). The nanocomposite films exhibited superior thermal stability to the pure polyamide membranes. In the current work, silica nanoparticle loading significantly modified the polyamide network structure, pore structure, and transport properties. The excellent membrane performance in terms of separation efficiency and productivity flux was also discussed. Zhang et al. prepared a novel isometric polyimide/silica hybrid material via solgel technique (Zhang, 2007). Initially, 3-[(4-phenylethynyl) phthalimide] propyl triethoxysilane was synthesised to modify the nanosilica precursor. Then, the isomeric polyimide/silica hybrid material was produced using isomeric polyimide resin solution and the modified nanosilica precursor after heat treatment. The isomeric polyimide/silica composite has much better thermal properties and nano-indenter properties than those of the isomeric polyimide.

In the current work, the commercial silica nanoparticles and self-synthesised spherical silica particles were successfully dispersed in the PAI polymer matrix after the surface modification process. The cationic surfactant cetyltrimethyl ammonium bromide (CTAB) was chosen to modify the silica nanoparticles. The amount of CTAB added in modifying the silica nanoparticles was increased from 0 to 3 wt%. After the surface modification process, the CTAB-modified silica nanoparticles showed better compatibility with the PAI polymer matrix. The results indicate that CTAB plays an important role in the preparation of silica/PAI composite film. The thermal stability improved and the decomposition temperature increased with increasing amounts of silica particles. The thermal expansion coefficient of the composite film was lower than that of the PAI polymer matrix, which is helpful in extending the life of the enameled wire.

## **2. Synthesis and characterisation of the spherical silica/PAI composite film**

Spherical silica/PAI composite films have been successfully prepared via simple ultrasonic blending. In the current study, the spherical silica particles were prepared according to the Stőber procedure, and the size was controlled to approximately 300 nm at room temperature. After the surface modification process, the commercial silica nanoparticles and the self-synthesised spherical silica particles were dispersed in separate PAI polymer matrices. The correlation of the silica particle size with the amount of silica dispersed in the PAI under optimal experimental conditions was discussed.

## **2.1 Preparation of the silica/PAI composite film**

558 Scanning Electron Microscopy

and mechanical properties of a filler (stiffness and strength, among others) play important roles in inorganic/organic composite material synthesis. In particular, the bond strength between the inorganic particles and the polymer matrix, influenced by the dispersion aid type or coupling agent used, should be improved (Kusakabe, 1996; Fuchigami, 2008;

Silica nanoparticles, as important inorganic materials, have emerged as an area of intense interest because of their special physical and chemical properties, such as their small size, strong surface energy, high scattered performance, and thermal resistance (Ouabbas, 2009; Lee, 2006; Bhagat, 2008; Oh, 2009; Xue, 2009). However, the applications of silica nanoparticles are largely limited because of their highly energetic hydrophilic surface, which causes the silica nanoparticles to easily agglomerate. Surface modification methods using different surfactant agents may resolve this limitation. Thus, the strong interface adhesion between the organic matrix and the silica nanoparticles is the key to the

Jadav et al. successfully synthesised a silica/polyamide nanocomposite film via interfacial polymerisation using two types of silica nanoparticles of 16 and 3 nm in size (Jadav, 2009). The nanocomposite films exhibited superior thermal stability to the pure polyamide membranes. In the current work, silica nanoparticle loading significantly modified the polyamide network structure, pore structure, and transport properties. The excellent membrane performance in terms of separation efficiency and productivity flux was also discussed. Zhang et al. prepared a novel isometric polyimide/silica hybrid material via solgel technique (Zhang, 2007). Initially, 3-[(4-phenylethynyl) phthalimide] propyl triethoxysilane was synthesised to modify the nanosilica precursor. Then, the isomeric polyimide/silica hybrid material was produced using isomeric polyimide resin solution and the modified nanosilica precursor after heat treatment. The isomeric polyimide/silica composite has much better thermal properties and nano-indenter properties than those of

In the current work, the commercial silica nanoparticles and self-synthesised spherical silica particles were successfully dispersed in the PAI polymer matrix after the surface modification process. The cationic surfactant cetyltrimethyl ammonium bromide (CTAB) was chosen to modify the silica nanoparticles. The amount of CTAB added in modifying the silica nanoparticles was increased from 0 to 3 wt%. After the surface modification process, the CTAB-modified silica nanoparticles showed better compatibility with the PAI polymer matrix. The results indicate that CTAB plays an important role in the preparation of silica/PAI composite film. The thermal stability improved and the decomposition temperature increased with increasing amounts of silica particles. The thermal expansion coefficient of the composite film was lower than that of the PAI polymer matrix, which is

**2. Synthesis and characterisation of the spherical silica/PAI composite film**  Spherical silica/PAI composite films have been successfully prepared via simple ultrasonic blending. In the current study, the spherical silica particles were prepared according to the Stőber procedure, and the size was controlled to approximately 300 nm at room temperature. After the surface modification process, the commercial silica nanoparticles and

Castellano, 2005; Alexandre, 2000; Wu, 2006; Zheng, 2007; Ohki, 2005).

application of silica nanoparticles as fillers.

helpful in extending the life of the enameled wire.

the isomeric polyimide.

## **2.1.1 Synthesis of the spherical silica submicron particles**

The spherical silica particles were prepared according to the Stőber procedure, which allows the preparation of monodispersed silica particles with particle sizes in the nanometer to submicron range. A 500 ml three-necked flask equipped with a mechanical stirrer was filled with 45 ml ethanol, 2 ml NH3 H2O, and 1 ml deionised H2O. The spherical silica synthesis was initiated by the rapid addition of 2 ml tetra-ethoxy-silane (TEOS) to a stirred solution. After the mixture was vigorously stirred for 2 h, spherical silica submicron particles of approximately 300 nm in size, were obtained at room temperature. Then, the silica particles were centrifuged at 10000 rpm for 30 min. The resultant silica particles were washed with ethanol and distilled H2O, and then modified with CTAB.

### **2.1.2 Surface modification of the silica particles**

CTAB can be directly added into the submicron silica solutions, and the CTAB amount can be increased from 0 to 3wt % at the optimal temperature of 60 °C. After the surface modification process, the modified silica particles were collected via centrifugation and then dried at 60 °C.

In the experiment, 100 ml deionised H2O and 1 g commercial silica nanoparticles were added to a flask and the solution was adjusted to pH 8 by the addition of 0.1 M NaOH. The silica nanoparticles were modified at 65 °C with the addition of CTAB under constant stirring. The dispersal state of the silica nanoparticles in the PAI matrix was improved by increasing the amount of CTAB from 0 to 3.0 wt%. After the surface modification process, the modified nanosilica particles were collected via suction filtration and then dried at 90 °C for 6 h.

## **2.1.3 Preparation of the silica/PAI composite film**

The nanosilica/PAI composite films were prepared via simple ultrasonic blending. Two grams of PAI powder were dissolved in 3 ml N,N-dimethyl form amide (DMF). The silica nanoparticles were added into the solution, and the amount of silica was increased from 2 to 10 wt%. The mixture was put under ultrasonic dispersion for approximately 3 h at room temperature. The mixture solution was then cast on a square glass plate (5 cm × 5 cm) using a bar coater. Bar coaters are primarily used in applying a variety of coatings or emulsions to a multitude of substrates. The bar coater used in the current experiment had a 0.2 mm diameter and was made of stainless steel wire, which resulted in more uniform silica/PAI composite films. The films were initially heated to remove the solvent in the vacuum oven. The temperature was controlled as follows: at 60 °C for 4 h, at 80 °C for 2 h, increased to 120 °C for 1 h, and finally, kept at 160 °C for 1 h. The experimental details of the silica/PAI composite film preparation are shown in Scheme 1.

Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire 561

by an FT-IR (OMNIC NICOLET 380) spectrometer. The spectra were measured in the range 4000–650 cm-1. A Scinco STA S-1500 simultaneous thermal analyser was then used to analyse the thermal stability of the nanosilica/PAI composite films. The samples were heated from 30 to 800 °C at 10 °C/min under air atmosphere. The coefficients of thermal expansion (CTE) of the silica/PAI composites films were evaluated using a Q 400 EM (U.S.A) thermomechanical analyser (5 °C/min from 25 to 300 °C, 50 mN). All the samples were

The FT-IR spectra of the silica/PAI composite films are shown in Fig. 1. The effect of the surfactant on the composite films was evaluated by increasing the CTAB dosage from 0 to 3 wt%. The amount of silica nanoparticles added to the PAI was 6 wt%. The character vibrations of the Si-O were observed at 1086, 945, and 796 cm-1, as shown in Fig. 1 (e). After the surface modification process, the typical stretching vibrations of the C-H were found at 2855 and 2928 cm-1, which resulted from the –CH2 and –CH3 in the CTAB. Figure 1 shows the typical characteristic bands of the PAI polymer matrix that were found, such as the N-H stretching band at 3317 cm-1, the amide C=O region at around 1710 cm-1, and the bands at 1771 and 1710 cm-1 associated with the imide carbonyl band. The bands are similar to one another because the same amount of silica was added into the composites. The characteristic stretching vibration of Si-O at 1086 cm-1 became wider when the silica nanoparticles were modified by CTAB. This peak broadening may be explained by the organic side-chain of CTAB grafted on the surface of the silica nanoparticles, which improved the interaction

Fig. 1. FT-IR spectra of (a) unmodified-silica/PAI nanocomposites, (b) 1% CTAB-silica/PAI

The fracture surface micrographs of the silica/PAI nanocomposite films are shown in Fig. 2. In Fig. 2 (a), some silica agglomerations are found in the micrograph when the unmodifiedsilica nanoparticles were added into the PAI matrix. The silica nanoparticles easily

nanocomposites, (c) 2% CTAB-silica/PAI nanocomposites, (d) 3% CTAB-silica/PAI

3 mm × 16 mm, cut from the original films using a razor blade.

between the silica nanoparticles and the PAI polymer matrix.

nanocomposites, and (e) CTAB-modified silica nanoparticles.

**2.2.1 CTAB effect on the synthesis of silica/PAI composite film** 

Scheme 1. Preparation of the silica/PAI nanocomposite films.

The spherical silica/PAI composite films were obtained under the same method. Spherical submicron silica particles do not easily agglomerate, so the amount of submicron silica added to PAI in the system was increased to 25%. The reactants of the submicron silica/PAI composites are listed in Table 1.


Table 1. Reactants of the different submicron silica/PAI composite samples.

#### **2.2 Characterisation of the silica/PAI composite film**

The fracture surfaces of the composite films were studied using a scanning electron microscope (SEM Hitachi S-4700, Hitachi Co.). Prior to SEM imaging, the samples were sputtered with thin layers of Pt-Pd. The silica/PAI nanocomposite films were characterised by an FT-IR (OMNIC NICOLET 380) spectrometer. The spectra were measured in the range 4000–650 cm-1. A Scinco STA S-1500 simultaneous thermal analyser was then used to analyse the thermal stability of the nanosilica/PAI composite films. The samples were heated from 30 to 800 °C at 10 °C/min under air atmosphere. The coefficients of thermal expansion (CTE) of the silica/PAI composites films were evaluated using a Q 400 EM (U.S.A) thermomechanical analyser (5 °C/min from 25 to 300 °C, 50 mN). All the samples were 3 mm × 16 mm, cut from the original films using a razor blade.

#### **2.2.1 CTAB effect on the synthesis of silica/PAI composite film**

560 Scanning Electron Microscopy

The spherical silica/PAI composite films were obtained under the same method. Spherical submicron silica particles do not easily agglomerate, so the amount of submicron silica added to PAI in the system was increased to 25%. The reactants of the submicron silica/PAI

(g)

The fracture surfaces of the composite films were studied using a scanning electron microscope (SEM Hitachi S-4700, Hitachi Co.). Prior to SEM imaging, the samples were sputtered with thin layers of Pt-Pd. The silica/PAI nanocomposite films were characterised

Composite 1 0.04 2.0 3 1.96 Composite 2 0.08 2.0 3 3.85 Composite 3 0.12 2.0 3 5.66 Composite 4 0.16 2.0 3.5 7.41 Composite 5 0.20 2.0 3.5 9.10 Composite 6 0.30 2.0 3.5 13.0 Composite 7 0.40 2.0 4 16.7 Composite 8 0.50 2.0 4 20.0

Table 1. Reactants of the different submicron silica/PAI composite samples.

**2.2 Characterisation of the silica/PAI composite film** 

DMF (ml)

Theoretical Weight Percent (wt%)

Scheme 1. Preparation of the silica/PAI nanocomposite films.

Reactant Sample Silica Particles (g) PAI

composites are listed in Table 1.

The FT-IR spectra of the silica/PAI composite films are shown in Fig. 1. The effect of the surfactant on the composite films was evaluated by increasing the CTAB dosage from 0 to 3 wt%. The amount of silica nanoparticles added to the PAI was 6 wt%. The character vibrations of the Si-O were observed at 1086, 945, and 796 cm-1, as shown in Fig. 1 (e). After the surface modification process, the typical stretching vibrations of the C-H were found at 2855 and 2928 cm-1, which resulted from the –CH2 and –CH3 in the CTAB. Figure 1 shows the typical characteristic bands of the PAI polymer matrix that were found, such as the N-H stretching band at 3317 cm-1, the amide C=O region at around 1710 cm-1, and the bands at 1771 and 1710 cm-1 associated with the imide carbonyl band. The bands are similar to one another because the same amount of silica was added into the composites. The characteristic stretching vibration of Si-O at 1086 cm-1 became wider when the silica nanoparticles were modified by CTAB. This peak broadening may be explained by the organic side-chain of CTAB grafted on the surface of the silica nanoparticles, which improved the interaction between the silica nanoparticles and the PAI polymer matrix.

Fig. 1. FT-IR spectra of (a) unmodified-silica/PAI nanocomposites, (b) 1% CTAB-silica/PAI nanocomposites, (c) 2% CTAB-silica/PAI nanocomposites, (d) 3% CTAB-silica/PAI nanocomposites, and (e) CTAB-modified silica nanoparticles.

The fracture surface micrographs of the silica/PAI nanocomposite films are shown in Fig. 2. In Fig. 2 (a), some silica agglomerations are found in the micrograph when the unmodifiedsilica nanoparticles were added into the PAI matrix. The silica nanoparticles easily

Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire 563

PAI matrix. The amount of silica in the composites was calculated based on the plots, and

Fig. 3. TGA plots of (a) unmodified-silica/PAI nanocomposites, (b) 1% CTAB-silica/PAI nanocomposites, (c) 2% CTAB-silica/PAI nanocomposites, and (d) 3% CTAB-silica/PAI

Thus, CTAB, with an optimal dosage of 3 wt%, was chosen to modify the nanosilica. The fracture surface micrographs of the composites show that the silica nanoparticles were welldispersed in the PAI matrix after the surface modification process. In the TGA plots of the silica/PAI nanocomposites, the thermal stability and the decomposition temperature increased with increasing CTAB. Therefore, CTAB is important in the preparation of the

The FT-IR spectra of the pure PAI and some silica/PAI nanocomposite films are shown in Fig. 4 The amount of silica nanoparticles added to PAI was changed from 2 to 10 wt%, and the dosage of CTAB added to the silica was 3 wt%. In the spectra shown in Fig. 4, the bands at 1771 and 1710 cm-1 are associated with the imide carbonyl band. Both bands are insensitive to the presence of the silica nanoparticles. The bands in the region from 945 to 650 cm-1 increased with the increase in silica content, caused by the presence of a broad band associated with the vibration of the Si-O bond. The N-H stretching band at 3317 cm-1 was slightly intensified with the increase in silica content, indicating the hydrogenbonded N-H groups in the PAI polymer and the Si-O-Si or Si-O-H groups of the silica nanoparticles. The characteristic band at 1594 cm-1 comes from the benzene-ring stretch and a contribution from the O-H bond in monomeric H2O, which also has a band at 1663 cm-1. All the characteristic peaks in the composites indicate that the interaction between the silica nanoparticles and PAI polymer matrix is sensitive to the amount of silica in the

**2.2.2 The effect of silica nanoparticle amount on the properties of silica/PAI** 

nanocomposites.

**composite film** 

composites.

silica/PAI nanocomposite film.

the result was about 5.6 wt%, which is in accordance with the theoretical data.

agglomerate because of their large surface-to-volume ratios and high surface tension. However, the CTAB-modified silica nanoparticles dispersed well in the PAI matrix with the increase in CTAB. The dispersal state of the silica nanoparticles improved when the silica nanoparticles were modified with 1 wt% CTAB, as shown in Fig. 2 (b). In Fig. 2 (c), the silica nanoparticles are almost monodispersed, and little agglomeration is observed when 2 wt% CTAB-modified silica nanoparticles were added into the PAI polymer. After the silica were modified with 3 wt% CTAB, the silica nanoparticles became monodispersed without any agglomerations, although the amount of silica nanoparticles added to the PAI was increased to 6 wt%, as shown in Fig. 2 (d). CTAB improves the dispersal state of the silica nanoparticles in a PAI polymer matrix, with an optimal dosage of 3 wt%.

Fig. 2. Fracture surface micrographs of (a) unmodified-silica/PAI nanocomposites, (b) 1% CTAB-silica/PAI nanocomposites, (c) 2% CTAB-silica/PAI nanocomposites, and (d) 3% CTAB-silica/PAI nanocomposites.

Fig. 3 shows the thermogravimetric analysis (TGA) plots of the different CTAB modifiedsilica/PAI nanocomposites films when 6 wt% nanosilica was added into the PAI matrix. Some differences are found in the weight loss curves shown in Fig. 3. When the temperature was increased to 475 °C, the silica/PAI composites began to decompose. Compared with the unmodified-silica/PAI composites, the decomposition temperature increased after the silica nanoparticles were modified by CTAB. The results indicate that the CTAB-modified silica particles improve the thermal stability of the PAI polymer matrix. In addition, the decomposition temperature of composite films increased with increasing CTAB dosage. When the amount of CTAB added to the silica was increased from 0 to 3 wt%, the decomposition temperature of the composite films increased from 646 to 658, 671, and 682 °C, respectively. A thermal decomposition process occurs when the temperature approaches 595 °C, as shown in Figs. 3 (b)-(d). Therefore, the interaction between the silica nanoparticles and PAI polymer matrix is enhanced after the silica are modified by CTAB.

CTAB improves not only the dispersal state of silica in the PAI matrix, but also the thermal stability of the composite film because of the better interaction between the nanosilica and

agglomerate because of their large surface-to-volume ratios and high surface tension. However, the CTAB-modified silica nanoparticles dispersed well in the PAI matrix with the increase in CTAB. The dispersal state of the silica nanoparticles improved when the silica nanoparticles were modified with 1 wt% CTAB, as shown in Fig. 2 (b). In Fig. 2 (c), the silica nanoparticles are almost monodispersed, and little agglomeration is observed when 2 wt% CTAB-modified silica nanoparticles were added into the PAI polymer. After the silica were modified with 3 wt% CTAB, the silica nanoparticles became monodispersed without any agglomerations, although the amount of silica nanoparticles added to the PAI was increased to 6 wt%, as shown in Fig. 2 (d). CTAB improves the dispersal state of the silica

nanoparticles in a PAI polymer matrix, with an optimal dosage of 3 wt%.

Fig. 2. Fracture surface micrographs of (a) unmodified-silica/PAI nanocomposites, (b) 1% CTAB-silica/PAI nanocomposites, (c) 2% CTAB-silica/PAI nanocomposites, and

and PAI polymer matrix is enhanced after the silica are modified by CTAB.

Fig. 3 shows the thermogravimetric analysis (TGA) plots of the different CTAB modifiedsilica/PAI nanocomposites films when 6 wt% nanosilica was added into the PAI matrix. Some differences are found in the weight loss curves shown in Fig. 3. When the temperature was increased to 475 °C, the silica/PAI composites began to decompose. Compared with the unmodified-silica/PAI composites, the decomposition temperature increased after the silica nanoparticles were modified by CTAB. The results indicate that the CTAB-modified silica particles improve the thermal stability of the PAI polymer matrix. In addition, the decomposition temperature of composite films increased with increasing CTAB dosage. When the amount of CTAB added to the silica was increased from 0 to 3 wt%, the decomposition temperature of the composite films increased from 646 to 658, 671, and 682 °C, respectively. A thermal decomposition process occurs when the temperature approaches 595 °C, as shown in Figs. 3 (b)-(d). Therefore, the interaction between the silica nanoparticles

CTAB improves not only the dispersal state of silica in the PAI matrix, but also the thermal stability of the composite film because of the better interaction between the nanosilica and

(d) 3% CTAB-silica/PAI nanocomposites.

PAI matrix. The amount of silica in the composites was calculated based on the plots, and the result was about 5.6 wt%, which is in accordance with the theoretical data.

Fig. 3. TGA plots of (a) unmodified-silica/PAI nanocomposites, (b) 1% CTAB-silica/PAI nanocomposites, (c) 2% CTAB-silica/PAI nanocomposites, and (d) 3% CTAB-silica/PAI nanocomposites.

Thus, CTAB, with an optimal dosage of 3 wt%, was chosen to modify the nanosilica. The fracture surface micrographs of the composites show that the silica nanoparticles were welldispersed in the PAI matrix after the surface modification process. In the TGA plots of the silica/PAI nanocomposites, the thermal stability and the decomposition temperature increased with increasing CTAB. Therefore, CTAB is important in the preparation of the silica/PAI nanocomposite film.

#### **2.2.2 The effect of silica nanoparticle amount on the properties of silica/PAI composite film**

The FT-IR spectra of the pure PAI and some silica/PAI nanocomposite films are shown in Fig. 4 The amount of silica nanoparticles added to PAI was changed from 2 to 10 wt%, and the dosage of CTAB added to the silica was 3 wt%. In the spectra shown in Fig. 4, the bands at 1771 and 1710 cm-1 are associated with the imide carbonyl band. Both bands are insensitive to the presence of the silica nanoparticles. The bands in the region from 945 to 650 cm-1 increased with the increase in silica content, caused by the presence of a broad band associated with the vibration of the Si-O bond. The N-H stretching band at 3317 cm-1 was slightly intensified with the increase in silica content, indicating the hydrogenbonded N-H groups in the PAI polymer and the Si-O-Si or Si-O-H groups of the silica nanoparticles. The characteristic band at 1594 cm-1 comes from the benzene-ring stretch and a contribution from the O-H bond in monomeric H2O, which also has a band at 1663 cm-1. All the characteristic peaks in the composites indicate that the interaction between the silica nanoparticles and PAI polymer matrix is sensitive to the amount of silica in the composites.

Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire 565

Fig. 5. Fracture surface micrographs of (a) pure PAI and silica/PAI nanocomposite, with the amount of silica added to PAI at (b) 2 wt%, (c) 4 wt%, (d) 6 wt%, (e) 8 wt%, and (f) 10 wt%.

Fig. 6. TGA plots of the (a) PAI matrix, and the silica/PAI nanocomposite films with the amount of silica added to PAI at (b) 2 wt%, (c) 4 wt%, (d) 6 wt%, (e) 8 wt%, and (f) 10 wt%.

CTE is an important parameter in evaluating the properties of enamel wire. A low CTE can reduce thermal stress build-up and prevent device failure through peeling and cracking at the interface between the polymer film and the copper. The CTE curves of the pure PAI film

Fig. 4. FT-IR spectra of the silica/PAI composites with the amount of silica nanoparticles added to PAI at (a) 2 wt%, (b) 4 wt%, (c) 6 wt%, (d) 8 wt%, and (e) 10 wt%.

The fractured surface micrographs of the pure PAI and several composite films are shown in Fig. 5. The fracture surfaces of the pure PAI film are uniform, and the continuous polymer phase is shown in Fig. 5 (a). Figs. 5 (b)–(f) show the fracture surfaces of the different silica/PAI composites. The amount of silica added to the PAI was increased from 2 to 10 wt%. The larger the amount of silica nanoparticles added to the PAI, the greater their amount found in the fracture surface micrographs. In Fig. 5 (f), when 10 wt% silica nanoparticles was added into the PAI, the silica nanoparticles remained monodispersed without any agglomerations. The results indicate that the CTAB-modified silica nanoparticles have better dispersal state in the PAI polymer matrix. In addition, the CTABmodified silica nanoparticles increased the silica nanoparticle content in the composites. Thus, the surface modification process is an effective method of preparing silica/PAI nanocomposites.

The thermal stability of the silica/PAI composite films was evaluated via TGA. The TGA plots of the PAI and the composites with the different amounts of silica nanoparticles are shown in Fig. 6. The amount of silica nanoparticles added to the PAI polymer matrix was increased from 0 to 10 wt%. The plots are shown in Figs. 6 (a)–(f). A weight loss is observed above 170 °C on all the TGA plots, which corresponds to water and solvent losses. When the temperature was increased to 450 °C, the PAI matrix began to decompose. The decomposition temperature increased when the silica nanoparticles were added into the PAI polymer matrix. At the same temperature, all the curves of the composites indicated that the composite weight loss was less than that of the pure PAI matrix. The silica/PAI composites have higher decomposition temperature when the PAI polymer matrix loses the same weight. The thermal stability of PAI was enhanced with the increase in the silica content. The amount of silica nanoparticles in the composites were calculated accurately in the TGA plots. The silica content in the composites based on Figs. 6 (b)–(f) are 1.9, 3.6, 5.8, 7.3, and 10.2 wt%, respectively.

Fig. 4. FT-IR spectra of the silica/PAI composites with the amount of silica nanoparticles

The fractured surface micrographs of the pure PAI and several composite films are shown in Fig. 5. The fracture surfaces of the pure PAI film are uniform, and the continuous polymer phase is shown in Fig. 5 (a). Figs. 5 (b)–(f) show the fracture surfaces of the different silica/PAI composites. The amount of silica added to the PAI was increased from 2 to 10 wt%. The larger the amount of silica nanoparticles added to the PAI, the greater their amount found in the fracture surface micrographs. In Fig. 5 (f), when 10 wt% silica nanoparticles was added into the PAI, the silica nanoparticles remained monodispersed without any agglomerations. The results indicate that the CTAB-modified silica nanoparticles have better dispersal state in the PAI polymer matrix. In addition, the CTABmodified silica nanoparticles increased the silica nanoparticle content in the composites. Thus, the surface modification process is an effective method of preparing silica/PAI

The thermal stability of the silica/PAI composite films was evaluated via TGA. The TGA plots of the PAI and the composites with the different amounts of silica nanoparticles are shown in Fig. 6. The amount of silica nanoparticles added to the PAI polymer matrix was increased from 0 to 10 wt%. The plots are shown in Figs. 6 (a)–(f). A weight loss is observed above 170 °C on all the TGA plots, which corresponds to water and solvent losses. When the temperature was increased to 450 °C, the PAI matrix began to decompose. The decomposition temperature increased when the silica nanoparticles were added into the PAI polymer matrix. At the same temperature, all the curves of the composites indicated that the composite weight loss was less than that of the pure PAI matrix. The silica/PAI composites have higher decomposition temperature when the PAI polymer matrix loses the same weight. The thermal stability of PAI was enhanced with the increase in the silica content. The amount of silica nanoparticles in the composites were calculated accurately in the TGA plots. The silica content in the composites based on Figs. 6 (b)–(f) are 1.9, 3.6, 5.8, 7.3, and

added to PAI at (a) 2 wt%, (b) 4 wt%, (c) 6 wt%, (d) 8 wt%, and (e) 10 wt%.

nanocomposites.

10.2 wt%, respectively.

Fig. 5. Fracture surface micrographs of (a) pure PAI and silica/PAI nanocomposite, with the amount of silica added to PAI at (b) 2 wt%, (c) 4 wt%, (d) 6 wt%, (e) 8 wt%, and (f) 10 wt%.

Fig. 6. TGA plots of the (a) PAI matrix, and the silica/PAI nanocomposite films with the amount of silica added to PAI at (b) 2 wt%, (c) 4 wt%, (d) 6 wt%, (e) 8 wt%, and (f) 10 wt%.

CTE is an important parameter in evaluating the properties of enamel wire. A low CTE can reduce thermal stress build-up and prevent device failure through peeling and cracking at the interface between the polymer film and the copper. The CTE curves of the pure PAI film

Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire 567

properties. The characterisation results of the spherical silica/PAI composites are discussed

Fig. 8 shows the SEM micrographs of (a) the spherical silica submicron particles and (b) the fracture surface of the pure PAI and several spherical silica/PAI composite films. In Fig. 8 (a), the mean diameter of the spherical silica nanoparticles is about 300 nm, and most are well-dispersed after the surface modification process. The fracture surfaces of the pure PAI film are uniform, and the continuous polymer phase is shown in Fig. 8 (b). Figs. 8 (c)–(f) show the fracture surfaces of the different silica/PAI composites. The amount of silica added to the PAI was increased from 2 to 8 wt%. When the submicron spherical silica was added into the PAI matrix, some prominent features were observed on the fracture surfaces of the composite films, as shown in Fig. 8 (c). The spherical silica particles were embedded in the PAI matrix, and the continuous PAI organic phase appeared when the amount of silica added to the PAI was 2 and 4 wt%, as shown in Figs. 8 (c) and (d). When the amount of submicron silica increased, more spherical silica particles were observed on the fracture surfaces of the composite films. The continuous organic PAI phase separated, as shown in Fig. 8 (e). Some alveolate pores were observed in Fig. 8(f), when 8 wt% spherical silica was added into the PAI matrix. These pores are caused by the removal of the submicron silica

More submicron spherical silica particles were subsequently added into the PAI matrix. The SEM micrographs of the composite film fracture surfaces are shown in Fig. 9. The micrographs show that the diameters of the submicron silica particles are uniform. In addition, the spherical silica particles are orderly arranged in the PAI matrix. An increase in the silica particles added into the films may result in a more compact framework. Partial discharge is a prime factor causing enamel wire breakdown, so the composite films caused a decrease in the erosion rate via reflection and scattering when the spherical silica particles were added into the PAI matrix. That is, the charged particles were reflected and scattered around the submicron silica, which slowed down the corrosion process. Therefore, the

particles from the PAI matrix when the composite films were broken.

Scheme 2. Mechanism of erosion suppression.

as follows.

and some composite films are shown in Fig. 7. The CTE value of the PAI films was 3.87 × 10- 5 m/m/°C, whereas that of the silica/PAI composite film decreased to 3.69 × 10-5 and 3.51 × 10-5m/m/°C when the amount of silica added to PAI was 4 and 6 wt%, respectively. The CTE values continuously decreased with increasing amount of the silica particles. In particular, the CTE value decreased to 3.35 × 10-5 m/m/°C when the amount of silica added to PAI was 10 wt%, as shown in Fig. 7 (d). Compared with the PAI polymer film, the silica/PAI composite films had lower CTE, which may be attributed to the rigidity and stiffness of the silica nanoparticles and the interaction between the silica and PAI polymer matrix. The rigidity and stiffness of the silica nanoparticles limit the polymer chain movement, resulting in the decrease of the PAI matrix thermal expansion.

Fig. 7. CTE curves of the (a) pure PAI film and the composite films with silica content at (b) 4 wt%, (c) 6 wt%, and (d) 10 wt%.

The thermal stability of the silica/PAI nanocomposites improved and the decomposition temperature increased when the amount of silica nanoparticles was increased. The lower CTE of the composite films can reduce the peeling and cracking at the interface between the polymer film and the copper. In the current system, the high thermal stability and low CTE show that the silica/PAI nanocomposite films can be widely used in the enamel wire industry.

#### **2.2.3 Effect of silica diameter on the properties of the silica/PAI composite film**

In the past years, innovative inorganic/organic composite technology has been used to develop wires with better inverter-surge-resistance and mechanical properties than those of conventional enamelled wires. Kikuchi et al. (Kikuchi, 2002) emphasised that the inorganic/organic composite film can decrease erosion rate by increasing creeping distance and decreasing collision energy via reflection or scattering, as shown in Scheme 2.

Submicron spherical silica particles were added into the PAI polymer matrix during the synthesis of the silica/PAI composite films to evaluate the effects of silica diameter on their

and some composite films are shown in Fig. 7. The CTE value of the PAI films was 3.87 × 10- 5 m/m/°C, whereas that of the silica/PAI composite film decreased to 3.69 × 10-5 and 3.51 × 10-5m/m/°C when the amount of silica added to PAI was 4 and 6 wt%, respectively. The CTE values continuously decreased with increasing amount of the silica particles. In particular, the CTE value decreased to 3.35 × 10-5 m/m/°C when the amount of silica added to PAI was 10 wt%, as shown in Fig. 7 (d). Compared with the PAI polymer film, the silica/PAI composite films had lower CTE, which may be attributed to the rigidity and stiffness of the silica nanoparticles and the interaction between the silica and PAI polymer matrix. The rigidity and stiffness of the silica nanoparticles limit the polymer chain

Fig. 7. CTE curves of the (a) pure PAI film and the composite films with silica content at (b)

The thermal stability of the silica/PAI nanocomposites improved and the decomposition temperature increased when the amount of silica nanoparticles was increased. The lower CTE of the composite films can reduce the peeling and cracking at the interface between the polymer film and the copper. In the current system, the high thermal stability and low CTE show that the silica/PAI nanocomposite films can be widely used in the enamel wire

**2.2.3 Effect of silica diameter on the properties of the silica/PAI composite film** 

and decreasing collision energy via reflection or scattering, as shown in Scheme 2.

In the past years, innovative inorganic/organic composite technology has been used to develop wires with better inverter-surge-resistance and mechanical properties than those of conventional enamelled wires. Kikuchi et al. (Kikuchi, 2002) emphasised that the inorganic/organic composite film can decrease erosion rate by increasing creeping distance

Submicron spherical silica particles were added into the PAI polymer matrix during the synthesis of the silica/PAI composite films to evaluate the effects of silica diameter on their

4 wt%, (c) 6 wt%, and (d) 10 wt%.

industry.

movement, resulting in the decrease of the PAI matrix thermal expansion.

properties. The characterisation results of the spherical silica/PAI composites are discussed as follows.

Fig. 8 shows the SEM micrographs of (a) the spherical silica submicron particles and (b) the fracture surface of the pure PAI and several spherical silica/PAI composite films. In Fig. 8 (a), the mean diameter of the spherical silica nanoparticles is about 300 nm, and most are well-dispersed after the surface modification process. The fracture surfaces of the pure PAI film are uniform, and the continuous polymer phase is shown in Fig. 8 (b). Figs. 8 (c)–(f) show the fracture surfaces of the different silica/PAI composites. The amount of silica added to the PAI was increased from 2 to 8 wt%. When the submicron spherical silica was added into the PAI matrix, some prominent features were observed on the fracture surfaces of the composite films, as shown in Fig. 8 (c). The spherical silica particles were embedded in the PAI matrix, and the continuous PAI organic phase appeared when the amount of silica added to the PAI was 2 and 4 wt%, as shown in Figs. 8 (c) and (d). When the amount of submicron silica increased, more spherical silica particles were observed on the fracture surfaces of the composite films. The continuous organic PAI phase separated, as shown in Fig. 8 (e). Some alveolate pores were observed in Fig. 8(f), when 8 wt% spherical silica was added into the PAI matrix. These pores are caused by the removal of the submicron silica particles from the PAI matrix when the composite films were broken.

Scheme 2. Mechanism of erosion suppression.

More submicron spherical silica particles were subsequently added into the PAI matrix. The SEM micrographs of the composite film fracture surfaces are shown in Fig. 9. The micrographs show that the diameters of the submicron silica particles are uniform. In addition, the spherical silica particles are orderly arranged in the PAI matrix. An increase in the silica particles added into the films may result in a more compact framework. Partial discharge is a prime factor causing enamel wire breakdown, so the composite films caused a decrease in the erosion rate via reflection and scattering when the spherical silica particles were added into the PAI matrix. That is, the charged particles were reflected and scattered around the submicron silica, which slowed down the corrosion process. Therefore, the

Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire 569

Fig. 9. SEM micrographs of (a) composite 5 with 10 wt% submicron silica, (b) composite 6 with 15 wt% submicron silica, (c) composite 7 with 20 wt% submicron silica, and (d)

Fig. 10. FT-IR spectra of (a) CTAB-modified submicron silica particles, (b) pure PAI, (c) composite 1 with 2 wt% silica, (d) composite 2 with 4 wt% silica, and (e) composite 4 with

The TGA plots of the PAI and the composites with the different amounts of submicron silica particles are shown in Fig. 11. Obvious weight loss is found in these plots. A weight loss is observed below 180 °C in all the TGA plots, corresponding to water and solvent losses. When the temperature was increased to 450 °C, the PAI matrix began to decompose with increasing decomposition temperature. The decomposition temperature increased from

composite 8 with 25 wt% submicron silica.

8 wt% silica.

particle discharge resistance is improved with the increase in silica content. When the amount of silica added to the PAI was increased to 25 wt%, as shown in Fig. 9 (d), the fracture surfaces of the composite films remained well-integrated without the obvious phase separation. The results indicate that such a simple method effectively increases the amount of inorganic silica particles in the PAI matrix after the surface modification process.

Fig. 8. SEM micrographs of (a) spherical silica submicron particles and fracture surfaces of (b) pure PAI film, (c) composite 1 with 2 wt% submicron silica, (d) composite 2 with 4 wt% submicron silica, (e) composite 3 with 6 wt% submicron silica, and (f) composite 4 with 8 wt% submicron silica.

The FT-IR spectra of CTAB-modified silica, pure PAI, and some submicron silica/PAI composite films are shown in Fig. 10. As shown in Fig. 3 (a), typical bands of silica were observed at 1086, 950, and 809 cm-1, indicating the stretching vibrations of Si-O. The stretching vibration peaks of C-H were found at 2861 and 2915cm-1, which came from the - CH2 and -CH3 in CTAB. The FT-IR spectra of PAI show the presence of the imide carbonyl band at 1776 and 1715 cm-1, and the peaks at 3320 and 1715 cm-1 come from the N-H stretching and C=O region, respectively. Other characteristic bands include the absorption at 1600 cm-1 caused by the benzene ring stretch and a contribution from the O-H bond in monomeric H2O, which also has a band at 1663 cm-1. The bands at 950–650 and 3320 cm-1 increased with the increase in silica content. Furthermore, the adsorption peak at 1086 cm-1 was intensified in the spectra of the composites. All the characteristic peaks in the composites were insensitive to the presence of the silica component, indicating good interaction between the spherical silica and the PAI polymer matrix.

particle discharge resistance is improved with the increase in silica content. When the amount of silica added to the PAI was increased to 25 wt%, as shown in Fig. 9 (d), the fracture surfaces of the composite films remained well-integrated without the obvious phase separation. The results indicate that such a simple method effectively increases the amount

of inorganic silica particles in the PAI matrix after the surface modification process.

Fig. 8. SEM micrographs of (a) spherical silica submicron particles and fracture surfaces of (b) pure PAI film, (c) composite 1 with 2 wt% submicron silica, (d) composite 2 with 4 wt% submicron silica, (e) composite 3 with 6 wt% submicron silica, and (f) composite 4 with

The FT-IR spectra of CTAB-modified silica, pure PAI, and some submicron silica/PAI composite films are shown in Fig. 10. As shown in Fig. 3 (a), typical bands of silica were observed at 1086, 950, and 809 cm-1, indicating the stretching vibrations of Si-O. The stretching vibration peaks of C-H were found at 2861 and 2915cm-1, which came from the - CH2 and -CH3 in CTAB. The FT-IR spectra of PAI show the presence of the imide carbonyl band at 1776 and 1715 cm-1, and the peaks at 3320 and 1715 cm-1 come from the N-H stretching and C=O region, respectively. Other characteristic bands include the absorption at 1600 cm-1 caused by the benzene ring stretch and a contribution from the O-H bond in monomeric H2O, which also has a band at 1663 cm-1. The bands at 950–650 and 3320 cm-1 increased with the increase in silica content. Furthermore, the adsorption peak at 1086 cm-1 was intensified in the spectra of the composites. All the characteristic peaks in the composites were insensitive to the presence of the silica component, indicating good

interaction between the spherical silica and the PAI polymer matrix.

8 wt% submicron silica.

Fig. 9. SEM micrographs of (a) composite 5 with 10 wt% submicron silica, (b) composite 6 with 15 wt% submicron silica, (c) composite 7 with 20 wt% submicron silica, and (d) composite 8 with 25 wt% submicron silica.

Fig. 10. FT-IR spectra of (a) CTAB-modified submicron silica particles, (b) pure PAI, (c) composite 1 with 2 wt% silica, (d) composite 2 with 4 wt% silica, and (e) composite 4 with 8 wt% silica.

The TGA plots of the PAI and the composites with the different amounts of submicron silica particles are shown in Fig. 11. Obvious weight loss is found in these plots. A weight loss is observed below 180 °C in all the TGA plots, corresponding to water and solvent losses. When the temperature was increased to 450 °C, the PAI matrix began to decompose with increasing decomposition temperature. The decomposition temperature increased from

Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire 571

Scheme 3. Interactions between the silica submicron particles and the PAI polymer matrix.

4 to 25 wt%.

Second, the thermal stability of the composite film was further improved by increasing the silica particles. However, the decomposition temperature was increased from 650 to 668 °C, the results of which did not exceed that of the composites shown in Fig. 11. These results indicate that the interaction between the silica and the PAI matrix changed when more silica submicron particles were added into the composite films. Initially, the thermal stability of PAI improved after the silica submicron particles were added into the PAI matrix. Given the thermal motion of molecules, the cohesion between the adjacent polymer chains and the partial resistance and friction from the incorporation of silica in the PAI matrix had to be overcome. When more silica submicron particles were added into the PAI polymer, the interaction between the silica and the PAI matrix weakened because of the more compact and orderly arrangement of silica in the PAI matrix, as shown in Scheme 3. In addition, the silica submicron particles were easier to break away from the fracture surfaces, as confirmed by the SEM micrographs shown in Fig. 9. As a result, the decomposition temperature of the composite film did not increase with increase in the amount of silica particles, when the amount of silica added to PAI was more than 10 wt%. However, the thermal stability and decomposition temperature of the composite film were obviously more prominent than those of the pure PAI polymer matrix. Considering the similar thermal stability and lower price of the enameled wire, the amount of silica submicron particles can be modulated from

Fig. 13. CTE curves of the (a) pure PAI film, (b) composite 2 with 4 wt% silica, (c) composite

3 with 6 wt% silica, and (d) composite 6 with 15 wt% silica.

640 to 678 °C after the silica particles were added to the PAI polymer matrix. The thermal stability of the PAI was increased by the incorporation of the submicron silica particles, as clearly shown in Fig. 11. The amount of silica particles in the composites were accurately calculated from the TGA plots. The silica content in the composites calculated based on the plots in Figs.11 (b)–(e) are 1.68, 3.80, 5.80, and 7.24 wt%. The calculated data are close to the theoretical values listed in Table 1.

Inorganic silica particles improve the thermal stabilities of the PAI polymer matrix and reduce the cost of enameled wire. To obtain the least expensive and most stable composite films, the amount of silica particles was increased from 10 to 25 wt%. In Fig. 12, the TGA curves are similar to the composite films shown in Fig. 11. The amount of silica particles was calculated based on the plots. The realistic data were 9.3, 13.7, 16.5, and 20.4 wt%.

Fig. 11. TGA plots of (a) pure PAI, (b) composite 1 with 2 wt% silica, (c) composite 2 with 4 wt% silica, (d) composite 3 with 6 wt% silica, and (e) composite 4 with 8 wt% silica.

Fig. 12. TGA plots of (a) composite 5 with 10 wt% silica, (b) composite 6 with 15 wt% silica, (c) composite 7 with 20 wt% silica, and (d) composite 8 with 25 wt% silica.

640 to 678 °C after the silica particles were added to the PAI polymer matrix. The thermal stability of the PAI was increased by the incorporation of the submicron silica particles, as clearly shown in Fig. 11. The amount of silica particles in the composites were accurately calculated from the TGA plots. The silica content in the composites calculated based on the plots in Figs.11 (b)–(e) are 1.68, 3.80, 5.80, and 7.24 wt%. The calculated data are close to the

Inorganic silica particles improve the thermal stabilities of the PAI polymer matrix and reduce the cost of enameled wire. To obtain the least expensive and most stable composite films, the amount of silica particles was increased from 10 to 25 wt%. In Fig. 12, the TGA curves are similar to the composite films shown in Fig. 11. The amount of silica particles was

calculated based on the plots. The realistic data were 9.3, 13.7, 16.5, and 20.4 wt%.

Fig. 11. TGA plots of (a) pure PAI, (b) composite 1 with 2 wt% silica, (c) composite 2 with 4 wt% silica, (d) composite 3 with 6 wt% silica, and (e) composite 4 with 8 wt% silica.

Fig. 12. TGA plots of (a) composite 5 with 10 wt% silica, (b) composite 6 with 15 wt% silica,

(c) composite 7 with 20 wt% silica, and (d) composite 8 with 25 wt% silica.

theoretical values listed in Table 1.

Scheme 3. Interactions between the silica submicron particles and the PAI polymer matrix.

Second, the thermal stability of the composite film was further improved by increasing the silica particles. However, the decomposition temperature was increased from 650 to 668 °C, the results of which did not exceed that of the composites shown in Fig. 11. These results indicate that the interaction between the silica and the PAI matrix changed when more silica submicron particles were added into the composite films. Initially, the thermal stability of PAI improved after the silica submicron particles were added into the PAI matrix. Given the thermal motion of molecules, the cohesion between the adjacent polymer chains and the partial resistance and friction from the incorporation of silica in the PAI matrix had to be overcome. When more silica submicron particles were added into the PAI polymer, the interaction between the silica and the PAI matrix weakened because of the more compact and orderly arrangement of silica in the PAI matrix, as shown in Scheme 3. In addition, the silica submicron particles were easier to break away from the fracture surfaces, as confirmed by the SEM micrographs shown in Fig. 9. As a result, the decomposition temperature of the composite film did not increase with increase in the amount of silica particles, when the amount of silica added to PAI was more than 10 wt%. However, the thermal stability and decomposition temperature of the composite film were obviously more prominent than those of the pure PAI polymer matrix. Considering the similar thermal stability and lower price of the enameled wire, the amount of silica submicron particles can be modulated from 4 to 25 wt%.

Fig. 13. CTE curves of the (a) pure PAI film, (b) composite 2 with 4 wt% silica, (c) composite 3 with 6 wt% silica, and (d) composite 6 with 15 wt% silica.

Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire 573

into the PAI matrix. The CTE value further decreased with the increase in the amount of silica nanoparticles. When the submicron silica particles were added into the PAI polymer matrix, a similar conclusion was reached. However, more submicron silica particles are well-dispersed in the PAI polymer matrix after the surface modification process because the submicron silica particles have better dispersal state. Considering the higher thermal stability, lower CTE value, and the lower cost, silica/PAI composite films can be widely

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) Funded by the Ministry of Education, Science and Technology. (No. 2011-0016699). Also, it was supported by Sanhak Fellowship program

Ahn, B. Y.; Seok, S. I.; Hong, S.; Oh, J.; Jung, H.; Chung, W. J. (2006). Optical properties of

Alexandre, M.; Dubois, P. (2000). Polymer-layered silicate nanocomposites: preparation,

Babooram, K.; Francis, B.; Bissessur, R.; Narain R. (2008). Synthesis and characterization of

Bhagat, S. D.; Kim, Y.; Suh, K.; Ahn, Y.; Yeo, J.; Han, J. (2008). Superhydrophobic silica

Butterworth, M. D.; Corradi, R.; Johal, J.; Lascelles, S. F.; Maeda, S.; Armes, S. P. (1995). Zeta

Castellano, M.; Conzatti, L.; Costa, G.; Falqui, L.; Turturro, A.; Valenti, B.; Negroni, F. (2005).

Chen, L. W.; Ho, K. S. (1997). Synthesis of Polyamide-imide by Blocked-Methylene

David, I. A.; Scherer, G. W. (1995). An Organic/Inorganic Single-Phase Composite.

Fuchigami, K.; Taguchi, Y.; Tanaka, M. (2008) Synthesis of spherical silica particles by sol-gel method and application. *Polymer for Advanced Technologies*, 19, pp. 977–983

Particles. *Journal of Colloid and Interface Science*, 174, 2, pp. 510-517

nanocrystals. *Optical Materials*, 28, 4, pp. 374-379

organic/inorganic nanocomposite sol-gel films containing LaPO4:Er,Yb

properties and uses of a new class of materials. *Materials Science and Engineering*, 28,

novel (amide–imide)-silica composites by the sol–gel process. *Composites Science and* 

aerogel powders with simultaneous surface modification, solvent exchange and sodium ion removal from hydrogels. *Microporous and Mesoporous Materials*, 112, 1-3,

Potential Measurements on Conducting Polymer-Inorganic Oxide Nanocomposite

Surface modification of silica: 1. Thermodynamic aspects and effect on elastomer

Diisocyanates. *Journal of Polymer Science Part A: Polymer Chemistry*, 35, 9, pp. 1711-

used in the enamel wire industry.

funded of Korea Sanhak foundation.

1-2, pp. 1-63

pp. 504-509

1717

*Technology*, 68, 3-4, pp. 617–624

reinforcement. *Polymer*, 46, 3, pp. 695-703

*Chemistry of Materials*, 7, pp. 1957-1967

**4. Acknowledgement** 

**5. References** 

Controlling the CTE value of the dielectrical PAI is important because copper is the typical choice for defining the circuit lines. The selection of composites with the CTE close to that of copper is very critical, and can prevent damage to the copper interconnection; thus improving the reliability of the integrated circuits.

The CTE curves of the pure PAI film and some composite films are shown in Fig. 13. The CTE value of the pure PAI film is 3.87 × 10-5m/m/°C, whereas that of the composite films in Figs. 13 (b)–(d) are 3.76 ×10-5, 3.57 × 10-5, and 3.25 × 10-5m/m/°C, respectively. The CTE values continuously decreased with the increasing amount of the silica particles. The CTE value decreased by 16% when the amount of silica added to the PAI was 15 wt%. Such a variation in CTE is explained by the rigidity and stiffness of the silica submicron particles, which limit the movement of the polymer chain, resulting in the decrease of the thermal expansion of the PAI matrix. When the temperature increased to 220 °C, it approached the glass transition temperature of PAI. The dimension change significantly decreased with increased silica content. Therefore, the silica submicron particles effectively decrease the CTE value of the PAI polymer matrix, which in turn increases the thermal stress build-up, resulting in device failure through peeling and cracking at the interface between the PAI polymer film and the copper.

The thermal stability and the CTE value of the spherical silica/PAI composite films significantly improved when the submicron silica particles were added into the PAI polymer matrix. The submicron silica particles were obtained through the sol-gel method. Most previous studies used a sol-gel process because of the diameter of the silica particles. In the present study, the diameter effect was easily controlled using the Stőber procedure. Compared with the silica nanoparticles, the submicron silica particles have better dispersal state and do not easily agglomerate. Thus, more submicron silica particles are welldispersed in the PAI polymer matrix after the surface modification process. Considering the higher thermal stability and lower CTE value, especially the lower cost, submicron silica/PAI composite films can be widely used in the enamel wire industry.

## **3. Conclusion**

In the current work, silica particles with two different diameters were successfully added into the PAI polymer matrix in the synthesis of silica/PAI composite films via simple ultrasonic blending. First, the effect of CTAB on the synthesis of the silica/PAI composite film was investigated. The optimal dosage of CTAB is 3 wt%. The fracture surface micrographs of the composites show that the silica nanoparticles are well-dispersed in the PAI matrix after the surface modification process. In the TGA plots of the silica/PAI nanocomposites, the thermal stability and the decomposition temperature obviously increased with increasing CTAB dosage. Therefore, CTAB improves not only the dispersal state of silica in the PAI matrix, but also the thermal stability of the composite film because of better interaction between the nanosilica and the PAI matrix.

When the amount of silica nanoparticles added to the PAI was increased from 2 to 10 wt%, the thermal stability of the silica/PAI nanocomposites improved. The decomposition temperature increased with the increase in the amount of silica nanoparticles. In particular, the CTE value decreased when the silica particles were added into the PAI matrix. The CTE value further decreased with the increase in the amount of silica nanoparticles. When the submicron silica particles were added into the PAI polymer matrix, a similar conclusion was reached. However, more submicron silica particles are well-dispersed in the PAI polymer matrix after the surface modification process because the submicron silica particles have better dispersal state. Considering the higher thermal stability, lower CTE value, and the lower cost, silica/PAI composite films can be widely used in the enamel wire industry.

## **4. Acknowledgement**

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) Funded by the Ministry of Education, Science and Technology. (No. 2011-0016699). Also, it was supported by Sanhak Fellowship program funded of Korea Sanhak foundation.

## **5. References**

572 Scanning Electron Microscopy

Controlling the CTE value of the dielectrical PAI is important because copper is the typical choice for defining the circuit lines. The selection of composites with the CTE close to that of copper is very critical, and can prevent damage to the copper interconnection; thus

The CTE curves of the pure PAI film and some composite films are shown in Fig. 13. The CTE value of the pure PAI film is 3.87 × 10-5m/m/°C, whereas that of the composite films in Figs. 13 (b)–(d) are 3.76 ×10-5, 3.57 × 10-5, and 3.25 × 10-5m/m/°C, respectively. The CTE values continuously decreased with the increasing amount of the silica particles. The CTE value decreased by 16% when the amount of silica added to the PAI was 15 wt%. Such a variation in CTE is explained by the rigidity and stiffness of the silica submicron particles, which limit the movement of the polymer chain, resulting in the decrease of the thermal expansion of the PAI matrix. When the temperature increased to 220 °C, it approached the glass transition temperature of PAI. The dimension change significantly decreased with increased silica content. Therefore, the silica submicron particles effectively decrease the CTE value of the PAI polymer matrix, which in turn increases the thermal stress build-up, resulting in device failure through peeling and cracking at the interface between the PAI

The thermal stability and the CTE value of the spherical silica/PAI composite films significantly improved when the submicron silica particles were added into the PAI polymer matrix. The submicron silica particles were obtained through the sol-gel method. Most previous studies used a sol-gel process because of the diameter of the silica particles. In the present study, the diameter effect was easily controlled using the Stőber procedure. Compared with the silica nanoparticles, the submicron silica particles have better dispersal state and do not easily agglomerate. Thus, more submicron silica particles are welldispersed in the PAI polymer matrix after the surface modification process. Considering the higher thermal stability and lower CTE value, especially the lower cost, submicron

In the current work, silica particles with two different diameters were successfully added into the PAI polymer matrix in the synthesis of silica/PAI composite films via simple ultrasonic blending. First, the effect of CTAB on the synthesis of the silica/PAI composite film was investigated. The optimal dosage of CTAB is 3 wt%. The fracture surface micrographs of the composites show that the silica nanoparticles are well-dispersed in the PAI matrix after the surface modification process. In the TGA plots of the silica/PAI nanocomposites, the thermal stability and the decomposition temperature obviously increased with increasing CTAB dosage. Therefore, CTAB improves not only the dispersal state of silica in the PAI matrix, but also the thermal stability of the composite film because

When the amount of silica nanoparticles added to the PAI was increased from 2 to 10 wt%, the thermal stability of the silica/PAI nanocomposites improved. The decomposition temperature increased with the increase in the amount of silica nanoparticles. In particular, the CTE value decreased when the silica particles were added

silica/PAI composite films can be widely used in the enamel wire industry.

of better interaction between the nanosilica and the PAI matrix.

improving the reliability of the integrated circuits.

polymer film and the copper.

**3. Conclusion** 


Synthesis and Characterisation of Silica/Polyamide-Imide Composite Film for Enamel Wire 575

Oh, C.; Lee, Y. G.; Jon, C. U.; Oh, S. G. (2009). Synthesis and characterization of hollow

Ohki, Y. (2005). Study on dielectric properties of LDPE-based nanocomposites by J-power

Ouabbas, Y.; Chamayou, A.; Galet, L.; Baron, M.; Thomas, G.; Grosseau, G.; Guilhot,

Ranade, A.; Souza, N.; Gnade, B. (2002). Exfoliated and intercalated polyamide-imide nanocomposites with montmorllonite. *Polymer*, 43, 13, pp. 3759-3766 Rangsunvigit, P.; Imsawatgul, P.; Na-ranong, N.; O'Haver, J. H.; Chavadej, S. (2008)

Rankin, S. E.; Macosko, C. W.; McCormick, A. V. (1998). Sol-Gel Polycondensation Kinetic

Rupnowski, P.; Gentz, M.; Kumosa, M. (2006). Mechanical response of a unidirectional

Stathatos, E.; Lianos, P.; Tsakiroglou, C. (2004). Highly efficient nanocrystalline titania films

Sun, S.; Li, C.; Zhang L.; Du, H.L.; Burnell-Gray J.S. (2006). Effects of surface modification of

Wu, J.; Yang, S.; Gao, S.; Hu, A.; Liu, J.; Fan, L. (2005). Preparation, morphology and

Wu, T.; Ke, Y. (2006). Preparation of silica–PS composite particles and their application in

Xu, J.; Wong, C.P. (2007). Characterization and properties of an organic–inorganic dielectric

Xue, L.; Li, J.; Fu, J.; Han, Y. (2009). Super-hydrophobicity of silica nanoparticles modified

Yang, Y.; Wang, P. (2006). Preparation and characterizations of a new PS/TiO2 hybrid

Yanagishita, H.; Kitamoto, D.; Haraya, K.; Nakane, T.; Okada, T.; Matsuda, H.; Idemoto,

chloride) composites. *European Polymer Journal*, 42, 7, pp. 1643-1652

Modeling: Methylethoxysilanes. *AIChE Journal*, 44, 5, pp. 1141-1156

*Science and Technology*, 66, 7-8, pp. 1045–1055

PET. *European Polymer Journal*, 42, 2, pp. 274-285

*Applied Science and Manufacturing*, 38, 1, pp. 13-19

membrane by sol–gel process. *Polymer*, 47, 8, pp. 2683–2688

*Materials*, 75, 3, pp. 255-260

41, 1, pp. 73-81

338, 1, pp. 15-19

systems. *IEEE Electrical Insulation Magazine*, 21, 3, pp. 55-56

and powder aging. *Powder Technology*, 190, 1-2, pp. 200-209

pp. 208-212

288-294

silica microspheres functionalized with magnetic particles using W/O emulsion method. *Colloids and Surfaces A: Physicochemical and Engineering Aspects*, 337, 1-3,

B. (2009). Surface modification of silica particles by dry coating: Characterization

Mixed surfactants for silica surface modification by admicellar polymerization using a continuous stirred tank reactor. *Chemical Engineering Journal*, 136, 2-3, pp.

graphite fiber /polyimide composite as a function of temperature. *Composites* 

made from organic/inorganic nanocomposite gels. *Microporous and Mesoporous* 

fumed silica on interfacial structures and mechanical properties of poly(vinyl

properties of nano-sized Al2O3/polyimide hybrid films. *European Polymer Journal*,

nanocomposite for embedded decoupling capacitor applications. *Composites Part A:* 

with vinyl groups. *Colloids and Surfaces A: Physicochemical and Engineering Aspects*,

Y.; Koura N. (2001). Separation performance of polyimide composite membrane prepared by dip coating process. *Journal of Membrane Science*, 188, 2, pp.165–172


Hossein, S. S.; Lia, Y.; Chunga, T.; Liu, Y. (2007). Enhanced gas separation performance of

Hwang, J.; Lee, B. I.; Klep, V.; Luzinov, I. (2008). Transparent hydrophobic organic– inorganic nanocomposite films. *Materials Research Bulletin*, 43, 10, pp. 2652-2657 Jadav, G. L.; Singh, P. S. (2009). Synthesis of novel silica-polyamide nanocomposite

Jiao, W. M.; Vidal, A.; Papirer, E.; Donnet, J.B. (1989). Modification of Silica Surfaces by

Kawakami, H.; Mikawa, M.; Nagaoka, S. (1996). Gas permeability and selectivity through

Kawakami, H.; Mikawa, M.; Nagaoka, S. (1998). Gas transport properties of asymmetric

Kawakami, H.; Nakajima, K.; Shimizu, H.; Nagaoka S. (2003). Gas permeation stability of

Kikuchi, H.; Yukimon, Y.; Itonaga, S. (2002). Inverter-surge-resistant enameled wire based on nano-composite insulating material. *Hitachi Cable Review*, 21, pp. 55-62 Kim, J. Y.; Mulmi, S.; Lee, C. H.; Park, H. B.; Chung, Y. S.; Lee, Y. M. (2006). Preparation of

Kusakabe, K.; Ichiki, K.; Hayashi, J.; Maeda, H.; Morooka, S. (1996). Preparation and

Lee, Y. L.; Du, Z. C.; Lin, W. X.; Yang, Y. M. (2006). Monolayer behavior of silica particles at

Ma, J.; Yang, Z.; Wang, X.; Qu, X.; Liu, J.; Lu, Y.; Hu, Z.; Yang, Z. (2007). Flexible bi-

Mosher, B. P.; Wu, C.; Sun, T.; Zeng, T. (2006). Particle-reinforced water-based organic-

separated structure. *Journal of Membrane Science*, 283, 1-2, pp. 172-181 Kim, T. K.; Kang, M.; Choi, Y. S .; Kim, H. K.; Lee, W.; Chang, H.; Seung, D. (2007).

for CO2 separation. *Journal of Membrane Science*, 115, 1, pp. 65-75

surface. *Journal of Colloid Interface Science*, 296, 1, pp. 233-241

structure. *Journal of Membrane Science*, 212, 1-2, pp. 195–203

302, 1-2, pp. 207–217

267

971

1-8

15, pp. 4305-4310

*Solids*, 352, 30-31, pp. 3295-3301

40, pp. 279-291

pp. 6636–6638

nanocomposite membranes using MgO nanoparticles. *Journal of Membrane Science*,

membrane with enhanced properties. *Journal of Membrane Science*, 328, 1-2, pp. 257-

Grafting of Alkyl Chains Part III. Particle/Particle Interactions: Rheology of Silica Suspensions in Low Molecular Weight Analogs of Elastomers. *Colloids and Surfaces*,

asymmetric polyimide membranes. *Journal of Appled Polymer Science*, 62, 7, pp. 965–

polyimide membrane with an ultrathin surface skin layer. *Macromolecules*, 31, 19,

asymmetric polyimide membrane with thin skin layer: effect of polyimide

organic–inorganic nanocomposite membrane using a reactive polymeric dispersant and compatibilizer: Proton and methanol transport with respect to nano-phase

Preparation of Nafion-sulfonated clay nanocomposite membrane for direct menthol fuel cells via a film coating process. *Journal of Power Sources*, 165, 1, pp.

characterization of silica-polyimide composite membranes coated on porous tubes

air/water interface:A comparison between chemical and physical modifications of

continuous mesostructured inorganic/polymer composite membranes. *Polymer*, 48,

inorgaic aocomposite coatings for tailored applications. *Journal of Non-Crystalline* 


**29** 

Areeya Aeimbhu

*Thailand* 

**Scanning Electron Microscope** 

**for Characterising of Micro- and** 

**Nanostructured Titanium Surfaces** 

*Department of Physics, Faculty of Science, Srinakharinwirot University, Bangkok,* 

Titanium and its alloys have been used broadly and successfully for numerous applications such as sport equipment [1], aerospace industry [2], marine application [3], medical applications [4] because of its optimum mechanical properties, outstanding corrosion resistance and bio-inert due to the presence of a thin layer of titanium oxide which a naturally formed onto the titanium surface [5-6]. This layer mainly consists of titanium dioxide or titania. Properties of oxide films covering titanium implant surfaces are a key role for a successful osseointegration [7-8], wear and corrosion resistance [9]. Moreover, titania has a large number of potentials in water photoelectrolysis and photocatalysis [10], sensors [11], wastewater remediation [12], automotive industry [13], industrial applications [14] and micro-optoelectronic applications [15]. However, there are some disadvantages of the native titanium oxide which has poor mechanical properties and easily fractured under small scale of fretting and wears conditions [16]. Therefore, many surface modification treatments for examples anodisation (anodic oxidation), cathodic electrodeposition and sol-gel reactions [17-22] have been studied in order to improve the performance of titanium. Moreover, surface properties such as topography, chemical composition and hydrophilicity have an effect on the mechanical stability of the implant-tissue interface. Various surface modification methods of titanium have been shown to improve interfacial interactions at the bone-implant interface and their clinical performance. Moreover, the biological performance of implantable titanium depends crucially on their surface topography in the micrometre (structures larger than 1 micron) and nanometre (structures smaller than 1 micron) range. Surface micro- and nano-topography can reduced inflammatory and guide direct osteoblast responses by altering adhesion, recruitment, movement, morphology, apoptosis and gene expression, and subsequently protein production [23-25]. An anodisation is a simple and an inexpensive technique to prepare thin film titania on titanium surface in different conditions and electrolytes such as acidic, basic, neutral, organic and inorganic which affect surface architecture and chemical composition [26-27]. Anodised oxide layer has thickness in the range of 20 to 180 nm which is thicker than a naturally formed oxide [28]. Moreover, this technique is presently used to achieve micro- and nano-topography surfaces. An anodisation is an electrolytic etching for coating the surface of a metal with an oxide layer which changes the microscopic texture of the surface and the crystal structure of the metal

**1. Introduction** 


## **Scanning Electron Microscope for Characterising of Micro- and Nanostructured Titanium Surfaces**

Areeya Aeimbhu *Department of Physics, Faculty of Science, Srinakharinwirot University, Bangkok, Thailand* 

#### **1. Introduction**

576 Scanning Electron Microscopy

Zhang, C.; Zhang, M.; Cao, H.; Zhang, Z.; Wang, Z.; Gao, L.; Ding, M. (2007). Synthesis and

Zheng, P.; Kong L. X.; Li, S.; Yin, C.; Huang M. F. (2007). Self-assembled natural

Zhong, S.; Li, C .; Xiao, X. (2002). Preparation and characterization of polyimide–silica

*and Technology*, 67, 3-4, pp. 380-389

199, 1-3, pp. 53–58

*Science and Technology*, 67, 15-16, pp. 3130-3139

properties of a novel isomeric polyimide/SiO2 hybrid material. *Composites Science* 

rubber/silica nanocomposites: Its preparation and characterization. *Composites* 

hybrid membranes on kieselguhr–mullite supports. *Journal of Membrane Science*,

Titanium and its alloys have been used broadly and successfully for numerous applications such as sport equipment [1], aerospace industry [2], marine application [3], medical applications [4] because of its optimum mechanical properties, outstanding corrosion resistance and bio-inert due to the presence of a thin layer of titanium oxide which a naturally formed onto the titanium surface [5-6]. This layer mainly consists of titanium dioxide or titania. Properties of oxide films covering titanium implant surfaces are a key role for a successful osseointegration [7-8], wear and corrosion resistance [9]. Moreover, titania has a large number of potentials in water photoelectrolysis and photocatalysis [10], sensors [11], wastewater remediation [12], automotive industry [13], industrial applications [14] and micro-optoelectronic applications [15]. However, there are some disadvantages of the native titanium oxide which has poor mechanical properties and easily fractured under small scale of fretting and wears conditions [16]. Therefore, many surface modification treatments for examples anodisation (anodic oxidation), cathodic electrodeposition and sol-gel reactions [17-22] have been studied in order to improve the performance of titanium. Moreover, surface properties such as topography, chemical composition and hydrophilicity have an effect on the mechanical stability of the implant-tissue interface. Various surface modification methods of titanium have been shown to improve interfacial interactions at the bone-implant interface and their clinical performance. Moreover, the biological performance of implantable titanium depends crucially on their surface topography in the micrometre (structures larger than 1 micron) and nanometre (structures smaller than 1 micron) range. Surface micro- and nano-topography can reduced inflammatory and guide direct osteoblast responses by altering adhesion, recruitment, movement, morphology, apoptosis and gene expression, and subsequently protein production [23-25]. An anodisation is a simple and an inexpensive technique to prepare thin film titania on titanium surface in different conditions and electrolytes such as acidic, basic, neutral, organic and inorganic which affect surface architecture and chemical composition [26-27]. Anodised oxide layer has thickness in the range of 20 to 180 nm which is thicker than a naturally formed oxide [28]. Moreover, this technique is presently used to achieve micro- and nano-topography surfaces. An anodisation is an electrolytic etching for coating the surface of a metal with an oxide layer which changes the microscopic texture of the surface and the crystal structure of the metal

Scanning Electron Microscope for Characterising of Micro- and Nanostructured Titanium Surfaces 579

Series 2 1 M - 30, 60, 120, 240, 360

Series 3 1 M 0.075 30, 60, 360, 480, 720,

Series 4 1 M 0.5, 1.5 and 2.0 60

In order to find out the effect of anodisation electrolyte and anodisation time on the forming of the oxide at constant voltage in an ambient temperature; the structure of morphology were investigated. Experimental with different anodisation conditions: the concentration of acetic acid, the concentration of hydrofluoric acid and an anodisation time were lead to get

Figure 1 displayed scanning electron microscope (SEM) photographs of an untreated and the anodised titanium surface at 20 volts in different electrolytes concentration. An untreated sample has no visible scratches or wrinkles (figure 1a). The effect of concentration of electrolytes on the surface morphology of the anodic oxide film formed on titanium was monitored. Based on the experimental observation, it was showed that the passive film formed on titanium surface in low concentrations of acetic acid range which from 0.001 to 0.1 M (figure 1(b-d)). The flower-like structures [29] were visible on the substrates which anodised in 1 M acetic acid. Moreover, upon increase concentration of acetic acid the surface morphologies shown the concentration of the flower was developed and increased as shown in figure 1e - 1f. The average diameter of the flowers formed in an acetic acid for 120

The effect of anodisation time on the morphology of an oxide film in 1M acetic acid was investigated. Figure 2(b-g) summarised the effect of anodisation time on the development of an anodised film in 1M acetic acid at 20 volts. Within 30 minutes (figure 2b), SEM image revealed the formation of a continuous of oxide film over the surface. Flower-like structures were developed at 60 minutes (figure 2c). The density of the flower oxides was observed with the increasing of anodisation time. The average diameter of the flowers increases from

The effect of hydrofluoric acid and anodisation time on the formation of titania nanotubes in 1M acetic acid with 0.075 wt% HF is studied. SEM image of the surface morphologies obtained in 1 M acetic acid with 0.075 %wt HF at different anodisation time are shown in figure 3(b-i).

Hydrofluoric acid (wt%)

10M - 120

Anidising time (minutes)

and 480

1440, 2160 and 2880

Acetic concentration (M)

an overall view of the formation process of titania nanotubes.

**3.2 Effect of anodisation time on the surface topography** 

**3.1 Effect of concentration of acetic acid on the surface topography** 

Series 1 0.001, 0.01, 0.1, 1 and

Table 1. Parameters used for anodisation

minutes was around 300 nanometres.

a few hundred nanometres to 1200 nanometres.

**3. Result and discussion** 

near the surface. An anodisation process accelerates the formation of an oxide coating under controlled conditions to provide the desired result. Since the coating is biocompatible as well as nontoxic, the process lends itself to achieve drastic improvement in implant performance. By adjusting the anodisation condition such as electrolyte, pH, voltage and time, micro- and nano-scale properties could be controlled.

In this article, the morphological of surface was studied by means of scanning electron microscope (SEM) after surface modification in order to evaluate qualitatively the effect of the anodisation conditions. For this purpose, the SEM uses to examine morphological of development of anodised film of titanium that were prepared under different controlled conditions.

## **2. Materials and methods**

#### **2.1 Preparation of surface**

Prior to anodisation, commercially pure titanium grade 2 were provided by Prolog Titanium Co., Ltd and cut into 1 cm × 1 cm squares. Titanium sheets were abraded mechanically using silicon carbide (SiC) abrasive papers (Buehler) number 120 to 2000 and rinsed with distilled water. The surface of sheets were sequentially polished to a mirror finished with aqueous alumina (Al2O3) 5, 1, 0.3 and 0.05 micron (Buehler, Alpha Micropolish II). Afterward clean with acetone in ultrasonic bath.

#### **2.2 Remove oxide**

Titanium surfaces were immersed in an acid mixture (2 ml 48% HF (48% in water) which purchased from Panreac + 3 ml 70% HNO3 (70% in water) which purchased from Fluka + 100 ml DI water) about 90 seconds to remove the naturally formed oxide layer and then immediately treated by anodisation. All electrolytes were prepared with reagent grade chemicals.

#### **2.3 Anodisation**

The electrochemical cell consists of a two electrodes system which graphite acting as the counter electrode (cathode) and titanium sheet acting as the working electrode (anode). The anode and cathode were connected by copper wires and were linked to a positive and negative port of a DC power supply (KMB 3002), respectively. During processing, the anode and cathode were kept parallel with a separation distance of about 1 cm, and were immersed into an electrolyte. Anodisation was performed under potentiostatic conditions at 20 volts. All experiments were carried out at room temperature. Test conditions applied vary in terms of the concentration of acetic acid, the concentration of hydrofluoric acid and an anodisation time. Different series of experiments have been performed by changing the associated experimental parameters as shown in Table 1. After the electrochemical process, the titanium sheets were immediately rinsed by deionised water and heated at 120°C for 30 minutes.

#### **2.4 Surface characterisation**

The surface morphology of the samples was observed by using Scanning electron microscope (SEM: JEOL 6300) with Dispersive X-Ray Spectroscopy (EDS) technique.



Table 1. Parameters used for anodisation

## **3. Result and discussion**

578 Scanning Electron Microscopy

near the surface. An anodisation process accelerates the formation of an oxide coating under controlled conditions to provide the desired result. Since the coating is biocompatible as well as nontoxic, the process lends itself to achieve drastic improvement in implant performance. By adjusting the anodisation condition such as electrolyte, pH, voltage and

In this article, the morphological of surface was studied by means of scanning electron microscope (SEM) after surface modification in order to evaluate qualitatively the effect of the anodisation conditions. For this purpose, the SEM uses to examine morphological of development of anodised film of titanium that were prepared under different controlled

Prior to anodisation, commercially pure titanium grade 2 were provided by Prolog Titanium Co., Ltd and cut into 1 cm × 1 cm squares. Titanium sheets were abraded mechanically using silicon carbide (SiC) abrasive papers (Buehler) number 120 to 2000 and rinsed with distilled water. The surface of sheets were sequentially polished to a mirror finished with aqueous alumina (Al2O3) 5, 1, 0.3 and 0.05 micron (Buehler, Alpha Micropolish II). Afterward clean

Titanium surfaces were immersed in an acid mixture (2 ml 48% HF (48% in water) which purchased from Panreac + 3 ml 70% HNO3 (70% in water) which purchased from Fluka + 100 ml DI water) about 90 seconds to remove the naturally formed oxide layer and then immediately treated by anodisation. All electrolytes were prepared with reagent grade

The electrochemical cell consists of a two electrodes system which graphite acting as the counter electrode (cathode) and titanium sheet acting as the working electrode (anode). The anode and cathode were connected by copper wires and were linked to a positive and negative port of a DC power supply (KMB 3002), respectively. During processing, the anode and cathode were kept parallel with a separation distance of about 1 cm, and were immersed into an electrolyte. Anodisation was performed under potentiostatic conditions at 20 volts. All experiments were carried out at room temperature. Test conditions applied vary in terms of the concentration of acetic acid, the concentration of hydrofluoric acid and an anodisation time. Different series of experiments have been performed by changing the associated experimental parameters as shown in Table 1. After the electrochemical process, the titanium sheets were immediately rinsed by deionised water and heated at 120°C for 30 minutes.

The surface morphology of the samples was observed by using Scanning electron

microscope (SEM: JEOL 6300) with Dispersive X-Ray Spectroscopy (EDS) technique.

time, micro- and nano-scale properties could be controlled.

conditions.

**2. Materials and methods 2.1 Preparation of surface** 

with acetone in ultrasonic bath.

**2.4 Surface characterisation** 

**2.2 Remove oxide** 

chemicals.

**2.3 Anodisation** 

In order to find out the effect of anodisation electrolyte and anodisation time on the forming of the oxide at constant voltage in an ambient temperature; the structure of morphology were investigated. Experimental with different anodisation conditions: the concentration of acetic acid, the concentration of hydrofluoric acid and an anodisation time were lead to get an overall view of the formation process of titania nanotubes.

#### **3.1 Effect of concentration of acetic acid on the surface topography**

Figure 1 displayed scanning electron microscope (SEM) photographs of an untreated and the anodised titanium surface at 20 volts in different electrolytes concentration. An untreated sample has no visible scratches or wrinkles (figure 1a). The effect of concentration of electrolytes on the surface morphology of the anodic oxide film formed on titanium was monitored. Based on the experimental observation, it was showed that the passive film formed on titanium surface in low concentrations of acetic acid range which from 0.001 to 0.1 M (figure 1(b-d)). The flower-like structures [29] were visible on the substrates which anodised in 1 M acetic acid. Moreover, upon increase concentration of acetic acid the surface morphologies shown the concentration of the flower was developed and increased as shown in figure 1e - 1f. The average diameter of the flowers formed in an acetic acid for 120 minutes was around 300 nanometres.

#### **3.2 Effect of anodisation time on the surface topography**

The effect of anodisation time on the morphology of an oxide film in 1M acetic acid was investigated. Figure 2(b-g) summarised the effect of anodisation time on the development of an anodised film in 1M acetic acid at 20 volts. Within 30 minutes (figure 2b), SEM image revealed the formation of a continuous of oxide film over the surface. Flower-like structures were developed at 60 minutes (figure 2c). The density of the flower oxides was observed with the increasing of anodisation time. The average diameter of the flowers increases from a few hundred nanometres to 1200 nanometres.

The effect of hydrofluoric acid and anodisation time on the formation of titania nanotubes in 1M acetic acid with 0.075 wt% HF is studied. SEM image of the surface morphologies obtained in 1 M acetic acid with 0.075 %wt HF at different anodisation time are shown in figure 3(b-i).

Scanning Electron Microscope for Characterising of Micro- and Nanostructured Titanium Surfaces 581

(a) (b)

(c) (d)

(e) (f)

(g)

Fig. 2. SEM images of titanium sheets (a) untreated surface and anodised in 1M acetic acid at 20 volts for (b) 30 minutes (c) 60 minutes (d) 120 minutes (e) 240 minutes (f) 360 minutes and

(g) 480 minutes at room temperature.

The results demonstrated at the initial stage of anodisation (Figure 3(b-c)), the tube-like features appeared over the titanium surface. For longer anodisation time, it can be seen from figure 3(e-f) that the architecture transformation from tube-like nanostructure to a sponge-like titania film with nanoholes. Further anodisation time, the surface of nanoholes is covered with the oxide film (figure 3(g-h)). When the anodisation time is prolonged to 2880 minutes, the discontinuous oxide film covered the titanium surface. The results confirm that fluorine containing electrolyte is a capable electrolyte for anodic formation of titania nanotubes.

Fig. 1. SEM images of titanium sheets (a) untreated surface and anodised in (b) 0.001 (c) 0.01 (d) 0.1 (e) 1 and (f) 10M acetic acid at 20 volts for 120 minutes at room temperature.

The results demonstrated at the initial stage of anodisation (Figure 3(b-c)), the tube-like features appeared over the titanium surface. For longer anodisation time, it can be seen from figure 3(e-f) that the architecture transformation from tube-like nanostructure to a sponge-like titania film with nanoholes. Further anodisation time, the surface of nanoholes is covered with the oxide film (figure 3(g-h)). When the anodisation time is prolonged to 2880 minutes, the discontinuous oxide film covered the titanium surface. The results confirm that fluorine

containing electrolyte is a capable electrolyte for anodic formation of titania nanotubes.

(a) (b)

(c) (d)

(e) (f)

Fig. 1. SEM images of titanium sheets (a) untreated surface and anodised in (b) 0.001 (c) 0.01

(d) 0.1 (e) 1 and (f) 10M acetic acid at 20 volts for 120 minutes at room temperature.

Fig. 2. SEM images of titanium sheets (a) untreated surface and anodised in 1M acetic acid at 20 volts for (b) 30 minutes (c) 60 minutes (d) 120 minutes (e) 240 minutes (f) 360 minutes and (g) 480 minutes at room temperature.

Scanning Electron Microscope for Characterising of Micro- and Nanostructured Titanium Surfaces 583

(i)

Fig. 3. SEM images of titanium sheets (a) untreated surface and anodised in 1M acetic acid with 0.075 wt% HF at 20 volts for (b) 30 minutes (c) 60 minutes (d) 360 minutes (e) 480 minutes (f) 720 minutes (g) 1440 minutes (h) 2160 minutes and (i) 2880 minutes at room

The surface topologies of titanium sheets anodised in electrolyte containing 0.1 M acetic acid with different concentration of HF: 0.075, 0.5, 1.5 and 2.0 wt% are shown in Figure 4. The anodisation was carried out at 20 volts. A network structure appears on the anodised titanium surfaces with concentration of 0.075 wt% HF (Figure 4b). Anodisation in 0.5 wt% and 1.5 wt% HF containing occur a highly ordered and uniform titanium oxide nanotube arrays (Figure 4 (d-e)). The average nanotube inner diameter is approximately 74.5 and 76.5 nanometre, respectively. As the concentration of hydrofluoric concentration was further increased to 2.0 wt%, the surface architecture developed sponge-like (Figure 4f). It points out that the concentration of hydrofluoric acid affect the morphology of titanium

The anodic growth of compact oxides on titanium substrate and the formation of nanotubes in fluoride-containing electrolytes is the result of key processes [30] which are (1) Fieldassisted oxidation of the titanium metal that leads to oxide growth at the surface due to interaction of titanium with O2- or OH- ions. An initial oxide layer formed on the substrate, these anions travel through the oxide layer reaching the titanium/oxide interface. (2) Titanium metal ion (Ti4+) migrate from the substrate at the titanium/oxide interface; Ti4+ cations will ejected from the titanium/oxide interface under application of an electric field that move towards the oxide/electrolyte interface. (3) Field-assisted dissolution of titanium metal ion at the oxide/electrolyte interface into the electrolyte. (4) Chemical dissolution of titanium metal ion and TiO2 due to etching away by fluoride ions. The reactions are [31]:

1. oxidation of titanium metal which releases titanium metal ions (Ti4+) and electron

2. interaction of titanium metal ions with O2- or OH- ions

2Ti → 2Ti4+ + 8e- (1)

Ti4+ + 4OH- → Ti(OH)4 (2)

**3.3 Effect of concentration of hydrofluoric acid on the surface topography** 

temperature.

surface.

at anode:

Fig. 3. SEM images of titanium sheets (a) untreated surface and anodised in 1M acetic acid with 0.075 wt% HF at 20 volts for (b) 30 minutes (c) 60 minutes (d) 360 minutes (e) 480 minutes (f) 720 minutes (g) 1440 minutes (h) 2160 minutes and (i) 2880 minutes at room temperature.

#### **3.3 Effect of concentration of hydrofluoric acid on the surface topography**

The surface topologies of titanium sheets anodised in electrolyte containing 0.1 M acetic acid with different concentration of HF: 0.075, 0.5, 1.5 and 2.0 wt% are shown in Figure 4. The anodisation was carried out at 20 volts. A network structure appears on the anodised titanium surfaces with concentration of 0.075 wt% HF (Figure 4b). Anodisation in 0.5 wt% and 1.5 wt% HF containing occur a highly ordered and uniform titanium oxide nanotube arrays (Figure 4 (d-e)). The average nanotube inner diameter is approximately 74.5 and 76.5 nanometre, respectively. As the concentration of hydrofluoric concentration was further increased to 2.0 wt%, the surface architecture developed sponge-like (Figure 4f). It points out that the concentration of hydrofluoric acid affect the morphology of titanium surface.

The anodic growth of compact oxides on titanium substrate and the formation of nanotubes in fluoride-containing electrolytes is the result of key processes [30] which are (1) Fieldassisted oxidation of the titanium metal that leads to oxide growth at the surface due to interaction of titanium with O2- or OH- ions. An initial oxide layer formed on the substrate, these anions travel through the oxide layer reaching the titanium/oxide interface. (2) Titanium metal ion (Ti4+) migrate from the substrate at the titanium/oxide interface; Ti4+ cations will ejected from the titanium/oxide interface under application of an electric field that move towards the oxide/electrolyte interface. (3) Field-assisted dissolution of titanium metal ion at the oxide/electrolyte interface into the electrolyte. (4) Chemical dissolution of titanium metal ion and TiO2 due to etching away by fluoride ions. The reactions are [31]:

at anode:

582 Scanning Electron Microscopy

(a) (b)

(c) (d)

(e) (f)

(g) Underlying (h)

nanoholes

1. oxidation of titanium metal which releases titanium metal ions (Ti4+) and electron

$$2\text{Ti} \rightarrow 2\text{Ti} \\ 4\text{\*} + 8\text{e} \tag{1}$$

2. interaction of titanium metal ions with O2- or OH- ions

$$\text{Ti}^{4+} + \text{4OH} \cdot \text{I} \to \text{Ti}(\text{OH})\_4 \tag{2}$$

Scanning Electron Microscope for Characterising of Micro- and Nanostructured Titanium Surfaces 585

(a) (b)

(c) (d)

(e)

Fig. 4. SEM images of titanium sheets (a) untreated surface and anodised in 1 M acetic acid with (b) 0.075 wt% HF (c) 0.5 wt% HF (d) 1.5 wt% HF (e) 2.0 wt% HF for 60 minutes at room

temperature.

$$\text{Ti}^{4+} + 2\text{O}^{2-} \rightarrow \text{TiO}\_2 \tag{3}$$

Equation (2) and (3) elucidate the hydrated anodic layer and the oxide layer. Further oxide is produced when the hydrated anodic layer releases water by a condensation reaction as the following equation:

$$\text{Ti(OH)}\_{4} \rightarrow \text{TiO}\_{2} + 2\text{ H}\_{2}\text{O} \tag{4}$$

at cathode:

$$8\text{H}^\* + 8\text{e}^\cdot \to 4\text{H}\_2\tag{5}$$

The overall process for anodic oxidation of titanium can be represented as:

$$\text{Ti} + 2\text{H}\_2\text{O} \rightarrow \text{TiO}\_2 + 2\text{H}\_2\tag{6}$$

In the presence of fluoride ions electrolyte, fluoride ions enters the Ti(OH)4 or anodic titanium oxide as the following equation:

$$\rm TiO\_2 + 6F^\cdot + 4H^+ \to TiF\_6 + H\_2O \tag{7}$$

$$\text{Ti(OH)}\_{4} + \text{6F} \rightarrow \text{TiF}\_{6} + \text{4OH} \cdot \text{N} \tag{8}$$

$$\text{Ti}4^{\circ} + \text{6F} \rightarrow \text{TiF}\_{6} \tag{9}$$

Equation (7-8) is the mechanism of the pit formation due to the localised dissolution of the oxide and hydrated anodic layer. Then these pits transfer to bigger pores and the pore density increases subsequent to uniformly pores over the titanium surface, with the TiO2 pores growing more and more deeply into the titanium metal. As described above, the formation of nanotubes govern by a competition between anodic oxidation and chemical dissolution of the oxide as soluble fluoride complexes.

The concentration of the HF added to the electrolyte was varied from 0.075 to 2.0 wt%. As a result, a large change in surface architecture was observed. An average diameter of the nanotubes increased as the fluoride content in electrolyte. As mentioned above, the formation of titania nanotubes determined by the oxide growth rate and the dissolution rate. A result shows that with 0.5 and 1.5 wt% HF, the dissolution rate was slow and resulted in small pore size. With increasing the concentration of HF, the dissolution rate increased and resulted in big pore. It is obvious that the dissolution rate was extremely high with 2 wt% HF because the morphology is not uniform.

The result of the chemical analysis by Energy Dispersive X-ray spectroscopy (EDS) indicated that anodised titanium in 1 M acetic acid with 0.075 wt% HF for 60 minutes at room temperature titania film are shown in figure 5b. The EDS measurements present an oxygen and titanium proportion of 34.914% and 65.086 %, respectively. The result reviewed that the chemical composition of anodic film was nonstoichiometric and the atomic ration of Ti/O is approximately 1.86. The presented of nonstoichiometric structure on the substrate layer implied that some defect exist in TiO2 nanotubes due to oxygen deficiencies which can cause the formation of crystallographic shear planes and active Ti-sites for the adsorption and chemisorptions of OH groups or other contaminants [32].

 Ti4+ + 2O2- → TiO2 (3) Equation (2) and (3) elucidate the hydrated anodic layer and the oxide layer. Further oxide is produced when the hydrated anodic layer releases water by a condensation reaction as the

Ti(OH)4 → TiO2 + 2 H2O (4)

8H+ + 8e- → 4H2 (5)

In the presence of fluoride ions electrolyte, fluoride ions enters the Ti(OH)4 or anodic

TiO2 + 6F- + 4H+ → TiF6- + H2O (7)

Ti(OH)4 + 6F- → TiF6- + 4OH- (8)

 Ti4+ + 6F- → TiF6 (9) Equation (7-8) is the mechanism of the pit formation due to the localised dissolution of the oxide and hydrated anodic layer. Then these pits transfer to bigger pores and the pore density increases subsequent to uniformly pores over the titanium surface, with the TiO2 pores growing more and more deeply into the titanium metal. As described above, the formation of nanotubes govern by a competition between anodic oxidation and chemical

The concentration of the HF added to the electrolyte was varied from 0.075 to 2.0 wt%. As a result, a large change in surface architecture was observed. An average diameter of the nanotubes increased as the fluoride content in electrolyte. As mentioned above, the formation of titania nanotubes determined by the oxide growth rate and the dissolution rate. A result shows that with 0.5 and 1.5 wt% HF, the dissolution rate was slow and resulted in small pore size. With increasing the concentration of HF, the dissolution rate increased and resulted in big pore. It is obvious that the dissolution rate was extremely high with 2 wt%

The result of the chemical analysis by Energy Dispersive X-ray spectroscopy (EDS) indicated that anodised titanium in 1 M acetic acid with 0.075 wt% HF for 60 minutes at room temperature titania film are shown in figure 5b. The EDS measurements present an oxygen and titanium proportion of 34.914% and 65.086 %, respectively. The result reviewed that the chemical composition of anodic film was nonstoichiometric and the atomic ration of Ti/O is approximately 1.86. The presented of nonstoichiometric structure on the substrate layer implied that some defect exist in TiO2 nanotubes due to oxygen deficiencies which can cause the formation of crystallographic shear planes and active Ti-sites for the adsorption and

Ti + 2H2O → TiO2 + 2H2 (6)

The overall process for anodic oxidation of titanium can be represented as:

following equation:

titanium oxide as the following equation:

dissolution of the oxide as soluble fluoride complexes.

chemisorptions of OH groups or other contaminants [32].

HF because the morphology is not uniform.

at cathode:

Fig. 4. SEM images of titanium sheets (a) untreated surface and anodised in 1 M acetic acid with (b) 0.075 wt% HF (c) 0.5 wt% HF (d) 1.5 wt% HF (e) 2.0 wt% HF for 60 minutes at room temperature.

Scanning Electron Microscope for Characterising of Micro- and Nanostructured Titanium Surfaces 587

SEM analysis can clearly showed that anodisation is a simple and economical method to synthesise various surface patterns and textures on the surface of a metallic titanium surface. Moreover, surface morphology is strongly affected by anodisation condition. SEM revealed that the anodisation condition caused micro- and nanomorphological alterations of titanium surface, whereas prolonged exposure to electrolyte resulted in micromorphological changes of the titanium surface. The above result clearly point out that the hydrofluoric acid play important role in controlling the formation of titania nanotubes. Moreover, by adding fluoride ions into the electrolyte, nanotubes can be fabricated under suitable conditions

[5] M Cortada, L L Giner, S Costa, F J Gil, D Rodríguez, J A Planell, *J of Mats Sci: Mats in* 

[6] A W E Hodgson, Y Mueller, D Forster, S Virtanen, *Electrochemical Acta*., 47 (2002) 1913.

[8] D Leonardis, A K Garg, G E Pecora, *Int. J. of Oral & Maxillofacial Implants*., 14 (1999) 1. [9] Y X Leng, J Y Chen, P Yang, H Sun, N Huang, *Surf & Coating Technol*., 166 (2003) 176. [10] D M Blake, P C Maness, Z Huang, E J Wolfrum, J Huang, *Separation and Purification* 

[11] J A Byrne, J W J Hamilton, T A McMurray, P S M Dunlop, V J A Donaldson, J Rankin, G

[13] Y Yamashita, I Takayama, H Fujii, T Yamazaki, *Nippon Steel Technical Report*., 85 (2002)

[15] R Vogel, P Meredith, I Kartini, M Harvey, J D Riches, A Bishop, N Heckenberg, M Trau,

[18] J M Lee, Y S Kim, C W Kim, K S Jang, Y J Lim, *J Korean Acad Prosthodont*., 42 (2004) 352. [19] D P Dowling, P V Kola, K Donnelly, T C Kelly, K Brumitt, L Lloyd, R Eloy, M Therin, N

[21] A M PeirÓ, E Brillas, J Peral, X Domènech, J A AyllÓn, *J of Mats Chem*., 12 (2002) 2769. [22] Y Li, J Hagen, W Schaffrath, P Otschi, D Haarer, *Solar Energy Material and Solar Cells*., 56

[24] F Variola, F Vetrone, L Richert, P Jedrzejowski, J.-H Yi, S Zalzal, S Clair, A Sarkissian, D.F Perepichka, J.D Wuest, F Rosei, A Nanci, *Small*., 5 (2009) 996-1006. [25] C Toth, G Szabó, L Kovács, K Vargha, J Barabás, Z Németh, *Smart Mater. Struct*., 11

**4. Conclusion** 

**5. References** 

[1] M Yamada, *Mats Sci & Eng A*., 213 (1996) 8. [2] R R Boyer, *Mats Sci & Eng* A., 213 (1996) 103. [3] I V Gorynin, *Mats Sci & Eng A*., 263 (1999) 112.

*Med.*, 11 (2000) 287.

*Methods*., 28 (1999) 1.

11.

(1999) 167.

(2002) 813-818.

[12] A S Stasinakis, *Global NEST J*., 10 (2008) 376.

[14] G F New, *Current Science*., 3 (1973) 133.

[4] M Balazic and J Kopac, *Int. J. Nano and Biomaterials*., 1 (2007) 3.

[7] T Albrektsson and B Albrektsson, *Acta Orthopaedica*., 58 (1987) 567.

H Rubinsztein-Dunlop, *ChemPhysChem*., 4 (2003) 595. [16] S A Brown and J E Lemons, *ASTM STP 1272, ASTM*. (Philadelphia, 1996).

[23] S Lavanus, G Louarn, P Layrolle, *Int. J. of Biomaterials*., 2010 (2010) 1-9.

Dale, D Al Rousan. *Clean Technol*., (2007) 242.

[17] X Liu, P K Chu, C Ding, *Mats Sci & Eng R*., 47 (2004) 49.

Weill, *Diamond and Related Mats*., 6 (1997) 390. [20] C C Chen, J H Chen, C G Chao, *J of Mats Sci*., 40 (2005) 4053.

Fig. 5. SEM image of (a) untreated surface and (b) anodised in 1 M acetic acid with 0.075 wt% HF for 60 minutes at room temperature then EDS analysis of the surfaces for energy dispersive analysis was 15 kV, Takeoff Angle 35.0° and Elapsed Livetime 10.0.

## **4. Conclusion**

586 Scanning Electron Microscopy

(a)

(b)

Fig. 5. SEM image of (a) untreated surface and (b) anodised in 1 M acetic acid with 0.075 wt% HF for 60 minutes at room temperature then EDS analysis of the surfaces for energy

dispersive analysis was 15 kV, Takeoff Angle 35.0° and Elapsed Livetime 10.0.

SEM analysis can clearly showed that anodisation is a simple and economical method to synthesise various surface patterns and textures on the surface of a metallic titanium surface. Moreover, surface morphology is strongly affected by anodisation condition. SEM revealed that the anodisation condition caused micro- and nanomorphological alterations of titanium surface, whereas prolonged exposure to electrolyte resulted in micromorphological changes of the titanium surface. The above result clearly point out that the hydrofluoric acid play important role in controlling the formation of titania nanotubes. Moreover, by adding fluoride ions into the electrolyte, nanotubes can be fabricated under suitable conditions

## **5. References**


**Part 5** 

**Thin Films, Membranes, Ceramic** 

