**1. Introduction**

58 Sintering of Ceramics – New Emerging Techniques

Yang, K.; He, J.; Su, Z.; Reppert, J. B.; Skove, M. J.; Tritt, T. M.; Rao, A. M. (2010). Inter-tube

Zhang, F. (2005). Strengthening and toughening mechanisms of carbon nanotubes

Zhang*,* F., Shen, J.; Sun, J.; Zhu, Y. Q.; Wang, G.; McCartney, G. (2005). Conversion of carbon nanotubes to diamond by a spark plasma sintering. *Carbon*. 43:1254-1258 Zhang, F.; Shen, J.; Sun, J. F.; McCartney, D. G. (2006). Direct synthesis of diamond from low

Zhang, F.; Burkel. E. (2010). Novel titanium manganese alloys and their macroporous foams

Zhang, F.; Adam, M.; Otterstein, E.; Burkel E. (2011). Pulsed electric field induced diamond

Zhang, F.; Mihoc, C.; Burke E. (2010). Spark plasma sintering assisted carbon conversion

Zhang, F.; Mihoc, C.; Ahmed, F.; Latte, C.; Burkel, E. (2011). Thermal stability of carbon

Zhang, F.; Ahmed, F.; Holzhüter, G.; Burkel, E. (2011). Growth of diamond from fullerene

Zhang, F.; Burkel, E.; Rott, G. (2011). Verfahren zur synthese von diamanten. *Deutsches* 

Zheng, Z.; Liao, L.; Yan, B.; Zhang, J. X.; Gong, H.; Shen, Z. X.; Yu, T. (2009). Enhanced field

C60 by spark plasma sintering. *Journal of Crystal Growth*.

sintered multi-wall carbon nanotube arrays. *Carbon*. 48:756-762

Harbin Institute of Technology

*Materials.* 20: 853–858

USA, October 17-21, 2010

doi:10.1016/j.jcrysgro.2011.11.01494

*Letters.* 510:109-114

*Patent*. P162-11

*Lett*. 4:1115–1119

purity carbon nanotubes. *Carbon*. 44 : 3136-3138

InTech, ISBN 978-953-7619, Rijeka, Croatia

bonding, graphene formation and anisotropic transport properties in spark plasma

reinforced WC nanocomposites synthesized by spark plasma sintering. *Ph.D thesis.*

for biomedical applications prepared by field assisted sintering. In: *Biomedical Engineering, Trends, Researches and Technologies*, Anthony N. Laskovski, pp. 203-224.

synthesis from carbon nanotubes with solvent catalysts. *Diamond and related* 

from various modifications to diamond. *Conference Proceedings of Materials Science and Technology (MS&T) 10,* pp. 2312-2317, ISBN 978-0-87339-756-8, Houston, TX,

nanotubes, fullerene and graphite under spark plasma sintering. *Chemical Physics* 

emission from argon plasma-treated ultra-sharp *a*-Fe2O3 nanoflakes. *Nanoscale Res* 

The successful development and commercialization of high performance ceramic materials has attracted much attention especially for multilayer substrates using the Low Temperature Co-fired Ceramic (LTCC) technology. This technology has become a popular technology for automobiles and wireless communications due to the advantages of the excellent combination of electrical, thermal, mechanical and chemical stability for a wide range of applications, thus allowing preparation of 3-dimensional circuits incorporating passive components within a multilayer construction (Matters-Kammerer et al., 2006; Zhou et al., 2008). This approach also allows the presence a number of interfaces and thus reduction of the overall substrate size and cost can be realized (Lo and Duh, 2002; Chen et al., 2004 and Zhu et al., 2007). The circuits are capable of withstanding sintering during processing temperatures up to 1000 °C. For telecommunication applications the usage of ceramic is implemented in Telecom control station and power supply circuits for the capability to dissipate excess heat and maintain dimensional control stability of the ceramic package. This is important where back-up power is required to maintain operation during primary power outages when cooling is restricted (Barlow and Elshabini, 2007). Another important parameter for wireless communication devices is the requirement to have low dielectric loss (tan δ ∼ 10-3 or less) for higher processing speed, higher dielectric constant (ε'>10) for miniaturization of the devices and higher integration density (≥ 3 g/cm3) (Kume et al., 2007; Long et al., 2009). For this reason, it is important to prepare high quality LTCC substrate/package whose properties are strongly dependent on microstructure, phase purity and sintering temperature (Xiang et al., 2002). Therefore the microstructure must be carefully controlled to get dense and fine grained ceramics in order to improve their properties and reliability in many applications (Hsu et al., 2003).

The starting point of the LTCC technology is the development of LTCC tape materials containing a glass-ceramic system that usually needs to show good compatibility with the paste system which acts as a conductive track for RF signal transportation from one location on the circuit to another. It should also have low energy loss in microwave applications to make sure high circuit performance can be achieved (Wang et al., 2009). One of the most important processes in LTCC manufacturing of multilayer LTCC substrates involves co-

The Effects of Sintering Temperature Variations on Microstructure Changes of LTCC Substrate 61

2008. Most of their studies have been performed on multiphase material systems and focused crystallization behavior of the systems. Wang and Zhou (2003) have studied the relation of densification and dielectric properties of glass ceramics with different compositions selected from the CaO-B2O3-SiO2 system. They have found that too high sintering temperature will ruin the properties and sintering with a heating speed of 10 °Cmin-1 at 850 °C for 60 min is an optimal technology for CBS system. They also notice that too much B2O3 will damage the dielectric properties of the glass ceramics (Wang and Zhou, 2003). More recently, Hsi and co-workers have further studied the relation of silver and LTCC. They demonstrate the easy method to decrease effectively the diffusivity of silver ions in LTCC dielectrics by adding 5 wt % of SiO2 (Hsi et al., 2011). However, not much literature has been found to investigate the microstructure changes of silver conductor and glass-ceramic substrate by changing the sintering temperature and this situation may not be very helpful in understanding the surface properties of the printed conductor and the microstructure of glass-ceramic system in this project. Nevertheless, the results obtained in this work can provide useful findings for researchers and

There are a few factors that determine the microstructure for both printed conductor and LTCC substrate such as the quality of the raw materials and the sintering temperature. The parameters of microstructure are the grain size, grain size distribution, pore size and pore size distribution and the density. The size, shape and also the distribution of grain size and pore size of the conductor and glass-ceramics will vary with preparation conditions and techniques. The objective of this work is to investigate the effect of sintering temperature variations on the microstructure of both metal and ceramic materials of the LTCC substrate

In this chapter, the work is constructed as follows; firstly, a simple transmission line pattern on the eight layers substrate is fabricated using a standard LTCC technology. Printed patterns were dried in a box oven for 10 minutes at about 70 °C to remove the organic solvent slowly before cofiring process in order to avoid blister, void formation, delamination, cracks and camber effect. These undesirable defects are usually formed due to the different shrinkage rates of LTCC tape and thick-film paste during firing (Hsueh and Evans, 1985; Hrovat et al., 2009). The film was then fired at various sintering temperatures of 800 °C to 900 °C at 25 °C increments. Generally at peak firing temperature the film achieves desired electrical properties and the ceramic composite also has undergone changes in physical properties. Further, sample characterizations were done such as crystal structure determination, linear shrinkage, density measurement and microstructure observation for both metal and LTCC substrate. After that sample analysis and detail discussion is carried

In the following sections, the various process stages will be described. The effect of process variables on the physical properties and microstructure of the finished product has always been a subject of great importance. The factors that have the most influence on these properties are the purity of the constituent oxides, their proportions and homogeneity in the

powder mix and the control of temperature and atmosphere during sintering.

industries.

for microwave applications.

out to explain the research findings.

**2. Experimental procedure** 

firing of metal and glass-ceramic. The main problem in the development of materials for LTCC integrated modules arises from the chemical incompatibility of the different materials within the module during the sintering process (Valant and Suvorov, 2000). So, mismatch in sintering kinetics, sintering stress, density variations between tape layers and non-uniform shrinkage of the individual tapes should be taken into consideration to avoid some defects such as delamination, blister, and camber form during a multilayer LTCC process. Therefore, the processes that result in the above mentioned defects must be completely understood before the processing parameters and material compositions are optimized.

Thick-film technology is a method whereby conductive, resistive and dielectric pastes are applied to a ceramic substrate. There are various available ways of producing thick-films such as dip coating, spin coating, screen printing and chemical vapor deposition technique down to 25 µm thickness. However the screen printing method is simpler, more convenient and the most cost effective method to transfer the desired thick-film pattern onto the substrate in order to realize interconnect films with a thickness ranging from 3-15 μm which depends upon the screen printing parameter. Highly conductive metals such as copper, silver and gold are typical electrode materials in LTCC components and modules. A silver (Ag) conductor is very attractive because the price of silver is much cheaper than that of others and can be fired in air. It also has some advantages such superior electrical conductivity and good thermodynamic stability below 200 °C. Above 200 °C it deoxidize to metallic silver (Imanaka, 2005; Bangali et al., 2008; Jean et al., 2004). For the LTCC process the dielectric ceramics and silver should be cofired simultaneously so it is important to carried out the sintering process below the melting point of the metal electrode such as Ag (961 °C) to reduce sintering mismatch between each other which form some defects for the whole substrate (Feteira and Sinclair, 2008; Chang and Jean, 1998; Long et al., 2009). With regard to the co-firing process technology and to get detailed understanding of LTCC process behavior, the mechanism of low temperature sintering was studied by Valant et al., in 2006 using BaTiO3 whose original sintering temperature was about 1250 °C -1300 °C. They used a small amount of Li2O as a sintering aid and found that small amount of sintering aid about 0.3 wt% was able to reduce the sintering temperature to 820 °C and ceramics with more than 95% of relative density could be produced.

The study of transition metal elements especially controlling the properties of silver is of great importance from the view point of industrial applications. It is not a simple task and demands thorough understanding of structure ⇔ property ⇔ processing relationship (Despande et al., 2005). The overall behavior of silver metal is influenced by the nature of atoms, the size of the grain and the nature of grain boundaries in addition to controlled chemical composition and preparation conditions. So the material specifications/characteristics of the thick-film can fulfill conductor paste requirements in a wide range of applications. Typically, a thick-film conductor paste consists of metal powder as a major component and glass and/or oxide (minor) dispersed in an organic medium (Sergent and Harper, 1995). The metal powder forms a continuous film on the substrate upon firing. The role of the glass and /or oxide in the conductor paste is to help the metal film adhere to the substrate. The investigation of the preparation of LTCC packages using glass + ceramic and glass ceramic system with high electrical conductivity such as gold and silver has been conducted by several researchers such as Shimada et al., 1983, Tummala, 1991, Imanaka and Kamehara, 1992, Jean et al., 2001 and Bangali et al., 2008. Most of their studies have been performed on multiphase material systems and focused crystallization behavior of the systems. Wang and Zhou (2003) have studied the relation of densification and dielectric properties of glass ceramics with different compositions selected from the CaO-B2O3-SiO2 system. They have found that too high sintering temperature will ruin the properties and sintering with a heating speed of 10 °Cmin-1 at 850 °C for 60 min is an optimal technology for CBS system. They also notice that too much B2O3 will damage the dielectric properties of the glass ceramics (Wang and Zhou, 2003). More recently, Hsi and co-workers have further studied the relation of silver and LTCC. They demonstrate the easy method to decrease effectively the diffusivity of silver ions in LTCC dielectrics by adding 5 wt % of SiO2 (Hsi et al., 2011). However, not much literature has been found to investigate the microstructure changes of silver conductor and glass-ceramic substrate by changing the sintering temperature and this situation may not be very helpful in understanding the surface properties of the printed conductor and the microstructure of glass-ceramic system in this project. Nevertheless, the results obtained in this work can provide useful findings for researchers and industries.

There are a few factors that determine the microstructure for both printed conductor and LTCC substrate such as the quality of the raw materials and the sintering temperature. The parameters of microstructure are the grain size, grain size distribution, pore size and pore size distribution and the density. The size, shape and also the distribution of grain size and pore size of the conductor and glass-ceramics will vary with preparation conditions and techniques. The objective of this work is to investigate the effect of sintering temperature variations on the microstructure of both metal and ceramic materials of the LTCC substrate for microwave applications.

In this chapter, the work is constructed as follows; firstly, a simple transmission line pattern on the eight layers substrate is fabricated using a standard LTCC technology. Printed patterns were dried in a box oven for 10 minutes at about 70 °C to remove the organic solvent slowly before cofiring process in order to avoid blister, void formation, delamination, cracks and camber effect. These undesirable defects are usually formed due to the different shrinkage rates of LTCC tape and thick-film paste during firing (Hsueh and Evans, 1985; Hrovat et al., 2009). The film was then fired at various sintering temperatures of 800 °C to 900 °C at 25 °C increments. Generally at peak firing temperature the film achieves desired electrical properties and the ceramic composite also has undergone changes in physical properties. Further, sample characterizations were done such as crystal structure determination, linear shrinkage, density measurement and microstructure observation for both metal and LTCC substrate. After that sample analysis and detail discussion is carried out to explain the research findings.

#### **2. Experimental procedure**

60 Sintering of Ceramics – New Emerging Techniques

firing of metal and glass-ceramic. The main problem in the development of materials for LTCC integrated modules arises from the chemical incompatibility of the different materials within the module during the sintering process (Valant and Suvorov, 2000). So, mismatch in sintering kinetics, sintering stress, density variations between tape layers and non-uniform shrinkage of the individual tapes should be taken into consideration to avoid some defects such as delamination, blister, and camber form during a multilayer LTCC process. Therefore, the processes that result in the above mentioned defects must be completely understood before the processing parameters and material compositions are optimized.

Thick-film technology is a method whereby conductive, resistive and dielectric pastes are applied to a ceramic substrate. There are various available ways of producing thick-films such as dip coating, spin coating, screen printing and chemical vapor deposition technique down to 25 µm thickness. However the screen printing method is simpler, more convenient and the most cost effective method to transfer the desired thick-film pattern onto the substrate in order to realize interconnect films with a thickness ranging from 3-15 μm which depends upon the screen printing parameter. Highly conductive metals such as copper, silver and gold are typical electrode materials in LTCC components and modules. A silver (Ag) conductor is very attractive because the price of silver is much cheaper than that of others and can be fired in air. It also has some advantages such superior electrical conductivity and good thermodynamic stability below 200 °C. Above 200 °C it deoxidize to metallic silver (Imanaka, 2005; Bangali et al., 2008; Jean et al., 2004). For the LTCC process the dielectric ceramics and silver should be cofired simultaneously so it is important to carried out the sintering process below the melting point of the metal electrode such as Ag (961 °C) to reduce sintering mismatch between each other which form some defects for the whole substrate (Feteira and Sinclair, 2008; Chang and Jean, 1998; Long et al., 2009). With regard to the co-firing process technology and to get detailed understanding of LTCC process behavior, the mechanism of low temperature sintering was studied by Valant et al., in 2006 using BaTiO3 whose original sintering temperature was about 1250 °C -1300 °C. They used a small amount of Li2O as a sintering aid and found that small amount of sintering aid about 0.3 wt% was able to reduce the sintering temperature to 820 °C and

ceramics with more than 95% of relative density could be produced.

The study of transition metal elements especially controlling the properties of silver is of great importance from the view point of industrial applications. It is not a simple task and demands thorough understanding of structure ⇔ property ⇔ processing relationship (Despande et al., 2005). The overall behavior of silver metal is influenced by the nature of atoms, the size of the grain and the nature of grain boundaries in addition to controlled chemical composition and preparation conditions. So the material specifications/characteristics of the thick-film can fulfill conductor paste requirements in a wide range of applications. Typically, a thick-film conductor paste consists of metal powder as a major component and glass and/or oxide (minor) dispersed in an organic medium (Sergent and Harper, 1995). The metal powder forms a continuous film on the substrate upon firing. The role of the glass and /or oxide in the conductor paste is to help the metal film adhere to the substrate. The investigation of the preparation of LTCC packages using glass + ceramic and glass ceramic system with high electrical conductivity such as gold and silver has been conducted by several researchers such as Shimada et al., 1983, Tummala, 1991, Imanaka and Kamehara, 1992, Jean et al., 2001 and Bangali et al.,

In the following sections, the various process stages will be described. The effect of process variables on the physical properties and microstructure of the finished product has always been a subject of great importance. The factors that have the most influence on these properties are the purity of the constituent oxides, their proportions and homogeneity in the powder mix and the control of temperature and atmosphere during sintering.

The Effects of Sintering Temperature Variations on Microstructure Changes of LTCC Substrate 63

A typical practice is to heat up the furnace to 450 °C with a gradient of about 2-5 °C/min for 1 hour to completely remove the organic solvent. In the next 2 hours the temperature is raised to about 850 °C at which the sintering process of the composite material is started. The temperature remains constant for 10 minutes to finish the sintering. The last stage is the cooling period which takes about 3 hours depending on the size of the product. This firing condition is selected because of the practical experience on thick-film circuit and suggested

The structural characteristics of the LTCC material and metal surface were measured using a Pan Analytical Diffractometer system operating at 45 kV and 40 mA and employing CuK<sup>α</sup> radiation with λ = 1.54060 Å. The scanning measurements were 2θ in a range from 2° to 80°, in steps of 0.05° of 2θ and a counting time of 25 s per step. The average crystallite size of all the samples is determined from the full width at half maximum (FWHM) of the (1 1 1) reflection peak in the XRD patterns by using the Debye Scherrer formula shown in equation

0.94<sup>λ</sup> <sup>D</sup>

where λ is the wavelength of the incident x-ray; β is the full width at half maxima; θ is the

The most important point to bear in mind in the sintering process is controlling the variation in dimensional changes. One of the methods to control the dimensional changes in terms of

P1 P2 P3

P4 P5 P6

P7 P8 P9

shrinkage is through the control part as shown in Fig. 2. The steps are as follows:

<sup>β</sup>cos<sup>θ</sup> <sup>=</sup> (1)

by the tape manufacturer.

**2.2 Substrate characterizations** 

(1), (Klug & Alexander, 1974).

Bragg's diffraction angle.

Fig. 2. Control part (Ferro).

**2.2.2 Shrinkage** 

**2.2.1 Crystal structure determination** 

The key stages in the fabrication of glass-ceramics are sintering process of various components together. During these processes the constituent atoms redistribute themselves in such a way as to minimize the free energy of the system. This involves a considerable movement of ions, their inter-diffusion to form a new phase, the minimization of the internal surface area and increase in grain size.

#### **2.1 LTCC multilayer substrate preparation**

In the present work, the commercial Ferro A6S was chosen and cut into 8 pieces for the required dimension (21 mm x 21 mm) using a die cutting machine (ATOM SE 20C). The multilayer stack was prepared by standard LTCC process flow as shown in Fig. 1. The thickness of the green tape is about 100 μm for each layer; the green tape was punched and screen printed with CN 33-391 for the surface and inner conductor and CN 33-407 for via fill. The printed pattern was then stacked using a manual stacker plate. The stacked substrate was then laminated by using an isostatic laminator system under pressure and temperature of 3000 psi (21 MPa) and 70 °C respectively to ensure that the layers of the stack are well adhered to each other and to form a compact multilayer substrate. The laminated substrate was placed in a tube furnace and fired using a sintering profile as suggested by the tape manufacturer. The substrates were sintered in the temperature range of 800 °C to 900 °C at 25 °C increments. This sintering is the most critical process step because during the firing step the material becomes a compact ceramic LTCC substrate and its properties are determined. Through sintering, a transformation from the original porous compact to a dense ceramic takes place (Kingery, 1976). At a high temperature, the particles are in an increased contact and the particles grow together to form crystallite grains.

Fig. 1. LTCC Multilayer Fabrication Process Flow.

A typical practice is to heat up the furnace to 450 °C with a gradient of about 2-5 °C/min for 1 hour to completely remove the organic solvent. In the next 2 hours the temperature is raised to about 850 °C at which the sintering process of the composite material is started. The temperature remains constant for 10 minutes to finish the sintering. The last stage is the cooling period which takes about 3 hours depending on the size of the product. This firing condition is selected because of the practical experience on thick-film circuit and suggested by the tape manufacturer.

#### **2.2 Substrate characterizations**

62 Sintering of Ceramics – New Emerging Techniques

The key stages in the fabrication of glass-ceramics are sintering process of various components together. During these processes the constituent atoms redistribute themselves in such a way as to minimize the free energy of the system. This involves a considerable movement of ions, their inter-diffusion to form a new phase, the minimization of the

In the present work, the commercial Ferro A6S was chosen and cut into 8 pieces for the required dimension (21 mm x 21 mm) using a die cutting machine (ATOM SE 20C). The multilayer stack was prepared by standard LTCC process flow as shown in Fig. 1. The thickness of the green tape is about 100 μm for each layer; the green tape was punched and screen printed with CN 33-391 for the surface and inner conductor and CN 33-407 for via fill. The printed pattern was then stacked using a manual stacker plate. The stacked substrate was then laminated by using an isostatic laminator system under pressure and temperature of 3000 psi (21 MPa) and 70 °C respectively to ensure that the layers of the stack are well adhered to each other and to form a compact multilayer substrate. The laminated substrate was placed in a tube furnace and fired using a sintering profile as suggested by the tape manufacturer. The substrates were sintered in the temperature range of 800 °C to 900 °C at 25 °C increments. This sintering is the most critical process step because during the firing step the material becomes a compact ceramic LTCC substrate and its properties are determined. Through sintering, a transformation from the original porous compact to a dense ceramic takes place (Kingery, 1976). At a high temperature, the particles are in an

increased contact and the particles grow together to form crystallite grains.

**AOI** Stacking

Sintering

Laminating

**AOI** 

**AOI** 

**AOI** 

internal surface area and increase in grain size.

**2.1 LTCC multilayer substrate preparation** 

Fig. 1. LTCC Multilayer Fabrication Process Flow.

Screen printing

Via punching

Via filling

Blanking

#### **2.2.1 Crystal structure determination**

The structural characteristics of the LTCC material and metal surface were measured using a Pan Analytical Diffractometer system operating at 45 kV and 40 mA and employing CuK<sup>α</sup> radiation with λ = 1.54060 Å. The scanning measurements were 2θ in a range from 2° to 80°, in steps of 0.05° of 2θ and a counting time of 25 s per step. The average crystallite size of all the samples is determined from the full width at half maximum (FWHM) of the (1 1 1) reflection peak in the XRD patterns by using the Debye Scherrer formula shown in equation (1), (Klug & Alexander, 1974).

$$\mathbf{D} = \frac{0.94\lambda}{\beta \cos \theta} \tag{1}$$

where λ is the wavelength of the incident x-ray; β is the full width at half maxima; θ is the Bragg's diffraction angle.

#### **2.2.2 Shrinkage**

The most important point to bear in mind in the sintering process is controlling the variation in dimensional changes. One of the methods to control the dimensional changes in terms of shrinkage is through the control part as shown in Fig. 2. The steps are as follows:

Fig. 2. Control part (Ferro).

The Effects of Sintering Temperature Variations on Microstructure Changes of LTCC Substrate 65

sintering the organic solvent evaporates and the oxides react to form crystallites, or grains of the required composition, the grains nucleating at discrete centers and growing outwards until the boundaries meet those of the neighboring crystallites. During this process, the density of the material rises; if this process were to yield perfect crystals meeting at perfect boundaries the density would rise to the theoretical maximum, i.e. the x-ray density, which is the material mass in a perfect unit crystal cell divided by the cell volume. In practice imperfections occur and the sintered mass has microscopic voids both within the grains and at the grain boundaries. The resulting density is referred to as the sintered density. The density of the sample was measured using the Archimedes principle shown in equation (3);

> *Wa <sup>w</sup> Ww* = ∗

 ρ

%Dth = (measure density / theoretical density) x 100% (4)

The theoretical density of Calcium Boron silicate (CaB2Si2O8) was calculated using equation

*N Ac V Nc <sup>A</sup>*

Where *NC* = No. of molecules per unit cell, A = molecular weight, *VC* = volume of one molecules and *NA* = Avogadro number. So the theoretical density of CaB2Si2O8 was calculated by taking the molecular weight of CaB2Si2O8 to be 245.87g. Since orthorhombic structure of CaB2Si2O8 have four formula units per unit cell, the molecular weight of one cell is (4) x (245.87) = 983.48. The volume of the orthorhombic structure unit cell is *a* x *b* x *c*. The volume of a mole of the material is therefore *NA* x a x b x c, where *NA* is Avogadro's number. The unit cell edge *a, b and c* (Ǻ) of CaB2Si2O8 = 8.7500, 8.0100 and 7.7200 therefore *a* 

[6.023 × 1023][541.08 × 10-24] = 325.92 cm3

*<sup>x</sup>* was calculated as above. The percentage of porosity of the

(6)

so, theoretical density is mass/volume equals to 983.48 g /325.89 = 3.018 g/cm3.

*P* 1

= − ρ

*x*

ρ

ρ

x *b* x c = 541.08 Ǻ3. As 1Ǻ3 = 10-24cm3, A x *a* x *b* x *c* is therefore:

ρ

is the measured density of the sample.

sample was calculated using the relation as below:

(3)

= (5)

ρ

The percentage theoretical density (%Dth) was calculated using this formula:

where Wa = weight of sample in air Ww = weight of sample in water ρ\*w = density of water = 1 gcm-3

below:

The theoretical density

where

ρ


$$\text{Shrinkage} = \left(\frac{\text{Length}\_{\text{beforeired}} - \text{Length}\_{\text{afterfree}}}{\text{Length}\_{\text{beforefree}}}\right) \times 100\,\% \tag{2}$$

For the substrate, linear shrinkage was measured along the compaction direction and the diametrical shrinkage from the geometry of the substrate. Repeatability and consistency of the shrinkage percentage must be the top criteria when designing the LTCC product because the shrinkage of LTCC substrate depends on the reactivity of the co-fired material containing ceramic oxide, glass, metal, organic solvent and also the firing conditions such as temperature, time and ambient air. Better reproducibility increases the uniformity of finished product characteristics and therefore increases the process yield. It is not an easy task because all process parameters (lamination, binder burnout, sintering) and material properties (high temperature reactivity, thermal expansion, etc.) must be matched (Rabe et al., 2005). Furthermore, the material quality of the finished product and process conditions also must be optimized in micro and macro structures in order to make sure work in progress are homogeneous at each process step (Imanaka, 2005). In order to achieve desired shrinkage data, the process engineer must establish control of the critical process variable. In commercial production, the designed shrinkage is generally between 12-16 % for XY direction and 20-25 % for Z direction.

#### **2.2.3 Microstructure observation**

The surface morphology and composition of the LTCC material and the printed material were observed by using a FEI NOVA Nano SEM 230 machine equipped with an Energy Dispersive X-ray (EDX) detector to reveal the microstructure of the resulting product. Most of the samples were imaged several times, with at least three pictures in each case from different areas of the sample. The thickness and width of conductive traces were measured by an optical microscope and SEM through the cross section view respectively. The average grain size was calculated using the line intercept method. The EDX spectrum is used to identify elements within a sample.

#### **2.2.4 Density measurement**

The properties of the final ceramic composite materials depend on the sintered density of the whole substrate. A stacked and laminated LTCC substrate before firing consists of a relatively porous compact of oxides in combination with a polymer solvent. During sintering the organic solvent evaporates and the oxides react to form crystallites, or grains of the required composition, the grains nucleating at discrete centers and growing outwards until the boundaries meet those of the neighboring crystallites. During this process, the density of the material rises; if this process were to yield perfect crystals meeting at perfect boundaries the density would rise to the theoretical maximum, i.e. the x-ray density, which is the material mass in a perfect unit crystal cell divided by the cell volume. In practice imperfections occur and the sintered mass has microscopic voids both within the grains and at the grain boundaries. The resulting density is referred to as the sintered density. The density of the sample was measured using the Archimedes principle shown in equation (3);

$$
\rho = \left(\frac{\mathcal{W}\_{\mathcal{Q}}}{\mathcal{W}\_{\mathcal{W}}}\right) \rho \*\_{w} \tag{3}
$$

where Wa = weight of sample in air Ww = weight of sample in water ρ\*w = density of water = 1 gcm-3

64 Sintering of Ceramics – New Emerging Techniques

1. The control part study main point is to study shrinkage variation. Measurements are taken before and after firing. Measure thickness (Z) of each laminate midway between each via and record data. There will be eight (8) measurements total. Next, measure the average width and length of each laminate by measuring from corner to adjacent corner

2. Next, measure the distance between each via. Record data, there are 12 measurements needed. Via holes 1,2 - 2,3 - 4,5 - 5,6 - 7,8 - 8,9 are for the X direction. Via holes 1,4 - 2,5 -

> Length Length Shrinkage x100% Length <sup>−</sup> <sup>=</sup>

For the substrate, linear shrinkage was measured along the compaction direction and the diametrical shrinkage from the geometry of the substrate. Repeatability and consistency of the shrinkage percentage must be the top criteria when designing the LTCC product because the shrinkage of LTCC substrate depends on the reactivity of the co-fired material containing ceramic oxide, glass, metal, organic solvent and also the firing conditions such as temperature, time and ambient air. Better reproducibility increases the uniformity of finished product characteristics and therefore increases the process yield. It is not an easy task because all process parameters (lamination, binder burnout, sintering) and material properties (high temperature reactivity, thermal expansion, etc.) must be matched (Rabe et al., 2005). Furthermore, the material quality of the finished product and process conditions also must be optimized in micro and macro structures in order to make sure work in progress are homogeneous at each process step (Imanaka, 2005). In order to achieve desired shrinkage data, the process engineer must establish control of the critical process variable. In commercial production, the designed shrinkage is generally between 12-16 % for XY

The surface morphology and composition of the LTCC material and the printed material were observed by using a FEI NOVA Nano SEM 230 machine equipped with an Energy Dispersive X-ray (EDX) detector to reveal the microstructure of the resulting product. Most of the samples were imaged several times, with at least three pictures in each case from different areas of the sample. The thickness and width of conductive traces were measured by an optical microscope and SEM through the cross section view respectively. The average grain size was calculated using the line intercept method. The EDX spectrum is used to

The properties of the final ceramic composite materials depend on the sintered density of the whole substrate. A stacked and laminated LTCC substrate before firing consists of a relatively porous compact of oxides in combination with a polymer solvent. During

before fired after fired beforefired

(2)

3. The shrinkage and density of sample test pattern are calculated using equation (2).

and record data.

3,6 - 4,7 - 5,8 - 6,9 are for the Y direction.

direction and 20-25 % for Z direction.

**2.2.3 Microstructure observation** 

identify elements within a sample.

**2.2.4 Density measurement** 

The percentage theoretical density (%Dth) was calculated using this formula:

$$\%D\_{\text{th}} = \text{(measure density / theoretical density)} \times 100\% \tag{4}$$

The theoretical density of Calcium Boron silicate (CaB2Si2O8) was calculated using equation below:

$$
\rho = \frac{N\_{\mathcal{C}} A}{V\_{\mathcal{C}} N\_A} \tag{5}
$$

Where *NC* = No. of molecules per unit cell, A = molecular weight, *VC* = volume of one molecules and *NA* = Avogadro number. So the theoretical density of CaB2Si2O8 was calculated by taking the molecular weight of CaB2Si2O8 to be 245.87g. Since orthorhombic structure of CaB2Si2O8 have four formula units per unit cell, the molecular weight of one cell is (4) x (245.87) = 983.48. The volume of the orthorhombic structure unit cell is *a* x *b* x *c*. The volume of a mole of the material is therefore *NA* x a x b x c, where *NA* is Avogadro's number. The unit cell edge *a, b and c* (Ǻ) of CaB2Si2O8 = 8.7500, 8.0100 and 7.7200 therefore *a*  x *b* x c = 541.08 Ǻ3. As 1Ǻ3 = 10-24cm3, A x *a* x *b* x *c* is therefore:

$$[6.023 \,\mathrm{\,\mathrm{\,\,\$^{1}\$^{1}\$^{2}\$}][541.08 \,\mathrm{\,\mathrm{\,\,\$^{1}\$^{2}\$^{1}\$}] = \$325.92 \,\mathrm{cm^{3}\$^{1}\$}}]$$

so, theoretical density is mass/volume equals to 983.48 g /325.89 = 3.018 g/cm3.

The theoretical density ρ *<sup>x</sup>* was calculated as above. The percentage of porosity of the sample was calculated using the relation as below:

$$P = 1 - \frac{\rho}{\rho\_X} \tag{6}$$

where ρis the measured density of the sample.

The Effects of Sintering Temperature Variations on Microstructure Changes of LTCC Substrate 67

**800 32.2207 2.77597 29.2565 825 32.1571 2.78132 29.2518 850 32.0970 2.78638 19.4982 875 32.0726 2.78845 25.0677 900 31.9778 2.7965 17.5432** 

The physical data of the LTCC substrate fired at various temperatures are tabulated in Table 2. The relationship of density with sintering temperatures is shown by Fig. 4. The results reveal decreasing trend of the density with sintering temperature with the highest density being 2.992 g/cm3 at 800 °C and the lowest density is 2.806 g/cm3 at 900 °C. The densities of all the samples were between 92 to 99 % of theoretical density (3.018 g/cm3). This trend seems to contradict the normal expected phenomenon that increasing sintering temperature should increase the density. As is well known the sintering process of a ceramic based material is the sintering of the powder compact into the final material. During this step, the porosity decreases and the microstructure of the material develop; this determines its final performance. During the sintering of a homogeneous material the porosity induced during the preparation of the green compact gradually decreases, depending on the powder morphology, agglomeration, the presence of liquid phase sintering and the sintering condition itself. However, the sintering of heterogeneous materials in the LTCC substrate, reactive sintering occurs in the concurrent process of reaction and densification during sintering. A variety of reactions are possible: oxidation-reduction, phase transition or solid solution formation. In this way reaction caused by impurities, additives or other product formed during heating which are often included in the normal sintering process may imply some sort of reactive sintering which usually generates additional porosity. This sintering process is complicated because the phase changes are involved. Thus the understanding of material behaviors such as binder burnout, densification mechanisms of LTCC, pore evolution and deformation of suspended LTCC is important in optimizing the fabrication process for multilayer LTCC substrates as well as for tailoring new LTCC systems

**(X)/%** 

**800 2.992 0.92 14.13 13.28 23.13 825 2.873 4.97 14.35 13.67 22.74 850 2.910 3.64 14.41 14.00 21.56 875 2.860 5.29 15.06 15.15 21.16 900 2.806 7.08 14.53 13.99 21.85**  Table 2. Physical properties of LTCC tape samples fired at various sintering temperature.

**Shrinkage (Y)/%** 

**Shrinkage (Z)/%** 

**Crystallite size (Dp)** 

**Temperature (**°**C) 2 Theta Dspacing**

**Sintering** 

Table 1. Parameter of XRD.

**3.2 Physical measurements** 

(Kemethmuller et al., 2007).

**Sample Density Porosity Shrinkage** 

 **(g/cm3) (%)** 

**3.2.1 Density** 
