Journal of The Electrochemical Society, 159 (3) B277-B284 (2012) B277 0013-4651/2012/159(3)/B277/8/$28.00 © The Electrochemical Society Thermodynamic Stability of Yttrium Alkaline Earth Borosilicate Glasses and Their Compatibility with Crofer for SOFC Gurbinder Kaur, Vishal Kumar, O. P. Pandey, and K. Singh z School of Physics and Materials Science, Thapar University, Patiala-147004, India Crystallization kinetics and stability are very important and essential parameters for glasses and glass ceramics. These parameters are highly dependent on initial composition as well as processing conditions. Thus, the role of modifier cation on thermal parameters have been studied for AO-SiO 2 -B 2 O 3 -Y 2 O 3 (A = Ca, Sr, Ba) glasses. Various kinetic parameters such as fluctuation free volume (f g ), change in bulk thermal expansion coefficient (α f ) and frequency factors (k y (T p ), k b (T p ), k f (T p )) have been calculated for better understanding of crystallization mechanics of present glasses. According to these parameters, CaO- SiO 2 -B 2 O 3 -Y 2 O 3 glass shows good thermodynamic and kinetic stability than other two glasses. The morphology of Crofer22APU/glass–ceramics diffusion couples (heat-treated at 850 ◦ C for 50 h) were analyzed by SEM and EPMA. The kinetic parameters and microstructural studies of glasses showed an excellent correlation with each other. CaY/Crofer22APU couple showed good interfacial adhesion along with minimum diffusion of ions on either side of the interface. © 2012 The Electrochemical Society. [DOI: 10.1149/2.037203jes] All rights reserved. Manuscript submitted October 7, 2011; revised manuscript received November 28, 2011. Published January 3, 2012. Efforts are being made to develop suitable sealing materials for solid oxide fuel cells (SOFC) which should be reliable, inexpensive and stable in the oxidizing as well as in reducing atmosphere. 1–3 Apart from this, it should also exhibit matching of thermal expansion co- efficient (TEC) with adjacent components of SOFC. 4, 5 Good quality glass sealants can be prepared by choosing proper initial constituents and their appropriate stochiometric proportion. It is essential to un- derstand the crystallization kinetics and thermodynamic behavior of glass sealants to achieve proper sealing and good chemical stabil- ity. Basically during operation of SOFCs, glass gets converted into glass ceramics. These crystalline phases, having different TEC, get segregated in the glass matrix causing thermal, residual and mechani- cal stresses in the glass composite. Moreover, during thermal cycling these internal stresses would be enhanced due to difference in TEC between the devitrified glass and other SOFC components. The crys- tallization kinetics of glass depends upon variety of parameters like activation energy, frequency and crystallization rate etc. Crystalliza- tion of glass is a heterogenous process occurring through interface at the phase boundary. It consists of molecular and atomic diffusion at the phase boundary and then leads to bulk crystallization. 6 Therefore in order to gain insight for the crystallization process, activation en- ergy of crystallization as well as glass transition has been interpreted with the help of theoretical models. All three models, which have been used in the present study, have some advantages and disadvantages over each other. For instance, Kissinger model is more generalized form to give information about the nucleation in glasses. Mohniyan model can give better idea about crystallization kinetic due to con- sideration of enthalpies of transformation. On the other hand, Augis and Benett model gives idea about the nucleation sites in terms of frequency factor besides activation energies. In addition to this, stability of glass sealant cannot be determined only by parameters like glass transition temperature (T g ) and crys- tallization temperature (T c ) as neither of them can be determined accurately. T g is a kinetic parameter and the underlying thermody- namic basis for glass formation is T g →T c as variation in temperature with time ( dT dt )→0. The properties of glasses are time dependent in glass transition region. When the time scale required for structural rearrangement of glasses becomes comparable to the experimental time scale, then their structural relaxation occurs. During the process of structural relaxation of glasses, the configuration of glasses tries to attain equilibrium. But at sufficiently low temperatures, the time scale becomes so long that any rapid change in temperature can “freeze-in” the liquid structure. 7 It is worthwhile to evaluate the thermal and kinetic stability of glasses during crystallization when they are subjected to thermal ag- z E-mail: kusingh@thapar.edu ing and reheating during their application as sealants. However, the determination of glass structure with variation in temperature is diffi- cult due to the absence of long range order in them. Moreover, glasses possess many possible metastable states having different atomic con- figuration slightly differing from each other. 8 The crystallization rate constant k takes into account both frequency and activation energy, so it is also a good criterion for the evaluation of glass stability as a sealant in SOFC. The k parameters can be used to evaluate the stability of glasses over a broader temperature range rather than at a single char- acteristic temperature. Therefore, three modified k criteria i.e. k y (T) = ν exp(−(E/RT)(T x −T g )/T m ), k b (T) = ν exp(−(E/RT)H r ), k f (T) = ν exp(−(E/RT)(T c −T f )/T f ) have been studied for comparisons 9–12 where T f is the inflection point temperature, T x is the crystallization onset temperature, T c is the peak crystallization temperature and T m is melting temperature of glass. In the present study, the major focus is on the chemical com- patibility of glasses with Crofer22APU metallic interconnect as a function of various theoretical parameters to check their applicability as sealants for SOFC. To gain insight into the interfacial interaction and its in-depth mechanism, other new parameters like fluctuation free volume and bulk thermal expansion coefficient have been calculated and correlated to experimental parameters. Along with these param- eters, crystallization constants as a function of different temperatures as well as different heating rates have been studied extensively. The diffusion couples of glass/ Crofer 22 APU (heat-treated at 850 ◦ C for 50 h) were investigated by using XRD, DTA, Dilatometery, SEM and EDS techniques. Experimental The glasses chosen for the present study are 30AO-40SiO 2 - 20B 2 O 3 -10Y 2 O 3 (A = Ca, Sr, Ba) and their compositions are listed in Table Ia. The glasses were prepared by taking required stochiometric amounts of different constituent oxides or carbonates of high purity (99.9%). These constituents were first mixed together using ball-mill in acetone medium. The powder obtained after ballmilling was melted at 1550 ◦ C and air quenched using copper plates. The quenched glass was further annealed at 500 ◦ C. More details of glass preparation are given in our earlier publication. 13 The final compositions of glasses were obtained using inductive coupled plasma (ICP) and are listed in Table Ib. The measurements were performed using the PerkinElmer Optima 7000 DV ICP-OES instrument (PerkinElmer, Inc., Shelton, CT, USA) equipped with WinLab32 for ICP, Version 4.0 software for measurement of all analyte wavelengths of interest. The transi- tion temperature measurements were performed on a Diamond Pyris TG/DTA (Perkin-Elmer) instrument in flowing nitrogen gas using 8–10 mg of glass powder in platinum crucibles. High purity (99.9%) ecsdl.org/site/terms_use address. Redistribution subject to ECS license or copyright; see 198.82.19.86 Downloaded on 2013-05-01 to IP