Post on 05-Sep-2016
transcript
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 3 8 3 9e3 8 4 7
Avai lab le a t www.sc iencedi rec t .com
journa l homepage : www.e lsev ier . com/ loca te /he
Microstructural study of Crofer 22 APU-glass interface forSOFC application
Bhupinder Kaur, K. Singh, O.P. Pandey*
School of Physics and Materials Science, Thapar University, Patiala, Punjab 147004, India
a r t i c l e i n f o
Article history:
Received 14 March 2011
Received in revised form
22 April 2011
Accepted 23 April 2011
Available online 11 June 2011
Keywords:
Crystallization kinetics
Hurby parameter
Microstructure
X-ray methods
* Corresponding author. Tel.: þ919888401777E-mail address: oppandey@thapar.edu (O
0360-3199/$ e see front matter Copyright ªdoi:10.1016/j.ijhydene.2011.04.160
a b s t r a c t
Glasses of composition (60�x)SiO2e15CaOe10Al2O3e(10 þ x)Na2Oe5TiO2 (x ¼ 0, 5, 10, 15)
were prepared. The crystallization kinetics of prepared glasses were investigated by
differential thermal analyzer, dilatometer and X-ray diffraction techniques. In order to
check the applicability of the prepared glasses as a sealant in solid oxide fuel cell, these
were deposited on Crofer 22 APU steel by slurry technique and thermally treated. The
interface of the glassesteel composites was further analyzed under scanning electron
microscope in conjunction with energy-dispersive X-ray spectroscopy and X-ray dot
mapping. All glass samples exhibit phase separation. The phase separation tendency
increases with increasing content of Na2O in glasses. The N-10 glass containing 10% Na2O
shows good adhesion with Crofer 22APU.
Copyright ª 2011, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights
reserved.
1. Introduction [6e11]. Moreover, these reported compositions of sealing
Solid oxide fuel cells (SOFCs) are energy conversion devices
that can convert chemical energy of the fuel directly into
electrical energy. This process is eco-friendly and efficient one
[1e3]. SOFCs have received great attention as a promising new
technology for power generation because they are able to
utilize a wide variety of hydrocarbon fuels. A planar design of
SOFC is simple, produces higher power densities with high
efficiency than other designs.
The typical design of planar SOFC requires hermetic seals
to prevent fuel-oxidant mixing and also electrical insulation
to the stack [4]. Glass and glass ceramics have been widely
used for sealing different components of SOFC. First and the
foremost requirements of a glass to be used as a seal is to
have excellent thermo-chemical and thermo-mechanical
stability in stringent oxidizing and reducing environments of
fuel cell [5]. Lanthanum containing borosilicate glasses have
been extensively studied as sealing materials for SOFC
..P. Pandey).2011, Hydrogen Energy P
glasses are mostly confined to composition where silica acts
as a glass former. Al2O3 in glasses acts as intermediate,
which behaves as a network former and network modifier in
the glass system. B2O3 is added in most of the glasses to
decrease viscosity and increase the wettability with other
components of SOFC. However, it also decreases the stability
of seal as it leads to phase separation in borosilicate glasses.
Furthermore, the presence of La2O3 additive in the glass
matrix leads to formation of the devitrified phase at
operating temperature of SOFC and the amount of that
phase increases with high temperature aging [12]. Apart
from B2O3, alkali metal oxide (Na2O) is used as a modifier to
increase the wettability of glasses. It leads to decrease in
glass transition temperature while the thermal expansion
coefficient increases in hydrothermal conditions of SOFC
[13]. This causes poor chemical stability of Na2O-SiO2 glasses
in SOFC environment. However, addition of CaO in a Na2O-
SiO2 glass increases the chemical resistance to water
ublications, LLC. Published by Elsevier Ltd. All rights reserved.
Table 1 e Glass compositions (mol %) with their label.
Glass label SiO2 Al2O3 TiO2 CaO Na2O
N-25 45 10 5 15 25
N-20 50 10 5 15 20
N-15 55 10 5 15 15
N-10 60 10 5 15 10
400 600 800 1000 1200
-50
-40
-30
-20
-10
0
Tp2
Tp2
Tp2
Tp2
Tp1Tp1
Tp1
End
o D
own
V
Temperature ( C)
N-10
N-20
N-15
N-25
Tp1
Tm
Tm
Tm
Tm
Fig. 1 e DTA curves of prepared glasses at the heating rate
of 10 �C/min.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 3 8 3 9e3 8 4 73840
induced phase separation [14,15]. Moreover, Al2O3 hinders the
formation of the hydrated open network structure surface
layer in a Na2OeCaOeSiO2 glasses [16]. Addition of polar
oxides such as PbO and transition metal oxides such as ZrO2
and TiO2 in these glasses improves the chemical stability [15].
The stability of glass sealant depends on the inter-diffusion
distance at the interface of glass-SOFC components, which
should be less than 10 mm. The diffusion distance of Cr from
ferritic steel into BaO-CaO borosilicate glasses increases from
10 to 15 mm with the thermal treatment time (from 10 h to
200 h) at 800 �C [17]. Conversely, a Na2O-CaO-silicate glass
reacts with pre-oxidized AISI 430 alloy and forms Cr, Mn, Fe
and O enriched interface layer of 1e2 mm thickness in
reducing atmosphere at 800 �C after 200 h [18]. Smeacetto
et al. [19] in their recent publication have also demonstrated
that glasses containing a low amount of sodium oxide can
be used as sealing material for SOFC.
Based on these studies, soda lime aluminosilicate glass has
been chosen in the present investigation as basic glass. TiO2
which acts as nucleating agent is added to improve the
chemical stability [20]. In the present work, Na2OeCaOe
Al2O3eSiO2eTiO2 glass system has been investigated in
detail. To estimate the applicability of these glasses as
sealants, their thermal properties, crystallization behavior,
thermal expansion coefficient after heat treatment and the
overall bonding characteristics of the glass with metallic
(Crofer22APU) interconnect were investigated.
2. Materials and methods
2.1. Glass preparation
Selected glass compositions for present study with their
designationaregiveninTable1.Glassesweresynthesizedusing
conventional splat quenching technique. Theminimumpurity
of all the ingredientsused to synthesize theseglasseswas99.8%.
For the synthesis of glasses, required amount of ingredients
were taken. These ingredients were ground in ball mill (Retsch
PM100) for 2 h in dry media. The ball milled mixtures were
melted at 1500 �C in a recrystallized alumina crucible.
Table 2 e Activation energy (kJ/mol) for the glass transition an
Sample identity Mahadevan method Kis
Eg Ep1 Ep2 Eg
N-25 271.11 386.11 215.55 254.86
N-20 58.80 51.29 68.45 43.17
N-15 149.38 216.61 124.04 134.40
N-10 100.77 142.73 167.32 83.78
The glass ingots after casting were kept in a preheated
furnace at 400 �C for 12 h to remove the stresses of the
quenched glasses.
2.2. Phase identification
The amorphous nature of samples was characterized by X-ray
diffraction (XRD) using X’Pert Pro, PAN alytical model of Phi-
lips, Netherlands. During the experiment, the scan speed was
0.02�/min. The crystalline phases obtained after heat treat-
ment was identified by the Joint Committee on Powder
Diffraction Standards (JCPDS-ICDD) files. Thermal behavior
and crystallization kinetics of the glass was investigated by
Differential Thermal Analysis (DTA) (DTA-TG, Perkin Elmer,
USA) using powder samples with alumina as a reference
material. For crystallization kinetics, 15 mg of glass samples
were scanned by DTA in nitrogen at different heating rates (a)
of 5, 10, 15, 20 �C/min from50 �C to 1200 �C. Thermal expansion
coefficient of the glasses was measured using Netzsch DIL 402
PC, UK, at the heating rate of 5 �C/min in the air environment.
2.3. Activation energy calculations
The activation energy of the glass transition (Eg) and crystal-
lization (Ec) can be obtained by the following relationship [21]:
ln a ¼ � ERT
þ constant (1)
d crystallizations with various methods.
singer method Augis and Bennett method
Ep1 Ep2 Eg Ep1 Ep2
369.21 196.47 262.98 265.03 258.59
33.92 48.14 242.57 236.49 263.59
199.99 104.96 260.47 260.76 255.27
124.07 146.77 249.55 254.84 256.42
1 .0 0 1 .0 1 1 .0 2 1 .0 3 1 .0 4
1 .5
2 .0
2 .5
3 .0
-1 2 .0
-1 1 .5
-1 1 .0
-1 0 .5-5 .5
-5 .0
-4 .5
-4 .0
1 .0 0 1 .0 1 1 .0 2 1 .0 3 1 .0 4ln
()
1 0 0 0 /T g (K -1)
a
ln (
/ Tg2 )
b
ln (
/Tg)
c
Fig. 2 e Activation energy plot with (a) Mahadevan (b)
Kissinger and (c) Augis and Bennett methods for N-25
glass.
10 15 20 25
0.5
1.0
1.5
2.0
2.5
3.0
3.5
S (K
)
Na2O (mol%)
Fig. 4 e The graphical representation of S parameter with
respect to Na2O (mol %) at heating rate 10 �C/min.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 3 8 3 9e3 8 4 7 3841
where E is the activation energy and R is the gas constant.
The second approach to evaluate the activation energy of
the glass transition (Eg) and crystallization (Ec) is Kissinger
equation as given below [22]:
ln
�a
T2
�¼ � E
RTþ constant (2)
10 15 20 25
0.6
1.2
1.8
15 C/min 10 C/min5 C/min
Na O content (mol%)2
Hr
Fig. 3 e The graphical representation of Hruby parameter
with respect to Na2O (mol %).
The slope of the plot ln (a=T2) versus 1=T (K�1) gives the value
of activation energy.
Additionally, activation energy was also calculated using
the method proposed by Augis and Bennett [23]:
ln�aT
�¼ � Ec
RTpþ ln Ko (3)
The activation energy of the glass transition and the crystal-
lization processes are calculated and compared with the
values as obtained from the Kissinger and Mahadevan
models.
From the point of view of technological application, the
glass should be thermally stable. A parameter usually
employed to estimate the glass stability is the thermal
stability (DT) [21], which is defined by the following equation:
DT ¼ Tc � Tg (4)
The larger difference between Tc and Tg, the higher is the
kinetic resistance to crystallization. In other words, the glass
is thermally stable.
100 200 300 400 500 600 700 800
0.000
0.001
0.002
0.003
0.004
0.005
0.006 N -25 N-20 N-15 N-10 N -10 GC Crofer
dL
/d
Lo
X
10
-6
K
-1
Temperature ( C)
Fig. 5 e Thermal expansion curves of prepared glasses,
N-10 glass ceramic and Crofer steel.
Table 3 e Properties of glasses and steel.
Properties
Sampleidentity
DTA data Dilatometric data
Glass transitiontemperature, Tg (�C)
Crystallizationtemperature, Tp (�C)
Glass transitiontemperature, Tg (�C)
Softeningtemperature, Ts (�C)
TEC (K�1) in the rangeof 200e550 �C
N-25 692 741, 864 562 589 8.88 � 10�6
N-20 616 705, 915 572 629 8.31 � 10�6
N-15 607 711, 910 628 657 8.34 � 10�6
N-10 708 831, 965 654 709 8.19 � 10�6
Heat treated N-10 e e e e 8.27 � 10�6
Crofer 22APU e e e e 8.42 � 10�6
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 3 8 3 9e3 8 4 73842
Another thermal stability parameter, S, proposed by Saad
and Poulain [24], is given by Eq. (5)
S ¼ �Tp � Tc
��Tc � Tg
��Tg (5)
The thermal stability parameter, S, reflects the resistance to
devitrification after the formation of the glass. (Tp-Tc) is
20 40 60 80
0
400
800
Inte
nsity
(Cou
nts)
2 (Deg.)
(a)
(b)
(c)
AlNa ( SiO4 )
SiO2
Fig. 7 e XRD patterns of (a) N-10 glass, (b) N-10 heat treated
glassat 900 �Cand (c)N-10heat treatedglassat 950 �C for 1h.
10 20 30 40 50 60 70 80 90 1000
400
800
1200
1600
N-10
N-15
N-20
Inte
nsit
y (C
ount
s)
2 (Deg.)
N-25
Fig. 6 e XRD patterns of prepared glasses.
related to the rate of devitrification transformation of the
glassy phases, while a high value of (Tc-Tg) delays the nucle-
ation process.
Hurby [25] has given a factor which combines both
nucleation and growth aspects of glass transformation and
is given by
Hr ¼ Tc � Tg
Tm � Tc(6)
Where Tm is the melting temperature of glass.
Fig. 8 e (a): SEM micrograph of interface between N-10
glass and Crofer steel showing the overall view of inter-
face. (b): Back scattered electron micrograph of interface
between N-10 glass and Crofer steel.
Fig. 9 e (a): EDS spectrum analyses of the glass of the diffusion couple marked 1 in Fig. 8(b). (b): EDS spectrum analyses of the
interface of the diffusion couple marked 2 in Fig. 8(b). (c): EDS spectrum analyses of the steel of the diffusion couple marked 3
in Fig. 8(c).
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 3 8 3 9e3 8 4 7 3843
2.4. Formation of the diffusion couple
The steel plates (Crofer) were cut in 1 � 1 cm2 to prepare the
diffusion couple. This piece was degreased by ultrasonic
cleaner in acetone media. After that, these pieces were dip in
a solution of 5% HNO3 for 5 min to dissolve any oxide layer
formation on the surface. The glass coating was applied onto
the cleaned substrate surface by a slurry method. The thick-
nessof the glass coating layerwas1mm.For coating, the slurry
was prepared from the mixture of the glass powder dispersed
in 5%PVA.Thediffusioncouplewas thermally treatedat 900 �Cfor 1 h, to check adhesion of the glass with Crofer 22APU. The
heat treated diffusion couple was characterized by scanning
electron microscopy (SEM, JEOL6400, Japan) to study the
interface in conjunction with energy-dispersive X-ray spec-
troscopy (EDS). The sample was mounted in epoxy resin and
ground flat by using 240, 400, 800 and 1200 grit abrasive papers
consecutively and then polished with diamond paste of one
mm to achieve a mirror-like surface finish. The sample was
etched with 0.05 N HCl solution containing few drops of HF for
1min before an examination under SEM. In order to check any
variation in chemical composition before and after heat
treatment, N-10 glass was further subjected to heat treatment
at 900 �C for 1 h. EDS analysis on polished surface of the heat
treated glass was carried out.
3. Results and discussion
3.1. Thermal properties of glass and glasseceramic
Fig. 1 shows the DTA curves of all the glasses. DTA curves
exhibit two exothermic peaks, which indicate phase
separation in the glasses. Basically phase separation occurs
due to the formation of a second glass matrix within the
glass matrix. Higher modifier contained glasses, in general,
exhibit phase separation tendency [26]. The glass transition
temperature (Tg) of glasses is observed in the range of
600e800 �C followed by two exothermic peaks, which
belongs to crystallization of the glasses. Endothermic peaks
Fig. 10 e X-ray dot mapping of Na, Al, Si, Cr and Fe of the interface of the diffusion couple.
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 3 8 3 9e3 8 4 73844
at around 1050e1200 �C are denoted for melting of the glasses
(Fig. 1). As a modifier (Na2O) increases the Tg, Tc and Tm
decrease in the present glasses. These observations are
similar to that of reported in sodium silicate glasses [27].
Based on the DTA results, the theoretical calculations were
made using different models as described in experimental
section. Calculated activation energy of these glasses with
various models is summarized in Table 2.
Kissinger and Mahadevan methods give nearly same value
of activation energy but Augis and Bennet’s method gives
relatively large value as compared to others. The average
value of activation energy of glasses except N-20 is large as
compared to another borosilicate system [9]. From Table 2, it
can be observed that N-20 glass shows very less value of
activation energy. It is well reported in literature that with
the increase of the modifier content the network weakens,
which will decrease the value of activation energy [26]. But
the value of activation energy follows the opposite trend in
this system. This anomaly arises due to the presence of
minor amount of SiO2 (Zeolite) phase in the microcrystalline
glass matrix containing higher Na2O. The increase in Na2O
content with a decrease of SiO2 content increases the phase
separation tendency in glasses as observed in Fig. 1. It can
be explained on the basis of the diffuse and broad
exothermic/endothermic peaks in DTA curve. However, the
trend observed for activation energy of all the glasses is
same. Fig. 2 depicts such variation for N-25 glass.
Asmentioned in experimental part, the difference between
Tg and Tc also indicate the thermal stability of the glasses.
According to this approach, N-10 glass is more stable as
compared to the other glasses as is evident from Fig. 1.
Fig. 3 shows the variation in Hr parameter for the second
crystallization peak of different compositions at different
heating rates. The Hr parameter of N-10 glass is higher than
other glasses as shown in Fig. 3. In general, the modifier
weakens the glass network and decreases the glass
transition temperature so the less content of Na2O in glass
composition shows better thermal stability [28]. Similarly,
Saad and Poulain have suggested a criterion to check the
stability of the glasses. According to this approach, N-10
glass exhibit higher S value which indicates the higher
stability of this particular glass. The S parameter with
respect to the Na2O is shown in Fig. 4.
The dilatometer studywas performed to know the variation
in a thermal expansion coefficient of all the pristine glasses,
which is shown in Fig. 5. N-10 glass shows a lower thermal
expansion coefficient as compared to the other glasses. The
thermal expansion coefficient originated due to asymmetric
potential well of the solids [29]. In the present glasses, the
observed TEC decreases as the modifier content decreases
except N-20. But the increasing trend of Ts value is observed
(Table 3). It is evident in XRD pattern of pristine glass as
shown in Fig. 6. The XRD pattern of higher modifier
containing glasses exhibits a sharp hump around 30� which is
due to zeolite phase [30]. The presence of this phase may lead
to decrease the TEC of the glasses since TEC of zeolite (SiO2) is
very low (0.5 � 10�6/K). In addition to TEC measurement on
pristine glasses, TEC of steel is also measured under a similar
condition to check the compatibility between selected N-10
glass and steel. The difference in TEC values of glasses and
steel is less than 1 � 10�6 which is necessary during SOFC
operation (Table 3). During SOFC operation glass gets
Fig. 11 e (a): Back scattered electronmicrograph of the N-10
glass. (b): EDS spectrum analyses of the glass marked in
Fig. 11(a).
Fig. 12 e (a): Back scattered electronmicrograph of the N-10
heat treated glass at 900 �C for 1 h (b): EDS spectrum anal-
yses of the glasseceramic marked in Fig. 12(a).
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 3 8 3 9e3 8 4 7 3845
converted to the glass ceramics. The presence of these
crystalline phases in the glass matrix changes the TEC.
Therefore, before the interfacial study, the selected glass was
heat treated for different duration and temperature. The heat
treatment is required to facilitate the maximum
crystallization for diffusion to occur. Ultimately, this will lead
to achieve good adhesion between steel and glass. Many
researchers have reported that phase separation leads less
activation energy of the crystallization, which is essential for
good adhesion with other metallic materials [31]. The peak
crystallization temperatures Tp1 and Tp2 of N-10 glass are
observed at 780 �C and 925 �C, respectively. Based on the
analysis of crystallization kinetics of the N-10 glass, the heat
treatment temperature selected for present study was 900 �C,which is just below the second crystallization temperature of
this particular glass. The heat treated glass exhibits two
different crystalline phases, i.e. aluminum sodium silicate
(JCPDS #00-002e0625) and zeolite (JCPDS #01-073e3414) as
shown in Fig. 7. The formation of zeolite (SiO2) phase in glass
ceramic is harmful to SOFC application since it has low
thermal expansion. However, overall thermal expansion
of glass ceramic increases slightly as compared to pristine
glass. As shown in Table 3, the variation in TEC values of glass
and glass ceramic with steel is within the limit (<2%). This
indicates that the glass may be useful for interfacial study
with Crofer [32]. Based on the experimental and theoretical
results, N-10 glass was selected to make the diffusion couple
between Crofer 22APU steel and N-10 glass.
The detailed microstructural study of the interface
between N-10 glass and Crofer 22APU steel was done to get
more insight about the adhesion and interfacial mechanism.
3.2. Analysis of glassesteel interface
To study the interfacial phenomenon, glassesteel interface
was analyzed under SEM. Cross-sectional microstructure of
the interface between N-10 glass and Crofer 22APU is shown
in Fig. 8 (a & b). The interface between glass and steel shows
good adhesion (Fig. 8(a)). However, some voids and porosity
are observed at the areas away from the interface (Fig. 8(b)).
The EDS analysis at the marked area as shown in Fig. 8(b)
was done to understand the diffusion and inter-diffusion of
the elements. This is shown in Fig. 9(a)e(c) for the areas
marked as 1, 2 and 3, respectively.
As observed from the EDS analysis, the maximum diffu-
sion of Cr has taken place from the steel side. On the other
hand, only Al3þ has diffused from the glass side to the inter-
face which is very less in amount. Basically, the interface
i n t e rn a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 3 8 3 9e3 8 4 73846
contained Cr3þ rich layer. Though its diffusion is within the
permissible limit, but it has got a higher tendency as
compared to other elements. The overall analysis indicates
that layers of chromium oxide and a reaction layer of Cr
diffusion into the glass identified by EDS were found at the
interface between the glass and the steel. For confirmation of
the diffusion from steel to glass and vice-versa, X-ray
elemental dot mapping of the interface was done. These
results are shown in Fig. 10. The SEM image of the interface
provides good adhesion of glass with steel as can be seen in
Fig. 10(a). The X-ray dot mapping of Na, Al, Si, Cr and Fe are
shown in Fig. 10(bef) respectively. These results further
confirmed the diffusion of Cr3þ which is more prominent
than other elements. However, the interface is smooth and
free from porosity, which is required for good sealing.
EDS analysis of as prepared and heat treated glass are
shown in Fig. 11(a and b) and Fig. 12(a and b), respectively. The
weight percentage of Na content in both the cases is about 4%.
There is no major change in the weight percentage of sodium
content before and after thermal treatment i.e. sodium is
present in the glass network.
4. Conclusions
Differential Thermal Analysis of all the glasses showed two
crystallization peaks due to phase separation in glasses. N-10
glass exhibitmaximum thermal stability as compared to other
glass compositions.
The TEC of N-10 glass ceramic is higher than the pristine
glass.
Heat treated glass (N-10) exhibit sodium rich (aluminum
sodium silicate) and sodium free (zeolite) phases which are
not detrimental for SOFC applications.
The diffusion couple between Crofer and glass (N-10) shows
the good and smooth interface which is formedmainly by the
diffusion of Cr3þ and Al3þ ions. Naþ ions are not present at the
interface.
Acknowledgment
The authors gratefully acknowledge the financial support
provided by UGC for this research work.
r e f e r e n c e s
[1] Steele BCH, Heinzel A. Materials for fuel-cell technologies.Nature 2001;414:345e52.
[2] Yamamoto O. Solid oxide fuel cells: fundamental aspects andprospects. Electrochim Acta 2000;45:2423e35.
[3] Kim Yongmin, Hong Seong-Ahn, Nama Suk Woo, Seo Seok-Ho, Yoo Young-Sung, Lee Sang-Houck, Development of 1 kWSOFC power package for dual-fuel operation. Int J Hydrogenenergy; doi:10.1016/j.ijhydene.2010.10.056.
[4] Fergus JW. Sealants for solid oxide fuel cells. J Power Sources2005;147:46e57.
[5] MahapatraMK, LuK. Glass-based seals for solid oxide fuel andelectrolyzer cells e a review. Mater Sci Eng R 2010;67:65e85.
[6] Xie X, Gao H. Calorimetric studies on the crystallization ofLi2SeB2O3 glasses. J Non-cryst Solids 1998;240:166e76.
[7] Yin Cheng, Hanning Xiao, Wenming Guo, Weiming Guo.Structure and crystallization kinetics of PbOeB2O3 glasses.Ceramics Int 2007;33:1341e7.
[8] Meinhardt KD, Kim DS, Chou YS, Weil KS. Synthesis andproperties of a barium aluminosilicate solid oxide fuel cellglasseceramic sealant. J Power Sources 2008;182:188e96.
[9] Arora Anu, Singh Kulvir, Pandey O.P., Thermal, structuraland crystallization kinetics of SiO2eBaOeZnOeB2O3eAl2O3
glass samples as a sealant for SOFC, Int J Hydrogen Energy;doi:10.1016/j.ijhydene.2011.03.036.
[10] Vishal Kumar, Anu Arora, Pandey OP, Singh K. Studies onthermal and structural properties of glasses as sealants forsolid oxide fuel cells. Int J Hydrogen Energy 2008;33:434e8.
[11] Yeong-Shyung Chou, Stevenson Jeffry W, Prabhakar Singh.Effect of aluminizing of Cr-containing ferritic alloys on theseal strength of a novel high-temperature solid oxide fuelcell sealing glass. J Power Sources 2008;185:1001e8.
[12] Sohn SB, Choi SY, Kim GH, Song HS, Kim GD. Suitable glass-ceramic sealant for planar solid-oxide fuel cells. J Am CeramSoc 2004;87:254e60.
[13] Tomozawa M, Takata M, Acocella J, Watson EB, Takamori T.Thermal properties of Na2O$3SiO2 glasses with high watercontent. J Non-cryst Solids 1983;56:343e8.
[14] Singh P, Vora SD. Vapor phase silica transport during SOFCoperation at 1000�C. Ceram Eng Sci Proc 2005;26:99e110.
[15] Scholze H, Nature Glass. Structure and properties. New York:Springer Verlag; 1991.
[16] Wassick TA, Doremus RH, Langford WA, Burman C.Hydration of soda-lime silicate glass, effect of alumina.J Non-cryst Solids 1983;54:139e51.
[17] Ghosh S, Sharma AD, Kundu P, Basu RN. Glass-Ceramicsealants for planar IT-SOFC: a bilayered approach for joiningelectrolyte and metallic interconnect. J Electrochem Soc2008;155:B473e8.
[18] Smeacetto F, Salvo M, Ferraris M, Casalegno V, Asinari P,Chrysanthou A. Characterization and performance of glass-ceramic sealant to join metallic interconnects to YSZ andanode-supported-electrolyte in planar SOFCs. J Eur CeramSoc 2008;28:2521e7.
[19] Smeacetto F, Salvo M, D’Herin Bytner FD, Leone P, Ferraris M.New glass and glass-ceramic sealants for planar solid oxidefuel cells. J Eur Ceram Soc 2010;30:933e40.
[20] Bahadur D, Lahl N, Singh K, Singheiser L, Hilpert K. Influenceof nucleating agents on the chemical Interaction ofMgOeAl2O3eSiO2eB2O3 glass sealants with components ofSOFCs. J Electrochem Soc 2004;151:A558e62.
[21] Mahadevan S, Giridhar A, Singh AK. Calorimetricmeasurements on as-sb-se glasses. J Non-Cryst Solids 1986;88:11e34.
[22] Kissinger HE. Reaction kinetics in differential thermalanalysis. Anal Chem 1957;29:1702e6.
[23] Augis JA, Bennett JE. Calculation of the Avrami parametersfor heterogeneous solid state reactions using a modificationof the Kissinger method. J Therm Anal 1978;13:283e92.
[24] Saad M, Poulain M. Glass forming ability criterion. Mater SciForum 1987;19-20:11e8.
[25] Hruby A. Evaluation of glass-forming tendency by means ofDTA. Czech J Phys B 1972;22:1187e93.
[26] Lahl N, Singh K, Singheiser L, Hilpert K. Crystallizationkinetics in AOeAl2O3eSiO2eB2O3 glass (A¼Ba, Ca, Mg).J Mater Sci 2000;35:3089e96.
[27] Ruzha Harizanova. Volksch Gunter, Russel Christian,crystallization and microstructure of glasses in the systemNa2O/MnO/SiO2/Fe2O3. Mater Res Bull 2011;46:81e6.
[28] Ma Mingsheng, Ni Wen, Wang Yali, Wang Zhongjie,Liu Fengmei. The effect of TiO2 on phase separation and
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n en e r g y 3 7 ( 2 0 1 2 ) 3 8 3 9e3 8 4 7 3847
crystallization of glass-ceramics in CaOeMgOeAl2O3
eSiO2eNa2O system. J Non-Cryst Solids 2008;354:5395e401.[29] Raghavan V., Materials science and engineering: a first
course, 5th edn.[30] Criado M, Fernandez-Jimenez A, Torre AG, de la,
Aranda MAG, Palomo A. An XRD study of the effect of theSiO2/Na2O ratio on the alkali activation of fly ash. CementConcrete Res 2007;37:671e9.
[31] Park J, Ozturk A. Effect of TiO2 addition on the crystallizationand tribological properties of MgOeCaOeSiO2eP2O5eFglasses. Thermochim Acta 2008;470:60e6.
[32] Ghosh S, Das Sharma A, Kundu P, Mahanty S, Basu RN.Development and characterizations ofBaOeCaOeAl2O3eSiO2 glasseceramic sealants forintermediate temperature solid oxide fuel cell application.J Non-Cryst Solids 2008;354:4081e8.