Project 2 Paper 1 Thakur

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Transformation kinetics for mullite in kaolin–Al2O3 ceramics

Downloaded from https://www.cambridge.org/core. University of Arizona, on 02 Apr 2020 at 19:02:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1557/JMR.2003.0186

Yung-Feng Chen
Department of Materials Science and Engineering, National Cheng Kung University,
1 Ta-Hsueh Road, Tainan 70101, Taiwan
Moo-Chin Wang
Department of Mechanical Engineering, National Kaohsiung University of Applied Sciences,
415 Chien-Kung Road, Kaohsiung 80782, Taiwan
Min-Hsiung Hon
Department of Materials Science and Engineering, National Cheng Kung University,
1 Ta-Hsueh Road, Tainan 70101, Taiwan

(Received 9 December 2002; accepted 4 March 2003)

Transformation kinetics of mullite formation in kaolin–Al2O3 ceramics was studied


by x-ray diffraction, transmission electron microscopy, and energy dispersion
spectrometry. The mullitization process of kaolin–Al2O3 ceramics is described by two
stages; one is the primary mullite transformation at 1273 to 1573 K, and the other is the
secondary mullite formation at 1573 to 1873 K. The activation energy of 1164.6 kJ mol−1
obtained for the secondary mullite formation is lower than 1356.9 kJ mol−1 for the
primary mullite transformation by the general form of the Johnson–Mehl–Avrami
equation. The lower value of growth morphology parameter strongly supports that
in the secondary mullite formation the added alumina is dissolved into glassy phase
and the mullite is then precipitated.

I. INTRODUCTION that is the rate-limited factor, is consistent with those


In recent years, mullite (3Al2O3 · 2SiO2) has been calculated using Aksay’s diffusion coefficient equa-
considered as a potentially important ceramic due to its tion.7,8 The rate of the mullite formation is determined by
low thermal expansion, excellent creep resistance, good the average diffusion coefficient of the slower cation Si4+
chemical and thermal stabilities, very low dielectric con- rather than the faster cation Al3+.
stant, and good optical properties. A variety of prepara- For the reactants homogeneously distributed with a
tion methods are used to synthesize mullite, such as shorter diffusion path, the mullite formation is nucleation
conventional mixing, precipitation, hydrolysis, hydro- rate-controlled from the monophasic gels with appar-
thermal synthesis, spray pyrolysis, and chemical vapor ent activation energy around 300 kJ mol−1.6,9,10 For the
deposition, in which various starting materials such as diffusion-limited powders7,8 and diphasic gels, a mass
alumina–silica minerals, hydroxides, sols, silicon alkox- transport of alumina and silica along the mullite grain
ide, aluminum alkoxide, etc., are used.1,2 interface11–13 or via nucleation and growth mechanism
A number of factors affect the chemical reaction for was proposed with a higher activation energy.14,15
mullite formation, such as particle size and the crystalline The activation energies of mullite formation corre-
form of the precursors, Al/Si ratio, degree of mixing, and sponding to different crystallization routes from solid
impurities.3 However, the diffusion for species plays reaction, diphasic gel, and monophasic gel were exam-
an important role on the mullite phase transformation for ined using the conventional isothermal3–7,9,12,15,16 or
nucleation and grain growth. Most of the crystallization nonisothermal method.11,12,14 The activation energy of
kinetics for mullite formation shows a multiple stage of mullite formation in solid reaction or diphasic gel is around
nucleation, nucleation–growth, and coalescence.4–6 1000 kJ mol−1, which is higher than that in the mono-
A possible mechanism proposed for mullite formation phasic gel system, namely 362 kJ mol−1.4,6,7–10,12,14–16
from raw materials is dominated by the diffusion of the However, the phase-transformation kinetics of kaolin–
different cations, Al3+ and Si4+.7 During the Al2O3 and Al2O3 ceramics has not been studied in detail.
SiO2 solid reaction, the counterdiffusion of Al3+ and Si4+ In this study, the phase transformation kinetics of
ions through the mullite interlayer determines the reac- kaolin–Al2O3 ceramics was investigated using x-ray
tion rate. The average diffusion coefficient value of Si4+, diffraction (XRD), transmission electron microscopy

J. Mater. Res., Vol. 18, No. 6, Jun 2003 © 2003 Materials Research Society 1355
Y-F. Chen et al.: Transformation kinetics for mullite in kaolin–Al2O3 ceramics

(TEM), and energy dispersion spectrometry (EDS). The The integrated intensity of the (121) reflection (2␪
objective of this study is to determine the phases evolu- from 40.5 to 41.5°) of the commercial mullite powder
Downloaded from https://www.cambridge.org/core. University of Arizona, on 02 Apr 2020 at 19:02:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1557/JMR.2003.0186

tion, the kinetics of phase transformation, and the micro- (Mullite, 98%, Alfa Aesar, Johnson Matthey Catalog
structure in kaolin–Al2O3 ceramics. Co., Inc., MA), which was leached with 20 wt.% hydro-
fluoric (HF) acid at room temperature for 12 h to remove
the impurities and calcined at 1873 K for 24 h, was used
II. EXPERIMENTAL as the standard for determining the relative phase content
A. Preparation of mullite. The integrated intensity of the reflection (113)
The raw material powders of kaolin and Al2O3 were (2␪ from 43° to 44°) of the ␣–Al2O3 powder was used as
supplied by Akima Co. (Berhad, Malaysia) (Ipoh, Ma- the standard for evaluating the relative phase content of
laysia, AKIMA-25, average size 1.3 ␮m) and Sumitomo ␣–Al2O3 in the sintered ceramics.
Chemical (Tokyo, Japan, AM-21, average size 4 ␮m), The morphology of the sintered samples and the
respectively, with their chemical compositions and prop- chemical compositions of glassy phase were observed
erties listed in Table I. In this study, the kaolin–Al2O3 and analyzed with TEM (Hitachi model HF-2000 field
ceramics contained 60 wt.% kaolin and 40 wt.% emission transmission electron microscope, Hitachi Ltd.,
Al2O3 powders added with 28.3 wt.% multiple-organics Tokyo, Japan, operated at 200 kV) and EDS (Noran
solution as binder, which consisted of 20.3 wt.% PEG- model Voyager1000, Noran Ltd., WI). The foils for
10k, 5.1 wt.% glycerin, 30.4 wt.% selosol 920, and TEM were prepared by slicing to a thickness of about
5.1 wt.% Celuna D-305 dissolved in 39.1 wt.% deionized 30 ␮m mechanically and ion-beam thinning to electron
water. A batch (1.0 kg) of admixed raw materials added transparency.
with multiple-organics solution was mixed by using a
III. RESULTS AND DISCUSSION
double Z-blade at 298 K for 30 min; subsequently, the
resulting feedstock was extrusion molded under a pres- A. Phases in sintered ceramics
sure of 18 MPa by a single-screw extruder at a tempera- The XRD pattern of the kaolin–Al2O3 samples sin-
ture of 298 K to obtain a pipe sample with radius, wall tered at 1573 K for 30 min is shown in Fig. 1(a). It was
thickness, and length of 3, 0.5, and 10 cm, respectively. found that the sintered kaolin–Al2O3 ceramics contain
In the process of thermal debinding, the water was ␣–Al2O3 and mullite as the major phases and cristobalite
preferentially extracted at 383 K for 30 min, and glyc- as the minor phase. Mullite and cristobalite, liberated as
erin, selosol 920, and celuna D-305 were extracted at a result of decomposition of kaolin, are converted above
about 443 K for 60 min. The samples were then heated to 1373 and 1473 K, respectively.17–19 Since the reaction
748 K at a heating rate of 1 K min−1 and held for 30 min between kaolin and the added Al2O3 powder is limited at
for PEG-10k debinding. Finally, the debinded samples
were sintered at 1273 to 1873 K for various periods of
time at a heating rate of 5 K min−1.

B. Characterization
The phases in sintered kaolin–Al2O3 samples were
identified by XRD with Cu K␣ radiation and a Ni fil-
ter, operated at 30 kV, 20 mA, and a scanning rate of
0.25° min−1 (model XD-D1, Shimadzu, Kyoto, Japan).

TABLE I. Chemical compositions of kaolin and Al2O3 powders.

Chemical composition (wt.%) Kaolin Al2O3


SiO2 54.0 0.02
Al2O3 32.0 99.9
Fe2O3 0.98 0.02
Na2O 0.25
TiO2 0.45
CaO 0.06
K2O 1.65
MgO 0.30
LOIa 10.31 0.1
Average particle size (␮m) 1.3 4
FIG. 1. XRD patterns of kaolin–Al2O3 ceramics sintered for 30 min at
a
LOI, loss on ignition. (a) 1573 K and (b) 1873 K (s, cristobalite; a, ␣–Al2O3; m, mullite).

1356 J. Mater. Res., Vol. 18, No. 6, Jun 2003


Y-F. Chen et al.: Transformation kinetics for mullite in kaolin–Al2O3 ceramics

a low temperature of 1573 K, the XRD pattern shows the of the mullite phase is reasonably sharp without anoma-
unreacted ␣–Al2O3 phase. Figure 1(b) shows the XRD lous intensity change for the XRD reflection peak (121).
Downloaded from https://www.cambridge.org/core. University of Arizona, on 02 Apr 2020 at 19:02:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1557/JMR.2003.0186

pattern of the kaolin–Al2O3 sample sintered at 1873 K The relative amount of mullite present in the sintered
for 30 min. It indicates that the XRD reflection intensity kaolin–Al2O3 sample was determined by comparing the
of mullite increases but the ␣–Al2O3 decreases by the integral intensity of the reflection (121) of the mullite
formation of mullite and dissolution of Al2O3 powder at with the standard.
the sintering temperature. Besides, the cristobalite phase Figure 2 shows the relative content of mullite in the
disappears, dissolves into an amorphous glassy phase, sintered kaolin–Al2O3 samples. It indicates that the rela-
and improves the mullite formation at 1873 K. tive content of mullite increased as the sintering tempera-
The transformation of kaolinite to mullite and glassy ture increased from 1273 to 1873 K. Besides, when the
phase has been reported by Brindly and Nakahira.17–19 sintering was conducted at a fixed temperature in this
When sintering temperature is at 1273 K, the transfor- range, the mullite content increased as the sintering time
mation from kaolin to mullite occurs according to the increased. When the samples were sintered at 1873 K for
following reactions: 5 h, about 88 wt.% of mullite was obtained. However,
673–773 K
when the sintering was conducted at 1273 K for 30 min,
Al2O3 ⭈ 2SiO2 ⭈ 2H2O (kaolinite) → there was only a small amount of mullite in sintered
Al2O3 ⭈ 2SiO2 (metakaolinite) + 2H2O , (1) samples due to the low sintering temperature.17–19,21,22
Figure 2 also indicates the formation behavior of mul-
∼1253 K
2(Al2O3 ⭈ 2SiO2) (metakaolinite) → lite. The amounts of mullite in the sintered ceramics in-
Si3Al4O12 (spinel) + SiO2 (amorphous) , (2) creased slightly when sintering temperature increased
from 1473 to 1573 K due to the nearly complete trans-
>1273 K formation of kaolin. As the samples were sintered at
3Si3Al4O12 (spinel) → 2(3Al2O3 ⭈ 2SiO2) 1673 and 1773 K for various periods of time, the mullite
(mullite) + 5SiO2 (amorphous) , (3) content in sintered ceramics suddenly increased by the
>1473 K dissolution of added Al2O3 into the glassy phase and
SiO2 (amorphous) → SiO2 (cristobalite) . (4) the secondary mullite precipitated.
The dissolution of added ␣–Al2O3, at different sinter-
The phase transformation and growth of mullite in kaolin
ing temperatures and for different lengths of time, in
ceramics by the nonisothermal method and TEM analysis
kaolin–Al2O3 ceramics is shown in Fig. 3. It reveals that
also were reported by Chen et al.20 The bulk nucleation
the amount of ␣–Al2O3 content decreased from 39.9 to
is dominant in mullite crystallization and the crystal
36.8 wt.% when the samples were sintered at 1573 K for
growth is controlled by diffusion in kaolin. Chen et al.21
15 to 300 min. However, the ␣–Al2O3 content decreased
also have pointed out that, in the kaolin–Al2O3 ceramics,
when the sintering temperature was greater than 1573 K.
the secondary mullite formed by solution–precipitation is
As the samples were sintered at 1873 K for greater than
mainly surrounded by the glassy phase and shows a
platelike morphology when the sintered temperature is
greater then 1573 K. The Al2O3 content in the secondary
mullite crystals increased from 59.69 to 70.41 wt.%
when the grain width increased from 15 to 40 nm, but the
lattice parameters in the orthorhombic structure of the a,
b, and c axes decreased from 8.653, 8.770, and 3.167 to
7.958, 8.064, and 2.926 Å, respectively.
In the kaolin–Al2O3–SiO2 system, Chen et al.22
pointed out that the glassy phase and SiO2 content de-
creased when sintering temperature was above 1573 K. A
similar result was obtained in this study. The mullite in
the orthorhombic structure was obtained, as indicated by
the splitting of (120) and (210) reflections at 2␪ ⳱ 25.97
and 26.27°, respectively, in Fig. 1(b), which was also
demonstrated by Li et al.23

B. Quantitative phase analysis


The commercial mullite powder was leached in
20 wt.% HF acid at room temperature for 12 h and cal- FIG. 2. Relative content of mullite in kaolin–Al2O3 ceramics as a
cined at 1873 K for longer than 24 h. The XRD pattern function of sintering temperature and time.

J. Mater. Res., Vol. 18, No. 6, Jun 2003 1357


Y-F. Chen et al.: Transformation kinetics for mullite in kaolin–Al2O3 ceramics

30 min, the ␣–Al2O3 content was below 4.4 wt.%, at


which temperature the added Al2O3 powder dissolved k = A exp − 冉 冊 Ea
, (6)
Downloaded from https://www.cambridge.org/core. University of Arizona, on 02 Apr 2020 at 19:02:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1557/JMR.2003.0186

RT
almost completely into the glassy phase. According to
Figs. 2 and 3, it indicates that the reaction between kaolin where Ea is the activation energy, R is the gas constant,
and the added Al2O3 powders is limited as the samples T is the absolute temperature, and A is a constant.
were sintered at 1273 to 1573 K and leads to only the The integrated form of Eq. (5) can be expressed by the
primary mullite formation.21,22,24–26 Johnson–Mehl–Avrami (JMA) equation28,29

冉 冊
According to the above discussion, the transformation
of kaolin–Al2O3 ceramics can be described by two 1
ln = 共kt兲n . (7)
stages; (i) the primary mullite transformation at 1273 to 1−x
1573 K, and (ii) the secondary mullite formation at tem-
peratures from 1573 to 1873 K. Equation (8) is obtained by taking the logarithms for both
sides of Eq. (7)
C. Kinetics of phase transformation in
kaolin–Al2O3 ceramics
Figure 4 shows the relationship between relative con-
冋 冉 冊册
ln ln
1
1−x
= n ln k + n ln t . (8)

tent of mullite and sintering time for the kaolin–Al2O3


ceramics by isothermal treatment at various tempera- When the left side of Eq. (8) ln{ln[1/(1 − x)]} is plotted
tures. It reveals that the mullite content in the sintered against ln t, as shown in Fig. 5, the straight line slope n
ceramics increased with the sintering temperature in- indicates the growth morphology parameter of the trans-
creasing from 1273 to 1873 K. When kaolin–Al2O3 forming phase. By Eq. (8) and Fig. 5, the growth mor-
samples were sintered at 1573 K for 7 h, the mullite phology parameter n for kaolin–Al2O3 ceramics is
content in the sintered samples was only 36.1 wt.%. This obtained and listed in Table II. Figure 5 also provides the
value was smaller than 50 wt.% for kaolin sintered at kinetic constant data k by the interception of the straight
1573 K for 30 min20 due to the hindrance caused by the line with the axis of ln{ln[1/(1 − x)]}.
added Al2O3 powder. Equation (9) is obtained by taking the logarithms for
To obtain the kinetics of phase transformation of both sides of Eq. (6)
kaolin–Al2O3 ceramics, the following rate equation is Ea
assumed:27 ln k = ln A − . (9)
RT
dx
= kntn−1 共1 − x兲 . (5) When the left side of ln k in Eq. (9) is plotted against the
dt
reciprocal of sintering temperature, as shown in Fig. 6, a
Here x is the mullite phase fraction at sintering time t, n
is the growth morphology parameter, and k is the rate
constant related to an Arrhenius-type equation:

FIG. 3. Relative content of the ␣–Al2O3 in kaolin–Al2O3 ceramics as FIG. 4. Relative content of mullite in kaolin–Al2O3 ceramics as a
a function of sintering temperature and time. function of sintering time and temperature.

1358 J. Mater. Res., Vol. 18, No. 6, Jun 2003


Y-F. Chen et al.: Transformation kinetics for mullite in kaolin–Al2O3 ceramics

straight line is obtained, and then the apparent activation obtained from Eq. (11) by taking the double logarithm of
energy can be calculated from the slope, as is listed in both sides and rearranging the terms of the equation:
Downloaded from https://www.cambridge.org/core. University of Arizona, on 02 Apr 2020 at 19:02:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1557/JMR.2003.0186

Table III.
The general form of the JMA equation can be as-
sumed by30,31
ln t =
EI + nEG 1
+
1
共n + 1兲R T n + 1
ln 冋
共n + 1兲 ln共1 − xt兲
gIoGon
册 . (12)

冋兰 册
Figure 7 shows the relationship between the logarithm
t=t
xt = 1 − exp − gIGn 共t − t⬘兲n dt⬘ , (10) of time required to reach 20 and 45 wt.% conversion of
t=0
mullite and the reciprocal of sintering temperature for the
kaolin–Al2O3 ceramics, respectively. If the nucleation
where xt is the degree of conversion for a given sintering
and growth of crystal begin at the time when the crys-
time, g is a geometrical factor, I denotes the rate of
tal is first detected by XRD,31 then xt ⳱ 0 at this time.
nucleation, G is the growth rate, and n is an integer which
There is an error involved in making this assumption
depends on the dimensionality of the growth mechanism.
since the diffractometer cannot accurately detect frac-
If the nucleation and growth rates depend on the tem-
tion of phase below approximately 3 wt.%. The slopes
perature but the growth rate is assumed to be independent
(EI + nEG)/[(n + 1)R] of the lines are obtained from
of time, Eq. (10) can be expressed as29,32
Fig. 7, and the activation energy is determined by general
1
1 − xt
= exp
1

g I G n t n+1 e−共EI+EG兲 Ⲑ RT
n+1 o o 册 , (11)
form of the JMA equation, as listed in Table III. These
values are greater than 531 kJ mol−1 as obtained by
Gualtieri et al.33
where EI and EG are the activation energies for nuclea- In Table III, it is found that the activation energy for
tion and crystal growth, respectively. Equation (12) is the primary mullite formation is higher than that for the
secondary mullite formation. The formation of mullite
from various raw materials directly depends on the de-
gree of mixing of Al3+, Si4+, and O2− components in
the raw materials7,8 (i.e., the chemical homogeneity
of the raw materials plays a key role in the mullite

FIG. 5. Relationship between ln{ln[1/(1 − x)]} and ln t of kaolin–


Al2O3 ceramics.

TABLE II. Values of the growth morphology parameter for kaolin–


Al2O3 ceramic sintered at various temperatures. FIG. 6. Plots of ln k versus 1/T for kaolin–Al2O3 ceramics.

Sintering temp. Growth morphology Average growth


(K) parameter, n morphology parameter
TABLE III. Activation energies (kJ mol−1) at various stages of mullite
1373
1473
1573
2.1
1.9
1.7
冎 1.9
(Primary mullite)
formation in kaolin–Al2O3 ceramics.

General form of JMA equation


1673
1773
1873
1.5
1.4
1.2
冎 1.4
(Secondary mullite)
JMA equation
609.6
Primary
1356.9
Secondary
1164.6

J. Mater. Res., Vol. 18, No. 6, Jun 2003 1359


Y-F. Chen et al.: Transformation kinetics for mullite in kaolin–Al2O3 ceramics
Downloaded from https://www.cambridge.org/core. University of Arizona, on 02 Apr 2020 at 19:02:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1557/JMR.2003.0186

FIG. 8. TEM morphology of kaolin–Al2O3 ceramics during sintering


at 1673 K for 30 min.

FIG. 7. Apparent activation energy for mullite transformation in crystal in liquid-phase sintering. In this study, the growth
kaolin–Al2O3 ceramics. morphology parameter n is close to two, indicating the
platelike precipitation. The growth morphology param-
formation). The natural minerals, such as kaolinite, ky- eter decreasing to 1.2 indicates that needlelike precipita-
anite, sillimanite, and topaz, etc., have a high degree of tion occurred as sintering temperature was increased.
mixing, a three-dimensional random network with infi- This result indicates that the mullite crystal grows to a
nite mixing of Al2O3 and SiO2 at the atomic level of high aspect ratio in kaolin transformation or liquid-phase
Al3+, Si4+, and O2− components.33 On the other hand, the sintering.17–19 The diffusion is necessary for grain
mechanical mixtures of solid particles exhibit a very low growth of the nuclei that the diffusion coefficient of ions
degree of homogeneity. So the unreacted Al2O3 powder in liquid-phase sintering will be the important factor in
dispersed in the kaolin–Al2O3 ceramics reduces the ho- mullite formation. Gualtieri et al.33 have pointed out that,
mogeneity; as a result, the activation energy for the pri- in kaolin–mullite reaction sequence, the growth morphol-
mary mullite formation is higher than the one for the ogy parameter decreased with the increasing sintering
secondary mullite formation. In this study, the activation temperature and that the lower value of n indicates that
energy of mullite formation obtained by general form of the chemical diffusion plays a non-negligible role in mul-
the JMA equation is also higher than that for the kaolin lite crystallization. This result can also be related to a
ceramics.20 Figure 3 shows ␣–Al2O3 content decreased change in the rate limiting mechanism of the reaction as
as the samples are sintered at secondary mullite forma- a function of temperature. The secondary mullite forma-
tion stage; that will increase the source of Al3+ ion in tion shows a value of n lower than two, which can be
glass for the mullite formation. Since the diffusion coef- related to a different crystallization mechanism.
ficient of Al3+ is greater than that of Si4+ in the glassy Figure 9 shows the TEM microstructure and EDS
phase,6,15 the activation energy of secondary mullite for- analysis of the kaolin–Al2O3 samples sintered at 1573
mation is lower than the primary one. and 1673 K for 30 min. From Fig. 9(a), it can be found
that the interface between Al2O3 particle and glassy
D. Glassy phase formation and microstructure in phase is still smooth and distinct because the added
kaolin–Al2O3 ceramics Al2O3 at 1573 K does not dissolve into the glassy phase.
The TEM micrograph of kaolin–Al2O3 ceramic sin- Figure 9(b) shows that the added Al2O3 has already dis-
tered at 1673 K for 30 min is shown in Fig. 8. It is found solved into the glassy phase at 1673 K, causing the
that these platelike mullite crystals have a broad size blurred interface to appear. The dissolution of Al2O3 into
distribution of 80–100 nm in length and 50–60 nm in the glassy phase is similar to the Al2O3–SiO2 powder
width. Sainz et al.26 have pointed out that the secondary mixture system,34 in which the curved surfaces of alu-
mullite formation took place when sintering temperature mina grains start to develop low-energy facets, indicating
was greater than 1673 K and the average grain size dis- the dissolution of Al2O3 into the siliceous phase.
tribution and the aspect ratio were a function of tempera- In Fig. 9, location a is the Al2O3 grain, locations b and
ture. The grain growth for both primary and secondary c are on the interface, and locations d and e are in the
mullites shows a broad size distribution of platelike glassy phase region. The Al and Si contents in Fig. 9(a),

1360 J. Mater. Res., Vol. 18, No. 6, Jun 2003


Y-F. Chen et al.: Transformation kinetics for mullite in kaolin–Al2O3 ceramics

at different locations, are listed in Table IV. It indicates


that the Al cation content suddenly decreases from 97.5
Downloaded from https://www.cambridge.org/core. University of Arizona, on 02 Apr 2020 at 19:02:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1557/JMR.2003.0186

to 51.3 at.% as location b switches to location c. How-


ever, when the location varies from d to e, the Al cation
content maintains at 59.5 to 52.7 at.%. This result reveals
that the added Al2O3 still does not react with the kaolin
at the interface.
Figure 9(b) shows that the interface between Al2O3
and glassy phase is not distinct. The Al cations content in
Fig. 9(b) remains at about 88.4 to 90.3 at.% when the
location varies from b to c, as listed in Table IV. How-
ever, when the location varies from d to e in Fig. 9(b), the
Al cation content decreases from 83.2 to 48.1 at.%. This
indicates the interaction of added Al2O3 with the glassy
phase at the interface. Besides melting, the primary mul-
lite rearrangement is aided by dissolution of the Al2O3 in
glassy phase. The preferential dissolution in glassy phase
of Al2O3 powders at the contact leads to the simultaneous
diffusion of the Al3+ and Si4+ ions.
The average chemical compositions of glassy phase
for kaolin–Al 2 O 3 ceramics sintered at 1573 and
1673 K for 30 min are listed in Table V. The mullite
formation will consume the Al3+ component more than
Si4+ when the temperature increases. According to
Figs. 2 and 3, the mullite increases 3.2 wt.% as sintering
temperature increases from 1573 to 1673 K for 30 min.
Although the Al2O3 content decreases 2.3 wt.% in
theory, it decreases 12.7 wt.% when sintering tempera-
ture increases from 1573 to 1673 K. It indicates that the
Al2O3 component does dissolve into the glassy phase.
Therefore, the Al at.% in glassy phase for kaolin–Al2O3
ceramics sintered at 1673 K for 30 min is higher than the
FIG. 9. TEM microstructure and EDS analysis of the kaolin–Al2O3 Al at.% sintered at 1573 K. The transitory liquid phase
samples sintered for 30 min at (a) 1573 K and (b) 1673 K. for secondary mullite formation may result either from
the presence of impurities in the kaolin raw materials
such as K2O or from a metastable eutectic composition
TABLE IV. Al and Si contents in the kaolin–Al2O3 ceramics at the for silica–alumina–alkali metal oxides (or transitory
different analysis locations in Fig. 9. metal oxide).35 In this study, there is about 1.65 wt.%
Locations and content (at.%) K2O in the raw materials of kaolin (Table I), and from
Sintering
the TEM–EDS analysis the glassy phase contents are
temp. (K) Element a b c d e
about 0.3 to 0.4 wt.%, as shown in Table IV. The tran-
Al 99.3 97.5 51.3 59.5 52.7 sitory metal oxide forms the liquid phase to promote
1573 Si 0.7 2.5 48.7 40.5 47.3
the mullite formation during the sintering process, and the
Al 100 88.4 90.3 83.2 48.1
1673 Si 0 11.6 9.7 16.8 51.9 mullite crystal grows from the platelike to needlelike
precipitation.

IV. CONCLUSION
TABLE V. Average chemical compositions of glassy phase for Transformation kinetics of mullite formation in
kaolin–Al2O3 ceramics as sintered at 1573 and 1673 K for 30 min.
kaolin–Al2O3 ceramics were studied using XRD, TEM,
Sintering
Chemical composition (at.%) and EDS. The mullitization process of kaolin–Al2O3 ce-
temp. (K) Aluminum Silicon Potassium ramics could be described by two stages; the primary
mullite transforms at temperatures from 1273 to 1573 K,
1573 33.5 66.1 0.4
1673 51.3 48.2 0.5
and the secondary mullite forms at temperatures from
1573 to 1873 K. The unreacted Al2O3 powder dispersed

J. Mater. Res., Vol. 18, No. 6, Jun 2003 1361


Y-F. Chen et al.: Transformation kinetics for mullite in kaolin–Al2O3 ceramics

in the kaolin–Al2O3 ceramics reduces the homogeneity 12. S.H. Hong and G.L. Messing, J. Am. Ceram. Soc. 80, 1551
and increases the activation energy of the first stage. The (1997).
Downloaded from https://www.cambridge.org/core. University of Arizona, on 02 Apr 2020 at 19:02:27, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1557/JMR.2003.0186

13. S.H Hong and G.L. Messing, J. Eur. Ceram. Soc. 19, 521 (1999).
lower value of growth morphology parameter for sec- 14. A.R. Boccaccini, T.K. Khalil, and M. Bücker, Mater. Lett. 38, 116
ondary mullite formation strongly supports that the (1999).
added alumina dissolved into glassy phase and precipi- 15. W.C. Wei and J.W. Halloran, J. Am. Ceram. Soc. 71, 581 (1998).
tated the mullite. The activation energy of the secondary 16. O. Matsuda, T. Watari, T. Torikai, Y. Yamasaki, and H. Katsuki,
mullite formation (1164.6 kJ mol−1) is lower than that of J. Ceram. Soc. Jpn. Int. Ed. 100, 719 (1992).
17. G.W. Brindly and M. Nakahira, J. Am. Ceram. Soc. 42, 311
the primary mullite formation (1356.9 kJ mol−1), by the (1959).
general form of the JMA equation. 18. G.W. Brindly and M. Nakahira, J. Am. Ceram. Soc. 42, 315
(1959).
19. G.W. Brindly and M. Nakahira, J. Am. Ceram. Soc. 40, 346
ACKNOWLEDGMENTS (1957).
20. Y.F. Chen, M.C. Wang, and M.H. Hon, J. Eur. Ceram. Soc (in
The financial support from the National Science Coun- press).
cil of Taiwan, Republic of China, as Grant No. 89-2216- 21. Y.F. Chen, M.C. Wang, and M.H. Hon, J. Mater. Res. (in press).
E-006-072 is appreciated. The authors sincerely thank 22. Y.F. Chen, M.C. Wang, and M.H. Hon, J. Ceram. Soc. Jpn. (sub-
Mr. H.Y. Yao for the assistance with the TEM–EDS mitted for publication).
analysis. 23. D.X. Li and W.J. Thomson, J. Mater. Res. 6, 819 (1991).
24. K.C Liu, G. Thomas, A. Caballero, J.S. Moya, and S.D. Aza, Acta
Metall. Mater. 42, 489 (1994).
REFERENCES 25. K. Hamano, H. Nakajima, F. Okuda, and M. Konuma, J. Ceram.
Soc. Jpn. Int. Ed. 102, 80 (1994).
1. I.A. Aksay, D.M. Dabbs, and M. Sarikaya, J. Am. Ceram. Soc. 74, 26. M.A. Sainz, F.J. Serrano, J.M. Amigo, J. Bastida, and
2343 (1991). A. Caballero, J. Eur. Ceram. Soc. 20, 403 (2000).
2. K. Okada and N. Otsuka, Ceram. Bull. Jpn. 70(10), 1633 (1991). 27. A. Marotta, A. Buri, and G.L. Valenti, J. Mater. Sci. 13, 2483
3. J.S. Lee and S.C. Yu, Mater. Res. Bull. 27(10), 405 (1992). (1978).
4. T. Takei, Y. Kamesshima, A. Yasumori, and K. Okada, J. Am. 28. M. Avrami, J. Chem. Phys. 7, 1103 (1939).
Ceram. Soc. 82, 2876 (1999). 29. S.B. Wen, N.C. Wu, S. Yand, and M.C. Wang, J. Mater. Res. 14,
5. B.O. Hildmann, H. Schneider, and M. Schmücker, J. Eur. Ceram. 3559 (1999).
Soc. 16, 287 (1996). 30. M. Avrami, J. Chem. Phys. 9, 177 (1941).
6. D.X. Li and W.J. Thomson, J. Am. Ceram. Soc. 73, 964 (1990). 31. S.W. Freiman and L.L. Hench, J. Am. Ceram. Soc. 51, 382
7. Y.M. Sung, Acta Mater. 48, 2157 (2000). (1987).
8. Y.M. Sung, J. Mater. Sci. Lett. 20, 1433 (2001). 32. M.C. Wang and M.H. Hon, J. Ceram. Soc. Jpn. 100, 1285 (1992).
9. E. Tkalcec, R. Nass, J. Schmauch, H. Schmidt, S. Kurajica, 33. A. Gualtieri, M. Bellotto, G. Aetioli, and S.M. Clark, Phys. Chem.
A. Bezjak, and H. Ivankovic, J. Non-Cryst. Solids 223, 57 (1998). Miner. 22, 215 (1995).
10. S. Sundaresan and I.A. Aksay, J. Am. Ceram. Soc. 74, 2388 34. H.J. Kleebe, F. Siegelin, T. Straubinger, and G. Ziegler, J. Eur.
(1991). Ceram. Soc. 21, 2521 (2001).
11. A.L. Campos, N.T. Silva, F.C.L. Melo, M.A.S. Oliveria, and 35. H. Schneider, K. Okada, and J.A. Pask, Mullite and Mullite Ce-
G.P. Thim, J. Non-Cryst. Solids. 304, 19 (2002). ramics (Wiley, New York, 1994), p. 100.

1362 J. Mater. Res., Vol. 18, No. 6, Jun 2003

You might also like