Direct evidence of temperature variation within ceramic powder compact during pulse electric current sintering

Article abstract of DOI:10.1111/j.1551-2916.2005.00722.x

Pulse electric current sintering (PECS) is a powerful technique for the preparation of nanoceramics. However, the temperature distribution within the ceramic powder compact during PECS is not uniform. In the present study, aluminum hydroxide powder is used as an in situ temperature indicator to determine the temperature uniformity. The phase evolution within the powder compact is taken to estimate its temperature distribution. The temperature is highest near the top surface of the compact; it then reduces with increasing distance away from the top surface of the compact. The temperature variation can be significantly reduced by inserting a carbon paper in between graphite punches and graphite mold and also by reducing the heating rate.

Full text of DOI:10.1111/j.1551-2916.2005.00722.x

J. Am. Ceram. Soc., 89 [2] 680–683 (2006)
DOI: 10.1111/j.1551-2916.2005.00722.x
r 2005 The American Ceramic Society
ournal
J
Direct Evidence of Temperature Variation Within Ceramic Powder
Compact During Pulse Electric Current Sintering
Dongming Zhang, Lianmeng Zhang, and Jingkun Guo
State Key Laboratory of Advanced Technology for Materials Synthesis and Processing, Wuhan University of
Technology, Wuhan, China
Wei-Hsing Tuan*,w
Department of Materials Science and Engineering, National Taiwan University, Taipei, Taiwan
Pulse electric current sintering (PECS) is a powerful technique
for the preparation of nanoceramics. However, the temperature
distribution within the ceramic powder compact during PECS is
not uniform. In the present study, aluminum hydroxide powder
is used as an in situ temperature indicator to determine the tem-
perature uniformity. The phase evolution within the powder
compact is taken to estimate its temperature distribution. The
temperature is highest near the top surface of the compact; it
then reduces with increasing distance away from the top surface
of the compact. The temperature variation can be significantly
reduced by inserting a carbon paper in between graphite punches
and graphite mold and also by reducing the heating rate.
study suggested that the coarsening behavior during PECS is
more complicated than expected.9 Zhou et al. indicated that a
higher heating rate induces a declining grain coarsening rate
throughout the sintering process.
More importantly, several previous studies indicated that the
microstructure of the specimens prepared by PECS is not uni-
form.6–8 Although the microstructure uniformity could be
improved by increasing the holding time at the peak
temperature,6,7 the trade-off would be the loss of ability of the
technique to produce nanostructures. The microstructural het-
erogeneity has been related to differential sintering induced dur-
ing PECS.8 The differential sintering is highly suspect because of
the nonuniform temperature distribution within the powder
compact. Nevertheless, direct evidence on temperature varia-
tion within the powder compact is still not yet available. The
microstructure uniformity is one of the most important criteria
to evaluate the potential of the PECS technique, and therefore,
determination of the detailed mechanism for the microstructural
inhomogeneity is essential.
In the present study, an in situ temperature indicator is used
to investigate the temperature distribution within the powder
compact. Aluminum hydroxide, which undergoes a series of
temperature-dependent phase transformations, is introduced
into the ceramic compact and treated as a temperature indica-
tor; the temperature distribution within the compact during
PECS can thus be estimated.
I. Introduction
ULSE electric current sintering (PECS, also known as spark
plasma sintering, SPS) is a potential technique for densifying
P
ceramic materials. Dense alumina has been prepared by apply-
ing this technique at 11501C for 10 min1 or at 12501C for 3 min.2
The technique is also capable of producing ceramic matrix com-
3
posites. It has been used to prepare dense Al2O3/3 vol% ZrO2
and Al2O3/SiC nanocomposites4 at 14501C within 5 min. Al-
though the densification rate during PECS is very fast, the abil-
ity to use PECS for tailoring the microstructure of ceramics is
very much questionable. The coarsening rate during PECS has
been reported to be extremely fast.5 Shen et al. had noticed that
the length of silicon nitride grains increases to five times its
original length within the first minute at the peak sintering tem-
perature; the coarsening then slows down in the next 10 min.
Regarding concerns of microstructural control, previous
studies all recognized the importance of processing parameters
during PECS. Systematic studies had been conducted2,6; the
peak temperature, pressure, and heating rate are suggested to be
the key parameters. Raising the peak temperature during PECS
typically gives rise to a larger final grain size. The increase of
pressure can shorten the time needed to reach full density. If the
dwell time at the peak temperature is not reduced accordingly, it
would typically result in a larger grain size. The role of the
heating rate in the coarsening behavior of grains remains un-
clear. The increase of heating rate has been reported to result in
finer grain sizes2,3,6–8; however, a fast heating rate has also been
documented as the key to enhancing grain growth.5 A recent
II. Experimental Procedure
Aluminum hydroxide powder (Zhengzhou Aluminum Co.,
Zheng Zhou, China) with a particle size ranging from 50 to
100 mm was used in the present study. The thermogravimetric
analysis (TGA) and differential thermal analysis (DTA) on the
powder were carried out from room temperature to 12001C using
a thermal analyzer (STA 449C, Netzsch Co., Selb, Germany).
The heating rate was 101C/min. Powder compacts with a diam-
eter of 25.4 mm were formed by uniaxial pressing at 30 MPa. A
box furnace was used for the heat treatment of the powder com-
pacts. According to the resulting TGA/DTA curves, the follow-
ing temperatures: 5801, 8001, 9001, and 12001C were chosen to
heat treat the powder compacts. The heating rate was 31C/min,
and the dwell time was 0.5 h. The phases in the heat-treated and
PECS specimens were analyzed using X-ray diffractometry
(XRD, PW1830, Philips Co., Eindhoven, the Netherlands).
The X ray first passed a slot that guided the X ray toward the
central area of the specimen. The detected area covered most of
the specimen surface, about 10 mm ꢀ 10 mm and 15 mm ꢀ
10 mm for a high and a low angle, respectively.
L. Klein—contributing editor
Manuscript No. 20621. Received May 31, 2005; approved August 4, 2005.
Supported by the National Natural Science Foundation, China (NSFC No. 50232020,
50220160657) and National Science Council, Taiwan (NSC92-2816-E-002-0004-6).
*Member, American Ceramic Society.
A graphite mold with a height of 60 mm, inner diameter of
20 mm, and an outer diameter of 50 mm was used as a container
wAuthor to whom correspondence should be addressed. e-mail: tuan@ccms.ntu.edu.tw
680

February 2006
Communications of the American Ceramic Society
681
for the specimen. A 0.1 mm-thick carbon sheet was inserted into
the small gap between the punches and mold unless otherwise
stated. Approximately 4 g of aluminum hydroxide powder was
then added into the graphite mold. An external load of 30 MPa
was applied throughout the whole PECS process. Sintering was
carried out in vacuum in an SPS furnace (Model SPS-1050,
Sumitomo Coal Mining Co., Tokyo, Japan). A 12 ms-on and 2
ms-off pulse sequence was used. The temperature was measured
by a K-type thermocouple inserted into a hole on the outer sur-
face of the mold. A hole with a depth of 2 mm was located at the
middle height of the outer surface. The peak temperature during
PECS as detected by the thermocouple was 9001C. This peak
temperature was chosen to limit the thermal radiation. The
heating rates were either 2601C/min or 801C/min. The current
used to generate a heating rate of 2601C/min was almost four
times that used to generate a heating rate of 801C/min. The
holding time at the peak temperature was 1 min. The final di-
mensions of the PECS specimens were 20 mm diameter and 9
mm height. For the phases at different depths in the specimens,
successive XRD patterns were obtained by removing consecu-
tive layers of materials, 0.3 mm- or 1.5 mm-thick, from the top
and bottom surfaces.
Fig. 1. Thermogravimetric analysis (TGA) and differential thermal
analysis (DTA) curves of the aluminum hydroxide powder.
The phase analysis reveals that the temperature on the top
surface is around 1001C higher than 9001C. The temperature of
the PECS specimen just under the top surface is slightly higher
than that on the top surface. A small amount of g-alumina is
found at the location 1.5 mm above the bottom surface, indi-
cating that the temperature at the place is slightly lower than
8001C. The temperature variation in terms of thickness in the
surface region (i.e., 0–1.5 mm from the top and bottom surfaces)
is larger than that in the central region (i.e., 1.5–7.5 mm from the
top surface).
III. Results and Discussion
The DTA/TGA shown in Fig. 1 suggest that the thermal de-
composition mainly takes place below 5501C. A huge endother-
mic peak covers a large temperature range, from 5801 to 12001C,
indicating that many phase changes take place within this tem-
perature range. The phase analysis was therefore carried out on
the specimens after heat treatment above 5801C.
Figure 2 shows the XRD patterns of the aluminum hydroxide
powder and of the compacts after conventional heat treatment
at various temperatures. The figure indicates that the aluminum
hydroxide powder is mainly composed of g-AlO(OH). After
heating at 5801C for 0.5 h, w-alumina and g-alumina are formed.
d-alumina is then formed at 8001C in the presence of w-alumina.
k-alumina is formed at 9001C on consumption of w-alumina.
When the powder compact is heated at 12001C for 0.5 h, a-alu-
mina is the only phase detected. Aluminum hydroxide thus fol-
lows the following transformation sequences as the powder
compact is heated to elevated temperatures:
The carbon paper separator was used in the PECS specimen
shown in Fig. 3(b). The specimen was also heated by PECS at
9001C for 1 min at a heating rate of 2601C/min. The carbon
paper had not adhered strongly to the specimen after PECS and
could be removed from the specimen surface by light brushing.
w-, d-, and k-alumina are found across the entire cross section of
the specimen, except that the intensity of k-alumina is the high-
est on the top surface and lowest at a location 1 mm above the
bottom surface. It indicates that a temperature variation still
existed near the top and bottom surfaces. The temperature var-
iation is smaller in the central region.
Comparing Figs. 3(a) and (b), it can be observed that the
presence of a carbon paper separator reduces the temperature
variation during the PECS process. Zavaliangos et al.10 had
suggested that the presence of a carbon paper separator could
improve temperature uniformity. The present study reinforces
this argument.
g-A1OðOHÞ
ꢂ
!
ꢁ580 C w-A12O3 þ g-A12O3
ꢂ
!
ꢁ800 C w-A12O3 þ d-A12O3
(1)
ꢂ
!
!
ꢁ900 C k-A12O3 þ d-A12O3
ꢂ
ꢁ1200 C a-A12O3
With the phase transformation sequences shown in Eq. (1), it is
then possible to apply the phase analysis to estimate the tem-
perature variation within the PECS powder compact.
Figure 3(a) shows the phases on the top surface and the bot-
tom surface of the specimen after PECS at 9001C for 1 min at a
heating rate of 2601C/min. Three test runs have been performed
for each PECS condition. The typical results are shown in
Fig. 3(a). For the specimens shown in the figure, no carbon pa-
per was inserted into the gap between the punches and mold.
The phases at depths of 0.3 and 1.5 mm under the top and bot-
tom surfaces of the PECS powder compact are also shown.
Apart from d- and k-alumina, a-alumina and graphite are also
found on the top surface of the specimen. After removing a layer
of 0.3 mm from the top surface, the intensity of a-alumina in-
creases and no graphite is found. The phases at a depth of 1.5
mm under the top surface are d- and k-alumina, which are char-
acteristic phases of the specimen heat treated at 9001C. Both
w- and d-alumina are found on the bottom surface and within a
0.3 mm-thick layer from the bottom surface.
Fig. 2. X-ray diffractometry patterns for the aluminum hydroxide pow-
der compacts after conventional heat treatment at various temperatures.

682
Communications of the American Ceramic Society
Vol. 89, No. 2
Fig. 4. X-ray diffractometry patterns for the aluminum hydroxide pow-
der compacts after pulse electric current sintering at 9001C for 1 min at a
heating rate of 801C/min. A carbon paper separator is used.
The present study demonstrates that a reduction of temper-
ature variation within the nonconducting ceramic compact is
possible. The most effective method is to introduce a soft elec-
trical conducting element, such as a carbon paper, to fill in the
gap between the punches and the mold. However, the use of
carbon paper is time consuming and might limit the industrial-
ization potential of the PECS technique.
The area analyzed by the XRD technique is relatively large,
covering most of the specimen surface. The temperature deter-
mined by the phase analysis can thus be treated as the average
value at a certain depth from the top surface. Furthermore, the
technique can only deliver information along the Z-direction of
the specimen but not in the X–Y direction.
The phase transformation is a function of the surrounding
temperature as well as a function of the heating rate. In the
present study, the phase transformation of aluminum hydroxide
is first determined by heating the specimens to various temper-
atures at a heating rate of 3 C/min in a box furnace. No phase
variation is noticed within the specimens. The phase transition
sequence as a function of temperature is therefore established. It
allows us to use the resulting phases as a temperature indicator
to estimate the temperature variation within the PECS speci-
mens. A higher heating rate was used during the PECS process.
The precise temperature is thus difficult to determine by using
the methodology proposed in the present study. Nevertheless,
the present methodology can provide an estimation of the tem-
perature difference at different depths. Furthermore, by using
the present methodology, it is possible to estimate each process-
ing parameter, such as peak temperature, pressure, and heating
rate, on the temperature uniformity within the powder compact
during PECS. The methodology can thus be used to improve the
application potential of the PECS technique.
Fig. 3. X-ray diffractometry (XRD) patterns for the aluminum hy-
droxide powder compacts after pulse electric current sintering (PECS) at
9001C for 1 min at a heating rate of 2601C/min. The phases at a certain
distance from the top and bottom surfaces are also shown. The estimat-
ed temperature for the XRD pattern is shown on the right-hand side of
each pattern. The symbol 0.3 top indicates that it is 0.3 mm from the top
surface. The thickness of the specimen is 9 mm. No carbon paper
separator is used in (a).
Figure 4 shows the XRD patterns of the top and bottom
surfaces prepared by the PECS at a heating rate of 801C/min.
The carbon paper separator was used for the specimen. Along
with d- and k-alumina, w-alumina is also detected, indicating
that the temperature on the top surface is between 8001 and
9001C. There is no w-alumina found on the bottom surface,
suggesting that the temperature at the bottom surface is around
8001C. Therefore, the temperature variation across the specimen
still existed. It is also worth noticing that the temperature of the
ceramic powder compact is lower than that of the graphite mold
when a slow heating rate is applied. It demonstrates that Joule
heating contributes to the temperature increase of the electric-
conducting mold. The heat is dissipated with the increase of
distance from the mold surface. The heat generated either by
spark plasma or electric discharge within the nonconducting ce-
ramic particles is relatively small if not negligible. As most Joule
heating is generated from the punches,10 the temperature near
the punches is the highest. The carbon paper is less rigid com-
pared with graphite punches, and therefore, the carbon paper
can provide a good contact between the punches and the powder
compact. The carbon paper is a good thermal conductor; the
temperature variation can be reduced through heat conduction
along the carbon paper.
IV. Conclusions
In the present study, a methodology is developed to estimate the
temperature uniformity within a ceramic powder compact dur-
ing PECS. A temperature variation is present within the alumina
powder compact during PECS. The temperature is the highest
near the top surface. The use of a carbon paper separator is an
effective method of improving temperature uniformity. The de-
crease of heating rate can also contribute to the reduction of
temperature variation within the powder compact.
Acknowledgment
The valuable comments given by Prof. Jay Shieh, Department of Material
Science & Engineering, National Taiwan University, were very helpful.

February 2006
Communications of the American Ceramic Society
683
6S. W. Wang, L. D. Chen, and T. Hirai, ‘‘Densification of Al2O3 Powder Using
Spark Plasma Sintering,’’ J. Mater. Res, 15 [4] 982–7 (2000).
References
1S. H. Risbud, C.-H. Shan, A. K. Mukherjee, M. J. Kim, J. S. Bow, and R. A.
Holl, ‘‘Retention of Nanostructure in Aluminum Oxide by Very Rapid Sintering at
11501C,’’ J. Mater. Res, 10 [2] 237–9 (1995).
7S. W. Wang, L. D. Chen, T. Hirai, and J. Guo, ‘‘Formation of Al2O3 Grains
With Different Size and Morphologies During the Pulse Electric Current Sintering
Process,’’ J. Mater. Res, 16 [12] 3514–7 (2001).
2Z. Shen, M. Johnsson, Z. Zhao, and M. Nygren, ‘‘Spark Plasma Sintering of
Alumina,’’ J. Am. Ceram. Soc, 85 [8] 1921–7 (2002).
8D. D. Jayaseelan, S. Ueno, T. Ohji, and S. Kanzaki, ‘‘Differential Sintering by
Improper Selection of Sintering Parameters During Pulse Electric Current Sinte-
ring,’’ J. Am. Ceram. Soc, 87 [1] 159–61 (2004).
3D. D. Jayaseelan, N. Kondo, D. A. Rani, S. Ueno, T. Ohji, and S. Kanzaki,
‘‘Pulse Electric Current Sintering of Al2O3/3vol%ZrO2 with Constrained Grains
and High Strength,’’ J. Am. Ceram. Soc, 85 [11] 2870–2 (2002).
4L. Gao, H. Z. Wang, J. S. Hong, H. Miyamoto, K. Miyamoto, Y. Nishikawa,
and S. D. D. L. Torre, ‘‘Mechanical Properties and Microstructure of Nano-SiC–
Al2O3 Composites Densified by Spark Plasma Sintering,’’ J. Eur. Ceram. Soc, 19
[2] 609–13 (1999).
9Y. Zhou, K. Hirao, Y. Yamauchi, and S. Kanzaki, ‘‘Densification and Grain
Growth in Pulse Electric Current Sintering of Alumina,’’ J. Eur. Ceram. Soc, 24,
3465–70 (2004).
10A. Zavaliangos, J. Zhang, M. Krammer, and J. R. Groza, ‘‘Temperature Ev-
olution During Field Activated Sintering,’’ Mater. Sci. & Eng, A 379 [1] 218–28
(2004).
&
5Z. Shen, Z. Zhao, H. Peng, and M. Nygren, ‘‘Formation of Tough Interlocking
Microstructures in Silicon Nitride Ceramics by Dynamic Ripening,’’ Nature, 417
[May 16] 266–9 (2002).