Effects of pressure application method on transparency of spark plasma sintered alumina

Article abstract of DOI:10.1111/j.1551-2916.2010.04274.x

Commercially available alumina powder was consolidated at 1150°C by spark plasma sintering at the heating rate of 100°C/min. The effects of the pressure application mode were examined with respect to microstructure, porosity, and transparency. A finite-element simulation was developed in order to understand the relationship between sample homogeneity and its temperature distribution. The effects of the temperature probing point on the microstructure were investigated. The application of two steps pressure was found effective to obtain homogeneously densified translucent alumina samples at high heating rate.

Full text of DOI:10.1111/j.1551-2916.2010.04274.x

J. Am. Ceram. Soc., 94 [5] 1405–1409 (2011)
DOI: 10.1111/j.1551-2916.2010.04274.x
r 2010 The American Ceramic Society
ournal
J
Effects of Pressure Application Method on Transparency of Spark
Plasma Sintered Alumina
Salvatore Grasso,z,y Chunfeng Hu,y Giovanni Maizza,z Byung-Nam Kim,y and Yoshio Sakka*,w,z,y
zGraduate School of Pure and Applied Sciences, University of Tsukuba, Ibaraki 305-0047, Japan
yWorld Premier International Research Center Initiative (WPI Initiative) on Materials Nanoarchitronics (MANA) and
Nano Ceramics Center, National Institute for Materials Science (NIMS), Ibaraki 305-0047, Japan
zDipartimento di Scienza dei Materiali ed Ingegneria Chimica, Politecnico di Torino, I-10129 Torino, Italy
Commercially available alumina powder was consolidated
at 11501C by spark plasma sintering at the heating rate of
1001C/min. The effects of the pressure application mode were
examined with respect to microstructure, porosity, and trans-
parency. A finite-element simulation was developed in order to
understand the relationship between sample homogeneity and its
temperature distribution. The effects of the temperature probing
point on the microstructure were investigated. The application of
two steps pressure was found effective to obtain homogeneously
densified translucent alumina samples at high heating rate.
For example, Kim and colleagues in order to achieve a trans-
parency of 47% used a sintering cycle of about 6 h (i.e., 21C/min),
such long processing time strongly limits the production rate.
Furthermore, the SPS is well known as a rapid sintering technol-
ogy, consequently the use of low heating rate would limit its
intrinsic advantage.9 The SPS processes are energetically efficient
because they can use two-thirds to four-fifths of the energy
needed by conventional HP.10 On the other hand, when slow
heating rate are used, the heat is rapidly extracted by the cooling
system and radiated from the graphite surfaces to the inner wall
of the SPS furnace. The reduction of the heating and dwelling
time can enormously contribute to reduce the heat dissipated by
the cooling system.
I. Introduction
The present work is divided in two parts. In the first part,
the effect of the temperature probing point with respect to the
microstructure and grain growth (Section III(1)) are analyzed.
In the second part, a combined experimental numerical simula-
tion is developed in order to understand the sintering mecha-
nism of alumina processed at high heating rate (Section III(2)).
The effects of the pressure application mode along with experi-
mental and numerical analysis are discussed.
S reported in literature, translucent alumina with an average
grain sizes of 0.5–0.6 mm and densities of about 99.8% has
A
been traditionally used as structural ceramic because of the very
high hardness (20–21 GPa1,2) and elevated three-point bending
strength (750–900 MPa1). The excellent mechanical properties
combined with in line transmittance exceeding 50 % make this
material particularly suitable for the production of ceramics for
lamp envelopes and other structural applications. Traditionally
translucent alumina is manufactured by hot isostatic pressing
(HIP), however recently, spark plasma sintering (SPS) has been
considered as an alternative method to obtain nearly full dense
fine-grained translucent alumina.3–7 Owing to the advantage of
rapid heating, the alumina ceramics obtained by SPS have a grain
size and density comparable to those of HIPed ones.8 Several
investigation focused on the consolidation of transparent alumina
by SPS. Kim et al.3,5 reported that in the case of slow heating of
21C/min rate was possible to obtain transparency as high as 47%,
conversely, in the case of 1001C/min the transmittance was as low
as 0.2%. The latter results has been confirmed in terms of porosity
by Aman and colleagues (see fig. 3 of Aman et al.7).
Most of the investigation reported that is rather difficult
to obtain translucent alumina by SPS at the heating rate of
1001C/min.3,5,7 At present, the effect of the heating rate on the
transparency of SPSed alumina has been not yet fully eluci-
dated. However, from a technological point of view, sintering
at high heating rate is strongly desired in order to: (i) increase
the productivity, (ii) decrease the specific energy consumption of
the consolidation process.
II. Experimental Procedure
Commercially available a-Al2O3 powder (TM-DAR, Taimei
Chemicals Co. Ltd., Tokyo, Japan), with a purity of 99.99%
and an average particle size of 0.2 mm, was used in this study.
The as-received powder was directly poured into a graphite die
without any special treatment or additives. The inner and outer
diameters of the graphite die were 30 and 70 mm respectively,
the height of the die was 60 mm. The height and the diameter of
the punches were both 30 mm. The two pairs of graphite spacers
(+80 height 40 mm and +120 height 20 mm) were pushed
between the water-cooled rams made up of steel (+120 height
20 mm). The temperature was accurately measured by two
pyrometers,11 the side pyrometer was focused on the die surface
(hereinafter, side pyrometer). The top pyrometer was focused on
nontrough hole (+9 mm) inside the punch (at the center of the
punch mid thickness). A drawing showing the temperature
probing point is given below in Fig. 5. The double pyrometers
configuration was at first calibrated. The SPS machine was
operated in temperature control mode, and the temperature
was measured at the same time by the top and side pyrometers.
In order to reduce the heat loss by radiation a graphite felt
(1 cm thick) was used.
J. Blendell—contributing editor
In all experiments an identical heating cycle was used: the
temperature was raised up to 7001C in 10 min consequently the
heating rate up to 11501C was 1001C/min. The dwelling times at
11501C were 30 and 60 min. The pressure was applied in two
ways, in the first case a constant pressure of 80 MPa was applied
for the entire duration of the sintering process (hereinafter,
Manuscript No. 27891. Received April 22, 2010; approved October 29, 2010.
This work was supported by World Premier International Research Center Initiative
(WPI Initiative), MEXT, Japan.
*Member, The American Ceramic Society.
wAuthor to whom correspondence should be addressed. e-mail: sakka.yoshio@nims.
go.jp
1405

1406
Journal of the American Ceramic Society—Grasso et al.
Vol. 94, No. 5
constant pressure). In the second case, the initial pressure of
35 MPa was applied, subsequently the pressure was increased in
3 min after the beginning of dwelling time (hereinafter, two steps
pressure). Heating was conducted using a sequence consisting of
12 DC pulses (40.8 ms) followed by power off for 6.8 ms.
Finally, we obtained a sintered disk with a diameter of 30 mm
and a thickness of 3 mm. The entire disk body was machined to
a disk of 30 mm diameter with a thickness of 1 mm, and mirror-
polished carefully on both sides using diamond slurry. The final
thickness of the sample was about 0.8 mm. The in-line trans-
mission was measured in the wavelength range from 0.24 to 1.6 mm
using a double-beam spectrophotometer (SolidSpec-3700DUV,
Shimadzu, Kyoto, Japan). The distance between the sample and
the detector was about 55 cm.
The fracture surfaces and thermally etched surfaces (9501C
for 1 h) were observed by means of a scanning electron micro-
scope (SEM) (JSM-7100, JEOL, Tokyo, Japan). The porosity
was measured on the SEM images of the polished and thermally
etched surfaces taken at a magnification of 10000 times. We did
not measure the absolute density because the conventional tech-
niques such as the Archimedes method are insensitive to ex-
tremely low porosity.
The grain size, observed on the SEM images of the polished
and thermally etched surfaces, was calculated by obtaining the
average cross-section area per grain and assuming spherical
grains. The measured grain size is an apparent one, so that it
was multiplied by 1.225 to determine the true grain size.8
A finite-element model was developed to predict the temper-
ature distribution of the alumina sample. The model permitted
to understand the relationship between temperature distribution
and densification inside the alumina samples. The model used an
integrated experimental/numerical methodology, which simul-
taneously permitted the developed SPS model to be reliably
tested against experiments. The SPS model coupled electrother-
mal and displacement fields; the contact multiphysics at the
sliding punch/die interface was modeled during powder sintering
using a moving mesh/moving boundary technique. The SPS
model used here assumes uniform displacement and stress
within the sample. The displacement history is taken from
experiments and it prescribes the average density of the sample.
Coupling between electrothermal field and displacement is made
through the evolving average density. The friction between
punch and die is not taken into account in the model. Model
details can be found in Maizza et al.12,13 The thermal conduc-
tivity of the pressed alumina powders assumed in the FEM
calculations was experimentally measured by a Xenon flash
apparatus (LFA447 Nanoflash, Netzsch, Selb, Germany).
III. Result and Discussions
(1) Effect of the Temperature Probing Point
Usually SPS machine are operated in temperature control mode.
The temperature is measured at a probing point while the volt-
age/current is continually adjusted to fulfill the desired heating/
cooling cycle. Especially in the case of high heating rate is im-
portant to choose the probing point in order to accurately heat
the powder up to the desired sintering temperature at pro-
grammed heating rate without overheating.
Figure 1 shows the voltage and temperature profiles in the
case of alumina samples sintered by controlling the temperature
from (a) the top and (b) side pyrometer. In the case of control
from the top, the top temperature did not exceeded 11501C. As
shown in Fig. 1(a), during the heating at temperature above
10001C, the temperature difference DT, between top and side
pyrometer was around 1431C. The temperature difference DT
during the dwelling time decreased to 151C.
On the contrary to the top pyrometer control, it was not
possible to accurately control the temperature in the case the
side pyrometer (Fig. 2(b)). At the end of the heating (e.g., before
the dwelling), the temperatures measured at the top and side
pyrometer were 12301 and 11901C, respectively. Such overheat-
ing was attributed to the thermal gradient DT and to the tem-
perature oscillation respect to the programmed temperature
cycle (i.e., 11501C). The voltage rapidly fluctuated during the
temperature oscillations due to the instability of the temperature
control.
By controlling the temperature from top, it was possible to
adjust the voltage in order to fulfill the programmed tempera-
ture cycle. In fact in the case of top pyrometer, unlike the side
one, no significant overheating was measured. The temperature
was directly measured nearby the heating source of the punch-
die assembly and consequently the voltage could be properly
adjusted by the SPS machine control.
As shown in Fig. 1, the probing point affected the stability of
automatic the SPS control voltage/current in terms of temper-
ature overheating and temperature oscillation respect to the
programmed temperature cycle. Depending on the temperature
probing point, top or side, the average grain sizes observed
in the center of the sample were 0.47 and 0.83 mm, respectively.
In the case of temperature controlled from the side (Fig. 1(b)),
the overheating enhanced severe grain growth (Fig. 2(b)), on the
contrary, in the case of top pyrometer (Fig. 1(a)) the accurate
temperature control limited the grain growth (Fig. 2(a)). In the
case of high heating rate SPS (i.e., 1001C/min), in order to
obtain finer microstructure and more accurate control of the
Fig. 1. Effect of the temperature probing point on the voltage and temperature profiles as function of time. SPS apparatus was controlled (a) from the
top pyrometer (b) from the side pyrometer.

May 2011
Effects of Pressure Application Method on SPSed Alumina
1407
Fig. 2. FESEM of intregranular fracture surface of alumina sample sintered at 11501C for 30 min heated at 1001C/min under constant pressure of
80 MPa. The temperature was controlled from the top (a) and from the side (b) pyrometers.
sintering temperature, the machine should be operated by con-
trolling the temperature from the top pyrometer.
reason of such significant dissimilarity of the transmittance can
be attributed to the residual porosity.14 In fact in the case of
sample Acenter the porosity was 1.26% which completely deteri-
orated the transmittance, instead in the case of sample B the
porosity was 0.20%.
(2) Combined Experimental and Modeling Analysis of the
Pressure Application Mode
The FEM simulation, developed in Maizza et al.,12,13 calcu-
lated the temperature distribution inside the sample and permit-
ted to understand the densification mechanism in the case of
constant pressure application. Figure 5 shows the temperature
distribution inside (a) the punch-die sample assembly and (b)
sample calculated before the holding at 11501C (i.e., at the end
of the heating cycle). Because the alumina behaves as electric
insulator, the Joule heating due to the current flowing inside the
sample is substantially negligible.15 Owing to the low thermal
conductivity of the pressed alumina powders (0.6 W/m ꢀ K)16,17
and the high heating rate (1001C/min), the heat was slowly con-
ducted from the punch-die assembly trough pressed alumina
powder and a significant thermal gradient of 421C was gener-
ated inside the sample (Fig. 5(b)).
Figure 3 shows a photograph of sintered alumina samples with a
thickness of about 0.8 mm on top of the text. The samples A–C
were sintered at 11501C at a heating rate of 1001C/min, the SPS
machine was controlled by the top pyrometer. The sample A was
sintered under 80 MPa constant pressure for 30 min. As shown in
the inset Figure 3 of sample A, the border of the sample (1.5 mm
width) hereinafter defined as Aborder, is translucent while the in-
ner part Acenter is opaque. In the case of samples B and C the
pressure was raised from 35 to 80 MPa 3 min after the beginning
of the dwelling time (two steps pressure). The holding time at the
sintering temperature were 30 and 60 min, respectively. The
transmittance measured at the center of the samples A–C along
with the grain size and residual porosity are given in Table I. The
grain size of sample B is shown in Fig. 4. It was not possible to
measure the transmittance of sample Aborder because the trans-
lucent area was not big enough to permit the measurement.
Despite the identical powder, sintering temperature, heating
rate and holding time, the transmittance and the porosity of
samples prepared by SPS are quite different depending on the
pressure application method. By comparing samples A and B
the in-line transmittance increased from 0.3% to 15.7%. The
The densification mechanism in the case of constant pressure
can be schematized as follows:
(1) The sample border was 421C hotter (Fig. 5(b)) than the
center due to the high heating rate and poor thermal conduc-
tivity of the pressed alumina powders. The latter promoted the
densification of the hotter region sample (sample Aborder).
(2) During the holding at 11501C, the sample Aborder and
Acenter densified up to 99.92% and 98.74%, respectively (see
Fig. 3. Photograph of alumina ceramics disks sintered by SPS at 11501C with a heating rate of 100 C/min. The sample shown in figure (a) was sintered
for 30 min under 80 MPa constant pressure. The sample (b) and (c) were sintered with pressure two steps pressure application for 30 and 60 min. The
samples are 3 cm in diameter and are on top of the text.

1408
Journal of the American Ceramic Society—Grasso et al.
Vol. 94, No. 5
Table I. Sintering Conditions (Sintering Temperature, Holding Time, Pressure Application Method), Grain Size, Residual Porosity,
and in Line Transmittance (Measured at a Wavelength of 645 nm) of the Sintered Samples
Sample name
Sintering conditions
Grain size (mm)
Porosity (%)
In line transmittance (%)
Sample Acenter
Sample Aborder
Sample B
1150 C, 30 min, constant pressure
0.47
0.39
0.41
0.47
1.26
0.08
0.20
0.14
0.33
—
15.7
19.8
1150 C, 30 min, 2 steps pressure
1150 C, 60 min, 2 steps pressure
Sample C
densifying fully. McWilliams and Zavaliangos20 investigated de-
veloped a FEM model which coupled the thermal-electric and
free (i.e., the compact is not constrained in a die or under pres-
sure) sintering phenomena. Similarly as in the present study, the
thermal gradient inside the sample generated non homoge-
neously densified samples.20
On the contrary to constant pressure, in the case of two steps
pressure (sample B and C Fig. 3) it was possible to obtain trans-
lucent alumina ceramics with grain size about 0.4 mm and final
density 499.8%. The initial pressure of 35 MPa was insufficient
to achieve high level of densification. After 3 min the beginning of
the dwelling time, the thermal gradient inside the sample decreased
down to 51C (as given by FEM simulation) and the pressure el-
evation up to 80 MPa permitted to obtain homogeneously densi-
fied translucent alumina as shown in Fig. 3. In the case of samples
B and C, by extending the holding time from 30 to 60 min the
porosity decreased from 0.2 to 0.14 while the in line transmittance
(for wavelength of 645 nm) increased from 15.6% to 19.8 %. No
significant difference in the microstructure along the sample radius
was observed. The pressure elevation 3 min after the beginning of
the dwelling permitted lower the thermal gradient inside the sam-
ple (e.g., 51C as given by FEM results) and offered an effective
means to obtain translucent alumina at high heating rate.
In the case of high heating rate SPS, two steps pressure and
constant pressure led to translucent and opaque samples, respec-
tively. Indeed, the samples obtained by two steps pressure
were more dense and more homogeneously densified than the
one obtained by constant pressure. The developed SPS model,
together with the two-step pressure method, is an aiding tool
to predict the temperature distribution inside the sample and to
elucidate the mechanism for obtaining homogeneously densified
tanslucent alumina ceramics.
Fig. 4. Microstructures of the sample B alumina sintered for 30 min at
11501C with heating rate of 11501C.
Table I). The highly dense sample Aborder prevented the further
shrinkage of the powder along the pressing direction.
Similarly to present investigation, Wang et al.18,19 reported
that the temperature gradient inside the sample resulted in differ-
ential densification which corresponded to a gradual decrease of
hardness from the border to center of the alumina samples.
In the case of the constant pressure, the sample Aborder, due to
the higher temperature, densified at a higher rate than the sam-
ple Acenter, consequently more dense sample border limited the
punch shrinkage and consequently inhibited the center from
Fig. 5. Temperature distribution inside (a) the punch-die sample assembly and (b) sample calculated before the dwelling time (i.e., at the end of the
heating cycle). Arrows plot shows the current distribution.

May 2011
Effects of Pressure Application Method on SPSed Alumina
1409
3B. Kim, K. Hiraga, K. Morita, and H. Yoshida, ‘‘Effects of Heating Rate on
Microstructure and Transparency of Spark-Plasma-Sintered Alumina,’’ J. Eur.
Ceram. Soc, 29, 323–7 (2009).
IV. Conclusion
In addition to previously reported investigations it was possible to
increase the transmittance of translucent alumina prepared by high
heating-rate SPS when the maximum pressure was applied during
a later stage of the sintering process. It is shown that the control
from the top pyrometer prevented powder overheating and con-
sequently the grain growth. The pressure application method was
crucial to obtain homogeneously densified translucent sample.
The application of constant pressure led to non-homoge-
neously densified sample due to the significant thermal gradient
generated during the heating. The border of the sample was
highly densified (i.e., porosity 0.08%) while the center was
porous (i.e., porosity 1.26%).
The two-step pressure method was effective to obtain homo-
geneously densified translucent alumina ceramics with grain size
of about 0.4 mm and porosity below 0.2%. The initial pressure of
35 MPa was insufficient to achieve high level of densification.
After 3 min the beginning of the dwelling time, the thermal gra-
dient inside the sample decreased down to 51C, consequently the
pressure elevation up to 80 MPa permitted to obtain homoge-
neously densified translucent samples.
4B. Kim, K. Hiraga, K. Morita, and H. Yoshida, ‘‘Spark Plasma Sintering of
Transparent Alumina,’’ Scr. Mater., 57, 607–10 (2007).
5B. Kim, K. Hiraga, K. Morita, H. Yoshida, T. Miyazaki, and Y. Kagawa,
‘‘Microstructure and Optical Properties of Transparent Alumina,’’ Acta Mater.,
57, 1319–26 (2009).
6Y. Aman, V. Garnie, and E. Djurado, ‘‘Influence of Green State Processes on
the Sintering Behaviour and the Subsequent Optical Properties of Spark Plasma
Sintered Alumina,’’ J. Eur. Ceram. Soc., 29, 3363–70 (2009).
7Y. Aman, V. Garnier, and E. Djurado, ‘‘A Screening Design Approach for the
Understanding of Spark Plasma Sintering Parameters: A Case of Translucent
Polycrystalline Undoped Alumina,’’ Int. J. Appl. Ceram. Technol., 7, 574–86
(2010).
8R. Apetz and M. P. B. Bruggen, ‘‘Transparent Alumina: A Light-Scattering
Model,’’ J. Am. Ceram. Soc., 86, 480–6 (2003).
9S. Grasso, Y. Sakka, and G. Maizza, ‘‘Electric Current Activated/Assisted
Sintering (ECAS): A Review of Patents 1906–2008,’’ Sci. Technol. Adv. Mater., 10,
053001 (2009).
10M. Tokita, ‘‘Large-Size Functionally Graded Materials Fabricated by Spark
Plasma Sintering (SPS) method,’’ Mater. Sci. Forum, 39, 423–5 (2003).
11K. Vanmeensel, A. Laptev, J. Hennicke, J. Vleugels, and O. Van der Biest,
‘‘Modelling of the Temperature Distribution during Field Assisted Sintering,’’
Acta Mater., 53, 4379–88 (2005).
12G. Maizza, S. Grasso, Y. Sakka, T. Noda, and O. Ohashi, ‘‘Relation between
Microstructure, Properties and Spark Plasma Sintering (SPS) Parameters of Pure
Ultrafine WC Powder,’’ Sci. Technol. Adv. Mater., 8, 644–54 (2007).
13G. Maizza, S. Grasso, and Y. Sakka, ‘‘Moving Finite-Element Mesh Model
for Aiding Spark Plasma Sintering in Current Control Mode of Pure Ultrafine WC
Powder,’’ J. Mater. Sci., 44, 1219–36 (2009).
The two-step pressure method enabled a significant improve-
ment of the degree of in line-translittance of high-heating rate
SPSed alumina. However, in the case of alumina sintered under
pressure not exceeding 80 MPa, it is another confirmation of
previously published results of this working group that high-rate
SPS does not achieve a complete elimination of pores on the
level requested for high transparency.
14S. Grasso, B. Kim, C. Hu, G. Maizza, and Y. Sakka, ‘‘Highly Transparent
Pure Alumina Fabricated by High-Pressure Spark Plasma Sintering,’’ J. Am.
Ceram. Soc., 93, 2460–2 (2010).
15H. Tomino, H. Wantanabe, and Y. Kondo, ‘‘Electric Current Path and Tem-
perature Distribution for Spark Sintering,’’ J. Jpn. Soc. Powder Powder Metall.,
44, 974–9 (1997).
16L. Braginsky, V. Shklover, H. Hofmann, and P. Bowen, ‘‘High-Temperature
Thermal Conductivity of Porous Al2O3 Nanostructures,’’ Phys. Rev. B, 70, 134201
(2004).
Acknowledgments
17W. N. dos Santos, ‘‘Effect of Moisture and Porosity on the Thermal Proper-
ties of a Conventional Refractory Concrete,’’ J. Eur. Ceram. Soc., 23, 745–55
(2003).
The authors are grateful to Dr. Toshiyuki Nishimura for providing the mea-
surement of the thermal conductivity of the Alumina used in the present study.
18S. W. Wang, L. D. Chen, T. Hirai, and Y. S. Kang, ‘‘Microstructure Inho-
mogeneity in Al2O3 Sintered Bodies Formed During the Plasma-Activated Sinte-
ring Process,’’ J. of Mat. Scie. Lett., 18, 1119–21 (1999).
References
1A. Krell, P. Blank, H. Ma, and T. Hutzler, ‘‘Transparent Sintered Corundum
with High Hardness and Strength,’’ J. Am. Ceram. Soc., 86, 12–8 (2003).
19S. W. Wang, L. D. Chen, and T. Hirai, ‘‘Densification of Al2O3 Powder using
Spark Plasma Sintering,’’ J. Mater. Res., 15, 982–7 (2000).
2A. Krell and S. Schadlich, ‘‘Nanoindentation Hardness of Submicrometer
Alumina Ceramics,’’ Mater. Sci. Eng. A, 307, 172–81 (2001).
20B. McWilliams and A. Zavaliangos, ‘‘Multi-Phenomena Simulation of Electric
¨
Field Assisted Sintering,’’ J. Mater. Sci., 43, 5031–5 (2008).
&