Full text of DOI:10.1557/mrs2001.115

www.mrs.org/publications/bulletin
tivity of over 200 W mꢀ1 Kꢀ1 at room tem-
perature. These facts, combined with the
excellent mechanical properties of Si3N4
ceramic, make it a serious candidate for
high-performance substrates. This article
summarizes recent results on the thermal
conductivity of Si3N4 and discusses the
extrinsic factors governing the thermal
conductivity of Si3N4 ceramic in terms of
microstructural parameters such as lattice
defects in single-crystal Si3N4 and the mor-
phology of grain-boundary secondary
phases, comparing Si3N4 with the cases of
SiC and AlN ceramics. In addition, high
thermal conductivities achieved in textured
Si3N4 are also introduced.
High Thermal
Conductivity Silicon
Nitride Ceramic
Kiyoshi Hirao, Koji Watari, Hiroyuki Hayashi,
and Mikito Kitayama
Processing and Microstructure
of Nonoxide Ceramics
In general, because of their low self-
diffusion coefficients even at high tem-
peratures, nonoxide ceramics with high
covalent bonding, such as Si3N4, AlN, and
SiC, are densified with the aid of sintering
additives. SiC is generally densified through
solid-phase sintering by adding small
amounts of additives such as B, C,11 and
Be compounds,12 although liquid-phase
sintering of SiC using oxides or carbides
has recently attracted attention as a method
for developing anisotropic grains that act
as reinforcements.13 On the other hand,
AlN and Si3N4 are densified by liquid-
phase sintering to lower the sintering
temperatures because these nitrides, in
particular Si3N4, are not very stable at
higher temperatures, as shown in Table I.
A wide variety of oxides, such as alkaline-
earth oxides and rare-earth oxides, are used
as sintering additives for AlN and Si3N4.
During sintering, these additives react
Introduction
Since the confirmation that nonmetallic
for alternative materials with both high
thermal conductivity and superior mechani-
cal properties in order to manage the ever-
increasing heat generated by electric devices.
Silicon nitride (Si3N4) ceramic, on the
other hand, is well known as a high-
temperature structural ceramic having
high strength and fracture toughness. The
excellent mechanical properties of Si3N4
result from its unique microstructure,
which is composed of hexagonal, rodlike
grains, bonded together and reinforcing
each other.3–5 Recently, a Si3N4 ceramic
single crystals with a diamond-like struc-
ture, such as SiC, BP, and AlN, have high
intrinsic thermal conductivities of over
ꢀ1
ꢀ1 1,2
300 W m K , a great deal of effort
has been focused on the development of
nonoxide polycrystalline ceramics with
high thermal conductivity. SiC and AlN
ceramics with high thermal conductivities
of over 200 W mꢀ1 Kꢀ1 have been success-
fully fabricated, and they are used com-
mercially as high thermal conductivity
substrates, heatsinks, and so on. However,
widespread use of these materials is still
restricted, owing to the low reliability that
arises from their poor mechanical proper-
ties. In addition, application of SiC ceram-
ics is further limited by their low electric
resistance and high dielectric constant.
Hence, electrical industries actively search
with thermal conductivities of over
ꢀ1
100 W m
K
ꢀ1 was reported.6–9 In addi-
tion, considering the similarity of chemi-
cal bonding and crystal structure of Si3N4
to those of SiC and AlN, Haggerty and
Lightfoot10 pointed out that single-crystal
Si3N4 has a high intrinsic thermal conduc-
Table I: Comparison of Si3N4 with AlN and SiC.
Si3N4
AlN
SiC
Single-Crystal Form
Crystallography
Decomposition temperature
Intrinsic thermal conductivity
ꢁ-Si3N4: hexagonal (P63)
1850ꢂC
200 W mꢀ1 Kꢀ1 or 320 W mꢀ1 Kꢀ1
(Reference 10)
Hexagonal (wurtzite)
2200ꢂC (sublimation)
320 W mꢀ1 kꢀ1 (Reference 1)
Polytypism (3C, 2H, 4H, 15R, etc.)
2800ꢂC (sublimation)
490 W mꢀ1 kꢀ1 (Reference 1)
Polycrystalline Form
General sintering process
Liquid-phase sintering
Mainly amorphous
Liquid-phase sintering
Mainly crystalline
Solid-phase sintering
(Liquid-phase sintering)
—
Secondary phase
(ꢀ1 W mꢀ1 Kꢀ1
Glass pocket
Thin film
)
(ꢀ10 W mꢀ1 Kꢀ1
)
Morphology of secondary phase
Dispersion (Thin film)
—
Highest experimental conductivity
Bending strength*
Fracture toughness*
155 W mꢀ1 Kꢀ1 (Reference 33)
600–1500 MPa
5–7 MPa m1/2
ꢀ260 W mꢀ1 Kꢀ1 (Reference 39)
300–350 MPa
3–4 MPa m1/2
270 W mꢀ1 Kꢀ1 (Reference 12)
380–800 MPa
3.5–4.5 MPa m1/2
*Data taken from catalog of commercially supplied materials.
MRS BULLETIN/JUNE 2001
451

High Thermal Conductivity Silicon Nitride Ceramic
Figure 1. Fractured surfaces of (a) Si3N4, (b) AlN, and (c) SiC ceramics.
can be discussed based on the results ob-
in aluminum sites that scatter phonons and
thereby lowers the thermal conductivity.1
Therefore, in developing high thermal
conductivity AlN ceramic, much attention
has been paid to researching the sintering
additives that are able to remove oxygen
from AlN crystals.16,19 Rare-earth oxides
are known to be suitable additives for pu-
rifying the AlN crystals during sintering.16,19
In the case of Si3N4 ceramic, it is well
recognized that the use of Al2O3, one of
the available sintering additives for Si3N4
ceramic, considerably lowers the thermal
conductivity because Al2O3 reacts with
Si3N4 to form the solid solution of ꢁ-Si3N4,
which has the chemical formula of
Si6ꢀzAlzOzN8ꢀz (referred to as ꢁ-Sialon,
0 ꢄ z ꢄ 4.2).7,20 ꢁ-Sialon is formed by the
simultaneous equivalent substitution of
Al-O for Si-N in ꢁ-Si3N4. However, there
have been no reports on whether oxygen
alone can be incorporated into the ꢁ-Si3N4
crystal lattice. Recently, Kitayama et al.21
succeeded in measuring the oxygen con-
with impurity oxides largely present as a
surface film on the nitride powder, Al2O3
in AlN and SiO2 in Si3N4, respectively, to
form a liquid phase, which enhances den-
sification through the rearrangement of
particles and a solution re-precipitation
reaction. In silicon nitride, there are two
crystalline phases designated as ꢃ and ꢁ,
which are the low- and high-temperature
phases, respectively. Since the ꢃ-to-ꢁ phase
transformation during liquid-phase sin-
tering promotes the development of elon-
gated ꢁ-Si3N4 grains,14 ꢃ-Si3N4 powder is
generally used as a starting raw powder.
After sintering, the liquid phase remains
in the corners of or along the matrix grains
as a glassy phase or crystalline phases (re-
ferred to as secondary phases), the volume
fractions of which are typically 5–10%.
Typical fractured surfaces of liquid-phase-
sintered Si3N4, as well as of liquid-phase-
sintered AlN and solid-phase-sintered
SiC, are shown in Figures 1a, 1b, and 1c,
respectively. SiC and AlN exhibit homo-
geneous microstructures composed of
equiaxial grains. In addition, SiC shows a
transgranular mode of fracture, owing to
the strong bonding between grains. Poor
fracture toughness of SiC and AlN is as-
cribed to such homogeneous microstruc-
ture. In contrast, Si3N4 is composed of
rodlike grains, reflecting the preferential
growth in the [001] direction in ꢁ-Si3N4 .15
These rodlike grains, in turn, act as re-
inforcements that promote crack-bridging
processes and toughen the Si3N4. Figure 2
illustrates the microstructural factors of
Si3N4 affecting thermal conductivity. Lat-
tice defects in ꢁ-Si3N4 grains and grain-
boundary phases are considered to be
extrinsic factors affecting the thermal con-
ductivity of Si3N4. Although the micro-
structures of AlN and Si3N4 are different,
the chemistry and processing of them are
similar in many ways, as summarized in
Table I. The extrinsic factors governing the
thermal conductivity of Si3N4, therefore,
tained for AlN, as described next.
Effect of Impurity Atoms
onThermal Conductivity
The conduction of heat in dielectric ce-
ramics is dominated by phonon transport:
at room temperature, a greater influence
in phonon scattering is attributed to im-
perfections in the crystal lattice, for example,
impurity atoms, interstitials, and vacan-
cies. Slack et al. have systematically inves-
tigated the effect of impurities on thermal
conductivities of nonmetallic crystals, and
have revealed the strong correlation be-
tween the concentration of lattice oxygen
and the thermal conductivity of single-
crystal AlN.1,2 Several additional studies
on polycrystalline AlN ceramic were car-
ried out,16–18 and it is now well recognized
that oxygen as an impurity is a dominant
factor lowering the thermal conductivity
of AlN ceramic. It is assumed that oxygen
dissolved in an AlN lattice creates vacancies
Figure 2. Microstructural factors of Si3N4 ceramic affecting thermal conductivity.
452
MRS BULLETIN/JUNE 2001

High Thermal Conductivity Silicon Nitride Ceramic
tent in the ꢁ-Si3N4 crystal lattice by the
hot-gas extraction method, revealing that
oxygen can be dissolved into the ꢁ-Si3N4
lattice up to at least 0.4% by mass. Based
on electron spin resonance analysis,22 they
also reported that the incorporation of
oxygen into the ꢁ-Si3N4 lattice occurs by
the dissolution of SiO2 in the following
plausible reaction:
2SiO2 l 2SiSi ꢅ 4ON ꢅ VSi ,
(1)
where ON and VSi denote oxygen in a nitro-
gen site and a vacant silicon site, respec-
tively. Applying these analytical methods
to hot-pressed Si3N4 with Y2O3 and SiO2 ,
they have established that the thermal
conductivity of ꢁ-Si3N4 ceramic increases
with decreasing oxygen content in the
Si3N4 single-crystal lattice. This investiga-
tion also indicated that the lattice oxygen
content varies with the Y2O3/SiO2 ratio.
As the grain growth of ꢁ-Si3N4 after phase
transformation occurs such that smaller
ꢁ-Si3N4 grains dissolve into the liquid
phase to re-precipitate on larger ꢁ-Si3N4
grains, the lattice oxygen content might
be closely related to the composition of
the liquid phase. Kitayama et al. have
explained the variation of lattice oxygen
with the Y2O3/SiO2 ratio, based on the
thermodynamic consideration of the grain-
boundary phase.22 In addition to the
Y2O3/SiO2 ratio, a family of rare-earth
oxides, Re2O3 (Re ꢀ Sc, Yb, Y, Nd, etc.),
affects the thermal conductivity of Si3N4
ceramic. It was revealed that as the ionic
radius of the rare-earth element decreases,
grain growth of ꢁ-Si3N4 is enhanced, lattice
oxygen of ꢁ-Si3N4 grains decreases, and
hence thermal conductivity of the speci-
men increases.23 The results indicated that
Yb2O3 and Y2O3 are suitable additives for
fabricating high thermal conductivity Si3N4,
compared with other rare-earth oxides.
Hayashi et al.24 have further investigated
the relation between lattice oxygen content
and thermal conductivity, using ꢁ-Si3N4
as a raw powder in order to exclude
the influence of phase transformation. In
their study, ꢁ-Si3N4 with concurrent ad-
dition of Yb2O3 and MgO was sintered at
1900ꢂC under 0.9 MPa nitrogen pressure.
Figure 3 shows the effect of sintering
time on the lattice oxygen content and
the thermal conductivity of sintered Si3N4.
The lattice oxygen, which in raw ꢁ-Si3N4
powders was about 0.4% by mass, de-
creased with increasing sintering time. As
a result, the thermal conductivity of the
specimen increased with sintering time,
Figure 3. Effect of sintering time on
the lattice oxygen content and the
thermal conductivity of Si3N4 with
2mol%Yb2O3-5mol%MgO sintered
at 1900ꢂC under 0.9 MPa N2.
with sintering time was observed, as
shown in Figure 4. It is therefore evident
that a decrease in lattice oxygen content,
namely, purification of ꢁ-Si3N4 grains, is
significantly affected by the solution re-
precipitation process, which results in
grain growth of ꢁ-Si3N4. Figure 5 summa-
rizes the relation between the lattice oxy-
gen content in ꢁ-Si3N4 and the thermal
resistivity of these specimens, along with
the results obtained by Kitayama et al.22
Although processing method, type of raw
Si3N4 powder, and additive system used
were quite different between these investi-
gations, there is a clear tendency for ther-
mal resistivity to decrease with decreasing
lattice oxygen content. The result suggests
that lattice oxygen content in ꢁ-Si3N4 is a
dominant extrinsic factor governing the
thermal conductivity of Si3N4 ceramic, as
has been demonstrated in AlN ceramic.
Moreover, by extrapolation, it can be said
that ꢁ-Si3N4 free from lattice oxygen should
exhibit at least a thermal conductivity of
Figure 4. Variation of microstructure of
Si3N4 with 2mol%Yb2O3-5mol%MgO
with sintering time. Specimens were
sintered at 1900ꢂC under 0.9 MPa N2.
ꢀ1
about 180 W mꢀ1 K .
Effect of Secondary Phases
onThermal Conductivity
The morphology of the secondary phases
in AlN and Si3N4 ceramics, in addition to
the lattice oxygen content, negatively af-
fects the thermal conductivity to some
extent because these phases have thermal
conductivities 1–2 orders of magnitude
lower than pure AlN or Si3N4 crystals. For
example, the thermal conductivities of
yttrium aluminate (typical secondary phase
in AlN with Y2O3 addition) and oxynitride
glass (typical secondary phase in Si3N4)
are about 10 W mꢀ1 Kꢀ1 and 1 W mꢀ1 K ,
ꢀ1
and a high thermal conductivity of about
respectively. Considering the effect of dis-
tribution of a secondary phase with low
thermal conductivity based on the simple
dispersed model,25 a continuous distribu-
Figure 5. Effect of lattice oxygen content
on thermal resistivity of Si3N4 fabricated
by gas-pressure sintering ( ) and
hot-pressing ( , ).
ꢀ1
120 W m
K
ꢀ1 could be achieved in a
specimen sintered for a very long time.
Considerable grain growth of ꢁ-Si3N4
MRS BULLETIN/JUNE 2001
453

High Thermal Conductivity Silicon Nitride Ceramic
tion around the matrix grains extensively
lowers the thermal conductivity, while an
isolated distribution does not appreciably
affect the thermal conductivity. In the case
of AlN ceramic, it has been confirmed that
the secondary phases have little influence
on the thermal conductivity because they
are present largely as isolated crystalline
phases, and that the amount of secondary
phase in AlN ceramic has a major influ-
ence on thermal conductivity only if the
lattice oxygen content in the AlN is very
low and the amount of secondary phase is
rather large.18,26,27
creases quickly as the size of the ꢁ-Si3N4
phase is pre-existing ꢁ particles, 36,37 and
after transformation, only some large ꢁ
grains selectively grow, particularly along
the c axis direction, by a solution re-
precipitation reaction.38 Figure 6 shows
the typical microstructure of seeded and
extruded Si3N4. The material exhibits high
anisotropy where large elongated grains,
grown from seeds, are almost unidirec-
tionally oriented parallel to the extruding
direction.35 Table II summarizes the thermal
conductivities, measured in three different
directions, for textured Si3N4 fabricated
by this method with different sintering
additives. As expected, for all of the speci-
mens, thermal conductivity is highest along
the grain alignment. In particular, highly
anisotropic Si3N4 with very long grains,
fabricated under extreme conditions, ex-
hibits a high thermal conductivity of about
150 W mꢀ1 Kꢀ1 in the direction parallel to
the grain alignment.33
grains decreases to less than 1 ꢆm.30 From
this viewpoint, in addition to the role of
purifying the ꢁ-Si3N4 lattice, grain growth
is necessary in order to improve the ther-
mal conductivity of Si3N4 ceramic. However,
it should be noted that grain-boundary
thin films have little effect on the thermal
conductivity of Si3N4 when the grain size
exceeds certain critical values (about 1 ꢆm).
HighThermal Conductivity in
Anisotropic Microstructures
Another approach for achieving high
thermal conductivity in Si3N4 ceramic is to
develop a textured microstructure in which
elongated ꢁ-Si3N4 grains are oriented al-
most unidirectionally. High thermal con-
ductivity along the grain orientation is
expected, compared with material with a
random distribution of ꢁ-Si3N4 grains, for
the following reasons: the alignment of
elongated ꢁ-Si3N4 grains, in turn, causes
the glassy phase to be distributed parallel
to the grain orientation, which minimizes
the negative effect of the glassy phase with
low thermal conductivity; in addition,
high-resolution thermoreflectance micros-
copy measurements of the extra large
ꢁ-Si3N4 grains reveal that the thermal dif-
fusivity and conductivity of ꢁ-Si3N4 crys-
tals along the caxis, namely, along the length
of the rodlike grains, are higher than those
along the a axis.31 Detailed microthermal
characterization of grains in ceramic ma-
terials is mentioned in the article by
Fournier in this issue.
Compared with AlN ceramic, the micro-
structure of Si3N4 ceramic is more compli-
cated, owing to the well-faceted grains
with a rodlike shape, as illustrated in Fig-
ure 2. Most of the secondary phase is gen-
erally present as a glassy phase because
the liquid phase containing SiO2 is dif-
ficult to crystallize perfectly. It should
be noted that the thermal conductivity of
ꢀ1
glass can be as low as 1 W mꢀ1 K , which
is about one order of magnitude lower
than the thermal conductivity of the crys-
talline secondary phase in AlN ceramic. In
addition, this glassy phase is distributed
not only as glass pockets surrounded by
ꢁ-Si3N4 grains, but also in the openings
between two grains, as most of rodlike
ꢁ-Si3N4 grains incline toward each other. It
is reasonable to think that the negative effect
of the secondary phase on thermal con-
ductivity is somewhat more pronounced
in Si3N4 ceramic than AlN ceramic.
In addition, grain-boundary thin films
with an equilibrium thickness (typically
1–2 nm) are generally present between
two-grain junctions in liquid-phase-sintered
Si3N4 ceramic.28 A calculation based on a
simple modified Wiener’s model29 for the
thermal conductivity of a composite mate-
rial predicts that the thermal conductivity
of ꢁ-Si3N4 ceramic having grain-boundary
films of a few nanometers’ thickness de-
Such anisotropic microstructures can
be fabricated by the combination of the
seeding of rodlike ꢁ-Si3N4 nuclei and a
forming process generating shear stress,
such as tape casting and extrusion.6,32–35
This process is based on the grain-growth
behavior of silicon nitride: during ꢃ-to-ꢁ
phase transformation, the preferential
nucleation site of the newly formed ꢁ-Si3N4
Figure 6. Microstructure of textured
Si3N4 ceramic.
Table II:Thermal Conductivity inThree Directions forTextured Si3N4 fabricated by Combined Seeding andTape Casting.
Thermal Conductivity
Annealing Conditions
(W mꢀ1 Kꢀ1
)
Sample No.
Sintering Additives
(temperature, time, pressure)
(x direction, y direction, z direction)
References
1
5 mass% Y2O3
2123 K
66 h
121, 75, 60
Hirao et al.6
0.9 MPa N2
2773 K
2 h
200 MPa N2
2423 K
4 h
2
3
5 mass% Y2O3
155, 67, 52
137, 97, 72
Watari et al.33
0.5 mol% Y2O3
0.5 mol% Nd2O3
Hirosaki et al.34
30 MPa N2
454
MRS BULLETIN/JUNE 2001

High Thermal Conductivity Silicon Nitride Ceramic
2. G.A. Slack, R.A. Tanzilli, R.O. Pohl, and J.W.
Vandersande, J. Phys. Chem. Solids 48 (1987)
p. 641.
3. E. Tani, S. Umebayashi, K. Kishi, K. Kobayashi,
and M. Nishijima, Am. Ceram. Soc. Bull. 65
(1986) p. 1311.
4. C.W. Li and J. Yamanis, Ceram. Eng. Sci. Proc.
10 (1989) p. 632.
5. P.F. Becher, J. Am. Ceram. Soc. 74 (1991) p. 255.
6. K. Hirao, K. Watari, M.E. Brito, M. Toriyama,
and S. Kanzaki, J. Am. Ceram. Soc. 79 (1996)
p. 2485.
7. N. Hirosaki, Y. Okamoto, M. Ando, F.
Munakata, and Y. Akimune, J. Am. Ceram. Soc.
79 (1996) p. 2878.
8. K. Watari, K. Hirao, M. Toriyama, and K.
Ishizaki, J. Am. Ceram. Soc. 82 (1999) p. 777.
9. N. Hirosaki, Y. Okamoto, F. Munakata, and
Y. Akimune, J. Eur. Ceram. Soc. 19 (1999) p. 2183.
10. J.S. Haggerty and A. Lightfoot, Ceram. Eng.
Sci. Proc. 16 (1995) p. 475.
21. M. Kitayama, K. Hirao, A. Tsuge, M.
Toriyama, and S. Kanzaki, J. Am. Ceram. Soc. 82
(1999) p. 3263.
22. M. Kitayama, K. Hirao, A. Tsuge, K. Watari,
M. Toriyama, and S. Kanzaki, J. Am. Ceram. Soc.
83 (2000) p. 1985.
23. M. Kitayama, K. Hirao, K. Watari, M.
Toriyama, and S. Kanzaki, J. Am. Ceram. Soc. 84
(2001) p. 353.
24. H. Hayashi, M. Kitayama, K. Hirao, Y.
Yamauchi, and S. Kanzaki, submitted to J. Am.
Ceram. Soc.
25. W.D. Kingery, J. Am. Ceram. Soc. 42 (1959)
p. 617.
26. H. Buhr and G. Müller, J. Eur. Ceram. Soc. 12
(1993) p. 271.
27. W.-J. Kim, D.K. Kim, and C.H. Kim, J. Am.
Ceram. Soc. 79 (1996) p. 1066.
28. H.-J. Kleebe, M.K. Cinibulk, R.M. Cannon,
and M. Rühle, J. Am. Ceram. Soc. 76 (1993) p. 1969.
29. O. Wiener, Phys. Z. 5 (1904) p. 332.
30. M. Kitayama, K. Hirao, K. Watari, M.
Toriyama, and S. Kanzaki, J. Am. Ceram. Soc. 82
(1999) p. 3105.
31. B. Li, L. Pottier, J.P. Roger, D. Fournier, K.
Watari, and K. Hirao, J. Eur. Ceram. Soc. 19
(1999) p. 1631.
32. K. Hirao, M. Ohashi, M.E. Brito, and S.
Kanzaki, J. Am. Ceram. Soc. 78 (1995) p. 1687.
33. K. Watari, K. Hirao, M.E. Brito, M. Toriyama,
and S. Kanzaki, J. Mater. Res. 14 (1999) p. 1538.
34. N. Hirosaki, M. Ando, Y. Okamoto, F.
Munakata, Y. Akimune, K. Hirao, K. Watari,
M.E. Brito, M. Toriyama, and S. Kanzaki,
J. Ceram. Soc. Jpn. 104 (1996) p. 1171.
35. H. Teshima, K. Hirao, M. Toriyama, and
S. Kanzaki, J. Ceram. Soc. Jpn. 107 (1999) p. 1216.
36. G. Petzow and M.J. Hoffmann, Mater. Sci.
Forum 113–115 (1993) p. 91.
Summary
The thermal conductivity of Si3N4 ceramic
at room temperature is basically governed
by the lattice oxygen content in the ꢁ-Si3N4
crystal. Therefore, in order to improve
thermal conductivity, it is important to pu-
rify the grains, as has been demonstrated
in AlN ceramic. The purification of ꢁ-Si3N4
grains can be achieved through a solution
re-precipitation reaction during sintering
using sintering additives that have a great
affinity for SiO2, such as Y2O3 and Yb2O3.
As far as the authors know, Si3N4 ceramic
with a high thermal conductivity of over
100 W mꢀ1 Kꢀ1 has been fabricated only by
gas-pressure sintering. From a commercial
point of view, it is very important to fabri-
cate high thermal conductivity Si3N4 ce-
ramic by a conventional sintering technique
such as pressureless sintering. Seeding,
combined with careful control of grain-
boundary composition, may be one of the
processing strategies for this purpose. In
addition, microstructure design for har-
monizing high thermal conductivity with
good mechanical and electrical properties
(such as low dielectric constant) is also
very important for the widespread use of
Si3N4 ceramic as a high thermal conduc-
tivity material.
11. S. Prochazka and R.J. Charles, Am. Ceram.
Soc. Bull. 52 (1973) p. 885.
12. Y. Takeda, K. Usami, K. Nakamura, S.
Ogihara, K. Maeda, T. Miyoshi, S. Shinozaki,
and M. Ura, Advances in Ceramics, Vol. 7, edited
by M.F. Yan and A.H. Heuer (American Ce-
ramic Society, Columbus, OH, 1984) p. 253.
13. N.P. Padture, J. Am. Ceram. Soc. 77 (1994)
p. 519.
14. F.F. Lange, J. Am. Ceram. Soc. 62 (1979)
p. 428.
15. C.M. Hwang, T.Y. Tien, and I.-W. Chen,
in Proc. Sintering ‘87, edited by S. Somiya,
M. Shimada, M. Yoshimura, and R. Watanabe
(Elsevier Applied Science Publishers, Essex,
1988) p. 1034.
16. A.V. Virkar, T.B. Jackson, and R.A. Cutler,
J. Am. Ceram. Soc. 72 (1989) p. 2031.
17. R.R. Lee, J. Am. Ceram. Soc. 74 (1991) p. 2242.
18. T.B. Jackson, A.V. Virl, K.L. More, R.B.
Dinwiddie Jr., and R.A. Cutler, J. Am. Ceram.
Soc. 80 (1997) p. 1421.
19. K. Watari, H.J. Hwang, M. Toriyama, and
S. Kanzaki, J. Mater. Res. 14 (1999) p. 1409.
20. K. Watari, Y. Seki, and K. Ishizaki, J. Ceram.
Soc. Jpn. 97 (1989) p. 56.
Acknowledgments
The authors gratefully acknowledge
helpful discussions with Drs. S. Kanzaki,
M. Toriyama (National Institute of Ad-
vanced Industrial Science and Technology),
and K. Maeda (Fine Ceramics Research
Association).
37. W. Dressler, H.-J. Kleebe, M.J. Hoffmann,
M. Rühle, and G. Petzow, J. Eur. Ceram. Soc. 16
(1996) p. 3.
38. H. Emoto and M. Mitomo, J. Eur. Ceram.
Soc. 17 (1997) p. 797.
References
39. H. Buhr, G. Müller, H. Wiggers, F. Aldinger,
P. Foley, and A. Roosen, J. Am. Ceram. Soc. 74
1. G.A. Slack, J. Phys. Chem. Solids 34 (1973)
p. 321.
(1991) p. 718.
ꢀ
Announcing...
Nonmembers and institutions may now purchase individual MRS proceedings articles
online at www.mrs.org/publications/epubs/
Thousands of papers will be available in the weeks to come.
Of course, MRS members continue to have free access to these papers
as part of their membership dues. To apply for membership,
visit www.mrs.org/membership/forms/
Or contact:
Member Services, Materials Research Society
506 Keystone Drive, Warrendale, PA 15086 USA
Tel: 724-779-3003 • Fax: 724-779-8313 • E-mail: info@mrs.org
ꢀ
www.mrs.org/publications/bulletin
MRS BULLETIN/JUNE 2001
455