Synthesis of Ti3SiC2 from powder blend of Ti, Si and TiC

Article abstract of DOI:10.1016/j.jallcom.2005.12.112

Titanium silicon carbide Ti₃SiC₂ was synthesized from a powder blend of Ti, Si and TiC at a molar ratio of Ti:Si:TiC = 2:2:3. The powder blend was pressed and shaped into a cylinder and then sintered at various temperatures in vacuum

Full text of DOI:10.1016/j.jallcom.2005.12.112

Journal of Alloys and Compounds 426 (2006) 263–267
Synthesis of Ti SiC from powder blend of Ti, Si and TiC
3
2
a,∗
a
b
a
Hitoshi Hashimoto , Zheng Ming Sun , Yosuke Inoue , Shuji Tada
a
Material Research Institute for Sustainable Development, National Institute of Advanced Industrial Science and Technology,
Shimoshidami, Moriyama-ku, Nagoya 463-8560, Japan
b
Student of Aichi Institute of Technology, Yagusa-cho, Toyota 470-0392, Japan
Received 11 November 2005; accepted 19 December 2005
Available online 27 March 2006
Abstract
Titanium silicon carbide Ti
3
2
SiC was synthesized from a powder blend of Ti, Si and TiC at a molar ratio of Ti:Si:TiC = 2:2:3. The powder blend
was pressed and shaped into a cylinder and then sintered at various temperatures in vacuum. X-ray diffraction analysis was conducted on the
cylinders sintered at various temperatures. Results of the XRD analysis revealed that Ti Si , the most thermodynamically stable phase in Ti–Si
system, was preferentially formed by the reaction between Ti and Si powders followed by the formation of TiSi from remaining Ti and Si powders
in a relatively low temperature range. These silicides reacted with TiC to form Ti SiC at higher temperatures. The final phases in the cylinders
sintered at temperatures over 1729 K were Ti SiC and Ti Si , which well agrees with those estimated by a Ti–Si–C phase diagram. On the other
5
3
2
3
2
3
2
5
3
hand, final phases in disks sintered under pressures of 14 MPa or more by using a carbon mold were found to be Ti
the equilibrium changes with the pressure.
3 2
SiC and TiC, which suggests
©
2006 Elsevier B.V. All rights reserved.
Keywords: Ceramics; Solid-state reactions
1
. Introduction
pure polycrystalline Ti3SiC2 by CVD. However the size was
very small (40 mm in length, 12 mm in width and 0.4 mm in
thickness). Only the solid-state reaction of powder blends can
offer a practical synthetic method for bulk Ti3SiC2. However,
this method mostly produces a considerable amount of TiC or
Ti–Si binary phases as byproducts. According to Lis et al. [6],
the best result was the formation of 82 vol.% of T3SiC2, with
at least 18 vol.% of TiC, by the solid-state reaction of elemental
Ti/Si/C powder blend of stoichiometric composition using self-
propagating high-temperature synthesis (SHS) or combination
of SHS and hot isostatic pressing (HIP). Barsoum and El-Raghy
[1] succeeded in synthesizing Ti3SiC2 with purity more than
99 vol.% by hot press or HIP of a Ti/SiC/C powder blend of the
stoichiometric composition, which led to a breakthrough in the
practical use of this material.
The authors [10] also succeeded in synthesizing Ti3SiC2 with
the purity more than 98 vol.% by pulse discharge sintering (PDS)
of a powder blend of Ti/Si/TiC. In the solid-state reaction of
the elemental Ti/Si/C powder blend, we thought, Ti–Si binary
phases and TiC were formed and then Ti3SiC2 was formed by
the reaction between these phases. According to the binary phase
diagrams [11], TiC has a wide compositional range on Ti-rich
side (Ti: 51.2–68 at.%). Therefore, Ti-rich TiC was formed by
Titanium silicon carbide, Ti3SiC2 shows a relatively
high elastic modulus (320 GPa) and oxidation resistance
−
8
2
−4 −1
(
parabolic rate constant of 2 × 10 kg m
s at 1273 K)
as a ceramic material, while it shows an electrical con-
6
−1 −1
m ) and a thermal conductivity
ductivity (4.5 × 10 ꢀ
−1
−1
(
43 W m
K ) both about two times larger than pure Ti, and
an excellent thermal shock resistance (bending strength is not
affected by quenching from 1400 C into water) and the machin-
◦
ability comparable to graphite [1]. As these attractive prop-
erties arise from its crystal structure itself, the fine control of
microstructure is not required, unlike most advanced materials.
Ti3SiC2 had been synthesized by Jeitschko and Nowotny [2]
in 1966 for the first time by heating a powder blend of TiH2, Si
◦
and graphite up to 2000 C in a graphite capsule. Since then,
Ti3SiC2 has been synthesized by chemical vapor deposition
(
CVD) [3], arc-melting [4] and solid-state reaction of powder
blends [5–9]. Goto and Hirai [3] successfully synthesized almost
∗
Corresponding author.
E-mail address: hitoshi-hashimoto@aist.go.jp (H. Hashimoto).
0
925-8388/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2005.12.112

2
64
H. Hashimoto et al. / Journal of Alloys and Compounds 426 (2006) 263–267
the reaction between the elemental Ti and C, which caused the
shortage of Ti for the formation of Ti3SiC2 and therefore the
residue of TiC. So we got an idea that use of a commercial TiC
powder having a very narrow compositional range as one of the
starting materials instead of C powder can avoid the shortage
of Ti for the formation of Ti3SiC2. We investigated the com-
position of a powder blend of Ti/Si/TiC for the formation of
pure Ti3SiC2 and found that a nonstoichiometric composition
(
Ti:Si:TiC = 2:2:3 in molar ratio) made Ti3SiC2 of 98 vol.% or
more.
In this report, the reaction route and the final phases were
investigated in the solid-state reaction of Ti/Si/TiC powder blend
with the nonstoichiometric compositions in order to obtain the
basis of pure Ti3SiC2 synthesis.
Fig. 2. Illustrated configuration of holes made on cylindrical powder compact
for temperature measurement by infrared thermometer and thermocouple (cross-
sectional view).
2
. Experimental
Commercially available Ti powder (99.9% pure, 10 ␮m in particle size), Si
powder (99.9%, −10 ␮m) and TiC powder (99%, 2 or 5 ␮m) were used in this
experiment. A chemical analysis was done on the TiC powder because that TiC
has the wide compositional range as mentioned above. The result showed that
the TiC powder contained 79.6 wt.% of Ti and 19.6 wt.% of C, corresponding to
TiC0.99. These powders were blended at a molar ratio of Ti:Si:TiC = 2:2:3 using
a Turbular mixer for 86.4 ks. A 7-g powder blend was pressed under a pressure
of 200 MPa and shaped into a cylinder of 14 mm in diameter and 17 mm in
height using a die set and a press machine. A hole of 5 mm in diameter and
and the temperature was measured by the thermometer and thermocouple. The
result is shown in Fig. 3. The thermometer indicates lower temperatures than the
thermocouple but a straight line can express the relation between two readings.
The calibration of thermometer was done in a temperatures range below 1423 K
because that a stainless steel sheath of the thermocouple is likely to react with
the powder compact at temperatures over 1423 K. Temperatures measured by
the thermometer were corrected using the straight line extrapolated to higher
temperatures over 1423 K, and the temperatures described in the following part
of this paper are the corrected values.
7
.5 mm in depth was made at the axial center of the cylindrical powder compact
for temperature measurement by an infrared thermometer. The powder compact
was heated in vacuum using a carbon mold and a pulse discharge sintering
machine (Sodick PAS-V) as shown in Fig. 1. The powder compact was insulated
electrically using alumina plates put on both ends of it. The compact was fixed
with an upper carbon punch put on it but no load was applied to the upper punch.
The load was applied to carbon blocks on the carbon mold to keep electrical
contact between electrodes of the sintering machine and the carbon mold. Thus
the electric current passed only through the blocks and the mold and heated them.
The powder compact was heated from outside by the mold under a very small
pressure (The load applied to the powder compact was merely the weight of the
upper punch.). The temperature of the powder compact was measured by the
infraredthermometerfocusedonbottomoftheholemadeonthecompact. Before
experiment, the thermometer was calibrated for the powder compact because the
thermometer had been already adjusted to carbon. The calibration was conducted
by using the compact on which another hole for the temperature measurement by
a K-type thermocouple was bored at right angle to the hole for the thermometer
The powder compact was heated in vacuum at a rate of 0.33 K/s up to 1307,
1412, 1518, 1570, 1623, 1729, 1834 and 1940 K, kept for 900 s and cooled in
the sintering machine. For comparison, a 17-g powder blend was sintered under
a pressure of 50 MPa at 1373, 1473, 1573, 1623 and 1673 K by using a carbon
mold having a 30-mm inner diameter. The temperature was measured by the
thermometer focused on the bottom of a hole (5 mm in diameter and 10 mm
in depth) made on the carbon mold at its axial center. The heating rate and
the period kept at the maximum temperature were the same as the sintering of
powder compact. As the sintered disc had reacted with the carbon mold to form
TiC near the surface of sintered disc when sintering at temperatures of 1623
and 1673 K in a preparative experiment, thin graphite sheets were put between
the powder blend and the carbon mold to prevent the reaction between the disc
and mold. The sintered disc reacted with carbon mold was so hard due to TiC
(
see Fig. 2). The compact was heated in the same way as shown in Fig. 1
Fig. 1. Schematic illustration for sintering experiment of cylindrical powder
compact in vacuum by a pulse discharge sintering machine.
Fig. 3. Relation between reading of infrared thermometer and that of thermo-
couple.

H. Hashimoto et al. / Journal of Alloys and Compounds 426 (2006) 263–267
265
Fig. 4. X-ray diffraction patterns of powder compact sintered in vacuum at 1307,
412, 1518, 1570, 1623 and 1729 K.
Fig. 6. Composition of starting powder blend plotted on a Ti–Si–C ternary phase
diagram at 1500 K.
1
formed near the surface that polishing was very difficult. The use of graphite
sheets successfully prevented the reaction and the polishing of sintered disc was
easy.
X-ray diffraction analysis was made on polished ends of the sintered com-
pacts and discs by using Philips X’pert MPD with Cu K␣ radiation to identify
the phases.
AccordingtoRileyetal. [12], theformationofTi5Si3 ispreceded
by the ␣/␤ transformation in Ti, which initiates interdiffusion of
Ti and Si, in the self-propagating high-temperature synthesis of
Ti Si3 from a powder blend of Ti and Si. As the ␣/␤ transfor-
5
mation occurs at 1138 K [11], it is reasonable that Ti Si3 was
5
already formed at 1307 K. (2) The formation of these titanium
silicides was terminated until 1412 K because of the complete
consumption of the elemental Ti and Si. It should be noted that
we blended the elemental Ti and Si powders at an equimolar ratio
3
. Result and discussion
X-ray diffraction patterns of the compacts sintered at 1307,
412, 1518, 1570, 1623 and 1729 K are shown in Fig. 4. Peaks
(
Ti:Si:TiC = 2:2:3). However, TiSi was not formed but Ti Si3
5
1
and TiSi2 were formed. In Ti–Si system, there are five inter-
metallic compounds, Ti3Si, Ti Si3, Ti Si4, TiSi and TiSi2 and
from the starting powders (Ti, Si and TiC) are seen together with
peaks from Ti Si3 when sintering at 1307 K. The peaks from Ti
and Si disappear at 1412 K and peaks from TiSi2 appear. In a
temperature range from 1518 to 1623 K, a rapid increase in the
intensity of peaks from Ti3SiC2 is seen, while the intensity of
peaks from TiC, Ti Si3 and TiSi2 decreases rapidly. To illustrate
the phase change with sintering temperature, relative intensity
of each main peak from Ti, Si, TiC, Ti Si3, TiSi2 and Ti3SiC2
is plotted against the temperature in Fig. 5. To analyze the reac-
tion route exactly, the in situ phase analysis during reaction is
required by using synchrotron radiation or neutron radiation.
However, we can estimate the reaction route roughly from this
figure as follows: (1) Ti Si3 was formed by the reaction between
the elemental Ti and Si and then the formation of TiSi2 started.
5
5
5
among them Ti Si3 is the most thermodynamically stable phase
5
with the highest formation enthalpy (ꢁH = −579 kJ/mol) [12].
It is presumed that the most stable phase, Ti Si3 was formed
5
primarily by the reaction between Ti and Si. TiSi and TiSi2 have
almost the same formation enthalpy, 130 and 135 kJ/mol, respec-
tively [13]. Therefore, the shortage of the elemental Ti powder
5
5
in the starting blend caused by the formation of Ti Si3 resulted
5
in the preferential formation of Si-rich compound, TiSi2. (3) The
formation of Ti3SiC2 by the reaction between these silicides and
TiC started at 1518 K and grew rapidly in the range from 1518
to 1623 K, which resulted in the rapid decrease in TiC and the
silicides. (4) The formation of Ti3SiC2 was almost terminated
5
until 1729 K and two phases, Ti3SiC2 and Ti Si3 remained until
5
1940 K as the final products.
The composition of starting powder blend is plotted on a
Ti–Si–C ternary phase diagram section at 1500 K [14] by symbol
X” in Fig. 6. The starting composition lies in a three-phase
region, Ti3SiC2–TiC–Ti Si3, very close to an immediate two-
“
5
phase region, Ti3SiC2–Ti Si3. Therefore the final products well
5
agree with those estimated by the phase equilibrium diagram.
On the other hand, the final products in the powder blend
sintered under the pressure of 50 MPa by using the carbon mold
were Ti3SiC2 and TiC as shown in Fig. 7. The final product
TiC could arise from the reaction between the sintered disc and
the graphite sheets. So the sintered disc was polished more than
1
mm and XRD pattern was measured again. As the XRD pattern
did not change, we believed that the final product TiC did not
arise from the reaction between the sintered disc and the graphite
Fig. 5. Variation of relative intensity of each main peak from Ti, Si, TiC, Ti5Si3,
TiSi2 and Ti3SiC2 with sintering temperature.

2
66
H. Hashimoto et al. / Journal of Alloys and Compounds 426 (2006) 263–267
Fig. 7. Variation of relative intensity of each main peak from TiC, Ti5Si3, TiSi2
and Ti3SiC2 with sintering temperature.
Fig. 9. Variation of volume fraction of Ti SiC in the sintered disc and density
3
2
of the disc with sintering pressure.
sheets. By comparing Fig. 7 with Fig. 5, the reaction route under
5
0 MPa seems to be the same as that under 0 MPa. That is, Ti Si3
5
the volume fraction could not be calculated for the disc sintered
under 0 MPa. From the figure, both the density and the volume
fraction seem to be almost independent on the sintering pressure
in a pressure range over 14 MPa. More detailed investigation on
the pressure effect should be done in the future.
According to Wu et al. [16] who investigated the reaction
route in the pressureless reactive sintering of a stoichiometric
powder blend of Ti/SiC/C by using neutron powder diffraction
was formed by the reaction between the elemental Ti and Si and
then the formation of TiSi2 started. Then Ti3SiC2 was formed
by the reaction between these silicides and TiC. Therefore, the
reaction route was not so affected by the pressure, while the final
products depended on the pressure.
Therefore, the effect of sintering pressure on the products in
sintered discs was investigated. The powder blend was sintered
at 1673 K for 900 s under a range of pressures from 0 to 60 MPa
by using the carbon mold. The variation of relative intensity
from each main peak from Ti3SiC2, TiC, Ti Si3 and TiSi2 with
sintering pressure is shown in Fig. 8. Ti3SiC2, TiC, Ti Si3 and
TiSi2 were seen in the disc sintered under 0 MPa, while only
Ti3SiC2 and TiC remained under 14 MPa or more. Thus Ti Si3
technique, intermediate phases TiCx and Ti Si3Cx (x ≤ 1) were
5
formed first. As C is soluble in Ti Si3 up to about 10 at.% in the
5
5
ternary equilibrium phase diagram [14], Ti Si3Cx is the same as
5
5
Ti Si3 inthiscase. Thesephaseswereconsumedintheformation
5
of Ti3SiC2 and TiCx remained. Thus Ti3SiC2 is formed by the
5
reaction between Ti Si3 and TiC and the final products tend to
5
tends to disappear under pressure, while TiC tends to remain.
Fig. 9 shows the variation of volume fraction of Ti3SiC2 in the
sintered disc and density of the disc with sintering pressure. The
volume fraction was calculated by the intensity of Ti3SiC2 and
TiC XRD peaks using the following formula [15]:
be Ti SiC2 and TiC for the stoichiometric powder blend. So
5
we prepared a stoichiometric powder blend of Ti/Si/TiC and
sintered in vacuum under 0 MPa. Fig. 10 shows the variation
of relative intensity of each main peak from Si, TiC, Ti Si3,
5
TiSi2 and Ti3SiC2 with sintering temperature. The reaction route
estimated from this figure is the same as the nonstoichiometric
powder blend. However, the final products are Ti3SiC2 and TiC.
The volume fraction of Ti3SiC2 calculated from the intensity
of Ti3SiC2 and TiC XRD peaks is 0.97 for 1729 K or more.
The stoichiometric powder blend produces Ti3SiC2 with a small
amount of TiC as the final products regardless of starting powder
combination.
Va = K/(K + I /Ia)
where Va is the fraction of Ti3SiC2, Ia and I the integrated inten-
sity of Ti3SiC2 (1 0 4) peak and TiC (2 0 0) peak, respectively,
and K is a constant having values of 1.8 for the calculation of
weight fraction and 1.95 for the volume fraction. As this formula
is applicable to composites consisting of Ti3SiC2 and TiC only,
(1)
b
b
Fig. 8. Variation of relative intensity of each main peak from TiC, Ti5Si3, TiSi2
and Ti3SiC2 with sintering pressure.
Fig. 10. Variation of relative intensity of each main peak from Si, TiC, Ti5Si3,
TiSi2 and Ti3SiC2 with sintering temperature.

H. Hashimoto et al. / Journal of Alloys and Compounds 426 (2006) 263–267
[3] T. Goto, T. Hirai, Mater. Res. Bull. 22 (1987) 1195–1201.
267
4
. Conclusions
[
[
4] S. Arunajatesan, A.H. Carim, J. Am. Ceram. Soc. 78 (1995) 667–672.
5] R. Pampuch, J. Lis, J. Piekarchuk, L. Stobierski, J. Mater. Synth. Process.
(1993) 93–100.
6] J. Lis, Y. Miamoto, R. Pampuch, K. Tanihata, Mater. Lett. 22 (1995)
63–168.
[7] J. Lis, R. Pampuch, T. Rudnik, Z. Wergzyn, Solid State Ionics 101–103
1997) 59–64.
Ti3SiC2 was synthesized from the powder blend of Ti, Si and
1
TiC at the molar ratio of Ti:Si:TiC = 2:2:3. From the result of
XRDanalysisonthecompactedpowderblendsinteredatvarious
temperatures in vacuum, the reaction route was estimated. As
[
1
(
a result, Ti Si3 was formed by the reaction between Ti and Si
5
[
[
8] A. Feng, T. Orling, Z.A. Munir, J. Mater. Res. 14 (1999) 925–939.
9] Y. Khoptiar, I. Gotman, J. Eur. Ceram. Soc. 23 (2003) 47–53.
and then the formation of TiSi2 started. These titanium silicides
reacted with TiC to form Ti3SiC2 in the next stage. The final
phases in the powder compacts sintered at temperatures over
[
10] Z.M. Sun, Z.F. Zhang, H. Hashimoto, T. Abe, Mater. Trans. 43 (2002)
28–431.
4
1
729 K were Ti3SiC2 and Ti Si3, which well agrees with those
[11] ASM’s Binary Alloy Phase Diagrams, 2nd ed., Plus Updates on CD-ROM,
Ver. 1.0, ASM International, 1996.
5
estimated by a Ti–Si–C phase diagram. On the other hand, the
final phases in the disks sintered under pressures of 14 MPa or
more by using a carbon mold were Ti3SiC2 and TiC, which
suggests the equilibrium changes with pressure.
[
[
12] D.P. Riley, C.P. Oliver, E.H. Kisi, Intermetallics 14 (2006) 33–38.
13] M. Doyama, M. Yabe, Data Handbook of Intermetallic Compounds, Sci-
ence Forum, Tokyo, 1989, p. 375.
[
[
[
14] P. Villars, S. Prince, H. Okamoto, Handbook of Ternary Alloy Phase Dia-
grams on CD-ROM, Ver. 1.0, ASM International, 1997.
15] Z.F. Zhang, Z.M. Sun, H. Hashimoto, Metall. Mater. Trans. 33A (2002)
References
3
321–3328.
16] E. Wu, E.H. Kisi, S.J. Kennedy, A.J. Studer, J. Am. Ceram. Soc. 84 (2001)
281–2288.
[
[
1] M.W. Barsoum, T. El-Raghy, J. Am. Ceram. Soc 79 (1996) 1953–1956.
2] W. Jeitschko, H. Nowotny, Monatasch Chem. 98 (1967) 329–337.
2