Ceramization of reflux-treated polymethylsilane precursors to silicon carbide

Article abstract of DOI:10.1111/j.1151-2916.2003.tb00004.x

The pyrolysis of polymethylsilane (PMS) in an argon gas environment with a flow rate of 1 L/min was investigated as a standard pyrolytic process, and the investigation showed Si - Si network formation at 573 K. Subsequently, various condensed PMS resins were prepared by adjusting pre-heat treatment or reflux conditions in the temperature range of 423-723 K. The effect of pre-heat treatment or refluxing on the ceramic yield at 1273 K was quantitatively evaluated. Structural evolution in the PMS resins prepared under various reflux conditions was investigated during pyrolysis up to 1873 K. The X-ray diffraction patterns of the pyrolysis products revealed crystallite growth of β-SiC and silicon at 1273-1473 K. β²⁹Si solid-state nuclear magnetic resonance with the single-pulse method was also conducted on, the pyrolysis products at 1273 K.

Full text of DOI:10.1111/j.1151-2916.2003.tb00004.x

J. Am. Ceram. Soc., 86 [2] 227–33 (2003)
journal
Ceramization of Reflux-Treated Polymethylsilane Precursors to
Silicon Carbide
Masaki Narisawa, Takashi Iseki, Yoshiro Katase, and Kiyohito Okamura*
Department of Metallurgy and Materials Science, Graduate School of Engineering, Osaka Prefecture University,
Sakai 599-8531, Japan
Kunio Oka and Takaaki Dohmaru
Institute for Advanced Science and Technology, Osaka Prefecture University, Sakai 599-8570, Japan
and a metal silicide,3,22–24 because the metal additive in the PMS
will react with excess silicon during pyrolysis.
The pyrolysis of polymethylsilane (PMS) in an argon gas
environment with a flow rate of 1 L/min was investigated as a
standard pyrolytic process, and the investigation showed
SiOSi network formation at 573 K. Subsequently, various
condensed PMS resins were prepared by adjusting pre-heat-
treatment or reflux conditions in the temperature range of
423–723 K. The effect of pre-heat treatment or refluxing on the
ceramic yield at 1273 K was quantitatively evaluated. Struc-
tural evolution in the PMS resins prepared under various
reflux conditions was investigated during pyrolysis up to 1873
K. The X-ray diffraction patterns of the pyrolysis products
revealed crystallite growth of ␤-SiC and silicon at 1273–1473
K. 29Si solid-state nuclear magnetic resonance with the single-
pulse method was also conducted on the pyrolysis products at
1273 K.
On the other hand, we recently found that the overall ceramic
yield of PMS was improved simply by heating the polymer at
423–723 K in a long vertical furnace with a reflux condenser
attached on the upper side of the sample cell.16,21 By adjusting the
reflux conditions, we obtained various types of condensed resins
from the PMS.
In this paper, the structural evolution of these condensed PMS
resins is investigated in terms of X-ray diffractometry (XRD)
patterns and solid-state 29Si nuclear magnetic resonance (NMR)
spectra obtained by the single-pulse method.2,4,5 Elemental anal-
yses on the resulting pyrolysis products are also described.
II. Experimental Procedure
(1) Preparation of PMS Samples and Pyrolysis Products
I. Introduction
The original PMS was prepared by a condensation reaction of
methyldichlorosilane with sodium, as described elsewhere.3,20 The
PMS obtained was a pale yellow, viscous liquid after evaporation
of the solvent (Mw ϭ 1780, Mw/Mn ϭ 2.0). The molecular
INCE the work of Schilling et al.1 on polymethylsilane (PMS)
synthesis in the mid-1980s, numerous studies have been
S
performed on PMS for use as silicon carbide (SiC) precursors.2–21
The initial interest in PMS was based on its use for tailoring SiC
materials without excess carbon by the polymer precursor method.
Toward this purpose, various chemical techniques for PMS man-
ufacture have been reported. In particular, the role of cross-linking
agents for PMS pyrolysis has been widely investigated. The use of
metallocene catalysts to promote a dehydrogenation reaction on
SiOH represented the earliest attempt.3 In recent years, the use of
boron compounds was reported to effectively increase residual
ceramic yield12,17,19 while maintaining the precursor processability.
It has become apparent that SiC materials derived from PMS
usually contain a considerable amount of excess silicon, because of
the evolution of hydrocarbon gas species during pyrolysis. How-
ever, the conversion of the PMS precursors from the amorphous to
the crystalline state, including phase segregation,4,5 has not been
characterized in detail. The formation of excess silicon phase at
high temperatures is considered to diminish the high-temperature
stability of the SiC ceramics derived from the PMS. However,
PMS combined with suitable metal sources is a promising precur-
sor system for new functional ceramic hybrids composed of SiC
1
structure of the synthesized PMS was characterized by H NMR
analysis to be [(CH3SiH)0.60–(CH3Si)0.40]n.
The synthesized PMS was placed in a quartz crucible (inner
diameter: 15 mm, depth: 100 mm), which was suspended in a
furnace equipped with molybdenum silicide (MoSi2) heaters. The
PMS then was heated from room temperature (RT) to the defined
temperatures (523–1273 K) at a rate of 5 K/min in an argon gas
flow. The argon flow rate was 1 L/min. The holding time at each
temperature was 2 h. For pyrolysis at 1473, 1673, and 1873 K, the
sample pyrolyzed at 1273 K was placed in a graphite crucible, and
the crucible was rapidly put into the graphite furnace that had been
preheated to the appointed temperatures. The argon flow rate again
was 1 L/min, and the holding time at each temperature was 15 min.
To investigate the effect of PMS condensation on the resulting
ceramic yield, the following two types of pre-heat treatment were
performed on the PMS.
(1) A quantity (1.0 g) of the synthesized PMS was placed in
the aforementioned quartz crucible suspended in the furnace and
heat-treated at 523–723 K (heating rate: 5 K/min, holding time:
2 h). The argon flow rate was 1 L/min. All of the volatile
compounds that formed from the PMS during the heat treatment
were flowed out under this condition.
(2) The PMS (1.0 g) was placed in a Pyrex† tube (diameter: 20
mm, length: 500 mm) to be cross-linked under reflux condi-
tions.16,21 The Pyrex tube was then placed inside a larger Pyrex
cell (diameter: 40 mm) equipped with a reflux condenser and a
vacuum pump. After evacuation for a sufficient time to remove
G. Grathwohl—contributing editor
Manuscript No. 187445. Received September 24, 2001; approved September 27,
2002.
Part of this study in the reflux treatment process for PMS resin condensation was
supported by Collaborative Research Project of Materials and Structures Laboratory,
Tokyo Institute of Technology.
*Member, American Ceramic Society.
†Corning, Inc., Corning, NY.
227

228
Journal of the American Ceramic Society—Narisawa et al.
Vol. 86, No. 2
1673, and 1873 K. Pyrolysis at 1073 and 1273 K was performed
using a furnace equipped with MoSi2 heaters (heating rate: 5
K/min, holding time: 2 h, argon flow rate: 1 L/min). Pyrolysis at
1473, 1673, and 1873 K was performed using the MoSi2 furnace
and the preheated graphite furnace, as described earlier.
(2) Measurements
Infrared (IR) spectra were obtained using a spectrometer (Mod-
el 1600, Perkin–Elmer, Inc., Norwalk, CT). NMR spectra for the
soluble samples also were recorded on a spectrometer (Model
Gemini 2000, Varian, Inc., Palo Alto, CA). Solid-state NMR for
29Si and 13C was conducted (Model Unity INOVA-300, Varian) at
75 and 59.6 MHz. A cross-polarization magic-angle-spinning
(CP-MAS) method was applied up to the pyrolysis temperature of
1073 K. In addition, the single-pulse method, with a delay time of
30 s, was used for the pyrolysis products at 1073, 1273, or 1473 K,
to obtain the 29Si NMR spectra.
Thermogravimetric (TG) analyses were conducted (Model TAS
200, Rigaku Co., Ltd., Tokyo, Japan) on 10 mg of the PMS
samples in an argon gas flow, at a heating rate of 10 K/min, up to
1273 K. Decomposition gas analyses during the step-by-step
pyrolysis were conducted using gas chromatography/mass spec-
trometry (Model QP5050A, Shimadzu Corp., Kyoto, Japan) and
gas chromatography (Model G-5000, Hitachi Co., Ltd., Tokyo,
Japan) equipped with a thermal conductivity detector (TCD) and a
flame ionization detector (FID). A quantitative estimation of the
individual gas species was performed in terms of standard methane
and hydrogen gas samples with defined partial pressures.
Fig. 1. Residual mass of PMS after simple pyrolysis in an argon gas flow.
oxygen and moisture in the cell, the cell was heated at 423–723 K
(heating rate: 5 K/min, holding time: 2 h). During the reflux
treatment, formation of mist in the inner Pyrex tube was observed
above 473 K.
After the heat treatment (steps (1) and (2) above), the condensed
resin had the appearance of a pale-yellow gum at 423–523 K and
maintained some elasticity; at 573–723 K, it was an orange, fragile
solid. The condensed resins were pyrolyzed at 1073, 1273, 1473,
The XRD patterns were investigated (Model RINT 1100,
Rigaku) using CuK␣ radiation. Elemental analyses of the pyro-
lyzed products at 1673 K were conducted by electron probe
Fig. 2. NMR spectra of PMS after simple pyrolysis in an argon gas flow: (a) 29Si NMR and (b) 13C NMR.

February 2003
Ceramization of Reflux-Treated Polymethylsilane Precursors to Silicon Carbide
229
Fig. 3. Gas-evolution profile of PMS during step-by-step pyrolysis.
microanalysis (EPMA; Model JXA-870, JEOL, Tokyo, Japan),
with the averaged values of five points. A polished cross section of
Nicalon‡ fiber (NL-200: SiC1.34O0.36) was used as the standard for
elemental silicon, carbon, and oxygen.
(2102 cmϪ1) and the SiH2-deformation band (terminal groups: 932
cmϪ1) are shown at 573 K. These results indicate the formation of
a SiOSi cross-link at 573 K, whereas the methyl groups have not
decomposed.
Above 573 K, the signals assigned to SiC2(H)Si (Ϫ38 ppm),
SiC3Si, or SiC3H (Ϫ15 ppm), and SiC4 (0 ppm) appear in the 29Si
NMR spectra,25–27 and the signal assigned to C(H2)Si2 (0 ppm)
appears in the 13C NMR spectra.25 At 723 K, these signals
assigned to SiC2(H)Si, SiC3Si (or SiC3H), and SiC4 in the 29Si
NMR spectra show a marked increase, while the signals assigned
to SiCSi3 disappear. At the same temperature, the signal assigned
to C(H2)Si2 in the 13C NMR spectra becomes dominant, while the
signal assigned to C(H3)Si disappears. This temperature range is
consistent with the large increase of the SiOCH2OSi band (1100
III. Results and Discussion
(1) Thermal Decomposition Process of PMS in an Argon
Gas Flow
The residual mass of the PMS after heat treatments in flowing
argon gas is shown in Fig. 1. The plotted residual masses at
individual temperatures are consistent with the reported TG curves
of PMS polymers with low ceramic yields (ϳ30 mass% at 1273
K).3,8 A large mass loss, ϳ 40%, occurred at 523–623 K, and the
mass loss up to 523 K was ϳ15%. Above 623 K, a small,
continuous mass loss was observed up to 723 K.
cm )
Ϫ1 2,28 in the IR spectra. These results indicate cleavage of the
SiOSi bonds at 573–723 K and the incorporation of methyl groups
into SiOSi to form SiOCH2OSi bridges (Kumada rearrangement
process).27–30
Above 723 K, several overlapping signals near Ϫ40 ppm to
ϳ40 ppm are convoluted to one broad signal (0 ppm) in the 29Si
NMR spectra. The shift of the CSi4 signal, from 0 ppm to 20 ppm
is also observed in the 13C NMR spectra. These changes in the
NMR spectra probably correspond to hydrogen loss during heat
treatment.
The 29Si and 13C NMR spectra of the pyrolysis products are
shown in Figs. 2(a) and (b), respectively. After pyrolysis at 573 K,
the signal assigned to SiC(H)Si2 (Ϫ67 ppm) disappears, and the
sharp signal assigned to SiCSi3 (Ϫ75 ppm) appears in the 29Si
NMR spectra.8,13 On the other hand, the signal assigned to
C(H3)Si (methyl groups: Ϫ10 ppm) in the 13C NMR spectra shows
a small down-field shift but is a dominant peak at 573 K.8,12,13 In
the measured IR spectra, decreases in the SiOH stretching band
Figure 3 shows gas evolution from the PMS during the
step-by-step pyrolysis. The large mass loss and the evolution of
gaseous silicon compounds at 523–623 K suggest fragmentation
‡Nippon Carbon Co., Tokyo, Japan.
Fig. 4. Mechanism of redistribution reaction via silyl radicals.

230
Journal of the American Ceramic Society—Narisawa et al.
Vol. 86, No. 2
Fig. 7. XRD patterns of the PMS pyrolysis products in a simple argon gas
flow.
Fig. 5. Residual mass of PMS with reflux heat treatment.
radicals formed by the cleavage of the PMS main chains. Thus,
hydrogen radicals form. Some of these radicals can attack SiOH
or SiOSi in the PMS again, while the remaining radicals form
stable species (hydrogen gas, methylsilane gases, and terminal
groups in silane oligomers). On the other hand, silylene interme-
diates are known to be stable when compared thermodynamically
with the silyl radicals.27,31,32 The insertion of the silylene into the
SiOH bonds is another promising chemical reaction to explain the
sudden formation of SiCSi3 at ϳ573 K. Silylene would form
easily, particularly at the terminal groups (OSi(CH3)H2) of the
PMS.
of PMS chains, whereas the 29Si NMR and 13C NMR spectra
suggest SiOSi cross-linking. The evolution of methane is substan-
tial at 623–823 K. Most of the methyl groups in the PMS have
been incorporated into SiOSi bonds to form SiOCH2OSi bridges
in the same temperature region, which were observed in the change of
the IR and NMR spectra. However, some of the methyl groups have
been eliminated in the form of methane. The evolution of hydrogen
occurs over a wide temperature range. In particular, the hydrogen-
evolution peak at 823–923 K is consistent with the conversion process
of the SiOCH2OSi bridges into a SiC network.
It may seem strange that SiOSi cross-linking is accompanied
by SiOSi cleavage at ϳ573 K. Numerous schemes have been
proposed for SiOSi cross-link formation in PMS, which should
depend on the chemical nature of the adopted cross-linking
processes. In this simple heat-treatment process, the redistribution
reaction mechanism proposed by Boury would be applicable,12
because the redistribution reaction always yields low-molecular-
weight components in addition to the SiOSi network.
The silyl radicals formed during cleavage of the PMS main
chains probably play a role in the redistribution reaction. Figure 4
shows the mechanism of the redistribution reaction in terms of the
radical behavior. In Fig. 4, SiOH groups in the PMS trap the silyl
(2) Effect of Reflux Treatment on PMS
Figure 5 shows the residual masses of the PMS samples after
reflux treatment. Up to 523 K, the residual masses after reflux
treatment are almost the same as those of PMS simply heat-treated
in an argon gas flow. Above 523 K, the residual masses are still
high under reflux treatment, whereas the values suddenly decrease
under simple heat treatment (Fig. 1). IR and NMR spectroscopic
analyses revealed that structural evolution in PMS accelerates with
reflux treatment, as compared with simple heat treatment at exactly
the same temperature. However, the observed difference in the
spectra was smaller than that caused by a 50 K difference in the
heat-treatment temperature.
Thus, it is natural to suppose that the main chemical reactions
under reflux conditions are almost the same as those under simple
heat treatment in flowing argon gas. Because the formed low-
molecular-weight components (silane oligomers) are cooled at the
upper side of the cell, they revert to a condensed resin to be
cross-linked again. This effective incorporation of the oligomers
should increase the residual mass.
When PMS is heat-treated once at Ͼ623 K, either under reflux
conditions or in a simple argon gas flow, the obtained resins do not
show large mass loss at higher temperatures. The residual masses
at 1273 K are usually Ͼ95%, based on the condensed resin masses.
Figure 6 summarizes the overall ceramic yields after the two-step
heat treatment (first step: 423–723 K, second step: 1273 K). The
overall ceramic yield becomes a stable value, ϳ75%, above 623 K
in the case of reflux treatment. Even under simple heat treatment
in flowing argon gas, the overall ceramic yield is improved by
holding at 523 or 573 K. However, the overall ceramic yield does
not exceed 50%. Holding at higher temperature, Ͼ573 K, in
flowing argon gas is not as effective as holding at 523 or 573 K for
improving ceramic yield. The overall ceramic yield tends to
decrease to ϳ40% under those conditions, although the value is
still higher than after direct pyrolysis (heating rate: 5 K/min from
RT to 1273 K, holding time: 2 h).
Fig. 6. Overall ceramic yield of PMS after heat treatment at 423–723 K
and pyrolysis at 1273 K ((- - -) direct pyrolysis in an argon gas flow with
a heating rate of 5 K/min from RT to 1273 K and a holding time of 2 h at
1273 K).

February 2003
Ceramization of Reflux-Treated Polymethylsilane Precursors to Silicon Carbide
231
Fig. 8. XRD patterns of the products derived from the refluxed PMS resins after pyrolysis at (a) 1273 K and (b) 1473 K.
(3) Conversion Process from Amorphous to Crystalline State
observed temperature region, all of the peaks in the XRD patterns
have been assigned to ␤-SiC and silicon.
The XRD patterns of the PMS pyrolyzed in an argon gas flow
are shown in Fig. 7. SiC crystallites and excess silicon are shown
at 1073 K. Up to 1273 K, the SiC crystallites are small (1–3 nm).
At 1473 K, the SiC crystallites show substantial growth. In the
Figures 8(a) and (b) show the XRD patterns of the pyrolysis
products derived from the refluxed PMS resins at 1273 and 1473
K, respectively. At 1273 K, all of the XRD patterns are broad,
showing a SiC amorphous structure with traces of silicon. The
difference in the pre-heat-treatment conditions has no influence on
Fig. 9. Apparent crystallite sizes estimated from the XRD (111) line
width in the pyrolysis products at 1273–1873 K (crystallite sizes (ϳ3 nm)
estimated for pyrolysis at 1273 K are not completely precise, because the
measured lines are highly broadened).
Fig. 10. Results of deconvolution of the 29Si NMR spectrum of the
pyrolysis product at 1273 K in a simple argon gas flow.

232
Journal of the American Ceramic Society—Narisawa et al.
Vol. 86, No. 2
Table I. Elemental Analyses of Pyrolysis Products†
Content (at.%)
Carbon
Sample condition
Silicon
Oxygen
SiCxOy formula‡
Original
523 K
623 K
723 K
54.16
51.05
51.69
52.36
42.10
44.52
44.10
43.97
3.74
4.43
4.21
3.67
SiC0.78O0.07
SiC0.87O0.09
SiC0.85C0.08
SiC0.84O0.07
†All samples were pyrolyzed at 1673 K under argon gas. ‡Five points averaged.
␣-SiC, although the evidence of ␣-SiC or stacking faults (peak or
shoulder at 33.6°) is absent in the XRD patterns.35 However,
quantitative estimation of the individual polytypes in the SiC
synthesized by solid-state NMR may give ambiguous results,
because of the long relaxation time of the silicon atom.36
Figure 11 shows the 29Si NMR spectra of the pyrolysis products
(1273 K) derived from the refluxed PMS resins. All of the spectra
consist of the previously explained four signals (R1, R2, R3, R4).
However, the spectrum of the product derived from the resin
refluxed at 523 K shows the most-broadened signals.
Apparently, the ceramic materials derived from the precursor
under reflux treatment at 523 K possess the most disordered
(amorphous-like) structure after pyrolysis. One of the important
factors for determining the crystallite sizes is the chemical com-
position of the pyrolysis product itself. Table I shows the EPMA
results for the elemental compositions of the pyrolysis products at
1673 K. The products at 1673 K exhibit the highest C/Si value,
0.87, with reflux treatment at 523 K and the lowest value, 0.78,
with no reflux treatment. The large amount of free silicon in the
product derived from the PMS without reflux treatment may
induce crystallite growth at 1273–1473 K. However, the C/Si
values are 0.85 or 0.84 in the materials with reflux treatment at 623
or 723 K, respectively. These values are close to that for reflux
treatment at 523 K. Therefore, crystallite growth probably is not
directly controlled by the amount of free silicon out of the SiOC
network.
Fig. 11. 29Si NMR spectra of the pyrolysis products at 1273 K derived
from the refluxed PMS resins.
The temperature at which the efficient cross-link starts is 523 K.
The observed phenomenon regarding the SiC crystallite sizes just
represents the same issue as the thermal history of organosilicon
precursors in the resulting ceramic crystallization. This is a very
difficult theme that encompasses various factors. In the case of
polycarbosilane or polysilazane pyrolysis, a sudden increase and
decrease in radical concentration, accompanied by hydrogen evo-
the observed XRD patterns. At 1473 K, all of the products show
substantial growth of ␤-SiC and silicon. However, the XRD
patterns of the products derived from the PMS resin refluxed at
523 K maintain a broad appearance, compared with those of the
other three samples.
Figure 9 shows the apparent SiC crystallite sizes in the pyrolysis
products derived from the refluxed PMS resins, estimated from the
half-height line width of the (111) line at 1273–1873 K. The
crystallites in the pyrolysis products derived from the resin
refluxed at 523 K always show the smallest sizes among the four
samples. On the other hand, the crystallites in the products derived
from the untreated PMS or the resin refluxed at 723 K are usually
larger. Crystallite growth of silicon decreases under reflux treat-
ment at 523 K.
lution, have been observed just before crystallite growth.37
A
precise estimation of the hydrogen content in the pyrolysis
products derived from the PMS resins at 1073–1273 K and an
evaluation of the remaining radical concentration during pyrolysis
probably are required to characterize the effect of pre-cross-linking
on the resultant crystallite growth at high temperatures.
In the 29Si solid-state NMR spectra for the products of pyrolysis
in flowing argon gas under the single-pulse method, the product at
1073 K shows a broad signal with a maximum at Ϫ10 ppm. This
result is consistent with the amorphous structure indicated in the
XRD patterns. However, at 1273 and 1473 K, the observed signals
are composed of several sharp signals that overlap each other.
Figure 10 shows the result of deconvolution for the observed
spectra under pyrolysis at 1273 K (R1: Ϫ13.1 ppm, R2: Ϫ15.2
ppm, R3: Ϫ20.1 ppm, R4: Ϫ24.0 ppm). The signal at Ϫ15.2 ppm
(R2) can be assigned to type A (second neighbors: 12Si, cubocta-
hedron) silicon atoms in ␤-SiC (3C). On the other hand, the signals
at Ϫ13.1, Ϫ20.1, and Ϫ24.0 ppm (R1, R3, R4) have been assigned
to type A, type B (second neighbors: 12Si, twinned cuboctahe-
dron), and type C (second neighbors: 12Si and 1C, capped
cuboctahedron) silicon atoms in ␣-SiC (6H, 15R,. . . ), respective-
ly.5,33,34 The difference in the peak positions of R1 (type A in ␣)
and R2 (type A in ␤) probably reflects the change in long-range
ordering in the pyrolysis products.
IV. Conclusions
Study of the thermal decomposition process of PMS in flowing
argon gas reveals the formation of a SiOSi cross-link at 573 K,
accompanied by cleavage of the PMS chains (redistribution
reaction). The formed SiOSi cross-link is converted to
SiOCH2OSi bridges at 723 K, by the Kumada rearrangement
process. The resulting network is gradually converted into a SiOC
network by losing hydrogen at high temperatures. Reflux treatment
above 523 K effectively increases overall ceramic yield, based on
the original PMS mass. In the environment of reflux treatment,
fragments formed by chain cleavage probably cool to cross-link
with residual resin at the bottom. The maximum ceramic yield
under reflux treatment is ϳ75% at a reflux temperature of 623 K.
Simple holding at 523 or 573 K, without the reflux condenser, also
effectively increases ceramic yield; however, the ceramic yield
under simple heat treatment is ϳ50%.
The growth of SiC and silicon crystallites in the pyrolysis
products derived from refluxed PMS resins is remarkable at
1273–1473 K. However, pyrolysis products derived from the resin
Surprisingly, some of the silicon atoms in such amorphous or
nanocrystalline SiC possess the same coordination as those in

February 2003
Ceramization of Reflux-Treated Polymethylsilane Precursors to Silicon Carbide
233
Pyrolysis (PIP) Processing Using Polymethylsilane,” J. Eur. Ceram. Soc., 20 [4]
441–51 (2000).
refluxed at 523 K possess the most disordered structure, as shown
by the XRD patterns of the pyrolysis products at 1473–1873 K and
the 29Si NMR spectra of the product at 1273 K.
19J.-G. Kho, D.-S. Min, and D.-P. Kim, “Polymethylsilane Post-treated with a
Polyborazine Promoter as a Precursor to SiC with High Ceramic Yield,” J. Mater. Sci.
Lett., 19 [4] 303–305 (2000).
20M. Narisawa, T. Yoshida, T. Iseki, Y. Katase, K. Okamura, K. Oka, and T.
Dohmaru, “␥-Ray Curing of Poly(methylsilane) and Poly(methylsilane-
dimethylsilane) for Improved Ceramic Yields,” Chem. Mater., 12 [9] 2686–92
(2000).
References
1C. L. Schilling, J. P. Wesson, and T. C. Williams, “Polycarbosilane Precursors for
Silicon Carbide,” Am. Ceram. Soc. Bull., 62 [8] 912–15 (1983).
2Z. Zhang, F. Babonneau, R. M. Laine, Y. Mu, J. F. Harrod, and J. A. Rahn,
“Poly(methylsilane)—A High Ceramic Yield Precursor to Silicon Carbide,” J. Am.
Ceram. Soc., 74 [3] 670–73 (1991).
21T. Iseki, M. Narisawa, Y. Katase, K. Okamura, K. Oka, and T. Dohmaru, “An
Efficient Cross-Linking Process of Polymethylsilane for SiC Ceramics,” Chem.
Mater., 13 [11] 4163–69 (2001).
22D. Seyferth and P. Czubarow, “Application of Preceramic Polymers in Powder
Metallurgy: Their Use as Low-Loss Binders and for the in Situ Formation of
Dispersed Ceramic Phases in the Metal Matrix,” Chem. Mater., 6 [1] 10–12 (1994).
23P. Greil, “Active-Filler-Controlled Pyrolysis of Preceramic Polymers,” J. Am.
Ceram. Soc., 78 [4] 835–48 (1995).
3D. Seyferth, T. G. Wood, H. J. Tracy, and J. L. Robisoe, “Near-Stoichiometric
Silicon Carbide from Economical Polysilane Precursor,” J. Am. Ceram. Soc., 75 [5]
1300–302 (1991).
4T. Kobayashi, T. Sakakura, T. Hayashi, M. Yumura, and M. Tanaka,
“Neodymium-Catalyzed Dehydrogenative Condensation of Methylsilane and Ceram-
ization of the Resulting Polymer,” Chem. Lett., 1157–60 (1992).
5R. M. Laine and F. Babonneau, “Preceramic Polymer Routes to Silicon Carbide,”
Chem. Mater., 5 [3] 260–79 (1993).
24D.-P. Kim, “Preparation of Electrically Conducting SiC/MoSi2 Composites from
Metal Mixtures of Preceramic Polymer,” Mater. Res. Bull., 36, 2497–505 (2001).
25G. D. Soraru, F. Babonneau, and J. D. Mackenzie, “Structural Evolutions from
Polycarbosilane to SiC Ceramics,” J. Mater. Sci., 25 [9] 3886–93 (1990).
26A. T. Hemida, M. Birot, J. P. Pillot, J. Dunogues, and R. Pailler, “Synthesis and
Characterization of New Precursors to Nearly Stoichiometric SiC Ceramics,” J.
Mater. Sci., 32 [13] 3475–83 (1997).
6Y. Mu, R. M. Laine, and J. F. Harrod, “Some Further Observations on
Polymethylsilane as a Precursor for Silicon Carbide,” Appl. Organomet. Chem., 8,
95–100 (1994).
7Z.-F. Zhang, C. S. Scotto, and R. M. Laine, “Pure Silicon Carbide Fibers from
Polymethylsilane,” Ceram. Eng. Sci. Proc., 15 [4] 152–61 (1994).
8M. F. Gozzi and I. V. P. Yoshida, “Thermal and Photochemical Conversion of
Poly(methylsilane) to Polycarbosilane,” Macromolecules, 28 [21] 7235–40 (1995).
9F. I. Hurwitz, T. A. Kacik, X.-Y. Bu, J. Masnovi, P. J. Heimann, and K. Beyene,
“Pyrolytic Conversion of Methyl- and Vinylsilane Polymers to Si-C Ceramics,” J.
Mater. Sci., 30 [12] 3130–36 (1995).
27Q. Liu, H.-J. Wu, R. Lewis, G. E. Maciel, and L. V. Interrnate, “Investigation of
the Pyrolytic Conversion of Poly(silylenemethylene) to Silicon Carbide,” Chem.
Mater., 11 [8] 2038–48 (1999).
28K. Shiina and K. Kumada, “Thermal Rearrangement of Hexamethyldisilane to
Trimethyl(dimethylsilylmethyl)silane,” J. Org. Chem., 23, 139 (1958).
29S. Yajima, Y. Hasegawa, J. Hayashi, and M. Iimura, “Synthesis of Continuous
Silicon Carbide Fiber with High Tensile Strength and High Young’s Modulus. Part 1.
Synthesis of Polycarbosilane as Precursor,” J. Mater. Sci., 13 [12] 2569–79 (1978).
30M. Birot, J.-P. Pillot, and J. Dunogue`s, “Comprehensive Chemistry of Polycar-
bosilanes, and Polycarbosilazanes as Precursors of Ceramics,” Chem. Rev., 95 [5]
1443–77 (1995).
10M. F. Gozzi and I. V. P. Yoshida, “Structural Evolution of
a
Poly(methylsilylene)/Tetra-Allylsilane Mixture,” Eur. Polym. J., 33 [8] 1301–306
(1997).
11P. Czubarow and D. Seyferth, “Application of Poly(methylsilane) and Nicalon®
Polycarbosilane Precursors as Binders for Metal/Ceramic Powders in Preparation of
Functionally Graded Materials,” J. Mater. Sci., 32 [8] 2121–30 (1997).
12B. Boury, N. Bryson, and G. Soula, “Borate-Catalyzed Thermolysis of Poly-
methylsilane,” Chem. Mater., 10 [1] 297–303 (1998).
31A. Tachibana, Y. Kurosaki, K. Yamaguchi, and T. Yamabe, “Quantum Chemical
Study of Silicon Carbide Formation,” J. Phys. Chem., 95 [18] 6849–54 (1991).
32R. J. P. Corriu, D. Leclercq, P. H. Mutin, J.-M. Planeix, and A. Vioux,
“Mechanism of Pyrolysis of Polycarbosilanes: Poly(silylenethylene) and Poly(di-
methylsilyl ethylene),” Organometallics, 12 [2] 454–62 (1993).
13P. Czubarow, T. Sugimoto, and D. Seyferth, “Sonochemical Synthesis of a
Poly(methylsilane), a Precursor for Near-Stoichiometric SiC,” Macromolecules, 31
[2] 229–38 (1998).
33J. S. Hartman, M. F. Richardson, B. L. Sheriff, and B. G. Winsborrow, “Magic
Angle Spinning NMR Studies of Silicon Carbide: Polytypes, Impurities, and Highly
Inefficient Spin-Lattice Relaxation,” J. Am. Chem. Soc., 109 [20] 6059–67 (1987).
34K. R. Carduner, S. S. Shinozaki, M. J. Rokosz, C. R. Peters, and T. J. Whalen,
“Characterization of ␤-Silicon Carbide by Silicon-29 Solid-State NMR, Transmission
Electron Microscopy, and Powder X-ray Diffraction,” J. Am. Ceram. Soc., 73 [8]
2281–86 (1990).
14M. F. Gozzi, M. D. Goncalves, and I. V. P. Yoshida, “Near-Stoichiometric
Silicon Carbide from a Poly(methylsilylene)/Tetra-Allylsilane Mixture,” J. Mater.
Sci., 34 [1] 155–59 (1999).
15K. W. Chew, A. Sellinger, and R. M. Laine, “Processing Aluminum Nitride
Silicon Carbide Composites via Polymer Infiltration and Pyrolysis of Polymethylsi-
lane, a Precursor to Stoichiometric Silicon Carbide,” J. Am. Ceram. Soc., 82 [4]
857–66 (1999).
35H. Tanaka and N. Iyi, “Polytype, Grain Growth, and Fracture Toughness of Metal
Boride Particulate SiC Composites,” J. Am. Ceram. Soc., 78 [5] 1223–29 (1995).
36H. Fujimori, N. Sato, K. Ioku, S. Goto, and T. Yamada, “29Si MAS NMR
Spin-Lattice Relaxation Study of ␣-, ␤-, and Amorphous Silicon Nitride,” J. Am.
Ceram. Soc., 83 [9] 2251–54 (2000).
16T. Iseki, M. Narisawa, K. Okamura, K. Oka, and T. Dohmaru, “Reflux
Heat-Treated Polymethylsilane as a Precursor to Silicon Carbide,” J. Mater. Sci. Lett.,
18 [3] 185–87 (1999).
17B. Boury, N. Bryson, and G. Soula, “Stoichiometric Silicon Carbide from
Borate-Catalyzed Polymethylsilane–Polyvinylsilane Formulations,” Appl. Organo-
met. Chem., 13, 419–30 (1999).
37M. Narisawa, M. Shimoda, K. Okamura, M. Sugimoto, and T. Seguchi,
“Reaction Mechanism of the Pyrolysis of Polycarbosilane and Polysilazane as
18M. A. Nechanicky, K. W. Chew, A. Sellinger, and R. M. Laine, “␣-Silicon
Carbide/␤-Silicon Carbide Particulate Composites via Polymer Infiltration and
Ceramic Precursors,” Bull. Chem. Soc. Jpn., 68 [4] 1098–1104 (1995).
Ⅺ