ArticlePDF Available

Production of an Interphase Coating of Polycarbosilane and Rolivsan Ceramic-Forming Compounds on Carbon Fiber

Authors:

Abstract and Figures

The conditions for depositing a ceramic-forming compound interphase coating on carbon fibers were investigated. Use of polycarbosilane and Rolivsan ceramic-forming compounds in hexane (<3 mass%) and polymer derived ceramic infiltration in an inert medium produced according to the optimal regime based on thermal kinetic calculations an even interphase coating that increased the thermal-oxidation resistance of the starting carbon fibers. The combustion products of carbon fibers with the deposited interphase were hollow SiO2 structures.
Content may be subject to copyright.
92
Fibre Chemistry, Vol. 51, No. 2, July, 2019 (Russian Original No. 2, March-April, 2019)
PRODUCTION OF AN INTERPHASE COATING OF POLYCARBOSILANE
AND ROLIVSAN CERAMIC-FORMING COMPOUNDS ON CARBON FIBER
M. A. Khaskov, E. A. Sul’yanova, UDC 546.287, 546.26
M. I. Valueva, and E. A. Davydova
The conditions for depositing a ceramic-forming compound interphase coating on carbon fibers were
investigated. Use of polycarbosilane and Rolivsan ceramic-forming compounds in hexane (<3 mass%)
and polymer derived ceramic infiltration in an inert medium produced according to the optimal regime
based on thermal kinetic calculations an even interphase coating that increased the thermal-oxidation
resistance of the starting carbon fibers. The combustion products of carbon fibers with the deposited
interphase were hollow SiO2 structures.
Carbon fibers [1] need to be coated with a high-temperature finish or so-called interphase coating (IC) if they are
used as a reinforcing filler for ceramic-matrix composites [2, 3] or polymer composites [4] with high formation temperatures
[5, 6]. ICs can be prepared from various nitrides, carbides, borides, etc. using various methods, e.g., vapor-phase deposition,
sol-gel technology, etc. [7]. Polymer derived ceramic (PDC) infiltration of a ceramic-forming compound avoids drawbacks
of IC application such as the use of highly toxic and volatile compounds and expensive equipment (chemical vapor deposition)
or high shrinkage during solvent removal (sol-gel method) [7]. Polycarbosilanes are ceramic-forming precursors of silicon
carbide compounds [8] and can be used to apply ICs by the PDC method [9]. It is noteworthy that curing pure polycarbosilane
in an inert medium produced low ceramic yields because volatile cyclic carbosilanes could form [10]. Formulations can be
treated with unsaturated compounds that can undergo radical hydrosilylation in order to increase the ceramic yield of pure
polycarbosilanes [11]. Several thermally reactive compounds with vinyl and methacrylate terminal groups, e.g., Rolivsans,
are capable of cross-linking and polycyclization and are attractive candidates for increasing the ceramic yield from
polycarbosilanes because Rolivsans are thermally and chemically resistant [12]. The present work is focused on the possibility
of applying an IC to carbon fibers using PDC with mixtures of polycarbosilanes and Rolivsans as preceramics.
The starting compounds were PCS-M polycarbosilanes (State Scientific Research Institute of Chemistry and
Technology of Organoelement Compounds), Rolivsans MV-1 (Institute of Macromolecular Compounds), and pure hexane.
Solutions of ceramic-forming compounds were prepared by dissolving a given amount of PCS-M and MV-1 in hexane
with constant stirring for 1 h. The contact angle at the air—solution interface of IC precursor—pyrolytic graphite
interface was measured on an OCA 15Pro (Dataphysics Instruments GmbH) using pyrolytic polished graphite (Medinzh,
Penza) as the substrate. A given volume of compound was placed onto the substrate. The change of contact angle over
time was measured using an automated camera. Measurements were made at 100-ms intervals. When the measurements
were finished, the change of contact angle with time was plotted. The ceramic-forming compound solution to be analyzed
was dried in a vacuum oven to residual pressure <100 Pa at room temperature. Differential scanning calorimetry (DSC)
and thermogravimetry (TG) studies used a STA 449 F3 Jupiter instrument (Netzsch) in Ar (80 mL/min) at various
heating rates (5, 10, and 20ºC/min). IR spectroscopy used a Tensor 27 FT-IR spectrometer (Bruker). For this, the tested
substance (1 mg) was mixed with calcined KBr (100 mg), ground in a ball mill for 1 min, and pressed into pellets on a
special hand press. Previously prepared carbon fibers were soaked with various concentrations of ceramic-forming
compounds (polycarbosilane:Rolivsan mass ratio, 2:1) in hexane [13]. Fibers with deposited coating were dried at room
temperature for 24 h and heat treated in flowing Ar according to a selected temperature program. Scanning electron
All-Russian Scientific Research Institute of Aviation Materials, Moscow; E-mail: khaskov@mail.ru. Translated
from Khimicheskie Volokna, No. 2, pp. 17-21, March—April, 2019.
0015-0541/19/5102-0092© 2019 Springer Science+Business Media LLC
DOI 10.1007/s10692-019-10052-1
93
microscopy of fibers with ICs used a Zeiss EVO MA 10 microscope. The resistance to thermal oxidation of starting
fibers and fibers with ICs of the same lengths (~100 mm) and mass were placed into 3-mm quartz tubes that were placed
into a muffle furnace that was heated according to a given temperature program.
Thermal analytical studies showed that annealing up to 350ºC of the selected ceramic-forming compound of
polycarbosilanes and Rolivsans was associated with two sequential exotherms and a slight mass loss (~3%) whereas
pyrolysis (heating above 350ºC) had characteristic sequential exotherms and endotherms with significant (~22%) mass
losses (Fig. 1a). Many of thermal effects of curing and pyrolysis were correlated with the rate of mass loss (DTG curve).
Mass losses at temperatures before the start of heat evolution may have been related to vaporization of residual solvent
(hexane). The samples were studied using IR spectroscopy after heat treatment at various temperatures to determine the
nature of the observed exotherms.
IR spectroscopy (Fig. 1b) found that the first exotherm during curing of the ceramic-forming compound was
related to polymerization of Rolivsan C=C bonds because HxC=CHy vibrations at ~1631, ~1637, and ~3088 cm–1
disappeared after heat treatment at 240ºC. The slight mass loss and diminished carbonyl (~1719 cm–1) and ester C–O
vibrations (~1247 and ~1167 cm–1) after heat treatment at 240ºC could be explained by vaporization of methacrylic acid
formed by chemical transformations in Rolivsan [12]. The slight weakening of Si–H (~2100 cm–1) and carbonyl vibrations
(~1719 cm–1) after heat treatment at 350ºC led to the hypothesis that C=O groups were reduced by hydrosilylation.
According to the literature [12], the appearance of vibrations for a six-membered cyclic anhydride at ~1803 and ~1759
cm–1 after heat treatment at 350ºC was due to curing of the Rolivsan structure. Heat treatment at 490ºC caused carbonyl
and anhydride vibrations to disappear completely and Si–H vibrations (~2100 cm–1) to decrease substantially because of
hydrosilylation [14]. Moreover, heat treatment at 490ºC diminished skeletal vibrations of aromatic rings (~1600 cm–1),
which could be explained by destruction of the cured Rolivsan structure and significant mass losses. Heating further to
900ºC led to endotherms due to pyrolysis, which was associated with considerable mass losses.
Thermokinetics calculations [15] and DSC and TG data at various heating rates were used to optimize the
temperature—time regime (TTR) for curing and pyrolysis [16]. The TTR was calculated considering the need for even
heat release/absorption and mass losses during curing and pyrolysis. Even heat release and uniform mass loss helped to
form ICs without defects resulting from local temperature and mass-transfer gradients during removal of reaction products
from the system.
The rates of change of the degrees of transformation and heat flow for DSC measurements were assumed to be
proportional for the calculations [Eq. (1)]; of degrees of transformation and mass loss, for TG data [Eq. (2)]:
dα/dt dH/dt,(1)
dα/dt dm/dt, (2)
where α is the degree of transformation; t, time; dH/dt, heat flow; dm/dt, rate of mass change.
100 200 300 400 500 600 700 800 3500 3000 2500 2000 1500 1000 500
Temperature, ºC Wavenumber, cm–1
100
90
80
70
60
50
40
30
TG/%
ÒÃ
1.0
0.8
0.6
0.4
0.2
0.0
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
-40 -30 -20 -10 0
Transmittance, %
DSC/(μW/mg)
DTG/(mass%/min)
EXO
Mass change: -9 %
Mass change: -1.5 %
Mass change: -1.8 %
Mass change: -1.5 %
Peak: 158.10 °C
Peak: 208.01 °C
Peak: 299.02 °C
Peak: 296.73 °C
Peak: 458.47 °C
Peak: 208.01 °CPeak: 477.90 °C
Peak: 488.43 °C
Peak: 671.19 °C
Peak: 693.01 °C
DTG
ÄÑÊ
Starting
240
°
C
350
°
C
490
°
C
ab
Fig. 1. TG, DTG, and DSC data during heating of ceramic-forming compound at 10ºC/min (a) and IR spectra of
ceramic-forming compound treated at various temperatures (b).
94
Figure 2a shows DSC data in the range 120-570ºC that were obtained at various heating rates and approximated
by a kinetic model with four sequential reactions. TG data in the range 40-600ºC were also approximated by a kinetic
model with four sequential reactions.
Four sequential reactions (SRs) were used as the model. Each of them obeyed the kinetic equation:
,[
[/ia n
i
RTE
ieA
dt
d]reagent starting
reagent] starting = (3)
where Ai is the pre-exponential factor; Ea, activation energy; ni, reaction order; i, reaction No.; t, time; T, temperature
(K); [starting reagenti], molar concentration of starting reagent i. Table 1 presents the obtained parameters of the
kinetic models.
0.8
0.6
0.4
0.2
0.0
DSC-signal, W/g
KÝÕÎ
150 250 350 450 550
Temperature, ºC
525
425
325
225
125
25
Temperature, ºC
100
90
80
70
60
50
40
30
20
10
0
Conversion, %
0 50 100 150 200 250 300
Time, min
1
2
3
3
2
1
Fig. 2. DSC curves taken at rates (ºC/min) of 5 (1), 10 (2), and 20 (3) and approximated by a kinetic model (a);
temperature—time regime of curing (1) and pyrolysis for even heat release (3) and mass loss (2).
ab
Table 1. Kinetic Model Parameters for Heat Release and Mass Loss
Parameter Pro cess
heat release mass loss
lg A1, с-1 9.4913 1. 15 31
E1, kJ/mol 107.7391 29.6733
n1 1.005 2. 2835
logA2, с-1 13.2 70 7 14 .87 16
E2, kJ/mol 170.9414 185.3986
n2 1.1466 2.1607
lgA3, с-1 15.312 13.3559
E3, kJ/mol 237.8711 220.6638
n3 1.0454 3.3698
logA4, с-1 14.9 88 7 -1.9711
E4, kJ/mol 249.893
4
20.2559
n4 2.2378 0. 128
SR1 0. 20 24 0. 13 85
SR2 0.0439 0. 07 65
SR3 0.3954 0. 55 1
SR4 0.358 0. 23 4
Q20°С
/
min, J/g 307.2553
Q10°С
/
min, J/g 355.6977
QС
/
min, J/g 429. 792 –
Δm20°С
/
min, % – –22.4855
Δm10°С
/
min, % – –21.4829
ΔmС
/
min, % – –19.4847
_______________
Note. QX°С/min is the total thermal effect for heating rate X°С/min; ΔmX°С/min, mass
change for heating rate X°С/min; SRi, contribution of reaction i to the total
thermal effect (mass loss).
95
The curing and pyrolysis regimes for a smooth change of degree of transformation was selected based on the
kinetic models and are shown in Fig. 2b. If the sample was heated according to the selected TTR (curve 1), a relatively
even mass decrease (curve 2) and heat evolution (curve 3) were observed. The rather large mass loss during heating of
the sample to 170ºC, as proposed earlier, was related to vaporization of residual solvent.
It is noteworthy that heat and mass exchange processes were not considered in calculating the regime because
the obtained sample was a thin film [17].
The technological parameters for depositing the ICs on the carbon fibers were optimized after studying the chemical
processes of curing and pyrolysis and selecting their TTRs. The contact angles between air (A), pyrolytic graphite (PG),
and ceramic-forming compound with a concentration <10 mass% did not differ from those between A—PG and hexane.
However, the coating was unevenly distributed and had ceramic clumps if the ceramic-forming compound concentration
was >5% (Fig. 3a). Also, traces of O2 in the Ar caused the coating to peel (Fig. 3b). O2 can cause combustion of carbon
from the PDC structure and the carbon fiber itself. Gaseous carbon oxides can promote foaming of the coating. Also, the
difference of the thermal expansion coefficients of silicon carbide (pyrolysis product of polycarbosilane) and silicon
dioxide (oxidation product of polycarbosilane in air) could affect exfoliation of the IC.
An even IC formed if a ceramic-forming compound with a concentration <3 mass% was used for curing and
pyrolysis in O2-free Ar according to the TTR determined based on the thermokinetics calculations (Fig. 4a). The nano-
sized thickness of the starting fiber increased insignificantly after depositing the IC. The IC was confirmed to be present
on the fiber by the increased resistance to thermal oxidation (see below) and the production of hollow whitish structures
with inner diameters equal to that of the starting fiber.
Table 2 presents results for oxidation in a muffle furnace of starting fibers and those with ICs that indicate the
IC decreases the oxidation rate and increases the resistance to thermal oxidation of the carbon fibers.
Combustion in air of carbon fibers with ICs formed hollow whitish structures. IR spectra of the combustion
products exhibited absorption bands with maxima at 3436, 1629, 1096, 806, and 467 cm–1. According to the literature
[18], the absorption band maxima at 1096 and 806 could be attributed to asymmetric and symmetric stretching vibrations
Fig. 3. Interphase coating produced from ceramic-forming compound
(5 mass%) in hexane (a) and in Ar with traces of O2 (b).
2 μm2 μm
a b
Table 2. Oxidation of Starting Carbon Fibers and Those with
Interphase Coating
Cycle No.* Sample mas s, %
st arting wit h IC
1
6
2 81
2 34 60
3 12 44
4 1 21
5 – 16
6 – 3
7 –
_______________
*Cycle includes heating in an o ven with static air to 600ºC over
90 min, holdin g at 600ºC for 60 min, and cooling to 150ºC over
90 min.
96
of Si–O–Si fragments, respectively. Absorption at 467 cm–1 could correspond to bending vibrations of O–Si–O fragments;
at 3436, to Si–OH vibrations in an amorphous SiO2 structure. Thus, IR spectroscopic data suggested that the combustion
products of carbon fibers with ICs were structures consisting of SiO2 particles.
We thank A. M. Shestakov, I. V. Zelenina, O. Yu. Sorokin, S. D. Sinyakov, and A. I. Gulyaev for assistance with the
work. The work was financially supported by the Russian Foundation for Basic Research in the framework of Project No.
17-03-01163.
REFERENCES
1. S.-J. Park, Carbon Fiber, Springer, Netherlands, 2015.
2. E. N. Kablov, D. V. Grashchenkov, et al., Steklo Keram., 4, 7 (2012).
3. S. S. Solntsev, V. S. Denisova, and V. A. Rozenenkova, Aviats. Mater. Tekhnol., (S), 329 (2017).
4. E. N. Kablov, B. V. Shchetanov, et al., Tr. VIAM, No. 2, 5 (2013).
5. E. N. Kablov, Vse Mater., No. 5, 7 (2007).
6. E. Ya. Beider, G. N. Petrova, and T. F. Izotova, Tr. VIAM: Elektron. Nauchn.-Tekh. Zh., No. 9, 7 (2014).
7. Department of Defense Handbook: Composite Materials Handbook, Vol. 5, Ceramic Matrix Composites, 2002.
8. S. S. Solntsev, V. A. Rozenenkova, and N. A. Mironova, Aviats. Mater. Tekhnol., (S), 359 (2012).
9. P. Colombo, G. Mera, et al., J. Am. Ceram. Soc., 93, 1805 (2010).
10. M. A. Khaskov, N. I. Shvets, and A. M. Shestakov, in: Proceedings of the Conference “High-Temperature
Ceramic Composites and Protective Coatings” [in Russian], VIAM, Moscow, 2014, p. 12.
11. A. M. Shestakov, N. I. Shvets, et al., Russ. J. Appl. Chem., 88, 1481 (2015).
12. B. A. Zaitsev and I. D. Shvabskaya, Russ. J. Appl. Chem., 83, 1270 (2010).
13. M. I. Valueva, I. V. Zelenina, et al., Tr. VIAM, No. 10, 9 (2017).
14. H. Q. Ly, R. Taylor, et al., J. Mater. Sci., 36, 4037 (2001).
15. E. Moukhina, J. Therm. Anal. Calorim., 109, 1203 (2012).
16. M. A. Khaskov, A. M. Shestakov, et al., Zh. Prikl. Khim., 91, 890 (2018).
17. M. A. Khaskov, Zh. Prikl. Khim., 89, 510 (2016).
18. S. Music, N. Filipovic-Vincekovic, and L. Sekovanic, Braz. J. Chem. Eng., 28, 89 (2011).
Fig. 4. Smooth interphase coating on carbon fibers produced by PDC process using polycarbosilanes
and Rolivsans (a) and IR spectrum of hollow whitish structures (combustion product of carbon fibers
with IP) (b).
2 μm3500 3000 2500 2000 1500 1000 500
Wavenumber, cm–1
-250 200 150 100 50 0
Transmittance, %
a b
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
The experimental conditions were optimized for the synthesis of amorphous SiO2 particles by the reaction of neutralization of sodium silicate solution with H2SO4 solution. Amorphous SiO2 particles were characterized by XRD, FT-IR, FE-SEM, EDS and microelectrophoresis. The amorphous peak was located at 2θ = 21.8º in the XRD pattern. Primary SiO2 particles were ~ 15 to ~ 30 nm in size and they aggregated into bigger particles. Amorphous SiO2 particles showed a specific surface area up to 130 m2g-1, dependent on the parameters of the precipitation process. The EDS spectrum of amorphous SiO2 particles did not show contamination with sulfate or other ions, which cannot be excluded in traces. pHzpc =1.7 was obtained by microelectrophoresis.
Article
Full-text available
The work considers the methods and techniques, allowing the assignment of the kinetic mechanisms to the chemical reactions evaluated from signals of thermoanalytical measurements. It describes which information about the kinetic mechanisms can be found from either model-free or model-based methods. The work considers the applicability of both methods and compares their results. The multiple-step reactions with well-separated peaks can be equally analyzed by both methods, but for overlapping peaks or for simultaneously running parallel reactions the model-free methods provide irrelevant results.
Article
This book contains eight chapters that discuss the manufacturing methods, surface treatment, composite interfaces, microstructure-property relationships with underlying fundamental physical and mechanical principles, and applications of carbon fibers and their composites. Recently, carbon-based materials have received much attention for their many potential applications. The carbon fibers are very strong, stiff, and lightweight, enabling the carbon materials to deliver improved performance in several applications such as aerospace, sports, automotive, wind energy, oil and gas, infrastructure, defense, and semiconductors. However, the use of carbon fibers in cost-sensitive, high-volume industrial applications is limited because of their relatively high costs. However, its production is expected to increase because of its widespread use in high-volume industrial applications; therefore, the methods used for manufacturing carbon fibers and carbon-fiber-reinforced composites and their structures and characteristics need to be investigated.
Article
A ceramic material of the composition Six Cy Nz was prepared by pyrolysis of a polymeric precursor based on polycarbosilane and bismaleimide in an inert medium at 1100°С. The thermochemical transformations of the polymeric product into the ceramic material and the properties of the final ceramic were studied. In the temperature interval 100–400°С, the chemical reactions yielding, along with the polycarbosilane/bismaleimide copolymer, also bismaleimide homopolymer occur in the systems. Pyrolysis of the ceramic-forming compounds obtained results in the formation of a ceramic whose yield is 20 wt % higher than the yield of the ceramic residue of polycarbosilane. The two-phase nature of the initial cured systems was assumed to be responsible for the low resistance of the ceramic material obtained to thermal oxidation. According to the data of synchronous thermal analysis, the weight loss the ceramic sample in air at 1200°С is 13 wt %.
Article
A commercial polycarbosilane (PCS) preceramic polymer has been characterised as-received and following curing under a variety of conditions. Elemental analysis, gel permeation chromatography (GPC), infra-red spectroscopy (FT-IR), simultaneous thermogravimetric analysis-differential thermal analysis (TG-DTA) and solid state nuclear magnetic resonance (NMR) have been employed. A number average molar mass of 1200 was found with a broad molar mass distribution ( [`(M)] \textw \overline M _{\text{w}} / [`(M)] \textn \overline M _{\text{n}} = 2.97). Elemental analysis gave an empirical formula of SiC2.2H5.3O0.3. IR and Solid state 29Si and 13C NMR spectra showed the presence of Si-O-Si, SiC4, SiC3H, Si-Si, Si-CH3 and Si-CH2 groups. Simultaneous TG-DTA performed under an argon flow showed that there was a weight gain which started at approximately 240 C. DTA showed an exotherm starting at this temperature showing that there was oxidation of the polymer even in an inert atmosphere. This is perhaps due to the oxygen in the PCS and there may also be some impurities in the inert atmosphere. Evidently the PCS is very sensitive to oxygen. Above 500 C, weight loss dominated although the exotherm continued to approximately 700 C. The effect of heating rate and dwell time at 200 C on the changes in the chemical composition during curing have been explored using IR and solid state NMR spectroscopies, and elemental analysis. The longer the cure time the higher was the weight gain and greater was the extent of the oxidation reactions. Elemental analysis showed that the ratio of H and C to Si decreased with holding time at the cure-temperature while the amount of oxygen increased. Use of a higher heating rate resulted in a lower weight gain when the same holding time was used. From this it is clear that curing starts below the holding temperature.
Article
The mechanism of structural transformation of divinylaromatic formulations containing methacrylates of secondary aromatic alcohols and diols into heat-resistant and strong glassy network polymers was examined. The structure and properties of the products were studied. The Rolivsans prepared demonstrate the possibility of chemical construction of high-temperature thermosets in which molecules with thermally stable ether groups and labile ester groups, providing excellent process properties of the formulation in the low-temperature step of the reactive forming, are capable of transforming into heat-resistant network polymers.
  • S.-J Park
S.-J. Park, Carbon Fiber, Springer, Netherlands, 2015.
  • E N Kablov
  • D V Grashchenkov
E. N. Kablov, D. V. Grashchenkov, et al., Steklo Keram., 4, 7 (2012).
  • S S Solntsev
  • V S Denisova
  • V A Rozenenkova
S. S. Solntsev, V. S. Denisova, and V. A. Rozenenkova, Aviats. Mater. Tekhnol., (S), 329 (2017).
  • E N Kablov
  • B V Shchetanov
E. N. Kablov, B. V. Shchetanov, et al., Tr. VIAM, No. 2, 5 (2013).