Chinese Journal of Aeronautics, (2020), 33(3): 1064–1073

Chinese Society of Aeronautics and Astronautics

& Beihang University
Chinese Journal of Aeronautics
[email protected]
www.sciencedirect.com

Methane pyrolysis in preparation of pyrolytic

carbon: Thermodynamic and kinetic analysis by
density functional theory
Chunxia HU a, He SHEN b, Shouyang ZHANG a, Hejun LI a,*

a
State Key Laboratory of Solidification Processing, C/C Composites Research Center, Northwestern Polytechnical University,
Xi’an 710072, China
b
Department of Mechanical Engineering, California State University, Los Angeles, CA 90032, USA

Received 19 November 2018; revised 6 December 2018; accepted 24 January 2019

Available online 28 May 2019

KEYWORDS Abstract The density functional theory has been successfully applied in analyzing pyrolytic carbon
Density functional theory; deposition by methane pyrolysis from the view of thermodynamics and kinetics based on a total
Kinetics; number of 39 elementary reactions. M06-2X/def2-TZVP level was employed to optimize species
Methane pyrolysis; structures and locate the transition states. The enthalpy changes and Gibbs free energy changes
Pyrolytic carbon; of all the reactions in the temperature range of 298.15–1800 K were derived with optimized species.
Thermodynamics Results show that the reacting temperature should be above 1200 K based on the equilibrium con-
stant analysis, which is consistent with the typical reaction temperature adopted in experiments.
Potential energy surface profiles report that radical attacking reactions have lower energy barriers
than those direct decomposition reactions, especially hydrogen radical attacking reactions. The
energy barriers of the first steps, dehydrogenations of methane and ethylene, are 272.4 kJ
mol1
and 288.9 kJ
mol1 at 1200 K, which are very close to the experimental activation energy for
methane pyrolysis. The most favorable decomposition reaction path is the path of hydrogen radical
attacking reactions. The highest energy barrier of the path at 1200 K is 185.7 kJ
mol1 presented by
the C–H bond breaking in ethynyl attacked by hydrogen radical.
Ó 2019 Chinese Society of Aeronautics and Astronautics. Production and hosting by Elsevier Ltd. This is
an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Carbon materials are showing a flourishing development with

* Corresponding author. various members, such as carbon nanotubes, graphene, fuller-
E-mail address: [email protected] (H. LI). ene and composites.1–4 Carbon fiber reinforced carbon matrix
Peer review under responsibility of Editorial Committee of CJA. (C/C) composites are widely used as rocket nozzles and air-
craft braking discs5–8 in aeronautic industry and further con-
sidered as a class of promising materials in electromagnetic
interference shielding for sensitive devices,9–11 because of their
Production and hosting by Elsevier

https://doi.org/10.1016/j.cja.2019.02.015
1000-9361 Ó 2019 Chinese Society of Aeronautics and Astronautics. Production and hosting by Elsevier Ltd.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Methane pyrolysis in preparation of pyrolytic carbon 1065

well physical and chemical properties. Indeed, these materials deposition. Thus, it is substantially important to study the
present low density, very good high-temperature mechanical small species energies and their reaction properties in growing
properties, high thermal conductivity and satisfactory electri- PyC for revealing the kinetics of PyC deposition accurately.
cal conductivity. Generally, such products are elaborated with In this work, various small species and elementary reactions
pyrolytic carbon (PyC) deposited in preformed carbon fibers in methane pyrolysis of PyC preparation are quantified with
by hydrocarbon pyrolysis using chemical vapor infiltration DFT accurately, aiming at getting the thermodynamic and
(CVI) technique. This process is a topic of great concerns kinetic properties of the process from the gas precursor to car-
despite it has been widely studied for several years, as its low bon. The reactions are proposed based on bond formations
kinetics and complex transport phenomena still lead to long and dissociations among main species, including dehydrogena-
durations and high costs of fabrications. Formations of PyC tion reactions, addition reactions and isomerization reactions.
with different textures, which have an important influence on Structures of species are first optimized at M06-2X/def2-TZVP
the holistic properties of C/C composites,12–15 are determined level and used in calculations of species thermodynamic
significantly by the kinetic of hydrocarbon pyrolysis. parameters (internal energy, enthalpy, Gibbs free energy and
A series of chemistry and kinetics of methane in CVI were entropy) and energy changes of reaction paths. The activation
studied in the late 19th century by Benzinger and Hüttinger. energy of each reaction path is provided by TS method based
The dissociation of methane was assumed to include five gas- on the transition state theory. Temperature effects on the
phase reactions and a surface reaction to form PyC.16 To make above parameters are also studied, revealing the different pre-
the assumption more reliable, thermodynamic calculations of ferred reaction paths at different temperatures.
gas phase equilibrium were done and more detailed elementary
reactions were considered,17 following a consecutive reaction 2. Reaction pathway design and calculation method
scheme:
k1 k2 Considering the experimental condition, the industry methane
C1 ! C2 ! C6þ
has a purity of 99.5%, the main impurity being ethylene (up to
#k3 #k4 ð1Þ
0.3%).18–20 Methane and ethylene, taken as the mother mole-
C1 C1 cules, directly decompose with the breaking of C–H bonds, fol-
where, Ci (i = 1, 2, 6+, 1) represents species with i carbon lowed by radical attacking reactions, leading to a huge amount
atoms and kj (j = 1, 2, 3, 4) is the reaction rate. of various radicals, aliphatic hydrocarbons and aromatic
Afterward, species concentrations during methane decom- hydrocarbons, as well as solid carbon. The present work will
position in preparing PyC were measured by experiments only focus on the species with no more than two carbon atoms
and used to enrich the reaction scheme specifically.18–20 In in order to limit the computer source of high-level calculations
addition, Li and Deutschmann developed a multi-step homo- in an acceptable time. The simplification of species is reason-
geneous reaction model and a heterogeneous surface reaction able as experimental results have shown that the main species
to simulate concentration ratios of intermediates in methane in methane pyrolysis are H2, CH4, C2H2, C2H4, C2H6 and
pyrolysis and PyC deposition in carbon fiber felts.21 Previ- C6H6, while the amount of C6H6 at a small residence time is
ously, the authors had summarized 909 elementary gas-phase zero.18–20 Thus several pathways of mother molecule pyrolysis
reactions and analyzed reaction pathways of methane in CVI are considered here. Table 1 lists the example reactions of
at the molecular level.22 With these researches, reaction routes methane and ethylene. More reactions of the radicals and
of methane pyrolysis in PyC deposition have been revealed. molecules are available in Table A1 in the Appendix A.
The resulting kinetic mechanisms were obtained either by sum- DFT at M06-2X/def2-TZVP level was employed to analyze
marizing the data in different researches or by using the the species and reactions with Gaussian software.25 M06-2X is
lumped experimental results and simulated the gas mole frac- chosen because it has been tested with databases of thermo-
tions and PyC deposition rates in the process, while the bottom chemistry, kinetics, bond lengths, vibrational frequencies and
of the kinetic modeling, the properties of species and related vibrational zero point energies and is recommended for appli-
chemical reactions, is picked with few caution. Liu et al.23 cations of main-group thermochemistry and kinetics.26–28 Fre-
studied the polycyclic aromatic hydrocarbon (PAH) radical quency analyses were conducted following the structure
formations and hydrogen transfer reactions with density func- optimization, which verified the stationary structures (with
tional theory (DFT) and revealed the radical energies and reac- all real frequencies). Thermodynamic parameters of all
tion rates in PAH deposition of PyC. Small species are not structures were performed at different temperatures
considered although they are easier to be produced with the (298.15–1800.0 K) with zero-point energy corrections. These
initial gas precursor and their depositions are dominant. The thermodynamic parameters satisfy29:
species and thermal rate constants of the CH4 + H ?
CH3 + H2 reaction were accurately calculated by DFT with
variational transition state theory and multidimensional semi- Table 1 Example reactions of methane and ethylene.
classical tunneling state theory24 without the discussion of pro-
Reactions of methane
cess applications. In the CVI process, PyC deposition is
influenced by what kind of carbon precursor is available and CH4 ? CH3 + H CH4 + H ? CH3 + H2
how the precursor and its radicals react. The hydrogen atom Reactions of ethylene
elimination of carbon precursors forms radicals and initiate
C2H4 isomerization C2H4 + H ? C2H3 + H2
a series reactions to grow species and PyC. And the isomeriza-
C2H4 ? C2H3 + H C2H4 + CH3 ? C2H3 + CH4
tion reactions and the addition reactions to form larger species C2H4 ? C2H2 + H2
compete with the dehydrogenation reactions of PyC
1066 C. HU et al.

E0 ¼ Eelec þ Ezpe ð2Þ of the species and transition states. No imaginary frequencies
are observed in the vibrational frequency analysis for the reac-
E ¼ E0 þ Evib þ Erot þ Etrans ð3Þ tants and products, which confirms that the obtained struc-
tures are true minimums on the potential surfaces. Sole
H ¼ E þ kB T ð4Þ imaginary frequency is shown for each transition state. An
IRC calculation has been done to determine the reactants
G ¼ H  TS ð5Þ and products that linked by the transition state. In Fig. 1,
the isomerization reactions expressed with species labels are
S ¼ S0 þ Svib þ Srot þ Strans ð6Þ
presented following each transition states. The calculated
Here, E, H, G and S are the internal energy, enthalpy, vibrational frequencies in cm1 of CH4 (Td, 1A1) are 1348.0,
Gibbs free energy and entropy, respectively. The subscripts 1567.5, 3068.4 and 3188.4 and those for C2H4 (D2h, 1Ag) are
0, elec, zpe, vib, rot, and trans stand for zero-point, electrical, 838.9, 1002.5, 1014.5, 1077.2, 1245.4, 1388.0, 1476.3, 1713.4,
zero-point-correction, vibration, rotation and transition. kB is 3154.5, 3169.5, 3231.8, 3257.9. Compared with the available
the Boltzmann constant and T is the temperature in Kelvin. experimental frequencies30 and theoretical results,31 the pre-
For each reaction path, the energy changes are the energy sent simulation results are in good agreements, especially when
differences between the reactants and products, since all the a scale factor (0.97132) is multiplied. The frequencies will be
atomic information has been cancelled out with the same atom scaled by the factor in the calculations of thermodynamic
number of each element on both sides of the reactions.29 The parameters.
equilibrium constant K is given by:
 
DG 3.2. Thermodynamic analysis of pyrolysis processes
K ¼ exp  ð7Þ
RT
Thermodynamic parameters of each elementary reaction in the
where DG is the relative Gibbs free energy change and R is the
pyrolysis process are obtained with its reactants and products.
gas constant.
Table 2 lists elementary reactions with any negative enthalpy
The transition state of each reaction path was located with
change DH or negative Gibbs free energy change DG at different
TS method and confirmed with frequency analysis (a sole
temperatures (298.15, 600, 900, 1200, 1500, 1800 K) at M06-
imaginary frequency) and intrinsic reaction coordinate (IRC)
2X/def2-TZVP level. The other reactions with positive DH
calculations. The energy analysis of transition states was done
and positive DG list in Table C1 in Appendix C, following the
at different temperatures, preparing for the kinetic study of the
sequence of DH at room temperature from large to small. For
pyrolysis process.
reactions with different structures, their thermodynamic data
are calculated separately. Among the total 39 elementary reac-
3. Results and discussion
tions, 11 reactions present negative enthalpy changes at all tem-
peratures, indicating an exothermal feature. As listed in the top
3.1. Structure and vibrational frequencies 11 rows of Table 2 according to the value of DH, these reactions
are mostly producing stable molecules from active radicals. The
The optimization results of structure, symmetry and electronic values of enthalpy changes in the other reactions are all posi-
state for the reactants, products and transition states calcu- tive, implying absorbing heat during the reactions. Thus, the
lated at M06-2X/def2-TZVP level are shown in Fig. 1 and mother molecule pyrolysis is generally an endothermic process
Fig. B1. Fig. 1 displays structures, symmetries and electronic as most of the reactions show endothermic properties.
states of isomers C2H2, C2H3 and C2H4 and their transition The Gibbs free energy change gives the information about
states at M06-2X/def2-TZVP level. Fig. B1 depicts the rest the spontaneity of the reaction and the conversion ratio of

Fig. 1 Structures, symmetries and electronic states for isomers C2H2, C2H3 and C2H4 and their transition states at M06-2X/def2-TZVP
level.
Methane pyrolysis in preparation of pyrolytic carbon 1067

Table 2 Elementary reactions with any negative enthalpy change DH or negative Gibbs free energy change DG at different
temperatures at M06-2X/def2-TZVP level (unit: kJ
mol1).
Elementary reactions Properties at different temperatures
Property T = 298.15 K T = 600 K T = 900 K T = 1200 K T = 1500 K T = 1800 K
CH + CH ? C2H2(S01) DH 968.8 971.2 971.0 969.8 968.0 965.7
DG 922.8 874.9 826.8 778.8 731.3 684.2
CH2 + CH2 ? C2H4(S06) DH 710.7 715.6 715.9 713.9 710.7 706.9
DG 659.6 605.5 550.2 495.2 440.9 387.2
CH3 + CH3 ? C2H6 DH 381.7 386.8 387.0 384.7 381.2 377.2
DG 333.5 282.3 229.8 177.7 126.4 75.8
C2H3(S04) + CH3 ? C2H2(S01) + CH4 DH 283.3 284.6 285.2 285.4 285.1 284.6
DG 270.5 256.9 242.9 228.8 214.7 200.7
C2H5 + CH3 ? C2H4(S06) + CH4 DH 283.0 287.7 290.5 292.2 293.4 294.2
DG 268.5 251.8 233.2 213.8 194.0 174.1
C2H3(S14) + H ? C2H2(S01) + H2 DH 277.3 275.9 276.2 277.5 279.2 281.1
DG 271.8 266.9 262.4 257.6 252.4 246.9
C2H5 + H ? C2H4(S06) + H2 DH 277.1 278.9 281.4 284.3 287.5 290.6
DG 269.8 261.7 252.6 242.6 231.8 220.3
C2H4(S07) ? C2H2(S01) + H2 DH 131.6 126.4 124.6 124.9 126.4 128.4
DG 159.5 190.0 222.4 254.9 287.3 319.3
CH + H ? C + H2 DH 96.4 96.4 96.7 97.3 98.0 98.7
DG 92.7 88.9 85.1 81.1 77.0 72.8
C2H6 + H ? C2H5 + H2 DH 10.2 7.7 7.5 8.9 10.9 13.3
DG 23.9 39.1 54.9 70.5 85.6 100.4
C2H6 + CH3 ? C2H5 + CH4 DH 16.2 16.4 16.6 16.7 16.8 16.9
DG 22.6 29.1 35.4 41.7 47.9 54.1
CH4 + H ? CH3 + H2 DH 6.0 8.7 9.1 7.9 5.9 3.5
DG 1.3 9.9 19.4 28.8 37.7 46.3
C2H4(S06) + H ? C2H3(S04) + H2 DH 27.0 28.7 28.4 26.9 24.8 22.4
DG 18.0 8.0 2.3 12.3 21.9 31.0
C2H6 ? C2H4(S06) + H2 DH 142.3 146.7 148.0 147.3 145.3 142.6
DG 106.3 67.9 28.1 11.9 51.4 90.6
C2H3(S15) ? C2H2(S02) + H DH 129.7 134.5 137.1 138.3 138.7 138.7
DG 99.8 67.5 33.3 1.5 36.5 71.5
CH3 + H ? CH2 + H2 DH 20.9 21.1 20.0 18.0 15.5 12.9
DG 16.2 11.2 6.4 2.2 1.4 4.6
C2H5 ? C2H3(S04) + H2 DH 179.5 183.0 183.9 183.0 181.0 178.3
DG 148.2 115.0 80.6 46.3 12.3 21.2
C2H3(S04) ? C2H2(S01) + H DH 152.2 157.4 160.8 163.0 164.5 165.5
DG 128.2 101.8 73.1 43.6 13.5 16.8

reactants to products33,34. Reactions with negative DG can much products could be produced. The equilibrium constants
happen spontaneously. All exothermal reactions show negative increase with the increase of temperature for endothermic reac-
values of DG above the ambient temperature as shown in tions and decrease for exothermic reactions. Thus with most of
Table 2. The endothermic reactions with any negative DG the reaction paths here being endothermic reactions, more
are also listed in Table 2 in the order of the temperature at products are going to be obtained at high temperatures at
which DG becomes negative. The reaction of methane attacked the equilibrium condition. At ambient temperature, the equi-
by hydrogen radical can occur spontaneously at room temper- librium constants of hydrogen abstraction reactions with
ature, while the reaction of ethylene attacked by hydrogen rad- methane and ethylene are so small (1070 and 1074) because
ical show negative Gibbs free energies above 900 K. The Gibbs of the large Gibbs free energy changes, indicating that the
free energy changes of C2H6 ? C2H4(S06) + H2 and decomposition process is inhibited and the amount of radicals
C2H3(S05) ? C2H2(S03) + H decrease with the increasing formed is very limited. These equilibrium constants increase
temperature and become negative above 1200 K. Similar dramatically when the temperature increases from room tem-
trends have been shown in the reactions of methyl attacked perature to 1800 K, being 106. The values of equilibrium con-
by hydrogen radical and ethyl cracking into vinyl and hydro- stants for different reactions vary in a very broad range, for
gen with the turning points at 1500 K and 1800 K respectively. example, 10252 at 298.15 K with the smallest value being
These show that a high temperature is beneficial for the 7.5  1091(C2H2(S01) ? C2H + H) and the largest being
process. 4.7  10161(CH + CH ? C2H 2(S01)). Most equilibrium
The conversion ratios of reactions at equilibrium calculated constants show a distinct variation with the increasing
with Eq. (7) give much more visualized information about how temperature and the value range for all reactions narrows
1068 C. HU et al.

quickly. Significant variations of equilibrium constants appear decomposition reaction paths at 298.15 K. Data in the square
at temperatures below 1200 K. When the temperature brackets are relative Gibbs free energies in kJ
mol1 calculated
increases above 1200 K, the value range of equilibrium con- at M06-2X/def2-TZVP level. Fig. 2(a) depicts the direct
stants is no larger than 50, indicating that the process has a decomposition of methane. Methane dissociates a hydrogen
quite well conversion. atom to form methyl with an energy change of 398.7 kJ
mol1,
which excellently agrees with the experimental result.37
3.3. Kinetic analysis of pyrolysis processes Following this, one C–H bond is breaking in each step, form-
ing C at last. There are no transition states for these bond-
The activation energies of the reaction paths are calculated at cleavage reactions, as no saddle point is found on the potential
M06-2X/def2-TZVP level based on the transition state the- energy surface. Thus BDEs, that need to be overcome during
ory.35,36 Fig. 2 illustrates the potential energy profiles of direct the reaction process, are approximately considered as

Fig. 2 Potential energy profiles of reactions in methane pyrolysis at room temperature.

Methane pyrolysis in preparation of pyrolytic carbon 1069

activation energies.38 The highest energy in the direct methane of ethylene, acetylene can also be formed by the association
decomposition is 416.2 kJ
mol1, for methyl to produce car- of two CH radicals. The dehydrogenation reactions give high
bene. Fig. 2(b) shows the direct decomposition of ethane. energy barriers, which are even higher than those in the direct
Ethane can be produced by an addition reaction between methane decomposition, implying that enough energy should
two methyl radicals and consumed by producing hydrogen be provided for direct decompositions of carbon deposition.
radicals or hydrogen molecules. The highest energy barrier is Apart from the direct decompositions, reactions with active
487.6 kJ
mol1 for the production of ethylene and hydrogen radicals are also available after a few direct decomposition
via TS14. Fig. 2(c) presents the direct decomposition of ethy- reactions have occurred. Fig. 3 gives potential energy surface
lene. Ethylene is either imported as the impurity of the mother profiles of reactions with methyl and hydrogen radicals at
gas or produced by an addition reaction of carbene. Ethylene 298.15 K. Data in square brackets are relative Gibbs free ener-
decomposes into vinyl and hydrogen radical with a high energy gies in kJ
mol1 calculated at M06-2X/def2-TZVP level.
barrier of 418.0 kJ
mol1, showing a well consistency with the According to the figure, methyl and hydrogen radicals have
previous calculation at G3(MP2)//B3PW91 level.39 Isomeriza- lowered the energy barrier in a large amount, indicating easy
tion of ethylene is much more preferred with a lower energy pathways for decompositions. The energy barrier of methane
barrier of 384.6 kJ
mol1 via TS04 for the direct ethylene attacked by hydrogen radical is 77.3 kJ
mol1, which is only
decomposition, followed by producing acetylene and hydro- 19.4% of that for the direct methane dehydrogenation. The
gen. The isomerization reactions of C2H3 and C2H2 have also energy barrier of the dehydrogenation for ethylene attacked
been expressed in Fig. 2(c) and shows that the most stable by hydrogen radical reduces 79.7% than that of the direct
structures are vinyl and acetylene. Fig. 2(d) is for the direct ethylene dehydrogenation. Usually, the reactions with hydro-
decomposition of acetylene. Except from the decomposition gen radicals express lower energy barriers that those with

Fig. 3 Potential energy surface profiles of reactions with active radicals at 298.15 K.

Fig. 4 The most favorable decomposition reaction paths with hydrogen radicals at 1200 K.
1070 C. HU et al.

methyl radicals. The energy barrier for ethane attacked by the highest energy barrier of 185.7 kJ
mol1 in the decomposi-
methyl is 1.3 times as large as that of ethane attacked by tion process.
hydrogen radical, implying that hydrogen radical is more
active than methyl radical. 4. Conclusions
With the increase of the temperature, the potential energy
profiles vary. The energy barriers of the first C–H bond cleav- (1) Methane decomposition in preparation of pyrolytic car-
age in methane and ethylene are 272.4 kJ
mol1 and bon was studied by an accurate chemistry model based
288.9 kJ
mol1 at 1200 K, which is very close to the experi- on the density functional theory with a total number
mental result for methane pyrolysis.17,19 For the direct of 39 elementary reactions. Stable species and transition
methane decomposition reactions, the highest value of relative states were optimized at M06-2X/def2-TZVP level.
Gibbs free energies at 1200 K is 308.0 kJ mol1 for carbene Molecular energies in the temperature range of 298.15–
dehydrogenation reaction and decreases 27.7% than that at 1800 K were calculated with the optimized structures
ambient temperature. The reductions for the energy barriers and scaled frequencies and used in the calculations of
of direct ethylene and ethane decomposition reactions are enthalpy changes and Gibbs free energy changes of
19.1% and 24.1%, given by C2H3(S05) dehydrogenation and reactions.
ethyl producing C2H3(S05). Ethylene is preferred to produce (2) Most of the enthalpy changes are positive, implying the
vinyl and hydrogen radical at 1200 K, which is quite different endothermic feature of the pyrolysis process. Analysis of
from the situation at 298.15 K. In direct acetylene decomposi- Gibbs free energy changes show that all the exothermic
tion reactions, the relative energy of the final product C2(S16) reactions can happen spontaneously at room tempera-
compared with that of acetylene at 1200 K is much lower than ture, as well as the reaction of methane attacked by
that at room temperature, which expresses a more stable state hydrogen radical. Well conversion ratios are obtained
at high temperatures. While the radical attacking reactions for all reactions above 1200 K, consistent with the typi-
show increasing energy barriers with the increase of the tem- cal reaction temperature employed in experiments.
perature, reactions with radicals are still more favorable than (3) Potential energy surface profiles of the pyrolysis process
the direct decomposition reactions and the hydrogen radical including direct decompositions and radical attacking
is still more active than methyl at 1200 K. Fig. 4 depicts the reactions at ambient temperature show that radical
most favorable decomposition reaction paths with hydrogen attacking reactions have lower energy barriers than
radicals at 1200 K. Data in square brackets along the paths those direct decomposition reactions, especially hydro-
are relative Gibbs free energies in kJ
mol1 calculated at gen radical attacking reactions. As the temperature
M06-2X/def2-TZVP level. Methane is attacked by hydrogen increases, the energy barriers of direct decomposition
radical to produce methyl via TS06 with an energy barrier of reactions decrease.
152.0 kJ
mol1. Then one C–H bond of methyl could be bro- (4) The first steps C–H bond cleavages in methane and ethy-
ken by hydrogen radical attacking via TS05 with an energy lene present energy barriers of 272.4 kJ
mol1 and
barrier of 152.9 kJ
mol1, showing the highest energy barrier 288.9 kJ
mol1 at 1200 K, which is very close to the
in C1 decomposition (Fig. 4(a)). There is only a small energy experimental activation energy for methane pyrolysis.
difference between the reactants and products in The most favorable decomposition reaction paths with
CH2 + H ? CH + H2, being 6.9 kJ
mol1. To release C from hydrogen radicals at 1200 K were illustrated with the
CH, the energy change is 81.1 kJ
mol1, making C the most highest energy barrier of 185.7 kJ
mol1 for the C–H
stable species among all the C1 species. Fig. 4(b) shows the bond breaking in ethynyl attacked by hydrogen radical.
most favorable decomposition reaction paths of C2 species
with hydrogen radicals at 1200 K. Ethane attacked by hydro-
gen radical forms ethyl via TS16 with an energy barrier of
Acknowledgements
136.9 kJ
mol1. An energy decrease is found in ethylene pro-
duced from vinyl attacked by hydrogen radical. Ethylene
This work was supported by the National Natural Science
attacked by hydrogen radical could dissociate a hydrogen
Foundation of China (51821091 and 51472203), the ‘‘111”
atom to form vinyl via TS12 with an energy barrier of
Project of China (Grant No. B08040), and the Research Fund
159.2 kJ
mol1. The most stable species in the C2 decomposi-
of State Key Laboratory of Solidification Processing (NWPU)
tion process is acetylene, producing from vinyl attacked by
of China (Grant No. 142-TZ-2016).
hydrogen radical with an energy reduction of 257.6 kJ
mol1.
Acetylene is attacked by hydrogen radical and forms C2H,
Appendix A. Reactions
showing an increasing energy of 92.8 kJ
mol1. C2H decom-
pose the last hydrogen atom to form C2(S16) via TS07 with

Table A1 Reactions of radicals and molecules except those with methane and
ethylene.
Reactions of C1 species
CH3 ? CH2 + H CH2 + CH2 ? C2H4
CH3 + H ? CH2 + H2 CH ? C + H
Methane pyrolysis in preparation of pyrolytic carbon 1071

Table A1 (continued)
CH3 + CH3 ? C2H6 CH + H ? C + H2
CH2 ? CH + H CH + CH ? C2H2
CH2 + H ? CH + H2
Reactions of C2 species
C2H6 ? C2H5 + H C2H3 + H ? C2H2 + H2
C2H6 ? C2H4 + H2 C2H3 + CH3 ? C2H2 + CH4
C2H6 + H ? C2H5 + H2 C2H2 isomerization
C2H6 + CH3 ? C2H5 + CH4 C2H2 ? C2H + H
C2H5 ? C2H + H C2H2 + H ? C2H + H2
C2H5 ? C2H3 + H2 C2H2 + CH3 ? C2H + CH4
C2H5 + H ? C2H4 + H2 C2H ? C2 + H
C2H5 + CH3 ? C2H4 + CH4 C2H + H ? C2 + H2
C2H3 isomerization C2H + CH3 ? C2 + CH4
C2H3 ? C2H2 + H

Appendix B. Species and transition states

Fig. B1 Structures, symmetries and electronic states for stable species without isomers and transition states of dehydrogenation
reactions at M06-2X/def2-TZVP level.
1072 C. HU et al.

Appendix C. Enthalpy changes and Gibbs free energy changes

Table C1 Elementary reactions and their positive enthalpy changes DH and positive Gibbs free energy changes DG at different
temperatures at M06-2X/def2-TZVP level (unit: kJ mol1).
Elementary reactions Properties at different temperatures
Property T = 298.15 K T = 600 K T = 900 K T = 1200 K T = 1500 K T = 1800 K
C2H2(S01) ? C2H + H DH 551.3 555.1 557.3 558.5 559.2 559.4
DG 514.4 475.4 435.0 394.0 352.8 311.5
C2H ? C2(S16) + H DH 498.7 501.3 503.2 504.5 505.2 505.4
DG 467.0 433.9 399.7 365.0 330.1 295.0
C2H4(S06) ? C2H3(S04) + H DH 456.5 462.0 465.4 467.4 468.5 469.0
DG 418.0 376.7 333.2 288.9 244.1 199.1
CH3 ? CH2 + H DH 450.4 454.4 457.0 458.5 459.2 459.5
DG 416.2 379.8 342.0 303.4 264.5 225.5
CH4 ? CH3 + H DH 435.5 442.0 446.1 448.4 449.6 450.1
DG 398.7 358.7 316.1 272.4 228.2 183.9
CH2 ? CH + H DH 433.4 437.5 440.6 443.2 445.1 446.6
DG 404.7 374.0 341.5 308.0 274.0 239.6
C2H6 ? C2H5 + H DH 419.3 425.6 429.4 431.6 432.8 433.3
DG 376.1 329.6 280.7 230.7 180.3 129.8
C2H5 ? C2H3(S05) + H2 DH 383.4 387.1 388.3 387.7 386.0 383.7
DG 351.2 316.9 281.4 245.8 210.5 175.6
C2H4(S06) ? C2H2(S02) + H2 DH 360.7 367.3 369.8 369.9 368.5 366.4
DG 320.8 277.4 231.8 185.8 139.9 94.3
CH ? C + H DH 333.1 336.9 340.3 343.2 345.7 347.9
DG 307.3 279.7 250.4 220.0 188.9 157.4
C2H4(S06) ? C2H4(S07) DH 310.9 312.5 313.8 314.8 315.6 316.3
DG 305.7 299.8 293.2 286.2 278.9 271.5
C2H3(S04) ? C2H3(S05) DH 203.9 204.1 204.3 204.7 205.0 205.3
DG 203.0 201.9 200.8 199.5 198.2 196.8
C2H2(S01) ? C2H2(S03) DH 177.9 177.5 177.0 176.6 176.1 175.5
DG 170.7 163.5 156.5 149.8 143.1 136.5
C2H5 ? C2H4(S06) + H DH 152.5 154.4 155.6 156.1 156.2 156.0
DG 130.2 106.9 82.9 58.6 34.2 9.8
C2H2(S01) + H ? C2H + H2 DH 121.8 121.8 120.3 118.1 115.5 112.8
DG 114.4 106.7 99.5 92.8 86.8 81.3
C2H2(S01) + CH3 ? C2H + CH4 DH 115.8 113.1 111.2 110.2 109.6 109.3
DG 115.7 116.7 118.9 121.6 124.6 127.6
C2H + H ? C2(S16) + H2 DH 69.2 68.0 66.2 64.0 61.5 58.8
DG 67.0 65.2 64.2 63.9 64.1 64.9
C2H + CH3 ? C2(S16) + CH4 DH 63.2 59.3 57.1 56.1 55.6 55.3
DG 68.3 75.2 83.6 92.7 101.9 111.2
C2H4(S06) + CH3 ? C2H3(S04) + CH4 DH 21.0 20.0 19.3 19.0 18.9 18.9
DG 19.3 18.0 17.2 16.5 15.9 15.3
CH2 + H ? CH + H2 DH 3.9 4.2 3.7 2.7 1.4 0.1
DG 4.7 5.3 6.0 6.9 8.1 9.5
C2H2(S03) ? C2H2(S02) DH 3.6 3.7 3.6 3.4 3.1 2.9
DG 3.9 4.2 4.4 4.8 5.1 5.6

References 4. Wang J, Zhang X, Miao Y, Li Y, Xi X, Zhong X, et al. The

influences of carbon nanotubes introduced in three different
1. Zhang H, Wu W, Yu X, Tong M, Zhou J, Cao J, et al. Structure phases of carbon fiber/pyrolytic carbon/silicon carbide composites
and electrocatalytic activity of flexible reduced graphene oxide/- on microstructure and properties of their composites. Carbon
fullerene smart fiber. Carbon 2019;142:411–9. 2018;129:409–14.
2. Wang C, Murugadoss V, Kong J, He Z, Mai X, Shao Q, et al. 5. Chowdhury P, Sehitoglu H, Rateick R. Damage tolerance of
Overview of carbon nanostructures and nanocomposites for carbon-carbon composites in aerospace application. Carbon
electromagnetic wave shielding. Carbon 2018;140:696–733. 2018;126:382–93.
3. Liu D, Song Y, Zhou Y, Song X, Fu W, Feng J. Brazing of C/C 6. Zaman W, Li K, Ikram S, Li W, Zhang D, Guo L. Morphology,
composite and Ti-6Al-4V with graphene strengthened AgCuTi thermal response and anti-ablation performance of 3D-four
filler: Effects of graphene on wettability, microstructure and directional pitch-based carbon/carbon composites. Corros Sci
mechanical properties. Chinese J Aeronaut 2018;31(7):1602–8. 2012;61:134–42.
Methane pyrolysis in preparation of pyrolytic carbon 1073

7. Li X, Li K, Li H, Wei J, Wang C. Microstructures and mechanical 23. Liu H, Yan L, Yue B, Li A. Hydrogen transfer reaction in
properties of carbon/carbon composites reinforced with carbon polycyclic aromatic hydrocarbon radicals. J Phys Chem A
nanofibers/nanotubes produced in situ. Carbon 2007;45(8):1662–8. 2014;118(25):4405–14.
8. Buckley JD, Edie DD. Carbon/carbon materials and compos- 24. Truong TN. A direct ab initio dynamics approach for calculating
ites. New Jersey: Noyes Publications; 1993. p. 267–81. thermal rate constants using variational transition state theory and
9. Jia Y, Li K, Xue L, Ren J, Zhang S, Li H. Mechanical and multidimensional semiclassical tunneling methods. An application
electromagnetic shielding performance of carbon fiber reinforced to the CH4 + H M CH3 + H2 reaction. J Chem Phys 1994;100
multilayered (PyC-SiC)n matrix composites. Carbon (11):8014–25.
2017;111:299–308. 25. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA,
10. Liu X, Yin X, Kong L, Li Q, Liu Y, Duan W, et al. Fabrication Cheeseman JR, et al. Gaussian 09, Revision D.01. Wallingford,
and electromagnetic interference shielding effectiveness of carbon CT: Gaussian Inc.; 2013.
nanotube reinforced carbon fiber/pyrolytic carbon composites. 26. Zhao Y, Truhlar DG. The M06 suite of density functionals for
Carbon 2014;68:501–10. main group thermochemistry, thermochemical kinetics, noncova-
11. Luo X, Chung DDL. Electromagnetic interference shielding using lent interactions, excited states, and transition elements: Two new
continuous carbon-fiber carbon-matrix and polymer-matrix com- functionals and systematic testing of four M06-class functionals
posites. Compos Part B Eng 1999;30(3):227–31. and 12 other functionals. Theor Chem Acc 2008;120(1):215–41.
12. Delhaes P. Chemical vapor deposition and infiltration processes of 27. Iversen KJ, Wilson DJD, Dutton JL. Comparison of the mech-
carbon materials. Carbon 2002;40(5):641–57. anism of borane, silane, and beryllium hydride ring insertion into
13. Xiong X, Huang B, Li J, Xu H. Friction behaviors of carbon/car- N-heterocyclic carbene C–N bonds: a computational study.
bon composites with different pyrolytic carbon textures. Carbon Organometallics 2013;32(21):6209–17.
2006;44(3):463–7. 28. Lawal MM, Govender T, Maguire GEM, Kruger HG, Honarpar-
14. Dong GL, Hüttinger KJ. Consideration of reaction mechanisms var B. DFT study of the acid-catalyzed esterification reaction
leading to pyrolytic carbon of different textures. Carbon 2002;40 mechanism of methanol with carboxylic acid and its halide
(14):2515–28. derivatives. Int J Quantum Chem 2018;118(4) e25497.
15. Li H, Li H, Li K, Wang Y, Zhang D, Lu J. Mechanical properties 29. Ochterski JW. Thermochemistry in Gaussian. 2000. p. 1–19.
improvement of carbon/carbon composites by two different 30. Malcolm W, Chase J. NIST-JANAF thermochemical tables. J
matrixes. J Mater Sci 2011;46(13):4667–74. Phys Chem Ref Data 1998, Monograph No.9:615, 676.
16. Benzinger W, Hüttinger KJ. Chemical vapor infiltration of 31. Yao X, Su K, Deng J, et al. Gas-phase reaction thermodynamics
pyrocarbon—III: the influence of increasing methane partial in preparation of pyrolytic carbon by propylene pyrolysis. Comput
pressure at increasing total pressure on infiltration rate and degree Mater Sci 2007;40(4):504–24.
of pore filling. Carbon 1999;37(2):181–93. 32. Alecu IM, Zheng J, Zhao Y, Truhlar DG. Computational
17. Benzinger W, Becker A, Hüttinger KJ. Chemistry and kinetics of thermochemistry: Scale factor databases and scale factors for
chemical vapour deposition of pyrocarbon: I. Fundamentals of vibrational frequencies obtained from electronic model chemis-
kinetics and chemical reaction engineering. Carbon 1996;34 tries. J Chem Theory Comput 2010;6(9):2872–87.
(8):957–66. 33. Tsushima S, Yang T, Suzuki A. Theoretical Gibbs free energy
18. Becker A, Hüttinger KJ. Chemistry and kinetics of chemical vapor study on UO2(H2O)2+ n and its hydrolysis products. Chem Phys
deposition of pyrocarbon — IV pyrocarbon deposition from Lett 2001;334(4):365–73.
methane in the low temperature regime. Carbon 1998;36 34. Lüttge A. Crystal dissolution kinetics and Gibbs free energy. J
(3):213–24. Electron Spectrosc Relat Phenom 2006;150(2):248–59.
19. Brüggert M, Hu Z, Hüttinger KJ. Chemistry and kinetics of 35. Steinfeld JI, Francisco JS, Hase WL. Chemical kinetics and
chemical vapor deposition of pyrocarbon: VI. Influence of dynamics. Englewood Cliffs (New Jersey): Prentice Hall; 1989. p.
temperature using methane as a carbon source. Carbon 1999;37 287–320.
(12):2021–30. 36. Truhlar DG, Garrett BC, Klippenstein SJ. Current status of
20. Antes J, Hu Z, Zhang W, Hüttinger KJ. Chemistry and kinetics of transition-state theory. J Phys Chem 1996;100(31):12771–800.
chemical vapour deposition of pyrocarbon: VII. Confirmation of 37. NIST chemistry webbook [Internet]. Available from: http://web-
the influence of the substrate surface area/reactor volume ratio. book.nist.gov/chemistry.
Carbon 1999;37(12):2031–9. 38. Jarvis MW, Daily JW, Carstensen HH, Dean AM, SHarman S,
21. Li A, Deutschmann O. Transient modeling of chemical vapor Dayton DC, et al. Direct detection of products from the pyrolysis
infiltration of methane using multi-step reaction and deposition of 2-phenethyl phenyl ether. J Phys Chem A 2011;115(4):428–38.
models. Chem Eng Sci 2007;62(18–20):4976–82. 39. Qu Y, Su K, Wang X, Liu Y, Zeng Q, Cheng L, et al. Reaction
22. Hu C, Li H, Zhang S, Li W. A molecular-level analysis of gas- pathways of propene pyrolysis. J Comput Chem 2010;31
phase reactions in chemical vapor deposition of carbon from (7):1421–42.
methane using a detailed kinetic model. J Mater Sci 2016;51
(8):3897–906.