Kinetic modeling of benzene decomposition near 1000 K: The effects of toluene impurity

Article abstract of DOI:10.1002/(SICI)1097-4601(1999)31:8<577::AID-KIN7>3.0.CO;2-K

Kinetic data of Brooks et al. on the decomposition of C₆H₆ near 1000 K have been analyzed by computer modeling. The observed overall 3/2-order kinetics could be accounted for by a mechanism composed of 4 key reactions involving H atoms and C₆H₅ radicals using recently acquired rate constants. However, the appearance of CH₄ and the enhanced H₂ yields could only be explained by invoking the reactions of approximately 0.1% of toluene present in the system as reported by the authors. Overall, the decomposition reaction is dominated by the unimolecular dissociation of C₆H₆ followed by the short chain process, H+C₆H₆ = C₆H₅+H₂ and C₆H₅+C₆H₆ = C₁₂H₁₀+H, which result in the dehydrogenation of C₆H₆, producing C₁₂H₁₀+H₂. In order to account for the yield of H₂ quantitatively, the displacement reaction, C₆H₅CH₂+C₆H₆ = CH₂(C₆H₅)₂+H, was invoked and modeled to have the approximate rate constant, 8.4×10¹¹ exp(-11800/T) cm³ mol^-1 s^-1.

Full text of DOI:10.1002/(SICI)1097-4601(1999)31:8<577::AID-KIN7>3.0.CO;2-K

Kinetic Modeling of
Benzene Decomposition
Near 1000 K: The Effects
of Toluene Impurity
M. G. BRIOUKOV, J. PARK, M. C. LIN
Department of Chemistry, Emory University, Atlanta, Georgia 30322
Received 29 September 1997; accepted 31 March 1999
ABSTRACT: Kinetic data of Brooks et al. [3] on the decomposition of C H near 1000 K have
6
6
been analyzed by computer modeling. The observed overall 3/2-order kinetics could be ac-
counted for by a mechanism composed of 4 key reactions involving H atoms and C H radicals
6
5
using recently acquired rate constants. However, the appearance of CH and the enhanced H
4
2
yields could only be explained by invoking the reactions of ϳ0.1% of toluene present in the
system as reported by the authors. Overall, the decomposition reaction is dominated by the
unimolecular dissociation of C H followed by the short chain process, H ϩ C H ϭ C H ϩ
6
6
6
6
6
5
H and C H ϩ C H ϭ C H ϩ H, which result in the dehydrogenation of C H , producing
2
6
5
6
6
12 10
6
6
C H ϩ H . In order to account for the yield of H quantitatively, the displacement reaction,
1
2
10
2
2
C H CH ϩ C H ϭ CH (C H ) ϩ H, was invoked and modeled to have the approximate rate
6
5
2
6
6
2
6
5 2
constant, 8.4 ϫ 1011 exp(Ϫ11800/T) cm mol s . ᭧ 1999 John Wiley & Sons, Inc. Int J Chem Kinet
3
Ϫ
1
Ϫ1
3
1: 577–582, 1999
INTRODUCTION
in the vicinity of 1000 K, which is controlled primarily
by the initiation reaction. In this lower temperature
regime (T Յ 1200 K), the ring-opening processes are
relatively unimportant and the overall mechanism in-
volves only a handful of reaction steps, which have
been recently studied experimentally and theoretically
with sufficient reliability. For example, for the first few
steps of the decomposition mechanism [3]:
The kinetics and mechanism of benzene decomposi-
tion reaction is relevant to the combustion chemistry
of hydrocarbons in conjunction with the formation of
soot in its incipient stages [1,2]. The thermal decom-
position reaction has been investigated by many lab-
oratories using a variety of techniques [3–8]. At tem-
peratures above 1200 K, the reaction becomes
complex and is dominated by ring-opening processes.
At temperatures between 900 and 1200 K, the dehy-
drogenation reactions producing H ϩ C H (bi-
C H !: C H ϩ H
(1)
(2)
(3)
(4)
6
6
6
5
2
12 10
C H ϩ H !: C H ϩ H
2
6
6
6
5
phenyl) are dominant [3].
This study is focused on the elucidation of the
mechanism for the dehydrogenation process occurring
C H ϩ C H !: C H ϩ H
6
5
6
6
12 10
C H ϩ C H !: C H
12 10
Correspondence to: M. G. Brioukov (chemmcl@emory.edu)
Contract grant sponsor: Basic Energy Sciences, Department of
Energy
6
5
6
5
their rate constants have been determined (see later)
and are employed in this study for kinetic modeling
Contract grant number: DE-FG02-97-ER14784
1999 John Wiley & Sons, Inc.
᭧
CCC 0538-8066/99/080577-06

5
78
BRIOUKOV, PARK, AND LIN
of H2 production measured by Brooks and co-
workers [3].
The results of this study indicate that the overall
rate constants for the production of hydrogen, which
obeys the 3/2-order rate law,
3/2
d[H ]/dt ϭ k [C H ]
2
6
6
with k ϭ 109.93 exp(Ϫ53,790/T) dm mol
3
Ϫ
1/2
s , is
Ϫ1
strongly influenced by the ϳ0.1% of toluene impurity
present in their reaction mixtures. The effect was ex-
amined by computer simulation and the results pre-
sented herein.
KINETIC MODELING
Figure 1 Function plot to test proposed mechanism at 993
K. Dashed line, kinetically modeled H yield excluding the
2
The kinetic data obtained from the pyrolysis of ben-
zene in a static reactor in the temperature range of
reversible reaction of reaction (1); dotted line, kinetically
modeled H yield including the reversible reaction of reac-
2
9
10–1157 K by Brooks and coworkers [3] were mod-
tion (1). The experimental data was reproduced from Figure
8
of Ref. 3.
eled with recently established rate constants for the
key elementary steps, including reaction (1)–(4) em-
ployed in their original analysis.
Steady-state assumption for [H] and [C H ] em-
13
Ϫ
55/T
3
Ϫ1
6
5
be represented by k ϭ 1.38 ϫ 10
s .
e
cm mol
4
ploying (1)–(4) led to the observed 3/2-order kinetic
see Fig. 1) for H production [3]:
Ϫ1
(
2
The employment of the “best” kinetic data (shown
in Table I) including our recent measurements on re-
action (7) and (13) [19,20] to account for the H pro-
1/2
3/2
d[H ]/dt ϭ k [C H ] ϩ k (k /k ) [C H ]
(I)
2
1
6
6
2
1
4
6
6
2
duction presented in Fig. 1 with equation (I), however,
leads to a significantly higher yield than the observed
one as indicated by the dashed line. The large devia-
tion results essentially from the exclusion of reverse
reactions, particularly (Ϫ1). The inclusion of all re-
verse processes in the scheme as required by modeling
with the CHEMKIN/SENKIN program [21,22] re-
provided that the initiation rate, k [C H ], is small in
1
6
6
comparison with the chain-propagation rate. The as-
sumption is valid on account of the high C–H bond
strength in benzene and zero intercept revealed by Fig-
ure 1. Accordingly equation (I) can in principle be
utilized to analyze the observed kinetic data and to
evaluate k or k , for example.
duces the H production below the experimental val-
1
2
2
At that time, the authors were unable to accomplish
this analysis because of the absence of reliable values
for all three rate constants. They have since been de-
termined directly by experiments or indirectly through
kinetic modeling or theoretical calculations. Reaction
ues as illustrated by the dotted line. For kinetic mod-
eling the thermodynamic property of C H5 was
6
obtained by using the recent ab initio result by Mebel
et al. [12] and those of CH (C H ) and CH(C H )
2
6
5 2
6
5 2
were obtained by using Melius’ BAC-MP2 results
[23]. All other thermodynamic properties were taken
directly from CHEMKIN/SENKIN thermodynamic li-
brary. Figure 2 further demonstrates the underesti-
(1) or its reverse process has been determined by sev-
eral groups [7–11]. Recently, reaction (2) was calcu-
lated by us with ab initio molecular orbital and statis-
tical theory [12]. The computed reverse rate constant
mation of kinetically modeled results in H production
2
(
k ) has been shown to agree with experimentally
by dashed (967 K) and dotted (910 K) lines. The
deviation between the experiment and the modeled re-
sult using the four reversible reactions may be attrib-
uted to the presence of the ϳ0.1% of C H CH im-
purity in the mixture as reported by the original
authors [3].
Ϫ
2
measured data [13]. Reaction (3), first studied by Fahr
and Stein [14], has been measured in its reverse direc-
tion by Manion and Tsang [15]. These data are kinet-
ically consistent. Recently, three independent sets of
data have been reported for reaction (4) [16–18]. The
result of Park and Lin [16], which covers a broader
6
5
3
Aside from the major toluene reactions responsible
for the enhanced initiation rate [reaction (Ϫ5) and
1
/T range, lies in the middle with the values which can

EFFECTS OF TOLUENE IMPURITY
Table I The Reactions and Rate Constants Used in the Modeling of Considered Kinetic Dataa
579
Reactions
A
n
E
Ref.
1
4
1
1
1
1
3
1
2
3
4
5
6
7
8
9
. C H ϩ H ϭ C H
7.83 ϫ 10
5.71 ϫ 10
4.10 ϫ 10
1.39 ϫ 10
1.50 ϫ 10
1.39 ϫ 10
2.70 ϫ 10
8.35 ϫ 10
1.00 ϫ 10
3.97 ϫ 10
6.62 ϫ 10
1.17 ϫ 10
3.90 ϫ 10
1.32 ϫ 10
1.00 ϫ 10
1.00 ϫ 10
2.51 ϫ 10
1.19 ϫ 10
5.40 ϫ 10
3.61 ϫ 10
0.0
2.43
0.0
0.0
0.0
0.0
4.02
0.0
0.0
3.4
3.0
0.0
4.57
3.0
0.0
0.0
0.4
0.0
0
26
12
15
16
27
20
19
6
5
6
6
. C H ϩ H ϭ C H ϩ H
6280
8780
111
0
6
5
2
6
6
3
3
4
3
. C H ϩ H ϭ C H ϩ C H
6
12
10
6
5
6
. 2C H ϭ C H
12 10
6
5
. C H CH ϩ H ϭ C H CH
3
6
6
2
6
5
. C H ϩ CH ϭ C H CH
3
30
6
5
3
6
5
Ϫ
1
. C H ϩ C H CH ϭ C H ϩ C H CH
2
Ϫ1590
23506
35000
3120
749
8000
10460
8040
0
6
5
6
5
3
6
6
6
5
1
8
2
2
1
Ϫ
4
1
. C H CH ϩ C H ϭ CH (C H ) ϩ H
This work
6
5
2
6
6
2
6
5 2
. CH (C H ) : C H ϩ H2
24
26
26
28
20
26
b
2
6
5 2
13 10
1
1
1
1
1
1
1
1
1
1
2
0. C H CH ϩ H ϭ C H CH ϩ H
1. C H CH ϩ H ϭ C H ϩ CH
3
2. CH ϩ C H CH ϭ C H CH ϩ CH
4
3. C H ϩ CH ϭ CH ϩ C H
6
4. H ϩ CH ϭ CH ϩ H
2
5. C H CH ϩ C H ϭ CH (C H )
5 2
6. CH(C H ) ϩ H ϭ CH (C H )
5 2
7. 2C H ϭ H ϭ C H
14 14
6 5 3 6 5 2 2
6
5
3
6
6
1
3
6
5
3
6
6
2
3
6
5
4
3
6
4
3
1
1
1
1
1
1
3
6
5
2
6
5
2
6
4
0
0
220
0
0
b
6
5 2
2
6
1
3
8
3
29
30
31
26
7
7
8. C H CH ϩ CH ϭ C H CH CH
3
9. 2H ϩ M ϭ H ϩ M
0. 2CH (ϩM) ϭ C H (ϩM)
6
5
2
3
6
5
2
Ϫ1.3
0.0
2
3
2
6
LOW / 1.770E ϩ 50 Ϫ9.670 6220.00/
TROE /.5325 151.00 1038.00 4970.00/
CH /2.00/C H /3.00/
4
6
6
2
1. CH ϩ H(ϩM) ϭ CH (ϩM)
1.27 ϫ 1016
Ϫ0.63
383
26
3
4
LOW / 2.477E ϩ 33 Ϫ4.760 2440.00/
TROE / .7830 74.00 2941.00 6964.00/
CH /2.00/C H /3.00/
4
6
6
a
n
6
Ϫ
1
Ϫ1
The tabulated rate constants are defined by k(T) ϭ AT exp(ϪE /RT), where E is in cal/mol and k(T) is in cm mol
s
for third-order
a
a
3
Ϫ1
Ϫ1
Ϫ1
reactions, cm mol
s
for second-order reactions and in s for first-order reactions. The equal sign in each reaction indicates that both
forward and reverse reactions are included in the modeling.
b
Reaction rate constants are kinetic estimations based on our literature review for combination reactions.
(
Ϫ6)] as well as the production of CH [reaction (Ϫ6)
The rate constants for these reactions are not known.
3
and (11)], C H CH may also inhibit the formation of
However, k has been estimated by Colket and Seery
6
5
3
9
H by the following reactions:
[24] in their modeling of the toluene decomposition.
Their estimated value is given in Table I and used in
our modeling. The rate constant for reaction (8), which
is more important than reaction (9), was obtained by
2
C H CH ϩ H ϭ C H CH
3
(5)
(6)
(7)
6
5
2
6
5
C H ϩ CH ϭ C H CH
adjusting k to match the experimental and predicted
6
5
3
6
5
3
8
concentrations of H given in Figures 1 and 2 of Ref.
2
C H ϩ C H CH ϭ C H ϩ C H CH
2
3 over the temperature range 910–1070 K assuming
the 0.1% toluene impurity in the reaction mixture. The
6
5
6
5
3
6
6
6
5
1
1
However, the inclusion of these reactions still led to a
slight underestimation of the H yield assuming that
result gave rise to the expression of k ϭ 8.35 ϫ 10
8
3
Ϫ1 Ϫ1
exp(Ϫ11830/T) cm mol
s
and are graphically pre-
2
the 0.1% of toluene impurity remained constant in the
mixture. In order to account for high concentration of
sented in Figure 3. By assuming the same amount of
C H CH impurity present in the mixture, we were
6
5
3
H yield we introduced the following two likely re-
2
able to reasonably predict the yields of H and CH as
2 4
actions:
shown in Figure 4. Similarly, the H yields measured
2
at two temperatures, 923 K and 973 K, as functions of
[C H ] could be accounted for by the model, assuming
C H CH ϩ C H ϭ CH (C H ) ϩ H
(8)
(9)
6
5
2
6
6
2
6
5 2
6
6
the 0.1% C H CH impurity (see Fig. 5). The slopes
6
5
3
CH (C H ) !: C H ϩ H2
of these plots, 1.56 and 1.61 at 923 K and 973 K,
2
6
5 2
13 10

5
80
BRIOUKOV, PARK, AND LIN
Figure 2 Yields of H with time at 967 K (᭹) and 910 K
Figure 4 The yields of H (᭺) and CH (ٗ) as functions
2 4
2
(
᭺) for a mixture of 31 Torr benzene and 296 Torr nitrogen.
of temperature. Solid curves are kinetically modeled results.
Reaction conditions: residence time ϭ 60 sec and 31 Torr
benzene and 296 Torr nitrogen with 0.1% toluene impurity.
The experimental data were taken from Figure 1 of Ref. 3.
Solid curves, modeling yields of hydrogen with 0.1% tolu-
ene impurity in the reaction mixture; dashed (967 K) and
dotted (910 K) lines are without toluene impurity.
respectively, confirm the 3/2-order global kinetics ob-
served experimentally. It should be noted that the ap-
We have also examined a potential effect of the
unimolecular decomposition of C H :
6
5
pearance of CH cannot be accounted for by the ring-
4
opening/fragmentation reactions of C H which are of
C H !: o-C H ϩ H
6 5 6 4
6
6
negligible importance at temperature near 1000 K. In-
terestingly, the 0.1% of toluene impurity is what is
which has been predicted theoretically to be the low-
est-energy decomposition path (with no reverse bar-
rier), instead of the commonly assumed ring-opening
reaction producing an open-chain l-C H ; the activa-
required for the enhanced yield of H and the yield of
2
CH measured in the experiment.
4
6
5
Figure 3 Arrhenius plot of the rate constant for the
C H CH ϩ C H ϭ CH (C H ) ϩ H reaction (k ). The rates
Figure 5 Plot of log [H ] against log [C H ] measured at
6
5
2
6
6
2
6
5 2
8
2
6
6
were obtained by adjusting k values to match the experi-
240-sec residence time. Solid lines are modeling results. To-
tal pressure of benzene and nitrogen with 0.1% toluene im-
purity is 300 Torr.
8
mental and predicted concentrations of H in Figures 1 and
2
2
of Ref. 3.

EFFECTS OF TOLUENE IMPURITY
581
[3] could be quantitatively accounted for with the four
key elementary reactions. However, the presence of
the CH product and the enhanced yields of H above
4
2
the predicted values could only be explained by in-
voking the reaction of the 0.1% toluene impurity in
the reaction mixture reported by the original authors.
This observation suggests that future studies of
benzene reaction kinetics should use ultrapure C H6
6
samples so as to avoid misleading results.
The authors are grateful for the support of this work from
the Basic Energy Sciences, Department of Energy, under
contract no. DE-FG02-97-ER14784.
BIBLIOGRAPHY
1
2
3
. Glassman, I. Combustion, 2nd ed.; Academic Press:
New York, 1986.
. Brockhorn, H. Soot Formation in Combustion;
Springer-Verlag: New York, 1993.
. Brooks, C. T.; Peacock, S. J.; Reuben, B. G. J Chem
Soc, Faraday Trans I 1979, 75(3), 652.
Figure 6 Plots of sensitivity coefficients for H (a) and
2
CH (b) were obtained at T ϭ 1016 K, 30.673 Torr benzene,
4
0
.327 Torr toluene and 296 Torr nitrogen.
4
5
6
. Slysh, R. S.; Kinney, C. R. J Phys Chem 1961, 65, 1044.
. Hou, K. C.; Palmer, H. N. J Phys Chem 1965, 69, 863.
. Laskin, A.; Lifshitz, A. 26th Symp. (Int.) on Combus-
tion, The Combustion Institute, 1996, 669.
. Asaba, T.; Fujii, N. 3rd Symp. (Int.) on Combustion,
The Combustion Institute, 1971, 155.
tion energy of the previous process was calculated to
be 78 kcal/mol [25] near 1000 K. However, this sec-
ondary reaction was found to be unimportant in the
present system.
7
8
. Rao, V. S.; Skinner, G. B. J Phys Chem 1984, 88, 5990.
SENSITIVITY ANALYSIS
9. Kielfer, J. H.; Mizerka, L. J.; Patel, M. R.; Wei, H.-C.
J Phys Chem 1985, 89, 2013.
1
1
0. Rao, V. S.; Skinner, G. B. J Phys Chem 1984, 88, 5990.
1. Hsu, D. S. Y.; Lin, C. Y.; Lin, M. C. 20th Symp. (Int.)
on Combustion, The Combustion Institute, 1984, 623.
2. Mebel, A. M.; Lin, M. C.; Yu, T.; Morokuma, K. J Phys
Chem A 1997, 101, 3189.
We have performed sensitivity analyses for the pro-
duction of H and CH ; the results of the analyses at
2
4
1
016 K are presented in Figure 6. Sensitivity coeffi-
1
1
cients less than 0.1 are not shown in the figure. For H2
formation shown in Figure 6(a), H-atom production
steps (2) and (8) dominate as expected. The removal
3. Park, J.; Dyakov, I. V.; Lin, M. C. J Phys Chem A 1997,
1
01, 8839.
of H by H ϩ C H CH (5), negatively affects the
6
5
2
14. Fahr, A.; Stein, S. E. 22nd Symp. (Int.) on Combustion,
The Combustion Institute, 1989, 1023.
yields of H2.
For CH , as revealed by Figure 6(b), reactions (8)
15. Manion, J. A.; Tsang, W. Eastern States Section: The
Combustion Institute Fall Technical Meeting on Chem-
ical and Physical Processes in Combustion, Dec. 1996.
4
and (11) have strong positive effects, whereas reac-
tions (5) and (10) negatively affect the yield of CH4
as can be expected.
1
1
6. Park, J.; Lin, M. C. J Phys Chem A 1997, 101, 14.
7. Horn, C.; Frank, P.; Tranter, R. S.; Schangg, J.; Gro-
theer, H.-H.; Just, Th. 26th Symp. (Int.) on Combustion,
The Combustion Institute, 1996, 575.
CONCLUSION
1
8. Heckmann, E.; Hippler, H.; Troe, J. 26th Symp. (Int.)
on Combustion, The Combustion Institute, 1996, 543.
The experimentally observed 3/2 kinetics for the de-
hydrogenation of C H near 1000 K by Brooks et al.
19. Park, J.; Chakraborty, D.; Bhusari, D.; Lin, M. D. J Phys
Chem in press.
6
6

5
82
BRIOUKOV, PARK, AND LIN
2
0. Tokmakov, I. V.; Park, J.; Gheyas, S. I.; Lin, M. D. J
Phys Chem in press.
1. Kee, R. J.; Rupley, F. M.; Miller, J. A. CHEMKIN-II:
A FORTRAN Chemical Kinetics Package, for the Anal-
ysis of Gas-Phase Chemical Kinetics, Sandia National
Laboratory Report No. SAND 89-8009, 1988.
2. Lutz, A. E.; Kee, R. J.; Miller, J. A. SENKIN: A FOR-
TRAN Program for Predicting Homogeneous Gas-
Phase Chemical Kinetics with Sensitivity Analysis,
Sandia National Laboratory Report No. SAND 87-8009,
26. Baulch, D. L.; Cobos, C. J.; Cox, R. A.; Esser, C.;
Frank, P.; Just, Th.; Kerr, J. A.; Pilling, M. J.; Troe, J.;
Walker, R. W.; Warnatz, J. J Phys Chem Ref Data 1992,
21, 411.
27. Tsang, W.; Kiefer, J. H. In The Chemical Dynamics and
Kinetics of Small Radicals, Pert, I; Liu, K.; Wager, A.
Eds.; Chapter 3, 1995; pp 58–119.
28. Dunlop, A. N.; Kominar, R. J.; Price, S. J. W. Can J
Chem 1970, 48, 1269.
29. Muller-Markgraf, W.; Troe, J. J Phys Chem 1988, 92,
4899.
2
2
1
989.
2
2
3. Melius, C. F. http://herzherg.sandia.gov/ϳmelius.
4. Colket, M. B.; Seery, D. J. 25th Symp. (Int.) on Com-
bustion, The Combustion Institute, 1994, 883.
5. Madden, L. K.; Moskaleva, L. V.; Kristyan, S.; Lin, M.
C. J Phys Chem A 1997, 101, 6790.
30. Braud, U.; Hippler, H.; Lindermann, L.; Troe, J. J Phys
Chem 1990, 94, 6305.
31. Cohen, N.; Westberg, K. R. J Phys Chem Ref Data
1983, 12, 531.
2