Detailed kinetics of cyclopentadiene decomposition studied in a shock tube

Article abstract of DOI:10.1002/(SICI)1097-4601(1997)29:7<505::AID-KIN4>3.0.CO;2-Y

Mixtures of cyclopentadiene diluted with argon were used to investigate its decomposition pattern in a single pulse shock tube. The temperatures ranged from 1080 to 1550 K and pressures behind the shock were between 1.7-9.6 atm. The cyclopentadiene concentrations ranged from 0.5 to 2%. Gas-chromatographic analysis was used to determine the product distribution The main products in order of abundance were acetylene, ethylene, methane, allene, propyne, butadiene, propylene, and benzene. The decomposition of cyclopentadiene was simulated with a kinetic scheme containing 44 species and 144 elementary reactions. This was later reduced to only 36 reactions The ring opening process of the cyclopentadienyl radical was found to be the crucial step in the mechanism. 1997 lohn Wiley and Sons, Inc.

Full text of DOI:10.1002/(SICI)1097-4601(1997)29:7<505::AID-KIN4>3.0.CO;2-Y

Detailed Kinetics of
Cyclopentadiene
Decomposition Studied
in a Shock Tube
ALEXANDER BURCAT, MICHAEL DVINYANINOV
Faculty of Aerospace Engineering, Technion—Israel Institute of Technology, Haifa 32000, Israel
Received 19 September 1995; revised 26 April 1996; accepted 17 October 1996
ABSTRACT: Mixtures of cyclopentadiene diluted with argon were used to investigate its de-
composition pattern in a single pulse shock tube. The temperatures ranged from 1080 to 1550
K and pressures behind the shock were between 1.7–9.6 atm. The cyclopentadiene concen-
trations ranged from 0.5 to 2%. Gas-chromatographic analysis was used to determine the
product distribution. The main products in order of abundance were acetylene, ethylene,
methane, allene, propyne, butadiene, propylene, and benzene. The decomposition of cy-
clopentadiene was simulated with a kinetic scheme containing 44 species and 144 elemen-
tary reactions. This was later reduced to only 36 reactions. The ring opening process of the cy-
clopentadienyl radical was found to be the crucial step in the mechanism. © 1997 John Wiley
and Sons, Inc. Int J Chem Kinet 29: 505–514, 1997.
performed a shock tube study, and found for the
reaction c-C5H6 L H ϩ c-C5H5 ;
k ϭ 2 ϫ 1015
INTRODUCTION
exp(Ϫ81000/RT ) sϪ1, for the temperature range
1100–2000 K and pressure range of 10–13 atm. But-
ler [2] and Bruisma et al. [3] performed experiments
in a flow reactor. According to Butler [2], the cy-
clopentadienyl radical undergoes ring cleavage and
there is a transition to straight chain aliphatic chem-
istry which is well understood. Butler [2] performed
only a qualitative survey but Bruisma et al. [3] inves-
tigated the flame speed and product distribution. Gey
et al. [4] published an ab-initio calculation with some
limited qualitative experiments performed in a regular
oven.
Cyclopentadiene was proposed as an important inter-
mediate in the process of simple aromatic hydrocar-
bon degradation.
The mechanism by which the cyclopentadienyl
radical reacts has not been studied in detail. It is
widely known that the decomposition of a fuel mole-
cule is the first step in many oxidative reactions.
Therefore decomposition studies were done in con-
junction with the cyclopentadiene oxidation [1,2].
There are four studies of cyclopentadiene de-
composition that we are aware of. Colket [1]
The formation of the cyclopentadienyl radical and
its consumption involves a transition from cyclic to
open chain kinetics. Thus understanding the role of
the cyclopentadienyl radical, is crucial to linking
these two chemical regimes which have been until
now connected only by global mechanisms.
Correspondence to: A. Burcat
Contract grant Sponsor: G.I.F. German Israeli Foundation for
Scientific Research and Development
Contract grant Sponsor: Ministry of absorption of Israel
Contract grant Sponsor: Technion Vice-President for Research
fund
© 1997 John Wiley & Sons, Inc.
CCC 0538-8066/97/070505-10

506
BURCAT AND DVINYANINOV
Table I Mixtures of CPD and Argon used for Pyrolysis Experiments
Series
Symbol
[CPD] %
[Ar] %
p1 ca. torr
P5 atm.
No. of Runs
1
2
3
4
᭝
᭪
᭞
ᮀ
1.01
0.50
2.01
0.70
98.99
99.50
97.99
99.30
75
75
75
1.8–3.0
1.7–2.5
2.1–3.2
5.6–9.6
58
45
32
27
225
which the temperature is calculated. A third piezo-
electric gauge, located on the end plate, records the
pressure. The pressure-time output is fed to a Nicolet
dual trace digital oscilloscope. Two 12 bit traces of
4098 points each, are recorded at 1 ␮s intervals. One
trace is connected to the end plate transducer and
records the experimental dwell time, while the other
is connected to the other two transducers in parallel
and records the shock speed. The temperature uncer-
tainty error is Ϯ20 K.
EXPERIMENTAL
Apparatus
The pyrolysis was studied in a 54 mm i.d. pressurized
driver, single pulse shock tube. This instrument
serves as a heater in the millisecond range where the
walls remain at room temperature. The shock tube is
made of stainless steel, and is 4 m long. The driven
section is 2.5 m long. Mylar diaphragms of different
gauges were burst to generate the shock wave. The
driver gas used was helium. Typical reaction times
were 2 milliseconds. At the end of the reaction time
an aerodynamic cooling of ca. 106 K/s is achieved.
The sampling section, 0.25 m long, contains three
Kistler 603A piezo-electric transducers. Two trans-
ducers, 0.2 m apart, measure the shock velocity from
Analysis and Mixture Handling
Post-shock gas mixtures are extracted from the test
section of the shock tube in preevacuated glass bulbs.
The products of the reaction consist of the stable
Figure 1 log k ϭ Ϫln([C5H6]t /([C5H6]0)/t vs. 104/T K, for
the four mixtures appearing in Table I. The open symbols
are the experiments. The closed symbols are the simulation
values for the same mixtures. The calculations with the re-
duced mechanism, coincide with full mechanism. The acti-
vation energies obtained for the experimental and simulated
points were 50 and 57.0 kcal/mole.
Figure 2 Results of gas chromatographic analysis for
acetylene on a logarithmic scale of ([C2H2]t /[C5H6]t) vs.
104/T K for the four mixtures in Table I. The black symbols
represent the simulated values obtained for the same mix-
ture. The activation energy obtained was 74 and 79
kcal/mole for the experimental and simulated points, re-
spectively.

CYCLOPENTADIENE DECOMPOSITION
507
Calculations
The reflected shock temperatures were calculated us-
ing standard conservation equations and the ideal gas
equation of state, assuming frozen chemistry. Ther-
modynamic data for all the species were taken from a
recent compilation [5].
RESULTS
162 runs were performed with mixtures of cyclopen-
tadiene diluted in argon. The 4 different mixtures are
listed in Table I.
The first-order rate constant of the overall decom-
position of CPD is shown in Figure 1 as log k vs. 1/T.
k is given by:
Ϫln([C5H6]t /[C5H6]0)
k ϭ
sϪ1.
t
Figures 2–9 show the concentrations of the products
relative to the final concentrations of cyclopentadiene
plotted on a logarithmic scale vs. 1/T.
Figure 3 Results of gas chromatographic analysis for eth-
ylene on a logarithmic scale of ([C2H4]t /[C5H6]t) vs. 104/T
K for the four mixtures in Table I. The black squares repre-
sent the simulated values obtained for the same mixture.
The activation energy was 58. and 73. kcal/mole for the ex-
perimental and simulated points, respectively.
log([product]t /[C5H6]t) vs. 1/T.
species found after the aerodynamic cooling. The re-
acted gas is injected into a HP 8900 series gas-chro-
matograph equipped with a flame ionization detector
and containing a 6 feet long Porapak N column.
The output of the ionization detector is transferred
to a Perkin-Elmer/Nelson Analytical PC Integrator.
The unit, performs on-line area calculations in com-
parison with standard samples.
Cyclopentadiene (CPD) tends to dimerise
[(C5H6)2]. The di-cyclo-pentadiene used was Riedel-
de-Haën AG listed as 95% pure. To obtain the
monomer, the dimer was heated in a flask under con-
stant pumping to get rid of lighter weight impurities,
and the vapors were distilled through a preheated
oven at 483ЊC directly into a preevacuated stainless
steel cylinder. The cylinder was then pressurized with
argon to 50 psia using Herzliya 99% pure argon. To
ensure thorough mixing of the gases, the mixtures
were left for 48 h before use. The concentration of
CPD in the cylinders did not change during a period
of 6 weeks. Gas chromatographic analyses showed
that the fuel in the mixture consisted of more than
99% C5H6 , ca. 0.2–0.4% CH4 (originating in the ar-
gon), less than 0.1% butadiene and less than 0.1% of
an unknown higher weight compound, probably a C6 .
Figure 4 Results of gas chromatographic analysis for
benzene on a logarithmic scale of ([C6H6]t /[C5H6]t) vs.
104/T K for the four mixtures in Table I. The black squares
represent the simulated values obtained for the same mix-
ture. The activation energy was 50. and 110. kcal/mole for
the experimental and simulated points, respectively.

508
BURCAT AND DVINYANINOV
The total carbon balance (including CPD) is
90–95% up to 1240 K, then it drops to 85% at 1300
K and from there it drops rapidly to 60% at 1370 K.
At these temperatures soot is formed, but was not an-
alyzed quantitatively.
COMPUTATIONS
Simulation of cyclopentadiene decomposition was
performed with the CHEMKIN [6] kinetic program.
The program input included forward reaction rates
and thermodynamic polynomials for all the partici-
pating species in addition to temperature, pressure,
and concentrations of the reactants. The program cal-
culates rate constant of the back reaction for every re-
action given, through the equilibrium constant using
the thermodynamic functions of the species involved.
The computed time is equal to the experimental con-
stant temperature dwell time. No calculations during
the cooling period were performed.
A kinetic scheme containing 144 reactions was
proposed. The reactions were gathered from well es-
tablished kinetic schemes. Those reaction rates have
Figure 5 Results of gas chromatographic analysis for al-
lene on a logarithmic scale of ([C3H4]t /[C5H6]t) vs. 104/T K
for the four mixtures in Table I. The black symbols repre-
sent the simulated values obtained for the same mixture.
The activation energy was 56.0 and 45. kcal/mole for the
experimental and simulated points, respectively.
Results are shown for acetylene, ethylene, benzene,
allene, propyne, propylene, methane, and 1,3-butadi-
ene. An additional unidentified C4 specie was also
recorded in quantities similar to butadiene. The data
from the four mixtures (Table I) are drawn with the
symbols shown in the table. The filled symbols repre-
sent kinetic calculations performed for mixtures of
the same composition and pressure. These computa-
tions are discussed further on.
The H/C ratio of CPD is 1.2 while the products re-
ported except C2H2 and C6H6 have a higher ratio. In
order to verify the H/C ratio balance all the concen-
trations of the products except CPD were summed up
according to the carbon content and again according
to the hydrogen content. Thus C2H4 was considered
first as 2/5 [C2H4] for the carbon sum and 4/6 [C2H4]
for the hydrogen content. This was repeated for all
the products. Summing up all products except CPD
we got the sum of [C]/6 and the sum of [H]/6 for
each experiment. These were plotted in Figure 10 one
against the other on a logarithmic scale. A perfect re-
sult would be a 45 degree line, and our result is ap-
proximately correct with a 10–15% carbon defi-
ciency at low temperatures. The reason for this may
be inaccuracies in the gas chromatographic calibra-
tion.
Figure 6 Results of gas chromatographic analysis for
propyne on a logarithmic scale of ([H4C3]t /[C5H6]t) vs.
104/T K for the four mixtures in Table I. The black symbols
represent the simulated values obtained for the same mix-
ture. The activation energy obtained was 70 and 75
kcal/mole for the experimental and simulated points, re-
spectively.

CYCLOPENTADIENE DECOMPOSITION
509
the cyclopentadiene at time t after the elimination of
a reaction was compared to the final concentration of
cyclopentadiene in the standard 144 step scheme.
The most sensitive reactions were gathered to a re-
duced scheme. This reduction method was defined by
Frenklach as “brut-force third degree reduced
scheme” [18]. The reduced scheme included 36 reac-
tions only, and showed results identical with the full
scheme. All the species recorded experimentally
could be reproduced with the reduced mechanism.
The sensitivity analysis was performed for total
cyclopentadiene decomposition and the individual
products acetylene, ethylene, benzene, allene,
propyne, propylene, and butadiene, and the results are
summarized in Tables II and III.
DISCUSSION
The modeling of a scheme containing 144 reactions
creates a problem of understanding the details in such
a large mechanism. The elimination of relatively
unimportant reactions is therefore essential.
Figure 7 Results of gas-chromatographic analysis for
propylene on a logarithmic scale of ([C3H6]t /[C5H6]t) vs.
104/T K for the four mixtures in Table I. The black symbols
represent the simulated values obtained for the same mix-
ture. The activation energy obtained was 41.0 and 38.
kcal/mole for the experimental and simulated points, re-
spectively.
The reduced mechanism of 36 reactions is shown
been verified by many kineticists, so their values are
well accepted [7,8,8(a)]. Additionally, specific reac-
tions were added from other sources. The first 37 re-
actions were taken from the scheme for methane oxi-
dation GRI-Mech developed for C1 —C2 species [7].
14 reactions were taken from a study by Butler [9].
The C3 —C4 model containing 54 reactions was
added to this scheme from Westbrook et al. [8,8(a)].
Westbrook does not identify between allene and
propyne. In this study this species was considered to
be propyne (CH3 —C#CH) and was identified in
the computer scheme as H4C3 . In addition 9 reactions
were added from Miller and Melius [10], and two re-
actions were taken from Frenklach et al. [11]. Finally,
10 reactions were taken from Dean [12], 3 reactions
were taken from Kern et al. [13,14], and one reaction
from Tsang and Hampton [15] and Warnatz [16]. The
thermodynamic data for the CHEMKIN program
were taken from a recent compilation [5] and the cy-
clopentadiene and derived radicals were taken from a
recent study [17].
Figure 8 Results of gas chromatographic analysis for 1,3-
butadiene on a logarithmic scale of ([C4H6]t /[C5H6]t) vs.
104/T K for the four mixtures in Table I. The black symbols
represent the simulated values obtained for the same mix-
ture. The black rhombuses represent the values obtained for
the reduced mechanism. The activation energy obtained
was 47 and 69 kcal/mole for the experimental and simu-
lated points, respectively.
A sensitivity analysis was performed in which the
scheme was tested by the elimination of each of the
reactions consecutively. The final concentration of

510
BURCAT AND DVINYANINOV
Table II and for the production of acetylene, ethylene,
benzene, allene, propyne, propylene, and butadiene at
the low and high temperature range are listed in Table
III.
The numbers appearing in the third and fifth col-
umn of Table II are the ratio of the decomposition of
cyclopentadiene in the absence [C]det and the pres-
ence [C]stan of the respective reaction of the 144 step
mechanism source. The numbers appearing in the
third and fifth column of Table III are the ratio of the
concentration at time t of the species in the absence
[C]det and the presence [C]stan of the respective reac-
tion in the standard 144 step mechanism source.
As expected the first reactions are the decomposi-
tion of cyclopentadiene, either by spontaneous loss of
a hydrogen atom or by a hydrogen atom attack
C5H6 EF C5H5 ϩ H
(1)
(2)
C5H6 ϩ H EF C5H5 ϩ H2
The sensitivity analysis confirms that these two reac-
tions are the most sensitive and the fifth most sensi-
tive reaction at the low temperature end (1116 K),
and as number one and four, respectively, at the
high temperature end (1409 K), when the cyclopenta-
diene decomposition is tested. The 0 number for
Reaction (1) means that the scheme cannot be calcu-
lated without this reaction. The rate constant for reac-
tion (1) that fits best our study is k ϭ 5 ϫ 1015
exp(Ϫ78682/RT ) sϪ1. But in the case of the oxidation
of cyclopentadiene [19] this value fits well the
low temperature end while at the higher temperature
end, around 2000 K, Dean’s [12] value k ϭ
Figure 9 Results of gas-chromatographic analysis for
methane on a logarithmic scale of ([CH4]t /[C5H6]t) vs.
104/T K for the four mixtures in Table I. The black symbols
represent the simulated values obtained for the same mix-
ture. The black rhombuses represent the values obtained for
a mechanism containing 0.5% of the total fuel. The activa-
tion energy obtained was 70 and 113 kcal/mole for the ex-
perimental and simulated points, respectively.
in Table IV. The results of the computer modeling us-
ing the full and reduced mechanism are shown in Fig-
ures 1–9 by filled symbols where the main decompo-
sition products are modeled. The results of the full
scheme are shown with filled circles and those of the
reduced mechanism with filled rhombuses. In most
cases both symbols fall one on the top of the other.
It can be seen that in Figures 1–9 the agreement
between the experiment and computation is very
good with the exception of methane. Our experiments
show excessive methane over the predicted mecha-
nism. In Figure 9 the presence of 0.5% CH4 of the
fuel concentration is plotted using black rhombuses.
It points out that this may be the cause of the existing
discrepancy.
In Table V a list of the experimental activation en-
ergies for the different products is compared with the
activation energy of the calculated points as they ap-
pear in Figures 1–9. The comparison is relatively
good to excellent except for benzene and methane.
A list of the most sensitive reactions in the de-
composition scheme of cyclopentadiene is shown in
Figure 10 The sum of [H]/6 vs. the sum of [C]/5 for all
the products except CPD in the four different mixtures from
Table I. A perfect agreement should be a 45° line.

CYCLOPENTADIENE DECOMPOSITION
511
Table II Results of Analysis Listing the Most Sensitive Reactions Regarding the
Cyclopentadiene Decomposition
T ϭ 1116 K
T ϭ 1409 K
[C]0 Ϫ [C]det
[C]0 Ϫ [C]det
[C]0 Ϫ [C]stan
Reaction #
Reaction
[C]0 Ϫ [C]stan
#
1
3
4
7
2
0.0
1
3
4
2
5
0.0
0.727
0.727
0.727
0.818
0.909
0.909
0.731
0.731
0.951
0.967
1.041
11
8
33
Table III Results of Analysis Listing the Most Sensitive Reactions for the Formation of
Cyclopentadiene Decomposition Products
T ϭ 1116 K
T ϭ 1409 K
Species
Reaction #
[C]det /[C]stan
Reaction #
[C]det /[C]stan
C2H2 formation
1
3, 4
7
2
0.0
1
3, 4
5
33
1
3, 4
12, 13
5
2
11
1
5
3
4
2
12
1
11
2
18
33
3, 4
5
0.0
0.262
0.624
1.128
0.0
3.18*10Ϫ3
0.323
0.395
1.443
2.366
0.0
0.241
0.753
1.17
C2H4 formation
C6H6 formation
1
0.0
12, 13
3, 4
7
8
2
1
3
5
9
2
7
1
7
0.192
0.297
0.592
1.586
1.876
0.0
1.15*10Ϫ3
0.0039
0.316
0.317
2.027
0.0
0.231
0.251
0.701
1.24
1.78*10Ϫ3
0.11
0.11
0.792
1.125
0.0
0.477
0.817
0.829
1.74
1.853
9.541
0.0
0.495
0.697
1.892
9.9
C3H4-allene
formation
3, 4
2
11
12, 13
1.266
H4C3-propyne
formation
1
28
3,4
7
12
1
0.0
1
0.029
0.268
0.273
1.211
0.0
0.029
0.579
1.46
11
28
13
5
C3H6 formation
C4H6 formation
1
0.0
8
3, 4
18
31, 32
11
1
0.313
0.588
1.63
1.79
0.0
3, 4
12, 13
2
1.56
0.0
1
12
13
3, 4
8
1.38*10Ϫ7
3.6*10Ϫ7
0.324
1.444
2.362
12, 13
3, 4
18
2
0.0042
0.169
0.534
1.5
2
11
2.042
511

512
BURCAT AND DVINYANINOV
Table IV Reduced Kinetic Scheme for the Decomposition of Cyclopentadiene
⌬r H
(1200 K)
Reaction
kcal/mole
A
n
E
Ref.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
C5H6 IJ C5H5 ϩ H
86.3
Ϫ20.30
33.3
Ϫ40.4
Ϫ34.5
Ϫ27.54
Ϫ5.2
5.00E ϩ 15
4.00E ϩ 13
5.00E ϩ 13
5.00E ϩ 12
2.00E ϩ 12
6.00E ϩ 12
1.10E ϩ 11
1.10E ϩ 11
3.10E ϩ 11
3.11E ϩ 11
1.00E ϩ 13
5.00E ϩ 13
1.00E ϩ 12
1.10E ϩ 45
1.00E ϩ 12
5.60E ϩ 12
1.02E ϩ 58
2.95E ϩ 32
1.52E ϩ 58
3.42E ϩ 52
8.38E ϩ 30
1.00E ϩ 13
5.01E ϩ 12
5.90E ϩ 47
6.60E ϩ 08
5.00E ϩ 08
2.80E ϩ 03
2.00E ϩ 14
5.00E ϩ 08
8.88E ϩ 29
3.00E ϩ 14
1.25E ϩ 14
3.00E ϩ 07
5.00E ϩ 13
1.32E ϩ 06
3.98E ϩ 19
0
0
0
0
0
0
0
0
78682
3000
33300
0
3000
0
5500
5500
5500
5500
0
This work
[12]
This work
This work
This work
[9]
This work
This work
This work
[12]
C5H6 ϩ H IJ C5H5 ϩ H2
␺
C5H5 IJ H5C5
C5H2 ϩ C3H3 IJ H5C5
␺
C3H3 ϩ C3H3 IJ C6H5 ϩ H
C5H6 ϩ C2H3 IJ C5H5 ϩ C2H4
C5H6 ϩ C3H3 IJ C5H5 ϩ C3H4
C5H6 ϩ C3H5 IJ C5H5 ϩ C3H6
C5H6 ϩ C6H5 IJ C5H5 ϩ C6H6
C5H6 ϩ CH3 IJ C5H5 ϩ CH4
C3H5 ϩ C5H5 IJ C5H6 ϩ C3H4
C3H5 ϩ C3H5 IJ C6H10
Ϫ4.2
Ϫ26.2
Ϫ21.46
Ϫ27.07
Ϫ66.6
24.17
Ϫ48.88
Ϫ30.93
Ϫ43.06
45.92
Ϫ86.46
14.96
Ϫ55.09
Ϫ40.05
60.3
0
0
0
0
0
Ϫ9.4
0
[12]
0
This work
This work
[12]
C6H10 IJ C4H6 ϩ C2H4
C2H2 ϩ CH3 IJ C3H5
C3H5 ϩ C3H5 IJ C3H6 ϩ C3H4
24000
35410
0
[12]
[7]
[12]
[12]
[12]
[12]
[12]
[8(a)]
[8(a)]
[12]
[7]
[10]
[10]
[22]
[10]
[12]
16. H ϩ C2H2 (ϩM) IJ C2H3 (ϩM)
0
2400
60160
25730
60600
27980
12820
45000
1500
107300
10840
5000
1400
62000
5000
98200
32960
37000
5000
0
␺
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
H7C5 IJ C5H6 ϩ H
Ϫ13.1
Ϫ5.8
Ϫ13.1
Ϫ12.2
Ϫ6.2
0
C3H5 ϩ C2H2 IJ C5H6 ϩ H
␺
␺
␺
␺
H7C5 IJ C5H7
C3H5 ϩ C2H2 IJ H7C5
C3H5 ϩ C2H2 IJ C5H7
C2H2 ϩ C2H2 IJ C4H3 ϩ H
C3H6 ϩ H IJ C3H5 ϩ H2
C3H6 IJ C5H5 ϩ H
H ϩ CH4 IJ CH3 ϩ H2
H4C3 ϩ H IJ C3H3 ϩ H2
Ϫ16.48
90.37
0
Ϫ9.48
1.6
2
2.9
0
1.3
␺
Ϫ12.65
Ϫ92.98
Ϫ1.5
Ϫ15.18
92.95
23.02
34.65
8.86
Ϫ115.3
7.24
C4H3 ϩ C2H2 IJ C6H5
␺
C3H4 IJ H4C3
C3H4 ϩ H IJ C3H3 ϩ H2
2
␺
H4C3 IJ C3H3 ϩ H
Ϫ4.41
0
0
2
0
C3H7 IJ CH3 ϩ C2H4
C3H7 IJ H ϩ C3H6
C6H6 ϩ H IJ C6H5 ϩ H2
C6H5 ϩ H IJ C6H6
[16]
[16]
[10]
[10]
[15]
[8(a)]
H ϩ C2H4 IJ C2H3 ϩ H2
2.53
Ϫ1
12240
98150
C4H6 IJ CH3
104.4
2
␺
—The species H5C5 represents the radical 3pentene-lyne-5yl. H4C3 represents propyne. H7C5 is the cyclopentene radical and C5H7 is the
linear 1,3-pentadien-5yl radical. C4H6 is 1,3 butadiene.
8.13 ϫ 1024TϪ2.981exp(Ϫ78682/RT ) sϪ1 fits better
with very little influence on our computations.
The cyclopentadienyl radical undergoes a ring
opening process, and this reaction is a key reaction in
the proposed mechanism. Without this value an
agreement between the calculated and the experimen-
tal results cannot be achieved.
Three cyclopentadienyl radicals can be formed:
2,4-cyclopentadiene-1yl, 1,3-cyclopentadiene-1yl, and
1,3-cyclopentadiene-2yl. The first is the only one elec-
tronically stabilized. Despite the stabilization the dou-
ble bonds are definitely shorter than the single bonds
as shown by calculations [17,20]. Thus its symmetry
C5H5 (cyclo) EF C5H5 (linear)
(3)
This reaction is the second most important reaction at
the low temperature end, and at high temperatures.
In the ring opening of cyclopentadiene, since the
conjugated C—C bond is unlikely to break, the most
probable linear species is 3-pentene-1yne-5yl
is C2␷ (␴ ϭ 2) and not D5h (␴ ϭ 10). This radical is
easier to obtain since the H—C bond on a double
bond carbon is stronger than on a primary carbon.
Therefore only the first possibility was taken into ac-
count [20].

CYCLOPENTADIENE DECOMPOSITION
513
Table V A Comparison of Experimental and Calculated Results
Experimental
Activation Energy
kcal/mole
Calculated
Activation Energy
kcal/mole
Species
Figure
Cyclopentadiene
Acetylene
Ethylene
Benzene
Allene
50
74
58
50
56
70
41
47
70
57
79
73
110
45
75
38
69
113
1
2
3
4
5
6
7
8
9
Propyne
Propylene
1,3-Butadiene
Methane
HC#C—CH"CH—CH2• although a hydrogen
atom must perform a two carbon displacement (see
Fig. 11). The thermochemistry of this linear radical
was estimated [5] with the help of the NIST 1994
Structures and Properties program [21]. The heat of
formation of the linear radical was approximated as
⌬f H298 ϭ 97.1 Ϯ 2.0 kcal.
C5H6 ϩ C3H5 EF C5H5 ϩ C3H6
(8)
The reaction of the cyclopentadienyl radical with
C3H5 to form cyclopentadiene and allene is also of
high importance.
C5H5 ϩ C3H5 EF C5H6 ϩ C3H4
(11)
The linear C5H5 radical can break into acetylene
and propargyl radical (in this scheme by the reverse
reaction):
Reactions 12 and 13 are important at low as well as at
high temperatures for the formation of ethylene, al-
lene, propyne, and propylene.
C2H2 ϩ C3H3 EF C5H5 (linear)
(4)
C3H5 ϩ C3H5 EF C6H10
C6H10 EF C4H6 ϩ C2H4
(12)
(13)
This reaction is the third most important reaction at
the low temperature end as well as at the high tem-
perature end. It is the third most important for the for-
mation of acetylene and propylene, low temperature
propyne, and high temperature ethylene and butadi-
ene, as well as number 4 for low temperature allene
and high temperature benzene.
Finally the isomerisation of allene 4 propyne is very
The propargyl radical is a precursor for the pro-
duction of benzene or phenyl radicals:
C3H3 ϩ C3H3 EF C6H5 ϩ H
(5)
This reaction plays an important role at high tempera-
tures but not at low temperatures except for benzene.
This reaction reported by Miller and Melius [10] had
to be slightly changed to fit our experimental values.
We reduced its preexponential factor and slightly in-
creased its activation energy.
The reactions of cyclopentadiene with vinyl,
propargyl, allyl, phenyl, and methyl radicals are the
next important steps. Specifically reaction (7) which
is the fourth in importance for the CPD decomposi-
tion acetylene and propyne formation, second for al-
lene formation, etc. but only at low temperatures.
C5H6 ϩ C3H3 EF C5H5 ϩ C3H4
(7)
Figure 11 The three possible cyclopentadienyl radicals
and the ring opening path.
Reaction (8) follows it in some cases

514
BURCAT AND DVINYANINOV
Phase Chemical Kinetics. Sandia Report SAND 89-
8009B. U.C-706, 1992.
important, and the rate constant of this reaction that
fits best is the one found experimentally [22].
7. R. V. Serkauskas, M. Frenklach, H. Wang, C. T. Bow-
man, G. P. Smith, D. M. Golden, W. C. Gardiner, and
V. Lissianski, An Optimized Detailed Chemical Mecha-
nism for Methane Combustion, Gas Research Institute
Topical Report 1994.
8. C. K. Westbrook, W. J. Pitz, M. M. Thornton, and P. C.
Malte, Comb. Flame, 72, 45 (1988); W. J. Pitz, C. K.
Westbrook, W. M. Proscia, and F. L. Dryer, A Compre-
hensive Chemical Kinetic Reaction Mechanism for the
Oxidation of N Butane, 20-th Comb. Symp., 1984, pp.
831–843.
9. J. L. Emdee, K. Brezinsky, and I. Glassman, J. Phys.
Chem., 96, 2151, (1992), F. N. Egolffopoulos, D. X.
Du, and C. K. Law, Comb. Sci. Tech., 83, 33 (1992).
10. J. A. Miller and C. F. Melius, Combust. Flame, 91, 21
(1992).
C3H4 (allene) EF C3H4 (propyne)
(28)
In conclusion, the reduced mechanism permits the
decipheration of the decomposition mechanism, and
points out the most important reaction in the process
which is the ring opening. It also reproduces almost
exactly the results obtained with the full mechanism.
The full mechanism, however, can reproduce the pro-
duction of additional species, many of which were
not identified in this study but were reported in exper-
iments using mixtures with higher concentration of
cyclopentadiene [2].
11. M. Frenklach, D. W. Clary, T. Yuan, W. C. Gardiner,
and S. E. Stein, Combust. Sci. Technol., 50, 79 (1986).
12. A.M. Dean, J. Phys. Chem., 94, 1432 (1990).
13. R. D. Kern, K. Xie, H. Chen, and J. H. Kiefer, High
Temperature Pyrolysis of Acetylene and Diacetylene
behind Reflected Shock Waves, 23 Combust Symp.,
(1990), pp. 69–75.
14. C. H. Wu, and R. D. Kern, J. Phys. Chem., 91, 6291
(1987).
15. W. Tsang and R. F. Hampson, J. Phys. Chem. Ref.
Data, 15, 1087 (1986).
This research was supported by a grant from the G.I.F. Ger-
man Israeli Foundation for Scientific Research and Devel-
opment. M. D. also acknowledges partial funding by a
grant from the Ministry of absorption of Israel, and by a
grant from the Technion Vice-President for Research fund.
The authors thank Dr. E. Olchansky and Dr. C. Sokolinsky
for computer simulation. Special thanks are extended to
Prof. Assa Lifshitz for very fruitful discussions.
16. J. Warnatz, Chemistry of High Temperature Combus-
tion of Alkanes up to Octane, 20-th Comb. Symp.,
1984, pp. 845–855.
BIBLIOGRAPHY
17. M. Karni, I. Oref, and A. Burcat, J. Phys. Chem. Ref.
Data, 20, 665 (1991).
1. M. B. Colket, The Pyrolysis of Cyclopentadiene, East-
ern States Section of Combustion Institute annual
meeting, Orlando, 1990, Paper 1.
2. R. G. Butler, A Flow Reactor Study of the Oxidation of
1,3-Cyclo-Pentadiene, MSc. Thesis, Princeton Univer-
sity, 1992.
3. O. S. L. Bruisma, P. J. J. Tromp, H. J. J. de Sauvage
Nolting, and J. A. Moulijn, Fuel, 67, 334 (1988).
4. E. Gey, B. Ondrushka, and G. Zimmermann, J. Prakt.
Chem, 329(3), 511 (1987).
18. M. Frenklach, Reduction of Chemical Reaction Models
in Numerical Approaches to Combustion Modeling,
Chap. 5, E. S. Oran and J. P. Boris, Eds., Progress in
Astronautics & Aeronautics, 1991, Vol. 135, pp.
129–154.
19. A. Burcat, M. Dvinyaninov, E. Olchanski, and Ch.
Sokolinski, Detailed Combustion Kinetics of Cyclopen-
tadiene, Combust. Flame, submitted.
20. C. F. Melius, BAC/MP4 ab-initio calculation database,
Sandia Co.
21. S. E. Stein, NIST Structures and Properties, Computer-
ized Database #25, Gaithersburg, Maryland, 1994.
22. A. Lifshitz, M. Frenklach, and A. Burcat, J. Phys.
Chem., 79, 1148 (1975).
5. A. Burcat and B. McBride, 1995 Thermodynamic
Database for Combustion and Air Pollution Use, Tech-
nion Aerospace Engineering (TAE) Report # 732, Jan-
uary 1995.
6. J. Kee, F. M. Rupley, and J. A. Miller, Chemkin-II A
Fortran Chemical Kinetic Package for Analysis of Gas