Experimental and modeling study of the oxidation of benzene

Article abstract of DOI:10.1002/kin.10148

This paper describes an experimental and modeling study of the oxidation of benzene. The low-temperature oxidation was studied in a continuous flow stirred tank reactor with carbon-containing products analyzed by gas chromatography. The following experimental conditions were used: 923 K, 1 atm, fuel equivalence ratios from 1.9 to 3.6, concentrations of benzene from 4 to 4.5%, and residence times ranging from 1 to 10 s corresponding to benzene conversion yields from 6 to 45%. The ignition delays of benzene-oxygen-argon mixtures with fuel equivalence ratios from 1 to 3 were measured behind shock waves. Reflected shock waves permitted to obtain temperatures from 1230 to 1970 K and pressures from 6.5 to 9.5 atm. A detailed mechanism has been proposed and allows us to reproduce satisfactorily our experimental results, as well as some data of the literature obtained in other conditions, such as in a plug flow reactor or in a laminar premixed flame. The main reaction paths have been determined for the four series of measurements by sensitivity and flux analyses.

Full text of DOI:10.1002/kin.10148

Experimental and Modeling
Study of the Oxidation
of Benzene
I. DA COSTA, R. FOURNET, F. BILLAUD, F. BATTIN-LECLERC
De´partement de Chimie Physique des Re´actions, UMR 7630 CNRS, INPL-ENSIC 1, rue Grandville, BP 451,
54001 Nancy Cedex, France
Received 16 September 2002; accepted 6 June 2003
DOI 10.1002/kin.10148
ABSTRACT: This paper describes an experimental and modeling study of the oxidation of ben-
zene. The low-temperature oxidation was studied in a continuous flow stirred tank reactor with
carbon-containing products analyzed by gas chromatography. The following experimental con-
ditions were used: 923 K, 1 atm, fuel equivalence ratios from 1.9 to 3.6, concentrations of ben-
zene from 4 to 4.5%, and residence times ranging from 1 to 10 s corresponding to benzene
conversion yields from 6 to 45%. The ignition delays of benzene–oxygen–argon mixtures with
fuel equivalence ratios from 1 to 3 were measured behind shock waves. Reflected shock waves
permitted to obtain temperatures from 1230 to 1970 K and pressures from 6.5 to 9.5 atm.
A detailed mechanism has been proposed and allows us to reproduce satisfactorily our ex-
perimental results, as well as some data of the literature obtained in other conditions, such
as in a plug flow reactor or in a laminar premixed flame. The main reaction paths have been
determined for the four series of measurements by sensitivity and flux analyses. ꢀ 2003 Wiley
C
Periodicals, Inc. Int J Chem Kinet 35: 503–524, 2003
reactions, are necessary to model the formation of the
broad spectrum of products observed.
INTRODUCTION
Fuels contain usually six main classes of or-
ganic compounds: paraffins, isoparaffins, olefins,
naphthenes, aromatics (including polyaromatic com-
pounds), and ethers. Commercial gasoline contains ap-
proximately 150 different molecules belonging to these
six classes of organic compounds. Since the end of the
last decade, a great number of experimental and theo-
reticalstudiesmadeitpossibletomodelthecombustion
of model molecules taken individually and to predict
theformationofcharacteristicpollutants. Nevertheless,
most studies concern paraffins and isoparaffins and to
a lesser extent, olefins and ethers [1]. Currently, the ex-
perimental and theoretical studies are directed toward
the two other families of organic compounds, with an
obvious interest for aromatic compounds. These last
compounds are present in significant amounts in gaso-
lines (∼35%) and diesel fuels (∼30%) [2] and are
The use of fuels in engines leads to a very significant
formation of toxics (e.g. carbon monoxide, aldehy-
des, butadiene, aromatics, soot) and air pollutants (e.g.
sulfur oxides, nitrogen oxides, unburned hydrocarbons
and oxygenates) that are responsible for acid rain and
the increase of the ozone concentration in the low at-
mosphere. The control of the influence of the formula-
tion of fuels on the formation of pollutants requires a
fundamental approach based on predictive models en-
abling us to quantify the formation of by-products dur-
ing the process of combustion. The complexity of the
chemistry of combustion is such that detailed chemical
kinetic models, involving a description of elementary
Correspondence to: F. Battin-Leclerc; e-mail: dcpr@ensic.inpl-
nancy.fr.
c
ꢀ
2003 Wiley Periodicals, Inc.

504
DA COSTA ET AL.
responsible for the formation of particles, which are
harmful for health. The reactions of benzene as the first
aromatic ring are of importance in building up polyaro-
matic hydrocarbons (PAHs), which are considered to
play a key role in the formation of soot particles.
Benzene is the smallest aromatic molecule and its
oxidation has consequently been the subject of a num-
ber of experimental studies in various types of reac-
tors (laminar premixed flames [3,4], shock tube [5],
plug flow reactors [6,7], jet-stirred reactors [8,9], and
in a flow reactor in the flue gas of a laminar flame of
methane [10]). Several kinetic mechanisms [4,7,9,11–
16] have also been proposed to model these experimen-
tal data. It is worth noting that all the experiments in re-
actors with no initial radicals pool [6–9] have been per-
formed using a low concentration of benzene (≤0.16%)
and, consequently, with no noticeable conversion for a
temperature lower than 1000 K. The double purpose of
this paper is then
the literature obtained in a flow reactor at 1100 K
[6] and in a near-sooting flat low-pressure lami-
nar flame [3].
EXPERIMENTAL STUDY OF THE
OXIDATION OF BENZENE
Experimental Procedure
This experimental study has been performed by us-
ing two different devices: a jet-stirred reactor and a
shock tube. Benzene (99.8% pure) was provided by
Aldrich. Oxygen (99.5% pure) and argon and helium
(both 99.995% pure) were supplied by Alphagaz-L’Air
Liquide.
Jet-Stirred Reactor. Figure1 presentsaschemeofthe
apparatus. Experiments were carried out in a continu-
ous jet-stirred reactor (internal volume 88 cm3) [15]
made of quartz and operated at constant temperature
(923 K) and pressure (1 atm). Residence times ranged
from 1 to 10 s. The heating of the reactor was achieved
by means of electrically insulated resistors directly
coiled around the vessel; temperature was measured
by using a thermocouple located inside the reactor in
a glass finger. To obtain a spatially homogeneous tem-
perature inside the reactor, reactants were preheated at
•
to present new experimental data for the oxida-
tion of benzene obtained both in a jet-stirred reac-
tor at a concentration of benzene from 4 to 4.5%
and a temperature of 923 K, and in a shock tube
for temperatures between 1230 and 1970 K.
•
to propose a mechanism able to model these ex-
perimental results, but also to reproduce data of
Figure 1 Scheme of the apparatus including a jet-stirred reactor.

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
505
a temperature close to the reaction temperature; corre-
sponding residence time in the preheating section was
approximately 1% of the total residence time. Temper-
ature gradients inside the reactor, when reaction was
occurring, were never larger than 5 K.
Gas-phase benzene was obtained by bubbling the
flow of carrier gas (helium) into liquid benzene main-
tained at a temperature below room temperature (here
at 280 K) to avoid the condensation of hydrocarbon
before entering the reactor. Gas flows of helium and
oxygen were measured by mass flow controllers; the
gas flow of benzene was deduced from chromatog-
raphy analyses. The benzene–helium–oxygen mix-
tures were mixed prior to their introduction into the
reactor.
hydrocarbons (CH4, C2 species, C3H6, C3H6,
and C4H6) were analyzed on a 30% Squalane on
Chromosorb P column with detection by flame
ionization detection and nitrogen as carrier gas.
The analysis of water, hydrogen, methanol, and
formaldehyde was not possible with our appara-
tuses. Calibrations were performed by analyzing
a range of samples containing known pressures
of each pure compound to quantify.
The compositions of the reacting flow were
helium/oxygen/benzene = 80.5:15.5:4 (equivalence
ratio ꢀ = 1.9) and 96.1:9.4:4.5 (ꢀ = 3.6) (mole). The
equivalenceratioꢀisdefinedasꢀ=7.5×(%benzene/
%oxygen) by using as a reference the total reaction
Becauseoftheformationofcompounds, whichwere
either gaseous or liquid at room temperature, the analy-
ses of the products leaving the reactor were performed
in two steps:
C6H6 + 7.5O2 = 6CO2 + 3H2O
To check the initial mixture compositions, a complete
cycle of analyses was performed, before each experi-
ment, withthereactantgasescirculatingintothereactor
at room temperature.
•
First, during a determined period of time, the out-
let flow was directed toward a trap maintained at
liquid nitrogen temperature. The short line be-
tween the reactor and the trap was heated up to
423 K to minimize condensation problems (this
temperature was low enough for the oxidation re-
action to be stopped). At the end of this period,
the trap was removed and, after the addition of
acetone and of an internal standard (o-xylene),
progressively heated up at room so as tempera-
ture so as to melt its solid content. The obtained
solution was then analyzed by gas chromatog-
raphy with mass spectrometer or flame ioniza-
tion detection by using a HP PONA capillary
column and helium or nitrogen as carrier gas.
In the conditions of this study, the column pro-
vided a good separation for the following prod-
ucts (with their measured retention time in min-
utes): benzene (8.9), toluene (12.1), ethyl ben-
zene (15.5), para-xylene (15.9), styrene (16.5),
benzaldehyde (18.7), phenol (19.2), benzylalco-
hol (20.5), and bibenzyl (31.1), with a limit of
detection around 10−5 in mole fraction. Calibra-
tions were performed by analyzing a range of
solutions containing known amounts of o-xylene
and of the compound to quantify.
Shock Tube. As it is already described in detail in pre-
vious papers [17–19], we will just recall here the main
features of this experimental device. The reaction (400
cm long, 7.8 cm i.d.) and the driver (90 cm long, 12.8
cm i.d.) parts of this stainless steel shock tube were sep-
arated by two terphane diaphragms, which were rup-
tured by decreasing suddenly the pressure in the space
separating them. The driver gas was helium. The inci-
dent and reflected shock velocities were measured by
four piezoelectric pressure transducers located along
the reaction section. The temperature and the pressure
of the test gas behind the incident and the reflected
shock waves were derived from the value of the in-
cident shock velocity by using ideal one-dimensional
shock equations.
The onset of ignition was detected by OH radical
emission at 306 nm through a quartz window with a
photomultiplier fitted with a monochromator at the end
of the reaction part. The last pressure transducer was
located at the same place along the axis of the tube as
the quartz window. The ignition delay time was defined
as the time interval between the pressure rise measured
by the last pressure transducer owing to the arrival of
the reflected shock wave and the rise of the optical
signalbythephotomultiplierupto10%ofitsmaximum
value.
•
In a second time, just before the removal of the
trap, the outlet flow was directed toward a wa-
ter cooler and a volumic flowmeter. Gas sam-
ples were directly obtained by expansion in an
evacuated volume. Chromatographs with ther-
malconductivitydetectorandhydrogenascarrier
gas were used to analyze O2, CO, and CO2 with
a Carbosphere packed column. Gas-phase light
Fresh reaction mixtures were manometrically pre-
pared every day and mixed using a recirculation pump.
Before each introduction of the reaction mixture (from
100 to 400 kPa), the reaction section was flushed with

506
DA COSTA ET AL.
pure argon and evacuated (below 10−2 kPa), so that the
residual gas was mainly argon.
Shock Tube. This study was performed in the follow-
ing experimental conditions (after the reflected shock):
•
•
•
Temperature (T) range from 1230 to 1970 K.
Pressure (P) range from 7.3 to 9.5 atm.
Experimental Results
Argon–benzene–oxygen mixtures (in molar per-
cent) were 1.25:18.75:80, 1.25:9.375:89.375,
1.25:6.26:92.49, 1.25:3.125:95.625, and 2.5:
18.75:78.75, respectively, which correspond to
four different equivalence ratios (ꢀ = 0.5, 1, 1.5,
and 3) and to two different concentrations of ben-
zene (1.25 and 2.5%) and permit to obtain corre-
sponding delays from 6 to 1980 ꢀs.
A discussion of these experimental results is given fur-
ther in the text with the comparisons with simulations.
Jet-Stirred Reactor. The main carbon-containing
products analyzed by gas chromatography were
phenol, carbon monoxide and dioxide, methane, C2
species (acetylene and ethylene were not separated),
propyne, propene, and 1,3-butadiene. For each
sample, the carbon balance was checked and an
agreement was obtained to within 5%. The formation
of cyclopentadiene was too low to be detected.
Table I presents the experimental results obtained
in a jet-stirred reactor. Figures 2 and 3 display the
evolution of the conversions of benzene and oxygen
and of the formations of product vs. residence time for
equivalence ratios equal to 1.9 and 3.6, respectively. A
decrease of the equivalence ratio of the mixtures en-
hances the rate of consumption of benzene, but does
not really change the repartition of the observed prod-
uct, except carbon monoxide, which is formed in higher
concentration at ꢀ equal to 1.9.
Table II and Fig. 4 present the experimental data thus
obtained.
For the five mixtures, ignition delays (τ) decrease
when T rises up and varies exponentially vs. 1000/T.
It is also shown that for a given T, ignition delays de-
crease with the equivalence ratio of the mixture for a
given concentration of benzene and with the concentra-
tion of hydrocarbon for a given equivalence ratio. The
slope of the log (τ) vs. 1000/T lines remains nearly con-
stant when the equivalence ratio varies, but decreases
with the concentration of benzene. The lower part of
Fig. 4 also presents the line obtained from the correla-
tion proposed by Burcat et al. [5], τ = 1.26 × 10−15
Table I Experimental Results Obtained in a Jet-Stirred Reactor at 923 K and 1 atm
Mole Fractions at Different Residence Times (s)
Compounds
1.6
2.2
3.3
4.4
5.5
6.6
8.8
(A) φ = 1.9; Xbenzene = 0.04
CH4
1.34 × 10−6 2.57 × 10−6 1.15 × 10−4 2.79 × 10−4 5.57 × 10−4 7.77 × 10−4
1.0 × 10−3
7.9 × 10−4
C2H2 + C2H4 1.48 × 10−6 3.39 × 10−6
1.8 × 10−4 3.16 × 10−4 5.46 × 10−4 6.74 × 10−4
C3H4
C3H6
1,3-C4H6
O2
1.98 × 10−7 3.93 × 10−7 8.11 × 10−6 7.72 × 10−6 7.78 × 10−6 8.17 × 10−6 8.74 × 10−6
2.2 × 10−7 6.24 × 10−7 1.61 × 10−5 2.46 × 10−5 3.97 × 10−5 4.26 × 10−5 4.84 × 10−5
0
0
2.97 × 10−6 8.23 × 10−6 8.43 × 10−6 1.23 × 10−5 1.55 × 10−5
0.155
0.155
0
0
0.0338
0
0.148
0.134
0.127
0.114
0.112
CO
0
1.69 × 10−2 2.94 × 10−2 4.66 × 10−2 5.58 × 10−2
6.6 × 10−2
6.3 × 10−3
0.0221
CO2
C6H6
C6H5OH
0
0.0375
0
1.68 × 10−3 5.17 × 10−3 5.61 × 10−3 8.85 × 10−3
0.0322
0.0305
0.0227
0.0232
1.03 × 10−3 1.81 × 10−3 1.69 × 10−3 1.87 × 10−3 1.72 × 10−3
(B) φ = 3.6; Xbenzene = 0.045
CH4
2.28 × 10−6 2.59 × 10−6 1.77 × 10−5 8.84 × 10−5 2.37 × 10−4 4.32 × 10−4 6.23 × 10−4
C2H2 + C2H4 9.71 × 10−7
1.6 × 10−6
2.4 × 10−5 1.42 × 10−4 3.02 × 10−4 4.79 × 10−4 6.02 × 10−4
1.62 × 10−6 4.34 × 10−6 5.32 × 10−6 5.63 × 10−6 5.92 × 10−6
3.41 × 10−6 8.66 × 10−6 1.65 × 10−5 3.46 × 10−5 3.99 × 10−5
1.99 × 10−6 5.35 × 10−6 6.79 × 10−6 1.32 × 10−5 1.77 × 10−5
C3H4
C3H6
1,3-C4H6
O2
0
0
0
0
0
0
0.0921
0.0897
0.0803
0.076
0.08
0.07
0.0629
CO
CO2
0
0
0
0
0
1.06 × 10−2 1.98 × 10−2 2.57 × 10−2 3.62 × 10−2
0
0.0397
0
1.13 × 10−3 3.22 × 10−3
2.8 × 10−3 9.28 × 10−3
C6H6
C6H5OH
0.042
0
0.0349
0.0361
0.035
0.0325
0.0287
1.4 × 10−3
2.77 × 10−5 8.61 × 10−4 1.35 × 10−3 1.56 × 10−3

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
507
Figure 2 Oxidation of benzene in a jet-stirred reactor at 923
K. Comparison between experimental (symbols) and com-
puted (lines) species vs. residence time at ꢀ = 1.9.
Figure 3 Oxidation of benzene in a jet-stirred reactor at 923
K. Comparison between experimental (symbols) and com-
puted (lines) species vs. residence time at ꢀ = 3.6.
× exp(40.6 kcal mol−1/RT) × [C6H6]0.42 × [O2]−1.70
× [Ar]0.44 s (concentrations in mol cm−3 units), based
on their experimental results, at ꢀ = 1 and at 1.25% of
benzene.
ture and are mainly those proposed by Baulch et
al. [22]andTsangandHampson[23]. TheC0 C2
reaction base was first presented in the paper of
Barbe´ et al. [21] and has been updated recently
[19]. The C0 C6 reaction base includes reac-
tions involving C3H2··, ·C3H3, C3H4 (allene and
propyne), ·C3H5 (three isomers), C3H6, C4H2,
·C4H3 (2 isomers), C4H4, ·C4H5 (5 isomers),
C4H6 (1,3-butadiene, 1,2-butadiene, methylcy-
clopropene, 1-butyne, and 2-butyne), as well
as the formation of benzene [18,19]. Pressure-
dependent rate constants follow the formalism
proposed by Troe [24] and efficiency coefficients
have been included. This reaction base was built
inordertomodelexperimentalresultsobtainedin
a jet-stirred reactor for methane and ethane [21],
profiles in laminar flames of methane, acetylene,
DESCRIPTION OF THE REACTION
MECHANISM
This mechanism, which is available on request, has
been written in the CHEMKIN II [20] format and in-
cludes three parts:
•
The C0 C6 reaction base, which is described in
recent papers [18,19]. This C0 C6 reaction base
was built from a review of the recent literature
and is an extension of our previous C0 C2 reac-
tion base [21]. This C0 C2 reaction base includes
the reactions of radicals or molecules including
carbon, hydrogen, and oxygen atoms and con-
taining less than three carbon atoms. The kinetic
data used in this base were taken from the litera-

508
DA COSTA ET AL.
Table II Mixture Compositions, Shock Conditions, and Ignition Delays for Benzene
Composition
(Mole Percent)
Composition
(Mole Percent)
V
V
P1
(atm)
P5
(atm)
T5
(K)
τ
P1
(atm)
P5
(atm)
T5
(K)
τ
(ꢀs)
C6H6
O2
(m s−1
)
(ꢀs) C6H6
O2
(m s−1
)
1.25 9.375 0.2105
0.2105
875
874
862
849
847
849
835
803
820
813
799
813
806
802
792
782
790
783
782
8.57
8.52
7.93
8.84
7.80
8.36
8.89
8.35
9.14
8.69
8.65
8.46
9.11
8.76
9.27
8.52
8.99
8.59
9.30
1579
1573
1536
1498
1491
1500
1456
1361
1411
1391
1351
1390
1372
1359
1332
1304
1326
1305
8
15
15
34
35
35
70
99
124
131
198
200
203
283
301
357
503
543
2.5
18.75 0.2237
0.2303
852
830
825
829
818
808
799
802
9.11
8.69
8.77
9.15
9.51
9.17
8.50
9.40
1362
1306
1293
1304
1277
1254
29
80
167
171
263
563
φ = 1
0.2039
0.2368
0.2105
0.2237
0.2500
0.2633
0.2704
0.2633
0.2763
0.2568
0.2829
0.2763
0.3033
0.2895
0.2961
0.2907
0.3166
φ = 1
0.2368
0.2434
0.2637
0.2632
0.2526
0.2763
1231 1167
1238
840
1.25
6.25 0.1908
0.2276
892
865
849
844
805
821
830
827
817
871
8.13
8.43
8.50
7.46
6.28
8.17
7.57
7.97
6.59
7.35
1659
1572
1523
1511
1391
1436
1468
1459
1428
1593
10
31
75
0.2237
φ = 1.5
0.2039
0.1974
0.2368
0.2171
0.2309
0.1980
0.1842
66
229
215
110
125
152
39
1303 1978
1.25 18.75 0.2251
0.2434
847
847
847
839
806
815
831
814
816
804
800
8.44
9.13
8.39
8.61
8.09
8.58
7.45
9.43
9.29
8.69
8.35
1418
1418
1418
1394
1304
1328
1372
1326
1332
1301
1290
17
23
31
42
46
60
86
111
151
168
257
1.25 3.125 0.1447
0.1513
972
951
923
7.78
7.65
7.36
1970
1893
1797
6
8.5
15
φ = 0.5
0.2237
0.2368
0.2500
0.2566
0.2105
0.2829
0.2763
0.2697
0.2632
φ = 3
0.1579
Note. P1 is the pressure of the mixture before the shock, V the speed of the incident wave, P5 and T5 are pressure and temperature behind the
reflected shock wave, and τ is the ignition delay time.
and 1,3-butadiene [18], and shock tube autoigni-
tion delays for acetylene, propyne, allene, 1,3-
butadiene [18], 1-butyne, and 2-butyne [19].
A primary mechanism containing 71 reactions, in
which only benzene and oxygen were considered
as molecular reactants.
A secondary mechanism including 64 reactions,
in which the reactants are the molecular products
formed by the primary mechanism.
mochemistry of cyclopentadienone, cyclopentadienol
and orthobenzoquinone molecules and of cyclopen-
tadienonyl, cyclopentadienoxy, and hydroxycyclopen-
tadienyl radicals, for which the calculation was not
possible by THERGAS, because some groups were
missing, polynomial coefficients have been derived
from the thermochemical data proposed by Alzueta
et al. [7]. The groups responsible for failing in apply-
ing THERGAS are CO (CO) (Cd), CO (Cd) (CO),
CO (Cd) (Cd), and C (Cd)/2 (h) (O). In the fu-
ture, these groups should be calculated by means of
thermodynamic data found in the literature [7] and in-
cluded in the software. It is worth noting that, in the
distributed version of THERGAS, the introduction of
new groups required to modify some data files, which
are not accessible by the user. Table III presents the
heats of formation used for aromatic species. It must
•
•
When it was possible, specific heats, heats of for-
mation, and entropies of the considered molecules
or radicals have been calculated using the soft-
ware THERGAS [25], based on the group and bond
additivity methods proposed by Benson [26], and
stored as 14 polynomial coefficients, according to the
CHEMKIN II formalism [20]. In the case of the ther-

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
509
The primary and secondary mechanisms for the ox-
idation of benzene are presented in Table IV, as well
as the references related to the used kinetic data. These
mechanisms, which involve 24 species not included in
the C0 C6 reaction base, have been built in order to be
more comprehensive than those proposed by Emdee
et al. [12] and Alzueta et al. [7] to model experimen-
tal results around 1000 K, e.g., termination steps of
resonance-stabilized radicals have been considered in
more details.
Primary Mechanism
The primary mechanism includes the reactions of
benzene molecules and of cyclohexadienyl, phenyl,
phenylperoxy, phenoxy, hydroxyphenoxy, cyclopenta-
dienyl, cyclopentadienoxy, and hydroxycyclopentadi-
enyl free radicals.
Reactions of benzene molecules involve the forma-
tion of phenyl radicals by unimolecular (1) (reaction
numbers are those of Table IV) and bimolecular (2) ini-
tiationsandbymetatheseswithH-abstractionby·H(8),
·O· (9), ·OH (10), ·HO2 (11), ·CH3 (12), ·C2H5 (13),
·a-C3H5 (14), ·n-C4H5 (14), and ·i-C4H5 (16) radicals.
Addition reactions are also considered to form ·C6H7
radicals (addition of ·H atoms (3)), phenoxy radicals
(addition of ·O· atoms (4)), phenol (addition of ·OH
radicals (5)), phenylacetylene (addition of ·C2H radi-
cals (6)), and styrene (addition of ·C2H3 radicals (7)).
Thedecompositionofbenzenemoleculetogivepropar-
gyl radicals is already included in the C0 C6 reaction
base through its reverse reaction. The kinetic value and
the products of the combination of propargyl radicals
is still a subject of discussion [51]. In the C0 C6 reac-
tion base, we have considered the formation of benzene
with a rate constant k = 1 × 1012 cm3 s−1 mol−1 close
to the value estimated by Stein et al. [52] (3 × 1012 cm3
Figure 4 Autoignition delays of benzene in a shock tube.
Semilog plot of experimental (symbols) and computed (lines)
ignition delays as a function of temperature behind the re-
flected shock wave for different equivalence ratios at 1.25%
of benzene (a) and for different benzene concentrations at ꢀ
= 1 (b). The correlation proposed by Burcat et al. [5] at ꢀ =
1 and at 1.25% of benzene is shown in (b).
be kept in mind that the precision obtained by using
group additivity methods to estimate heats of formation
is around 2 kcal mol−1 for molecules and 4 kcal mol−1
for radicals [26].
Table III Heats of Formation for Aromatic Species at
298 K in kcal mol−1
s
−1 mol−1). Recent measurements [51] proposed a rate
between 4.5 × 1012 and 9 × 1012cm3s−1mol−1 for the
formation of linear and cyclic C6H6 and found that the
contribution of the formation of phenyl radicals and H·
atoms was less than 10%.
Species
ꢁHf (298 K)
C6H6
C6H5OH
C5H6
19.8
−23.1
32
Reactions of resonance-stabilized cyclohexadienyl
radicals include their linearization (17) as proposed by
Weissman and Benson [32], their oxidation with O2
(18) as usual allylic radicals, and their termination reac-
tions with other radicals (20–23). Their decomposition
by beta-scission to produce acetylene and ·n-C4H5 rad-
icals is already considered in the C0 C6 reaction base
(written in the reverse direction).
The linearization of phenyl radicals (24) has been
considered as proposed by Braun-Unkoff et al. [33].
The rate constants of the decomposition reactions of the
linear ·C6H5 isomer (25–26) have been reevaluated in
C5H5OH∗
C5H4O∗
C6H5OO
C6H5
−9
13.2
31
81
12.8
61
15.9
67
−25
C6H5O
C5H5
C5H4OH∗
C5H3O∗
OC6H4O∗
Note. The heats of formation have been calculated by the software
THERGAS [25], except for the compounds distinguished by “∗”, for
which they have been proposed by Alzueta et al. [7].

510
DA COSTA ET AL.
Table IV Primary and Secondary Mechanisms for the Oxidation of Benzene
Reactions
A
n
Ea
References
Wang [27]
Reaction No.
PRIMARY MECHANISM
Reactions of benzene molecules
Umimolecular initiation
C6H5 + H(+M) ꢀ C6H6(+M)
k
k0
1.0 × 1014
6.6 × 1075
0.0
−16.3
0.0
7.0
(1)
(2)
∞
Troe’s factors: 1.0, 0.1, 585, 6113
Efficiency factors: O2, 0.4; CO, 0.75; CO2, 1.5; H2O, 6.5; N2, .4; Ar, 0.35; He, 0.35; C6H6, 3.0
Bimolecular initiation
C6H6 + O2 ꢀ C6H5 + HO2
6.0 × 1013
0.0
63.4
Alzueta [7]
Additions
C6H6 + H ꢀ C6H7
3.2 × 1013
2.8 × 1013
1.3 × 1013
5.0 × 1013
7.9 × 1011
0.0
0.0
0.0
0.0
0.0
3.2
4.9
10.6
0.0
Mebel [28]
Nicovich [29]
Baulch [22]
Wang [27]
(3)
(4)
(5)
(6)
(7)
C6H6 + O ꢀ C6H5O + H
C6H6 + OH × C6H5OH + H
C6H6 + C2H ꢀ C6H5C2H + H
C6H6 + C2H3 ꢀ styrene + H
Metatheses
6.4
Wang [27]
C6H6 + H ꢀ C6H5 + H2
C6H6 + O ꢀ C6H5 + OH
C6H6 + OH ꢀ C6H5 + H2O
C6H6 + HO2 ꢀ C6H5 + H2O2
C6H6 + CH3 ꢀ C6H5 + CH4
C6H6 + C2H5 ꢀ C6H5 + C2H6
C6H6 + a-C3H5 ꢀ C6H5 + C3H6
6.0 × 108
2.0 × 1013
1.6 × 108
5.5 × 1012
2.0 × 1012
6.3 × 1011
6.3 × 1011
1.0
0.0
1.42
0.0
0.0
0.0
0.0
0.0
0.0
16.8
14.7
Mebel [28]
Lindstedt [13]
(8)
(9)
1.45 Baulch [30]
28.9
(10)
(11)
(12)
(13)
(14)
(15)
(16)
Baulch [22]
Zhang [31]
Zhang [31]
Estimateda
Estimateda
Estimateda
15.0
15.0
20.0
15.0
20.0
C6H6 + n-C4H5 ꢀ C6H5 + 1,3-C4H6 6.3 × 1011
C6H6 + i-C4H5 ꢀ C6H5 + 1,3-C4H6
6.3 × 1011
Reactions of cyclohexadienyl radicals (resonance-stabilized)
Isomerization
C6H7 ꢀ l-C6H7
2.5 × 1014
7.9 × 1011
1.0 × 1014
0.7
0.0
0.0
41.8
9.9
Weissman [32]
Estimatedb
(17)
(18)
(19)
Oxidation
C6H7 + O2 ꢀ C6H6 + HO2
Combinations
C6H7 + H ꢀ C6H8
0.0
Estimatedc
Disproportionations
C6H7 + H ꢀ C6H6 + H2
C6H7 + OH ꢀ C6H6 + H2O
C6H7 + CH3 ꢀ C6H6 + CH4
C6H7 + C6H7 ꢀ C6H6 + C6H8
Reactions of phenyl radicals
3.3 × 1012
1.0 × 1013
3.0 × 1012
8.4 × 1010
0.0
0.0
0.0
0.0
Ristori [9]
Estimatedd
Estimatedd
Estimatedd
(20)
(21)
(22)
(23)
−0.32
−0.1
0.0
−0.3
Isomerization and decompositions by beta-scission
C6H5 ꢀ l-C6H5
5.0 × 1013
0.0
0.0
0.0
72.5
51.0
41.0
Braun-Unkhoff [33]
Estimatede
Estimatede
(24)
(25)
(26)
l-C6H5 ꢀ C2H2 + C4H3
l-C6H5 ꢀ l-C6H4 + H
Reactions with O2
2.0 × 1013
2.0 × 1012
C6H5 + O2 ꢀ C6H5O + O
C6H5 + O2 ꢀ OC6H4O + H
C6H5 + O2 ꢀ C6H5O2
Additions
2.6 × 1013
3.0 × 1013
2.2 × 1019
0.0
0.0
−2.5
6.1
9.0
0.0
Frank [34]
Frank [34]
Estimated f
(27)
(28)
(29)
C6H5 + C2H2 ꢀ C6H5C2H + H
4.0 × 1013
0.0
−0.074
10.1
7.5
Marinov [35]
Wang [27]
(30)
(31)
C6H5 + C6H6 biphenyl + H
5.6 × 1012
Combinations (followed in some cases by decompositions)
C6H5 + O ꢀ C5H5 + CO
C6H5 + OH ꢀ C6H5OH
Toluene ꢀ C6H5 + CH3
C6H5 + CHO ꢀ C6H5CHO
1.0 × 1014
1.0 × 1013
1.0 × 1016
5.0 × 1012
0.0
0.0
0.0
0.0
0.0
0.0
97.0
0.0
Frank [34]
Calculatedg
Colket [36]
Calculatedg
(32)
(33)
(34)
(35)
Continued

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
511
Table IV Continued
Reactions
A
n
Ea
0.0
0.0
0.0
0.0
7.9
References
Reaction No.
C6H5 + C2H3 ꢀ styrene
C6H5 + C2H5 ꢀ ethylbenzene
C6H5 + HO2 ꢀ C6H5O + OH
C6H5 + C3H3 ꢀ C6H5C3H3
C6H5 + C6H5 ꢀ biphenyl
Disproportionations
5.0 × 1012
5.0 × 1012
5.0 × 1012
3.0 × 1012
3.8 × 1031
0.0
0.0
0.0
0.0
−5.75
Calculatedg
Calculatedg
Calculatedg
D’Anna [37]
Wang [27]
(36)
(37)
(38)
(39)
(40)
C6H5 + OH ꢀ C6H5O + H
C6H5 + C6H7 ꢀ C6H6 + C6H6
Reactions of peroxyphenyl radicals
Decomposition by beta-scission
C6H5O2 ꢀ OC6H4O + H
C6H5O2 ꢀ C5H4O + CHO
5.0 × 1013
0.0
0.0
0.0
0.0
Alzueta [7]
Shandross [4]
(41)
(42)
1.0 × 1012
2.0 × 1013
2.0 × 1013
0.0
0.0
30.0
30.0
Estimatedh
Estimatedh
(43)
(44)
Reactions of phenoxy radicals (resonance-stabilized)
CO elimination
C6H5O ꢀ CO + C5H5
2.5 × 1011
0.0
0.0
43.8
0.0
Baulch [30]
(45)
(46)
Terminations
C6H5O + H(+M) ꢀ C6H5OH(+M)
k
1.0 × 1014
1.0 × 1094
Estimatedc
Wang [27]
∞
k0
−21.84 13.9
Troe’s factors: 0.043, 304, 60000, 5896
Efficiency factors: O2, 0.4; CO, 0.75; CO2, 1.5; H2O, 6.5; N2, .4; Ar, 0.35; He, 0.35; C6H6, 3.0
C6H5O + H ꢀ C5H6 + CO
C6H5O + O ꢀ OC6H4OH
C6H5O + O ꢀ OC6H4O + H
C6H5O + O ꢀ C5H5 + CO2
1.1 × 1053
2.6 × 1010
8.5 × 1013
1.0 × 1013
−10.7
41.4
0.8
0.0
0.0
Tan [15]
Lin [38]
Alzueta [7]
Alzueta [7]
(47)
(48)
(49)
(50)
0.47
0.0
0.0
Reactions of HOC6H4O· radicals (resonance-stabilized)
CO elimination
OC6H4OH ꢀ CO + C5H4OH
7.4 × 1011
0.0
43.8
Estimatedi
(51)
Reactions of cyclopentadienyl radicals (resonance-stabilized)
Isomerization and decompositions by beta-scission
C5H5 ꢀ l-C5H5
1.0 × 1014
1.0 × 1014
2.0 × 1013
0.0
0.0
0.0
45.5
0.0
50.0
Braun [33]
Estimatedc
Estimatede
(52)
(53)
(54)
l-C5H5 + H ꢀ l-C5H6
l-C5H5 ꢀ C3H3 + C2H2
Reactions with O2
C5H5 + O2 ꢀ CHO + C4H4O
1.2 × 1019
−2.48 11.0
Zhong [39]
Estimatedc
(55)
Combinations
C5H5 + H ꢀ C5H6
1.0 × 1014
5.8 × 1013
3.2 × 1013
1.0 × 1013
3.0 × 1012
2.0 × 1012
0.0
−0.02
−0.17
0.0
0.0
0.0
0.0
(56)
(57)
(58)
(59)
(60)
(61)
C5H5 + O ꢀ C5H4O + H
C5H5 + O ꢀ 2C2H2 + CHO
C5H5 + OH ꢀ C5H5OH
C5H5 + HO2 ꢀ C5H5O + OH
C5H5 + C5H5 ꢀ C10H10
Disproportionations
0.02 Zhong [39]
0.44 Zhong [39]
0.0
0.0
0.0
Calculatedg
Calculatedg
Calculatedg
C5H5 + HO2 ꢀ C5H6 + O2
Reactions of cyclopentadienoxy radicals
Decompositions by beta-scission
C5H5O ꢀ 2C2H2 + CHO
C5H5O ꢀ C5H4O + H
Combinations
2.5 × 109
1.0
3.5
Estimatedj
(62)
2.0 × 1013
2.0 × 1013
0.0
0.0
30.0
30.0
Estimatedk
Estimatedk
(63)
(64)
C5H5O + H ꢀ C5H5OH
1.0 × 1014
0.0
0.0
Estimatedc
(65)
(66)
Reactions of hydroxyclopentadienyl (resonance-stabilized)
Oxidation
C5H4OH + O2 ꢀ C5H4O + HO2 1.0 × 1013
0.0
0.0
5.0
0.0
Estimatedl
Estimatedc
Combinations
C5H4OH + H ꢀ C5H5OH
1.0 × 1014
(67)
Continued

512
DA COSTA ET AL.
Table IV Continued
Reactions
A
n
Ea
References
Reaction No.
C5H4OH + O ꢀ CO2 + C2H2 + C2H3
C5H4OH + HO2 ꢀ OH + CO2 + C2H3 + C2H2
Disproportionations
C5H4OH + HO2 ꢀ C5H5OH + O2
C5H4OH + C6H5O ꢀ C5H4O + C6H5OH
SECONDARY MECHANISM
3.2 × 1013
−0.17
0.44
0.0
Estimatedm
Estimatedm
(68)
(69)
3.0 × 1012
0.0
2.5 × 109
1.0 × 1012
1.0
0.0
3.5
0.0
Estimatedm
Estimatedn
(70)
(71)
Reactions of ortho-benzoquinone molecules
OC6H4O ꢀ C5H4O + CO
1.0 × 1012
5.2 × 1013
1.6 × 106
4.4 × 106
0.0
0.0
2.5
2.0
40.0
Frank [34]
Estimatedo
Estimatedo
Estimatedo
(72)
(74)
(75)
(76)
OC6H4O + H ꢀ 2CO + C2H2 + C2H3
OC6H4O + H ꢀ H2 + 2CO + C2H2 + C2H
OC6H4O + OH ꢀ H2O + 2CO + C2H2 + C2H
Reactions of phenol molecules and derived radicals
C6H5OH ꢀ C5H6 + CO
3.2
9.8
1.4
1.0 × 1012
1.0 × 1013
1.6 × 1013
1.2 × 1014
1.3 × 1013
1.4 × 108
1.0 × 1012
1.8 × 1011
4.9 × 1012
4.9 × 1011
4.9 × 1011
4.9 × 1011
1.7 × 1014
2.0 × 1013
1.4 × 1013
4.0 × 1011
2.0 × 1012
2.1 × 1013
1.0 × 1014
0.0
0.0
0.0
0.0
0.0
1.4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
61.2
38.8
3.4
12.4
2.9
−0.96
10.0
7.7
4.4
9.4
9.4
9.4
16.0
14.7
4.6
28.9
15.0
6.1
Horn [40]
(77)
(78)
(79)
(80)
(81)
(82)
(83)
(84)
(85)
(86)
(87)
(88)
(89)
(90)
(91)
(92)
(93)
(94)
(95)
C6H5OH + O2 ꢀ HO2 + C6H5O
C6H5OH + O ꢀ OC6H4OH + H
C6H5OH + H ꢀ C6H5O + H2
Estimatedp
Estimatedq
Alzueta [7]
Alzueta [7]
Shandross [4]
Alzueta [7]
Mulcahy [41]
Alzueta [7]
Estimatedr
Estimatedr
Estimatedr
Shandross [4]
Estimateds
Shandross [4]
Estimateds
Estimateds
Estimatedt
Estimatedc
C6H5OH + O ꢀ C6H5O + OH
C6H5OH + OH ꢀ C6H5O + H2O
C6H5OH + HO2 ꢀ C6H5O + H2O2
C6H5OH + CH3 ꢀ C6H5O + CH4
C6H5OH + C6H5 ꢀ C6H5O + C6H6
C6H5OH + C5H5 ꢀ C6H5O + C5H6
C6H5OH + a-C3H5 ꢀ C6H5O + C3H6
C6H5OH + i-C4H5 ꢀ C6H5O + 1,3-C4H6
C6H5OH + H ꢀ C6H4OH + H2
C6H5OH + O ꢀ C6H4OH + OH
C6H5OH + OH ꢀ C6H4OH + H2O
C6H5OH + HO2 ꢀ C6H4OH + H2O2
C6H5OH + CH3 ꢀ C6H4OH + CH4
C6H4OH + O2 ꢀ OC6H4OH + O
C6H4OH + H ꢀ C6H5OH
0.0
Reactions of cyclopentadiene molecules and derived radicals
C5H6 + O2 ꢀ C5H5 + HO2
1.4 × 1012
0.0
0.0
31.6
3.2
0.59
2.0
Estimatedu
Estimatedo
Zhong [39]
Roy [42]
(96)
(97)
(98)
C5H6 + H ꢀ C5H7
5.2 × 1013
8.9 × 1012
2.8 × 1013
4.8 × 104
3.1 × 106
1.1 × 104
1.8 × 10−1
6.0 × 1012
6.0 × 1012
6.0 × 1012
2.0 × 1013
7.9 × 1011
C5H6 + O ꢀ C5H5O + H
−0.15
C5H6 + H=C5H5 + H2
0.0
(99)
C5H6 + O ꢀ C5H5 + OH
2.7
2.0
2.6
4.0
0.0
0.0
0.0
0.0
1.1
0.0
12.9
0.0
0.0
0.0
0.0
35.5
5.0
Zhong [39]
Zhong [39]
Zhong [39]
Zhong [39]
Estimatedv
Estimatedv
Emdee [12]
Estimatedw
Estimatedx
(100)
(101)
(102)
(103)
(104)
(105)
(106)
(107)
(108)
C5H6 + OH ꢀ C5H5 + H2O
C5H6 + HO2 ꢀ C5H5 + H2O2
C5H6 + CH3 ꢀ C5H5 + CH4
C5H6 + a-C3H5 ꢀ C5H5 + C3H6
C5H6 + n-C4H5 ꢀ C5H5 + 1,3-C4H6
C5H6 + i-C4H5 ꢀ C5H5 + 1,3-C4H6
C5H7 ꢀ C2H2 + a-C3H5
C5H7 + O2 ꢀ C5H6 + HO2
0.0
Reactions of cyclopentadione molecules and derived radicals
C5H4O ꢀ 2C2H2 + CO
5.7 × 1032
−6.76
0.0
0.0
0.0
0.0
1.42
0.0
−13.47
68.5
3.2
2.0
Alzueta [7]
Estimatedo
Alzueta [7]
Alzueta [7]
Alzueta [7]
Alzueta [7]
Estimatedy
Alzueta [7]
(109)
(110)
(111)
(112)
(113)
(114)
(115)
(116)
C5H4O + H ꢀ CO + n-C4H5
C5H4O + O ꢀ C4H4 + CO2
C5H4O + H ꢀ C5H3O + H2
C5H4O + O ꢀ C5H3O + OH
C5H4O + OH ꢀ C5H3O + H2O
C5H3O ꢀ C2H2 + CO + C2H
C5H3O + O2 ꢀ CO2 + C2H2 + CHCO
2.6 × 1013
1.0 × 1013
2.0 × 1012
1.4 × 1013
1.1 × 108
2.0 × 1013
9.7 × 1058
8.1
14.7
1.45
51.0
38.2
Continued

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
513
Table IV Continued
Reactions
A
n
Ea
References
Reaction No.
Reactions of cyclopentadienol molecules
C5H5OH + H ꢀ C5H5O + H2
4.0 × 1013
1.0 × 1013
1.0 × 1013
1.4 × 1013
4.8 × 104
1.5 × 106
0.0
6.1
4.
1.7
2.0
1.1
0.0
Alzueta [7]
Alzueta [7]
Alzueta [7]
Estimatedz
Estimatedz
Estimatedz
(117)
(118)
(119)
(120)
(121)
(122)
C5H5OH + O ꢀ C5H5O + OH
C5H5OH + OH ꢀ C5H5O + H2O
C5H5OH + H ꢀ C5H4OH + H2
C5H5OH + O ꢀ C5H4OH + OH
C5H5OH + OH ꢀ C5H4OH + H2O
Reactions of vinylketene molecules
C4H4O + H ꢀ C2H3 + CH2CO
C4H4O + H ꢀ C2H4Z + CHCO
C4H4O + H ꢀ s-C3H5 + CO
0.0
0.0
0.0
2.7
2.0
1.3 × 1013
1.3 × 1013
1.3 × 1013
1.4 × 1012
1.4 × 1012
6.0 × 104
6.0 × 104
8.2 × 105
4.1 × 105
2.2 × 106
1.1 × 106
1.2 × 1011
6.0 × 1010
0.0
0.0
0.0
0.0
0.0
2.56
2.56
2.5
2.5
2.0
2.0
0.7
0.7
3.0
3.0
1.5
−1.0
−1.0
−1.1
−1.1
12.3
9.8
2.8
1.5
8.7
8.7
Estimatedo
Estimatedo
Estimatedo
Estimatedo
Estimatedo
Estimatedo
Estimatedo
Estimatedo
Estimatedo
Estimatedo
Estimatedo
Estimatedo
Estimatedo
(123)
(124)
(125)
(126)
(127)
(128)
(129)
(130)
(131)
(132)
(133)
(134)
(135)
C4H4O + OH ꢀ C2H3CHO + CHO
C4H4O + OH ꢀ CO2 + a-C3H5
C4H4O + O ꢀ CH2CHO + CHCO
C4H4O + O ꢀ 2CH2CO
C4H4O + H ꢀ C2H2 + CHCO + H2
C4H4O + H ꢀ C3H3 + CO + H2
C4H4O + OH ꢀ C2H2 + CHCO + H2O
C4H4O + OH ꢀ C3H3 + CO + H2O
C4H4O + O ꢀ C2H2 + CHCO + OH
C4H4O + O ꢀ C3H3 + CO + OH
Note. The rate constants are given at 1 atm (k = A T n exp(−Ea/RT)) in cc, mol, s, kcal units. Reference numbers are given in brackets when
they appear for the first time. This mechanism should be used together with our C0 C6 reaction base [17].
a Rate constant taken equal to that of the H-abstraction with ethyl radicals with an activation energy 5 kcal mol−1 higher for resonance-
stabilized radicals.
b Rate constant for the oxidation estimated using correlations proposed in the case of allylic radicals by Heyberger et al. [43].
cRate constant taken equal to that of the recombination of ·H atoms with alkyl radicals as proposed by Allara and Shaw [44].
d Rate constant taken equal to that of the similar disproportionation of allyl radicals as proposed by Tsang [45].
e Estimated from the rate constants of the reverse addition reaction with an activation energy of 7 kcal mol−1 for the addition of ·C2H or
·C3H3 radicals and of 3 kcal mol−1 for the addition of ·H atoms; A-factor is an average value for beta-scissions [43].
(l-C6H5 = ch///c/ch//ch/ch//ch., l-C6H4 = ch///c/ch//ch/c///ch, l-C5H5 = ch2(.)/ch//ch/c///ch).
f Rate constant taken equal to that of the addition of oxygen to alkyl radicals as proposed by Warth et al. [46].
g Values calculated using software KINGAS [47] based on thermochemical kinetics methods [26]. Activation energies of recombination are
taken to be zero and A factors are estimated from modified collision theory [46].
h This reaction goes through an internal addition followed by beta-scissions involving the breaking of a C C or a C H bond. A-factor is an
average value for beta-scissions [43] and activation energies have been adjusted to obtain a reactivity at low temperature, which is in agreement
with the experimental results.
i Rate constant taken equal to that of the same reaction in the case of phenoxy radicals.
j Rate constant estimated from that of the same reaction in the case of allyl radicals [43].
k Rate constant for beta-scissions involving the breaking of a C C or a C H bond estimated by analogy with that of reactions (43) and (44).
l Rate parameters estimated from those proposed by Miyoshi et al. [48] and Alzueta et al. [7] (see text).
m Rate constant taken equal to that of the similar reaction in the case of cyclopentadienyl radicals.
n Rate constant taken equal to that of reaction (43).
o Rate constant for the addition of ·H atoms or the H-abstraction estimated using correlations proposed in the case of alkenes [43].
p A-factor taken from Alzueta et al. [7] and activation energy taken equal to the reaction enthalpy as proposed by Ingham et al. [49].
q Rate constant taken equal to that of the addition of ·O· atoms to toluene as proposed by Hoffmann et al. [50].
r Rate constant estimated from that of the H-abstraction with phenyl radicals, with A divided by 10 and an activation energy 5 kcal mol−1
higher to take into account the resonance stabilization of the involved radicals.
s Rate constant taken equal to that of the similar H-abstraction from benzene.
t Rate constant taken equal to that of the same reaction in the case of phenyl radicals.
u The rate constant of this bimolecular initiation with oxygen molecule has been calculated as proposed by Ingham et al. [49]: A is taken
equal to 2 × 7 × 1012 cm3 mol−1 s−1, as there are two allylic hydrogen atoms abstractable and the activation energy to the reaction enthalpy.
v Rate constant taken equal to that of reaction (106).
w Rate constant for the beta-scission with the breaking of a C-C bond estimated using correlations proposed in the case of alkenyl radicals
[43].
x Rate constant for the oxidation estimated using correlations proposed in the case of alkyl radicals proposed by Warth et al. [46].
y Rate constant taken equal to that of reaction (25).
z Rate constant taken equal to that of the same reaction in the case of cyclopentadiene molecules, with A-factor divided by 2 as only one
H-atom is available.

514
DA COSTA ET AL.
order to be usable in all the temperature ranges studied,
but are close to the values proposed by Braun-Unkoff
et al. [33] at 1800 K, the temperature from which
the flux of these reactions is no more negligible. We
have not taken into account the formation of o-benzyne
fromphenylradicals, whichhasbeenrecentlyproposed
based on theoretical calculations [53].
radicals (46–50). The reaction of phenoxy radicals with
·O· atoms has been experimentally studied at room
temperature by Buth et al. [55], who have identified the
formation of benzoquinone (49) and cyclopentadiene
(50). Ab initio calculations of Lin and Mebel [38] have
provided evidence for a possible third channel, involv-
ing the formation of hydroxyphenoxy radicals (48). In
the case of these less abundant hydroxyphenoxy rad-
icals, we have only considered the unimolecular CO
elimination (51).
For the resonance cyclopentadienyl radicals, we
have considered their linearization (52) as proposed by
Braun-Unkoff et al. [33] and Burcat and Dvinyaninov
[56], with reevaluated rate parameters for the follow-
ing decompositions (53–54), the reaction with oxygen
molecules (55), and termination steps with other rad-
icals (56–62). The reactions with oxygen atoms and
molecules are those proposed by Zhong and Bozzelli
[39] in their theoretical study of the reactions of cy-
clopentadienyl radicals. The reactions of cyclopenta-
dienyl radicals have also been studied by Roy et al.
[42,57]. The C10H10 compound obtained by combi-
nation of two cyclopentadienyl radicals (61) is dihy-
drofulvalene. At low temperature, the direct reaction
leadingfromtworesonance-stabilizedradicalstonaph-
thalene and two reactive H atoms, as proposed by Roy
et al. [42], would increase too much the reactivity
of the system and cannot then be taken into account.
The formation of naphthalene could only derive from
an H-abstraction from dihydrofulvalene by small rad-
icals, which are present in high concentration in
flame.
As proposed by Frank et al. [34], phenyl radicals
can react with oxygen to form phenoxy radicals and
·O· atoms (27) or benzoquinone and H· atoms (28);
according to Alzueta et al. [7], we have considered that
reaction (28) leads mainly to ortho-benzoquinone. But
to be able to reproduce our results at low temperature,
we have also considered the formation of phenylper-
oxy radicals (29). The used rate parameters lead to a
rate constant in agreement with the value measured by
Yu and Lin [54] at 473 K. Other reactions of phenyl
radicals include additions to acetylene (30) and ben-
zene (31) molecules and termination steps with other
radicals (32–42). It is worth noting that the formation
of peroxy radicals is a reversible step and that the im-
portance of the reverse reaction strongly depends on
thermochemical data, which should then be calculated
by a more accurate method than the groups additivity.
In addition, the contributions of the three channels to
the global reaction of phenyl radicals with oxygen de-
pend on pressure. At high pressure, peroxy radicals can
easily be stabilized, while at lower pressure, the forma-
tion of benzoquinone or phenoxy radicals are favored.
Very little is known about the reaction channels of
phenylperoxy radicals. We have postulated that they
canreactbyinternaladditionaccordingtothefollowing
scheme:
to form ortho-benzoquinone and H· atoms (43) or cy-
clopentadienone and CHO· radicals (44).
The reactions of oxygenated C5 cyclic radicals are
not yet well defined. Cyclopentadienoxy radicals can
decompose by beta-scission involving the breaking of
a C C (63) or a C H bond (64) or combine with
Resonance-stabilized phenoxy radicals can react ei-
ther by the classical CO elimination to form cyclopen-
tadienyl radicals (45) or by termination steps with other

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
515
H· atoms (65). Resonance-stabilized hydroxycy-
clopentadienyl radicals can react with oxygen
molecules to form cyclopentadienone molecules (66)
or by termination steps similar to those related to cy-
clopentadienyl radicals (67–71). The reaction of hy-
droxyalkylradicalstoformketonemoleculeshavebeen
studied by Miyoshi et al. [48] who proposed an average
of the ring and the following beta-scission involving
the breaking of a C C bond and the opening of the
ring to form carbon monoxide and n-C5H5 radicals
(110). Cyclopentadienonyl radicals can react with oxy-
gen molecule (115) as proposed by Alzueta et al. [7],
but we have also considered the decomposition by beta-
scission involving the breaking of a C C bond and the
opening of the ring to form carbon monoxide, acety-
lene, and ·CHCO radicals (116).
rate constant, k = 1 × 1013
s
−1 at 296 K. In the case of
these cyclic resonance-stabilized hydroxy radicals, we
have considered an activation energy of 10 kcal mol−1
.
Cyclopentadienol can react by H-abstractions to
form either cyclopentadienoxy radicals (117–119) as
proposed by Alzueta et al. [7] or the resonance-
stabilized hydroxycyclopentadienyl radicals (120–
122).
Secondary Mechanism
The secondary mechanism includes the reactions
of ortho-benzoquinone, phenol, cyclopentadiene, cy-
clopentadienone, andvinylketenemolecules, whichare
produced in the primary mechanisms.
Ortho-benzoquinone can decompose to give cy-
clopentadienone and CO (72), as proposed by Frank
et al. [34], or react by addition of H· atoms (74) or by
H-abstractions with the subsequent decompositions of
the obtained radicals (75–76) with rate constant derived
from those proposed for alkenes [43].
The reactions of vinylketene (CH2 CHCH C O)
were not considered in the C0 C6 reaction base [17].
They have been deduced from those of alkenes [43]
and included the reactions of H· (123–125) and ·O·
(128–129) atoms and of ·OH (126–127) radicals by
addition to the double bonds or by H-abstraction (130–
135).
In the case of the simulations in flame, secondary
mechanisms of toluene, ethylbenzene, and styrene have
been added. In the case of a flame, these secondary
mechanisms had a small influence on the results, while
this effect was negligible in other types of reactor, even
in fuel-rich conditions, because very small amounts of
toluene (formed by the reverse of reaction 34), ethyl-
benzene (formed by reaction 37), or styrene (formed by
reaction 36) were formed. As our purpose was to in-
vestigate the chemistry of the oxidation of benzene and
not the formation of PAHs, we have neglected the for-
mation of aromatic compounds heavier than biphenyl
and C6H5C3H3.
Alzueta et al. [7] have experimentally studied and
modeled the pyrolysis and oxidation of phenol in
plug-flow conditions between 900 and 1450 K. For
this compound, we have considered the molecular
decomposition to give cyclopentadiene and carbon
monoxide (77), the unimolecular initiation with oxy-
gen molecules (78), the addition of ·O· atoms (79),
and the H-abstractions, which can lead either to phe-
noxy radicals (80–88) or to C6H4·OH radicals (89–93).
As the formation of the resonance-stabilized phenoxy
radicals is easy, we have considered the H-abstraction
from phenol with many radicals, including the main
small resonance-stabilized radicals, i.e. cyclopentadi-
enyl, allyl, and i-C4H5 radicals. C6H4·OH radicals can
react with oxygen (94) in the same way as phenyl rad-
ical or combine with H· atoms (95).
COMPARISON BETWEEN COMPUTED
AND EXPERIMENTAL RESULTS
Simulations were performed using the softwares of
CHEMKIN II [18]. We have tried to reproduce the
experimental data described in this paper (jet-stirred
reactor and shock tube), but to extend the validity
of the proposed mechanism, we have also attempted
to model results of the literature obtained in a flow
tube [6] and in a near-sooting laminar flame [3]. The
mechanism presented here before has been used for all
the different performed simulations.
Cyclopentadiene can react by unimolecular initia-
tion with oxygen molecules (96), by additions of H·
(97) and ·O· (98) atoms, and by H-abstractions to form
cyclopentadienyl radicals (99–106). By analogy with
the reaction of alkenyl radicals [40], the C5H7· radicals
obtained by additions of H· atoms can react by beta–
scission involving the breaking of a C C bond (107)
or by reaction with oxygen molecules to give back cy-
clopentadiene (108).
As proposed by Alzueta et al. [7], cyclopenta-
dienone can react by molecular decomposition to give
acetylene and carbon monoxide (109), by addition of
·O· (111) atoms and by H-abstractions to form cy-
clopentadienonylradicals(112–114). But, wehavealso
considered the addition of H· atoms to the double bonds
Perfectly Stirred Reactor
As the jet-stirred reactor has been proved to provide
one of the very few types of gas-phase perfectly stirred
reactor [58], we have modeld our results using the PSR

516
DA COSTA ET AL.
code of CHEMKIN II. Figures 2 and 3 display com-
parisons between the experimental (symbols) and com-
puted (lines) results for the experiments performed in
a jet-stirred reactor at ꢀ equal to 1.9 and 3.6, respec-
tively. These figures show that our mechanism leads to
an acceptable agreement for both equivalence ratios.
The conversion of reactants are well reproduced in both
cases, while the production of phenol is systematically
overpredicted, whatever the residence time, betraying
probably condensation problems for this compound be-
fore the entrance in the trap. For both equivalence ra-
tios, the formation of carbon monoxide, carbon diox-
ide, methane, C2 compounds, propyne, propene, and
butadiene is overpredicted at short residence times, but
the shape of the curves and the values at longer resi-
dence times are globally well reproduced. At 923 K,
at an equivalence ratio of 1.9, and at a residence time
of 4 s, C2 compounds contain 53% acetylene and 47%
ethylene.
cases. For the shortness of the paper, simulations at ꢀ
equal to 1 are not presented, but they lead to a similar
agreement. The model can satisfactorily reproduce the
consumption of benzene and the formation of phenol,
cyclopentadiene, and C4 compounds (99% vinylacety-
lene and 0.9% 1,3-butadiene, at ꢀ equal to 0.76, at
a residence time of 0.075 s) in both conditions. The
production of carbon monoxide is well simulated at ꢀ
equal to 0.76, while it is overestimated by a factor 2
for the longest residence times at ꢀ equal to 1.3. The
formation of C2 compounds (8% acetylene and 92%
ethylene, at ꢀ equal to 0.76, at a residence time of
0.075 s) is slightly underpredicted at ꢀ equal to 0.76,
while it is overpredicted by a factor up to 3 at ꢀ equal
to 1.3.
Ourmodelallowsustoobtainbetterpredictionsthan
the mechanisms proposed by Emdee et al. [12], espe-
cially for the production of carbon monoxide, reflecting
the fact that we have a more detailed submechanism for
Shock Tube
Computed results for the shock tube appear in Fig. 4.
It is worth noting that the experimental OH emission at
306 nm is related to electronically exited OH concen-
tration and is not directly proportional to OH radical
concentration. Nevertheless, previous work [18] has
shown a correct agreement between the shapes of the
profiles of experimental emission and calculated OH
concentration during the rise of the signal, which is
the important part of the curve for the determination
of ignition delays. Both experimentally and theoreti-
cally, the ignition delay is determined at 10% of the
maximum OH peak. Model results reproduce the sys-
tematic trends of the measurements, including varia-
tions in ignition delays with temperature, equivalence
ratio, and concentration of benzene. Agreement dete-
riorates slightly for ꢀ equal to 0.5, where computed
delay times are about two times higher than the exper-
imental values. At ꢀ equal to 1 and 1.25% of benzene,
the computed line is closer to the experimental results
presented in this paper than to the line derived from the
correlation proposed by Burcat et al. [5].
Flow Reactor
Lovell et al. [6] have studied the oxidation of benzene
in a flow reactor at 1100 K, atmospheric pressure, with
nitrogen as bath gas, for a concentration of benzene of
1580 ppm and for three equivalence ratios, 0.76, 1, and
3. Figures 5 and 6 display comparisons between the
experimental (symbols) and computed (lines) results
at ꢀ equal to 0.76 and 1.3, respectively, and show that
a globally correct agreement can be observed in both
Figure5 Oxidationofbenzeneinflowreactorat1100K[6].
Comparison between experimental (symbols) and computed
(lines) species vs. residence time at ꢀ = 0.76.

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
517
molecular beam mass spectrometer system. They have
usedmixturescontaining13.5%ofbenzeneatanequiv-
alence ratio of 1.8 and a cold gas velocity of 0.5 m s−1
at room temperature.
To model correctly a laminar premixed flame, it is
necessary to consider the transport of species, momen-
tum and energy, in a multicomponent gaseous mixture.
CHEMKIN II [20] takes into account these phenomena
by means of multicomponent transport properties (dif-
fusion coefficients, viscosities, and thermal diffusion
coefficients). These coefficients are calculated using
relations involving the ratio ε/kB (with ε the Lennard–
Jones well depth and kB the Boltzmann constant), the
Lennard–Jones collision diameter σ, the dipole mo-
ment µ, and the polarizability α. A file containing these
parameters for several simple species is provided in the
CHEMKIN II package [20]. However, in the case of the
oxidation of benzene, it has been necessary to estimate
transport data, especially the Lennard–Jones parame-
ters (ε/kB and σ), for large molecules. For this, we used
correlations found in the literature [59]:
σ3 Pc/Tc = 13.56 + 9.6ω + 6.26ω2 − 10ω3
(ε/kB)/Tc = 0.753 − 0.468ω − 0.277ω2 + 0.426ω3
where Pc (atm), Tc (K), and ω represent, respec-
tively, the critical pressure, the critical temperature,
and the acentric factor. These last parameters can be
found in the literature for a large range of molecules
or can be estimated by means of the Joback group
contributions [60]. Table V summarizes, for several
molecules involved in the oxidation of benzene, the
overall parameters used in the calculation of Lennard–
Jones parameters. The coefficients of the molecule,
which has the closest chemical structure, were used for
radicals.
Figure 7 presents a comparison between these ex-
perimental data (symbols) and the predictions (lines)
given by our model. The major stable species, benzene,
oxygen, carbon monoxide, carbon dioxide, acetylene,
phenol, cyclopentadiene, hydrogen, and water, are well
predicted, while the maximum of vinylacetylene is un-
derpredicted. The profiles of H· and ·OH radicals are
well represented, but those of ·HO2 (up to factor 2) and
phenyl (up to factor 4) radicals are overestimated. The
agreement obtained here is globally similar to that of
other recent modeling work [9,13–15].
Figure6 Oxidationofbenzeneinflowreactorat1100K[6].
Comparison between experimental (symbols) and computed
(lines) species vs. residence time at ꢀ = 1.3.
C5 species. It is worth noting that in our case no time
shift has been necessary. These data have been also
modeled by Zhong and Bozzelli [39] in order to test
their submechanism for the oxidation of cyclopentadi-
ene. In all the studied conditions, their agreement for
the profiles of benzene and carbon monoxide is good,
but they overpredict the formation of cyclopentadiene
by a factor 2, even for short residence times, and they
do not present the formation of phenol, C2, and C4
compounds. An example of simulation of these exper-
iments in lean mixture is presented by Alzueta et al.
[7], but the agreement is much poorer than what can be
obtained by Zhong and Bozzelli [39] and than what is
shown here.
Premixed Flame
ANALYSIS OF THE MECHANISM
Bittner and Howard [3] have studied a near-sooting
laminar premixed benzene–oxygen–argon flame stabi-
lized on a burner at 2.67 kPa and analyzed by using a
It is worth noting that the fluxes and sensitivity analy-
ses presented hereafter have been performed with the

518
DA COSTA ET AL.
Table V Critical Temperature and Pressure, Acentric Factor, and Lennard–Jones Parameters
Species
Critical Temperature Tc (K)
Critical Pressure Pc (atm)
Acentric Factor ω
ε/kb (K)
σ (A)
C6H6
561.9
746
694.1
593.8
650
616.9
647.4
768
493
532
687
625
48.6
60
60.5
41.6
44
36.9
39.4
31.8
44.6
49.3
54.5
52
0.216
0.486
0.439
0.258
0.237
0.32
0.243
0.416
0.174
0.2
361.5
379.7
368.0
368.8
410.9
363.3
407.2
415.5
328.1
346.7
361.4
375.1
281.8
269.9
5.7
6.1
5.9
6.1
6.2
6.6
6.4
7.6
5.5
5.5
6.1
5.9
6.0
5.1
OC6H4O
C6H5OH
Toluene
C6H5C2H
Ethylbenzene
Styrene
Biphenyl
l-C5H6
C5H6
C
10H10
0.448
0.298
0.614
0.286
C5H4O
C5H5OH
C4H4O
612.8
445
54.6
55.7
Figure 7 Near-sooting laminar premixed flame of benzene [3]. Comparison between experimental (symbols) and computed
(lines) species vs. distance to the burner at ꢀ = 1.8.

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
519
mechanism described previously and that the accuracy
of conclusions derived from them can only reflect the
accuracy of the proposed reaction channels and of the
used thermodynamic and kinetic data.
in a laminar flame at 1710 K (equivalence ratio of 1.8,
distance to the burner of 0.75 cm corresponding to a
85% conversion of benzene), respectively.
Figures 8 and 9a present fluxes and sensitivity
analysis computed in a jet-stirred reactor at 923 K, an
equivalence ratio of 1.9, and a residence time of 4 s
corresponding to a 30% conversion of benzene. Fig-
ure 9b presents a sensitivity analysis computed around
1300 K and a shock tube at 70% conversion of benzene.
Figure 8b demonstrates the determinant role for igni-
tion delays of termination (e.g., the recombinations of
·H atoms with phenoxy or cyclopentadienyl radicals)
and branching (e.g., the branching reaction between ·H
atoms and oxygen molecules) reactions. Figures 10
and 11 display fluxes computed in a flow tube at 1100
K (equivalence ratio of 0.76, residence time of 0.075
s corresponding to a 47% conversion of benzene) and
Main Reaction Pathways
of Benzene Molecules
In all cases, benzene reacts mainly by two channels,
addition of ·O· atoms to give phenoxy radicals (4) and
metatheses, i.e. H-abstraction by H· atoms, OH·, and
phenoxy radicals, to form phenyl radicals (8, 10, −85).
In a perfectly stirred reactor, at 923 K, metatheses ac-
count for 66% of the benzene consumption, while 32%
is due to the addition of ·O· atoms. The ratio between
these two channels does not depend much on tempera-
ture or equivalence ratio. Above 1800 K and in a flame,
benzene slightly decomposes to propargyl radicals (re-
action of the C0 C6 reaction base [19]).
Figure 8 Oxidation of benzene in a jet-stirred reactor at 923 K. Flux analysis at ꢀ = 1.9 for a residence time of 4 s.

520
DA COSTA ET AL.
Figure 9 Oxidation of benzene in a jet-stirred reactor and in a shock tube. Sensitivity analyses related to (a) the conversion
of benzene in a jet-stirred reactor, at 923 K, ꢀ = 1.9 and a residence time of 4 s (only reactions, with an absolute value of the
sensitivity coefficient above 0.024, are shown) and to (b) the mole fraction of hydroxyl radicals in a shock tube at an initial
temperature of 1300 K, at ꢀ = 1 and a residence time corresponding to 50% conversion of benzene (only reactions, with an
absolute value of the sensitivity coefficient above 0.06, are shown).
tor and in a shock tube. It is worth noting that the
rate constants of some important reactions of phe-
noxy radicals, such as their reactions with ·HO2 rad-
icals (−78) or the disproportionation with hydrox-
ycyclopentadienyl radicals (51), are only based on
estimations.
The abstractions of the hydroxylic H-atoms by H·
(80) and ·O· (81) atoms, OH· (82), and, at low temper-
ature, ·HO2 radicals (83) to give back phenoxy radical
are important reactions of phenol. In a perfectly stirred
reactor, at 923 K, these reactions account for 80% of
the phenol consumption. This is due to the fact that
phenoxy radicals are resonance-stabilized. At 1710 K,
the major reaction is the molecular decompositions to
Reactions Deriving from the Formation
of Phenoxy Radicals
Phenoxy radicals can react with ·H atoms (46) or, at
low temperature, with ·HO2 radicals (−78) or ben-
zene molecules (−85), to form phenol or eliminate
carbon monoxide (45) to give cyclopentadienyl rad-
icals. A minor channel involves the formation of
benzoquinone by addition of ·O· atoms (49). The
importance of the CO elimination increases with tem-
perature, while that of termination steps decreases.
Figure 9 shows that the reactions forming or consum-
ing resonance-stabilized phenoxy radicals have im-
portant sensitivity coefficients in a jet-stirred reac-

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
521
Figure 10 Oxidation of benzene in flow reactor at 1100 K. Flux analysis at ꢀ = 0.76 at a residence time of 0.075 s.
give cyclopentadiene and carbon monoxide (77). Other
reactions include the addition of ·O· atoms to form hy-
droxyphenoxy radicals (79) and the abstraction of aro-
matic H-atoms by OH· radicals to give ·C6H4OH rad-
icals (91), which react rapidly with oxygen molecules
to give hydroxyphenoxy radicals (94). Hydroxyphe-
noxy radicals decompose by CO elimination to give
the resonance-stabilized hydroxycyclopentadienyl rad-
icals (51), which react mainly by disproportionation
with phenoxy radicals and lead to cyclopentadienone
(71).
radicals (27), the formation of ·H atoms and of ben-
zoquinone (28), and the production of phenylperoxy
radicals, are of importance (29). Of course, the forma-
tion of phenylperoxy radicals decreases with tempera-
ture. Phenylperoxy radicals decompose to give either
·H atoms or benzoquinone (43) or ·HCO radicals and
cyclopentadienone (44). Cyclopentadienone is also the
main product of the decomposition of benzoquinone
by CO elimination (72).
Above 1800 K and in a low-pressure flame, a small
fraction of phenyl radicals decompose to give acetylene
and n-C4H3· radicals (reaction of the C0 C6 reaction
base [19]) or are linearized (24) and decomposed to
produce also acetylene and n-C4H3· radicals (25).
Reactions Deriving from the Formation
of Phenyl Radicals
In the studied conditions, phenyl radicals react mainly
with oxygen molecules. In perfectly stirred and flow
reactors (below 1100 K), the three considered chan-
nels, the branching step to give ·O· atoms and phenoxy
Reactions of C5 Compounds
In a perfectly stirred reactor, at 923 K and for an
equivalence ratio equal to 1.9, the main reaction of

522
DA COSTA ET AL.
Figure 11 Near-sooting laminar premixed flame of benzene. Flux analysis at ꢀ = 1.8, a distance to the burner equal to 0.75
cm, and a temperature of 1710 K.
cyclopentadienyl radicals is with oxygen to produce
·HCO radicals and vinylketene (55). The four main re-
actions of vinylketene are with ·H atoms to produce
CH2CO and vinyl radicals (123), ethylene and ·CHCO
radicals (124), or carbon monoxide and s-C3H5 radi-
cals (125), and with ·OH radicals to give carbon diox-
ide and allyl radical (127). The decomposition of s-
C3H5 radicals by breaking of a C H bond leads to the
formation of propyne (reaction of the C0 C6 reaction
base [19]). The reaction of allyl radicals with phenol
explains the major part of the production of propene
(87).
In a flow reactor and in a shock tube, other reac-
tions of cyclopentadienyl radicals include the forma-
tion of cyclopentadiene by combination with ·H atoms
(56) and that of ·OH and cyclopentadienoxy radicals
by combination with ·HO2 radicals (60). The major
reaction of cyclopentadienyl radicals in a flame is its
linearization (52) and the decomposition of the lin-
ear isomer to give acetylene and propargyl radicals
(54). Figure 9b shows the important role of resonance-
stabilized cyclopentadienyl radicals for ignition delays.
The rate constants of most reactions of cyclopentadi-
enyl radicals are based on calculations or estimations.
The very few direct measurements of rate constants for
these radicals have been obtained at high temperature
in a shock tube [33,57].
Cyclopentadiene reacts mainly by metatheses with
·H atoms (99) and ·OH (101), and allyl (104) and phe-
noxy (−86) radicals to give back cyclopentadienyl rad-
icals. The major reactions of cyclopentadienoxy rad-
icals are decompositions to give ·HCO radicals and
acetylene (63) and ·H atoms and cyclopentadienone
(64).

EXPERIMENTAL AND MODELING STUDY OF THE OXIDATION OF BENZENE
523
Cyclopentadienone reacts mainly with ·H atoms to
produce n-C4H5· and carbon monoxide (110). The re-
action of n-C4H5· radicals with oxygen leads to viny-
lacetylene (110), which can react with ·H atoms to
give i-C4H5· radicals (reactions of the C0 C6 reaction
base [19]), which can react with phenol to produce 1,3-
butadiene (88). Other reactions of cyclopentadienone
include the reaction with ·O· atoms to give vinylacety-
lene and carbon dioxide (111) and the H-abstraction
with ·OH radicals to produce cyclopentadienonyl rad-
icals (114), which react with oxygen molecules and
lead to ·HCCO radicals, carbon dioxide, and acetylene
(116).
It is worth noting that, in all the studied condi-
tions, the reactions of oxygenated C5 cyclic radicals,
which are not yet well defined, are important stages
to explain the formation of small degradation prod-
ucts such as carbon monoxide and C2 and C4 com-
pounds. That could explain the problems encountered
in our simulations to reproduce some of these light
compounds.
3. Bittner, J. D.; Howard, J. B. Symp (Int) Combust, [Proc]
1981, 8, 1105.
4. Shandross, R. A.; Longwell, J. P.; Howard, J. B. Int Com-
bust Proc 1996, 26, 711.
5. Burcat, A.; Snyder, C.; Brabbs, T. NASA TM-87312,
1986.
6. Lovell, A. B.; Brezinsky, K.; Glassman, I. Symp (Int)
Combust, [Proc] 1988, 22, 1063.
7. Alzueta, M. U.; Glarborg, P.; Dam-Johansen, K. Int J
Chem Kinet 2000, 32, 498.
8. Chai, Y.; Pfefferle, L. D. Fuel 1998, 77, 313.
9. Ristori, A.; Dagaut, P.; El Bakali, A.; Pengloan, G.;
Cathonnet, M. Combust Sci Technol 2001, 167, 223.
10. Scho¨bel, A.; Glass, A.G.; Krebs, L.; Braun-Unkhoff, M.;
Wahl, C.; Frank, P. Chemosphere 2001, 42, 591.
11. Bittker, D. A. Combust Sci Technol 1991, 79, 49.
12. Emdee, J. L.; Brezinsky, K.; Glassman, I. J Phys Chem
1992, 96, 2151.
13. Lindstedt, R. P.; Skevis, G. Combust Flame 1994, 99,
551.
14. Zhang, H. Y.; McKinnon, J. T. Combust Sci Technol
1995, 107, 261.
15. Tan, Y.; Frank, P. Symp (Int) Combust, [Proc] 1996, 26,
677.
16. Richter, H.; Grieco, W. J.; Howard, J. Combust Flame
1999, 119, 1.
17. Bauge´, J. C.; Battin-Leclerc, F.; Baronnet, F. Int J Chem
Kinet 1998, 30, 629.
18. Fournet, R.; Bauge´, J. C.; Battin-Leclerc, F. Int J Chem
Kinet 1999, 31, 361.
19. Belmekki, N.; Glaude, P. A.; Da Costa, I.; Fournet, R.;
Battin-Leclerc, F. Int J Chem Kinet 2002, 34, 172.
20. Kee, R. J.; Rupley, F. M.; Miller, J. A. Sandia Laborato-
ries Report, SAND 89-8009B, 1993.
21. Barbe´, P.; Battin-Leclerc, F.; Coˆme, G. M. J Chim Phys
1995, 92, 1666.
22. Baulch, D. L.; Cobos, C. J.; Cox, R. A.; Frank, P.;
Hayman, G. D.; Just, T.; Kerr, J. A.; Murrells, T. P.;
Pilling, M. J.; Troe, J.; Walker, R. W.; Warnatz, J. Com-
bust Flame 1994, 98, 59.
23. Tsang, W.; Hampson, R. F. J Phys Chem Ref Data 1986,
15, 1087.
24. Troe, J. Ber Bunsen-Ges Phys Chem 1974, 78, 478.
25. Muller, C.; Michel, V.; Scacchi, G.; Coˆme, G. M. J Chim
Phys 1995, 92, 1154.
CONCLUSION
This work has allowed us to present new experimental
results for the oxidation of benzene in a jet-stirred re-
actor at a temperature of 923 K and in a shock tube at
temperatures between 1230 and 1970 K and to propose
a detailed mechanism able to reproduce these data, but
also previously published experiments in a flow reactor
and in a near-sooting laminar flame [3,6].
The reactions of importance in this mechanism have
been determined by using flux and sensitivity analyses.
Amongst them, the reactions of oxygenated C5 cyclic
radicals are yet very uncertain and are a necessary sub-
ject of studies for specialists of elementary reactions to
progress toward a better modeling of the oxidation of
aromatic compounds.
Despite these uncertainties in the chemistry of C5
compounds, the mechanism proposed here constitutes
a first basis in the further development of models for
the oxidation of other aromatic compounds, such as
toluene or xylene.
26. Benson, S. W. ThermochemicalKinetics, 2nded.;Wiley:
New York, 1976.
27. Wang, H.; Frenklach, M. Combust Flame 1994, 110,
173.
28. Mebel, A. M.; Lin, M. C.; Yu, T.; Yorokuma, K. J Phys
Chem A 1997, 101, 3189.
BIBLIOGRAPHY
29. Nicovich, J. M.; Gump, C.; Ravishankara, A. R. J Phys
Chem 1982, 86, 1684.
1. Walker, R. W.; Morley, C. In Comprehensive Chemi-
cal Kinetics; Pilling, M. J. (Ed.); Elsevier: Amsterdam,
1997; p. 1.
2. Guibet, J. C. FuelsandEngines;Publicationsdel’Institut
Franc¸ais du Pe´trole, Editions Technip, 1999; Vols. 1 and
2.
30. Baulch, D. L.; Cobos, C. J.; Cox, R. A.; Esser, C.; Frank,
P.; Just, T.; Kerr, J. A.; Pilling, M. J.; Troe, J.; Walker, R.
W.; Warnatz, J. J Phys Chem Ref Data 1992, 21, 411.
31. Zhang, H. X.; Ahonkhai, S. I.; Back, M. H. Can J Chem
1989, 67, 1541.

524
DA COSTA ET AL.
32. Weissman, M.; Benson, S. W. Prog Energy Combust Sci
1989, 15, 273.
33. Braun-Unkhoff, M.; Frank, P.; Just, Th. Symp (Int) Com-
bust, [Proc] 1988, 22, 1023.
34. Frank, P.; Herzler, J.; Just, Th.; Wahl, C. Symp (Int)
Combust, [Proc] 1994, 25, 833.
35. Marinov, M. N.; Pitz, W. J.; Westbrook, C. K.; Castaldi,
M. J.; Senkan, S. M. Combust Sci Technol 1996, 116,
211.
46. Warth, V.; Stef, N.; Glaude, P. A.; Battin-Leclerc, F.;
Scacchi, G.; Coˆme, G. M. Combust Flame 1998, 114,
81.
47. Bloch-Michel, V. PhD Thesis, Institut National Poly-
technique de Lorraine, Nancy, France, 1995.
48. Miyoshi, A.; Matsui, H.; Washida, N. J Phys Chem 1990,
94, 3016.
49. Ingham, T.; Walker, R. W.; Woolford, R. E. Symp (Int)
Combust, [Proc] 1994, 25, 767.
36. Colket, M. B.; Seery, D. J. Symp (Int) Combust, [Proc]
1994, 25, 883.
50. Hoffmann, A.; Klatt, M.; Wagner, H. Gg. Z Chem Phys
(Neue Folge) 1990, 168, 1.
37. D’Anna, A.; Violi, A. Symp (Int) Combust, [Proc] 1998,
27, 425.
51. Scherer, S.; Just, Th.; Frank, P. Symp (Int) Combust,
[Proc] 2000, 28, 1511.
38. Lin, M. C.; Mebel, A. M. M. J Phys Org Chem 1995, 8,
407.
52. Stein, S. E.; Walker, J. A.; Suryan, M. M.; Fahr, A. Int
Combust Proc 1990, 23, 85.
39. Zhong, X.; Bozzelli, J. W. J Phys Chem A 1998, 102,
3537.
53. Wang, H.; Laskin, A.; Moriarty, N. W.; Frenklach, M.
Int Combust Proc 2000, 28, 1545.
40. Horn, C.; Roy, K.; Frank, P.; Slutsky, V. G.; Just, Th. Int
Combust Proc 1998, 27, 321.
41. Mulcahy, M. F. R.; Williams, D. J. Aust J Chem 1965,
18, 20.
42. Roy, K.; Horn, C.; Frank, P.; Just, Th. Int Combust Proc
1998, 27, 329.
43. Heyberger, B.; Belmekki, N.; Conraud, V.; Glaude, P. A.;
Fournet, R.; Battin-Leclerc, F. Int J Chem Kinet 2002,
34, 172.
54. Yu, T.; Lin, M. C. J Am Chem Soc 1994, 85, 2262.
55. Buth, R.; Hoyermann, K.; Seeba, J. Symp (Int) Combust,
[Proc] 1994, 25, 841.
56. Burcat, A.; Dvinyaninov, M. Int J Chem Kinet 1996, 29,
505.
57. Roy, K.; Braun-Unkhoff, M.; Frank, P.; Just, Th. Int J
Chem Kinet 2001, 33, 821.
58. Matras, D.; Villermaux, J. Chem Eng Sci 1973, 28, 129.
59. Gosse, J. Techniques de l’inge´nieur K 1991, 425, 1.
60. Reid, R. C.; Prausnitz, J. M.; Poling, B. E. The Properties
of Gases and Liquids, 4th ed.; McGraw-Hill: New York,
1987.
44. Allara, D. L.; Shaw, R. J Phys Chem Ref Data 1980, 9,
3.
45. Tsang, W. J Phys Chem Ref Data 1991, 20, 221.