Decane oxidation in a shock tube

Article abstract of DOI:10.1002/kin.20204

The ignition delay of n-decane and oxygen diluted in argon was investigated for a series of mixtures ranging from 0.49 to 1.5% decane and 4.16 to 23.25% O₂ diluted in argon. The temperature range was 1239-1616 K and the pressure range was 1.82-10.0 atm. All experiments were performed in a heated shock-tube. An overall ignition delay equation was deduced for 144 experiments: τ = 10^-12 exp(+34240/RT)|C₁₀H₂₂| ^0.60[O₂]^-1.305[Ar]_0.08 s. Product distribution from preignition periods were measured. Detailed simulation schemes available in the literature were checked and a corrected model is proposed that fits well our experiments.

Full text of DOI:10.1002/kin.20204

Decane Oxidation in a
Shock Tube
ERNA OLCHANSKI, ALEXANDER BURCAT
Faculty of Aerospace Engineering, Technion—I.I.T., Haifa 32000, Israel
Received 22 June 2004; revised 24 May 2006; accepted 25 May 2006
DOI 10.1002/kin.20204
Published online in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: The ignition delay of n-decane and oxygen diluted in argon was investigated for a
series of mixtures ranging from 0.49 to 1.5% decane and 4.16 to 23.25% O2 diluted in argon.
The temperature range was 1239–1616 K and the pressure range was 1.82–10.0 atm. All experi-
ments were performed in a heated shock-tube. An overall ignition delay equation was deduced
for 144 experiments: τ = 10−12 exp(+34240/RT )[C10H22
]
0.60[O2]−1.305[Ar]0.08 s. Product dis-
tribution from preignition periods were measured. Detailed simulation schemes available in
the literature were checked and a corrected model is proposed that fits well our experiments.
ꢀ
C
2006 Wiley Periodicals, Inc. Int J Chem Kinet 38: 703–713, 2006
Ignition delay measurements were performed in a
heated shock tube by Adomeit and coworkers [12] for
n-decane/air mixtures at 12 and 50 bar pressure and
temperature range (T5) of 700–1300 K. The purpose
of these experiments was mainly to describe a two-
step oxidation process at high- and low-temperature
regions. They found a negative temperature trend in
their ignition experiments. The number of experiments,
variety of the test-mixtures, and pressure conditions
were not sufficient for statistical analysis. These re-
sults were used for decane oxidation modeling by
three groups [3,7,8]. One group [7] validated their
n-decane autoignition mechanism against these exper-
imental data in the temperature range 1000–1300 K.
It should be noticed that in the lower part of the
temperature interval (1000–1150 K), the model over-
predicted the ignition delay time. Another group [8]
proposed an oxidation mechanism, which fitted well
the experimental data at temperatures 700–1250 K
with good prediction of the temperature’s negative
trend region. In all the modeling studies, the models
were compared directly against individual experiments
ratherthan anoverall experimental expression(notsup-
plied up to now), which applies to a broader range
of pressures, temperatures, and compositions. Such a
correlation was proposed by Lifshitz [10] more than
INTRODUCTION
Normal decane (C10H22) is one of the most obvious
ingredients of diesel fuel although its concentration in
diesel oil distillates does not exceed 3%. It was also
defined as one of three ingredients (the others being
methyl naphthalene and normal heptane) and some-
timesasoneoftwoingredients(theotherbeingtoluene)
that their blend should be used to simulate diesel
fuel.
A short list of experimental and numerical stud-
ies of n-decane combustion has been reported lately
[1–9] and these include Dagaut et al. [1], Westbrook
and coworkers [2], Warnatz and coworkers [3], Battin-
Leclerc and coworkers [4,5], Lindstedt and Maurice
[6], Dryer and coworkers [7], Peters and coworkers
[8], and Wilson and coworkers [9]. In these studies,
oxidation mechanisms were validated against different
types of n-decane experiments in the gas phase, mainly
jet stirred reactors and premixed flame propagation
measurements. Ignition delay times were reported at a
minimum.
Correspondence to: A. Burcat; e-mail: aer0201@tx.technion.
ac.il.
c
ꢀ
2006 Wiley Periodicals, Inc.

704
OLCHANSKI AND BURCAT
30 years ago and since then it has been extensively
used. It should be mentioned the work of Davidson
and coworkers [11] where the authors try to corre-
late four sets of stoichiometric mixtures for four dif-
ferent normal alkanes (propane, butane, heptane, and
decane), with oxygen diluted in argon and statisti-
cally regress them together using Lifshitz’s [10] for-
mula to obtain a general correlation for autoignition of
n-alkanes. Theauthorsproposedasingleexpressionfor
all four alkanes’ ignition delay time with an activation
energy of 46 kcal/mol. They used their own experi-
ments for decane ignition, which were performed with
only one stoichiometric test-mixture of decane–O2–Ar
under a narrow range of temperature and pressure
levels.
The purpose of the present study is to provide a set
of experiments for n-decane that can be presented with
an overall ignition delay correlation (Lifshitz formula)
in order to compare it to other hydrocarbons on similar
basis. The second purpose is to present a mechanism
that will fit the experimental data. As it is well known,
the single pulse shock tube experiments are the closest
possible representation of the PSR (perfect stirred re-
actor) concept. Any other kind of combustion kinetic
experiment such as flame measurements and even plug
flow reactors includes some hard-to-define nonchem-
ical parameters such as flow, heat transfer, and wall
reactions.
The sampling section of 0.25 long contains three
Kistler 603A piezoelectric transducers. Two transduc-
ers 0.20 m apart measure the shock velocity from
which the temperature is calculated. A third piezoelec-
tric gauge, located on the end plate, records the ignition
delay time. The pressure outputs are fed to digital oscil-
loscopes that record 12-bit traces of 4098 points, each
at 1 ꢀs intervals.
The reflected shock temperatures were calculated
using standard conservation equations and the ideal gas
equation of state assuming frozen chemistry.
Thermodynamic data for all the species were
taken from Burcat’s compilation [13]. The temperature
uncertainty error was 20 K.
The materials for these experiments were n-decane
product of Aldrich 99+% pure. The oxygen and argon
of pure quality (98%) were purchased from Oxygen
Stores, Haifa. The materials were used as it is with-
out further purification. Twenty liter heated glass bulbs
were used to store the mixtures. The mixtures were
left for more than 24 h to mix. These mixtures served
as stock for the individual experiments. The analysis
of postshock gases was performed by taking samples
of the gas into preevacuated glass bulbs of ∼100 cc
and analyzed on a Hewlett-Packard 5900 model gas
chromatograph equipped with a flame ionization de-
tector and containing a 6 feet long PORAPACK N
column and a PACKARD 427 gas chromatograph
withaPORAPACK Qcolumn. APerkin-Elmer/Nelson
Analytical PC Integrator was connected to the flame
ionization detector of both instruments. The gas inlets
and the mixture in the bulb were heated during the
analyses with the PACKARD 427 instrument.
Some researchers claim that a mechanism is better if
it fits as many experimental systems as possible. In the
end, they represent a mechanism that is a compromise
among the different available results. The question is
“What is the purpose of such a mechanism?” Our idea
is to obtain the best simulation of a PSR experiment in
order to obtain the best description of PSR conditions.
RESULTS
EXPERIMENTAL
The oxidation of decane was investigated by mea-
suring the ignition delay of n-decane–oxygen–argon
blends in a single pulse shock tube. End plate pres-
sure traces were used to measure the ignition delay as
defined above. Mixtures of 0.5–1.5 mol% decane and
4.2–23 mol% oxygen diluted in argon were used at
pressures between 1.8 and 9.4 atm. Table I is a list of
the mixtures and conditions used in the experiments
as well as representative shocks from each mixture.
In Fig. 1, a least square statistical analysis of all the
144 experiments is presented in order for them to sat-
isfy the Lifshitz equation [10] of the ignition delay
time
Ignition delay experiments were performed behind re-
flected shock waves in a heated single pulse shock tube.
This instrument serves as a heater in the millisecond
range, where the walls remain at constant temperature
(100◦C in our case).
Ignition delay times, as measured in this shock-tube
study, are the time intervals between the rapid rise of
the pressure due to the shock arrival and the onset of
the pressure due to the ignition.
The shock tube is made of a 4 m long stainless
steel tube. The driven section is 2.5 m long. Mylar
diaphragms of different gauges were burst to gener-
ate the shock wave. Maximal ignition delay and cool-
ing times behind the reflected shocks were 2 ms. The
aerodynamic cooling achieved was ∼106 K/s.
ꢀ
ꢁ
E
τ (s) = A exp
+
[C10H22]a[O2]b[Ar]c
RT
International Journal of Chemical Kinetics DOI 10.1002/kin

DECANE OXIDATION IN A SHOCK TUBE
705
Table I Representative Experiments of n-Decane–Oxygen–Argon Mixtures
Mixture No. Symbol [Dec] (%) [O2] (%) Num Exp.a Num Stat.b Experiment No. P1 (Torr) P5 (atm) T5 (K) τ (ꢀs)
1
ꢀ
1.5
23.25
31
20
8
15
29
35
40
43
49
55
59
66
68
96
99
106
111
114
117
124
128
135
139
145
150
157
159
163
167
177
182
197
51
70
70
251
240
71
2.70
3.24
3.82
7.25
6.25
2.57
1.82
7.01
9.29
3.13
2.37
7.57
8.09
1.81
2.28
6.63
9.11
2.31
3.17
8.2
1448
1362
1478
1384
1282
1532
1282
1270
1473
1592
1352
1432
1492
1291
1471
1239
1463
1343
1616
1427
1594
1560
1418
1481
1268
1523
1284
1452
1237
1535
60
118
36
101
314
66
657
299
35
ꢁ
h
2
0.52
8.4
19
17
ꢀꢀ
70
3
5
6
7
8
9
+
0.82
0.49
0.74
0.81
0.75
0.78
11.0
7.5
20
20
20
20
20
27
18
17
15
17
18
22
241
240
70
41
258
82
72
ꢀ
241
239
70
45
667
120
321
30
263
21
216
70
180
641
60
496
110
792
58
490
25
69
∇
12.7
4.16
7.40
17.6
242
241
72
70
r
242
241
71
10.0
2.84
2.28
8.86
6.60
2.71
1.96
2.73
2.03
3.08
68
ꢁ
241
241
70
70
70
ꢂ
70
70
a The actual number of experiments performed.
b The number of experiments used for the statistical analysis. The rest were discarded.
The overall ignition delay of decane based on a σ = 2
specie at temperature T normalized by the initial
decane concentration ([Specie]T /[C10H22]0) versus
1/T .
0.2
spread of 144 experiments is τ = 10−12
exp(+34240/RT )[C10H22]0.60 0.06[O2]−1.305
0.042
0.05
[Ar]0.08
s. The overall ignition delay of de-
cane based on a σ = 3 spread of 168 experiments is
0.09
τ = 10−11.9 0.4 exp(+34600/RT )[C10H22 ]0.66
MODELING AND CALCULATIONS
[O2]−1.33 0.06[Ar]0.055
s.
0.068
Some experiments were conducted so that the
quenching rarefaction wave arrived before ignition oc-
curred. These experiments were called “preignition”
shocks. The postshock preignition samples of gas were
gathered and analyzed for products in a gas chromato-
graph. The postshock products detected were CO2,
CH4, C2H4, C2H6, C3H6, C4H6, 1-C4H8, 1-C5H10,
1-C6H12, 1-C7H14, and 1-C8H16. No higher prod-
ucts than C8 were detected except for decane itself.
In Figs. 2 and 3, the different intermediate oxida-
tion ingredients found in the preignition gas chro-
matographic analysis are shown as concentration of
The modeling of n-decane worth mentioning is the
Westbrook’s[2], Battin-Leclerc’s[4,5], Lindstedt’s[6],
and Wilson’s [9]. We have tried most of these mod-
els one way or another and found that all of them
do not fit our experiments and give calculations that
are far lower from the main experimental line and/or
have a steeper activation energy. Of special interest is
the work of Battin-Leclerc and coworkers [4,5], which
was achieved using an automatic mechanism genera-
tor Exgas [3]. Some researchers envision such com-
puter programs as achieving final/absolute answers to
kinetic problems.
International Journal of Chemical Kinetics DOI 10.1002/kin

706
OLCHANSKI AND BURCAT
ꢁ
ꢂ
C2H3 + O2
CH O + CHO,
2
ꢀ
ꢁ
1015
k = 4.58 × 1016T −1.39 exp −
RT
ꢁ
ꢂ
C2H4 + CH3
C H + CH ,
2
2
4
ꢀ
ꢁ
9200
k = 2.27 × 105T 2.0 exp −
RT
The second part of the decane-ignition scheme includes
reactions of hydrocarbons from C6 to C22 that are pre-
sented in Table II.
As initiation reactions in our kinetic scheme, we
have considered the following steps of n-decane
pyrolysis:
ꢁ
ꢂ
C10H22
C10H22
C10H22
C10H22
C H + C H
2
5
8
17
15
13
ꢁ
ꢂ
C H + C H
3
7
7
ꢁ
ꢂ
C H + C H
4
9
6
Figure 1 Ignition delay times of 144 experiments with
the notation from Table I. The statistically reduced least
squares satisfy the equation τ = 10−12 exp(+34240/RT )
ꢁ
ꢂ
C H11 + C5H11
5
[C10H22]
0.60[O2]−1.305[Ar]0.08 s.
with kinetic parameters close to Bikas and Peters [8]:
activation energy (81–82 kcal/mol) and preexponential
factor multiplied by 1.5–2.5 values. It should be no-
ticed that the activation energy in Battin-Leclerc and
coworkers [4,5] had higher values ∼83–87 kcal/mol
whereasLindstedtandMaurice[6]usedamuchsmaller
value for the decane pyrolysis stages, in the range of
∼68 kcal/mol. The initiation also includes the forma-
tion of three different decyl radicals (reactions 5–7 in
Table II).
The kinetic parameters of H-abstraction reactions,
followed by decyl radical formation, are proposed in
our mechanism mainly by analogy for n-heptane oxi-
dation used by Chakir et al. [18]. Some corrections are
made only on the preexponential factor (see Table II)
while the activation energy and n-parameter for tem-
perature remained without changes.
As explained in [18], the n-heptane possesses four
distinguishable H-atom sites; therefore, four heptyl
radicals are produced by H atom abstraction. The
rate constants of these reactions depend on the na-
ture of the H-abstraction site and the number of the
hydrogen atoms present in the site. As for n-pentane
H-abstraction reactions [19a], the kinetic data used
in this case were derived from recommendations of
Tamura [19b].
The problem with this automatic reaction generator
is that it gives too much attention to all the possibili-
ties of radical reactions whether they exist or not. This
causes the models to be blown up beyond proportions
to thousands of reactions that cannot be followed by
humans because of the human limitations. The conse-
quences are obvious.
The proposed mechanism was built on previous pro-
posed schemes. To keep the mechanism in manageable
conditions, we had to reduce the number of possible
isomer products, and to adjust some of the rates so
that the final scheme will not only predict the exper-
imental Lifshitz equation of n-decane but also fit the
experimental concentration of products (Figs. 2 and 3).
Equations were chosen for rate change following sensi-
tivity analysis tests that were conducted at a multitude
of temperature and conditions.
Our kinetic scheme contains 433 reactions and 69
species. The first 270 reactions are from n-pentane
oxidation model [15], which is basically Westbrook’s
scheme [16,17]. We eliminated 2- and 3-pentyl radi-
cals and 2-pentene from this mechanism as explained
above. Only 1-pentyl and 1-pentene remained in our
n-decane mechanism. In the following reactions of
C–C2 submechanism, theparametersfromthenewGRI
(ver. 3.0) are used:
ꢁ
In [18], for n-C7H16 + H 1-C H15 + H2, A =
ꢂ
7
2.81 × 107 is given. For n-C7H16 + H 2-C H
+
ꢁ
ꢂ
7
15
H2, A = 0.91 × 107 is given. The same is for 3-C7H15,
and for 4-C7H15, A = 4.5 × 106 is given. These val-
ues do not follow exactly the rules of Tamura, because
they were fit for the [18] model. H abstraction by other
ꢁ
ꢂ
H + O2
O + OH,
ꢀ
ꢁ
17040
k = 2.5 × 1016T −0.67 exp −
RT
International Journal of Chemical Kinetics DOI 10.1002/kin

DECANE OXIDATION IN A SHOCK TUBE
707
Figure 2 Product distribution of CH4, C2H4, C3H6, and C4H8 species vs. calculated parameters from our model represented
by the solid inverted triangles and the least square line. Concentration of specie at temperature T normalized by the initial decane
concentration. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
radicals in their paper (CH3, HO2, C2H5) follow ex-
actly Tamura’s recommendations (A-factor according
to the number of the H-atom in every abstraction site).
Therearefivedifferentdecyl-radicals, whichmaybe
formed in n-decane oxidation, although in our work we
consider as explained above only three of them, which
we considered more important (n-C10H21, 2-C10H21,
and 3-C10H21).
In our mechanism, the same “n” and “Ea” were used
as in [18], becausetheH-abstraction siteforC10H22 and
C7H16 are similar, but the preexponential factor A for
1-C10H21 was slightly changed (A = 3.5 × 107). Taking
into account that the 4- and 5-decyls, that were omitted,
may increase the quantity of 2- and 3-decyls, the A-
factor for 2- and 3-decyl was doubled (0.91 × 107) × 2.
As a result A = 1.9 × 107.
The parameters for reactions of H-abstraction with
H-radical are taken from Battin-Leclerc and cowork-
ers [4,5]. There are great differences in kinetic pa-
rameters for H-abstraction reactions with CH3-radical
in the articles considered for n-decane oxidation
[4,5,8] and we used preexponential factors in our
scheme from [18], multiplying by 2. The preexponen-
tial factor for H-abstraction reaction by HO2-radical
with the formation of n-decyl is taken from [4,5]
whereas the values for 2-decyl and 3-decyl forma-
tion are very different in references [4,5] and [8],
namely −4 × 1012 and −1014, respectively. We used
A-factors for these reactions for corresponding radi-
cals by analogy to n-heptane [18].
The rate constants for isomerization of decyl-
radicals are taken according to [18], which originate
International Journal of Chemical Kinetics DOI 10.1002/kin

708
OLCHANSKI AND BURCAT
Figure 3 Product distribution of C5H10, C6H12, C7H14, and C8H16 species vs. calculated parameters from our model repre-
sented by the solid inverted triangles and the least square line. [Color figure can be viewed in the online issue, which is available
at www.interscience.wiley.com.]
in Westbrook’s work, and are very close to those of
Battin-Leclerc and coworkers [4,5].
H-abstraction reactions from decenes by H, O, OH,
and CH3-radicals are proposed to lead to only one
alkenyl radical (C10H19) for both decenes (C10H20 and
C10H20-2). This radical is supposed to decompose in
analogy to C7H13 in the n-heptane oxidation process.
The kinetic parameters for its decomposition into two
smaller parts are taken from the corresponding reac-
tions of C7H13 [18,20].
As we do not take into consideration decyl-radicals
on carbons 4 and 5 in our modeling, we suppress for-
mation of some intermediate species resulting from
these radicals. To compensate the formation of some
olefins, which appear in our oxidation experiments
before ignition at temperature 1020–1070 K, we
Inourscheme, itisproposedthatdecenesareformed
in reaction of decyl radicals with O2 followed by HO2
formation. We take into consideration, in order to re-
duce redundancy, only two decenes with double bond
on the first and second C-atoms. Kinetic parameters
for decene formation are taken by analogy for corre-
sponding heptyl reactions [18] with the correction of
the A-factor. The decenes are consumed by thermal de-
composition into smaller radicals and olefins and by ad-
ditive reactions with O and OH, which are followed by
rupture of C C bonds. The rates of these reactions for
bothgroupsaretakenfrom[18]for1,2-and3-heptenes.
International Journal of Chemical Kinetics DOI 10.1002/kin

DECANE OXIDATION IN A SHOCK TUBE
709
Table II Kinetic Scheme of C10–C6 Used in Our Mechanism
Reaction
A (mol/cm3)
N
Ea (kcal/mol)
Remarks
ꢁ
C10H22 C H + C H
6.50E+16
8.50E+16
4.50E+16
4.3E+16
0.0
0.0
0.0
0.0
0.0
0.0
0.0
2.0
2.0
2.0
2.4
2.5
2.5
0.97
1.61
1.61
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
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
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.25
82.0
81.0
82.0
81.0
100.0
100.0
100.0
7.7
5.0
5.0
1.59
5.0
5.0
1.59
0.0
0.0
11.6
9.5
9.5
19.4
17.0
17.0
49.0
47.6
47.6
28.8
28.3
29.1
33.0
40.4
40.4
40.4
40.4
11.1
18.1
20.0
2.0
*b, see text
*b, see text
*b, see text
*b, see text
*a-139
*a-140
*a-141
[3]
[3]
[3]
*a-12, A*3
*a-13
*a-14
*a-16
*a-17
*a-18
*a-20, A*2
*a-21, A*2
*a-22, A*2
*a-24
*a-25
*a-26
ꢂ
ꢁ
2
5
8
17
15
13
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
10H22 C H + C H
ꢂ
ꢁ
3
7
7
10H22 C H + C H
ꢂ
ꢁ
4
9
6
10H22 C H11 + C5H11
ꢂ
ꢁ
5
10H22
10H22
10H22
C
C
C
10H21 + H
1.00E+15
1.00E+15
1.00E+15
3.51E+07
1.91E+07
1.91E+07
7.60E+06
6.40E+05
6.40E+05
1.05E+10
4.70E+07
4.70E+07
6.00E+12
3.20E+12
3.20E+12
1.12E+13
6.80E+12
6.80E+12
2.51E+13
3.98E+13
4.00E+13
7.52E+12
1.00E+13
1.00E+13
8.0 E+12
1.00E+13
1.00E+13
1.00E+13
1.00E+13
2.00E+11
2.00E+11
2.00E+11
2.00E+12
2.00E+12
3.00E+12
3.00E+12
2.52E+16
1.60E+16
8.00E+13
4.00E+13
2.00E+13
2.00E+11
2.50E+12
2.82E+13
2.82E+13
2.82E+13
2.82E+13
2.82E+13
2.82E+13
5.00E+13
1.29E+09
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
10H21-2 + H
10H21-3 + H
ꢁ
ꢂ
10H22 + H
10H22 + H
10H22 + H
10H22 + O
10H22 + O
10H22 + O
10H22 + OH
10H22 + OH
10H22 + OH
C
C
C
C
C
C
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
10H21 + H2
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
10H21-2 + H2
10H21-3 + H2
10H21 + OH
10H21-2 + OH
10H21-3 + OH
ꢁ
ꢂ
C
C
C
C
C
10H21 + H2O
10H21-2 + H2O
10H21-3 + H2O
ꢁ
10H22 + CH3
10H22 + CH3
10H22 + CH3
10H22 + HO2
10H22 + HO2
10H22 + HO2
10H21 + CH4
ꢂ
ꢁ
10H21-2 + CH4
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
C
10H21-3 + CH4
C
C
C
10H21 + H2O2
10H21-2 + H2O2
10H21-3 + H2O2
10H22 + O2
10H22 + O2
10H22 + O2
ꢁ
C
C
C
10H21 + HO2
[3]
10H21-2 + HO2
Est, see text
Est
10H21-3 + HO2
10H21 C H + C H
[3], A/3
[3], A/3
[3], A/3
[3], A/3
*a-57
*a-58
*a-59
*a-60
*a-63
ꢂ
2
4
8 17
ꢁ
10H21-2 C H + C H
ꢂ
ꢁ
3
6
7
15
10H21-3 C H + C H
ꢂ
ꢁ
4
8
6
13
10H21-3 C H18 + CH3
ꢂ
9
ꢁ
ꢂ
10H21
C
10H20 + H
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
10H21-2
10H21-2
10H21-3
C
C
C
10H20 + H
10H20-2 + H
10H20-2 + H
ꢁ
10H21
10H21
C
C
10H21-2
10H21-3
ꢂ
ꢁ
*a-64
*a-66
ꢂ
ꢁ
ꢂ
10H21-2
C
10H21-3
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
10H21 + O2
10H21-2 + O2
10H21-2 + O2
10H21-3 + O2
ꢁ
C
10H20 + HO2
*a-67, A*2
*a-68, A*2
*a-69, A*1.5
*a-70, A*1.5
*a-73
*a-74
*a-77, see text
*a-80
*a-83
*a-86
*a-89, A/2
*a-92, see text
*a-93
*a-94
*a-93
*a-93
*a-93
*a-95
*a-102
C
C
C
10H20 + HO2
4.5
10H20-2 + HO2
10H20-2 + HO2
4.25
4.25
71.1
69.3
3.4
4.0
2.6
6.8
0.0
5.2
5.2
5.2
5.2
5.2
5.2
7.8
0.7
10H20 C H15 + C3H5
ꢂ
7
ꢁ
10H20-2 C H13 + C4H7
ꢂ
10H20 + H
10H20 + O
6
C
C
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
10H19 + H2
10H19 + OH
C10H20 + OH
C
C
C
10H19 + H2O
C
C
C
C
C
C
C
C
C
C
10H20 + CH3
10H20 + HO2
ꢁ
10H19 + CH4
ꢁ
ꢂ
10H19 + H2O2
10H20 + O C H + C H16 + OH
ꢂ
2
3
8
ꢁ
10H20 + O C H + C H14 + OH
ꢂ
ꢁ
3
5
7
10H20 + O C H + C H12 + OH
ꢂ
ꢁ
4
7
6
10H20 + O C H + C H + OH
ꢂ
ꢁ
5
9
5 10
10H20 + O C H + C H + OH
ꢂ
ꢁ
4
8
6 11
10H20 + O C H + C H + OH
ꢂ
ꢁ
3
6
7 13
10H20 + O C H + C H + OH
ꢂ
2
4
8 15
ꢁ
ꢂ
10H20 + OH C H + C H14 + H2O
3
5
7
Continued
International Journal of Chemical Kinetics DOI 10.1002/kin

710
OLCHANSKI AND BURCAT
Table II Continued
Reaction
A (mol/cm3)
N
Ea (kcal/mol)
Remarks
*a-103
ꢁ
C10H20 + OH C H + C H12 + H2O
1.29E+09
1.29E+09
1.29E+09
1.29E+09
1.29E+09
4.27E+09
0.80E+14
4.00E+13
2.00E+13
2.00E+11
2.50E+12
2.82E+13
2.82E+13
2.82E+13
1.16E+09
1.16E+09
1.16E+09
2.1E+06
2.1E+06
3.0E+06
3.0E+06
2.5E+05
2.5E+05
0.2E+04
0.16E+04
1.0E+13
0.5E+13
7.8E+01
7.8E+01
2.82E+13
2.52E+13
1.0E+13
2.0E+16
8.0E+13
4.00E+13
2.00E+13
2.00E+11
2.50E+12
2.82E+13
2.82E+13
2.82E+13
2.82E+13
1.29E+09
1.29E+09
1.29E+09
1.29E+09
2.0E+12
2.0E+12
2.0E+12
2.52E+13
1.00E+12
1.00E+13
2E+16
1.25
1.05
1.25
1.25
1.25
1.05
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.25
1.25
1.25
2.4
2.4
2.0
2.0
2.6
2.6
2.6
2.6
0.0
0.0
3.3
3.3
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
1.25
1.25
1.25
1.25
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.7
0.7
0.7
0.7
0.7
1.81
3.4
4.0
2.6
6.8
0.0
5.2
5.2
5.2
0.7
0.7
0.7
4.47
4.47
−1.312
−1.312
1.91
1.91
13.9
13.9
47.6
47.6
18.16
18.16
5.2
30.0
32.0
68.0
3.4
4.0
2.6
6.8
0.0
5.2
5.2
5.2
5.2
0.7
0.7
ꢂ
ꢁ
4
7
6
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
10H20 + OH C H + C H + H2O
*a-104
*a-101
*a-103
*a-103
*a-107
*a-78, see text
*a-81
*a-84
*a-87
*a-90
*a-96
ꢂ
ꢁ
5
9
5 10
10H20 + OH C H + C H + H2O
ꢂ
ꢁ
2
3
8 16
10H20 + OH C H + C H + H2O
ꢂ
ꢁ
4
8
6 11
10H20 + OH C H + C H + H2O
ꢂ
ꢁ
3
6
7 13
10H20 + OH C H + C H + H2O
ꢂ
2
4
15
10H20-2 + H
10H20-2 + O
10H20-2 + OH
C
C
10H19 +8H2
ꢁ
ꢂ
ꢁ
ꢂ
ꢁ
ꢂ
10H19 + OH
C
10H19 + H2O
ꢁ
ꢂ
ꢁ
ꢂ
10H20-2 + CH3
10H20-2 + HO2
ꢁ
C
C
10H19 + CH4
10H19 + H2O2
10H20-2 + O C H + C4H8 + OH
ꢂ
6
11
ꢁ
10H20-2 + O C H + C3H6 + OH
*a-97
*a-97
ꢂ
ꢁ
7
13
10H20-2 + O C H + C5H9 + OH
ꢂ
5
10
ꢁ
10H20-2 + OH C H11 + C4H8 + H2O
*a-105
*a-106
*a-105
*d-260
*d-246
*d-262
*d-250
*d-261
*d-248
*d-264
*d-255
*d-263, A/4
*d-253, A/2
*d-266
*d-266
*a-97
ꢂ
ꢁ
6
10H20-2 + OH C H + C3H6 + H2O
ꢂ
ꢁ
7
13
10H20-2 + OH C H + C5H9 + H2O
ꢂ
10H22 + H → H2 +5C5H11 + C5H10
10
10H22 + H → H2 + C2H5 + C8H16
10H22 + OH → H2O + C5H11 + C5H10
10H22 + OH → H2O + C2H5 + C8H16
10H22 + O → OH + C5H11 + C5H10
10H22 + O → OH + C2H5 + C8H16
10H22 + HO2 → H2O2 + C5H11 + C5H10
10H22 + HO2 → H2O2 + C2H5 + C8H16
10H22 + O2 → HO2 + C5H11 + C5H10
10H22 + O2 → HO2 + C2H5 + C8H16
10H22 + C3H5 → C3H6 + C2H5 + C8H16
10H22 + C3H5 → C3H6 + C5H11 + C5H10
ꢁ
ꢂ
10H20-2 + O C H + C H14 + OH
3
5
7
ꢁ
10H19 C H + C H
*a-110
*a-112
ꢂ
ꢁ
3
5
7
14
13
10H19 C H + C H
ꢂ
4
6
6
ꢁ
ꢂ
C9H18 C H + C H
4
9
5 9
ꢁ
C9H18 + H C H17 + H2
*a-77
*a-80
*a-83
*a-86
*a-89, A/2
*a-92
*a-93
*a-94
*a-93
*a-102
*a-103
*a-102
*a-101
*a-138
*a-138
*a-138
[3]
*a-67
*a-57
*d-260 to 1-C6H12
*d-210
*d-212
ꢂ
ꢁ
9
C9H18 + O C H + OH
ꢂ
9
17
ꢁ
ꢂ
C9H18 + OH C H17 + H2O
9
ꢁ
C9H18 + CH3 C H17 + CH4
ꢂ
ꢁ
9
C9H18 + HO2 C H + H2O2
ꢂ
9
17
ꢁ
C9H18 + O C H + C H14 + OH
ꢂ
ꢁ
2
3
7
C9H18 + O C H + C H12 + OH
ꢂ
ꢁ
3
5
6
C9H18 + O C H + C H + OH
ꢂ
ꢁ
4
7
5 10
C9H18 + O C H + C H + OH
ꢂ
5
9
4 8
ꢁ
C9H18 + OH C H + C H12 + H2O
ꢂ
ꢁ
3
5
6
C9H18 + OH C H + C H + H2O
ꢂ
ꢁ
4
7
5 10
C9H18 + OH C H + C H + H O
0.7
0.7
ꢂ
ꢁ
5
9
4
8
2
C9H18 + OH C H + C H14 + H2O
ꢂ
2
3
7
ꢁ
C9H17 C H + C H
30.0
30.0
30.0
28.8
2.0
40.4
71.0
57.0
1.302
ꢂ
ꢁ
4
7
5
10
12
14
13
C9H17 C H + C H
ꢂ
ꢁ
3
5
6
C9H17 C H + C H
ꢂ
ꢁ
2
3
7
C8H17 C H + C H
ꢂ
2
4
ꢁ
ꢂ
6
C8H17 + O2 C H16 + HO2
8
ꢁ
C8H17 C H16 + H
ꢂ
ꢁ
8
C8H16 C H + C H
ꢂ
ꢁ
3
5
5
11
C8H16 C H + C H
4E+14
ꢂ
C8H16 + H →8C5H140 + C2H4 + CH3
7.23E+12
4
8
Continued
International Journal of Chemical Kinetics DOI 10.1002/kin

DECANE OXIDATION IN A SHOCK TUBE
711
Table II Continued
Reaction
A (mol/cm3)
N
Ea (kcal/mol)
Remarks
C8H16 + H → H2 + C4H6 + C4H9
C8H16 + O → OH + C3H6 + C5H9
C8H16 + O → OH + C2H3 + C6H12
C8H16 + OH → H2O + C4H6 + C4H9
1.15E+05
4.76E+04
1.3E+06
2.08E+06
2.5E+12
2.0E+12
2.0E+13
8.0E+13
2.0E+11
1.0E+11
1.0E+11
2.52E+15
4.00E+13
1.60E+15
1.0E+12
2.8E+13
2.8E+13
2.8E+13
5.0E+13
6.5E+09
1.5E+09
1.29E+09
4.27E+09
1.60E+13
2.52E+13
1.0E+12
1.0E+13
2.5E+13
1.0E+13
1.0E+13
1.0E+12
2.0E+13
8.0E+13
2.0E+11
1.0E+11
1.0E+11
1.0E+13
2.80E+13
4.00E+13
2.52E+16
1.0E+11
2.8E+13
5.0E+13
1.0E+11
1.0E+11
6.5E+09
1.5E+09
2.15E+10
1.50E+12
5.0E+12
5.0E+11
2.5
2.71
2.4
2.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.25
1.25
1.25
1.05
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
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.25
1.25
1.05
0.0
0.0
0.0
2.49
2.11
4.47
−0.3
30.0
30.0
2.6
3.4
6.8
0.0
0.0
71.1
4.0
69.3
0.0
5.2
5.2
5.2
7.85
0.7
0.7
0.7
1.81
28.3
28.8
2.0
40.4
30.0
32.0
38.11
2.0
2.6
3.4
6.8
0.0
*d-214
*d-220
*d-216
*d-222
*a-138, A/2
*a-138, A/2
[19]-88
[19]-82
[19]-91
*a-135
*a-136
[18]-73, A/10
[19]-85
[18]-74, A/10
[18], A/5
[19]-97
ꢁ
C8H15 C H + C H
ꢂ
2
3
6
12
ꢁ
C8H15 C H + C H
ꢂ
3
5
ꢁ
ꢂ
5
10
C7H14 + OH C H13 + H2O
7
ꢁ
ꢂ
C7H14 + H C H13 + H2
7
ꢁ
ꢂ
C7H14 + CH3 C H13 + CH4
7
ꢁ
C7H14 + OH C H13 + CH2O
ꢂ
ꢁ
6
C7H14 + OH C H11 + CH3CHO
ꢂ
5
ꢁ
C7H14 C H + C H
ꢂ
3
5
4 9
ꢁ
ꢂ
C7H14 + O C H13 + OH
7
ꢁ
C7H14 C H + C H
ꢂ
3
7
4 7
ꢁ
ꢂ
C7H14 + HO2 C H13 + H2O2
7
ꢁ
C7H14 + O C H + C H10 + OH
ꢂ
ꢁ
2
3
5
C7H14 + O C H + C H + OH
[19]-98
[19]-99
ꢂ
ꢁ
3
5
4 8
C7H14 + O C H + C H + OH
ꢂ
ꢁ
4
7
3 6
C7H14 + O C H + C H + OH
[19]-100
[19]-106
[19]-107
[19]-108
[19]-109
[18]-53
[4]
[19]-130
[18]-57
[18]-110
[18]-112
[18]-113
[8]-313
[18]-124
[18]-122
[18]-123
[18]-135
[18]-136
[18]-121, A/5
[8]-307
ꢂ
5
9
2 4
ꢁ
C7H14 + OH C H + C H10 + H2O
ꢂ
ꢁ
2
3
5
C7H14 + OH C H + C H + H O
ꢂ
ꢁ
3
5
4
8
2
C7H14 + OH C H + C H + H O
ꢂ
ꢁ
4
7
3
6
2
C7H14 + OH C H + C H + H O
ꢂ
5
9
2
4
2
ꢁ
C7H15 C H + C H
ꢂ
ꢁ
3
6
4
9
C7H15 C H + C H
ꢂ
2
4
5
11
ꢁ
ꢂ
C7H15 + O2 C H14 + HO2
7
ꢁ
C7H15 C H14 + H
ꢂ
ꢁ
7
C7H13 C H + C H
ꢂ
ꢁ
3
5
4
8
C7H13 C H + C H
ꢂ
ꢁ
4
6
3
7
C6H13 C H12 + H
ꢂ
6
ꢁ
ꢂ
C6H13 + O2 C H12 + HO2
6
ꢁ
ꢂ
C6H12 + OH C H11 + H2O
6
ꢁ
ꢂ
C6H12 + H C H11 + H2
6
ꢁ
ꢂ
C6H12 + CH3 C H11 + CH4
6
ꢁ
C6H12 + OH C H11 + CH2O
ꢂ
ꢁ
5
C6H12 + OH C H + CH CHO
0.0
56.80
5.2
ꢂ
4
9
3
ꢁ
C6H12 C H + C H
ꢂ
3
6
3 6
ꢁ
C6H12 + O C H + C H + OH
ꢂ
ꢁ
3
5
3 6
C6H12 + O C H11 + OH
4.0
[18]-123
[8]-302
[18]-126
[8]-308
ꢂ
6
ꢁ
C6H12 C H + C H
71.1
17.0
5.2
7.85
0.0
0.0
0.7
0.7
1.81
28.8
30.0
32.0
ꢂ
3
7
3 5
ꢁ
ꢂ
C6H12 + HO2 C H11 + H2O2
6
ꢁ
C6H12 + O C H + C H + OH
ꢂ
ꢁ
2
3
4 8
C6H12 + O C H + C H + OH
[8]-306
ꢂ
ꢁ
4
7
2
4
C6H12 + O CHO + C H
[18]-130
[18]-131
[18]-132
[18]-133, A/4
[18]-134
[4], A/10
[18]-137, A/10
[18]-138, A/10
ꢂ
ꢁ
5
11
C6H12 + O CH + CO + C H
ꢂ
3
4 9
ꢁ
C6H12 + OH C H + C H + H O
ꢂ
ꢁ
2
3
4
8
2
C6H12 + OH C H + C H + H O
ꢂ
ꢁ
3
5
3
6
2
C6H12 + OH C H + C H + H O
ꢂ
4
7
2
4
2
ꢁ
C6H13 C H + C H
ꢂ
ꢁ
4
9
2
4
6
6
C6H11 C H + C H
ꢂ
ꢁ
3
5
3
C6H11 C H + C H
ꢂ
2
5
4
*a: in analogy to reaction number written in ref. [18].
*b: in analogy to reaction number written in ref. [8].
*c: in analogy to reaction number written in ref. [4,5].
*d: in analogy to reaction number written in ref. [20].
International Journal of Chemical Kinetics DOI 10.1002/kin

712
OLCHANSKI AND BURCAT
were kept, according to [6]∗. The reaction of C6 and C7
were taken from their previous work on n-heptane com-
bustion [20]. These reactions were combined with the
C3–C5 submechanism of our scheme with correspond-
ing necessary changes in the A-parameter dimensions.
The oxidation of small hydrocarbons (C–C2) was pre-
sented in two variants: the first was GRI (Fig. 4, line 3)
and the second was Westbrook’s original scheme, with
changes in three reaction parameters as described in our
mechanism (Fig. 4, line 4). As it can be seen in Fig. 4,
the line 1 representing our model fits very well the ex-
periments of ignition delay time. Lindstedt and Mau-
rice scheme with C–C2 from Westbrook [16,17] (Fig. 4,
line 4) is parallel slightly below the experimental data
and our model line. Although the difference is rela-
tively small for ignition delay time experiments, at low
temperatures we get large differences with Lindstedt–
Maurice scheme in oxidation intermediate product
species; thus, the concentration of small hydrocarbons
is by 1–2 orders larger than the experimental data. Our
mechanism fits sufficiently well the experimental data
for oxidation products, as shown in Figs. 2 and 3.
In Fig. 5, we present a comparison of our model and
the experimental ignition delay equation as a function
Figure 4 Comparison of experimental data with different
mechanisms taken from the literature. (1) Our model; (2)
Battin-Leclerc’s (Exgas) model, see text; (3) Lindstedt and
Maurice + GRI model; (4) Lindstedt and Maurice and West-
brook’s model with changes of three of Westbrook’s para-
meters. [Color figure can be viewed in the online issue, which
is available at www.interscience.wiley.com.]
have included several semiempirical reactions into our
mechanism, proposed earlier for n-heptane [20]. In
these reactions, the decane-molecule reacts with radi-
cals H, OH, and HO2, which abstract H from alkane,
and then the alkenyl-radical breaks into an olefin and
a radical. Kinetic parameters for these semiempirical
reactions are taken by analogy to n-heptane transfor-
mation through 3,4-heptyl [21].
Earlier proposed schemes are compared with our
experimental data as shown in Fig. 4.
We used three mechanisms to validate our experi-
mental data: Battin-Leclerc’s scheme [4,5], Lindstedt
and Maurice scheme [6], and our mechanism described
above. All the mechanisms were calculated using the
same thermodynamic database [13].
In Battin-Leclerc’s automatic scheme [5], we elim-
inated all reactions where peroxy-radicals or com-
pounds (including keto-peroxy) were formed, and
several hundreds reactions connected with different
radical isomerizations. We did that because these com-
pounds exist only at low-temperature cool flame con-
ditions but do not exist at our experimental conditions.
In the compared mechanism, 500 reactions remained
from the original of 1700+ reactions. This simulation
is shown as line 2 on Fig. 4.
Figure 5 Our calculated model compared with the over-
all experimental ignition delay equation. [Color figure
can be viewed in the online issue, which is available at
www.interscience.wiley.com.]
∗In the works of Lindstedt and Maurice [6,19], there is a question
about the dimensions of the A-parameter reported. They use one
value for their own reactions, but the reactions taken from other
papers are given as they appear in the original work, not according to
the dimensions listed in the tables. We have used the reactions listed
by Lindstedt and Maurice with their dimensions, thus, we multiplied
by 103 to adjust them to our dimensions.
In Lindstedt–Maurice scheme [6], all the reactions
with H-abstraction and formation of five decyl radicals,
decane, and decyl pyrolysis and decyl isomerization
International Journal of Chemical Kinetics DOI 10.1002/kin

DECANE OXIDATION IN A SHOCK TUBE
713
of τ, the ignition delay versus the fuel concentra-
tion. Here, we get a good estimation of the lean and
the rich limits of combustion of n-decane and we are
reminded again that the Lifshitz equation cannot be
used outside the experimental window, but the mod-
eling scheme will show us the correct behavior. The
correct width of the U-shaped line can be seen only
with a full mechanism like the one we used and any
attempt to reduce the mechanism means a narrower
U [15].
5. Battin-Leclerc, F.; Fournet, R.; Glaude, P. A.; Judenherc,
B.; Warth, V.; Coˆme, G. M.; Scacchi, G. Proc Combust
Inst 2000, 28, 1597–1605.
6. Lindstedt, R. P.;Maurice, L. Q. JPropulsionPower2000,
16, 187–195.
7. Zeppieri, S. P.; Klotz, S. D.; Dryer, F. L. Proc Combust
Inst 2000, 28, 1587–1595.
8. Bikas, G.; Peters, N. Combust Flame 2001, 126, 1456–
1475.
9. Eliott, I.; Ingham, D. B.; Kyne, A. G.; Mera, N. S.;
Pourkashanian, M.; Wilson, C. W. Proc Combust Inst
2004, 30, to be published.
10. Lifshitz, A. Ignition Delay Times in Handbook of
Shock-Waves; Academic Press: New York, 2001; Vol. 3,
pp. 211–256.
CONCLUSION
11. Horning, D. C.; Davidson, D. F.; Hanson R. K. J Propul-
sion Power 2001, 18, 363–371.
An experimental study of ignition of n-decane was
presented and compared with a number of proposed
kinetic mechanisms found in the literature. None of
the models fitted well our findings, and none of the
experimental studies from the literature could be com-
pared with ours because of differences of presentation.
In order to fit a mechanism to our experiments, many
changes in individual rates were necessary. Of spe-
cial disappointment was the Exgas automatic model
that turned out to be the least fitting to our experi-
ments.
12. Feiweger, K.; Pfahl, U.; Adomeit, G. In Proceedings
of the 20th International Symposium on Shock Waves,
Pasadena, CA, July 1995 (1996); Vol. 2, pp. 929–934.
13. Burcat, A. ftp://ftp.technion.ac.il/pub/supported/aetdd/
thermodynamics.
14. Steele, W. V.; Chirico, R. D.; Cowell, A. B.; Knipmeyer,
S. E.; Nguyen, A. J. Chem Eng Data 2002, 47, 667–688;
ibid 700–714; ibid 725–739.
15. Burcat, A.; Dvinyaninov, M. In Shock Waves@
Marseille, Vol. 2, Proccedings of 19th Symposium
(International) on Shock Wave Marseille, 1993; Burn,
R.; Dumitrescu, L. V. (Eds.); Springer-Verlag: Berlin,
1995, pp. 197–202.
16. Westbrook, C. K.; Pitz, W. J.; Thornton, M. M.; Malte,
P. C. Comb Flame 1988, 72, 45–62.
BIBLIOGRAPHY
17. Pitz, W. J.; Westbrook, C. K.; Leppard, W. R. SAE Trans-
action, Section 4, paper #912315, 1991.
18. Chakir, A.; Belliman, M.; Boettner, J. C; Cathonet, M.
Int J Chem Kinet 1992, 24, 385–410.
19. (a) Chakir, A.; Bellimam, M.; Cathonet, M.; Boettner,
J. C. Combust Sci Technol 1991, 77, 239; (b) Tamura,
T. M. Lawrence Livermore National Laboratory, Report,
1987.
20. Lindstedt, R. P.; Maurice, L. Q. Combust Sci Technol
1995, 107, 317–353.
1. Dagaut, P.; Reuillion, M.; Boettner, J. C.; Cathonnet, M.
Proc Combust Inst 1994, 25, 919–926.
2. Ranzi, E.; Sogaro, A.; Gaffuri, P.; Pennati, G.; West-
brook, C. K.; Pitz, W. J. Combust Flame 1994,
99, 201–211. See http://www-cms.llnl.gov/combustion/
combustion home.html.
3. Nehse, M.; Warnatz, J.; Chevalier, C. Proc Combust Inst
1996, 26, 773–780.
4. Glaude, P. A.; Warth, V.; Fournet, R.; Battin-Leclerc, F.;
Scacchi, G.; Coˆme, G. M. Int J Chem Kinet 1998, 30,
949–959.
21. Held, T. J.; Marchese, A. J.; Dryer, F. L. Combust Sci
Technol 1997, 123, 107–146.
International Journal of Chemical Kinetics DOI 10.1002/kin