Mechanism and rate of the reaction CH3+ O - Revisited

Article abstract of DOI:10.1039/b419137d

The primary products and the rate of the reaction of methyl radicals with oxygen atoms in the gas phase at room temperature have been studied using three different experimental arrangements: (A) laser flash photolysis to produce CH₃ and O from the precursors CH₃I and SO₂ (the educts and the products were detected by quantitative FTIR spectroscopy); (B) the coupling of a conventional discharge flow reactor via a molecular sampling system to a mass spectrometer with electron impact ionization, which allowed the determination of labile and stable species; (C) laser induced multiphoton ionization combined with a TOP mass spectrometer-molecular beam sampling-flow reactor, which was used for the specific and sensitive detection of the CH ₃, CD₃, C₂H₅ and C₂D ₅ radicals and the determination of rate coefficients. The branching ratio of the reaction channels was determined by the experimental arrangements (A) and (B) leading to CH₃ + O → HCHO + H (55 ± 5)% → CO + H₂ + H (45 ± 5)%. The rate coefficients of the normal and deuterated methyl and ethyl radicals with atomic oxygen showed no isotope effect: k(CD₃ + O)k(CH₃ + O) = 0.99 ± 0.12, k(C₂D₅ + O)k(C₂H₅ + O) = 1.01 ± 0.07 (statistical error, 95% confidence level). The absolute rate coefficient of the reaction CH₃ + O was derived with reference to the reaction C₂H₅ + O (k = 1.04 × 10¹⁴ cm³ mol^-1 s^-1) leading to k(CH₃ + O) = (7.6 ± 1.4) × 10¹³ cm³ mol^-1 s^-1. The Owner Societies 2005.

Full text of DOI:10.1039/b419137d

View Article Online / Journal Homepage / Table of Contents for this issue
R E S E A R C H P A P E R
Mechanism and rate of the reaction CH þ O—revisited
3
a
b
b
b
b
W. Hack, M. Hold, K. Hoyermann, J. Wehmeyer and T. Zeuch*
a
Max-Planck-Institut fu¨r Biophysikalische Chemie, Go¨ttingen, Germany
Institut fu¨r Physikalische Chemie, Georg-August-Universita¨t, Go¨ttingen, Germany.
E-mail: jkupfer@gwdg.de
b
Received 21st December 2004, Accepted 15th March 2005
First published as an Advance Article on the web 7th April 2005
The primary products and the rate of the reaction of methyl radicals with oxygen atoms in the gas phase at room
temperature have been studied using three different experimental arrangements: (A) laser flash photolysis to
produce CH and O from the precursors CH I and SO (the educts and the products were detected by
3
3
2
quantitative FTIR spectroscopy); (B) the coupling of a conventional discharge flow reactor via a molecular
sampling system to a mass spectrometer with electron impact ionization, which allowed the determination of
labile and stable species; (C) laser induced multiphoton ionization combined with a TOF mass spectrometer–
molecular beam sampling–flow reactor, which was used for the specific and sensitive detection of the CH , CD ,
3
3
C
2
H
5
and C
2 5
D radicals and the determination of rate coefficients. The branching ratio of the reaction channels
was determined by the experimental arrangements (A) and (B) leading to CH þ O - HCHO þ H (55 ꢀ 5)%
3
-
CO þ H
2
þ H (45 ꢀ 5)%. The rate coefficients of the normal and deuterated methyl and ethyl radicals with
atomic oxygen showed no isotope effect: k(CD þ O)/k(CH þ O) ¼ 0.99 ꢀ 0.12, k(C D þ O)/k(C H þ O) ¼
3
3
2
5
2
5
1
.01 ꢀ 0.07 (statistical error, 95% confidence level). The absolute rate coefficient of the reaction CH
3
þ O was
1
4
3
ꢂ1 ꢂ1
derived with reference to the reaction C H þ O (k ¼ 1.04 ꢁ 10 cm mol
s
) leading to k(CH þ O) ¼
2
5
3
1
3
3
ꢂ1 ꢂ1
(
7.6 ꢀ 1.4) ꢁ 10 cm mol
s .
1
and Preses in respect to the rate coefficient. Data compila-
0
14
Introduction
1
5,16
13
3
ꢂ1 ꢂ1
s .
tions
recommend k ¼ 8.4 ꢁ 10 cm mol
Methyl radicals are essential radicals in the oxidation of
hydrocarbons, which is obvious by considering different sce-
narios. Natural gas for heating purposes consists mainly of
methane, which is primarily attacked in an oxidation process
by atoms (H, O) and radicals (OH) to yield methyl radicals. At
high temperatures the global mechanism of the oxidation of
higher alkanes includes, as the main route, the thermal decom-
position of the higher alkyl radicals to give methyl and ethyl
radicals (see for example ref. 1). In a combustion process the
concentration of oxygen atoms is often well below that of
molecular oxygen but the rate coefficients of k(O þ CH , C H )
The importance of reaction (1) for a quantitative description
of the oxidation of alkanes, where the methane oxidation is the
main building block in the hierarchy, motivates additional
measurements of this fundamental reaction. Since all the
experimental techniques which have been used in previous
studies have their own virtues and deficiencies, an additional
study with the experimental arrangements discussed in the
experimental section can be considered as a useful and critical
test of the data reported so far.
The objectives of the present study are to elucidate the
reaction mechanism by measuring the primary products of
both channels (a ) and (a ) using different experimental ar-
1 2
rangements and to determine the rate coefficient of the overall
reaction (1).
3
2
5
may exceed those of k(O
magnitude. Therefore in flames, for example, the fate of the
methyl radical is strongly governed by the attack of the oxygen
2
þ CH
3 2 5
, C H ) by several orders of
atoms. Two aspects of the title reaction CH þ O are of vital
3
importance for the understanding and modeling of the com-
bustion of hydrocarbons: the primary products and the rate
coefficient of this reaction. Numerous theoretical and experi-
mental studies report on the mechanism
Experimental
Three independent experimental arrangements and procedures
have been applied. In order to keep the paper short only a brief
description is given here.
CH
3
þ O - CH
3
O*
(1)
Arrangement (A)
CH O* - HCHO þ H
(a1)
3
A laser flash photolysis cell (pressure: 3 mbar to 1 bar) with an
internal multiple path mirror arrangement is directly coupled
to a FTIR spectrometer for a quantitative concentration
-
CO þ H
2
þ H
2
(a )
and on the rate coefficient.
A short but incomplete summary is listed in Table 1.
Previous studies state an overwhelming contribution of (a
determination of the reactants. The labile species CH
are produced from the precursors CH I and SO by co-photo-
lysis at 193 nm. As the determination of the formed HCHO
3
and O
3
2
1
7,18
1
)
to the total reaction (495%, refs. 2–5), recent studies, how-
is crucial for the evaluation of the branching ratio and as the
absolute calibration of HCHO is susceptible to errors caused
by losses at the wall and by polymerization, an extensive
calibration program was performed. The consumed amount
of the precursor CH I was determined and all C atom-bearing
3
products were identified and calibrated. The recovery of the
ever, have reported lower contributions for (a ) and higher for
1
(
a
2
), (refs. 6, 7, 9–11, 13 and 15). The reported rate coefficients
as measured or evaluated otherwise show values between ca.
1
3
3
ꢂ1 ꢂ1
s . A more recent theoretical study
states agreement with the experimental studies of Fockenberg
(
7–11) ꢁ 10 cm mol
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 1 9 7 7 – 1 9 8 4
1977

View Article Online
Table 1 The reaction CH
3
þ O: previous results
Table 2 The reaction mechanism of CH3 þ O, data compiled from
refs. 15 and 25
3
Channel fraction k/cm mol
ꢂ1 ꢂ1
s
Channel fraction
HCHO (%)
Ref.
3 ꢂ1 ꢂ1
k/cm mol s
CO (%)
at 289 K
Reaction
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
3
3
3
3
3
2
2
2
0
1
2
0
3
2
3
3
4
95
00
00
00
—
—
—
2
3
4
5
6
7
9
CH
CH
CH
CH
CH
CH
CH
CH
CH
3
3
3
3
3
3
3
3
3
þ O
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
HCHO þ H
4.2 ꢁ 10
3.4 ꢁ 10
2.8 ꢁ 10
5.0 ꢁ 10
5.0 ꢁ 10
7.0 ꢁ 10
4.7 ꢁ 10
4.0 ꢁ 10
4.4 ꢁ 10
1.0 ꢁ 10
5.5 ꢁ 10
3.0 ꢁ 10
1.5 ꢁ 10
6.0 ꢁ 10
1.8 ꢁ 10
7.2 ꢁ 10
3.0 ꢁ 10
3.2 ꢁ 10
2.0 ꢁ 10
1.1 ꢁ 10
1.0 ꢁ 10
1.5 ꢁ 10
2.4 ꢁ 10
1
3
1
1
1
8.43 ꢁ 10 (ꢀ20%)
þ O
CO þ H
2
þ H
—
—
—
—
—
þ CH
3
C
2
H
6
þ H
CH
CH
CH
CH
CH
CH
4
3
3
3
3
2
—
—
—
40 ꢀ 10
17 ꢀ 11
18 ꢀ 4
15 ꢀ 6
45 ꢀ 6
—
þ OH
þ I
OH
I
14
1.07 ꢁ 10 (ꢀ17%)
—
I þ O
I þ H
I þ OH
þ IO
13
8
4 ꢀ 12
7.34 ꢁ 10 (ꢀ13%)
10
11
þ HI
13
5
5 ꢀ 6
8.00 ꢁ 10 (ꢀ15%)
I þ H
2
O
13
7.83 ꢁ 10 (ꢀ15%)
—
—
—
12
13
HCHO þ O
HCHO þ OH
HCHO þ H
OH þ CO þ H
2
14
—
—
1.02 ꢁ 10 (ꢀ20%)
H
O þ CO þ H
CO þ H þ H
CH
þ O
HCHO þ OH
13
9.26 ꢁ 10 (ꢀ10%)
14
15
2
13
100
60
—
8.43 ꢁ 10 (ca. ꢀ50%)
CH
CH
CH
3
3
3
O þ O
O þ O
O þ H
3
2
1
3
40
8.4 ꢁ 10
3,6,16
HCHO þ H
2
IO þ O
O
O
O
O
2
2
2
2
þ I
þ I
total carbon mass was 95–98%. Moreover, the absorption
cross sections of SO and CH I at the laser emission line
193.2 nm, half-width 1.3 nm) were carefully measured in our
12
12
13
12
IO þ IO
IO þ IO
OH þ O
OH þ OH
OH þ H
2
2
3
þ I þ I
þ H
(
reaction cell (known path length, laser-beam energy monitored
ꢂ2
H
2
O þ O
10
11
13
for different substance concentrations, SO : 252 mol m
ꢂ2
,
I: 48 mol m , see also ref. 18) in order to determine the
2
2
H O
CH
absolute O atom and CH radical start concentrations in the
3
OH þ C H
C H þ H O
2 4 2
2
6
2
5
2
3
CH I þ CH I
2
2
C H I
photolytic experiments. By using this arrangement a well
defined reaction system could be prepared and final product
concentrations could be modeled with a simple kinetic scheme.
(
(
3 2
CH I, SO ), of the primary products, and of the by-products
Arrangement (B)
C H , CH , C H , S O, (SO) ).
2
6
4
2
4
2
2
The hydrocarbons C
by-products are explained by the reactions listed in Table 2.
Their contributions were quantified at the characteristic wave-
2
H
6
4 2 4
, CH , C H and their formation as
A combination of two independently operated conventional
discharge flow reactors is connected to the ion source of a
magnetic deflection mass spectrometer with electron impact
ionization by a molecular beam sampling. The specific detec-
tion of labile and stable species was performed by the mass
spectrometry at low-energy electron impact ionization (4.5–
ꢂ1
length regions (CH around 3000 cm , rotational resolved;
4
ꢂ1
ꢂ1
ꢂ1
C
2
H
6
: 2950–3000 cm , rotational resolved, C
H
2 4
: 954 cm .)
A peak at 1160 cm is assigned to S O as peaks at 1169/679
2
ꢂ1
21
cm are known from calculations and the peak at 2060 is
2
9.5 eV); by adopting such a value an optimal signal-to-noise
also assigned to a sulfur-bearing by-product such as (SO) or
2
an overtone of S O.
2
ratio was achieved at low mass spectroscopic interference (i.e.
fragment ion formation). A sensitive detection is achieved by a
synchronous single ion counting technique, where the atom/
radical producing microwave discharge is switched ‘‘on/off’’,
thus preparing the condition of reaction ‘‘on/off’’. CH radicals
3
are formed via the reaction of F atoms with CH . F and O
4
Fig. 1 shows an example of the extracted product spectra of
CO (Fig. 1c) and HCHO (Fig. 1b) obtained by subtracting the
contributions of the calibrated by-product concentrations from
the reduced product spectrum (Fig. 1a). As illustrated in
Fig. 1d the net absorption in the fingerprint region of C–C
atoms are produced directly by discharging highly diluted
1
ꢂ1
9
compounds (around 850–1150 cm ) levels off indicating that
2 2
F /He and O /He mixtures.
these compounds were correctly taken into account. The
deduced CO spectrum from the overall spectrum matches the
pure CO spectrum. The CO yield was determined by recording
spectra for each pressure relevant to the experimental condi-
tions and the product spectra were compared to this spectra–
concentration set. By this, saturation effects due to the limited
spectral resolution of the rotational CO-bands are avoided.
The extracted HCHO spectrum from the measurements (Fig.
Arrangement (C)
Rate coefficients were determined by a combination of dis-
charge flow reactors and a TOF-mass spectrometer where the
radicals are detected specifically and sensitively via the laser
induced multiphoton ionization. Chemicals were of commer-
cial grade: (He( Z 99.995%), Ar ( Z 99.6%, SO ( Z 99.98%),
2
0
2
1
line strength of a HCHO spectrum reported in the literature
b) coincides perfectly with the features of spectral position and
22
CO( Z 98%), CH4 ( Z 99.995%), O2 ( Z 99.995%), He/F2
(
3
99.995%/98%), all Messer-Griesheim; CFCl ( Z 99.5%),
ꢂ1 ꢂ1 ꢂ1 ꢂ1
(
2843 cm , 2783 cm , 1746 cm , 1500 cm ). This is taken as
CH I ( Z 99%), Fluka; CD ( Z 98 atom% D), C D ( Z 99
3
4
2
6
evidence that all carbon containing species were correctly as-
signed and quantified. Therefore the absolute HCHO concen-
tration was derived from the total carbon mass balance, leading
to an experimental calibration factor of the HCHO concentra-
tion. It is noteworthy to state that this calibration factor has
been applied also to the determination of the HCHO yield of the
reaction C H þ O (2): C H + O - C H O*
atom% D), Aldrich).
Results and discussion
Primary products
The determination of the primary products by arrangement
ꢂ1
2
5
2
5
2
5
(
after the co-photolysis of CH I/SO
A) is based on the IR spectra in the region 600–5000 cm
mixtures. In order to
C H O* - HCHO + CH
3
(2a1)
3
2
2
5
unravel the spectra, characteristic regions with minimal spec-
tral interferences were chosen for identifying and quantifying
the contributions of the precursors and their consumption
-
CH CHO + H
3
(2a2)
C
H
2 5
+ O - C
2
H
4
+ OH
(2b)
1
978
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 1 9 7 7 – 1 9 8 4
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
Fig. 2 The reaction CH
CO with increasing conversion of the precursors (p ¼ 4.0 mbar, T ¼
3
þ O: formation of the products HCHO and
2
0
93 K, initial mixture strength of [CH ¼ 0.40/1.40; 0.60/0.20;
3
I]
0
/[SO
2
]
0
.60/0.30 mbar, [CH ] /[O] ratios were 0.05, 0.57, 0.38).
3
0
0
product analysis, the influence of consecutive and parallel
reactions on the branching ratio (a )/(a ) was studied by
1
2
modeling the total reaction system along the scheme of Table
. In Fig. 3 concentration/time profiles are given. Such simula-
2
tions with the initial experimental conditions (see Fig. 2) show
that the primary branching ratio (a )/(a ) at short reaction
times deviates even for the large initial excess of O atoms from
the measured one (‘‘final product analysis’’) by less than 2%,
which is well within the experimental error. Therefore reactions
such as the O atom or H atom attack on HCHO, which could
Fig. 1 The reaction CH
and CO as measured by FTIR spectroscopy; (b) and (c) extracted
spectra of HCHO and CO, respectively; (d) fingerprint region of C–C
3
þ O: (a) formation of the products HCHO
1
2
species (p ¼ 4.0 mbar, T ¼ 293 K, p(CH
3
I) ¼ 0.4 mbar, p(SO ) ¼ 1.4
2
mbar, p(Ar) ¼ 2.2 mbar, 200 laser shots, energy around 80 mJ per shot
when coupled to the cell, averaged over 200 scans).
1 2
potentially influence the ratio (a )/(a ), show no significant
impact under the chosen conditions and reactions of IO with
2
6
The branching ratio of the reaction channels HCHO þ CH
xa )/CH ) has been reported in the literature to
CHO þ H (xa
be: (xa )/(xa
) ¼ 0.80 (experimental), 0.91 (theoretical).
Our experimental value, based on the calibration factor
derived here, is (x(a ))/(x(a
)) ¼ 0.73. This agreement in the
HCHO formation for the reaction C
þ O gives indirect
support for the determination of the HCHO yield for the
reaction CH
þ O. In Figs. 2a, 2b and 2c the formation of
the products HCHO and CO is displayed as a function of the
consumption of the precursor CH I for different excess ratios
of the initial CH I and SO concentrations. No systematic
3
molecules are too slow to do so. It is noteworthy that also the
final concentrations of CH and C are correctly predicted
(
1
3
2
4
H
2 6
2
3
19
1
2
1
8
1
2
2
H
5
3
3
3
2
dependence of the product yield of HCHO and CO on the
applied initial concentration ratio [CH ] /[O] is observed, thus
an averaging of the product yields seems to be adequate
3
0
0
leading to (a )/(a ) ¼ 55.8/44.2.
1
2
The experiments were performed with small and large excess
of O over CH . As additional CH is quickly formed via
3
3
CH
3
I þ O - CH
3
þ IO, ref. 24, the actual conditions after
Fig. 3 The reaction CH
profiles of species assuming the reactions of Table 2 and the experi-
3
þ O: modeling of the concentration/time
5
an O atom excess for [CH
ms resulted in a CH excess for [CH I] /[SO ] ¼ 0.6/0.2 and
3 0 2 0
3
3
0
2 0
I]
determination of the branching ratio is based on the final
/[SO
]
¼ 0.4/1.4. Since the
mental conditions of Fig. 2 ([CH
3
I] /[SO
0
2
]
0
¼ 0.40 mbar/1.4 mbar,
ꢂ2
[CH ] ¼ 0.9 ꢁ 10 mbar, [O] ¼ 0.18 mbar).
3
0
0
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 1 9 7 7 – 1 9 8 4
1979

View Article Online
by the model for low and high initial O atom excess. Together
with the scatter of the measurements (statistical regression
errors, 95% confidence interval, each graph was taken and
averaged over the three experiments) this leads to error mar-
As already emphasized in the introduction our presented
experimental study is intended to add new experimental data
on the important reaction CH
speculate on the virtues and deficiencies of the experimental
techniques applied so far. Moreover, we do not feel qualified to
3
þ O and we hesitate to
gins of (a
By independent measurements via the experimental set up
1
)/(a
2
) ¼ (55.8 ꢀ 5.6)/(44.2 ꢀ 5.6).
5
,9
analyze the theoretical investigations published until now.
But we point to the high density of states of the chemical highly
(
B), the formation of CO together with the consumption of
CH radicals by O atoms was followed by time-resolved mass
activated CH O* making a precise theoretical quantification of
3
3
spectrometry, applying low absolute radical concentrations
ꢂ10
channel fractions a challenging problem (see refs. 9 and 14).
ꢂ3
(
o10
mol cm ) and small detection times (o2 ms). The
quantitative determination of the formation of CO was per-
formed at an ionization energy of 24 eV at m/z ¼ 28, leading to
a good compromise with respect to sensitivity, specifity and
Rate of reaction
The rate coefficient of the reaction of methyl radicals with
oxygen atoms was measured with reference to the reaction of
ethyl radicals with oxygen atoms, as the latter is reported as a
1
1
1
fragmentation (e.g. HCHO - HCO , CO ). Absolute CH
3
concentrations were derived from the controlled mass flow of
CH and the consumption by the F atoms. At the mass peak of
18
high quality rate coefficient.
The competitive method for the determination of the ratio of
the rate coefficients of the general type
4
CO (m/z ¼ 28) only a minor contribution of HCHO (o6%)
was observed. A contribution of C H was absent as C H
6
2
6
2
formation according to CH
3
þ CH
3
- C
2
H
6
was suppressed
by the low absolute CH concentrations. A small background
R þ O - products
(i)
1
3
R
2
þ O - products
can be accomplished by the simple relation
/ki0 ¼ ln {[R /[R ]}/ln{[R /[R
(i0)
signal at m/z ¼ 28 was attributed to air leaks (N
2
) and to CO
produced via CH þ O with O atoms from the discharge of the
3
2
(F /He) mixture with an air impurity. The extensive mass
k
i
1
]
0
1
2
]
0
2
]}
(1)
spectroscopic calibration procedures for the stable species are
based on Ar as an internal reference standard. In order to
illustrate the data reduction and the extent of corrections an
example is given in Fig. 4. The evaluation of the production of
CO per reacted CH3 radical with O atoms gave (þD[CO])/
where [R] and [R] are the concentrations of the radical R in
0
the absence and in the presence of oxygen atoms, respectively.
The precautions to be taken are the absence of other radical
removing reactions (radical combination and cross combina-
tion) and other radical forming reactions (products of the
radical and atomic oxygen reaction). Radical combination
reactions can be greatly suppressed when using very low
absolute radical concentrations as in the sensitive and specific
(
(
ꢂD[CH
3
]) ¼ (46 ꢀ 6)%. Thus the branching ratio of (a
2
)/
P
(a
i
)) ¼ (46 ꢀ 6)% is in good agreement with the results
obtained by the flash photolysis–FTIR study.
Our presented determination of the branching ratio (a
2
1
)/(a )
as obtained by the two independent experimental arrange-
ments differing in the reactor (static photolysis cell/flow reac-
tor) and the analytical tool (FTIR spectroscopy/mass
6
spectrometry), matches the results of Seakins and Leone.
2
Their value of 40% for channel (a ) was derived from time-
resolved FTIR emission spectrocopy of CO*. This technique
probes the nascent formation of products such as CO* or OH*
as convincingly demonstrated and discussed by Leone et al. in
6
,9,27,28
several papers.
tion of the CO formed via route (a
However, lower values for the contribu-
) were obtained from
2
experiments with other techniques ((17 ꢀ 11)%, flow reactor-
7
photoionization mass spectrometer (TOFMS-PI), (18 ꢀ 4)%,
8
photolysis cell-diode laser absorption, (15 ꢀ 6)%, flow reac-
1
0
9
tor-TOFMS-PI ). A theoretical study favours these lower
values for the efficiency of CO formation.
Fig. 4 The reaction CH
spectrometric arrangement (B). Ion signals at m/z ¼ 28 for the
following conditions in the flow reactor: (a) presence of CH radicals
discharge ‘‘on’’ for F - CH þ HF;
/He - F/He, reaction F þ CH
discharge ‘‘off’’ for O /He), ‘‘background’’ BKG; (b) presence of CH
radicals and O atoms (discharge ‘‘on’’ for F
/He - F/He, reaction F þ
CH - CH /He - O/He; (c) difference
of the ion signals of (b) ꢂ (a), ‘‘corrected for background signal’’; (d)
correction of signal (c) due to fragment ion formation of HCHO at
m/z ¼ 28, ‘‘net CO formation’’.
3
þ O: formation of CO as studied by the mass
Fig. 5 The reactions CH
the products by mass spectrometry after laser induced multiphoton
ionization. (a) CH
radical at l ¼ 450.6 nm (laser energy: 18 mJ; p ¼
1.35 mbar; T ¼ 298 K, x(CH ) ¼ 3.8 mol%, x(F ) ¼ 0.15 mol%; (b)
CH
radical at l ¼ 340.8 (laser energy: 8.5 mJ, p ¼ 1.73 mbar,
x(CH
difference of the mass spectra for the condition ‘‘CH
absence of O atoms, spectrum (b)’’; (d) as (c), CH
(x(CD
) ¼ 0.52 mol%).
3
/CD
3
þ O: detection of the radicals and of
3
(
2
4
3
3
2
3
4
2
2
3
4
3
þ HF; discharge ‘‘on’’ for O
2
4
) ¼ 0.52 mol%, x(F
2
) ¼ 0.14 mol%, x(O
2
) ¼ 0.04 mol%); (c)
3
þ O’’ and ‘‘CH
3
,
3
replaced by CD
3
4
1
980
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 1 9 7 7 – 1 9 8 4
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
detection of the CH
multiphoton ionization. The interference between CH
consumption and formation can be avoided by the isotopic
substitution of C by C as explained below.
3
and C
2
H
5
radicals via laser induced
3
radical
2
H
5
2 5
D
Additionally, the rate coefficients of the atomic oxygen
reactions of the fully deuterated CD3 and C D are to be
2
5
determined. As shown in Figs. 5a and 5b, CH
detected by mass spectrometry and laser induced multiphoton
3
radicals are
2
ionization at the wavelength l ¼ 450.8 nm (3d E’’, 3 þ 1
2
29
Rempi band) and l ¼ 340.8 nm (3p A ’’, 2 þ 1 Rempi band ),
2
showing some ion fragmentation at the lower wavelength. Fig.
5c gives the signal difference of the mass spectra for the
conditions of the presence of the CH radicals with and with-
3
out O atoms, i.e. a negative value indicates consumption, a
positive one formation due to reaction. The consumption of
CH radicals (m/z ¼ 15) is accompanied by a signal increase at
3
1
m/z ¼ 29. This is assigned to a fragment ion HCO (m/z ¼ 29)
1
1
produced by the ionization of HCHO - HCO þ H. The
equivalent measurements of the deuterated methyl radical CD
are displayed in Fig. 5d with a signal decrease at m/z ¼ 18
1
3
1
1
(
CD
3
) and an increase at m/z ¼ 30 (DCO , from DCDO
-
CDO þ D). It is to be noticed that for the conditions of the
Fig. 7 The reactions C
radicals and the products by mass spectrometry after laser induced
2
H
5
þ O and C
2
D
5
þ O: detection of the
reaction CH
and for CD þ O no signal is observed at m/z ¼ 15 (Fig. 5d).
3
þ O, no signal is observed at m/z ¼ 18 (Fig. 5c)
3
multiphoton ionization. (a) C
.5 mJ; p ¼ 1.78 mbar, T ¼ 298 K, x(C
mol%), (b) C
radical at l ¼ 340.8 nm (laser energy ¼ 8.5 mJ,
p ¼ 1.5 0 mbar, x(C H ) ¼ 1.10 mol%, x(F ) ¼ 0.17 mol%), (c)
2
H
5
radical at l ¼ 406.0 nm (laser energy ¼
Therefore no mass spectrometric interference is present for the
determination of the ratio of the rate coefficients of the
8
2
H
6
) ¼ 0.88 mol%, x(F ) ¼ 0.33
2
2
H
5
reactions (1) and CD
3
þ O (1D), k
1
/k1D.
2
6
2
In Fig. 6 the measurements are displayed according to eqn.
1), showing the expected line through the origin with a slope of
difference of the mass spectra for the condition ‘‘C
‘C , absence of O atoms spectrum (b)’’ (conditions of (b), x(O
.09 mol%), (d) C replaced by C
(p ¼ 1.73 mbar, x(C
.64 mol%, x(F ) ¼ 0.18 mol%).
) ¼ 0.14 mol%, x(O
2
H
5
þ O’’ and
) ¼
) ¼
‘
2
H
5
2
(
0
0
H
2 5
2
D
5
2 6
D
k
confidence level).
1
/k1D ¼ 1.01 ꢀ 0.12, (error limits: statistical error, 95%
2
2
This means that the reactions (1) and (1D) are equally fast
and no isotope effect is observed. A similar result (k
1
/k1D
¼
1
3
tion of the rate coefficients. The consumption of C
4) is accompanied by the formation of CD
(m/z ¼ 18),
þ O,
2
D
5
(m/z ¼
1
.07 ꢀ 0.05) is reported by Washida and Inomata.
Equivalent studies were performed for the reference reac-
tions C þ O (2D). Fig. 7a shows the
þ O (2), and C
nearly fragmentionless ionization of C radicals at l 4 400
3
3
1
DCDO (m/z ¼ 30, DCO from DCDO, see also CH
3
1
H
2 5
2
D
5
CD
CD
3
þ O, Fig. 5) and CD
3
CDO (m/z ¼ 46, CD
3
CO from
2
H
5
18
3
CDO, Fig. 7d. These findings reflect the mechanism:
nm (here shown at l ¼ 406 nm). Care has to be taken in the
spectral region of the ionization of NO being produced in small
amount due to air leaks in the experimental set up if the
microwave discharge is on (signal of NO at m/z ¼ 30 at 380–
C
D
2 5
þ O - C
2
D
5
O* C
2
D
5
O* - DCDO þ CD
þ O - C þ OD
3
- CD
CDO þ D C
D
5
D
2 4
3
2
3
83, 429–431, 452–454 nm). At a lower wavelength (l ¼ 340.8
3 3
nm), the wavelength for the specific detection of CH and CD
It is noted that no background signal is present at m/z ¼ 15 (the
radicals, a rich fragmentation pattern is observed (Fig. 7b)
m/z ¼ 29, 27, 26, 25, 24, 15, 14). The signal differences of the
mass spectra for the reaction C
mass peak of CH
mass peak of C
3
) and that the mass peak at m/z ¼ 34 (the
(
2
D
) is not masked by other species or fragment
5
H
5
þ O and C
2
D
5
þ O (Fig. 7c
ions. In the study of the reaction C
consumption of C
(m/z ¼ 29) is obscured by HCO from
HCHO (m/z ¼ 29), but the formation of CH radicals (m/z ¼
5) is clearly detected. In order to determine the ratio of the
rate coefficients k(C þ O) a wavelength of
þ O)/(k(C
l ¼ 435 nm was chosen. Here the mass peaks m/z ¼ 34 (C D )
H
2 5
þ O (Fig. 7c) the
2
1
and 7d) allow the identification of some primary products of
these reactions and the experimental choice for the determina-
H
2 5
3
1
D
2 5
2 5
H
2
5
and m/z ¼ 29 (C
2 5
H ) do not interfere with products or
1
fragment ions (HCO from HCHO) and no background of
NO is present. Measurements being displayed in Fig. 8 resulted
in k(C D þ O)/(k(C H þ O) ¼ 1.01 ꢀ 0.07 (linear regression,
2
5
2
5
9
5% confidence level). This means that no isotope effect is
observed within the experimental error.
The rate of the reaction CH
þ O was measured with
reference to C D þ O. As demonstrated by Figs. 5 and 7
3
2
5
the mass peaks m/z ¼ 15 and m/z ¼ 34 at l ¼ 340.8 nm can be
3
2 5
unequivocally assigned to the radicals CH and C D thus
allowing here the measurement of the consumption of these
radicals. The measurements with different absolute radical
concentrations, achieved using a low and very low F atom
ꢂ10
ꢂ11
ꢂ3
mol cm ,
Fig. 6 The reactions CH /CD
coefficients of the reactions CD
of the radicals for increasing amounts of O atoms. I
3
3
þ O: determination of the relative rate
þ O and CH þ O by the consumption
(R), I(R): ion
concentration (1.46 ꢁ 10
and 2.55 ꢁ 10
þ O)/k(C
3
3
respectively) gave the values (see Fig. 9) k(CH
D
5
þ
3
2
0
O) ¼ 0.707 ꢀ 0.08 (‘‘very low’’) and 0.592 ꢀ 0.02 (‘‘low’’).
signal of the radical in the absence and in the presence of O atoms,
respectively. Arrangement (C): l ¼ 340.5 nm; laser energy ¼ 5.5 mJ;
At very low radical concentrations the condition of suppres-
sing secondary and competitive reactions is approximated at
the expense of a larger scatter of the measurements. This
p ¼ 1.76 mbar; T ¼ 298 K, x(CH
4
) ¼ x(CD
4
) ¼ 0.96 mol%, x(F ) ¼
2
0
.14 mol%, x(O
2
) ¼ 0–0.27 mol%.
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 1 9 7 7 – 1 9 8 4
1981

View Article Online
Table 3 The reaction CH
compiled from refs. 15 and 25
3
þ O and C
2
D
5
þ O: the mechanism. Data
3
k/cm mol
ꢂ1 ꢂ1
Reaction
s
1
3
3
3
3
3
3
3
3
3
3
3
CH
CH
4
þ F
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
CH
3
þ HF
4.72 ꢁ 10
1
3
þ CH
3
C
2
C
2
C
4
C
2
C
2
C
4
C
3
H
D
H
D
H
H
H
6
2.11 ꢁ 10
1
C
C
C
C
C
C
2
2
2
2
2
2
D
D
D
H
H
H
6
þ F
þ C
5
þ DF
6.00 ꢁ 10
1
5
5
6
5
5
2
D
5
10
1.02 ꢁ 10
1
þ CH
3
5
5
CH
3
2.85 ꢁ 10
1
þ F
þ HF
6.00 ꢁ 10
1
þ C
2
H
5
10
8
1.02 ꢁ 10
1
þ CH
3
2.85 ꢁ 10
1
CH
CH
CD
CD
CD
CD
3
3
3
3
2
2
þ O
þ O
HCHO þ H
CO þ H þ H
DCDO þ D
CO þ D þ D
HCDO þ D
CO þ D þ H
HCHO þ CH
CH
4.16 ꢁ 10
1
2
3.44 ꢁ 10
1
þ O
4.16 ꢁ 10
Fig. 8 The reactions C
relative rate coefficients of the reactions C
the consumption of the radicals for increasing amounts of O atoms.
2
H
5
þ O and C
2
D
5
þ O: determination of the
13
þ O
2
3.44 ꢁ 10
4.15 ꢁ 10
3.40 ꢁ 10
3.39 ꢁ 10
4.58 ꢁ 10
2.50 ꢁ 10
3.39 ꢁ 10
4.58 ꢁ 10
2.50 ꢁ 10
1.61 ꢁ 10
1.61 ꢁ 10
1.00 ꢁ 10
3.32 ꢁ 10
3.61 ꢁ 10
3.61 ꢁ 10
3.61 ꢁ 10
2
D
5
þ O and C
2
H
5
þ O by
13
H þ O
H þ O
13
2
Arrangement (C): l ¼ 435.0 nm, laser energy ¼ 13 mJ, p ¼ 1.73 mbar,
13
13
13
13
13
13
11
11
14
12
13
13
13
C
C
C
C
C
C
C
C
2
2
2
2
2
2
2
2
H
H
H
D
D
D
H
D
5
5
5
5
5
5
6
6
þ O
þ O
þ O
þ O
þ O
þ O
3
T ¼ 298 K, x(C
2
H
6
) ¼ x(C
2
D
6
) ¼ 0.69 mol%, x(F ) ¼ 0.15 mol%;
2
CHO þ H
x(O
2
) ¼ 0–0.22 mol%.
3
C
2
H
4
þ OH
DCDO þ CD
CD
3
3
CDO þ D
C
2
C
2
C
2
D
H
D
4
5
5
þ OD
kinetically well-controlled value is chosen as the basis for the
further evaluation.
In order to assess the influence of parallel and consecutive
þ OH
þ OD
þ H
þ D
2
O
O
2
CH
CH
3
þ D
þ H
CH
CH
CD
CD
CH
2
4
3
3
3
D þ H
1
reactions on the measured value of k /k2D, simulations of the
3
C
C
C
2
2
2
D
D
H
5
þ D
þ H
þ H
þ CD
þ CD
þ CH
3
concentration profiles of a number of species have been
performed (see Table 3 and Fig. 10). The low concentrations
of the indicative species (e.g. CH C D , C H ) formed in the
5
5
2
H
3
3
2
5
2
6
interfering reactions suggest a marginal influence on the rate
determination. Quantitatively this leads to a correction of the
Application
apparent ratio (k
1
/k2D) by a factor of 1.027 (i.e. 2.7% higher).
The directly measured rate coefficient k is reported as (1.04 ꢀ
The methane–air flame is the simplest hydrocarbon–air flame
and as such has attracted attention for several reasons (natural
gas combustion, safety management in mines, flame structure
investigations—microprobes, molecular beam mass spectro-
metry, laser based diagnostics—and modelling studies of this
basic flame being used as a model for more complex fuel
flames). An interesting account of the development of the
knowledge of methane-flame structure and the inherent che-
mical processes is given in ref. 30 (see also ref. 31, and the
discussion therein). In the context of the present study a few
details are mentioned. It is generally accepted that the primary
attack of CH is by atoms and radicals leading to CH radicals,
2
1
4
3
ꢂ1 ꢂ1
0
.1) ꢁ 10 cm mol
s (ref. 18). Thereby the absolute value
of k is deduced from k ¼ (k /k ) ꢃ (k /k ) ꢃ k and, together
1
1
1
2D
2D
2
2
1
3
3
ꢂ1
with the combined error, as k
s
1
¼ (7.6 ꢀ 1.4) ꢁ 10 cm mol
ꢂ1
.
This value is in remarkable good agreement with the latest
1
0
experimental value of Fockenberg and Preses (k(300 K) ¼
1
3
3
ꢂ1 ꢂ1
s ) and the quantum chemical
(
7.34 ꢀ 0.92) ꢁ 10 cm mol
1
4
13
3
calculation of Harding et al. (k(300 K) ¼ 9.26 ꢁ 10 cm
ꢂ1 ꢂ1
mol
s ). Moreover, the negligible isotope effect of the rate
coefficient k(CD þ O) is confirmed.
3
4
3
which in lean and stoichiometric flames, react with O atoms.
From the early formation of CO in a flame two possible
pathways have been suggested: HCHO þ M - CO þ
Fig. 9 The reactions CH
relative rate coefficients of the reactions CH
consumption of the radicals for increasing amounts of O atoms
3
þ O and C
2
D
5
þ O: determination of the
3
þ O and C þ O by the
2 5
D
(
(
arrangement (C), l ¼ 340.5 nm). ‘‘Very low radical concentration’’
ꢂ10
empty circles): p ¼ 1.50 mbar, T ¼ 298 K, [CH
4
]
and [O]
0
¼ 5.05 ꢁ 10
,
Fig. 10 The reactions CH
3
þ O and C
2
D
5
þ O: modeling of the
ꢂ10
ꢂ11
ꢂ10
[
C
2
D
6
]
0
¼ 4.04 ꢁ 10 , [F]
0
¼ 2.55 ꢁ 10
0
¼ 0–1 ꢁ 10
concentration/time profiles of the species assuming the reactions of
]
ꢂ3
mol cm . ‘‘Low radical concentration’’ (filled squares): p ¼ 1.73 mbar,
Table 3 and the experimental conditions of Fig. 9 ([CH
4
0
¼ 5.05 ꢁ
¼ 3.06 ꢁ
ꢂ10
ꢂ10
ꢂ10
ꢂ10
ꢂ11
T ¼ 298 K, [CH
4
]
0
¼ 3.61 ꢁ 10 , [C
2
D
mol cm
6
]
0
¼ 4.38 ꢁ 10 , [F]
0
¼
10 , [C
2
D
6
]
0
¼).4.04 ꢁ 10 , [F]
ꢂ3
mol cm
0
¼ 2.55 ꢁ 10 and [O]
0
ꢂ10
ꢂ10
ꢂ3
ꢂ11
1
.46 ꢁ 10
and [O]
0
¼ 0–1.25 ꢁ 10
.
10
1
982
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 1 9 7 7 – 1 9 8 4
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
Fig. 13 Simulations of HCHO, HCO (offset 0.008, scaling factor 100)
and CH (offset 0.0017, scaling factor 1000) mole fraction profiles in an
atmospheric stoichiometric CH
above the burner for branching ratios k
10% (case I) and 50% (case II); (k
mol
4
–air flame as a function of height
(a1)/(k
(a1) þ k (a2) ¼ 7.6 ꢁ 10 cm
ꢂ1 ꢂ1
s . Profiles for case I are always higher than for case II.
1
1
(a1) þ k (a2)) ¼
2
13
3
1
2
intermediate HCHO under flame conditions is obvious since
there is a factor of almost three between the maximum HCHO
concentration for the 100% and the 20% case. A high sensitivity
to the CO channel fraction is also seen in the concentration
profiles of the important intermediate species HCO and CH. Here
calculations were performed with CO channel fractions of 10%,
case I, and 50%, case II, being, respectively, the lower and upper
limit given in ref. 10 and in this study. As HCO is mainly formed
by HCHO decomposition, its maximum HCHO concentration is
lowered to the same extent as expected when calculation are
performed for the cases I and II. But also minor species, not
4
Fig. 11 Atmospheric stoichiometric laminar CH –air flame: integral
chemical flow analysis. The thickness of the arrows indicates the
significance of a reaction path: 100–75% (very thick); 75–50% (thick);
50–25% (thin) and 25–5% (very thin).
H2 and a route via HCO. Whereas early measurements on
CH
þ O indicate the sole formation of HCHO, IR chemilu-
minescence studies of CO(v ¼ 1, 2, ...) gave a 40% route to CO
3
6
1
6
þ H þ H (being adopted in the GRI mechanism ) and
2
suggest a mechanism which is broadly validated for flame
speeds, structures and ignition delay times of C1–C4 hydro-
directly formed from CH or HCHO, are affected. The CH
3
3
2
carbons. This mechanism has been applied to perform an
integral reaction flow analysis of a stoichiometric CH –air
flame at atmospheric pressure (see Fig. 11, ref. 33) in which
the presented data on the rate coefficient of the reaction CH
concentration for example is reduced by 20% when applying
case I instead of II. These results are illustrated in Fig. 13. The CO
concentration profiles show only minor differences with a slightly
higher and earlier CO peak in case II. But the flow analysis
4
3
þ
O and the channel branching have been included. The simula-
tion shows that in the first step 95% of the fuel loses an H atom
and forms CH of which 50% is directly oxidized by O atoms.
reveals that in this case (II) 50% of CO is formed from CH
only 12% is predicted in case I.
3
while
The flame-structure analysis shows that the reaction flow of
the important intermediate species CH , HCO, and HCHO is
3
Fig. 11 shows the results of flame-structure simulations with
respect to the HCHO concentrations. Five simulations were
performed with fractions of the HCHO forming channel of
3
strongly influenced by the CO channel fraction of the title
reaction. But also minor species such as CH are coupled to the
CO channel fraction. All calculations point to the necessity of
the inclusion of a kinetically verified mechanism and rate of the
1
00%, 80%, 60%, 40% and 20%. In all cases the total rate
coefficient k(CH
þ O - products) is taken from this study. The
predicted HCHO concentration profiles are shown in Fig. 12. The
high sensitivity of the two reaction routes (O þ CH - HCHO þ
H/CO þ H þ H) for predicting concentrations of the key
3
reaction CH
3
þ O.
Details of the mechanism, the modeling of measured species
concentrations in flames, the calculation of burning velocities,
and ignition delay times are given in refs. 32 and 33.
3
2
Acknowledgements
This work was financially supported by the European Com-
munity IHP programme ‘‘Marie Curie Fellowships’’ through
contract number HPMT-CT-2000–00080 (FP-5). Funding by
the Deutsche Forschungsgemeinschaft (SFB 357 ‘‘Molekulare
Mechanismen unimolekularer Prozesse’’) is also gratefully
acknowledged.
References
1
J. Warnatz, U. Maas and R. W. Dibble, Combustion, Springer,
Berlin, 1996.
2
3
K. Hoyermann and R. Sievert, Proc. Combust. Inst., 1978, 17, 517.
I. R. Slagle, D. Sarzynski and D. Gutman, J. Phys. Chem., 1987,
9
R. Zellner, D. Artmann, J. Karthauser, P. Rhasa and G. Weibr-
ing, J. Chem. Soc., Faraday Trans. 2, 1988, 84, 549.
A. M. Dean and P. R. Westmoreland, Int. J. Chem. Kinet., 1987,
19, 207.
1, 4375.
Fig. 12 Simulations of HCHO concentration profiles in an atmo-
spheric stoichiometric CH –air flame as a function of height above the
burner for different branching ratios k (a1)/(k (a2)) ¼ 100%,
(a1) þ k
0%, 60%, 40%, 20%; k (a2) ¼ 7.6 ꢁ 10 cm mol
(a1) þ k
4
5
¨
¨
4
1
1
2
13
3
ꢂ1 ꢂ1
s .
8
1
2
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 1 9 7 7 – 1 9 8 4
1983

View Article Online
6
7
P. W. Seakins and S. R. Leone, J. Phys. Chem., 1992, 96, 4478.
C. Fockenberg, G. E. Hall, J. M. Preses, T. J. Sears and J. T.
Muckerman, J. Phys. Chem. A, 1999, 103, 5722.
M. Preses, C. Fockenberg and G. W. Flynn, J. Phys. Chem. A,
2000, 104, 6758.
T. P. Marcy, R. R. Diaz, D. Heard, S. R. Leone, L. B. Harding
and S. J. Klippenstein, J. Phys. Chem. A, 2001, 105, 8361.
C. Fockenberg and J. M. Preses, J. Phys. Chem. A, 2002, 106,
21 T. Muller, P. H. Vaccaro, F. Perez-Bernal and F. Iachello,
J. Chem. Phys., 1999, 111, 5038.
22 T. Shimanouchi, Tables of Molecular Vibrational Frequencies Con-
solidated Volume I, National Bureau of Standards, 1972, 1–160.
´
23 I. R. Slagle, D. Sarzynski, D. Gutman, J. A. Miller and C. F.
Melius, J. Chem. Soc., Faraday Trans., 1988, 84, 491.
24 M. K. Gilles, A. A. Turnipseed, R. K. Talukkdar, Y. Rudich, P.
W. Villalta, L. G. Huey, J. B. Burkholder and A. R. Ravishan-
kara, J. Phys. Chem., 1996, 100, 14005.
25 NIST Standard Reference Database 17, NIST Chemical Kinetics
Database: Version 2Q98e, 1998.
26 F. Maguin, A. Mellouki, G. Laverdet, G. Poulet and G. Le Bras,
Int. J. Chem. Kinet., 1991, 23, 237.
27 J. W. Lindner, R. A. Loomis, J. J. Klaasen and S. R. Leone,
J. Chem. Phys., 1998, 108, 1944.
28 J. P. Reid, T. P. Marcy, S. Kuehn and S. R. Leone, J. Chem. Phys.,
2000, 113, 4572.
29 J. W. Hudgens, in Advances in Multi-photon Processes and Spec-
troscopy, ed. S. H. Lin, World Scientific, Singapore, vol. 4, 1988,
p. 171.
30 R. M. Fristrom, Flame Structure and Processes, Oxford University
Press, New York, 1995, p. 323.
31 J. Peeters and G. Mahnen, Proc. Combust. Inst., 1973, 14, 133–
146.
32 K. Hoyermann, F. Mauß and T. Zeuch, Phys. Chem. Chem. Phys.,
2004, 6, 3824.
33 T. Zeuch, Reaktionskinetik von Verbrennungsprozessen in der
Gasphase: Spektroskopische Untersuchungen der Geschwindig-
keit, Reaktionsprodukte und Mechanismen von Elementarreak-
tionen und die Modellierung der Oxidation von Kohlen-
wasserstoffen mit detaillierten Reaktionsmechanismen, , PhD
8
9
0
1
2
3
1
1
1
1
2924.
W. Hack, K. Hoyermann, M. Olzmann and T. Zeuch, Proc.
Combust. Inst., 2002, 29, 1255.
H. Oser, D. Walter, N. D. Strothard, O. Grotheer and H. H.
Grotheer, Chem. Phys. Lett., 1991, 181, 521.
C. N. Washida and S. Inomata, 16th Symposium on Gas Kinetics,
the Gas Kinetics Group of the Faraday Division of the Royal
Society of Chemistry, Cambridge, UK, 2000, p. C6.
L. B. Harding, S. J. Klippenstein and Y. Georgieskii, Proc.
Combust. Inst., 2004, 30, 985–993.
D. L. Baulch, C. J. Cobos, R. A. Cox, P. Frank, G. Hayman, T.
Just, J. A. Kerr, T. Murrells, M. J. Pilling, J. Troe, R. W. Walker
and J. Warnatz, J. Phys. Chem. Ref. Data, 1994, 23, 847.
G. P. Smith, D. M. Golden, M. Frenklach, N. W. Moriarty, B.
Eiteneer, M. Goldenberg, C. T. Bowman, R. K. Hanson, S. Song,
W. C. Gardiner, Jr., V. V. Lissianski and Z. Quin, http://
www.me.berkeley.edu/gri_mech/.
W. Hack, K. Hoyermann, C. Kersten, M. Olzmann and B.
Viskolcz, Phys. Chem. Chem. Phys., 2001, 3, 2365.
W. Hack, K. Hoyermann, M. Olzmann and T. Zeuch, Proc.
Combust. Inst., 2002, 29, 1247.
K. Hoyermann, M. Olzmann, J. Seeba and B. Viskolcz, J. Phys.
Chem. A, 1999, 103, 5692.
K. Hoyermann and F. Nacke, Proc. Combust. Inst., 1996, 26, 505.
1
1
4
5
1
6
1
1
1
2
7
8
9
0
¨
thesis, University of Gottingen, Germany, 2003.
1
984
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 1 9 7 7 – 1 9 8 4
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5