Formation of HO2 and OH in photolytically initiated oxidation of dimethyl ether

Article abstract of DOI:10.1016/j.cplett.2006.04.116

Time-resolved measurements of HO₂ and OH have been conducted in 355 nm photolysis of dimethyl ether/Cl₂/O₂ mixture at elevated temperature, using near-infrared frequency modulation spectroscopy. A part of OH was found to be produced at a timescale of several microseconds by the methoxymethyl with O₂ reaction, while HO₂ is formed mainly in milliseconds with the yield increasing up to 60% between 500 and 600 K. It was rationalized that HO₂ is not a direct product of the O₂ adduct decomposition, but a secondary product through HCHO + OH reaction. Another pathway through HCO formation from the adduct is also discussed.

Full text of DOI:10.1016/j.cplett.2006.04.116

Chemical Physics Letters 425 (2006) 179–184
www.elsevier.com/locate/cplett
Formation of HO2 and OH in photolytically initiated oxidation
of dimethyl ether
a
c
b
Kotaro Suzaki , Nozomu Kanno b,c, Kenichi Tonokura , Mitsuo Koshi ,
d
a,
*
Kentaro Tsuchiya , Atsumu Tezaki
a
Department of Mechanical Engineering, School of Engineering, The University of Tokyo, Hongo, Bunkyo, Tokyo, Japan
b
Department of Chemical System Engineering, School of Engineering, The University of Tokyo, Tokyo, Japan
c
Environmental Science Center, The University of Tokyo, Tokyo, Japan
National Institute of Advanced Industrial Science and Technology (AIST), Japan
d
Received 19 October 2005; in final form 3 April 2006
Available online 16 May 2006
Abstract
Time-resolved measurements of HO2 and OH have been conducted in 355 nm photolysis of dimethyl ether/Cl2/O2 mixture at elevated
temperature, using near-infrared frequency modulation spectroscopy. A part of OH was found to be produced at a timescale of several
microseconds by the methoxymethyl with O2 reaction, while HO2 is formed mainly in milliseconds with the yield increasing up to 60%
between 500 and 600 K. It was rationalized that HO2 is not a direct product of the O2 adduct decomposition, but a secondary product
through HCHO + OH reaction. Another pathway through HCO formation from the adduct is also discussed.
ꢀ 2006 Elsevier B.V. All rights reserved.
R þ O2 ! RO2 ! ðQOOH !Þalkene þ HO2:
ð2Þ
1. Introduction
It was argued that the slow HO2 formation dominant at
elevated temperature is primarily from alkylperoxy radi-
cal(RO2) decomposition and the hydroperoxy alkyl radi-
cal(QOOH) route is of minor importance in the cases of
ethyl, n-propyl and iso-propyl, as supported by quantum-
chemical calculations. Substantial contribution of the
QOOH route leading to more OH formation is expected
in butyl and larger alkyls as reaction (3)
The reactions of alkyl radicals with molecular oxygen
are important in ‘low temperature combustion’ region, par-
ticularly on autoignition properties. The general descrip-
tion for alkyl(R) + O2 reactions was established by the
end of last century [1,2].
In the last couple of years, Taatjes, Klippenstein and co-
workers have provided further insights to the R + O2 reac-
tions using direct time-resolved measurements of HO2 and
RRKM-master equation theory for ethyl to butyl [3–8],
neopentyl [9] and cyclo-propyl [10] as the alkyl radicals.
Their notable results include the presence of a direct HO2
forming channel;
R þ O2 ! RO2 ! QOOH ! O ꢀ hetercycle þ OH
ð3Þ
and in practice, the reduction of the HO2 yield was recog-
nized in their butane experiment [8].
Dimethyl ether (DME) is a diesel alternative fuel exhib-
iting outstanding ignitability as this size of hydrocarbon
molecule. Detailed reaction mechanisms accounting for
DME property have been proposed by Dagaut et al. [11]
and Curran et al. [12,13], in which the methoxymethyl
(CH3OCH2) + O2 reaction and the following reproduction
of OH constitute a growing chain reaction mechanism effec-
tive in the low temperature combustion region. With regard
R þ O2 ! alkene þ HO2;
ð1Þ
distinguished from the slow formation via decomposition
of intermediates;
*
Corresponding author. Fax.: +81 3 5841 6293.
E-mail address: tezaki@comb.t.u-tokyo.ac.jp (A. Tezaki).
0009-2614/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2006.04.116

180
K. Suzaki et al. / Chemical Physics Letters 425 (2006) 179–184
to elementary reaction studies, the CH3OCH2 + O2 system
has been investigated between 240 and 700 K, by observing
species using UV absorption and mass-spectrometric sam-
pling [14–18], through which the main pathway following
the CH3OCH2 + O2 reaction in the reduced pressure exper-
iments is considered to be a specific case of reaction (3) as
High ratios of initial [DME]/[Cl2] and [O2]/[Cl2] were main-
tained in order to assure prompt completion of reaction (6)
and 100% yield of reaction (7) to [Cl]0 by avoiding
CH3OCH2 + Cl and CH3OCH2 + Cl2 reactions.
The quantity of HO2 relative to that of initial radical is
known by comparing with the reference reaction of
CH3OH/Cl2/O2 mixture, i.e.
CH3OCH2 þ O2 ! CH3OCH2OO
! CH2OCH2OOH
Cl þ CH3OH ! HCl þ CH2OH
CH2OH þ O2 ! HO2 þ HCHO
ð8Þ
ð9Þ
! OH þ 2HCHO:
ð4Þ
The OH reference is obtained by adding NO to the HO2
reference, in which 100% conversion through
The dominance of this pathway has been supported by
quantum chemical calculations [19–21] and is consistent
with product analysis finding formaldehyde as a major
product [14,15]. Ethylene oxide, which is a product corre-
sponding to the reaction (2), was not detected between
room temperature and 450 K. However, the possibility of
HO2 formation via RO2/QOOH decomposition or other
indirect pathway at higher temperature is yet to be con-
firmed. In this study, we examine the product formation
in a DME oxidation system detecting HO2 and OH using
a near-IR modulation technique at the temperatures up
to 650 K. The pathway forming HO2 is discussed with
kinetics modeling and quantum chemistry calculations.
HO2 þ NO ! OH þ NO2
ð10Þ
takes place. The reference data are acquired after every
DME run with the same temperature, total density and
the mixture composition.
A typical condition is [Cl2] = 2 · 1014, [O2] = 1.2 · 1016,
[DME] = [CH3OH] = 1 · 1015, and a total gas density =
1.1 · 1018 all in molecule cmꢀ3 with He buffer at tempera-
ture range 296–650 K in 50 K increments up to 500 K and
25 K increments over 500 K. The photolysis yield for the Cl
atom was estimated as ca. 5% of the Cl2 at a typical 355 nm
fluence of about 130 mJ cmꢀ2 and at the typical condition.
Photolysis repetition was set at 1 Hz to assure the gas
replacement at the total flow rate of 5200 cm3/min.
2. Experimental
The setup of the frequency modulation spectroscopy in a
Herriott type multi-pass cell has been described elsewhere
[22]. Briefly, a CW diode laser (New Focus, Velocity
6327) is two-tone modulated at 600 2.6 MHz using an
electro-optic modulator (EOM; New Focus, 4223 M). The
beam is passed through the Herriott cell, detected by a
photoreceiver (New Focus, 1811MFS), and demodulated
by mixing with a frequency-doubled fraction of 2.6 MHz
modulation signal. The Herriott cell consists of two concave
mirrors placed co-axially at a distance of 1.5 m, between
which 24 reflections are established. The center region
about 60 cm is the temperature controlled observation
region where the probe path is overlapped with the co-axial
photolysis beam from a frequency tripled Nd-YAG laser.
Consequent effective path length is approximately 15 m.
The temperature is measured by thermocouples placed at
the center and 40 cm off-center regions in the cell. The signal
is usually averaged 1000 times with a digital oscilloscope. In
this study, HO2 was detected at 7013.520 cmꢀ1 of the A–X
(00–000) 1ꢀ0Q2(18) transition [23] and OH was detected at
6971.291 cmꢀ1 of Q1e(1.5) of the vibrational overtone [24].
The detection limit of HO2 concentration with 1000 times
3. Results and discussion
Fig. 1 shows typical time profiles of (a) HO2 and (b) OH
observed at 600 K in the photolysis of Cl2/O2/CH3OCH3
mixture and the methanol reference. The HO2 reference
(a) HO2
Reference
DME
0.00
0.01
0.02
0.03
0.04
Time /s
(b) OH
Reference
DME
Reference model
average is estimated to be 1 · 1012 molecule cmꢀ3
.
The reaction is initiated by pulsed photolysis in a mix-
ture of Cl2, O2 and DME with He buffer to establish the
following reaction sequence:
0.0
0.2
0.4
0.6
0.8
1.0
Time /ms
Cl2 þ hm ! 2Cl
Cl þ CH3OCH3 ! HCl þ CH3OCH2
CH3OCH2 þ O2 ! products
ð5Þ
ð6Þ
ð7Þ
Fig. 1. Time profiles of (a) HO2 and (b) OH FM signals at 600 K in the
photolysis of CH3OCH3/Cl2/O2 mixture and reference mixtures. The
references are CH3OH/Cl2/O2 for HO2 and CH3OH/Cl2/O2/NO for OH.

K. Suzaki et al. / Chemical Physics Letters 425 (2006) 179–184
181
profile shows a prompt rise at the time of photolysis fol-
(a) HO2 from CH3OH at 600 K
1.5
1.0
0.5
0.0
lowed by a gradual decrease at the rate of the HO2 + HO2
self-reaction. The profile from DME shows a slower
increase, which is similar to those reported for C2–C4 alkyl
radicals + O2 [4–8]. On the other hand, the OH profile
from DME shows a rapid rise whose rate is equivalent to
that of the reference. This rapid OH formation was also
reported for ethyl, propyl and neopentane [9,25], however,
the amount close to half of the reference is apparently lar-
ger than those reported.
It is necessary to correct the HO2 profile to obtain the
overall HO2 yield relative to the initial chlorine atom
amount according to the established procedure [4]. The
precise rate constant of the HO2 self-reaction necessary
for the correction is obtained from the reference profile,
but the rate constant of HO2–RO2 reaction, which should
be given separately, is unknown. In this study, we set the
initial radical concentration as low as possible to minimize
the HO2 consumption and corrected the HO2 signal from
DME only for the self-reaction. As shown in Fig. 2b, the
extent of correction is not very significant.
raw signal
corrected signal
-0.01
0.00
0.01
0.02
0.03
0.04
Time /s
(b) HO2 from CH3OCH3 at 600 K
1.5
1.0
0.5
0.0
raw signal
corrected signal
Curran 98 original moder [12]
Present model
-0.01
0.00
0.01
0.02
0.03
0.04
Time /s
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
(c) OH from CH3OCH3 at 600 K
experimental signal
Curran 98 original model [12]
Present model
As superimposed in Fig. 1b, the OH reference profile is
reproduced by the reaction mechanism reported in Ref. [9],
in which OH is consumed through reactions with CH3OH,
NO, OH, etc. The OH profile from DME is scaled assum-
ing that the reference peak corresponds to the initial chlo-
rine atom concentration.
-0.2
0.0
0.2
0.4
0.6
0.8
0.04
0.8
Time /ms
1.0
(d) HO2 from CH3OCH2 at 550 K
0.8
0.6
0.4
0.2
0.0
-0.2
A mechanism effective at 600 K can be extracted from
the Curran et al. model [12] as:
-0.01
0.00
0.01
0.02
0.03
CH3OCH2 þ O2ðþMÞ () CH3OCH2O2ðþMÞ
CH3OCH2O2ðþMÞ () CH2OCH2OOHðþMÞ
CH2OCH2OOHðþMÞ ! 2HCHO þ OHðþMÞ
OH þ HCHO ! H2O þ HCO
ð11Þ
ð12Þ
ð13Þ
ð14Þ
ð15Þ
ð16Þ
ð17Þ
Time /s
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
(e) OH from CH3OCH2 at 550 K
HCO þ O2 ! CO þ HO2
CH3OCH3 þ OH ! CH3OCH2 þ H2O
CH3OCH2ðþMÞ ! CH3 þ HCHOðþMÞ
-0.2
0.0
0.2
0.4
0.6
Time /s
Fig. 2. Time profiles of HO2 from (a) CH3OH/Cl2/O2, (b) CH3OCH3/Cl2/
O2 mixtures at 600 K with and without corrections for HO2 + HO2
reactions, (c) time profiles of OH at 600 K, (d) time profiles of HO2 and (e)
time profiles of OH at 550 K. The bold lines are present models and the
dash lines are Curran model’s.
The above reactions together with reactions (5) and (6)
essentially represent the OH and HO2 formations in the
Curran et al. mechanism at the current experimental condi-
tions, although we used the full Curran et al. mechanism
for quantitative completeness. As drawn in Fig. 2b and c,
this model with the original set of rate constants does not
reproduce the amount of HO2 and the rapid rise of OH
at 600 K. Fig. 2 also shows HO2 and OH profiles in the
case of 550 K, where the HO2 yield of the original model
is in better agreement with the measurement. However,
the calculated HO2 profile resulted in considerably slower
rise than that in the measurement. As well, the OH profile
is in poor agreement. The Curran et al. model has been
established to represent experiments of over atmospheric
pressures therefore we need to revise mechanism for the
reduced pressure experiments as follows:
The thermal decomposition of CH3OCH2 (17) is most
responsible to the HO2 yield decreasing above 550 K.
According to the high and low pressure limit rate constants
reported by Louck et al. [26] we set the constant as
k17 = 5.0 · 109· exp(ꢀ75.8 kJ/mol/RT) sꢀ1 for the current
density of 1.1 · 1018 molecule/cm3, which is about 10 times
lower than the original at 600 K. The O2 addition to
CH3OCH2 (11) was experimentally examined [15,16] and
the pressure dependence was theoretically evaluated by
Yamada et al. [19]. The evaluated value at the current den-
sity is close to that of the original Curran et al. model, and
we used the original as it is. The reactions (12) and (13)
Unimolecular reactions, for which high pressure limits
have been given in the original model, were modified.

182
K. Suzaki et al. / Chemical Physics Letters 425 (2006) 179–184
were also treated by Yamada et al. [19] as pressure depen-
dent. Since the product of these series reactions is respon-
sible for the observed time profile of HO2, we fixed the
rate constant k12 as 2.18 · 109 · exp(ꢀ66.3 kJ/mol/RT) sꢀ1
according to Yamada et al., which is about ten times lower
than that of the original, and fit the rate constant k13.
In order to reproduce the rapid rise of OH, the direct
forming pathway of OH from CH3OCH2 is assumed, i.e.;
mated k13 is almost 10 times higher than that of Yamada
et al. value, but still lower than the Curran et al. original
value by a factor of 0.2. In the case of neopentane [9], which
apparently has no direct HO2 formation pathway, the reac-
tion of OH with RO2 was proposed to be the main pathway
for the HO2 formation. In the current case of DME, the rate
constant of OH + RO2 is assumed as 1.2 · 10ꢀ10 cm3 sꢀ1
.
Without this reaction the rise of HO2 and the OH decay is
not predicted well as shown in Fig. 3. Fig. 3 also shows
CH3OCH2 þ O2 ! OH þ 2HCHO
ð18Þ
the influence of k13, k18 and kOHþRO and OH profiles.
2
The profiles calculated from the set of estimated rate con-
stants are compared with those from increased and
decreased constants for each reaction by a factor of 2. The
rate constant of the direct OH formation reaction (18)
affects OH profiles strongly and the amount of OH affects
that of the HO2. The rate constant of OH + RO2 has an
effect on initial rise of HO2 only and does not affect the
amount of HO2. Even at this rate of OH + RO2, the
OH + HCHO pathway accounts for 80% of the HO2
formation.
As shown in Fig. 2, the HO2 and OH profiles are pre-
dicted well when the rate constants are assumed as
k13 = 1.25 · 1013 · exp (ꢀ84.1 kJ/mol/RT) sꢀ1, and k18
=
8.3 · 10ꢀ12 · exp(ꢀ8.8 kJ/mol/RT) cm3 sꢀ1. These Arrhe-
nius parameters are obtained from fitting to HO2 and OH
profiles between 500 and 600 K. The direct OH formation
(18) competes with the recombination (11) and the branch-
ing ratio determines the peak OH intensity. The 550 K pro-
files are also predicted well by the modified model as shown
in Fig. 2d for HO2 and (e) for OH. It is noted that the esti-
1.2
1.2
(a) k13 dependence of HO2
(b) k13 dependence of OH
1.0
1.0
Exp
Present model
2 k13
0.8
0.8
0.6
0.4
0.2
0.0
-0.2
0.6
1/2 k13
0.4
0.2
0.0
-0.2
-0.01
0.00
0.01
0.02
0.03
0.04
0.04
0.04
-0.2
0.0
0.2
Time /ms
0.4
0.6
0.6
0.6
0.8
0.8
0.8
Time /s
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
(c) k18 dependence of HO2
(d) k18 dependence of OH
Exp
Present model
2 k18
1/2 k18
-0.01
0.00
0.01
0.02
0.03
-0.2
0.0
0.2
Time /ms
0.4
Time /s
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
(e) RO2+OH dependence of
(f) RO2+OH dependence of
OH
Exp
Present model
2 k
HO2
RO2+OH
1/2 k
Without RO2+OH
RO2+OH
-0.01
0.00
0.01
0.02
0.03
-0.2
0.0
0.2
0.4
Time /s
Time /ms
Fig. 3. The HO2 and OH profiles changing the value of k13, k18, kRO þOH. The lines are present model profiles, the dash lines are twice and the dash-dot
2
lines are half as much as the estimated value. The dot lines in (e), (f) are the profiles without the RO2 + OH reaction.

K. Suzaki et al. / Chemical Physics Letters 425 (2006) 179–184
183
Potential energy diagrams for the CH3OCH2 + O2
systems are shown in Fig. 4. Our current calcula-
tions using G2M (CC1) //B3PW91/cc-pTVZ level of theory
(MRMP2//CASSCF/cc-pVTZ for TS4) are compared with
those reported by Andersen et al. [20,21] using UDFT/
B3LYP and by Yamada et al. [19] using MP2(full)/6-
31 G(d,p). The pathways forming OH + 2HCHO and
HO2 + c-C2H4O in our case and those by Yamada et al.
are in good agreement, whereas the results of Anderson
et al. tend to lie higher. The barrier of the direct HO2 form-
ing channel (TS3) is considerably above the CH3O-
CH2 + O2 level consistently in all cases; consequently this
channel is unlikely to contribute.
represented by giving the branching ratio of 5% for this chan-
nel with a modification of k13 at a factor of 0.95. Even in this
case the major route to HO2 formation is via OH + HCHO.
It should be noted that the OH production further consumes
DME via reaction 16 repeatedly, so that the nominal yield of
HO2 relative to initial Cl concentration is enhanced.
Fig. 5 shows temperature dependence of the HO2 yield
from DME. The HO2 yield is reasonably predicted by this
modified model through the investigated temperature
0.6
0.5
Exp
Curran 98 original model
Present model
We found another pathway leading to HO2 formation
via;
0.4
0.3
0.2
0.1
0.0
CH2OCH2O2H ! HOCH2OCH2O
! HCHO þ HCO þ H2O;
ð19Þ
where HCO readily reacts with O2 in the current experimen-
tal conditions. The energy of the final HCHO + HCO +
H2O is considerably lower than those of C2H4O + HO2
and 2HCHO + OH. The highest barrier of this channel
(TS4) is 13 kJ/mol above that of the OH channel (TS2) in this
calculation; hence it may contribute a few percent in the
CH2OCH2O2H decomposition. The HO2 profiles are also
300
350
400
450
500
550
600
Temperature /K
Fig. 5. Temperature dependence of HO2 yield from DME against the
initial Cl atom concentration. Solid circles are experiments. Solid lines are
the result of present model and dashed lines are that of Curran et al. model.
Fig. 4. Potential energy diagrams of CH3OCH2 + O2 system with geometry of TS4 and TS5. Thick black lines are results of present calculation, dashed
lines are Andersen et al. [20] and gray lines are Yamada et al. [19].

184
K. Suzaki et al. / Chemical Physics Letters 425 (2006) 179–184
[2] R.W. Walker, C. Morley, Basic Chemistry in Combustion, in: M.J.
Pilling (Ed.), Low-Temperature Combustion and Autoignition, Else-
vier, Amsterdam, 1997, p. 1.
[3] J.A. Miller, S.J. Klippenstein, S.H. Robertson, Proc. Combust. Inst.
28 (2000) 1479.
[4] E.P. Clifford, J.T. Farrell, J.D. DeSain, C.A. Taatjes, J. Phys. Chem.
A 104 (2000) 11549.
[5] J.D. DeSain, S.J. Klippenstein, J.M. Miller, C.A. Taatjes, J. Phys.
Chem. A 104 (2000) 4415.
range. It is notable that the HO2 yield begins to rise at the
temperature about 50 K lower but the maximum yield
remains lower, compared with the results of ethane and
propane [4,6]. As the barrier height for QOOH isomariza-
tion from RO2 in the case of DME is lower than that of
direct HO2 formation from RO2 in the cases of ethane
and propane, the OH formation channel is favored at
investigated temperature and HO2 is formed by secondary
reactions (14) and (15) as described above.
[6] J.D. DeSain, E.P. Clifford, C.A. Taatjes, J. Phys. Chem. A 105 (2001)
3205.
It was deduced in this study that the formation of HO2
from methoxymethyl + O2 is via secondary reaction of
OH + HCHO. It is contrasted from the cases of C2–C4
alkyl radicals, where HO2 is formed directly from RO2 spe-
cies. Neopentyl radical also forms HO2 secondarily via
OH + RO2. Despite the different pathways, all these reac-
tions show remarkably similar behavior of HO2 formation,
such as the slower rise than the rate of R + O2 reaction and
the dramatic increase of yield with increasing temperature.
It is because the overall rate is governed by the unimolecu-
lar decomposition of RO2/QOOH species whichever it
directly produces HO2 or OH.
[7] J.D. DeSain, C.A. Taatjes, J. Phys. Chem.
6646.
[8] J.D. DeSain, C.A. Taatjes, J.M. Miller, S.J. Klippenstein, D.K.
Hahn, Faraday Discuss. 119 (2001) 101.
[9] J.D. DeSain, S.J. Klippenstein, C.A. Taatjes, Phys. Chem. Chem.
Phys. 5 (2003) 1584.
[10] J.D. DeSain, S.J. Klippenstein, C.A. Taatjes, J. Phys. Chem. A 107
(2003) 1992.
[11] P. Dagaut, C. Daly, J.M. Simmie, M. Cathonet, Proc. Combust. Inst.
27 (1998) 361.
[12] H.J. Curran, W.J. Pitz, C.K. Westbrook, P. Dagaut, J.C. Boettner,
M. Cathonnet, Int. J. Chem. Kinet. 30 (1998) 229.
[13] H.J. Curran, S.L. Fischer, F.L. Dryer, Int. J. Chem. Kinet. 32 (2000)
713.
[14] J. Sehested, T. Mogelberg, T.J. Wallington, E. Kaiser, O.J. Nielsen, J.
Phys. Chem. 100 (1996) 17218.
[15] J. Sehested, K. Sehested, J. Platz, H. Egsgaard, O.J. Nielsen, Int. J.
Chem. Kinet. 29 (1997) 627.
A 105 (2001)
Acknowledgements
The authors thank Dr. Craig A. Taatjes and Mr. Leon-
ard Jusinski at Sandia National Laboratories for their help
in introducing the experimental technique. Helpful discus-
sions with Dr. Taatjes and Dr. Stephen J. Klippenstein at
Argonne National Laboratories are also appreciated. This
study was supported by the Ministry of Education, Science,
Sports and Culture of Japan (MEXT), a Grant-in-Aid for
Exploratory Research, 15656053, 2003, and a Grant-in-Aid
for Scientific Research in the priority areas ‘A new avenue
of radical chain reactions in atmospheric and combustion
chemistry’. K. Suzaki was supported by MEXT, the 21st
Century COE Program ‘Mechanical System Innovation’.
[16] K. Hoyermann, F. Nacke, Proc. Comb. Inst. 26 (1996) 505.
[17] M.E. Jenkin, G.D. Haymann, T.J. Wallington, M.D. Hurley,
J.C. Ball, O.J. Nielsen, T. Ellermann, J. Phys. Chem. 97 (1993)
11712.
[18] M.M. Maricq, J.J. Szente, J.D. Hybl, J. Phys. Chem. A 101 (1997)
5155.
[19] T. Yamada, J.W. Bozzelli, T.H. Lay, Int. J. Chem. Kinet. 32 (2000)
435.
[20] A. Anderson, E.A. Carter, Isr. J. Chem. 42 (2002) 245.
[21] A. Anderson, E.A. Carter, J. Phys. Chem. A 107 (2003) 9463.
[22] N. Kanno, K. Tonokura, A. Tezaki, M. Koshi, J. Mol. Spectrocs. 229
(2005) 193.
[23] E.H. Fink, D.A. Ramsay, J. Molec. Spectrocs. 185 (1997) 304.
[24] M.C. Abrams, S.P. Davis, L.P. Rao, R. Engleman Jr., J.W. Brault,
The Astrophys. J. Suppl. Series. 93 (1994) 351.
[25] J.D. DeSain, S.J. Klippenstein, J.A. Miller, C.A. Taatjes, J. Phys.
Chem. A 107 (2003) 4415.
References
[26] L.F. Loucks, K.J. Laidler, Can. J. Chem. 45 (1967) 2767.
[1] C.K. Westbrook, Proc. Combust. Inst. 24 (2000) 1563.