Low temperature rate coefficients for the reactions of CN and C2H radicals with allene (CH2=C=CH2) and methyl acetylene (CH3CCH)

Article abstract of DOI:10.1016/S0009-2614(01)00682-0

Using a continuous flow CRESU (Cinétique de Réaction en Ecoulement Supersonique Uniforme or Reaction Kinetics in Uniform Supersonic Flow) apparatus, rate coefficients have been measured for the reactions of the cyanogen (CN) and ethynyl (C₂H) radicals with allene (CH₂=C=CH₂) and methyl acetylene (CH₃CCH) at temperatures from 295 down to 15 K for the reactions of CN and down to 63 K for those of C₂H. All four reactions occur at rates close to the collision-determined limit. The results are compared with those obtained earlier for the reactions of other alkenes and alkynes, and, in the accompanying Letter by Vakhtin et al., with results for C₂H+CH₂=C=CH₂ and C₂H+CH₃CCH obtained at 103 K using a pulsed Laval apparatus. The implications of these latest results for the chemistry of interstellar clouds and planetary atmospheres are discussed.

Full text of DOI:10.1016/S0009-2614(01)00682-0

24 August 2001
Chemical Physics Letters 344 02001) 310±316
www.elsevier.com/locate/cplett
Low temperature rate coecients for the reactions
of CN and C2H radicals with allene ꢀCH2@C@CH2† and
methyl acetylene ꢀCH3CBCH†
David Carty, Valery Le Page, Ian R. Sims *, Ian W.M. Smith
School of Chemistry, The University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
Received 5 April 2001; in ®nal form 22 May 2001
Abstract
ꢀ
ꢀ
Using a continuous ¯ow CRESU 0Cinetique de Reaction en Ecoulement Supersonique Uniforme or Reaction Ki-
netics in Uniform Supersonic Flow) apparatus, rate coecients have been measured for the reactions of the cyanogen
0CN) and ethynyl ꢀC2H† radicals with allene ꢀCH2@C@CH2† and methyl acetylene ꢀCH3CBCH† at temperatures from
295 down to 15 K for the reactions of CN and down to 63 K for those of C2H. All four reactions occur at rates close to
the collision-determined limit. The results are compared with those obtained earlier for the reactions of other alkenes
and alkynes, and, in the accompanying Letter by Vakhtin et al., with results for C2H ‡ CH2@C@CH2 and
C2H ‡ CH3CBCH obtained at 103 K using a pulsed Laval apparatus. The implications of these latest results for the
chemistry of interstellar clouds and planetary atmospheres are discussed. Ó 2001 Elsevier Science B.V. All rights
reserved.
1. Introduction
of the convergent±divergent nozzle. The results of
the experiments, which demonstrate that many
neutral±neutral reactions can remain fast down to
very low temperatures, have led to a major re-
evaluation of the processes that may contribute to
the chemistry leading to molecular formation in
dense interstellar clouds [2].
Amongst the reactions whose kinetics have been
studied down to very low temperatures are those
of the CN [3±5] and C2H radicals [6]. In 1993,
Smith and Sims [7] proposed that the reactions of
these radicals with alkynes could provide a route
to the synthesis of the cyanopolyynes, which are
found widely in interstellar environments. This
proposal was based on 0a) the rapidity of the re-
actions of CN and C2H with alkynes at room
temperature, 0b) early CRESU measurements that
demonstrated that the reaction of CN radicals
Over the past decade, rate coecients for al-
most 40 elementary reactions between electrically
neutral species have been measured at tempera-
tures as low as 13 K [1]. These experiments employ
continuous ¯ow CRESU 0Cinetique de Reaction
en Ecoulement Supersonique Uniforme or Reac-
tion Kinetics in Uniform Supersonic Flow) appa-
ratuses in Rennes and Birmingham, in which
expansion of a gas mixture through a Laval nozzle
generates a cold, uniform and relatively dense su-
personic ¯ow of gas downstream from the nozzle,
with a temperature which is de®ned by the design
ꢀ
ꢀ
* Corresponding author. Fax: +44-121-414-4426.
E-mail address: i.r.sims@bham.ac.uk 0I.R. Sims).
0009-2614/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved.
PI I : S 0 0 0 9 - 2 6 1 4 0 0 1 ) 0 0 6 8 2 - 0

D. Carty et al. / Chemical Physics Letters 344 '2001) 310±316
311
with C2H2 occurred at essentially the collision-
limited rate below 100 K, and 0c) the likelihood
that CN and C2H would react with a given alkyne
at similar rates because the two radicals are iso-
electronic. Later measurements of the rate coe-
cients for the reactions of C2H with C2H2; C2H4
and C3H6 [6] con®rmed that reactions of C2H with
unsaturated hydrocarbons occur at rates compa-
rably fast to those of CN radicals. Moreover,
modelling calculations [8,9] have shown that neu-
tral chemistry is important in producing cyano-
polyynes under the conditions prevalent in dense
interstellar clouds.
Chastaing et al. [6] added to the original argu-
ments of Smith and Sims [7] by pointing out that
the rates of addition of a given free radical to
unsaturated hydrocarbons have been correlated to
the ionisation energy of the unsaturated species
[10]. As the higher alkenes and alkynes have lower
ionisation energies than C2H4 and C2H2 [11], re-
spectively, then it is to be expected that higher
alkenes and alkynes will react at least as fast as
C2H4 and C2H2 with CN and C2H radicals. The
measurements on the reactions of CN and C2H
with CH2@C@CH2 and CH3CBCH that are re-
ported in the present Letter provide a partial test
of this proposal.
A second reason for the present measurements
is to extend a comparison of the rates of the same
reactions that are measured in a continuous ¯ow
CRESU apparatus with those obtained in an ap-
paratus that uses a pulsed Laval nozzle. A com-
parison for reactions of OH radicals has already
been published [12]; the Letter accompanying this
one [13] reports rate coecients for the reactions
of C2H with C2H2, O2, CH2@C@CH2 and
CH3CBCH at 103 K. These data are compared
with earlier measurements from our laboratory [6]
on the ®rst two of these reactions and our present
and 15 K, in the case of the reactions of CN, and
296 and 63 K, in the case of the reactions of C2H.
A full description of the Birmingham CRESU
apparatus has been given elsewhere [14]. Further-
more, the methods used to study the kinetics of
these reactions of CN and C2H closely follow
those used previously [3±6].
However, in contrast with earlier experiments
where the photolysis of NCNO was used to gen-
erate CN radicals, in the present experiments ICN
was photolysed at 266 nm with the fourth har-
monic output of a Nd:YAG laser 0Spectron La-
sers). A fraction of the carrier gas ¯ow was passed
over crystals of ICN before this ¯ow of gas entered
the gas reservoir upstream of the Laval nozzle.
Photodissociation of ICN at this wavelength pro-
duces CN radicals that are overwhelmingly in the
m ˆ 0 level of the X2R ground vibronic state but
‡
over a wide range of rotational levels [15]. CN
radicals were detected by exciting them in the 00,0)
2
band of the CN ꢀA R À X2R † system at ꢁ388
nm, and observing the ¯uorescence in the 00,1)
band at ꢁ420 nm through a narrow band inter-
ference ®lter centred at ꢁ420 nm with a 10 nm
FWHM bandwidth. Kinetic decays were observed
by recording the variation in the laser-induced
¯uorescence 0LIF) signal as a function of the time
delay between the pulses from the photolysis and
probe lasers. These traces of LIF signal versus time
delay were ®tted to single exponential functions,
the ®t being started at time delays sucient to
allow for rotational relaxation. This procedure
yielded pseudo-®rst-order rate coecients related
to the rate of loss of CN radicals.
‡
‡
C2H radicals were generated by pulsed laser
photolysis of C2H2 at 193 nm produced from
transitions in ArF in one of three excimer lasers
0Lambda Physik, Compex 102, Compex 202 or
LPX 210i) operating with a pulse frequency of 10
Hz 0Compex 102 and 202) and 30 Hz 0LPX 210i)
and a pulse duration of ꢁ10 ns. The progress of
C2H reactions was monitored by observing
results
C2H ‡ CH3CBCH.
for
C2H ‡ CH2@C@CH2
and
‡
chemiluminescence from CH ꢀA2D †, which is
2. Experimental
formed as a minor product of the reaction between
C2H and O2 [16]. Details of this method and a
discussion of possible complications are discussed
at length by Chastaing et al. [6]. One of its disad-
vantages is that, to study any reaction of C2H, it is
Rate coecients have been measured for the
reactions of CN and C2H with CH2@C@CH2 and
CH3CBCH at several temperatures between 294

312
D. Carty et al. / Chemical Physics Letters 344 '2001) 310±316
necessary to have both C2H2 and O2 present, one
as the photochemical precursor of the radicals and
the other to generate the chemiluminescent mar-
that these errors are unlikely to exceed 10%. This
estimate has been combined with the 95% con®-
dence limits for statistical error in Table 1a±d and
in Figs. 1 and 2.
‡
ker, CH ꢀA2D †, as well as any other reagent such
as allene or methyl acetylene. The `background
rate', i.e., that with no allene or methyl acetylene
present, is substantial and means that relatively
large concentrations of any third reagent must be
added in order to change the ®rst-order decay rates
appreciably. In the present experiments, reliable
kinetic measurements could not be made on the
C2H ‡ CH2@C@CH2 and C2H ‡ CH3CBCH re-
actions below 63 K because at the concentrations
of allene and methyl acetylene needed to accelerate
the rate of removal of C2H over the background
rate, there was signi®cant loss of species due to the
formation of dimers and higher complexes of the
three molecules present.
In all four cases, the variation of the rate coef-
®cient with temperature is slight. Indeed, for both
reactions of the CN radical, di€erences between the
rate coecients for CH2@C@CH2 and CH3CBCH
are negligible, as are their temperature-depen-
dences. For astrochemical modelling purposes, we
recommend that a temperature independÀen1 t value
of k ˆ ꢀ4:1 Æ 0:5†  10À10 cm3 molecule sÀ1 is
used. For both reactions of the C2H radical, the rate
coecients show a negative dependence on tem-
perature. Non-linear least squares ®ts of our data to
n
the expression k ˆ AꢀT=298 K† in the range
T ˆ 63±296 K gave the following results:
C2H ‡ allene :
k ˆ ꢀ2:0 Æ 0:6†
Àꢀ0:4Æ0:3†
 10À10ꢀT =298 K†
C2H ‡ methyl acetylene :
k ˆ ꢀ2:1 Æ 1:6†
cm3 moleculeÀ1 sÀ1
3. Results
Second-order rate coecients ꢀk† for a given
reaction were obtained by measuring pseudo-®rst-
order rate coecients ꢀk1st† for the disappearance
of the radical, by the methods described in the
previous section, for several 0ꢁ10) concentrations
of C3H4 0CH2@C@CH2 or CH3CBCH) and then
plotting k1st against ‰C3H4Š. The gradient of these
lines yields the value of k appropriate to the par-
ticular reagents and the selected temperature. Us-
ing di€erent nozzles, measurements were made at
several temperatures on each of the four reactions.
Our results are summarised in Table 1a±d and
Figs. 1 and 2. In Fig. 1 the temperature-depen-
dence of the rate coecients for the reaction of CN
and C2H radicals with CH2@C@CH2 are dis-
Àꢀ0:3Æ0:7†
 10À10ꢀT =298 K†
cm3 moleculeÀ1 sÀ1
Average absolute statistical deviations over the
range of experiments were 0.1 andÀ10.3, respec-
s
À1. Based
tively, in units of 10À10 cm3 molecule
on previous measurements of radical-unsaturated
hydrocarbon molecule reactions [5,6,17±19] it is
unlikely that the observed negative temperature-
dependence will continue to temperatures signi®-
cantly below our experimental range. We therefore
recommend for astrochemical models of dense
interstellar clouds, where the temperature is usu-
ally in the range 10±100 K, a temperature inde-
played on
a
log±log plot of k=10À10 cm3
pendent
value
of
k ˆ ꢀ3 Æ 1†  10À10 cm3
moleculeÀ1 sÀ1 versus T=K. Fig. 2 shows the cor-
responding data for the reactions of the CN and
C2H radicals with CH3CBCH. Standard errors
from the second-order plots have been multiplied
by the appropriate Student's t-factor for 95%
con®dence limits. This corresponds to the statisti-
cal error. In addition, some systematic errors may
have arisen from, for example, inaccuracies in the
calibration of ¯ow controllers or in the determi-
nation of the total gas density, but it is estimated
moleculeÀ1 sÀ1 for both reactions.
4. Discussion
The Letter accompanying this one reports the
measurement of rate coecients for the reactions
of C2H radicals with C2H2; O2; CH2@C@CH2 and
CH3CBCH at 103 K [13]. Cooling to this tem-
perature is achieved using a pulsed Laval nozzle.

D. Carty et al. / Chemical Physics Letters 344 '2001) 310±316
313
Table 1
Rate coecients k for the reactions:
T/K
M
‰MŠ=1016
Range of
‰C3H4Š=1014
No. of
points
k=10À10
0molecule cmÀ3
)
0cm3 moleculeÀ1 sÀ1
)
0molecule cmÀ3
)
0a) CN radicals with allene
294
200
112
54
Ar
N2
Ar
Ar
He
He
8.56
5.83
2.79
5.37
4.65
5.05
0.00±1.65
0.00±4.65
0.00±2.58
0.00±5.31
0.40±3.97
0.00±0.86
11
11
12
11
10
9
3:9 Æ 0:4a
4:3 Æ 0:5
4:5 Æ 0:5
4:0 Æ 0:5
3:9 Æ 0:6
4:4 Æ 1:1
27
15
0b) CN radicals with methyl acetylene
294
200
112
54
Ar
N2
Ar
Ar
He
He
8.56
5.83
2.79
5.37
4.65
5.05
0.00±1.58
0.00±4.53
0.00±2.52
0.00±5.19
0.39±3.88
0.00±1.26
11
11
11
12
10
8
4:0 Æ 0:4
4:3 Æ 0:5
4:4 Æ 0:5
3:5 Æ 0:4
3:6 Æ 0:7
3:8 Æ 0:7
27
15
T/K
M
‰MŠ=1016
‰C2H2Š=1013
‰O2Š=1014
Range of
‰C3H4Š=1014
No. of
points
k=10À10
0molecule cmÀ3
)
0molecule cmÀ3
)
0molecule cmÀ3
)
0cm3 moleculeÀ1 sÀ1
)
0molecule cmÀ3
)
0c) C2H radicals with allene
296
149
112
63
Ar
He
Ar
He
23.0
1.42
30.7
2.53
8.79
4.19
38.4
2.12
6.46
4.37
1.1±11.0
0.2±2.1
0.2±2.5
0.1±1.5
7
9
1:8 Æ 0:3
2:8 Æ 0:3
2:9 Æ 0:3
3:5 Æ 0:4
2.71
2.03
10
10
0d) C2H radicals with methyl acetylene
296
149
112
63
Ar
He
Ar
He
24.0
1.42
32.1
2.33
40.2
1.95
6.46
4.37
1.1±7.9
0.2±1.9
0.2±2.2
0.1±1.4
10
10
11
10
1:7 Æ 0:3
2:9 Æ 0:3
2:9 Æ 0:3
2:9 Æ 0:3
2.71
2.03
8.80
4.19
a Errors 0here and throughout the tables) are quoted at the 95% con®dence limit, and include a 10% contribution from possible
systematic error.
Results for the reactions with C3H4 isomers are
shown in Figs. 1 and 2. The results for all four of
these reactions are in satisfactory agreement with
those obtained using our continuous ¯ow CRESU
apparatus and reported in this Letter and previ-
ously [6].
The slightly smaller rate coecients found for
the reactions of C2H compared with the corre-
sponding ones for reactions of CN is consistent
with what has been found before for the reactions
of CN and C2H with C2H4 and C2H2 [6]. There
have been no previous measurements of low tem-
perature rate coecients for the reactions of CN
with CH2@C@CH2 and CH3CBCH. As Fig. 1
shows, our results are consistent with rate coe-
cients determined by Lin and co-workers at room
temperature and above [20]. For the reactions of
C2H with CH2@C@CH2 and CH3CBCH, rate
coecients have been determined at temperatures
down to 155 K by Hoobler and Leone [21]. These
data are also shown on Figs. 1 and 2. As with our
results, they found little di€erence in the values of
k for the two isomers of C3H4. Moreover, their
data for these two reactions are reasonably con-
sistent with ours.
Our experiments measure only the total rate of
removal of the radical species ꢀCN; C2H† by the
C3H4 reactants. There have been no studies of the
products of these reactions at the low temperatures
or collision energies appropriate to dense inter-

314
D. Carty et al. / Chemical Physics Letters 344 '2001) 310±316
Fig. 1. Rate coecients k for the reactions of CN and C2H
radicals with allene plotted on a log±log scale against temper-
ature. The ®lled symbols show the results of this work: ꢀd†
denoting rate coecients for CN + allene; and ꢀj† denoting
rate coecients for C2H ‡ allene. For the CN + allene reaction,
the horizontal dashed line represents our recommendation for
the rate coecient, k ˆ 4:1  10À10 cm3 moleculeÀ1 sÀ1 and 0n)
shows the results of Butter®eld et al. over the range
T ˆ 297±673K [20]. For the C2H ‡ allene reaction, ꢀÆ show
the results of Hoobler and Leone over the range
T ˆ 159±297 K [21] and 0)) is the result of Vakhtin et al. from
the accompanying Letter [13]. The solid line shows the result of
Fig. 2. Rate coecients k for the reactions of CN and C2H
radicals with methyl acetylene plotted on a log±log scale against
temperature. The ®lled symbols show the results of this work:
ꢀd† denoting rate coecients for CN + methyl acetylene; and
ꢀj† denoting rate coecients for C2H ‡ methyl acetylene. For
the CN + methyl acetylene reaction, the horizontal dashed line
represents our recommendation for the rate coecient,
k ˆ 4:1  10À10 cm3 moleculeÀ1 s1 and 0s) shows the room
temperature result of Sayah et al. [29]. For the
C2H ‡ methyl acetylene reaction, ꢀÆ show the results of
Hoobler and Leone over the range T ˆ 155±298 K [21] and 0))
is the result of Vakhtin et al. from the accompanying Letter
[13]. The solid line shows the result of a non-linear least-squares
a non-linear least-squares ®t to our C2H ‡ allene data giving
À0:4
k ˆ 2:0  10À10 ꢀT=298 K†
cm3 moleculeÀ1 sÀ1 with an av-
®t to our C2H ‡ methyl acetylene data giving k ˆ 2:1  10À10
À0:3
cm3 moleculeÀ1 sÀ1 with an average absolute
erage absolute statistical deviation of 0:1 Â 10À10 cm3
moleculeÀ1 sÀ1 over the range of experiments. The dotted line
ꢀÁ Á ÁÁ† represents the extrapolation of this ®t below the lowest
temperature at which data were obtained.
ꢀT=298 K†
statistical deviation of 0:3 Â 10À10 cm3 moleculeÀ1 sÀ1 over the
range of experiments. The dotted line ꢀÁ Á ÁÁ† represents the ex-
trapolation of this ®t below the lowest temperature at which
data were obtained.
stellar clouds or the atmospheres of planets such as
Titan 0ꢁ0:1±1 kJ molÀ1), and quantitative product
branching ratios are unavailable at any tempera-
ture. However, crossed-molecular beam measure-
ments have been performed at much higher
collision energies 0in the range 20±40 kJ molÀ1) by
Lee, Kaiser and co-workers, coupled in most cases
with potential energy surface calculations [22±25].
In all cases, they concluded that the barrier-free
addition of the radical CN or C2H followed by loss
of an H-atom is either the only or a major path-
way. For CN + allene, cyanoallene ꢀH2C@C@
CHCN† was inferred as the major product [22],
whereas for CN + methyl acetylene, both cya-
nopropyne ꢀCH3CBCCN† and cyanoallene were
observed [22,23]. The C2H ‡ methyl acetylene re-
action is found at
a
collision energy of
38:8 kJ molÀ1 to yield mainly methyl diacetylene
ꢀCH3CBCCBCH†, along with the minor product
ethynyl allene ꢀH2C@C@CHC2H† [24].
In Table 2 we list the rate coecients at 50 K
for the reactions of CN and C2H with alkenes and
alkynes that can be inferred from the present and
previous [5,6] CRESU experiments. In all cases,
the reaction is rapid with rate coecients close to
the collision-determined limit. The rate coecients
for the C2H reactions are consistently smaller, by
about a factor of 2, than those for CN. Table 2
also gives the ionisation energies of the unsatu-
rated hydrocarbons including those for the

D. Carty et al. / Chemical Physics Letters 344 '2001) 310±316
315
Table 2
Comparison of rate coecients for the reactions of CN and C2H radicals with alkenes and alkynes at T ˆ 50 K, estimated from this
work, that of Sims et al. [5] and that of Chastaing et al. [6]. Ionisation energies for the unsaturated hydrocarbons are also listed [11]
kCNꢀT ꢂ 50 K†=10À10
kC HꢀT ꢂ 50 K†=10À10
Ionisation energy
0eV)
2
0cm3 moleculeÀ1 sÀ1
)
0cm3 moleculeÀ1 sÀ1
)
HCBCH
CH3CBCH
H2C@CH2
CH3CH@CH2
CH2@C@CH2
HCBC±CBCH
HCBC±CC±CBCH
HCBCCN
4.7
4.1
4.9
±
1.9
3.6
2.3
3.4
3.9
±
11.40
10.36
10.51
9.73
4.1
±
9.69
10.18
9.5
11.64
±
±
±
±
polyynes C4H2 and C6H2, for which no low tem-
perature kinetic data are available but which may
be synthesised in interstellar clouds by reactions of
C2H radicals with smaller species:
of CN with cyanoacetylene ꢀHCBCCN† and
found it to be ꢀ1:70 Æ 0:08†  10À11 cm3
moleculeÀ1 À1. They also inferred that it is likely
s
that this reaction has a small activation energy
0ꢁ6 kJ molÀ1). If they are correct, then this reac-
tion would be too slow to play a signi®cant role in
the chemistry of dense interstellar clouds where the
temperature is usually 650 K. We note 0see Table
2) that the ionisation energy of HCBCCN is
higher than that for C2H2 so the existence of a
small activation energy for this reaction would not
be inconsistent with the correlation of rates with
ionisation energies that has been used to predict
fast rates at low temperatures for the reactions of
CN and C2H with higher alkynes. Measurements
of rate coecients for these reactions at low tem-
peratures are sorely needed.
C2H ‡ CmHn ! Cm‡2Hn ‡ H
and may be converted to cyanopolyynes by reac-
tion with CN radicals:
CN ‡ CmHn ! CmHnÀ1CN ‡ H
as proposed by Smith and Sims [7]. Seki et al. have
measured a room temperature rate coecient for
CN ‡ C4H2 of ꢀ4:2 Æ 0:2†  10À10 cm3 moleculeÀ1
sÀ1 [26], and as C4H2 and C6H2 both have lower
ionisation energies than C2H2 it is likely that they
react at low temperatures with both CN and C2H
radicals at rates at least comparable to those for
the reactions of C2H2. More recently, Fukuzawa
et al. [27] have performed ab initio quantum cal-
culations on the reactions of C2H and CN with
polyynes, and concluded that these reactions pro-
ceed without barriers, and, under the cold condi-
tions of molecular clouds, lead exclusively to the
products shown above.
Our results indicate that reactions of CN and
C2H with unsaturated hydrocarbons occur rapidly
at low temperatures and should be included in
chemical models of dense interstellar clouds and
possibly in those of other astrophysical environ-
ments, such as the atmosphere of Titan. An in-
teresting question that remains unresolved is
whether CN and C2H radicals react rapidly at low
temperatures with cyanopolyynes themselves.
Halpern and co-workers [28] have measured the
room temperature rate coecient for the reaction
Acknowledgements
We are grateful to EPSRC for substantial re-
search grants to construct and extend the CRESU
apparatus in Birmingham and for support of this
work by the European Commission through a
TMR Research Network grant, contract FMRX-
CT97-0132 0DG12-MIHT), for research on As-
trophysical Chemistry. DC also thanks EPSRC for
the award of a research studentship.
References
[1] I.W.M. Smith, B.R. Rowe, Acc. Chem. Res. 33 02000) 261.
[2] E. Herbst, Ann. Rev. Phys. Chem. 46 01995) 27.

316
D. Carty et al. / Chemical Physics Letters 344 '2001) 310±316
[3] I.R. Sims, J.L. Que€elec, A. Defrance, C. Rebrion-Rowe,
D. Travers, P. Bocherel, B.R. Rowe, I.W.M. Smith, J.
Chem. Phys. 100 01994) 4229.
[17] I.R. Sims, in: R.G. Compton, G. Hancock 0Eds.), Research
in Chemical Kinetics, Blackwell Science, Oxford, 1997, p.
121.
[4] I.R. Sims, J.L. Que€elec, A. Defrance, C. Rebrion-Rowe,
D. Travers, B.R. Rowe, I.W.M. Smith, J. Chem. Phys. 97
01992) 8798.
[18] A. Canosa, I.R. Sims, D. Travers, I.W.M. Smith, B.R.
Rowe, Astron. Astrophys. 323 01997) 644.
[19] I.R. Sims, I.W.M. Smith, P. Bocherel, A. Defrance, D.
Travers, B.R. Rowe, J. Chem. Soc.-Faraday Trans. 90
01994) 1473.
[5] I.R. Sims, J.L. Que€elec, D. Travers, B.R. Rowe, L.B.

Herbert, J. Karthauser, I.W.M. Smith, Chem. Phys. Lett.
211 01993) 461.
[6] D. Chastaing, P.L. James, I.R. Sims, I.W.M. Smith,
Faraday Discuss. 01998) 165.
[20] M.T. Butter®eld, T. Yu, M.C. Lin, Chem. Phys. 169 01993)
129.
[21] R.J. Hoobler, S.R. Leone, J. Phys. Chem. A 103 01999)
1342.
[7] I.W.M. Smith, I.R. Sims, J. Chem. Soc. Faraday Trans. 89
01993) 2166.
[22] N. Balucani, O. Asvany, L.C.L. Huang, Y.T. Lee, R.I.
Kaiser, Y. Osamura, H.F. Bettinger, Astrophys. J. 545
02000) 892.
[8] E. Herbst, H.H. Lee, D.A. Howe, T.J. Millar, Mon. Not.
Roy. Astron. Soc. 268 01994) 335.
[9] R.P.A. Bettens, H.H. Lee, E. Herbst, Astrophys. J. 443
01995) 664.
[23] L.C.L. Huang, N. Balucani, Y.T. Lee, R.I. Kaiser,
Y. Osamura, J. Chem. Phys. 111 01999) 2857.
[24] R.I. Kaiser, C.C. Chiong, O. Asvany, Y.T. Lee, F. Stahl,
P.V. Schleyer, H.F. Schaefer, J. Chem. Phys. 114 02001)
3488.
[10] E. Martinez, B. Cabanas, A. Aranda, J. Albaladejo, R.P.
Wayne, J. Chem. Soc.-Faraday Trans. 93 01997) 2043.
[11] S.G. Lias, J.E. Bartmess, J.F. Liebman, J.L. Holmes, R.D.
Levin, W.G. Mallard, J. Phys. Chem. Ref. Data 17 01988) 1.
[12] A.B. Vakhtin, D. Heard, I.W.M. Smith, S.R. Leone, J.
Phys. Chem. 2001, in press.
[25] F. Stahl, P.V. Schleyer, H.F. Bettinger, R.I. Kaiser, Y.T.
Lee, H.F. Schaefer, J. Chem. Phys. 114 02001) 3476.
[26] K. Seki, M. Yagi, M.Q. He, J.B. Halpern, H. Okabe,
Chem. Phys. Lett. 258 01996) 657.
[13] A.B. Vakhtin, D. Heard, I.W.M. Smith, S.R. Leone,
Chem. Phys. Lett. 2001, accompanying paper, in press.
[14] P.L. James, I.R. Sims, I.W.M. Smith, M.H. Alexander,
M.B. Yang, J. Chem. Phys. 109 01998) 3882.
[15] W.H. Fisher, R. Eng, T. Carrington, C.H. Dugan, S.V.
Filseth, C.M. Sadowski, Chem. Phys. 89 01984) 457.
[16] K. Devriendt, H. VanLook, B. Ceursters, J. Peeters, Chem.
Phys. Lett. 261 01996) 450.
[27] K. Fukuzawa, Y. Osamura, H.F. Schaefer, Astrophys.
J. 505 01998) 278.
[28] J.B. Halpern, G.E. Miller, H. Okabe, Chem. Phys. Lett.
155 01989) 347.
[29] N. Sayah, X.C. Li, J.F. Caballero, W.M. Jackson,
J. Photochem. Photobiol. A-Chem. 45 01988) 177.