Kinetic study of the reaction of chlorine atoms with chloroform in the gas phase

Article abstract of DOI:10.1016/j.cplett.2008.12.081

The kinetics of the gas-phase reactions of chlorine atoms with H-chloroform and D-chloroform was studied experimentally. The relative rate method was applied using Cl + CH₃Br as the reference reaction. The rate constants for H-abstraction from

Full text of DOI:10.1016/j.cplett.2008.12.081

Chemical Physics Letters 469 (2009) 250–254
Contents lists available at ScienceDirect
Chemical Physics Letters
journal homepage: www.elsevier.com/locate/cplett
Kinetic study of the reaction of chlorine atoms with chloroform in the gas phase
Agnieszka A. Gola, Dariusz Sarzy n´ ski, Andrzej Dry s´ , Jerzy T. Jodkowski *
Department of Physical Chemistry, Wroclaw Medical University, pl. Nankiera 1, 50-140 Wroclaw, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
The kinetics of the gas-phase reactions of chlorine atoms with H-chloroform and D-chloroform was stud-
Received 20 October 2008
In final form 22 December 2008
Available online 30 December 2008
ied experimentally. The relative rate method was applied using Cl + CH
rate constants for H-abstraction from CHCl (k ) and D-abstraction from CDCl
temperature range of 297–527 K and at total pressure of 100 Torr. The derived temperature dependencies
3
Br as the reference reaction. The
3
H
3
D
(k ) were measured in the
ꢁ
12
of the rate constants are given by k
H
= (4.8 ± 0.5) ꢀ 10
ꢀ exp(ꢁ1160 ± 30/T) and k
D
= (6.4 ± 2.1) ꢀ
ꢁ12
3
ꢁ1 ꢁ1
10
ꢀ exp(ꢁ1660 ± 110/T) cm molecule
s
. The kinetic isotope effect described by the ratio k
H
/k
D
was (0.77 ± 0.25) ꢀ exp(500 ± 95/T).
Ó 2009 Elsevier B.V. All rights reserved.
1
. Introduction
In this study, we present measurements of the rate constant for
the hydrogen abstraction reaction (1) using the relative rate tech-
nique. The experiments were performed in the temperature range
Chloroalkanes and products of their atmospheric reactions are
considered toxic and biocumulative species. Chloroform is among
the most abundant chlorine-containing compounds in the atmo-
sphere. Ninety percent or more of the atmospheric CHCl is emit-
3
H
of 297–527 K to examine the temperature dependence of k and
the values of the activation energy available in literature. The reac-
tion of deuterated chloroform
ted from natural sources such as the oceans, soil, and termites.
The rest is a result of such industrial activity as paper manufactur-
ing, water treatment, and waste incineration [1–5]. The products of
CDCl
3
þ Cl ! CCl
3
þ DCl
ð2Þ
was also included in our kinetic investigation. The independent
the atmospheric destruction of CHCl
catalytic atmospheric reaction cycles responsible for the depletion
of the ozone layer. The stratospheric lifetime of CHCl is estimated
to be over 3 years [5,6]. The chemical degradation initiated by OH
radicals is a major loss pathway for atmospheric chloroform [6,7].
However, in marine areas and polar regions, especially during sun-
rise events, reactions with Cl atoms can also be of some importance
3
may be involved in various
measurement of the rate constant k for reaction (2) enabled us
D
to estimate the kinetic isotope effect (KIE). Values of KIE provide
information useful for interpreting the stable isotope composition
of the organic compounds in the atmosphere. The experimental
data on the kinetics of isotopomers of CHCl are very limited. To
3
the best of our knowledge, only the old study by Clyne and Walker
[9] showed measurements of the kinetic isotope rate constant ratio
3
[
5]. The reaction of chloroform with chlorine atoms
H D
k /k in a wide temperature range. The hydrogen abstraction
reaction
CHCl þ HCl
has been the subject of several experimental [8–17] and theoretical
studies [18–21]. Results of the experimental investigations show,
however, substantial scattering in the values of the measured rate
3
þ Cl ! CCl
3
ð1Þ
CH Br þ Cl ! CH
3
2
Br þ HCl
ð3Þ
was selected as the reference reaction in our investigations. The rate
constant k3 for reaction (3) and its temperature dependence are
well known and have been estimated with satisfactorily accuracy
[24].
H
constant k for reaction (1). Distinct differences occur in values
for either the pre-exponential factor or the activation energy. Esti-
mates of the rate constant k
H
at room temperature cover a range
ꢁ
13
3
ꢁ1 ꢁ1
of (0.4–3.2) ꢀ 10
cm molecule
s . Despite the discrepancies
2
. Experimental
in the values of kinetic measurements, the most recent IUPAC and
NASA evaluations of the kinetic data recommend very similar val-
ꢁ13
ꢁ13
The experiments were carried out in a greaseless static system
ues of the rate constant, i.e. k
H
of 1.1 ꢀ 10
[22] and 1.2 ꢀ 10
3
3
ꢁ1 ꢁ1
using a cylindrical Pyrex reactor of ca. 250 cm (dead volume: ꢂ2%)
placed in an aluminium heating block. The gas-phase reactions of
chlorine atoms with chloroform and D-chloroform were investi-
gated using the reaction with bromomethane as the reference at
six temperatures in the temperature range of 297–527 K. The tem-
perature was maintained by a power regulator connected to a tem-
cm molecule
s
[23] at 298 K, respectively.
*
Corresponding author. Fax: +48 71 7840230.
E-mail address: jurek@kchfiz.am.wroc.pl (J.T. Jodkowski).
0
009-2614/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2008.12.081

A.A. Gola et al. / Chemical Physics Letters 469 (2009) 250–254
251
perature regulator equipped with a Pt-100 resistance sensor placed
inside the chamber of the heating block. A chrome–nickel thermo-
couple located in the middle of the side wall of the reactor was
used to measure the temperature. The temperature of the reaction
cell was constant to within 0.5° during the experiment. The reac-
tants were introduced into the cell by expansion, starting from
where [AH]
0
and [BH]
0
are the starting concentrations of the reac-
tants AH and BH and [AH]
ven time t. The plot of ln([AH]
linear, which enables an estimation of the rate constant ratio k
as the slope. Consequently, the rate constant k can easily be de-
rived if k is known with sufficient accuracy.
The results of measurements of the concentrations of the reac-
tant AH (CHCl or CDCl ) and the reference reactant BH (CH Br) at
the same time are analyzed according to Eq. (4). The CCl and
CH Br radicals formed in competing reaction channels may inter-
t
and [BH]
t
are the concentrations at a gi-
) vs. ln([BH] /[BH] ) should be
/k
0
/[AH]
t
0
t
A
B
A
B
CHCl
finally filled up with N
All the experiments were carried out at a total pressure of
00 Torr with the partial pressure of Cl varying from 7 to 10 Torr.
The partial pressure of CHCl /CDCl and CH Br was in the range of
.5–4 Torr. The mixtures of Cl in N were prepared (at least 48 h
before the first use) from pure Cl (>99.5%) and N (>99.99%) and
stored in a 4L blackened Pyrex bulb. The reactants CHCl , CDCl
and CH Br were degassing using the freeze-pump-thaw method
3
or CDCl
3
followed by CH
3 2 2
Br and a mixture of Cl in N , and
2
to a total pressure of 100 Torr.
3
3
3
3
1
2
2
fere with each other or initiate different subsequent reactions.
Eq. (4) is exact assuming an absence of fast radical processes in
the reaction system resulting in changes in the concentrations of
3
3
3
0
2
2
2
2
3
,
CHCl
magnitude of the rate constant or the concentrations of reactants.
The reverse reactions CCl Br + HCl
+ HCl/DCl (ꢁ1,ꢁ2) and CH
(ꢁ3), which reproduce the parent compounds CHCl /CDCl and
CH Br, are very slow [22,23] and are hence considered unimpor-
tant processes even at a higher degree of conversion of the reac-
tants. The major fate of the CCl and CH Br radicals formed is
reaction with molecular chlorine, Cl , because of the large excess
of Cl in the reaction system. The influence of the secondary radical
3 3 3
, CDCl , and CH Br. The reaction rate depends on either the
3
3
and stored under vacuum in light-tight containers. The pressure
was measured by a Model 127 A MKS Baratron capacitance
manometer equipped with two gauges (0–100 and 0–1000 Torr).
A Xe arc lamp (Osram XBO 150W/1 OFR) served as the light source.
The optical train consisted of a manually operated shutter, a con-
densing lens, a variable-width slit, and a monochromator with
3
2
3
3
3
3
2
2
the Czerny–Turner optical system (dispersing element
a
2
1
200 line/mm grating and aperture F/4, range: 200–900 nm with
reactions on the kinetics of the primary H-abstraction reactions (1–
3) is negligible under the reaction conditions applied in our
investigations.
The preliminary tests showed that ca. 10 min of mixing time
was sufficient to reach ambient temperature by the reactants in
a band variability of 2–20 nm) of a Hitachi MPF-4 fluorescence
spectrophotometer. The light from the monochromator was intro-
duced into the cylindrical reactor through its bottom window and,
after passing through the reactor, was reflected back by the mirror
placed above the upper window of the reactor. The reaction cell,
feed lines, and optical assembly were housed in a light-tight
enclosure to prevent photolysis initiated by room or stray light.
The absence of such reactions was confirmed by repeated blank
analyses. Chlorine atoms were generated in their ground state
1
0
0
0
.0
.8
.6
.4
X=CHCl3
2
2
(
Cl( P3/2) > 99%, Cl( P1/2) > 1%, at 298 K) [25] by the photolysis of
Cl at 420 nm. The irradiation time (2–20 min) and the slit width
2–20 nm) varied depending on the reaction temperature in order
X=CDCl3
2
(
to obtain appropriate conversion of reactants.
The loss of reactants was monitored by thermal conductivity
gas chromatography (HP, Model 5890 Series II) using
2
a
5 m ꢀ 0.53 mm ParaPlot Q Chrompack capillary column with
temperature–time programming between 70 and 200 °C with he-
lium as the carrier gas. Before and after the experiments, calibra-
tion of the reactive peak area vs. the pressure for each reactant
at any temperature was determined by GC analysis. The reactants
used in this study and their minimal purity were bromomethane
(
99.7%), H-chloroform (99.7%), and D-chloroform (99.7%) and were
from Aldrich. The reaction products were purified three times
using the freeze-pump-thaw method.
3
. Results and discussion
Expressions describing the kinetics of the competitive reactions
of two reactants with the same reactive species are the basis of the
relative rate method used. Let us consider the H-abstraction reac-
tion from two reactants, AH and BH, by chlorine atoms
0.2
AH þ Cl ! A þ HCl ðk
BH þ Cl ! B þ HCl ðk
where k and k denote the rate constants describing the rate of for-
A
Þ
B
Þ
A
B
mation of the channel reaction products. Assuming that reactants
AH and BH are only consumed in reaction with chlorine atoms, their
concentrations change during the reactions according to the
relation:
0
.0
0
.0
1.0
2.0
3.0
4.0
ln([CH Br]0/[CH Br]t)
3 3
½
AHꢃ
k
k
A
½BHꢃ
Fig. 1. Relative rate data obtained at 527 K and a pressure of 100 Torr for the
0
0
ln
¼
ꢄ ln
ð4Þ
3 3 3
reactions of Cl with CHCl (d) and CDCl (j) using CH Br as the reference
½
AHꢃ
B
½BHꢃ
t
t
compound.

2
52
A.A. Gola et al. / Chemical Physics Letters 469 (2009) 250–254
Table 1
cursor ensured us that chloroform is chemically and
photochemically stable in the reaction chamber. It is not produced
by the oxidation of impurities and is not adsorbed on the cell walls.
The values of the rate constants for reactions (1) and (2) were
determined at six temperatures between 297 and 527 K at a nearly
constant total pressure of 100 Torr. In Fig. 1 are shown sample ki-
netic data obtained from the experiments plotted according to Eq.
The measured rate constant ratios k
and derived values of the kinetic isotope effect k
H
/k
3
and k
D
/k
3
, the absolute values of k
H D
and k ,
H
/k
D
.
T (K)
k
/k
H 3
1013 ꢀ k
H
k
D
/k
3
1014 ꢀ k
D
H D
k /k
3
ꢁ1 ꢁ1
3
ꢁ1 ꢁ1
(
cm molecule
s
)
(cm molecule
s )
2
3
3
3
4
5
97.3 0.240 1.04 (0.4–3.2)a
26.9 0.247 1.47
37.0 0.226 1.49
83.3 0.238 2.33
48.6 0.244 3.77
26.9 0.235 5.56
0.0512
0.0772
0.0704
0.0817
0.117
2.21
4.59
4.62
8.00
18.1
25.1
4.7
3.2
3.2
2.9
2.1
2.2
(
4) for the reaction of Cl with CHCl
spect to CH Br at 527 K. The kinetic expression obtained by Piety
et al. [24] describing the temperature dependence of the rate con-
3 3
and CDCl measured with re-
3
0.106
3
a
stant for the reaction CH
cule
in our experiments. The results of the measurements were ana-
lyzed using a weighted least squares procedure which included
uncertainties in reactant concentrations and allowed a zero-point
3
Br + Cl ? CH
2
Br + HCl of k
3
/cm mole-
H
The range of estimated values of k at room temperature from Refs. [8–17].
ꢁ1
ꢁ1
ꢁ12
1.42
s
= 3.32 ꢀ 10
ꢀ (T/298)
ꢀ exp(ꢁ605/T) was utilized
the cell and confirmed that dark reactions were never observed.
Possible photolysis or thermal decomposition of the organic reac-
tants were taken into account. The mixtures of organics were irra-
diated in the absence of Cl
than 60 min. Neither photolysis nor thermal dissociation reactions
of CHCl /CDCl and CH Br were observed. Any kinetic experiment
at a temperature above 295 K was preceded by tests for dark reac-
tions. The mixtures of organic species and molecular chlorine were
prepared and placed in the dark for at least 60 min. At tempera-
tures below 448 K, the reaction of the organic species with molec-
ular chlorine was negligible in the absence of photolytic light. No
destruction of the reactants was observed. At the highest temper-
ature of this study (527 K), thermal dissociation of Cl
occurs, which allowed us to use the thermolysis of Cl
tional source of chlorine atoms. A blind test with the radical pre-
offset. Values of the ratios k
H 3 D 3
/k and k /k as well as the absolute
values of k and k at six temperatures (297, 326, 337, 383, 448,
H
D
2
at the highest temperature for more
and 527 K) are gathered in Table 1.
H D
Fig. 2 shows the Arrhenius plot of k and k . There is little scat-
tering of experimental points around the regression lines, which
indicates a satisfactory repeatability of the experiments. The tem-
perature dependence of the rate constant k can be expressed in
H
the temperature range of 297–527 K as:
3
3
3
¼ ð4:8 ꢅ 0:5Þ ꢀ 10ꢁ12 ꢀ expðꢁ1160 ꢅ 30=TÞ cm molecule
3
ꢁ1 ꢁ1
k
H
s
ð5Þ
2
molecules
2
as an addi-
with the 2
r
error limits. The value of (1.0 ± 0.2) ꢀ 10
ꢁ13
cm mole-
derived from the above expression at room temper-
3
ꢁ1 ꢁ1
cule
s
of k
H
ature is in good agreement with available kinetic data. Our value of
ꢁ13
ꢁ13
k
H
is only slightly lower than those of 1.1 ꢀ 10 , 1.2 ꢀ 10 , and
-
-
-
12.0
12.5
13.0
CHCl +Cl
3
-11.5
CDCl +Cl
3
Knox (1962)
Clyne & Walker (1973)
Jeoung et al. (1991)
Beichert et al. (1995)
Catoire et al. (1996)
Talhaoui et al. (1996)
Orlando (1999)
Bryukov et al. (2002)
this study
-12.0
-12.5
-13.0
-13.5
-
13.5
14.0
-
1
.5
2.0
2.5
3.0
3.5
1
2
3
4
5
1
000K/T
1000 K/T
Fig. 2. Arrhenius plot for the reactions of Cl atoms with CHCl
3
(d) and CDCl
3
(j) in
3
Fig. 3. Arrhenius plot for the CHCl + Cl reaction comparing available results of
the temperature range from 297 to 527 K.
kinetic measurements. The solid line corresponds to the plot of Eq. (5).

A.A. Gola et al. / Chemical Physics Letters 469 (2009) 250–254
253
ꢁ
13
3
ꢁ1 ꢁ1
ꢁ13
3
ꢁ1 ꢁ1
1
.4 ꢀ 10
cm molecule
s
obtained by Catoire et al. [15], Beic-
4.2 ꢀ 10
cm molecule
s
by Bryukov et al. [17] and confirms
hert et al. [12] and Brahan et al. [13], and Orlando [16], respectively.
the reliability of the rate constant k and its dependence on tem-
perature derived in this study.
The kinetic analysis of the experiments performed for the reac-
tion of deuterated chloroform with chlorine leads to the expression
H
ꢁ
13
It is also very close to the values of 1.1 ꢀ 10
[22] and
ꢁ13
3
ꢁ1 ꢁ1
1
.2 ꢀ 10
cm molecule
s
[23] recommended by recent ki-
netic data evaluations. However, these evaluations probably do
not take into account results of the newest study by Bryukov
¼ ð6:4ꢅ2:1Þꢀ10ꢁ ꢀexpðꢁ1660ꢅ110=TÞ cm molecule
12
3
ꢁ1 ꢁ1
s
ꢁ
14
3
ꢁ1 ꢁ1
K
D
et al. [17], who obtained a value of 9.5 ꢀ 10
cm molecule
s
at room temperature. This value is also very close to that found in
our study. Only the results of the oldest and probably least credible
experiments [8,9] deviate distinctly from our estimates. The values
ð6Þ
which is also valid in the temperature range of 297–527 K. The
abstraction of deuterium from CDCl proceeds distinctly more
slowly than the abstraction of a hydrogen atom from unsubstituted
ꢁ
14
3
ꢁ1 ꢁ1
3
of 4.1 ꢀ 10
cm molecule
s
obtained by Knox et al. [8] and
by Clyde and Walker [9] mark out
ꢁ13
3
ꢁ1 ꢁ1
3
.2 ꢀ 10
cm molecule
s
ꢁ14
3
chloroform. The value of
k
D
of (2.6 ± 0.9) ꢀ 10
cm mole-
the lower and upper limits of the reported experimental
measurements.
ꢁ1 ꢁ1
cule
s
at room temperature is one fourth of that of k at
H
2
98 K. The significance of the kinetic isotope effect declines with
rising temperature. At 500 K the value of KIE only slightly exceeds
. The temperature dependence of KIE, described by the ratio
/k , can be expressed in the form
The magnitude of the exponential parameter in Eq. (5) indicates
a low value of the activation energy for the reaction studied. This
implies a weak dependence of k
are only a few measurements of k
higher than ambient, so the temperature dependence of k
recognized. Results of former kinetic measurements are compared
in Fig. 3. The most credible investigations were performed by Bryu-
kov et al. [17], who did a thorough study of the reaction kinetics in
2
H
on temperature. However, there
at temperatures considerably
is less
k
H
D
H
k
H
=k ¼ ð0:77 ꢅ 0:25Þ ꢀ expð500 ꢅ 95=TÞ
The derived values of KIE are in line with the results of Clyne
and Walker [9]. As can be seen in Fig. 4, the agreement is good,
especially at higher temperatures. The largest difference in the val-
ues of KIE occurs at the lowest temperatures of both studies, i.e. at
room temperature. However, even in this case the difference does
not exceed 20% of the KIE value. This supports the reliability of the
derived temperature dependence of KIE.
D
ð7Þ
H
ꢁ13
3
the range of 297–854 K. Our value of 2.6 ꢀ 10
cm mole-
ꢁ
1 ꢁ1
cule
s
derived from Eq. (5) at 400 K is in good agreement with
ꢁ13
the values of 2.4 ꢀ 10
of Bryukov et al. [17] and
of Talhaoui et al. [14]. At a temper-
ꢁ
13
3
ꢁ1 ꢁ1
2
.5 ꢀ 10
cm molecule
s
ꢁ
13
3
ature of 500 K our k
H
reached a value of 4.7 ꢀ 10
cm mole-
ꢁ1
ꢁ1
cule
s
.
This is also in line with the estimate of
4
. Conclusion
5
The rate constants and their temperature dependence for the
3 3
reactions of chlorine atoms with CHCl and CDCl were estimated
using the relative rate method. Isotopic substitution distinctly
changes the reaction rate. The rate constant for the reaction of D-
chloroform is consequently several times lower than that of H-
3
abstraction from CHCl . The kinetic isotope effect distinctly de-
pends on temperature. The results of our investigations are in very
good accordance with the recently reported measurements and
values recommended by kinetic data evaluations. The analytical ki-
netic expressions derived in this study allow a successful descrip-
4
3
tion of the reaction kinetics and KIE in
a wide range of
temperature, which has significant importance for modelling of
the kinetics of complex reaction systems in the gas phase.
References
[
[
[
1] M.A.K. Khalil, R.A. Rasmussen, S.D. Hoyt, Tellus B 35 (1983) 266.
2] Y.L. Yung, M.B. McElroy, S.C. Wofsy, Geophys. Res. Lett. 2 (1975) 397.
3] M.A.K. Khalil, R.A. Rasmussen, J.R.J. Fremch, J.A. Holt, J. Geophys. Res. 95 (1990)
3
619.
4] E.J. Hoekstra, J.H. Duyzer, E.W.B. de Leer, U.A.T. Brinkman, Atmos. Environ. 35
2001) 61.
5] B.J. Finnlayson-Pitts, J.N. Pitts Jr., Chemistry of the Upper and Lower
Atmosphere, Academic Press, San Diego, 2000.
6] T.P. Kindler, W.L. Chameides, P.H. Wine, D.M. Cunnold, F.N. Alyea, J.A. Franklin,
J. Geophys. Res. 100 (1995) 1235.
7] M.K.W. Ko et al., Very short lived halogen and sulfur substances (Chapter 2), in:
Scientific Assessment of Ozone Depletion, Global Ozone Research and
Monitoring Project, World Meteorological Organization, Geneva, Switzerland,
[
[
[
[
(
2
1
2
003.
Clyne&Walker
this study
[
[
8] J.H. Knox, Trans. Faraday Soc. 58 (1962) 275.
9] M.A.A. Clyne, R.F. Walker, J. Chem. Soc. Faraday Trans. I 69 (1973) 1547.
[10] T.N. Bell, K.A. Perkins, P.G. Perkins, J. Phys. Chem. 81 (1977) 2610.
11] S.C. Jeoung, K.Y. Choo, S.W. Benson, J. Phys. Chem. 95 (1991) 7282.
12] P. Beichert et al., J. Phys. Chem. 99 (1995) 13156.
13] K.M. Brahan, A.D. Hewitt, G.D. Boone, S.A. Hewitt, Int. J. Chem. Kinet. 28 (1996).
14] A. Talhaoui, B. Louis, B. Meriaux, P. Devolder, J.P. Sawerysyn, J. Phys. Chem. 100
(1996) 2107.
[
[
[
[
3
00
400
500
T (K)
[
15] V. Catoire, R. Lesclaux, W.F. Schneider, T.J. Wallington, J. Phys. Chem. 100
(1996) 14356.
H D
Fig. 4. Comparison of the derived values of the kinetic isotope effect k /k (d –
experimental points). The full line represents the plot of Eq. (7)) with results
obtained by Clyne and Walker [9].
[16] J.J. Orlando, Int. J. Chem. Kinet. 31 (1999) 515.
17] M.G. Bryukov, I.R. Slagle, V.D. Knyazev, J. Phys. Chem. A 106 (2002) 10532.
[

2
54
A.A. Gola et al. / Chemical Physics Letters 469 (2009) 250–254
[
[
[
18] S.M. Senkan, D. Quam, J. Phys. Chem. 96 (1992) 10837.
19] Q.S. Li, D.X. Xu, S.W. Zhang, Chem. Phys. Lett. 384 (2004) 20.
20] J.F. Xiao, Z.S. Li, Y.H. Ding, J.Y. Liu, X.R. Huang, C.C. Sun, J. Phys. Chem. A 106
[23] S.P. Sander et al., Chemical kinetics and photochemical data for use in
atmospheric studies, NASA Panel for Data Evaluation, JPL Publication 06-2,
Pasadena, CA, 2006.
(
2002) 320.
[24] C.A. Piety, R. Soller, J.M. Nicovich, M.L. McKee, P.H. Wine, Chem. Phys. 231
(1998) 155.
[
[
21] V.D. Knyazev, J. Phys. Chem. A 107 (2003) 11082.
22] R. Atkinson et al., Summary of evaluated kinetic and photochemical data for
atmospheric chemistry, IUPAC Subcommittee on Gas Kinetic Data Evaluation
for Atmospheric Chemistry, NIST, Gaitherburg, Maryland, 2006.
[25] A.R. Ravishankara, P.H. Wine, J. Chem. Phys. 72 (1980) 25.