Kinetics of the reaction of Cl atoms with CHCl3 over the temperature range 253-313 K

Article abstract of DOI:10.1016/j.cplett.2010.06.026

The reaction CHCl₃ + Cl → CCl₃ + HCl was studied in the atmospherically relevant temperature range from 253 to 313 K and in 930 mbar of N₂ diluent using the relative rate method. A temperature dependent reaction rate constant, valid in the temperature range 220-330 K, was determined by a fit to the result of the present study and that of Orlando (1999); k = (3.77 ± 0.32) × 10^-12 exp((-1011 ± 24)/T) cm³ molecule^-1 s^-1.

Full text of DOI:10.1016/j.cplett.2010.06.026

Chemical Physics Letters 494 (2010) 160–162
Contents lists available at ScienceDirect
Chemical Physics Letters
journal homepage: www.elsevier.com/locate/cplett
Kinetics of the reaction of Cl atoms with CHCl over the temperature range
3
253–313 K
*
Elna J.K. Nilsson , Janus Hoff, Ole John Nielsen, Matthew S. Johnson
Copenhagen Center for Atmospheric Research, Department of Chemistry, University of Copenhagen, Universitetsparken 5, 2100 Copenhagen, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
The reaction CHCl + Cl ? CCl + HCl was studied in the atmospherically relevant temperature range from
3
3
Received 29 April 2010
In final form 8 June 2010
Available online 11 June 2010
2
253 to 313 K and in 930 mbar of N diluent using the relative rate method. A temperature dependent reac-
tionrateconstant,validinthetemperaturerange220–330 K,wasdeterminedbyafittotheresultofthe pres-
ꢁ12
3
ꢁ1 ꢁ1
ent study and that of Orlando (1999); k = (3.77 ± 0.32) ꢀ 10
exp((ꢁ1011 ± 24)/T) cm molecule
s .
Ó 2010 Elsevier B.V. All rights reserved.
ꢁ
12
3
1
. Introduction
ing the Arrhenius expression: k = 3.31 ꢀ 10
exp(ꢁ990/T) cm
ꢁ
1 ꢁ1
molecule
s . A slightly different temperature dependence
Chloroform is present in the atmosphere with an average global
is currently recommended by IUPAC [13]: k(220–500 K) = 2.4 ꢀ
ꢁ12
3
ꢁ1 ꢁ1
atmospheric concentration of about 18.5 pptv, with seasonal, alti-
tudinal, and latitudinal variability [1]. Due to its reaction with hy-
droxyl radicals chloroform has a relatively short lifetime in the
atmosphere, and thus high variability in concentration. Global
atmospheric chloroform budgets have been presented by several
groups [1–3]. The inventories of McCulloch and of Khalil and
Rasmussen estimate the total emission of chloroform to be
10
exp(ꢁ920/T) cm molecule
s
.
The recommendations
build on the studies by Orlando in the range 222–298 K [14] and
Bryukov et al. in the range 297–854 K [15]. Sparse data and dis-
crepancies in the literature call for a new determination of the
temperature dependent rate coefficients of the title reaction.
Recently Gola et al. published a study where they used the relative
rate method to investigate the temperature dependence of the
reaction in the range 297–527 K [16]. In the present study we have
studied the reaction in the atmospherically relevant range of
253–313 K.
ꢁ
1
ꢁ1
6
60 ± 220 Gg yr and 350–600 Gg yr , respectively.
Sources of chloroform to the atmosphere are both natural and
anthropogenic. The pulp and paper industry, and drinking and
wastewater treatment are significant anthropogenic sources, while
a smaller contribution comes from the pharmaceutical industry
2. Experimental method
[
1,2,4–6]. The dominant natural source is an as yet unidentified,
probably biological, process in the oceans, while soil processes con-
stitute the second most important contribution [2]. The uncertain-
ties in the natural sources are large.
In relative rate studies the rate of a reaction is determined rel-
ative to the rate of a well known reference reaction. The method
relies on the assumption that no reactions other than the investi-
gated one consume the reactants to a significant degree. If this is
true the following expression is valid
While chloroform is not classified as a greenhouse gas or con-
sidered to be a threat to the ozone layer, its reactivity is important
to our understanding of the behavior of the halogenated methane
derivatives. Reaction with Cl atoms is an often used method in
investigations of the oxidation of atmospherically relevant haloge-
nated hydrocarbons in smog chambers. For the analysis of such
experiments well determined rate coefficients are needed. A num-
ber of absolute and relative rate studies have been reported for the
½
Aꢂ
k
k
A
½Bꢂ
ln
0
0
ln
¼
ð1Þ
½
Aꢂt
B
½Bꢂt
, [A] , [B]
where, [A]
0
t
0
and [B]
t
, are the concentrations of the com-
pounds at times 0 and t. In a plot of ln([A]
B] ) the slope will equal the relative rate of the reaction.
The rate of the title reaction, Eq. (R1), was determined relative
0 t 0
/[A] ) versus ln([B] /
[
t
3
CHCl + Cl reaction at room temperature [7–11], unfortunately the
to reactions Eqs. (R2) and (R3).
results from different studies show significant scatter. The value
for the rate coefficient at 298 K currently recommended by JPL is
CHCl þ Cl ! CCl þ HCl
3
3
ðR1Þ
ðR2Þ
ðR3Þ
ꢁ
13
3
ꢁ1 ꢁ1
1
[
.2 ꢀ 10
cm molecule
s . The latest JPL data evaluation
CH
CH
2
F
2
þ Cl ! CHF þ HCl
2
12] recommend a temperature dependent rate coefficient follow-
4
þ Cl ! CH þ HCl
3
Cl atoms were produced by the photolysis of Cl
2
using broad-
*
Corresponding author. Present address: Department of Physics, Combustion
band UV-A lamps from OSRAM, kmax ꢃ 350 nm, leading to the pro-
Physics, Lund University, Box 118, SE-22100 Lund, Sweden.
E-mail address: elna.heimdal_nilsson@forbrf.lth.se (E.J.K. Nilsson).
duction of ground state chlorine atoms
0
009-2614/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2010.06.026

E.J.K. Nilsson et al. / Chemical Physics Letters 494 (2010) 160–162
161
Cl
2
þ h
m
! 2Cl
ðR4Þ
gas and experiments were done at a total pressure of 930 mbar.
Chemicals were purified by three freeze–pump–thaw cycles. FTIR
spectra were analyzed using a non-linear least squares fitting pro-
cedure developed by Griffith [18]. The errors in the spectral fit
were used to determine the error in the relative rate according to
Partial pressures of chloroform and the reference compounds
were in the range of 10–90 bar, and partial pressures of Cl were
l
2
1
0–20 times higher than the total pressure of chloroform and refer-
ence compound in each experiment. Details on the experiments are
presented in Table 1. In Ref. [17] the photolysis capacity of the exper-
imental system is calculated to be 0.005 s per lamp at room tem-
perature. Up to eight lamps were used in the present study which
the method of York [19]. Reference spectra of CH
taken from the HITRAN database [20], while for CHCl
4
and HCl were
2 2
and CH F
ꢁ1
3
reference spectra were recorded using the same conditions as
employed in the kinetics experiments.
1
3
14
ꢁ3
result in Cl concentrations in the range 1 ꢀ 10 to 5 ꢀ 10 cm
at 298 K. At colder temperatures the lamps are less efficient, at
2
68 K it is estimated to be about 1/3 of the capacity at 298 K.
To check for possible photolytic removal of chloroform and
3. Results and discussion
methane, experiments were performed where no Cl
in the chamber. A reaction mixture of CHCl and CH
bath gas was irradiated for several minutes and no decrease in
the reactants was detected. Experiments where also performed
where the reaction mixture was left in the dark, to check for pos-
sible reactions with Cl
change in concentration of the reactants, CHCl
HCl was probably the result of reaction of Cl
in the system, possibly on the walls.
The relative rate experiments were carried out in a photochem-
ical reactor at the University of Copenhagen. The reaction chamber
is a quartz tube with stainless steel end flanges; it is 2 m long and
has a volume of 101.4 l. The chamber is confined in a temperature
controlled housing enabling experiments in the temperature range
2
was present
in nitrogen
2 2
Fig. 1 shows a relative rate plot of chloroform versus CH F ;
data from two experiments at 313 K are included. This type of plot
is constructed for all experiments to yield the relative reaction
rates at the different temperatures. Table 2 lists the relative reac-
tion rates obtained at each temperature, as well as the absolute
rates calculated using the known reaction rates of the reference
compounds. The result at each temperature is based on two exper-
iments. The errors given for the rate constants in last column of Ta-
ble 2 include uncertainties due to uncertainties in the rate
constants of the reference reactions. The uncertainty in reference
reactions is the largest source of error. Rate coefficients used for
3
4
2
. Small amounts of HCl were seen, but no
, CH or CH
with some impurities
3
2
F
2
4
.
2
ꢁ
17
2
the reference reactions are k(CH
2
F
2
+ Cl) = 1.19 ꢀ 10
N
exp
3
ꢁ1 ꢁ1
ꢁ19
(ꢁ1023/T) cm molecule
s
[21] and k(CH
4
+ Cl) = 5.69 ꢀ 10
2
.49
3
ꢁ1 ꢁ1
N
exp (ꢁ609/T) cm molecule
s
[15].
2
40–330 K. The experimental setup is described in detail by Nils-
son et al. [17]. The spectra were recorded using a Bruker IFS
6v/S infrared spectrometer with a liquid nitrogen cooled MCT
detector. For the experiments with Eq. (R3) as reference reaction
In Fig. 2 the results are plotted together with the results of pre-
vious studies in the temperature range 220–330 K. The result of the
present work is in agreement with the temperature dependence
study by Orlando [14], when Orlando’s results have been put on
an absolute scale using the same rate constant for the reference
6
ꢁ1
a resolution of 0.125 cm was used to capture the fine structure
of the methane spectra, while for experiments with Eq. (R2) as
the reference reaction a resolution of 0.5 cm
reaction CH
4
+ Cl, as used in the present work [15]. In addition to
Br + Cl
ꢁ
1
was sufficient.
CH + Cl, it is interesting to note that Orlando also used CH
4
3
Reactant and reference compounds were monitored using absorp-
as a reference reaction and when the data or Orlando were recalcu-
lated the agreement between the result obtained using the two
reference reactions improved. In this case the JPL value was taken
ꢁ1
tion over the following wavenumber ranges (cm ): CHCl
3
, 2390–
2
500; CH , 3000–3200; CH , 1380–1480. For each spectrum 32
4
2 2
F
scans were co-added to give an acceptable signal to noise ratio,
and the interferograms were Fourier transformed using boxcar
apodization. During each experiment the reaction mixture was
photolyzed 4–15 times, using up to eight lamps. Each photolysis
step was followed by a stabilization time, typically 2 min, before
the FTIR spectrum was recorded. The experiments were terminated
when around 30–70% of the reactant species were consumed.
Experiments were performed at temperatures of 303, 283, 298
and 268 K with Eq. (R2) as the reference, and at 313, 273 and
3 3
for the CH Br + Cl rate [12]. Gola et al. used CH Br + Cl as the ref-
erence reaction, but their relative rate data are not in agreement
with Orlando at 298 K, the only temperature where the two studies
overlap [14,16].
The room temperature relative rate studies, by Beichert et al.
[7], Brahan et al. [8] and Catoire et al. [9], all used Eq. (R3) as the
reference reaction, and the rate coefficients shown in Fig. 2 are
recalculated using the rate of the Eq. (R3) determined by Bryukov
et al. [15]. The rate constant determined using absolute methods at
2
53 K with Eq. (R3) as reference. Nitrogen was used as the bath
2
1.5
1
Table 1
Experimental details; reactant concentrations and photolysis time. The photolysis
time is normalized to one lamp.
T (K)
[CHCl
bar)
3
]
[Ref.]
bar)
Ref.
compound
t
photolysis
(s)
(l
(l
2
2
2
2
2
2
2
2
2
2
3
3
3
3
53
53
68
68
73
73
83
83
98
98
08
08
13
13
22
21
30
40
14
15
10
30
50
10
20
21
20
60
13
13
50
10
60
80
40
90
50
50
13
17
40
55
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
4
4
2
2
4
4
2
2
2
2
2
2
4
4
870
900
500
360
390
480
170
240
540
180
180
240
120
180
F
F
2
2
0.5
0
F
F
F
F
F
F
2
2
2
2
2
2
0
0.1
0.2
0.3
0.4
0.5
0.6
ln([CH F ] /[CH F ] )
2
2 0
2 2 t
Fig. 1. Loss of CHCl
reaction with Cl atoms, at 313 K. Error bars of spectral fits are within symbols.
3
versus loss of the reference compound CH
2
F
2
following

1
62
E.J.K. Nilsson et al. / Chemical Physics Letters 494 (2010) 160–162
Table 2
This work
Gola et al. (2009)
Bryukov et al. (2002)
Orlando (1999)
Talhaoui et al. (1996)
Bryukov et al. (2002)
This work
Rate constant ratios for the title reaction relative to the two reference compounds,
and the value of the rate coefficients based on an average of two determinations.
Errors for the relative rates are based on the error on the slope of the relative rate plot,
while in the error bar in the rate constants in last column also includes uncertainties
due to uncertainties in the reference reaction rate constants.
10
T (K)
k
k
CHCl3þCl
REF+Cl
/
Ref.
compound
k
(10
CHCl3 þCl
13
ꢁ
cm3 molecules
ꢁ1 ꢁ1
s )
253
268
273
283
298
308
313
1.39 ± 0.01
4.64 ± 0.07
1.23 ± 0.10
4.33 ± 0.06
3.65 ± 0.04
1.14 ± 0.10
3.36 ± 0.02
CH
4
CH
2
CH
4
CH
2
CH
2
CH
4
CH
2
0.688 ± 0.079
0.873 ± 0.131
0.877 ± 0.133
1.11 ± 0.17
1.25 ± 0.19
1.41 ± 0.21
1.49 ± 0.22
F
2
1
F
F
2
2
F
2
1
1.5
2
2.5
1
3
3.5
4
4.5
000/T, K-1
This work
3
Fig. 3. Arrhenius plot for k(CHCl + Cl) in the temperature range 200–1000 K. The
Gola et al. (2009)
Bryukov et al. (2002)
Orlando (1999)
results of five different temperature studies, including the present one, are plotted
in the figure. The full drawn line is the temperature dependent rate coefficient
determined from a fit to the present data set and that of Orlando [14] in the range
220–330 K. The dash-dotted line is the three parameter fit by Bryukov et al. [15],
valid at higher temperatures.
Talhaoui et al. (1996)
JPL
1
IUPAC
This work
Jeoung et al. (1991)
Beichert et al. (1995)
Beichert et al. (1995)
Brahan et al. (1996)
Catoire et al. (1996)
2 2
hydrocarbon reactions with chlorine. In addition the CH F + Cl
reaction has recently been reexamined using FTIR and GCMS anal-
ysis in separate reactor systems and the results are in excellent
agreement with each other and with an earlier measurement at
room temperature [21].
While the temperature dependent rate expression presented in
this work is valid in the temperature range 220–330 K, the expres-
sion by Bryukov et al. [15] should be used at higher temperatures.
3
3.5
4
4.5
000/T, K-1
5
5.5
1
Acknowledgements
3
Fig. 2. Arrhenius plot for k(CHCl + Cl) in the range 220–330 K. Results from the
present and previous studies plotted with error bars, excluding errors due to
uncertainties in reference reactions. The JPL [12] and IUPAC [13] recommendations
mentioned in the introduction are plotted as lines. Fulldrawn line is a fit to the data
of the present study and the results by Orlando [14] and is expressed by
The authors would like to acknowledge the Copenhagen Center
for Atmospheric Research, supported by the Villum Kahn Rasmus-
sen foundation and the Danish Natural Science Research
Foundation.
ꢁ
12
3
ꢁ1 ꢁ1
k = (3.77 ± 0.32) ꢀ 10
exp (–(1011 ± 23)/T) cm molecule
s . The results of
the relative rate study by Orlando [14] have been recalculated using the same rate
4
constant expression for the reference reaction, CH + Cl, as used in the present
study.
References
[
1] M.A.K. Khalil, R.A. Rasmussen, Atmos. Environ. 33 (1999) 1151.
2
98 K have lower values and show a large spread. The current
[2] A. McCulloch, Chemosphere 50 (2003) 1291.
[
[
3] D.R. Worton, W.T. Sturges, J. Schwander, R. Mulvaney, J.M. Barnola, J.
Chappellaz, Atmos. Chem. Phys. 6 (2006) 2847.
4] M.L. Aucott, A. McCulloch, T.E. Graedel, G. Kleiman, P. Midgley, Y.F. Li, J.
Geophys. Res. Atmos. 104 (1999) 8405.
recommendations by JPL [12] and IUPAC [13] are plotted as dashed
and dotted lines in Fig. 2. Since both the data set of the present
study, and that of Orlando [14] are slightly higher than the recom-
mendations a fit to the two data sets was done. The fit resulted in a
[5] E.J. Hoekstra, J.H. Duyzer, E.W.B. de Leer, U.A.T. Brinkman, Atmos. Environ. 35
(2001) 61.
[
[
ꢁ12
temperature dependent rate constant of k = (3.77 ± 0.32) ꢀ 10
6] Y.L. Yung, M.B. McElroy, S.C. Wofsy, Geophys. Res. Lett. 2 (1975).
7] P. Beichert et al., J. Phys. Chem. 99 (1995) 13156.
3
ꢁ1 ꢁ1
exp (ꢁ(1011 ± 23)/T) cm molecule
s , where uncertainties are
due to the fit. The expression is represented by the full drawn line
in Fig. 2.
[8] K.M. Brahan, A.D. Hewitt, G.D. Boone, S.A. Hewitt, Int. J. Chem. Kinet. 28 (1996)
97.
9] V. Catoire, R. Lesclaux, W.F. Schneider, T.J. Wallington, J. Phys. Chem. 100
1996) 14356.
[10] S.C. Jeoung, K.Y. Choo, S.W. Benson, J. Phys. Chem. 95 (1991) 7282.
3
[
Fig. 3 shows the result of the present study plotted together
with literature data over a wider temperature range. Again the full
drawn line represents a fit to the experimental data from the pres-
ent study and that of Orlando [14]. The most extensive data set is
that of Bryukov et al. [15], reaching from 297 to 854 K, represented
by filled triangles and a dashed line.
For temperatures where there is overlap between the present
study and that by Gola et al. [16] our results are about 20 % higher.
Gola et al. have not provided error bars for the rate constants, and
one source of error in their work is the rate constant for their ref-
(
[
11] A. Talhaoui, F. Louis, B. Meriaux, P. Devolder, J.P. Sawerysyn, J. Phys. Chem. 100
1996) 2107.
12] S.P. Sander et al., Chemical Kinetics and Photochemical Data for Use in
Atmospheric Studies. Evaluation Number 15, National Aeronautics and Space
Administration, Jet Propulsion Laboratory, California Institute of Technology,
Pasadena, California, 2006.
13] R. Atkinson et al., Atmos. Chem. Phys. 6 (2006) 3625.
[14] J.J. Orlando, Int. J. Chem. Kinet. 31 (1999) 515.
15] M.G. Bryukov, I.R. Slagle, V.D. Knyazev, J. Phys. Chem. A 106 (2002) 10532.
16] A.A. Gola, D. Sarzynski, A. Drys, J.T. Jodkowski, Chem. Phys. Lett. 469 (2009)
(
[
[
[
[
250.
erence reaction CH
3
Br + Cl. There has only been one study of the
[17] E.J.K. Nilsson, C. Eskebjerg, M.S. Johnson, Atmos. Environ. 43 (2009) 3029.
[18] D.W.T. Griffith, Appl. Spectrosc. 50 (1996) 59.
19] D. York, N.M. Evensen, M.L. Martinez, J.D. Delgado, Am. J. Phys. 72 (2004).
20] L.S. Rothman et al., J. Quant. Spectrosc. Radiat. Transf. 96 (2005) 139.
21] E.J.K. Nilsson, M.S. Johnson, O.J. Nielsen, E.W. Kaiser, T.J. Wallington, Int. J.
Chem. Kinet. 41 (2009) 401.
reference reaction over the full temperature interval, which makes
the reference reaction not better determined than the studied reac-
tion. A strength of the present work is that one of our reference
[
[
[
4
reactions is CH + Cl, which is one of the most extensively studied