Rate coefficients for the gas-phase reactions of chlorine atoms with cyclic ethers at 298 K

Article abstract of DOI:10.1002/kin.20765

Rate coefficients of reactions of Cl atoms with cyclic ethers, tetrahydropyran (THP), tetrahydrofuran (THF), and dihydrofurans (2,5-DHF and 2,3-DHF) have been measured at 298 K using a relative rate method. The relative rate ratios for THP and THF are 0.80 ± 0.05 and 0.80 ± 0.08, respectively, with n-hexane as the reference molecule. The relative rate ratios for THF and 2,5-DHF with n-pentane as the reference molecule are 0.95 ± 0.07 and 1.73 ± 0.06, respectively, and for 2,5-DHF with 1-butene as reference is 1.38 ± 0.05. The average values of the rate coefficients are (2.52 ± 0.36), (2.50 ± 0.39), and (4.48 ± 0.59) × 10^-10 cm³ molecule^-1 s^-1 for THP, THF, and 2,5-DHF, respectively. The errors quoted here for relative rate ratios are 2σ of the statistical variation in different sets of experiments. These errors, combined with the reported errors of the reference rate coefficients using the statistical error propagation equation, are the quoted errors for the rate coefficients. In the case of 2,3-DHF, after correcting for the dark reaction with CH₃COCl and assuming no interference from other radical reactions, a relative rate ratio of 0.85 ± 0.16 is obtained with respect to cycloheptene, corresponding to a rate coefficient of (4.52 ± 0.99) × 10^-10 cm³ molecule^-1 s^-1. Unlike cyclic hydrocarbons, there is no increase with increasing number of CH₂ groups in these cyclic ethers whereas there is an increase in the rate coefficient with unsaturation in the ring. An attempt is also made to correlate the rate coefficients of cyclic hydrocarbons and ethers with the molecular size as well as HOMO energy.

Full text of DOI:10.1002/kin.20765

Rate Coefficients for the
Gas-Phase Reactions of
Chlorine Atoms with Cyclic
Ethers at 298 K
H. D. ALWE, M. WALAWALKAR, A. SHARMA, K. K. PUSHPA, S. DHANYA, P. D. NAIK
Radiation & Photochemistry Division, Bhabha Atomic Research Centre, Trombay, Mumbai 400 085, India
Received 6 March 2012; revised 17 August 2012; 14 September 2012; accepted 19 October 2012
DOI 10.1002/kin.20765
Published online 26 March 2013 in Wiley Online Library (wileyonlinelibrary.com).
ABSTRACT: Rate coefficients of reactions of Cl atoms with cyclic ethers, tetrahydropyran (THP),
tetrahydrofuran (THF), and dihydrofurans (2,5-DHF and 2,3-DHF) have been measured at
298 K using a relative rate method. The relative rate ratios for THP and THF are 0.80 0.05 and
0.80 0.08, respectively, with n-hexane as the reference molecule. The relative rate ratios for
THF and 2,5-DHF with n-pentane as the reference molecule are 0.95 0.07 and 1.73 0.06,
respectively, and for 2,5-DHF with 1-butene as reference is 1.38 0.05. The average values of
the rate coefficients are (2.52 0.36), (2.50 0.39), and (4.48 0.59) × 10−10 cm3 molecule−1
s−1 for THP, THF, and 2,5-DHF, respectively. The errors quoted here for relative rate ratios are
2σ of the statistical variation in different sets of experiments. These errors, combined with
the reported errors of the reference rate coefficients using the statistical error propagation
equation, are the quoted errors for the rate coefficients. In the case of 2,3-DHF, after correcting
for the dark reaction with CH3COCl and assuming no interference from other radical reactions,
a relative rate ratio of 0.85 0.16 is obtained with respect to cycloheptene, corresponding to
a rate coefficient of (4.52 0.99) × 10−10 cm3 molecule−1
s
−1. Unlike cyclic hydrocarbons,
there is no increase with increasing number of CH2 groups in these cyclic ethers whereas there
is an increase in the rate coefficient with unsaturation in the ring. An attempt is also made to
correlate the rate coefficients of cyclic hydrocarbons and ethers with the molecular size as well
as HOMO energy. ꢀC 2013 Wiley Periodicals, Inc. Int J Chem Kinet 45: 295–305, 2013
also known to contribute significantly to the oxidative
capacity of the troposphere. Although the global es-
timate of the Cl atom concentration in the northern
hemisphere is only about 103 atoms cm−3, orders of
magnitude lower than the ambient OH concentration
(106 atoms cm−3) [1], it is significantly higher in con-
ditions of the marine boundary layer (MBL), coastal
regions, and also in arctic region, reaching a peak con-
centration of about 105 atoms cm−3, especially during
sunrise [2]. The generation of Cl atoms from inorganic
INTRODUCTION
Most of the volatile organic compounds (VOCs) that
are released into the troposphere react with hydroxyl
radical (OH), the major tropospheric oxidant. Chlo-
rine atoms (Cl), which react rapidly with VOCs, are
Correspondence to: S. Dhanya; e-mail: sdhanya@barc.gov.in.
2013 Wiley Periodicals, Inc.
C
ꢀ

296
ALWE ET AL.
chlorides via heterogeneous reactions is attributed to
be responsible for this high concentration in these ar-
eas. A recent field study suggests that Cl atoms may
be formed to a similar level, even in midcontinental re-
gions, thousands of kilometers away from the coastal
area, probably due to the presence of anthropogenic
pollutants [3]. Cl atoms react with most of the organic
molecules with rate coefficients that are higher, by an
order of magnitude or more, as compared to the OH
radical. Hence, despite the lower concentration, the
reaction of Cl atoms contributes significantly to the
tropospheric degradation of VOCs in the conditions
prevalent in the MBL and highly polluted industrial
and urban areas.
are known to react very fast with OH radicals, their
reactivity being higher than that of the corresponding
hydrocarbons. The recent relative rate measurements
of the rate coefficients of the reactions of Cl atoms
with THP [16,17] and THF [17] show the reactivity
of THP to be lower than that of THF in spite of the
increased number of CH2 groups, and no clear trend is
observed in the reactivity of Cl atoms with respect to
the corresponding hydrocarbons. In the present study,
the rate coefficients for the reaction of Cl atoms with
THF and THP are measured at 298 K along with that
of dihydrofurans, namely 2,5-dihydrofuran (2,5-DHF,
also known as 3-oxolene) and 2,3-dihydrofuran (2,3-
DHF or 2-oxolene), to see the effect of unsaturation.
Reference molecules used are different from those used
in the previous study.
Recently, considerable effort has been put into the
measurement of the rate coefficients of reactions of
Cl atoms with different classes of VOCs. The reac-
tion mechanism is expected to be similar to that of
the OH radical, as suggested by a linear correlation
of the rate coefficients for Cl atom reactions to those
of the OH radical reaction [4,5]. The reactivity of OH
with a large number of VOCs is very well correlated
by structure activity relations (SARs) [6–9], whereas
efforts to develop similar SARs for Cl reactivity are
limited [10,11]. The reactivity of both OH and Cl to-
ward unsaturated molecules is generally higher than
that of the saturated ones, due to the higher rate co-
efficients of addition reactions [6,11]. However, our
recent measurement of rate coefficients of reactions of
Cl atoms with unsaturated cyclic hydrocarbons at room
temperature [12,13] showed that their rate coefficients
increase with the number of carbon atoms (from 5 to
8), but do not change significantly with unsaturation.
Considering the importance of Cl atom reactions in the
troposphere, it is desirable to understand these trends
in their rate coefficients as compared to that of OH rad-
icals, which requires measurement of rate coefficients
of more molecules with different functional groups. In
the present study, rate coefficients of the reaction of the
Cl atom with cyclic ethers, including two unsaturated
molecules, are measured and compared with cyclic hy-
drocarbons, to understand the effect of the O atom and
unsaturation in the ring on the Cl reactivity. These mea-
surements also contribute in estimating the role of Cl
atom reactions toward their tropospheric degradation.
Cyclic ethers such as oxolane and oxane, commonly
known as tetrahydrofuran (THF) and tetrahydro-2H-
pyran (THP), are used widely in industry. A range of
cyclic ethers have been detected and analyzed during
the low-temperature oxidation of n-alkanes [14]. Sub-
stituted dihydrofurans are also reported to be important
intermediates in the formation of aerosols from the at-
mospheric oxidation of hydrocarbons, through their
reactions with OH radicals [15]. These cyclic ethers
EXPERIMENTAL
Rate coefficients at room temperature (298 K) were
determined using a relative rate method by comparing
the rate of decrease in the concentration of the cyclic
ethers (CE) due to their reaction with Cl atoms to that
of a reference molecule (R) with the known rate coef-
ficient:
CE + Cl → products
R + Cl → products
(1)
(2)
Assuming that the CE and the reference compounds
react only with Cl atoms, the fractional loss of CE and
R, is related by the standard expression
ꢀ
ꢁ
ꢀ
ꢁ
ꢀ
ꢁ
(CE)
(CE)t
(R)
(R)t
kCE
kR
t0
t0
ln
=
ln
(3)
where [CE]t0, [CE]t and [R]t0, [R]t are the concentra-
tions of the cyclic ethers and the reference compound at
time 0 and t, respectively, and kCE and kR are the rate co-
efficients for reactions 1 and 2, respectively. Thus, the
plots of logarithms of the ratios of fractional changes
in the concentrations of both CE and R at specific time
t give a straight line with zero intercept and the slope
of kCE/kR. The rate coefficient kCE is calculated using
the known value of kR.
All the experiments were performed at 298
2 K. The details of the experimental setup are
given elsewhere [12,13]. The reaction mixture, con-
sisting of CE (200–250 ppm), reference molecule
(200–300 ppm), the source molecule for Cl gener-
ation (Cl2/CCl3COCl), and buffer gas (N2/air), was
International Journal of Chemical Kinetics DOI 10.1002/kin.20765

RATE COEFFICIENTS FOR THE GAS-PHASE REACTIONS OF CHLORINE ATOMS
297
prepared in a Pyrex/quartz reactor cell of volume 3 L,
having a sealed port for taking out samples for con-
centration measurement. The total pressure was main-
tained at 800 3 Torr. Chlorine atoms were generated
by the in situ photolysis of Cl2 at 350 nm in the pyrex
reactor, in the case of THF and THP. Since 2,5-DHF
showed the dark reaction upon addition of molecular
chlorine, Cl atoms were generated by photolysis of
trichloroacetyl chloride (CCl3COCl) at 254 nm in a
quartz reactor. Monochromatic UV lamps, either from
Sankyo Denki or Rayonet, were used for photolysis
at 254 and 350 nm. The mixture was photolyzed for
a period of 4–5 min, in steps of 40–50 s, and deple-
tion in the concentrations of the CEs, and the reference
molecules was followed after each step. This is done
by measuring the concentrations of CE and reference,
before and after photolysis, using a gas chromatograph
(Shimadzu GC-2014 with a flame ionization detector),
by introducing sample aliquots of 500 μL manually
for analysis, using a gas tight syringe (SGE-500F-
GT). Either SE-30 (10%, 2 m × 3 mm) or Porapak-Q
(1 m × 6 mm) stainless steel columns were employed
for the separation with suitable temperature conditions.
Before and after each photolysis, the prepared reac-
tion mixture was allowed to reach equilibrium, which
was confirmed by reproducibility of the concentra-
tion measured during consecutive gas chromatographic
samples. Hexane, pentane, butane, and 1-butene were
the different reference molecules used, except for 2,3-
DHF, based on their comparable rate coefficients and
the least interference during the gas chromatography
(GC) analysis, either from the molecule or from the
products of their reactions with Cl atoms.
Since 2,3-DHF was found to react rapidly with
CCl3COCl even at low concentrations of about 50 ppm,
a different procedure was adopted. The rate of reaction
of 2,3-DHF with CH3COCl was found to be lower than
that with CCl3COCl and hence photolysis of CH3COCl
at 254 nm was used as the source molecule for the Cl
atom. The utility of this molecule as the source of the
Cl atom was tested with butane, 1-butene, and cyclo-
heptene with n-hexane as the reference molecule, and
the measured relative ratios and rate coefficients were
found to match with the reported values [12,18,19].
1-Butene and n-hexane could not be used as refer-
ence molecules for determining the rate coefficient of
2,3-DHF, since their depletion rate due to the reac-
tion with the Cl atom was found to be slower than the
rate of depletion of 2,3-DHF due to the dark reaction
with CH3COCl. Hence cycloheptene was used as a
reference molecule. The reaction mixture consisted of
CH3COCl (100–120 ppm), 2,3-DHF (∼100 ppm), cy-
cloheptene (∼80 ppm), and N2 (buffer gas, keeping the
total pressure as 800 Torr). The mixture was subjected
to a single step of photolysis, with duration of pho-
tolysis varying in each experiment. Concentrations of
DHF and cycloheptene were monitored by GC before
and after photolysis. The total duration of experiment
was kept the same. The blank experiments were carried
out for the same duration to obtain the loss of 2,3-DHF
due to the dark reaction.
Some of the major products were characterized by a
gas chromatograph coupled with a mass spectrometer
(GCMS QP2010, Shimadzu), using different columns,
namely HT-8, BP-X-50, AB1-MS (30 m × 0.25 mm
× 0.25 μm), for the separation.
Nitrogen gas of purity >99.9% (INOX Air Prod-
ucts, Mumbai, India) and ultrapure air (zero grade;
Chemtron Science, Mumbai, India) were used as buffer
gases. THF (spectrograde) was obtained from Thomas
Baker (Mumbai, India) with >99.9% purity and 2,3-
DHF from Sigma Aldrich (Steinheim, Germany) with
99% purity. All other reagents were obtained from
Sigma Aldrich with purities >97%. Samples of the
compounds were stored in evacuated glass vessels and
subjected to freeze–pump–thaw cycles prior to use.
RESULTS
Prior to the determination of rate coefficients, the sta-
bility of the reaction mixture with respect to wall losses
and dark reactions was checked for about 7 h, which is
more than the total duration of a relative rate measure-
ment. The loss of the CEs through direct photolysis,
in the absence of chlorine precursors, was also inves-
tigated, and no significant decrease was observed. In
addition, reactions of the Cl atom with each CEs as
well as reference molecules were carried out individ-
ually to ensure that the products of the reactions did
not interfere with the GC measurement of the concen-
tration of the CEs and the reference molecules. Based
on this, n-pentane, n-hexane, and 1-butene, which have
rate coefficients comparable to that of CEs, were cho-
sen as the reference molecules. The rate coefficients
for the reaction of n-hexane and n-pentane with Cl
atoms were taken to be (3.15
0.40) × 10−10 and
(2.62 0.34) × 10−10 cm3 molecule−1
s
−1, respec-
tively. These values were obtained after modifying the
values reported by Atkinson and Aschmann [18], us-
ing the best current IUPAC estimate, (2.05
0.25)
× 10−10 cm3 molecule−1 s−1 [19], as the rate co-
efficient of their reference reaction, namely Cl + n-
butane. There is no IUPAC recommended value in
the case of 1-butene, where the reported values are
in the range of (2.89–3.38) × 10−10 cm3 molecule−1
s−1 [11,20–22]. Here, (3.21
0.41) × 10−10 cm3
molecule−1 s−1 is used as the rate coefficient for
International Journal of Chemical Kinetics DOI 10.1002/kin.20765

298
ALWE ET AL.
t
t
Figure 1 A typical plot of the relative decrease in the concentration of THP and n-hexane due to the reaction with the Cl atom
d
at 298 2 K (ꢀ) in N2 and ( ) in air. (The Y axis for o is marginally shifted, shown at the right). The source of the Cl atom is
Cl2 gas.
the reaction of the Cl atom with 1-butene, calculated
on the basis of the experimental relative rate ratio
from the linear regression analysis of the data in each
experiment, arising from the scattering of the points
due to random errors in sampling and concentration
measurements, is found to be small (1–2%), in most of
the cases, as compared to the variation in independent
measurements. In a few cases, where it is more than
5%, both the errors are added quadratically. The error
arising from continuous sampling from the reaction
mixture is also very small, <1%, and hence not added
to the total error. Since absolute concentrations are not
measured and care was taken to passivate the syringe
and use the same syringe throughout an experiment,
systematic errors in these measurements are consid-
ered to be negligible. These errors in the relative rate
ratios are combined with the errors reported for the re-
spective reference rate coefficients, using the statistical
error propagation equation and are quoted for kCE in
Table I. Except in the case of 2,5-DHF, with 1-butene
as reference, the relative rate ratios are same in nitro-
gen and air. However, the average rate coefficients for
each molecule, shown in bold in Table I, is the average
from all data sets in N2. (The experiments in air were
not considered.) The logarithms of fractional depletion
of 2,3-DHF and cycloheptene, after correcting for the
loss of 2,3-DHF due to the dark reaction, are plotted
in Fig. 4. The treatment of the data is slightly different
due to the presence of the dark reaction. The data plot-
ted in Fig. 4 are obtained from 13 independent runs,
each with single photolysis of varying duration, and
with respect to n-hexane [20], 1.02
0.02, and the
above rate coefficient for n-hexane. Trichloroacetyl
chloride (CCl3COCl) was used as the Cl atom source
in the case of 2,5-DHF instead of molecular chlorine,
due to the latter’s dark reaction with 2,5-DHF. To en-
sure that the change in a photolytic precursor does
not affect the relative rate ratios, due to secondary re-
actions of the counterfragment, relative rate measure-
ments with CCl3COCl as a photolytic precursor were
also carried out for THF and the ratios were found to
be the same as those with Cl2 as a source of the Cl
atom.
The typical plots of the logarithms of the fractional
loss of CE versus that of the reference molecules, ac-
cording to Eq. (1), are given in Figs. 1, 2, and 3 for
THP, THF and 2,5-DHF, respectively. The plots are
linear, with near zero intercept, in all experiments,
suggesting the absence of any complications due to
secondary reactions. The slopes of the linear least-
ꢂ
square fits, kCE kR, and the calculated rate coefficients
for the three molecules (kCE) with different reference
molecules, in either N2 or air, are listed in Table I.
The ratios given are average slopes from a number of
measurements (indicated in parentheses), and the er-
ror is two times the standard deviation, 2σ, calculated
from the distribution in the slope values of independent
runs, using the standard formula. The error in the slope,
International Journal of Chemical Kinetics DOI 10.1002/kin.20765

RATE COEFFICIENTS FOR THE GAS-PHASE REACTIONS OF CHLORINE ATOMS
299
t
t
Figure 2 Plot of the relative decrease in the concentration of THF due to the reaction with the Cl atom at 298 2 K; with
n-pentane as the reference molecule in N2 (ꢀ) and in air (ꢁ), The source of the Cl atom is Cl2 gas. The linear fit shown is for
d
d
(ꢀ). ( ): with n-hexane as the reference molecule and CCl3COCl as the Cl atom source in N2. The Y axis for is marginally
shifted, shown at the right.
all are corrected for the dark reaction, as mentioned in
the Experimental section. The ratio of the rate coef-
ficients and its error, given in Table I, corresponds to
the slope and 2σ from the linear regression analysis of
these points, as shown in Fig. 4. The rate coefficient is
estimated assuming no interference from other radical
reactions. The rate coefficient of the reaction of the Cl
atom with cycloheptene is (5.32 0.57) × 10−10 cm3
DISCUSSION
The average values of the rate coefficients for the Cl
atom reactions from the present work are listed in
Table II, along with the previously reported average
values for THP and THF. There is a good agree-
ment, within experimental errors, among the values of
THP and THF, though the present value is marginally
higher in the case of THP [16,17]. The value re-
ported for THP by Ballesteros et al. [16] is an aver-
age of the rate coefficients measured relative to ethene,
propene, 1-butene, and 1,3-butadiene. Considering the
IUPAC-recommended value [19] for 1-propene (2.7 ×
molecule−1 −1 [12].
s
Although the absence of heterogeneous chemistry
cannot be completely ruled out in the present experi-
mental conditions, the major products observed here
match very well with those reported for THP in a
200 L Teflon bag [16], namely tetrahydro-4H-pyran-
4-ol, tetrahydro-4H-pyran-4-one, tetrahydro-2H-
pyran-2-one, 2-methyl tetrahydrofuran, tetrahydro-
2-furanone, and ring-opened product 3-chloropropyl
formate. Similar products, formed by H atom ab-
straction, are also observed in the case of Cl-initiated
oxidation of THF. In the case of 2,5-DHF, Cl
atom addition products, 3,4-dichlorotetrahydrofuran,
2,3-dichlorotetrahydrofuran, 4-chloro-3,6-dihydro-2-
H-pyran, and 2,3-dichloro1-propanol were observed
along with the products of H abstraction reactions,
furan-2(5H)-one, 3-hydroxydihydro-2-(3H)-furanone,
and furan.
10−10 cm3 molecule−1 −1) and the corrected value
s
for 1-butene used in the present study (3.21 × 10−10
cm3 molecule−1 −1), the rate coefficients reported by
s
Ballesteros et al. [16] are modified to 2.57 × 10−10 and
2.38 × 10−10 cm3 molecule−1 s−1, respectively, closer
to the present determination (Table II). The value re-
ported by Giri and Roscoe for THP [17] is an average
of the rate coefficients measured in the temperature
range of 280–357 K with reference to THF and in-
cludes the error that they reported for THF. However,
the rate coefficient that is reported for THF, with ref-
erence to isobutene, does not include the uncertainty
in the value of the reference rate coefficient. Although
International Journal of Chemical Kinetics DOI 10.1002/kin.20765

300
ALWE ET AL.
t
t
Figure 3 Plot of the relative decrease in the concentration of 2,5-DHF due to the reaction with the Cl atom at 298 2 K; with
t
n-pentane as the reference molecule in N2 (ꢀ) and in air (ꢁ); with 1-butene as the reference molecule in N2 ( ). The source of
t
the Cl atom is CCl3COCl. The Y axis for ꢀ and is marginally shifted, shown at the right.
their rate coefficients for THP and THF are the same
within the combined error estimates, they observed the
rate coefficient of THF to be consistently about 30%
larger than that of THP, in spite of a lower number of
hydrogen atoms. However, the present results indicate
very similar rate coefficients for the reactions of THF
and THP with Cl atoms. The rate coefficients for the
reactions of DHFs with the Cl atom are almost 1.7–1.8
times higher than that of THF, indicating an increase
due to the presence of unsaturation.
The products of the reactions of Cl atoms observed
for THF and THP in our study are very similar to
those reported by Ballesteros et al. [16] for THP.
Even though the theoretical calculations of Ballesteros
et al. [16] predict the predominance of abstraction of
the α H atom, they observed products arising due to
H atom abstraction from all the possible positions in
the ring. In the case of unsaturated molecule, 2,5-DHF,
Cl-added products as well as those resulting from H ab-
straction are formed. These observations indicate that
because Cl atoms are so highly reactive, there is less
selectivity in their reactions.
rate coefficients of the reaction of Cl increases with
increasing number of CH2 groups [12]. It was also
observed that increasing unsaturation, i.e. the num-
ber of double bonds, does not influence the reactivity
significantly in six-membered ring compounds, cyclo-
hexane, cyclohexene, and cyclohexadiene [13]. For es-
timating the rate coefficients of Cl atom reactions with
alkanes and alkenes, SARs have been developed ear-
lier [7,10,11,24]. In general, these SARs consider the
total rate coefficients as the summation of group rate
coefficients, modified by substituent factors to take care
of the neighboring group effects. Ezell et al. [11] have
shown that these SARs are to be fairly in agreement
with experimental values in the case of alkenes up to
C5. However, these have not been applied in the case
of cyclic alkanes and alkenes. The calculated values of
rate coefficients in saturated molecules, cyclopentane,
cyclohexane, (2.90 and 3.51 × 10−10 cm3 molecule−1
s
−1), using these group additivity values [7,11],
agree with the experimental values, 3.05 and 3.24 ×
10−10 cm3 molecule−1 −1. However, there is con-
s
siderable discrepancy between the calculated and
experimental values in the case of cyclic alkenes
such as cyclopentene (4.30 and 3.05 × 10−10 cm3
Although the rate coefficients determined here are
sums of rate coefficients of individual reaction chan-
nels, an attempt is made to compare the total reactivity
of these cyclic ethers with that of cyclic hydrocarbons.
The earlier study on cyclic molecules showed that the
molecule−1
s
−1), cyclohexene (4.88 and 3.97 × 10−10
cm3 molecule−1 −1), and cyclohexadiene (6.70 and
s
4.06 × 10−10 cm3 molecule−1
s
−1). The trend shows
International Journal of Chemical Kinetics DOI 10.1002/kin.20765

RATE COEFFICIENTS FOR THE GAS-PHASE REACTIONS OF CHLORINE ATOMS
301
Table I Summary of the Measured Relative Rate Ratios and the Derived Rate Coefficients for the Reactions of the Cl
Atom with Cyclic Ethers, at Different Experimental Conditions. T = 298 K, P = 800 Torr
Cyclic Ether
Source of Cl
Reference
(Concentration ×
(Concentration × 10−15 (Concentration × 10−15
k298 × 1010(cm3
ꢂ
10−15 molecules cm−3
molecules cm−3
)
molecules cm−3
)
Buffer Gas
kCE kR
molecule−1 s−1
)
THP
6.0–7.2
6.0–6.5
Cl2 (95–100)
Cl2 (95–100)
n-Hexane (6.5–7.5)
n-Hexane (6.2–6.7)
Average in N2
N2 (3)
Air (2)
0.80 0.05
0.82 0.03
2.52 0.36
2.58 0.34
2.52 0.36
THF
4.6–6.0
5.7–6.5
6.0–6.5
5.2 –5.4
Cl2 (30–40)
Cl2 (30–40)
Cl2 (45–52)
CCl3COCl(20–22)
n-Pentane (6.2–8.8)
N2 (7)
Air (2)
N2 (3)
N2 (2)
0.95 0.07
0.97 0.10
0.79 0.07
0.81 0.11
2.49 0.37
2.54 0.42
2.49 0.39
2.55 0.47
2.50 0.39
n-Pentane (6.2–8.8)
n-Hexane (6.2–8.8)
n-Hexane (6.2–8.8)
Average in N2
2,5-DHF
2.9–3.1
2.9–3.1
2.9–3.1
3.1–3.4
CCl3COCl(13–16)
CCl3COCl(13–16)
CCl3COCl(13–16)
CCl3COCl(13–16)
n-Pentane (6.2–8.8)
n-Pentane (6.2–8.8)
1-Butene (3.9–4.9)
1-Butene (4.7–5.2)
N2 (3)
Air (4)
N2 (3)
Air (4)
1.73 0.06
1.70 0.07
1.38 0.05
1.63 0.16
4.53 0.59
4.45 0.61
4.43 0.59
5.23 0.83
4.48 0.59
Average in N2
2,3-DHF
2.9–3.4
CH3COCl(3.0–4.7)
cycloheptene(1.6–2.3)
N2(13)
0.85 0.16
4.52 0.99
The rate coefficients of reactions of reference molecules with the Cl atom at 298 K: k(Cl + n-hexane) = (3.15
0.40) × 10−10, and
k(Cl + n-pentane)= (2.62 0.34) × 10−10 cm3 molecule−1 s−1 (both from [18], corrected using the IUPAC recommended value for k(Cl +
n-butane) [19]). k(Cl + 1-butene) = (3.21 0.41) × 10−10 [20] and k(Cl + cycloheptene) = (5.32 0.57) × 10−10 cm3 molecule−1 s−1 [12],
both corrected, using the above value for k(Cl + n-hexane).
that the rate coefficient of Cl addition to the double
bond is less in the case of cyclic alkenes as compared
to straight chain alkenes, used to derive these values.
The rate coefficient of addition, estimated based on
the benzene yield in the case of cyclohexadiene (1.4
coefficients is seen in the reactions of THP and THF
with OH (25% larger for THF) and Br (70% larger for
THF). Giri and Roscoe [17] have noted that in the case
of the reactions of the Br atom, the higher rate coeffi-
cient of THF as compared to that of THP is attributed
to the difference in the activation energies [17]. Being
independent of temperature, the same reason is not ap-
plicable to rate coefficients of Cl atom reactions with
THP and THF. It is possible that only the CH2 group
α to the O atom contributes to the reactivity and the
additional CH2 groups do not have any effect. How-
ever, the product analysis shows that abstraction also
takes place from CH2 groups other than those at α po-
sition. Thus, it is difficult to give a clear explanation for
the similarity of the rate coefficients of THP and THF.
The higher rate coefficients of DHFs as compared to
THF imply that the addition rate coefficient is higher
than two times the rate coefficient for abstraction. A
slightly enhanced rate coefficient for the abstraction
from the allylic CH2 group may also contribute to this
increase.
× 10−10 cm3 molecule−1
s
−1), is only about half of
the recommended rate coefficient of addition based on
acyclic alkenes (2.98 × 10−10 cm3 molecule−1 s−1
)
[13].
The neighboring group substituent effect for the
ether functional group is estimated to be about 8 for
OH reactions [6,7], leading to enhancement of rate co-
efficient of abstraction. This is assigned to be due to the
formation of a hydrogen-bonded prereactive complex,
as suggested by a negative temperature dependence of
the rate coefficients [25]. Although Cl atom reactions
with THP [17] and 1,4-dioxane [26] are also reported
to have very small or no dependence on temperature,
there is no increase in the rate coefficient of THP as
compared to that of cyclopentane. It is also observed
that the measured rate coefficient of THP is the same
(present work) or marginally lower [17] as compared
to that of THF. This contradicts with the SAR by Ezell
et al. [11], which predicts an increase in the rate coeffi-
cient with more number of CH2 groups. The reason for
this difference is not clear, but a similar trend in rate
The lack of activation barriers for the reactions, as
suggested by the very small or negative dependence on
temperature for the rate coefficients of the reaction of
the Cl atom with alkanes and alkenes [7,21,23] as well
as cyclic ethers [17,26], implies that these reaction rate
International Journal of Chemical Kinetics DOI 10.1002/kin.20765

302
ALWE ET AL.
t
t
Figure 4 Plot of the relative decrease in the concentration of 2,3-DHF due to the reaction with the Cl atom at 298 2 K; with
cycloheptene as the reference molecule in N2, after correction for the loss due to the dark reaction with CH3COCl.
coefficients are almost in the collision-controlled lim-
its. The measured rate coefficient of reaction of Cl with
cyclooctane also suggests the reactions to be approach-
ing the collision-controlled limits for cyclohexane and
cyclooctane, since the increase with the number of
CH2 groups is not as high as in the case of smaller
cyclic alkanes [27]. The collision-controlled rate co-
efficient is proportional to the collision cross section,
a function of molecular size and velocity, a function
of mass and temperature. Hence, it is possible that the
presence of unsaturation or other functional features
in the molecule may not affect the rate coefficient sig-
nificantly. In cyclic hydrocarbons, it was observed that
unsaturation does not change the rate coefficients [13].
However, in the present study, the rate coefficient of
the reaction of the Cl atom with DHFs is almost 1.8
times higher than that of THF, the corresponding sat-
urated ether, though the molecular size remains the
same.
The above discussion shows that different factors,
such as size of the molecule and the presence of func-
tional groups of different reactivities, may influence the
overall rate coefficients in the case of reactions of Cl
atoms that have high rate coefficients. This dependence
Table II Summary of the Rate Coefficients of Reactions of Cl with Cyclic Ethers from the Present Work and from the
Literature
k298 × 1010
Cyclic Ether
THP
(cm3 molecule−1 s−1
)
Reference
Present work
Comments Reference Molecule/Error in k298
2.52 0.36
2.03 0.82
n-hexane; includes error in kR.
Roscoe and Giri [17]
THF, from [17], includes error in kR; k298 is average of
the k in the temperature range 280–360 K.
Ethene, propene, 1-butene, 1,3-butadiene; includes the
errors in kR.
2.21 0.32
[16]
THF
2.50 0.39
2.71 0.34
Present work
Roscoe and Giri [17]
n-Hexane, n-pentane; includes the error in kR.
k298 is average of the k in the temperature range
280–360 K. Not inclusive of the error in kR.
n-Pentane, 1-butene; includes the error in kR.
Cycloheptene, includes the error in kR.
2,5-DHF
2,3-DHF
4.48 0.59
4.52 0.99
Present work
Present work
International Journal of Chemical Kinetics DOI 10.1002/kin.20765

RATE COEFFICIENTS FOR THE GAS-PHASE REACTIONS OF CHLORINE ATOMS
303
Size
t
Figure 5 Variation of rate coefficients of reactions of cyclic hydrocarbons/cyclic ethers with OH ( ) and Cl (ꢁ): (A) with size
parameter and (B) with HOMO energy. The rate coefficients of reactions of Cl with cyclic hydrocarbons are marked by ꢀ, to
differentiate from those of cyclic ethers. The straight line fitting in (A) is for only cyclic hydrocarbons, whereas in (B) it is for
all the molecules. Cyclic hydrocarbons considered are C5H10, C6H12 [30,31], C8H16 [27,30], C5H8, C6H10, C7H12 [12,30],
C6H8, and C8H12 [13,30].
is further investigated and compared to that of OH rad-
ical reactivity. Fast reactions of radicals with organic
molecules can be treated on the basis of the frontier
orbital interactions. In the case of electrophiles such as
OH and Cl, the important interaction is with the highest
occupied molecular orbital (HOMO) of the molecule.
The presence of more reactive functional groups raises
the HOMO energy, and the values of HOMO energy
are found to correlate very well with the rate coeffi-
cients of OH radical reactions with organic molecules
[8] and also with the rate coefficients of Cl atoms
with fluorocompounds such as fluoroalcohols [28].
Figure 5 shows the variation of rate coefficients with
both size parameter and HOMO energy. The molec-
ular size and HOMO energies are calculated using
Gaussian 92 program [29]. The geometry optimiza-
tion and energy calculations were carried out at the
MP2/6-31G+ level. Size parameters were calculated
as square of the sum of largest interatomic distance
divided by μ1/2 (reduced mass) to take care of the vari-
ation in the molecular velocity. The rate coefficients of
OH radical reactions are from Moriarty et al. [25] for
cyclic ethers and from Atkinson and Arey for cyclic
hydrocarbons [30].
From Fig. 5A, it can be seen that the correlation
of logarithmic value of the rate coefficients of all the
cyclic hydrocarbons and ethers together with size pa-
rameter is quite poor for OH reactions, which have
rate coefficients much lower than collision-controlled
limits. In the case of rate coefficients of Cl atom re-
actions, in cyclic hydrocarbons alone, there is a good
correlation with size parameter (a straight line fit is
shown in the figure), if the value for cyclooctane, mea-
sured at low-pressure conditions (1 Torr) [27], is not
considered. This correlation may be a consequence of
the following facts: The increase in size corresponds
to the increase in the number of CH2 groups and
hence the number of abstractable H atoms, leading
to an increase in the rate coefficient. When there is
unsaturation within the ring of same size or the num-
ber of carbon atoms, abstraction possibility from two
CH2 groups is replaced by the possibility of an ad-
dition reaction. If the rate coefficient of addition is
equal to almost two times that of the abstraction rate
coefficient per the CH2 group, as observed in cyclo-
hexadiene [13], the total rate coefficient may remain
same as the saturated molecule. However, when cyclic
ethers and hydrocarbons are considered together, the
International Journal of Chemical Kinetics DOI 10.1002/kin.20765

304
ALWE ET AL.
correlation is poor, showing that even though the
rate coefficients are very high and reach collision-
controlled limits, they are still dependent on the pres-
ence of different functional groups that can influence
the reactivity. It is also possible that the poor depen-
dence on size only reflects the deviation from the clas-
sical collision theory, as the molecules and radicals are
not really hard spheres and are subjected to intermolec-
ular interaction potential.
Figure 5B clearly shows that the rate coefficients
of the reactions of these molecules with OH (in the
logarithmic scale) have a good linear dependence on
HOMO energy. This includes even the cyclic ethers, in
which case the increased reactivity with OH is assigned
to the prereactive complex formation, involving hydro-
gen bonding between the incoming OH group and the
O atom in the ether. It can be seen from Fig. 5B that the
rate coefficients of Cl atoms also correlate well with
the HOMO energy. The correlation coefficient for the
linear regression of the variation of rate coefficients
of Cl with HOMO is not as good as that in the case
of rate coefficients of OH reactions, even though the
variations in the former are minor, all the values be-
ing in the 10−10 range, as compared to that in latter,
distributed in a wide range, from 10−12 to 10−10. How-
ever, the rate coefficients of CEs and hydrocarbons,
considered together, are correlated better when plotted
against HOMO energy, as compared to the plot against
size parameter.
CONCLUSION
The rate coefficients of reactions of Cl atom with
cyclic ethers, THP, THF, and DHF, have been deter-
mined at 298
2 K, using the relative rate method
and compared to those of cyclic hydrocarbons, to see
the effect of the O atom and unsaturation on the ring.
The rate coefficients determined for THF and THP are
in agreement with the recent literature values, deter-
mined using different reference molecules. The results
confirm that there is no increase in the rate coeffi-
cient of THP as compared to that of THF, whereas
the rate coefficients increase with an increase in CH2
groups in cyclic hydrocarbons. Unlike cyclic hydro-
carbons, there is also an increase in the rate coefficient
of the Cl atom reaction with unsaturation in cyclic
ether, in DHF. A good correlation with molecular size
parameter is observed for the rate coefficients of Cl
atom reactions with cyclic hydrocarbons alone, but not
when cyclic ethers and hydrocarbons are considered
together. The logarithmic values of rate coefficients
of the Cl atom reaction of CEs and hydrocarbons are
found to be better correlated with HOMO energy of the
molecule.
BIBLIOGRAPHY
1. Rudolph, J.; Koppmann, R. Plass-Dulmer, Ch. Atmos
Environ 1996, 30, 1887–1894.
The tropospheric lifetimes of THP and THF, with
respect to their reactions with Cl and OH, have already
been compared by Giri and Roscoe [17]. They have
estimated the tropospheric lifetimes for the removal of
THF and THP by Cl to be in the range of 10–16 h in
the conditions of marine environment, which are com-
parable to the lifetimes for the removal by OH, 14 and
21 h for THF and THP, respectively. Lifetimes for the
removal of THF and THP by Cl, estimated based on the
present rate coefficients, are 11 h for both, almost same
as that estimated earlier, indicating that the reactions of
Cl and OH have similar importance for the removal of
these molecules in the MBL. The concentration of Cl
and OH, used for the above estimation, are the same as
those considered by Giri and Roscoe [17], i.e., 105 and
106 molecules cm−3, respectively. In the case of unsat-
urated molecules, DHFs, the rate coefficient with Cl
is only increased about 1.8 times as compared to that
of THF, whereas the reactivity with OH is expected
to be an order of magnitude higher than THF and an
additional pathway of the reaction with O3 is also pos-
sible, making them the most prominent pathway for
the degradation of DHF, even under the conditions of
the MBL.
2. Spicer, C. W.; Chapman, E. G.; Finlayson-Pitts, B. J.;
Plastridge, R. A.; Hubbe, J. M.; Fast, J. D.; Berkowitz,
C. M. Nature 1998, 394, 353–355.
3. Thornton, J. A.; Kercher, J. P.; Riedel, T. P.; Wagner, N.
L.; Cozic, J.; Holloway, J. S.; Dube, W. P.; Wolfe, G.
M.; Quinn, P. K.; Middlebrook, A. M.; Alexander, B.;
Brown, S. S. Nature 2010, 464, 271–274.
4. Andersen, M. P. S.; Nielsen, O. J.; Wallington, T. J.;
Hurley, M. D.; DeMore, W. B. J Phys Chem A 2005,
109, 3926–3934.
5. Garib, A.; Timerghazin, Q. K.; Ariya, P. A. Can J Chem
2006, 84, 1686–1695.
6. Kwok, E. S. C.; Atkinson, R. Atmos Environ 1995, 29,
1685–1695.
7. Atkinson, R. J Phys Chem Ref Data 1997, 26, 215–290.
8. King, M. D.; Canosa-Mas, C. E.; Wayne, R. P. Phys
Chem Chem Phys 1999, 1, 2231–2238.
9. McGillen, M. C.; Crosier, J.; Percival, C.J.; Sanchez-
Reyna, G.; Shallcross, D. E. Chemosphere 2006, 65,
2035–2044.
10. Tyndall, G. S.; Orlando, J. J.; Wallington, T. J.; Dill, M.;
Kaiser, E. W. Int J Chem Kinet 1997, 29, 43–55.
11. Ezell, M. J.; Wang, W.; Ezell, A. A.; Soskin, G.;
Finlayson-Pitts, B. J. Phys Chem Chem Phys 2002, 4,
5813–5820.
International Journal of Chemical Kinetics DOI 10.1002/kin.20765

RATE COEFFICIENTS FOR THE GAS-PHASE REACTIONS OF CHLORINE ATOMS
305
12. Sharma, A.; Pushpa, K. K.; Dhanya, S.; Naik, P. D.;
Bajaj, P. N. Int J Chem Kinet 2010, 42, 98–105.
13. Sharma, A.; Pushpa, K. K.; Dhanya, S.; Naik, P. D.;
Bajaj, P. N. Int J Chem Kinet 2011, 43, 431–440.
14. Herbinet, O.; Bax, S.; Glaude, P-A.;. Carre´, V.; Battin-
Leclerc, F. Fuel 2011, 90, 528–535.
22. Ceacero-Vega, A. A.; Ballesteros, B.; Albaladejo, J.;
Bejan, I.; Barnes, I. Chem Phys Lett 2009, 484,
10–13.
23. Ohta, T. Int J Chem Kinet 1984, 16, 1495–1503.
24. Senkan, S. M.; Quam, D. J Phys Chem 1992, 96, 10837–
10842.
15. Liu, S.; Day, D. A.; Shields, J. E.; Russell, L. M. Atmos
Chem Phys Discuss 2011, 11, 7189–7233.
16. Ballesteros, B.; Ceacero-Vega, A. A.; Garzo´n, B.A.;
Jime´nez, E.; Albaladejo, J. J. Photochem Photobiol A:
Chem 2009, 208, 186–194.
17. Giri, B. R.; Roscoe, J. M. J Phys Chem A 2010, 114,
8369–8375.
18. Atkinson, R.; Aschmann, S. M. Int J Chem Kinet 1985,
17, 33–41.
19. Evaluated kinetic data from IUPAC Sub Committee
for Gas Kinetic Data Evaluation. Available at www
.iupac-kinetic.ch.cam.ac.uk/. Accessed 17 September
2010.
20. Coquet, S.; Ariya, P. A. Int J Chem Kinet. 2000, 32,
478–484.
25. Moriarty, J.; Sidebottom, H.; Wenger, J.; Mellouki,
A.; Le Bras, G. J Phys Chem A 2003, 107, 1499–
1505.
26. Giri, B. R.; Roscoe, J. M.; Gonzalez-Garcia, N.; Olz-
mann, M.; Lo, J. M. H.; Marriott, R. A. J Phys Chem
2011, 115, 5105–5111.
27. Aranda, A.; D´ıaz-de-Mera, Y.; Bravo, I.; Morales, L.
Environ Sci Pollut Res 2006, 14, 176–181.
28. Papadimitriou, V. C.; Papanastasiou, D. K.; Ste-
fanopoulose, V. G.; Zaras, A. M.; Lazarou, Y. G.; Pa-
pagiannakopoulos, P. J Phys Chem 2007, 111, 11608–
11617.
29. Frisch, M. J., et al. Gaussian 92, Revision E 1; Gaussian,
Inc.: Pittisburgh, PA, 1992.
30. Atkinson, R.; Arey, J. Chem Rev 2003, 103, 4605–4638.
31. Anderson, R. S.; Huang, L.; Iannone, R.; Rudolph J. J
Phys Chem A 2007, 111, 495–504.
21. Orlando, J. J.; Tyndall, G. S.; Apel, E. C.; Riemer, D.
D.; Paulson, S. E. Int J Chem Kinet 2003, 35, 334–353
International Journal of Chemical Kinetics DOI 10.1002/kin.20765