A kinetics and mechanistic study of the atmospherically relevant reaction between molecular chlorine and dimethyl sulfide (DMS)

Article abstract of DOI:10.1039/b415566a

A gas-phase kinetics study of the atmospherically important reaction between Cl₂ and dimethyl sulfide (DMS) Cl₂ + CH ₃SCH₃ → products (1) has been made using a flow-tube interfaced to a photoelectron spectrometer. The rate constant for this reaction has been measured at 1.6 and 3.0 torr at T = (294 ± 2) K as (3.4 ± 0.7) × 10^-14 cm³ molecule^-1 s ^-1. Reaction (1) has been found to proceed via an intermediate, (CH₃)₂SCl₂, to give CH₃SCH ₂Cl and HCl as the products. The mechanism of this reaction and the structure of the intermediate were investigated using electronic structure calculations. A comparison of the mechanisms of the reactions between Cl atoms and DMS, and Cl₂ and DMS has been made and the relevance of the results to atmospheric chemistry is discussed. The Owner Societies 2005.

Full text of DOI:10.1039/b415566a

View Article Online / Journal Homepage / Table of Contents for this issue
R E S E A R C H P A P E R
A kinetics and mechanistic study of the atmospherically relevant
reaction between molecular chlorine and dimethyl sulfide (DMS)
J. M. Dyke,*a M. V. Ghosh,a D. J. Kinnison,a G. Levita,a A. Morrisa and D. E. Shallcrossb
a School of Chemistry, University of Southampton, Southampton, UK SO17 1BJ
b Bristol Biogeochemistry Research Centre, School of Chemistry, University of Bristol,
Bristol, UK BS8 1TS
Received 8th October 2004, Accepted 6th December 2004
First published as an Advance Article on the web 5th January 2005
A gas-phase kinetics study of the atmospherically important reaction between Cl2 and dimethyl sulfide (DMS)
Cl2 þ CH3SCH3 - products (1)
has been made using a flow-tube interfaced to a photoelectron spectrometer. The rate constant for this reaction
has been measured at 1.6 and 3.0 torr at T ¼ (294 ꢀ 2) K as (3.4 ꢀ 0.7) ꢁ 10ꢂ14 cm3 moleculeꢂ1
s
ꢂ1. Reaction
(1) has been found to proceed via an intermediate, (CH3)2SCl2, to give CH3SCH2Cl and HCl as the products.
The mechanism of this reaction and the structure of the intermediate were investigated using electronic structure
calculations. A comparison of the mechanisms of the reactions between Cl atoms and DMS, and Cl2 and DMS
has been made and the relevance of the results to atmospheric chemistry is discussed.
emissions are extremely important. The primary step of
Introduction
DMS oxidation in the atmosphere is predominantly reaction
with the OH radical during the day and the NO3 radical at
night. Subsequent oxidation in the atmosphere leads to for-
mation of species such as SO2, H2SO4 and CH3SO3H (methane
sulfonic acid or MSA). These species may contribute signifi-
cantly to the acidity of the atmosphere and in the case of
sulfuric acid to cloud formation.9
This paper reports the first study in which a flow-tube has been
interfaced to a photoelectron spectrometer to allow a gas-
phase kinetics study to be performed on an atmospherically
important reaction. Ultraviolet photoelectron spectroscopy
(PES) has the advantage over mass spectrometry in that it
does not suffer from fragmentation problems. It has the
potential to observe reactants, intermediates and products
and to measure branching ratios between reaction channels
which give different intermediates and/or final products. An
example of a reaction studied by PES in Southampton in which
different intermediates were observed is the reaction between F
atoms and propane which gives rise to the n-propyl and iso-
propyl radicals as primary products.1 The only previous use of
PES in a kinetics study is the work of Wang et al., who studied
the unimolecular thermal rearrangements of a number of
Recently molecular chlorine has been observed in coastal
marine air. This is produced at night, as well as during the day,
from heterogeneous reactions of ozone with wet sea-salt and is
enhanced by the presence of ferric ions.10 Employing a high-
pressure chemical ionization mass spectrometry technique,
Spicer et al.11 measured Cl2 levels ranging from o15 to 150
pptv. Night-time Cl2 mixing ratios were in the range 40–150
pptv, with the highest value being observed near midnight,
which dropped to 15 pptv in daylight. A modelling study
conducted in this work found that, shortly after sunrise, the
oxidation rate of DMS by Cl, produced by photolysis of Cl2,
could under favourable conditions, be an order of magnitude
higher than the rate of oxidation by OH. The rate constant for
the reaction between Cl atoms and DMS at 298 K and 1 torr
pressure has been measured as (6.9 ꢀ 1.3) ꢁ 10ꢂ11 cm3
moleculeꢂ1 sꢂ1 using discharge-flow mass spectrometry,12
and at 3 torr pressure and at 297 K it has been measured as
(1.8 ꢀ 0.3) ꢁ10ꢂ10 cm3 moleculeꢂ1 sꢂ1 using time-resolved
resonance detection of Cl.13 A more recent study using cavity
ring-down spectroscopy14 determined the rate constant at
atmospheric pressure as (3.6 ꢀ 0.2) ꢁ10ꢂ10 cm3 moleculeꢂ1
sꢂ1 and a low-pressure limit rate constant consistent with that
found in refs. 12 and 13. It has been found that the Cl þ DMS
reaction, reaction (2), proceeds via two routes, a direct strip-
ping channel (2a) that proceeds without a barrier to give
CH3SCH2 and HCl and an addition channel (2b) to give
(CH3)2SCl.13–16
2
compounds,2–4 notably isomerisation of CH3NC to CH3CN
and isomerisation of SSF2 to FSSF.3
The reaction studied in this work is the atmospherically
important reaction between molecular chlorine and dimethyl
sulfide (DMS)
Cl2 þ CH3SCH3 - products
(1)
This work, as well as related research with photoelectron, IR
and UV/visible spectroscopy and ab initio molecular orbital
calculations,5 has shown that this reaction proceeds via a
covalently bound intermediate, (CH3)2SCl2, and to give
CH3SCH2Cl and HCl as the products.
The sulfur cycle in the earth’s atmosphere has been the
subject of intensive investigation in recent years because of
the need to assess the contribution of anthropogenically pro-
duced sulfur to acid rain, visibility reduction and climate
modification. Anthropogenic emissions of sulfur to the atmos-
phere are dominated by SO2 whereas natural (biogenic) sulfur
emissions are thought to be dominated by dimethyl sulfide
derived from oceanic phytoplankton.6–9 At present, anthropo-
genic emissions of sulfur dominate; however, these emissions
are predominantly in the northern hemisphere. In the southern
hemisphere and in particular the southern oceans, natural
Cl þ CH3SCH3 - CH3SCH2 þ HCl
Cl þ CH3SCH3 þ M - (CH3)2SCl þ M
(CH3)2SCl - CH3 þ CH3SCl
(2a)
(2b)
(2c)
866
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 8 6 6 – 8 7 3
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
Decomposition of the adduct to produce CH3SCl and CH3
(reaction 2c) is very slow, as it has a high activation energy
barrier, and is not competitive with CH3SCH2 and HCl
production.16
The flow-tube
The flow-tube used consists of a 50 cm long stainless steel tube
with an inner diameter of 3 cm. It was pumped by a SV 200
pump (Leybold) that provides linear flow velocities from
3 msꢂ1 up to 40 msꢂ1 of the carrier gas (helium). The total
gas pressure in the flow-tube was measured downstream by a
capacitance manometer (MKS Baratron, 10 torr range). In this
study the pressure within the flow reactor was changed between
1.6 and 3.0 torr and linear flow velocities between 3 and 8 msꢂ1
were employed. The character of the flow in the flow-tube was
laminar, as calculated Reynolds numbers were below 2000.
Under these conditions the flow is characterised by a parabolic
velocity profile across the flow-tube; the molecules close to the
walls experience a higher viscous drag than those in the middle
of the flow-tube and so have a lower velocity.
The kinetic experiments presented in this work were per-
formed under pseudo-first-order conditions. Chlorine, the
reactant in excess (1014–1015 molecules cmꢂ3), was added
through the movable injector (70 cm long) and DMS (concen-
trations less than 1013 molecules cmꢂ3) entered the flow tube
through the fixed side arm. Contact times were between 20 and
150 ms. The pseudo-first-order rate constant was determined
by measuring the relative concentration of DMS, from the
intensity of its first photoelectron band, with the movable
injector at several different positions while the chlorine partial
pressure was in excess and held constant.18 This was then
repeated at different chlorine partial pressures. The flow rates
of all the gases were regulated using mass-flow controllers
(MKS, Type 1179A), which were calibrated for each individual
gas mixture used in these experiments. The flow rates of the
carrier gas (helium) were in the range of B0.3–0.9 SLM and
were much greater than the reactant gas flow rates. The
movable injector (14 mm outer diameter) was also a stainless
steel tube and its external surface was Teflon lined.
Given the known presence of both DMS and Cl2 in the
marine boundary layer, it was thought valuable to measure the
rate constant of reaction (1) at room temperature and deter-
mine its products, using a recently developed instrument which
interfaces a flow-tube to a UV photoelectron spectrometer via
several stages of differential pumping. In addition, to comple-
ment the experimental work, molecular orbital calculations
were used to determine relative energies of intermediates and
their decomposition products, and study the mechanism of
reaction (1). The rate constant of this reaction has not been
measured previously, although a discharge-flow electron im-
pact mass spectrometric study17 gave an upper limit for the rate
constant of 8 ꢁ 10ꢂ14 cm3 moleculeꢂ1 sꢂ1 at room temperature.
Experimental details
The apparatus used to study reaction (1) consists of a stainless
steel flow-tube connected via a differential pumping system to a
photoelectron spectrometer designed to study short-lived spe-
cies in the gas-phase. All experiments were performed at total
pressures of 1.6 and 3.0 torr and at room temperature (294 ꢀ 2)
K. Helium was used as the carrier gas in all experiments. Fig. 1
is a block diagram of the central part of the apparatus, which
shows the flow-tube positioned at an angle of 151 to the
vertical, attached to the ionization chamber of the photoelec-
tron spectrometer and aligned perpendicular to the photon
source.
Sampling system
A two stage differential pumping system was constructed19 to
sample a small fraction of the flow tube mixture into the
ionization chamber of the photoelectron spectrometer. The
experimental sampling system consists of two stainless steel
chambers separated by thin discs with small holes in their
centres. The hole sizes were selected, with the pumping system
used, to allow a gradual decrease in pressure from the flow tube
(B3 torr) down to 10ꢂ5 torr in the ionization chamber of the
photoelectron spectrometer.
The flow tube (at 1.6–3.0 torr total pressure) was sampled
through a hole of 2.5 mm diameter drilled in the centre of a
Teflon disc situated in Chamber 1 (1 in Fig. 1) that is pumped
by a SV 200 rotary pump. It is assumed that once the gas is
sampled from the flow tube into Chamber 1, no further
reaction occurs. Chamber 2 (2 in Fig. 1) is pumped by a second
rotary pump (D25B Leybold) to a pressure of approximately
10ꢂ1 torr. At the end of chamber 2 there is a second small hole
co-axial with the first that is 5 mm diameter, drilled in the
centre of a stainless steel disc leading into chamber 3 (3 in
Fig. 1). Chamber 3 is pumped by a third rotary pump (D25B
Leybold) and is maintained at a pressure r10ꢂ2 torr. At the
end of the third chamber there is a final stainless steel disc with
a hole in the middle of 2 mm diameter (co-axial with the
previous two holes).
Detection system
Fig. 1 Schematic diagram of the flow-tube used in this work, using
detection with a photoelectron spectrometer. In this figure: A—Ana-
lyser Chamber, I—Ionization Chamber, C—Channeltron, D—diffu-
sion pump, ꢀV Hemisphere Voltages, F1—exit focus, F2—entrance
focus, 1—Chamber 1 of flow-tube (total pressure B3 torr), 2,3—
Chamber 2 and 3 of the differential pumping system (approx. pressures
B10ꢂ1 torr and r10ꢂ2 torr, respectively).
The end of the flow tube is situated in the ionization chamber
of the photoelectron spectrometer20,21 and is aligned perpendi-
cular to the radiation source, the helium discharge lamp. The
photon beam was B1.5 cm below the 2 mm hole of Chamber 3
(see Fig. 1).
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 8 6 6 – 8 7 3
867

View Article Online
All photoelectron spectra were recorded using He Ia radia-
tion (21.22 eV) on a single detector photoelectron spectro-
meter. Typical resolution under normal operating conditions
was 25–30 meV as measured for the (3p)ꢂ1 ionization of argon.
Experiments were carried out with the flow tube interfaced
to the photoelectron spectrometer as shown in Fig. 1 with the
three sampling holes, diameters of 2.5, 5.0 and 2.0 mm, to
evaluate the minimum detectable partial pressures in the flow-
tube of Ar, DMS and Cl2 from the measured intensities of their
photoelectron bands. Also it was important to check the range
of partial pressures of the gases in the flow tube over which the
photoelectron signal, as measured with this set-up, was linear.
The minimum detection limits for Ar, DMS and Cl2 were
estimated as 1 ꢁ 1011, 3.5 ꢁ 1011 and 4 ꢁ 1011 molecules cmꢂ3
,
respectively. Also for these three gases the pressure range over
which the photoelectron signal was linear with pressure in the
flow tube was 1 ꢁ 1011–1.1 ꢁ 1014, 3.5 ꢁ 1011–1.6 ꢁ 1013 and
4 ꢁ 1011–1 ꢁ 1014 molecules cmꢂ3, respectively.
(These experiments set the detection limit as a signal: noise
ratio of 2 with an integration time of 1 s).
With this apparatus absolute photoionization cross-sections
can be evaluated. This was achieved by recording PE spectra of
a sample gas (such as DMS, Cl2 and CH3SCH2Cl) at a known
partial pressure in helium and PE spectra, recorded under the
same conditions on the same day, of a known partial pressure
of a second gas such as Ar in helium, for which the photo-
ionization cross-section (s) and angular distribution parameter
(b) are known at the HeI photon energy. The results of these
measurements will be presented in a separate paper.5
The ionization energy (IE) scale of spectra recorded during
kinetics experiments were calibrated using the first vertical
ionization energy (VIE) of DMS (8.72 eV)22 and either the
lowest spin–orbit component of the first band of HCl (12.75
eV)23 or the VIE of the second band of Cl2 (14.43 eV)24 when
the HCl signal was either undetectable or out of scale.
Preparation of reactant gases
Each reagent, DMS (99þ%, Aldrich) or chlorine (99.9%, Air
Products) was mixed with He and stored in 6 L Pyrex bulbs.
The mixtures (typically 10% Cl2 in He and 1% DMS in He)
were made by passing a gas of known pressure into an
evacuated bulb and then filling the bulb to atmospheric
pressure with He (BOC, CP grade). The gases were handled
within a Pyrex manifold, with pressures being measured by 10,
100 and 1000 torr capacitance manometers (MKS). The freeze-
pump-thaw method was used to purify DMS and chlorine. The
He carrier gas was flowed through two molecular sieves to
ensure removal of water, CO2 and hydrocarbon impurities
before entering the flow tube.
Fig. 2 Spectra recorded at different mixing distances for the Cl2
þ
DMS reaction. (a—6 cm mixing distance, b—13 cm mixing distance,
c—25 cm mixing distance). It can be seen that the bands of DMS (first
band at 8.72 eV) and Cl2 (first band at 11.65 eV) decrease and the
bands of CH3SCH2Cl (bands at 9.18 and 10.97 eV) and HCl (first band
at 12.75 eV VIE) increase with mixing distance (reaction time) in the
flow-tube. The initial partial pressures of DMS and Cl2 in this experi-
ment were approximately the same. The mixing distances shown in the
caption are the distances of the movable injector in Chamber 1 above
the small hole (2.5 mm diam.) between Chamber 1 and Chamber 2
(see Fig. 1). Bands associated with a reaction intermediate are labelled
as ‘Intermed.’
Results
Initial survey spectra
Although qualitative studies of the products of reaction (1)
observed as a function of time have been made on another
photoelectron spectrometer, and the results will be reported
separately,5 it is useful to present survey spectra obtained with
the spectrometer used in this work. In these survey experiments
the initial concentrations of DMS and Cl2 were kept constant
and were approximately the same. The distance of the central
injector from the sampling hole between Chambers 1 and 2 was
changed in the range 0–35 cm and photoelectron spectra were
recorded at each mixing distance. Some representative results
are shown in Fig. 2. Inspection of Fig. 2 indicates that the first
bands of DMS (vertical ionization energy (VIE) 8.72 eV) and
Cl2 (VIE 11.65 eV) decrease with reaction time whereas the
bands associated with the products, HCl (VIE 12.75 eV) and
CH3SCH2Cl (VIE 9.18 eV), increase with reaction time. Also,
observed are bands associated with a long-lived reaction inter-
mediate at 9.69 and 10.62 eV vertical ionization energy. The
bands associated with the reaction intermediate were found in
this and other studies to maximize and start to decrease before
the product bands reach maximum intensity, and were assigned
on the basis of photoelectron and IR spectroscopic evidence to
the intermediate (CH3)2SCl2.5
Rate constant measurement
Reaction (1) was studied with [Cl2] in excess over [DMS] at two
different pressures, 1.6 and 3.0 torr. Flow velocities in the flow
tube were approximately 4 msꢂ1 at 1.6 torr and 8 msꢂ1 at
868
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 8 6 6 – 8 7 3
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
3.0 torr. Helium was used as the carrier gas. In order to
measure the rate constant of reaction (1), the most intense
band of DMS observed at 8.72 eV VIE was used to monitor the
change in DMS concentration after the addition of Cl2 as a
function of the mixing distance. In a typical experiment, DMS
mixed in helium was allowed to pass through the flow-tube for
about 30 min before the addition of Cl2 in helium since DMS is
relatively viscous and it was essential to reach stable conditions
where a constant PE signal of DMS was obtained. Cl2 was then
added and the intensity of DMS was seen to drop. The helium
carrier flow was reduced to maintain the fixed pressure in the
flow tube. The injector was repositioned at a different mixing
distance and the experiment repeated. The pseudo-first order
rate constant at a known concentration of Cl2 is given by the
ꢀ
ꢁ
½DMSꢃ
½DMSꢃ0t
different positions of the injector. Examples of the pseudo-first
order plots are shown in Fig. 3.
gradient of the plot of ln
against the contact time for
Fig. 4 Second-order plot for the reaction of Cl2 with DMS at a total
pressure of 1.6 torr.
Wall losses were reduced by using an internal Teflon coating
in the flow-tube. The wall loss rate constant, kwall, was esti-
mated by introducing DMS through the injector in the absence
of chlorine at different mixing points above the 2.5 mm
sampling hole at the end of the flow-tube. In this way the
contact time of DMS with the walls of the flow-tube could be
changed and an estimate of the wall loss rate constant, kw,
pseudo-first order rate constants for axial and radial diffusion
using the method described by Keyser25 with estimated values
of the diffusion coefficient of DMS in helium at 1.6 torr of
164.6 cm2 sꢂ1 and at 3.0 torr of 91.1 cm2 sꢂ1. These corrections
gave rise to a correction in the pseudo-first order rate constant
of less than 20% with the axial diffusion correction being an
order of magnitude larger than the radial diffusion correction.
Plotting the corrected pseudo-first order rate constants
against the molecular chlorine concentration yields the second
order rate constant. The results obtained are (3.39 ꢀ 0.52) ꢁ
10ꢂ14 cm3 moleculeꢂ1 sꢂ1 at 1.6 torr, Fig. 4, and (3.41 ꢀ
0.70) ꢁ 10ꢂ14 cm3 moleculeꢂ1 sꢂ1 at 3.0 torr, Fig. 5, where the
errors quoted are the statistical standard errors of the slopes
(both values were determined at a temperature of (294 ꢀ 2) K).
A comparison of the results obtained at 1.6 and 3.0 torr is
given in Fig. 6. Overall the agreement between the two data sets
is good, although an insufficient pressure range has been
studied to estimate the pressure dependence fully. The values
obtained for the second order rate constant at the two pres-
sures are in good agreement with each other and it is reassuring
that they are less than the upper limit value estimated by
Butkovskaya et al.17 of 8 ꢁ 10ꢂ14 cm3 moleculeꢂ1 sꢂ1 by
discharge-flow electron impact mass spectrometry.
ꢀ
ꢁ
½DMSꢃ
½DMSꢃ00
time (where [DMS]0 is the concentration of DMS at zero
could be made by plotting ln
as a function of contact
t
0
contact time and [DMS]t is the concentration of DMS at time
t after the wall loss). It was found that the measured wall loss
rate constant was very small, kwall D 0, suggesting that
heterogeneous losses were much smaller than the homogeneous
ꢀ
ꢁ
½DMSꢃ
½DMSꢃ0t
corrected for this effect. Corrections were also made to the
losses. For each point in Figs. 3(a) and (b), ln
has been
Ab initio calculations
In order to investigate the mechanism of reaction (1), ab initio
calculations have been carried out at the MP2 level using aug-
cc-pVDZ (aVDZ) basis sets26 with the GAUSSIAN 03
Fig. 3 Typical pseudo-first order plots for reaction (1) at (a) 1.6 and
(b) 3.0 torr, respectively obtained from DMS decay kinetics in excess of
Cl2. (The values listed on each graph are the concentrations (molecules
cmꢂ3) of Cl2 used in each experiment). Each point has been corrected
for wall losses (see text).
Fig. 5 Second-order plot for the reaction of Cl2 with DMS at a total
pressure of 3 torr.
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 8 6 6 – 8 7 3
869

View Article Online
minimum energy structures were located on the potential sur-
face (see Fig. 7), most notably a reaction intermediate
(CH3)2SCl2 (see Fig. 8). In separate experiments,5 it was found
that the photoelectron spectrum obtained for a reaction inter-
mediate from reaction (1), bands at 9.69 and 10.62 eV in Fig. 2
are part of this spectrum, could not be assigned to reactant or
product-type intermediates, DMS:Cl2 or CH3SCH2Cl:HCl,
but could be assigned to (CH3)2SCl2. There is also IR evidence
to support this conclusion.5 Computed minimum energy struc-
tures of the reactant intermediate (DMS:Cl2, Cs), the inter-
mediate ((CH3)2SCl2) and the product intermediate
(CH3SCH2Cl:HCl trans) are shown in Fig. 8 and the computed
structures of the transition states are shown in Fig. 9.
Based on the schematic potential energy diagram shown in
Fig. 7, the reaction proceeds from the reactants, DMS and Cl2,
to a reactant-type intermediate DMS:Cl2 then to a structure of
the type (CH3)2SCl:Cl. It then passes over a transition state
(TS1) to (CH3)2SCl2, the intermediate associated with the
photoelectron bands at 9.69 and 10.62 eV in Fig. 2. This then
decomposes via TS2, which is lower than TS1, to the products
via a product-type intermediate. There is also a route which by-
passes the (CH3)2SCl2 intermediate via TS3 but this is higher in
energy than TS1. Fig. 7 indicates that the rate determining step
of reaction (1) is passage over the TS1. This has an energy
relative to the reactants of þ4.5 kcal molꢂ1 (DHz ¼ 4.89 kcal
molꢂ1, DSz ¼ ꢂ37.8 cal Kꢂ1 molꢂ1). Use of these values for
DHz and DSz with the standard transition state expression gives
1 ꢁ 10ꢂ20 cm3 moleculeꢂ1 sꢂ1 for the rate constant, clearly
much lower than the experimentally measured rate constant of
(3.4 ꢀ 0.7) ꢁ 10ꢂ14 cm3 moleculeꢂ1 sꢂ1. This indicates that the
energy of TS1 relative to the reactants is too high (probably by
E1–2 kcal molꢂ1) and the computed entropy change from the
reactants to TS1 is too negative. This almost certainly arises
because a modest basis set has been used in the calculations
performed in this work. Improved calculations would involve
using a larger basis set. These should lead to improved values
of DHz and DSz from the reagents to the TS1.
Fig. 6 Combined second-order plots for the reaction of Cl2 with DMS
at a total pressure of 1.6 (n) and 3.0 torr (K).
programme.27 A search was made for minimum energy struc-
tures and transition states which are inter-connected on the
potential energy surface between the reagents (Cl2 and DMS)
and the products (CH3SCH2Cl and HCl). Then for the mini-
mum energy structures and transition states located at the
MP2/aug-cc-pVDZ level, fixed point CCSD(T) calculations
were performed to provide improved values of the total
energies. The MOLPRO programme28 was used for the
CCSD(T) calculations. For all these CCSD(T) calculations,
the T1 diagnostic was acceptably small. In this procedure, the
MP2 calculations include some dynamic electron correlation
and the CCSD(T) calculations include higher order dynamic
electron correlation.
Transition states were characterised via harmonic frequency
analysis and connected to minimum energy structures through
IRC (intrinsic reaction co-ordinate) calculations. The station-
ary points located on the potential energy surface are repre-
sented in Fig. 7 where the computed energies of the stationary
points obtained are shown as bold horizontal lines and the
transition states are labelled TSn. As expected, the reactant-
type intermediate DMS:Cl2 and product-type intermediate
CH3SCH2Cl:HCl were located as minimum energy structures.
However, apart from the reactants and products, nine other
Discussion
It is interesting to compare reaction (1) with the reaction
between Cl and DMS, reaction (2). Both reactions involve
Fig. 7 Relative electronic energy diagram for reaction (1). Minimum energy structures and transition states were located at the MP2/ aug-cc-pVDZ
level. Then fixed point CCSD(T) calculations were performed to provide improved values of the total energies.
870
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 8 6 6 – 8 7 3
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
Fig. 8 Computed minimum energy structures of the reactant-inter-
mediate (DMS:Cl2, Cs), the intermediate (CH3)2SCl2, and the product-
intermediate CH3SCH2Cl:HCl trans.
formation of a complex, (CH3)2SCl2 in the case of reaction (1)
and (CH3)2SCl in the case of reaction (2). The (CH3)2SCl
intermediate formed in reaction (2) has been detected by
electron impact mass spectrometry,12 UV-visible absorption
spectroscopy15 and cavity-ring down laser spectroscopy14
whereas the (CH3)2SCl2 intermediate formed in reaction (2)
has been observed in this work by photoelectron spectroscopy.
A comparison of the computed molecular structures and
geometrical parameters for these two complexes is made in
Fig. 10. In this figure, the parameters for the (CH3)2SCl
complex are taken from ref. 16, where they were computed at
the UMP2/DZP level, and the parameters for (CH3)2SCl2 were
computed in this work at the MP2/aug-cc-pVDZ level. These
structures are surprisingly similar with the C–S–Cl angle equal
to 91.91 in both cases. The Cl–S distance is longer and the C–S
distance shorter in (CH3)2SCl than in (CH3)2SCl2 (see Fig. 10).
SCl4 is also expected to have a very similar geometry with a
pseudo trigonal-bipyramidal co-ordination geometry at the S
atom. The Cl–S–Cl bond angle has been computed to be 169.71
for the axial Cl–S–Cl unit.29
Fig. 9 Transition states located on the potential surface of reaction (1)
(see text and Fig. 7).
10ꢂ10 cm3 moleculeꢂ1 ꢂ1. The low pressure limit value is
s
(6.9 ꢀ 1.3) ꢁ 10ꢂ11 cm3 moleculeꢂ1 sꢂ1 12,13
.
Reaction (2)
proceeds by two channels, (2a) the hydrogen abstraction
channel, which is pressure independent, and (2b) the addition
channel which is pressure dependent.
Stickel et al.,13 in a study of reaction (2) using time-resolved
detection of Cl atoms, reported that the branching ratio of
reaction (2a) approaches unity with decreasing total pressure
and k2 increases with pressure. Both reactions (2a) and (2b) are
expected to have large rate constants, as they proceed without
a barrier, whereas reaction 2c, decomposition of (CH3)2SCl to
CH3 and CH3SCl, is very slow because it has a high activation
energy barrier of E18 kcal molꢂ1 16
.
Reaction (1), unlike reaction (2), has only one pathway
which can be written as:-
Reaction (2) is much faster than reaction (1) with the most
recent measurement of the rate constant for this reaction being
made by cavity-ring-down spectroscopy,14 with a rate constant
at atmospheric pressure being obtained as k2 ¼ (3.6 ꢀ 0.2) ꢁ
Cl2 þ CH3SCH3 þ M - (CH3)2SCl2 þ M
(CH3)2SCl2 - CH3SCH2Cl þ HCl
(1a)
(1b)
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 8 6 6 – 8 7 3
871

View Article Online
The reactive halogen species involved will include Cl atoms,
BrO radicals37 and possibly IO radicals.38
This present work has shown that the interaction between
Cl2 and DMS is not very slow, and that it produces molecular
products CH3SCH2Cl and HCl. There are field observations
that Cl2 levels can reach 150 ppt at night. If it is assumed that
the average level of Cl2 at night is 50 ppt, then the lifetime of
DMS at night with respect to reaction (1) is around 6 h. If it is
assumed that DMS levels are on average 50 ppt at night (field
measurements suggest that this is perfectly reasonable in the
southern ocean, for example) and that the DMS and Cl2 levels
are being replenished at night to maintain these levels, over a
6 h period, around 40 ppt of CH3SCH2Cl would be generated,
assuming that CH3SCH2Cl simply builds up at night. In the
morning, photolysis of Cl2 will be rapid and reaction (1) will be
ineffective as Cl2 levels fall rapidly, but the CH3SCH2Cl, which
has built up during the night, can either react with OH or be
photolysed. The fate of CH3SCH2Cl will be discussed more
fully in forthcoming papers; however, work from this labora-
tory suggests that photolysis to yield CH3S and CH2Cl will be
quite rapid, with a CH3SCH2Cl lifetime of several hours. CH3S
is known to undergo rapid oxidation to CH3SO2 via reaction
with NO2 or O3 and CH3SO2 is known to undergo thermal
decomposition to yield SO2 and CH3. Therefore, the night-time
interaction between Cl2 and DMS may well provide a mechan-
ism to speed up SO2 production in the day and go some way to
explain the discrepancy between DMS decay rates and SO2
production rates during the day. If the levels of Cl2 observed by
Spicer and co-workers11 are representative of the open ocean,
then reaction (1) could play an important role in DMS oxida-
tion and the coupling between halogens and DMS cannot be
ignored in climate studies.
Fig. 10 Computed structures for (A) the (CH3)2SCl and (B) the
(CH3)2SCl2 intermediates. The (CH3)2SCl structure was taken from
ref. 16 computed at the UMP2/DZP level and the (CH3)2SCl2 structure
was computed in this work at the MP2/aug-cc-pVDZ level.
Clearly it would be valuable to establish the pressure depen-
dence of k1, and the value of k1 at atmospheric pressure will be
of particular importance in atmospheric modelling calcula-
tions. In view of the relative values of k1 and k2, reaction (1)
will not be important in the atmosphere during the day, when
photolysis of Cl2 to Cl will occur, but it will be important at
night when conversion of DMS to CH3SCH2Cl will take place
via reaction (1), as discussed in the next section.
Conclusions
A flow-tube has been interfaced to a photoelectron spectro-
meter for the first time in order to measure rate constants of
reactions of importance in atmospheric chemistry. The reac-
tion between Cl2 and DMS was the first reaction to be studied
in this way and its rate constant was measured at 1.6 and
3.0 torr at room temperature, (294 ꢀ 2) K, returning a value of
k1 ¼ (3.4 ꢀ 0.7) ꢁ10ꢂ14 cm3 moleculeꢂ1 sꢂ1
.
Photoelectron spectra recorded as a function of reaction
time and supporting molecular orbital calculations showed
that the reaction proceeds through an intermediate,
(CH3)2SCl2, which has a lifetime of E30 ms under the condi-
tions used, with subsequent production of CH3SCH2Cl and
HCl. The atmospheric implications of these results have been
briefly discussed.
Atmospheric implications
The oxidation of DMS has been highlighted by Charlson and
co-workers30 as being a potentially important natural pathway
for the generation of cloud condensation nuclei (CCN), and
could provide a key negative feedback to the earth’s radiative
balance. Crucial to the assessment of this ‘CLAW’ hypothesis
is an understanding of factors that control the flux of DMS
from the oceans to the atmosphere; these include sea surface
temperature, surface wind speed, CO2 levels, nutrient levels
etc., some or all of which may change significantly in the wake
of climate change. However, once in the atmosphere, in order
for DMS to affect aerosol loading the rate of its oxidation and
subsequent conversion to SO2 and then H2SO4 is crucial. There
have been many field and laboratory studies in the last decade
(e.g. refs. 31–34) and yet there are still considerable uncertain-
ties concerning the oxidation of DMS.35 However, it appears
from field studies in particular that DMS oxidation in the
atmosphere is more rapid than can be accounted for using
existing models, or more correctly, the production of SO2
(assumed from DMS) is more rapid than expected. The
dominant oxidant for DMS is thought to be the OH radical,
and although there is evidence that the NO3 radical can be an
important oxidant at night in coastal areas,36 its significance in
the open ocean is thought to be very small. Halogen chemistry
has recently been considered as a possibility to account for the
increased oxidation rate,35 and may well be part of the puzzle.
Acknowledgements
The authors gratefully acknowledge support of this work from
NERC via a COSMAS grant, which also supported M. V. G
with a postdoctoral fellowship. G. L. was supported by a
postgraduate studentship from the EU Reactive Intermediates
Network. The authors also gratefully acknowledge advice from
and discussions with Dr C. E. Canosa-Mas and Prof. R. P.
Wayne of Oxford University and Dr E. P. F. Lee of South-
ampton University. The Biogeochemistry Research Centre at
Bristol University is a joint initiative between the School
of Chemistry and the Departments of Earth Science and
Geography.
References
1
J. M. Dyke, A. R. Ellis, N. Jonathan and A. Morris, J. Chem.
Soc., Faraday Trans. 2, 1985, 81, 1573.
D. Wang, X. Qian and J. Peng, Chem. Phys. Lett., 1996, 258, 149.
2
872
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 8 6 6 – 8 7 3
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5

View Article Online
3
4
X. Cao, X. Qian, C. Qiao and D. Wang, Chem. Phys. Lett., 1999,
299, 322.
(a) D. Wang and X. Qian, J. Electron Spectrosc. Relat. Phenom.,
1997, 83, 41; (b) D. X. Wang, S. Li, D. Wang and Y. Li, J. Electron
Spectrosc. Rel. Phenom., 1994, 70, 173.
J. M. Dyke, M. V. Ghosh, M. Goubet, G. Levita, K. Miqueu,
J. S. Ogden and D. E. Shallcross, unpublished work.
C. F. Cullis and M. M. Hirschler, Atmos. Environ., 1980, 14,
1263.
H. Rodhe and I. Isaksen, J. Geophys. Res., 1980, 85, 7401.
T. S. Bates, B. K. Lamb, A. Guenther, J. Dignon and R. E.
Stoiber, J. Atmos. Chem., 1992, 14, 315.
J. Chem. Phys., 1993, 98, 1358; (c) T. H. Dunning, Jr., J. Chem.
Phys., 1989, 90, 1007.
27 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M.
A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K.
N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi,
V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.
A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Strat-
mann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J.
J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels,
M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul,
S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M.
A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P.
M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez
and J. A. Pople, GAUSSIAN 03, (Revision B.01), Gaussian,
Inc., Pittsburgh PA, 2003.
5
6
7
8
9
H. Berresheim, P. H. Wine and D. D. Davis, in Composition,
Chemistry and Climate of the Atmosphere, van Nostrand Reinhold,
New York, 1995, p. 251.
10 Y. Sadanaga, J. Hirokawa and H. Akimoto, Geophys. Res. Lett.,
2001, 28, 4433.
11 C. W. Spicer, E. G. Chapman, B. J. Finlayson-Pitts, R. A.
Platridge, J. M. Hubbe, D. J. Fast and C. M. Berkowitz, Nature,
1998, 394, 353.
12 Y. Diaz-de-Mera, A. Aranda, D. Rodrigues, R. Lopez, B. Caba-
nas and E. Martinez, J. Phys. Chem. A, 2002, 106, 8627.
13 R. E. Stickel, J. M. Nicovich, S. Wang, Z. Zhao and P. H. Wine,
J. Phys. Chem., 1992, 96, 9875.
14 S. Enami, Y. Nakano, S. Hashimoto, M. Kawasaki, S. Aloisio and
J. S. Francisco, J. Phys. Chem. A, 2004, 108, 7785.
15 S. P. Urbanski and P. H. Wine, J. Phys. Chem. A, 1999, 103,
10935.
28 MOLPRO is a package of ab initio programs written by H.-J.
Werner and P. J. Knowles, with contributions from J. Allmf, R. D.
Amos, A. Berning, D. L. Cooper, M. J. O. Deegan, A. J. Dobbyn,
F. Eckert, S. T. Elbert, C. Hampel, R. Lindh, A. W. Lloyd, W.
Meyer, A. Nicklass, K. Peterson, R. Pitzer, A. J. Stone, P. R.
Taylor, M. E. Mura, P. Pulay, M. Schutz, H. Stoll and T.
Thorsteinsson.
16 S. M. Resende and W. B. Almeida, J. Phys. Chem. A, 1997, 101,
9738.
29 Y. Drozdova, R. Steudel, W. Koch, K. Miaskiewicz and I. A.
Topol, Chem. Eur. J., 1999, 6, 1936.
17 N. I. Butkovskaya, G. Poulet and G. LeBras, J. Phys. Chem.,
1995, 99, 4536.
18 C. Howard, J. Phys. Chem, 1979, 83, 3.
19 D. J. Hucknall and A. Morris, Vacuum Technology, Calculations
in Chemistry, Royal Society of Chemistry, Cambridge, UK, 2003.
20 J. M. Dyke, A. Morris and N. Jonathan, Int. Rev. Phys. Chem.,
1982, 2, 3.
30 R. J. Charlson, J. E. Lovelock, M. O. Andreae and S. G. Warren,
Nature (London), 1987, 326, 655.
31 W. J. Bruyn, M. Harvey, J. M. Cainey and E. S. Saltzman,
J. Atmos. Chem., 2002, 41, 89.
32 G. Chen, D. D. Davis, P. Kasibhatla, B. R. Bandy, D. C.
Thornton, B. J. Huebert, A. D. Clarke and B. W. Blomquist,
J. Atmos. Chem., 2000, 37, 137.
21 J. M. Dyke, N. Jonathan and A. Morris, Electron Spectroscopy,
Academic Press, London, 1979, vol. 3, 189.
33 J. Sciare, E. Baboukas and N. Mihalopoulos, J. Atmos. Chem.,
2001, 39, 281.
22 K. KimuraS. Katsumata, Y. Achiba, T. Yamazaki and S. Iwata,
Handbook of He I Photoelectron Spectra, Japan Scientific Societies
Press, Halsted Press, Tokyo, 1981.
23 D. W. Turner, C. Baker, A. D. Baker and C. R. Brundle,
Molecular Photoelectron Spectroscopy, Wiley-Interscience,
London, 1970.
24 A. B. Cornford, D. C. Frost, C. A. McDowell, J. L. Ragle and
I. A. Stenhouse, J. Chem. Phys., 1971, 54, 2651.
25 L. F. Keyser, J. Phys. Chem., 1984, 88, 4750.
26 (a) R. A. Kendall, T. H. Dunning, Jr. and R. J. Harrison, J. Chem.
Phys., 1992, 96, 6796; (b) D. E. Woon and T. H. Dunning, Jr.,
34 G. Knight and J. N. Crowley, Phys. Chem. Chem. Phys., 2001, 3,
393.
35 D. D. Lucas and R. G. Prinn, J. Geophys. Res., 2002, 107, DOI:
10.1029/2001JD000843.
36 H. Bardouki, H. Berresheim, J. Sciare, G. Kouvarakis, J. Schnei-
der and N. Mihalopoulos, Atmos. Chem. Phys., 2003, 3, 1871.
37 T. Ingham, D. Bauer, R. Sander, P. J. Crutzen and J. N. Crowley,
J. Phys. Chem., 1999, 103, 7199.
38 J. D. James, R. M. Harrison, N. H. Savage, A. G. Allen, J. L.
Grenfell, B. J. Allan, J. M. C. Plane, C. N. Hewitt, B. Davison and
L. Robertson, J. Geophys. Res., 2000, 105(26), 379.
T h i s j o u r n a l i s & T h e O w n e r S o c i e t i e s 2 0 0 5
P h y s . C h e m . C h e m . P h y s . , 2 0 0 5 , 7 , 8 6 6 – 8 7 3
873