Atmospheric DMSO degradation in the gas phase: Cl-DMSO reaction. Temperature dependence and products

Article abstract of DOI:10.1021/es010169s

The reactions of Cl atoms and ClO radicals with CH₃-SOCH₃ (DMSO) have been studied using the discharge flow method with direct detection of DMSO, ClO, and products by mass spectrometry. The absolute rate constant at room temperature

Full text of DOI:10.1021/es010169s

Environ. Sci. Technol. 2002, 36, 1226-1230
(
4), and periodical evaluations of sulfur compounds levels
Atmospheric DMSO Degradation in
the Gas Phase: Cl-DMSO
Reaction. Temperature Dependence
and Products
in the atmosphere are published (5).
Dimethyl sulfoxide has been observed in the marine
boundary layer (6) and in rainwater samples and aerosols
(
7). Furthermore DMSO has been identified as a product of
the OH initiated oxidation of DMS in laboratory chamber
studies (8-10). The addition of OH to DMS to form the CH S-
OH)CH adduct and further reaction with O is considered
3
(
3
2
as the main production source of DMSO in chamber studies
and in the atmosphere. Recently (11), it has been reported
that the reactions of DMS with halogen oxides, mainly BrO,
may be an additional source of DMSO in the remote marine
boundary layer. DMSO is also found in the seawater as a
result of bacterial transformation of DMS, but this is not
expected to represent a direct source since it is highly soluble
in the liquid phase.
E R N E S T O M A R T IÄ N E Z , *
A L F O N S O A R A N D A ,
YO L A N D A D IÄ A Z D E M E R A ,
D I A N A R O D R IÄ G U E Z ,
M . R E YE S L OÄ P E Z , A N D J O S E´ A L B A L A D E J O
Facultad de Ciencias Qu ´ı micas, Universidad de
Castilla-La Mancha, Avda, Camilo Jos e´ Cela, n° 10,
1
3071-Ciudad Real, Spain
The fate of DMSO in the atmosphere is poorly understood.
Its removal from the gas phase may be due to a combination
of heterogeneous processes and homogeneous reactions with
The reactions of Cl atoms and ClO radicals with CH3-
SOCH3 (DMSO) have been studied using the discharge
flow method with direct detection of DMSO, ClO, and products
by mass spectrometry. The absolute rate constant at
room temperature measured for reaction 1, (CH3)2SO + Cl
3 3
OH, NO , O , and Cl atoms. In relation to the heterogeneous
fate of DMSO it is recognized as an important precursor of
MSA (12), although recent observations disagree on the
uptake of DMSO onto the particles. Putaud et al. (13) estimate
low losses of DMSO from the gas phase due to its small
Henry’s constant, while for Sciare et al. (7) the DMSO could
account for 40% of the MSA observed in rainwater.
-
11
3
-1
f products, was k(1) ) (1.7 ( 0.3) × 10 cm molecule
. For reaction 2, (CH3)2SO + ClO f products, only an upper
-
1
s
Recent investigations about DMSO-OH reactions have
been reported (9, 10, 14-16) with proposed rate constants
-
14
3
-1
limit could be established, k(2) e 6 × 10 cm molecule
-
1
s
. Reaction 1 has been found to proceed through
-10
3
-1 -1
ranging from 0.6 to 1 × 10 cm molecules
s . Neverthe-
adduct formation and further decomposition involving the
cleavage of the C-S bound. The pressure effect on the Cl-
DMSO reaction from 0.5 to 3 Torr was negligible, and the
temperature dependence in the range 273-335 K was
also very slight. The results obtained are related to previous
studies of sulfur compounds, and the atmospheric
implications are also discussed in relation to the homogeneous
sinks of DMSO. Tropospheric lifetimes of DMSO based
on average Cl and ClO concentrations and the measured
rate constants have been calculated showing that the
contribution of reaction 1 must be of minor relevance in
the marine boundary layer. Reaction 2 is so slow that it does
not play any role within the atmospheric sulfur chemistry.
less a complete agreement in relation to the products of the
reaction is not found. From Hynes and Wine (14) and
Urbanski et al. (15) a yield of unity is expected in CH
CH S(O)OH, (methanesulfinic acid, MSIA) after OH addition
to DMSO, while Sorensen et al. (9) have observed DMSO
a main product. The NO
temperature rate constant of (5.0 ( 3.8) × 10
molecules (16), and the kinetics of DMSO with O
3
and
3
2
as
3
-DMSO reaction shows a room-
13 3
-
cm
-
1 -1
s
3
seems
s . Up to now,
-
19
3
-1 -1
to be very slow k < 1 × 10 cm molecule
only a previous relative measurement of the reaction Cl-
DMSO has been reported with a room-temperature kinetic
rate constant of k ) (7.4 ( 1.0) × 10-11 cm
3
molecule-1
-1
s
(16). In that work, yields of 28 ( 12% on SO and 8 ( 2% in
2
2
DMSO were found, the latter argued as a result of DMSO
secondary reactions with oxidizing intermediates such as
peroxy radicals or ClO.
Recently (17) it has been found that heterogeneous
reactions of nitrogen oxides on sea salt particles can lead to
significant concentrations of photolytic precursors of Cl
atoms in the marine boundary layer. It seems that the main
Introduction
DMS has been recognized as the main natural source of sulfur
in the atmosphere. It is produced in oceanic waters by both
benthic and planktonic marine organisms (1) suggesting that
it may be found in the surface’s ocean. It is thought to
originate from the decomposition of dimethyl-sulfoniopro-
pionate produced by marine organisms in particular phy-
toplankton (2). In the remote marine boundary layer, where
atmospheric source of atomic chlorine is the ClNO
by heterogeneous reaction of N vapor with moist NaCl
(18). Since global average estimates of Cl atoms concentra-
2
generated
2
O
5
3
4
-3
tions range from 10 to 10 atom cm (19, 20) in marine
areas, the effect on the removal of DMSO should be evaluated
and compared to that of the OH-DMSO reaction, presumably
the most relevant sink. The relative rate of these two reactions
may define the global distribution of products from the
atmospheric oxidation of DMS under local conditions.
These precedents together with the unknowns on the
kinetic behavior of the system DMSO-Cl, like the not fulfilled
mass balance or the temperature dependence or the influence
of pressure, make further investigations necessary. In this
work we report the first absolute kinetic study of the Cl-
DMSO reaction, its temperature dependence in the range
contributions from anthropogenic SO
phase oxidation of DMS is thought to be the most important
source of oxidized sulfur compounds such as SO , meth-
anesulfonic acid (MSA), dimethyl sulfoxide, dimethyl sulfone,
and H SO . Although sulfur compounds are minor constitu-
2
are small, the gas-
2
2
4
ents of the Earth’s Atmosphere, they play a significant role
in various global scale phenomena, including the climate (3)
and the generation of acid rains. In this sense, excellent
reviews on laboratory and field research are now available
2
73 to 335 K for pressures from 0.5 to 3.0 Torr, the observations
*
Corresponding author phone: +34 926 29 53 00; fax: + 34 926
2
9 53 18; e-mail: emartine@qifi-cr.uclm.es.
of the products of reaction, and the possible mechanism.
1
2 2 6
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 6, 2002
10.1021/es010169s CCC: $22.00

2002 Am erican Chem ical Society
Published on Web 02/08/2002

measuring the NO
NO signal from an absolute calibration. In these calibration
experiments, to generate ClO, Cl atoms were used in excess
over O to avoid the reaction
2
formed at m/ e ) 46 using a reference
2
3
NO + O f NO + O (6)
3
2
2
-
14
cm molecule s-1 (22)
3
-1
k ) 1.8 × 10
which would contribute to the NO
titration. O
2
signal disturbing the ClO
FIGURE 1. Schematic diagram of the experimental system.
3
was produced externally by passing pure O
through an ozone generator (Ozogas, 5 g/ h) and purified
and stored in a liquid N trap before use. DMSO (diluted in
2
Some results concerning the reaction between ClO and
DMSO, which had not been studied previously, are also
reported for room-temperature and low-pressure conditions.
2
He) was introduced through the movable injector.
The concentrations of reactants in the reactor were
calculated from their pressure decrease rates in calibrated
volume bulbs and the flows of the carrier gas, helium, were
regulated and measured by mass flow controllers. For reaction
Experimental Section
The kinetic experiments were carried out using the discharge-
flow technique and a detection system based on mass
spectrometry. The apparatus is shown schematically in Figure
2
, DMSO was stored in a glass trap in the liquid phase with
a known volume and ended in a calibrated capillary. Different
concentrations could be obtained by bubbling different flows
of He through the trap and measuring the decrease in the
capillary and the time of bubbling.
1
. The reactor consists on a 2.7 cm internal diameter and 100
cm length Pyrex flow tube with a jacket for the thermostating
liquid circulation (SilOil). A movable injector (120 cm length,
1
cm i.d.) was used for the introduction of the reactants. The
The chemicals used had the following purities: helium
inner surfaces of the reactor and the movable injector were
coated with halocarbon wax to reduce heterogeneous losses
of atoms and radicals. The reactor was pumped by means
(
Praxair, 99.999%), passed through a N
the reactor; O (Praxair, 99.999%), Cl (Praxair, >99.8%), CH
SOCH (Aldrich, 99.9%), C Br (Aldrich, 98%), C Cl (Fluka,
g99.5%), NO (Praxair, g99%), NO (Praxair, g99.5%). The
2
trap before entering
2
2
3
-
3
2
H
3
2 3
H
3
-1
of a rotary pump (70 m h ), and the pressure was measured
by means of a 10 Torr full-scale capacitance gauge. The
2
reactants were purified from trap-to-trap distillation before
use.
+
quadrupole mass spectrometer (VG Smart IQ ) was equipped
with electron impact ion source and a Channeltron electron
multiplier. The energy of the ionising electrons was typically
Results
E ) 40 eV. The high vacuum chamber was pumped by means
e
Reaction DMSO + Cl. For reaction 1 the kinetic experiments
were conducted under pseudo-first-order conditions with
Cl atoms in great excess over DMSO with a ratio [Cl]/ [DMSO]
from 10 to 150. In most of the experiments the concentration
-1
of a turbo-molecular pump (Varian V 550 L‚s ), and a copper
device cooler with liquid nitrogen surrounding the electron
impact source was used to reduce the background spectrum.
9
The detection limits for DMSO and ClO were 2 × 10 and 7
1
1
-3
of DMSO was close to 2 × 10 molecule cm and Cl atoms
9
-3
×
10 molecule cm , respectively.
1
2
13
-3
ranged from 1 × 10 to 1 × 10 atom cm . Although an
exact measurement of the DMSO was not required it could
be estimated from the rate of its pressure decay in the storage
balloon. Cl atoms were introduced in the reactor through a
fixed port, while DMSO and bromoethene or chloroethene
for Cl calibration entered the flow tube through the sliding
injector (Figure 1). The flow velocities were fixed around
Cl atoms, precursors of ClO radicals in our experiments,
were generated by passing a Cl
microwave discharge (2450 MHz) in a Pyrex sidearm coated
with phosphoric acid in order to increase the Cl dissociation
yield, which could be checked by mass spectrometry at m/ e
70, 72. The decrease of the Cl signal gave a measure of
2
/ He mixture through a
2
)
2
the efficiency of the process, from 15 to 50% depending on
the experimental conditions, pressures, and fluxes. Cl atoms
concentrations were measured from the titration reaction
2
500 cm/ s giving times of reaction up to 30 ms.
For a bimolecular reaction between Cl and DMSO, the
integrated rate constant that applies to our experimental
conditions with Cl atoms in excess is
Cl + BrCH ) CH f Br + ClCH ) CH (3)
2
2
-
10
cm molecule s-1 (21)
3
-1
k ) 1.4 × 10
Ln([DMSO] / [DMSO]) ) k′t with k′) k [Cl] (1)
o
following the consumption of BrCHdCH
rising signal at m/ e ) 62 (ClCHdCH ). Reference signals of
bromoethene and chloroethene were obtained from known
flows from their respective volume-calibrated storage bal-
loons. Generally a good agreement was found between these
two measurements and the direct calculation from the
2
(m/ e ) 107) or the
The temporal profiles of DMSO were followed versus time,
as it may be seen in Figure 2, where some typical kinetic runs
using different [Cl] are shown. The experimental data follow
the model, and the pseudo-first-order kinetic rate constants
may be calculated from linear regression. Before and after
each kinetic run, additional experiments were conducted in
which the losses of reactants were evaluated separately.
DMSO was expected to be lost on the walls of the reactor due
to its low vapor pressure (0.37 Torr at 293 K). For the
experiments concerning reaction 1, with low [DMSO], no
heterogeneous losses were observed in the presence or
2
2
decrease of Cl .
ClO radicals were produced in an excess of O
a complete consumption of the Cl atoms
3
to ensure
Cl + O f ClO + O (4)
3
2
-
11
cm molecule s-1 (22)
3
-1
k ) 1.2 × 10
absence of other coreactants, like Cl
2
, see Figure 1; an upper
-
1
limit of k
W
< 1 s was estimated. In the case of Cl atoms,
and were detected at m/ e ) 51. ClO signal was calibrated in
complementary experiments by the titration reaction
they entered the injector through a fixed position but they
were exposed to different lengths of injector. The Cl atoms
were titrated versus time (injector position) with bromo-
ethene and chloroethene coming from the sliding injector
as described in the Experimental Section. From these
ClO + NO f NO + Cl (5)
2
-
11
cm molecule s-1 (22)
3
-1
k ) 1.7 × 10
-
1
calculations Cl losses in the injector walls of k ) 2-3 s were
with an excess of NO introduced through the injector and
measured, which do not disturb the pseudo-first-order model
VOL. 36, NO. 6, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 2 7

TABLE 1. Measured Second-Order Rate Constants k(1) for the Reaction Cl + DMSO f Products at Different Temperatures and
a
Pressures
P
T
0.5 Torr
298 K
(1.6 ( 0.2) × 10
1.0 Torr
273 K
(1.6 ( 0.2) × 10
1.0 Torr
298 K
(1.7 ( 0.3) × 10
1.0 Torr
335 K
(1.4 ( 0.3) × 10
3.0 Torr
298 K
(2.0 ( 0.3) × 10
k/(cm m olecule-1 s-1
3
)
-11
-11
-11
-11
-11
a
Errors are σ.
FIGURE 3. Pseudo-first-order rate coefficients, k′, versus [Cl] for
the Cl + DMSO reaction at 1.0 Torr and different temperatures: T
)
335 K(4), 298 K (O), and 273 K (9). The solid lines are the linear
least-squares fits.
FIGURE 2. Pseudo-first-order plots at different chlorine concentra-
-
3
tions obtained for reaction 1. [Cl] in units of atom cm : [Cl] ) 7.0
1
2
12
12
12
×
10 (+), 4.6 × 10 (b), 3.0 × 10 (*), 1.0 × 10 (2), and 0 (0).
The solid lines are the linear least-squares fits.
high levels of the background signals coming from the
reactants, only qualitative considerations are possible. The
sensitivity for the different products was not high enough to
afford a quantification of the yield of the possible channels
of reaction. Among the main observations it should be noticed
that no production of HCl as a function of time (with a
since Cl concentrations remained essentially constant during
the experiment. Anyway, they were taken into account for
the calculations.
To derive the rate constant of reaction 1, a correction on
the eq 1 was applied to take into account the axial diffusion
of DMSO. The experimental pseudo-first-order kinetic rate
constants, k′exp, obtained as described before, were corrected
by (23)
1
1
-3
detection limit of 2 × 10 molecule cm ) was observed,
what would suggest that the H-abstraction channel is
negligible. Two weak peaks at m/ e ) 113 and 115, assigned
3
5
37
to (CH SOCH3 Cl) and (CH SOCH3 Cl), with increasing
3
3
intensities versus time could be seen. Peaks at m/ e ) 98, 100,
2
35
37
5
0, 52 which could be assigned to CH
3
SO Cl, CH
3
SO Cl,
k′ ) k′exp (1+ k′1exp × DDMSO/ He/ v ) (2)
3
5
37
CH
3
Cl, and CH
3
Cl also showed temporal dependent
profiles.
where v is the linear velocity of the gas mixture in the reactor
-
1
(
(
(
cm s ) and DDMSO/ He is the effective diffusion coefficient
cm s-1) calculated from the volumes of atomic diffusion
24). As DDMSO/ He is pressure and temperature dependent it
Reaction DMSO + ClO. No previous data were available
in the literature concerning reaction 2. In this work it has
been found to be very slow and so it required a significant
increase of reactants to enable the study. However, the
efficiency of the Cl discharge decreased with increasing
2
concentrations and so [ClO] could not be enhanced high
enough to obtain a temporal decay of DMSO. The conditions
2
was calculated for each experimental conditions. At room
2
-1
temperature and 1 Torr pressure DDMSO/ He ) 270 cm s . In
most of the experiments the corrections on k′ due to the
axial diffusion was around 2%. The second-order rate
constants, k, were then calculated by plotting the pseudo-
first-order constant versus the corresponding Cl concentra-
tion (eq 1) and applying weighted least-squares fittings. A
linear dependence on [Cl] with zero intercept was always
found. Typical results obtained for different temperatures
and pressures are shown in Figure 3, and the results are
summarized in Table 1.
1
3
were then inverted, with DMSO concentrations (0.2 × 10 -4
1
4
-3
× 10 molecule cm ) in excess over ClO which was
monitored at m/ e ) 51, 53. Under these experimental
conditions, the loss rate of DMSO in the wall of the reactor
-
1
was typically 2-3 s , very close to those obtained for the
ClO losses on the injector wall. The kinetic data were analyzed
like in the case of reaction 1. However the results were not
so trustworthy due to the high uncertainties in the measure-
ments of DMSO concentrations, see Figure 4. Although the
pseudo-first-order rate constants clearly depend on the ClO
concentration, significant contributions from secondary
chemistry could be masked within these high error bars and
In the range of temperatures studied (273-335 K), no
significant changes were observed in the rate constants within
the experimental error, as it may be concluded from Table
1
. In all the cases those parameters possibly affected by
temperature, such as wall losses or diffusion coefficients were
taken into account. Concerning the pressure dependence of
the reaction it was checked from 0.5 to 3 Torr at room
temperature. The results show that reaction 1 is not sensitive
to pressure changes within this limited range.
-
14
so our measured rate constant for reaction 2, k(2) e 6 × 10
3
-1 -1
cm molecule at room temperature and 1.0 Torr, should
s
be considered only as an upper limit value. k(2) measured
in this work is of the same order of magnitude than the
corresponding rate constant obtained for the reaction of ClO
Further experiments were also conducted to identify the
products and intermediates of reaction 1 increasing the
concentrations of the reactants. In these experiments,
-
15
cm3 molecule
-1 -1
with DMS (k ) 9.5 ( 2.0 × 10
s ) (22).
Up to now, no more data concerning ClO reactions toward
sulfur compounds are available in the literature, and so
quantitative trends cannot be stated.
1
2
14
-3
[
[
DMSO] ranged from 1 × 10 to 3 × 10 molecule cm and
Cl] from 1 × 10 to 9 × 10 atom cm . Due to the relatively
1
2
12
-3
1
2 2 8
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 6, 2002

TABLE 2. Tropospheric Chemical Lifetimes, τ, Estimated from Average Remote Concentrations and Gas-Phase Reactions
concn/molecule cm-
3
kDMS/cm molecule s-1
3
-1
τ(DMS)
kDMSO/cm molecule s-1
3
-1
τ(DMSO)
5
6
1
3
6
a
b
-12 f
-11 h
OH
NO3
O3
Cl
ClO
9.7 × 10
7.4 × 10
7.4 × 10
5.0 × 10
1.0 × 10
6.1 × 10
2 days
2 days
>15 days
14 days
3.3 years
8.7 × 10
3.3 h
3 days
>156 days
136 days
g6 m onths
-12 f
-13 i
1.0 × 10
5.0 × 10
1 c
d
-18 f
-19 i
<1 × 10
<1 × 10
-10 g
-11 j
3.3 × 10
1.7 × 10
e
-5 f
-14 j
9.5 × 10
e 6 × 10
a
From ref 29. b From ref 30. c From ref 31. d From refs 19 and 20. From ref 32. f From ref 22. g From ref 26. h From ref 15. i From ref 16.
e
j
Measured in this work.
From the mass spectroscopic peaks described in the
results section, an addition mechanism involving an adduct
formation is proposed. The adduct, supposed to be relatively
stable since it maybe followed under the time reaction scale,
may decompose into CH
by the fact that CH SO could not be identified, which suggests
that the formation of CH Cl occurs after the adduct decom-
position by subsequent reaction of CH with the Cl in excess
not completely dissociated in the microwave discharge).
The high lifetime of the Cl-DMSO adduct is not surprising
if we take into account the calculations of McKee (27), who
predicted that a relatively stable adduct of the type CH S-
and a binding energy of -7.4 kcal mol at 0 K,
3 3
SOCl and CH . This is confirmed
3
3
3
2
(
3
-
1
(
Cl)OCH
3
could be formed as an intermediate in reaction 1. In a related
case, like the Cl-DMS reaction (26), a long-lived Cl-DMS
adduct was also found which did not decompose within the
time scale of the experiments (several milliseconds). Our
observations are also consistent with other previous results
concerning DMSO-OH reactions. While DMS shows both
abstraction and reversible adduct formation, the reactions
of DMSO follow predominantly an addition mechanism (9,
10, 15). Furthermore, the fragments found in this work by
mass spectrometry confirm the results of Urbanski et al. (15)
FIGURE 4. Pseudo-first-order rate coefficients, k’, versus [DMSO]
for the ClO + DMSO reaction at 1.0 Torr and room temperature. The
horizontal bars represent the errors (σ) in the measurements of
[DMSO].
Discussion
In this article we report the first absolute study of the kinetics
and mechanism of reaction 1 under low-pressure conditions.
Previously to this work only a reference (16) reporting the
results obtained in chamber experiments at atmospheric
3
who showed a yield of 100% in CH and MSIA from the
pressure and room temperature in the presence of O
2
was
cleavage of a C-S bond in the DMSO-OH adduct. Our results
are also supported by bond dissociation energy (BDE)
calculations. The source of the difference in Cl reaction
mechanism between DMS and DMSO probably lies in the
sulfur-carbon BDE. As C-S BDE is estimated to be (308.4 (
5.0 kJ mol ) in DMS and (236.0 ( 16.7 kJ mol ) in DMSO,
the bond is more labile in the second case and the adduct
may undergo direct dissociation. Nevertheless, a small
branching ratio (<0.1) in the abstraction channel should not
be ruled out since the background at m/ e ) 36, 38 was
relatively high and so, if the production of HCl is small it
would not contribute significantly enough at its parent peaks
as to be detected.
available. In that study the kinetic rate constant, k298K ) (7.4
-
11
3
-1 -1
(
1.0) × 10
formation of SO
8 ( 2)%, respectively was found. DMSO
cm molecule
s
, was obtained, and the
with yields of (28 ( 12)% and
was argued to
2
and DMSO
2
(
2
-
1
-1
proceed not from a direct reaction of DMSO + Cl atoms but
from DMSO reactions with long-lived intermediates or
secondary chemistry. Only a minor part of the products was
identified, and the large time scale of their experiments made
it possible that secondary chemistry was responsible for
additional losses of DMSO that would increase the apparent
rate constant. This could explain the disagreement with the
results obtained in this work, k(1) ) (1.7 ( 0.3) × 10 cm3
-11
-
1
-1
molecule
s
, where the fast discharge flow technique was
An estimation of the gas-phase lifetime of DMSO may be
obtained from the reaction rates at room temperature with
the tropospheric agents such as OH, NO , O , Cl, ClO, taking
allowed to obtain direct information from the elementary
process.
3
3
From the data obtained in this study, reaction 1 may be
considered as not affected by a change of temperature, with
a near zero energy activation, which is consistent with the
data available for other reactions of Cl atoms with sulfur
compounds
into account their average concentrations in remote areas
and the measured room temperature rate constants, Table
2. Considering that DMSO is produced from the OH-DMS
reaction mainly in marine areas, where DMS is emitted by
the phytoplankton, and that heterogeneous reactions of
nitrogen oxides on sea salt particles can lead to photolytic
precursors of Cl atoms, the local Cl concentrations may be
as high as 10 -10 atom cm (19, 28). So, in these marine
areas, the calculated DMSO lifetime from the reaction with
Cl atoms may be as low as 6.8 days, comparable to that from
Cl + H S f HCl + SH
2
4
5
-3
-
11
k ) 3.7 × 10
exp[(210 (
00)/ T] cm molecule s-1 (25)
3
-1
1
3
NO (3 days). Nevertheless, even under these local conditions
Cl + CH SH f CH S + HCl
3
3
OH constitutes the main DMSO sink. Using an average Cl
-
10
k ) 1.2 × 10
exp[(150 (
3
concentration of 5.0 × 10 , Table 2, the DMSO lifetime lies
0)/ T] cm molecule s-1 (25)
3
-1
5
in the order of 136 days. From our results (Table 1) the
Troposphere’s temperature profile would not affect this sink
of sulfur species. So, although Cl atoms scavenge DMS
efficiently, Table 2, giving lifetimes of few days, comparable
For Cl addition processes, like the system Cl-DMS at
high pressures, a very slight temperature dependence was
also found by Stickel (26) in the range 240-421 K.
3
to those of OH, NO , and ozone, the gas-phase chemistry of
VOL. 36, NO. 6, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1 2 2 9

DMSO is dominated by the OH initiated processes and, in
a less extent, by reaction with NO radical. The effect of
(10) Patroescu, I.; Barnes, I.; Becker, K.; Mihalopoulos, N. Atmos.
Environ. 1999, 33, 25.
3
(
11) Ingham, T.; Bauer, D.; Sander, R.; Crutzen, P. J.; Crowley, J. N.
J. Phys. Chem. A 1999, 103, 7199.
reaction 1 in the removal of DMSO should be considered in
mechanistic models of sulfur atmospheric chemistry only
under high Cl concentrations and as a minor homogeneous
sink. Lifetimes of DMS and DMSO in the order of several
months are calculated due to ClO reactions, Table 2, showing
that reaction 2 has a negligible effect on the removal of DMSO.
Finally, for a complete evaluation of DMSO processes,
the heterogeneous losses must be considered and compared
to homogeneous reactions. Up to now this is not possible
due to the disagreement concerning the uptake rate of DMSO
onto particles (12, 13). So, further investigations in this sense
are required to quantify the relative importance of homo-
geneous and heterogeneous processes.
(
12) Jefferson, A.; Tanner, D.; Eisele, F.; Davis, D.; Cheng, G.; Crawford,
J.; Huey, J.; Torres, A.; Berresheim, H. J. Geophys. Res. 1998, 103,
1
647.
(13) Putaud, J.; Davidson, D.; Watts, S.; Mihalopoulos, N.; Nguyen,
B.; Hewitt, C.; Atmos. Environ. 1999, 33, 647.
(
(
14) Hynes, A.; Wine, P. J. Atmos. Chem. 1996, 24, 23.
15) Urbanski, S.; Stickel, R.; Wine, P. J. Phys. Chem. A 1998, 102,
1
0522.
(
(
16) Falbe-Hansen, H.; Sørensen, S.; Jensen, N.; Pedersen, T.; Hjorth,
J. Atmos. Environ. 2000, 34, 1543.
17) Composition chemistry and climate of the Atmosphere; Singh,
H. B., Ed.; Van Ostrand Reinhold: New York, 1995; p 216; ISBN
0-442-01264-0.
(
(
18) Behnke, W.; Zetzsch, C. J. Aerosol Sci. 1990, 21, Suppl. 1, 5229.
19) Pszenny, A.; Jacob, W.; Fan, D.; Maben, J.; Zetwo, M.; Springer-
Young, M.; Galloway, J. Geophys. Res. Lett. 1993, 20, 699.
20) Wingenter, O.; Kubo, M.; Lake, N.; Smith, T.; Rowland, F. J.
Geophys. Res. 1996, 101, 4331.
Acknowledgments
This research was supported by the European Commission
within the program “Energy, Environment and Sustainable
Development”, contract EVK2-CT-1999-00033, and the Span-
ish Ministry of Science and Technology contract PB97-0432.
We also thank Mr. Manuel Palencia and Mr. Francisco Maigler
for their technical help in mounting and repairing the
experimental system.
(
(21) Park, J. Y.; Slagle, I. R.; Gutman, D. J. Phys. Chem. 1983, 87, 1812.
(22) De More, W. B.; Sander, S. P.; Golden, D. M.; Hampson, R. F.;
Kurylo; M. J. Howard, C. J.; Ravishankara, A. R.; Molina, M. J.;
Kolb, C. E. Chemical Kinetics and Photochemical Data for Use
in Stratospheric Modeling; NASA, JPL, California Institute of
Technology: Pasadena, CA, 1997.
(
(
23) Kaufman, F. J. Phys. Chem. 1984, 88, 1409.
24) Gilliland, E. R. Ind. Eng. Chem. 1934, 26, 681.
Literature Cited
(
(
(
1) ) Dacey, J.; Wakeham, S. Science 1986, 233, 1314.
2) Andreae, M. O. Marine Chem. 1990, 30, 1-29.
(25) Nicovich, J.; Wang, S.; Wine, P. Int. J. Chem. Kinet. 1995, 27, 359.
(26) Stickel, R.; Nicovich, J.; Wang, S.; Zhao, Z.; Wine, P. J. Phys.
Chem. 1992, 96, 9875.
3) Charlson, R.; Lovelock, J. E.; Andrae, M. O.; Warren, S. G. Nature
1
987, 326, 655.
(
(
(
27) McKee, M. L. J. Phys. Chem. 1993, 97, 10971.
(
(
4) S-centered radicals; Alfassi, Z. B., Ed.; Wiley and Sons: 1999.
5) James, J. D.; Harrison, R. M.; Savage, N. H.; Allen, A. G.; Grenfell,
J. L.; Allan, B. J.; Plane, J. M. C.; Hewitt, C. N.; Davison, B.;
Robertson, L. J. Geophys. Res. Atmos. 2000, 105 (D21), 26379.
6) Berresheim, H.; Huey, J. W.; Thorn, R. P.; Eisele, F.; Tanner, D.
J.; Jefferson, A. J. J. Geophys. Res. 1998, 103, 1629.
7) Sciare, J.; Baboukas, E.; Hancy, R.; Mihalopoulos, N.; Nguyen,
B. J. Atmos. Chem. 1998, 30, 229.
8) Barnes, I.; Becker, K.; Patroescu, I. Atmos. Environ. 1996, 30,
28) Vogt, R.; Crutzen, P. J.; Sander, R. Nature 1996, 383, 327.
29) Prinn, R.; Weiss, R.; Miller, B.; Huang, J.; Alyea, F.; Cunnold, D.;
Fraser, P.; Hartley, D.; Simmons, D. Science 1995, 269, 187.
30) Noxon, J. J. Geophys. Res. 1983, 88, 11017.
31) Logan, J. J. Geophys Res. 1985, 90, 10463.
32) Singh, H.; Kasting J. J. Atmos. Chem. 1988, 7, 261.
(
(
(
(
(
(
(
Received for review June 13, 2001. Revised manuscript re-
ceived November 26, 2001. Accepted December 17, 2001.
1
805.
9) Sørensen, F.; Falbe-Hansen, H.; Mangoni, M.; Hjorth, J.; Jensen,
N. J. Atmos. Chem. 1996, 24, 299.
ES010169S
1
2 3 0
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 6, 2002