Kinetics and mechanism of the oxidation of dimethyl sulfide by hydroperoxides in aqueous medium: Study on the potential contribution of liquid-phase oxidation of dimethyl sulfide in the atmosphere

Article abstract of DOI:10.1039/a700722a

Conventional and multi-wavelength stopped-flow spectrophotometry has been used to study the kinetics of the oxidation of dimethyl sulfide (DMS) by hydroperoxides, ROOH = hydrogen peroxide (H₂O₂), peroxo formic acid (HCO₃H), peroxo acetic acid (CH₃CO₃H), peroxo nitrous acid (ONOOH), peroxo monosulfuric acid anion (PMS = HSO₅^-), and the H₂O₂ analogue hypochlorous acid (HOCl), in aqueous solution in the pH range 0-14 at 293 K and I = 1.0 _M. The reaction between DMS and ROOH and between dimethyl sulfoxide (DMSO) and ROOH is a second-order process, leading to DMSO and dimethyl sulfone (DMSO₂), respectively. It was shown by gas chromatography that, except for the oxidant ONOOH, DMSO and DMSO₂ are the only oxidation products. It follows from the pH dependence of the second-order rate constant k₂ that both the hydroperoxide ROOH, rate constant k_ROOH, and its anion ROO^-, rate constant k_ROO, oxidize DMS and DMSO, respectively. The data for k_ROOH (dm³ mol^-1 s^-1) and for k_ROO (dm³ mol^-1 s^-1) at 293 K for the formation of DMSO and DMSO₂ are presented. For the oxidation of DMS k_ROOH > k_ROO; for the step DMS → DMSO, k_ROOH ranges from 4780 (PMS) to 0.018 (H₂O₂), whereas k_ROO lies in the range 88 (PMS) to 0.0018 (H₂O₂); for the step DMSO → DMSO₂, k_ROOH ranges from 349 (HOCl) to 2.7 × 10^-6 (H₂O₂), whereas k_ROO lies in the range 18 (PMS) to 8.4 × 10^-5 (H₂O₂). A mechanistic interpretation of the oxidation reactions, based on the ambifunctional character of both ROOH and DMSO, is presented. The relevance of in-cloud oxidation of DMS by atmospheric hydroperoxides such as CH₃CO₃H and HCO₃H is discussed and substantiated.

Full text of DOI:10.1039/a700722a

View Article Online / Journal Homepage / Table of Contents for this issue
Kinetics and mechanism of the oxidation of dimethyl sulÐde by
hydroperoxides in aqueous medium
Study on the potential contribution of liquid-phase oxidation of dimethyl sulÐde
in the atmosphere
Peter Amels, Horst Elias* and Klaus-Jurgen Wannowius
Ž
Anorganische Chemie III, Institut f ur Anorganische Chemie, T echnische Hochschule Darmstadt,
Petersenstrasse 18, D-64287 Darmstadt, Germany
Conventional and multi-wavelength stopped-Ñow spectrophotometry has been used to study the kinetics of the oxidation of
dimethyl sulÐde (DMS) by hydroperoxides, ROOH \ hydrogen peroxide (H O ), peroxo formic acid (HCO H), peroxo acetic
2
2
3
acid (CH CO H), peroxo nitrous acid (ONOOH), peroxo monosulfuric acid anion (PMS \ HSO ~), and the H O analogue
3
3
5
2 2
hypochlorous acid (HOCl), in aqueous solution in the pH range 0È14 at 293 K and I \ 1.0 M. The reaction between DMS and
ROOH and between dimethyl sulfoxide (DMSO) and ROOH is a second-order process, leading to DMSO and dimethyl sulfone
(DMSO ), respectively. It was shown by gas chromatography that, except for the oxidant ONOOH, DMSO and DMSO are the
2
2
only oxidation products. It follows from the pH dependence of the second-order rate constant k that both the hydroperoxide
2
ROOH, rate constant k
, and its anion ROO~, rate constant k
, oxidize DMS and DMSO, respectively. The data for
ROOH
ROO
k
(dm3 mol~1 s~1) and for k
(dm3 mol~1 s~1) at 293 K for the formation of DMSO and DMSO are presented. For the
ROOH
oxidation of DMS k
ROO
2
[ k
; for the step DMS ] DMSO, k
ranges from 4780 (PMS) to 0.018 (H O ), whereas k
lies
ROOH
ROO
ROOH
2
2
ROO
in the range 88 (PMS) to 0.0018 (H O ); for the step DMSO ] DMSO , k
whereas k
ranges from 349 (HOCl) to 2.7 ] 10~6 (H O ),
2
2
2
ROOH
2 2
lies in the range 18 (PMS) to 8.4 ] 10~5 (H O ). A mechanistic interpretation of the oxidation reactions, based on
ROO
2 2
the ambifunctional character of both ROOH and DMSO, is presented. The relevance of in-cloud oxidation of DMS by atmo-
spheric hydroperoxides such as CH CO H and HCO H is discussed and substantiated.
3
3
3
DMS, emitted steadily by oceanic phytoplankton at a rate of
ca. 50 ] 1012 g of sulfur per year, is the major natural source
of sulfur in the troposphere. Much e†ort has been focused on
the understanding of the atmospheric oxidation of DMS,1 but
many uncertainties still remain. They are primarily due to a
lack of understanding of the mechanism of the atmospheric
more than 200-fold greater reactivity of the SO 2~ anion
5
compared with that of the PMS anion HSO ~.
5
As pointed out, DMSO is the Ðrst product of the aqueous-
phase oxidation of DMS by hydroperoxides, followed by
DMSO as the further product of oxidation. This fact and the
2
occurrence of DMSO in rain8h10 and snow11 and of DMSO
2
gas-phase oxidation, as initiated by OH and NO radicals.2
in rain12 prompted us to study the kinetics of the aqueous-
phase oxidation of DMS and DMSO according to reactions
(1) and (2), respectively, with a variety of hydroperoxides
ROOH, including H O , HCO H, CH CO H, ONOOH,
3
The extent of branching to the likely major end-products
sulfur dioxide, sulfur trioxide and methanesulfonic acid (MSA)
is unclear and so is, consequently, the contribution of atmo-
spheric oxidation of DMS to the formation of acid rain and
the e†ect of DMS on the climate.
It is well known3 that, in solution, the oxidation of DMS by
hydroperoxides leads to DMSO which is, more slowly than
2
2
3
3
3
PMS and the H O analogue HOCl.¤ The study was carried
2
2
out in order to provide kinetic data which allow us to
compare the aqueous-phase reactivity towards DMS of the
natural atmospheric oxidants, O , H O and O , with the
reactivity of peroxo compounds, the atmospheric existence of
which has been proven or suggested.
2
2
2
3
DMS, further oxidized to DMSO [see reactions (1) and (2)].
2
CH wSwCH ] RwOOH ] CH wSOwCH ] ROH
3
3
3
3
(1)
Experimental
CH wSOwCH ] RwOOH ] CH wSO wCH ] ROH
Chemicals and solutions
3
3
3
2
3
(2)
DMS and DMSO (reagent grade, Merck); H O (85%,
2
2
Peroxid-Chemie); CH CO H (15% in acetic acid, Peroxid-
The kinetics of the DMS oxidation by H O in aqueous solu-
3
3
2
2
Chemie); “OxoneÏ (1.88 KHSO , 0.93 KHSO , 1.16 K SO ;
tion according to reaction (1) (R \ H) were studied by
Adewuyi and Carmichael in 0.05È0.2 M H SO and at pH 2, 6
5
4
2
4
Aldrich); NaOCl (alkaline solution, Aldrich); chloroacetic acid
(Aldrich); TRIS (99%, Biomol) were used. The standard
chemicals were of reagent grade quality and doubly distilled
water was used.
2
4
and 7.4 They found that, (i) the rate of reaction (1) is Ðrst
order in both DMS and H O and, (ii) at pH \ 1, the rate of
2
2
DMSO formation is acid catalysed. Brimblecombe et al.5
reported on the catalytic e†ect of sea salt on the rate of DMS
oxidation by H O .
The concentration of stock solutions of DMS and DMSO
in water was controlled spectrophotometrically (e \ 1270
215
dm3 mol~1 cm~1 for DMS; e \ 530 dm3 mol~1 cm~1 for
215
2
2
Pandurengan and Maruthamuthu6 studied the kinetics of
the oxidation of DMSO by PMS according to reaction (2)
(RwOOH \ ~O SwOOH) in aqueous H SO and at pH 4.6
DMSO). H O was determined by titration with CeIV.
2
2
3
2
4
and 7.0. Betterton7 investigated the kinetics of the oxidation
of DMS (and diethyl sulÐde) by PMS in aqueous solution as a
function of pH, temperature and ionic strength. He reported a
¤ The chlorine atom is, in a broader sense, isoelectronic with the
OH radical. HOOH and HOCl are thus electronically equivalent
compounds.
J. Chem. Soc., Faraday T rans., 1997, 93(15), 2537È2544
2537

View Article Online
bu†er (4È5.5), phosphate bu†er (6È7.5), TRIS bu†er (7.5È9),
carbonate bu†er (9È11) and NaOH (11È14).
Gas chromatography
The analysis of the reaction products DMSO and DMSO
2
was carried out as described by Betterton.7 Before injection,
the samples were de-ionized by passing them through ion-
exchange cartridges (Merck, Adsorbex SAX/SCX).
Kinetic measurements
Fig. 1 Dependence of k \ f ([H O ] ) in the system DMSÈH O
Slow reactions were monitored by conventional spectropho-
tometry, faster ones by stopped-Ñow spectrophotometry or
exp
2
2 0
2 2
at pH 4.4 and computer-Ðt of the data to eqn. (V). [DMS] \ 2.75
0
] 10~4 M, I \ 0.3 M (Na SO ), T \ 293 K.
2
4
multi-wavelength
stopped-Ñow
spectrophotometry,
as
described earlier.16 Monitoring was performed at suitable
Peroxo nitrous acid was prepared by the reaction of
NaNO with H O in perchloric acid according to Keith and
wavelengths in the range 205È360 nm. The absorbanceÈtime
2
2 2
data were computer-Ðtted to eqn. (II) or (III) to obtain k and
Powell.13 The alkaline solution of sodium peroxo nitrite was
2
k
, respectively.
stored at [28 ¡C. Its concentration was determined spectro-
exp
photometrically (e \ 1670 dm3 mol~1 cm~1).14
302
A \ (A [ A )/(1 ] [substrate] k t) ] A
(II)
Peroxo formic acid was obtained by equilibrating formic
0
=
0 2
=
=
acid (96%) with H O (85%) according to reaction (3). The
A \ (A [ A )exp([k t) ] A
(III)
2
2
0
=
exp
concentration of HCO H and of CH CO H was determined
3
3
3
In several experiments the ratio z \ [ROOH]/[substrate] (or
z \ [substrate]/[ROOH]) lay in the range 1.5È10. In these
cases eqn. (IV) was applied with b \ k [substrate] (1 [ z~1)
H O ] RwCO H H H O ] RwCO H
(3)
2
2
2
2
3
potentiometrically.” The solutions of “oxoneÏ are not very
stable and had to be freshly prepared before each kinetic
experiment. To avoid metal-catalysed decomposition, all the
ROOH solutions were set to [EDTA] B 5 ] 10~5 M. The
concentration of the solution of NaOCl was controlled
2
0
Mor b \ k [ROOH] (1 [ z~1)N.
2
0
A \ (A [ A )(z [ 1)exp([bt)/[z [ exp([bt)] ] A (IV)
0
=
=
Eqn. (V) and (VI) were used to determine k and k (k \ rate
2
d
d
spectrophotometrically (e \ 345 dm3 mol~1 cm~1).15
constant for decomposition of ROOH).
292
k
k
\ k [excess partner]
(V)
exp
exp
2
pH measurements and bu†ers
\ k [excess partner] ] k
(VI)
2
d
The pH data were measured with a glass electrode and
relationship (I) was used to calculate the proton concentration
(x \ ] 0.1 at I \ 1.0 M and 293 K).
In the system DMSÈPMS at pH [ 8, the absorbanceÈtime
data were biphasic and eqn. (VII) was applied to obtain k
exp1
and k
and the amplitudes a and a .
exp2
A \ a exp([k
1
2
[ log[H`] \ pH ] x
(I)
t) ] a exp([k
t) ] A
(VII)
The adjustment of the pH was achieved with HClO (pH 0È
1.5), chloroacetate bu†er (1.5È3), formate bu†er (3È4), acetate
1
exp1
2
exp2
=
4
Results
It was found for the various systems ROOHÈDMS and
ROOHÈDMSO that substrate oxidation follows the rate law
(VIII).
rate \ k [substrate][ROOH]
(VIII)
2
0
Since hydroperoxides ROOH can lose or add a proton
according to reactions (4) and (5), respectively, the second-
order rate constant k was found to be pH-dependent.
2
Ka
ROOH A8B ROO~ ] H`
(4)
(5)
KH
ROOH ` A8B ROOH ] H`
2
Rate law (IX) considers the three possible pathways and eqn.
(X) describes the pH dependence of k .
2
[d[substrate]/dt \ (k
[ROOH] ] k [ROO~]
ROOH
ROO
] kROOH2[ROOH `])[substrate] (IX)
K /[H`]
2
Fig. 2 pH dependence of second-order rate constant k for the oxi-
2
dation of DMS by H O at 293 K. (K) Data reported by Adewuyi
k \ (k
] k
2
2
2
ROOH
ROO a
and Carmichael in ref. 4; (ÈÈ) see text.
] kROOH2[H`]/K )/(1 ] K /[H`] ] [H`]/K ) (X)
In the system HOOHÈDMS, variation of the bu†er concen-
H
a
H
” For R \ H and R \ CH , the reverse of reaction (3) is slow,
3
which allows the simultaneous determination of the acid and the
peroxo acid by potentiometric titration.
tration in the range 0.003È0.03 M and of the ionic strength in
2538
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
the range 0.03È1.0 M led, within the limits of error, to the same
value for k .
2
Oxidation of DMS and DMSO by hydrogen peroxide
In the system DMSÈH O , k
increases linearly with
2
2
exp
[H O ] (see Fig. 1), which leads to the second-order rate
2
2 0
constant k and conÐrms rate law (VIII). Fig. 2 shows the pH
2
dependence of k , with a plateau at pH 2È9, a decrease at
2
pH [ 9 and an increase at pH \ 1.5 (The k data reported by
2
Adewuyi and Carmichael4 are also shown in Fig. 2; for pH 2
and 6, they are higher by a factor of ca. 2). Fitting of eqn. (X)
to the k data shown in Fig. 2 led to k
\ (1.8 ^ 0.1)
Fig. 3 Dependence of
k
\ f ([DMSO] ) in the system
2
HOOH
exp
0
] 10~2 dm3 mol~1 s~1, k
\ (1.8 ^ 0.2) ] 10~3 dm3
DMSOÈH O at pH 13 and computer-Ðt of the data to eqn. (VI).
HOO
2
2
mol~1 s~1, pK \ 11.0 ^ 0.1 (literature:17 11.65 ^ 0.03 at 298
[H O ] \ 3.0 ] 10~3 M, I \ 1.0 M (NaClO ), T \ 293 K.
a
2
2 0
4
K and I \ 0), and kHOOH2/K \ 0.21 ^ 0.03 dm6 mol~2 s~1.
H
The species HOOH ` has been postulated and discussed
2
by several authors as an exceptionally powerful oxidant for
the oxidation of hydrocarbons18 and organic sulÐdes.19,20
There are uncertainties, however, concerning the strength of
the acid HOOH `. An early estimate for the pK of
2
a
HOOH `, as given by Evans and Uri,21 was pK \ [3. In
2
H
1964, Alder and Whiting18 presented data for pK which are
H
based on the estimate that H O is a factor of 106 less basic
2
2
than water. These data range from [7.7 for water to [11 for
sulfuric acid.18 The pH proÐle for the oxidation of DMS by
H O does not level o† at very low pH (see Fig. 2), which
2
2
means that, unfortunately, the Ðtting procedure can only
provide the composite parameter
mol~2 s~1 (Adewuyi and Carmichael:4 kHOOH2/K \ 0.21 dm6
k
HOOH2/K \ 0.21 dm6
H
H
mol~2 s~1). This number allows the estimate that, even for the
largest pK value of HOOH ` reported (pK \ [321), rate
H
2
H
constant kHOOH2 would be as high as 210 dm3 mol~1 s~1. This
would mean that, with regard to the reactivity towards DMS,
the species HOOH ` is at least a factor of ca. 104 and 105
2
more reactive than H O and its anion HOO~, respectively.
2
2
Alternatively, one might assume that the reaction of the
species HOOH ` is di†usion-controlled (kHOOH2 B 1010 dm3
2
mol~1 s~1). This would make HOOH ` an extremely strong
Fig. 4 pH dependence of second-order rate constant k for the oxi-
2
2
acid with a pK of ca. [11.
dation of DMSO by di†erent hydroperoxides ROOH at 293 K. (ÈÈ)
a
The reaction of DMSO with H O is very slow, with a half
and (È È È) Computer-Ðt of the data to eqn. (XI); (Ç) data of the
present study obtained for the oxidation of DMSO by PMS; (|) data
reported in ref. 6 for the oxidation of DMSO by PMS; (K) data
reported in ref. 7 for the oxidation of DMS by PMS.
2
2
life of many days at 293 K. The tedious kinetic investigation
of this reaction was therefore restricted to measurements at
pH 2.0 and 13. As found for the system DMSÈH O , the rate
2
2
2
of oxidation of DMSO by an excess of H O at pH 2.0
2
follows rate law (VIII) with k \ (2.75 ^ 0.03) ] 10~6 dm3
2
mol~1 s~1 at 293 K. The data obtained at pH 13 are shown in
Fig. 3. They lead to k \ (8.7 ^ 0.4) ] 10~5 dm3 mol~1 s~1
2
and k \ 0.4 ] 10~5 s~1 for the slow decomposition of the
d
species HOO~ in alkaline solution. Rate constants k for the
2
oxidation of DMSO at pH 2.0 and 13 are shown in Fig. 4.
The dashed line connecting these two points was obtained by
computer-Ðtting them with eqn. (XI) for K \ 1.0 ] 10~11 M
a
(see system DMSÈH O ).
2
2
k \ (k
] k
K /[H`])/(1 ] K /[H`])
(XI)
2
ROOH
ROO
a
a
Despite the obvious lack of data it is interesting to see that the
peroxo anion HOO~ is much more reactive towards DMSO
than H O . This behaviour is opposite to that observed for
2
2
the oxidation of DMS (see Fig. 2).
Oxidation of DMS and DMSO by peroxo formic acid and
peroxo acetic acid
The pH proÐles of k for the oxidation of DMS and DMSO
2
by peroxo formic acid are shown in Fig. 5. As found for
HO ~, peroxo formic acid anion, HCO ~, is more reactive
Fig. 5 pH dependence of second-order rate constant k for the oxi-
2
3
2
towards DMSO than the acid, HCO H. For the oxidation of
dation of DMS and DMSO by peroxo formic acid at 293 K. (ÈÈ)
3
Computer-Ðt of the data to eqn. (XI).
DMS with peroxo formic acid, the opposite result is obtained.
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
2539

View Article Online
Fig. 6 pH dependence of second-order rate constant k for the oxi-
Fig. 8 pH dependence of second-order rate constant k for the oxi-
2
2
dation of DMS by various oxidants at 293 K. (ÈÈ) Computer-Ðt of
dation of DMS and DMSO by peroxo nitrous acid at 293 K. (ÈÈ)
the data to eqn. (XI); Ðtting of the data for HOCl: see text.
Computer-Ðt of the data to eqn. (XI).
The kinetic pattern found for the oxidation of DMS and
DMSO by CH CO H corresponds to that found for HCO H.
3
3
3
Fig. 6 and Fig. 4 show the corresponding pH dependence of
k . One learns that, compared with HCO H, CH CO H is
2
3
3
3
clearly a less reactive oxidant for both DMS and DMSO.
Oxidation of DMS and DMSO by PMS
In acidic media, the kinetics of the oxidation of DMS and
DMSO by the PMS anion correspond to those found for
H O , HCO H and CH CO H. In alkaline media, the
2
2
3
3
3
change of the absorbance with time is biphasic (Fig. 7). The
experimental rate constants k and k increase linearly
exp1
exp2
with the concentration of the excess partner PMS and the
resulting second-order rate constants k di†er by a factor of
2
ca. 5.
At pH 12.7, PMS is deprotonated and present in the form
Fig. 7 Change in absorbance at 223 nm with time in the system
DMSÈPMS at pH 12.7 and computer-Ðt of the data (ÈÈ) to eqn.
(VII). [DMS] \ 5 ] 10~4 M, [PMS] \ 2.56 ] 10~3 M, I \ 1.0 M
of the anion SO 2~ [pK (HSO ~) \ 9.4].22 The spectral
5
a
5
changes associated with the two-step oxidation of DMS by
PMS at pH 12.7 are in agreement with the interpretation that
0
0
(Na SO ), T \ 293 K.
2
4
the powerful oxidant SO 2~ Ðrst oxidizes DMS to DMSO
5
(k ) and then, in a consecutive second step (k ), DMSO
Table 1 Results of the gas chromatographic product analysis in the
system DMSÈROOHa
exp1
exp2
to DMSO . In acidic media, the consecutive DMSO oxida-
tion by the anion HSO ~ is obviously much slower. The pH
proÐles of k obtained for the DMS oxidation (see Fig. 6) and
2
5
ROOH
[DMSO] (%)b
[DMSO ] (%)b
&[S] (%)c
2
2
the DMSO oxidation (see Fig. 4) are in line with this interpre-
pH 2
tation. The two rate constants di†er by almost Ðve orders of
magnitude at low pH and by less than one order of magnitude
at high pH.
HCOwOOH
33 ^ 3
26 ^ 1
0
69 ^ 5
76 ^ 11
103 ^ 4
3 ^ 1
102 ^ 6
102 ^ 11
103 ^ 4
105 ^ 13
98 ^ 3
CH COwOOH
3
~O SwOOH
3
At this point it is important to state that, in a study carried
out recently,7 the fast oxidation of DMS by PMS was obvi-
ONwOOH
102 ^ 13
HOCl
0
98 ^ 3
ously overlooked. The second-order rate constant, k \ 0.13
pH 11
2
^ 0.09 dm3 mol~1 s~1 (298 K), reported for the oxidation of
HCOwOOH
14 ^ 1
27 ^ 2
0
94 ^ 1
81 ^ 8
90 ^ 5
64 ^ 3
84 ^ 2
108 ^ 2
108 ^ 8
90 ^ 5
DMS by the anion HSO ~,7 is incorrectly assigned. As shown
CH COwOOH
5
3
in Fig. 4, this rate constant has to be ascribed to the oxidation
~O SwOOH
3
ONwOOH
31 ^ 2
95 ^ 4
of DMSO by the anion HSO ~, the rate constant for the oxi-
5
HOCl
17 ^ 3
101 ^ 4
dation of DMS being a factor of almost 105 higher (see Fig. 6).
a Conditions: [DMS] \ 10~3 M, [ROOH] \ 6 ] 10~3 M; 293 K.
0
0
b Percentage of DMSO and DMSO , respectively, as referenced to
Oxidation of DMS and DMSO by peroxo nitrous acid
2
[DMS] 4 100%. The numbers listed represent the mean of the
0
Peroxo nitrous acid, ONwOOH, an isomer of nitric acid, de-
cays in acidic solution rather rapidly to form HNO .16,23,24
results obtained for four to six samples. c Sum of the relative fractions
of DMSO and DMSO .
2
3
2540
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
Table 2 Results of the gas chromatographic product analysis in the
system DMSOÈROOHa
ROOH
[DMSO] (%)b
[DMSO ] (%)b
&[S] (%)c
2
pH 2
HCOwOOH
0
0
94 ^ 2
92 ^ 3
94 ^ 2
92 ^ 3
CH COwOOH
3
~O SwOOH
0
35 ^ 1
0
124 ^ 2
11 ^ 1
124 ^ 2
46 ^ 2
3
ONwOOH
HOCl
100 ^ 7
100 ^ 7
pH 11
HCOwOOH
0
0
99 ^ 5
95 ^ 2
115 ^ 2
79 ^ 3
96 ^ 2
99 ^ 5
95 ^ 2
CH COwOOH
3
~O SwOOH
0
33 ^ 2
0
115 ^ 2
112 ^ 4
96 ^ 2
3
ONwOOH
HOCl
a Conditions: [DMSO] \ 10~3 M, [ROOH] \ 6 ] 10~3 M; 293 K.
0
0
b Percentage of DMSO and DMSO , respectively, as referenced to
2
[DMSO] 4 100%. The numbers listed represent the mean of the
results obtained for fourÈsix samples. c Sum of the relative fractions of
0
DMSO and DMSO .
2
Fig. 10 pH dependence of second-order rate constant k for the oxi-
2
dation of DMSO by hypochlorous acid and peroxo nitrous acid at
293 K. (ÈÈ) Computer-Ðt of the data to eqn. (XI).
In the pH range 1.5È5, the half-life of ONwOOH is ca. 1 s at
293 K. At low concentrations of DMS, the oxidation of DMS
by ONwOOH competes, therefore, with the decay of
ONwOOH. The analysis of the absorbanceÈtime data
according to eqn. (VI) leads to k and k . As shown in Fig. 8
range 14È11, where log k increases linearly with log[H`]
2
(k D [H`]). A plausible explanation for this sort of depen-
2
d
2
and 6, ONwOOH is a highly reactive oxidant for DMS. The
dence is the assumption that HOCl instead of the anion OCl~
anion ONwOO~ is considerably less reactive, the ratio
is the species oxidizing DMS according to reaction (7).
k
/k
being ca. 104 : 1.
ONOOH ONOO
As mentioned above, the anion ONwOO~ is much more
Ka
HOCl A8B H` ] OCl~
(6)
(7)
stable than ONwOOH. Fig. 9 shows a series of spectra docu-
menting the relatively slow consumption of the peroxo nitrite
anion in the course of the oxidation of DMSO at pH 12.7.
The pH proÐle for the DMSO oxidation by ONOOH is
shown in Fig. 10.
kHOCl
DMS ] HOCl ÈÈÈ DMSO ] HCl
In this case, relationship (XII) describes the pH dependence of
rate constant k and allows the determination of k
, since
2
HOCl
K (HOCl) is known.
Oxidation of DMS and DMSO by hypochlorous acid
a
k \ k
/K (HOCl)[H`]
(XII)
HOCl, electronically equivalent to HOOH,¤ was found to be
a very powerful oxidant for DMS (see Fig. 6). The fast second-
order oxidation process could be monitored only in the pH
2
HOCl
a
The second-order oxidation of DMSO by HOCl is slow
enough to be followed from pH 14 down to pH 3 (see Fig. 10).
Product analysis
The gas chromatographic determination of the products
DMSO and DMSO was carried out in the system DMSÈ
2
ROOH and DMSOÈROOH, respectively, with a sixfold
excess of ROOH at pH 2 and 11. The reaction time chosen
was long enough to guarantee complete oxidation of the sub-
strate DMS and DMSO. The numbers obtained for the rela-
tive fraction of DMSO and DMSO have rather high limits of
2
error. This may be due, in part, to the deionization procedure
to which the samples were subjected before the GC analysis.
As shown in Tables 1 and 2, for most of the oxidants the
sum of the fractions of DMSO and DMSO is, within the
2
limits of error, around 100% or at least approaching 100%.
The fact that, for a given oxidant ROOH, the relative frac-
Fig. 9 Multi-scan monitoring of the oxidation of DMSO by peroxo
nitrite anion at pH 12.7. Time interval between consecutive spectra:
*t \ 0.95 s; [DMSO] \ 1.0 M, [ONwOO~] B 1 ] 10~4 M, I \ 1.0
tions of DMSO and DMSO are di†erent at pH 2 and pH 11
2
(see Table 1), reÑects the di†erent reactivity of the species
0
0
M (NaClO ), T \ 293 K.
ROOH and its anion ROO~.
There is no obvious explanation for the Ðnding that the
4
oxidation of DMSO by PMS produces clearly more DMSO
than would be expected (see Table 2).
2
The results obtained with the unstable oxidant peroxo
nitrous acid are somewhat atypical. First, in addition to
¤ The chlorine atom is, in a broader sense, isoelectronic with the
OH radical. HOOH and HOCl are thus electronically equivalent
compounds.
DMSO and DMSO a third, yet unidentiÐed, component was
2
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
2541

View Article Online
2542
J. Chem. Soc., Faraday T rans., 1997, V ol. 93

View Article Online
found in the gas chromatograms. Secondly, at pH 2 the frac-
log k
\ apK (ROH) ] b
(XIII)
ROOH
a
tions of DMSO and DMSO add up to only 47% (see Table
2
It means that the electronic properties of the group RO a†ect
the electron density at the reacting site of the molecule ROOH
in such a way that an electron-withdrawing group RO (i.e. a
stronger “acidÏ ROH) enhances the rate of oxygen transfer and
splitting o† of the anion RO~ according to reaction (8).
2). This might indicate that, in the case of ONwOOH,
DMSO is not only oxidized to DMSO but also to some
2
other compounds, which are not being detected under the
given GC conditions. In this connection it is interesting to
note that Pryor et al.25 reported the formation of formalde-
hyde as a product of the oxidation of DMSO by peroxo
nitrous acid at pH 7.
rds
(CH ) S ] HOOR ÈÈÈ (CH ) SOH` ] RO~
3 2
3 2
fast
ÈÈÈ (CH ) SO ] ROH
(8)
Discussion
3 2
The results prove that the oxidation of DMS and DMSO by
hydroperoxides ROOH in aqueous solution is a second-order
process according to rate law (VIII).
The species DMS and DMSO are not subject to proto-
nation in aqueous solution. The observed pH dependence of
A low pK (ROH) makes the anion RO~ a good leaving
a
group.
It is found, for the oxidation of DMS, that k
\ k
ROO
ROOH
and that k
also follows the pK of ROH (see Fig. 11).
ROO
a
These Ðndings indicate that, compared with the protonated
oxygen in ROOH, the deprotonated oxygen in ROO~ is a
considerably poorer electrophile for interaction with the
nucleophilic sulfur in DMS according to reaction (8).
k
has therefore to be ascribed to the deprotonation of the
2
species ROOH according to reaction (4) and follows relation-
ship (XI). Both the hydroperoxide ROOH and its anion
ROO~ oxidize DMS and DMSO, respectively, albeit at di†er-
ent rates. The data for k
are summarized in Table 3.
From a mechanistic point of view, it is interesting to see
and k
and for pK (ROOH)
that, for ROOH \ HOOH, HCOwOOH, CH COwOOH
ROOH
ROO
a
3
and ~O SwOOH( \ PMS), the oxidation of DMSO by the
3
For the oxidation of DMS to DMSO it is found that
. In the case of HOOH, HCO H, CH CO H
and PMS, the hydroperoxide ROOH is, by one or two orders
anion ROO~ is faster than that by the neutral oxidant
k
[ k
ROOH, i.e. k
[ k
(see Table 3). In contrast to DMS,
ROOH
ROO
3
3
3
ROO
ROOH
the molecule DMSO with its polarized SwO bond
of magnitude, more reactive than the anion ROO~. For
[(CH ) S`wO~] can obviously act as an O-based nucleo-
3 2
peroxo nitrous acid the ratio k
/k
is [104. The e†ect of
phile or an S-based electrophile, depending on the nature of
ROOH ROO
the group R on the reactivity of ROOH towards DMS is very
pronounced. As demonstrated in Fig. 6, PMS, ONOOH,
HCO H and CH CO H react four to Ðve orders of magni-
the reaction partner. DMSO apparently recognizes the anions
of the above mentioned hydroperoxides, as well as the hydro-
peroxides ROOH themselves, to be so nucleophilic that it
interacts with its electrophilic sulfur site according to reaction
(9).
3
3
3
tude faster than H O . The rate constants k
and k
for
2
2
ROOH
ROO
the oxidation of DMS (see Table 3) have in common that both
k
and k are smallest for H O (high pK ) and largest
ROOH
ROO
2
2
a
O~
for PMS (low pK ).
a
=
In trying to correlate the rate constants with the acidity of
K
ROO~ ] OS(CH ) A8B ROOwS(CH )
(9)
either ROOH or ROH one Ðnds that log k
correlate satisfyingly with pK (ROH) (see Fig. 11), whereas the
correlation with pK (ROOH) is much less convincing. The
correlation with pK (ROH), as described by eqn. (XIII), is a
and log k
3 2
3 2
ROOH
ROO
a
O~
a
a
=
rds
ROOwS(CH ) ÈÈÈ RO~ ] O S(CH )
(10)
3 2
2
3 2
linear free energy relationship, which was also reported by
Edwards31 for the oxidation of bromide ions by various
hydroperoxides ROOH.
Since ROOH is less nucleophilic than ROO~, equilibrium (9)
is less favoured for ROOH than for ROO~, reaction (10) is,
most probably, slower and, hence, k
[ k
. As found for
ROO
ROOH
the system DMSÈROOH, the correlation of log k
and log
ROOH
k
for the oxidation of DMSO is better with pK (ROH)
than with pK (ROOH). This Ðnding again indicates that
breaking of the OwO bond and release of the RO~ ion is the
ROO
a
a
rate-determining step.
The results obtained for the oxidation of DMSO to
DMSO by peroxo nitrous acid and hypochlorous acid are
2
di†erent, in the sense that k
\ k
(see Fig. 10 and Table
ROO
ROOH
3). One has to conclude that, as analogously described for
DMS, the nucleophilic oxygen site of DMSO interacts with
the most electrophilic site in ONwOOH and HOCl, respec-
tively, in the initial step.
In summary, DMSO and the hydroperoxides studied are
obviously ambident or ambifunctional species. This means
that the mode of attack in the reaction between these two
species is coupled to the charge distribution in the hydro-
peroxide ROOH, as controlled by R.
Fig. 11 Linear free energy relationship between second-order rate
Atmospheric relevance
constant
k
( \ k
and
k
,
respectively) and pK (ROH).
ROOH
ROO
a
pK (ROH) refers to the pK of the acid ROH corresponding to the
a
a
There is no doubt that the oxidation of DMS in the tropo-
sphere occurs primarily in the gas phase.1 Liquid-phase oxida-
tion of DMS in cloud water has been discussed7,32 but there
hydroperoxide ROOH; (…) DMSÈROOH, (=) DMSÈROO~, (L)
DMSOÈROOH, (K) DMSOÈROO~; k for the oxidation of
ONOO
DMS (in brackets) not included in the Ðtting procedure.
J. Chem. Soc., Faraday T rans., 1997, V ol. 93
2543

View Article Online
is, to our knowledge, no experimental evidence proving the
formation of DMSO in cloud water. The presence of DMSO
in rain8h10 and snow11 has, however, been documented.
The major argument used to exclude the consideration of
liquid-phase oxidation of DMS in the lower atmosphere is its
limited solubility in water. The Henry constant for DMS is
indeed very small (H \ 0.69 mol dm~3 atm~1; 293 K). One
has to consider, however, that the solubility of the oxidation
product, DMSO, in water is inÐnitely high. It is conceivable,
therefore, that the oxidation of DMS takes place at the surface
of cloud water droplets, with the DMSO thus formed being
readily taken up by the water droplet. As discussed in detail
by Lee and Zhou32 for the in-cloud oxidation of DMS by
ozone, such a process could contribute to the overall atmo-
spheric oxidation of DMS, as long as the time constant for the
DMS transport from the gas phase to the liquid phase is
shorter or comparable to the time constant for the oxidation
of DMS in the liquid phase.
overall discussion of loss processes for DMS in the atmo-
sphere.
This paper is in honour of Prof. R. G. Wilkins. This work was
supported by the Deutsche Forschungsgemeinschaft
(Sonderforschungsbereich 233), Verband der Chemischen
Industrie e.V. and Otto-Rohm-Stiftung. The authors appre-
ciate stimulating discussions with Prof. M. O. Andreae (MPI
fur Chemie, Mainz).
References
1
2
DimethylsulÐde: Oceans, Atmosphere and Climate, ed. G. Restelli
and G. Angeletti, Kluwer, Dordrecht, 1993.
S. B. Barone, A. A. Turnipseed and A. R. Ravishankara, Faraday
Discuss., 1995, 100, 39.
A. G. Davies, Organic Peroxides, Butterworths, London, 1961.
Y. G. Adewuyi and G. R. Carmichael, Environ. Sci. T echnol.,
1986, 20, 1017.
3
4
Betterton7 calculated characteristic times for the depletion
5
6
P. Brimblecombe, S. Watts and D. Shooter, Proc. 194th ACS
Meeting, American Chemical Society, Washington 1987, p. 96.
T. Pandurengan and P. Maruthamuthu, Bull. Chem. Soc. Jpn.,
1981, 54, 3551.
of DMS in aqueous droplets by the oxidants ~O SwOOH,
3
HOOH and O at pH 6 and 298 K. His calculations were
3
based on a concentration of 2.5 lM for PMS and H O and
2
2
0.2 nM for O (20 ppbv). The lifetimes obtained were 36 days
7
8
9
E. A. Betterton, Environ. Sci. T echnol., 1992, 26, 527.
M. O. Andreae and W. R. Barnard, Anal. Chem., 1983, 55, 608.
R. P. Kiene and G. Gerard, Marine Chemistry, 1994, 47, 1.
3
for PMS, 136 days for H O and \7 h for O . As stated
above, the second-order rate constant used in these calcu-
2
2
3
10 R. G. Ridgeway Jr., D. C. Thornton and A. R. Bandy, J. Atmos.
lations for the oxidation of DMS by PMS (k \ 0.13 dm3
PMS
Chem., 1992, 14, 53.
mol~1 s~1; 298 K) was incorrectly assigned. The correct value
11 M. O. Andreae, L imnol. Oceanogr., 1980, 25, 1054.
12 A. A. P. Pszenny, R. G. Harvey, C. J. Brown, R. F. Lang, W. C.
Keene, J. N. Galloway and J. T. Merrill, Global Biogeochem.
Cycles, 1990, 4, 367.
13 W. G. Keith and R. E. Powell, J. Chem. Soc. A, 1969, 90.
14 M. N. Hughes and H. G. Nicklin, J. Chem. Soc. A, 1968, 450.
15 J. C. Morris, J. Phys. Chem., 1966, 70, 3798.
16 C. Drexler, H. Elias. B. Fecher and K. J. Wannowius, Fresenius J.
Anal. Chem., 1991, 340, 605.
17 R. M. Smith and A. E. Martell, Critical Stability Constants,
Plenum Press, New York, 1976, vol. 4.
18 R. W. Alder and M. C. Whiting, J. Chem. Soc., 1964, 4707.
19 S. D. Ross, J. Am. Chem. Soc., 1946, 68, 1484.
20 C. G. Overberger and R. W. Cummins, J. Am. Chem. Soc., 1953,
75, 1484.
21 M. G. Evans and N. Uri, T rans. Faraday Soc., 1949, 45, 224.
22 C. Marsh and J. O. Edwards, Prog. React. Kinet., 1989, 15, 35.
23 D. J. Benton and P. Moore, J. Chem. Soc. A, 1970, 3178.
24 M. N. Hughes and H. G. Nicklin, J. Chem. Soc. A, 1970, 925.
25 W. A. Pryor, R. Cucto, X. Jin, W. H. Koppenol, M. Ngu-
Schwemlein, G. L. Squadrito and R. M. Uppu, Free Radical Biol.
Med., 1995, 18, 75.
is k \ 4.78 ] 103 dm3 mol~1 s~1 (293 K; see Table 3),
PMS
which reduces the calculated lifetime of DMS for the oxida-
tion by PMS from 36 days to 1.4 min. This means that, for the
concentrations chosen, PMS is a much more efficient oxidant
even than O .
3
Lee and Zhou32 calculated, for an ozone level of 30 ppb, an
atmospheric lifetime of DMS of ca. 3 days against the in-cloud
O -DMS reaction. They pointed out that the aqueous-phase
3
3
O
reaction pathway is thus comparable to the OH pathway
for the oxidation of DMS in the gas phase. The present study
provides second-order rate constants for the liquid phase oxi-
dation of DMS at pH 6 which range from 310 dm3 mol~1 s~1
for peroxo acetic acid to 4.78 ] 103 dm3 mol~1 s~1 for PMS
(see Table 3). This means that, within the framework of the
model calculations carried out by Betterton, a PMS concen-
tration of ca. 1 nM and a concentration of peroxo acetic acid
of ca. 10 nM would be sufficient to limit the atmospheric life-
time of DMS to ca. 3 days. The cloud-water chemistry model
presented by Jacob33 predicts that the PMS anion may occur
in remote marine and continental clouds at levels approaching
2.5 lM and Kleinman34 calculated a gas-phase concentration
of ca. 1 ppbv of peroxo acetic acid, which corresponds to a
cloud-water concentration of ca. 0.5 lM (H \ 468 mol dm~3
atm~1 for CH CO H; 298 K).
26 A. F. Ghiron and R. C. Thompson, Inorg. Chem., 1989, 28, 3647.
27 H. Elias, B. Fecher and K. J. Wannowius, to be published.
28 A. J. Everett and G. J. Minko†, T rans. Faraday Soc., 1953, 49,
410.
29 H. Elias, U. Gotz and K. J. Wannowius, Atmos. Environ., 1994,
28, 439.
3
3
30 A. Skrabal, Z. Elektrochem. Angew. Phys. Chem., 1942, 48, 314.
31 J. O. Edwards, in Peroxide Reaction Mechanisms, ed. J. O.
Edwards, Wiley-Interscience, New York, 1962, pp. 67È106.
32 Y-N. Lee and X. Zhou, J. Geophys. Res., 1994, 99, 3597.
33 D. J. Jacob, J. Geophys. Res. D, 1986, 91, 9807.
34 L. I. Kleinman, J. Geophys. Res., 1986, 91, 10 889.
35 T. J. Kelly, P. H. Daum and S. E. Schwartz, J. Geophys. Res.,
1985, 90, 7861.
Kelly et al.35 measured the concentration of peroxides in
cloud water and rain and reported concentrations of total per-
oxide of 1È70 lM. They found, for some samples, that, after
selective destruction of H O , there was substantial signal,
2
2
possibly indicating the presence of organic peroxides.
In conclusion, the liquid-phase oxidation of DMS by
hydroperoxides such as peroxo formic acid, peroxo acetic
acid, peroxo monosulfuric acid anion and peroxo nitrous acid
represents a reaction channel which is not to be ignored in the
Paper 7/00722A; Received 31st January, 1997
2544
J. Chem. Soc., Faraday T rans., 1997, V ol. 93