A R T I C L E S
Korang et al.
oxygen and two sulfides.1,2 However, a unimolecular mechanism
was suggested in more recent studies.5–9 Due to the large
hydrophobic regions of sulfoxides 5 and 11, increased aggrega-
tion might be expected to occur in water. Though the concentra-
tion range was limited, the UV-absorption spectra of 5 and 11
showed no concentration dependence, which suggests that
association before absorption of the light was limited.
free energy associated with electron transfer from HO- to the
excited state of 11 (∆Get) was approximated as -4.1 kcal/mol
exergonic.
∆Get ) 23.06([E°(HO•/HO-) - E°(11/11•-)] - Wd - ∆E00)
(1)
The pKa of (CH3)2SOH• was reported as approximately 17,
which suggests 20 would be protonated at the examined pH
values to form the hydroxysulfuranyl radical 21.32,33 While
persistent sulfuranyl radicals have been reported, hydroxysul-
furanyl radicals are generally not stable, and their fate in aqueous
solutions has consistently been reported as the heterolytic
cleavage to form R2S•+ and HO-.32–34 Thus, after protonation
of 20, the subsequently formed 21 would be expected to
decompose to the radical cation 22 and HO-.
Most of the results in this work argue against the role of a
dimer in the photoinduced deoxygenation of the sulfoxides 5
and 11. The starting sulfoxides and all of the observed
photoproducts appeared to be inert to oxidation by singlet
oxygen (1O2) generated by using methylene blue as a photo-
sensitizer. Also, no evidence for 1O2 was observed when benzyl
phenyl sulfide was used as a trap during the photolysis of 11.
The oxidation of the benzoic acid 18 was found to be inert to
1O2.31 Thus, none of the results observed during this study
suggested the involvement of singlet oxygen (1O2) as a potential
oxidant during the photolysis of 5 and 11 in anaerobic or aerobic
conditions. Thus, degradation of a sulfoxide exciplex leading
to 1O2 and two sulfides was largely ruled out as a major process.
Potentially, degradation of the sulfoxide exciplex could lead
to ground-state molecular oxygen and two sulfides. Assuming
the excited state of 11 does not selectively associate with itself,
the direct irradiation of 11 in the presence of the sulfoxide 14
would be expected to yield a sulfoxide dimer of 11 and 14.
Degradation of this dimer would lead to molecular oxygen and
the sulfides 8 and 15; however, 15 was not observed. Also, over
the limited range examined, the quantum yields were largely
independent of the starting sulfoxide concentrations. While these
results do not completely rule out the possibility of a sulfoxide
dimer, the involvement of a sulfoxide dimer along the photo-
chemical pathway to deoxygenation was viewed as unlikely.
The photodeoxygenation of sulfoxide 11 to sulfide 8 while
isolated in a matrix indicates that a unimolecular process was
possible for sulfoxides 5 and 11. As expected, the corresponding
sulfides were the major photoproducts observed during irradia-
tion of 5 and 11 in the matrix. The formation of formyl sulfide
12, but not formyl sulfide 13, during irradiations in the EPA
matrix suggested that proximity to the sulfoxide functional group
influences the susceptibility of the benzylic carbons to oxidation
by a unimolecular mechanism. Clearly, conditions in the matrix
were quite different than aqueous solutions, and yet, the extent
of oxidation of the hydroxymethyl substituents of 5 and 11 in
water was quite unexpected. The change in the product and
quantum yields between neutral and basic pH was consistent
with two distinct mechanisms being involved in the photode-
oxygenation.
Similar photochemical mechanisms for 5 and 11 were
expected, and thus for simplicity, the focus of the discussion
will be sulfoxide 11. As shown in Scheme 7, three initial events
from the excited state of 11 were considered. The quantum
yields reported in this paper at high pH were comparable to
those reported for the bimolecular photoreduction of diphenyl
sulfoxide in alcohol/alkoxide solvent systems.12 The bimolecular
photoreduction of 11 was expected to begin with an electron
transfer from HO- to the excited state (11*) to form the radical
anion (20). The photoinduced change in free energy for the
electron-transfer from HO- to 11* can be estimated from
the Rehm-Weller equation (eq 1). Since HO• has no charge,
the energy required to separate an ion pair (Wd) was ap-
proximated as zero. The excitation energy of 11 (∆E00) was
approximated as 3.75 eV. Using 1.77 V vs NHE for E°(HO•/
HO-)13 and -1.80 V vs NHE for E°(11/11•-), the change in
The oxidation potential of 8 (E°(8•+/8) 1.60 vs NHE)
measured in acetonitrile suggests the radical cation 22 can
oxidize water (E°(O2/H2O) ) 0.52 V vs NHE at pH 12).27
Oxidation of water by 22 would be expected to yield sulfide 8,
oxygen, and protons. Since the oxidation of water is a four-
electron process, the oxidation of water by the radical cation
22 is expected to be slow; however, the exergonic nature of
the reaction is anticipated to drive the reaction. The expected
drop in the pH may not have been detected due to the low yield
of 8 in basic conditions. If it is assumed that the oxidation of
water by 22 leads to 8, the dominance of 13 at basic conditions
presumably arises from a competing kinetic process that is
accelerated by HO-.
Deprotonation of the hydroxy group of radical cation 22,
which would be more prevalent at high pH, could lead to an
alkoxy radical. This alkoxy radical intermediate could dispro-
portionate with sulfide 8 to yield the formylsulfide 13 and sulfide
8. However, this alkoxy radical intermediate would also be
expected to undergo a disproportionation reaction with sulfoxide
11 to yield 2-formyl-8-hydroxydibenzothiophene S-oxide, which
was not detected even when the photolysis was carried out to
high conversion. Thus, it seems unlikely that the formation of
13 is due to a bimolecular disproportionation reaction.
A possible explanation for the increased formation of
formylsulfide 13 at high pH is that the radical cation 22 could
potentially oxidize HO- to O•- (E°(O•-/HO-) ) 1.64 V vs
NHE).13 Though, it should be pointed out that this process would
be slightly endergonic. Exposure of sulfide 8 to HO• generated
by the Fenton reaction yielded 13. Thus, the oxidation of HO-
to O•- and a proton could lead to the formylsulfide 13 through
a reaction between the nascent sulfide 8 and O•-/HO• (HO• pKa
) 11.9).13 However, the process would not be expected to go
to completion, and the conversion of 8 to 13 by O•- would
require an additional oxidant. Thus, thorough removal of
molecular oxygen by freeze-pump-thaw cycles would have
been expected to decrease the yield of 13, and yet, the opposite
was observed. Also, the oxidation of HO- by 22 would become
increasingly endergonic as the pH of the solution decreased,
and thus, this process would not be expected to occur at neutral
pH.
An alternative mechanism to explain the increased formation
of formyl sulfide 13 is that the abstraction of a benzylic
(32) Merenyi, G.; Lind, J.; Engman, L. J. Phys. Chem. 1996, 100 (21),
8875–8881.
(33) Chaudhri, S.; Mohan, H.; Anklam, E.; Asmus, K. J. Chem. Soc., Perkin
Trans. 2 1996, (3), 383–390.
(34) Perkins, C. W.; CLarkson, R. B.; Martin, J. C. J. Am. Chem. Soc.
1986, 108, 3206–3210.
9
4472 J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010