The photooxygenation of benzyl, heteroarylmethyl, and allyl sulfides

Article abstract of DOI:10.1002/(SICI)1099-0690(199907)1999:7<1723::AID-EJOC1723>3.0.CO;2-R

The photosensitised oxidation of benzyl ethyl sulfides in aprotic solvents (benzene or acetonitrile) gives the corresponding aldehydes under mild conditions. This is a general reaction which applies to benzyl derivatives containing either electron-donating or electron-withdrawing substituents and furthermore to hereto analogues such as 2-pyridinylmethyl sulfide (not to the 3-indolylmethyl sulfide, since reaction at the heterocycle moiety competes) as well as to allyl sulfides. In a protic solvent (methanol) these sulfides give the sulfoxides instead (except for the nitrobenzyl derivatives, where the aldehyde remains the major product). Among the α-substituted sulfides tested, the α-phenylbenzyl and the 3- cyclohexenyl sulfide give the corresponding ketone (the latter in a low yield), but the α-methylbenzyl sulfide gives the sulfoxide as the main product. The rate for singlet oxygen quenching and for chemical reaction have been measured for representative benzyl sulfides. The reaction is discussed in the frame of the currently accepted mechanism for sulfide photooxygenation. The key step for oxidative C-S bond cleavage appears to be hydrogen transfer from the activated α position in the first formed intermediate, the persulfoxide. This reaction is inhibited in methanol where the persulfoxide is hydrogen-bonded.

Full text of DOI:10.1002/(SICI)1099-0690(199907)1999:7<1723::AID-EJOC1723>3.0.CO;2-R

FULL PAPER
The Photooxygenation of Benzyl, Heteroarylmethyl, and Allyl Sulfides
Sergio M. Bonesi,[a] Rosangela Torriani,[a] Mariella Mella,[a] and Angelo Albini*[a]
Keywords: Oxygenations / Singlet Oxygen / Sulfides / Photochemistry
The photosensitised oxidation of benzyl ethyl sulfides in
and the 3-cyclohexenyl sulfide give the corresponding
aprotic solvents (benzene or acetonitrile) gives the ketone (the latter in a low yield), but the α-methylbenzyl
corresponding aldehydes under mild conditions. This is a sulfide gives the sulfoxide as the main product. The rate for
general reaction which applies to benzyl derivatives
containing either electron-donating or electron-withdrawing
substituents and furthermore to hetero analogues such as 2-
singlet oxygen quenching and for chemical reaction have
been measured for representative benzyl sulfides. The
reaction is discussed in the frame of the currently accepted
pyridinylmethyl sulfide (not to the 3-indolylmethyl sulfide, mechanism for sulfide photooxygenation. The key step for
since reaction at the heterocycle moiety competes) as well as
to allyl sulfides. In a protic solvent (methanol) these sulfides
give the sulfoxides instead (except for the nitrobenzyl
derivatives, where the aldehyde remains the major product).
Among the α-substituted sulfides tested, the α-phenylbenzyl
oxidative C–S bond cleavage appears to be hydrogen
transfer from the activated α position in the first formed
intermediate, the persulfoxide. This reaction is inhibited in
methanol where the persulfoxide is hydrogen-bonded.
Introduction
Results
Explorative Studies
Sulfides are well known to undergo smooth dye-photo-
sensitised oxidation to give sulfoxides (the major products)
and sulfones.[1] This reaction has been the subject of de-
tailed mechanistic investigations. However, CϪS bond
A series of alkyl ethyl sulfides (1) with various alkyl
groups was considered. This included benzyl (1a؊f), allyl
(1g, h) (in both cases α-unsubstituted as well as α-substi-
tuted derivatives were considered), as well as heteroaryl-
methyl sulfides (1i, j). In explorative studies, 0.01  oxygen-
equilibrated solutions of the sulfides in benzene, acetonitrile
or methanol containing triphenylporphine (for the first sol-
vent) or Rose Bengal as the sensitizer were irradiated in the
visible (under this condition, light was absorbed only by
the sensitiser).
From all of the benzyl sulfides (1aϪd, 1f) the main prod-
ucts from the oxygenation both in benzene and in aceto-
nitrile were the corresponding aryl aldehydes (2aϪd) and
ketones (2f) resulting from oxidative CϪS bond cleavage,
with a minor amount of the sulfoxides 3 and sulfones 4
arising from sulfur atom oxygenation (Scheme 1). In the
case of 1a the time evolution of the photooxygenation was
followed, and it was noticed that sulfoxide 3a was negligible
at low conversion, and its amount increased during the con-
version, though it always remained a minor product. A
trace (Յ 0.5%) of the corresponding bibenzyl was formed
from most sulfides. An exception to the general behaviour
was the α-methylbenzyl sulfide 1e, from which only a small
amount of acetophenone (2e) was formed in benzene, while
the main product was sulfoxide 3e.
cleavage occurs in some cases, as first shown by Corey and
[2]
`
Ouannes in the case of benzyl sulfides. These authors
found that dibenzyl sulfide and some alkyl benzyl sulfides
give benzaldehyde (45Ϫ75%) when reacting with singlet
oxygen, produced both by radiofrequency electrodeless dis-
charge or by dye-sensitised irradiation, along with sulfox-
ides and sulfones (25Ϫ55% overall). This reaction was con-
firmed and extended to 9-fluorenyl sulfide by Ando and co-
workers,[3][4] who, also on the basis of the results they had
obtained with some thiazolidines,[5] suggested that persul-
foxides and S-hydroperoxysulfonium ylides are the key in-
termediates. However, the scope of this oxidative cleavage
of sulfides has not been explored. Extension of this reaction
is of interest in view of the mild condition typical of photo-
sensitised oxygenation as a method for the preparation of
aldehydes or at any rate as a synthetic path via sulfides.
Thus, we carried out an explorative study on the photoox-
ygenation of various substituted benzyl sulfides and ex-
tended it to heteroarylmethyl and allyl sulfides. The me-
dium dependence of this reaction was also studied. It
turned out that photooxidation in aprotic media is a
reasonably general method for the preparation of aldehydes
from sulfides (and, with some limitations, also of ketones
from the corresponding α-substituted derivatives), whereas
sulfoxides are generally the main products in protic sol-
vents.
With 1a and 1b the product distribution changed in
methanol and the corresponding sulfoxides 3a, b became
the main products. The change in product distribution oc-
curred in neat methanol or with a large amount (e.g. 30%)
of MeOH as co-solvent with benzene, while when only a
small (e.g. 0.5%) amount was added the products were the
same as in neat benzene. Furthermore, this change did not
occur with all of the benzyl sulfides: With the two nitrated
[a]
Department of Organic Chemistry, University of Pavia,
viale Taramelli 10, I-27100 Pavia, Italy
Fax: (internat.) ϩ 39-0382/507323
E-mail: albini@chifis.unipv.it
Eur. J. Org. Chem. 1999, 1723Ϫ1728
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
1434Ϫ193X/99/0707Ϫ1723 $ 17.50ϩ.50/0
1723

S. M. Bonesi, R. Torriani, M. Mella, A. Albini
FULL PAPER
Scheme 1. Products obtained by the dye-sensitised oxidation of ethyl sulfides 1
Table 1. Products obtained by the dye-sensitized oxidation of ethyl sulfides 1[a]
Substrate
R
RЈ
Solvent
% Sulfide converted
Products (% yield)
2
3
4
a. Benzyl sulfides
1a
Ph
H
H
H
Benzene
MeCN
MeOH
Benzene
MeCN
MeOH
Benzene
MeCN
MeOH
Benzene
MeOH
Benzene
MeOH
Benzene
MeOH
24
35
87
36
40
83
15
22
20
31
34
14
55
20
75
80
76
18
85
78
10
83
75
44
70
30
10
2
3
8
8
5
65
2
10
8
5
7
Ϫ
Ϫ
Ϫ
7
Ϫ
5
Ϫ
22
Ϫ
1b
1c
4-MeOC6H4
4-NO2C6H4
4
55
7
4
3
1d
1e
1f
2- NO2C6H4
H
7
Ϫ
70
65
9
Ph
Ph
Me
Ph
60
40
3
b. Allyl sulfides
1g
PhCHϭCH
H
Benzene
MeOH
Benzene
MeOH
60
78
25
87
45
5
20
7
7
4
15
Ϫ
70
6
1h
ϪCHϭCHϪ[CH2]3Ϫ
20
c. Heteroarylmethyl sulfides
1i
2-Pyridyl
3-Indolyl
H
H
Benzene
MeOH
Benzene
MeOH
45
55
27
55
70
3
10
3
25
74
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
1j
Ϫ
[a]
As obtained by 10 min of irradiation of a 0.01  solution.
derivatives 1c and 1d and with the benzhydryl sulfide 1f the Preparative Experiments
sulfoxide was not formed and the aldehydes or ketones 2
were the main products also in methanol. In this solvent
Since, as mentioned above, aldehydes were in several
the yield of 2 was lower, probably because secondary photo- cases obtained in a fair yield in benzene or MeCN, it was
decoposition of the primarily formed carbonyl derivatives tested whether this could lead to a preparatively useful pro-
was significant in this case.
cedure for aldehydes. Thus, the irradiation was carried out
The same medium-dependent reaction course as with on 100 mL of a 0.1  solution of sulfides 1a, b, c, i in an
benzyl sulfides 1a, b was observed also with the cinnamyl immersion-well apparatus by means of a mercury arc. The
sulfide 1g, with the 2-pyridylmethyl sulfide 1i and, though oxygenation was carried out up to almost complete conver-
with low overall yield, with the aliphatic allyl sulfide 1h. sion, the solvent was evaporated and the residue was bulb-
The 3-indolyl sulfide 1j gave only 10% of the aldehyde 2j. In to-bulb distilled under reduced pressure. In this way benzal-
order to compare the reactivity of the substrates, the results dehyde (2a) and its p-methoxy (2b) and p-nitro (2c) deriva-
obtained by 10 min irradiation Ϫ which involved a conver- tives as well as 2-pyridinecarbaldehyde (2i) were obtained
sion ranging from 14 to 87% Ϫ are gathered in Table 1.
from the corresponding sulfides in a 0.5- to 1-g amount and
1724
Eur. J. Org. Chem. 1999, 1723Ϫ1728

The Photooxygenation of Benzyl, Heteroarylmethyl, and Allyl Sulfides
FULL PAPER
Table 2. Rate constants for the reaction of singlet oxygen with the
With some of these sulfides, the course of the reaction
changes in protic solvents, where the sulfoxides Ϫ i.e. the
same products which are obtained from dialkyl sulfides Ϫ
are the main products. With the nitro derivatives 1c and 1d,
however, the aldehydes remain the main products also in
methanol, though with a lower yield probably due to sec-
ondary photodecomposition.
This oxidative cleavage and the characteristic solvent de-
pendence are not limited to benzyl sulfides. Thus, aldehydes
or sulfoxides depending on the solvent are obtained also
from a hetero analogue such as the 2-pyridylmethyl sulfide
1i (from the 3-indolylmethyl derivative 1j the yield of alde-
sulfides 1a؊c
Substrate
Solvent
kT [Ϫ1sϪ1 [a]
]
kr [Ϫ1sϪ1 [b]
]
1a
Benzene
MeCN
MeOH
Benzene
MeCN
MeOH
Benzene
MeCN
MeOH
1.14ϫ107
1.22
0.55ϫ107
0.61
0.94
1.2
1b
1c
1.66
0.52
1.37
0.34
1.40
1.65
0.17
0.10
0.28
0.23
0.13
0.14
[a]
Total quenching rate constant, as determined by measuring of hyde is poor; however, see below) and from allyl derivatives,
[b]
the singlet oxygen lifetime through the emission at 1270 nm. Ϫ
with a satisfactorily yield from cinnamyl sulfide 1g and a
Chemical reaction rate constant, as measured in competition exper-
iments in the presence of octaline, kr ϭ 1.84·106 Ϫ1sϪ1, ref.[6] Ave-
poor yield from the 3-cyclohexenyl sulfide 1h. As for the
two α-substituted benzyl sulfides, the benzhydryl sulfide 1f
gives ketone 2f, while with the α-methylbenzyl sulfide 1e
sulfoxide 3e is the main product with only 10% of 2e. In
some cases the sulfone is also a significant product (note-
worthy, 22% of 4e from 1e).
rage of at least 3 experiments within 5%.
with a yield of 52, 60, 62, and 55%, respectively, through a
simple procedure that may be easily scaled up (see Exper-
imental Section).
The present results with benzyl sulfides and related de-
rivatives can be discussed with reference to the generally
accepted mechanism for the photosensitised oxidation of
dialkyl sulfides.[1] These give the sulfoxides under all con-
ditions, but the efficiency of the reaction is known to be
medium-dependent. More precisely, the total rate of singlet
oxygen quenching (kT ϭ kq ϩ kr, see Equations 1, 2) does
not change, while the rate of the chemical reaction (kr) is
much higher in protic than in aprotic solvents.[1a,7,8] This is
currently explained by admitting the intermediacy of per-
sulfoxide 5 (see Scheme 2), although such a species has
never been directly detected. In aprotic media this mainly
collapses to unchanged sulfide and triplet oxygen, and in
part (kr Յ 5% kd, see Scheme 2) rearranges to a further in-
termediate, usually assumed to be the thiadioxirane 6 (see
below, however). This in turn transfers an oxygen atom to
a sulfide giving two sulfoxide molecules. Protic solvents, on
the other hand, preclude unproductive decay and make for-
mation of the sulfoxide much more effective.[1a,7,8]
Kinetic Data
In order to obtain mechanistic information the two key
kinetic parameters, viz the total (physical ϩ chemical,
Equations 1, 2) quenching rate of singlet oxygen by the sulf-
ides (kT ϭ kq ϩ kr) and the rate of chemical reaction (kr)
were measured in representative cases, viz for the three
benzyl sulfides 1a؊c.
3
1O2 ϩ R2S Ǟ O2 ϩ R2S
(1, rate constant kq)
(2, rate constant kr)
1O2 ϩ R2S Ǟ Products
The value of kT was directly determined by laser flash
photolysis through the quenching of the singlet oxygen pro-
duced by phenalenone sensitisation and that of kr by de-
termining the competitive reaction of the same sulfides and
octaline (assumed to add singlet oxygen at a solvent-inde-
pendent rate) in low-conversion experiments according to
the method introduced by Foote.[6] The data in three sol-
vents, benzene, acetonitrile and methanol, are reported in
Table 2.
Discussion
Inspection of Table 1 shows that the oxidative cleavage
first reported by Corey[2] is in fact a general reaction of
benzyl sulfides, tolerant of ring substitution. All of the α-
unsubstituted derivatives give moderate to good yields of
Scheme 2. Mechanism for the photooxidation of alkyl sulfides
Table 2 shows that quenching of singlet oxygen by benzyl
the aldehydes when the irradiation is carried out in aprotic ethyl sulfide 1a occurs at a rate (kT ϭ 1.14·107 Ϫ1sϪ1 in
(both polar and apolar) solvents such as benzene or aceto- benzene, see Table 2) in the same order as that of simple
nitrile. This may be a convenient way to prepare aldehydes, alkyl sulfides (e.g., for Et2S: kT ϭ 1.7, 2.0, or 3.04·107
since visible light absorbed only by the sensitizer is used Ϫ

Ϫ1sϪ1 has been reported).[8Ϫ11] Singlet oxygen is an elec-
so that no direct photochemical reaction of the substrate trophile and reacts faster with electron-rich substrates. Ac-
takes place Ϫ and the reaction occurs under the peculiarly cordingly, ring substitution affects the kT rate in the ex-
mild condition of photochemical reactions. As shown in a pected way with these benzyl sulfides, 4-MeO > H > 4-O2N,
number of representative cases, this preparation of alde- see Table 2. The methylene group interposed between the
hydes can be conveniently carried out on a gram scale.
ring and the sulfur atom reduces the effect relative to that
Eur. J. Org. Chem. 1999, 1723Ϫ1728
1725

S. M. Bonesi, R. Torriani, M. Mella, A. Albini
FULL PAPER
observed for phenyl sulfides (a methoxy group in 4 in- of the persulfoxide, as in general of 1,3-dipoles.[15a] Note-
creases kT by a factor of 3.5 in thioanisole,[12] by a factor worthy, the ratio kr/kT, which may be taken as a rough esti-
of 1.45 in benzyl sulfide 1). Table 2 shows that quenching mate of the propensity of the persulfoxide to react chemi-
is little solvent-dependent, as is generally assumed[8] to be cally rather than decay to the components, only marginally
the case for alkyl sulfides.
increases both in the series 1b < 1a < 1c and in polar
The chemical reaction of benzyl sulfides differs from that MeCN relative to benzene. Thus, polar factors have a slight
of dialkyl sulfides both quantitatively and qualitatively. effect on the rearrangement efficiency, and it is the homo-
With Et2S, as an example, the product of sensitised oxygen- lytic strength of the CϪH bond that matters (only in the
ation remains the sulfoxide under all conditions, but its rate case of the less activated 1c polar acetonitrile does double
of formation (kr) strongly increases (by a factor of ca. 15, the rate of reaction with respect to benzene, which provides
from 3.8·106 to Ն 5.6·107 Ϫ1sϪ1) in going from benzene to evidence for some polar effect). A different homolytic frag-
methanol.[8] With the benzyl sulfides in Table 2 the rate of mentation of the ylide reasonably originates the observed
reaction kr accounts for a large fraction of the total quench- trace of bibenzyl.
ing (kT) in benzene (32% with 1b, 70% with 1c, see Table 2,
instead of Յ 5% of Et2S). Furthermore, the kr increase in
MeOH is moderate (by a factor of 2 with 1a, 3 with 1b, 1.4
with 1c) but involves a change in the product distribution
(except for 1c).
As for the mechanism of CϪS cleavage, Ando showed
that the sensitised oxidation of a series of thiazolidines
leads to α-hydroperoxysulfides via a strongly oxidising in-
termediate, suggested to be an S-hydroperoxysulfonium yl- Scheme 4. Alternative paths from the persulfoxide
ide.[4] He found that the oxidation of benzyl sulfide also
produces an oxidising intermediate (it reacts with Me2S)
The easy formation of the ylide explains most of the re-
and therefore suggested that benzaldehyde is formed from sults, but two points need further consideration. First, the
the persulfoxide via ylide 7 and hydroperoxide 8 as shown α-methyl sulfide 1e gives the sulfoxide as the main product,
in Scheme 3.[4] This is reasonable; in accord with the inter- not the ketone (while the α-phenyl sulfide 1f behaves as the
mediacy of the ylide, we observe that the proportion of sulf- α-unsubstituted derivatives). It may be that the methyl
oxide 3a from 1a in apolar solvents grows somewhat with group disfavours the conformation suitable for hydrogen
increasing conversion, probably by secondary oxidation of transfer in the persulfoxide and makes this process too slow
1a by 7.
to compete with some other path leading to the sulfoxide.
We are currently carrying out a computational analysis of
the persulfoxide conformation in order to support this hy-
pothesis.[15b]
Alternatively, it may be that the ylide is formed but does
not rearrange to the hydroperoxide. As mentioned above,
calculations suggest that a viable path for the ylide is re-
arrangement to the sulfone.[13] With the present benzyl de-
rivatives sulfones are often significant products (5Ϫ10%)
and this is the second noteworthy point. The fact that the
sulfide for which the maximum yield of sulfone obtained is
1f (22% 4f in benzene), for which a decrease of the cleavage
to the ketone to 60% (from 80% with 1a) is observed, sug-
Scheme 3. Mechanism for the photooxidation of benzyl alkyl sulfi-
des
Recent computations[13] generalized the argument, sug- gests that this is indeed an alternative path from the ylide.
gesting that the persulfoxide from dimethyl sulfide may also In protic (not in polar) solvents the efficiency of the reac-
rearrange to the corresponding ylide, not to the thiadioxir- tion increases and S-oxygenation becomes the main process
ane as previously assumed, and that this may be the path with most of the above sulfides. Mechanistic studies on the
leading to the sulfone (for this point there is also some ex- oxygenation of Et2S in methanol by Foote[7] and Clen-
perimental evidence)[14] as well as to the α-hydroperoxide nan[8][16] suggested that under these conditions the solvent
(Scheme 4). It remained unclear whether the ylide may also forms a hydrogen bond with the persulfoxide (formula 9 in
be involved in the formation of the sulfoxide. As mentioned Scheme 3) or adds to it (to give a hydroperoxy alkoxy sulf-
above, with dialkyl sulfides all of these paths are unimpor- ide. The former hypothesis is simpler and explains the pre-
tant, however. It appears that with the present substrates sent results. Indeed, it is reasonable that when the outer
the weak (benzylic or allylic) α-CϪH bond makes re- oxygen atom of the persulfoxide is involved in a hydrogen
arrangement to ylide 7 faster and thus a better competitor bond, intramolecular hydrogen-transfer to give 7 does not
for decay of the persulfoxide to the components. This can occur, and the reaction rather is electrophilic oxygen trans-
be understood with reference to the mixed zwitterion Ϫ di- fer to a second molecule of sulfide (to give ultimately two
radical nature (see Scheme 3, mesomeric formulae 5 ↔ 5Ј) molecules of sulfoxide, see Scheme 3).
1726
Eur. J. Org. Chem. 1999, 1723Ϫ1728

The Photooxygenation of Benzyl, Heteroarylmethyl, and Allyl Sulfides
FULL PAPER
Some points are worth noticing with respect to the pro- drogen transfer the predominant reaction from the persulf-
tonation of the persulfoxide. This is a weak base, as indi- oxide. The highly effective oxygenation of benzyl sulfides
cated by the fact that the change in the product distribution and their heteroaryl analogues discussed here may be rel-
(from aldehydes to sulfoxides) requires a large amount of evant for biological systems.
MeOH as co-solvent, while 0.5% has no effect. This indi-
cates that the intermediate is quite short-lived or it is pro-
Experimental Section
tonated at a rate below diffusion control. For the case of
1a, the latter hypothesis has been supported through a more
detailed examination of this point.[17] With the nitro-substi-
tuted derivatives 1c and 1d, the electron-withdrawing sub-
stituent is expected to limit charge localisation of the nega-
tive charge on the outer oxygen atom and thus to make the
persulfoxide still less basic. In fact, with these derivatives
the rate of photooxygenation in MeOH increases less than
with 1a and the benzaldehyde remains an important prod-
uct.
Finally, two of the sulfides we considered, 1h and 1j, did
not undergo any of the above reactions in a satisfactorily
yield. This is due the occurrence of alternative processes.
These were not investigated, but may have been expected in
these cases. As for 1h, Clennan demonstrated that in ali-
phatic allyl sulfides the CϭC bond participates in the reac-
tion[16b,c] and different products are formed. Remarkably,
with a cinnamyl rather than an aliphatic allyl group both
of the general reactions, oxidative cleavage in aprotic and
S-oxidation in protic media, occur unperturbed (see 1g in
Table 1). In the case of 1g again a different process is ex-
pected to intervene because of the known sensitivity of the
indole ring to singlet oxygen.[18]
General Remarks: Spectroscopic-grade solvents were used. Benzene
and acetonitrile were distilled from calcium hydride before use. The
sulfides 1a,[19] 1b,[20] 1c,[21] 1d, e,[22] 1f,[23] 1g,[24] 1h,[25] and 1i[26]
were prepared from the condensation of the appropriate chlorides
or bromides and sulfides according to published procedures. The
heterocyclic derivative 1j[27] was prepared from indole as reported.
Samples of the sulfoxides
3a؊i[28Ϫ30] and sulfones
4a؊i[20,22,24,30Ϫ34] for the comparison with the photoproducts were
prepared by 3-chloroperbenzoic acid or acetic acid/hydrogen per-
oxide oxidation according to the general procedure.[35] Rose
Bengal, tetraphenylporphine and 1,2,3,4,5,6,7,8-octahydronaph-
thalene were commercial products (Aldrich).
Photochemical Reactions: Explorative irradiations (see Table 1) in
the presence of the appropriate sensitiser were carried out in rub-
ber-stoppered Pyrex tubes for 10 min while passing a stream of
oxygen saturated with the solvent through the solutions. Four phos-
phor-coated 15-W lamps (Applied Photophysics) emitting from 350
to 700 nm were used as the light source. The course of the reaction
was followed by VPC and/or HPLC. Ϫ Kinetic measurements were
carried out in the same way in the presence of octaline and the
reactions were discontinued at Յ 20% conversion. The solution was
treated with triphenylphosphane before analysis in order to reduce
the hydroperoxide to the allylic alcohol and then analysed as above.
The data reported in Table 2 are the average of at least three meas-
urements within 5%.
Benzaldehyde: A solution of 1.52 g of benzyl ethyl sulfide (1a,
0.1 ) and 50 mg of TPP in 100 mL of benzene was irradiated for
5 h in a water-cooled immersion-well apparatus fitted with a high-
pressure mercury arc (Helios Italquartz, 150 W) through an uranyl
filter (λtr > 400 nm) while passing a slow stream of oxygen through
the solution. The solvent was removed by rotary evaporation and
the residue distilled to give 0.55 g (52% yield) benzaldehyde. The
reaction was carried out analogously with sulfides 1b, c, and i to
give the corresponding aldehydes in the 60, 62, and 55% yield,
respectively.
Conclusion
The photooxidative cleavage of benzyl sulfides first re-
ported by Corey is in fact a general reaction for all sub-
strates containing an activated α-hydrogen atom. Benzyl
(independently from ring substitution), heteroarylmethyl
(provided that the heterocycle is not itself reactive with 1O2)
and phenylallyl sulfides all give the corresponding alde-
hydes in a fair yield by reaction with singlet oxygen in ap-
rotic solvents. The production of ketones from α-substi-
tuted sulfides appears to be subjected to some limitations.
Analytical Methods: Gas-chromatographic determinations were
carried out by means of a Hewlett Packard Mod. 5890 with an
At least for aldehydes, the generality and simplicity of the Innowax column (30 m ϫ 0.25 mm, 0.5 µm film thickness) and He
as the carrier gas. The injector pressure was constant (12 ψ). The
temperature program was 50Ϫ140 °C (5 °C/min), 140Ϫ250°C
(25 °C/min, 7 min final time). For mass determinations this was
equipped with a Mod. 5971 mass detector.
reaction, the mild conditions involved and the wide applica-
bility due to the resistance of most organic functions to 1O2
and the lack of influence of ring-substitution suggest that
the above procedure may be proposed as a method for their
preparation from sulfides. This may be a mild alternative to
other known methods for the overall conversion of a halide
to an aldehyde.
In methanol, reaction with singlet oxygen in most cases
leads to sulfoxides. This reaction may also be of some inter-
est, for example in view of the fact that S-oxidation occurs
under mild conditions and, in the case of the pyri-
dinemethyl sulfide 1i, with no competition by N-oxidation.
The observed chemistry fits in the frame of the generally
accepted mechanism for sulfide oxidation, with the differ-
Photophysical Measurements: Rate constants for the quenching of
singlet oxygen were measured by Prof. G. Aloisi and F. Elisei at the
Dept. of Chemistry, University of Perugia. This was done by laser
pulse spectroscopy (Spectra Physics Nd:YAG laser), by measuring
the emission lifetime at 1270 nm after sensitization by phena-
lenone.[36]
Acknowledgments
S. M. B. (permanent address: Department of Organic Chemistry,
ence that an activated α-CϪH bond makes homolytic hy- University of Buenos Aires, Ciudad Universitaria, CP1428 Buenos
Eur. J. Org. Chem. 1999, 1723Ϫ1728 1727

S. M. Bonesi, R. Torriani, M. Mella, A. Albini
FULL PAPER
[14]
K. Ishiguro, M. Hayashi, Y. Sawaki, J. Am. Chem. Soc. 1996,
118, 7265.
Aires, Argentina) thanks FOMEC for a fellowship. R. T. thanks
Prochimica for a fellowship. Dr. P. P. Rossi (Prochimica) is thanked
for discussing this project. Financial support by Consiglio Nazio-
nale delle Ricerche, Rome, is gratefully acknowledged.
[15] [15a]
[15b]
P. C. Hiberty, Isr. J. Chem. 1983, 23, 10. Ϫ
S. M.
Bonesi, M. Freccero, A. Albini J. Phys. Org. Chem., submitted.
[16] [16a]
E. L. Clennan, K. Yang, J. Org. Chem. 1992, 57, 4477. Ϫ
[16b]
E. L. Clennan, X. Chen, J. Am. Chem. Soc. 1989, 111,
[16c]
5787. Ϫ
111, 8212.
E. L. Clennan, X. Chen, J. Am. Chem. Soc. 1989,
[17]
[18]
[1] [1a]
[1b]
S. M. Bonesi, M. Mella, N. dЈAlessandro, G. G. Aloisi, M.
Vanossi, A. Albini, J. Org. Chem. 1998, 63, 9946.
H. W. Wasserman, B. H. Lipshutz, in: Singlet Oxygen (Eds.: H.
H. Wasserman, R. W. Murray), Academic Press, New York,
1979, p. 430.
E. L. Clennan, Sulfur Rep. 1996, 19, 171. Ϫ
W. Ando,
T. Takada, in: Singlet O2 (Ed.: A. A. Frimer), CRC, Boca Ra-
[1c]
ton, 1985, vol.3, p. 1. Ϫ
W. Ando, Sulfur Rep. 1981, 1, 143.
C. S. Foote, in: Singlet Oxygen (Eds.: H. H. Wasserman, R. W.
Murray), Academic Press, New York, 1979, p. 139. Ϫ [1d] A. A.
[19]
[20]
[21]
[1e]
G. Modena, Gazz. Chim. Ital. 1960, 89, 834.
Gorman, Adv.Photochem. 1992, 17, 217. Ϫ
F. Jensen, in:
F. R. Lewis, S. Archer, J. Am. Chem. Soc. 1951, 73, 2109
H. Hellmann, D. Eberle, Justus Liebigs Ann. Chem. 1963, 662,
188.
Advances in Oxygenated Processes (Ed.: A. L. Baumstark), JAI
[1f]
Press, Greenwich CT, 1995, vol. 4, p. 49. Ϫ
C. H. Krauch Angew.Chem. 1962, 74, 510.
G. O. Schenck,
[22]
[23]
[24]
[25]
[26]
[2]
[3]
W. E. Bacon, W. M. LeSuer, J. Am. Chem. Soc. 1954, 76, 670.
M. Dagonneau, J. Vialle, Bull. Soc. Chim. Fr. 1972, 2062
E. A. Fehnel, P. R. Resnick, J. Org. Chem. 1955, 20, 996.
L. Batenau, F. W. Shipley, J. Chem. Soc. 1955, 1996.
M. Winter, F. Gautshi, I. Flament, M. Stoll (Firmenich S. A.),
US Patent 3,952,026, 1976; Chem. Abstr. 1976, 85, 94216a.
F. Poppelsdorf, S. J. Holt, J. Chem. Soc. 1954, 1124.
A. Cerniani, G. Modena, P. E. Todesco, Gazz. Chim. Ital. 1960,
90, 3.
`
E. J. Corey, C. Ouannes, Tetrahedron Lett. 1976, 4263.
W. Ando, T. Nagashima, K. Saito, S. Kohmoto, J. Chem. Soc.,
Chem. Commun. 1979, 154.
[4]
T. Akasaka, A. Sakurai, W. Ando, J. Am. Chem. Soc. 1991,
113, 2696.
[5] [5a]
T. Takata, K. Hoshino, E. Takeuchi, Y. Tamura, W. Ando,
[27]
[28]
Tetrahedron Lett. 1984, 25, 4767. Ϫ [5b] T. Takata, K. Ishibashi,
W. Ando, W. Tetrahedron Lett. 1985, 26, 4609.
R. Higgins, C. S. Foote, H. Cheng, Adv. Chem. Ser. 1968, 77,
102.
[6]
[7]
[29]
Y. Kodama, S. Zushi, K. Nishihata, M. Nishio, J. Uzawa, J.
Chem. Soc., Perkin Trans. 2 1980, 1307
J. J. Liang, C. L. Gu, C. S. Foote, J. Am. Chem. Soc. 1983,
[30]
[31]
`
S. Becker, Y. Fort, P. Caubere, P. J. Org. Chem. 1990, 55, 6194
105, 4717.
J. Buechi, M. Prost, H. Eichenberger, R. Lieberherr, Helv. Chim.
Acta 1952, 35, 1527.
[8]
[9]
E. L. Clennan, A. Greer, J. Org. Chem. 1996, 61, 4793.
E. L. Clennan, H. Zhang, J. Am. Chem. Soc. 1995, 117, 4218.
M. L. Kacher, C. S. Foote, Photochem. Photobiol. 1979, 29, 765.
F. Wilkinson, in: Singlet Oxygen. Reaction with Organic Com-
pounds and Polymers (Eds.: B. Ranby, J. F. Rabek), Wiley, Lon-
don, 1978, p. 27.
[32]
[33]
F. L. Riley, E. Rothstein, J. Chem. Soc. 1964, 3860.
N. Furukawa, S. Ogawa, T. Kawai, S. Oae, J. Chem. Soc., Perkin
Trans. 2 1984, 1833.
[10]
[11]
[34]
[35]
J. J. Licari, G. Dougherty, J. Am. Chem. Soc. 1954, 76, 4039.
J. Drabowicz, M. Mikolajczyk, Org. Prep. Proced. Int. 1982,
14, 45.
[12]
B. M. Monroe, Photochem. Photobiol. 1979, 29, 761.
[13] [13a] F. Jensen, A. Greer, E. L. Clennan, J. Am. Chem. Soc. 1998,
R. Schmidt, J. Photochem. Photobiol. A: Chem. 1994, 79, 11
Received November 30, 1998
[36]
[13b]
120, 4439. Ϫ
3963.
M. L. McKee, J. Am. Chem. Soc. 1998, 120,
[O98541]
1728
Eur. J. Org. Chem. 1999, 1723Ϫ1728