Singlet oxygen oxidation of selenides to selenoxides

Article abstract of DOI:10.1016/S0040-4039(02)01246-7

The presence of water and a carbonate proved to be highly beneficial for singlet oxygen oxidation of selenides to selenoxides.

Full text of DOI:10.1016/S0040-4039(02)01246-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 6255–6257
Singlet oxygen oxidation of selenides to selenoxides
Alain Kriefa,* and Fre´de´ric Loneza,b
aLaboratoire de Chimie Organique de Synthe`se, De´partement de Chimie, Faculte´s Universitaires Notre-Dame de la Paix,
61 rue de Bruxelles, Namur B-5000, Belgium
bFond pour la Recherche Scientifique dans l’Industrie et l’Agriculture (F.R.I.A.), 5 rue d’Egmont, Bruxelles B-1000, Belgium
Accepted 27 June 2002
Abstract—The presence of water and a carbonate proved to be highly beneficial for singlet oxygen oxidation of selenides to
selenoxides. © 2002 Elsevier Science Ltd. All rights reserved.
Several years ago, we disclosed1 that singlet oxygen,
produced by shining light on molecular oxygen and
trace amount of rose bengal, is able to efficiently oxi-
dize in methanol, selenides to selenoxides (Scheme 1).1
to prove that selenoxides are valuable co-oxidants in
the SeOAD reaction.3b
As an extension of this work, we also disclosed3b that
the rate of the SeOAD reaction was highly dependent
upon the nature of the selenoxide. Dialkyl selenoxides
are less efficient than alkyl aryl selenoxides3b,c and the
presence of an electron-withdrawing group on the aro-
matic ring proved to be highly beneficial.
Since that time, this reaction was successfully used (i)
by Abadjoglou2 for the reoxidation of Os(VI) to
Os(VIII) implied in the dihydroxylation of olefins and
(ii) by us3a for its asymmetric version (SeAD reaction,
Scheme 2). The SeAD reaction is very close to the
Sharpless AD reaction4 but uses oxygen in place of
potassium ferricyanate as the co-oxidant (Scheme 2).3a
It was obvious that these findings would have an
important impact on the SeAD reaction. In this case,
the problem was expected to be more complex due to
the fact that those aryl selenides which possess an
electron-withdrawing group on their aromatic ring are
those which are usually the most difficult to oxidize.
Consequently it was important to know which of the
two steps (oxidation of the selenide or reduction of the
selenoxide) is the rate-limiting one.
It is not clear whether the selenoxide, or a putative
peroxoselenide intermediate, is the effective oxidant.
However, in a parallel work, which involved a stoichio-
metric amount of selenoxides, we were effectively able
We decided to get a more detailed picture of the
reactivity of singlet oxygen toward selenides and
planned to extend this reaction to the whole series of
selenoxides used in the previous work3b and to carry
out kinetic comparative studies.
Scheme 1. Oxidation of selenides to selenoxides using singlet
oxygen.
We first performed the reaction on benzyl phenyl sele-
nide (10−2 M, oxygen, 10−4 M rose bengal, methanol,
Ph
Me
Ph
K2OsO2(OH)4,(DHQD)2PHAL, K2CO3, 8 % PhSeBn
aqueous t-BuOH, 12°C, O2, 0.3% rose bengal, hν
OH
Me
OH
Scheme 2. Asymmetric dihydroxylation of a-methyl styrene (SeAD reaction) using selenides and singlet oxygen.3a
* Corresponding author. Tel.: +32 81 724539; fax: +32 81 724536; e-mail: alain.krief@fundp.ac.be
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)01246-7

6256
A. Krief, F. Lonez / Tetrahedron Letters 43 (2002) 6255–6257
Table 1. Reaction of selenides 1 with singlet oxygen under different experimental conditions according to Scheme 1
Conditions
I
II
III
Ar
Ph
R1
Ta
2/1 MeOH
2/1b1+H2O
2/1b2+H2O+K2CO3
99/1
a
b
c
d
e
f
PhCH2
4
16
4
4
4
12/88
50/50
0/100
21/79
43/57
0/100
0/100
67/33
98/2
0/100
84/16
99/1
p-NO2-Ph
p-MeO-Ph
Ph
PhCH2
PhCH2
Me
4/96
99/1
100/0
44/56
95/5
Ph
Ph
4
16
3/97
6/94
g
a T refers to the reaction time quoted in hours.
b These reactions have been carried out in 0.01 M solution (1) in MeOH/H2O (6/3) or (2) in MeOH/H2O (6/3) in the presence of 0.5 M equiv.
of K2CO3.
light,† 20°C)1 under conditions closely related to those
we previously used and were surprised to get after 4 h
reaction only a small amount of the corresponding
selenoxide (2a/1a 12/88, Scheme 1, Table 1 entry Ia)
whereas in our original paper we reported a much
better yield of benzyl phenyl selenoxide after a much
shorter reaction time (70%, 2.5 h).1 It was moreover not
an isolated case as, except for methyl phenyl selenide
(Table 1, entry Ie) which proved to be a little more
prone to oxidation, all other aryl selenides proved to be
reluctant to react (Table 1, entry I). All the modifica-
tions we tried (lamp,‡,§ quality of oxygen, careful drying
and distillation of methanol) unfortunately did not
improve the results.
thesis of a few other selenoxides by performing the
reaction in methanol/water (6/3) (Table 1, entry II
compare to entry I). We were unable to oxidize p-nitro-
benzyl phenyl selenide but we have produced, although
in small amount (6%), diphenyl selenoxide from
diphenyl selenide (Table 1, entry g). It is interesting to
recall that it has been firmly established that singlet
oxygen is unable to oxidize the analogous diphenyl
sulfide.5,6
It is not obvious to rationalize the role of water in this
reaction since it was believed, until recently, to be a
singlet oxygen scavenger.6 Nevertheless a very recent
paper stresses the ‘‘unusual regiochemical dependence’’
of the singlet oxygen ene-reaction involving tiglic acid
and related esters towards the polarity of the solvent.7
Interestingly, these reactions have been carried out in
methanol and in methanol/water (60/40) and differ-
ences have been noticed.7
We finally planned to carry out this reaction in differ-
ent solvents.
Little choice remains since several non-polar solvents
such as toluene or methylene dichloride do not dissolve
rose bengal enough to allow the reaction. We tried
unsuccessfully to perform the reaction (i) in other alco-
holic solvents such as ethanol, i-propanol and t-
butanol or (ii) in acetonitrile or in acetone (Table 2,
entry I).
In relation to very recent findings,8 singlet oxygen could
also have produced hydrogen peroxide which is an
efficient oxidant of selenides to selenoxides. However the
inaptitude of the newly developed method to oxidize
We, however, surprisingly found§ that benzyl phenyl
selenoxide is produced at a faster rate if the reaction is
carried out in wet methanol (Table 1, entry IIa, com-
pare to entry Ia). A larger amount of water increases
substantially the reaction rate up to a certain point
(MeOH/H2O [2a/1a]: 100/0 [12/88]; 90/10 [15/85]; 80/20
[33/67]; 70/30 [55/45]; 66/33 [67/33]) from which a sharp
decrease occurs due to the insolubility of the selenide in
the medium (MeOH/H2O [2a/1a]: 60/40 [60/40]; 50/50
[37/63]). We have then successfully performed the syn-
Table 2. Reaction of benzyl phenyl selenide 1a with sin-
glet oxygen in different solvents and with different addi-
tives according to Scheme 1
Conditionsa
Solvent
I
II
III
2/1b1 Neat
2/1b2+H2O 2/1b3+H2O
+K2CO3
a
b
c
d
e
f
MeOH
EtOH
i-PrOH
t-BuOH
MeCN
Me2CꢀO
12/88
0/100
0/100
0/100
0/100
0/100
67/33
10/90
0/100
0/100
0/100
0/100
99/1
97/3
65/35
39/61
61/39
30/70
† In our original work we used a lamp which is no longer available.1
‡ We have performed these reactions using (i) two halogen spots
(Osram Haloline 230 V, 500 W emitting mainly in the visible region
and providing each one 9500 lumens, (ii) four economic cold light
lamps (Osram Delux EL, 230 V, 23 W) providing 1580 lumens each,
and (iii) two sodium lamps (Philips SDW-T, 230 V, 50 W providing
2300 lumens each. All three sources provide very closely related
results on benzyl phenyl selenide.
a In all cases, reaction time is 4 h.
b These reactions have been carried out in 0.01 M solution in (1) the
solvent (2) an aqueous solution (solvent/H2O: 6/3) or (3) an
aqueous solution (solvent/H2O: 6/3) in the presence of 0.5 M equiv.
of K2CO3.
§ For convenience the Philips SDW-T lamp has been mostly used in
this work.

A. Krief, F. Lonez / Tetrahedron Letters 43 (2002) 6255–6257
6257
Table 3. Reaction of benzyl phenyl selenide with singlet oxygen in different solvents and with different additives
Time (h)
PhSMe ratio
PhS(O)Me ratio
PhS(O)2Me ratio
Anhydrous methanol
4
16
4
55
04
00
02
39
74
89
90
6
22
11
8
Aqueous methanol
Aqueous methanol+K2CO3
4
p-nitro-phenyl benzyl selenide sheds some doubt on this
hypothesis since we have independently found that
p-nitro-phenyl benzyl selenoxide is effectively generated
on reaction with hydrogen peroxide under closely
related conditions.9
obviously does not contain acidic hydrogens. We are
actively working to identify the role of water and
potassium carbonate in these reactions.
Interestingly, preliminary results tend to show that
sulfides exhibit a closely related behavior although both
the sulfoxides and the sulfones are simultaneously pro-
duced whatever conditions are used. For example, pho-
toxidation of methyl phenyl sulfide in methanol/water
produces the corresponding sulfoxides and sulfones
more efficiently than in methanol alone (Table 3). Pho-
toxidation of diphenyl sulfide however still remains a
problem.6
Singlet oxygen could also participate in the reaction by
stabilizing or trapping one of the multiple postulated
intermediates which finally leads to the selenoxide. We
do not yet have tangible arguments in favor of one of
the hypotheses disclosed above or any other alternative
explanation. Anyhow, the reaction is slower when per-
formed in deuterated methanol (MeOD)¶ instead of
methanol, attesting thus the important role of the
solvent.
References
Even more interesting is the fact that the effect we have
just reported is substantially enhanced by carrying out
the reaction in aqueous methanolic solution containing
potassium carbonate (Table 1, entry III). The best
results have been observed when the reaction is per-
formed with a 0.5 M equiv. of this inorganic com-
pound. Two extreme cases merit discussion: on the one
hand these conditions still do not allow the oxidation of
p-nitro-phenyl benzyl selenide (Table 1, entry IIIc), and
on the other hand they permit the efficient oxidation of
diphenyl selenide to diphenyl selenoxide (Table 1, entry
IIIg).
1. Hevesi, L.; Krief, A. Angew. Chem. 1976, 88, 413.
2. Abatjoglou, A. G.; Bryant, D. R. Tetrahedron Lett. 1981,
22, 2051.
3. (a) Krief, A.; Colaux, C.; Dumont, W. Tetrahedron Lett.
1997, 38, 3315; (b) Krief, A.; Castillo-Colaux, C. Synlett
2001, 501; (c) Krief, A.; Castillo-Colaux, C. Chem. Com-
mun. 2002, 558.
4. Kolb, H. C.; VanNieuwenhze, M. S.; Sharpless, K. B.
Chem. Rev. 1994, 94, 2483.
5. Jensen, F.; Greer, A.; Clennan, E. L. J. Am. Chem. Soc.
1998, 120, 4439.
6. (a) Singlet Oxygen; Frimer, A. A., Ed.; CRC Press: Boca
Raton, Florida, 1985; Vols. 1–4; (b) Ando, W.; Takata, T.
In Singlet Oxygen; Frimer, A. A., Ed. Photoxidation of
sulfur compounds. CRC Press: Boca Raton, Florida, 1985;
Vol. 3, p. 1.
7. Stensaas, K. L.; Payne, J. A.; Ivancic, A. N.; Bajaj, A.
Tetrahedron Lett. 2002, 43, 25.
8. Wentworth, P.; Jones, L. H.; Wentworth, A. D.; Zhu, X.;
Larsen, N. A.; Wilson, I. A.; Xu, X.; Goddard, W. A., III;
Janda, K. D.; Eschenmoser, A.; Lerner, R. A. Science
2001, 293, 1806.
The observations reported apply to a wide range of
alcoholic solvents for which increasing reactivity is
found if the reaction is carried out in the presence of
water or aqueous potassium carbonate solution (Table
2, entries II, III). These results express the trends and
do not reflect the best conditions for selenoxide synthe-
sis. For example, in t-butanol, benzyl phenyl selenide is
more efficiently oxidized if the reaction is performed
with 3 instead of 0.5 M equiv. of potassium carbonate
(Table 3).
9. (a) Krief, A.; Laval, A.-M. Bull. Soc. Chim. Fr. 1997, 869;
(b) Grieco, P. A.; Gilman, S.; Nishizawa, M. J. Org.
Chem. 1976, 41, 1485; (c) Sharpless, K. B.; Young, M. W.
J. Org. Chem. 1975, 40, 947.
The real role of potassium carbonate in these reactions
has not yet been comprehended. Evidence that it does
not play the role of a base comes from the observation
that it favors the oxidation of diphenyl selenide which
¶ Only a trace amount of benzyl phenyl selenoxide has been once
obtained instead of a 12% yield when the reaction is carried for 4 h
in MeOD in place of MeOH.