Diselenide-assisted sulfuration of dienes

Article abstract of DOI:10.1016/j.tetlet.2004.10.085

Various diselenides assist in the sulfuration of dienes giving cyclic di- and tetrasulfides as main products. The reaction requires a 2-fold excess of diselenides to be efficient. Catalytic amounts of diselenides result in lower yields. This is likely due to secondary reactions (polymerization, aromatization) occurring during extended reaction times under catalytic conditions. It was verified that the sulfur-transferring properties of diselenatetrasulfides are virtually identical to those of diselenides combined with sulfur. Contrary to previous claims, not only the cyclic diselenatetrasulfide but also linear diselenatetrasulfides (RSeS_nSeR) transfer sulfur to dienes. A mechanism is proposed and its implications to the nature of diatomic sulfur are discussed.

Full text of DOI:10.1016/j.tetlet.2004.10.085

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 9181–9184
Diselenide-assisted sulfuration of dienes
Andrzej Z. Rys,a Yihua Hou,a Imad A. Abu-Yousef b and David N. Harppa,*
aDepartment of Chemistry, McGill University, Montreal, Quebec, Canada H3A 2K6
bDepartment of Chemistry, American University of Sharjah, PO Box 26666, Sharjah, United Arab Emirates
Received 21 September 2004; revised 15 October 2004; accepted 18 October 2004
Available online 2 November 2004
Abstract—Various diselenides assist in the sulfuration of dienes giving cyclic di- and tetrasulfides as main products. The reaction
requires a 2-fold excess of diselenides to be efficient. Catalytic amounts of diselenides result in lower yields. This is likely due to
secondary reactions (polymerization, aromatization) occurring during extended reaction times under catalytic conditions. It was
verified that the sulfur-transferring properties of diselenatetrasulfides are virtually identical to those of diselenides combined with
sulfur. Contrary to previous claims, not only the cyclic diselenatetrasulfide but also linear diselenatetrasulfides (RSeSnSeR) transfer
sulfur to dienes. A mechanism is proposed and its implications to the nature of diatomic sulfur are discussed.
Ó 2004 Elsevier Ltd. All rights reserved.
Singlet diatomic sulfur is generally seen as a selectively
reacting species converting dienes into disulfides in a
Diels–Alder fashion.1 One of the more interesting ways
to transfer sulfur to dienes 1 was reported by Schmidt
and Go¨rl.2 It involves use of a cyclic seven-membered
ring 1,4-diselenatetrasulfide 2 as a 2-sulfur transfer rea-
gent. Upon heating, heterocycle 2 contracts, releasing a
two-sulfur unit (possibly 1S2) subsequently trapped by a
diene 1 (R = alkyl, aryl) providing 3 accompanied by
diselenide 4 (Eq. 1).
sulfides as unsuitable for sulfur transfer purposes, no de-
tails were provided.
Recently, we have developed a simple and efficient meth-
od to prepare both cyclic 2 and acyclic diselenatetrasul-
fides, 5a–f from corresponding diselenides 4, 6a–f and
trityl thiosulfenyl chloride Ph3CS2Cl (Eq. 2).3
This easy access to such interesting compounds allowed
us to study them as sulfur transfer reagents; we have
examined this area and report our findings. In a compar-
ative study, chlorobenzene solutions of 2,3-diphenylbuta-
diene (1A) and various 1,4-diselenatetrasulfides 5 were
refluxed under conditions similar to these reported by
Schmidt (Eq. 3). To maximize the yields, a 2-fold excess
diselenatetrasulfides 5 to diene was used instead of stoi-
chiometric amounts; the results are presented in Table 1.
S
S
R
R
R
R
Se
Se
Se Se
S
S
+
+
ð1Þ
4
1
3
2
While the reaction was not particularly efficient and re-
sulted in the formation of other products like thiophe-
nes, for years it has been considered a model method
for generating diatomic sulfur.1
a R' = Ph
b R' = PhCH2
Ph3CS2Cl
c R' = O2NC6H4CH2
R'SeSeR'
6a-f
R'SeS2SeR'
ð2Þ
d R' = MeOC6H4CH2
e R' = MeOC6H4
f R' = O2NC6H4
5a-f
We were interested to extend the original Go¨rl and Sch-
midt method to linear diselenatetrasulfides. Although
the authors commented on the linear 1,4-diselenatetra-
The most striking observation was the significant
formation of trapped products in the case of linear
diselenatetrasulfides 5 (entries 1–3). Yields of cyclic
disulfides exceeding 40% were typically observed and
were higher than those obtained for cyclic diselenatetra-
sulfide 2.
Keywords: Sulfur transfer reactions; Diselenatetrasulfides; Diatomic
sulfur; Cyclic disulfides.
*
Corresponding author. Tel.: +1 514 398 6685; fax: +1 514 398
3797; e-mail: david.harpp@mcgill.ca
0040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.10.085

9182
A. Z. Rys et al. / Tetrahedron Letters 45 (2004) 9181–9184
Table 1. Sulfuration of 2,3-diphenylbutadiene (1A, R = Ph) with diselenatetrasulfides. Yields determined with bibenzyl as the internal standard
Entry
Time [h]
Diselenatetrasulfide
Equivalents of 5 (2)
% 3 + 7
3/7
% unreacted diene
% of 8 and other products
1
2
3
4
4
4
5a (PhSeS)2
5b (PhCH2SeS)2
5c (O2N–C6H4CH2SeS)2
2
2
2
42
56
44
5
10
10
10
9
48
35
55
2
Se
S
S
4
4
2
2
51
0.7
21
28
Se
Se
5a
6
1.5
4
1
1
23
32
0.5
0.7
75
49
2
19
S
S
2
Se
a Results obtained from the reaction repeated under the reported conditions.
Ph
S
Ph
Ph
Ph
Ph
Ph
Ph
R'Se-SS-SeR'
S
S
S
S
other
+
+
S
+
∆
products
ð3Þ
Ph
S
3A
7A
8A
1A
100
90
80
70
60
50
40
30
20
10
0
In our hands, sulfuration of 2,3-diphenylbutadiene (1A)
with 2 gave only 23% of sulfur transfer products 3A and
7A (entry 5) with not much better results after 4h (entry
6).4 Curiously, it was only under our conditions that a
substantial amount of 3A formed (entry 4).
Se2 + S8
PhC H2SeSeCH2Ph + S8
Ph
Ph
Se2S2
PhC H2SeSSSeCH2Ph
%
Another remarkable difference between cyclic and acy-
clic diselenatetrasulfides was the relatively high yield
of tetrasulfide (7A) using reagent 2. These yields were
expressed as disulfide/tetrasulfide ratios (determined
by NMR) as the compounds are difficult to separate.
Although important mechanistically, a high 3/7 ratio is
not essential from a preparative point of view as
tetrasulfides 7 are easily converted to disulfides 3 by a
treatment of the crude reaction mixture with triphenyl-
phosphine as required.5
Ph
S
S
Ph
0
50
100
150
200
250
300
min
Figure 1. Sulfuration with pure PhCH2SeSSSeCH2Ph (5b) and with a
mixture of PhCH2SeSeCH2Ph (6b) + elemental sulfur.
In general, diselenatetrasulfides are thermally unstable.
Their decomposition concomitant with the formation
of various selenachalcogenides and S8 led to the premise
that S2 could be generated that way.2 Using 1H NMR to
follow the reaction progress, we observed that the bulk
of diselenatetrasulfides were transformed into the corre-
sponding diselenides in less than 4min, that is, long be-
fore much of the diene reacted. In spite of that, the
sulfuration continued suggesting that the diselenides
might be also involved in that process. To test this poss-
ibility, 2,3-diphenylbutadiene (1A) was treated with a
mixture of 2equiv of dibenzyl diselenide (6b) and 4equiv
of elemental sulfur (building blocks of diselenatetra-
sulfide 5b). The results compared to those obtained with
compound 5b are shown in Figure 1. Surprisingly, the
results with original sulfur transfer reagent were almost
identical to those where a mixture of dibenzyl diselenide
and elemental sulfur was used. This can be reasonably
explained by invoking a very rapid convergence of the
reactive selenium species in both reaction mixtures.
the Se–Se bond of diselenides was also observed but to
a lesser extent resulting in a formation of up to 10% of
various RSeSnSeR species. While all diselenides includ-
ing the dibenzyl ones, underwent sulfur insertion to
varying degrees, only the latter were observed to lose
selenium. Thus, upon loss of selenium, diselenide 6b
equilibrated with the corresponding monoselenide
PhCH2SeCH2Ph to reach 26% and 32%, respectively,
in both reaction mixtures.
The optimal sulfuration conditions employ 2equiv of
diselenides. There was no benefit in increasing this
amount. On the other hand, 1equiv of diselenide re-
sulted in both lower reaction rate and ca. 5% lower over-
all yield. Catalytic amounts of diselenides gave expected
products but were inefficient, resulting in a lower sulfu-
ration rate and more importantly, in lower yields. A
comparison of the efficiencies of various diselenides as
sulfur transfer reagents under optimized conditions is
presented in Figure 2.
Thus, a major part of diselenatetrasulfide 5b converted
rapidly (in less than 5min) to the corresponding disele-
nide 6b. The reverse process in which sulfur inserted to
The slowest rate (and lowest yield) was observed in the
case of cyclic diselenide 4. Benzylic diselenides 6b and

A. Z. Rys et al. / Tetrahedron Letters 45 (2004) 9181–9184
9183
with 2equiv of diethyl disulfide. Under our conditions,
diene 1A remained unchanged.
60
50
40
30
20
10
0
Finally, the efficiency of various diselenides in sulfur
transfer reactions was tested with dienes 1A–D (Eq. 5).
The results are shown in Table 2.
6b
6c
6e
6f
4
R
R
R
R
R
R''SSR''
SR''
SeR'
SR''
SR''
R''SSR''
hν
ð4Þ
R'SeSeR' cat.
R
hν
0
50
100
150
200
250
300
350
R
R
R
[min]
PPh3
6a-f, S8
PhCl, ∆
S
Sn
S
S
R1
R1
R1
Figure 2. The combined yields of di- and tetrasulfide obtained from
the sulfur transfer in the presence of various diselenides 6b, 6c, 6e, 6f,
and 4.
ð5Þ
1A-D
3A-D
n = 1, 3
A: R, R1 = Ph; B: R, R1 = PhCH2, C: R, R1 = Me;
D: R = Me2C=CHCH2CH2; R1 = H
6c were the fastest. Aromatic diselenides 6e and 6f were
less active than the benzylic ones but significantly more
so than 4. Interestingly, although the substituents at-
tached to the aromatic ring did not significantly affect
the reaction rate, the sulfuration yields were determined
by them. Nitro-substituents (6c and 6f) resulted in a de-
crease as compared to their more electron-rich counter-
parts 6b and 6e.
The sulfuration yields were in the 40–65% range with
dienes containing aromatic groups showing the best re-
sults (including a new dibenzyl substituted disulfide
3B).9 A relatively low yield of disulfide 3C might be ex-
plained by a high volatility of the corresponding diene
1C preventing it from reacting efficiently under the ap-
plied conditions.
Similarly as in the case of diselenatetrasulfides, signifi-
cant amounts of other products formed. Their 1H
NMR spectra show broad peaks in 3.9–4.3ppm region
suggestive of some oligomeric species. The addition of
a 24-equivalent excess of sulfur blocked oligomer pro-
duction and resulted in the formation of cleaner prod-
ucts. For the efficient separation of a desired product
by column chromatography the diselenide should have
a different polarity from that of the product (for the
choice of diselenides, see Table 2).
The role of light in sulfuration of dienes with diselenide/
S8 mixtures was tested. To this end, two reactions were
carried out; one in the dark and another one irradiated
by a 300W sun lamp (50cm from the flask). In fact, both
rate and yield were virtually the same as those of the
control experiment indicating a marginal role of light
in the sulfuration.
Diselenides are known to form radicals but they add to
dienes with difficulty.6 However, when a photolytic
addition to dienes is carried out in the presence of disul-
fides, the formation of mixed sulfide selenide adducts re-
sults.6 As the C–Se bonds dissociate easily, the mixed
adducts were easily converted to corresponding thio-
ethers (Eq. 4).7
S
S
S
10
An attempt to extend the method to norbornene 9 was
thwarted by a low, 25% yield of the corresponding
In our control experiment, paralleling dithiolation re-
ported by Ogawa et al.,7 we replaced elemental sulfur
Table 2. Yields of sulfuration products obtained from various dienes and norbornene8
Sulfuration product
6
Equivalents of sulfur
Time [h]
Yielda,b [%]
Ph
a
c
e
24
6
5
5
5
50
42 (53)
(48)
S
S
3A
3B
Ph
6
PhCH2
a
c
e
24
6
2
2
2
67
S
S
(50)
(65)
PhCH2
6
e
e
6
24
2
2
38
41
S
S
3C
3D
e
e
6
24
2
2
(28)
36
S
S
a Tetrasulfides were desulfurized with PPh3 prior to the purification.
b Values in brackets show the amounts of product in cases when complete purification could not be achieved—determined with the internal standard.

9184
A. Z. Rys et al. / Tetrahedron Letters 45 (2004) 9181–9184
R
R
R
S
Sn
∆
S
Sn
S8
R' Se Se R'
R'-Se
R
S
Sn
S8
D
R' Se
A
R
R
R
Se
'R Se
R'Se
R'-Se
R'-Se
R'
Se R'
B
C
R
S
Sn
Se
R' Se Sn Se R'
S
Sn
R'
Se R'
R' Se
R' Se
Scheme 1.
trisulfide 10 obtained under optimized conditions (2equiv
of PhSeSePh and 24-excess of elemental sulfur).10
References and notes
1. Steliou, K.; Gareau, Y.; Harpp, D. N. J. Am. Chem. Soc.
1984, 106, 799–801; Steliou, K.; Salama, P.; Brodeur, D.;
Gareau, Y. J. Am. Chem. Soc. 1987, 109, 926–927; For
reviews see: Harpp, D. N. Phosphorus, Sulfur, and Silicon
and Rel. El 1997, 120, 41–59.
2. Schmidt, M.; Go¨rl, U. Angew. Chem., Int. Ed. Engl. 1987,
26, 887–888.
3. Hou, Y.; Rys, A. Z.; Abu-Yousef, I. A.; Harpp, D. N.
Tetrahedron Lett. 2003, 44, 4279–4282.
4. Likely, there is a typographic error in Ref. 2, as the
reaction times for 2,3-dimethylbutadiene and 2,3-diphen-
ylbutadiene seem to be reversed. The latter, known for its
lower reactivity in D–A reactions, was reportedly treated
with compound 2 for only 1.5h while the former for 4h.
5. Rys, A. Z.; Harpp, D. N. Tetrahedron Lett. 1998, 39,
9139–9142.
6. Ogawa, A.; Tanaka, H.; Yokoyama, H.; Obayashi, R.;
Yokoyama, K.; Sonoda, N. J. Org. Chem. 1992, 57, 111–
115.
The sulfuration appears to be a thermally controlled
homolytic process, that is, only slightly enhanced by
light (up to 5% higher sulfur transfer yields). Most
likely, it implies a cleavage of the diselenides to form
reactive radicals (R0SeÆ) as reported by Ogawa6,7 and
Chu.11 Other radicals likely do not form as we do not
observe any R0SnR0-type product. The selenide radicals
can generate radical sulfur species that might follow
three reaction paths (Scheme 1). In path A, sulfur is
transferred to dienes to form the expected sulfuration
products. In path B, unsymmetric addition results in
the formation of a mixed adduct similarly to what was
reported by Ogawa et al.7 This path seems to be redi-
rected to the formation of sulfur transfer products when
massive amounts of elemental sulfur are added. Finally,
in path C, the sulfur radical species reversibly converts
to diselenapolysulfides. It is by following this path that
diselenatetrasulfides convert to radical sulfur species.
Benzylic diselenides tend to lose selenium upon heating
(path D). It seems that this process is not a prerequisite
in sulfuration as even diselenides that do not deselenize
(aromatic 6a, 6e and 6f and cyclic 4) efficiently assist in
sulfuration.
7. Ogawa, A.; Obayashi, R.; Sonoda, N.; Hirao, T. Tetra-
hedron Lett. 1998, 39, 1577–1578.
8. Typical procedure: a mixture of a diene 0.4mmol, disele-
nide 0.8mmol and 12.8mmol (410mg) of sulfur in 4mL of
PhCl was refluxed for 2–5h under nitrogen. The reaction
mixture was poured into 15mL of MeOH, cooled in
the freezer and filtered. Five millimoles of PPh3 were
dissolved in the solution upon warming. After 10min the
excess of PPh3 was eliminated by addition of 1.2mmol of
sulfur, warming and stirring at rt for 1h. The solvent was
then removed, the residue was extracted five times with
hexanes (total ca. 25mL). Solvent was removed and the
residue was chromatographed with 0–1% of EtOAc in
hexanes.
9. Properties of 3B: white solid; 79mg (67%) yield; mp 37–
38°C; 1H NMR (CDCl3): d 3.31 (s, 4H); 3.62 (s, 4H); 7.1–
7.4 (m, 10H); 13C NMR (CDCl3): d 32.9; 39.9; 126.4;
128.3; 128.6; 130.4; 138.6. HRMS: calc. for C18H18S2
298.08499, obt. 298.08437.
Once formed, the sulfuration products 3 and 7 can un-
dergo some subsequent reactions decreasing their yield.
Aromatization of 3 and 7 under the sulfuration condi-
tions was reported earlier.5 In this work we have usually
observed very small amounts of 7. A separate experi-
ment helped us to determine that this effect is due to
presence of diselenides that attack 7 converting it to
thiophene and some unidentified polymeric species with
only 20% left after keeping it under typical reaction
conditions.
10. Norbornene trisulfide is known to undergo polymeriza-
tion, likely initiated when treated with PPh3; see: Baran,
T.; Duda, A.; Penczek, S. J. Polym. Sci. 1984, 22A, 1085–
1095.
Acknowledgements
11. Chu, J. Y. C.; Lewicki, J. W. J. Org. Chem. 1977, 42,
2491–2493.
We thank the NSERC of Canada for financial support.