Synthesis of Disulfides from the Palladium(0)-Catalyzed Reactions of Sulfenyl Halides and Organostannanes

Full text of DOI:10.1021/jo951378w

J . Org. Chem. 1996, 61, 1533-1536
1533
Sch em e 1
Syn th esis of Disu lfid es fr om th e
P a lla d iu m (0)-Ca ta lyzed Rea ction s of
Su lfen yl Ha lid es a n d Or ga n osta n n a n es
J ohn C. Cochran,* Susan R. Friedman, and
J ohn P. Frazier
Department of Chemistry, Colgate University,
Hamilton, New York 13346
Received J uly 27, 1995
(Revised Manuscript Received
December 8, 1995)
In tr od u ction
Palladium(0 or II)-catalyzed coupling reactions have
acquired a well-established status in the synthetic or-
ganic repertoire.1 A recent survey by Farina and co-
workers2 revealed that a majority of the transition metal-
mediated cross-coupling reactions reported in 1992
involved palladium catalysis. While most examples
involve the formation of carbon-carbon bonds, pal-
ladium-catalyzed formation of carbon-silicon, carbon-
germanium, and carbon-tin bonds have also been re-
ported. These latter reactions usually involve palladium-
catalyzed additions of an organometallic or organobi-
metallic species to a carbon-carbon unsaturated system.1c
Labadie,3 in 1989, showed that the palladium-catalyzed
cross-coupling process could be extended to reactions
between vinylstannanes and sulfonyl chlorides to give
vinyl sulfones. The reaction was characterized by short
reaction times, high yields, and varied functionality.
Similarly, Quintard and co-workers4 reported a pal-
ladium-catalyzed coupling reaction between acyl chlo-
rides and vinylstannanes to produce R,â-unsaturated
ketones. Since the sulfur of the sulfonyl chloride and the
carbon of the acyl halide exhibit similar electrophilic
character in the displacement of the stannyl group from
the vinylstannane, and since the sulfur of sulfenyl halides
is electrophilic in cleavage reactions with vinylstan-
nanes,5 we chose to determine whether vinyl sulfides
could be obtained from the palladium-catalyzed coupling
of sulfenyl chlorides and vinylstannanes. To the con-
trary, the palladium-catalyzed reaction between sulfenyl
chlorides and vinylstannanes leads to high yields of the
corresponding disulfide, a trialkytin chloride, and vinyl
chloride. We show that a tetraalkylstannane or a
hexaalkyldistannane is both necessary and as efficacious
in the production of the disulfide.
followed by exchange of a ligand on palladium with a new
ligand, usually from a stannane, and finally a reductive
elimination of two coupled ligands as the product.1 We
envision a similar process for the formation of disulfides,
as outlined in Scheme 1. The very reactive sulfenyl
halide, 1-6a , undergoes an oxidative addition to pal-
ladium analogous to that of an aryl halide in the usual
Stille reaction. This is followed by a ligand exchange step
in which a second sulfur bonds to palladium, releasing a
molecule of chlorine. This step is analogous to ligand
exchange in which a substituent on tin replaces the
halogen and a molecule of tin halide is formed. The two
sulfur groups then undergo reductive elimination to form
the disulfide, 1-5b (eq 1). The chlorine formed in the
ligand exchange step is then available to reverse the
process by converting the disulfide back to sulfenyl
chloride, a well-known reaction6 (eq 2). However, orga-
notin compounds are very reactive to electrophilic cleav-
age by halogens7 and effectively scavenge the chlorine
by reaction to give the chlorostannane and an alkyl
chloride (eq 3). The results of reactions with several
chlorine scavengers are given in Table 1.
Resu lts a n d Discu ssion
The focus of this study is the palladium(0)-catalyzed
coupling of sulfenyl halides, resulting in the production
of symmetrical disulfides. Stille-type coupling reactions
are characterized by a sequence of steps involving first
an oxidative addition of the substrate to palladium,
2RSCl Pd(0)8 (RS)2 + Cl
(RS)2 + Cl2 f 2RSCl
(1)
(2)
(3)
R4′Sn + Cl2 f R3′SnCl + R′
(1) For reviews, see: (a) Stille, J . K. Angew. Chem., Int. Ed. Engl.
1986, 25, 508. (b) Scott, W. J . ; McMurry, J . E. Acc. Chem. Res. 1988,
21, 47. (c) Mitchel, T. N. Synthesis 1992, 158.
(2) Farina, V.; Krishman, B.; Marshall,D. R.; Roth, G. P. J . Org.
Chem. 1993, 58, 5434.
(3) Labadie, S. S. J . Org. Chem. 1989, 54, 2496.
(4) Parrain, J .-L.; Beaudet, I.; Ducheˆne, A.; Watrelot, S.; Quintard,
J .-P. Tetrahedron Lett. 1993, 34, 5445.
Using 2,4-dinitrobenzenesulfenyl chloride as the pro-
totype, the first entry in Table 1 shows that a low but
measurable yield of disulfide can be obtained from a THF
(5) (a) Wardell, J . L. J . Chem. Soc., Dalton Trans. 1975, 1786. (b)
Wardell, J . L.; Ahmed, S. J . Organomet. Chem. 1974, 78, 395. (c)
Kutyrev, G. A.; Kapura, A. A.; Cherkasov, R. A.; Pudovik, A. N. Zh.
Obshch. Khim. 1985, 55, 1919.
(6) For a review of sulfenyl halide syntheses, see: Ku¨hle, E.
Synthesis 1970, 561.
(7) Pereyre, M.; Quintard, J .-P.; Rahm, A. Tin in Organic Synthesis;
Butterworths: London, 1987; p 314.
0022-3263/96/1961-1533$12.00/0 © 1996 American Chemical Society

1534 J . Org. Chem., Vol. 61, No. 4, 1996
Notes
Ta ble 1. Ch lor in e-Sca ven gin g Rea ction s
F igu r e 1. Plot of percent reactions of 2a and 6a with Me4Sn
and Pd(0) in THF-d8 vs time. See Experimental Section for
details.
1
mane (14), H NMR spectra of reactions run in THF-d8
contain a singlet peak at 3.03 ppm, assigned to methyl
chloride. This peak disappears from the spectrum when
the solvent is removed under vacuum or the solution
flushed with argon. Also this peak does not appear in
the product spectra when the chlorine scavengers are
hexamethyldistannane (9) or tributylvinylstannane (10).
Production of the related product, chlorotrimethylstan-
nane, can be directly correlated with the decrease of its
precursor, tetramethylstannane or hexamethyldistan-
nane.
The ligand exchange step of the catalytic cycle has been
shown to be the rate-determining step.10 In the Stille
coupling reaction this step constitutes a transmetalation
by electrophilic cleavage of a carbon-tin bond. The
reactivity of groups on tin follows the usual order of ease
of electrophilic cleavage.11 In the reactions at hand, this
step involves exchange of a chlorine for an RS group from
the sulfenyl halide. The chlorine atoms then form a
molecule of chlorine. The chlorine is either scavenged
by the stannane or converts a molecule of disulfide back
to the sulfenyl chloride.
The proposal of the ligand exchange step is based,
experimentally, on two considerations. Figure 1 shows
a plot of 1H NMR peak areas for chlorotrimethylstannane
and bromotrimethylstannane as a function of time.
These products arise from the coupling of 2-nitroben-
zenesulfenyl chloride (2a ) and 2-nitrobenzenesulfenyl
bromide (6a ), respectively, in the presence of tetrameth-
ylstannane. The plot shows that the sulfenyl chloride is
both more reactive and proceeds to a higher yield than
the bromide. Since bromine is a softer ligand, it is to be
expected that the ligand exchange step would be slower
in the case of the sulfenyl bromide. Second, a comparison
a
Molar ratio of sulfenyl halide to chlorine scavenger. b Disulfide
product. c Addition product, mp 116-117 °C [lit.13 mp 115-117
°C].
solution of the sulfenyl chloride and the palladium
catalyst. This yield is about doubled when argon is
bubbled through the reaction mixture. Addition of an
organostannane to the system results, in most cases, in
a dramatic increase in the yield of disulfide. Compounds
investigated are tetramethylstannane (7), tetraethyl-
stannane (8), hexamethyldistannane (9), tributylvinyl-
stannane (10), tetracyclohexylstannane (11), tetraphen-
ylstannane (12), cyclohexene (13), and tetramethylger-
mane (14). Comparison of the yields in entries 3, 6, and
12 to those in entries 4, 7, and 13, respectively, suggests
that an excess of chlorine scavenger has little effect on
the yield of disulfide. The low yields in entries 9 and 10
are undoubtedly due to low solubility of stannanes 11
and 12 in THF. With cyclohexene (13), the primary
reaction is addition of the sulfenyl chloride to the double
bond.8 No 1,2-dichlorocyclohexane was detected in the
reaction mixture. Also the extent of reaction is consider-
ably diminished when tetramethylgermane (14) is the
scavenger compared to tetramethylstannane. This is
consistent with their relative reactivities to chlorine.9
Several observations are consistent with the stepwise
process illustrated in Scheme 1. When the chlorine
scavenger is tetramethylstannane (7) or tetramethylger-
(8) Wigzell, J . M.; Wardell, J . L. J . Organomet. Chem. 1982, 235,
37.
(10) Negishi, E.; Takahashi, T.; Baba, S.; Van Horn, D. E.; Okukado,
N. J . Am. Chem. Soc. 1987, 109, 2393.
(9) Coates, G. E. Organo-Metallic Compounds; J . Wiley: New York,
1960; p 164.
(11) Reference 1a, p 520.

Notes
J . Org. Chem., Vol. 61, No. 4, 1996 1535
Ta ble 2. Yield s of Disu lfid es
Exp er im en ta l Section
Gen er a l. Tetrakis(triphenylphosphine)palladium(0), THF-
d8, tributylvinylstannane, hexamethyldistannane, tetraphenyl-
stannane, 2,4-dinitrobenzenesulfenyl chloride, triphenylmethane-
sulfenyl chloride, 2,4,5-trichlorophenyl disulfide, 2-nitrophenyl
disulfide, and 2-nitrobenzenesulfenyl chloride were obtained
from Aldrich and used without further purification. Tetra-
methylstannane, tetraethylstannane, and tetramethylgermane
were obtained from Gelest and used without further purification.
Tetracyclohexylstannane was prepared from the Grignard re-
agent of chlorocyclohexane and tin(IV) chloride. THF was
distilled from sodium/benzophenone ketyl. Carbon tetrachloride
was dried over molecular sieves. 1H and 13C NMR spectra were
obtained on a Bruker AC-250 spectrometer at 250 and 62.9 MHz,
respectively, and referenced to TMS and CDCl3 or acetone-d6,
respectively. Elemental analyses were performed by Galbraith
Laboratories, Knoxville, TN. All melting points are uncorrected.
2,4,5-Tr ich lor oben zen esu lfen yl Ch lor id e (4a ). In a 250
mL flask, fitted with a reflux condenser and gas dispersion tube,
were placed 1.53 g (3.60 mmol) of 2,4,5-trichlorophenyl disulfide,
30 mg of iodine and 100 mL of dry carbon tetrachloride.
Chlorine gas was bubbled through the solution until the color
turned to a pale yellow. The excess chlorine and the carbon
tetrachloride were removed on a rotary evaporator. The result-
ing solid was purified by column chromatography on silica gel
using methylene chloride as the eluent. The sulfenyl chloride
was eluted after unreacted disulfide. The yield was 1.45 g (5.85
mmol, 81%): mp 35.5-37.0 °C; 1H NMR (CDCl3) δ 7.50 (s, 1H),
7.65 (s, 1H); 13C NMR (CDCl3) δ 128.97, 129.10, 130.66, 132.51,
133.67, 135.03. Anal. Calcd for C6H2SCl4: C, 29.06; H, 0.81.
Found: C, 29.12; H, 0.80. The sample was protected from light
during workup and on storage.
2-Nitr oben zen esu lfen yl Br om id e (6a ). In a 500 mL flask,
fitted with a reflux condenser, addition funnel, and magnetic
stirrer, were placed 5.00 g (16.2 mmol) of 2-nitrophenyl disulfide
and 200 mL of dry carbon tetrachloride. To this mixture were
added, at room temperature, 2.60 g (16.2 mmol) of bromine in
50 mL of dry carbon tetrachloride and a catalytic amount of
iodine (30 mg). The mixture was heated to reflux. After 24 h,
a sample was removed and evaporated to dryness. An 1H NMR
spectrum indicated about 40% conversion to product. After 72
h, the reaction was stopped and the volatile components were
a
Yield as based on recovery of starting sulfenyl chloride 5a .
The coproducts are Me3SnBr (1H NMR 0.72 ppm, s, 2J SnH ) 65.1/
b
62.3 Hz) and MeBr (1H NMR 2.68 ppm, s).
of entries 3 or 4 with 8 in Table 1 shows that the
vinylstannane (10) gives a slightly lower yield than
tetramethylstannane (7). Vinyl groups, bonded to tin,
are about 700 times more reactive in transmetalation
than are alkyl groups.11 Thus if the tin compound is
involved in the ligand exchange process, as opposed to
acting as a chlorine scavenger, 10 should give a compa-
rable yield to 7, which it does not.
Table 2 list yields for the six sulfenyl halides using 7
as the halogen scavenger. It is clear that the reaction is
facilitated by electron withdrawing groups operating near
the sulfenyl halide. Triphenylmethanesulfenyl chloride
was significantly less reactive, and to obtain a decent
yield, a longer reaction time and higher temperature were
necessary.
removed on
a rotary evaporator. The resulting solid was
recrystallized from carbon tetrachloride yielding unreacted
disulfide. The mother liquor was evaporated to dryness and
yielded 5.40 g (71%) of 2-nitrobenzenesulfenyl bromide: 1H NMR
(CDCl3) δ 7.44 (t, 1 H), 7.73 (t, 1 H), 8.04 (d, 1 H), 8.31 (d, 1 H);
13C NMR (CDCl3) δ 125.50, 126.68, 129.09, 134.52, 135.37,
137.70; mp 85.5-87 °C (lit.12 mp 83-85 °C).
Syn th etic Cou p lin g Rea ction s. Com p ou n d s 1-4. Gen -
er a l. In a 10 mL pear-shaped flask were mixed 1.00 mmol of
the sulfenyl halide, 0.500 mmol of tetramethylstannane, 2 mol
% of tetrakis(triphenylphosphine)palladium(0), and 4 mL of dry
THF. The catalyst was added last, and an argon atmosphere
was maintained in the flask. The contents of the flask were
stirred for 4 h at room temperature. Further workup is
described below.
1
Finally it should be noted that the H NMR spectrum
of each crude reaction product, except 5b, gave a singlet
peak at about 3.5 ppm. This peak is attributed to a MeS
group and constitutes less than 3% of the product
mixture. The presence of this peak suggests that the
normal Stille coupling reaction (eq 4) is present but
competes poorly with the disulfide coupling process. This
peak was absent when the palladium(0) catalyst was
withheld and also did not appear in reactions with 9 as
the stannane.
2,4-Din itr op h en yl Disu lfid e (1b). During the reaction
period a yellow precipitate appeared. After 4 h the reaction
mixture was cooled to -2 °C for 0.5 h. The product was filtered
on a sintered glass funnel and dried to constant weight, yielding
188 mg (95%) of 1b: 1H NMR (DMSO-d6) δ 8.17 (d, 2H), 8.43
(dd, 2H), 9.01 (d, 2H); 13C NMR (DMSO-d6) δ 121.45, 128.43,
128.70, 140.71, 145.42, 145.97; mp 307 °C dec [lit.13 mp 280 °C
dec].
Isola t ion a n d Ch a r a ct er iza t ion of Com p ou n d s 2-4b .
After 4 h the THF was removed on a rotary evaporator and the
resulting solid dissolved in methylene chloride. This solution
was washed twice with 2 mL portions of water and dried over
magnesium sulfate, and the methylene chloride was removed
on a rotary evaporator.
RSCl + Me4Sn Pd(0)8 RSMe + Me3SnCl
(4)
In summary, we report a unique coupling process in
which sulfenyl halides are converted to disulfides in
excellent yield. The overall process is similar to a Stille
coupling reaction but differs in the function of the
stannane.
(12) Hunter, W. H.; Sorenson, B. E. J . Am. Chem. Soc. 1932, 54,
3368.
(13) Kharasch, N.; Wehrmeister, H. L.; Tigerman, H. J . Am. Chem.
Soc. 1947, 69, 1612.

1536 J . Org. Chem., Vol. 61, No. 4, 1996
Notes
2-Nitr op h en yl Disu lfid e (2b). Removal of the methylene
chloride in a tared flask yielded 135 mg (88%) of 2b. Further
purification was obtained by recrystallization from benzene: 1H
NMR (CDCl3) δ 7.41 (dd, 2H), 7.61 (dd, 2H), 7.86 (d, 2H), 8.34
(d, 2H); 13C NMR (CDCl3) δ 126.37, 127.00, 127.17, 134.52,
134.88, 146.10; mp 193-194 °C [lit.12 mp 192.5-193.5 °C]. The
yield of 2b from 2-nitrobenzenesulfenyl bromide (6a ) was 83%.
4-Nitr op h en yl Disu lfid e (3b). Removal of the methylene
chloride in a tared flask yielded 129 mg (84%) of 3b. Further
purification was obtained by recrystallization from absolute
ethanol: 1H NMR (CDCl3) δ 7.62 (d, 4H, J ) 9.00 Hz), 8.20 (d,
4H, J ) 8.98 Hz); 13C NMR (CDCl3) δ 124.39, 126.32, 143.99,
146.90; mp 182-182.5 °C [lit.14 mp 182-183 °C].
2,4,5-Tr ich lor op h en yl Disu lfid e (4b). Removal of the
methylene chloride in a tared flask yielded 180 mg (86%) of 4b.
Further purification was obtained by recrystallization from
benzene: 1H NMR (CDCl3) δ 7.50 (s, 2H), 7.61 (s, 2H); 13C NMR
(CDCl3) δ 128.78, 130.91, 131.07, 132.29, 132.49, 134.07; mp
144.5-145 °C [lit.15 mp 146.5-147.5 °C].
Samples 5-7 contained starting sulfenyl chloride 5a and
samples 10-13 contained the product triphenylmethyl disulfide
(5b). The solvent from the pooled samples 10-13 was evapo-
rated to constant weight in a tared flask and yielded 170 mg
(62%) of product. On the basis of recovered sulfenyl chloride,
the yield was 73%: 1H NMR (CDCl3) δ 7.29 (m, 36 H); 13C NMR
(acetone-d6) δ 69.94, 128.35, 129.12, 129.63, 149.48; mp 158-
159 °C [lit.16 mp 156-157 °C].
P a lla d iu m -Ca ta lyzed Rea ction s of 2,4-Din itr oben zen e-
su lfen yl Ch lor id e w ith Ch lor in e Sca ven ger s 7-14.
A
solution of 1.00 mmol of 1a , 1.00 or 0.500 mmol of chlorine
scavenger, 7-14, 2 mol % of tetrakis(triphenylphosphene)-
palladium(0), and 4 mL of dry THF were mixed in a flask. The
mixture was stirred overnight, and the solid disulfide (1b) was
filtered on a Hirsch funnel. The solid was washed with THF
and dried in a vacuum desiccator. Yields are listed in Table 2.
The THF filtrates were evaporated to dryness, and the 1H NMR
spectrum of the solid showed the presence of the alkyltin or
germanium halide.
Tr ip h en ylm eth yl Disu lfid e (5b). A solution of 311 mg
(1.00 mmol) of triphenylmethanesulfenyl chloride, 89 mg (0.500
mmol) of tetramethylstannane, 23 mg (∼2 mol %) of tetrakis-
(triphenylphosphine)palladium(0), and 4 mL of dry THF was
mixed in a 10 mL flask fitted with a magnetic stirrer and septum
cap. An atmosphere of argon was maintained. The flask was
heated in a water bath to 40 °C for 24 h and then the THF
removed on a rotary evaporator. The black solid was dissolved
in methylene chloride and chromatographed on a silica gel
column (3 × 20 cm) using methylene chloride as the eluent.
Fifteen 7.5 mL samples were collected and analyzed by TLC.
Rea ctivity Stu d ies of 2a a n d 6a . In an NMR tube were
mixed 0.100 mmol of 2a or 6a , 0.0500 mmol of tetramethylstan-
nane, 2 mol % of tetrakis(triphenylphosphine)palladium(0), and
0.5 mL of THF-d8. The reaction was monitored by 1H NMR over
a period of 4 h. The reaction mixture was maintained at 25 °C.
The extent of reaction was determined by the normalized areas
of the tetramethylstannane, chlorotrimethylstannane, and bro-
motrimethylstannane peaks at 0.10, 0.62, and 0.72 ppm, re-
spectively. The results are shown in Figure 1.
Ack n ow led gm en t. This project was supported in
part by the Colgate University Research Council. This
support is gratefully acknowledged.
(14) Hoggarth, E.; Sexton, W. A. J . Chem. Soc. 1947, 815.
(15) Lukashevich, V. O. Dokl. Akad. Nauk SSSR 1955, 103, 627;
Chem. Abstr. 1956, 50, 5556.
(16) Sosnovsky, G.; Krogh, J . A. Liebigs Ann. Chem. 1982, 121.
J O951378W