to produce the expected Wittig product, 2-bromo-3-styryl-
thiophene, in 61% yield as the major product (Scheme 1 eq
10 mL of acetonitrile, and the solution was cooled to 0 °C.
When larger amounts of the phosphonium salts were used,
200 mL of the solvent was used to dissolve the reagents.
The base was then slowly added to solution. The reaction
mixture was then stirred for 1 h (2 h when 200 mL of solvent
was used) while allowing it to warm to room temperature.
The solvent was then evaporated, and the products were
separated from the triphenylphosphine byproduct and the
unreacted starting material, using silica gel chromatography
with hexanes as eluent, and characterized by NMR. The
purity of the products was determined by GC-MS. The E:Z
ratio was determined by integrating the alkenyl H peaks in
Scheme 1. Homocoupling vs Wittig Reaction
1
the H NMR spectrum (except in the case of the homocou-
pling reaction with 6 when the product distribution was
assigned based on the peak integrations of GC). Table 1 lists
the results of the reactions of various phosphonium salts with
bases in acetonitrile. When 2 and its 2-bromo derivative 1
were treated with base in acetonitrile, 60-70% yields of the
homocoupling products were obtained. Reaction of 2 with
NaH in acetonitrile gave nearly quantitative yields of the
product. Castle and co-workers synthesized 1,2-bis(3,3′-
benzo[b]thienyl)ethane by the Wadsworth-Emmons reaction
of diethyl 3-benzo[b]thienylphosphonate with benzo[b]-
thiophene-3-carbaldehyde using NaH as a base in 83%
yield.12 However, using the homocoupling reaction of 3 with
t-BuOK, we obtained the desired product in a comparable
78% yield. When 6 was stirred with t-BuOK in acetonitrile
for 1 h, only 29% E-stilbene was obtained.13 When 5 was
subjected to similar conditions, only 21% yield of the
homocoupling product was obtained. In both cases, large
quantities of unreacted starting phosphonium salts were
recovered. On the basis of these observations, it appears that
the homocoupling reaction is specific to 2 and its derivatives.
However, when a larger volume of the solvent was used,
the homocoupling yields increased substantially. An ad-
ditional equivalent of the base also helped improve the yields.
For example, stilbene was obtained in 64% yield from 6
when 2 equiv of base and 200 mL of acetonitrile were used.
Similarly 30-60% of homocoupling products were obtained
from 7 and 8 when 200 mL of solvent was used. When 2
equiv of the NaH was used, quantitative yields of the
homocoupling products were obtained from 7. However,
when an electron withdrawing group was present in the
aromatic ring, e.g. 4, no homocoupling product could be
detected. This lack of reactivity could be attributed to the
lower nucleophilicity of the resulting ylide.
1). This is consistent with literature reports of the reaction
of 1 with aldehydes in THF as solvent.9 However, when 1
was treated with the base and reacted with acetophenone or
3-nitroacetophenone in acetonitrile, no products due to the
Wittig reaction could be detected. Following analysis of the
reaction mixture when 3-nitroacetophenone was used, only
the homocoupling product and triphenylphosphine were
detected (Scheme 1, eq 2). When the reaction was carried
out in tert-butyl alcohol, the expected Wittig product was
obtained (Scheme 1, eq 3). TLC analysis of the reaction
mixture for eqs 1 and 3 indicated the presence of the
homocoupling products in trace amounts. We wanted to
exploit the results in eq 2 as a convenient way to synthesize
symmetric ethenyldithiophenes.
The phosphonium salts (except 6, which was obtained
commercially) were synthesized from the corresponding
methylthiophenes. The methylthiophenes were brominated
with N-bromosuccinimide in the presence of benzoyl per-
oxide, and then treated with triphenylphosphine.9,10 For the
synthesis of compound 4, 3-methylthiophene was first
nitrated, using literature procedures,11 followed by bromi-
nation and treatment with triphenylphosphine.
The phosphonium salt 1 was found to be stable in
acetonitrile, which ruled out a direct interaction of 1 with
solvent. However, in the presence of a base, we found that
the starting material decomposed leading to the formation
of the homocoupling product and triphenylphosphine. Typi-
cally, the phosphonium salt (50-70 mg) was dissolved in
Scheme 2 describes a possible mechanism for this reaction.
The phosphonium salt reacts with the base to form the ylide
A. The ylide then reacts with another equivalent of the
phosphonium salt to form the intermediate B, which then
(12) Kudo, H.; Tedjamulia, M. L.; Castle, R. N.; Lee, M. L. J. Heterocycl.
Chem. 1984, 21, 185-192.
(9) Archer, W. J.; Cook, R.; Taylor, R. J. Chem. Soc., Perkin Trans. 2
1983, 813-819.
(10) (a) Campaigne, E. E.; Tullar, B. F. Org. Synth. 1953, 33, 96-98.
(b) Munro, D. P.; Sharp, J. T. J. Chem. Soc., Perkin Trans. 1 1980, 1718-
1723.
(11) (a) Threadgill, M. D.; Webb, P.; O’Neill, P.; Naylor, M. A.;
Stephens, M. A.; Stratford, I. J.; Cole, S.; Adams, G. E.; Fielden, E. M. J.
Med. Chem. 1991, 34, 2112-2120. (b) Snyder, H. R.; Carpino, L. A.; Zack,
J. F.; Mills, J. F. J. Am. Chem. Soc. 1957, 79, 2556-2559.
(13) GC-MS analysis of the isolated product showed one major peak.
The EI-MS of this peak corresponds to the MS of authentic stilbene samples.
1H NMR of the isolated product when compared to the spectra of authentic
compounds (The Aldrich Library of NMR Spectra, 2nd ed.; Pouchert, C.
J., Ed.; Aldrich Chemical Co., Inc.: Milwaukee, WI, 1983; Vol. 1, p 752
A-B) indicated it to be trans-stilbene while the cis-stilbene could not be
1
detected in H NMR.
(14) Vogel, A. I. Textbook of Practical Organic Chemistry, 4th ed.;
Longman Group Limited: London, UK, 1978; p 1371.
4086
Org. Lett., Vol. 8, No. 18, 2006