7
910
B. Joli6et, D. Uguen / Tetrahedron Letters 43 (2002) 7907–7911
Scheme 5. Suggested mechanism for the 1c–e–4c–e conversion.
means be considered as a possible alternative to existing
methods for preparing stilbenes unless, as demonstrated
identification of methylcyclopentane in the products
formed by treatment of 5-hexenyl p-tolylsulfone with
sodium in boiling toluene (Stetter, H.; Lehman, K. A.
Liebigs Ann. 1973, 499–507).
1
2e
by Eisch, a short UV irradiation or added nickel salts
would be used. A few interesting points emerge from
this study however. Oxygen appears to play no other
role in these reactions than to induce the formation of
an aldehyde. Further, unlike alkoxymethylsulfones, the
treatment of which by Na·Hg in aprotic medium gener-
ates almost exclusively free radicals, the reduction of
benzylic sulfones under the same conditions affords
successively a benzyl anion, a sulfone anion, then a
stilbene: an olefination process best explained by
assuming that the formed sulfone carbanion slowly
decomposes into, and couples with a carbene. Efforts
are now being made to support more firmly this view
by trapping experiments.
7. For other conditions for preparing stilbenes from a ben-
zylsulfone, see: Vedejs, E.; Dolphin, J. M.; Stolle, W. T.
J. Am. Chem. Soc. 1979, 101, 249–251 (for a closely
related procedure, though not using sulfones, see:
Katritzky, A. R.; Tymoshenko, D. O.; Belyakov, S. A. J.
Org. Chem. 1999, 64, 3332–3334).
8. Protocol for the Na·Hg reduction of sulfones in aprotic
medium: Finely ground, freshly prepared 6% Na·Hg (4.2
g) and sulfone 1d (0.64 g; 2.4 mmol) were placed in a
flask equipped with a condenser connected to an argon/
−
3
vacuum line. The flask was evacuated to 10 Torr and
then filled with argon. THF (15 ml) was added with a
syringe and the resulting suspension was perfectly
degassed (three ‘freeze and thaw’ cycles), then brought to
reflux for 5 days with magnetic stirring. After cooling,
ether (50 ml) was added and the resulting mixture was
filtered on paper. The clear filtrate thus obtained was
References
1
2
. Magnus, P. D. Tetrahedron 1977, 33, 2019–2045.
. (a) Dabby, R. E.; Kenyon, J.; Mason, R. F. J. Chem.
Soc. 1952, 4881–4882. Similar results are obtained by
using magnesium (Lee, G. H.; Lee, H. K. Tetrahedron
Lett. 1995, 36, 5607–5608); (b) for other reduction condi-
tions, see: Julia, M.; Uguen, D. Bull. Soc. Chim. Fr. 1976,
washed with brine (3×25 ml), dried (K CO ), then evapo-
2
3
rated to give a pasty solid (0.3 g). Chromatography of
this residue on silica gel (CH Cl /hexane) afforded suc-
2
2
cessively anisol 3d (102 mg; 34%), the stilbene 4d (50 mg;
7%), and finally unreacted 1d (41%). Selected data: 4d:
1
13
Mp 121°C; C NMR: 55.41, 114.19, 126.27, 127.5,
30.57, 159.09.
. (a) Guijarro, D.; Yus, M. Tetrahedron Lett. 1994, 35,
965–2968 (for a review, see: Najera, C.; Yus, M. Tetra-
5
13–518.
. (a) Horner, L.; Neumann, H. Chem. Ber. 1965, 98, 1715–
721. Since a mercury electrode was used, these authors
1
3
9
1
2
suggest that a tetralkylammonium amalgam is the effec-
tive reagent and thus that these electrochemical reduc-
tions proceed in the same way as those making use of
Na·Hg as the reducing species; the standard potential for
sodium amalgam has been estimated to ca. −2 V versus
SCE (Mairanovsky, V. G. Angew. Chem. 1976, 283–294);
hedron 1999, 55, 10547–10658). By way of illustration,
adding a mixture of t-butyl benzylsulfone and TMSCl to
excess biphenyl–lithium in THF results in the almost
quantitative formation of benzyl trimethylsilane (Ver-
peaux, J.-N.; Uguen, D., unpublished results).
1
0. (a) Chu, T. L.; Weismann, T. J. J. Am. Chem. Soc. 1956,
(b) Simonet, J.; Jeminet, G. Bull. Soc. Chim. Fr. 1971,
78, 23–26; (b) Darling, S. D.; Devgan, O. N.; Cosgrove,
1
08, 2754–2760.
4
. The rate constant for the reduction of a primary alkyl
R. E. J. Am. Chem. Soc. 1970, 92, 696–697; (c) Kamata,
S.; Uyeo, S.; Haga, N.; Nagata, W. Synth. Commun.
1973, 3, 265–272; (d) solutions of the TMBLi reagent
were prepared as described in Ref. 10b by stirring for 2 h
at −15°C trimesitylborane (1 mmol) and lithium wire (1
mmol) in THF (10 ml) until disappearance of the metal,
then cooling to −78°C before being treated by a solution
of the sulfone 1c (1 mmol) and an internal standard
(decane) in THF (5 ml), with (or without) added t-BuOH
(1.1 mmol). In both cases, the reaction mixture was
radical by sodium/naphtalene in THF has been estimated
9
−1 −1
to 1.6×10 mol l
s
(Barton, J. F.; Garst, F. E. J. Am.
Chem. Soc. 1974, 96, 523–529).
. Amatore, C.; Bayachou, M.; Boutejengout, F.; Verpeaux,
J. N. Bull. Soc. Chim. Fr. 1993, 130, 371–381. The ease
5
(i.e. the rate) with which a radical is reduced into the
corresponding anion should be correlated to its nucle-
ophilicity (res electrophilicity), as defined by Mulliken,
and not to its stability, as stated in this otherwise valu-
able paper.
diluted with pentane, washed with brine, dried (K CO ),
2 3
6
. (a) Uguen, D. Dissertation Thesis (Paris, 1977). For an
application of this reaction to the preparation of crown
ethers, see: Julia, M.; Uguen, D.; Zhang, D. Austr. J.
Chem. 1995, 48, 279–290; (b) Formation of radicals under
these aprotic reduction conditions is supported by the
and analysed (GLC–Mass). In one case (experiment with
D O), the solvents were subsequently evaporated and the
2
residue was chromatographed on silica gel (hexane/
AcOEt) to afford the TMB reagent, then 1c–d (character-
1
ised by mass and H NMR).