Synthesis of Allylsilanes
SCHEME 5. Red u ctive Silyla tion of Th ioeth er 11
probably comes from the anomalous deprotonation of
12.13 For comparison, the same reaction conditions per-
formed on thioether 5 led exclusively to an adduct sub-
stituted by both a thioether and a TMS moiety, indicating
that the introduction of the methyl groups decreases
dramatically the anomalous deprotonation of allylthio-
ethers.
benzyl group of compound 23 was removed during the
course of this reaction.
Effective Activa tion Meth od a n d Sp ecificity of
DBB a s a n Electr on Tr a n sp or ter . Next, we studied a
parameter that had not been examined yet: the condi-
tions of formation of the reducing agent LiDBB (Table
6). To prepare this reagent, an excess of lithium and a
catalytic amount of DBB were sonicated for less than 5
min at room temperature to afford a dark-green solution.
This was then immediately cooled to the desired tem-
perature (conditions A). In the second procedure (condi-
tions B) the time of LiDBB preparation was extended to
1 h. In conditions C, the medium was sonicated for 5 min
and stirred at room temperature for 1 h. When LiDBB
was prepared within 5 min (conditions A), the silyl
thioether 32 was never formed (entries 1-4 and 10). On
the other hand, when it was prepared according to
conditions B, we observed a substantial amount of adduct
32 at -42 °C in non-Barbier-type conditions (entry 7).
This reaction was exclusive at -78 °C (entry 6).
One possible reason for these observations which came
to mind is that THF might be cleaved into 4-lithiobutox-
ide, which would deprotonate the thioethers. Such a
reductive cleavage of THF has been reported, albeit
under more drastic conditions. Eisch was the first to note
that THF could be cleaved by lithium in the presence of
an arene.15 This author showed that refluxing a mixture
of biphenyl and lithium in THF affords some butanol
after hydrolysis of the reaction system. At that temper-
ature, 4-lithiobutoxide decomposed, and its formation was
not quantified. Later, Cohen showed that, in the presence
of boron trifluoride etherate, LiDBB cleaves THF instan-
taneously at a temperature as low as -78 °C.16 This reac-
tivity stems from the complexation of the oxygen atom
by this strong Lewis acid. More recently, Yus and col-
leagues reported that the dilithium naphthalene dianion
reacts slowly with THF at room temperature for 24 h,
affording after hydrolysis small amounts of butanol as
well as products of condensation of naphthalene with
THF.17
As previously, increasing the temperature to -42 °C
totally suppressed this deprotonation: the expected
allylsilanes 12 and 14 were the sole ones to be observed
(entry 2). At -11 °C, the overall yield was again dimin-
ished (to 21%) (entry 3). Only the silyl thioether 16 and
some bibenzyl were isolated, suggesting a homolytic
cleavage of the radical anion as outlined in Scheme 5.
Under Barbier-type conditions at -78 and -42 °C, we
observed uniquely the formation of the expected allylsi-
lanes 12 and 14, with, in particular, an improved
regioselectivity at -42 °C (compare entries 4 and 5 to
entries 1 and 2). In this latter case (entry 5), a 84% yield
of a 8:92 ratio of regioisomers 12 and 14 was observed.
At -11 °C, we isolated again only the debenzylated
adduct 16. These examples provide a second demonstra-
tion of the possibility to orientate selectively the reductive
silylation to different major products (14 and 16) depend-
ing on reaction conditions.
Comparing the regioselectivity of the reductive silyla-
tion of 1, 5, and 11, we can conclude that the structure
of the substrate, the temperature, and the presence or
the absence of chlorosilane during the reductive lithiation
step are three critical factors that cannot be considered
independently from one another, probably because each
of these three factors will affect the aggregation state of
the lithiated intermediates and its reactivity. Another
effect that may intervene is the intramolecular π-com-
plexation of the lithium cation to an alcene moiety. Such
an interaction has been reported by Ro¨lle and Hoff-
mann.14
The main common point that came out of this study is
that Barbier-type conditions at -42 °C prevent the
anomalous deprotonation and give the best yield and
selectivity. We settled therefore on this condition to
examine the reactivity of several allylthioethers (Table
5). We observed that the trimethylsilyl group was pref-
erentially introduced at the least substituted position
with a regioselectivity ranging from 100:0 (in the case of
26, entry 4) to 56:44 (in the case of 23, entry 3). The low
selectivity of this latter case is due to a similar level of
substitution of allylsilanes 24 and 25. As expected, the
When we analyzed a solution of LiDBB prepared
according to conditions B, we were pleased to notice the
(14) Ro¨lle, T.; Hoffmann, R. W. J . Chem. Soc., Perkin Trans. 2 1995,
1953-1954.
(15) Eisch, J . J . J . Org. Chem. 1963, 28, 707-710.
(16) (a) Mudryk, B.; Cohen, T. J . Am. Chem. Soc. 1991, 113, 1866-
1867. THF cleavage can also be achieved with a catalytic amount of
DBB or naphthalene, see: (b) Ramon, D. J .; Yus, M. Tetrahedron 1992,
48, 3585-3588. For other synthetic applications of THF cleavage,
see: (c) Oikawa, M.; Oikawa, H.; Ichihara, A. Tetrahedron 1995, 51,
11457-11464. (d) Dvorak, C. A.; Rawal, V. H. J . Chem. Soc., Chem.
Commun. 1997, 2381-2382.
(13) Similar deprotonation by strong bases has been reported, see
for example: (a) Marr, F.; Fro¨hlich, R.; Hoppe, D. Tetrahedron:
Asymmetry 2000, 13, 2587-2592. (b) Schlosser, M.; Zellner, A.; Leroux,
F. Synthesis 2001, 1830-1836.
(17) Yus, M.; Herrera, R. P.; Guijarro, A. Chem. Eur. J . 2002, 8,
2574-2584.
J . Org. Chem, Vol. 69, No. 22, 2004 7595