Communications
Table 2: Substrate scope with TBSCl.
the binding of the catalyst to the substrate is entropically
Entry[a]
Product
Yield [%]
79
ee [%]
neutral (because an alcohol is released upon binding of
substrate). Consequently, the subsequent silylation step is
intramolecular but does not have to pay an entropic penalty
for bringing the catalyst and diol substrate together. Further-
more, the enantioselectivity in the reaction could be derived
from the binding step, functionalization step, or a combina-
tion of the two.
1[b]
89
2[c]
3[b]
4[c]
87
88
86
90
95
92
To initiate these studies, we developed a catalyst structure
that is easy to synthesize and modify. The catalysts in Table 1
are readily synthesized in two steps from amino alcohols. As
depicted in Table 1, when employing catalyst 1 in the
desymmetrization of 1,2-cyclopentane diol, the monopro-
tected product forms in 17% yield with À9% ee (Table 1,
entry 1). Using 2 and 3, which are derived from valine and
tert-leucine, respectively, affords the product in modest yield
and improved enantioselectivity (Table 1, entries 2 and 3). By
optimizing the less expensive valine core structure, it was
found that addition of a substituent adjacent to the imidazole
ring dramatically affects the selectivity of the reaction.
Silylation with 4a and 4b show a significant matched and
mismatched relationship, with 4b yielding a product with
excellent enantioselectivity (Table 1, entry 5);[14] and 4a
furnishes the opposite product enantiomer in low selectivity
(Table 1, entry 4). The high enantioselectivity in the silylation
reaction is notable given the fact that the reaction is
performed at room temperature. An explanation for the
high selectivities at noncryogenic temperatures is the rigid
nature of covalent bonding.
5[d]
6[e]
7[f]
82
93
78
90
86
90
[a] All reactions performed with 20 mol% 4b and 3 mol% PMP·HCl, and
are an average of two runs. [b] TBSCl (4 equiv), PMP (1.2 equiv), RT,
24 h. [c] TBSCl (2 equiv), PMP (1.2 equiv), RT, 12 h. [d] TBSCl (4 equiv),
PMP (2 equiv), 48C, 24 h. [e] TBSCl (4 equiv), PMP (2 equiv), RT, 24 h.
[f] TBSCl (4 equiv), PMP (2 equiv), 48C, 36 h.
hexane-1,2-diol, 1,2,3,4-tetrahydro-2,3-naphthalenediol, and
cyclohexene-1,2-diol provide comparable yields and enantio-
selectivities to those of the five-membered ring substrates
(Table 2, entries 2–4). Seven- and eight-membered rings also
yield the desired product in 90% and 86% ee, respectively
(Table 2, entry 5 and 6). The substrate scope was expanded to
an acyclic substrate, butane-2,3-diol, thus affording the
product in 90% ee and 78% yield (Table 2, entry 7). The
silyl transfer onto both trans-cyclohexane-1,2-diol and (R,S)-
pentane-2,4-diol using 4b as the catalyst yields the products in
less than 5% yield; furthermore, cis-4-cyclopentane-1,3-diol
affords the desired product in 26% yield and 15% ee. Given
that the proposed mechanism involves intramolecular trans-
fer or deprotonation, a strong proximity effect might account
for these observations. Though constraining in terms of
substrate scope, a catalyst that recognizes both chirality and
proximity has the potential to be a powerful reagent for site-
selective reactions.
Having optimized the catalyst structure, we investigated
the substrate scope.
A tetrahydrofuran-based substrate
affords the product in good yield and enantioselectivity
(Table 2, entry 1). The six-membered ring substrates, cyclo-
Table 1: Optimization of silyl transfer reaction.
With these encouraging results, we also investigated the
generality of the silyl reagents that can be transferred. The
smaller triethylsilyl group is transferred effectively with
excellent enantioselectivity (90% ee) and yield (94%;
Table 3, entry 1) after a reaction time of 1 hour at room
temperature. The sterically larger silyl reagent tert-butyldi-
phenyl silyl chloride (TBDPSCl) affords the product in
90% ee and good yield (75%; Table 3, entry 2), although
extended reaction times are necessary to achieve high
conversion. The reaction time can be shortened by increasing
the concentration of the reaction mixture (1.0m in diol), the
result being only a small decrease in the ee value (Table 3,
entry 2). Silylation with the very reactive dimethylphenyl silyl
chloride (DMPSCl) results in a decrease in enantioselectivity
to 79% (Table 3, entry 3); presumably, the lowered selectivity
is a result of background silylation.
Entry
Catalyst
Yield [%][a]
ee [%][b]
1
2
3
4
1
2
3
4a
4b
4b
17
19
20
25
À9
34
40
À16
97 (94)[c]
92
5
84 (92)[c]
76
6[d]
[a] Reaction performed at 0.20 mmol substrate (0.4m). Yields deter-
mined by GC analysis using an internal standard, 1,3,5-trimethoxyben-
zene. [b] Enantiomeric excess (ee) determined by GC analysis. [c] Reac-
tion performed at 0.40 mmol substrate. Yield of product given in
parentheses. [d] Reaction performed 0.2m of diol in tBuOH.
PMP=pentamethylpiperidine, TBS=tert-butyldimethylsilyl, THF=te-
trahydrofuran.
8168
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 8167 –8171