3242
C. Song et al. / Tetrahedron Letters 46 (2005) 3241–3244
ruthenium complexes for arylboronic acid additions
to enones.3 Planar chiral [2.2]paracyclophane ligands
include diphosphines,10 oxazoline-phosphines,11 oxazo-
line-imidazolium,12 oxazoline-selenides,13 oxazoline-
alcohols,14 and Schiff base-phenols for hydrogenation,
allylic substitution, and organozinc addition reactions.15
Dimeric chiral [2.2]paracyclophanes are rare and their
use as catalysts has not been reported previously.16
(11% ee) and tetrabutylammonium bromide (TBAB)
gave no reaction. Use of 0.1 mol % ruthenium, with
0.24 mol % 3, the reaction rate was very slow, 98 h,
and the yields and selectivities were reduced, 86%, 93%
ee. Use of greater amounts of the catalyst, for example,
1 mol % ruthenium with 2.4 mol % 3, showed no signifi-
cant improvement, 97%, 98% ee. Use of 1 equiv of
diphenylsilane gave a low yield, 56% (24 h) and the
selectivity was reduced, 90% ee. With 2 equiv, the yield
was reduced, 89% (18 h) and the selectivity was the
same, 97% ee. Use of 3 equiv gave no improvement over
the optimal 2.5 equiv shown.
The synthesis of the bis-paracyclophane ligands 1–3
begins with the known compound Sp-pseudo-ortho-
bromoamino [2.2]paracyclophane17 as previously re-
ported.3 The ligands were used under the conditions of
Uemura using conveniently available ruthenium dichlo-
ride tris(triphenylphosphine) (0.5 mol %) with added sil-
ver triflate (1 mol %) in THF at room temperature
(Table 1). With ligand 1, acetophenone was reduced, fol-
lowing work-up with HCl and purification by chroma-
tography, in 98% yield and 90% ee (chiral HPLC).18
Ligands 2 and 3 both gave S-phenylethanol in excellent
selectivity, 98% ee. Use of [RuCl2(PhH)]2 with 3 also
gave a high yield (97%, 24 h) but with reduced selectiv-
ity, 92% ee. The other test substrate at this time was 3-
methylpropylphenylketone, which was reduced using 3
in 97% ee (not shown). Ligand 2 was less selective,
90% ee in this case. Ligand 3 was selected for further
investigation to optimize the conditions and reduce
other substrates. Various solvents were explored with 3
and THF proved to be superior. Use of chloroform low-
ered the selectivity to 77% ee. Toluene and benzene gave
low yields (36% and 20%) and the selectivity was very
low. Dioxane was also low yielding, 58%, however the
selectivity was not greatly reduced, 92% ee. The ratio
of the metal to the ligand was also explored. At 1:1 ratio,
the reactivity was high, 96% in 16 h, and the selectivity
was compromised, 83% ee. At 1:2 ratio, the yield
(97%, 20 h) and the selectivity improved, 96% ee. The
optimal ratio at 1:2.4 (98%, 97% ee) is superior to 1:3
where both the yield (67%) and the selectivity (91% ee)
begin to be lowered.
Other methyl ketones were used under the optimized
conditions using the RuÆ3 complex as catalyst (Table
2). Cyclohexylmethyl ketone was one the of few sub-
strates that did not show high selectivity, 58% ee. All
other arylmethyl ketones, including ortho-substituted
substrates gave excellent yields and selectivities. More
electron rich, methyl and methoxy aryl ketones reacted
with high rates. p-Trifluoromethyl acetophenone
required 24 h and gave an 81% yield with only 77% ee.
This ligand 3–ruthenium combination also gave excel-
lent yields and selectivities for more hindered aryl
ketone substrates (Table 3). Ethyl, isopropyl, and sec-
butyl phenyl ketones were highly selective, 90–93% ee.
Even cyclohexylphenyl ketone reacted in 90% yield with
88% ee. Cyclic aryl ketones, indanone and benzocyclo-
hexanone also reacted with high selectivity, 93% and
92% ee.
While the origin of the stereoinduction remains difficult
to elucidate at this time, the nature of the Ru–ligand for-
mation and the step responsible for the selectivity may
be compared to mechanistic findings reported by Prock
and Giering for the related rhodium based process
(Scheme 2). Key findings in this and previous studies
determined that the reaction is first order in silane and
rhodium and there is a saturation effect with acetophe-
none only when ketone is present at high concentra-
tion.19 In the case of ruthenium with 3, excess ligand
may be needed to favor formation of mono-NHC com-
plex A. Silane then adds to the 14 e complex to give the
Various additives were also explored along with catalyst
and silane loadings for the process. Use of copper(II) tri-
flate, in place of silver(I) triflate, gave high reactivity,
97% (20 h) with lowered selectivity, 92% ee. Use of 18-
c-6 with potassium carbonate gave very low selectivity
Table 2. Reduction of methyl ketones
1.2 mol%
OH
3
O
Table 1. Hydrosilylation with cyclophane ligands
4
RuCl2(PPh3)3 0.5 mol%
R
R
OH
Ligand
O
AgOTf, THF, rt
/ H2O, HCl
5
+ Ph2SiH2
2.5 equiv.
RuCl2(PPh3)3 0.5 mol%
4
5
AgOTf, THF, rt
/ H2O, HCl
R =
Time (h)
% Yield
% ee
+ Ph2SiH2
2.5 equiv.
Ph
c-Hexyl
2-PhEt
16
36
15
20
15
16
36
20
12
12
48
98
93
95
98
92
96
91
90
98
91
81
97
58
93
96
96
94
93
97
96
93
77
Ligand
Ru/lig. ratio
Time (h)
% Yield
% ee
1
2
3
3
3
3
3
1:2.4
1:2.4
1:2.4
1:2.4
1:1
16
16
16
24
16
20
24
98
98
98
86
96
97
67
90
97
97
77a
83
96
91
2-Naphthyl
o-MeOPh
o-Tolyl
o-BrPh
o-ClPh
1:2
1:3
p-Tolyl
p-MeOPh
p-F3CPh
a CHCl3 used as solvent.