phine ligands, BINAP (L1), MeO-BIPHEP (L2), and
SIPHOS-PE (L3), were examined, low yields (<15%) and
low levels of enantioselectivity (23–61% ee) were observed
(Table 1, entries 1, 3, and 5). When the reactions were car-
ried out at 1008C, a common reaction temperature for reac-
tions involving phosphine ligands, the results were worse
(Table 1, entries 2, 4, and 6). Then we turned our attention
to the chiral diene ligands. The use of chiral diene L4 gave
the product with a very high ee value (99%), albeit in mod-
erate yield (50%; Table 1, entry 7). High yields could be ob-
tained by using chiral ligands L5 and L6, however, low
levels of enantioselectivity were obtained (<65% ee;
Table 1, entries 8 and 9). Using diene L4, attempts to further
improve the yield by varying solvent, base, and temperature
proved unsuccessful (Table 1, entries 10–15). We reasoned
that the low yield probably originates from the deactivation
of the rhodium catalyst, through its binding to the alkene
moiety of adduct 3aa, and the side products generated
through the hydrolysis of trifluoroborate 2a. To our delight,
when freshly prepared [{Rh(OH)(L4)}2] was used as a more
active catalyst,[12] the yield significantly improved and
reached 97% (Table 1, entry 16). Even with the use of a
lower catalyst loading (1 mol% Rh), the product was still
obtained in 79% yield with 99% ee (Table 1, entry 18).
With optimized reaction conditions established (Table 1,
entry 17), the reaction scope was examined (Scheme 2). The
reactions of a variety of alkenyltrifluoroborates with lactam
1a were successful and products were obtained in moderate
to excellent yields (43–97%), with excellent levels of enan-
tioselectivity (97–99% ee). Both steric and electronic prop-
erties of alkenyltrifluoroborates affected the yield and the
enantioselectivity. The use of substituted alkenyltrifluorobo-
rates gave products that were generally obtained in higher
yield, therefore showing that increased steric hindrance is
beneficial to the reaction. For example, very high yields
were observed in the conjugate addition reaction of 2,2-di-
methyl-substituted and 1,2,2-trimethyl-substituted alkenyltri-
fluoroborates (3aa and 3ac). Alkenyltrifluoroborates that
were relatively electron rich gave higher yields. In compari-
son with the conjugate addition reaction of 2-phenyl-alke-
nyltrifluoroborate, the reaction of a substrate containing an
electron-donating methoxy group on the phenyl ring gave a
slightly higher yield (85%; compare 3ah with 3ai); however,
reaction of a substrate containing an electron-withdrawing
chloride group on the phenyl ring gave a lower yield (66%;
compare 3ah with 3aj). When a six-membered lactam was
used, the levels of enantioselectivity were slightly lower al-
though the yields remained high (3ba, 3bb, and 3bh). When
the reaction of a,b-unsaturated lactones was investigated, a
different combination of solvent and base was found to be
optimal. The combination of toluene and Et3N was replaced
by the combination of dioxane and KF and in doing so, the
yields were satisfactory, and again the levels of enantioselec-
tivity observed for 5-membered lactones were higher than
those observed for 6-membered lactones (compare 3ch with
3dh). As expected, the use of the more reactive a,b-unsatu-
rated ketones led to high yields, although only the 5-mem-
bered 2-cyclopentenone gave excellent levels of enantiose-
lectivity (98–99%; compare 3ed, 3eh, and 3 fh).
Table 1. Optimization of reaction conditions.[a]
Entry Catalyst
T [8C] Base
Yield [%][e] ee [%][f]
1
2
3
4
5
6
7
8
L1/
N
G
Et3N
Et3N
8
4
4
23
53
9
–
61
–
99
64
15[g]
95
95
96
95
96
71
99
99
99
L1/
L2/
L2/
L3/
L3/
L4/
L5/
L6/
L4/
L4/
L4/
L4/
Et3N
Et3N
Et3N
Et3N
Et3N
Et3N
Et3N
DIEA
pyridine
KOH
Et3N
Et3N
Et3N
Et3N
–
trace
13
trace
50
69
92
46
8
43
38
9
10
11
12
13
14
15[b]
16[c]
17[c]
18[d]
L4/
L4/
37
58
97
42
[{Rh(OH)(L4)}2]
[{Rh(OH)(L4)}2]
[{Rh(OH)(L4)}2]
RT
RT
RT
Et3N
79
[a] Unless noted otherwise, reactions were carried out with 1a
(0.2 mmol), 2a (0.4 mmol), [{RhCl(C2H4)2}2] (0.005 mmol), ligand
AHCTUNGTRENNUNG
(0.011 mmol), catalytic amount of KOH (0.011 mmol) and 2 equiv of
base in toluene/H2O (10/1, 2.2 mL). [b] 1,4-Dioxane was used instead of
toluene. [c] [{Rh(OH)(L4)}2] (0.005 mmol) was used as rhodium catalyst
and without adding KOH. [d] [{Rh(OH)(L4)}2] (0.001 mmol) was used as
rhodium catalyst and without adding KOH. [e] Yield of isolated product.
[f] Determined by HPLC analysis, using a chiral stationary phase. [g] The
sense of asymmetric induction was opposite that observed in reactions as-
sociated with other entries in this table. DIEA=N,N-diisopropylethyla-
mine.
The absolute configurations of 3ch, 3eh, and 3 fh were as-
signed as R by comparison of their optical rotation with
those in the literature;[13] the configuration of the products
is in agreement with the stereochemistry-defining model for
the arylation of enones that was described by Hayashi
et al.[14] Using the conjugate addition reaction of potassium
2-(E)-phenylethenyltrifluoroborate and [5H]-furan-2-one in
the presence of the rhodium catalyst bearing the ligand
(S,S)-L4 as an example, the alkenylrhodium intermediate
prefers to coordinate to the Re face of the lactone to mini-
mize steric repulsions between the phenyl groups on the
&
2
&
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!