Angewandte
Chemie
to be an excellent ligand and the mixture of hexane/toluene
(2:1) was clearly the best choice of solvent for inducing
enantioselectivity and increasing the reactivity (Table 1,
entries 3–9). Generally, we found that lower temperature
and higher hydrogen pressure improved the diastereomeric
ratio. Finally, we could obtain a good diastereomeric ratio
(5:1) and excellent enantioselectivity at 58C under 120 bar of
H2 with 5 mol% of catalyst (Table 1, entry 9). Furthermore,
the ratio could be elevated to 9:1 after recrystallization. By
comparison of the NMR spectrum of 2a with known
compounds, the main product was determined to be cis-
configured.
With optimized conditions in hand, we tested a variety of
2-substituted flavones and chromones to gain insight into the
versatility of our catalytic system (Scheme 3). Fortunately, for
most of the 2-alkyl-substituted chromones including many
with primary and secondary alkyl chains, the reaction
proceeds with full conversion, moderate to good diastereo-
selectivity, and high enantioselectivity for the major product
(Scheme 3, 1a–i). We noticed that the diastereomeric ratio
decreased slightly when the length or substitution of the chain
was increased, while full conversion to the desired product
was maintained. Furthermore, we studied the influence of the
substitution on the carbocyclic ring of the chromone (1h, 1i).
In these cases good diastereoselectivity and high enantiose-
lectivity were observed, as well. Changing the position of the
substituent to position 3 (1q) led to
Scheme 2. Hydrogenation of chromones and flavones to optically
active flavanones, flavanols, chromanones, and chromanols.
We commenced our study with the hydrogenation of 2-
methyl-4H-chromen-4-one (1a) to optimize the reaction
conditions. Fortunately, when we applied our Ru catalyst
formed from ICy·HCl (3a; ICy = N,N’-dicyclohexylimidazol-
2-ylidene) for the hydrogenation of 1a, we exclusively
obtained the corresponding 2-methyl-4H-chroman-4-ol (2a)
with a d.r. of 1:1 (Table 1, entry 1). Encouraged by this result,
we tested whether our previously developed chiral ruthe-
nium–NHC complex, formed from the imidazolinium salt 3b,
could control the diastereoselectivity and enantioselectivity in
the hydrogenation of 1a. When we submitted 1a to the
hydrogenation process at 80 bar of H2 and 608C, the desired
product 2a was formed with full conversion, low diastereo-
selectivity (d.r. = 1.2:1), and good enantioselectivity (for the
major product, e.r. = 93:7) (Table 1, entry 2). We subse-
quently screened a variety of NHC ligands, solvents, pres-
sures, and temperatures. NHC ligand derived from 3b proved
a drop in enantioselectivity but full
conversion and higher diastereose-
Table 1: Optimization of the reaction conditions for the asymmetric hydrogenation of 2-methyl-4H-
chromen-4-one (1a).[a]
lectivity (8.3:1) than with the regio-
isomer 1a. Our method could also
be applied in the hydrogenation of
thiochromenone (2r).
When 2-substituted flavones
were employed as substrates, we
surprisingly found that only small
amounts of the desired hydrogen-
ated products were formed under
the above-mentioned conditions.
Fortunately, the hydrogenation of
flavone 1j proceeded smoothly with
full conversion and good enantiose-
lectivity to give the flavanol 2j
when the reaction was conducted
with 10 mol% of catalyst at room
Entry
L
T [8C]
p(H2) [bar]
Solv.
d.r.[b]
e.r.[c]
Yield[d] [%]
1
2
3
4
5
6
7
3a
3b
3b
3c
3d
3b
3b
3b
3b
60
60
25
25
25
25
25
5
80
80
80
80
80
80
120
120
120
tol.
tol.
tol.
tol.
1:1
–
>99
>99
>99
>99
trace
>99
>99
51%
>99
1.2:1
2.5:1
1.8:1
–
93:7
95.5:4.5
92:8
–
97:3
97:3
98:2
98.5:1.5
tol.
hex.
hex.
hex.
hex./tol.
(2:1)
2.7:1
3.0:1
4.5:1
5:1
8[e]
9[e]
5
(9.3:1)[f]
temperature.
As
shown
in
[a] General conditions: [Ru(cod)(2-methylallyl)2] (0.015 mmol), KOtBu (0.045 mmol), and NHC ligand
(0.03 mmol) were stirred at 708C in toluene or hexane (2 mL) for 16 h, after which the reaction mixture
was added to 1a (0.3 mmol), and hydrogenation was performed under the conditions listed in each case
for 24 h. [b] d.r. was determined by 1H NMR analysis. [c] e.r. given for main product was determined by
HPLC on a chiral stationary phase. [d] Yields pertain to isolated product. [e] The reaction time was
prolonged to 36 h. [f] After recrystallization from n-hexane/iPrOH; tol.=toluene, hex.=hexane.
Scheme 3, under the modified con-
ditions, in most cases, the reaction
of 2-substituted flavones gave the
desired flavanols in high yields and
good enantioselectivity for the cis
product. We found that the reactiv-
ity changes only slightly with the
electronic properties of the sub-
stituents: When the phenyl ring
bears electron-withdrawing groups
or electron-rich groups, such as
fluorine (1m), trifluoromethyl
(1n), and methoxy (1o) in para
Angew. Chem. Int. Ed. 2013, 52, 8454 –8458
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8455