Communications
The results presented herein are complementary to those
obtained with earlier protocols based on [Ru(p-cymene)Cl2]2
and amino acid hydroxyamide ligands, which demonstrated
high efficiency, low catalyst loading, and short reaction times
for the reduction of aryl alkyl ketones. Nevertheless, these pro-
tocols showed limitations for sterically hindered ketones,
which were reduced with high enantioselectivity, albeit in low
yields. The current protocol circumvents this problem and de-
livers sterically demanding secondary alcohols in moderate to
high yields with good to excellent ee values. The major differ-
ences between the current and previous catalytic systems are
changes to the solvent system and the amount of added base.
These two factors appear to have a most positive influence on
the lifetime of the catalyst, which in contrast to earlier studies
show catalytic activity even after a reaction time of 24 h. The
extended catalyst lifetime is likely the reason for the high
yields, also for sterically demanding substrates. We previously
successfully used LiCl as an additive in the ATH of prochiral ke-
tones, for which lithium ions play an important role in generat-
ing a tight transition state for hydride transfer. However, in
contrast to these observations, the current protocol shows en-
hanced activity and selectivity if potassium ions are used as
either the counterion of the base or as an additive to the reac-
tion mixture. The larger potassium ion would allow for a more
flexible transition state, which would better fit the increased
size of hindered ketones. As a consequence, sterically demand-
ing substrates are reduced in higher yields and with higher
enantioselectivity. This mechanistic insight could open up fur-
ther applications and modifications of the current catalytic
system.
Table 3. Investigation of the effect of different counterions.[a]
Entry
Base
Additive
Conv.[b]
[%]
ee[c]
[%]
1
2
3
4
5
6
7
8
9
10
LiOtBu
LiOtBu
NaOtBu
NaOtBu
KOtBu
KOtBu
KOtBu
KOtBu[f]
KOtBu[f]
KOtBu[f]
–
72
70
72
20
99
19
90
60
56
70
93
91
94
78
95
45
92
90
90
95
[12-crown-4][d]
–
[15-crown-5][d]
–
[18-crown-6][d]
[18-crown-6][e]
–
LiCl[g]
KCl[g]
[a] Reaction conditions: [Ru(p-cymene)Cl2]2 (1 mol%), ligand 6 (2.2 mol%),
ketone 1a (1 mmol, 0.25m reaction solution), dry ethanol and dry THF
(3:2) as solvent, base (20 mol%). All reactions were performed at 408C for
24 h. [b] Conversion was determined by 1H NMR spectroscopy. [c] The ee
was measured by GLC analysis on a chiral stationary phase (CP Chirasil
DEX CB). [d] 40 mol%. [e] 2 mol%. [f] 7 mol%. [g] 10 mol%.
different complexation agents were undertaken (Table 3). The
bases used were tert-butoxide salts containing cations of differ-
ent sizes, LiOtBu, NaOtBu, and KOtBu, in combination with the
crown ethers [12-crown-4], [15-crown-5], and [18-crown-6]
(Table 3, entries 1–6). In general, the addition of crown ethers
to the reaction mixture decreased both the conversion and the
enantioselectivity, which strongly indicates that the cation
plays a crucial role in the reaction. No inhibition of the catalyst
was observed if small amounts (2 mol%) of the crown ether
were added, which shows that the trends observed are not
a result of catalyst inhibition by the additive (Table 3, entry 7).
There was only a small difference in conversion and enantiose-
lectivity for the reactions performed with the use of the lithium
base (Table 3, entries 1 and 2), which is in line with previous re-
sults with the use of [12-crown-4] to trap lithium ions.[13] In re-
actions containing sodium or potassium tert-butoxide, signifi-
cant decreases in both the conversion and enantioselectivity
were observed if the crown ethers were present in the reaction
mixture (Table 3, entries 3–6). The largest difference was seen
by using KOtBu, for which the addition of [18-crown-6] re-
duced the ee from 95 to 45%.
Experimental Section
General
All reactions were performed under a nitrogen atmosphere with
oven-dried glassware.
General procedure for the asymmetric transfer hydrogena-
tion of sterically hindered ketones
The catalyst precursor [Ru(p-cymene)Cl2]2 (6.2 mg, 0.01 mmol) was
treated under vacuum in a capped vial for 10 min. Dry THF
(1.60 mL) and dry ethanol (1.80 mL) were added, followed by
a 0.11m stock solution of ligand 6 in dry ethanol (0.20 mL,
0.022 mmol, 2.2 mol%)[18] and the ketone (1.0 mmol). The resulting
mixture was stirred for 15 min at 408C. The reaction was initiated
by the addition of a 0.5m stock solution of KOtBu in dry ethanol
(0.40 mL, 0.20 mmol, 20 mol%). Aliquots were withdrawn at suit-
able intervals (see details in the tables) and were then pressed
though a pad of silica with ethyl acetate as the eluent. The result-
ing solutions were analyzed by 1H NMR spectroscopy, GLC on
a chiral stationary phase (CP Chirasil DEX CB), or HPLC on a chiral
stationary phase (AD, AS, OB, and ODH columns).
In previous studies on this particular catalyst system we
demonstrated that a minimum of 3 equivalents of base in rela-
tion to the amount of ligand was necessary for a successful re-
action outcome.[12a] Hence, approximately 7 mol% base would
be sufficient in the current setup, but this was not the case
(Table 3, entry 8). The addition of lithium chloride to the reac-
tion mixture did not increase the conversion or enantioselec-
tivity (Table 3, entry 9). On the contrary, the addition of potassi-
um chloride resulted in a significant improvement in the enan-
tioselectivity (Table 3, entry 10), which indicates that potassium
plays an essential role in the reaction, in line with the proposal
in Scheme 4b.
To isolate the product, the crude material was filtered through
a pad of silica with ethyl acetate as the eluent to remove metal
residues and the ligand, and this was followed by concentration
and purification by column chromatography.
ChemCatChem 2015, 7, 3445 – 3449
3448
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim