ACS Catalysis
Letter
a
It is noteworthy that alkylation of the hydroxyl groups in
catalyst F led to a drastic decrease in enantioselectivity, which
implies that the hydroxyl groups play an important role in
controlling the enantioselectivity, possibly through hydrogen
bonding with the enolate intermediate.9a
Scheme 5. Scope of Cyclic Ketones
During the course of the catalyst screening, we observed the
formation of side products in the reaction, which led to low
yield of the desired product, even when there was complete
consumption of 1a. Isolation and characterization of the side
products revealed that they were acetophenone and benzoate
salt. In fact, a closely related transformation was studied in
detail by Ogata and co-workers, who proposed the mechanism
of α-cleavage of the peroxide intermediate in the presence of
base (Scheme 3).12 Hence, we set about optimizing the
reaction conditions to minimize the side product formation and
to increase the yield of the desired α-hydroxy ketone.
With PTC D as the optimal catalyst, different solvents were
screened, and it was found that reaction in benzene gave
slightly better enantioselectivity compared to toluene (Table 2,
entry 3), while poorer results were obtained with diethyl ether
and dichloromethane (Table 2, entries 1 and 2). Even though
triethyl phosphite is normally used for reducing the peroxide
intermediate to the product,13 we decided to screen other
reductants in hopes of increasing product yield. It was found
that phosphines could also be used for this role,13a and 1,2-
bis(diphenylphosphino)ethane (DPPE) was able to furnish the
product in good yield, although with diminished enantiose-
lectivity (Table 2, entry 7). Gratifyingly, by decreasing the
amount of DPPE used to 0.5 equiv, we were able to obtain the
product with an improved yield and no compromise in
enantioselectivity compared to using P(OEt)3 (Table 2, entry 8
vs entry 3). Lowering the reaction temperature to 10 °C further
improved the enantiomeric ratio (Table 2, entry 9). Very
importantly, the reaction can also be conducted in air in place
of O2, although a lower yield and slightly diminished
enantioselectivity were obtained (Table 2, entry 10).
a
General conditions: 3 (0.1 mmol), D (5 mol %), P(OEt)3 (0.1
mmol), 50% aq NaOH (0.25 mL), PhH (0.1 M) at room temperature.
b
c
Reaction performed in PhH (0.2 M). Reaction performed in PhH
d
(0.067 M) and 50% aq NaOH (0.30 mL). Reaction using D (10 mol
%), 50% aq NaOH (0.30 mL) in PhH (0.067 M).
With the optimized conditions in hand, we explored the
substrate scope of the α-hydroxylation reaction (Scheme 4).
Substitution on the phenyl ring generally decreased the
enantioselectivity of the products (2a−2e). Those bearing
electron-withdrawing substituents could also be accessed, albeit
with lower yield (2c and 2f) due to more facile α-cleavage of
the intermediate. On the other hand, it was observed that
increasing the steric bulk on the α-carbon led to improved
yields of the products (2e, 2g), probably due to the decreased
rate of α-cleavage of the intermediate. We were pleased to find
that the α-hydroxylation reaction can also be applied to enones,
yielding the synthetically versatile products with good
enantiomeric ratios, although the yields were modest (2h−
2j). Substrates bearing dialkyl substitutions at the α-position
were also examined. The level of efficiency and selectivity were
unfortunately much lower (results not shown).
Encouraged by the results obtained with acyclic substrates,
we applied dimeric catalyst D to the α-hydroxylation of cyclic
ketones, in particular α-substituted tetralones (Scheme 5). To
our great delight, the hydroxylated products were obtained with
good to excellent enantioselectivities and generally good yields
even when the reactions were performed at room temperature
using only 5 mol % of catalyst and triethyl phosphite as the
reductant. A variety of alkyl substituents at the α-position were
well tolerated (4a-4d, 4f, 4g), whereas substitutions on the
aromatic ring decreased the enantiomeric ratios slightly (4e,
4h). Although indanones can also be hydroxylated with good
carbonyl compounds remained elusive. Herein we report our
recent discovery of an efficient and operationally simple
method for the synthesis of enantioenriched acyclic and cyclic
α-hydroxy ketones using oxygen. A cinchona alkaloid-derived
dimeric phase-transfer catalyst that can be easily prepared in
one step proved to be highly efficient and enantioselective.
We initiated our studies using acyclic ketone 1a as the model
substrate (Table 1), which was previously used by the Davis
group in their studies on enantioselective α-hydroxylation using
chiral N-sulfonyloxaziridines.10 Cinchona alkaloid-based PTCs
A to C that have found wide application in asymmetric phase-
transfer catalysis were screened, but they proved to be
disappointing because they gave the product 2a in low yields
and enantiomeric ratios (entries 1−3). We then turned to
catalyst D, which was originally reported by Park and Jew.11 It
can be easily synthesized in high yields in a single step from
commercially available starting materials. To our delight, use of
just 5 mol % of D afforded the product in good enantiomeric
ratio (92:8 er) albeit with a modest isolated yield of 51% (Table
1, entry 4). Encouraged by this result, we screened catalyst E,
which is a variant of D but has a para-substitution pattern on
the linker as well as F, in which the hydroxyl groups on the
cinchona alkaloids were allylated. Both these modifications
failed to improve the yield or enantiomeric ratio of the product
(Table 1, entries 5 and 6). Thus, PTC D was chosen as the
catalyst for further optimization of the hydroxylation reaction.
3611
ACS Catal. 2015, 5, 3609−3612
Sim, Sui-Boon Derek
