Enantioselective reduction of aryl ketones using LiBH4 and TarB-X: A chiral Lewis acid

Article abstract of DOI:10.1016/S0040-4039(02)00652-4

High enantioselectivities are obtained using a tartaric acid-derived boronate ester in combination with lithium borohydride for asymmetric reduction of aryl ketones. The chiral Lewis acid, TarB-X, is easily prepared in 1 h, and the resulting alcohols are

Full text of DOI:10.1016/S0040-4039(02)00652-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 3649–3652
Enantioselective reduction of aryl ketones using LiBH4 and
TarB-X: a chiral Lewis acid†
Jeff T. Suri, Truong Vu, Arturo Hernandez, Julie Congdon and Bakthan Singaram*
Department of Chemistry and Biochemistry, University of California, Santa Cruz, 1156 High Street, Santa Cruz, CA 95064,
USA
Received 28 February 2002; accepted 28 March 2002
Abstract—High enantioselectivities are obtained using a tartaric acid-derived boronate ester in combination with lithium
borohydride for asymmetric reduction of aryl ketones. The chiral Lewis acid, TarB-X, is easily prepared in 1 h, and the resulting
alcohols are obtained in enantiomeric excesses of 88–99%. © 2002 Elsevier Science Ltd. All rights reserved.
Enantioselective reduction of prochiral ketones to their
corresponding optically active alcohols is one of the
most powerful methodologies utilized in asymmetric
synthesis.1 A number of systems have been reported to
effect enantioselectivity, and a great majority involve
prochiral ketones involving chiral borane reagents,
have emerged.2,5 For example, Corey’s systematic study
of oxazaborolidines led to the development of the CBS
catalyst, which when used catalytically affords products
in 99% ee.6
boron reagents.2 Borane (BH3) and borohydride (BH4 )
−
containing compounds are the two fundamental classes
of boron-based reducing agents. Prior to use in asym-
metric synthesis, they were known to be key reagents in
the reduction of specific functionalities.3 Once it was
demonstrated that optically pure materials could be
obtained via chiral borane and borohydride reagents,
much attention was focused on finding the best chiral
precursor. In 1981, Itsuno et al. reported the first
promising result. Utilizing a mixture of chiral amino
alcohols and BH3:THF, the authors obtained enan-
tiomeric excesses in the range of 10–73% ee.4 Since
then, numerous studies on the asymmetric reduction of
Chiral borohydrides have also found utility in asym-
metric synthesis. Most notable are the reagents pre-
pared via modification of NaBH4. Specific examples
include sodium borohydride modified with monosac-
charide derivatives,7 amino alcohols,8 b-oxoaldimine
cobalt (II) complex,9 mandelic acid,10 amino acids,11
and tartaric acid.12
In connection with our program on LABs,13 we became
interested in asymmetric reductions involving borohy-
drides. After a survey of the literature, a recent report
caught our attention. Nozaki and co-workers reported
the preparation of bisboronate esters from naturally
occurring (L)-tartaric acid and demonstrated their effec-
O
tiveness as chiral Lewis acids in the reduction of amino-
substituted aryl ketones.14 Employing a series of
reducing agents, they found that the best results were
obtained using a stoichiometric amount of LiBH4.
Although the procedure required stoichiometric
amounts of all reagents, it utilized inexpensive chiral
auxiliaries and reagents that could be recycled. More-
over, the previous success of tartaric acid-derived cata-
lysts,12,15 further demonstrated the potential of such a
system, prompting us to pursue the generality of Noza-
ki’s methodology.
CO2H
CO2H
HO
HO
CaH2
OH
OH
O
O
OH
OH
B
B
THF
reflux, 1 h
X
X
O
TarB-X
1
Scheme 1. Preparation of TarB-X.
Keywords: enantioselective; asymmetric; chiral Lewis acid; tartaric
acid; boronic acid; reduction.
* Corresponding author. Tel.: 831-459-3154; fax: 831-459-2935;
e-mail: singaram@chemistry.ucsc.edu
†
Initially we attempted to repeat Nozaki’s work. When 2
This paper is cordially dedicated to Professor Herbert C. Brown on
the occasion of his 90th birthday.
equiv. of phenylboronic acid were reacted with (L)-tar-
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00652-4

3650
J. T. Suri et al. / Tetrahedron Letters 43 (2002) 3649–3652
The generality of the asymmetric reducing system was
determined via reduction of various aryl ketones with
LiBH4 and 1h under the optimum conditions. As sum-
marized in Table 3, excellent enantioselectivities were
obtained for most substrates with the highest selectivity
being achieved in the reduction of acetophenone, tetra-
lone, indanone, and 2%-acetonaphthone.
taric acid in refluxing toluene/THF, decomposition of
tartaric acid was apparent after 1 h, effectively turning
the solution black. Because toluene was required to
azeotropically remove water, high temperature reflux
was necessary. This proved to be problematic since
tartaric acid rapidly degrades above 100°C. As a result,
we opted to develop a procedure using milder condi-
tions. Surprisingly, we found that 2 equiv. of boronic
acid were not necessary to produce an active reagent.
Instead, only 1 equiv. of arylboronic acid was required
to generate a compound that functioned as a highly
selective chiral Lewis acid in the reduction of aryl
ketones with lithium borohydride. Herein we report the
preparation, characterization, and synthetic utility of
this novel tartaric acid-derived Lewis acid: TarB-X 1.
A variety of boronate esters were prepared according to
To ensure that the unprotected carboxylic acid was a
necessary component of TarB-X, a separate experiment
was carried out. The protected tartrate boronate ester 2
(Fig. 2) was prepared from dimethyltartrate and 4-
chlorophenylboronic acid. Reduction of acetophenone
in the presence of 2 under standard conditions,22 gave
the corresponding alcohol in only 7% enantiomeric
Scheme 1. In the presence of CaH2, (L)-tartaric acid
reacts readily with arylboronic acids generating, after 1
h in refluxing THF, the cyclic boronate esters 1 in
nearly quantitative yield.16,17 To ensure that these com-
pounds were in fact novel,18 and uniquely different
from Nozaki’s, they were carefully characterized using
1H, 13C, and 11B NMR,19 as well as FT-IR spectroscopy
and active hydride estimation. As indicated in Fig. 1,
(L)-tartaric acid contains two equivalent methine pro-
tons with a chemical shift of 4.3 ppm in THF-d8 relative
to TMS.
As indicated in Fig. 1, (L)-tartaric acid contains two
equivalent methine of 3-nitrophenylboronic acid the
methine protons shift to 5.0 ppm. However, the peak
remains a singlet, indicating that the molecule is C2
symmetric with respect to the appended boronic acid.
The relative integration of aromatic protons to methine
protons also matches the monoboronate adduct, giving
a 2:1 ratio, respectively. The 13C NMR spectrum, in
analogy to Fig. 1, shows the expected peaks. Finally,
the presence of free carboxylic acid groups in 1 was
verified via active hydride estimation.20 Upon addition
of LiBH4 to a solution of 1h in THF, 2 equiv. of H2(g)
were generated. FT-IR validated the structure, giving a
CꢀO stretch of 1736 cm−1 and an O–H stretch of 3427
cm−1, which is typical of carboxylic acids.
Figure 1. 1H NMR of (a) (
NO2·1h in THF-d8.
L)-tartaric acid and (b) TarB-
Table 1. Enantioselective reduction of acetophenone as a
function of arylboronic acida
OH
(R)
O
LiBH4
1a-h
THF
RT, 30 min
premix 30 min
The reduction of acetophenone with LiBH4 was carried
out utilizing 1a–h as chiral Lewis acids (Table 1).
Dropwise addition of LiBH4 to a premixed solution of
1:acetophenone (1:1) gave, after acid/base workup and
purification, (R)-1-phenylethanol in high yield with
enantiomeric excesses reaching 94% ee. The boronic
acid was also efficiently recovered using 3 M NaOH.
1
X
%
eeb,c,d
a
b
c
H
2-F
4-F
71
90
90
13
91
76
90
94
d
e
f
g
h
2,3,4,5,6-F
4-CF3
3,5-CF3
4-Cl
To optimize the selectivity, different sets of conditions
were explored using acetophenone as the model com-
pound. At low and high temperatures, the asymmetric
induction decreased in comparison to the ambient tem-
perature reaction (Table 2). Tetrahydrofuran was the
preferred solvent although similar ees were obtained in
acetonitrile. Overall, the optimum results required 2
equiv. of TarB–NO2 1h, 1 equiv. of LiBH4, and 1 equiv.
of acetophenone; under these conditions (R)-1-
phenylethanol was obtained in 99% ee.
3-NO2
a Reactions carried out at room temperature according to Ref. 21,
method A.
b Determined by GC analysis on a Supelco Beta-Dex 120 column.
c All products were of the (R) configuration as determined by chirop-
tical comparison.
d All reactions gave 100% conversion to the alcohol according to GC.

J. T. Suri et al. / Tetrahedron Letters 43 (2002) 3649–3652
3651
Table 2. Enantioselective reduction of acetophenone under different conditions with Lewis acid 1a
b
1
1 (mmol)
LiBH4 (mmol)
Acetophenone (mmol)
Solvent
°C
%
eec,d
g
g
g
g
g
h
h
h
h
3
6
6
6
12
6
6
6
12
6
3e
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
THF
THF
Et2Of
CH3CNg
THF
THF
THF
THF
THF
23
23
23
23
23
−15
0
40
50
88
32
90
99
74
83
85
99
23
a Reactions carried out as described in Ref. 21, method A, unless otherwise noted.
b All reactions gave 100% conversion to the alcohol according to GC unless otherwise noted.
c Determined by GC analysis on a Supelco Beta-Dex 120 column.
d All products were of the (R) configuration as determined by chiroptical comparison.
e Only 50% conversion to the alcohol according to GC.
f 1g was prepared in Et2O as in Ref. 17, method A, and reduction performed in Et2O as in Ref. 21, method A.
g 1g was prepared in CH3CN as in Ref. 17, method A, and reduction performed in CH3CN as in Ref. 21, method A.
Table 3. Reduction of aryl ketones with LiBH4 /1ha
Ar
B
O
O
CO2CH3
O
O
Cl
B
O
O
CO2CH3
B
O
O
Ar
2
3
Figure 2. Protected (L)-tartaric acid derivatives.
To the best of our knowledge there have been no
reports of simple boronate esters derived from unpro-
tected (
boron appended to (
L
)-tartaric acid. Most of the research involving
)-tartaric acid has focused on
L
tartrate and tartramide modified boronates,15 such as
Nozaki’s 3.14 Therefore, TarB-X, with its free car-
boxylic acid groups, represents a new type of chiral
Lewis acid.12c In addition, because the reagent is easily
prepared, induces high enantioselectivity, and can be
essentially fully recovered, the implications for its use in
both industry and academia appear to be quite
promising.
References
1. Asymmetric Synthesis; Morrison, J. D., Ed.; Academic
Press: New York, 1983; Vol. 2, Chapters 2 and 3.
2. For a comprehensive review on asymmetric boron-cata-
lyzed reactions, see: Deloux, L.; Srebnik, M. Chem. Rev.
1993, 93, 763.
3. Brown, H. C.; Krishnamurthy, S. Tetrahedron 1979, 35,
567.
4. (a) Hirao, A.; Itsuno, S.; Nakahama, S.; Yamazaki, N. J.
Chem. Soc., Chem. Commun. 1981, 315; (b) Itsuno, S.;
Hirao, A.; Nakahama, S.; Yamazaki, N. J. Chem. Soc.,
Perkin Trans. 1 1983, 1673.
5. (a) Brown, H. C.; Ramachandran, P. V. Pure Appl.
Chem. 1994, 66, 201; (b) Ramachandran, P. V.; Brown,
H. C.; Pitre, S. Org. Lett. 2001, 3, 17.
a
Reactions carried out as described in reference 21, method B.
b All reactions gave 100% conversion to the alcohol according to GC.
c Determined by GC analysis on a Supelco Beta-Dex 120 column.
d Values in parentheses were determined via optical rotation; all
products were of the (R) configuration.
e Enantiomers did not separate on GC.
6. For a comprehensive review on oxazaborolidines, see:
Corey, E. J.; Helal, C. J. Angew. Chem., Int. Ed. 1998, 37,
1986.
excess. Consequently, the free carboxylic acid is
essential in order to achieve high asymmetric induc-
tion.

3652
J. T. Suri et al. / Tetrahedron Letters 43 (2002) 3649–3652
7. (a) Hirao, A.; Mochizuki, H.; Nakahama, S.; Yamazaki,
bottom flask, sealed with a septum, and used without
further purification. Method B: Same as above except the
solvent was removed in vacuo and the product isolated
(typically isolated yields were >97%); a stock solution (0.3
M in THF) was prepared, stored in an ampoule at 4°C,
and used in the reduction step.21 The purity of 1a–h was
estimated by NMR to be >98% in all cases with the
N. J. Org. Chem. 1979, 44, 1720; (b) Hirao, A.; Naka-
hama, S.; Mochizuii, D.; Itsuno, S.; Ohowa, M.;
Yamazaki, N. J. Chem. Soc. Jpn. 1979, 807; (c) Hirao, A.
Bull. Chem. Soc. Jpn. 1981, 54, 1424; (d) Morrison, J. D.
J. Org. Chem. 1980, 45, 4229; (e) Hirao, A. J. Org. Chem.
1980, 45, 4231; (f) Hirao, A.; Itsuno, S.; Owa, M.;
Nagami, S.; Mochizuki, H; Zoorov, H. H. A.; Naka-
hama, S.; Yamazaki, N. J. Chem. Soc., Perkin Trans. 1
1981, 900.
minor impurity being due to unreacted (L)-tartaric acid.
18. Nozaki et al.14 reported that their preparation of bis-
boronate esters of tartaric acid contained no free car-
boxylic acid groups (Fig. 2, 3).
8. Itsuno, S.; Sakurai, Y.; Shimuzu, K.; Ito, K. J. Chem.
Soc., Perkin Trans. 1 1989, 1548.
19. In THF relative to BF3:Et2O as an external standard two
peaks are observed: one broad peak at 30 ppm and
another broad peak at 10 ppm.
20. H2(g) was measured in a hydride meter as described in:
Brown, H. C.; Kramer, G. W.; Levy, A. B.; Midland, M.
M. Organic Synthesis via Boranes; Wiley-Interscience:
New York, 1975; Chapter 9.
9. (a) Nagata, T.; Yorozu, K.; Yamada, T.; Mukaiyama, T.
Angew. Chem., Int. Ed. Engl. 1995, 34, 2147; (b) Sugi, K.
D.; Nagata, T.; Yamada, T.; Mukaiyama, T. Chem. Lett.
1997, 493.
10. Nasipuri, D.; Sarkar, A.; Konar, S. K.; Ghosh, A. Ind. J.
Chem. 1982, 21B, 212.
11. (a) Umino, N.; Iwakuma, T.; Itoh, N. Chem. Pharm.
Bull. 1979, 27, 1479; (b) Yamada, K.; Takeda, M.;
Iwakuma, T. J. Chem. Soc., Perkin Trans. 1 1983, 265.
12. (a) Yatagai, M.; Ohnukim, T. J. Chem. Soc., Perkin
Trans. 1 1990, 1826; (b) Polyak, F. D.; Solodin, I. V.;
Dorofeeva, T. V. Synth. Commun. 1991, 21, 1137; (c)
Furuta, K.; Maruyama, T.; Yamamoto, H. J. Am. Chem.
Soc. 1991, 113, 1041. The authors reported a tartaric acid
derivative with one free carboxylic acid group.
13. Lithum aminoborohydrides, see: (a) Thomas, S.; Collins,
C. J.; Cuzeens, J. R.; Spiciarich, D.; Goralski, C. T.;
Singaram, B. J. Org. Chem. 2001, 66, 1999; (b) Fisher, G.
B.; Fuller, J. C.; Harrison, J.; Alvarez, S.; Burkhardt, E.
R.; Goralski, C. T.; Singaram, B. J. Org. Chem. 1994, 59,
6378.
14. Nozaki, K.; Kobori, K.; Uemura, T.; Tsutsumi, T.;
Takaya, H.; Hiyama, T. Bull. Chem. Soc. Jpn. 1999, 72,
1109.
15. Gawronski, J.; Gawronska, K. Tartaric and Malic Acids
in Synthesis: a Source Book of Building Blocks, Ligands,
Auxiliaries, and Resolving Agents; John Wiley & Sons:
New York, 1999; Chapter 8.
16. Arylboronic acids are known to form boroxines upon
dehydration; the borixine facilitates facile attack by tar-
taric acid to afford the boronate ester, see: Ishira, K.;
Kurihara, H.; Matsumoto, M.; Yamamoto, H. J. Am.
Chem. Soc. 1998, 120, 6920.
17. Representative procedures for the preparation of TarB-X
1. Method A: An oven-dried 50 mL round bottom flask
with a sidearm was equipped with a magnetic stirring bar
and reflux condenser, and cooled to room temperature
under argon. The flask was charged with arylboronic acid
21. Representative procedure for the reduction of ketone.
Method A: The filtrate containing TarB-NO2 1h (15 mL
of a 0.4 M soln in THF), isolated from the first reac-
tion,17 (method A) was kept under a blanket of argon while
acetophenone (6 mmol, 0.721 g) was added via syringe;
the solution was stirred at the appropriate temperature
for 30 min. LiBH4 (3 mL of a 2 M solution in THF) was
carefully added dropwise over 10 min and the reaction
mixture was kept stirring for 30 min or until TLC (hex-
ane:ethyl acetate, 2:1) showed no more starting material.
The solution was quenched (CAUTION!, hydrogen evo-
lution) with 5 mL of H2O, 5 mL of 3 M HCl, and
brought to pH 12 with 3 M NaOH solution. The aqueous
layer was extracted with diethyl ether (4×30 mL) and
re-acidified with 12 M HCl to recover the boronic acid,
which was dried under vacuum and washed with pentane
(4.9 mmol, 0.891 g, 80% recovery). The combined organic
layers were washed with brine (30 mL) and dried over
anhydrous MgSO4. After removal of the solvents the
crude yellow oil was distilled under reduced pressure,
collected with diethyl ether and dried under high vacuum
for 1 h to yield 1-phenylethanol (5.3 mmol, 89%), [h]2D0=
+42.3° (c 5, MeOH), which corresponds to 94% ee of
(R)-1-phenylethanol; this was confirmed by GC analysis
on a Supelco Beta-Dex 120 column. Method B: TarB–
NO2 1h (12 mmol, 40 mL of a prepared 0.3 M stock
solution in THF) was kept under argon while substrate (6
mmol) was added via syringe. The mixture was stirred at
room temperature for 30 min. LiBH4 (3 mL of a 2 M
solution in THF) was carefully added dropwise over 10
min and the reaction mixture was kept stirring for 30 min
or until TLC (hexane:ethyl acetate, 2:1) showed no more
starting material. The remaining workup was the same as
above to yield the corresponding alcohol.
(6 mmol), (L)-tartaric acid (6 mmol), and CaH2 (12
mmol), and subsequently purged with argon. Anhydrous
THF, distilled from sodium–benzophenone ketyl (15
mL), was added via syringe, and the suspension heated to
reflux for 1 h. The reaction mixture was cooled to room
temperature, transferred to a fritted funnel using a dou-
ble-ended needle, and carefully filtered under argon. The
filtrate was collected into an oven-dried 50 mL round
22. Compound 2 (6 mmol) was prepared according to Ref.
17, method A and used in the LiBH4 (6 mmol) reduction
of acetophenone (6 mmol) as in Ref. 21, method A. GC
analysis of alcohol showed 100% conversion and 7% ee of
the (R)-isomer.