Role of NaBH4 stabilizer in the oxazaborolidine-catalyzed asymmetric reduction of ketones with BH3-THF

Article abstract of DOI:10.1021/jo016257c

When stabilized BH_3-THF (BTHF) was added to a mixture of ketone and tetrahydro-1-methyl-3,3-diphenyl-1H,3H-pyrrolo[1,2-c] [1,3,2]oxazaborole (MeCBS-ozaxaborolidine, MeCBS) catalyst 1, low enantioselectivities resulted. Several relative rate experiments showed that a borohydride species in BTHF catalyzed the nonselective borane reduction of ketones, effectively competing with enantioselective MeCBS reduction of ketones, lowering the overall selectivity of the reaction. Improved enantioselectivities in the reaction are obtained by reversing the mode of addition (ketone to BTHF and catalyst), lowering the concentration of NaBH₄ stabilizer in the BTHF solution (87- 95% ee) and increasing the concentration or addition rate of BTHF. Decreased reaction temperature and increased catalyst loading only slightly improved the selectivity of the reaction. Upon reaction parameter optimization, simultaneous addition of substrate and BTHF to MeCBS catalyst stabilizer resulted in the highest overall enantioselectivities (96% ee) and diminished the effect of the borohydride. Alternatively, the addition of Lewis acids such as BF_3-THF to the reaction mixture effectively destroyed the NaBH₄ stabilizer in BTHF solutions, restoring the enantioselectivity to acceptable levels.

Full text of DOI:10.1021/jo016257c

2
970
J . Org. Chem. 2002, 67, 2970-2976
Role of Na BH
4
Sta bilizer in th e Oxa za bor olid in e-Ca ta lyzed
Asym m etr ic Red u ction of Keton es w ith BH3-THF
Shawn M. Nettles, Karl Matos,* Elizabeth R. Burkhardt, Dave R. Rouda, and
†
J oseph A. Corella
Callery Chemical Company, 1420 Mars-Evans City Road, Evans City, Pennsylvania 16033
Karl.Matos@MSAnet.com
Received October 31, 2001
When stabilized BH3-THF (BTHF) was added to a mixture of ketone and tetrahydro-1-methyl-
3
,3-diphenyl-1H,3H-pyrrolo[1,2-c][1,3,2]oxazaborole (MeCBS-ozaxaborolidine, MeCBS) catalyst 1,
low enantioselectivities resulted. Several relative rate experiments showed that a borohydride
species in BTHF catalyzed the nonselective borane reduction of ketones, effectively competing with
enantioselective MeCBS reduction of ketones, lowering the overall selectivity of the reaction.
Improved enantioselectivities in the reaction are obtained by reversing the mode of addition (ketone
to BTHF and catalyst), lowering the concentration of NaBH
4
stabilizer in the BTHF solution (87-
9
5% ee) and increasing the concentration or addition rate of BTHF. Decreased reaction temperature
and increased catalyst loading only slightly improved the selectivity of the reaction. Upon reaction
parameter optimization, simultaneous addition of substrate and BTHF to MeCBS catalyst stabilizer
resulted in the highest overall enantioselectivities (96% ee) and diminished the effect of the
borohydride. Alternatively, the addition of Lewis acids such as BF3-THF to the reaction mixture
effectively destroyed the NaBH
acceptable levels.
4
stabilizer in BTHF solutions, restoring the enantioselectivity to
In tr od u ction
and/or moisture. Corey and Helal also promoted the
advantages of other borane sources over BH3-THF for
Borane-tetrahydrofuran complex (BTHF) is a very
this asymmetric reduction due to the sensitivity of BTHF
good reducing agent for aldehydes, ketones, amides,
to moisture and oxygen.3
1
carboxylic acids, and other functional groups. Due to the
Another problem with BTHF in the CBS reaction is
the reduction of other functional groups within the
ketonic substrate, especially when the substrate is added
to the borane/catalyst mixture. Recently, King and co-
workers reported hydroboration byproducts (3-10%)
when they carried out the oxazaborolidine-catalyzed
borane reduction of LTD4 antagonist L-699,392 used for
asthma treatment. Due to the over-reduction problems
with the CBS process, they changed the process to employ
commercial availability of BTHF along with its selectivity
and cleanliness as reducing agent (only easily removed
borate byproduct is produced), BTHF is a very useful
reagent for pharmaceutical and industrial applications.2
The enantioselective reduction of prochiral ketones
with BTHF catalyzed by tetrahydro-1-methyl-3,3-di-
phenyl-1H,3H-pyrrolo[1,2-c][1,3,2]oxazaborole [(R)- or
(
S)-MeCBS-oxazaborolidine reagent, (R)- or (S)-MeCBS]
catalyst 1 is an excellent tool for the synthesis of alcohols
9
chlorodiisopinocampheylborane for the reduction.
3
in high enantiomeric excess. This process has been
The variable results and potential over-reduction
problems when BTHF is used for these CBS reductions
represent serious limitations for potential applications
of this chemistry. Therefore, we chose to find optimal
conditions for this process that would give consistently
high enantioselectivities with BTHF as the borane source.
successfully applied to the asymmetric synthesis of
various natural products and pharmaceuticals such as
4
5
6
(
R)-denopamine, (S)-R-damascone, (R)-fluoxetine, and
7
ginkgolide B.
Early in the examination of this catalyst, J ones and
co-workers reported variable results when they utilized
commercial BTHF as the borane source together with the
MeCBS catalyst to synthesize the anhydrase inhibitor
Resu lts a n d Discu ssion
8
MK-0417. They attributed the erratic selectivities to the
A way to decrease or eliminate substrate over-reduc-
tion is by adding the borane to the substrate and catalyst.
Through this mode of addition, borane is not in excess,
thus lowering the chance for reduction of other functional
groups. Borane addition would also be required in cases
of low substrate solubility.
propensity of BTHF to decompose in the presence of air
*
To whom correspondence should be addressed. Fax (412) 967-4188.
Davos Chemical Corporation, 464 Hudson Terrace, Englewood
†
Cliffs, NJ 07632.
1) Brown, H. C.; Zaidlewics, M. In Encyclopedia of Reagents for
Organic Synthesis; Wiley: New York, 1995; Vol 1, p 638.
2) (a) Wilkerson, W. W.; Rodgers, J . D. US Patent 5,508,400. (b)
Sohar, P.; Mathre, D. J .; Blacklock, T. J . EP 617037.
(
(
(8) J ones, T. K.; Mohar, J . J .; Xavier, L. C.; Blacklock, T. J .; Mathre,
D. J .; Sohar, P.; J ones, T. T.; Reamer, R. A.; Roberts, F. G.; Grabowski,
E. J . J . J . Org. Chem. 1991, 56, 763.
(9) King, A. O.; Corley, E. G.; Anderson, R. K.; Larsen, R. D.;
Verhoeven, T. R.; Reider, P. J .; Xiang, Y. B.; Belley, M.; Leblanc, Y.;
Labelle, M.; Prasit, P.; Zamboni, R. J . J . Org. Chem. 1993, 58, 3731.
(
(
(
(
(
3) Corey, E. J .; Helal, C. J . Angew Chem. Int. Ed. 1998, 37, 1986.
4) Corey, E. J .; Link, J . O. J . Org. Chem. 1991, 56, 442.
5) Mori, K.; Amaike, M.; Itou, M. Tetrahedron 1988, 29, 3423.
6) Corey, E. J .; Reichard, G. A. Tetrahedron Lett. 1989, 30, 5207.
7) Corey, E. J .; Gavai, A. V. Tetrahedron Lett. 1988, 29, 3201.
1
0.1021/jo016257c CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/05/2002

Oxazaborolidine-Catalyzed Reduction
J . Org. Chem., Vol. 67, No. 9, 2002 2971
Ta ble 1. Keton e Red u ction w ith BTHF a n d (R)-MeCBS
entry
ketone
acetophenone
% eea,b
% eeb,c
1
2
3
4
5
65
33
37
66
46
95
61
61
81
91
3-methyl-2-butanone
cyclohexyl methyl ketone
R-tetralone
3,3-dimethylbutanone
a
Commercial 1 M BTHF in THF was added to a mixture of
b
ketone and (R)-MeCBS. Enantiomeric excess was measured with
c
chiral column J & W 30 m × 0.25 mm CDX-B. Nonstabilized 1
M BTHF in THF was added to a mixture of ketone and (R)-
MeCBS.
F igu r e 1. Reduction of acetophenone with 9 nonstabilized
BTHF, [ NaBH -stabilized BTHF, and 2 nonstabilized BTHF
4
and (R)-MeCBS. The disappearance of acetophenone was
monitored by GC.
To determine general reaction conditions to minimize
over-reduction byproducts, we studied the reduction of
acetophenone with commercial 1 M BTHF (stabilized
with ∼0.005 M NaBH
4
, >98% purity)10 and MeCBS
catalyst. We were greatly disappointed that the addi-
tion of commercial 1 M BTHF to a mixture of aceto-
phenone and the MeCBS catalyst (5 mol % relative to
ketone) at ambient temperature gave 1-phenethanol with
low enantioselectivity (60% ee). Similar low enantio-
selectivities were obtained when commercial 1 M BTHF
was used in the MeCBS reduction of other ketones (see
Table 1).11
tions accelerated the BTHF reduction of ketones.14 They
found that the use of commercial BTHF for ketone
reduction resulted in ∼300 times faster reaction than
with freshly prepared BTHF solution. In a related paper,
they determined the kinetics of the borane reduction of
pinacolone catalyzed by phenyl oxazaborolidine deriva-
15
tives. Although they found that the presence of NaBH
4
stabilizer in commercial BTHF solutions accelerated the
direct borane reduction and oxazaborolidine-catalyzed
reduction, the enantioselectivity of the reduction was not
Since others have attributed such low selectivities to
8
the purity of the borane, we chose to address this issue.
For the reductions summarized in Table 1, we used
16
discussed.
1
1
commercial BTHF solution (>98% purity by B NMR
spectroscopy) that had been stored cold (0-5 °C). Under
these storage conditions, 11B NMR analysis has revealed
its stable nature for periods of at least 1 year. However,
if 1 M BTHF solutions are left at ambient temperature,
degradation by about 10% occurs in 40-50 d. The
To further investigate the role of borohydride in CBS-
catalyzed reductions, we prepared and used a 1 M BTHF
solution without the NaBH stabilizer. Higher enantio-
4
selectivities were obtained in the MeCBS-catalyzed re-
duction of the same ketones with this unstabilized BTHF
solution (see Table 1).
Evidently, a borohydride species present in commercial
BTHF is participating in nonselective ketone reduction
or interacting with oxazaborolidine catalyst to lower the
enantioselectivity of the CBS reduction. To rationalize
the lower selectivity, one must learn more about the
mechanisms involved and the relative rate of reduction
of the pathways. Since J ockel’s experiments were with a
higher amount of borohydride and under different condi-
tions, we decided to study the relative rate of reduction
of acetophenone with (1) nonstabilized BTHF; (2) NaBH -
4
stabilized commercial BTHF; and (3) nonstabilized BTHF/
MeCBS (5 mol % relative to ketone) mixture.
In each experiment, the BTHF solution was rapidly
added to a solution of acetophenone in THF at 0 °C. A
slight exotherm of approximately 6 °C was noted when
the nonstabilized BTHF was added to the ketone. On the
contrary, a very large exotherm was observed upon
addition of either commercial BTHF (to approximately
degradation products are di-n-butoxyborane, (n-BuO)
2
BH,
and tri-n-butylborate ((n-BuO)
3
B).12
To evaluate the effect of these degradation products
on selectivity, commercial BTHF kept at room temper-
ature for 99 days (84% hydride activity) was tested under
the same ketone reduction conditions. A higher selectivity
(82% ee compared to 65% ee) was observed in the
acetophenone reduction. The increased selectivity could
be related to (1) borate species, (2) alkoxyborane species,
or (3) borohydride species generated or destroyed during
the degradation process.
To better understand the increased selectivity with
degraded BTHF, we spiked commercial BTHF with
dibutoxyborane (10 mol % relative to BTHF) and carried
out the CBS reduction. The enantioselectivity using this
spiked BTHF solution was 66% ee, essentially the same
as unspiked BTHF. Therefore, we eliminated this species
as a contributor to the observed selectivity employing
commercial BTHF.
2
5 °C) or BTHF/MeCBS catalyst (to approximately 35 °C)
Recently, J ockel and Schmidt demonstrated that the
to the ketone.
1
3
borohydride concentration in commercial BTHF solu-
A plot of acetophenone disappearance versus time for
the three kinetic experiments showed that the MeCBS
reduction is the fastest (see Figure 1). In the absence of
MeCBS or borohydride, 50% of the acetophenone had
reacted with unstabilized BTHF in ∼8 min. The reduc-
tion in the presence of borohydride stabilizer was much
(
10) For the experiments, 1 and 2 M borane-tetrahydrofuran
solutions (>98% pure) were supplied from Callery Chemical Co. These
solutions are prepared by bubbling 99.5% pure diborane gas into
tetrahydrofuran, followed by addition of the required amounts of
4
NaBH stabilizer. This preparation eliminates any traces of boron
trifluoride in the solution, preventing the partial destruction of
stabilizer.
(
11) In all experiments, the addition rate of borane to ketone was
(13) The commercial source was Aldrich Chemical Co.
(14) J ockel, H.; Schmidt, R. J . Chem. Soc., Perkin Trans. 2 1997,
2719.
(15) Schmidt, R.; J ockel, H.; Schmalz, H.-G.; J ope, H. J . Chem. Soc.,
Perkin Trans. 2 1997, 2125.
(16) J ockel, H.; Schmidt, R.; J ope, H.; Schmalz, H.-G. J . Chem. Soc.,
Perkin Trans. 2 2000, 69.
kept uniform at 0.11 mL/min (0.11 mmol/min for 1 M and 0.22 mmol/
min for 2 M BTHF) using a syringe pump. The ratio of borane to ketone
was 0.67 to 1. MeCBS prepared in toluene was 5 mol % relative to
ketone. The percent enantiomeric excess (% ee) of product alcohol was
determined via chiral gas chromatography.
(12) Burkhardt, E. R.; Corella, J . A. US Patent 6,048,985.

2
972 J . Org. Chem., Vol. 67, No. 9, 2002
Nettles et al.
Ta ble 2. Keton e Red u ction w ith Na BH4-Sta bilized BTHF
a n d (R)-MeCBS
%
eeb
sodium
borohydride
concentration (M)
acetophenone
reduction
pinacolone
reduction
a
entry
1
2
3
4
5
6
nonec
0.0005
0.0025
0.0035
0.004
95
91
85
91
88
66
62
46
63
65
0.005
a
NaBH4-stabilized 1 M BTHF in THF was added to a mixture
of ketone and (R)-MeCBS (5 mol % relative to ketone). b Enantio-
meric excess was measured with chiral column J & W 30 m ×
F igu r e 2. Addition of 1 M BTHF with increasing amounts of
NaBH4 to 2 acetophenone or [ pinacolone.
c
0
.25 mm CDX-B. Nonstabilized 1 M BTHF in THF was added to
a mixture of ketone and (R)-MeCBS.
faster, with a t1/2 estimated at ∼0.3 min. However, in the
presence of MeCBS catalyst, the t1/2 could not be esti-
mated accurately, since the initial sample at 5 s showed
only 0.23% acetophenone remaining unreduced. The
exotherm and the rapid decrease in ketone with stabi-
lized BTHF solution suggests that a species in the
stabilized BTHF solution is playing a catalytic role in the
nonselective acetophenone reduction.
The kinetic measurements of J ockel et al. using stop-
flow techniques demonstrated that the borohydride-
catalyzed ketone reduction is 300 times faster than the
uncatalyzed reduction.13 Although a direct comparison
of our data with J ockel’s data cannot be made, even very
small amounts of borohydride in BTHF participate in
nonselective reduction and compete with the enantio-
selective MeCBS-catalyzed process, thus lowering the
overall enantioselectivity of the reduction.
F igu r e 3. Shelf life of unstabilized 1 M BTHF: 2 1 M BTHF
at 0 °C and [ 1 M BTHF at -20 °C.
To investigate the borohydride species in BTHF solu-
tion responsible for the drop in selectivity, we analyzed
the commercial 1 M BTHF by 11B NMR spectroscopy. The
1
1
spectrum revealed the presence of only NaB
δ -28.0) as a minor component in the BTHF solution.
Presumably, this NaB is produced from NaBH and
equiv of borane after loss of hydrogen. Addition of ∼0.7
mol % of NaB to nonstabilized BTHF prior to aceto-
3 8
H ( B NMR
3
H
8
4
2
F igu r e 4. Shelf life of unstabilized 2 M BTHF: [ 2 M BTHF
at 0 °C, 2 2 M BTHF at -20 °C, and 9 2 M BTHF at 0 °C
3
H
8
4
with 0.0035 M NaBH .
phenone reduction in the presence of 5 mol % MeCBS
gave an enantioselectivity of only 69% ee. This result
clearly demonstrates that a soluble anionic B-H- spe-
amount of NaBH4 stabilizer present in BTHF and the
enantioselectivity (see Figure 2).
cies in THF such as NaB
overall selectivity of the MeCBS reduction.
In another experiment, NaBH (5 mol % relative to
3 8
H can negatively impact the
The higher selectivities obtained with no stabilizer or
lower concentrations of borohydride point toward com-
mercialization of these BTHF solutions, provided the
shelf life is adequately long. Refrigerated storage shelf
life studies on these BTHF solutions, as well as un-
stabilized BTHF, indicated very little decomposition over
1 year at 0 °C. The decomposition was further suppressed
at -20 °C, indicating a shelf life longer than 1 year (see
Figures 3 and 4).
4
ketone) was added to ketone and MeCBS in THF followed
by addition of unstabilized BTHF. The enantioselectivity
of the reduction (89% ee) correlates to nonselective
reduction of ketone by two hydrides of the added NaBH
but the effect here was apparently not catalytic.
4
,
Throughout this paper, we will refer to sodium boro-
hydride (NaBH ) rather than NaB , because NaBH
is the added stabilizing agent. Since NaB is the only
borohydride species observed, we assume that the “boro-
4
3
H
8
4
3
H
8
Mod e of Ad d ition . The mode of addition can have a
dramatic effect on the overall enantioselectivity of a
reduction. The preferred mode of addition, studied and
hydride” concentration is directly related to NaBH
added.
4
rationalized by Merck co-workers, is ketone to BH /
3
MeCBS.17 Their mechanistic investigation pointed to
Bor oh ydr ide Con cen tr ation . To determine the mini-
mum concentration of NaBH that could be utilized to
4
several reaction intermediates capable of chiral reduction
but at a lower enantioselectivity than the MeCBS-
borane complex 2 (see Scheme 1). In the absence of
borohydride, these competing pathways can also lower
stabilize 1 M BTHF and still achieve a high enantio-
selectivity in the MeCBS reduction, we reduced pina-
colone and acetophenone with 1 M BTHF solutions
containing increasing amounts of NaBH
A plot of selectivity versus concentration of NaBH
BTHF showed a linear, inverse relationship between the
4
(see Table 2).
4
in
(
17) King, A. O.; Mathre, D. J .; Tschaen, D. M.; Shinkai, I. ACS
Symp. Ser. 1996, 641, 98.

Oxazaborolidine-Catalyzed Reduction
J . Org. Chem., Vol. 67, No. 9, 2002 2973
Sch em e 1
F igu r e 5. Addition of 2 acetophenone or [ pinacolone to 1
M BTHF with increasing NaBH
MeCBS at 23 °C.
4
concentrations and (R)-
Ta ble 3. Keton e Red u ction w ith Na BH4-Sta bilized BTHF
a n d (R)-MeCBS
%
eeb
sodium
borohydride
concentration (M)
acetophenone
reduction
pinacolone
reduction
a
entry
1
2
3
4
5
6
nonec
0.0005
0.0025
0.0035
0.004
95
94
95
93
87
92
91
90
83
82
0.005
F igu r e 6. Addition of 2 M BTHF with increasing NaBH
concentrations to acetophenone and (R)-MeCBS at 23 °C.
4
a
Ketone was added to a mixture of NaBH4-stabilized 1 M BTHF
in THF and (R)-MeCBS (5 mol % relative to ketone). b Enantio-
meric excess was measured with chiral column J & W 30 m ×
over.19 A number of other researchers have also observed
c
0
1
.25 mm CDX-B. Ketone was added to a mixture of nonstabilized
a substrate dependence on the stereoselectivity. As the
rate of catalyst turnover slows, the opportunity for
nonselective ketone reduction by “borohydride” increases.
Con cen tr a tion of BTHF Com p lex. We found that
the enantioselectivities of the MeCBS reduction of aceto-
phenone with commercial 2 M BTHF (80% ee) were
superior to that of commercial 1 M BTHF (65% ee).
Commercial 2 M BTHF solutions are stabilized with
M BTHF in THF and (R)-MeCBS.
the enantioselectivity, especially with ketones of specific
structural types that react slower.
As expected, higher enantioselectivities in the MeCBS
reduction were obtained when acetophenone (89% ee) or
pinacolone (82% ee) was slowly added to a mixture of
commercial BTHF and MeCBS (see Table 3 and Figure
∼
0.008 M NaBH
amount per borane, “BH
4
, which is close to the same relative
”, as in 1 M BTHF. The higher
1
8
5
).
3
Following the trend observed, even better selectivities
were obtained when acetophenone or pinacolone was
added to MeCBS catalyst and BTHF solutions with lower
selectivity appears related to the actual addition rate of
borane. For 2 M BTHF, the addition rate was 0.22 mmol/
min compared to 0.11 mmol/min for 1 M BTHF.
Our correlation between enantioselectivity and increas-
ing BTHF concentration agrees very well with J ockel’s
results.16 They found that the reaction was first-order in
catalyst and the rate of the catalytic reaction increases
with increasing borane concentration. On the basis of
J ockel’s and our results, the higher concentration of
BTHF is presumably contributing to a more rapid
turnover of the MeCBS catalyst.
4
concentrations of NaBH (see Table 3 and Figure 5).
During the ketone addition, excess BTHF in the
reaction solution may play a role in the MeCBS catalytic
cycle by driving the intermediates in the catalytic cycle
back to active MeCBS-borane complex 2 through a six-
membered-hydride bridged complex 5 (see Scheme 1).
Comparison of the two modes of addition strongly sug-
gests that the equilibration between the intermediate
complexes and borane (BH
nonselective reduction presumably catalyzed by boro-
hydride species.
3
) was critical to minimize the
The data shown in Figure 6 for reduction of aceto-
phenone using 2 M BTHF also confirms the linear trend
seen with 1 M BTHF of higher enantioselectivity with
lower borohydride concentration.
We noticed the slope is steeper for pinacolone than
acetophenone, showing that some steric and/or electronic
factors are involved in the reduction pathways. Research-
ers at Merck were able to observe intermediates in the
The optimum conditions for high enantioselective
ketone reductions were demonstrated with simultaneous
addition experiments. Recently, Tillyer and co-workers
reported better selectivities in the synthesis of cyclic
amino alcohols using simultaneous addition of the ketone
catalytic cycle (see Scheme 1) by 11B NMR and showed
evidence for ketone structure effecting catalyst turn-
2
0
2
and BH3-SMe as the borane source to the catalyst.
Under these reaction conditions, the MeCBS to ketone
(18) The BTHF and MeCBS catalyst were stirred for 10 min before
ketone addition to ensure equilibration of the MeCBS-borane complex,
since a control experiment where ketone addition was immediately
added gave 84% ee.
(19) Douglas, A. W.; Tschaen, D. M.; Reamer, R. A.; Shi, Y.-J .
Tetrahedron: Asymmetry 1996, 7, 1303.

2
974 J . Org. Chem., Vol. 67, No. 9, 2002
Nettles et al.
F igu r e 7. Effect of temperature on the % ee: 2 commercial
M BTHF added to mixture of acetophenone and (R)-MeCBS
and [ acetophenone added to a mixture of commercial 1 M
BTHF and (R)-MeCBS.
F igu r e 8. Effect of the catalyst concentration on % ee:
commercial 1 M BTHF added to mixture of acetophenone and
(R)-MeCBS (mol % relative to ketone).
1
variable. When commercial BTHF was added to ketone
and MeCBS, higher enantioselectivity was obtained with
5% loading over 2.5%, but >5% MeCBS did not signifi-
cantly increase the selectivity (see Figure 8). Stone also
reported a similar nonlinear effect with 1-10 mol % of
PhCBS using borane-1,4-thioxane complex.2
ratio is probably near 1:1, which may help to minimize
the influence of the borohydride species on the overall
selectivity.21 Indeed, we found that by simultaneously
adding commercial 1 M BTHF and acetophenone to
MeCBS in THF, high enantioselectivity was obtained
4
(
95% ee).22 Addition of the acetophenone two times faster
With this mode of addition, the MeCBS catalytic
than BTHF, or vice versa, gave equally high enantio-
selectivities (both 96% ee), demonstrating that equal
rates were not critical. Simultaneous addition of BTHF
and ketone to a continuous reactor with supported
oxazaborolidine catalyst, as reported by BASF using
system is borane (BH ) starved, resulting in only a small
amount of MeCBS actually participating in the catalytic
3
cycle. As a consequence, the higher amounts of MeCBS
do not overcome the nonselective “borohydride” pathway.
Since temperature and catalyst-load optimization did
not sufficiently increase the enantioselectivity, an alter-
native process to increase the selectivity in the MeCBS
with commercial BTHF was sought, especially in the case
where the ketone substrate has limited solubility, ruling
out simultaneous addition techniques.
2
3
borane-dimethyl sulfide complex, would be advanta-
geous for reducing costs and environmental impact.
Rea ction Tem p er a tu r e. Because other groups had
2
4
observed temperature effects on enantioselectivity, we
explored the effect of this variable on the selectivity of
MeCBS-catalyzed reduction. When commercial BTHF
was added to ketone, the results showed decreasing
enantioselectivity as the temperatures was lowered from
Ad d ition of Lew is Acid s. Researchers at Merck have
used additives to achieve a high degree of enantioselec-
21
tivity in the oxazaborolidine-catalyzed ketone reduction.
3
5 to 0 °C and then increasing selectivity below 0 °C to
Triethylamine was used with stoichiometric methyl-
oxazaborolidine-borane complex to trap the mono-
alkoxyborane product as an amine complex, thus inhibit-
ing less selective reduction of ketone by the mono-
alkoxyborane. 2-Propanol was also used to intercept the
monoalkoxyborane, yielding a mixed dialkoxyborane that
about -25 °C (see Figure 7). As reported by several
groups,25 the selectivity dropped below -25 °C. The
higher enantioselectivity at temperatures between 0 to
-
25 °C may reflect the faster regeneration of active
catalyst (BH /MeCBS) or a slowing in the rate of ketone
reduction by the BTHF/“NaBH ” system.
3
26
4
is very slow to reduce ketones. Both of these additives
Although higher enantioselectivities were observed
when acetophenone was added to a mixture of com-
mercial BTHF and MeCBS, the effect of temperature was
less dramatic (see Figure 7).
Ca ta lyst Loa d in g. We also investigated the MeCBS-
catalyzed reduction while changing the catalyst-loading
make less than efficient use of the borane source.
If simultaneous addition mode is not feasible, the
borohydride species in BTHF solutions must be removed
or deactivated to improve the overall enantioselectivity
of the MeCBS-catalyzed reduction. The initial observa-
tion that degraded BTHF gives a higher selectivity is not
a viable option for commercial utility. The reaction of
(
20) Tillyer, R. D.; Boudreau, C.; Tschaen, D.; Dolling, U.-H.; Reider,
P. J . Tetrahedron Lett. 1995, 36, 4337.
21) Cai and co-workers reported very high enantioselectivities in
4
Lewis acids with NaBH is a well-known process to make
diborane, but would a Lewis acid react with the NaB
observed or other trace “borohydrides”?
3 8
H
(
the reduction of representative ketones using oxazaborolidine-BH
3
complex as the stoichiometric reducing agent: Cai, D.; Tschaen, D.;
Shi, Y.-J .; Verhoeven, T. R.; Reamer, R. A.; Douglas, A. W. Tetrahedron
Lett. 1993, 34, 3243.
Upon screening several Lewis acids as additives, we
found that the selectivity of the MeCBS reduction was
effectively restored (see Table 4). Of the Lewis acids
investigated, BF3-THF complex proved to be the best,
with an enantioselectivity of 90% ee for the 1-phenyl-
ethanol product.
(
22) 3 mmol of 1 M BTHF and 5 mmol of ketone were added over
0 min.
23) (a) Woltinger, J .; Bommarius, A. S.; Drauz, K.; Wandrey, C.
3
(
Org. Process Res. Dev. 2001, 5, 241. (b) Giffels, G.; Felder, M.; Kragl,
U.; Wandrey, C.; Bommarius, A.; Bolm, C.; Derrien, N.; Drauz, K. US
Patent 6,180,837.
Further optimization of the conditions to utilize BF
THF in the reaction by adding the BTHF solution to a
mixture of ketone, MeCBS, and BF -THF gave a slightly
3
-
(
24) Stone, G. B. Tetrahedron: Asymmetry 1994, 5, 465.
(25) Literature reports have shown that, depending on the borane
source, the selectivity of the MeCBS reduction is influenced by the
reaction temperature. See: (a) Corey, E. J .; Bakshi, R. K.; Shibata, S.
J . Am. Chem. Soc. 1987, 109, 5551; (b) Mathre, D. J .; Thompson, A.
S.; Douglas, A. W.; Hoogsteen, K.; Carrol, J . D.; Corley, E. G.;
Grabowski, J . J . Org. Chem. 1993, 58, 2880; (c) Stone, G. B. Tetra-
hedron: Asymmetry 1994, 5, 465; (d) J iang, Y. Z.; Qin, Y.; Mi, A. Q.
Chin. Chem. Lett. 1995, 6, 9.
3
(26) (a) Shi, Y.-J .; Cai, D.; Dolling, U.-H.; Douglas, A. W.; Tschaen,
D. M.; Verhoeven, T. R. Tetrahedron Lett. 1994, 35, 6409. (b) Tschaen,
D. M.; Abrahamson, L.; Cai, D.; Desmond, R.; Dolling, U.-H.; Frey, L.;
Krady, S.; Shi, Y.-J .; Verhoeven, T. R. J . Org. Chem. 1995, 60, 4324.

Oxazaborolidine-Catalyzed Reduction
J . Org. Chem., Vol. 67, No. 9, 2002 2975
Ta ble 4. Effect of Lew is Acid s on th e En a n tioselectivity
of th e (R)-MeCBS Red u ction of Acetop h en on e w ith
Com m er cia l 1 M BTHF Solu tion s
The mechanistic implications of this work merit further
study of this important asymmetric reduction process.
entry
Lewis acid
% eea,b
Exp er im en ta l Section
1
2
3
4
5
ZrCl4
AlCl3
FeCl3
BF3
85.6
83.0
86.0
90.4
82.6
Gen er a l Con sid er a tion s. All reactions were carried out
in predried glassware (1 h, 150 °C) under a dry nitrogen
atmosphere. Tetrahydrofuran (THF) and ketones were dried
over 4 Å molecular sieves. BF3-THF was distilled under
TiCl4
a
Commercial 1 M BTHF solution (∼0.005 M NaBH4) in THF
vacuum prior to use. Sodium octahydrotriborate (NaB
3 8
H ) was
prepared according to a literature method.30 BTHF solution
and the Lewis acid (∼3-8 mol %) was added to a mixture of ketone
b
and (R)-MeCBS. Enantiomeric excess was measured with chiral
(
>98% pure containing ∼0.005 M NaBH for 1 M and ∼0.008
4
column J & W 30 m × 0.25 mm CDX-B.
M for 2 M), 99.5% pure diborane gas, and (R)-MeCBS (1 M in
toluene) were used directly as obtained from commercial
sources unless otherwise noted. Unstabilized BTHF solutions
were prepared by bubbling pure diborane into tetrahydrofuran.
better selectivity of 93% ee. The flexibility of where the
BF -THF additive can be introduced into the reaction
3
BTHF solutions containing <0.005 M NaBH
by addition of pure diborane gas followed by addition of the
required amounts of NaBH . Chiral alcohol products were
4
were prepared
shows the utility of this Lewis acid additive to increase
the selectivity of the MeCBS reaction.
To demonstrate that the Lewis acid, BF3-THF, was not
enhancing the selectivity of the reduction, we carried out
the MeCBS reduction of acetophenone with nonstabilized
4
analyzed via gas chromatography techniques.
Red u ction of Acetop h en on e, BTHF Ad d ed to Aceto-
p h en on e/THF /MeCBS. To a solution of acetophenone (2.1 g,
17.1 mmol) in THF (17 mL) was added (R)-MeCBS in toluene
(0.86 mL of 1.0 M, 0.86 mmol). A BTHF solution in THF (10.3
mL of 1.0 M, 10.3 mmol) was then slowly added at ambient
temperature over 30 min. After completing the addition of
BTHF, the reaction mixture was allowed to stir for 10 min
before quenching with a solution of HCl in H O (10 mL of 2
2
M, 20 mmol). Diethyl ether (20 mL) was added, and the organic
phase was washed with saturated aqueous solutions of KCl
1
M BTHF in the presence of 8 mol % of BF3-THF
complex. The selectivity observed under these reaction
conditions was 95% ee.27 Evidently, the BF3-THF serves
to remove the borohydride and not to catalyze the
reduction.
The addition of NaBF
4
, a byproduct of the reaction,28
did not significantly affect the enantioselectivity of the
reduction (93% ee).29 This result does not rule out the
(3 × 8 mL), NaHCO (3 × 12 mL), and KCl (3 × 8 mL). The
3
possibility that other byproducts from the BF
3
addition
2 4
organic phase was then dried over Na SO , filtered, and
analyzed by chiral GC to determine optical purity of the
may slightly lower the enantioselectivity as compared to
the reduction with nonstabilized BTHF solutions (95%
ee).
1
-phenylethanol product.
Redu ction of Acetoph en on e, Acetoph en on e/THF Added
to BTHF /MeCBS. To a BTHF solution in THF (10.3 mL of
.0 M, 10.3 mmol) was added (R)-MeCBS in toluene (0.86 mL
1
Con clu sion
of 1.0 M, 0.86 mmol). After stirring the BTHF/MeCBS reaction
mixture for 10 min, acetophenone (2.1 g, 17.1 mmol) in THF
(17 mL) was then slowly added to the reaction flask at ambient
temperature over 30 min. After completing the addition of
acetophenone in THF, the reaction solution was allowed to stir
for 10 min before quenching with a solution of HCl in H O (10
2
mL of 2 M, 20 mmol). Diethyl ether (20 mL) was added, and
the organic phase was washed with saturated aqueous solu-
In conclusion, a soluble borohydride species in BTHF
solution was detrimental to the enantioselective reduc-
tion of ketones using the oxazaborolidine catalyst. From
the relative rate studies, the MeCBS-catalyzed reduction
pathway is much faster than the “borohydride”-catalyzed
pathway. However, the borohydride participates in a
nonselective reduction, effectively competing with the
enantioselective process, resulting in an overall drop in
enantioselectivity.
tions of KCl (3 × 8 mL), NaHCO
3
(3 × 12 mL), and KCl (3 ×
8
mL). The organic phase was then dried over Na
2
SO , filtered,
4
and analyzed by chiral GC to determine the optical purity of
the 1-phenylethanol product.
Upon reaction parameter optimization, the nonselec-
tive borohydride reduction pathway was dramatically
minimized via simultaneous addition of substrate and
BTHF to MeCBS catalyst. During the simultaneous
addition, the ketone/MeCBS catalyst ratio is close to
stoichiometric, as opposed to during BTHF addition to
ketone and 5 mol % MeCBS, where the catalytic cycle
may be deficient in borane.
When the ketonic substrate has limited solubility,
3
requiring BTHF addition, the use of BF as an additive
proved very effectively to restore the enantioselectivity
to acceptable levels. On the basis of our control experi-
ments, the BF additive was reacting with the boro-
3
Red u ction of Acetop h en on e, Sim u lta n eou s Ad d ition
of BTHF /Keton e to MeCBS. To a solution of (R)-MeCBS in
toluene (0.25 mL of 1.0M, 0.25 mmol) were added simulta-
neously BTHF solution in THF (3.2 mL of 0.94 M, 3 mmol)
and acetophenone (0.60 g, 5 mmol) in THF (4.4 mL) at ambient
temperature over 30 min. After completing the addition of the
BTHF and ketone solutions, the reaction solution was allowed
to stir for 10 min before quenching with a HCl solution in H O
2
(10 mL of 2 M, 20 mmol). Diethyl ether (20 mL) was added,
and the organic phase was washed with saturated aqueous
solutions of KCl (3 × 8 mL), NaHCO
3
(3 × 12 mL), and KCl (3
×
8 mL). The organic phase was then dried over Na
2
SO ,
4
filtered, and analyzed by chiral GC to determine the optical
purity of the 1-phenylethanol product.
Red u ction of Acetop h en on e, BTHF /BF 3-THF Ad d ed
t o Acet op h en on e/TH F /MeCBS. To a solution of aceto-
phenone (2.0 g, 16.6 mmol) and THF (14.7 mL) was added (R)-
MeCBS in toluene (0.83 mL of 1.0 M, 0.83 mmol). BF3-THF
hydride species and not catalyzing the enantioselective
process.
(
0.1 g, 0.81 mmol) was then added to the BTHF solution. A
(
27) The mode of addition for this experiment was acetophenone
added to the BTHF/MeCBS/BF -THF mixture.
28) Sodium tetrafluoroborate is also a byproduct of in situ generated
borane from sodium borohydride and boron trifluoride.
29) The mode of addition for this experiment was acetophenone
added to unstabilized 1 M BTHF/MeCBS/NaBF mixture.
BTHF solution in THF (10.0 mL of 1.0 M, 10.0 mmol) solution
3
(
(
(30) Dewkett, W. J .; Grace, M.; Beall, H. Inorg. Synth. 1974, 15,
4
111.

2
976 J . Org. Chem., Vol. 67, No. 9, 2002
Nettles et al.
was then slowly added at ambient temperature over 30 min.
After completing the addition of the BTHF solution, the
reaction solution was allowed to stir for 10 min before
analyzed by chiral GC to determine the optical purity of the
1-phenylethanol product.
quenching with a HCl solution in H
mmol). Diethyl ether (20 mL) was added, and the organic
phase was washed with saturated aqueous solutions of KCl
2
O (10 mL of 2 M, 20
Ack n ow led gm en t. The authors are grateful to Mr.
Charles S. Bello for GC analysis of secondary alcohol
products.
(
3 × 8 mL), NaHCO
3
(3 × 12 mL), and KCl (3 × 8 mL). The
organic phase was then dried over Na
2
SO , filtered, and
4
J O016257C