The influence of fluorine on the asymmetric reduction of fluoromethyl ketones

Article abstract of DOI:10.1016/j.jfluchem.2007.04.015

A comparative study of the asymmetric reduction of representative aryl and alkyl α-fluoro- and α-chloromethyl ketones using (-)-diisopinocampheylchloroborane [(-)-DIP-Chloride] and (-)-B-isopinocampheyl-9-borabicyclo[3.3.1]nonane [R-Alpine-Borane] has been made. It was observed that DIP-Chloride is superior in terms of the rate and enantioselectivity for both classes of halo-ketones. While the reduction of monofluoroacetone and trifluoroacetone with DIP-Chloride provided the product alcohols in 61% ee and 96% ee, respectively, the reduction of difluoroacetone yielded only 5% ee. The influence of a lone halogen atom was not observed for monochloroacetone, all of which point towards a chelating effect of monofluoroacetone on the Lewis acidic chloroborane.

Full text of DOI:10.1016/j.jfluchem.2007.04.015

Journal of Fluorine Chemistry 128 (2007) 844–850
www.elsevier.com/locate/fluor
The influence of fluorine on the asymmetric reduction
of fluoromethyl ketones
*
P. Veeraraghavan Ramachandran , Baoqing Gong, Aleksandar V. Teodorovic
´
Herbert C. Brown Center for Borane Research, Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette,
IN 47907-2084, United States
Received 22 February 2007; received in revised form 10 April 2007; accepted 10 April 2007
Available online 19 April 2007
Abstract
A comparative study of the asymmetric reduction of representative aryl and alkyl a-fluoro- and a-chloromethyl ketones using (ꢀ)-
diisopinocampheylchloroborane [(ꢀ)-DIP-ChlorideTM] and (ꢀ)-B-isopinocampheyl-9-borabicyclo[3.3.1]nonane [R-Alpine-Borane1] has been
made. It was observed that DIP-ChlorideTM is superior in terms of the rate and enantioselectivity for both classes of halo-ketones. While the
reduction of monofluoroacetone and trifluoroacetone with DIP-ChlorideTM provided the product alcohols in 61% ee and 96% ee, respectively, the
reduction of difluoroacetone yielded only 5% ee. The influence of a lone halogen atom was not observed for monochloroacetone, all of which point
towards a chelating effect of monofluoroacetone on the Lewis acidic chloroborane.
# 2007 Published by Elsevier B.V.
Keywords: Fluoroketones; Chloroketones; Asymmetric reduction; DIP-ChlorideTM; Enantioselectivity
1. Introduction
with chloroacetophenones. Herein, we report the results of our
study, which points to a fluorine-mediated chelation [8] with
the stronger Lewis acidic reagent 1.
Examination of the effects of fluorine substitution in
organic molecules is important for the successful development
of pharmaceuticals, agrochemicals and novel materials [1].
Optically active fluorinated alcohols are important building
blocks or end-products in organic synthesis [2] and the steric
and electronic effects of fluorine substitution in ketones play a
vital role in reactions involving them [3]. The reduction of
fluoroacetophenones with (ꢀ)-B-chlorodiisopinocampheyl-
borane [(ꢀ)-DIP-ChlorideTM, 1] [4] and B-isopinocam-
pheyl-9-borabicyclo[3.3.1]nonane [R-Alpine-Borane1, 2]
[5] (Fig. 1) reported by us earlier [6] showed that the reagent
1 has more influence on the enantiomeric excesses achieved for
the product fluoro-alcohols [7]. To further understand the
influence of fluorine atoms on the reductions, we undertook a
study of the rate and enantioselectivities in the reduction of the
simplest of fluoroketones, mono-, di- and trifluoroacetones
with 1 and 2 and compared them with chloroacetones. The
earlier results on fluoroacetophenones were also compared
2. Results and discussion
We began the study with the comparison of the reduction of
chloro- and fluoroacetophenones with 1 and 2. The reduction of
2,2-dichloroacetophenone (5a) with 1 in ethyl ether (Et2O) at
ꢀ25 8C is very slow, complete in only 5 days. The usual
diethanolamine workup provides 2,2-dichloro-1-phenethanol
(6a) in 78% yield and 82% ee in the (R)-isomer (Eq. (2)) [9]. This
contrasts with the fast reduction (0.5 h) of 2-chloroacetophenone
(3a) with 1, which also provides (R)-2-chloro-1-phenethanol
(4a) in 95% ee (Eq. (1)) [10]. It is worth noting that, in the case of
fluoro-substituted acetophenones, the difluoro derivative (5b)
reacted faster than the monofluoroacetophenone (3b) [6a].
Although, the two chlorine atoms in 5a should make the carbonyl
more reactive than in 3a, the retarded rate of the reduction could
be due to the increased steric bulk as is the norm with DIP-
ChlorideTM reductions. The reaction of 2,2,2-trichloroacetophe-
none (7a) with 1 is even slower at ꢀ25 8C, and was conducted
under neat condition, at rt which require 22 days for near
completion to provide (S)-2,2,2-trichloro-1-phenethanol (8a) (of
* Corresponding author. Tel.: +1 765 494 5303; fax: +1 765 494 0239.
E-mail address: chandran@purdue.edu (P.V. Ramachandran).
0022-1139/$ – see front matter # 2007 Published by Elsevier B.V.
doi:10.1016/j.jfluchem.2007.04.015

P.V. Ramachandran et al. / Journal of Fluorine Chemistry 128 (2007) 844–850
845
unreacted 5a is recovered and the product 6a with (R)-
configuration is obtained in 83% ee. The reduction of 7a with 2
is even slower, only 45% complete in 22 days (Eq. (6)). The
product 8a obtained after recovering 55% of 7a showed an ee of
27% in the (R)-isomer, similar to the reduction of 7b with 2
(32% ee) [6a]. The (R)-configuration of the trichlorophenetha-
nol 8a and trifluorophenethanol 8b from Alpine-Borane
reduction is in contrast to the (S)-8a and (S)-8b obtained from
Fig. 1. Pinane-based asymmetric reducing agents.
opposite stereochemistry in contrast to 4a and 6a) in 91% ee (Eq.
(3)). This is comparable to the product from the reduction of
2,2,2-trifluoroacetophenone (7b) when the reaction was com-
plete in 24 h under neat condition at rt and provided the (S)-
alcohol in 90% ee (Eq. (3)) [11]. The decreased rate in the case of
7a might be due to the steric requirements of the three chlorine
atoms.
reduction of 7a and 7b, respectively, with DIP-ChlorideTM
.
Clearly, the net effect of the steric and electronic factors of the
halogens in the ketone affects the rate of the reduction and
influences the configuration of the product as well. The results
of the reduction of all of the above aryl (phenyl) halomethyl
ketones with 1 and 2 are summarized in Table 1.
We now chose chloroacetone (9a), 1,1-dichloroacetone
(11a) and 1,1,1-trichloroacetone (13a) for this study since a
methyl group should be sterically smaller than the chlor-
omethyl groups and the effect of the sterics versus electronics
can probably be differentiated. In addition, the differences in
the % ee and the configuration of the product alcohols from the
reduction of the chloracetones and corresponding fluoroace-
tones, 1-fluoroacetone (9b), 1,1-difluoroacetone (11b) and
1,1,1-trifluoroacetone (13b), should provide a better picture of
the electronic and/or steric effects due to the absence of the
phenyl ring.
The reduction with 1 is believed to follow a six-membered
transition state [4] with the hydride transfer as the rate-limiting
step. The coordination of the carbonyl oxygen to the boron
atom of the reagent is also equally important. Our results show
that the trichloromethyl group is probably sterically more
demanding than even a tert-butyl group since the reduction of
pivalophenone is relatively faster than the reduction of 7a.
Corey et al. have made a detailed analysis of the structure and
reduction of a series of trichloromethyl ketones including tert-
butyl and adamantyl trichloromethyl ketones and have
concluded that the trichloromethyl group effectively screens
the lone pair of the carbonyl oxygen more than even the rigid
adamantyl group [12].
Reagent 1 reduces 9a in Et2O at ꢀ25 8C within 90 min to
provide (R)-1-chloro-2-propanol (10a) in 65% yield and 18%
ee (Eq. (7)). Surprisingly, the configuration of the product
shows that the methyl group acts as the enantiocontrolling
larger group in the tentative transition state model, probably
due to the electronic effect of the chlorine atom on the reagent
1. In comparison, the corresponding reduction of 9a with 2
provides the (S)-alcohol in 16% ee, with the chloromethyl
acting as the larger group. However, the reduction of the
fluoroacetone, 9b, with 1 in Et2O at ꢀ25 8C is complete in 1 h
providing 1-fluoro-2-propanol (10b) in a relatively high 61% ee
in the (S)-isomer (Eq. (8)) [13], which is quite unexpected when
compared with the 4% ee obtained for the reduction of 2-
butanone [4b] and 18% ee obtained for of 10a. A similar effect
of a lone fluorine atom was also noticed in the reduction of
The reduction of chloroacetophenones with 2 was then
undertaken to compare with the above results and, probably,
delineate the electronic and steric effects in this reduction.
Since both 1 and 2 have similar reaction mechanism, any
difference in the rate and configuration may be attributed to the
electronic environment of the reagent due to the chlorine atom
in 1. Unlike the relatively fast reduction of 3a with 2, complete
in 3 days, to produce 4a in 96% ee (R) (Eq. (4)), the presence of
two chlorine atoms in 5a retards the rate and the reduction is
only 70% complete in 22 days (Eq. (5)). Thirty percent of

846
P.V. Ramachandran et al. / Journal of Fluorine Chemistry 128 (2007) 844–850
Table 1
Asymmetric reduction of a-fluoro- and a-chloroacetophenones with (ꢀ)-DIP-Chloride and R-Alpine-Borane
Entry
PhCOR
Reagent
Reaction condition
Haloalcohol
Controlb
#
R
Solv.
Temp.
Time
Yield (%)
ee (%)a
Conf.
1
2
3ac
3ad
3be
3be
5a
CH2Cl
CH2Cl
CH2F
CH2F
CHCl2
CHCl2
CHF2
CHF2
CCl3
1
2
1
2
1
2
1
2
1
2
1
2
Et2O
None
Et2O
None
Et2O
None
Et2O
None
Et2O
None
None
None
ꢀ25 8C
0.5 h
60
75
80
80
78
70
90
69
65
62
90
57
95
96
95
89
82
83
85
97
91
27
90
32
R
R
R
R
R
R
R
R
S
Ph
rt
8 h
Ph
3
ꢀ25 8C
1 h
Ph
4
rt
1 h
Ph
5
ꢀ25 8C
5 days
22 daysf
0.5 h
Ph
6
5a
rt
Ph
7
5be
5be
7a
ꢀ25 8C
Ph
8
rt
4 days
22 days
22 daysg
24 h
Ph
9
ꢀ25 8C
CCl3
Ph
10
11
12
7a
CCl3
rt
rt
rt
R
S
7be
7be
CF3
CF3
Ph
CF3
45 daysh
R
a
ee determined as their MTPA or TFA derivative on a capillary GC.
b
c
d
e
f
Enantiocontrolling group based on the proposed mechanism of the reduction, ref. [4b].
From ref. [5c].
From ref. [4b].
From ref. [6a].
For 70% reaction, 30% ketone was recovered.
For 45% reaction. 55% ketone was recovered.
For 90% reaction.
g
h
monofluoromethyl acetylenic ketones [14] and 1-fluoro-2-
octanone with 1 [6a] (Fig. 2). It is worthy to note that the
interaction between the fluorine atom of the ketone 9b and the
chlorine atom of the reagent 1 has more effect on the chiral
outcome than the interaction between the chlorine atoms of the
ketone 9a and 1. This could be due to the chelating effect of the
fluorine and the stronger Lewis acidic 1 (Fig. 3). This effect is
further substantiated by the lack of dramatic configurational
changes in Alpine-Borane reductions of fluoroketones as well.
difluoro compound compared to the a-fluoro compound was
also observed for the reduction of a-fluorooctanones with 1
when the % ee changed from 40% (R) to 32% (S) (Fig. 2) [6a].
However, there is not much change in ee between the products
from the reduction of monochloro- and dichloroacetone (9a and
11a, respectively) with 2.
The reduction of trichloroacetone (13a) with 1 is very
slow, compared to mono- and dichloroacetones at ꢀ25 8C in
Et2O. Even under neat condition, at rt, the reaction took 3
days for completion and the product is obtained in 78% yield
and 91% ee (Eq. (10)). Yet, the reaction rate (3 days) is faster
compared to the reduction of the phenyl analog (7a) (22
days), the reasons of which are unclear, at present. As is
normal with the reduction of perfluoroalkyl ketones with 1,
the reduction of 13b, at ꢀ25 8C, provides the (S)-alcohol in
96% ee.
The reduction of dichloroacetone 11a with 1 is complete
within 2 h and provides the product 1,1-dichloroethanol (12a)
in 72% yield and 15% ee (Eq. (9)). In comparison, the reduction
of 1,1-difluoroacetone (11b) with 1 provides the alcohol in only
5% ee. A similar change in enantioselectivity for an a,a-
Fig. 3. Transition state model depicting the chelation effect in the reduction of
fluoroacetone with DIP-Chloride.
Fig. 2. Theeffect offluorineon%eeinreductionsoffluoromethylketoneswith1.

P.V. Ramachandran et al. / Journal of Fluorine Chemistry 128 (2007) 844–850
847
Table 2
Asymmetric reduction of a-fluoro- and a-chloroacetones with (ꢀ)-DIP-Chloride and R-Alpine-Borane
Entry
MeCOR
Reagent
Reaction condition
Haloalcohol
Yield (%)
Controlb
#
R
Solv.
Temp.
Time
ee (%)a
Conf.
1
2
9a
9a
9b
9b
11a
11a
11b
11b
13a
13a
13bd
13b
CH2Cl
CH2Cl
CH2F
CH2F
CHCl2
CHCl2
CHF2
CHF2
CCl3
1
2
1
2
1
2
1
2
1
2
1
2
Et2O
None
Et2O
None
Et2O
None
Et2O
None
Et2O
None
None
None
ꢀ25 8C
1.5 h
3 days
1 h
65
68
80
68
72
75
80
54
78
70
72
60
18
16
61
13
15
62
5
R
Me
rt
S
CH2Cl
CH2F
CH2F
Me
3
ꢀ25 8C
Sc
Sc
Rc
Sc
Rc
Sc
Sc
S
4
rt
3 days
2 h
5
ꢀ25 8C
6
rt
4 days
2 h
CHCl2
Me
7
ꢀ25 8C
8
rt
3 days
3 days
22 days
4 h
16
91
24
90
82
CHF2
CCl3
CCl3
CF3
9
ꢀ25 8C
10
11
12
CCl3
rt
rt
rt
CF3
S
CF3
4 days
S
CF3
a
ee determined as their MTPA, benzoate or TFA derivative on a capillary GC.
Based on the proposed mechanism of the reduction (ref. [4b]).
Based on analogy.
b
c
d
From ref. [11].
To unequivocally delineate the influence of the chlorine of
the reagent on the reduction, the haloacetones were now
reduced with 2. The reduction of 9a under neat conditions, at rt,
is complete within 3 days and the product is obtained in 68%
yield and 13% ee, with the same stereochemistry as obtained
from the reduction with 1 (Eq. (11)). In contrast to the reduction
of 9b with 1 in 61% ee, the reduction with 2 provided only 13%
ee for 10b with the same configuration.
The rate of the reduction of 13a with 2 is slower than the
dichloro analog 11a, complete within 22 days, producing 14a in
70% yield and 24% ee. The product reveals the same
configuration of the product obtained from a reduction of
13a with 1. We believe this to be the (S)-isomer. On the other
hand, reduction of 13b with 2 is little faster, complete in 7 days,
and provides 60% of the product alcohol in 82% ee in the (S)-
isomer (Eq. (13)).
The results of the reduction of all of the above alkyl
halomethyl ketones with 1 and 2 are summarized in Table 2.
The reduction of 11a with 2 is complete in 4 days and the
product is obtained in 75% yield and 62% ee (Eq. (12)). As in
the case of 9a, the products from the reduction of 11a with 1 and
2 had the opposite configurations. On the basis of the earlier
results, we believe that we have obtained (R)-12a from the
reduction with 1 and (S)-12a from the reduction with 2. Again,
the differences in the % ee for the reduction of 11a with 1 and 2,
an overall change of 77% (15% R to 62% S), can probably be
accounted for only by considering the electronic environment
of reagent 1. The reduction of difluoroacetone with 2, complete
in 3 days, provides the product of opposite configuration in 16%
ee. Again on the basis of earlier results, we believe that we have
obtained (R)-12b with 1 and (S)-12b from reduction with 2.
3. Conclusion
In summary, we have studied the asymmetric reduction of a
representative series of aromatic and aliphatic a-halomethyl
ketones to understand the steric and/or electronic influences of
the fluorine atom on chiral reductions. DIP-ChlorideTM and
Alpine-Borane1 are used as the reagents in this study as they
differ in their electronic environments. As usual, all chloro- and
fluorosubstituted acetophenones are reduced with 1 in very high
ee. In the aliphatic series, 1 reduces trihaloketones in excellent
ee with the trihalomethyl group acting as the enantio-controller.
Surprisingly, we noticed a dramatic influence of the fluorine
atom in the reduction of monofluoroacetone, possibly due to a

848
P.V. Ramachandran et al. / Journal of Fluorine Chemistry 128 (2007) 844–850
chelating effect with the Lewis acidic chloroborane. This effect
is not noticeable in the reductions using 2 and in monochlor-
oacetone reduction with 1. Further molecular modeling studies
will provide additional insight into this mechanism of
fluoroketones.
was added dropwise. The mixture was then warmed to rt and
stirred for 3 h. The solid boron-amine complex was filtered,
washed with pentane and the combined filtrate was
concentrated. The residue was passed through a silica gel
pad to separate a-pinene and the product (pentane and
CH2Cl2 as eluents). The fraction containing the product was
concentrated and distilled using a Kugelrohr apparatus. The
alcohol was further purified by preparative GC using an SE-
30 column. The rotation was measured. The MTPA, MCF or
TFA derivative of the alcohol was prepared by the standard
procedure. Racemic alcohols of the ketones were obtained by
reduction with NaBH4. All of the racemic alcohols were
converted to the MTPA or MCF derivative and analyzed using
a GC fitted with an appropriate capillary column to obtain the
diastereomeric pairs of peaks. Then the derivatives of
optically active alcohols were analyzed to obtain the
enantiomeric excess.
4. Experimental
Unless otherwise noted, all manipulations were carried out
under an inert atmosphere using flame-dried glassware.
Techniques for handling air-sensitive compounds have been
previously described [15]. Anhydrous ethyl ether (Mallinck-
rodt) was used as received. 2,2-Dichloroacetophenone [16],
2,2,2-trichloroacetophenone [17] and 1,1-difluoroacetone [18]
were prepared according to the literature procedure. All other
materials, including reagents 1 and 2 were obtained from
Aldrich Chemical Co. (R)-(+)-a-Methoxy-a-(trifluoromethyl)-
phenylacetic acid (MTPA) was obtained from Aldrich
Chemical Co. and converted to the acid chloride using
Mosher’s procedure [19].
5.2. Reduction of chloroalkyl ketones with (R)-Alpine-
Borane
1
The H, 13C and 19F nuclear magnetic resonance (NMR)
spectra were plotted on a Varian Gemini-300 spectrometer
(300, 75 and 282 MHz, respectively) with a Nalorac-quad
5.2.1. General procedure
A 12 mmol of a 0.5 M THF solution of 2 was added to a
50 mL round-bottomed flask fitted as described above and the
solvent was removed under aspirator. The halomethyl ketone
(10 mmol) was then added and the mixture was stirred at rt. The
reaction was followed by 11B NMR of an aliquot dissolved in
ee. When the reaction was complete (11B NMR: 52 ppm),
acetaldehyde (3 mmol) was added at 8C and stirred at rt for
30 min. ee (20 mL) was then added to the reaction mixture
followed by enthanolamine (12 mmol) and stirred for 1 h. The
solid complex was filtered and washed with pentane. The
filtrate was concentrated and distilled to yield the alcohol.
Further purification and analysis was carried out as described
for reduction with 1.
1
probe. H NMR spectra were obtained using CDCl3 as the
solvent with either tetramethylsilane (TMS: 0 ppm) or
chloroform (CHCl3: 7.2 ppm) as the internal standard. 19F
NMR spectra were recorded in CDCl3 using CFCl3 or
1
trifluoroacetic acid (TFA) as the internal standard. H NMR
data are reported as chemical shifts (d ppm), multiplicity (s,
singlet; d, doublet; t, triplet; q, quartet; m, multiplet), coupling
constant (Hz) and integration. Chromatography was performed
on 40–60 mm silica gel (230–400 mesh). Analyses of the MTPA
[19] or MCF [20] derivatives were performed on a Hewlett-
Packard 5890A gas chromatograph using a Supelcowax glass
capillary column (15 m) or a SPB-5 capillary column (30 m), at
appropriate temperatures, and integrated using a Hewlett-
Packard 3390 A integrator. Analysis of the trifluoroacetates or
benzoates of alcohols were performed using a Chiraldex-GTA
[21] (23 m) capillary column.
5.3. 2,2-Dichloro-1-phenylethanol, 6a
5.3.1. (a) From reduction with (ꢀ)-1
2,2-Dichloroacetophenone (10 mmol) was treated with (ꢀ)-
DIP-Chloride (11 mmol) in ee at ꢀ25 8C. The reaction was
complete in 5 days. Workup as described in the general
5. Experimental procedure and analytical data for
products
21
procedure provided 6a in 78% yield. ½aꢁD ¼ ꢀ30:17 (ca. 3.0,
5.1. Reduction of chloroalkyl ketones and fluoroalkyl
ketones with (ꢀ)-DIP-ChlorideTM
CH2Cl2) which correspond to the (R)-isomer [22]. The GC
analysis of its TFA derivative on a Chiraldex-GTA capillary
1
column showed an ee of 82%. H NMR d (CDCl3) 3.02 (sb,
5.1.1. General procedure
1H), 4.93–4.96 (m, 1H), 5.80 (d, J = 5.4 Hz, 1H), 7.30–7.40 (m,
5H). 13C NMR d (CDCl3) 76.16, 78.55, 127.03, 128.34, 128.86,
137.25.
An oven-dried, 50-mL round-bottom flask equipped with a
side arm, magnetic bar and a connecting tube was cooled to rt
in a stream of nitrogen. (ꢀ)-DIP-Chloride (3.52 g, 11 mmol)
was transferred to the flask in a glove bag and dissolved in ee
(15 mL). The solution was cooled to ꢀ25 8C, and the ketone
(10 mmol) was added, dropwise, using a syringe. The
reaction was followed by 11B NMR spectrometry after
aliquots were methanolyzed at ꢀ25 8C at periodic intervals.
When the reaction was complete (11B: d 32 ppm), the mixture
was warmed to 0 8C and diethanolamine (2.1 mL, 22 mmol)
5.3.2. (b) From reduction with 2
A 10 mmol of 5a was treated with (R)-Alpine-Borane
(12 mmol) at rt in neat condition. The reaction was complete
in 22 days. Workup as described above provided 6a in 70%
yield. The GC analysis of the TFA derivative on a Chiraldex-
GTA capillary column showed an ee of 83% in the (R)-
isomer.

P.V. Ramachandran et al. / Journal of Fluorine Chemistry 128 (2007) 844–850
849
5.4. 2,2,2-Trichloro-1-phenylethanol, 8a
5.6.2. (b) From reduction with 2
1,1,-Dichloroacetone (10 mmol) was treated with (R)-
Alpine-Borane (12 mmol) at rt under neat condition. The
reaction was complete in 4 days. Workup as described above
5.4.1. (a) From reduction with (ꢀ)-1
2,2,2-Trichloroacetophenone (10 mmol) was added to solid
(ꢀ)-DIP-Chloride (11 mmol) at rt. The reaction was complete
in 22 days. Workup as described above and distillation using a
Kugelrohr apparatus (pot temperature 100–110 8C/0.5 mmHg)
provided 8a in 65% yield, after a recovery of 20% of 6a.
21
provided 12a in 75% yield. ½aꢁD ¼ ꢀ2:92 (ca. 3.2, CHCl3). The
GC analysis of the MTPA derivative on a SPB-5 capillary
column showed an ee of 62% in the opposite isomer as
compared to the product from (ꢀ)-DIP-ChlorideTM reduction
described above.
25
½aꢁD ¼ þ35:80 (ca. 3.0, CH2Cl2) correspond to the (S)-isomer
[22]. GC analysis of the MTPA ester on a SPB-5 capillary
1
column showed an ee of 91%. H NMR d (CDCl3) 3.31 (d,
5.7. 2,2,2-Trichloro-1-propanol, 14a
J = 4.1 Hz, 1H), 5.20 (d, 4.1 Hz, 1H), 7.35–7.40 (m, Ph, 3H),
7.58–7.62 (m, Ph, 2H). 13C NMR d (CDCl3) 84.31, 102.89,
127.72 129.11, 129.36, 134.80.
5.7.1. (a) From reduction with (ꢀ)-1
1,1,1-Trichloroacetone (10 mmol) was added to solid (ꢀ)-
DIP-Chloride (11 mmol) at rt. The reaction was complete in 3
days. Workup as described above (pot temperature 110–120 8C/
5.4.2. (b) From reduction with 2
21
2,2,2,-Trichloroacetophenone (10 mmol) was treated with
12 mmol (R)-Alpine-Borane without solvent at rt. The reaction
was complete in 22 days. Workup as described above provided
8a in 62% yield. The GC analysis of the MTPA derivative on a
SPB-5 capillary column showed an ee of 27% in the (R)-
isomer.
25 mmHg) provided 15a in 78% yield. ½aꢁD ¼ ꢀ4:68 (ca. 1.26,
CHCl3). The GC analysis of the TFA derivative on a Chiraldex-
GTA capillary column showed an ee of 91%. 1H NMR d
(CDCl3) 1.53 (d, J = 6.1 Hz, 3H), 2.90 (sb, 1H), 4.22–4.32 (m,
1H). 13C NMR d (CDCl3) 17.75, 79.18, 104.41.
5.7.2. From reduction with 2
5.5. 2-Chloro-1-propanol, 10a
1,1,1-Trichloroacetone (10 mmol) was treated with (R)-
Alpine-Borane (12 mmol) at rt under neat condition. The
reaction was complete in 22 days. Workup as above provided
14a in 70% yield. The GC analysis of the TFA derivative on a
Chiraldex-GTA capillary column showed an ee of 24% in the
same isomer as obtained from (ꢀ)-DIP-ChlorideTM reduction.
5.5.1. (a) From reduction with (ꢀ)-1
1-Chloroacetone (10 mmol) was treated with (ꢀ)-DIP-
Chloride (11 mmol) in ee at ꢀ25 8C. The reaction was
complete in 1.5 h. Workup as described above and distillation
using a Kugelrohr apparatus (pot temperature 75–85 8C/
25
25 mmHg) provided 10a in 65% yield, ½aꢁD ¼ ꢀ3:49 (ca.
5.8. (R)-1-Fluoro-2-propanol, 10b
5.6 CHCl3) which correspond to the (R)-isomer [23]. GC
analysis as the MCF derivative on a SPB-5 capillary column
showed an ee of 18%. 1H NMR d (CDCl3) 1.28 (d, J = 6.3 Hz,
3H), 2.35 (sb, 1H), 3.46 (dd, J = 11.0, 7.0 Hz, 1H), 3.66 (dd,
J = 11.0, 3.6 Hz, 1H), 3.95–4.05 (m, 1H). 13C NMR d (CDCl3)
20.06, 51.21, 67.54.
5.8.1. (a) From reduction with 1
2-Fluoroacetone (10 mmol) was treated with (ꢀ)-DIP-
Chloride (11 mmol) in ee at ꢀ25 8C. The reaction was
complete in 1 h. The usual workup as described in the general
procedure bp 98–105 8C (literature [24] bp 95–110 8C)
21
provided 10b in 80% yield, ½aꢁD ¼ ꢀ10:21 (ca. 1.9, CH3OH).
5.5.2. (b) From reduction with 2
1H NMR d (CDCl3) 1.19 (dd, J = 6.5, 1.7 Hz, 3H), 2.42 (s, 1H),
4.00–4.15 (m, 1H), 4.26 (ddd, J = 48.0, 9.2, 7.0 Hz, 1H), 4.69
(ddd, J = 46.9, 9.2, 3.1 Hz, 1H). 13C NMR d (CDCl3) 17.42 (d,
J = 7.9 Hz), 66.49 (d, J = 19.1 Hz), 87.79 (d, J = 168.5 Hz).
The above product was converted to the benzoate ester in
1-Chloroacetone (10 mmol) was treated with 12 mmol of
(R)-Alpine-Borane under neat condition at rt. The reaction was
complete in 3 days. Workup as above provided 10b in 68%
yield. GC analysis of the MCF derivative on a SPB-5 capillary
column showed an ee of 16% in the (S) isomer.
21
92% yield. ½aꢁD ¼ ꢀ4:88 (ca. 1.0, CHCl3) corresponds to the
(S) isomer [13]. The GC analysis of the benzoate ester on a
Chiꢀra1 ldex-GTA capillary column showed an ee of 61%. IR:
max
5.6. 2,2-Dichloro-1-propanol, 12a
cm
v
(neat) 1715 (C O). 1H NMR d (CDCl3) 1.41 (dd, J = 6.6,
5.6.1. (a) From reduction with (ꢀ)-1
1.4 Hz, 3H), 4.42–4.67 (m, 2H), 5.30–5.45 (m, 1H), 7.40–7.60
(m, Ph, 3H), 8.04–8.10 (m, Ph, 2H). 13C NMR d (CDCl3) 15.82
(d, J = 6.8 Hz), 70.02 (d, J = 19.8 Hz), 85.09 (d, J = 174.3 Hz),
128.84, 130.17, 130.58, 133.57, 166.42. 19F NMR d (CDCl3)
ꢀ228.81 (dt, J = 48.4, 20.9 Hz).
1,1-Dichloroacetone (10 mmol) was treated with (ꢀ)-DIP-
Chloride (11 mmol) in ee at ꢀ25 8C. The reaction was
complete in 2 h. Workup as described above using a Kugelrohr
apparatus (pot temperature 95–105 8C/25 mmHg) provided
12a in 72% yield. GC analysis as the MTPA derivative on a
1
SPB-5 capillary column revealed an ee of 15%. H NMR d
5.8.2. (b) From reduction with 2
(CDCl3) 1.39 (d, J = 6.3 Hz, 3H), 2.46 (sb, 1H), 4.02–4.11 (m,
1H), 5.67 (d, J = 4.4 Hz, 1H). 13C NMR d (CDCl3) 17.89,
72.49, 77.15.
The reduction of 9b with 2 was complete in 3 days. Workup
provided the product alcohol in 68% yield. The analysis of the
MTPA derivative on a Supelcowax capillary column showed

850
P.V. Ramachandran et al. / Journal of Fluorine Chemistry 128 (2007) 844–850
(f) P.V. Ramachandran, K.J. Padiya, V. Rauniyar, M.V.R. Reddy, H.C.
Brown, Tetrahedron Lett. 45 (2004) 1015–1017;
the product to be of 13% ee in the same isomer (S) as obtained
from the reaction with 1.
(g) P.V. Ramachandran, K.J. Padiya, M.V.R. Reddy, H.C. Brown, J.
Fluorine Chem. 125 (2004) 579–583;
5.9. (S)-1,1-Difluoro-2-propanol, 12b
(h) P.V. Ramachandran, K.J. Padiya, V. Rauniyar, M.V.R. Reddy, H.C.
Brown, J. Fluorine Chem. 125 (2004) 615–620.
[3] For a review on asymmetric aldol reactions of fluoroketones, see: V.
Soloshonok in ref. [2a], Chapter 7.
5.9.1. (a) From reduction with 1
The reduction of 11b with 1 was complete in 2 h. Workup as
ꢀ1
[4] (a) H.C. Brown, P.V. Ramachandran, in: A.F. Abdel-Magid (Ed.), ACS
Symposium Series, vol. 641, American Chemical Society, Washington,
DC, 1996, , Chapter 5;
cm
above provided 12b in 80% yield. bp 85–90 8C. IR v
neat:
max
3375 (OH); 1H NMR d (ppm) (CDCl3): 1.27 (t, J = 6.6 Hz, 3H,
CH3), 2.67 (br s, 1H, CHOH), 3.8–4.0 (m, 1H, H-C-OH), 5.59
(dt, J = 56.3, 4.1 Hz, 1H, CHF2); MS EI: m/z: 51 (M–
CH3CHOH)+ (100%); CI: m/z: 97 (MH)+ (100%). The GC
analysis of the MTPA derivative on a SPB-5 capillary column
showed the product to be of 5% ee.
(b) H.C. Brown, J. Chandrasekharan, P.V. Ramachandran, J. Am. Chem.
Soc. 110 (1988) 1539.
[5] (a) M.M. Midland, S.A. Zderic, J. Am. Chem. Soc. 104 (1982) 525;
(b) M.M. Midland, Chem. Rev. 89 (1989) 1553–1561;
(c) H.C. Brown, G.G. Pai, J. Org. Chem. 50 (1985) 1384.
´
[6] (a) P.V. Ramachandran, B. Gong, A.V. Teodorovie, H.C. Brown, Tetra-
hedron: Asymmetry 5 (1994) 1075;
(b) For a review, see: P.V. Ramachandran, H.C. Brown in Ref. [2], Chapter
6, 2006, pp. 179–228.
5.9.2. (b) From reduction with 2
The reduction of 11b with 2 was complete in 3 days. Workup
provided the product alcohol in 54% yield. The GC analysis of
the MTPA derivative on a Supelcowax capillary column
showed the product to be of 16% ee in the opposite isomer as
compared to the product from reduction with 1. We believe that
we obtain the (R)-isomer based on analogy with the reduction
of 7c with 2.
[7] Mosher has discussed the electronic and steric influences in the reduction
of trifluoromethyl ketones using asymmetric Grignard reagents almost 40
years ago, 2007.
(a) C. Aaron, D. Dull, J.L. Schmiegel, D. Jaeger, Y.O. Ohashi, H.S.
Mosher, J. Org. Chem. 32 (1967) 2797;
(b) J.D. Morrison, H.S. Mosher, Asymmetric Organic Reactions, Amer-
ican Chemical Society, Washington, DC, 1976, Chapter 5.
[8] (a) T. Hanamoto, T. Fuchikami, J. Org. Chem. 55 (1990) 4969–4971;
(b) Y. Itoh, M. Yamanaka, K. Mikami, J. Am. Chem. Soc. 126 (2004)
13174–13175;
5.10. (S)-(ꢀ)-1,1,1-trifluoro-2-propanol, 14b
(c) P.K. Mohanta, T.A. Davis, J.R. Gooch, R.A. Flowers II, J. Am. Chem.
Soc. 127 (2005) 11896–11897.
5.10.1. From reduction with 2
The reduction of 13b with 2 was complete in 4 days. Workup
provided 6f in 60% yield. bp 76–77 8C (literature [25] bp 77.8–
[9] R.S. Cahn, C. Ingold, V. Prelog, Angew. Chem. Int. Ed. 5 (1966) 385, The
R configuration is an artifact of Cahn-Ingold-Prelog rules.
[10] M. Srebnik, P.V. Ramachandran, H.C. Brown, J. Org. Chem. 53 (1988)
2916–2920.
22
77.9 8C). ½aꢁD ¼ ꢀ5:23ꢂ (neat) which corresponds to an optical
25
´
[11] P.V. Ramachandran, A.V. Teodorovic, H.C. Brown, Tetrahedron 49 (1993)
purity of 80.4% in the (S) isomer based on the reported
maximum rotation [11] of ½aꢁD ¼ ꢀ6:5ꢂ (neat). The GC
1725.
[12] (a) E.J. Corey, J.O. Link, S. Sarshar, Y Shao, Tetrahedron Lett. 33 (1992)
7103;
analysis of the MCF derivative on a SPB-5 capillary column
showed the product to be of 82% ee in the (S)-isomer.
(b) E.J. Corey, J.O. Link, R.K. Bakshi, Tetrahedron Lett. 33 (1992) 7107;
(c) E.J. Corey, J.O. Link, J. Am. Chem. Soc. 114 (1992) 1906;
(d) E.J. Corey, J.O. Link, Tetrahedron Lett. 33 (1992) 3431.
[13] P. Bravo, G. Resnati, Tetrahedron Lett. 28 (1987) 4865.
Acknowledgement
´
[14] P.V. Ramachandran, B. Gong, A.V. Teodorovic, H.C. Brown, Tetrahedron:
We gratefully acknowledge Herbert C. Brown Center for
Borane Research for support of this research.
Asym. 5 (1994) 1061.
[15] H.C. Brown, G.W. Kramer, A.B. Levy, M.M. Midland, Organic Synthesis
via Boranes, Wiley-Interscience, New York, 1975, Chapter 9.
[16] J.C. Nobrega, S.M.C. Goncalves, C. Peppe, Synth. Commun. 32 (2002)
3711.
References
[17] E.J. Corey, J.O. Link, Y. Shao, Tetrahedron Lett. 33 (1992) 3435.
[18] G.C. Barrett, D.M. Hall, M.K. Hargreaves, B. Moderai, J. Chem. Soc. (C)
(1971) 279.
[1] X.L. Qiu, W.D. Meng, F.L. Qing, Tetrahedron 60 (2004) 6711–6745;
(b) F.M.D. Ismail, J. Fluorine Chem. 118 (2002) 27;
(c) Y. Kumar, S. Muzammil, S. Tayyab, J. Biochem. 138 (2005) 335–341;
(d) A.G. Myers, J.K. Barbay, B. Zhong, J. Am. Chem. Soc. 123 (2001)
7207–7219.
[19] J.A. Dale, D.L. Dull, H.S. Mosher, J. Org. Chem. 34 (1969) 2543.
[20] J.W. Westley, B. Halpern, J. Org. Chem. 33 (1968) 3978.
[21] Chiraldex-GTA is the trademark of Alltech Associates Inc.
[22] R. Weidmann, A.R. Schoofs, A.C.R. Horeau, Seances Acad. Sci. Ser. 2
294 (1982) 319.
[2] (a) V.A. Soloshonok (Ed.), Enantiocontrolled Synthesis of Fluoro-organic
Compounds, Wiley, 1999;
(b) P.V. Ramachandran (Ed.), Asymmetric Fluoroorganic Chemistry,
American Chemical Society, Washington, DC, 2000;
(c) S. Sun, A. Adejare, Curr. Top. Med. Chem. 6 (2006) 1457–1464;
(d) C. Presenti, F. Viani, ChemBioChem (2004) 590;
(e) J.A. Sikorski, J. Med. Chem. 49 (2006) 1–22;
[23] T. Sakai, K. Wada, T. Murakami, K. Kohra, N. Imajo, Y. Ooga, S. Tsuboi,
A. Takeda, M. Utaka, Bull. Chem. Soc. Jpn. 65 (1992) 631.
[24] E.D. Bergmann, S. Cohen, J. Chem. Soc. (1958) 2259.
[25] J.W.C. Crawford, J. Chem. Soc. C. (1967) 2332.