Enantiomeric resolution of fluorous mixture by chiral CD columns: Asymmetric reduction of a mixture of fluorous ketones

Article abstract of DOI:10.1016/S0040-4039(03)01539-9

Asymmetric reduction of a mixture of four fluorous ketone analogues was carried out with (R)-oxazaborolidine as a catalyst. The fluorous mixture products were resolved by a reverse phase HPLC with chiral β-cyclodextrin column to result in good separation of the enantiomers.

Full text of DOI:10.1016/S0040-4039(03)01539-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 6221–6225
Enantiomeric resolution of fluorous mixture by chiral CD
columns: asymmetric reduction of a mixture of fluorous ketones
Yutaka Nakamura,a,* Seiji Takeuchi,a,* Kazuo Okumura,a Yoshiaki Ohgo,a Hiroshi Matsuzawab and
Koichi Mikamib,*
aNiigata University of Pharmacy and Applied Life Sciences, 5-13-2 Kamishin’ei-cho, Niigata 950-2081, Japan
bDepartment of Applied Chemistry, Tokyo Institute of Technology, 2-12-1 Ookayama, Tokyo 152-8552, Japan
Received 16 May 2003; revised 20 June 2003; accepted 20 June 2003
Abstract—Asymmetric reduction of a mixture of four fluorous ketone analogues was carried out with (R)-oxazaborolidine as a
catalyst. The fluorous mixture products were resolved by a reverse phase HPLC with chiral b-cyclodextrin column to result in
good separation of the enantiomers.
© 2003 Elsevier Ltd. All rights reserved.
Curran and co-workers made an epoch-making discov-
ery in 1999 that a fluorocarbon bonded silica gel HPLC
a synthesis of a library including 560 mappicine ana-
logues in miligram order of isolated samples via several
steps of mixture (internal parallel) and split-parallel
(double parallel) reactions, ‘demixing’ the products and
finally removing the fluorous tags. On the other hand,
Mikami and co-workers have recently discovered a high
separating ability of b-cyclodextrin (b-CD) HPLC
columns (SUMICHIRAL® OA-7000 series) for
fluorous compounds such as benzoic acid and benzene-
sulfonic acid fluoroalkyl esters.3 The chiral CD columns
demonstrated more efficient separation of the com-
column
(silica-OSi(CH3)2(CH2)2CnF2n+1),
Fluofix®
120E, was able to separate molecules that had fluorous
tags of CH2CH2CnF2n+1 (n=3–10) in a very regular way
depending upon their fluorine contents; the more fluori-
nes a molecule possesses, the slower it moves in a
‘reverse-phase’ mode.1 Based upon this discovery they
created a new strategic technique of combinatorial
chemistry, ‘Fluorous Mixture Synthesis’.2 Seven
fluorous tags, CH2CH2CnF2n+1 (n=3, 4, 6–10), enabled
Scheme 1. Enantiomeric resolution of fluorous mixture by chiral b-CD columns.
* Corresponding authors. Tel.: +81-25-268-1247; fax: +81-25-268-1230 (S.T.); Tel.: +81-3-5734-2142; fax: +81-3-5734-2776 (K.M.); e-mail: takeuchi@
niigata-pharm.ac.jp; kmikami@o.cc.titech.ac.jp
0040-4039/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)01539-9

6222
Y. Nakamura et al. / Tetrahedron Letters 44 (2003) 6221–6225
pounds bearing short (CF3) to long (C9F19) tags as
compared to Fluofix 120E.
fluorous tags were synthesized individually, and these
esters were mixed and then resolved. It was found that
the mixture of racemic O-benzoylmandelate derivatives
was resolved using OA-7500 as a b-CD column. In this
case, esters were eluted first from O-(p-fluoroben-
zoyl)mandelate 1a bearing C1 tag. O-Benzoylmandelate
1c bearing C7 tag showed the best resolution (Rs)
amongst all the esters.
Takeuchi and co-workers have reported a different kind
of mixture reaction, an asymmetric addition of ZnEt2
to aromatic aldehydes using a fluorous chiral BINOL-
Ti complexes as catalysts.4 Five kinds of aromatic
aldehydes were reacted with ZnEt2 in one-pot and the
product mixture was separated from the catalyst by a
fluorous reverse phase silica gel and then analyzed by
gas chromatography with chiral column. The products
and their enantiomers were well separated to lead to a
simultaneous evaluation of yields and % ee for each
product. In contrast to the results, however, attempts to
separate the products and their enantiomers simulta-
neously by HPLC with chiral columns were unsuccess-
ful. Therefore, this approach is useful for analysis of a
mixture of products but not suitable for resolution
thereof. Curran group’s new strategy of ‘Fluorous Mix-
ture Synthesis’ in combination with Mikami’s strategy
of discrimination of fluorous chains by chiral CD
column prompted us to undertake an asymmetric
fluorous mixture resolution because it will provide us a
useful method not only for analysis of enantioselectivi-
ties of the products but also for isolation of the pure
enantiomers (Scheme 1).
Keeping it in our mind, we next examined the Corey–
Itsuno asymmetric borane reduction for fluorous
ketonic substrates using (R)-oxazaborolidine as a cata-
lyst (Scheme 3). The fluorous groups are skeletal parts
of the molecules and are unable to be removed after the
reaction. Thus the substrates just seem to have a role to
test our idea for the time being but these kinds of chiral
compounds will become useful for fluorous chemistry in
the long run.
The fluorous ketones 3a–d were prepared in good yields
via two steps. At the outset, each substrate underwent
independently the asymmetric reduction at 25°C in
THF for 15 min under argon.5 The products 2a–d were
analyzed by HPLC with DAICEL CHIRALCEL OD-
H under normal phase separation conditions to get
enantioselectivity of each product and then analyzed by
HPLC with Fluofix 120E under the reverse phase sepa-
ration conditions by both CD and UV detectors to
obtain g value (the ratio of absorption coefficient of
CD to that of UV) of each product for super high-
throughput screening (SHTS).6 From the enantioselec-
First, we checked efficiency in resolution of a racemic
mixture of substrates with different length of fluorous
tags with b-CD columns (Scheme 2). Three O-benzoyl-
mandelate derivatives bearing different length of
Scheme 2. Resolution of O-benzoylmandelate derivatives 1a–c with b-CD column. Column: OA-7500: mobile phase, methanol–
water (0 to 60 min, 75/25 up to 85/15, v/v); flow rate, 0.5 mL/min; detector, absorption at 254 nm; column temperature, 20°C.

Y. Nakamura et al. / Tetrahedron Letters 44 (2003) 6221–6225
6223
Scheme 3.
tivity and the g value of each product, g values for
enantiomerically pure products (entry 1 in Table 1)
were calculated. The four products were clearly sepa-
rated by HPLC with Fluofix 120E under the conditions
shown in Figure 1. The measurement of CD and UV
absorption coefficient for each product mentioned
above was carried out under the same separating condi-
tions as those in Figure 1. Next we examined the
reaction of a mixture of the four substrates 3a–d and
the product mixture was analyzed by HPLC with
Fluofix 120E to get g value for each product (Fig. 1).
The enantiomeric excesses were evaluated by the g
values obtained above for enantiomerically pure prod-
ucts. The enantioselectivities were almost the same as
those obtained by the independent reactions as shown
in Table 1 (entry 2).
will be able to assess the enantioselectivity of each
product in the mixture product and to get enantiomeri-
cally enriched sample of each product as long as each
product can be separated clearly by HPLC with an
effective column such as Fluofix 120E. This technique
must be very useful for finding the most efficient reac-
tion and optimal conditions to provide isolated prod-
ucts of the highest enantioselectivities.
Finally we tried to separate the products and their
enantiomers simultaneously by using HPLC with CD
chiral columns. Each enantiomer pair of the four
racemic fluorous alcohols was separated by SUMICHI-
RAL OA-7500 column although the separation was not
perfect, and was not separable at all by OA-7000 and
OA-7100 columns (Fig. 2). When the product mixture
was analyzed by the chiral column, somewhat lower
values of enantioselectivity were obtained as compared
to those evaluated by g values, probably because of
incomplete separation of the enantiomer pairs (entry 3
in Table 1).
As seen above, when we take the procedure, that is,
determination of the products’ enantiomeric excesses by
HPLC or GC analysis independently and then correlate
the % ees to the g values of the products in advance, we
Figure 1. HPLC chromatograms of alcohols 2a–d. Column: Fluofix 120E; mobile phase: CH3OH–H2O (0 to 30 min, 80/20
up to 100/0 v/v); flow rate: 1.0 mL/min; detector: 268 nm CD and UV.

6224
Y. Nakamura et al. / Tetrahedron Letters 44 (2003) 6221–6225
Table 1. Enantioselective borane reduction of perfluoroalkyl ketones 3a–d catalyzed by chiral oxazaborolidine
Entry
Yield (%)
% ee
2c
2a
2b
2c
2d
2a
2b
2d
1a
2b
3c
98
99
N.D.
97
99
N.D.
97
100
N.D.
95
99
N.D.
88
85
82
85
85
75
87
83
77
84
87
71
a The reactions were carried out independently. The yields were isolated yields. The ees were determined by HPLC analysis using DAICEL
CHIRALCEL OD-H (hexane/2-propanol=99/5, 1.0 mL/min, 254 nm UV).
b The yields and ees were determined by HPLC analysis with Fluofix 120E by using BTF as an internal standard (80% MeOH/H2O to 100%
MeOH for 30 min, 1.0 mL/min, 268 nm UV and CD). Ee was calculated based on the dissymmetric factor (g).
c Determined by HPLC analysis using SUMICHIRAL OA-7500.
Figure 2. HPLC chromatograms of recemic alcohols 2a–d. Top: CH3CN/H2O=55/45, 0.5 mL/min, 254 nm UV. Middle and
bottom: 0 to 60 min, CH3CN/H2O=60/40 up to CH3CN/H2O=80/20; 60 to 80 min. CH3CN/H2O=80/20, 0.5 mL/min, 254 nm
UV.

Y. Nakamura et al. / Tetrahedron Letters 44 (2003) 6221–6225
6225
As seen from the HPLC chart, it is expected that
enantiomer pairs of another series of the fluorous alco-
hol 2e–h (n=3, 5, 7, 9, respectively) will be separable by
the chiral column and their peaks will appear before
and between the peaks of 2a–d without any serious
overlap. We are now preparing many kinds of such
fluorous substrates and testing separation of the prod-
ucts and their enantiomers by using reverse phase
HPLC chiral columns including CD ones in order to
find more efficient chiral columns and/or analytical
conditions for complete separation.
4. Nakamura, Y.; Takeuchi, S.; Okumura, K.; Ohgo, Y.;
Curran, D. P. Tetrahedron 2002, 58, 3963–3969.
5. A typical procedure for entry 1 in Table 1: To a solution
of ketone 3a (176 mg, 0.5 mmol) and (R)-2-methyl-CBS-
oxazaborolidine (1 M in toluene, 0.05 mL, 0.05 mmol) in
THF (2 mL) was added borane–THF (1.0 M in THF, 0.5
mL, 0.5 mmol) at 25°C under argon. After stirring at that
temperature for 15 min, the reaction mixture was
quenched with saturated aqueous NH4Cl solution (5 mL)
and then extracted with ether (3×10 mL). The combined
organic layer was washed with brine (10 mL), dried over
anhydrous MgSO4, and then concentrated in vacuo. The
residue was purified on a silica gel column with hexane/
ether (10/1) to afford alcohol (−)-2a (173 mg, 98% yield) in
88% ee as a colorless solid.
6. (a) Ding, K.; Ishii, A.; Mikami, K.; Angew. Chem., Int. Ed.
Engl. 1999, 38, 497–501; (b) Mikami, K.; Ding, K.; Ishii,
A.; Tanaka, A.; Sawada, N.; Kudo, A. Chromatography
1999, 20, 65–69; (c) Reetz, M. T.; Kuhling, K. M.; Hin-
richs, H.; Deege, A. Chirality 2000, 12, 479–482; (d) Ange-
laud, R.; Matsumoto, Y.; Korenaga, T.; Kudo, K.; Senda,
M.; Mikami, K. Chirality 2000, 12, 544–547; (e) Mikami,
K.; Angelaud, R.; Ding, K.; Ishii, A.; Tanaka, A.; Sawada,
N.; Kudo, K.; Senda, M. Chem. Eur. J. 2001, 7, 730–737.
Also see: (f) Drake, A.; Gould, J. M.; Mason, S. F. J.
Chromatogr. 1980, 202, 239–245; (g) Betrucci, C.; Sal-
vadori, P.; Guimaraes, L. J. Chromatogr. A 1994, 666,
535–539. For a review, see: (h) Reetz, M. T. Angew.
Chem., Int. Ed. Engl. 2001, 40, 284. Also see: (i)
Mannschreck, A. Chirality 1992, 4, 163.
Acknowledgements
The authors would like to thank Professor Dennis P.
Curran, University of Pittsburgh for his helpful sugges-
tions. They also thank Noguchi Research Institute for
their financial help as Noguchi Fluorous Project’s
Fund. This work was partially supported by Grant-in-
Aid for Scientific Research from the Ministry of Educa-
tion, Science, Sports and Culture.
References
1. Curran, D. P.; Luo, Z. J. Am. Chem. Soc. 1999, 121,
9069–9072.
2. Zhang, W.; Luo, Z.; Chen, C. H.-T.; Curran, D. P. J. Am.
Chem. Soc. 2002, 124, 10443–10450.
3. Matsuzawa, H.; Mikami, K. Synlett 2002, 1607.