Kinetic resolution of secondary alcohols. Enantioselective acylation mediated by a chiral (dimethylamino)pyridine derivative

Full text of DOI:10.1021/ja953631f

J. Am. Chem. Soc. 1996, 118, 1809-1810
1809
adduct. Metalation at C2 with lithium tetramethylpiperidide
(LiTMP) followed by reaction with pivaloyl chloride then gave
the ketone 1 (61%) together with recovered DMAP (10%) and
the 2,6-dipivaloyl derivative (15%). Reduction of 1 using (-)-
Kinetic Resolution of Secondary Alcohols.
Enantioselective Acylation Mediated by a Chiral
(Dimethylamino)pyridine Derivative
Edwin Vedejs* and Xinhai Chen
Chemistry Department, UniVersity of Wisconsin
Madison, Wisconsin 53706
ReceiVed October 30, 1995
Kinetic resolution of certain chiral alcohols or their ester
derivatives can be carried out using the lipase/esterase family
of acyl transfer catalysts.1 If the enantiomers differ in relative
reactivity by a factor (s) of 100 or more, >40% recovery of
each enantiomer is theoretically possible with >90% ee.1a
However, lower enantioselectivities s in the range of 10-30
are often encountered, and purified yields can be considerably
lower. In such cases the less reactive enantiomer can still be
obtained with high ee by forcing the conversion well past the
theoretical 50% optimum because eventual destruction of the
more reactive enantiomer compensates for “errors” in enanti-
oselective recognition. However, higher ee is achieved at the
cost of decreased material recovery.1a
Some progress has been made with enantioselective non-
enzymatic acylating agents.2-4 The best selectivities to date
were reported by Evans et al. using a 10-fold excess of racemic
alkoxides ArCH(CH3)OMgBr and a chiral N-acyl imide as the
stoichiometric acyl donor.3 At ca. 10% conversion of the
alkoxide, this system gave up to 90% enantiomer excess in the
product esters, corresponding to s (calculated ratio of rate
constants for the more reactive vs the less reactive enantiomer)5
in the range of 20-30. The best s values reported using a chiral
nonenzymatic catalyst and an achiral acyl donor as the sto-
ichiometric reagent are considerably lower.2a,b,e,h,4 Under condi-
tions where catalyst turnover is demonstrated, only one example
is known where s is greater than 10.4
B-chlorodiisopinocampheylborane (ipc2BCl)8 produced 2 (71%;
96% ee), and recrystallization afforded material with >99% ee
according to HPLC assay. Methylation (CH3I/KH/18-crown-
6) then produced the key reagent 3 (absolute configuration
assigned by analogy to the pyridine reduction precedent of Bolm
et al.).8
Treatment of 3 with the commercially available chloroformate
4 generated the corresponding pyridinium salt 5 as evidenced
by characteristic 1H NMR downfield shifts for the ring protons
(δ 8.06, 6.56, 6.49 ppm for 3; 8.29, 6.69, 6.67 ppm for 5 in
CD3CN). The solution containing 5 did not acylate representa-
tive secondary alcohols at room temperature. However, the
addition of a tertiary amine together with a Lewis acid
(anhydrous ZnCl2 or MgBr2) initiated a slow acyl transfer
reaction (15-40 h for consumption of 5 using 2 equiv of the
racemic secondary alcohol 6, ca. 20 °C), resulting in the
formation of the mixed carbonate 7. Product assay was
1
performed by H NMR on the mixture of 6 and 7 to measure
percent conversion, and the ee of esters 7 was established by
HPLC assay on a chiral support after purification of 7 and
saponification to the original alcohol 6 (see supporting informa-
tion). In all entries, the material balance (6 + 7) was at least
90%, calculated from the isolated yield of purified 7 and from
the ratio of 6:7 determined by NMR. According to this evidence
(Table 1), several of the mixed carbonate esters 7 were formed
with >90% enantiomeric purity at conversions in the 20-42%
range.9 Several entries in Table 1 report s values well above
30 (calculated from product ee and percent conversion),5 the
best results observed to date with any nonenzymatic acylating
agent.
As expected, the ee of unreacted 6a was improved by
doubling the amount of reagents to increase the percent
conversion (96% ee at 71% conversion; 60% ee at 42%
conversion). However, the improvement was less than calcu-
lated (>99% ee) based on s ) 42.5 The reason for the
discrepancy was not identified, but possible explanations include
minor racemization of 6a, integral error in the percent conver-
sion, or interference by the product 7a in the enantioselective
acylation.10 The increased conversion experiments were not
pursued further because they require a larger proportion of the
valuable reagent 5. Thus, it is more practical to control
conversion so that product (not unreacted substrate 6) ee and
We now describe chiral acyl transfer agents based on the
p-(dimethylamino)pyridine (DMAP) nucleus.6 The new re-
agents must be employed in stoichiometric amounts, but the
chiral DMAP derivatives are recovered unchanged at the end
of the reaction and can be reused. By analogy to Kessar’s results
with pyridine,7 DMAP was activated by conversion into the BF3
(1) Reviews: (a) Sih, C. J.; Wu, S.-H. Top. Stereochem. 1989, 19, 63.
Chen, C.-S.; Sih, C. J. Angew. Chem., Int. Ed. Engl. 1989, 28, 695. (b)
Klibanov, A. M. Acc. Chem. Res. 1990, 23, 114. (c) Ward, S. C. Chem.
ReV. 1990, 90, 1. (d) Drueckhammer, D. G.; Hennen, W. J.; Pederson, R.
L.; Barbas, C. F., III; Gautheron, C. M.; Krach, T.; Wong, C.-H. Synthesis
1991, 499. (e) Haraldsson, G. G. The Application of Lipases in Organic
Synthesis. In The Chemistry of Functional Groups, Suppl. B, The Chemistry
of Acid DeriVatiVes; Patai, S., Ed.; John Wiley & Sons: Chichester, 1992;
p 1395. (f) Roberts, S. M. Chimia 1993, 47, 85.
(2) (a) Wegler, R. Justus Liebigs Ann. Chem. 1932, 498, 62. Wegler, R.
Justus Liebigs Ann. Chem. 1933, 506, 77. Wegler, R. Justus Liebigs Ann.
Chem. 1934, 510, 72. Wegler, R.; Ru¨ber, A. Chem. Ber. 1935, 68, 1055.
(b) Bird, C. W. Tetrahedron 1962, 18, 1. (c) Weidmann, R.; Horeau, A.
Bull. Soc. Chim. Fr. 1967, 117. (d) Mukaiyama, T.; Tomioka, I.; Shimizu,
M. Chem. Lett. 1984, 49. (e) Stegman, W.; Uebelhart, P.; Heimgartner, H.;
Schmid, H. Tetrahedron Lett. 1978, 3091. (f) Duhamel, L.; Herman, T.
Tetrahedron Lett. 1985, 26, 3099. (g) Ichikawa, J.; Asami, M.; Mukaiyama,
T. Chem. Lett. 1984, 949. (h) Potapov, V. M.; Dem’yanovich, V. M.;
Klebnikov, V. A.; Korovina, T. G. Zh. Org. Khim. 1986, 22, 1218 (1095
English translation); see ref 4, footnote 4f for similar results; the same system
was studied with different results by Weidert et al.: Weidert, P. J.; Geyer,
E.; Horner, L. Liebigs Ann. Chem. 1989, 533. (i) Ishihara, K.; Kubota, M.;
Yamamoto, H. Synlett 1994, 611.
(8) Bolm, C.; Ewald, M.; Felder, M.; Schlingloff, G. Chem. Ber. 1992,
125, 1169.
(3) Evans, D. A.; Anderson, J. C.; Taylor, M. K. Tetrahedron Lett. 1993,
34, 5563.
(9) Procedure for kinetic resolution of 1-arylethanol derivatives: to the
solution of 3 (0.15 mmol) in CH2Cl2 (1 mL) at 0 °C was added a CH2Cl2
solution of 4 (0.14 mmol, Aldrich). The mixture was warmed to room
temperature and stirred for 2 h. A solution of anhydrous (fused) ZnCl2 (0.3
mmol, 0.5 M) in diethyl ether was then added. After 10 min, the racemic
6 (0.3 mmol) and triethylamine or PMP (0.45 mmol) were added
sequentially. The solution was stirred at room temperature under N2 for
the specified time (Table 1), and the neutral products were separated by
flash chromatography prior to assay.
(4) Vedejs, E.; Daugulis, O.; Diver, S. T. J. Org. Chem., in press.
(5) Kagan, H. B.; Fiaud, J. C. Top. Stereochem. 1988, 18, 249.
(6) Steglich, W.; Ho¨fle, G. Angew. Chem., Int. Ed. Engl. 1969, 8, 981.
Ho¨fle, G.; Steglich, W.; Vorbru¨ggen, H. Angew. Chem., Int. Ed. Engl. 1978,
17, 569.
(7) Kessar, S. V.; Singh, P.; Singh, K. N.; Dutt, M. J. Chem. Soc., Chem.
Commun. 1991, 570.
0002-7863/96/1518-1809$12.00/0 © 1996 American Chemical Society

1810 J. Am. Chem. Soc., Vol. 118, No. 7, 1996
Communications to the Editor
Table 1. Enantioselective Acylation of Secondary Alcohols with 5
hindered amine PMP (1,2,2,6,6-pentamethylpiperidine) can still
function as a basic catalyst without binding the Lewis acid as
strongly as does Et3N. A smaller accelerating effect was
observed for PMP with MgBr2 as the Lewis acid. Since the
MgBr2 experiments were inherently faster than with ZnCl2 as
the Lewis acid, the PMP/MgBr2 conditions were not explored
in detail. All of the reactions were sensitive to moisture and to
the purity of reactants, and failure to control these variables
resulted in decreased enantioselectivity.
Isobutyl chloroformate gave minimal product ee with the
usual combination of 3/ZnCl2/Et3N and 6a as the substrate, and
an attempt to use tert-butyl chloroformate in a similar experi-
ment failed because of decomposition of the chloroformate.
Simple acyl chlorides (pivaloyl; benzoyl) were ineffective in
place of 4 (nearly racemic ester products). Zinc triflate was
comparable to ZnCl2 or MgBr2, but the other Lewis acids tested
gave slow acylation (CeCl3; SnCl4; Yb(OTf)3) or caused the
formation of side products (TiCl4; ZrCl4; AlCl3). Boron
trifluoride etherate did promote acyl transfer from 5, but
selectivity was low (s < 2), and the sense of enantioselection
with 6a was opposite to that seen with ZnCl2 or MgBr2 under
the usual conditions. These results leave open the possibility
that the activated form of 5 is a zinc or magnesium complex
with the metal ion chelated by the carbonyl and ether oxygens,
but much remains to be learned about this complicated system,
and we will not speculate further regarding the nature of the
acyl transfer step or other mechanistic issues.
Because 3 is the starting material for the above process and
because it is recovered unchanged after aqueous workup (no
change in ee after five repeated acylation cycles), the chiral
DMAP derivative satisfies one of the definitions of a catalyst.
Of course, 3 is used in stoichiometric quantity and does not
turn over under the acylation conditions. Preliminary attempts
to achieve turnover by altering stoichiometry and the order of
mixing the reagents were not successful, but many options
remain to be explored.
In summary, we have described a nonenzymatic system for
kinetic resolution of secondary benzylic alcohols by enantio-
selective acylation that approaches some of the lipase methods
in enantioselectivity. Enantiomerically enriched products can
be obtained in several cases in 20-44% yield (based on the
racemic starting material; 50% theoretical yield limit) with ca.
90% ee. These results are not competitive with the best lipase
experiments, but they suggest that nonenzymatic acyl transfer
alternatives may be capable of acceptable levels of selectivity.
Experiments intended to improve practicality and to achieve
catalyst turnover are in progress.
a HPLC assay. b Enantioselectivity calculated from percent conver-
sion and product ee;5 error in s: see footnote 10. c Configuration
established by comparison with commercial material. d TEA ) triethyl-
f
amine; PMP ) 1,2,6-pentamethylpiperidine. e Reference 11. Reference
12.
recovery are maximized, in contrast to typical lipase
experiments.
Neither the tertiary amine nor the Lewis acid were sufficient
to induce substantial conversion of 5 to the mixed carbonate 7
at room temperature unless both were present. Addition of the
Lewis acid prior to the addition of 4 resulted in nearly racemic
product, perhaps due to catalysis of the reaction between 4 and
6 by the tertiary amine. Relatively fast acylations were observed
using 5 activated by the ZnCl2/PMP combination compared to
the ZnCl2/Et3N combination. This is probably because the
(10) Discrepancies were also encountered with some of the entries in
Table 1 when s values calculated from the ee of recovered 6 were compared
with s based on ee of the product (7).5 Due to the logarithmic nature of the
calculation, as little as 5% undetected impurity, side reaction, racemization,
or integral error (used to measure percent conversion) could account for
the discrepancy in s values. At most, traces (<1%) of racemization were
detected by exposing 6e to zinc chloride + (PMP), or with added 3, or
with PMP‚HCl added (HPLC assay). Racemization cannot be ruled out,
but error in percent conversion is the more likely source of the discrepancy.
However, racemization of product ester or the recovered alcohol may explain
why marginal results were obtained with the sensitive 1-(2-methoxyphenyl)-
ethanol (maximum 15-30% ee for the ester as well as the recovered
alcohol).
(11) Ohkuma, T.; Ooka, H.; Hashiguchi, S.; Ikariya, T.; Noyori, R. J.
Am. Chem. Soc. 1995, 117, 2675.
(12) The absolute configuration was assigned by chemical shift com-
parisons for the Mosher ester with data reported by Fuganti et al.: Fuganti,
C.; Grasselli, P.; Spreafico, F.; Zirotti, C.; Casati, P. J. Chem. Res., Synop.
1985, 22.
Acknowledgment. This work was supported by the National
Science Foundation.
Supporting Information Available: Experimental details for
preparation of 1-3, characterization data for 1-3 and 7, detailed
procedures for acylation; and HPLC assay methods for enantiomers of
6 (8 pages). This material is contained in many libraries on microfiche,
immediately follows this article in the microfilm version of the journal,
can be ordered from the ACS, and can be downloaded from the Internet;
see any current masthead page for ordering information and Internet
access instructions.
JA953631F