Lipase-catalyzed resolution of esters of 4-chloro-3-hydroxybutanoic acid: Effects of the alkoxy group and solvent on the enantiomeric ratio

Article abstract of DOI:10.1016/S0957-4166(99)00126-3

Various lipases have been investigated for their potential use as catalysts for the resolution of esters of 4-chloro-3-hydroxybutanoic acid via transesterification in organic solvents. Rhizomucor miehei lipase was found to be the most efficient lipase, with the enantiomeric ratio (E) being dependent upon of the nature of the alkoxy group of the ester and the resolution medium. Higher E-values were obtained when transesterification was performed in benzene or carbon tetrachloride than was the case in hexane. In mixtures of benzene and hexane the trend in E-values followed a linear relationship.

Full text of DOI:10.1016/S0957-4166(99)00126-3

Pergamon
Tetrahedron: Asymmetry 10 (1999) 1401–1412
Lipase-catalyzed resolution of esters of 4-chloro-3-hydroxy-
butanoic acid: effects of the alkoxy group and solvent on the
enantiomeric ratio
Bård Helge Hoff and Thorleif Anthonsen ∗
Department of Chemistry, Norwegian University of Science and Technology, N-7491 Trondheim, Norway
Received 11 March 1999; accepted 1 April 1999
Abstract
Various lipases have been investigated for their potential use as catalysts for the resolution of esters of 4-chloro-
3-hydroxybutanoic acid via transesterification in organic solvents. Rhizomucor miehei lipase was found to be the
most efficient lipase, with the enantiomeric ratio (E) being dependent upon of the nature of the alkoxy group
of the ester and the resolution medium. Higher E-values were obtained when transesterification was performed
in benzene or carbon tetrachloride than was the case in hexane. In mixtures of benzene and hexane the trend in
E-values followed a linear relationship. © 1999 Elsevier Science Ltd. All rights reserved.
1. Introduction
The (R)-enantiomers of alkyl 4-chloro-3-hydroxybutanoates are valuable synthons for the produc-
tion of pharmaceuticals, such as L-carnitine1,2 and (R)-4-amino-3-hydroxybutanoic acid (GABOB).
Methods for producing enantiomerically enriched alkyl 4-chloro-3-hydroxybutanoates include reduction
of the corresponding 3-oxobutanoates using either microorganisms3–7 or chiral diphosphine ruthenium
catalysts.8–10 However, lipase-catalyzed resolutions have also been reported.7,11–13 In contrast to asym-
metric synthesis, resolution may give access to both enantiomers in the same process, and may give a
high enantiomeric excess (ee) of the two products even if the E-value is not excellent. The drawback,
however, is that the maximum yield is only 50%.
One of the benefits of using low-water reaction systems is the shift of equilibrium, such that hydrolysis
is avoided. Since the substrates in the present work are bifunctional, it was essential to exclude hydrolysis
and/or alcoholysis of the starting ester and direct acylation of the C-3 hydroxy group. For lipase-catalyzed
kinetic resolutions in organic media, the choice of solvent often affects the enantiomeric ratio.14,15 In this
∗
Corresponding author. E-mail: thorleif.anthonsen@chembio.ntnu.no
0957-4166/99/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(99)00126-3

1402
B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
study, the effect of the solvent on the E-value is investigated using Rhizomucor miehei lipase (RML),
one of the more popular biocatalysts for performing kinetic resolutions.16 RML has been shown to be
active at water activities as low as 0.000117 and it is relatively stable toward acetaldehyde.18 The solvent
effect on E-values in the resolution of 1-(2-furyl) ethanol19 and 1-phenylethanol,16 using vinyl acetate as
an acyl donor, has been studied previously. Dichloromethane was found to be the best solvent for ring
opening of 4-substituted 2-phenyloxazolin-5-one with 1-butanol as the acyl acceptor.20
2. Results and discussion
Racemic substrates 1a, 1b, 1c and 1d were synthesized by various routes (Scheme 1). Esterification of
3-butenoic acid (2, R1=H) with benzyl alcohol or cyclohexanol gave 2b and 2c, respectively. Subsequent
epoxidation gave 3b and 3c which, on further hydrochlorination, gave 1b and 1c. Since ethyl 4-chloro-
3-oxobutanoate 4a is commercially available, 1a was easily obtained by sodium borohydride reduction.
The tert-butyl ester 1d was not obtainable by these methods, however, treatment of diketene 5 with
chlorine gas in CCl4 followed by addition of tert-butyl alcohol/pyridine gave 4d.21 The reaction gave
several side-products, the most abundant being tert-butyl 2,4-dichloro-3-oxobutanoate 6 as revealed by
GC–MS and NMR.
Scheme 1.
The yield of 4d was increased by adding chlorine to diketene, in contrast to the reverse order of
addition. The reduction of 4d to yield 1d using NaBH4 gave 10–16% of tert-butyl 3,4-epoxybutanoate
3d and minor amounts of tert-butyl 3-hydroxybutanoate (co-elution, GLC).
Since enzymatic reactions usually proceed with higher enantiomeric ratios when the site of reaction
is close to the stereocenter, resolutions of 4-chloro-3-hydroxybutanoates 1 were chosen to be performed
through transesterification of the secondary hydroxy group (Scheme 2).
In an enzyme-catalyzed transesterification reaction, an acyl donor is needed in order to acylate
(esterify) the enzyme. Since the present substrates are bifunctional, hydroxy esters, the ester group of
the substrate is also a possible acyl donor. This may lead to formation of side-products and thus lower
the yield. This problem was addressed firstly by letting ethyl 3-hydroxybutanoate react with itself using
lipase B from Candida antarctica (CALB) as a catalyst in the absence of an acyl donor. After 24 h, 25%
of the substrate had reacted to form dimers indicating that the substrate may indeed function as an acyl
donor. Therefore, it was judged necessary to use an acyl donor which is a much better acylating agent

B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
1403
Scheme 2.
than the substrate to be resolved. Vinyl alkanoates are known to be rapid acylating agents, and reactions
with these acyl donors did not lead to any detectable dimerization. When 2-chloro-, 2,2,2-trichloro or
2,2,2-trifluoro butanoates were used as acyl donors, small amounts of dimers were formed. Dimerization
of 1a, 1b and 1c was investigated in the absence of an acyl donor in a benzene solution using RML
as the catalyst. Substrates 1a and 1b dimerized quite readily, converting by 13 and 16%, respectively,
after 48 h, while 1c only dimerized to 3% during the same time. The reaction was particularly simple to
monitor for 1b since it liberated benzyl alcohol. The dimerization reaction was found to reduce the ee
of the resolution. A slight preference for the (R)-enantiomer was observed in the dimerization process
as opposed to that observed using a regular acyl donor. In the resolution of 1b using an acyl donor, no
benzyl alcohol was detected in either of the solvents. Since 1b dimerizes most readily, it is likely that no
dimerization took place for 1a–1d under the same conditions. Thus, the different E-values in the different
solvents cannot be attributed to different rates of dimerization in the different solvents.
Initially, several lipases were investigated for the resolution of ethyl 4-chloro-3-hydroxybutanoate
1a and phenylmethyl 4-chloro-3-hydroxybutanoate 1b in hexane using vinyl acetate as the acyl donor
(Table 1).
Table 1
Resolution of substrates 1a and 1b using different lipases in hexane; acyl donor vinyl acetate
All the lipases tested gave low E-values for the resolution of 1a. RML gave the best result with an
E-value of 8. For the resolution of 1b the E-values generally increased compared to the resolution of
1a, for which RML and Aspergillus lipase gave the highest E-values. CALB showed almost the same
E-value for both substrates.
Based on these results it was decided to focus on RML as the catalyst. Resolutions of 1a were
performed in different solvents using vinyl acetate as acyl donor. The results are summarized in Table 2.
The reaction medium affected the enantiomeric ratio as well as the rate of reaction. The more
hydrophobic solvents gave a higher rate of reaction. An increase in the E-value was observed on going
from hexane to tert-BuOMe, toluene, benzene and CCl4. In THF, dioxane and CHCl3, the reaction rates
were low, probably due to inactivation of the enzyme.22 E-values for the resolution of 1a in these solvents

1404
B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
Table 2
Resolution of 1a in different solvents using vinyl acetate as the acyl donor and RML as the catalyst
were not higher than for reactions in more hydrophobic solvents, and they were, therefore, excluded from
this study.
When different acylating agents were tested, it became evident that using vinyl propanoate, instead of
vinyl acetate as the acyl donor, increased the rate of reaction considerably. Hence, vinyl propanoate was
chosen as the acyl donor for further studies.
Resolutions of 1a, 1b, 1c and 1d by transesterification using vinyl propanoate as the acyl donor in
various solvents were performed using RML as the catalyst (Table 3). Preparative resolutions were
performed in order to determine the absolute configuration of the enantiomers. The (R)-enantiomers
of 1b, 1c and 1d, were isolated in high enantiomeric purity, ee >96–99%.
Table 3
E-values obtained during resolution of hydroxyesters 1a, 1b, 1c and 1d using Rhizomucor miehei
lipase, RML, and vinyl propanoate or lipase B from Candida antarctica, CALB, and vinyl acetate in
different solvents. Since eep-values were not easily determined, the E-values of 1d were determined
from ees-values and the degree of conversion using internal standards. Hence these E-values are not
as accurate as the others
2.1. Effects of solvent on the enantiomeric ratio
From Table 3 it is evident that there is an effect of solvent on the E-value. Reactions in hexane gave,
for 1a–1c, lower enantiomeric ratios than reactions in other solvents. In order to investigate the nature
of this effect, transesterifications were performed in mixtures of hexane in benzene (0, 50, 80, 95 and
100%), using ethyl 4-chloro-3-hydroxybutanoate 1a as the substrate. Fig. 1 shows ees and eep versus
the degree of conversion for the resolution of 1a. Each set of curves represents one E-value obtained for
each of the three solvent systems: benzene (E=15), hexane:benzene, 1:1 (E=10), and hexane (E=5). The
E-value decreased linearly with the increased amount of hexane (Fig. 2). The same effect was found for
the resolution of phenylmethyl 4-chloro-3-hydroxybutanoate 1b. The observed effect of solvent on the
enantiomeric ratio is obviously not a general trend for this lipase. Hexane was one of the better solvents

B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
1405
Figure 1. Effect of solvent on the enantiomeric ratio in resolution of 1a. Plot of ees and eep versus degree of conversion.
Squares=benzene; triangles=benzene:hexane 1:1; circles=hexane. Filled symbols=product fraction; open symbols=substrate
fraction
Figure 2. Enantiomeric ratio (E) obtained in various mixtures of hexane in benzene (0, 50, 80, 95 and 100%, respectively) in
resolution of 1a
tested in terms of E-values for the resolution of 1-(2-furyl)ethanol, while benzene, toluene and CCl4 gave
lower E-values.19
Various explanations have been offered for the effect of solvent on the E-value, such as change of
enzyme conformation23 or the fact that solvent molecules have different affinities for sites inside the

1406
B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
Figure 3. Resolution of 1a=circles, 1c=triangles and 1d=squares, using RML, vinyl propanoate and hexane. Filled symbols are
product fraction while empty symbols are substrate fraction
active site.24 Since there is a linear change of the E-value with increasing amounts of hexane in benzene,
we think that this indicates a more general solvation effect of the substrate.
The rate of reaction depended strongly on the nature of the solvent. Reactions in hexane gave high
rates compared to reactions performed in other solvents. Moreover, for resolution in solvents other than
hexane, the rate of reaction after ca. 40% conversion was extremely slow when performed on a small
scale. This problem was less serious when the resolution was performed on a preparative scale, in a
reaction vessel with a large head space.
2.2. Effect of the alkoxy group
The reason for varying the alkyl group of the ester was to improve the enantiomeric ratio by
introduction of a more bulky group, thereby increasing enantiomeric discrimination. Moreover, the
competing dimerization would be slowed down and, in the case of tert-butyl ester, be negligible.
The effect of the alkoxy group on the E-value can be read from Table 3. As can be seen, the effect is
solvent-dependent, but tert-butyl gave the highest E-value independent of the solvent used. For resolution
of 1a–1c, reactions in benzene and CCl4 gave an increase in E-value going from 1a -→ 1b -→ 1c, while in
tert-BuOMe the E-value seemed independent of the three alkoxy groups. For resolutions in hexane and
toluene an increase in E-value was observed going from 1a to 1b, while no clear difference in E-value
was observed for the resolutions of 1b and 1c in these solvents. The plot of ee versus conversion is shown
in Fig. 3 for the resolution of 1a, 1c and 1d in hexane. The resolutions of 1a and 1d in hexane were further
analyzed by plotting the remaining percentage of the individual substrate enantiomers versus time (see
Fig. 4). The (R)-enantiomers appear to react equally fast, and hence the increase in E-values, when going
from 1a to 1d, is caused by a slower rate of reaction of the (S)-enantiomer of 1d.

B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
1407
Figure 4. Percent remaining of each enantiomer plotted against reaction time in resolution of 1a=circles and 1d=squares. Filled
symbols are slow reacting enantiomers and empty symbols are fast reacting enantiomers
3. Experimental
3.1. General
3-Butenoic acid, ethyl (R)-4-chloro-3-hydroxybutanoate, ethyl 4-chloro-3-oxobutanoate, m-
chloroperoxybenzoic acid, DCC, 4-pyrrolidinopyridine, diketene (stab. w. cupric sulfate) and
cyclohexanol were purchased from Fluka; vinyl propanoate and benzylalcohol were from Aldrich.
The pet. ether had a boiling range of 60–80°C. Column chromatography was performed using silica
gel 60 from Fluka. Immobilized Rhizomucor miehei lipase (Lipozyme IM, Novo-Nordisk) had specific
activity 60 B.i.u./g.
3.2. Analyses
Optical rotation was determined using an AA-10 automatic polarimeter from Optical Activity Ltd,
and concentrations are given in g/100 mL. Chiral analyses were performed using CP-Chiracil-Dex
CB columns from Chrompack, and a G-TA column from Astec. 1a–1c: for detailed description see
the literature.25 For 1d: derivatized to its TMSi-derivative and analyzed on CP-Chiracil-Dex CB, DF:
0.25 mm, 8 psi, split: 85 mL/min, temp. prog.: 110–124, 1°C/min, t1: 10.70, t2: 11.04, Rs: 2.3. NMR
spectroscopy was performed in CDCl3 solutions, using Bruker DPX 300 and 400 instruments, operating
1
at 300 and 400 MHz for H and 75 and 100 MHz for 13C, respectively. Chemical shifts are in ppm rel.
to TMS and coupling constants in hertz. Enantiomeric ratios, E and equilibrium constants, Keq were
calculated using the computer program E and K calculator version 2.03.26,27

1408
B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
3.3. Transesterifications, small scale
Substrate (1.31×10−4 mol) was dissolved in solvent (3 mL) and vinyl propanoate (5 equiv.) was
added. The reaction was started by adding immobilized Rhizomucor miehei lipase (20 mg) to the reaction
mixture. The reactions were performed in a shaker incubator at 30°C. Chiral GLC analysis gave ees- and
eep-values from which conversion, c, was calculated, c=ees/(ees+eep). For substrate 1a, the conversion
was also measured using n-hexadecane as an internal standard, giving the same values as obtained
using the above-mentioned method. For substrate 1d, the conversion was measured using tetradecane
as the internal standard. In control experiments without enzyme, no acylation was observed using vinyl
propanoate as the acyl donor.
3.4. Dimerization
The degree of dimerization was detected by monitoring the decrease of the starting substrate, using
tetradecane as internal standard.
3.5. Ethyl 4-chloro-3-hydroxybutanoate 1a
Ethyl 4-chloro-3-oxobutanoate 4a (4.73 g, 26.1 mmol) was dissolved in EtOH (50 mL) and cooled to
0°C and NaBH4 (0.5 equiv.) was added. The reaction mixture was stirred for 1.5 h at room temperature
and HCl (0.1 M) was added. When the evolution of H2 had ceased, EtOH was removed in vacuo. The
concentrated reaction mixture was extracted with Et2O (4×75 mL), the organic fraction washed twice
with saline water (saturated), dried over MgSO4 and concentrated in vacuo. To remove yellow impurities
the substrate was purified by column chromatography using CH2Cl2:acetone, 20:1, giving 3.41 g (78%).
1H NMR, ethoxy group: 1.28 (t, 3H) and 4.18 (q, 2H) J=7.1, acyl part ABMXY-syst.: 2.60 (1H, A), 2.65
(1H, B), 3.59 (1H, X), 3.61 (1H, Y), 4.26 (m, 1H, M), JAB 16.5, JAM 7.4, JBM 4.9, JXY 11.2, JXM 5.5,
JYM 5.2, 3.34 (b, 1H); 13C NMR, 14.5, 38.9, 48.5, 61.4, 68.3 and 172.2.
3.6. Phenylmethyl 3-butenoate 2b
3-Butenoic acid (2, R1=H) (5 g, 58 mmol) and benzylalcohol (1 equiv.) were dissolved in CH2Cl2 (260
mL). The reaction mixture was cooled to 0°C and DCC (1 equiv.) and 4-pyrrolidinopyridine (0.78 g, 5.3
mmol) were added. The reaction was stirred at 0°C for 1 h, and further at room temperature for 20 h.
The reaction mixture was filtered to remove solid N,N-dicyclohexyl urea and concentrated in vacuo. The
residue was dissolved in hexane and more solid material precipitated. The reaction mixture was further
worked up by extraction with water (3×150 mL), 5% AcOH (2×100 mL) and water (2×100 mL). The
product was purified by column chromatography using pet. ether:acetone, 8:2, yielding 6.34 g (62%) of
phenylmethyl 3-butenoate 2b. 1H NMR, acyl part: 3.14 (td, 2H, J=6.9 and 1.4), 5.15 (m, 1H), 5.18 (m,
1H), 5.95 (m, 1H), benzyl part: 5.12 (s, 2H), 7.29–7.39 (m, 5H); 13C NMR, 39.5, 66.9, 119.1, 128.6,
128.7, 129.0, 130.5, 136.3 and 171.8.
3.7. Phenylmethyl 3,4-epoxybutanoate 3b
Phenylmethyl 3-butenoate 2b (5.06 g, 28.7 mmol) was dissolved in CHCl3 (200 mL) and m-chloro
peroxybenzoic acid (70% purity, 9.3 g) was added. The reaction was left stirring at room tempera-
ture for 7 days. The reaction mixture was extracted with phosphate buffer (pH 7.5, 2×100 mL) and

B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
1409
Na2CO3/NaHCO3 solution (pH 10, 4×100 mL). The product was further purified by column chromato-
graphy using pet. ether:acetone, 2:1, yielding 3b, 3.94 g (71%). 1H NMR, acyl part ABMPX-syst.: 2.59
(dd, 1H, P), 2.63 (1H, A), 2.66 (1H, B), 2.87 (dd, 1H, X), 3.33 (m, 1H, M), JAB 16.5, JAM 6.1, JBM 5.9,
JPX 4.8, JPM 2.6, JXM 4.4, benzyl part: 5.19 (s, 2H) and 7.35–7.40 (m, 5H); 13C NMR, 38.4, 47.0, 48.3,
67.0, 128.7, 128.8, 129.0, 136.1 and 170.6.
3.8. Phenylmethyl 4-chloro-3-hydroxybutanoate 1b
LiCl (2.14 g, 50.4 mmol) and CuCl2 (3.4 g, 25.2 mmol) were added to dry THF (50 mL). The mixture
was stirred for 20 min at room temperature and 3b (1.90 g, 9.9 mmol) in THF (10 mL) was added. The
reaction mixture was stirred at room temperature under N2 for 18 h. The reaction was quenched with
phosphate buffer (pH 7.0, 50 mL) and THF was removed in vacuo. The residue was diluted with water
(25 mL) and extracted with Et2O (5×50 mL). The organic phase was washed with water (2×50 mL),
dried over MgSO4, and concentrated in vacuo. The product was purified by column chromatography
using pet. ether:CHCl3:acetone, 3:7:1, yielding 1.95 g (86%) of 1b. 1H NMR, acyl part ABMXY-syst.:
2.70 (1H, A), 2.73 (1H, B), 3.62 (1H, X), 3.64 (1H, Y), 4.30 (m, 1H, M), JAB 16.6, JAM 7.1, JBM 5.1,
JXY 11.2, JXM 5.5, JYM 5.1, 3.05 (b, 1H), benzyl part: 5.19 (s, 2H), 7.30–7.43 (m, 5H); 13C NMR, 38.9,
48.5, 67.2, 68.3, 128.7, 128.9, 129.1, 135.7 and 172.0.
3.9. Cyclohexyl 3-butenoate 2c
The cyclohexyl ester was synthesized as described for 2b (Section 3.6) using cyclohexanol to yield
3.14 g (32%) of 2c. 1H NMR cyclohexyl part: 1.20–1.85 (m, 10H), 4.75 (m, 1H), acyl part: 3.04 (td, 2H,
J=7.0 and 1.3), 5.10 (m, 1H), 5.14 (m, 1H), 5.90 (m, 1H); 13C NMR, 24.1, 25.7, 31.9, 39.9, 73.2, 118.5,
131.0 and 171.3.
3.10. Cyclohexyl 3,4-epoxybutanoate 3c
1
The epoxide 3c was synthesized from 2c as described in Section 3.7. H NMR, cyclohexyl part:
1.23–1.88 (m, 10H), 4.83 (m, 1H), acyl part ABMPX-syst.: 2.53 (1H, A), 2.57 (1H, B), 2.56 (1H, P),
2.84 (1H, X), 3.29 (m, 1H), JAB 16.2, JAM 5.5, JBM 6.1, JPX 4.4, JXM 4.4; 13C NMR, 24.1, 25.7, 32.0,
38.8, 47.1, 48.5, 73.6 and 170.2.
3.11. Cyclohexyl 4-chloro-3-hydroxybutanoate 1c
The hydroxy ester 1c was synthesized from 3c as described in Section 3.8. The column chromato-
graphy step for 3c could be omitted, leading to 3.05 g, 74% overall yield from 2c. 1H NMR, cyclohexyl
part: 1.24–1.89 (m, 10H), 4.84 (m, 1H), acyl part ABMXY-syst.: 2.63 (1H, A), 2.66 (1H, B), 3.62 (1H,
X), 3.64 (1H, Y), 4.27 (m, 1H, M), JAB 16.4, JAM7.4, JBM 4.9, JXY 11.2, JXM 5.5, JYM 5.2, 3.2 (b, 1H);
13C NMR, 24.0, 25.6, 31.9, 39.2, 48.6, 68.4, 73.9 and 171.6.
3.12. tert-Butyl 4-chloro-3-oxobutanoate 4d
To a solution of diketene (1.65 g, 19.7 mmol) in CCl4 (6 mL) cooled to −20 to −25°C was added a
cooled solution of Cl2 (1 equiv.) in CCl4 (9 mL). The reaction was stirred at this temperature for 30 min.
tert-Butyl alcohol (1 equiv.) and pyridine (1 equiv.) in CCl4 (9 mL) were subsequently added at −5°C.

1410
B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
The reaction was stirred for 1 h. After removal of solid precipitate and evaporation of solvent the reaction
mixture was diluted in EtOAc and washed twice with water. After drying over MgSO4, the product was
purified by column chromatography using CH2Cl2 yielding 1.49 g (39%) of 4d. 1H NMR, keto tautomer:
1.50 (s, 9H), 3.59 (s, 2H), 4.23 (s, 2H), enol tautomer: 1.52 (s), 4.00 (s), 5.25 (s) and 12.21 (OH); 13C
NMR, keto tautomer: 28.3, 47.9, 48.5, 83.1, 166.0 and 196.3.
3.13. tert-Butyl 3,4-dichloro-3-oxobutanoate 6
1
tert-Butyl 3,4-dichloro-3-oxobutanoate 6 was isolated under the conditions used to purify 4d. H
NMR, keto tautomer: 1.53 (s, 9H), CH2Cl AB-syst.: 4.45 (1H), 4.49 (1H), JAB=16.2, 5.01 (1H), enol
tautomer: 1.57, 4.32, 12.35; 13C NMR: 28.0, 46.2, 60.1, 85.7, 163.4 and 192.0.
3.14. tert-Butyl 4-chloro-3-hydroxybutanoate 1d
The hydroxy ester 1d was synthesized from 4d as described in Section 3.5. Reaction time was 1 h. After
column chromatography using CH2Cl2:pet. ether:acetone, 7:3:1, tert-butyl 4-chloro-3-hydroxybutanoate
1d was isolated yielding 1.00 g (72%). 1H NMR, tert-butyl group: 1.50 (s, 9H), acyl part ABMXY-syst.:
2.56 (1H, A), 2.59 (1H, B), 3.60 (1H, X), 3.62 (1H, Y), 4.23 (m, 1H, M), JAB 16.6, JAM 7.0, JBM 5.0,
JXY 11.0, JXM 5.0, JYM 5.5, 3.27 (b, 1H); 13C NMR, 28.5, 39.9, 48.5, 68.5, 82.1 and 171.6.
3.15. tert-Butyl 3,4-epoxybutanoate 3d
1
The epoxide 3d was isolated under the conditions used for purification of 1d. H NMR, tert-butyl
group: 1.48 (s, 9H), acyl part ABMPX-syst.: 2.46 (1H, A), 2.52 (1H, B), 2.55 (1H, P), 2.84 (1H, X), 3.26
(m, 1H), JAB 16.2, JAM 5.4, JBM 6.1, JPX 4.9, JPM 2.7; 13C NMR, 28.4, 39.6, 47.1, 48.6, 81.6 and 170.1
3.16. Preparative resolutions
Substrate and vinyl propanoate (5 equiv.) were dissolved in the organic solvent. The reaction was
started by adding immobilized RML (300 mg) in the case of 1a–1c and 100 mg for 1d. The reaction
was shaken at room temperature until the desired conversion was reached. The enzyme was filtered
off, and the residue concentrated in vacuo. The alcohol and propanoate were separated by column
chromatography using CH2Cl2:acetone, 20:1.
3.17. Resolution of ethyl 4-chloro-3-hydroxybutanoate 1a
Resolution of 1a (2.17 g, 13.2 mmol) was performed using benzene (200 mL) as the solvent. The
reaction was stopped after 5 days when a conversion of 30% was reached, yielding (S)-7a, 0.76 g (24%),
25
1
ee 86%, [α] =−13.7 (c 1.02, CHCl3). H NMR, propanoate: 1.15 (t, 3H), 2.36 (q, 2H), J=7.5, for
D
ethoxy: 1.26 (t, 3H), 4.16 (q, 2H), J=7.1, acyl part ABMXY-syst.: 2.73 (1H, A), 2.77 (1H, B), 3.71 (1H,
X), 3.75 (1H, Y), 5.41 (m, 1H, M), JAB 16.3, JAM 6.9, JBM 6.1, JXY 11.8, JMX 4.7, JMY 4.8; 13C NMR,
9.3, 14.2, 27.8, 36.7, 45.4, 61.2, 69.3, 169.9 and 173.6.

B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
1411
3.18. Resolution of phenylmethyl 4-chloro-3-hydroxybutanoate 1b
Resolution of 1b (1.03 g, 4.5 mmol) was performed in benzene:hexane, 7:1 (80 mL), as the solvent.
After 5 days the reaction was stopped. Phenylmethyl (R)-(+)-4-chloro-3-hydroxybutanoate (R)-1b was
isolated yielding 0.30 g (29%), ee 96%, [α]25=+17.3 (c 5.02, CHCl3), [α]25=+17.8 (c 1.5, CHCl3).
D
D
Phenylmethyl (S)-(−)-4-chloro-3-propanoyloxybutanoate (S)-7b was isolated in a yield of 0.55 g (43%),
25
1
ee 77%, [α] =−8.2 (c 1.5, CHCl3). H NMR, propanoate ABX3-syst.: 1.12 (t, 3H, X), 2.28 (1H, A),
D
2.33 (1H, B), JAB 16.6, JAX 7.5, JBX 7.5, acyl part ABMXY-syst.: 2.82 (1H, A), 2.84 (1H, B), 3.72 (1H,
X), 3.75 (1H, Y), 5.43 (m, 1H, M), JAB 16.3, JAM 6.9, JBM 6.1, JXY 11.7, JXM 4.6, JYM 4.9, benzyl
group: 5.16 (s, 2H) and 7.32–7.41 (m, 5H); 13C NMR, 9.4, 27.8, 36.8, 45.5, 67.1, 69.4, 128.7, 128.8,
129.0, 136.0, 169.9 and 173.7.
3.19. Resolution of cyclohexyl 4-chloro-3-hydroxybutanoate 1c
Resolution of 1c (1.09 g, 4.9 mmol) was performed in CCl4 (150 mL) as the solvent. The re-
action was stopped after 4 days. Cyclohexyl (R)-(+)-4-chloro-3-hydroxybutanoate, (R)-1c, was iso-
lated in 0.48 g (44%) yield, ee 96%, [α]25=+17.6 (c 1.02, CHCl3). Cyclohexyl (S)-(−)-4-chloro-3-
D
propanoyloxybutanoate, (S)-7c, was isolated in a yield of 0.61 g (45%), ee 89%, [α]25=−9.7 (c 1.03,
D
1
CHCl3). H NMR, propanoate ABX3-syst.: 1.16 (t, 3H, J=7.6) and 2.36 (m, 2H), cyclohexyl group:
1.20–1.87 (m, 10H) and 4.79 (m, 1H), acyl part ABMXY-syst.: 2.74 (1H, A), 2.77 (1H, B), 5.43 (m, 1H,
M), 3.72 (1H, X), 3.76 (1H, Y), JAB 16.1, JAM 6.9, JBM 6.2, JXY 11.8, JXM 4.6, JYM 4.8; 13C NMR, 9.3,
24.0, 25.6, 27.8, 31.8, 37.1, 45.4, 69.4, 73.6, 169.3 and 173.6.
3.20. Resolution of tert-butyl 4-chloro-3-hydroxybutanoate 1d
Resolution of 1d (488 mg, 2.51 mmol) was performed using hexane (60 mL) as the solvent. The
reaction was stopped after five days. tert-Butyl (R)-(+)4-chloro-3-hydroxybutanoate, (R)-1d was isolated
in a yield of 205 mg (42%), ee 99%, [α]25=+22.0 (c 1.00, CHCl3). tert-Butyl (S)-(−)-4-chloro-3-
D
propanoyloxybutanoate, (S)-7d, was isolated in a yield of 301 mg (48%), ee >97%, [α]25=−8.5 (c 1.00,
D
CHCl3). 1H NMR, propanoate ABX3-syst.: 1.17 (t, 3H, J=7.6), 2.37 (m, 2H), tert-butyl group: 1.46 (s,
9H), acyl part ABMXY-syst.: 2.67 (1H, A), 2.70 (1H, B), 5.40 (m, 1H, M), 3.70 (1H, X), 3.74 (1H, Y),
JAB 16.1, JAM 7.2, JBM 6.0, JXY 11.7, JXM 4.7, JYM 4.7; 13C NMR, 9.4, 27.8, 28.3, 38.0, 45.5, 69.5,
169.1 and 173.6.
3.21. Absolute configurations
The faster reacting enantiomer of ethyl 4-chloro-3-hydroxybutanoate was identified by co-injection
of commercially available ethyl (R)-4-chloro-3-hydroxybutanoate (R)-1a with the racemic sample. The
(R)-enantiomer of phenylmethyl 4-chloro-3-hydroxybutanoate (R)-1b has been described previously.1
The reported specific rotation of +8.7 (c 5.26, CHCl3) indicates, however, that the enantiomeric excess
of this product was only moderate. The absolute configuration of the enantiomers of cyclohexyl 4-
chloro-3-hydroxybutanoate has not been described. (R)- and (S)-1c were verified by synthesizing 1c by
alcoholysis of racemic ethyl 4-chloro-3-hydroxybutanoate 1a and ethyl (R)-4-chloro-3-hydroxybutanoate
(R)-1a on a small scale. Cyclohexanol was used as an acyl acceptor (5 equiv.) in hexane, and lipase B
from Candida antarctica was used as the catalyst. In the resolution of ethyl 4-chloro-3-hydroxybutanoate
1a by alcoholysis the (S)-enantiomer was transformed to products faster than the (R)-enantiomer (Chiral

1412
B. H. Hoff, T. Anthonsen / Tetrahedron: Asymmetry 10 (1999) 1401–1412
GLC), E=6. It is therefore concluded that the most abundant enantiomer seen in the chiral analysis of
1c must be (S)-1c. Cyclohexyl (R)-4-chloro-3-hydroxybutanoate (R)-1c formed in the alcoholysis of
ethyl (R)-4-chloro-3-hydroxybutanoate (R)-1a coinjected with the racemic sample, and eluted with the
enantiomer which reacts slower in these transesterifications. The specific rotation of (R)-tert-butyl-4-
chloro-3-hydroxybutanoate (R)-1d was compared to the one previously reported, +18.09 (c 5.1, CHCl3).2
Acknowledgements
We gratefully acknowledge a fellowship to BHH from the Norwegian Research Council (NFR). We
also want to thank Novo-Nordisk A/S for a gift of Rhizomucor miehei lipase and lipase B from Candida
antarctica, and Julie Jackson for linguistic help and comments.
References
1. Sih, C. J. US Patent 4,710,468; 1987.
2. Tinti, M. O. Eur. Patent 0208662 B1; 1986.
3. Santaniello, E.; Casati, R.; Milani, F. J. Chem. Res. (S) 1984, 132–133.
4. Tricone, A.; Nicolaus, B.; Lama, L.; Marsiglia, F.; Gambacorta, A. Biotechnol. Lett. 1991, 13, 31–34.
5. Fuganti, C.; Grasselli, P.; Casati, P.; Carmeno, M. Tetrahedron Lett. 1985, 26, 101–104.
6. Zhou, B.-N.; Gopalan, A. S.; van Middelsworth, F.; Shieh, W.-R.; Sih, C. J. J. Am. Chem. Soc. 1983, 105, 5925–5926.
7. Aragozzini, F.; Valenti, M.; Santaniello, E.; Ferraboschi, P.; Grisenti, P. Biocatalysis 1992, 5, 325–332.
8. Genet, J. P.; Pinel, C.; Ratovelomanana-Vidal, V.; Mallart, S.; Pfister, X.; Bischoff, L.; Cano De Andrade, M. C.; Darses,
S.; Galopin, C. Tetrahedron: Asymmetry 1994, 5, 675–690.
9. Trost, B. M.; Hanson, P. R. Tetrahedron Lett. 1994, 35, 8119–8122.
10. Kitamura, M.; Ohkuma, T.; Takaya, H.; Noyori, R. Tetrahedron Lett. 1988, 29, 1555–1556.
11. Santaniello, E.; Ferrabochi, P.; Grisenti, P.; Manzocchi, A.; Trave, S. Gazz. Chim. Ital. 1989, 581–584.
12. Garcia, M. J.; Rebolledo, F.; Gotor, V. Tetrahedron: Asymmetry 1993, 4, 2199–2210.
13. Sakaki, J.; Sakoda, H.; Sugita, Y.; Sato, M.; Kaneko, C. Tetrahedron: Asymmetry 1991, 2, 343–346.
14. Anthonsen, T.; Jongejan, J. A. Meth. Enzymol. 1997, 286, 473–495.
15. Anthonsen, T.; Hoff, B. H. Chem. Phys. Lipids 1998, 93, 199–207.
16. Alcántara, A. R.; de Fuentes, I. E.; Sinisterra, J. V. Chem. Phys. Lipids 1998, 93, 169–184.
17. Valivety, R. H.; Halling, P. J.; Macrae, A. R. FEBS Lett. 1992, 301, 258–260.
18. Weber, H. K.; Zuegg, J.; Faber, K.; Pleiss, J. J. Mol. Catal. B 1997, 3, 131–138.
19. Kaminska, J.; Górnicka, I.; Sikora, M.; Góra, J. Tetrahedron: Asymmetry 1996, 7, 907–910.
20. Bevinakatti, H. S.; Banerji, A. A.; Newadkar, R. V.; Mokashi, A. A. Tetrahedron: Asymmetry 1992, 3, 1505–1508.
21. Claus, K. Liebigs Ann. Chem. 1980, 494–502.
22. Dellamora-Oriz, G. M.; Martins, R. C.; Rocha, W. L.; Dias, A. P. Biotechnol. Appl. Biochem. 1997, 26, 31–37.
23. Ueji, S.; Fujino, R.; Okubo, N.; Miyazawa, T.; Kurita, S.; Kitadani, M.; Muromatsu, A. Biotechnol. Lett. 1992, 14,
163–168.
24. Nakamura, K.; Takebe, Y.; Kitayama, T.; Ohono, A. Tetrahedron Lett. 1991, 32, 4941–4044.
25. Hoff, B. H.; Anthonsen, T. Fresenius J. Anal. Chem. 1999, accepted for publication.
26. Anthonsen, H. W.; Hoff, B. H.; Anthonsen, T. Tetrahedron: Asymmetry 1996, 7, 2633–2638.
27. Anthonsen, H. W. http://bendik@kje.ntnu.no 1996–7.