Enantiomer selective acylation of racemic alcohols by lipases in continuous-flow bioreactors

Article abstract of DOI:10.1016/j.tetasy.2008.01.002

Continuous-flow mode enantiomer selective acetylations of racemic 1-phenylethanol, 1-cyclohexylethanol, and 1-phenylpropan-2-ol (rac-1a-c, respectively) with vinyl acetate were performed in small stainless steel packed-bed reactors filled with different commercial lipase preparations. In several lipase-filled columns, highly enantiomer selective (E > 100) kinetic resolutions of these alcohols were achieved. In most cases, comparison of the continuous-flow and batch mode (shake flask) biotransformations indicated similar enantiomer selectivities (E) but higher productivities (specific reaction rate: r) in the corresponding continuous-flow reaction. The effect of temperature (0-60 °C) and pressure (1-120 bar) on the continuous-flow acetylation of racemic 1-phenylpropan-2-ol was investigated in an immobilized Candida antarctica lipase B (CaLB) filled reactor. Pressure had no significant effect on r and E. Expectedly, a monotonous increase of specific reaction rate (r) was observed within this temperature range. Most surprisingly, the enantiomer selectivity had a maximum (E ～ 25, at 20 °C) and a minimum (E ～ 7, at 50 °C). The continuous-flow reactions in CaLB-filled columns were successfully applied for preparative scale kinetic resolutions of rac-1a-c.

Full text of DOI:10.1016/j.tetasy.2008.01.002

Available online at www.sciencedirect.com
Tetrahedron: Asymmetry 19 (2008) 237–246
Enantiomer selective acylation of racemic alcohols by
lipases in continuous-flow bioreactors
a
b
b
a,
*
¨
Csaba Csaj a´ gi, G a´ bor Szatzker, Enik o} Rita T o} ke, L a´ szl o´ Urge,
a
b,
*
Ferenc Darvas and L a´ szl o´ Poppe
a
ThalesNano Nanotechnology Inc., Graphisoft Park, H-1031 Budapest, Z a´ hony u. 7., Hungary
Department of Organic Chemistry and Technology, and Research Group for Alkaloid Chemistry of the Hungarian Academy of Sciences,
Budapest University of Technology and Economics, H-1111 Budapest, M }u egyetem rkp. 3, Hungary
b
Received 27 November 2007; accepted 4 January 2008
Abstract—Continuous-flow mode enantiomer selective acetylations of racemic 1-phenylethanol, 1-cyclohexylethanol, and 1-phenylpro-
pan-2-ol (rac-1a–c, respectively) with vinyl acetate were performed in small stainless steel packed-bed reactors filled with different com-
mercial lipase preparations. In several lipase-filled columns, highly enantiomer selective (E > 100) kinetic resolutions of these alcohols
were achieved. In most cases, comparison of the continuous-flow and batch mode (shake flask) biotransformations indicated similar
enantiomer selectivities (E) but higher productivities (specific reaction rate: r) in the corresponding continuous-flow reaction. The effect
of temperature (0–60 °C) and pressure (1–120 bar) on the continuous-flow acetylation of racemic 1-phenylpropan-2-ol was investigated
in an immobilized Candida antarctica lipase B (CaLB) filled reactor. Pressure had no significant effect on r and E. Expectedly, a monot-
onous increase of specific reaction rate (r) was observed within this temperature range. Most surprisingly, the enantiomer selectivity had
a maximum (E ꢀ 25, at 20 °C) and a minimum (E ꢀ 7, at 50 °C). The continuous-flow reactions in CaLB-filled columns were successfully
applied for preparative scale kinetic resolutions of rac-1a–c.
Ó 2008 Elsevier Ltd. All rights reserved.
1
. Introduction
continuous systems for food and related applications.
For example, esterification of glycerol with (poly)unsatu-
rated fatty acids, interesterification of butterfat, produc-
9
10
Biocatalysts are predominantly suited for the production of
enantiomerically pure stereoisomers. The production of
optically active intermediates is an area of growing demand
in the pharmaceutical and fine chemical industry, and here
biocatalysis has developed from a niche technology to a
1
1,12
tion of structured lipids by acidolysis,
alcoholysis for
1
3,14
15
production biodiesel
monoglyceride production, and hydrolyzes of rice bran-
or lubricant, glycerolysis for
1
6
1
7
18
oil or the oil content of wastewater were investigated
in continuous systems.
1
–3
widely used manufacturing method.
Biocatalytic steps
are already being used industrially to manufacture a wide
range of products, including drugs, agricultural chemicals,
The vast majority of the enzymatic enantioselective pro-
4
,5
1,3,6,7
organics, fine chemicals and plastics. Hydrolases, espe-
cially lipases have proved to be versatile biocatalysts for
cesses was performed in batch mode.
Surprisingly,
there are only a few examples of hydrolase-catalyzed enan-
tioselective processes carried out in continuous-flow sys-
tems. Most of these studies were aimed at the biocatalytic
syntheses of chiral pharmaceutical intermediates and were
performed on a relatively large scale using immobilized
6
synthetic biotransformations. Lipases were used in various
stereoselective biotransformations such as kinetic resolu-
6,7
tions, deracemisations, and dynamic kinetic resolutions.8
4
,19
2
Numerous lipase-catalyzed reactions—mostly involving
triglycerides and related compounds—were studied in
lipases in packed-bed reactor.
For example, the kinetic
0
resolution of racemic ibuprofen or 2-(4-chlorophenoxy)
2
1
propionic acid were performed by continuous-mode
esterification. A study on the esterification of racemic ibu-
profen in packed-bed plug-flow reactor (PFR) compared
with a continuous stirred tank reactor (CSTR) revealed
*
Corresponding authors. Tel.: +36 1 8818500; fax: +36 1 8818501 (L.U.),
tel.: +36 1 4632229; fax: +36 1 4633297 (L.P.); e-mail addresses:
laszlo.urge@thalesnano.com; poppe@mail.bme.hu
0957-4166/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2008.01.002

238
C. Csaj a´ gi et al. / Tetrahedron: Asymmetry 19 (2008) 237–246
that PFR was the most appropriate in terms of minimizing
the residence time. Enantiomer selective hydrolyses of
2. Results and discussion
2
2
2
3
racemic 1-chloro-2-acetoxy-3-(1-naphthyloxy)-propane,
One of the most investigated secondary alcohols in enzyme
2
4
25
1–3
naproxen methyl or ethoxyethyl esters or a-methyl-b-
catalyzed kinetic resolutions is 1-phenylethanol rac-1a, a
2
6
propiothiolactone were also performed in continuous
systems. Batchwise loop reactor systems were used for
two-step production of (R)- and (S)-ethyl-3-hydroxybuty-
small molecule containing a rigid aromatic ring and a
methyl group adjacent to the asymmetric carbon center.
This alcohol and its saturated analogue 1-cyclohexyletha-
nol rac-1b proved to be versatile test substrates in our
recent work for screening the enantiomer selectivities of
2
7
rate (HEB) involving acetylation and ethanolysis.
3
6
Only a few systematic studies on lipase-catalyzed kinetic
resolutions of secondary alcohols in continuous-flow reac-
tor systems have been performed. For example, kinetic
resolutions of racemic 1-phenylethanol with Candida
antarctica lipase B in organic solvent were compared in
batch stirred tank reactor (BSTR) and recirculated
lipases produced by solid state fermentation. In addition,
1-phenylpropane-2-ol rac-1c—representing a valuable class
of intermediates for benzodiazepine drugs—also proved to
be a valuable test substrate in kinetic resolutions with com-
3
7
mercial lipases. These typical secondary alcohols—the
small and relatively rigid 1-phenylethanol rac-1a, the satu-
rated, even more bulky 1-cyclohexylethanol rac-1b and the
similar but conformationally more flexible secondary alco-
hol 1-phenylpropane-2-ol rac-1c—were chosen as sub-
strates for this study. The kinetic resolutions of these
racemic secondary alcohols rac-1a–c were performed in a
heat and pressure resistant stainless steel column (Cat-
2
8
packed-bed reactor (RPBR). The kinetic resolutions of
2
9
racemic 2-methyl-1-pentanol with Pseudomonas sp. lipase,
1
3
0
-methoxy-2-propanol with C. antarctica lipase B, and
other simple secondary alcohols (2-heptanol, 2-octanol
3
1
and 2-nonanol) with Burkholderia cepacia lipase (BCL)
were systematically studied in continuously operated
packed-bed reactors (PBR). In a study on the kinetic reso-
lution of racemic 1-phenylethanol with vinyl propionate,
an immobilized Mucor miehei lipase entrapped in photo-
crosslinkable resins was applied for 72 days of continuous
TM
Cart ) using a sophisticated continuous-flow bench-top
TM
lab reactor system (X-Cube ) allowing precise control
ꢁ
1
and variation of flow rate (0.1–5 mL min ), temperature
(0–60 °C) and back pressure (1–120 bar) (Scheme 1).
3
2
production of enantiomerically pure (S)-alcohol.
Recent changes have influenced not only the production
side of pharmaceutical, fine chemical and agrochemical
industry but also generated a need for preparing an in-
creased number of compounds in research laboratories.
Thus, the search for new technologies that allow the auto-
mation of synthetic processes has gained a major inter-
OAc
p
R
OAc
rac-2a-c
3
3,34
R
est.
Despite the strong trend for automation, the
major part of high-throughput chemistry is still carried
T
out in batches,1
,3,6,7
2a-c
Ac O
(R)-
whereas flow-through processes are
2
rather restricted to production processes. Consequently,
the main advantages of the flow-through approach—such
as facile automation, reproducibility, safety, and process
reliability—are not much exploited at research phase.
Thus, technologies are missing that allow rapid transfer
of methods from the laboratory scale to production plant
scale without time-consuming adaptation and optimiza-
tion. Continuous-flow processes can overcome these
restrictions and allow the rapid preparation of compounds
with minimum workup. Many of these approaches use
immobilized reagents and catalysts, which are embedded
Et N
3
+
OH
OH
R
R
vinyl acetate
hexane - THF
rac-1a-c
(S)-1a-c
3
3
in a structured flow-through reactor. Ideal reactors are
those with high space time/yield to increase the efficiency
of the biotransformation. A packed-bed of immobilized
enzyme has clear advantages in that voidage is low: 34%
1
,2
a
b
c
3
5
compared to over 80–90% for a stirred tank reactor.
R
In this study, our aim was to perform a systematic compar-
ison of lipase-catalyzed kinetic resolutions of several sec-
ondary alcohols in continuous-flow mode and shake flask
batch mode using immobilized and non-immobilized
commercial lipases. In addition, we also considered it
worthwhile investigating the much less studied influence
of the temperature and pressure on the reactivity and selec-
tivity of the lipase-catalyzed continuous-flow kinetic
resolutions.
Scheme 1. Kinetic resolution of racemic secondary alcohols rac-1a–c in
continuous-flow enzyme reactors [.: pump; ꢀꢀ:temperature control unit;
ꢂ: pressure regulation valve].
First, the typical substrate 1-phenylethanol rac-1a has been
investigated in a continuous-flow reactor filled with lipase
B from C. antarctica (CaLB, immobilized on an acrylic

C. Csaj a´ gi et al. / Tetrahedron: Asymmetry 19 (2008) 237–246
239
resin). In previous screens for enantiomer selective
(R)-2a
(
S)-2a
3
6,38
lipases,
vinyl acetate proved to be a proper acylating
agent in hexane–THF solution. Therefore, this reaction
system has been chosen for this study as well. Our first
investigations (5–50 mg mL substrate concentration at
(R)-1a
(S)-1a
ꢁ
1
ꢁ
1
0
.1–1 mL min
flow rate) revealed that almost perfect
ꢁ
1
kinetic resolution of a 10 mg mL solution of racemic 1-
phenylethanol rac-1a in hexane–THF–vinyl acetate 2:1:1
mixture at 25 °C can be performed in a small column
(
0.82 mL) filled with immobilized CaLB (0.4 g) at a flow
rac-1a + rac-2a
ꢁ1
rate of 0.1 mL min . Analysis of the outflow from the col-
umn (at each 10 min after starting) indicated that the sta-
tionary state of the continuous-flow reaction [50%
conversion to (R)-acetate, (R)-2a (99.2% ee) and (S)-alco-
hol, (S)-1a (98.8% ee)] was reached after 30 min (Fig. 1).
continuous
(60 min)
Conversion [%]
batch
(24 h)
[%]
ee 1a [%]
ee 2a [%]
batch
(8 h)
100
8
0
0
0
0
batch
4 h)
(
6
4
2
7
8
9
R
t [min]
Figure 2. Comparison of kinetic resolutions of racemic 1-phenylethanol
rac-1a with CaLB (on acrylic resin) in a continuous-flow enzyme reactor
and shake flask batch reaction using enantioselective GC analysis. Data
curves are excerpts from the following GC’s (from up to down): a mixture
of rac-1a and rac-2a; continuous-flow reaction at 60 min after starting;
batch reaction at 24 h, 8 h, and 4 h.
0
1
0
20
30
40
50
60
Time [min]
Figure 1. The time course of a kinetic resolution of racemic 1-phenyleth-
anol rac-1a in a continuous-flow enzyme reactor filled with CaLB (on
acrylic resin).
immobilized CaLB were kept constant for testing other
commercial lipase preparations in continuous-flow reactors
(Table 1A). The reactions under comparable conditions
(10 mg mL rac-1a in 2 mL of hexane–THF–vinyl acetate
2:1:1 solution; enzyme, 20 mg; 25 °C; 1000 rpm) were also
performed in shake flask batch reactions with the same
enzymes (Table 1B).
The reaction with immobilized CaLB was also performed
in shake flask batch reaction under comparable conditions
ꢁ1
ꢁ
1
(
2
10 mg mL rac-1a in 2 mL of hexane–THF–vinyl acetate
:1:1 solution; CaLB, 20 mg; 25 °C; 1000 rpm). As the
enantiomer selective GC analyses of the samples from the
two reactions revealed (Fig. 2), the batch reaction at 4 h
or 8 h reached significantly lower conversions than the
desired 50%, thus the slower reacting alcohol enantiomer
For the comparison of the continuous-flow and batch reac-
tions, several of their features can be considered. One of the
general features of the kinetic resolutions is the well-char-
acterized degree of enantiomer selectivity (E), which, under
irreversible reaction conditions, can be calculated from two
of the following three properties: the conversion (c), the
(
1
S)-1a contained significant amounts of its antipode (R)-
a. The batch reaction gave comparable results to the con-
tinuously operating system only after 24 h (Fig. 2).
This indicates one of the main advantages of the packed-
bed continuous system over the batch reaction: highly
enantiopure samples of the slower reacting enantiomer
can be obtained much faster by the continuously operating
system (quantities satisfactory for GC, NMR or specific
rotation measurements can be obtained in less than 2 h
vs the 24 h required for the batch reaction).
enantiomeric excess of the product (ee ) and enantiomeric
P
3
9,40
excess of the remaining substrate (eeS).
Thus, the value
3
9
39
of E can be determined from c and ee , from c and eeS
P
4
0
or from ee and ee . In addition to their selectivities, the
P
S
productivity of the reactions can also be characterized by
the so-called specific reaction rate (r), which indicates the
amount of the product that can be formed in a minute
by 1 g of enzyme. The specific reaction rate for a continu-
4
1
For the sake of comparability, the conditions
ous-flow system (rflow) can be calculated from the concen-
ꢁ
1
ꢁ1
ꢁ1
(
0.1 mL min flow; 10 mg mL rac-1a in hexane–THF–
tration of the product ([P] (lmol mL )), the flow rate (f
ꢁ
1
vinyl acetate 2:1:1 solution; 25 °C; 0.82 mL column filled
with ca. 0.4 g of enzyme) found to be optimal for the
(mL min )), and the mass of the applied enzyme (m (g))
e
according to Eq. 1.

240
C. Csaj a´ gi et al. / Tetrahedron: Asymmetry 19 (2008) 237–246
Table 1. Kinetic resolution of racemic 1-phenylethanol rac-1a by acylation with vinyl acetate catalyzed by various lipases: panel A: reactions in
continuous-flow packed-bed reactor at stationary state; panel B: reactions in shake flask
a
a
a
b
c
ꢁ1 ꢁ1
No.
Enzyme (in packed-bed column)
c (%)
ee(S)-1a (%)
ee(R)-2a (%)
E
r (lmol min
g
)
Panel A
1
2
3
4
5
6
7
8
CaLB
50
50
19
18
52
24
15
6
98.8
96.8
31.1
28.9
99.8
39.6
21.7
13.6
99.2
97.4
99.1
98.9
94.6
99.6
53.2
97.2
ꢅ200
>200
>200
>100
>100
ꢅ200
4.0
10.2
10.2
4.0
3.7
10.6
5.0
Lipase, Pseudomonas cepacia, IM
Lipozyme Mucor miehei, IM
Lipozyme TL IM
Amano lipase AK
Amano lipase PS
Lipase, Candida rugosa
Lipase, porcine pancreas
TM
3.1
1.2
81
d
d
d
b
e
ꢁ1 ꢁ1
No.
Enzyme (in shake flask)
c
(%)
ee(S)-1a (%)
ee(R)-2a (%)
E
r (lmol min
g
)
Panel B
9
CaLB
51
57
42
31
75
44
16
16
99.6
99.5
78.3
55.4
99.0
84.4
22.4
32.1
98.3
80.6
98.7
97.7
47.0
99.4
53.7
98.5
ꢅ200
53
>200
>100
13
5.7
7.5
4.1
2.5
17.3
4.4
1.1
1.0
10
11
12
13
14
15
16
Lipase, Pseudomonas cepacia, IM
Lipozyme Mucor miehei, IM
Lipozyme TL IM
Amano lipase AK
Amano lipase PS
Lipase, Candida rugosa
Lipase, porcine pancreas
TM
ꢅ200
4.1
>100
a
ꢁ1
ꢁ1
Flow reaction at stationary state [25 °C, 0.1 mL min flow of hexane–THF–vinyl acetate 2:1:1 mixture containing 10 mg mL (0.082 M) rac-1a in
TM
enzyme filled CatCart column; sample taken at 60 min after starting]. The ee values of (R)-2a and (S)-1a and c were determined by GC.
Enantiomer selectivity (E) was calculated from ee(S)-1a and ee(R)-2a (and confirmed by independent calculation c and ee(R)-2a
b
40
39
,
data not shown). Due to
sensitivity to experimental errors, E values calculated in the 100–250 range are reported as >100, values in the 250–500 range are reported as >200 and
values calculated above 500 are given as ꢅ200.
c
Specific reaction rate (r) calculated for the flow system according to Eq. 1.
d
ꢁ1
Shake flask batch reaction at 24 h [25 °C, 1000 rpm, sealed amber glass vial, 2 mL of hexane–THF–vinyl acetate 2:1:1 mixture, 10 mg mL (0.082 M)
rac-1a and enzyme (20 mg)].
Specific reaction rate (r) calculated for the batch system according to Eq. 2.
e
ꢀ
ꢁ
½
Pꢃ ꢄ f lmol
are in accordance with the previous considerations: the voi-
dage in packed-bed immobilized enzyme reactor is much
lower (ꢀ34%) compared to over 90% for the stirred tank
r
flow
¼
ð1Þ
m
e
min ꢄ g
3
5
(shake flask) reactor.
A stirred (or shake flask) batch reaction can also be char-
acterized by the specific reaction rate (rbatch), which can
be calculated from the amount of the product (n (lmol)),
the reaction time (t (min)), and the mass of the applied
enzyme (m (g)) according to Eq. 2.
In most cases, no significant differences were found in the
enantiomer selectivities (E) of the enzymes in the kinetic
resolution of racemic 1-phenylethanol rac-1a when using
continuous-flow or batch reactions (Table 1A and B,
respectively). However, in the cases of immobilized lipase
from P. cepacia (Table 1, Nos. 2 and 10) and Lipase AK
P
e
ꢀ
ꢁ
n
P
lmol
r
batch
¼
ð2Þ
t ꢄ m
e
min ꢄg
(Table 1, Nos. 5 and 13), the reactions performed in contin-
uous-flow system showed higher enantiomer selectivities
(E > 200 vs 53; and E > 100 vs 13, respectively). On the
other hand, the lipase from porcine pancreas was more
selective in the batch mode operation for this particular
case than in the continuous-flow system (E > 100 vs 81, Ta-
ble 1, Nos. 16 and 8, respectively).
Although the above defined specific reaction rate can be
calculated at any degree of conversion (c), rigorous com-
parisons between the productivity of a continuous-flow
reaction and its batch mode counterpart using their r val-
ues can only be made at the same degree of conversions
(
as the rate of the product formation is not a linear func-
tion of c).
There were immobilized (CaLB and lipase from P. cepacia,
Table 1, Nos. 1 and 2) as well as powder form lipases (Li-
pase AK, Table 1, No. 5) among the biocatalysts, which
proved to be productive and selective enough for an effi-
cient kinetic resolution of rac-1a in continuous-flow mode.
It is worth mentioning that the commercial non-immobi-
lized lipase powders (Table 1, Nos. 5–8) performed equally
well in the continuous-flow packed-bed reactors as the
immobilized lipases (Table 1, Nos. 1–4). As the existing lit-
For the first look, it is obvious that the productivity of a
given enzyme is higher in the continuous-flow system than
in the corresponding batch reaction. For instance, r
flow
ꢁ
ꢁ
1
ꢁ1
ꢁ1
(
(
10.2 lmol min
5.7 lmol min
g
) for CaLB is about twice of rbatch
) at c ꢀ 50% (Table 1, Nos. 1 and 9).
1
g
An even higher difference was observed (Table 1, Nos. 7
and 15) for the lipase from Candida rugosa at c = 15–
ꢁ1
ꢁ1
ꢁ
1
6%, where rflow (3.1 lmol min
g
) was almost three
erature examples of hydrolase-catalyzed biotransforma-
1
ꢁ1
10–27
times larger than rbatch (1.1 lmol min
g
). Our findings
tions in continuous-mode
applied almost exclusively

C. Csaj a´ gi et al. / Tetrahedron: Asymmetry 19 (2008) 237–246
241
the immobilized forms of the enzymes, our results with the
lyophilized powder forms of the lipases broaden the appli-
cability of these enzymes in continuous-mode biotransfor-
mations performed in non-aqueous solvent systems.
drop in the enantiomer selectivity compared to the powder
form enzyme (E > 100 vs 18; Table 2, Nos. 6 and 2,
respectively).
Although it was thought that the higher degree of confor-
mational mobility of the lastly investigated racemic 1-phen-
ylpropan-2-ol rac-1c will result in lower degree of
enantiomer selectivity compared to the less mobile 1-phen-
ylethanol 1a, most commercial lipases retained their high
enantiomer selectivities (Table 3) except CaLB (Table 3,
Nos. 1 and 9). Interestingly, in the case of the kinetic reso-
lution of rac-1c, there is no remarkable difference in the
productivities in favor of the continuous-flow mode reac-
tions (Table 3). The results with this substrate rac-1c also
indicate that there are immobilized (lipase from P. cepacia,
Table 3, No. 2) as well as powder form lipases (Lipase AK,
Table 3, No. 5), which are capable of performing efficient
kinetic resolution in continuous-flow mode. The compari-
son of Lipase PS and immobilized lipase from P. cepacia
is interesting within this series as well. In kinetic resolutions
of racemic 1-phenylpropan-2-ol rac-1c, immobilization of
the lipase from P. cepacia on ceramic particles resulted in
a dramatic enhancement in productivity (r = 9.2 vs 1.1; Ta-
ble 3, Nos. 2 and 6, respectively) but in this case without
decrease in the enantiomer selectivity (E > 100 for both
preparations, Table 3, Nos. 2 and 6).
Next, the kinetic resolutions of racemic 1-cyclohexyletha-
nol rac-1b with the same commercial lipases were studied
under continuous-flow (Table 2A) and batch mode condi-
tions (Table 2B). The results with this more bulky substrate
rac-1b were a bit surprising, since in a recently reported
screen for enantiomer selective lipases, significantly more
preparations exhibited high enantiomer selectivity
(
E > 100) in kinetic resolution of racemic 1-cyclohexyletha-
nol rac-1b than in the resolution of racemic 1-phenyletha-
3
6
nol rac-1a (14 vs 7 from the 38 investigated). Among
the investigated enzymes only CaLB was selective and pro-
ductive enough for continuous-flow application (Table 2,
No. 1). The other enzyme with satisfactory enantiomer
selectivity (E > 100) was the powder form of Lipase PS
(
(
Table 2, No. 6). However, the low specific rate
ꢁ1
ꢁ1
r = 1.8 lmol min
g ) for this enzyme toward 1-cyclo-
hexylethanol 1b is not enough for an effective kinetic reso-
lution process. In this series of reactions, the comparison of
Lipase PS and immobilized lipase from P. cepacia is also
interesting. According to our best knowledge, Lipase PS
is the powder form of the lipase from P. cepacia, which is
also commercially available in immobilized form (on cera-
mic particles). In kinetic resolutions of racemic 1-cyclo-
hexylethanol (rac-1b), immobilization of the lipase from
P. cepacia on ceramic particles resulted in a significant
After the analysis of the kinetic resolutions in continuous-
flow bioreactors on a small scale, preparative biotransfor-
mations were also performed with the CaLB-filled bioreac-
Table 2. Kinetic resolution of racemic 1-cyclohexylethanol rac-1b by acylation with vinyl acetate catalyzed by various lipases: panel A: reactions in
continuous-flow packed-bed reactor at stationary state; panel B: reactions in shake flask
a
a
b
c
ꢁ1 ꢁ1
No.
Enzyme (in packed-bed column)
c (%)
ee(R)-2b (%)
E
r (lmol min
g
)
Panel A
1
2
3
4
5
6
7
8
CaLB
44
50
5
1
36
9
99.8
89.6
96.1
86.3
94.8
98.9
84.4
95.0
ꢅ200
18
51
14
37
>100
12
39
8.6
9.7
0.9
0.1
7.0
1.8
0.3
0.8
Lipase, Pseudomonas cepacia, IM
Lipozyme Mucor miehei, IM
Lipozyme TL IM
Amano lipase AK
Amano lipase PS
Lipase, Candida rugosa
Lipase, porcine pancreas
TM
2
4
d
d
b
e
ꢁ1 ꢁ1
No.
Enzyme (in shake flasks)
c
(%)
ee(R)-2b (%)
E
r (lmol min
g
)
Panel B
9
CaLB
50
55
18
7
57
10
43
10
98.8
77.1
95.4
91.1
73.6
87.3
61.7
86.0
>100
7.8
42
22
6.6
15
5.3
6.6
1.2
0.4
7.3
0.6
4.1
0.6
10
11
12
13
14
15
16
Lipase, Pseudomonas cepacia, IM
Lipozyme Mucor miehei, IM
Lipozyme TL IM
Amano lipase AK
Amano lipase PS
Lipase, Candida rugosa
Lipase, porcine pancreas
TM
4.2
13
a
ꢁ1
ꢁ1
Flow reaction at stationary state [25 °C, 0.1 mL min flow of hexane–THF–vinyl acetate 2:1:1 mixture containing 10 mg mL (0.078 M) rac-1b in
TM
enzyme filled CatCart column; sample taken at 60 min after starting]. The ee value of (R)-2b and c were determined by GC.
Enantiomer selectivity (E) was calculated from c and ee(R)-2b
b
39
.
Due to sensitivity to experimental errors, E values calculated in the 100–250 range are
reported as >100, values in the 250–500 range are reported as >200 and values calculated above 500 are given as ꢅ200.
c
Specific reaction rate (r) calculated for the flow system according to Eq. 1.
d
ꢁ1
Shake flask batch reaction at 24 h [25 °C, 1000 rpm, sealed amber glass vial, 2 mL of hexane–THF–vinyl acetate 2:1:1 mixture, 10 mg mL (0.078 M)
rac-1b and enzyme (20 mg)].
Specific reaction rate (r) calculated for the batch system according to Eq. 2.
e

242
C. Csaj a´ gi et al. / Tetrahedron: Asymmetry 19 (2008) 237–246
Table 3. Kinetic resolution of racemic 1-phenylpropan-2-ol (rac-1c) by acylation with vinyl acetate catalyzed by various lipases: panel A: reactions in
continuous-flow packed-bed reactor at stationary state; panel B: reactions in shake flask
a
a
a
b
c
ꢁ1 ꢁ1
No.
Enzyme (in packed-bed column)
c (%)
ee(S)-1c (%)
ee(R)-2c (%)
E
r (lmol min
g
)
Panel A
1
2
3
4
5
6
7
8
CaLB
42
50
24
16
38
6
57.2
92.4
30.8
17.4
61.0
5.7
84.9
94.3
98.9
98.4
98.7
98.6
0.3
23
7.7
9.2
4.4
2.9
7.0
1.1
1.4
0.5
Lipase, Pseudomonas cepacia, IM
Lipozyme Mucor miehei, IM
Lipozyme TL IM
Amano lipase AK
Amano lipase PS
Lipase, Candida rugosa
Lipase, porcine pancreas
>100
>100
>100
>200
>100
1.0
TM
8
3
2.6
2.1
89.4
18
d
d
d
b
e
ꢁ1 ꢁ1
No.
Enzyme (in shake flask)
c
(%)
ee(S)-1c (%)
ee(R)-2c (%)
E
r (lmol min
g
)
Panel B
9
CaLB
71
52
49
49
52
44
16
14
99.2
99.4
93.4
97.1
99.5
80.1
2.4
39.4
89.7
97.9
95.7
89.7
97.6
12.4
87.5
10.7
>100
>200
>100
>100
>100
1.3
12.5
5.6
4.8
4.9
5.5
4.0
1.0
0.9
10
11
12
13
14
15
16
Lipase Pseudomonas cepacia
Lipozyme Mucor miehei
Lipozyme TL IM
Amano lipase AK
Amano lipase PS
Lipase Candida rugosa
Lipase, porcine pancreas
TM
7.6
20
a
ꢁ1
ꢁ1
Flow reaction at stationary state [25 °C, 0.1 mL min flow of hexane–THF–vinyl acetate 2:1:1 mixture containing 10 mg mL (0.073 M) rac-1c in
TM
enzyme filled CatCart column; sample taken at 60 min after starting]. The ee values of (R)-2c and (S)-1c and c were determined by GC.
Enantiomer selectivity (E) was calculated from ee(S)-1c and ee(R)-2c (and confirmed by independent calculation c and ee(R)-2c
b
40
39
,
data not shown). Due to
sensitivity to experimental errors, E values calculated in the 100–250 range are reported as >100, values in the 250–500 range are reported as >200.
Specific reaction rate (r) calculated for the flow system according to Eq. 1.
Shake flask batch reaction at 24 h [25 °C, 1000 rpm, sealed amber glass vial, 2 mL of hexane–THF–vinyl acetate 2:1:1 mixture, 10 mg mL (0.073 M)
c
d
ꢁ1
rac-1c and enzyme (20 mg)].
Specific reaction rate (r) calculated for the batch system according to Eq. 2.
e
tor. To demonstrate the usefulness of this enzyme reactor,
three preparative biotransformations (20 mL solution of
CaLB-filled column (Table 4: E ꢅ 200, >200, and 22 for
rac-1a, rac-1b, and rac-1c, respectively), were quite compa-
rable to the results on shorter term runs (E ꢅ 200,ꢅ200
and 23 in Table 1, No. 1 for rac-1a; Table 2, No. 1 for
rac-1b; Table 3, No. 1 for rac-1c, respectively). The rela-
tively low yields for (S)-1b and (R)-2b are because of a re-
peated chromatographic separation which was necessary
due to the presence of a labile side product (close to the
acetate (R)-2b on normal silica TLC, forming in less than
ꢁ
1
1
0 mg mL rac-1a–c in hexane–THF–vinyl acetate 2:1:1
ꢁ
1
solvent mixture; 0.1 mL min , 25 °C; 1 bar) were carried
out using the same column with a short acetone washing
protocol (ꢀ10 times volume acetone as the dead volume
of the bioreactor) between the runs with different substrates
(
Table 4). The preparative scale kinetic resolution results,
with the racemic alcohols in the commercial immobilized
5% amount) which was not identified. It can be assumed
that this side product was the acetal of the forming acetal-
dehyde with 1b. Negligible traces of similar by-products
were also removed during the chromatographic separations
of the (S)-1a/(R)-2a and (S)-1c/(R)-2c mixtures. In princi-
ple, the formation of these side products can be avoided
by using isopropenyl instead of vinyl acetate. In this case,
however, the cost of the higher purity is a drop in the pro-
ductivity (as the formation of the same acyl enzyme inter-
mediate, which is acylating the racemic alcohol, is slower
from isopropenyl acetate than from vinyl acetate).
Table 4. Preparative kinetic resolutions of racemic secondary alcohols
rac-1a–c with vinyl acetate in immobilized CaLB-filled continuous-flow
reactor
a
b
25
c
d
Compound
Yield (%)
ee (%)
½aꢃD
E
(
(
(
(
(
(
S)-1a
R)-2a
S)-1b
R)-2b
S)-1c
R)-2c
40
48
26
41
34
41
98.5
99.1
77.4
99.0
56.4
85.1
ꢁ62.8
+125.3
+2.0
+7.1
+4.9
ꢅ200
>200
22
ꢁ23.3
a
Yields refer to products isolated from the outflow (20 mL, each reaction)
As the immobilized CaLB exhibited moderate enantiomer
selectivity in the kinetic resolution of racemic 1-phenylpro-
pan-2-ol (rac-1c, E = 22 at 25 °, 1 bar, Table 3, No. 1), this
reaction has been chosen for studying the effect of pressure
and temperature on the kinetic resolution in continuous-
flow mode. In this series, temperature (0–60 °C) and pres-
sure (1–120 bar) were varied. For gaining impression on
the operation stability of the biocatalyst under these condi-
tions, all reactions were performed in the same bioreactor.
ꢁ1
of continuous-flow reactions (10 mg mL rac-1a–c in hexane–THF–
ꢁ1
TM
vinyl acetate 2:1:1 mixture; 0.1 mL min flow; CaLB-filled CatCart
column; 25 °C; 1 bar).
The ee values were determined by enantioselective GC.
Specific rotations: (c 1.0, chloroform).
Enantiomer selectivity (E) was calculated from ee(S)-1 and ee(R)-2
b
c
d
40
.
Due
to sensitivity to experimental errors, E values calculated in the 250–500
range are reported as >200 and values calculated above 500 are given as
ꢅ200.

C. Csaj a´ gi et al. / Tetrahedron: Asymmetry 19 (2008) 237–246
243
After each sub series (variation of the pressure at a given
temperature), the column was tested in the standard room
temperature reaction resulting significantly unchanged
r
r
(1 bar)
(120 bar)
E
E
E
(1 bar)
or
results (25 °C, c ꢀ 42%, ee
ꢀ 57% ee
ꢀ 85%) even
(
S)-1
(R)-2c
-1 -1
r
[µmol min g ]
(120 bar)
after operating at 60 °C and 120 bar. It should be noted
here that such precise recovery of activity with 0.4 g en-
zyme is hardly possible with batch mode reactions due to
the mechanical damage of the biocatalyst even on shaking
and due to the loss of biocatalyst during filtrations.
30
2
2
5
0
Analysis of the data revealed that pressure has no signifi-
cant influence on the productivity (r) or enantiomer selec-
tivity (E) of the reaction (Fig. 3). The apparent slight
variation of the enantiomer selectivity at 0 °C on pressure
(
1–120 bar; E between 22 and 24, the amplitude of this var-
iation is close to the experimental errors) might be due to
the largest changes in the viscosity of the reaction media
at the lowest temperature. In the other cases, pressure
has negligible influence on the productivity (r) or enantio-
mer selectivity (E).
1
0
5
0
r
r
(0 ºC)
(30 ºC)
(0 ºC)
E
0
10
20
30
40
50
60
or
r
[µmol min-1 g-1]
E
Temperature [°C]
E
(30 ºC)
Figure 4. Dependence of the specific reaction rate (r) and enantiomer
selectivity (E) of the kinetic resolution of racemic 1-phenylpropan-2-ol rac-
2
2
1
5
1
c on the temperature in a continuous-flow enzyme reactor filled with
immobilized CaLB at two selected pressures (1 bar and 120 bar).
0
5
enantiomer selectivity (E) with increasing temperature, it
went through a maximum first between 10 and 20 °C
(E = 24 at 20 °C, 1 bar; or E = 26 at 10 °C, 120 bar) and
a minimum later at 50 °C (E = 7 at 50 °C, at 1 or
120 bar). This indicates that the enantiomer selectivity (E)
of this process can be determined by several factors (enthal-
py, entropy, viscosity, diffusion) varying differently with
change in temperature.
5
0
3
. Conclusion
Our results indicate that immobilized as well as lyophilized
powder forms of lipases can be effectively used in continu-
ous-flow mode kinetic resolutions of racemic alcohols rac-
1a–c in non-aqueous solvent systems. In accordance with
the previous findings, the productivity (specific rate, r) of
the lipases was higher in continuous-flow reactors than in
batch mode systems, whereas the enantiomer selectivities
1
20
40
60
80
100
120
Pressure [bar]
Figure 3. Dependence of the specific reaction rate (r) and enantiomer
selectivity (E) of the kinetic resolution of racemic 1-phenylpropan-2-ol rac-
c on the pressure in a continuous-flow enzyme reactor filled with
1
immobilized CaLB at two selected temperatures (0 °C and 30 °C).
(E) were not much different. From a high throughput
On the other hand, temperature had significant effect both
on the productivity (r) and the enantiomer selectivity (E)
screening point of view, it is important that application
of a lipase-filled bioreactor operated in continuous-flow
mode can produce much faster the highly enantiomerically
enriched form of the slower reacting enantiomer in
amounts required for proper characterization (IR, MS,
NMR, specific rotation) than the corresponding batch
mode reactions.
(
Fig. 4). As expected, the specific reaction rate increased
monotonously with the increasing temperature (r = 5.0–
3.0 at 1 bar) and this trend was not influenced by the pres-
1
sure (r = 5.0–12.9 at 120 bar).
Influence of the temperature on the enantiomer selectivity,
however, exhibited more interesting behavior (Fig. 4). In-
stead of the most expected monotonous decrease of the
Investigation of the effect of temperature (0–60 °C) and
pressure (1–120 bar) on the continuous-flow mode kinetic

244
C. Csaj a´ gi et al. / Tetrahedron: Asymmetry 19 (2008) 237–246
resolution of racemic 1-phenylpropan-2-ol (rac-1c) with vi-
nyl acetate in a bioreactor filled with the moderately enan-
tiomer selective immobilized lipase B from C. antarctica
indicated apparently no effect of the pressure on the selec-
tivity (E) and the productivity (r). As expected, productiv-
ity (r) increased monotonously with the temperature.
Surprisingly, no such simple trend regarding the selectivity
4.3. Packed-bed lipase columns
Enzymes listed above (commercial and precipitated prepa-
rations) were packed into stainless steel CatCart columns
TM
according to the proprietary filling process of ThalesNano
Inc. Before packing, the columns were washed with dis-
tilled water, ethanol, n-hexane and acetone in ultrasonic
cleaner. For the continuous flow enzymatic applications,
the columns were sealed by silver metal filter membranes
[Sterlitech Silver Membrane Filter from Sigma–Aldrich,
Z623237, pore size 0.45 lm; pure metallic silver, 99.97%
with no extractable or detectable contaminants] due to
the known benefits of Ag (bacteriostatic). The sealings were
made by PTFE. The filling mass of the enzymes was
(
E) has been found. According to our best knowledge, this
is the first report on a case when the enantioselectivity (E)
of a biotransformation have a local maximum (E ꢀ 25 at
2
1
0 °C, 1 bar) and a local minimum (E = 7 at 50 °C,
bar) within a temperature range.
4
00 ± 30 mg. At least three columns per enzyme (separate
4
. Experimental
columns for each substrate, rac-1a–c) were packed for this
study.
4
.1. Methods
4.4. Enantiomer selective acetylation of racemic alcohols
rac-1a–c in batch mode
The NMR spectra were recorded in CDCl on a Bruker
3
DRX-500 spectrometer and are reported in ppm on the
d scale. GC analyses were carried out on HP 5890 or
Agilent 4890D instruments equipped with FID detector
and Hydrodex-b-PM column (50 m ꢄ 0.25 mm ꢄ 0.25 lm
film with permethylated b-cyclodextrin; Macherey&Nagel)
or Hydrodex-b-6-TBDM column (25 m ꢄ 0.25 mm ꢄ
To a solution of racemic alcohol (rac-1a–c, 20 mg) in hex-
ane–THF–vinyl acetate 2:1:1 mixture (2 mL), the enzyme
(20 mg) was added in a sealed amber glass vial and the
resulting mixture was shaken (1000 rpm) at 25 °C for
2
2
4 h. For GC analysis, samples (10 lL) were taken at 1,
, 8, and 24 h. Results of the batch mode reactions are col-
0
.25 lm film with heptakis-(2,3-di-O-methyl-6-O-t-butyldi-
methylsilyl)-b-cyclodextrine; Macherey&Nagel) using H2
carrier gas (oven: 100 °C, injector: 250 °C, detector: 250,
head pressure: 10 psi, 50:1 split ratio). Optical rotations
were determined on a Perkin Elmer 241 polarimeter.
TLC was carried out on Kieselgel 60F254 (Merck) sheets.
Spots were visualized by treatment with 5% ethanolic phos-
phomolybdic acid solution and heating of the dried plates.
The continuous flow bioreactions were performed by X-
lected in Tables 1B, 2B and 3B.
4
.4.1. GC data for 1a and 2a. tr (50 m Hydrodex b-PM
column, 130 °C, 10 min, 26 psi), min: 7.68, (R)-1a; 7.82,
S)-1a; 8.76, (S)-2a; 8.90, (R)-2a; GC molar response factor
(
of acetate–alcohol (rac-2a/rac-1a): 1.24 [used for calcula-
tions of c and E in Tables 1 and 4.
TM
TM
Cube laboratory flow reactor (X-Cube —trademark of
ThalesNano, Inc.; Ser. No.: 002/2006) equipped with en-
4
.4.2. GC data for 1b and 2b. tr (25 m Hydrodex b-PM
ꢁ1
column, 100–112 °C, 1 °C min , 12 psi), min: 7.5, (R)-1b;
7
TM
TM
zyme filled CatCart
columns (CatCart —registered
.6, (S)-2b; 8.2, (R)-2b; 8.9, (S)-1b; GC molar response fac-
trademark of ThalesNano Inc.: stainless steel (INOX 316
L); inner diameter, 4 mm; total length, 70 mm; packed
length, 65 mm; inner volume, 0.816 mL).
tor of acetate/alcohol (rac-2b/rac-1b): 1.23 [used for calcu-
lations of c and E in Tables 2 and 4].
4
.4.3. GC data for 1c and 2c. tr (Hydrodex 6-b-TBDM
ꢁ1
4
.2. Chemicals and enzymes
column, 80–120°C, 2 °C min , 12 psi) min: 13.9, (S)-2c;
1
4.6, (S)-1c 15.1, (R)-1c; 15.8, (R)-2c; GC molar response
Racemic 1-phenylethanol rac-1a, 1-cyclohexylethanol rac-
b, 1-phenylpropan-2-ol rac-1c and vinyl acetate were
factor of acetate–alcohol (rac-2c/rac-1c): 1.23 [used for cal-
culations of c and E in Tables 3 and 4].
1
products of Aldrich. As standards for GC analysis, racemic
acetates rac-2a–c were prepared by chemical acetylation
4.5. Enantiomer selective acetylation of racemic alcohols
rac-1a–c in continuous-flow bioreactors
(
Ac O, Et N, D, 20 mg scale) from the racemic alcohols
2
3
rac-2a–c. All further chemicals and solvents were of analyt-
ical grade or higher and purchased from Fluka, Aldrich or
Sigma. Lipases for this study are the following: Lipase
from Pseudomonas cepacia (Fluka, 17261; immobilized on
ceramic particles); Lipozyme TL IM (Novozymes; immobi-
lized lipase from Thermomyces lanuginosus); Lipozyme ,
immobilized from M. miehei (Fluka, 62350; trademark of
Novozymes); Lipase (B lipase) from C. antarctica, CaLB
ꢁ
1
The solution of racemic alcohol (10 mg mL of rac-1a–c:
0.082 M for rac-1a, 0.078 M for rac-1b, 0.073 M for rac-
1c) in hexane–THF–vinyl acetate 2:1:1 mixture was
pumped through the enzyme filled column thermostatted
TM
ꢁ1
to 25 °C at a flow rate of 0.1 mL min without choking
(no measurable back-pressure). For checking the station-
ary state, samples were analyzed by GC at every 10 min be-
tween the start and 60 min. After a run, the columns were
(
Sigma, L 4777; immobilized on acrylic resin); Lipase from
ꢁ
1
porcine pancreas, Type II (Sigma, L 3126); Lipase from C.
rugosa, Type VII (Sigma, L 1754); Lipase PS from P. cepa-
cia, Amano (Aldrich, 534641); Lipase AK from Pseudomo-
nas fluorescens, Amano (Aldrich, 534730).
routinely washed with acetone (0.5 mL min , 10 min) and
stored in a refrigerator (4 °C). In this way, most of the col-
umns were used repeatedly (3–20 times) without significant
change in their activity. In Tables 1A, 2A, and 3A, data at

C. Csaj a´ gi et al. / Tetrahedron: Asymmetry 19 (2008) 237–246
245
6
0 min after starting the continuous-flow bioreactions are
4.5.6. (R)-1-Phenylpropan-2-yl acetate, (R)-2c. 22Colorless
shown.
oil; yield: 108.3 mg, 41%; 85.1% ee by GC; ½aꢃ ¼ ꢁ23:3
D
2
D
5
50
(
c 1.0, CHCl ) {lit.: ½aꢃ ¼ ꢁ10:9 (c 1, Et O), 92% ee };
3
2
1
Experiments on the influence of the temperature (0–60 °C)
and the pressure (1–120 bar) on the enzyme catalyzed
kinetic resolution of rac-1c in continuous-flow reactor were
performed similarly, except pressurizing (1–120 bar) and
thermostatting (0–60 °C) the CaLB-filled CatCart column
in the X-Cube reactor (one column was used for all
H NMR: 1.24 (d, J = 6.4 Hz, 3H), 2.01 (s, 3H), 2.78
(dd, J1 = 6.4 Hz, J2 = 13.2 Hz, 1H), 2.93 (dd,
J = 6.4 Hz, J = 13.2 Hz, 1H), 5.05 (m, 1H), 7.22 (m,
1
2
3H), 7.30 (m, 2H). NMR data agreed with the spectra
TM
49
reported for racemic 1-phenylpropan-2-yl acetate.
TM
measurements in this series). In these experiments, samples
were analyzed at 30 min after starting the reactions. The
results of these experiments are shown in Figures 3 and 4.
Acknowledgment
L.P. thanks the financial support from Hungarian OTKA
Foundation (T-048854).
For preparative scale kinetic resolutions, 20 mL solution of
ꢁ
1
the racemic alcohol (10 mg mL of rac-1a–c: 0.082 M for
rac-1a, 0.078 M for rac-1b, 0.073 M for rac-1c) in hexane–
THF–vinyl acetate 2:1:1 mixture was pumped through the
TM
CaLB-filled CatCart column (one column regenerated
References
with acetone as described in the first section was used for
all measurements in this series) thermostatted to 25 °C at
1
. Faber, K. Biotransformations in Organic Chemistry, 4th ed.;
Springer: Berlin, 2004.
ꢁ1
a flow rate of 0.1 mL min with no measurable back pres-
sure. The solvent was distilled off from the collected solu-
tion by rotary evaporation and the residue was separated
by chromatography (silica gel, hexane–acetone 10:0.5, v/
v) to give alcohol [(S)-1a–c] and acetate [(R)-2a–c]. The re-
sults of these experiments are collected in Table 4.
2. Rehm, H.-J.; Reed, G.; P u¨ hler, A.; Stadler, P.; Kelly, D. R.,
2nd ed. In Biotechnology: Biotransformations I and II; Wiley-
VCH: Weinheim, 1998; Vols. 8a and 8b.
3
. Poppe, L.; Nov a´ k, L. Selective Biocatalysis: A Synthetic
Approach; VCH: Weinheim-New York, 1992.
. Industrial Biotransformations; Liese, A., Seelbach, K., Wan-
drey, C., Eds., 2nd ed.; Wiley-VCH: Weinheim-New York,
4
4
.5.1. (S)-1-Phenylethanol, (S)-1a. Colorless oil; yield:
22
2
006.
8
0.0 mg, 40%; 98.5% ee by GC; ½aꢃ ¼ ꢁ62:8 (c 1.0,
D
5. Panke, S.; Held, M.; Wubbolts, M.; Schmid, A.; Hollmann,
F.; Park, J. B.; B u¨ hler, B. Curr. Opin. Biotechnol. 2004, 15,
272–279.
6. Bornschauer, U. T.; Kazlauskas, R. J. Hydrolases in Organic
Synthesis: Regio- and Stereoselective Biotransformations;
Wiley-VCH: Weinheim-New York, 2006.
2
D
2
36
CHCl ) {lit.: ½aꢃ ¼ ꢁ57:7 (c 1.0, CHCl ), 89% ee ;
3
3
2
5
42
½
aꢃ ¼ ꢁ55:1 (c 1.63, CHCl ) ; [a] = ꢁ53.5 (c 1.13,
D
3
D
4
3
CHCl ) }; NMR data agreed with the reported
3
3
6,43
spectra.
7
. Ghanem, A.; Aboul-Enein, H. Y. Chirality 2005, 17, 1–15.
4
.5.2. (R)-1-Phenylethyl acetate, (R)-2a. Colorless oil;
22
8. Turner, N. J. Curr. Opin. Chem. Biol. 2004, 8, 114–119.
9. Arcos, J. A.; Garcia, H. S.; Hill, C. G. Biotechnol. Bioeng.
2000, 68, 563–670.
10. Rønne, T. H.; Yang, T.; Mu, H.; Jacobsen, C.; Xu, X. J.
Agric. Food Chem. 2005, 53, 5617–5624.
yield: 129.3 mg, 48%; 99.1% ee by GC; ½aꢃ ¼ þ125:3 (c
D
4
4
1
.0, CHCl ) {lit.:[a] = + 105.1 (c 1.3, CHCl ) ;
3 D 3
22
1
36
½
aꢃ ¼ þ125 (c 1.0, CHCl ), 97% ee }; H NMR: 1.57
D
3
(
d, J = 6.9 Hz), 2.10 (s, 3H), 5.93 (q, J = 6.9 Hz, 1H),
7
.30–7.40 (m, 5H); NMR data agreed with the reported
1
1. Kim, B. H.; Akoh, C. C. J. Agric. Food Chem. 2006, 54, 5132–
5141.
3
6
spectra.
1
2. Gonz a´ lez Moreno, P. A.; Robles Medina, A.; Camacho
Rubio, F.; Camacho P a´ ez, B.; Molina Grima, E. Biotechnol.
Prog. 2004, 20, 1044–1052.
3. Al-Zuhair, S. Biotechnol. Prog. 2005, 21, 1442–1448.
14. Royon, D.; Daz, M.; Ellenrieder, G.; Locatelli, S. Bioresour.
Technol. 2007, 98, 648–653.
4
.5.3. (S)-1-Cyclohexylethanol, (S)-1b. 22Colorless oil;
yield: 53.8 mg, 26%; 77.4% ee by GC; ½aꢃ ¼ þ2:0 (c 1.0,
D
2
D
0
45
1
CHCl ) {lit.: ½aꢃ ¼ þ3:5 (c 3.1, CHCl ) }; NMR data
3
3
4
6
agreed with the reported spectra.
1
5. Dossat, V.; Combes, D.; Marty, A. J. Biotechnol. 2002, 97,
17–124.
4
.5.4. (R)-1-Cyclohexylethyl acetate, (R)-2b. Colorless oil;
1
22
yield: 110.8 mg, 41%; 99.0% ee by GC; ½aꢃ ¼ þ7:1 (c 1.0,
D
16. Yang, T.; Rebsdorf, M.; Engelrud, U.; Xu, X. J. Agric. Food
Chem. 2005, 53, 1475–1481.
17. Murty, V. R.; Bhat, J.; Muniswaran, P. K. Biotechnol. Lett.
2004, 26, 563–567.
2
D
0
47
22
D
CHCl ) {lit.: ½aꢃ ¼ þ6:6 (c 2.6, CHCl ) ; ½aꢃ ¼ þ7:2 (c
3
3
3
6
1
1
3
1
.0, CHCl ), 97% ee }; H NMR: 1.09 (d, J = 6.4 Hz,
3
H), 1.65 (m, 11H), 1.96 (s, 3H), 4.75 (q, J = 6.4 Hz,
H). NMR data agreed with the reported spectra.36
18. Jeganathan, J.; Nakhla, G.; Bassi, A. J. Hazard. Mater. 2007,
1
45, 127–135.
1
9. Patel, R. N.; Banerjee, A.; Szarka, L. J. J. Am. Oil Chem. Soc.
1996, 73, 1363–1375.
4
6
{
.5.5. (S)-1-Phenylpropan-2-ol, (S)-1c. Colorless oil; yield:
22
8.5 mg, 34%; 56.4% ee by GC; ½aꢃ ¼ þ4:9 c 1.0, CHCl )
D
3
2
D
5
48
1
20. Chen, J. C.; Tsai, S. W. Biotechnol. Prog. 2000, 16, 986–992.
½aꢃ ¼ þ41:8 (c 1, benzene) }; H NMR: 1.23 (d,
2
1. Ujang, Z.; Husain, W. H.; Seng, M. C.; Rashid, A. H. A.
J = 6.2 Hz, 3H), 2.05 (br s, 1H), 2.69 (dd, J = 6.3 Hz,
J = 12.9 Hz, 1H), 2.82 (dd, J = 6.3 Hz, J = 12.9 Hz,
1
1
Process Biochem. 2003, 38, 1483–1488.
2
1
2
22. Sanchez, A.; Valero, F.; Laufente, J.; Sola, C. Enzyme
Microb. Technol. 2000, 27, 157–166.
H), 4.07 (m, 1H), 7.25 (m, 3H), 7.33 (m, 2H); NMR data
agreed with the spectra reported for racemic 1-phenylpro-
pan-2-ol.
23. Mohapatra, S. C.; Hsu, J. T. Biotechnol. Bioeng. 1999, 64,
213–220.
4
9

246
C. Csaj a´ gi et al. / Tetrahedron: Asymmetry 19 (2008) 237–246
2
2
2
2
4. Xin, J. Y.; Li, S. B.; Xu, Y.; Wang, L. L. Biotechnol. Bioeng.
000, 68, 78–83.
5. Battistel, E.; Bianchi, D.; Cesti, P.; Pina, C. Biotechnol.
Bioeng. 1991, 38, 659–664.
6. Sekyoo, J.; Hwang, B. Y.; Juhan, K.; Kim, B. G. J. Mol.
Catal. B: Enzym. 2000, 10, 597–604.
7. Fishman, A.; Eroshov, M.; Dee-Noor, S. S.; Van Mil, J.;
Cogan, U.; Effenberger, R. Biotechnol. Bioeng. 2001, 74, 256–
37. Kiss, V.; Egri, G.; B a´ lint, J.; Ling, I.; Bark o´ czi, J.; Fogassy,
2
E. Tetrahedron: Asymmetry 2006, 17, 2220–2234.
38. B o´ dai, V.; Peredi, R.; B a´ lint, J.; Egri, G.; Nov a´ k, L.; Szak a´ cs,
G.; Poppe, L. Adv. Synth. Catal. 2003, 345, 811–818.
39. Chen, C. S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am.
Chem. Soc. 1982, 104, 7294–7299.
40. Rakels, J. L. L.; Straathof, A. J. J.; Heijnen, J. J. Enzyme
Microb. Technol. 1993, 15, 1051–1056.
2
63.
8. Chualee, S.; Roji Sarmidi, M. J. Mol. Catal. B: Enzym. 2004,
8, 111–119.
41. Tomotani, E. J.; Vitolo, M. J. Food Eng. 2007, 80, 662–
667.
42. Nakamura, K.; Masuda, T. J. Org. Chem. 1998, 63, 8957–
8964.
43. Nakamura, K.; Fujii, M.; Ida, Y. J. Chem. Soc., Perkin
Trans. 1 2000, 19, 3205–3211.
2
2
3
3
3
2
9. Indlekofer, M.; Brotz, F.; Bauer, A.; Reuss, M. Biotechnol.
Bioeng. 1996, 52, 459–471.
0. Berendsen, W. R.; Lapin, A.; Reuss, M. Chem. Eng. Sci.
2
007, 62, 2375–2385.
1. Yanagishita, H.; Sakaki, K.; Hirata, H. J. Oleo Sci. 2007, 56,
37–148.
44. Faraldos, J.; Arroyo, E.; Herradon, B. Synlett 1997, 367–
370.
45. Li, G.; Kabalka, G. W. J. Organomet. Chem. 1999, 581, 66–
69.
46. Janssen, A. J. M.; Klunder, A. J. H.; Zwanenburg, B.
Tetrahedron 1991, 47, 7645–7662.
47. Naemura, K.; Fukuda, R.; Murata, M.; Konishi, M.; Hirose,
K.; Tobe, Y. Tetrahedron: Asymmetry 1995, 6, 2385–2394.
48. Pickord, R. H.; Kenyon, J. J. Chem. Soc. 1914, 1115–
1117.
1
2. Frings, K.; Koch, M.; Hartmeier, W. Enzyme Microb.
Technol. 1999, 25, 303–309.
3
3
3. Jas, G.; Kirchning, A. Chem. Eur. J. 2003, 9, 5708–5723.
4. Kirchning, A.; Solodenko, W.; Mennecke, K. Chem. Eur. J.
2
006, 12, 5972–5990.
3
5. Woodley, J. M.; Lilly, M. D. In Appl. Biocatal.; Cabral, J. M.
S., Best, D., Boros, L., Tramper, J., Eds.; Harwood
Academic: London, 1994; pp 371–393.
49. Hatzakis, N. S.; Smonou, I. Bioorg. Chem. 2005, 33, 325–
337.
50. Bianchi, D.; Cesti, P.; Battistel, E. J. Org. Chem. 1988, 53,
5531–5534.
3
6. Nagy, V.; To}ke, E. R.; Chee Keong, L.; Szatzker, G.;
Ibrahim, D.; Che-Omar, I.; Szak a´ cs, G.; Poppe, L. J. Mol.
Catal. B: Enzym. 2006, 39, 141–148.