Simple and effective method for the deracemization of ethyl 1-hydroxyphosphinate using biocatalysts with lipolytic activity

Article abstract of DOI:10.1016/j.tetasy.2006.10.041

The kinetic resolution of ethyl (1-hydroxyethyl)phenylphosphinate was carried by enzymatic esterification with vinyl butyrate or by hydrolysis of ethyl (1-butyryloxyethyl)phenylphosphinate either by the use of lipases or whole cells of microorganisms (bac

Full text of DOI:10.1016/j.tetasy.2006.10.041

Tetrahedron: Asymmetry 17 (2006) 2870–2875
Simple and effective method for the deracemization of ethyl
1-hydroxyphosphinate using biocatalysts with lipolytic activity
Paulina Majewska,* Paweł Kafarski and Barbara Lejczak
Wrocław University of Technology, Department of Bioorganic Chemistry, Wybrzez_e Wyspian´skiego 27, 50-370 Wrocław, Poland
Received 11 August 2006; accepted 26 October 2006
Abstract—The kinetic resolution of ethyl (1-hydroxyethyl)phenylphosphinate was carried by enzymatic esterification with vinyl butyrate
or by hydrolysis of ethyl (1-butyryloxyethyl)phenylphosphinate either by the use of lipases or whole cells of microorganisms (bacteria
and fungi). Since no stereodifferentiation of the phosphinate moiety was observed, the biotransformations gave diastereoisomers of ethyl
(1-hydroxyethyl)phenylphosphinate with excellent enantiomeric excesses above 98%, while the percentage conversion ranged from 10%
to 49% depending upon the procedure and conditions of the biocatalysis.
ꢀ 2006 Elsevier Ltd. All rights reserved.
1. Introduction
oxyethyl)phenylphosphinate 2 and transesterification of 1
using lipases of different origin.
Green chemistry is a field of science, which has developed
very dynamically over the last few years. One of its major
techniques is the application of biocatalysts, namely
enzymes or whole cells (mostly microbial ones), to obtain
desired products of various interests.1 A range of methods
are based upon the improvement of classic kinetic resolu-
tion processes. An example of such a process is deracemiza-
tion, followed by repeated resolution, dynamic resolution
and follow-up stereoinversion reactions.2
2. Results and discussion
All the three tested methods of the biotransformation of
ethyl (1-hydroxyethyl)phenylphosphinate appeared to be
a simple and effective means for the resolution of its
stereoisomers.
Enzymatic hydrolysis of compound 2 (Scheme 1) gave the
major pair of diastereoisomers (SP,S) and (RP,S) of com-
pound 1 with excellent enantiomeric excess >98% and the
maximum possible yield of 50% (Table 1).
a-Heteroatom substituted phosphinic acids are of particu-
lar interest due to their usefulness in the development of
catalytic antibodies3 and pharmacologically active sub-
stances.4 Therefore, there is a growing interest in the pre-
paration of such phosphonates in enantiomerically pure
forms and biotransformations are one of the most promis-
ing means to reach that objective.
Experimental data showed that hydrolysis of the discussed
substrate stopped when the chemical yield reached a value
of 50%, which is typical for a kinetic resolution. The only
exception was observed when lipase from Rhizopus sp. was
used as a biocatalyst. In this case, the reaction progress
exceeded 50%, and consequently the enantiomeric purity of
the product was eroded (Table 1). However, a satisfactory
result was obtained when shortening the reaction time to
18 h.
The purpose of this study was to obtain, biocatalytically,
ethyl (1-hydroxyethyl)phenylphosphinate 1 with good
enantiomeric excess in the most simple and effective way
while offering the possibility of scale up. Two different
strategies were examined. Hydrolysis of ethyl (1-butyryl-
Hydrolysis of compound 2 (Scheme 1) by whole cells of
the chosen microorganisms gave the same (SP,S) and
(RP,S) pair of diastereoisomeric ethyl (1-hydroxyethyl)-
phenylphosphinate 1 with good enantiomeric excess and
*
Corresponding author. Tel.: +48 713202597; e-mail: paulina.
majewska@pwr.wroc.pl
0957-4166/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2006.10.041

P. Majewska et al. / Tetrahedron: Asymmetry 17 (2006) 2870–2875
2871
O
P
OBut
C
O
P
OBut
+
C
EtO
Ph
H
Ph
H
CH3
EtO
CH3
O
P
OBut
(
RP
,
R
)-2
2
R)-
(
SP
,
Lipase
C
H
CH3
EtO
or microorganisms`
cells
Ph
O
P
OH
C
O
P
OH
C
+
EtO
CH3
Ph
CH3
2
Ph
H
EtO
H
(SP
,S)-1
(
RP
,
S
)-1
Scheme 1.
of the biocatalytic process. Summing up, the results pre-
sented in this work clearly show that lipases and micro-
organisms are able to convert substrates bearing an
(S)-configuration at the carbon atom adjacent to the phos-
phorus, whereas the configuration of the phosphinate atom
is meaningless here.
Table 1. Results of the enzymatic hydrolysis of ethyl 1-
butyryloxyphosphinate
Lipase
Time
(h)
Conversion
(%)
ee of alcohol
(%)
(SP,R) (RP,R) (RP,S) (SP,S)
(RP,S) (SP,S)
Rhizopus niveus
24
95
32
34
32
34
>98
>98
>98
>98
2.1. Determination of the absolute configuration
Rhizopus sp.
18
24
95
46
54
65
47
54
55
>98
61
45
>98
72
74
Spectroscopic data for the two stereoisomers of compound
1 (Fig. 1) were described previously in the literature.5 The
authors of this paper defined the pair of isomers shown
in Figure 1 as (RP,R)- and (SP,S)-isomers. However, our
studies indicate that this configurational assignment is
not correct.
Mucor javanicus
24
95
49
46
49
48
>98
>98
>98
>98
Pseudomonas cepaciaa 95
Mucor circinelloidesa
10
40
46
40
>98
>98
>98
>98
21
a Immobilized lipase.
O
P
OH
C
O
P
OH
C
Ph
H
EtO
CH3
EtO
CH3
Ph
H
satisfactory chemical yield. However, the results compara-
ble to the enzymatic hydrolysis were obtained only when
Bacillus subtilis cells were used as a biocatalyst (Table 2).
Figure 1. Two isomers of the ethyl (1-hydroxyethyl)phenylphosphinate 1
obtained by Shioji.5
The reverse reaction, an enzymatic esterification of
compound 1 (Scheme 2), allowed us to obtain the pair of
(SP,S)- and (RP,S)-isomers of compound 2 with enantio-
meric excesses above 98% (Table 3). The stereochemistry
of the phosphorus atom had no influence on the course
It is obvious that the proper evaluation of the absolute
configuration at the asymmetric phosphorus atom should
consider the double-bonded oxygen as a priority over the
ester oxygen atom. If so, the correct configuration of the
discussed isomers has to be set as (SP,R) and (RP,S), not
(RP,R) and (SP,S). The pair of the stereoisomers (shown
in Fig. 1 after Shioji et al.)5 of compound 1 was obtained
according to the Haynes procedure,6 which results in the
steric course of this reaction. Their configurational assign-
ment additionally strengthens our opinion about incorrect
configurational assignment being given in the literature. If
this is so, it is possible to link configuration on the basis of
31P NMR spectra of compound 1 as shown in Figure 2.
This assignment is also confirmed by the differences in
the 1H and 13C NMR spectra observed between two diaste-
reomeric pairs. These spectra for (SP,R)- and (RP,S)-
isomers (Fig. 2) are identical with those described by
Shioji.5
Table 2. Results of hydrolysis of ethyl 1-butyryloxyphosphinate by
microorganisms
Microorganism
Time
(h)
Conversion
(%)
ee of alcohol
(%)
(SP,R) (RP,R) (RP,S) (SP,S)
(RP,S) (SP,S)
Pseudomonas aeruginosa 22
20
32
7
17
>98
>98
>98
68
71
Bacillus subtilis
20
24
30
41
33
41
>98
>98
>98
79
Acinetobacter baumannii 48
38
60
36
43
>98
20
68
66
72
The absolute configuration was additionally analyzed by
the method reported by Mosher.7 Thus, a purified mixture
Serratia liquefaciens
94
17
9
49
39

2872
P. Majewska et al. / Tetrahedron: Asymmetry 17 (2006) 2870–2875
O
P
OH
C
O
P
OH
C
+
EtO
H
Ph
H
Ph
CH3
EtO
CH3
O
P
OH
C
1
(
RP
,
R
)-
(
SP
,R
)-1
Lipase
OBut
EtO
H
Ph
CH3
O
P
OBut
C
O
P
OBut
C
+
Ph
CH3
EtO
CH3
1
EtO
H
Ph
H
2
,S)-
(SP
,S
)-2
(RP
Scheme 2.
Table 3. Results of the enzymatic esterification of ethyl 1-hydroxyphosphinate
Lipase
Enzyme amount (mg)
Time (h)
Conversion (%)
ee of ester (%)
(SP,R) (RP,S)
(RP,R) (SP,S)
(RP,S)
(SP,S)
Candida cylindracea
Aspergillus niger
Rhizopus niveus
200
200
200
20
20
20
72
168
168
96
24
24
44
44
<5
48
40
26
43
42
44
<5
46
45
43
47
91
91
—
38
>98
>98
>98
79
>98
—
Mucor javanicus
51
Porcine pancreas
Pseudomonas cepaciaa
Mucor circinelloidesa
>98
>98
>98
20
24
a Immobilized lipase.
using S(+)MTPA-Cl (chemical yield of 85%). After the
reaction was complete, the mixture of products consisted
1
of four isomers of (R)-MTPA ester 3. H NMR chemi-
cal shifts of the methyl groups of Mosher ester of hydroxy-
phosphonate (CH3–CH–P) of these isomers were assigned
as follows: (SP,S,R)—1.37; (RP,R,R)—1.43; (RP,S,R)—
1.43; (SP,R,R)—1.53. In this study, a mixture of isomers
of the ratio 0.53:0.18:1.00:0.35 was used. The signals from
this methyl group of (S)-configuration are situated upfield
compared to (R)-isomers. This derives from the anisotropic
effect exerted by the phenyl group (Fig. 3) and thus derives
directly from the Mosher model. Since there are overlap-
ping signals at 1.43 ppm, the additional assignment was
made using 2D 1H–31P HMQC as a proof. 1H NMR chem-
ical shifts of phosphinate ester methyl group (CH3–CH2–
O–P) were as follows: (SP,S,R)—1.34; (RP,R,R)—1.29;
(RP,S,R)—1.32; (SP,R,R)—1.26. Thus, the signals of this
group of (R)-configuration are situated upfield when com-
pared with the respective (S)-isomers (Fig. 3).
The absolute configuration of the asymmetric carbon atom
of compound 1 (obtained by hydrolysis or esterification),
which was previously defined using the Mosher method,
was then assigned in 31P NMR spectra recorded with
the addition of quinine as a chiral discriminator (Figs. 4
and 5).
Figure 2. 31P NMR spectra of ethyl 1-hydroxyethylphenylphosphinate
diastereomers.
According to both 31P NMR spectra, recorded with the
addition of quinine, it can be concluded that the signals
deriving from the (R)-isomers at the phosphorus atom
are situated upfield compared to the corresponding (S)-iso-
mers (Figs. 4 and 5).
of isomers of compound 1 (molar ratio of: (SP,R):
(RP,S):(SP,S):(RP,R) = 0.35:1.00:0.53:0.18), obtained after
hydrolysis by the lipase from Rhizopus sp., was acylated

P. Majewska et al. / Tetrahedron: Asymmetry 17 (2006) 2870–2875
2873
O
P
O
P
O
O
EtO
Ph
Ph
EtO
OEt
Ph
Ph
OEt
Me
Ph
Me
Ph
P
P
Me
Me
MeO
upfield
relative to
upfield
relative to
MeO
Ph
MeO
MeO
Ph
CF3
CF3
CF3
)-3
CF3
(RP
,
S
,
R
)-3
(SP
,
R,
R
)-3
(SP
,
S,
R
(RP
,
R,
R
)-3
O
P
O
P
O
O
EtO
Ph
Ph
EtO
OEt
Ph
Ph
Me
Ph
Me
Ph
P
P
OEt
Me
Me
upfield
relative to
upfield
relative to
MeO
Ph
MeO
MeO
Ph
MeO
CF3
CF3
)-3
CF3
RP
CF3
)-3
(SP
,
R,
R
)-3
(RP
,
S,
R
(
,
R
,
R
)-3
(SP
,
S,
R
Figure 3. Signals used for the determination of absolute configuration of a-carbon atom of ethyl (1-hydroxyethyl)phenylphosphinate 1 by Mosher’s
procedure.
Figure 4. 1H NMR spectrum of 1 with quinine after hydrolysis of 2 by
lipase from Rhizopus sp.—24 h.
Figure 5. 1H NMR spectrum of 1 with quinine after esterification of 1 by
lipase from Mucor javanicus—68 h.
3. Experimental
(Sigma), Aspergillus niger (Fluka), Rhizopus niveus (Fluka),
Rhizopus sp. (Serva), Mucor javanicus (Fluka), Penicillium
roqueforti (Fluka) and porcine pancreas (Sigma). Also
lipase from Pseudomonas cepacia immobilized in Sol–gel-
AK (Fluka) and that from Mucor circinelloides immobi-
lized in situ. NMR spectra were recorded in CDCl3 using
Bruker AvanceTM 600 operating at 600.58 MHz for 1H,
243.12 MHz for 31P and 151.02 MHz for 13C or using
3.1. Materials and methods
All materials were obtained from commercial suppliers:
Sigma, Aldrich, Fluka, POCh, Serva, and used without
purification. LC was performed on silica gel 60 (60–230
mesh). The sources of lipases were Candida cylindracea

2874
P. Majewska et al. / Tetrahedron: Asymmetry 17 (2006) 2870–2875
Bruker Avance DRX 300 system operating at 300.13 MHz
J = 121.2, CHP), 128.56, 128.64, 132.40, 132.47 (aromatic),
132.84 (d, J = 2.2, aromatic), 172.36 (CO); for (RP,R) and
(SP,S) isomers—31P NMR (243.1 MHz, CDCl3): d 37.53;
1H NMR (600 MHz, CDCl3): d 0.82 (t, J = 7.4 Hz, 3H,
CH2CH2CH3), 1.32 (t, J = 7.0 Hz, 3H, OCH2CH3), 1.42
(dd, J = 7.1 15.5 Hz, 3H, CHCH3), 1.47–1.54 (m, 2H,
CH2CH2CH3), 2.13–2.23 (m, 2H, CH2CH2CH3), 3.99–
4.22 (m, 2H, OCH2CH3), 5.45–5.50 (m, 1H, CHP), 7.47–
7.51 (m, 2H, aromatic), 7.57–7.60 (m, 1H, aromatic),
7.80–7.84 (m, 2H, aromatic); 13C NMR (151.0 MHz,
CDCl3): d 13.42 (CH2CH2CH3), 14.25 (OCH2CH3), 16.47
(d, J = 6.1, PCHCH3), 18.28 (CH2CH2CH3), 36.01
(CH2CH2CH3), 61.40 (d, J = 6.5 Hz, POCH2), 65.94 (d,
J = 123.4 Hz, CHP), 128.45, 128.53, 132.50, 132.56 (aro-
matic), 132.79 (d, J = 3.0, aromatic) 172.36 (CO).
1
1
for H and 121.50 MHz for 31P. TMS for H spectra and
CHCl3 for 13C spectra were used as internal standards,
while 85% phosphoric acid in H2O was used as the external
standard for 31P measurements.
3.2. Synthetic procedures and spectral data of the chemical
compounds
3.2.1. Ethyl (1-hydroxyethyl)phenylphosphinate 1. Com-
pound 1 was synthesized according to the Texier-Boullet
method:8 5 g of aluminium oxide was ground with 5 g of
potassium fluoride. Then 20 mmol of ethylphenylphosphi-
nate and 20 mmol of the acetic aldehyde were added and
the reaction mixture left at room temperature for 48 h.
The mixture was then extracted with methylene dichloride
(4 · 50 ml). Product 1 was purified by means of silica gel
liquid chromatography using methylene dichloride and
ethyl acetate (5:3 v/v) as eluent (Rf (1) = 0.15). This was
obtained with 54% yield as a mixture of diastereomers in
a 5:3 molar ratio. For isomers (RP,S) and (SP,R)—31P
3.3. Enzymatic reactions—general procedures
Enzymatic hydrolysis of ethyl (1-butyryloxyethyl)phenyl-
phosphinate 2 was carried out in a biphasic system
(3.8 ml) consisting of 0.05 M phosphate buffer, pH 7
(3.0 ml) and a mixture of diisopropyl ether (0.2 ml) with
n-hexane (0.6 ml). After the addition of 0.2 mM of sub-
strate and 100 mg of suitable lipase (see Table 1) the reac-
tions were carried out at room temperature with shaking
(150 rpm). The reaction was stopped after certain periods
of time, and the product was extracted twice with 15 ml
of ethyl acetate and the organic phase dried over anhy-
drous magnesium sulfate. After filtration, the organic sol-
vent was removed by evaporation and the obtained ethyl
phosphinate analyzed with 31P NMR using quinine as a
chiral discriminator.9
NMR (243.1 MHz, CDCl3):
d
41.82; 1H NMR
(600 MHz, CDCl3): d 1.32 (dd, J = 6.9, 17.6 Hz, 3H,
CHCH3), 1.34 (t, J = 7.1 Hz, 3H, CH2CH3), 3.96–4.06
(m, 2H, OCH2), 4.13–4.20 (m, 1H, CHP), 7.46–7.50 (m,
2H, aromatic), 7.55–7.58 (m, 1H, aromatic), 7.78–7.85
(m, 2H, aromatic); 13C NMR (151.0 MHz, CDCl3): d
16.57 (d, J = 5.9, CHCH3), 17.09 (d, J = 3.4, CH2CH3),
61.55 (d, J = 7.5, OCH2), 66.12 (d, J = 115.5, CHP),
128.48, 128.55, 132.56, 132.63 (aromatic); For isomers
(RP,R) and (SP,S)—31P NMR (243.1 MHz, CDCl3): d
1
41.16; H NMR (600 MHz, CDCl3): d 1.43 (dd, J = 7.1
16.5 Hz, 3H, CHCH3), 1.33 (t, J = 7.0 Hz, 3H, CH2CH3),
4.13–4.20 (m, 3H, CHP and OCH2), 7.46–7.50 (m, 2H,
aromatic), 7.55–7.58 (m, 1H, aromatic), 7.78–7.85 (m,
2H, aromatic); 13C NMR (151.0 MHz, CDCl3): d 16.43,
16.53 (d, J = 6.3 Hz, CHCH3), 17.09, 61.48 (d, J =
7.3 Hz, CH2CH3), 65.89 (d, J = 116.5 Hz, OCH2), 128.49,
128.57, 132.43, 132.49, 132.56 (aromatic).
Enzymatic transesterification of ethyl (1-hydroxy-
ethyl)phenylphosphinate 1 was carried out in diisopropyl
ether (2 ml) with the addition of 20 mg of powdered mole-
˚
cular sieves (3 A mesh). 0.02 mM of the substrate, 20 or
200 mg of suitable lipase and 0.165 mM of vinyl butyrate
were added (see Table 2). The reactions were carried out
at 36 ꢁC in a shaker (150 rpm). The reaction was stopped
after certain periods of time by filtration followed by evap-
oration of the organic layer. The resulting product was
purified by HPLC (C-18 column, gradient: 40% acetonitrile
in water–70% acetonitrile in water, retention time of 2:
11.2 min) and analyzed by HPLC (CHIRALPAK AD,
Diacel, 10% 2-propanol in n-hexane, retention time:
(SP,S)—7.6 min; (RP,R)—8.2 min; (RP,S)—8.5 min and
(SP,R)—9.7 min).
3.2.2. Ethyl (1-butyryloxyethyl)phenylphosphinate 2.
Compound 1 was converted to 2 according to the following
procedure: 10 mmol of 1 was added to 100 ml of the reac-
tion media containing chloroform and triethylamine (10:1
v/v), followed by the addition of 11 mmol of butyryl chlo-
ride. The resulting mixture was stirred for 24 h at room
temperature. a-Butyryloxyphosphinate was purified by
means of column chromatography as described above (Rf
(2) = 0.51). A mixture of diastereomers in a 5:3 molar ratio
was obtained with 45% yield. For (SP,R)- and (RP,S) iso-
3.4. Microorganisms, growth and whole cell biotransforma-
tion conditions
1
mers—31P NMR (243.1 MHz, CDCl3): d 37.11; H NMR
(600 MHz, CDCl3):
d
0.89 (t, J = 7.4 Hz, 3H,
CH2CH2CH3), 1.36 (t, J = 7.0 Hz, 3H, OCH2CH3), 1.42
(dd, J = 7.1, 15.5 Hz, 3H, CHCH3), 1.55–1.61 (m, 2H,
CH2CH2CH3), 2.21–2.31 (m, 2H, CH2CH2CH3), 3.99–
4.22 (m, 2H, OCH2CH3), 5.31–5.35 (m, 1H, CHP), 7.47–
7.51 (m, 2H, aromatic), 7.57–7.60 (m, 1H, aromatic),
7.80–7.84 (m, 2H, aromatic); 13C NMR (151.0 MHz,
CDCl3): d 13.53 (CH2CH2CH3), 14.21 (OCH2CH3), 16.58
(d, J = 5.9, PCHCH3), 18.28 (CH2CH2CH3), 36.04
(CH2CH2CH3), 61.54 (d, J = 6.6, POCH2), 66.81 (d,
Pseudomonas aeruginosa, Bacillus subtilis, Serratia liquefac-
iens and Acinetobacter baumannii are from our own collec-
tion. These microorganisms were identified by Deutsche
Sammlung fur Mikroorganismen und Zellkulturen, Braun-
¨
schweig, Germany. In order to assure the most vigorous
growth and the highest lipolytic activity, several media
were tested. The medium of choice consisted of 10 g of
starch soluble, 1 g of yeast extract, 5 g of (NH4)2SO4, 2 g
of K2HPO4, 100 ll of tributyrin and 1000 ml of distilled

P. Majewska et al. / Tetrahedron: Asymmetry 17 (2006) 2870–2875
2875
water. Microorganisms were incubated at 26 ꢁC with shak-
References
ing at 150 rpm for 3 days after which the cells were centri-
fuged at 3000 rpm for 10 min and washed twice in 0.017 M
phosphate buffer, pH 7.0. Biotransformations were per-
formed in 100 ml solution of 0.017 M phosphate buffer,
pH 7.0, after addition of 50 ll of substrate with shaking
at 150 rpm at room temperature. After biotransformation,
the biomass was centrifuged, the supernatant was extracted
twice with ethyl acetate and dried over anhydrous magne-
sium sulfate. After filtration, the organic solvent was evap-
orated and the product analyzed by means of 31P NMR
spectroscopy using quinine as a chiral discriminator.9
´
´
1. Kiełbasinski, P.; Omelanczuk, J.; Mikołajczyk, M. Tetra-
hedron: Asymmetry 1998, 9, 3283.
2. Strauss, U. T.; Felfer, U.; Faber, K. Tetrahedron: Asymmetry
1999, 10, 107.
3. Martin, M. T.; Angeles, T. S.; Sugasawara, R.; Aman, N. I.;
Napper, A. D.; Darsley, M. J.; Sanchez, R. I.; Booth, P.;
Titmas, R. C. J. Am. Chem. Soc. 1994, 116, 6508.
4. Collinsova, M.; Jiracek, J. J. Curr. Med. Chem. 2000, 7, 629.
5. Shioji, K.; Toshiro, A.; Shibata, S.; Okuma, K. Tetrahedron
Lett. 2003, 44, 1103.
6. Haynes, R. K.; Lam, W. W.-L.; Yeung, L.-L. Tetrahedron
Lett. 1996, 37, 4729.
7. Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512.
8. Texier-Boullet, F.; Lequitte, M. Tetrahedron Lett. 1986, 27,
3515.
Acknowledgement
_
´
´
9. Zymanczyk-Duda, E.; Skwarczynski, M.; Lejczak, B.; Kafar-
This work was financed from Grant 3 T09A 05427 given by
the Ministry of Science and Higher Education.
ski, P. Tetrahedron: Asymmetry 1996, 7, 1277.