Enzyme-promoted kinetic resolution of racemic, P-chiral phosphonyl and phosphorylacetates

Article abstract of DOI:10.1016/S0957-4166(98)00258-4

A series of racemic methyl phosphonyl- and phosphorylacetates were hydrolyzed in the presence of porcine liver esterase (PLE) under kinetic resolution conditions to give the corresponding P-chiral phosphonyl- and phosphorylacetic acids and recovered esters in moderate to high enantiomeric purity (up to 95% ee). The Jones PLE active site model was applied to explain the enantioselectivity of this reaction.

Full text of DOI:10.1016/S0957-4166(98)00258-4

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 9 (1998) 2641–2650
Enzyme-promoted kinetic resolution of racemic, P-chiral
phosphonyl and phosphorylacetates
Piotr Kiełbasin´ski,a,∗ Piotr Góralczyk,a Marian Mikołajczyk,a Michał W. Wieczorek b and
Wiesław R. Majzner b
aCentre of Molecular and Macromolecular Studies, Polish Academy of Sciences, Sienkiewicza 112, 90-363 Łódz´, Poland
bInstitute of General Food Chemistry, Technical University, Stefanowskiego 4/10, 90-924 Łódz´, Poland
Received 19 May 1998; accepted 25 June 1998
Abstract
A series of racemic methyl phosphonyl- and phosphorylacetates were hydrolyzed in the presence of porcine
liver esterase (PLE) under kinetic resolution conditions to give the corresponding P-chiral phosphonyl- and
phosphorylacetic acids and recovered esters in moderate to high enantiomeric purity (up to 95% ee). The Jones
PLE active site model was applied to explain the enantioselectivity of this reaction. © 1998 Elsevier Science Ltd.
All rights reserved.
1. Introduction
Enzymes, the proteins created by nature to catalyze chemical reactions occurring in living organisms,
have recently become common reagents for the transformation of man-made organic compounds.1–3
As enzymes are intrinsically chiral, they have mainly been applied in the synthesis of optically active
molecules, using both racemic and prochiral substrates. In addition to hundreds of examples of enzyme-
promoted syntheses of non-racemic C-chiral products, in recent years the preparation of several classes
of heteroorganic compounds, with a sole center of chirality located on the heteroatom, such, as sulfur,
silicon, germanium and phosphorus, has also been achieved.4
Within this subject area we have demonstrated the applicability of enzymatic methods in the prepa-
ration of optically active sulfoxides5 and phosphine oxides.6,7 In the latter case, a series of racemic
methyl phosphinylacetates was hydrolyzed in the presence of porcine liver esterase (PLE) under kinetic
resolution conditions to give the corresponding P-chiral phosphinylacetic acids and recovered esters in
72–100% enantiomeric purity. Stereoselectivity of this reaction was explained7 in terms of the Jones’
PLE active site model,8 originally developed for substrates with a stereogenic carbon.
∗
Corresponding author. E-mail: piokiel@bilbo.cbmm.lodz.pl
0957-4166/98/$ - see front matter © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00258-4

2642
P. Kiełbasin´ski et al. / Tetrahedron: Asymmetry 9 (1998) 2641–2650
The aim of our present study was to determine the scope of the above reaction and to gain more
data which would allow a general explanation of the stereoselectivity observed to be formulated. Herein
we would like to disclose the detailed results of the investigations on the PLE promoted hydrolysis of
phosphonyl- and phosphorylacetates.9
2. Results and discussion
The racemic phosphonyl- and phosphorylacetates 1a–i were subjected to hydrolysis in the presence of
PLE at 30°C in phosphate buffer, using an automatic titrator to maintain the appropiate pH value. The
reaction was generally carried out until 50% conversion was reached, although in some cases, when the
hydrolysis was very slow, it was stopped earlier. The reaction was then quenched by additon of excess
acetone and cooling the solution to below 0°C for a few hours. The precipitate formed was filtered off
and the acetone evaporated. The unreacted ester was extracted with dichloromethane and the aqueous
layer acidified with sulfuric acid to pH≈3 and lyophilized. Since the acids 2* formed were generally
difficult to purify, they were reesterified with an excess of methanol in the presence of a catalytic amount
of H2SO4. Both enantiomerically enriched esters thus obtained were purified by column chromatography
(Eq. 1, Table 1).
(1)
1
The enantiomeric excess (ee) values of 1 were determined by means of H-NMR spectra of their
complexes with (−)-(S)-t-butylphenylphosphinothoic acid which has recently been widely used as a
versatile chiral solvating agent.10 An inspection of Table 1 reveals that the ee values of the products
obtained are generally moderate, the exceptions being 1a and 1i, where ee exceeds 90%. This is in
contrast to the high stereoselectivity observed in the case of phosphinylacetates described previously.6,7
As the lower stereoselectivity is accompanied by a severe drop in the reaction rate, a competing, non-
enzymatic hydrolysis had to be excluded. To this end, a background reaction was conducted under the
enzymatic reaction conditions applied (pH up to 7.6) and showed that no hydrolysis of 1c, used as an
example, took place after 140 hours.
Nevertheless, it was necessary to establish the absolute configuration of the products in order to find out
whether there is any general relationship between the substrate structure and product configuration. This
was achieved for the unreacted esters 1*a–e and acids 2*a–e. Thus, the absolute configuration of 2*a was
known from the literature to be (+)-(R).11 The absolute configuration of 2*b was determined on the basis
of chemical correlation, consisting of decarboxylation of (+)-2*b, using the conditions described earlier

P. Kiełbasin´ski et al. / Tetrahedron: Asymmetry 9 (1998) 2641–2650
2643
Table 1
Enzymatic hydrolysis of phosphonyl- and phosphorylacetates 1

2644
P. Kiełbasin´ski et al. / Tetrahedron: Asymmetry 9 (1998) 2641–2650
Fig. 1. Thermal ellipsoidal view with the atom numbering scheme of the salt of (+)-2*b with (−)-(S)-α-phenylethylamine.
Ellipsoids are shown with 50% probability
for decarboxylation of α-carboxymethylsulfonium salts,12 to the corresponding phosphinate 3 of known
absolute configuration13 (Eq. 2), and confirmed by an X-ray analysis performed for the diastereomerically
pure salt of (+)-2*b with (−)-(S)-α-phenylethylamine (Fig. 1, Table 2).
(2)
In turn, the absolute configurations of 1*c, d, e were ascribed by comparison of their CD spectra with
those of 1*a and 1*b. It turned out that the CD curves of all the levorotatory unreacted esters were of the
same shape and exhibited the same sign of Cotton effect. The same was valid for all the dextrorotatory
esters obtained by re-esterification of acids 2*a–e, the sign of the Cotton effect being obviously opposite
in this case (Fig. 2).
Since all the compounds under discussion are closely related (the phosphorus atom is in each case
connected with three identical substituents, the fourth one, i.e. an alkoxy group, differing only by
the length or branching of the alkyl chain), one can assume with a high probability that the products
exhibiting the same sign of specific rotation have the same absolute configuration at phosphorus. Thus,
all the levorotatory esters 1*a–e have (S) and the dextrorotatory ones (R) configuration. Moreover, an
inspection of Table 1 clearly shows that in all cases the (R) enantiomers are recognized by PLE and
hence are preferentially hydrolyzed.
The above regularity makes it possible to consider application of the Jones’ model8 to the phosphonyl-
acetates investigated, as was done previously by us for phosphinylacetates7 (Fig. 3).

P. Kiełbasin´ski et al. / Tetrahedron: Asymmetry 9 (1998) 2641–2650
2645
Table 2
Crystal data and experimental details
Thus, the (R) enantiomers which undergo hydrolysis faster than the (S) ones, should be accommodated
in the ‘pockets’ of the enzyme active site as follows.
The methoxycarbonyl group should be located within the spherical locus of the catalytically active
serine function, the large phenyl group in the large hydrophobic pocket (HL), the alkoxy group in
the small hydrophobic pocket (HS) and finally the phosphoryl oxygen in the back polar pocket (PB)
(Fig. 3a). As far as the location of the three former groups is quite obvious, there is some uncertainty

2646
P. Kiełbasin´ski et al. / Tetrahedron: Asymmetry 9 (1998) 2641–2650
Fig. 2. CD spectra (in MeOH) of chiral esters 1*
Fig. 3. Binding orientations of 1*a–e in the active site of PLE: (a) of the enantiomer preferably hydrolyzed; (b) of the enantiomer
recovered
as to whether the phosphoryl oxygen may alternatively be accommodated in the front polar pocket PF
(Fig. 3b), which would lead to the preferential hydrolysis of the (S) enantiomers, and in the end result
in a lower stereoselection. However, in our opinion it is not the case. Thus, according to the Jones’
predictions, the PF pocket is mostly used to bind the second nonhydrolyzed ester function of diester
substrates, but it can accept nonpolar groups also. On the other hand, the PB site interacts well, among
others, with carbonyl functions, but is too polar to accept hydrophobic moieties. That is why the strongly
polar P_O group should be preferentially accommodated in the PB pocket, hence favoring the model
shown in Fig. 3a. Moreover, since the stereoselectivity is very high for 1a (R=Me) and becomes lower
on passing to the larger alkyl groups, this is rather the decreasing size difference between the ‘large’ and
the ‘small’ nonpolar substituents which seems responsible for the lower ee values of the products.

P. Kiełbasin´ski et al. / Tetrahedron: Asymmetry 9 (1998) 2641–2650
2647
3. Experimental
3.1. General
Phosphate buffer solutions were purchased from Aldrich. An ammonium sulfate suspension of porcine
liver esterase was used which was purchased either from FLUKA or SIGMA. A 0.2 M solution of NaOH
was used in an automatic titrator to adjust pH. NMR spectra were recorded at 300 MHz or 500 MHz for
1H and 121 MHz for 31P with CDCl3 as a solvent. Optical rotations were measured on a Perkin–Elmer
241 MC polarimeter. Column chromatography was carried out using Merck 60 F254 silica gel. TLC was
performed on Merck 60 F254 silica gel plates.
3.2. Synthesis of substrates 1
All the substrates were prepared via the Arbuzov reaction of methyl bromoacetate with the appropriate
trivalent phosphorus esters (see below), according to the following general procedure. To methyl
bromoacetate (0.1 mol), stirred, heated to 80°C and purged with a continuous stream of argon, the
appropriate trivalent phosphorus ester was added very slowly. The reaction was exothermic. After the
addition was completed the mixture was heated at 90°C for an additional hour and then submitted to
fractional distillation. All products gave satisfactory combustion analysis results (C, H, P ꢀ0.4%).
3.3. Methyl methoxy(phenyl)phosphonylacetate 1a
Obtained from dimethyl phenylphosphonite and methyl bromoacetate as a colorless liquid after
1
distillation (bp 140°C/0.2 mmHg) in 74% yield. 31P NMR: δ 36.0. H NMR: δ 3.10 (d, JP–H=17.7,
2H), 3.61 (d, JP–H=0.54, 3H), 3.72 (d, JP–H=11.4, 3H), 7.40–7.85 (m, 5H).
3.4. Methyl ethoxy(phenyl)phosphonylacetate 1b
Obtained from diethyl phenylphosphonite and methyl bromoacetate as a colorless liquid after distil-
lation (bp 132°C/0.1 mmHg) in 86% yield. 31P NMR: δ 34.3. 1H NMR: δ 1.31 (t, 3H), 3.11 (d, JP–H=17.7,
2H), 3.60 (s, 3H), 3.9–4.2 (m, 2H), 7.40–7.80 (m, 5H).
3.5. Methyl n-propoxy(phenyl)phosphonylacetate 1c
Obtained from di-n-propyl phenylphosphonite and methyl bromoacetate as a colorless oil after bulb-
to-bulb distillation (bp 150°C/0.2 mmHg) in 88% yield. 31P NMR: δ 34.2. 1H NMR: δ 0.95 (t, 3H), 1.71
(sxt, 2H), 3.13 (d, JP–H=17.7, 2H), 3.63 (d, JP–H=0.53, 3H), 3.5–4.2 (m, 2H), 7.45–7.80 (m, 5H).
3.6. Methyl isopropoxy(phenyl)phosphonylacetate 1d
Obtained from di-isopropyl phenylphosphonite and methyl bromoacetate as a colorless oil after bulb-
to-bulb distillation (bp 150°C/0.2 mmHg) in 75% yield. 31P NMR: δ 33.0. 1H NMR: δ 1.24 (d, 3H), 1.41
(d, 3H), 3.13 (d, JP–H=17.7, 2H), 3.63 (s, 3H), 4.55–4.75 (m, 1H), 7.45–7.80 (m, 5H).

2648
P. Kiełbasin´ski et al. / Tetrahedron: Asymmetry 9 (1998) 2641–2650
3.7. Methyl n-butoxy(phenyl)phosphonylacetate 1e
Obtained from di-n-butyl phenylphosphonite and methyl bromoacetate as a colorless oil after bulb-
1
to-bulb distillation (bp 150°C/0.2 mmHg) in 72% yield. 31P NMR: δ 34.2. H NMR: δ 0.91 (t, 3H),
1.25–1.50 (m, 2H), 1.55–1.75 (m, 2H), 3.12 (d, JP–H=17.7, 2H), 3.63 (s, 3H), 3.85–4.45 (m, 2H),
7.40–7.80 (m, 5H).
3.8. Methyl methoxy(ethyl)phosphonylacetate 1f
Obtained from dimethyl ethylphosphonite and methyl bromoacetate as a colorless liquid after frac-
tional distillation (bp 60°C/0.1 mmHg) in 60% yield. 31P NMR: δ 51.5. 1H NMR: δ 1.15 (dt, JH–H=7.67,
JP–H=19.6, 3H), 1.75–1.95 (m, 2H), 2.19 (d, JP–H=16.9, 2H), 3.68 (s, 3H), 3.70 (d, JP–H=10.93, 3H).
3.9. Methyl ethoxy(ethyl)phosphonylacetate 1g
Obtained from diethyl ethylphosphonite and methyl bromoacetate as a colorless liquid after fractional
1
distillation (bp 70°C/0.25 mmHg) in 84% yield. 31P NMR: δ 49.3. H NMR: δ 1.16 (dt, JH–H=7.67,
JP–H=19.43, 3H), 1.29 (t, 3H), 1.8–2.0 (m, 2H), 2.92 (d, JP–H=16.9, 2H), 3.70 (s, 3H), 3.95–4.25 (m, 2H).
3.10. Methyl ethoxy(phenoxy)phosphorylacetate 1h
Obtained from diethyl phenylphosphite and methyl bromoacetate as a colorless oil after fractional
1
distillation (bp 105°C/0.1 mmHg) in 62% yield. 31P NMR: δ 17.0. H NMR: δ 1.31 (dt, JH–H=7.1,
JP–H=0.54, 3H), 3.10 (d, JP–H=21.6, 2H), 3.73 (d, JP–H=0.52, 3H), 4.1–4.4 (m, 2H), 7.1–7.4 (m, 5H).
3.11. Methyl methoxy(diethylamido)phosphorylacetate 1i
Obtained from dimethyl diethylamidophosphite and methyl bromoacetate as a colorless liquid after
1
fractional distillation (bp 85°C/0.3 mmHg) in 43% yield. 31P NMR: δ 26.0. H NMR: δ 1.07 (t, 6H),
2.87 (2×AB, 2H), 2.97–3.10 (m, 4H), 3.59 (d, JP–H=11.46, 3H), 3.68 (d, JP–H=0.57, 3H).
3.12. General procedure for the enzymatic hydrolysis of 1
To a stirred solution of ester 1 (1 mmol) in a phosphate buffer (∼20 ml), at 30°C, PLE (∼50 µL) was
added. The pH was maintained by a continuous addition of 0.2 M aqueous NaOH using an automatic
titrator. When 50% conversion was reached (i.e. 2.5 mL of NaOH added dropwise) the reaction was
quenched by adding 200 mL of acetone and cooling in a freezer for ca. 3 h (in some cases the reaction
was deliberately stopped at different conversions, see Table 1). The mixture was filtered through Celite
and the acetone was evaporated. The residue was extracted with CH2Cl2 (3×10 mL) and the combined
organic layers dried over MgSO4. The solvent was evaporated and the crude product purified by silica
gel chromatography (AcOEt:hexane gradient as eluent) to give pure unreacted esters 1* (see Table 1).
The remaining aqueous layer was acidified with H2SO4 (pH ca. 3.0) and lyophilized. The residue was
either extracted with CHCl3, dried and evaporated to give acids 2*, or dissolved in excess methanol and
left at room temperature for three days, during which time acids 2* were re-esterified to the methyl esters
1* of opposite absolute configuration. Methanol was evaporated, the residue extracted with CH2Cl2,
washed with a small amount of water and dried over MgSO4. After evaporation of the solvent, the residue

P. Kiełbasin´ski et al. / Tetrahedron: Asymmetry 9 (1998) 2641–2650
2649
was purified by column chromatography (as above) to give the esters 1*. Specfic rotations and the ee
values of the products obtained are collected in Table 1.
3.13. Decarboxylation of 2*b
Decarboxylation was performed according to the procedure described previously.12 Thus, 2*b [0.488
20
g, 2.1 mmol, [α]D +9.2, (c 1.2, MeOH)] was placed in toluene (5 mL) and acetone was added until
the solid had dissolved. Then Bu3N (1 mL) was added and the mixture was refluxed for 5 days. The
solvents were evaporated and the residue was separated on preparative TLC (MeOH:CHCl3=1:15). (+)-
20
(R)-Ethyl methylphenylphosphinate 3 was obtained (0.1 g, 25%), [α]D +28.7, (c 1.15, benzene) [lit.13
for (+)-(R)-3 [α]D20 +49 (benzene)].
3.14. The salt of (+)-2a and (−)-(S)-α-phenylethylamine
The diastereomerically pure salt of (+)-2a and (−)-(S)-α-phenylethylamine was obtained after re-
peated crystallization of the appropriate diastereomeric mixture from benzene. Diastereomeric purity was
established by 1H NMR. The corresponding optical rotation values were as follows: of the diastereomer-
20
ically pure salt: [α]D +11.4 (c 0.43, MeOH); of 2*b released from the above salt using ion exchange
resin: [α]D20 +9.2 (c 0.26, benzene); [α]D20 +19.2 (c 0.26, MeOH); lit.11: [α]D20 +9.1 (c 3.0, benzene).
3.15. Crystal structure of the salt (+)-2*b with (−)-(S)-α-phenylethylamine
Crystal and molecular structure of the salt of (+)-2*b with (−)-(S)-α-phenylethylamine was deter-
mined using data collected at room temperature on a CAD4 diffractometer with graphite monochroma-
tized CuKα radiation. The compound crystallizes in the orthorhombic system, in space group P212121
with the unit cell consisting of 4 anions and 4 cations. Crystal data and experimental details are shown
in Table 2. The lattice constants were refined by a least-squares fit of 25 reflections in the θ range
20.35–36.08°. The decline in intensities of three control reflections (−2,6,−5; −4,−2,3; −2,6,7) was
5.3% during 46.2 hours of exposure time; the DECAY correction was applied.14 An empirical absorption
correction was applied by the use of the ψ-scan method (EAC program).14,15 A total of 3419 observed
reflections with I>0σ(I) were used to solve the structure by direct methods and to refine it16,17 by
full matrix least-squares using F2. Hydrogen atoms were found on a difference Fourier map and were
refined isotropically, except hydrogens attached to the C4 that were placed geometrically at idealized
positions, set as riding with C–H distance free to refine, and fixed thermal parameters equal to 1.3 times
the equivalent isotropic thermal parameter of the parent atom (rotation about the C4–C3 bond was also
allowed). Anisotropic thermal parameters were refined for all nonhydrogen atoms. The final refinement
converged at R=0.0580 for 304 refined parameters and 3316 observed reflections with I>2σ(I).
The authors have deposited all crystallographic data for this structure with the Cambridge Crystallo-
graphic Data Centre.18
Acknowledgements
Financial support by the Committee of Scientific Research (KBN), Grant No 3T09A 101 10 for P.K.,
is gratefully acknowledged.

2650
P. Kiełbasin´ski et al. / Tetrahedron: Asymmetry 9 (1998) 2641–2650
References
1. Enzyme Catalysis in Organic Chemistry — A Comprehensive Handbook, Drauz, K.; Waldmann, H., Eds; VCH: Weinheim,
1995.
2. Preparative Biotransformations, Roberts, S. M., Ed.; Wiley & Sons: Chichester. Consecutively edited and supplemented
starting from 1990.
3. Faber, K. Biotransformations in Organic Chemistry, 3rd Edn; Springer Verlag: Berlin, 1996.
4. For a review see: Kiełbasin´ski, P.; Mikołajczyk, M. Main Group Chem. News, 1997, 5, 7.
˙
5. Mikołajczyk, M.; Kiełbasin´ski, P.; Zurawin´ski, R.; Wieczorek, M. W.; Błaszczyk, J. Synlett, 1994, 127.
˙
6. Kiełbasin´ski, P.; Zurawin´ski, R.; Pietrusiewicz, K. M.; Zabłocka, M.; Mikołajczyk M. Tetrahedron Lett., 1994, 35, 7081.
˙
7. Kiełbasin´ski, P.; Zurawin´ski, R.; Pietrusiewicz, K. M.; Zabłocka, M.; Mikołajczyk M. Polish J. Chem., 1998, 72, 564.
8. Toone, E. J.; Werth, M. J.; Jones, J. B. J. Am. Chem. Soc., 1990, 112, 4946; Provencher, L.; Jones, J. B. J. Org. Chem.,
1994, 59, 2729.
9. Preliminary results were presented at the XIIIth ICPC, Jerusalem, 1995: Kiełbasin´ski, P.; Mikołajczyk M. Phosphorus,
Sulfur, Silicon, 1996, 109–110, 497.
10. For recent applications see: Omelan´czuk, J.; Mikołajczyk M. Tetrahedron: Asymmetry, 1996, 7, 2687; Drabowicz, J.;
Dudzin´ski, B.; Mikołajczyk, M.; Colonna, S.; Gaggero, N. Tetrahedron: Asymmetry, 1997, 8, 2267 and references therein.
11. Gałdecki, Z.; Główka, M. L.; Musierowicz, S.; Michalski, J. Polish J. Chem., 1980, 54, 539.
12. Kelstrup, E.; Kjaer, A.; Abrahamsson, S.; Dahlen, B. J. Chem. Soc., Chem. Commun., 1975, 629.
13. DeBruin, K. E.; Mislow, K. J. Am. Chem. Soc., 1969, 91, 7393.
14. Frenz, B. A. SDP — Structure Determination Package, Enraf-Nonius, Delft: Holland, 1984.
15. North, A. C. T.; Philips, D. C.; Mathews, F. S. Acta Cryst., 1968, A24, 351.
16. Sheldrick, G. M.; Kruger, G. M.; Goddard, R. SHELXS-86. Crystallographic Computing 5, Oxford, 1985.
17. Sheldrick, G. M. SHELXS-97.
18. University Chemical Laboratory, Cambridge Crystallographic Data Centre, Lensfield Read, Cambridge CD2 1EW, UK.
19. Flack, H. D. Acta Cryst., 1983, A39, 876–881.