Enantioselective ester hydrolysis catalyzed by imprinted polymers. 2

J . Org. Chem. 2000, 65, 4009-4027

4009

En a n tioselective Ester Hyd r olysis Ca ta lyzed by Im p r in ted

P olym er s. 2^§,|

Bo¨rje Sellergren,*^,†Rohini N. Karmalkar,^†and Kenneth J . Shea^‡

Department of Inorganic Chemistry and Analytical Chemistry, J ohannes Gutenberg University,

Mainz, Duesbergweg 10-14, D-55099, Mainz, Germany, and Department of Chemistry,

University of California, Irvine, California 92717

Received J anuary 6, 2000

Highly cross-linked network polymers prepared by molecular imprinting catalyzed enantioselectively

the hydrolysis of N-tert-butoxycarbonyl phenylalanine-p-nitrophenyl ester (BOCPheONP). The

templates were designed to allow incorporation of the key catalytic elements, found in the proteolytic

enzyme chymotrypsin, into the polymer active sites. Three model systems were evaluated. These

were constructed from a chiral phosphonate analogue of phenylalanine (series A, C) or ^L-

phenylalanine (series B) attached by a labile ester linkage to an imidazole-containing vinyl monomer.

Free radical copolymerization of the template with methacrylic acid (MAA) and ethylene glycol

dimethacrylate (EDMA) gave a highly cross-linked network polymer. The templates could be

liberated from the polymers by hydrolysis, giving catalytically active sites envisaged to contain an

enantioselective binding site, a site complementary to a transition state like structure (series A,

C), and a hydroxyl, imidazole, and carboxylic acid group at hydrogen bond distance. As predicted,

the enantiomer of BOCPheONP complementary to the configuration of the template was

preferentially hydrolyzed with ^D-selectivity for the series A polymers (kD/kL ) 1.9) and L-selectivity

for the series B polymers (k^L/kD ) 1.2). The maximum rate enhancement, when compared with a

control polymer, prepared using a benzoyl-substituted imidazole monomer as template, was 2.5,

and comparing with the imidazole monomer in solution, a maximum rate enhancement of 10 was

observed. The catalytic activity was higher for polymers subjected to the nucleophilic treatment.

This was explained by a higher site density and flexibility of the polymer matrix caused by this

treatment. In a comparison of template rebinding to polymers imprinted with a template containing

either a carboxylate (planar ground state structure) or a phosphonate (tetrahedral transition state

like structure) functionality, it was observed that imprinted polymers are able to discriminate

between a transition state like and a ground state structure for transesterification. However the

influence of transition state stabilization on the observed rate enhancements remains obscure. Only

at acidic pH’s was catalysis observed, whereas at basic pH’s the polymers inhibit the reaction. At

a later stage, the catalytic activity of the polymers for nonactivated ^D- and L-phenylalanine ethyl

esters was investigated. A rate enhancement of up to 3 was observed when compared to the blank.

Most important, however, the polymers imprinted with a ^Dtemplate preferentially hydrolyzed the

D-ethyl ester and exhibited saturation kinetics.

In tr od u ction

cessful. Recently however, promising rate enhancements

were reported for the hydrolysis of nonactivated esters

using imprinted polymers as heterogeneous catalysts.¹⁰

Still, only a few attempts have been made to achieve

stereoselective hydrolysis using molecularly imprinted

polymers.^8c,d,11The problems in achieving stereoselec-

tivity is related to the common use of nondirectional

hydrophobic binding forces as well as the lack of design

The ability of enzymes to catalyze reactions with high

stereoselectivity and specificity has led to several ap-

plications of enzymes in the chemical and pharmaceutical

industry.^1,2For instance lipases have been successfully

used for stereoselective acylation and deacylation of

racemates.³The development of catalytic antibodies has

added the possibility of tailor-making catalysts for a given

reaction.^1,4Examples of catalytic antibodies exhibiting

lipase activity with either R or S specificity have been

described.⁴Attempts to mimic hydrolytic enzymes using

either synthetic low molecular,^1,5polymeric,^6,7or im-

printed polymeric^8,9model systems have been less suc-

(5) (a) Breslow, R. Science 1982, 218, 532-537. (b) Bender, M. L.

In Enzyme Mechanisms; Page, M. I., Williams, A., Eds.; Royal Society

of Chemistry: London, 1987; Chapter 4.

(6) (a) Klotz, I. M. In Enzyme Mechanisms; Page, M. I., Williams,

A., Eds.; Royal Society of Chemistry: London, 1987; Chapter 2. (b)

Suh, J . Acc. Chem. Res. 1992, 25, 273.

(7) Fife, W. K. Trends Polym. Sci. 1995, 3, 214.

^§Dedicated to Professor Gu¨nter Wulff on the occasion of his 65th

birthday.

(8) (a) Leonhardt, A.; Mosbach, K. React. Polym. 1987, 6, 285. (b)

Robinson, D. K.; Mosbach, K. J . Chem. Soc., Chem. Commun. 1989,

969. (c) Morihara, K.; Kurokuwa, M.; Kamata, Y.; Shimada, T. J . Chem.

Soc., Chem. Commun. 1992, 358.13. (d) Ohkube, K.; Fukanoshi, Y.;

Urata, Y.; Hirota, S.; Usui, S.; Sagawa, T. J . Chem. Soc., Chem.

Commun. 1995, 2143. (e) Karmalkar R. N.; Kulkarni, M. G.; Mashelkar,

R. A. Macromolecules 1996, 29, 1366. (f) Lele B. S.; Kulkarni, M. G.;

Mashelkar, R. A. Polymer 1999, 40, 4063.

^|For part 1, see ref 11.

^†J ohannes Gutenberg University.

^‡University of California.

(1) Kirby, A. Angew. Chem. 1996, 108, 770.

(2) J ones, J . B. Desantis, G. Acc. Chem. Res. 1999, 32, 99.

(3) Whitesides, G. M., Wong, C. H. Angew. Chem., Int. Ed. Engl.

1985, 24, 617.

(4) (a) Schultz, P. G. Angew. Chem., Int. Ed. Engl. 1989, 28, 1283.

(b) J anda, K. D.; Benkovic, S. J .; Lerner, R. A. Science 1989, 244, 437.

(9) Davis, M. E.; Katz. A.; Ahmed, W. R. Chem. Mater. 1996, 8, 1820.

(10) Wulff G.; Gross, T.; Scho¨nfeld, R. Angew. Chem., Int. Ed. Engl.

1997, 36, 1963.

Published on Web 06/08/2000

4010 J . Org. Chem., Vol. 65, No. 13, 2000

Sellergren et al.

Sch em e 1

elements in the preparation of the catalyst.^6,7A stereo-

specific binding site is more readily achieved using

directed binding forces such as hydrogen bonds.¹

imidazole and a carboxyl group at hydrogen bond dis-

tance to the nucleophile. We have earlier reported pre-

dictable enantioselective ester hydrolysis of tert-butoxy-

carbonyl phenylalanine-p-nitrophenyl ester (BOCPheONP)

using these polymers (see Scheme 1).¹¹

Our goal was to use the binding energy and selectivity,

observed using imprinted polymers,¹²to achieve stereo-

selective esterolytic catalysis.¹¹By allowing a network

polymer to form in the presence of a suitable template

molecule and subsequently liberating the template, a

polymer containing selective binding sites can be pre-

pared. Focusing on the mimic of chymotrypsin, we have

synthesized several templates designed to provide an

active site incorporating the elements believed to be

responsible for the catalytic action of chymotrypsin:^13,14

(1) a stereoselective binding site, (2) a site able to stabilize

a tetrahedral intermediate, and (3) a nucleophile in

proximity of the reactive carbonyl group as well as an

This paper describes the synthesis of an extended

series of template monomers, their incorporation into

highly cross-linked methacrylate polymers, removal of

the template from the polymers, and finally the evalua-

tion of the esterase activity of monomers and polymers.

At a later stage, ^D- and L-phenylalanine ethyl ester was

used as substrates to investigate the polymers ability to

catalyze the hydrolysis of nonactivated esters.

Exp er im en ta l Section

Gen er a l P r oced u r es. Unless otherwise noted materials

were obtained from Aldrich and used without further purifica-

tion. Standard laboratory reagents and solvents were purified

according to standard procedures and the reactions carried out

under inert atmosphere (N₂). The NMR spectra were recorded

with a 300 MHz (GE, QE-300) instrument. ¹³C NMR spectra

were obtained at 75.4 MHz. Chemical shifts are reported in δ

units utilizing TMS as an internal reference for ¹H NMR and

the solvent signal for ¹³C NMR. The FTIR spectra were

(11) Sellergren, B.; Shea, K. J . Tetrahedron Asymmetry 1994, 5,

1403.

(12) (a) Molecular and Ionic Recognition with Imprinted Polymers;

ACS Symposium Series 703; Bartsch R. A., Maeda, M., Eds.; Oxford

University Press: Oxford, 1998. (b) Wulff, G. Angew. Chem., Int. Ed.

Engl. 1995, 34, 1812-1832. (c) Sellergren, B. Angew. Chem., Int. Ed.

2000, 39, 1031-1037.

(13) Fersht A. R. Enzyme Structure and Mechanisms; Freeman:

New York, 1985.

(14) Bruice, T., Lightstone, F. C. Acc. Chem. Res. 1999, 32, 127.

Enantioselective Ester Hydrolysis

J . Org. Chem., Vol. 65, No. 13, 2000 4011

Sch em e 2

recorded on a Bomem Michelson FTIR spectrometer, and the

solid state ¹³C NMR spectra were kindly recorded by Professor

Thomas Hjertberg and Dr. Staffan Schantz at Chalmers

Institute of Technology, Sweden.

(tri-n-butyl)tin (0.65 mL, 2.24 mmol), and a trace of 2,6-di-

tert-butyl-4-methylphenol in toluene (15 mL) under nitrogen,

and the solution was brought to reflux resulting in a color

change from yellow to red. The reaction was followed on GC

(methylsilica column, T_c) 180 °C) by measuring the consump-

tion of vinyl(tributyl)tin (R_f) 0.8) and the appearance of

product peaks at k′ ) 1.38 and 1.70. After 5 h all of the reagent

had been consumed. After filtration the solvent was evaporated

at room temperature and the oil dissolved in 10 mL of

acetonitrile. This was then washed with hexane and then

evaporated giving 1.00 g of an oil that crystallized slowly.

Recrystallization from benzene/hexane gave 0.47 g (53%) of

light yellow crystals. Mp: 124-127 °C. ¹H NMR (CDCl₃): δ

5.25 (d, J ) 10.9 Hz, 1H, vinyl), 5.74 (d, J ) 17.6 Hz, 1H, vinyl),

6.68 (dd, J ) 10.9, 17.7 Hz, 1H, vinyl), 7.02-7.11 (m, 2H, Ph),

7.19-7.27 (m, 5H, Ph), 7.42-7.53 (m, 5H, Ph), 7.65-7.80 (m,

3H, Ph). ¹³C NMR (CDCl₃): δ 113.67, 119.94, 121.47, 124.07,

126.54, 126.99, 127.15, 127.20, 127.24, 127.64, 127.85, 128.56,

129.16, 129.78, 132.44, 132.61, 133.97, 134.11, 134.60, 143.60,

146.26, 162.73, 165.47. HRMS calcd for C₂₅H₁₈N₂O₃: 394.1317.

Found: 394.1309.

Mon om er Syn th esis (Ser ies A) (Sch em e 6). 2-(2′-Hy-

d r oxy-5′-br om op h en yl)im id a zole (5). p-Bromosalicylalde-

hyde (10 g, 0.05 mol) and glyoxal (trimertetrahydrate) (10.5

g, 0.05 mol) in 200 mL of glacial acetic acid were brought to

reflux under nitrogen, and then ammonium acetate (36 g, 0.05

mol) was added followed by continued reflux for 4 h. After

cooling to room temperature the dark brown solution was

poured into 2 L of water, giving a dark brown precipitate. After

filtration on Celite, the deep red filtrate was made basic with

aqueous ammonia, and a yellow precipitate was formed.

Filtration and drying gave 7.9 g (66%) of 5 as a light brown

solid. Purification on a silica column failed, but active carbon

may be applied however with some loss of product. The purity

was sufficient for the following reaction. TLC: R_f) 0.5 (silica

plates, CHCl₃/MeOH: 9/1), mp 240-242 °C. ¹H NMR (DMSO-

d₆+ 3 drops CD₃COOD): δ 0.10 (bs, OH, NH), 6.82 (d, J )

8.7 Hz, 1H, H3′), 7.13 (s, 2H, H4,5), 7.23 (d, J ) 8.5, 1H,H4′),

8.03 (s, 1H, H6′). ¹³C NMR (DMSO-d₆+ 3 drops CD₃COOD):

δ 110.68 (C5′), 115.97 (C1′), 119.48 (C3′), 122.32 (C4,5), 126.78

(C6′), 132.64 (C4′), 145.12 (C2′), 156.19 (C2). HRMS calcd for

C₉H₇N₂OBr: 237.9742. Found: 237.9723.

N-Ben zoyl-2-(2′-ben zoxy-5′-br om oph en yl)im idazole (6).

To 5 (1 g, 4.2 mmol) and triethylamine (1.2 mL, 8.4 mmol) in

dry THF was added benzoyl chloride (1.0 mL, 8.4 mmol), and

the mixture stirred at room temperature overnight. After

filtration of the salt the filtrate was evaporated to dryness,

giving crude 6 as a yellow solid, which was recrystallized by

dissolving the solid in 8 mL of benzene and adding 100 mL of

hexane; 1.37 g (67%) of light yellow crystals precipitated. Mp:

132-134 °C. ¹H NMR (CDCl₃): δ 7.05-7.14 (m, 2H), 7.22-

7.28 (m, 5H), 7.48-7.56 (m, 5H), 7.78-7.86 (m, 3H). ¹³C NMR

(CDCl₃): δ 118.75, 121.02, 123.94, 126.66, 128.20, 128.25,

128.29, 129.28, 129.89, 130,37, 130,83, 133.08, 133.56, 133.78,

2-(2′-Hyd r oxy-5′-eth en ylp h en yl)im id a zole (7). To a so-

lution of A3 (0.59 g, 1.49 mmol) in ethanol (3 mL) was added

sodium dissolved in ethanol (3 mg/mL). A yellow color ap-

peared, and a strongly UV-active product spot appeared

immediately on TLC (silica plates, CHCl₃/MeOH: 9:1, R_f)

0.6). Ethanol was evaporated and the resulting oil dissolved

in aqueous NaOH (25 mL, 0.2 M). This was extracted with

hexane. To the aqueous phase was added 1 M HCl until a

yellow precipitate was formed at pH 10-11. The precipitate

went back into solution at pH below 4-5. The pH was adjusted

to 7-8 followed by extraction of the product into ethyl acetate.

Drying on MgSO₄and evaporation gave crude 7 as a brown

solid. This was purified by column chromatography (silica, 70

g using CHCl₃/MeOH, 9:1, as eluent), giving 0.20 g (72%) of

pure 7 as a light beige solid. HPLC: RP-18 column using

MeOH/[phosphate buffer 0.05 M, pH 2.5]: 1:2 (v/v) as eluent

gave a single peak (k′ = 1.5). Mp: 131-134 °C. ¹H NMR

(CDCl₃+ 3 drops CD₃OD): δ 5.10 (bs, 2H, NH,OH), 5.12 (d, J

) 10.9 Hz, 1H, vinyl), 5.65 (d, J ) 17.5 Hz, 1H, vinyl), 6.56

(dd, J ) 10.9, 17.6 Hz, 1H, vinyl), 6.96 (d, J ) 8.5 Hz, 1H, H3,

Ph), 7.08 (s, 2H, H4,5, imidazole), 7.30 (d, J ) 8.5 Hz, 1H, H4,

133.82, 143.86, 147.02, 163.73, 166.55. HRMS calcd for C₂₃H₁₅

BrN₂O₃: 446.0266. Found: 446.0249.

-

N-Ben zoyl-2-(2′-b en zoxy-5′-et h en ylp h en yl)im id a zole

(A3). Palladium(tetrakistriphenylphosphine) (0.052 g, 0.045

mmol) was added to a solution of 6 (1.0 g, 2.24 mmol), vinyl-

4012 J . Org. Chem., Vol. 65, No. 13, 2000

Sellergren et al.

Sch em e 3

Ph), 7.77 (s, 1H, H6,Ph). ¹³C NMR (CDCl₃+ 3 drops CD₃OD):

δ 111.09 (vinyl), 113.54 (C1′), 116.84 (C3′), 121.02 (C4,5),

121.99 (C6′), 127.36 (C4′), 128.84 (C5′), 135.94 (vinyl), 146.15

(C2′), 156.03 (C2). HRMS calcd for C₁₁H₁₀N₂O: 186.0793.

Found: 186.0794.

Tem p la te Syn th esis (Ser ies A) (Sch em e 5). S- and R-1-

amino-2-phenylethylphosphonate (1S and 1R) were synthe-

sized following the procedures by Oleksyszyn et al.¹⁵and

Kafarski et al.¹⁶The diastereomers were separated by convert-

ing into tartrate salts and fractional crystallization.¹⁷The

amino group was protected using bis-tert-butoxycarboxylic

anhydride (2S) and converted to the diethyl ester (3S) using

triethylorthoformate and then to the monoester (4S) by

alkaline hydrolysis.

followed on HPLC (RP-18 column, eluent: MeOH/[potassium

phosphate buffer, 0.05 M, pH 3]: 2:1 (v/v)) due to product

decomposition. After 15 h ca. 50% conversion was observed.

Additional BOP reagent (221 mg, 0.5 mmol) and triethylamine

(70 µL, 0.5 mmol) were therefore added, and the reaction was

allowed to proceed for a total of 24 h, at which point over 80%

conversion was observed. Two adjacent spots on TLC (silica

plates, eluent: CHCl₃/MeOH: 9:1, R_f) 0.5) as well as on

HPLC (k′ ) 1.6) indicated the presence of diastereomers. Brine

(10 mL) was added and the product extracted into ethyl

acetate. The organic phase was washed with 1 M HCl,

saturated NaHCO₃, and brine, dried on MgSO₄, and evapo-

rated, giving crude product as a yellow oil. Rapid purification

was done by flash chromatography on silica gel (eluent: CHCl₃/

MeOH, 9:1), giving 270 mg (91%) of A1 as an oil that solidified

to a light yellow foam upon further evacuation. The hygro-

S-Eth yl, 2-(2′-Im id a zolyl)-4-eth en ylp h en yl (1-(N-ter t-

Bu toxycar bon ylam in o)-2-ph en yl)eth ylph osph on ate (A1).

4S (197 mg, 0.6 mmol), 7 (112 mg, 0.6 mmol), triethylamine

(167 µL, 1.2 mmol), and Castros BOP reagent¹⁸(benzotriazol-

scopic product was stored dry under nitrogen in freezer. [R]²³

D

1

) +4.33° [c 2, CH₂Cl₂]. H NMR (CDCl₃+ 1 drop CD₃OD): δ

1-yloxy-tris(dimethylamino)phosphonium

hexafluorophos-

0.93, 1.14, 1.15 (3t, J ) 7.0 Hz, 3H, -CH₂CH₃, I_1/2/3) 40/30/

30), 1.24, 1.26 (2s, 9H, -C(CH₃)₃, I_1/2) 63/37), 2.80-2.88 (m,

1H, -CH₂-Ph), 3.07-3.14 (m, 1H, -CH₂Ph), 3.70-3.88 (m,

2H, -CH₂CH₃), 4.03-4.09 (m, 1H, -CH<), 4.38-4.45 (m, 1H,

-NH-), 5.24, 5.25 (2d, J ) 10.9 Hz, 1H, vinyl, I_1/2) 63/38),

5.74 (d, J ) 17.6 Hz, 1H, vinyl), 6.66 (dd, J ) 17.5, 10.9 Hz,

1H, vinyl), 7.11-7.47 (m, 10H, PhIm), 7.99 (s, 1H, H1-Im).

¹³C NMR (CDCl₃+ 1 drop CD₃OD): δ 15.66 15.94 (2d, J )

5.0,5.4 Hz, -O-CH₂-CH₃), 27.88 (C(CH₃)₃), 34.38, 34.45

(-CH₂-Ph), 47-50 (d, C-P), 64.45, 64.55 (2d, J ) 29.7, 29.1

Hz, -O-CH₂-), 80.43 (-C(CH₃)₃), 114.61, 114.78 (vinyl),

phate) (Sigma) (265 mg, 0.6 mmol) in CH₂Cl₂(10 mL) were

stirred at room temperature under nitrogen. The reaction was

(15) Oleksyszyn, J .; Subotkowska, L; Mastalerz, P. Synthesis 1979,

985-986.

(16) Kafarski, P.; Lejczak, B.; Szewcyk, J . Can. J . Chem. 1983, 61,

2425.

(17) Prepared by reaction of tartaric acid and benzoyl chloride as

described by: Butler, C. L.; Cretcher, L. J . Am. Chem. Soc. 1933, 55,

2605.

(18) Castro, B.; Evin, G.; Salve, C.; Seyer, R. Synthesis 1977, 413.

Enantioselective Ester Hydrolysis

J . Org. Chem., Vol. 65, No. 13, 2000 4013

Sch em e 4

Sch em e 5

+

–

121.00, 121.03, 121.07, 126.75, 126.94, 127.11, 127.26, 127.71,

128.30, 128.85, 128.93, 135.01, 135.05 (-Ph, -PhIm), 135.90

(vinyl), 142.04 (-CO-). HRMS calcd for C₂₆H₃₂N₃O₅P: 497.2080.

Found: 497.2065. Anal. Calcd for C₂₆H₃₂N₃O₅P: C 62.7; H 6.5;

N 8.4; P 6.2. Found: C 60.6, H 6.4, N 8.3; P 6.5.

S-Eth yl, 2-(2′-Im id a zolyl)-4-eth en ylp h en yl 1-Am in o-2-

p h en yleth ylp h osp h on a te (A2). A1 (200 mg, 0.4 mmol) was

dissolved in an ice cold mixture of freshly distilled trifluoro-

acetic acid (TFA, distilled in the presence of TFA anhydride

0.05%) and CH₂Cl₂(1:1, v/v) and stirred for 30 min at room

temperature. The TFA and CH₂Cl₂were evaporated and the

resulting oil taken down four times in anhydrous CH₂Cl₂and

one time in ether, giving 210 mg (87%) of the TFA salt of A2

as a light brown solid. This can be converted to the free base

by rapid extraction in cold CH₂Cl₂/NaHCO₃(aq) (saturated).

For the TFA salt: [R]²³) +4.35° [c 2, CH₂Cl₂]. TLC (CHCl₃/

D

MeOH, 9:1) R_f) 0.5 (two close spots colored purple by

1

ninhydrin). H NMR (CDCl₃+ 1 drop CD₃OD): δ 0.83, 1.05,

4014 J . Org. Chem., Vol. 65, No. 13, 2000

Sellergren et al.

Sch em e 6

3

1.19, 1.27, (3t+1m, J ) 7.0, 6.8, 7.0 Hz, 3H, -O-CH₂-CH₃,

I_1/2/3) 50/32/18), 3.21-3.33 (m, 2H, -O-CH₂-), 3.55-3.80

(2m, 2H, -CH₂-Ph), 3.95-4.20 (m, 1H, -CH-P), 5.35 (d, J

) 10.9 Hz, 1H, vinyl), 5.81 (d, J ) 17.6 Hz, 1H, vinyl), 5.8-

7.0 (bs, 5H, exchangeable), 6.66 (dd, J ) 10.9, 17.6 Hz, 1H,

vinyl), 7.22-7.87 (m, 10H, Ph, Im). ¹³C NMR (CDCl₃+ 1 drop

CD₃OD): δ 15.39, 15.46, 15.52, 15.59 (-O-CH₂-CH₃), 33.88,

33.93, 33.95, (-CH₂-Ph), 48-50 (d, C-P), 65.69, 65.78, 65.90

(-O-CH₂-), 108.65, 113.12, 115.98, 116.03, 116.75, 116.85,

117.17, 118.22, 120.06, 121.93, 122.81, 124.78, 127.76, 127.89,

127.94, 128.62, 128.90, 128.96, 129.11, 129.13, 129.28, 130.28,

130.86, 131.14, 133.29, 133.51, 133.63, 133.71, 134.67, 136.43,

136.73, 139.96, 140.00 (aromatic), 144.81, 144.95 (-CO-),

) 66/15/19), 2.85 (m, 2H, -CH₂-Im), 3.53-3.63 (m, 2H,

-CH₂OH), 4.20-4.22 (m, 1H, -CH<), 5.33 (s, 1H, zH-vinyl),

5.62 (s, 1H, eH-vinyl), 6.87 (s, 1H, H4-Im), 7.63 (s, 1H, H2-

Im). ¹³C NMR (CD₃OD + D₂O): δ 18.85 (-CH₃), 28.97 (-CH₂-

Im), 52.84 (-CH<), 64.08 (-CH₂OH), 118.54 (H₂Cd), 120.66

(C4-Im), 134.93 (C5-Im), 135.97 (dC(CH₃)-CO), 141.19 (C2-

Im), 171.42 (-CO-). HRMS calcd for C₁₀H₁₅N₃O₂: 209.1164.

Found: 209.1150.

Tem p la te Syn th esis (Ser ies B). S-O-(N-ter t-Bu toxy-

ca r bon yl-^L-p h en yla la n yl)-2-(N-m eth a cr yloyla m in o)-3-(5′-

im id a zolyl)p r op a n ol (B1). tert-Butoxycarbonyl-^L-phenyl-

alanine (133 mg, 0.5 mmol), dicyclohexylcarbodiimide (DCCI)

(124 mg, 0.6 mmol), B4 (105 mg, 0.5 mmol), and (dimethyl-

amino)pyridine (DMAP) (12 mg, 0.1 mmol) were dissolved in

order in DMF (5 mL) and allowed to react overnight. After 15

min a precipitate was formed. After 12 h, additional DCCI was

added (42 mg, 0.2 mmol) and the reaction allowed to proceed

for an additional 6 h. EtOAc (50 mL) was added followed by

extraction with 1 M NaHCO₃, drying of the organic phase

(MgSO₄), and evaporation, giving 0.244 g of crude product as

an oil. Purification by column chromatography (silica, eluent:

CHCl₃/MeOH, 2:1), removing unreacted DCCI adduct of BOC-

155.18, 160.38, 160.85 (CF₃CO). FAB-MS calcd for C₂₁H₂₄

-

N₃O₃P: 398.1633, Found: 398. Anal. Calcd for C₂₅H₂₆F₆N₃O₇P

+ 6H₂O: C 42, H 3.66, N 5.87, F 15.94, P 4.33. Found: C 41.98,

H 3.98, N 5.24, F 15.02, P 4.11.

Mon om er Syn th esis (Ser ies B) (Sch em e 7). S-2-Amino-

3-(5′-imidazolyl)propanol + 2HCl (9) was prepared from ^L-

histidine methyl ester‚2HCl (8), which was prepared as

described by Davis.¹⁹The carboxylic group was reduced using

sodium borohydride and the product isolated as the hydro-

chloride salt.²⁰

L-Phe, gave 64 mg (28%) of B1 as a clear oil. [R]²³) -2.37°

D

1

S-2-(N-Meth a cr yloyla m in o)-3-(5′-im id a zolyl)p r op a n ol-

(N-m eth a cr yloyl-^L-h istid in ol) (B4). To 9 (1.3 g, 6.1 mmol)

in NaOH(aq) (20 mL, 1 M) was added dropwise a solution of

methacryloyl anhydride (0.9 mL, 16.1 mmol) in dioxane (5 mL).

After 1 h stirring, additional anhydride (0.5 equiv, 3.05 mmol)

and NaOH (6 mL, 1 M) were added. The reaction could be

followed on TLC (silica plates, eluent: CHCl₃/MeOH/H₂O/

HOAc, 6:4:1:1) as the disappearance of the ninhydrin-active

spot of 9 (R_f) 0.3) and the appearance of a UV-active spot (R_f

) 0.6). After 7 h, the pH was adjusted to 2-3 and the solution

extracted with ether. Then the pH was adjusted to 7 and the

solution lyophilized. The product was dissolved in ethanol, and

the salts were removed by filtration. Evaporation gave crude

B4 as an oil that was purified by column chromatography on

silica (eluent: MeOH). This gave 1.03 g (81%) of B4 as a clear

oil that crystallized slowly to a hygroscopic white solid. This

can be recrystallized from THF/MeOH. Mp: 144-148 °C

[c 2, CHCl₃]. H NMR (CDCl₃): δ 1.42 (s, 9H, -C(CH₃)₃), 1.97

(s, 3H, CH₃-Cd), 2.85 (m, 2H, -CH₂-Ph), 3.08 (m, 2H,

-CH₂-Im), 4.10 (m, 2H, -CH₂-O-), 4.48 (m, 2H, -CH<),

5.23 (d, J ) 7.4 Hz, 1H, -NH-), 5.36 (s, 1H, H₂Cd), 5.78 (s,

1H, H₂Cd), 6.83 (s, 1H, H5-Im), 7.17-7.35 (m, 5H, Ph), 7.42

(bs, 1H, H1-Im), 7.57 (s, 1H, H2-Im). ¹³C NMR (CDCl₃): δ

18.50 (H₃C-Cd), 47.97 (-CH₂-Im), 54.73 (-CH<His), 64.80

(-CH₂-O-), 64.84 (-CH<Phe), 80.08 (-C(CH₃)₃), 120.34

(C4-Im), 127.00 (C4-Ph), 128.48 (H₂Cd), 128.56 (C3-Ph),

129.12 (C2-Ph), 129.28 (C1 Ph), 135.08 (dCCH₃CO), 135.92

(C5-Im), 139.38 (C2-Im), 155.33 (-CO- BOC), 168.26 (-CO-

amide), 171.98 (-CO- ester). HRMS calcd for C₂₄H₃₂N₄O₅:

456.2372. Found: 456.2357.

S-O-(^L-P h en yla la n yl)-2-(N-m et h a cr yloyla m in o)-3-(5′-

im id a zolyl)p r op a n ol (B2). B1 (25 mg, 55 µmol) was dis-

solved in ice cold CHCl₃/TFA, 1:1 (v/v) (4 mL), and left to react

at room temperature for 30 min. After evaporation the

resulting oil was taken down twice in CHCl₃. Addition of ether

resulted in a white precipitate. The ether was removed with

pasteur pipet and the solid dried, 27 mg (84%). The free base

could be obtained by extraction into EtOAc with 1 M NaHCO₃.

Drying (MgSO₄) and evaporation gave the free base as an oil,

(polymerization). [R]²³) -16.9° [c 2, MeOH]. ¹H NMR

D

(CD₃OD + 1 drop D₂O): δ 1.88, 1.90, 1.92 (3s, 3H, -CH₃, I_1/2/3

(19) Davis, N. C. J . Biol. Chem. 1956, 223, 935.

(20) Bentley, K. W.; Crease, E. H. Org. Prep. Proc. Int. 1973, 5, 5.

Enantioselective Ester Hydrolysis

J . Org. Chem., Vol. 65, No. 13, 2000 4015

Sch em e 7

which gave a purple ninhydrin reaction on TLC (silica plates,

eluent: CHCl₃/MeOH/HOAc/H₂O, 6:4:1:1, R_f) 0.6). Charac-

terization of the TFA salt of B2: [R]²³_D) -14.7° [c 1.35,CHCl₃/

MeOH, 3:1]. ¹H NMR (CDCl₃): δ 1.80 (s, 3H, -CH₃), 2.86-

3.06 (m, 2H, -CH₂Ph), 3.23 (m, 2H, -CH₂-Im), 4.00-4.45 (m,

9H, -CH₂-O-, -CH<, -NH₃+, ImH2+), 5.29 (s, 1H, H₂Cd),

5.60 (s, 1H, H₂Cd), 7.15-7.32 (m, 6H, Ph, H4-Im), 8.37 (s,

1H, H2-Im). ¹³C NMR (CDCl₃): δ 14.89 (-CH₃), 25.45 (-CH₂-

Im), 36.25 (-CH₂-Ph), 48.73 (-CH<His), 65.79 (-CH₂-

O-), 66.68 (-CH<Phe), 116.74, 120.80, 127.85, 128.94, 129.11,

130.46, 132.28, 133.34, 138.66 (Ph, Im, vinyl), 167.79 (-CO-

NH-), 169.48 (-COO-). FAB-MS calcd for C₁₉H₂₄N₄O₃: 356.

Found: 356.2.

10 Hz, -NH-, I_1/2) 76/24), 7.05-7.38 (m, 15H, -Ph). ¹³C

NMR (CDCl₃): δ 27.27, 27.62, 28.00 (-C(CH₃)₃), 35.97, 36.03

(-CH₂-Ph), 48.53 (d, J ) 156 Hz, -C-P), 79.91 (-C(CH₃)₃),

120.33, 120.39, 120.50, 120.55, (C2,6-PhO), 125.10, 125.27

(C4-PhO), 126.7, 128.29, 129.26, 129.53, 129.69, (C2,6-Bzl),

135.91, 136.10 (C1-Bzl), 149.87, 150.00, 150.19, 150.32 (C1-

PhO), 154.70, 154.80 (CO-). HRMS calcd for C₂₅H₃₈NO₅P:

453.1705. Found: 453.1687.

Dieth yl (1-(N-ter t-Bu toxyca r bon yla m in o)-2-p h en yl)-

eth ylp h osp h on a te (3RS). 10 (5.18 g, 11.3 mmol) dissolved

in ethanol (30 mL) was added to a solution of sodium (1.3 g,

57 µmol) in ethanol (30 mL). The reaction was allowed to

proceed at room temperature for 2 h. Then ether was added

to 500 mL and extraction done with 1 M NaOH and brine.

After drying (MgSO₄) and evaporation 3RS was obtained as a

light yellow oil (3.6 g, 89%). ¹H NMR (CDCl₃): δ 1.18, 1.31

(m, 15H, -C(CH₃)₃, -CH₂-CH₃), 2.85-2.93 (m, 1H, -CH₂-

Ph), 3.15-3.28 (m, 1H, -CH₂-Ph), 4.05-4.20 (m, 4H, -CH₂-

CH₃), 4.25-4.45 (m, 1H, -CH<), 4.74 (d, J ) 10.1 Hz, -NH-

), 7.15-7.31 (m, 5H, -Ph). ¹³C NMR (CDCl₃): δ 16.25, 16.32,

16.38 (-CH₂-CH₃), 28.08 (-C(CH₃)₃), 36.06, 36.12 (-CH₂-

Ph), 47.72 (d, J ) 156.1 Hz, C-P), 62.31, 62.62, 62.71 (-O-

CH₂-), 79.81 (-C(CH₃)₃), 126.57 (C4-Ph), 128.26 (C3,5-Ph),

129.20 (C2,6-Ph), 136.59, 136.77 (C1-Ph), 154.87, 154.96

Tem p la te Syn th esis (Ser ies C) (Sch em e 8). Dip h en yl

(1-(N-ter t-Bu t oxyca r b on yla m in o)-2-p h en yl)et h ylp h os-

p h on a te (10). A solution of B (3.4 g, 9.6 mmol) (obtained as

intermediate in the synthesis of 1R and 1S) and BOCOBOC

(2.30 g, 10.5 mmol) in THF (50 mL) was refluxed overnight.

The reaction was followed on TLC (silica plates, eluent: CHCl₃/

MeOH, 9:1) with a product spot at R_f) 0.8. Additional

BOCOBOC was added (0.44 g, 2 mmol) and the reaction

allowed to continue for an additional 5 h. After filtration and

evaporation the residue was dissolved in CHCl₃(50 mL) and

washed with 1 M HCl, 1 M NaHCO₃, and brine. The organic

phase was dried (MgSO₄) and evaporated, giving 5.3 g of crude

10 as an oil that solidified overnight to a light yellow solid.

This was recrystallized in hexane, giving 3.25 g (75%) of 10

(-CO-). HRMS calcd for

357.1722.

C₁₇H₂₈NO₅P: 357.1705. Found:

H, Eth yl (1-(N-ter t-Bu toxyca r bon yla m in o)-2-p h en yl)-

eth ylp h osp h on a te (4RS). A solution of 3RS (3.6 g, 10 mmol)

in 2 M NaOH (15 mL) and ethanol (30 mL) was heated at 70

°C overnight. Water was added and the solution extracted with

ether. The aqueous phase was acidified to pH 1.2 and the

1

as a white solid. Mp: 100-103 °C. H NMR (CDCl₃): δ 1.12,

1.27, 1.50 (3s, 9H, -C(CH₃)₃, I_1/2/3: 17/70/13), 2.80-3.10 (m,

1H, -CH₂-Ph), 3.30-3.45 (m, 1H, -CH₂-Ph), 4.35-4.60,

4.60-4.88 (2m, 1H, -CH<, I_1/2) 24/76), 5.19, 5.42 (2d, J )

4016 J . Org. Chem., Vol. 65, No. 13, 2000

Sellergren et al.

Sch em e 8

product extracted into EtOAc. After drying (MgSO₄) and

evaporation, 2.48 g (75%) of 4RS was obtained as a white solid.

Mp: 130-132 °C. ¹H NMR (CDCl₃): δ 1.20-1.35 (m, 12H,

-CH₃), 2.77-2.89 (m, 1H, -CH₂-Ph), 3.17-3.29 (m, 1H,

-CH₂-Ph), 4.08-4.21 (m, 2H, -CH₂-CH₃), 4.30-4.42 (bm,

vinyl, I_1/2) 56/44), 5.65, 5.17 (2d, J ) 17.6 Hz, 1H, vinyl, I_1/2

) 42/58), 6.65, 6.66 (2dd, J ) 11.0, 10.8, 17.5 Hz, 1H, vinyl),

7.13-7.43 (m, 9H, Ph). ¹³C NMR (CDCl₃): δ 15.91, 15.98

(-CH₃), 27.75, 27.78 (-C(CH₃)₃), 35.66, 35.73, 35.79 (-CH₂-

Ph), 47.74, 48.03 (2d, J ) 158.1, 157.4 Hz, C-P), 63.11, 63.21

(2d, J ) 22.1 Hz, -O-CH₂-), 79.47, 79.57 (-C(CH₃)₃), 113.34,

113.45, 113.52 (vinyl), 120.07, 120.12, 120.23, 120.28 (C2,6-

p-vinylphenol), 127.99 (C3,5-Ph), 128.97 (C2,6-Ph), 134.09,

134.27 (C4-p-vinylphenol), 135.38 (C1-Ph), 135.98, 136.17

(vinyl), 149.32, 149.36, 149.44 (C1-p-vinylphenol), 154.49,

154.52, 154.59, 154.61 (-CO-). HRMS calcd for C₂₃H₃₀NO₅P:

431.1861. Found: 431.1853.

1H, -CH<), 5.17, 5.95 (2bd, J ) 9.0 Hz, 1H, -NH-, I_1/2

)

2.4). ¹³C NMR (CDCl₃): δ 16.11, 16.19, 16.28 (-CH₂CH₃),

27.75, 27.86, 28.03 (-C(CH₃)₃), 35.64, 35.69, 35.85, 35.89, 35.92

(-CH₂-Ph), 47.89, 50.53 (2D, J ) 158.1, 155.8 Hz, C-P),

62.29, 62.34, 62.52, 62.61 (-O-CH₂-CH₃), 79.73, 80.85,

(-C(CH₃)₃), 126.43 (C4-Ph), 128.17 (C3,5-Ph), 129.13 (C2,6-

Ph), 136.60, 136.79, 137.16 (C1-Ph), 155.08, 155.17, 155.97

(-CO-). HRMS calcd for C₁₅H₂₄NO₅P: 329.1392. Found:

329.1383.

S-Eth yl, 4-Eth en ylp h en yl (1-(N-ter t-Bu toxyca r bon yl-

a m in o)-2-p h en yl)eth ylp h osp h on a te (C1S) (BOP P r oce-

d u r e). Prepared as described for C1 but starting from 4S.

Eth yl, 4-Eth en ylp h en yl (1-(N-ter t-Bu toxyca r bon yla m i-

n o)-2-p h en yl)eth ylp h osp h on a te (C1). 4-Vinylphenol was

prepared in 50.8% yield as described by Overberger et al.²¹by

decarboxylation and sublimation of p-hydroxycinnamic acid.

4RS (0.66 g, 2 mmol), p-vinylphenol (0.24 g, 2 mmol), triethyl-

amine (0.56 mL, 4 mmol), and Castros BOP reagent¹⁸(Sigma)

(0.88 g, 2 mmol) were dissolved successively in CH₂Cl₂(15 mL).

The reaction was followed on HPLC (RP-18 column, eluent:

MeOH/[potassium phosphate buffer, 0.05 M, pH 3], 4:3 (v/v)).

After 20 h an additional 0.5 equiv of BOP and triethylamine

were added, and the reaction was left overnight, leading to

high conversion according to the HPLC analysis. Brine (50 mL)

was added and the product extracted into EtOAc. The organic

phase was washed with 1 M HCl, NaHCO₃(saturated), and

brine. After evaporation 0.81 g of a dark brown solid was

obtained. The product was purified by column chromatography

(silica, eluent: CHCl₃/MeOH, 9:1), giving 0.88 g (100%) of C1

as a red green oil. ¹H NMR (CDCl₃): δ 1.15-1.30 (m,3H,

-CH₂-CH₃), 1.28, 1.32 (2s, 9H, -C(CH₃)₃), I_1/2) 50/50), 2.88-

3.00 (m, 1H, -CH₂-Ph), 3.24-3.35 (m, 1H, -CH₂-Ph), 4.15-

4.27 (m, 2H, -O-CH₂-CH₃), 4.50-4.60 (m, 1H, -CH<), 5.05-

5.15 (m, 1H, -NH-), 5.20, 5.22 (2d, J ) 11.0, 10.9 Hz, 1H,

Yield: 37 mg (57%). [R]²³) +10.2° [c 2, CHCl₃]. ¹H NMR

D

(CDCl₃): δ 1.13-1.26 (m, 3H, -CH₂CH₃), 1.28, 1.32 (2s, 9H,

-C(CH₃)₃, I_1/2) 43/57), 2.85-3.00 (m, 1H, -CH₂-Ph), 3.25-

3.34 (m, 1H, -CH₂-Ph), 4.16-4.27 (m, 2H, -O-CH₂-), 4.45-

4.58 (m, 1H, -CH<), 4.83-4.91 (m, 1H, -NH-), 5.21, 5.22

(2d, J ) 10.9 Hz, 1H, vinyl, I_1/2) 30/70), 5.66, 5.67 (2d, J )

17.6 Hz, 1H, vinyl, I_1/2) 42/58), 7.13-7.42 (m, 9H, Ph+p-

vinylphenol).

S-Eth yl, 4-Eth en ylp h en yl (1-(N-ter t-Bu toxyca r bon yl-

a m in o)-2-p h en yl)eth ylp h osp h on a te (C1S) (BOP Cl P r o-

ced u r e).²²4S (50 mg, 0.15 mmol), BOPCl (Aldrich) (bis(2-oxo-

3-oxazolidinyl)phosphinic chloride) (42 mg, 0.17 mmol), tri-

ethylamine (0.046 mL, 0.33 mmol), and p-vinylphenol (18 mg,

0.15 mmol) were dissolved successively in CH₂Cl₂(1 mL) and

stirred at room temperature overnight. The reaction was

followed by TLC (silica plates, CHCl₃/MeOH, 9:1). After

addition of an additional 1 equiv of triethylamine and BOPCl

and again stirring overnight a high conversion was observed.

The mixture was washed with 0.5 M NaHCO₃, 0.1 M HCl, and

H₂O and the organic phase dried (K₂CO₃) and evaporated. This

gave 52 mg of an oil. This was purified by column chromatog-

(21) Overberger, C. G.; Salamone, J . C.; Yaroslavsky, S. J . Am.

Chem. Soc. 1967, 89, 6231.

(22) Diago-Meseguer, J .; Palomo-Coll, L.; Fernandez-Lizarbe, J . R.;

Zugaza-Bilbao, A. Synthesis 1980, 547.

Enantioselective Ester Hydrolysis

J . Org. Chem., Vol. 65, No. 13, 2000 4017

Ta ble 1. Im p r in ted P olym er Ser ies A a n d B a n d th e

Am ou n t of Tem p la te Rem oved ^a

raphy (silica, eluent: CHCl₃/MeOH, 9:1), giving 51 mg (80%)

of a light yellow oil. [R]²³) +2.0° [c 3.5, CDCl₃]. ¹H NMR

D

(CDCl₃): δ 1.13-1.27 (m, 3H, -CH₂CH₃), 1.28, 1.32 (2s, 9H,

-C(CH₃)₃, I_1/2) 35/65), 2.88-2.93 (m, 1H, -CH₂-Ph), 3.25-

3.34 (m, 1H, -CH₂Ph), 4.16-4.24 (m, 2H, -O-CH₂-), 4.45-

4.60 (m, 1H, -CH<), 4.75-4.82 (m, 1H, -NH-), 5.22 (d, J )

10.9 Hz, 1H, vinyl), 5.67, 5.68 (2d, J ) 17.5 Hz, 1H, vinyl, I_1/2

) 33/67), 6.67 (dd, J ) 17.6, 10.9 Hz, 1H, vinyl), 7.12-7.38

(m, 9H, Ph+p-vinylphenol). ¹³C NMR (CDCl₃): δ 16.21, 16.28

(-CH₂CH₃), 28.08 (-C(CH₃)₃), 36.05, 36.11, 36.19 (-CH₂-Ph),

47.98 (2d, J ) 158 Hz, C-P), 63.41, 63.51 (2d, J ) 16.0 Hz,

-O-CH₂-), 79.9, 80.0 (-C(CH₃)₃), 113.67, 113.76 (vinyl),

120.40, 120.53, 120.58 (C2,6-p-vinylphenol), 126.69 (C4-Ph),

127.31, 127.43 (C3,5-p-vinylphenol), 128.31 (C3,5-Ph), 129.27

(C2,6-Ph), 134.61 (C4-p-vinylphenol), 135.68 (C1-Ph), 136.20,

136.38 (vinyl), 149.58, 149.70 (C1-p-vinylphenol), 154.73,

removal (%)

CHCl₃/MeOH MeOH ∆ CsF OH^-

polymer

template

A1

(1)

(2)

(3)

(4)

P A1

P A2

P A3

P A4

P B1

P B2

P B3

P B4

11

57

23

75

98

89

97

A2

A3 + GlyOEt

GlyOEt

B1

18 (60)^b

34

25 (73)^b

70

9

12

9

B2

B3 + GlyOEt

B4 + GlyOEt

a

The polymers were prepared using the templates shown in

Schemes 2 and 3 by free radical polymerization for 24 h using

AIBN (1 mol %) as photochemical initiator at 7 °C. The monomer

mixture consisted of the template (0.5 mol %), MAA (10 mol %),

and EDMA (90 mol %) in CHCl₃(F_m) volume porogen/(volume

porogen + volume monomers) ) 0.56). In polymers P A3, P A4,

P B3, and P B4 an equimolar amount of glycine ethyl ester

(GlyOEt) (0.5 mol %) was added to the template. The template

splitting after the various treatments is calculated from the

amount of template monomer originally added. After treatment 1

and 2 this was determined with ¹H NMR analysis of the extracts,

while after treatment 3 and 4 it was determined from elemental

analysis on phosphorus. The phosphorus analysis gave the fol-

lowing results. P A1: After treatment 1 and 2 (MeOH ∆) 0.010%,

3 (CsF) 0.0015%, 4 (OH^-) 0.0020%. P A2: 1 and 2 (MeOH ∆)

154.83 (-CO-). HRMS calcd for

Found: 431.1848.

C₂₃H₃₀NO₅P: 431.1861.

Eth yl, 4-Eth en ylp h en yl 1-Am in o-2-p h en yleth ylp h os-

p h on a te (C2). C1 (0.35 g, 0.8 mmol) was dissolved under ice

cooling in TFA/CH₂Cl₂(1:1, v/v) (5 mL) and then stirred for

30 min at room temperature. The residue obtained after

evaporation was taken down in ether four times, giving 0.39

g of a red oil. This was dissolved in CH₂Cl₂(25 mL) and washed

with NaHCO₃(saturated) (2 × 25 mL). The organic phase was

dried (MgSO₄) and evaporated, giving 0.21 g (80%) of C2 as a

yellow oil pure on TLC (silica, CHCl₃/MeOH, 9:1, R_f) 0.5).

This was stored dry in a freezer. The product can also be

isolated as the more stable oxalate salt by slowly adding the

free base as an etheral solution to an ether solution containing

dissolved anhydrous oxalic acid.^{23 1}H NMR (CDCl₃): δ 1.29

(2t, J ) 7.1 Hz, 3H, I_1/2) 50/50), 2.71-2.84 (m, 1H, -CH₂-

Ph), 3.28-3.35 (m, 1H, -CH₂-Ph), 3.39-3.48 (m, 1H, -CH<),

4.15-4.30 (m, 2H, O-CH₂-), 5.23 (d, J ) 11 Hz, vinyl), 7.18-

7.42 (m, 9H, Ph+p-vinylphenol). ¹³C NMR (CDCl₃): δ 16.27,

16.34 (-CH₃), 37.55, 37.62 (-CH₂-Ph), 50.24, 50.31 (2d, J )

153.1 Hz, C-P), 63.18, 63.24 (-O-CH₂-), 113.62 (vinyl),

120.48 (C2,6-p-vinylphenol), 126.71 (C4-Ph), 127.32 (C3,5-p-

vinylphenol), 128.49 (C3,5-Ph), 129.10, 129.12 (C2,6-Ph),

134.33 (C4-p-vinylphenol), 135.59 (C1-Ph), 137.30, 137.50

(vinyl), 149.95, 150.03 (C1-p-vinylphenol). HRMS calcd for

b

0.032%, 3 (CsF) 0.0015%, 4 (OH^-) 0.0092%. Value referring to

splitting of GlyOEt.

4-Eth en ylp h en yl-4-ch lor oben zoa te (C3). To a solution

of p-vinylphenol (1.2 g, 10 mmol) and triethylamine (1.5 mL,

11 mmol) in ice cold THF was added p-chlorobenzoyl chloride

(1.4 mL, 11 mmol) under stirring. The reaction was followed

on TLC (silica, eluent: CH₂Cl₂/hexane, R_f) 0.7). After comple-

tion, purification was done by aqueous washes with 1 M HCl,

saturated NaHCO₃, and brine. After drying (MgSO₄) and

evaporation 1.9 g (74%) of product was obtained as a white

1

solid. Mp: 88-90 °C. H NMR (CDCl₃): δ 5.25, (d, J ) 10.9

Hz, 1H, vinyl), 5.72 (d, J ) 17.6 Hz, 1H, vinyl), 6.71 (dd, J )

17.6, 10.9 Hz, 1H, vinyl), 7.14 (d, J ) 8.6 Hz, 2H, H3′,5′),

7.44,7.45 (2d, J ) 8.6 Hz, 4H, H2′,6′,3,5), 8.11 (d, J ) 8.6 Hz,

2H, H2,6). ¹³C NMR (CDCl₃): δ 114.10 (vinyl), 121.58 (C2′,6′),

127.18, 127.84, 128.85, 131.44, 135.45, 135.75, 140.04, 150.18,

C

₁₈H₂₂NO₃P: 331.1337. Found: 331.1316.

Eth yl, 3(4)-Eth en ylben zyl-(1-(N-ter t-Bu toxyca r bon yl-

a m in o)-2-p h en yl)eth ylp h osp h on a te (11). 4RS (0.33 g, 1

mmol), 3(4)-vinylbenzyl alcohol (0.13 g, 1 mmol) (prepared as

described in the Supporting Information), BOP (0.44 g, 1

mmol), and triethylamine (0.21 mL, 2 mmol) were dissolved

successively in CH₂Cl₂and stirred at room temperature. The

reaction was followed on HPLC (RP18, eluent: MeOH/potas-

sium phosphate buffer 0.05 M, pH 3], 1:1 (v/v), k′(product) )

6.9). After 24 h an additional 0.5 equiv of BOP and triethy-

lamine were added and left again overnight. High conversion

was obtained according to HPLC. The color changed from light

yellow to a red color. Ethyl acetate (50 mL) was added and

the solution washed with brine, 1 M HCl, NaHCO₃, and finally

brine. After drying (MgSO₄) and evaporation 0.56 g of a red

oil was obtained. The product was purified by column chro-

matography (silica, eluent: CHCl₃/MeOH, 9:1). This gave 0.40

g (100%) of 11 as an orange oil. ¹H NMR (CDCl₃): δ 1.15-

1.44 (m, 12H, -CH₃), 1.82-2.88 (m,1H, -CH₂-Ph), 3.15-3.25

(m, 1H, -CH₂-Ph), 4.02-4.18 (m, 2H, -O-CH₂-), 4.37-4.46

(m, 1H, -CH<), 4.85 (d, J ) 10 Hz, -NH-), 5.03-5.15 (m,

2H, -O-CH₂-Ph), 5.27 (d, J ) 10.9 Hz, 1H, vinyl), 5.75, 5.76

(2d, J ) 17.7 Hz, 1H, vinyl), 6.70 (dd, J ) 10.9, 17.6 Hz, 1H,

vinyl), 7.20-7.50 (m, 9H, Ph). ¹³C NMR (CDCl₃): δ 15.92,

15.99, 16.04 (-CH₂-CH₃), 27.81 (-C(CH₃)₃), 35.69, 35.73

(-CH₂-Ph), 47.60 (d, J ) 156.1 Hz, C-P), 62.32, 62.47, 62.55

(-O-CH₂-CH₃), 67.19, 67.27 (-O-CH₂-Ph), 79.53 (-C(CH₃)₃),

114.13 (vinyl), 125.43, 125.49, 125.92, 126.06, 126.34, 126.99,

127.05, 127.99, 128.48, 128.92, 135.94, 136.02, 136.24, 136.42,

137.54 (aromatic+vinyl), 154.62, 154.70 (-CO-). HRMS calcd

for C₂₄H₃₂NO₅P: 445.2018. Found: 445.2037.

164.16. HRMS calcd for

258.0442.

C₁₅H₁₁ClO₂: 258.0447. Found:

Molecu la r Im p r in tin g of ^L-P h en yla la n in e Eth yl Ester

(13) a n d S-Dieth yl-1-Am in o-2-P h en yleth ylp h osp h on a te

(12S). The templates were converted to the free amine by

extraction into ethyl acetate in bicarbonate, and 0.8 mmol of

the oils was dissolved in 4.6 mL of methylene chloride. To this

was added 3.1 mL (16.5 mmol) of EDMA, 0.28 mL (3.3 mmol)

of MAA, and 0.034 g (0.2 mmol) of AIBN. The homogeneous

solutions were then transferred to polymerization tubes and

subjected to three freeze-thaw-degas cycles before sealing.

They were then photopolymerized as described below at 15

°C for 20 h. The tubes were turned 180° after 10 min, 40 min,

and 10 h to achieve an even exposure. After polymerization

the transparent polymers were crushed and Soxhlet extracted

in MeOH for 24 h followed by sieving (25-38 µm) and packing

into HPLC columns (10 cm × 0.5 cm). The chromatographic

performance was then investigated at room temperature using

HPLC at 1 mL/min and MeCN/H₂O/HOAc, 96.25:1.25:2.5, as

eluent and detection at 267 nm if not otherwise mentioned.

The recovery of template after Soxhlet extraction was deter-

mined using ^L-phenylalanine anilide as internal standard on

RP-18 column with eluent: MeOH/3% HOAc, 1:1 (v/v).

Im p r in ted P olym er s: Ser ies A a n d B. P olym er P r ep a -

r a tion . The polymers were prepared using the templates

indicated in Table 1 and shown in Schemes 1-3. In a typical

preparation (P A1) template A1 (50 mg, 0.1 mmol) was dis-

solved in CHCl₃(3.76 mL) and MAA (0.17 mL, 2 mmol), and

EDMA (3.39 mL, 18 mmol) and the initiator AIBN (33 mg,

(23) Kowalik, J .; Kupezyk-Subotkowska, L.; Mastalerz, P. Synthesis

1981, 57.

4018 J . Org. Chem., Vol. 65, No. 13, 2000

Ta ble 2. Im p r in ted P olym er Ser ies C a n d th e Am ou n t of Tem p la te Rem oved ^a

splitting (%)

MeOH/NaOH(aq) (2)

Sellergren et al.

monomers (mol %)

CsF (1)

polymer

templ

VIm

MAA

EDMA

NMR 40 h

HPLC 17 h

NMR 3.5 h

78

NMR 17 h

Anal. 17 h

PC11

PC12

PC21

PC22

PC31

PC32

C1

C2

C3

12

6

12

6

12

6

87

19

10

100

89

100

>60

6

72

51

55

56

93

69

68

a

The polymers were prepared as described in the Experimental Section by free radical polymerization of the indicated monomer mixtures

in the presence of template (1 mol %) using benzene (F_m) 0.56) as porogen and AIBN (1 mol %) as photochemical initiator. VIm )

vinylimidazole, MAA ) methacrylic acid, EDMA ) ethylene glycol dimethacrylate. Template splitting was carried out by treatment with

CsF in MeOH (20 mg/mL) at 60 °C (treatment 1) or MeOH/10 M NaOH, 4:1 (v/v), (treatment 2) overnight. The splitting was followed by

¹H NMR, HPLC, or elemental analysis on phosphorus (Anal.) as described in the Experimental Section and was calculated from the

amount of template added as monomer.

0.2 mmol) in 1 mL of CHCl₃were added followed by transfer

of the mixture to a thick-walled glass tube. Oxygen was

removed by two freeze-thaw cycles (or by helium bubbling)

and the tube sealed under vacuum. The polymerization was

then started by placing the tubes at ca. 10 cm distance from a

medium-pressure mercury vapor lamp (Canrad Hanovia, 550

W, 33 W in λ range 320-400 nm) and allowed to proceed at 7

°C. The tubes were turned at regular intervals for symmetric

exposure and left for 24 h. In the preparation of P A2 the poor

stability of template A2 made it necessary to synthesize it just

prior to polymerization. It was then rapidly converted to the

free base form by cold extraction into CHCl₃with 1 M

NaHCO₃. After drying (MgSO₄) a slightly colored oil was

(m, 1H, -CH<), 7.1-7.4 (m, 5H, -Ph). Template A3: δ 7.2-

7.8 (m, 10H, Ph). GlyOEt: δ 1.1-1.3 (t, 3H, -O-CH₂-CH₃),

3.5-4.0 (m, 2H, -CH₂-), 4.0 (q, 2H, -OCH₂-). Templates B1

and B2: δ 1.5 (s, 9H, tBut), 2.5-3.5 (m, 2H, -CH₂-Ph), 3.5-

4.5 (m, 1H, -CH<), 7.0-7.5 (m, 5H, -Ph). Template B3: δ

7.9-8.1 (m, 2H, H2,6-Ph).

Elem en ta l An a lysis. Polymer samples were dried for 24

h in a vacuum oven at 50 °C and then submitted for P-analysis.

The splitting percentage was calculated based on a total

incorporation of template corresponding to 0.081% P.

Sp ectr oscop ic Ch a r a cter iza tion of P olym er s. The poly-

mers subjected to template splitting procedure (4) were

1

characterized by IR and H NMR spectroscopy. FTIR (KBr) of

1

obtained, appearing pure on H NMR and TLC. This was taken

P A1: 3530 (br), 2977, 1739, 1637, 1468, 1394, 1261, 1155, 947,

869, 814, 753, 654, 597, 520, 468. P B4: 3446 (br), 2973, 1734,

1637, 1470, 1394, 1261, 1160, 949, 869, 814, 752, 653, 465.

CP-MAS ¹³C NMR of P A1: δ 177, 167, 137, 127, 63, 55, 46,

24, 19 ppm. Level of unsaturation: I₁₆₇/I₁₇₇: P A1 10%, P A2

14%, P A3 11%.

directly into the polymerization, where in order to save time

the degassing was done by He bubbling. Blank samples lacking

initiator were used in order to check stability of the templates

by TLC (CHCl₃/MeOH, 9:1) under the polymerization condi-

tions.

Wor k u p a n d Tem p la te Sp littin g. The polymers were

crushed and passed through a 250 µm sieve, and the splitting

of templates was attempted in order as follows: (1) The

polymers were placed in flasks and shaken with 100 mL of

CHCl₃/MeOH, 1:1 (v/v), for 5 h followed by filtration into

Soxhlet extraction timbles. The filtrate was evaporated and

saved for analysis. (2) Soxhlet extraction was performed in

MeOH overnight. The extract was evaporated and saved for

analysis. Then either of the following procedures were fol-

lowed: (3) To 0.5 g of polymer was added 5 mL of a solution of

CsF (32 mg/mL) in MeOH, and the mixture shaken at room

temperature and then heated at 60 °C on a hot plate for 48 h.

The supernatant was removed (saved for analysis), the poly-

mers were washed with MeOH and dried, and the MeOH wash

was combined with the supernatant, evaporated, and saved

for analysis. (4) To 0.5 g of polymer was added 10 mL of

Na₂CO₃(0.5 M)/MeOH, 1:1 (v/v), and the mixtures were

shaken overnight. The supernatant was removed and the

polymers were washed and dried as described under (3). (5)

To 0.5 g of polymer was added 10 mL of NaOH (1 M)/MeOH,

1:1 (v/v), and the mixtures were shaken overnight at either

room temperature or 60 °C. The supernatant was removed and

the polymers were washed and dried as described under (3).

All wash solutions were combined, evaporated, and saved for

analysis. The analysis was done by TLC, ¹H NMR, and

elemental analysis on phosphorus.

Im p r in ted P olym er s: Ser ies C (Sch em e 4). The poly-

mers were prepared using the monomer mixtures shown in

Table 2. Vinylimidazole was prepared according to Overberger

et al.²⁴In a typical preparation (P C11) C1 (86 mg, 0.2 mmol),

benzene (3.7 mL), MAA (0.20 mL, 2.4 mmol), and EDMA (3.28

mL, 17.4 mmol) were added to a 50 mL thick-walled glass tube.

After ensuring complete solubility of the template, a small

sample of this mixture was removed in order to check template

stability. The initiator, AIBN (33 mg, 0.2 mmol) in benzene (1

mL), was added. Degassing was done by two freeze-thaw

cycles and the polymerization started by placing the polymers

at a 10 cm distance from a medium-pressure mercury vapor

lamp (Canrad Hanovia, 550 W, 33 W in λ range 320-400 nm)

in a thermostated water bath (15 °C). The tubes were turned

at regular intervals for symmetric exposure and left for 24 h.

The template stability was checked by TLC (CHCl₃/MeOH, 9:1)

on the solution samples lacking initiator. No side reactions

could here be detected.

Ch oice of Sp littin g Con d ition s: Mod el Stu d y. Ba se-

Ca ta lyzed Rea ction . To samples of either template C1,

benzyl ester 11, or diethyl p-vinylphenylphosphate (20 µmol)

was added, MeOH/[NaOH 1 M], 1:1 (1 mL), and the hydrolysis

was followed by HPLC (RP-18, MeOH/[potassium phosphate

buffer 0.05 M, pH 3], 4:3 (v/v)).

Acid - a n d TBAF -Ca ta lyzed Rea ction . To samples (20

µmol) of templates C2, C3, diethyl 1-amino-3-phenylethylphos-

phonate (12), and EDMA was added MeOH/[HCl 2 M], 1:1 (v/

v), or 1 mL of tetrabutylammonium fluoride (TBAF) (0.5 M)

in THF/MeOH, 1:1 (v/v). The reactions were followed on HPLC

(RP-18, MeOH/[potassium phosphate buffer 0.05 M, pH 7.5],

4:3 (v/v)).

An a lysis of Sp littin g P r od u cts. TLC. This was done on

silica gel plates in CHCl₃/MeOH, 9:1, with UV and ninhydrin

development.

¹H NMR. To the oily residues were added CDCl₃and a few

drops of CD₃OD for solubilization. Then 1 equiv (based on

100% splitting) of dioxane was added as internal standard,

and the integrals of the template characteristic signals were

compared with the dioxane signal at δ 3.7 ppm corresponding

to 8H. The signals that were used in the comparison were for

templates A1 and A2: δ 1.25 (m, 3H, -O-CH₂-CH₃), 2.7-

3.3 (m, 2H, -CH₂-Ph), 4.0-4.3 (m, 2H, -O-CH₂-), 4.2-4.5

CsF -Ca ta lyzed Rea ction . To templates C1 or C3 (20 µmol)

was added cesium fluoride (CsF) (60 µmol) in dry MeOH (4

(24) Overberger, C. G.; Vorheimer, N. J . Am. Chem. Soc. 1963, 85,

951.

Enantioselective Ester Hydrolysis

J . Org. Chem., Vol. 65, No. 13, 2000 4019

mL). The reaction was followed on TLC (silica plates: CHCl₃/

MeOH, 9:1) and HPLC (RP-18: MeOH/H₂O, 3:1).

mobile phase was MeCN/[potassium phosphate buffer 0.01 M,

pH 6], 65:35 (v/v), and the flow rate was 1.5 mL/min. The

wavelength was chosen as the λ_maxof PNP, which varies with

pH. Thus at pH 6 detection was done at 310 nm. The peaks of

the substrate BOCPheONP (k′ ) 1.7) and the product (k′ )

0.6) were well separated. A linear response was verified by

injecting various concentrations of BOCPheONP and PNP and

IP for PNP determined. To get a true IP for PNP, the

supernatant was analyzed after several weeks reaction, thus

compensating for nonspecific absorption of PNP to the polymer.

Moreover the UV spectra obtained at this point confirmed that

no change in pH had occurred, which would have resulted in

a change in ꢀ and λ_max. Above pH 7 the reactions were

performed in stoppered quartz cuvettes and followed by

monitoring the increase in UV absorption at 400 nm due to

the formation of PNP. This technique could not be used at low

pH due to interference of fine particles in the spectral window

used here.

Wor k u p a n d Sp littin g of Tem p la te. The polymers were

crushed and Soxhlet extracted in methanol for 24 h. After

drying under vacuum at 50 °C for ca. 10 h they were

mechanically sieved to a particle size of 38-250 µm. The fine

particles were removed by repeated sedimentation in metha-

nol. The amount of template removed was analyzed by HPLC

(RP 18, MeOH/[potassium phosphate buffer 0.05 M, pH 3], 4:3

(v/v)). No template or unreacted monomer was detected in the

methanol extracts.

Tr ea tm en t A: Aqu eou s Ba se in MeOH. To 1.5 g of each

polymer in a 25 mL screwcap vial was added 20 mL of MeOH/

10% NaOH, 4:1 (v/v), followed by gentle shaking overnight.

Polymers P C31 and P C32 immediately turned light yellow.

After 17 h, the supernatant was removed and neutralized with

HCl before analysis. Polymer P C11 was subjected to further

splitting treatment at elevated temperature. The splitting yield

was checked by HPLC and by ¹H NMR analysis. Before the

¹H NMR analysis the polymers were washed as follows: 1 ×

15 mL of MeOH/H₂O, 4:1, 5 × 15 mL of MeOH/[0.1 M HCl],

4:1 (pH acidic), 2 × 15 mL of MeOH/H₂O, 4:1 (pH neutral), 1

× 15 mL of MeOH. The wash solutions were combined with

the hydrolyzate and neutralized followed by evaporation. The

residues were then dissolved in 5 mL of MeOH, and salts were

separated, evaporated, and dissolved in CD₃OD/D₂O contain-

ing 3-trimethylsilylpropane sulfonic acid sodium salt as in-

ternal standard.

Eva lu a tion of th e Ra te Da ta . Since the catalyst is in

excess of the substrate (theoretical amount of sites 0.5 mM in

the PA series and 0.15 mM in PB series; substrate concentra-

tion 0.025 mM), the hydrolysis can be assumed to follow

pseudo-first-order kinetics. This can be expressed in terms of

HPLC peak integrals for either the substrate (IS) or the

product (IP) as

ln IS_t) ln IS_o- kt

(1)

Tr ea tm en t B: Cesiu m F lu or id e (CsF ) in MeOH. To dry

polymer (0.4 g) was added dry MeOH (4 mL) containing CsF

(20 mg/mL). The mixtures were heated to 60 °C, and the

splitting was followed on HPLC (RP-18, MeOH/H₂O, 3:1 (v/

v)) and by ¹H NMR. In the ¹H NMR investigation the

supernatant was removed after 40 h and the polymer washed

repeatedly with MeOH. The wash solutions were pooled with

the CsF fraction, evaporated, and then dissolved in 2 mL of

CDCl₃containing dioxane as internal standard (20 µmol). The

integrals were compared: for dioxane, δ 3.7 (8H); for P C11-

P C22, δ 4-4.3 (2H), 2.6-3.3 (2H); and for P C31,32, δ 7.9-

8.1 (2H).

or

ln(IP_inf- IP_t) ) ln IP_inf- kt

(2)

where k is the first-order rate constant, IS_tand IS_oare the

substrate peak integral at time t and at the start. IP_infis the

product peak integral at infinte time.

Thus by plotting the natural logarithm of the peak integrals

versus time for both the substrate and the product, the pseudo-

first-order rate constants are obtained from the slopes of the

graphs. In the case where the reactions were followed to high

conversion (>3 half-lives) the given k is the average of these

rates. Otherwise since ꢀ_(BOCPheONP)< ꢀ_(PNP), k is most accurately

obtained by following the product formation. Reproducibility

was verified by repeating several runs twice and the error

determined. The lines were drawn based on at least four data

points, and the correlation coefficient was larger than 0.98.

For the hydrolysis of phenylalanine ethyl esters the reaction

was followed by the appearance of the product ^D(L) phenyl-

alanine and also the by the conversion of the ester.

Hom ogen eou s Ca ta lyzed Rea ction s. Homogeneous reac-

tions using the functionalized monomers 7 and B4 as soluble

catalysts were followed. To 2 mL of an MeCN/aqueous mixture

in a stoppered quartz cuvette containing 7, B4 (0.7 mM), or

no catalyst was added substrate (BOCPheONP) to a concen-

tration of 0.025 mM. The amount of added catalyst corresponds

to the theoretical amount of catalytic groups bound to the

polymer. The reaction was performed at room temperature and

followed spectrophotometrically using a Perkin-Elmer Lambda

4A UV-vis spectrophotometer by single-point measurements

at regular time intervals. Prior to start, the spectra of the

substrate, product (PNP), and the catalyst were recorded in

order to choose a wavelength where absorption overlap of these

species is at a minimum. Monomer B4 showed a weakly pH-

dependent λ_max) 220 nm (ꢀ ) 3.0 mM^-1cm^-1). No absorbance

was seen above λ ) 280 nm. Monomer 7 however absorbed

strongly at λ ) 320 nm with a shoulder at λ ) 350 nm.

P olym er -Ca ta lyzed Rea ction s. Dry polymer (50 mg) was

weighed into 4 mL screwcap vials. MeCN or MeCN/aqueous

reaction medium was added and the mixture left for a few

hours for equilibration. After gentle shaking the polymer was

allowed to settle for a few minutes and the supernatant

removed with a pasteur pipet by careful suction from the

bottom of the vial. This procedure was repeated 2-3 times in

order to remove all fine particles and ensure a pH unaffected

by ionization of pendant acid groups of the polymer. Finally

fresh reaction medium was added to a total volume of 2 mL.

If not otherwise stated, the reaction was started by adding 20

µL of a freshly prepared solution of substrate (2.5 mM) in dry

acetonitrile. This gave a final concentration of 0.025 mM. The

vials were gently shaken at room temperature, and the

reaction was followed by HPLC by injecting 20 µL of the

supernatant after settling of the polymer. A reversed-phase

column (C-18), a variable-wavelength UV detector (Waters),

and an integrating recorder were used (Hewlett-Packard). The

Resu lts a n d Discu ssion

Ca ta lyst Design . The templates were designed with

the objective of incorporating the key catalytic elements,

found in the proteolytic enzyme chymotrypsin, into the

polymer active sites. After template removal we antici-

pated that template A1 would provide an active site

complementary to the ^Denantiomer of BOCPheOEt (see

Scheme 1), equipped with a phenolimidazole catalytic

group that would be situated in proximity to the reactive

carbonyl group of a bound substrate. Complementary

hydrogen bonds between carboxylic acid groups of the site

and hydrogen-bonding groups of the substrate are re-

sponsible for the binding selectivity¹²and the tetrahedral

phosphonate function provides a site complementary to

a transition state like structure (vide supra). Note that

an additional carboxylic acid group hydrogen bonded to

the imidazole group²⁵(Scheme 1) completes the catalytic

triad of functional groups found in chymotrypsin.^13,14

Moreover due to the relative basicity of the syn and anti

(25) Lindemann, R.; Zundel, G. J . Chem. Soc., Faraday Trans. 2

1977, 73, 788.

4020 J . Org. Chem., Vol. 65, No. 13, 2000

Sellergren et al.

lone pair of the carboxylic acid group, it is likely to be

positioned syn to the imidazole group, providing an

additional catalytic advantage.²⁶In the control polymer

P A3 however (Scheme 2) the carboxylic acid groups

responsible for substrate binding are expected to be

randomly distributed in the polymer (ion-paired with

GlyOEt), tetrahedral complementarity is absent, and

with the achiral template A3 no stereoselectivity is

expected.²⁷Control polymer P A2, equal to P A1 except

for the absence of the BOC group in the template, we

believed would additionally probe the structural comple-

mentarity of the site, and the simplest control polymer,

P A4, would reveal whether nonspecific catalysis takes

place. In series B polymer catalysts however (Scheme 3)

P B1 is expected to be complementary to BOC-^L-PheONP

with an active site containing a chiral pendant ^L-

histidinole group and carboxylic acid groups positioned

to bind BOC-^L-PheONP. P B2 as well as P A2 will test

the spatial requirement of the substrate, and P B3 and

P B4 the importance of the binding site as well as the

influence of carboxylic acid groups neighboring the imid-

azole group. Finally series C (Scheme 4) was designed

to investigate the importance of the imidazole group

regarding its position and general role in catalysis.

In Schemes 2-4 are seen the three groups of synthetic

template targets and the result of polymerization and

nucleophilic splitting of the template from the polymer.

In each group the first two entries correspond to the

actual mimic and the last entries to blank control

polymers. A convergent synthetic strategy was chosen

where the chiral phosphonate (Scheme 5), racemic phos-

phonate (Scheme 8), and the functional monomers

(Schemes 6 and 7) were first synthesized followed by a

final condensation step and deprotection.

Syn th esis of th e Ch ir a l P h osp h on a te Moiety (4S)

(Sch em e 5). 1R and 1S were synthesized in three steps

as reported elsewhere.^15,16Starting with condensation of

phenylacetaldehyde, benzylcarbamate, and triphenyl-

phoshite catalyzed by acetic acid giving diphenyl-N-

carbobenzyloxy-1-amino-2-phenyl-ethylphosphonate (A),

the latter was deprotected to the free amine (B) and then

reacted with dibenzoyltartaric anhydride, giving the

corresponding N-dibenzoyltartrate (C). The diastereo-

mers of the latter could easily be resolved in two

crystallization steps followed by HBr aqueous hydrolysis

of both the N and PO protecting groups. The amino acids

1R and 1S were finally purified by either cation exchange

chromatography or precipitation with propylene oxide.

The latter procedure was preferable since it was less

time-consuming and gave a higher yield. 1S was N

protected by BOC²⁸in good yield, giving the BOC-

protected P-amino acid 2S (Scheme 5). This was O

alkylated with triethylorthoformate (TEOF) to give the

diethylphosphonate 3S,²⁹which was monohydrolyzed by

base to the acid ester 4S.³⁰The latter was used as a

common precursor for the synthesis of the chiral transi-

tion state analogue template assemblies. A more versatile

synthesis of racemic 4 was developed (Scheme 8). B was

here directly BOC protected, giving 10 followed by

transesterification with sodium ethoxide, giving 3RS, and

monohydrolysis as with the chiral precursor to give 4RS.

As judged from the doubling of some ¹H NMR signals

the above BOC-protected compounds seemed to be present

as at least two stable conformers with a relative abun-

dacy of approximately 5/2.

Syn th esis of th e Hyd r oxyim id a zole m on om er s 7

a n d B4 (Sch em es 6 a n d 7). The functionalized mono-

mers were synthesized as shown in Schemes 6 and 7. The

first attempt to synthesize the phenolimidazole mono-

mer departed from 5-bromosalicylic acid following a

procedure by Rogers and Bruice.³¹The acid was converted

to the methyl ester followed by reaction with ethylene-

diamine to give the imidazoline. Dehydrogenation of the

imidazoline to give the imidazole using Pd/C or barium

manganate failed. Moreover attempts to protect the

phenol hydroxy group by TBDMS, TBDPS-silyl, or MOM

ether formation failed. Instead as shown in Scheme 6,

5-bromosalicylaldehyde was converted to the 2-substi-

tuted imidazole 5 by a condensation reaction with am-

monium acetate and the trimer of glyoxal.³²Direct Pd-

catalyzed coupling of the vinyl group was not possible

probably due to the presence of the basic nitrogen.³³

Benzoyl chloride allowed a facile protection of both the

phenol and imidazole functions, which could not be

achieved with other protecting groups such as acetyl, silyl

(not stable), or MOM ethers (leaves imidazole unpro-

tected). The Pd coupling³³gave A3, which was depro-

tected to give the vinylphenol imidazole monomer 7. In

the ¹³C NMR the signals from the C4 and C5 carbons of

the imidazole ring were broadened, probably because of

slow rotation around the phenylimidazole bond due to

conjugation and intramolecular hydrogen bonding.³¹By

adding a drop of CD₃COOD, the signal was sharpened.

The histidinole monomer B4 (Scheme 7) was synthesized

starting with esterification of ^L-histidine, giving the

methyl ester 8. This was reduced to the alcohol by sodium

borohydride, giving 9,³⁴which was finally N acylated

under Schotten-Bauman conditions with methacrylic

acid anhydride to give B4.

Th e F in a l Cou p lin g Step . Several condensing agents

were tried in the final coupling step (Schemes 5, 7, 8). In

the phosphonate coupling, DCC was inefficient while

Castros BOP reagent¹⁸gave a high yield and a fast

reaction. In the coupling of the phosphonate precursor

4S with p-vinylphenol (see Scheme 8 for coupling with

4R,S) the use of BOP and BOPCl as condensing agents

was compared. While both reactions proceeded in a high

yield, the optical purity of the BOPCl reaction was lower.

Apparently some racemization occurred here. The BOP

reagent was therefore used in all couplings of the

phosphonate precursor. The spectra of the phosphonate

products A1 and A2 of Scheme 5 were complicated by

the existence of multiple conformers, by the presence of

two diastereomers due to the newly formed stereogenic

(26) Gandour, R. D. Bioorg. Chem. 1981, 10, 169.

(27) Note that A1 is a better hydrogen bond acceptor than A3, both

in terms of the carbamate and phosphonate (A1) versus the ester group

(A3) of the binding site and in terms of the imidazole (A1) versus the

benzoylimidazole (A3) group of the cataytic group (see ref 25).

(28) Of the common N-protecting groups available, the acid labile

BOC group is suitable here. FMOC protection proceeded without

problem, but it was not compatible with the following base treatment,

and Cbz protection was not suitable due to the hydrogenation required

in deprotection which would reduce the vinyl group.

(31) Rogers, G. A.; Bruice, T. C. J . Am. Chem. Soc. 1974, 96, 2463.

(32) For a similar reaction with diketones: Lombardino, J . G. J .

Heterocycl. Chem. 1973, 10, 697.

(33) Krolski, M. E.; Renaldo, A. F.; Rudisill, D. E.; Stille, J . K. J .

Org. Chem. 1988, 53, 1170.

(34) Andersson, L.; Ekberg, B.; Mosbach, K. Tetrahedron. Lett. 1985,

26, 3623.

(29) Bartlett, P. A.; Lamden, L. A. Bioorg. Chem. 1986, 14, 356.

(30) Yamauchi, K.; Kinoshita, M.; Imoto, M. Bull. Chem. Soc. J pn.

1972, 45, 2528.

Enantioselective Ester Hydrolysis

J . Org. Chem., Vol. 65, No. 13, 2000 4021

Ta ble 3. Discr im in a tion betw een Ch ir a l P h osp h on a te- a n d Ca r boxyla te-Con ta in in g Su bstr a tes by Th eir

Com p lem en ta r y P olym er s^a

a

The polymers were prepared as described in the Experimental Section. In the chromatographic evaluation 25 nmol of either enantiomer

was applied using as eluent MeCN/H₂O/HOAc, 96.3:1.2:2.5 (v/v %), and a flow rate of 0.4 mL/min. k′_maxrefers to the capacity factor of the

most retained enantiomer which is complementary to the configuration of the template. a ) separation factor ) k′_max/k′_min

.

center at phosphorus, and finally by CP and HP cou-

plings. Nevertheless adequate characterization by HRMS,

elemental analysis, and NMR indicated a high purity of

the products. A1 and A2 were also hygroscopic and

unstable and had to be kept cool and dry.

Coupling of the phosphonate and the ^L-histidinole

monomer B4 failed. Therefore the phosphonate was

replaced with BOC-^L-phenylalanine and the coupling

with B4 again attempted. Using DCCI as condensing

agent and DMAP as catalyst, B1 could be obtained in

poor yield.

Im p r in tin g w ith Ca r boxyla te (Gr ou n d Sta te) a n d

P h osp h on a te (Tr a n sition Sta te) Tem p la tes. The

mechanism of ester hydrolysis or transesterification is

believed to involve a high-energy tetrahedral oxyanion

intermediate (transition state like). Rate enhancements

caused by esterolytic enzymes can to a large part be

ascribed to eletrostatic stabilization of this structure.^13,14

In some cases the transition state theory has been proven

by a correlation between the enzymatic rate enhancement

and the ratio of binding constants of stable transition

state analogues and the substrate in the ground state.^14,35

As stable transition state (TS) analogues, phosphonates

have been used.

To probe the ability of imprinted polymers to discrimi-

nate between a transition state like structure and a

ground state structure, we therefore imprinted the simple

ethylesters 13L (^L-PheOEt) and 12S (Table 3). Note that

template 12S ressembles the tetrahedral intermediate

formed upon transesterification of 13D with ethanol

(with the exception of a negatively charged oxyanion).

R- and S-diethyl-1-amino-2-phenylethylphosphonate (12R,

12S) (oxalate salts) were synthesized as described else-

where. (Scheme 5).³⁶Although the triethylorthoformate

reaction was relatively unclean, the product could be

purified by precipitation as the oxalate salt. Templated

polymers were prepared by photopolymerization using

either 12S or 13L as templates, as described in the

Experimental Section. The polymers were then evaluated

in the chromatographic mode regarding their ability to

discriminate between the enantiomers of both templates

(Table 3).

The enantiomers of the templates were efficiently

retained and resolved only on their complementary

polymers. This shows that an imprinted polymer can be

prepared that binds a tetrahedral (transition state like)

structure in preference to a planar structure (ground

state). The separation factors observed for the template

on both polymers are very similar despite PO being a

stronger hydrogen bond acceptor than CO. This may

reflect the influence of an intramolecular hydrogen bond³⁷

which is stronger in 12 than in 13 and which in turn

would reduce the extent of intermolecular hydrogen

bonding to MAA. On the other hand this intramolecular

hydrogen bond would limit the number of stable con-

formers which could lead to more well-defined recognition

sites. The absence of discrimination of ^D- and L-13 on the

polymer imprinted with 12S supports this argument.

This polymer also showed weaker binding of its template

than the 13L imprinted polymer, possibly a result of the

weaker intermolecular hydrogen bonding to MAA. The

13L imprinted polymer resolved to some extent 12R,S,

indicating a larger conformational freedom for the tem-

plate 13L. Attempts to use the polymer imprinted with

12S for the catalysis of transesterification of 13 failed.

P r ep a r a t ion of E st er a se-Mim ick in g P olym er s.

P olym er Syn th esis. The polymers were prepared by

free radical photoinitiated polymerization of the monomer

mixtures indicated in Tables 1 and 2 using the templates

(35) For a review see: Kraut, J . Science 1988, 242, 533-540.

(36) Kafarski, P.; Lejczak, B. Synthesis 1980, 307.

(37) Gilmore, W. F.; McBride, M. A. J . Pharm. Sci. 1972, 1087.

4022 J . Org. Chem., Vol. 65, No. 13, 2000

Sellergren et al.

shown in Schemes 1-4. As judged by TLC, the templates

A1-A3 containing the phenolimidazole function were not

completely stable under the polymerization conditions.

The poor stability was also noted during the BOP-

catalyzed synthesis of these derivatives. Since the de-

composition might be catalyzed by the TLC silica gel

plate, an estimate of the extent of decomposition was

extracts. As expected, all spectra showed that the split-

ting resulted in transesterifications giving the corre-

sponding methyl esters. The absence of vinyl protons

showed that all monomers including the templates have

reacted. The higher splitting yield of P A2 may be a result

of its poor stability, implying that it has partly decom-

posed prior to polymerization. This is seen from the

stability tests prior to polymerization (vide supra). The

splitting yields of P A3 and P B3 based on the NMR data

probably do not represent the true values. Loss of the

volatile methyl benzoate upon evaporation in the sample

preparation for the NMR analysis results in values that

are too low. The values resulting from the HPLC analysis

are therefore more reliable (values in parantheses in

Table 1).

1

done with H NMR. The stability of templates A1 (as TFA

salt and as free base) and A3 in CDCl₃both in the

presence and absence of MAA was checked. For A1 (TFA

salt) this showed a decomposition of 8% in the absence

and 15% in the presence of MAA after 2.5 h. After 17 h

about 20% of A1 as TFA salt and as free base had

decomposed. A3, however, appeared stable under these

conditions. The poor stability of template A1 made it

necessary to synthesize it fresh just prior to polymeri-

zation.

Tem p la te Rem ova l by F lu or id e (3) or Aqu eou s

Ba se (4) Tr ea tm en t. Ser ies A a n d B. In both treat-

ments 3 and 4 TLC showed only a weak ninhydrin-active

spot corresponding to BOCPheOMe (P B1) and PheOMe

Tem p la te Rem ova l: Mod el Stu d y. To remove the

template from the polymers, they were subjected to

various nucleophilic treatments. These were chosen on

the basis of initial attempts to cleave templates B2, C1-

3, diethyl-p-vinylphenyl phosphate, and the benzyl ester

11 in homogeneous solution. Base treatment (MeOH/

[NaOH 1 M], 1/1 (v/v)) led to quantitative hydrolysis of

B2, C1, and diethyl-p-vinylphenyl phosphate whithin 3

h, while 11 reacted considerably slower, giving a 60%

conversion after 3 h and a 90% conversion after 18 h.

About 10% of 11 hydrolyzed by cleavage of the ethyl ester

linkage. Note that the cross-linker EDMA was hydrolyzed

under these conditions. No acid-catalyzed reaction was

observed at room temperature, and only after prolonged

heating did some of the C2 react. EDMA was also stable

under the acidic splitting conditions. However fluoride-

catalyzed transesterification³⁸was a more efficient way

of cleaving the templates. As evidenced by the appear-

ance of nonpolar products, tetrabutylammonium fluoride

(TBAF) catalyzed efficiently the transesterification of C2

to the corresponding methyl ester with complete conver-

sion after 2 h. C3 reacted only slowly but relatively fast

at elevated temperature. Cesium fluoride (CsF) was also

efficient in catalyzing the transesterification. After over-

night reaction at room temperature B2 and C3 had been

completely converted, while only 55% of C1 was trans-

esterified according to HPLC. This was detected as the

release of p-vinylphenol and the absence of polar prod-

ucts. After 48 h at room temperature C1 showed 86%

conversion, while complete conversion was seen within

a few hours when heating the samples to 60 °C.

Thus aqueous base or fluoride treatment is preferably

used for cleavage of the phenylphosphonate linkage. As

noted above and in the BOP-catalyzed synthesis of A1

and A2, the phenolimidazole-containing templates re-

quire considerably milder conditions for cleavage.

Tem p la te Rem ova l by Meth a n ol Extr a ction (P r o-

ced u r es 1 a n d 2). Only templates from polymer series

A and B were partly removed by methanol extraction

(Table 1). A strongly UV- and ninhydrin-active spot with

R_f) 0.6 is seen on TLC of the extracts of P A2 and P B2.

A weak ninhydrin-active spot with R_f< 0.1 was also seen

in P A2. A UV-active spot was seen in P A1. These spots

are most likely due to products resulting from the

transesterification discussed above (see Schemes 2-4).

This was confirmed by the ¹H NMR spectra of the

1

(P B2). H NMR analysis did not show any measurable

amount of template. Instead in treatment 4 a substantial

amount of MAA was released. Possibly this is due to

hydrolysis of pendant unreacted methacrylate groups.

Elemental analysis of residual phosphorus in the poly-

mers indicated in contrast to the NMR data that almost

quantitative splitting of the template had taken place.

The reason for the disagreement between these data is

unclear.

Ser ies C. Aqueous base-catalyzed template removal

1

was followed with HPLC (RP-18) and H NMR analysis

on the supernatants. For P C31 and P C32 HPLC data

showed a template splitting of ca. 70%, while only ca.

20% was observed for the remaining polymers. The

additional heat treatment did not result in any further

increase in splitting. A considerable amount of MAA had

also been released from these polymers. The NMR data

after the aqueous base treatment showed a considerably

higher splitting yield probably due to the extensive

washing of the polymers. The splitting result obtained

from the elemental analysis also shows that the yield of

splitting is higher than that indicated by the HPLC

analysis. In the CsF-catalyzed template removal the

HPLC data showed that ca. 50% of methyl-p-chloro-

benzoate (product from P C31-2 splitting) had been

1

released and that no MAA was detected. Moreover in H

NMR no vinylic protons were seen, indicating a high level

of template incorporation. The low yield from P C31-2

can, as in series A above, be explained by evaporation of

the methylbenzoate formed. Transesterification was in

all cases indicated by appearance of methyl ester signals

at around 4 ppm.

Ch a r a ct er iza t ion of t h e P olym er s. P A1-3 and

P C31-2 showed a yellow color where the intensity

depended on the pH of the surrounding medium. From

being light yellow colored at low and neutral pH they

adopt a clear yellow color at pH over 10-11, the color

becoming darker with time. The color change correlates

well with the pK_aof the phenol group,²⁴indicating

accessibility of the phenol function. The intensity de-

creased in the order P A3, P A2, P A1, probably reflecting

the yield of splitting of the template. Indeed after the

aqueous base and CsF treatment the difference in

intensity between the polymers was less pronounced.

When adding a small amount of Cu(OAc)₂at pH 12,

known to coordinate nitrogen bases,³⁹a clear green color

is formed in P A1-3 again with intensity decreasing in

(38) Ogilvie, K. K.; Beaucage, S. L.; Theriault, N.; Entwistle, D. W.

J . Am. Chem. Soc. 1977, 99, 1277.

Enantioselective Ester Hydrolysis

J . Org. Chem., Vol. 65, No. 13, 2000 4023

the order P A3, P A2, P A1. No color was formed in the

other polymers upon this treatment. The polymers not

mentioned here remained white upon the above treat-

ments. Depending on the nucleophilic treatment applied,

the polymers showed different swelling factors in aceto-

nitrile. The polymers subjected to aqueous base treament

showed clearly a larger swelling than the polymers

treated by either only methanol extraction or CsF treat-

ment. This is probably due to hydrolytic cleavage of cross-

links in the polymer leading to an increased flexibility

in the polymer backbone (vide infra).

The polymers were finally characterized by IR and ¹³C

NMR spectroscopy. The FTIR spectra supported the

correlation between the yield of template removal and

the color intensity of the materials. Thus P A3 showed

the largest intensity of the band arising from the free

-OH stretch at approximately 3550 cm^-1assigned to

carboxylic acid groups and liberated phenol groups of the

active sites. Unfortunately, the low concentration of

aromatic residues in the polymer (0.5 mol %) prevented

quantitative analysis of incorporated template by solid

state ¹³C NMR. These spectra revealed however a level

of residual unsaturation in the polymers between 10 and

15%.

Mon om er -Ca ta lyzed Ester Hyd r olysis. As a first

step in the kinetic evaluation the homogeneous hydroly-

sis of BOCPheONP in the presence of the functionalized

monomers was studied. The hereby obtained information

about the catalytic role of the individual functional

groups serves as a guideline in the following kinetic

evaluation in the presence of the polymer catalysts. Only

in the presence of the imidazole functionalized monomers

(series A and B, monomers: 7 and B4) were appreciable

rate enhancements observed (Figure 1). The observed

rate, given as the pseudo-first-order rate constant k_obs

,

F igu r e 1. (a) Rate of hydrolysis of BOC-L-PheONP (0.025

mM) in MeCN/[potassium phosphate buffer 0.05 M], 7:3 (v/v)

in the presence of soluble monomeric catalysts 7 and B4 (0.7

mM) at varying pH. k_obs) observed pseudo-first-order rate

constant. bg ) background rate. (b) k_cat) catalytic rate

constant ) k_obs- k_bg, ꢀ₃₅₀) extinction coefficient of 7 at 350

nm.

increased with increasing pH both in the background

noncatalyzed reaction and in the presence of catalyst

(Figure 1a). Also the catalytic rate k_cat(observed rate

minus background rate) increased with increasing pH

(Figure 1b), although this was more pronounced in the

reactions catalyzed by 7. In this case the increase in k_cat

was accompanied by an increase in the UV absorbance

of a shoulder at λ ) 350 nm.

Sch em e 9

Kinetics of ester hydrolysis catalyzed by 4(5)-hydroxy-

phenylimidazole (4(5)HPI)⁴⁰or 2-hydroxyphenylimidazole

(2HPI)³¹(analogues of 7) (Scheme 9) have been exten-

sively studied. In the hydrolysis of the activated esters

p-nitrophenylacetate (PNPA) or p-nitrophenyltoluate

(PNPT) in the presence of 4(5)HPI the rate was of first

order with respect to catalyst and increased with increas-

ing pH, indicating involvement of the phenoxide group

in the catalytic reaction mechanism.⁴⁰This was further

supported by a correlation between the rate and the pK_a

of the phenol group.³¹In water the phenoxide of 2HPI

was associated with a λ_maxat 325 nm.³¹The correlation

between the increase in k_catand ꢀ₃₅₀of 7 with pH (Figure

1b) is therefore evidence for involvement of phenoxide

in the catalytic reaction mechanism at higher pH values.

This probably takes place by either nucleophilic attack

by phenoxide followed by intramolecular attack by imid-

azole and hydrolysis of the formed imidazolide or phen-

oxide-assisted nucleophilic catalysis by imidazole.⁴⁰In the

presence of B4 the rate enhancement is ascribed to the

imidazole group, which apparently is more efficient than

the imidazole group in 7 in catalyzing the hydrolysis of

BOCPheONP (compare the relative rates at neutral pH).

For activated substrates nucleophilic displacement of

ONP by imidazole is the exclusive pathway.³¹

P olym er -Ca ta lyzed Ester Hyd r olysis. p H Dep en d -

en ce. In Figure 2a the same experiment as in Figure 1

but in the presence of the polymer catalysts is shown.

Only at pH’s up to 8 is catalysis observed, while at higher

pH the polymers all inhibit the reaction. At high pH the

polymers are abundant in negatively charged carboxylic

acid groups, suggesting electrostatic exclusion of nucleo-

philic OH^-from polymer-bound substrate as a possible

cause of the inhibition. In a study of the effect of mobile

phase pH on chiral separations using imprinted polymers

(39) For example see: Chin, J .; J ubian, V. J . Chem. Soc., Chem.

Commun. 1989, 839.

(40) Overberger, C. G.; Shen, C. M. J . Am. Chem. Soc. 1971, 93,

6992.

4024 J . Org. Chem., Vol. 65, No. 13, 2000

Sellergren et al.

the bound template. In series A the catalytic activity of

P A1 increased with the amount of template removed

from the polymer (Table 4). After the methanol extraction

(treatment 1 and 2) the control polymer P A2 was more

efficient than P A1 in the catalysis, probably reflecting

the difference in the amount of template removed from

the polymers. As seen in Table 1, this treatment removed

only 23% A1 from P A1, while about 75% A2 was removed

from P A2. After the CsF treatment however the level of

template removal should be similar for all polymers. The

rate in the presence of P A1 has now more than doubled

compared to after treatment 1 and 2, while that in the

presence of P A3 remained approximately the same. This

is evidence that the polymeric active sites generated by

template removal are responsible for the rate enhance-

ments observed here. Interestingly after treatment with

aqueous base at room temperature P A1 gave rise to an

additional more than doubling of the rate, while the

remaining polymers in the A series showed only modest

increases. Note that the level of template removal is

about the same as after the CsF treatment. A significant

enantioselectivity is here also observed, with the enan-

tiomer complementary to the configuration of A1 being

preferentially hydrolyzed. After 48 h treatment in aque-

ous sodium carbonate an even larger selectivity was

observed. In presence of P A1 BOC-^D-PheONP was here

hydrolyzed 1.9 times faster than the ^Lenantiomer.

Comparing P A1 and P A3, a rate enhancement of 2.5 was

observed, and compared to the soluble catalyst a 7-fold

rate enhancement was observed. Using aqueous sodium

hydroxide in methanol, a similar result is obtained. The

selectivity of the binding site is here revealed by the

control polymer P A2 (prepared from template lacking the

BOC group), which is clearly less efficient than P A1 in

catalyzing the hydrolysis. When the hydrolytic treatment

was carried out at elevated temperature, enantioselec-

tivity was completely lost. At the same time the rate

enhancement relative to the control polymer P A3 was

unaffected. Apparently the functional groups of the sites

are still positioned for substrate binding, while the

stereochemical complementarity has been lost. It should

be noted that in the chromatographic enantiomer separa-

tions on polymers imprinted with ^L-phenylalanine anilide

the selectivity decreased significantly (R 5.6 f 2.3)

(swelling factor: 1.55 f 1.70) upon aqueous base treat-

ment at room temperature.⁴²

F igu r e 2. (a) Rate of hydrolysis of BOC-D-PheONP (0.025

mM) in MeCN/[potassium phosphate buffer 0.05 M], 7:3 (v/v)

(2 mL) in the presence of series A polymer catalysts (50 mg)

at varying pH. (b) Ratio of the observed pseudo-first-order rate

constants (k_obs) for some of the catalysts in (a).

containing carboxylic acid groups, we observed that

selective binding occurred preferentially at low pH,

suggesting a difference in pK_abetween the selective and

nonselective sites.⁴¹At low pH the carboxylic acid groups

of the sites should be mainly protonated and able to more

efficiently form hydrogen bonds to the substrate. Inter-

estingly, by comparing the observed rates in the presence

of the template-imprinted polymers with the reference

polymers or the soluble catalysts the relative rate in-

creased with decreasing pH (Figure 2b). A relatively low

pH is thus required for a selective polymer-catalyzed

ester hydrolysis to occur in this system. As expected,

carboxylic acid groups appear to be responsible for the

substrate binding to the active sites. At pH 4.5 the

imidazole group should be protonated (Scheme 9) and

may therefore function as a general acid catalyst assist-

ing the water attack on the substrate.³¹

The larger rate enhancements exhibited by the poly-

mers subjected to aqueous base treatment may be related

to the flexibility of the polymer backbone reflected in the

polymer swelling factor. The base-treated polymers swelled

more in acetonitrile (series A: SF ) 2.33 mL/mL (after

heating for 24 h at 60 °C, SF ) 2.47 mL/ml); series C:

SF ) 1.70 mL/mL) than those subjected to CsF treatment

(series C: SF ) 1.41 mL/mL) or those only subjected to

methanol extraction (series A: SF ) 2.00 mL/mL; series

C: SF ) 1.49 mL/mL). In the aqueous base treatment of

cross-linked methacrylate polymers an increase in swell-

ing has been explained by hydrolysis of ester cross-links,

leading to a more flexible polymer structure.⁴³Possibly

the aqueous base treatment applied gives rise to a more

flexible polymer network where the binding sites can

more easily accommodate the transition state of the

Effect of Nu cleop h ilic Tr ea tm en ts of th e P oly-

m er s on th e Obser ved Ra te En h a n cem en ts. Ser ies

A. The catalytic activity of the polymers was sensitive

to the nucleophilic treatments applied in order to remove

(42) Sellergren, B.; Shea, K. J . Chromatogr. A 1993, 635, 31.

(43) J elinkova, M.; Shataeva, L. K.; Tischenko, G. A.; Svec, F.

Reactive. Polym. 1989, 11, 253.

(41) Sellergren, B.; Shea, K. J . Chromatogr. A 1993, 654, 17-28.

Enantioselective Ester Hydrolysis

J . Org. Chem., Vol. 65, No. 13, 2000 4025

Ta ble 4. P seu d o-F ir st-Or d er Ra te Con sta n ts for th e Hyd r olysis of Boc-D(L)-P h eONP in th e P r esen ce of Ser ies A

P olym er Ca ta lysts Su bjected to Nu cleop h ilic Tr ea tm en ts^c

treatment

catalyst

k_D× 10³(min^-1

)

k_D(PA1)/k_D

k_D/k_L

MeOH ∆ (1+2)

P A1

P A2

P A3

P A4

7

0.026

1.00

0.68

1.08

1.85

1.9

1.04

0.038

0.024

0.014

1.00

-

CsF (3)

P A1

P A2

P A3

P A4

7

P A1

P A3

7

P A1

P A3

7

P A1

P A2

P A3

P A4

7

0.057

0.046

0.035

0.022

1.00

1.24

1.63

2.59

4

1.00

1.06

1.00

0.014

Na₂CO₃(aq) 24 h

Na₂CO₃(aq) 48 h

NaOH(aq)

0.140 ((0.01)^a

1.37 ((0.05)^a

1.00

0.080 ((0.004)^a

0.014

1.75 ((0.05)^a

10 ((0.7)^a

1.00

0.122 (0.410)^b

0.048 (0.190)^b

0.014 (0.125)^b

0.102

1.85 (1.63)^b

1.00 (1.00)^b

2.54 (2.10)^b

7 (3.3)^b

1.00

1.60

1.00

0.056

0.046

0.022

0.014

1.82

2.22

4.64

6

NaOH(aq) 60 °C

P A1

P A3

7

0.108

0.046

0.014

1.00

2.35

6

1.00

a

b

Average of two independent experiments with error in parantheses. Reaction carried out at 40 °C. ^cThe hydrolysis of BOC-D- or

L-PheONP (0.025 mM) in MeCN/[potassium phosphate buffer 0.05 M, pH 4.5], 1:1 (v/v), in the presence of the various catalysts was

followed by monitoring the formation of p-nitrophenol by HPLC (see Experimental Section). 7 was added to a concentration corresponding

roughly to the site concentration determined from the amount of template removed from the polymers.

Ta ble 5. P seu d o-F ir st-Or d er Ra te Con sta n ts for th e

Hyd r olysis of Boc-^L(D)-P h eONP in th e P r esen ce of Ser ies

B P olym er Ca ta lysts Su bjected to Nu cleop h ilic

Tr ea tm en ts

treatment

catalyst k_L× 10³(min^-1

)

k_L(PB1)/k_Lk_L/k_D

MeOH ∆ (1+2)

P B1

P B2

P B3

P B4

B4

P B1

P B2

P B3

P B4

B4

0.329

0.258

0.243

0.276

0.039

0.342

0.155

0.199

0.138

0.039

0.304

0.236

0.039

1.0

1.3

1.4

1.2

8.2

1.0

2.2

1.7

2.5

8.8

1.0

1.3

7.8

1.18

1.08

1.07

1.06

CsF (3)

1.11

1.07

1.04

1.01

1.07

Na₂CO₃(aq) 24 h

P B1

P B3

B4

a

The hydrolysis of BOC-L- or D-PheONP (0.025 mM) in MeCN/

[potassium phosphate buffer 0.05 M, pH 6], 1:1 (v/v), in the

presence of the various catalysts was followed by monitoring the

formation of p-nitrophenol by HPLC (see Experimental Section).

B4 was added to a concentration corresponding roughly to the site

concentration determined from the amount of template removed

from the polymers.

F igu r e 3. Evaluation of the observed pseudo-first-order rate

constants for the hydrolysis of BOC-^L-PheONP (0.025 mM) in

MeCN/[potassium phosphate buffer 0.05 M, pH 6], 7:3 (v/v) (2

mL), in the presence of CsF-treated series B polymer catalysts

(50 mg).

reaction. Note that conformational flexibility is a common

requirement in enzyme catalysis.^13,14

Ser ies B. The rates of hydrolysis in the presence of

the series B catalysts responded in a different way to the

various nucleophilic treatments (Table 5). P B1, being the

complement to BOC-L-PheONP, was in all cases most

efficient in catalyzing the hydrolysis of the ^Lenantiomer.

A significant enantioselectivity in preference for the ^L

enantiomer was observed. Comparing the rate enhance-

ments after each treatment, no difference was noted in

the presence of P B1, while upon CsF treatment the rates

in the presence of the other polymers decreased slightly.

This resulted in a slightly larger relative rate enhance-

ment (Figure 3) of 2.5 for P B1 relative to P B4 and 8.8

for P B1 relative to soluble catalyst after the CsF treat-

ment. The other notable effect of the various treatments

was that in contrast to the series A polymers the

enantioselectivity decreased upon the more severe nu-

cleophilic treatments. Note that the soluble chiral cata-

lyst by itself exhibits some enantioselectivity in the

catalysis.

Th e Role of th e Active Site F u n ction a l Gr ou p s.

In Table 6 CsF-treated catalytic polymers lacking the

imidazole group (P C11 and P C31), containing randomly

distributed imidazole groups (P C12 and P C32), or hav-

ing an imidazole group in proximity to the binding site

(P A1, P A3 and P B1, P B3) have been compared. Due to

the various concentrations of active sites in the polymers,

second-order rate constants have been calculated (based

on the amount of imidazole present in the monomer

4026 J . Org. Chem., Vol. 65, No. 13, 2000

Sellergren et al.

Ta ble 6. Secon d -Or d er Ra te Con sta n ts for th e Hyd r olysis of Boc-D(L)-P h eONP in th e P r esen ce of Va r iou s P olym er

Ca ta lysts^a

a

The hydrolysis of the complementary enantiomer of BOCPheONP (0.025 mM) in MeCN/[potassium phosphate 0.05 M, pH 4.5 ((a) pH

6)], 1:1 (v/v), in the presence of esterase-mimicking polymers P C11, P C12, P A1, and P B1 and corresponding reference polymers P C31,

P C32, P A3, and P B3 was followed on HPLC as described in the Experimental Section. The second-order rate constants (k₂) were calculated

from the theoretical concentration of accessible imidazole groups in the polymers or in the absence of imidazole groups, from the theoretical

concentration of accessible sites.

mixture). The imidazole group is apparently required for

catalysis since only small rate enhancements and no

template-induced catalysis are seen using P C11 and

P C31. Introducing a randomly distributed imidazole

group, a larger rate enhancement is seen but still no

template-induced effect. Only when the imidazole group

is implanted in conjunction with the site are template-

induced rate enhancements observed (P A and P B).

Clearly the positioning of the imidazole group in proxim-

ity to the binding site is necessary in the design of a

template-imprinted polymeric esterase model.

Ra te of Hyd r olysis ver su s Su bstr a te Con cen tr a -

t ion . The rate of hydrolysis of BOC-^D-PheONP in the

presence of aqueous base-treated P A1 and P A3 was

studied up to substrate concentrations of 10 mM (Figure

4). The rate enhancement of P A1 relative to P A3 and

soluble catalyst persists at higher substrate concentra-

tions. Only a weak saturation tendency is seen, implying

F igu r e 4. Rate of hydrolysis of BOC-D-PheONP in MeCN/

[potassium phosphate buffer 0.05 M, pH 4.5], 7:3 (v/v) (2 mL),

in the presence of series A polymer catalysts (50 mg) as a

function of substrate concentration. The reactions were fol-

lowed over 3 days and the rates calculated as the average of

the rate of substrate consumption and the rate of product

formation.

that the substrate binding is quite weak. An Eadie-

Hofstee plot of the kinetic data gave for P A1 K_m) 5.4

((0.4) mM, k_cat) 1.40 × 10^-3min^-1and for P A3 K_m

)

10.8 ((0.1) mM, k_cat) 1.14 ((0.04) × 10^-3min^-1. The

K_mvalues agreed well with those that could be estimated

from the substrate peak integrals at the start of the

reaction. Apparently the rate enhancements are mainly

the result of a stronger substrate binding to the comple-

mentary polymers. At a substrate concentration of 10 mM

turnover kinetics was observed. Over the time course

followed the soluble catalyst performed 0.7 turnovers.

Meanwhile P A1 and P A3 had performed 5 and 3

turnovers, respectively.

P olym er -Ca ta lyzed Hyd r olysis of ^L(D)-P h en yl Ala -

n in e Eth yl Ester (L(D)-P h eOEt). The hydrolysis of

Enantioselective Ester Hydrolysis

J . Org. Chem., Vol. 65, No. 13, 2000 4027

Ta ble 7. P seu d o-F ir st-Or d er Ra te Con sta n ts for th e

Hyd r olysis of ^D(L)-P h eOEt^a

k_Dobs× 10⁵

k_Lobs× 10⁵

k_Dobs

k_BLD

/

k_Lobs

k_BLL

/

k_D/

k_L

b

catalyst

(min^-1

)

(min^-1

)

P A1

P A2

P A3

2.50

2.23

1.97

1.92

1.86

3.05

2.68

2.37

2.43

2.37

2.29

1.44

1.26

1.08

a

The hydrolysis of D(L)-PheOEt (0.025 mM) was monitored in

MeCN/[potassium phosphate 0.05 M, pH 7.4], 1:1 (v/v), in the

presence of esterase-mimicking polymers as described in the

b

Experimental Section. Ratio of first-order rate constants after

subtraction of the rate constants for the background reaction.

(PheOEt), a nonactivated substrate, was studied at pH

7.4 using the polymer catalysts P A1, P A2, and P A3 (see

Table 7). Polymer P A2 has the exact size complementa-

rity to ^D-PheOEt. A rate enhancement up to 3 was

observed when compared to the blank rate. P A2 hydro-

lyzed ^D-PheOEt 1.3 times and polymer P A1 1.4 times

faster than the ^Lester. The Michaelis-Menten kinetic

studies for the hydrolysis of D- and L-ethyl esters was

carried out for polymers P A1 and P A2. The concentration

was varied from 0.5 to 5 mM.(Figure 5). The K_mand k_cat

values for polymer P A1 for L-PheOEt are 1.96 ((0.05)

mM and 1.91 × 10^-5min^-1, respectively. The correspond-

ing values for polymer P A2 are K_m) 2.32 ((0.06) mM

and 1.82 × 10^-5min^-1. The values for the hydrolysis of

D-PheOEt are for P A1 K_m) 1.92 ((0.05) mM, k_cat) 2.32

-1

× 10^-5min

and for polymer P A2 K_m) 2.04 ((0.04)

mM and k_cat) 2.07 × 10^-5min^-1. The ratio of the

resulting apparent second-order rate constants k_cat/K_mfor

D-PheOEt over L-PheOEt is informative about the D-

enantioselectivity of the polymers within a larger con-

centration interval. These figures are 1.24 for polymer

P A1 and 1.30 for P A2, which is in agreement with the

expected structural complementarity.

F igu r e 5. Lineweaver-Burk plots of the initial reaction rates

versus the substrate concentration for the hydrolysis of ^L-

phenyalanine ethyl ester in the presence of (a) P A1 and (b)

P A2. The hydrolysis reactions were followed for over 30%

conversion, and the rates were calculated as the average of

the rate of product formation for duplicate samples after

subtracting the blank rates.

Con clu sion s

Fully predictable enantioselective catalysis of ester

hydrolysis has been achieved with template-imprinted

polymers containing similar catalytic elements believed

to be responsible for the catalytic action of chymotrypsin.

The catalysts described here are truly tailor-made since

the enantioselectivity in all cases could be predicted. They

furthermore provide yet an example of an imprinting

strategy combining noncovalent and covalent binding

forces between the template and the functionalized

monomers. This allows simultaneous introduction of

different functional groups into the sites. Hydrogen

bonding, involving carboxylic acid groups at the sites, is

the main driving force in the stereoselective binding step

and an imidazole group associated with the binding site

responsible for the catalytic action. For the first time the

enantioselective hydrolysis of a nonactivated ester is

reported. The polymer catalysts were found to be cata-

lytically active and retained the enantioselectivity over

a period of about 10 years. This clearly highlights the

advantage of polymer mimics over the other less stable

soluble models. The influence of transition state comple-

mentarity is still unknown, but imprinted polymers are

apparentely capable of discriminating between a planar

and tetrahedral structure similar to the substrate and

template A1, respectively. It may be of interest that the

optimum pH for the selective catalysis of the activated

ester is at the acidic side. This is quite unusual in

esterolytic reactions and may be of general synthetic

interest when the substrate contains for instance base

labile protecting groups.

Although modest, the rate enhancements reported here

should be considered as a first step in the development

of imprinted polymers for enantioselective catalysis. The

major problem with the present system is the rather

weak substrate binding. Introduction of additional strong-

er binding interactions directed both toward the acyl part

and the leaving group of the substrate are guidelines for

our future efforts to achieve an efficient mimic of chy-

motrypsin.

Ack n ow led gm en t. R.K. acknowledges the financial

support from the Graduierten-Kolleg for Chemistry and

Physics of Supramolecular Systems, University of Mainz.

Su p p or tin g In for m a tion Ava ila ble: Detailed synthetic

procedures and supporting structural characterization not

included in the Experimental Section are available for the

compounds described in this work . This material is available

free of charge via the Internet at http://pubs.acs.org.

J O000014N

Article Doi

Enantioselective ester hydrolysis catalyzed by imprinted polymers. 2

DOI: 10.1021/jo000014n

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jo000014n

Full text of DOI:10.1021/jo000014n

Products guided by the article

R&D Labs maybe for 289506-18-9

Relevant to this article

Hot Product