Penicillin Acylase Catalysed Synthesis of Ampicillin in Hydrophilic Organic Solvents

Article abstract of DOI:10.1002/adsc.200303023

Penicillin acylase (EC 3.5.1.11) from Alcaligenes faecalis, immobilised as a cross-linked enzyme aggregate (CLEA), catalysed the synthesis of ampicillin in water-miscible organic solvents at low water concentrations. Below 4% water (v/v) no reaction was observed, showing the crucial role of water in maintaining the activity of penicillin acylase. The initial value of S/H was strongly affected by the nature of the solvent, but the effect of the water content was slight in the studied range of 4 to 15%. A reaction in acetonitrile containing 8% water afforded ampicillin in up to 86% yield.

Full text of DOI:10.1002/adsc.200303023

FULL PAPERS
Penicillin Acylase Catalysed Synthesis of Ampicillin in
Hydrophilic Organic Solvents
Luuk M. van Langen, Natasja H. P. Oosthoek, Fred van Rantwijk, Roger A.
Sheldon*
Laboratoryof Biocatalysis and Organic Chemistry, Julianalaan 136, 2628 BL Delft, The Netherlands
Fax: ( ‡ 31)-15-278-1415, e-mail: boc-secretariat@tnw.tudelft.nl
Received: January28, 2003; Accepted: April 5, 2003
Abstract: Penicillin acylase (EC 3.5.1.11) from Alca- nature of the solvent, but the effect of the water
ligenes faecalis, immobilised as a cross-linked enzyme content was slight in the studied range of 4 to 15%. A
aggregate (CLEA), catalysed the synthesis of ampi- reaction in acetonitrile containing 8% water afforded
cillin in water-miscible organic solvents at low water ampicillin in up to 86% yield.
concentrations. Below 4% water (v/v) no reaction was
observed, showing the crucial role of water in Keywords: ampicillin synthesis; cross-linked enzyme
maintaining the activityof penicillin aclyase. The
initial value of S/H was stronglyaffected bythe
aggregate; enzyme catalysis; immobilisation; organic
solvents; penicillin acylase
Introduction
(50:50).[8] Removing all of the water, as is routinely
done in lipase-catalysed amine acylation,[9] would ob-
The industrial synthesis of penicillins and cephalospor- viate the hydrolysis problem entirely. However, such
ins is undergoing a transition from predominantly reaction conditions are not tolerated verywell by
chemistry-based procedures to biological and biocata- penicillin acylase, although it still has acyl transfer
lytic ones.[1,2] Thus, the chemical hydrolysis of penicillin activityin toluene at reduced water activity( aW ˆ
G into the keybuilding block 6-aminopenicillanic acid
(6-APA) has largelybeen replaced byan enzymatic
0.5).[10] We have recentlydemonstrated the facile
immobilization of E. coli penicillin acylase as a cross-
procedure,[3] which utilises the unique chemoselectivity linked enzyme aggregate (CLEA), which mediated the
of penicillin acylase (EC 3.5.1.11). In contrast, the synthesis of ampicillin in various organic solvents in the
industrial transformation of the b-lactam intermediates presence of 5% (v/v) of water.[11]
6-APA and 7-deacetoxycephalosporanic acid into semi-
In our continued investigations of the enzymatic
synthetic penicillins and cephalosporins still largely synthesis of ampicillin in organic media at low water
involves chemical procedures, although there is a activitywe turned our attention to the relatively
definite trend towards their replacement bycleaner
unknown penicillin acylase from Alcaligenes faeca-
biocatalytic procedures.[4]
lis[12,13] on account of its stability[13] and pH tolerance.[14]
The enzymatic synthesis of a b-lactam antibiotic such We now report the effects of the nature of the solvent
as ampicillin is not trivial, since a simple condensation of and the water content on the course of the enzymatic
d-phenylglycine (d-PG) and 6-APA is thermodynami- coupling of d-PGA and 6-APA using a CLEA of
[5,6]
callyunfavourable.
In consequence, all viable syn- A. faecalis penicillin acylase. Our goal was to increase
thesis schemes involve an activated side-chain donor, the synthesis/hydrolysis ratio by reducing the amount of
such as d-phenylglycine amide (d-PGA), which acylates water in the system, thereby providing a more econom-
6-APA in the presence of penicillin acylase in a kineti- icallyviable process.
callycontrolled reaction; [7] water is the reaction medium
of choice. A major problem of such schemes is the
competing irreversible hydrolysis of the acyl donor as Results and Discussion
well as the product, leading to unreactive d-PG (see
Figure 1).
We investigated the synthesis of ampicillin from d-PGA
One approach to reducing the competing hydrolysis is, and 6-APA in the presence of an A. faecalis pencillin
obviously, to replace the water by an organic solvent. acylase CLEA[15] in three water-miscible organic sol-
Improved yields of ampicillin were indeed obtained vents: tert-butyl alcohol, 1,2-dimethoxyethane (DME)
when the reaction was performed in glycol-water and acetonitrile (ACN). tert-Butyl is a standard solvent
Adv. Synth. Catal. 2003, 345, 797 801
DOI: 10.1002/adsc.200303023
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
797

Luuk M. van Langen et al.
FULL PAPERS
for the lipase-mediated acylation of strongly hydrophilic ampicillin and d-PG formed), which is a convenient
reactants, such as carbohydrates; DME and ACN were yardstick for measuring the efficiency of a kinetically
chosen becausetheygave the best resultsin the synthesis
controlled process,[4] was calculated.
Ampicillin synthesis took place in all of these solvents
of ampicillin in the presence of a CLEA of E. coli
penicillin acylase.[11] The effects of water were measured if the water content was at least 4% (v/v), but no reaction
byadding various amounts, ranging from 2 15% (v/v);
estimated values of the water activity( aw) are included crucial for the catalytic action of A. faecalis penicillin
in Table 1. acylase as was also observed in the case of E. coli
took place at 2% water. The presence of water is clearly
[16]
To assess the kinetic performance of the biocatalyst, penicillin acylase,[10b] but, contraryto expectations,
the initial values of the synthesis and hydrolysis rate there is no obvious relationship with the water activity.
were measured (see Table 1). From these, the initial In fact, the minimum water concentration required for
synthesis/hydrolysis ratio (S/H, the molar ratio of activityis verysimilar in the solvents used, although the
corresponding water activities diverge widely. We can
onlyspeculate on the cause of these phenomena. The
loss of the bridging water molecule that activates the
catalytic serine[17] could possiblybe involved. We also
note that, even in 80% (v/v) methanol, water does not
mix at a molecular level but mainlyis present as strings
or clusters of molecules.[18] Possiblythe required 4%
water corresponds with a transition from oligomeric
water into a molecular dispersion in these media.
The nature of the solvent markedlyinfluencedthe rate
of ampicillin synthesis. At ꢀ 8% water the synthesis
rate increased in the order tert-butyl alcohol < DME <
ACN. The influence of the water content on the rate was
dependent on the solvent used. The synthesis and
hydrolysis rates in tert-butyl alcohol increased strongly
when the water content was increased from 4 to 15%. In
contrast, the reaction rates in DME medium were hardly
affected bythe water concentration, whereas in ACN
the synthesis rate passed through a maximum at 8%
water.
The nature of the solvent stronglyaffected the initial
value of S/H, which increased from 0.6 in tert-butyl
alcohol to 2 in DME and 4 in ACN tert-butyl alcohol.[19]
We have previouslyobserved a similar effect of the
solvent on S/H with the penicillin acylase from
E. coli.[11,21] There is no obvious relationship with any
molecular propertyof these solvents, such as logP [22] or
Figure 1. Kinetically controlled enzymatic synthesis of ampi-
cillin.
[23]
the polarityindex
(see Table 2). Onlythe relatively
Table 1. Effect of the medium on the enzymatic synthesis of ampicillin[a]
tert-Butyl alcohol
DME
ACN
S/H aw vsynth
H2O
aw vsynth
vhydrol
S/H aw vsynth
vhydrol
vhydrol
[mmolUÀ1hÀ1] [mmolUÀ1hÀ1]
S/H
[vol %]
[mmolUÀ1hÀ1] [mmolUÀ1hÀ1]
[mmolUÀ1hÀ1] [mmolUÀ1hÀ1]
2
4
6
8
10
15
0.3 0.00
0.00
0.23
0.23
0.38
0.85
1.40
0.1 0.00
0.00
0.23
0.28
0.31
0.30
0.30
0.4 0.00
0.00
0.23
0.17
0.24
0.10
0.075
0.5 0.09
0.6 0.09
0.7 0.23
0.8 0.47
0.9 0.70
0.4 0.2 0.37
0.4 0.3 0.60
0.6 0.4 0.66
0.5 0.5 0.60
0.5 0.6 0.60
1.6 0.7 0.64
2.1 0.8 0.71
2.1 0.9 0.87
2.0 0.9 0.40
2.0 1.0 0.30
2.7
4.2
3.6
4.0
4.0
[a]
Initial rates. Water activities have been estimated using the UNIFAC computer programme and are not corrected for water
uptake bythe reagents. Reaction conditions: 150 mM d-PGA, 50 mM 6-APA and 130 U immobilised penicillin acylase in
20 mL solvent at 08C.
798
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2003, 345, 797 801

Penicillin Acylase Catalysed Synthesis of Ampicillin
FULL PAPERS
high dielectric constant of ACN[24] could be significant in
We subsequentlyexamined the synthetic potential of
this respect, as it affects the balancing of coulombic and our reaction system by allowing the reaction to proceed
hydrophobic interactions in the enzyme. In contrast, the to maximum conversion. These experiments were per-
influence of the water content on S/H was slight. A lower formed using DME and ACN, which had given the best
water content would be expected to favour synthesis results at lower conversions. A maximum yield of 86%
over hydrolysis but in fact S/H even tended to decline at was obtained in ACN containing 8% water; the yield was
the lowest water content. Averysimilar observation has
been made in the comparable kineticallycontrolled
82% in DME containing 4% water (see Figure 2).
Analogous to what is observed in aqueous systems, the
alcoholysis of disaccharides in the presence of glyco- value of S/H declined as the reaction proceeded.[27] This
sidases.[25] It also is relevant to note that Candida latter effect is ascribed to the accumulation of ampicillin
antarctica lipase, for example, avidlyreacts with even
trace amounts of water, which maycause an appreciable
in the reaction medium, causing an increasing competi-
tion of ampicillin with d-PGA for the active site of
hydrolytic side-reaction even under nearly anhydrous penicillin acylase. Moreover, the concentration of the
conditions.[26]
nucleophile (6-APA) declines in the course of the
We next considered the issue of the limited solubility reaction, depressing the synthesis rate versus the
of the nucleophile 6-APA in the reaction media. In a primaryhydrolysis rate (see Figure 1).
saturated solution the thermodynamic activity of 6-APA
A slow net hydrolysis of ampicillin took place after its
will be at a maximum value, equal to that in the maximum concentration had been reached, in particular
crystalline phase. If the 6-APA is completely dissolved when the reaction was performed in ACN. This is
its thermodynamic activity will be lower, which could characteristicallyobserved in a kineticallycontrolled
result in a lower S/H. We found that the initial 6-APA synthesis and shows that the biocatalyst is still active
concentration, 50 mM, exceeded its solubilityin DME
after 300 min in ACN medium. In DME medium the
or ACN containing < 8% water, whereas at higher hydrolysis of ampicillin, as well as that of d-PGA, was
water concentrations all of the 6-APAwas dissolved. No much less pronounced in the final stage of the reaction.
corresponding effect on vsyn or S/H became apparent, Inactivation of the biocatalyst was not involved as we
however (see Table 1).
found, in an independent experiment, that the enzy-
matic hydrolysis rate of d-PGA in DME-water (96:4) in
the absence of 6-APA was comparable to the low rate
found in the synthetic experiment after reaching max-
imum ampicillin concentration (data not shown). It
would seem that, in DME medium, hydrolysis of the
acyl-enzyme complex requires the binding of 6-APA in
the active site, which is a well-recognised kinetic path-
Table 2. Some properties of the solvents.
Log P[b]
ET
[a]
N[c]
Solvent
e0
[28]
tert-Butyl alcohol
1,2-Dimethoxyethane
Acetonitrile
12.47
7.20
35.94
0.35
À 0.21
À 0.34
0.39
0.16
0.46
way. Consequently, the hydrolysis is slow at low 6-
APA concentration. A verysimilar effect was observed
in the enzymatic synthesis of ampicillin in 50% aqueous
1,2-propanediol or 1,3-butanediol.[8] We also showed
that the penicillin acylase could be recycled: more than
80% of the initial activitywas recovered upon rehydra-
tion of the catalyst (see Table 4 in the Experimental
Section).
[a]
[b]
[c]
At 258C, data taken from ref.[24]
See ref.[22]
Normalised
polarityindex
(water
ˆ 1.00,
tetra-
methylsilane ˆ 0.00) according to Reichardt.[23]
Figure 2. Penicillin acylase catalysed synthesis of ampicillin in organic solvents. A: ACN-water (92:8, v/v); B: DME-water
~
(96:4, v/v). Reaction conditions: 800 U penicillin acylase, further as described below Table 1. Legend: d-PGA: ; 6-APA: &;
*
ampicillin: ; d-PG: ^.
Adv. Synth. Catal. 2003, 345, 797 801
asc.wiley-vch.de
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
799

Luuk M. van Langen et al.
FULL PAPERS
Table 3. The catalytic efficiency of penicillin acylase in several systems
Reaction
Source
Formulation
Solvent
Temp. [ 8C] pH Turnover rate [sÀ1] Ref.
Pen G hydrolysis
Pen G hydrolysis
Ampicillin synthesis
Ampicillin synthesis
Ampicillin synthesis
A. faecalis native
water
water
25
25
10
10
0
7.5 54
7.5 50
[14]
[14]
[29]
E. coli
E. coli
E. coli
native
Assemblase[a] water
6.3
0.6
CLEA
water
ACN-water (92: 8)
7
2.3
1
A. faecalis CLEA
[30]
Phenylacetyl transfer E. coli
on celite rods toluene-water (aw 0.48) 30
0.3
[10c]
[a]
[15]
A covalentlybound preparation of penicillin acylase on gelatin-chitosan.
The Netherlands). d-(À)-Phenylglycine amide was provided
byDSM (Geleen, The Netherlands). The cross-linking agents
dimethyl adipimidate dihydrochloride and 25% glutaralde-
hyde solution in water were obtained from Aldrich. A
suspension of A. faecalis penicillin acylase CLEA (483
U mLÀ1) was prepared according to a published procedure.[15]
As regards the biocatalyst turnover rate, it is com-
monlyobserved that enzymes exhibit low activityin
organic media. To put this issue into perspective, we
have compared (see Table 3) the turnover rate of the
penicillin acylases from A. faecalis and E. coli in a
number of reactions, each at optimum conditions. Thus,
the activity of the penicillin acylases in the hydrolysis of
penicillin G indeed is much higher than that of the
A. faecalis pencillin acylase CLEA in the synthesis of
ampicillin from d-PGA in an organic solvent. However,
the latter is comparable with that in an aqueous medium
using a carrier-bound E. coli enzyme,[29] but a CLEA of
the same enzyme was four times as active.[30] We
attribute this latter effect mainlyto diffusion limitations
in the carrier. Finally, a celite-immobilised penicillin
acylase in low-water toluene medium was much less
active than our CLEA preparation in ACN-water
(92:8). We tentativelyconclude that the cataltyic
efficiency of a penicillin acylase CLEA in the synthesis
of ampicillin in organic media is competitive with that in
an aqueous medium.
Penicillin Acylase Assay
The activity of penicillin acylase was assayed in the hydrolysis
of a 2% solution of penicillin G potassium salt in 0.1 M
phosphate buffer at 348C and pH 8.0, which was maintained by
automatic titration with NaOH. One unit (U) of penicillin
acylase will liberate one mmol of phenylacetic acid per min.
Activity Recovery of the Biocatalyst
Table 4. Recovery of the hydrolytic activity of A. faecalis
penicillin acylase CLEA after 2 h in the reaction mixture.
Reaction medium
Recovered activity[a] [%]
tert-Butyl alcohol
DME
ACN
93
82
78
Conclusion
[a]
In the standard hydrolytic assay after rehydration in 0.1 M
phosphate buffer pH 7.
Penicillin acylase from A. faecalis is active in water-
miscible solvents at low water content. The enzyme
catalysed the synthesis of ampicillin in these solvents at a
rate that is comparable with the same reaction in water.
Comparison of several solvents and water contents
showed that the ratio between ampicillin synthesis and
formation of the hydrolytic side product d-PG varied
stronglywith the nature of the solvent but the water
content had almost no effect on this ratio. The discovery
that a suitable immobilised penicillin acylase is active in
low-water organic media provides possibilities for new
applications of this readilyavailable enzyme.
Enzymatic Ampicillin Synthesis
To 20 mL of an organic solvent-water mixture 6-aminopeni-
cillanic acid (6-APA, 1 mmol), d-(À)-phenylglycine amide
(3 mmol) (PGA) and of penicillin acylase CLEA (130 U) were
added. The reaction mixture was stirred in a closed reaction
vessel at 08C.
Analysis
Experimental Section
Aliquots of the reaction mixture were taken during the
reaction, dissolved in water-ethanol (3:1, v/v) and acidified
with phosphoric acid; 1,3-dimethoxybenzene was added as
internal standard. Samples were analysed by HPLC using a
Waters M6000 pump, a 5 m, 3 Â 150 mm Nucleosil C18 column
and a Shimadzu SPD 6A UV detector at 215 nm. The eluent
Chemicals
Penicillin acylase solution from A. faecalis and 6-aminopeni-
cillanic acid were obtained from (DSM Anti-Infectives Delft,
800
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2003, 345, 797 801

Penicillin Acylase Catalysed Synthesis of Ampicillin
FULL PAPERS
was prepared byadjusting a 0.68 g L À1 solution of KH2PO4 in
acetonitrile-water (30:70, v/v) containing 0.68 g LÀ1 sodium
dodecyl sulphate to pH 3.0 with phosphoric acid. The flow rate
was 1.0 mL minÀ1.
[12] M. Kulhanek, M. Tadra, Folia Microbiol. 1968, 13, 340
¬
345.
[13] R. M. D. Verhaert, A. M. Riemens, J. M. van der Laan, J.
van Duin, W. J. Quax, Appl. Env. Microb. 1997, 63,
3412 3418.
œ
[14] V. Svedas, D. Guranda, L. van Langen, F. van Rantwijk,
Acknowledgements
R. Sheldon, FEBS Lett. 1997, 417, 414 418.
[15] L. M. van Langen, M. H. A. Janssen, N. H. P. Oosthoek,
œ
Generous donations of A. faecalis penicillin acylase by DSM
Anti-infectives (Delft, The Netherlands) are gratefully acknowl-
edged. Thanks are due to Prof. Dr. Vytas Svedas and Dr. Maxim
Youshko of the Lomonosov Moscow State University for
assistance with the kinetic model. The authors express their
thanks to Prof. Dr. Alle Bruggink and Dr. Erik de Vroom of
DSM for helpful discussions and encouragement. This work
was financially supported by the Netherlands Ministry of
Economic Affairs (contract grant number SSC 95010) and
DSM Life Science Products.
S. R. M. Pereira, V. K. Svedas, F. van Rantwijk, R. A.
Sheldon, Biotechnol. Bioeng. 2002, 79, 224 228.
[16] P. J. Halling, Enzyme Microb. Technol. 1994, 16, 178
206.
[17] H. J. Duggleby, S. P. Tolley, C. P. Hill, E. J. Dodson, G.
Dodson, P. C. E. Moody, Nature 1995, 373, 264 268.
[18] S. Dixit, J. Crain, W. K. C. Poon, J. L. Finney, A. K.
Soper, Nature 2002, 416, 829 831.
œ
[19] For comparison we calculated the course of the ampi-
cillin synthesis in aqueous medium at otherwise similar
œ
conditions, using the kinetic model drawn up byS vedas
et al. (see ref[20]):
References and Notes
A. faecalis enzyme: initial S/H ˆ 0.96, maximum con-
version 28%;
[1] A. Bruggink, P. D. Roy, in Synthesis of b-Lactam Anti-
biotics, (Ed.: A. Bruggink), 2001, Kluwer, Dordrecht,
pp. 12 55.
[2] M. A. Wegman, M. H. A. Janssen, F. van Rantwijk, R. A.
Sheldon, Adv. Synth. Catal. 2001, 343, 559 576.
[3] J. Verweij, E. de Vroom, Recl. Trav. Chim Pays-Bas 1993,
112, 66 81.
[4] A. Bruggink, E. C. Roos, E. de Vroom, Org. Proc Res.
Dev. 1998, 2, 128 133.
[5] A. L. Margolin, V. K. Svedas, I. V. Berezin, Biochim.
E. coli enzyme: initial S/H ˆ 2.7, maximum conversion
43%.
œ
[20] M. I. Yousko, V. K. Svedas, Biochemistry (Moscow)
2000, 65, 1367 1375; M. I. Youshko, G. G. Chilov, T. A.
œ
Shcherbakova, V. K. Svedas, Biochim. Biophys Acta
2002, 1599, 134 140.
[21] L. Cao, L. M. van Langen, F. van Rantwijk, R. A.
Sheldon, J. Mol. Catal. B: Enzym. 2001, 11, 665 670.
[22] C. Hansch, A. Leo, D. Hoekman, Exploring QSAR;
hydrophobic, electronic, and steric constants, American
Chemical Society, Washington, 1995.
[23] C. Reichardt, Chem. Rev. 1994, 94, 2319 2358.
[24] J. A. Riddick, W. B. Bunger, T. K. Sakano, Organic
Solvents. Physical properties and methods of purification,
Wiley, New York, 4th edn., 1986.
[25] See: F. van Rantwijk, M. Woudenberg-van Oosterom,
R. A. Sheldon, J. Mol Catal. B: Enzym. 1999, 6, 511
532.
[26] See, for example: M. C. de Zoete, A. C. Kock-van Dalen,
F. van Rantwijk, R. A. Sheldon, J. Mol. Catal. B: Enzym.
1996, 2, 19 25.
œ
Biophys. Acta 1980, 616, 283 289; M. B. Diender, A. J. J.
Straathof, L. A. M. van der Wielen, C. Ras, J. J. Heijnen,
J. Mol. Catal. B: Enzym. 1998, 5, 249 253; C. G. P. H.
»
Schroen, V. A. Nierstrasz, P. J. Kroon, R. Bosma,
A. E. M. Janssen, H. H. Beeftink, J. Tramper, Enzyme
Microb. Technol. 1999, 24, 498 506.
[6] Such a procedure was thought to be impossible but was
recentlyshown to be feasible when the product is
precipitated as its zinc salt, see: R. V. Ulijn, L. De
Martin, P. J. Halling, B. D. Moore, A. E. M. Janssen, J.
Biotechnol. 2002, 99, 215 222.
[7] V. Kasche, Enzyme. Microb. Technol. 1986, 8, 4 16.
[8] A. Illanes, A. Fajardo, J. Mol. Catal. B: Enzym. 2001, 11,
587 595; A. Illanes, S. Anjari, R. Arrieta, C. Aguirre,
Appl. Biochem. Biotechnol. 2002, 97, 165 179.
[9] F. van Rantwijk, M. A. P. J. Hacking, R. A. Sheldon,
Monatsh. Chem. 2000, 131, 549 569.
[10] a) C. Ebert, L. Gardossi, P. Linda, Tetrahedron Lett.
1996, 37, 9377 9380; b) C. Ebert, L. Gardossi, P. Linda,
J. Mol. Catal. B: Enzym. 1998, 5, 241 244; c) L. De
Martin, C. Ebert, G. Garau, L. Gardossi, P. Linda, J. Mol.
Catal. B: Enzym. 1999, 6, 437 445.
[27] M. I. Youshko, L. M. van Langen, E. de Vroom, F.
œ
van Rantwijk, R. A. Sheldon, V. K. Svedas, Biotechnol.
Bioeng. 2001, 73, 426 430.
[28] This is a well-recognised hydrolysis pathway; see, for
example: M. I. Youshko, G. G. Chilov, T. A. Shcherba-
œ
kova, V. K. Svedas, Biochim. Biophys. Acta 2002, 1599,
134 140.
[29] L. M. van Langen, E. de Vroom, F. van Rantwijk, R. A.
Sheldon, Green Chem. 2001, 3, 316 319.
[30] L. M. van Langen, Thesis, Delft Universityof Technol-
ogy, 2001, p 103.
[11] L. Cao, F. van Rantwijk, R. A. Sheldon, Org. Lett. 2000,
2, 1361 1364.
Adv. Synth. Catal. 2003, 345, 797 801
asc.wiley-vch.de
¹ 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
801