Aldehyde-based racemization in the dynamic kinetic resolution of N-heterocyclic α-amino esters using Candida antarctica lipase A

Article abstract of DOI:10.1016/j.tet.2003.10.103

The present research introduces approaches for the dynamic kinetic resolution of the methyl esters of proline and pipecolic acid. As the result, a method was developed which is based on the acylation of the secondary amino group of the amino esters with vinyl butanoate by Candida antarctica lipase A. In the optimized method, acetaldehyde as a racemizing agent is released in situ from vinyl butanoate in the presence of triethylamine, allowing ca. 90% of the racemic proline and 70% of the pipecolic acid methyl esters to be acylated in the forms of highly enantiopure (ee=97%) butanamides with the S-absolute configurations.

Full text of DOI:10.1016/j.tet.2003.10.103

Tetrahedron 60 (2004) 671–677
Aldehyde-based racemization in the dynamic kinetic resolution of
N-heterocyclic a-amino esters using Candida antarctica lipase A
*
Arto Liljeblad, Anu Kiviniemi and Liisa T. Kanerva
¨
Laboratory of Synthetic Drug Chemistry, Department of Chemistry, University of Turku, Lemminkaisenkatu 2, FIN-20520 Turku, Finland
Received 11 July 2003; revised 15 August 2003; accepted 17 October 2003
Abstract—The present research introduces approaches for the dynamic kinetic resolution of the methyl esters of proline and pipecolic acid.
As the result, a method was developed which is based on the acylation of the secondary amino group of the amino esters with vinyl butanoate
by Candida antarctica lipase A. In the optimized method, acetaldehyde as a racemizing agent is released in situ from vinyl butanoate in the
presence of triethylamine, allowing ca. 90% of the racemic proline and 70% of the pipecolic acid methyl esters to be acylated in the forms of
highly enantiopure (ee¼97%) butanamides with the S-absolute configurations.
q 2003 Elsevier Ltd. All rights reserved.
1. Introduction
elevated temperatures.6,9,10 Traditional racemization by
aldehydes is assisted by metal catalysts.11 Metal catalysts
also alone are able to provoke racemization particularly in
alkaline solutions.12
N-Heterocyclic amino acids, such as proline and pipecolic
acid, are valuable building blocks of many pharmaceuti-
cals.1 In our previous paper the enzymatic kinetic resolution
of methyl pipecolinate (rac-2; Scheme 1) as a non-
proteinogenic a-amino ester was studied.2 As the most
important result, Candida antarctica lipase A (CAL-A) was
reported to catalyze the acylation of the secondary ring
nitrogen with 2,2,2-trifluoroethyl esters in a highly enantio-
selective manner in organic solvents (enantioselectivity
ratio E.100).
Several biocatalytic dynamic kinetic resolutions of amino
acids have been published including both enzymatic and
chemical racemization steps. Especially oxazolones, thia-
zolones and hydantoins have been used as a starting
material.13 Among amino acid racemases, hydantoin and
N-acylamino acid racemases have previously gained
significance in industrial dynamic kinetic resolution pro-
cesses.14 Regarding aldehyde-based methods, salicylalde-
hyde and pyridoxal have been utilized for the racemization
of phenylglycine methyl ester during the lipase-catalyzed
ammoniolysis in organic solvents.15,16 The pyridoxal
5-phosphate-catalyzed racemization of amino esters
together with alcalase-catalyzed resolution is also known.7
Moreover, Schiff bases of a-amino esters have served as
starting materials for enzyme-catalyzed ester hydrolysis.8,17
Dynamic kinetic resolution enables the transformation of a
racemate to a new product as one enantiomer. This becomes
possible through the racemization of one of the enantiomers
whereas the other reacts to a stable product. Most racemizing
methods of a-amino acids are based on the acidic a-hydrogen
at an amino acid or its derivative. This allows enol formation
under acidic and enolate formation under basic conditions.
However, free amino acids are generally difficult to racemize.
Derivatization of an amino acid as a Schiff base, hydantoin or
oxazolone results in a more acidic a-proton, and accordingly
promotes racemization.3–5 Acidic and basic conditions are
known to aid aldehyde-promoted racemization through a
Schiff base.6–8 The previously published racemizations of
amino acids or esters including the racemization of (S)-proline
were effectively performed in acetic acid or in the mixture of
acetone–acetic acid using various aldehydes or ketones at
The above results prompted us to start work in order to
develop an aldehyde-based dynamic kinetic resolution
method for N-heterocyclic amino esters rac-1 and rac-2 in
the presence of CAL-A and an acyl donor (Scheme 1). In
order to study the general usability of CAL-A, the normal
kinetic resolution of rac-1 and rac-2 as secondary amines
and that of some common a-amino esters as primary amines
with 2,2,2-trifluoroethyl butanoate have been investigated.
Racemization of (S)-1 and (S)-2 in the presence of additives
(acids or bases) and aldehydes has been studied. Finally, the
dynamic kinetic resolution method was optimized and the
preparative scale experiment performed using racemic
proline methyl ester as a substrate.
Keywords: Dynamic kinetic resolution; Proline methyl ester; Pipecolic acid
methyl ester; Candida antarctica lipase A; Aldehyde-based racemization.
*
Corresponding author. Tel.: þ358-2333-6773; fax: þ358-2333-7955;
e-mail address: lkanerva@utu.fi
0040–4020/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2003.10.103

672
A. Liljeblad et al. / Tetrahedron 60 (2004) 671–677
Scheme 1.
2. Results and discussion
racemic compound with 2,2,2-trifluoroethyl butanoate in
TBME smoothly proceeded to 53% conversion in 3 h. On
the above basis and because 2,2,2-trifluoroethyl butanoate
as an acyl donor (entries 16 and 19) was more effective than
the corresponding acetate (entries 18 and 21) the studies
were continued using butanoates as achiral acyl donors and
rac-1 and rac-2 as substrates.
In order to develop a satisfactory method for dynamic
kinetic resolution where the product enantiomer is formed
with high yield and enantiomeric excess, the resolution
reaction should be highly enantioselective and the less
reactive enantiomer rapidly racemized under the conditions
where the product of an enzyme-catalyzed reaction remains
stable and the enzyme catalytically active. Under such
conditions, the starting material is always practically
racemic, allowing the maximal enantiopurity (as determined
by the E value of the corresponding kinetic resolution) to be
reached at the theoretical zero conversion for the more
reactive enantiomer throughout the reaction. Thus, it can be
calculated that in order to achieve the product with
ee.95%, the E value for kinetic resolution should be 40
or higher.18
2.2. Racemization experiments
Racemization of amino esters with aldehydes is a process
where an aldehyde reacts with the starting material in the
formation of carbinolamine 5 which in the presence of an
acid forms a Schiff base 6 (Scheme 1). The subsequent base-
catalyzed release of a proton is responsible for the
deterioration of enantiopurity in the formation of the
carbanion intermediate 7.3,7,8 When racemization is planned
to proceed in situ at the same time with the enzymatic
acylation mild reaction conditions are necessary for the
enzyme to stay active. In the present work, methyl ester
(S)-1 (0,1 M, ee¼99%) and acetaldehyde (0.1 M) were
chosen for racemization in TBME with and without
additives (Table 2). Carboxylic acids and ammonium
acetate (entries 2–4) clearly favour racemization while
bases such as triethylamine (entry 6) do not have an impact
over the effect of acetaldehyde alone (entry 1). As another
observation, the amount of 1 clearly decreases in the
presence of carboxylic acids (entries 2 and 3). Salt
formation between 1 and the acid is one explanation for
the decrease. Indeed, some precipitation was observed in the
mixture of 1 or 2 (0.1 M) and acetic or butanoic acid (0.1 M)
in TBME at room temperature while the precipitate
disappeared at elevated temperature (48 8C). Another
possible explanation for the loss of an amino ester is the
stabilization of a species such as 6 through salt formation.
The loss of 1 when no additives were added shows that part
of the starting material is in the form of a Schiff base (entry
1). The formation of various condensation products
(although not detected by the present GC method) cannot
be totally ruled out either.
2.1. Kinetic resolution
We have shown earlier that CAL-A catalyzes highly
enantioselective N-acylations between b-amino esters and
achiral esters with absolute chemoselectivity while some
other common lipases tend to lead interesterification
products.19 – 22 We have also shown that CAL-A is
exceptional among lipases by catalyzing acylations of the
secondary amino group of (S)-2.2 In order to find the general
usability of CAL-A for the resolution of a-amino esters,
various methyl esters (Table 1) were subjected to enzymatic
resolution under the best conditions of the previous works.
According to the results in Table 1, enantioselectivity
requirements for a successful dynamic kinetic resolution are
properly fulfilled only for the acylation of sterically
hindered N-heterocyclic amino esters 1 and 2 with 2,2,2-
trifluoroethyl esters (entries 16–21). Aspartic acid dimethyl
ester as a primary amine is also acceptable with E¼33 and
with a calculated theoretical ee¼94% at zero conversion for
the (S)-amide product (entry 1). However, the observed high
stability against racemization did not persuade us to
continue studies for its dynamic kinetic resolution. Thus,
there was only a small drop in ee from 99 to 73% when the
(S)-amino ester (0.1 M) was incubated in TBME (tert-butyl
methyl ether) for 24 h while the enzymatic acylation of the
Different aldehydes were next tested for the acetic acid-
catalyzed racemization of (S)-1 in TBME (Table 3).

A. Liljeblad et al. / Tetrahedron 60 (2004) 671–677
673
Table 1. CAL-A-catalyzed acylation at room temperature of the methyl esters of a-amino acids (0.1 M) with trifluoroethyl butanoate (0.2 M) in TBME (A) and
in CH3CN (B) and with butyl butanoate as a solvent and an acyl donor (C)
Entry
Methyl ester of
Conditions
E
N-Acylated product formed (%)
Time (h)
1
2
3
4
5
6
7
8
aspartic acid
aspartic acid
aspartic acid
glutamic acid
glutamic acid
glutamic acid
methionine
methionine
methionine
phenylglycine
phenylglycine
phenylglycine
valine
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
C
A
A
B
A
33^2
20^1
17^1
12^1
12^1
13
10^1
7^1
8^0
7^0
1.5^0.1
6^0
15^1
9^0
18^0
.100
.100
.100
.100
.100
.100
16a
19a
9a
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
9
37b
33b
53c
49c
100c
90c
77c
54c
35c
38c
44c
24c
50c
44c
2
9
10
11
12
13
14
15
16
17
18
19
20
21
valine
valine
proline (1)
proline (1)
prolined (1)
pipecolic acide (2)
pipecolic acide (2)
pipecolic acidd,e (2)
49c
39c
29c
22
24
a
5 mg/mL of the enzyme preparation.
10 mg/mL of the enzyme preparation.
75 mg/mL of the enzyme preparation.
CH3CO2CH2CF3 as an acyl donor.
Ref. 2.
b
c
d
e
Acetaldehyde (entry 2) is clearly most favourable for the
purpose. Increasing acetic acid (entries 1–4, Table 4) and
acetaldehyde (entries 5–9) concentrations have a positive
effect on racemization although acetic acid (0.1 M, entry 5)
alone is not effective. Increasing acetaldehyde concen-
trations also decrease the amount of 1 in the presence of
acetic acid (0.1 M; entries 5–9). This indicates improved
possibilities for the adducts like 6 and/or improved
possibility for side reactions. As the best compromise,
1 equiv. of acetaldehyde and acetic acid with respect to the
starting material were chosen for the subsequent dynamic
kinetic resolution experiments.
rac-1 and rac-2 with 2,2,2-trifluoroethyl butanoate and
CAL-A smoothly proceeds in a highly enantioselective
manner with E.100 in TBME (entries 16 and 19). High
enantiopurity is also observed for the formed (S)-3 when
vinyl butanoate is used in the place of 2,2,2-trifluoroethyl
butanoate although the acylation of rac-1 then is approach-
ing toward the conditions of dynamic kinetic resolution due
to the acetaldehyde produced from vinyl alcohol (Fig. 1;
Table 5, entry 2). Thus, after an hour 75% of the racemate
(X) was transformed to the product (B) with ee(S)-3¼98%
and at this point the reaction mixture contained only 4% of
the unreacted starting material. The enantiopurity of (S)-4 is
also excellent (entry 8) for the acylation of rac-2 with vinyl
butanoate, but the yield of the product only slightly
increases after 40% is reached (Fig. 2). It is clear according
to Figures 1 and 2 that a considerable amount of the starting
material has disappeared somewhere.
Contrary to the above results for proline methyl ester, the
racemization of (S)-2 proceeds slowly in the presence of
acetaldehyde (0.1 M) and various acetic acid concentrations
in TBME (Table 4, entries 10–12). Elevated temperature
(48 8C) favours racemization and also leads to the enhanced
disappearance of the starting material (entries 13–15).
Three different strategies were now chosen for improving
the yield of the dynamic kinetic resolution of rac-1 and rac-
2. In method I, acetic acid and acetaldehyde (both 0.1 M)
were added to the reaction mixture where 2,2,2-trifluoro-
ethyl butanoate served as an acyl donor. In methods II and
III vinyl butanoate served as an acyl donor and as the in situ
2.3. Dynamic kinetic resolution
As shown in Table 1, the traditional kinetic resolution of
Table 2. Effect of additives on the racemization of (S)-1a (0.1 M) with
acetaldehyde (0.1 M) in TBME after 24 h at room temperature
Table 3. Effect of different aldehydes on the racemization of (S)-1a (0.1 M)
in the presence of acetic acid (0.1 M) in TBME after 2 h at room
temperature
Entry
Additive (0.1 M)
ee (%)
1 left (%)
Entry
Aldehyde (0.1 M)
ee (%)
1 left (%)
1
2
3
4
5
6
No additive
Acetic acid
Butanoic acid
CH3CO22 NHþ4
DMAPb
72
0
0
3
72
76
92
86
85
96
98
93
1
2
3
4
5
Pivalaldehyde
Acetaldehyde
Pyridoxal
Benzaldehyde
4-Methylbenzaldehyde
96
0
97
10
4
100
95
29
93
Triethylamine
89
a
Original ee(S)-1¼99%.
b
a
N,N-Dimethylaminopyridine.
Original ee(S)-1¼99%.

674
A. Liljeblad et al. / Tetrahedron 60 (2004) 671–677
Table 4. Effects of acetaldehyde and acetic acid concentrations on the
racemization of (S)-1a (0.1 M) and (S)-2b (0.1 M) in TBME after 2 h at
room temperature
preparation, always giving some butanoic acid in the
reaction mixture independent of which method is
considered.
Entry
Acetic acid
(M)
Acetaldehyde
(M)
ee(S)-1 or (S)-2
(%)
1 or 2 left
(%)
In accordance with the fast racemization in the case of (S)-1
with acetic acid (Table 4, entry 4), rac-1 (0.1 M) was
effectively transformed to (S)-3 in TBME using either
method I or II (Table 5, entries 1, 3, and 4). The same
methods were useless for the dynamic kinetic resolution of
rac-2 (entries 7 and 9). Accordingly, the reaction by method
I almost stopped at ca. 30% yield for (S)-4 after 16 h and
thereafter was slowly approaching to 43% yield at 75% total
conversion in 3–4 days. The reaction closely followed the
progression curves shown in Figure 2 where no acid was
added. There was no increase in the amount of (S)-4 either
when normal kinetic resolution was allowed to proceed first
to 48% conversion (ee(R)-2¼91% and ee(S)-4¼99%) before
acetic acid (0.1 M) and acetaldehyde (0.1 M) were added.
Instead and as expected, the less reactive enantiomer was
subject to slow racemization (ee(R)-2¼73%, ee(S)-4¼99%
after 70 h). Under the conditions of method II, the
enzymatic acylation of rac-2 stopped when only 25% of
(S)-4 was obtained (ee(S)-4¼96%) at the point where 33%
(ee(R)-2¼3) of the starting material was left unreacted (entry
9). Replacing acetaldehyde with benzaldehyde did not
alleviate the problem and the acylation of rac-2 in the
presence of acetic acid (0.1 M) stopped at 28% yield for the
product. The above-suggested improved existence of stable
adducts like 6 with an added acid (or that formed from an
acyl donor) and the precipitation of amino esters as acetate
(or butanoate) salts can explain where at least part of the
original substrate disappeared.
1
2
3
4
0
0.1
0.1
0.1
0.1
1; 86
1; 79
1; 27
1; 0
95
96
96
95
0.01
0.05
0.1
5
6
7
8
9
0.1
0.1
0.1
0.1
0.1
0
0.05
0.1
0.25
0.5
1; 99
1; 66
1; 0
1; 1
1; 2
97
100
95
86
46
10
11
12
13
14
15
0
0.1
0.5
0.1
0.1
0.1
0.1
0.1
0.1
2; 95
2; 93
2; 83
2; 90
2; 82
2; 76
100
98
93
100
97
74
0.1c
0.25c
0.5c
a
Original ee(S)-1¼99%.
Original (S)-2¼95%.
Temperature 48 8C.
b
c
source of acetaldehyde. In method II, the reaction proceeded
in the presence of acetic acid (0.1 M) with the purpose to
enhance racemization. In method III, triethylamine (or
ammonium acetate) was added to the mixture of a substrate
(0.1 M) and vinyl butanoate (0.2 or 0.4 M) in TBME in
order to enhance racemization and to bind the liberated
butanoic acid. It is worth mentioning that 2,2,2-trifluor-
oethyl and vinyl butanoates as activated esters are partly
hydrolyzed by water in the seemingly dry enzyme
Figure 1. Progression curves for the formation of (S)-3 (B) and disappearance of rac-1 (X) when rac-1 (0.1 M) reacts with vinyl butanoate (0.2 M) in TBME in
the presence of CAL-A preparation (75 mg/mL).

A. Liljeblad et al. / Tetrahedron 60 (2004) 671–677
675
Table 5. Dynamic kinetic resolution of rac-1 and rac-2 (both 0.1 M) with PrCO2R in TBME in the presence of CAL-A-preparation (75 mg/mL) and additives
(0.1 M)
Entry
Substrate
R
ee(S)-1 or (S)-2
(S)-3 or (S)-4 formeda (%)
Time (h)
Temperature (8C)
Additive
b
1
2
3
4
5
6
7
8
rac-1
rac-1
rac-1
rac-1
rac-1
rac-1
rac-2
rac-2
rac-2
rac-2
rac-2
rac-2
CH2CF3
99
98
96
97
97
97
97
97
96
97
97
97
97
75
82
88
86
86
43
39
25
50
61
69
25
1
3
3
1
25
25
25
48
25
25
48
25
48
25
48
56
CH3CO2H
None
CH3CO2H
CH3CO2H
CH3CO22 NHþ4
TEA
CH3CO2H
None
CH3CO2H
TEA
TEA
b
b
b
c
c
CHvCH2
CHvCH2
CHvCH2
CHvCH2
CHvCH2
1
b
CH2CF3
187
24
45
24
24
24
b
b
c
c
c
CHvCH2
CHvCH2
CHvCH2
CHvCH2
CHvCH2
9
10
11
12
TEA
a
Values indicate maximum yields obtained in the reaction.
0.2 M.
0.4 M.
b
c
Finally in method III we decided to use base-induced
racemization of the Schiff base (although the racemization
tests of Table 2 were not promising) and to neutralize the
formed butanoic acid by adding triethyl amine (0.1 M) in
the enzymatic reaction mixture of rac-1 and rac-2 with
vinyl butanoate in TBME. At this stage the concentration of
vinyl butanoate was also doubled to 0.4 M. Surprisingly, the
addition of triethyl amine caused ca. 10% enhancements in
the product yield of (S)-3 (Table 5, entry 6) compared to the
case where no additives were present (entry 2). The same
result was obtained when 0.05 and 0.1 M triethylamine was
used. The presence of ammonium acetate affected in the
same way (entry 5). For the acylation of rac-2, the positive
effect of triethyl amine was even more pronounced and still
increased at elevated temperature. Finally, ca. 70% of rac-2
was transformed to (S)-4 by refluxing the mixture of rac-2
(0.1 M), vinyl butanoate (0.4 M) and CAL-A preparation
(75 mg/mL) in TBME in the presence of triethyl amine
(0.1 M) (entry 12).
From an economical point of view, vinyl acetate should be
preferred over vinyl butanoate as an achiral acyl donor.
When vinyl butanoate was replaced with vinyl acetate in the
presence of triethyl amine 84% of initially rac-1 was
transformed to (S)-3 with only 92% ee in 24 h at 48 8C. The
reaction at room temperature stopped when 77% product
was yielded. Due to higher enantiopurity, the dynamic
kinetic resolution of rac-1 in gram-scale was finally
performed with vinyl butanoate as described in Section 4
using method III.
Figure 2. Progression curves for the formation of (S)-4 (B) and disappearance of rac-2 (X) when rac-2 (0.1 M) reacts with vinyl butanoate (0.2 M) in TBME in
the presence of CAL-A preparation (75 mg/mL).

676
A. Liljeblad et al. / Tetrahedron 60 (2004) 671–677
3. Conclusions
determined by synthesizing an enantiopure amino ester
from a commercially available amino acid and by compar-
ing the peak of the enantiomer with the peaks of a racemic
compound in a chromatogram.
The previously reported2 exceptional property of lipase A
from Candida antarctica to catalyze highly enantioselective
N-acylations of N-heterocyclic amino esters as secondary
amines has been confirmed in the present work. This
property was used for dynamic kinetic resolution studies
where the methyl esters of proline (rac-1) and pipecolic acid
(rac-2) have served as racemic substrates and vinyl
butanoate and 2,2,2-trifluoroethyl butanoate as achiral acyl
donors in TBME (Tables 1 and 5). Negligible to moderate
enantioselectivity with E values from 1 to 33 did not give
reasons to use CAL-A for the dynamic kinetic resolution of
various natural a-amino esters through the acylation of
primary amino groups (Table 1).
CAL-A was the product of Roche (Chirazyme L5, lyo.).
Before use, CAL-A (20% (w/w)) was adsorbed on Celite by
dissolving the enzyme (5 g) and sucrose (3 g) in Tris–HCl
buffer (250 mL, 20 mM, pH 7.8) followed by the addition of
celite (17 g). The mixture was dried by letting water
evaporate. The enzyme preparation gave the initial velocity
of 0.028 mmol/min/g for the acylation of racemic valine
methyl ester (0.1 M) with trifluoroethyl butanoate (0.2 M) in
TBME (E¼15.5^0.2).
1H and 13C NMR spectra were measured in CDCl3 on a Jeol
Lambda 400 Spectrometer tetramethylsilane being as an
internal standard. MS-spectrum was recorded on a VG
Analytical 7070E instrument equipped with a VAXstation
3100 M76 computer. Optical rotation was measured using a
Jasco DIP-360 polarimeter. The determination of E was
based on equation E¼ln[(12c)(12eeS)]/ln[(12c)(1þ
eeS)].18 Using linear regression E was achieved as the
slope of a line.
For the enzymatic acylation of rac-1 with vinyl esters the
reaction was clearly approaching the conditions of dynamic
kinetic resolution with 75% yield for (S)-3, showing the
aldehyde-based racemization ability of the less reactive
(R)-1 under the acylation conditions (Scheme 1; Fig. 1;
Table 5, entry 2). For the acylation of rac-2 the reaction
tended to stop when only 40% of the starting material was
transformed to (S)-4. Three different strategies were chosen
for improving the yields. In methods I and II, acetic acid
served to catalyze aldehyde-based racemization. The
methods failed badly with rac-2 as the substrate. Salt
formation between the acid and amino esters and/or
intermediates such as 6 was suggested to explain the
observed loss of the substrate and accordingly the low yield
for the produced amide (S)-4. In method III, triethyl amine
was added to the acylation mixture of rac-1 and rac-2 with
vinyl butanoate in order to bind the butanoic acid which is
liberated when the acyl donor is enzymatically hydrolyzed
by the water in the enzyme preparation. In methods II and
III the use of vinyl butanoate allows the releasing
acetaldehyde to racemize the amino ester in situ. Under
optimized conditions, ca. 90% of the racemic proline and
70% of the pipecolic acid methyl esters were acylated to
highly enantiopure (ee¼97%) butanamides with the
S-absolute configurations.
Typical reaction volume for enzymatic reactions was 1–
3 mL. Substrate (0.1 M) and an acyl donor (0.2 M) were
dissolved in TBME and an acid or base and acetaldehyde
were added. CAL-A preparation (5–75 mg/mL) started the
reaction. The progress of the reactions was followed by GC
on Chrompack CP-Chirasil-DEX CB or Chrompack CP-
Chirasil-L-Valine column by taking samples (0.1 mL) at
intervals and derivatizing them with acetic anhydride
(butanoate as an acyl donor) or butanoic anhydride (acetate
as an acyl donor) and N,N-dimethylaminopyridine in
pyridine (1% solution). N,N-dimethylaminopyridine solu-
tion was not used for those samples where aldehyde-induced
racemization was present. Quantitative analysis of the
reactions was performed by using dihexyl ether or
hexadecane as internal standards (0.1 or 0.2 M).
4.1.1. Enzymatic gram-scale resolution. Before perform-
ing a gram scale reaction the amount of the enzyme
preparation was optimized and as the best compromise
25 mg/mL was chosen. Vinyl butanoate (5.9 mL,
46.5 mmol) and rac-1 (1.50 g, 11.6 mmol) were dissolved
in TBME (116 mL). Addition of triethylamine (1.6 mL,
11.6 mmol) and CAL-A on Celite (2.904 g, 25 mg/mL)
started the reaction. The reaction was stopped when 90% of
rac-1 was transformed to (S)-3 after 5 h by filtering off the
enzyme. The crude product was purified by column
chromatography on Silicagel. After first purification
(EtOAc–petroleum ether (3/7)) the product still contained
butanoic acid. It was removed by dissolving the crude
product in methanol and by adding thionyl chloride
(200 mL) in an ice bath. After evaporation the product
was repeatedly purified by column chromatography yielding
(S)-3 (viscous liquid, 1.79 g, 78%, 9.0 mmol, ee¼97%,
[a]2D0 296.4 (c¼1.04, MeOH)).
4. Experimental
4.1. Materials and methods
Racemic proline, butyl butanoate and (S)-glutamic acid
were obtained from Acros, methyl pipecolinate hydrochlo-
ride, valine, (S)-valine methyl ester hydrochloride, racemic
and (R)-phenylglycine, racemic and (R)-methionine, tert-
butyl methyl ether and dihexyl ether from Aldrich, methanol
and petroleum ether from J. T. Baker and thionyl chloride
¨
from Riedel De-Haen. Vinyl butanoate, hexadecane, ethyl
acetate, glutamic acid and (S)-proline were the products of
Fluka. Dimethyl aspartate was a product of Sigma. All
solvents were of the best analytical grade. (S)-methyl
pipecolinate (ee¼95%) was the product of the enzymatic
kinetic resolution.2 Amino acid methyl esters were syn-
thesized by esterifying the acid with thionyl chloride in
methanol followed by bubbling with ammonia in chloro-
form. Trifluoroethyl esters were synthesized from trifluoro-
ethanol and an acid chloride. Absolute configurations were
1
(S)-3. H NMR: d (ppm) 0.99 (t, J¼7.4 Hz, 3H, CH3CH2),
1.69 (m, 7.6 Hz, 2H, CH3CH2), 1.90 (m, 2H, CH2CON),
2.0–2.4 (m, 4H, CH2CH2CHN), 3.53 (m, 2H, CH2N), 3.72

A. Liljeblad et al. / Tetrahedron 60 (2004) 671–677
677
(s, 3H, CO2CH3), 4.49 (m, J¼4.0, 8.5 Hz, 1H, CH), 13C
NMR: d (ppm) 13.7 (CH3CH2), 17.9 (CH3CH2), 24.7
(CH2CH2CH), 29.1 (CH2CH), 36.2 (CH2CO), 46.9 (CH2N),
52.0 (CO2CH3), 58.4 (CH), 171.8 (CO), 172.9 (CO). HRMS
calculated for C10H17NO3 Mþ¼199.1208. Found
Mþ¼199.1216.
`
9. Barry, L. G.; Pugniere, M.; Castro, B. Int. J. Pept. Protein Res.
1993, 41, 323–325.
10. Grigg, R.; Gunaratne, H. Q. N. Tetrahedron Lett. 1983, 24,
4457–4460.
11. Gillard, R. D.; O’Brien, P. J. Chem. Soc. Dalton Trans. 1978,
1444–1447.
12. Buckingham, D. A.; Marzilli, L. G.; Sargeson, A. M. J. Am.
Chem. Soc. 1967, 89, 5133–5138.
Acknowledgements
13. Williams, J. M. J.; Parker, R. J.; Neri, C. In Enzymatic Kinetic
Resolution. Enzyme Catalysis in Organic Synthesis; Drauz, K.,
Waldmann, H., Eds.; Wiley-VCH: Weinheim, 2002; Vol. 1, pp
287–312.
Thanks are due to the Technology Development Center in
Finland (TEKES) for financial support.
14. May, O.; Verseck, S.; Bommarius, A.; Drauz, K. Org. Process
Res. Dev. 2002, 6, 452–457.
References and notes
15. Hacking, M. A. P. J.; Wegman, M. A.; Rops, J.; van Rantwijk,
F.; Sheldon, R. A. J. Mol. Catal. B: Enzym. 1998, 5, 155–157.
16. Wegman, M. A.; Hacking, M. A. P. J.; Rops, J.; Pereira, P.;
van Rantwijk, F.; Sheldon, R. A. Tetrahedron: Asymmetry
1999, 10, 1739–1750.
1. Petersen, M.; Sauter, M. Chimia 1999, 53, 608–612.
2. Liljeblad, A.; Lindborg, J.; Kanerva, A.; Katajisto, J.;
Kanerva, L. T. Tetrahedron Lett. 2002, 43, 2471–2474.
3. Rios, A.; Crugeiras, J.; Amyes, T. L.; Richard, J. P. J. Am.
Chem. Soc. 2001, 123, 7949–7950.
17. Belokon, Y. N.; Kochetkov, K. A.; Plieva, F. M.; Ikonnikov,
N. S.; Maleev, V. I.; Parmar, V. S.; Kumar, R.; Lozinsky, V. I.
Appl. Biochem. Biotechnol. 2000, 84, 97–106.
18. Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, J. Am. Chem.
Soc. 1982, 104, 7294–7299.
4. Pietzsch, M.; Syldatk, C. In Hydrolysis and Formation of
Hydantoins. Enzyme Catalysis in Organic Synthesis; Drauz,
K., Waldmann, H., Eds.; Wiley-VCH: Weinheim, 2002; Vol.
2, pp 761–799.
19. Liljeblad, A.; Kanerva, L. T. Tetrahedron: Asymmetry 1999,
10, 4405–4415.
5. De Jersey, J.; Zerner, B. Biochemistry 1969, 8, 1967–1974.
6. Yamada, S.; Hongo, C.; Yoshioka, R.; Chibata, I. J. Org.
Chem. 1983, 48, 843–846.
´ ´
20. Gedey, S.; Liljeblad, A.; Lazar, L.; Fu¨lop, F.; Kanerva, L. T.
¨
Tetrahedron: Asymmetry 2001, 12, 105–110.
´ ´
¨
21. Kanerva, L. T.; Csomos, P.; Sundholm, O.; Bernath, G.; Fu¨lop,
7. Chen, S.-T.; Huang, W.-H.; Wang, K.-T. J. Org. Chem. 1994,
59, 7580–7581.
F. Tetrahedron: Asymmetry 1996, 7, 1705–1716.
¨
8. Parmar, V. S.; Singh, A.; Bisht, K. S.; Kumar, N.; Belokon,
Y. N.; Kochetkov, K. A.; Ikonnikov, N. S.; Orlova, S. A.;
Tararov, V. I.; Saveleva, T. F. J. Org. Chem. 1996, 61,
1223–1227.
´ ´ ´
22. Solymar, M.; Liljeblad, A.; Lazar, L.; Fu¨lop, F.; Kanerva, L. T.
Tetrahedron: Asymmetry 2002, 13, 1923–1928.