Chemoenzymatic approach to enantiopure piperidine-based β-amino esters in organic solvents

Article abstract of DOI:10.1016/j.tetasy.2007.01.027

This research concentrates on the enantioselectivities of lipase-catalysed reactions with methyl esters of 2-piperidylacetic acid and 3-piperidinecarboxylic acid derivatives. N-Acetylated 2-piperidylacetic acid methyl ester displayed good enantioselectivi

Full text of DOI:10.1016/j.tetasy.2007.01.027

Tetrahedron: Asymmetry 18 (2007) 181–191
Chemoenzymatic approach to enantiopure piperidine-based
b-amino esters in organic solvents
Arto Liljeblad, Hanna-Maija Kavenius, Petri Tahtinen and Liisa T. Kanerva*
¨
Department of Pharmacology, Drug Development and Therapeutics/Laboratory of Synthetic Drug Chemistry and
Department of Chemistry, University of Turku, Lemminka¨isenkatu 5 C, FIN-20520 Turku, Finland
Received 10 November 2006; accepted 22 January 2007
Abstract—This research concentrates on the enantioselectivities of lipase-catalysed reactions with methyl esters of 2-piperidylacetic acid
and 3-piperidinecarboxylic acid derivatives. N-Acetylated 2-piperidylacetic acid methyl ester displayed good enantioselectivity (E = 66)
in a 1:1 mixture of diisopropyl ether and butyl butanoate in the presence of lipase PS-C II from Burkholderia cepacia. The reaction is
known as interesterification with butyl butanoate rather than alcoholysis with the butanol, because butyl butanoate has to be first hydro-
lysed or go through alcoholysis with MeOH in order to release butanol. Other N-protective groups (Boc, Ns, Fmoc and Bzn) gave excel-
lent enantioselectivity (E >200) under the same conditions, and a gram-scale resolution was performed with N-Boc-2-piperidylacetic acid
methyl ester. Reaction with a 3-piperidylcarboxylic acid derivative took place with disappointingly low enantioselectivity (E = 4), with
Candida antarctica lipase B being the best of the lipases screened.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
the simple ping-pong bi-bi mechanism.5 Our previously
proposed hypothesis was that the alcohol obtained from
2,2,2-trifluoroethyl ester through enzymatic action does
not leave the enzyme, but rather stays around the catalytic
site.4,6 A similar tendency for interesterification has also
been observed with some other lipases, for example, with
lipase PS (Burkholderia cepacia lipase).5
Enantiopure N-heterocyclic amino acids are important
structural units in many natural products and valuable
intermediates and starting materials in the preparation of
pharmaceutically important compounds.1 Our previous
work with N-heterocyclic a-amino acids opened a route to
highly efficient kinetic and dynamic kinetic resolution of
proline and pipecolic acid methyl esters through CAL-A-
catalysed (Candida antarctica lipase A) N-acylation at the
secondary ring nitrogen of the compounds.2,3 Under exactly
the same conditions with the same acyl donors, CAL-B (C.
antarctica lipase B) catalysed reactions specifically at the
methyl ester function in a reaction, which is known as inter-
esterification.2 Interesterification has recently been success-
fully applied in the activation of various b-amino acids as
2,2,2-trifluoroethyl esters for b-peptide synthesis by using
2,2,2-trifluoroethyl butanoate as the in situ source of the
alcohol.4 Typical to these reactions, alcoholysis with
2,2,2-trifluoroethanol did not lead to the desired 2,2,2-tri-
fluoroethyl ester formation, suggesting more complicated
mechanistic details than can be anticipated on the basis of
Inspired by these observations, we prepared piperidyl-based
b-amino acid derivatives rac-1a–f and rac-2a–c and studied
their lipase-catalysed kinetic resolution in organic solvents
(Schemes 1–3). Various N-protective groups were used,
because the nature of a group is often critical when amino
acids are used for synthetic applications. N-Acylation of
rac-1a and rac-2a by CAL-A (A), and aminolysis (B), alco-
holysis (C) and interesterification (D) at the ester function
of rac-1b (also 1c for aminolysis) have been used as model
reactions in order to determine the conditions and the best
reaction type for gram-scale resolution. The highly enantio-
selective hydrolysis of rac-1c by lipase PS,7 the hydrolysis of
methyl N-benzyl-2-piperidylacetic acid by CAL-B8 and the
practically non-selective pig liver esterase-catalysed hydro-
lysis of rac-2b9 have been described earlier. However, the
possibilities allowed by lipase-catalysed reactions in organic
solvents are not known so far. Herein, we report the poten-
tial and challenges in interesterification, when the product is
*
Corresponding author. Tel.: +358 2 333 6773; fax: +358 2 333 7955;
e-mail: lkanerva@utu.fi
0957-4166/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2007.01.027

182
A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
formed from a mixture of butyl butanoate and a b-amino
ester in the presence of a lipase in an organic solvent.
Throughout this article we distinguish the terms alcoholysis
[a reaction between an ester (RCO2R1) and an alcohol
(R2OH)] and interesterification [a reaction whereby two
esters (RCO2R1 and R3CO2R2) have changed their alkyl
groups in the products formed], although in both reactions,
the alcohol R2OH evidently serves as a nucleophile in the
deacylation step of the bi-bi ping-pong mechanism of serine
hydrolases, including lipases.10
key step of the syntheses was the catalytic hydrogenation
of 2-pyridylacetic acid 3 and nicotinic acid 6. The acids
were quantitatively transformed into the corresponding
rac-4 and rac-7 by a known method.9
The structures of the compounds prepared were established
by 1H and 13C NMR spectroscopic methods. Although
compounds 1c and 2b are previously known, their stereo-
chemistry has not been analysed in detail.7–9 NMR spectro-
scopic studies revealed two isomers in equilibrium with
each other for 1b, 1c, 1e, 1f, 2b and 2c. The isomers corre-
spond to the E- and Z-isomers with respect to the partial
double bond of the amide moiety (Scheme 4). For the N-
acetyl derivatives 1b and 2b, the E- and Z-isomers gave sep-
arate NMR spectra at 25 ꢁC, whereas a lower temperature
(ꢀ50 ꢁC used) was needed for the carboxamide derivatives.
Thus, the rotation of the N-substituent is less restricted,
indicating a lower double bond degree in the latter.
2. Results and discussion
2.1. Preparation of racemic compounds
Racemic amino esters 1a–f and 2a–c were synthesised
according to the routes shown in Schemes 1 and 2. The
H2 (g) / PtO2
CO2H
CO2H
H2O / HCl
N
N
Cl-
H2
4
3
CO2Me
N
H
NH3
CHCl3
SOCl2
MeOH
1a
CbzCl, DMAP,
Ac2O, DMAP,
TEA
CO2Me
1f
N
Cbz
CO2Me
CO2Me
TEA, CHCl3
N
Ac
N
Cl-
H2
1b
5
NaH
Boc2O
DMAP
TEA
FmocCl
THF
NsCl
TEA
CH2Cl2
CO2Me
1e
CO2Me
1c
N
Fmoc
N
Boc
CO2Me
1d
N
Ns
Scheme 1. Preparation of racemic 2-piperidylacetic acid derivatives 1a–f.
CO2Me
N
H
NH3
2a
CHCl3
CO2Me
CO2H
CO2H
CO2Me
H2 / PtO2
H2O / HCl
SOCl2
MeOH
Ac2O, DMAP
TEA
N
N
N
N
H2Cl-
6
7
H2Cl-
8
2b
O
Boc2O
DMAP
TEA
CO2Me
N
Boc
2c
Scheme 2. Preparation of racemic 3-piperidinecarboxylic acid derivatives 2a–c.

A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
183
+
CO2Bu
CO2Bu
MeO2C
MeO2C
CO2Me
N
H
(R)-9a
N
H
N
COPr
(S)-1a
C. Butanolysis
BuOH
A. Acylation
PrCO2CH2CF3
enzyme
enzyme
X
+
D. Inter-
N
R
N
R
esterification
PrCO2Bu
enzyme
(R)-1b-f
(S)-9b-f
CO2Me
N
R
rac-1a-f
B. Aminolysis
+
CONHR'
MeO2C
R'-NH2, TBME
N
R
N
R
enzyme
R' = iPr or Bu
(R)-1b
(S)-10b
(R)-1c
(S)-10c
Scheme 3. Enzymatic resolution strategies.
the hydrogens on the ring carbons. The equatorial position
of H-2 in the 1,2-disubstituted compounds 1b–f is shown
by the lack of a large 3JH-2,H-3ax in each case and, therefore,
the substituent at C-2 must remain axial. In 1a, the methyl
acetate substituent is equatorial and H-2 is axial, which is
4
5
6
3
2
CO2Me
CO2Me
1
N
N
O
O
3
shown by the large JH-2,H-3ax (11.0 Hz). Similarly, the
E
Z
methyl carboxylate substituent in 2a–c occupies the
3
Scheme 4. E/Z Isomeric equilibrium of 1b. Note, that in the carboxamide
derivatives 1c, 1e, 1f and 2c, the E/Z forms are the other way round than
in the acetamide derivatives 1b and 2b.
equatorial position, as shown by the large JH-2ax,H-3 and
3JH-3,H-4ax values (9.0–10.7 Hz).
2.2. Enzymatic kinetic resolution of 1a–f
The molar ratios of the E- and Z-isomers are collected in
Table 1. For N-acetyl derivatives 1b and 2b, the E- and
Z-isomers are easily recognised by their H-2, H-6eq, C-2
and C-6 chemical shifts. In the Z-isomers, the equatorial
H-2 is deshielded and C-2 shielded by the adjacent car-
bonyl group, whereas in the E-isomers H-6eq is deshielded
and C-6 is shielded likewise as the carbonyl group is turned
adjacent to them. In 1c, 1e, 1f and 2c, the characterisation
of the isomers is not as unambiguous, as their N-substitu-
ent possesses two oxygens deshielding both H-2eq and H-
6eq. Attempts to characterise the E- and Z-isomers of the
carboxamide derivatives with NOE measurements did not
result in further evidence.
2.2.1. N-Acylation. On the basis of the previous stud-
ies,2,3,6 the CAL-A-catalysed kinetic resolution of 1a and
2a was expected to be successful through N-acylation with
2,2,2-trifluoroethyl butanoate (Scheme 3, method A. Acyl-
ation). No reaction was observed, however, in the presence
of CAL-A. The butanolysis of 1a in neat butanol with the
majority of the other 12 lipases or lipase preparations
(shown in Section 4, method C. Butanolysis) did not pro-
ceed either. CAL-B (10 mg mLꢀ1), as an exception, gave
(R)-selective alcoholysis (67% conversion after 23 h with
ee(S)-1a = 89% and ee(R)-9 = 43%) in a reaction, which was
significantly contaminated by a chemical background reac-
tion (13% in 23 h in the absence of CAL-B). The back-
ground reaction was diminished and the reactivity greatly
enhanced (78% conversion after 40 min) by the use of butyl
butanoate as a solvent in place of butanol (method D.
Interesterification). Unfortunately the change of solvent
did not improve the enantioselectivity. As a consequence,
our attention turned to N-protected 1b–f as substrates.
Table 1. The mole ratios of the E- and Z-isomers as determined from the
peak area integrals of selected, well separated 1H or 13C NMR signals
Compound
Time (ꢁC)
E/Z
1b
1c
1e
1f
+25
ꢀ50
ꢀ50
ꢀ50
+25
ꢀ50
53:47
57:43
56:44
55:45
53:47
71:29
2.2.2. Aminolysis. Lipase-catalysed aminolysis has suc-
cessfully been used in the kinetic resolution of some
carboxylic acid derivatives and in asymmetrisation of
meso-dicarboxylic acids.11–13 Aminolysis is a fascinating
possibility for lipase catalysis, because the amide product
obtained is generally stable in the presence of lipases.
2b
2c
The piperidine ring remains in a chair conformation in
each compound, as shown by the 3J couplings between

184
A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
Table 2. Lipase (80 mg mLꢀ1)-catalysed aminolyses of 1b and 1c (0.1 M) with isopropyl- and butylamines (0.2 M) in TBME at room temperature; reaction
time 24 h
Entry
Compound
Enzyme
Amine
ees (%)
eep (%)
Conversion (%)
E
1
2
3
4
5
6
7
8
9
1b
1b
1b
1b
1b
1b
1b
1c
1c
1c
1c
CAL-B
CAL-B
PS-C II
PS-C II
PS-C II
PS-D
iPrNH2
BuNH2
2
<1
7
6
4
1
2
1
1
65
12
99
>99
>99
99
>99
<1
3
4
6
6
3
1
2
69
2
8
5
1
192
93
24
167
iPrNH2
iPrNH2
iPrNH2
iPrNH2
iPrNH2
iPrNH2
a
b
AK-C
>200
1
CAL-B
CAL-B
PS-C II
PS-C II
BuNH2
iPrNH2
BuNH2
69
>99
>99
6
10
11
8
7
>200
150
7
a iPrNH2 0.1 M.
b iPrNH2 0.5 M.
Herein, the aminolysis of 1b and 1c was studied by using
butyl- and isopropylamines as nucleophiles in the presence
of selected lipase preparations in TBME (Scheme 3, Table
2). CAL-B displayed both low reactivity (conversion
reached after a certain time) and enantioselectivity with
the substrates used (entries 1, 2, 8 and 9); aminolysis of
1b with isopropylamine being an exception with relatively
good reactivity (entry 8). Lipase PS (entries 3–6, 10 and
11) and AK (Pseudomonas fluorescens lipase; entry 7) prep-
arations gave reactions with excellent enantioselectivities,
but the reactions were too slow to be of practical value.
Attempts to enhance the reactivity by changing the amine
concentration failed (entries 3–5). As a conclusion, amino-
lysis as a reaction type was rejected.
For this reason, the ‘E’-values were all determined before
40% conversion as described in Section 4.3.
Interesterification of 1b in neat butyl butanoate (Table 3,
entry 5) with lipase PS-C II was fast, compared to that in
neat BuOH (entry 1), and it was possible to roughly double
the conversion at 47 ꢁC (entry 6). The enzyme preparation
started to become inactive at 80 ꢁC (entry 7). The effect of
temperature on the enantioselectivity was significant, ‘E’
decreasing with increasing temperature. As a compromise
between reactivity and enantioselectivity, 47 ꢁC was used
for all further interesterification optimisations. As shown
in Figure 1, increasing DIPE contents in butyl butanoate
positively affected enzymatic reactivity, ‘E’ being highest
when the 9:1 mixture of DIPE and butyl butanoate was
used. Interestingly, the (S)-enantiomer reacted faster in
the case of N-protected amino esters 1b–f, whereas it was
the (R)-enantiomer in the case of 1a (Scheme 3).
2.2.3. Alcoholysis/interesterification. BuOH (alcoholysis),
butyl butanoate (interesterification) and their mixtures
with diisopropyl ether (DIPE) were chosen as the reac-
tion media according to our previous studies with
lipases.2–6,14,15 N-Acetylated rac-1b was used as a substrate
for optimisation experiments. From the lipases screened,
only lipase PS-C II (lipase PS immobilised on Toyonite
200M)16 catalysed the alcoholysis of rac-1b in neat BuOH
(Table 3, entry 1). The reaction was extremely slow. A
higher reactivity was obtained by adding DIPE into the
reaction mixture (entries 2–4). This can be explained by sol-
vent effects and by the previously observed inhibitory effect
caused by BuOH.17,18 The effect of BuOH concentration on
the enantioselectivity was minimal, as measured by ‘E’. We
have used ‘E’ rather than E, because the enzymatic reaction
of (S)-9b with the MeOH produced back to (S)-1b can be
significant, especially at low (0.2 M) BuOH concentrations.
For the lipase-catalysed reaction (RCO2R1 + R2OH !
RCO2R2 + R1OH), the first mechanistic step includes acyl-
ation of the serine hydroxyl of the active site with RCO2R1
as an acyl donor and the liberation of the first product
R1OH.10 In the second mechanistic step, the formed acyl-
enzyme intermediate reacts with R2OH as an added nucleo-
phile, leading to the formation of the second product and
the liberation of the free enzyme. The interesterification
mixture in the present work contains two acyl donors
(rac-1b and butyl butanoate) without an added nucleo-
phile. Thus, the acyl donors liberate MeOH and BuOH,
respectively, in the formation of acyl-enzyme intermedi-
ates. This is possible only if the water in the seemingly
Table 3. Lipase PS-C II-catalysed (75 mg mLꢀ1) alcoholysis/interesterification of 1b (0.1 M); reaction time 48 h
Entry
Medium
Temp (ꢁC)
ee(R)-1b (%)
ee(S)-9b (%)
Conversion (%)
‘E’
a
a
a
1
2
3
4
5
6
7
BuOH (neat)
23
23
23
23
23
47
80
—
—
1
30
35
37
10
19
7
—
BuOH (0.6 M) in DIPE
BuOH (0.4 M) in DIPE
BuOH (0.2 M) in DIPE
PrCO2Bu (neat)
PrCO2Bu (neat)
PrCO2Bu (neat)
41
51
56
12
23
4
97
97
96
99
95
55
95
92
69 12
138
42
3
5
5
8
2
1
a Due to slow reaction ee and ‘E’ was not obtained.

A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
185
60
50
40
30
20
10
0
enzyme intermediate. The observed competitive inhibition
by BuOH is not against this hypothesis.17,18
In the next step of the optimisation, rac-1b was subjected to
interesterification in various mixtures of DIPE and butyl
butanoate in the presence of lipase PS-C II (Fig. 1).
According to the above consideration, there are enzyme
molecules containing MeOH in addition to BuOH for the
lipase PS-C II-catalysed interesterification of rac-1b with
butyl butanoate. This explains why ee(R)-1b did not exceed
the value 86% in the 9:1 mixture of DIPE–butyl butanoate
and why the reaction tended to stop near 50% conversion
(Fig. 1 (j) and 2 (j and h); Table 4, entry 1). BuOH
was added into the reaction mixture in order to suppress
possible methanolysis and/or hydrolysis of (S)-9b into the
starting material, to lead the reaction to higher conversions
and to obtain (R)-1b with better enantiopurity (Fig. 2, d
and s; Table 4, entry 2). Finally, a 1:1 mixture of DIPE
and butyl butanoate was chosen for the conditions, where
the effect of MeOH seemed to be suppressed (entry 3).
When BuOH was added to the reaction mixture, reactivi-
ties decreased somewhat while the enantioselectivity
increased, the ‘E’ values being independent of the added
alcohol concentration (entries 3–6).
DIPE:PrCO2Bu
9:1 ("E"=84 11)
c
3:1 ("E"=64 5)
1:1 ("E"=66 2)
1:3 ("E"=58 4)
1:9 ("E"=54 1)
0:1 ("E"=42 2)
0
20
40
60
80
100
120
140
160
t (h)
Figure 1. Progression curves for the lipase PS-C II-catalysed
(75 mg mLꢀ1) interesterification of rac-1b (0.1 M) in DIPE–PrCO2Bu at
47 ꢁC.
dry enzyme preparation acts as a nucleophile, hydrolysing
the acyl-enzyme intermediates. After this initiation step,
there are three competitive nucleophiles (water, MeOH
and BuOH). Our hypothesis is that the liberating BuOH
(and MeOH) must stay bound at the active site or its close
vicinity, where it can replace water or be inserted into a
possible water network. Thus, the alcohol concentration
for the reaction is at least equimolar to that of the acyl-
The N-protected substrates rac-1b–f were finally screened
under the optimised conditions in the 1:1 mixture of DIPE
and butyl butanoate at 47 ꢁC, and the results are shown
in Figure 3. The replacement of an acetyl group with
tert-butoxycarbonyl (Boc, 1c), o-nitrotoluenesulfonyl [Ns
Table 4. Effect of BuOH on the lipase PS-C II-catalysed (75 mg mLꢀ1) interesterification of 1b (0.1 M) in the mixture of DIPE and butyl butanoate at
47 ꢁC
Entry
DIPE–PrCO2Bu
BuOH (M)
t (h)
ee(R)-1b (%)
ee(S)-9b (%)
Conversion (%)
‘E’a
1
2
3
4
5
6
9:1
9:1
1:1
1:1
1:1
1:1
0
0.2
0
0.2
0.4
0.6
72
144
24
24
24
86
90
23
22
18
14
91
87
97
98
98
98
49
51
19
18
16
12
84 11
89
66
5
2
2
1
3
99
101
100
24
a E values calculated at conversions less than 40%.
100
95
90
85
80
75
70
65
60
55
50
45
40
35
30
DIPE:PrCO2Bu 9:1
DIPE:PrCO2Bu 9:1
DIPE:PrCO2Bu 1:1
DIPE:PrCO2Bu 1:1
DIPE:PrCO2Bu 1:1 + 0.2 M BuOH
DIPE:PrCO2Bu 1:1 + 0.2 M BuOH
15
20
25
30
35
40
45
50
55
60
65
70
75
c
(%)
Figure 2. Enantiomeric excesses of the substrate and the product versus conversion in various DIPE–PrCO2Bu–BuOH mixtures.

186
A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
resolution time significantly. Thus, one should be critical
of the necessity of using BuOH as a conucleophile.
50
40
30
20
10
0
2.3. Enzymatic kinetic resolution of 2a–c
The lipase PS-C II-catalysed alcoholysis and interesterifica-
tion methods described above were used for studies with
rac-2a–c (Scheme 5, Table 5). Both reactions took place
with low enantioselectivity (entries 4–6). In order to find
a more usable catalyst for the kinetic resolution, enzyme
screening for the alcoholysis of rac-2b in neat butanol
was performed. CAL-B, as the best enzyme, gave a low
‘E’ of 4 and, as a consequence, a gram-scale resolution
was not performed (entries 1–3).
1b Ac ("
1c Boc ("
1d Ns ("
E" = 66 2)
E
" > 200)
c
E" > 200)
1e Fmoc ("E" > 200)
1f Cbz ("
E
" > 200)
0
50
100
150
200
250
300
t (h)
Figure 3. Lipase PS-C II-catalysed (75 mg mLꢀ1) interesterifications of 2-
piperidylacetic acid methyl ester derivatives 1b–f (0.1 M) in PrCO2Bu–
DIPE (1:1).
3. Conclusion
Research aimed at the chemoenzymatic preparation of the
enantiomers of piperidyl-based b-amino acid derivatives
through the lipase-catalysed kinetic resolution of rac-1a–f
and rac-2a–c has been reported. The N-acylation of rac-
1a and rac-2a (A), and the aminolysis (B), alcoholysis (C)
and interesterification (D) at the ester function of rac-1b
(also 1c for aminolysis), as model reactions, revealed that
interesterification in the presence of lipase PS-C II from
B. cepacia is optimal in a 1:1 mixture of DIPE and butyl
butanoate yielding an ‘E’ = 66. Under these conditions,
interesterification of rac-1b–f was highly enantioselective
(‘E’ >200). It was possible to perform kinetic resolutions
with all three substrates, rac-1d (N-nosyl protection) being
an exception due to very low reactivity. Due to the highest
reactivity, a gram-scale resolution was performed using
N-Boc-protected rac-1c as the substrate, which conse-
quently led to the unreacted (R)-1c (ee = 94%) and the
interesterification product (S)-9c (ee = 99%) at 49% con-
version. A method for the kinetic resolution of the nicotinic
acid derived 3-piperidylacetic acid derivatives 2 was not
found in this study.
(nosyl), 1d], 9-fluorenylmethoxycarbonyl (Fmoc, 1e) and
benzyloxycarbonyl (Cbz, 1f) groups enhanced the enantio-
selectivity substantially (E increased from 66 for rac-1b to
>200 for rac-1c–f). Enzymatic reactivities of 1b–f differed
considerably from each other. With N-nosyl protection
(j), the reaction was too slow to be of any practical
importance. rac-1c gave a relatively fast reaction to yield
a mixture of (R)-1c and (S)-9c (.).
Finally, rac-1c (0.1 M; Fig. 3, .), as the most reactive of
the present substrates, was subjected to gram-scale resolu-
tion by lipase PS-C II-catalysed (75 mg mLꢀ1) interesterifi-
cation in DIPE–PrCO2Bu (1:1) in the presence of BuOH
(0.2 M) at 47 ꢁC. The reaction was stopped at 49% conver-
sion, giving (R)-1c (ee = 94%) and (S)-9c (ee = 99%) with
high chemical yields as shown in Section 4. BuOH was
added to make sure that the methanolysis of (S)-9c was
suppressed. The addition of BuOH (0.2 M) increased the
CO2Bu
MeO2C
CO2Me
Enzyme
+
BuOH or PrCO2Bu
solvent
N
R
N
R
N
R
(R)- or (S)-11a-c
(R)- or (S)-2a-c
2a (R=H)
2b (R=Ac)
2c (R=Boc)
Scheme 5. Enzymatic resolution of 2a–c.
Table 5. Lipase-catalysed (75 mg mLꢀ1) reactions of rac-2a–c, (0.1 M) at 23 ꢁC
Entry
Compound
Enzyme
Solvent
Time (h)
ees (%)
eep (%)
Conversion (%)
‘E’
a
1
2
3
4
5
6
2a
2b
2b
2b
2b
2c
CAL-B
CAL-B
CAL-B
PS-C II
PS-C II
PS-C II
BuOH (neat)
BuOH (neat)
PrCO2Bu–DIPE (1:1)
BuOH (neat)
PrCO2Bu–DIPE (1:1)
PrCO2Bu–DIPE (1:1)
24
24
24
24
24
90
32
89
6
<1
11
31
—
41
83
68
7
40
56
3
4
1
1
2
2
19
3
6
16
24
a Not determined.

A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
187
To explain why interesterification was highly enantioselec-
tive and relatively fast, we proposed a hypothesis whereby
BuOH is released from butyl butanoate through the lipase-
catalysed hydrolysis by water in the dry enzyme prepara-
tion, and where the released BuOH stayed at the active site
either replacing active site water or being part of the water
network. Accordingly, BuOH concentration is high at the
active site, although its total concentration in the reaction
mixture is low. This hypothesis also explains the well-
known phenomenon that water in the seemingly dry
enzyme preparations effectively hydrolyses various carb-
oxylic acid substrates in dry organic solvents.
The absolute configurations were determined by compar-
ing the specific rotation of 1c with a literature value.21
The absolute configuration of the reactive enantiomer
was supposed to stay the same with other protective
groups.
4.2. Preparation of racemic starting materials
4.2.1. 2-Piperidylacetic acid derivatives 1a–f.
4.2.1.1. 2-Piperidylacetic acid methyl ester hydrochloride
5. The synthesis is based on known methods.9 2-Pyridyl-
acetic acid 3 (9.01 g, 51.9 mmol) was dissolved in diluted
hydrochloric acid (0.71 M). Hydrogenation (5 bar) took
place for 50 h at room temperature in the presence of
PtO2 (200 mg, 0.88 mmol). Filtering of the catalyst and
evaporation of the solvent afforded the crude oily product
4 (9.32 g, 51.9 mmol).
4. Experimental
4.1. Materials and methods
In the second step 2-piperidylacetic acid hydrochloride 4
(2.00 g, 11.1 mmol) was dissolved in methanol (100 mL),
and thionyl chloride (1.3 mL, 18 mmol) was added drop-
wise in an ice bath. The reaction was mixed for half an
hour in an ice bath, one hour at room temperature and
3 h by refluxing. Evaporation of the solvent yielded 5 as
a white solid (2.15 g, 11.1 mmol). 1H NMR (500 MHz,
D2O) d (ppm): 3.74 (s, 3H, CH3), 3.55 (m, 1H, H-2),
3.43 (dddd, JH-6ax = ꢀ12.8, JH-5ax = 4.1, JH-5eq = 2.2 and
JH-4eq = 2.0 Hz, 1H, H-6eq), 3.04 (ddd, JH-6eq = ꢀ12.8,
JH-5ax = 12.7 and JH-5eq = 2.5 Hz, 1H, H-6ax), 2.79 (dd,
1H, CH2CO2), 2.78 (dd, 1H, CH2CO2), 1.96 (m, 1H,
H-3eq), 1.89 (m, 1H, H-5eq), 1.87 (m, 1H, H-4eq), 1.64
(m, 1H, H-5ax), 1.56 (m, 1H, H-4ax), 1.54 (m, 1H,
H-3ax). 13C NMR (500 MHz, D2O) d (ppm): 175.36
(CO2), 56.20 (C-2), 55.60 (CH3), 48.00 (C-6), 40.14
(CH2CO2), 31.06 (C-3), 24.62 (C-5), 24.32 (C-4).
Reagents were purchased from Aldrich, Fluka and Acros.
Solvents with the highest analytical grade were obtained
from Aldrich, J. T. Baker and Lab-Scan. Amano provided
lipases AH, PS, PS-C II and PS-D I from B. cepacia, lipases
AK and AK-C I from P. fluorescens and lipase R from
Penicillium roquefort. Lipases from Thermomyces lanuginos
(Lipozyme TL IM), Rhizomucor miehei (Lipozyme RM
IM) and C. antarctica A and B (CAL-A = Novozymeꢂ
735, CAL-B = Novozymeꢂ 435) were the products of
Novo Nordisk. Lipases from porcine pancreas (PPL) and
Candida rugosa (CRL, type VII) were obtained from
Sigma. Esterase from porcine liver (Chirazymeꢂ E-1,
PLE) was the product of Roche.
MS-spectra were measured with VG 7070E (EI) or
ZapSpec-oa Tof (EI, FAB) and optical rotations with
1
PerkinElmer polarimeter. The solution state H and 13C
nuclear magnetic resonance (NMR) spectra were
recorded using
a
Bruker Avance 500 spectrometer
4.2.1.2. 2-Piperidylacetic acid methyl ester 1a. 2-Piper-
idylacetic acid hydrochloride 4 (608 mg, 3.14 mmol) was
dissolved in chloroform. Gaseous ammonia was bubbled
into the solution. Precipitated ammonium chloride was fil-
tered off. Evaporation of the solvent in vacuum gave 1a
equipped with BBI-5 mm-Zgrad-ATM probe operating
at 500.13 and 125.77 MHz, respectively. The spectra were
recorded either at +25 ꢁC or at ꢀ50 ꢁC with a non-spin-
ning sample in 5 mm NMR tubes using CDCl3 as a sol-
1
1
vent, except for 5 D2O was used. The H spectra of the
(360 mg, 2.29 mmol, 73%). H NMR (500 MHz, CDCl3)
CDCl3 samples were referenced internally to the residual
protonated solvent signal (CHCl3, 7.27 ppm) and 13C
spectra to the solvent signal (CDCl3, 77.10 ppm). TSP
was used as the reference compound in the D2O samples
d (ppm): 3.66 (s, 3H, CH3), 3.02 (dddd, JH-6ax = ꢀ11.9,
JH-5ax = 4.0, JH-5eq = 2.6 and JH-4eq = 1.7 Hz, 1H, H-
6eq), 2.89 (dddd, JH-3ax = 11.0, JH-7b = 8.6, JH-7a = 4.4
and JH-3eq = 2.6 Hz, 1H, H-2), 2.64 (ddd, JH-6eq = ꢀ11.9,
JH-5ax = 12.2 and JH-5eq = 2.7 Hz, 1H, H-6ax), 2.37 (dd,
JH-7b = ꢀ15.9 and JH-2 = 4.4 Hz, 1H, H-7a, CH2CO2),
2.35 (dd, JH-7a = ꢀ15.9 and JH-2 = 8.6 Hz, 1H, H-7b,
CH2CO2), 1.76 (dddddd, JH-4ax = ꢀ13.4, JH-5ax = 3.9,
JH-3ax = 3.9, JH-3eq = 3.1, JH-5eq = 2.8 and JH-6eq = 1.7 Hz,
1
1
(0.00 ppm for both H and 13C). In some cases, the H
spectra were analysed employing the PERCH iteration
19,20
software for the extraction of d and JH,H
.
The cor-
rect assignment of chemical shifts was confirmed by
application of two-dimensional correlation measure-
ments, including gradient selected DQF-COSY measure-
ments (cosygpmfqf pulse program), gradient selected
1H,13C HSQC measurements (hsqcetgpsi2 pulse program
1H, H-4eq), 1.60 (ddddd, J
= ꢀ12.8, JH-4ax = 3.7,
H-3ax
JH-4eq = 3.1, JH-2 = 2.6 and JH-5eq = 1.5 Hz, 1H, H-3eq),
1.57 (dddddd, JH-5ax = ꢀ13.1, JH-4ax = 4.0, JH-4eq = 2.8,
JH-6ax = 2.7, JH-6eq = 2.6 and JH-3eq = 1.5 Hz, 1H, H-
1
optimised for 145 Hz JCH couplings), gradient selected
NOESY measurements (noesygpph pulse program with
5eq), 1.39 (ddddd, JH-5eq = ꢀ13.1, JH-4ax = 12.9, JH-6ax
=
1
a 0.3 s mixing time) and gradient selected H,13C HMBC
12.2, JH-6eq = 4.0 and JH-4eq = 3.9 Hz, 1H, H-5ax),
1.35 (ddddd, JH-4eq = ꢀ13.4, JH-5ax = 12.9, JH-3ax = 12.8,
JH-5eq = 4.0 and JH-3eq = 3.7 Hz, 1H, H-4ax), 1.15 (dddd,
measurements (hmbcgplpndqf pulse program optimised
n
for 10 Hz JCH couplings). The chemical shifts and cou-
pling constants are reported at +25 ꢁC, unless otherwise
mentioned.
JH-3eq = ꢀ12.8, JH-4ax = 12.8, JH-2 = 11.0 and JH-4eq
=
3.9 Hz, 1H, H-3ax). 13C NMR (500 MHz, CDCl3) d

188
A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
(ppm): 172.85 (CO2), 53.37 (C-2), 51.58 (CH3), 46.86 (C-6),
41.46 (CH2CO2), 32.62 (C-3), 26.01 (C-5), 24.61 (C-4).
HRMS: M+ = 257.1626, C13H23NO4 requires M+ =
257.1627.
4.2.1.3. N-Acetyl-2-piperidylacetic acid methyl ester
1b. 2-Piperidylacetic acid methyl ester hydrochloride 5
(2.00 g, 10.3 mmol) was dissolved in chloroform (50 mL).
The addition of triethylamine (TEA, 7.4 mL, 52 mmol),
DMAP (4-N,N-dimethylaminopyridine, 64 mg, 0.52 mmol)
and acetic anhydride (2.1 mL, 22 mmol) started the reac-
tion. After 24 h, the reaction was stopped by adding meth-
anol (20 mL). Evaporation of the solvent and purification
by column chromatography with silica gel (acetone–petro-
leum ether 3:7–1:1) yielded oily product 1b (1.64 g,
8.2 mmol, 80%). 1H NMR (500 MHz, CDCl3) d (ppm):
Major isomer (53%): 4.54 (br d, JH-6ax = ꢀ13.0 Hz, 1H,
H-6eq), 4.44 (br m, 1H, H-2), 3.66 (s, 3H, CO2CH3), 2.77
(dd, JH-7b = ꢀ15.3 and JH-2 = 8.0 Hz, 1H, H-7a, CH2CO2),
2.61 (dd, 1H, H-7b, CH2CO2), 2.58 (m, 1H, H-6ax), 2.14 (s,
3H, COCH3), 1.71–1.48 (m, 5H, H-3ax, H-3eq, H-4ax, H-
4eq and H-5eq), 1.36 (m, 1H, H-5ax). Minor isomer (47%):
5.17 (br m, 1H, H-2), 3.63 (s, 3H, CO2CH3), 3.61 (br m,
1H, H-6eq), 3.12 (ddd, JH6-eq = ꢀ13.4, JH-5ax = 13.4 and
JH-5eq = 2.8 Hz, 1H, H-6ax), 2.59 (dd, 1H, H-7a,
CH2CO2), 2.52 (dd, JH-7a = ꢀ14.0 and JH-2 = 7.8 Hz, 1H,
H-7b, CH2CO2), 2.05 (s, 3H, COCH3), 1.71–1.48 (m, 5H,
H-3ax, H-3eq, H-4ax, H-4eq and H-5eq), 1.41 (m, 1H,
H-5ax). 13C NMR (500 MHz, CDCl3) d (ppm): Major iso-
mer (53%): 171.52 (CO2), 169.48 (NCO), 51.94 (CO2CH3),
50.73 (C-2), 36.57 (C-6), 35.51 (CH2CO2), 29.31 (C-3),
25.24 (C-5), 21.38 (COCH3), 19.04 (C-4). Minor isomer
(47%): 171.65 (CO2), 169.48 (NCO), 51.78 (CO2CH3),
45.26 (C-2), 42.07 (C-6), 34.74 (CH2CO2), 27.96 (C-3),
25.82 (C-5), 21.86 (COCH3), 18.83 (C-4). HRMS:
M+ = 199.1208, C10H17NO3 requires M+ = 199.1212.
4.2.1.5. N-Ns-2-piperidylacetic acid methyl ester
1d. The reaction was carried out as described for 1b, ex-
cept that acetic anhydride was replaced by nosyl chloride
(1.5 equiv) and chloroform with dichloromethane. DMAP
was not used as a catalyst. The reaction yielded 1d (795 mg,
2.3 mmol, 90%). 1H NMR (500 MHz, CDCl3) d (ppm):
0
0
0
8.10 (m, JH-3 ¼ 7:9, JH-2 ¼ 1:5 and JH-1 ¼ 0:3 Hz, 1H,
H-40, CHCNO2), 7.70 (m, JH-1 ¼ 8:0, JH-3 ¼ 7:8 and
0
0
JH-4 ¼ 1:5 Hz, 1H, H-20, SCCHCH), 7.69 (m, JH-4 ¼ 7:9,
0
0
JH-2 ¼ 7:8 and JH-1 ¼ 1:2 Hz, 1H, H-30, CHCHCNO2),
0
0
0
0
0
7.65 (m, JH-2 ¼ 8:0, JH-3 ¼ 1:2 and JH-4 ¼ 0:3 Hz, 1H,
H-10, SCCH), 4.48 (m, 1H, H-2), 3.80 (br d, JH-6ax
=
ꢀ14.0 Hz, 1H, H-6eq), 3.60 (s, 3H, CH3), 3.06 (ddd,
JH-6eq = ꢀ14.0, JH-5ax = 13.0 and JH-5eq = 2.6 Hz, 1H, H-
6ax), 2.73 (dd, JH-7b = ꢀ15.2 and JH-2 = 9.1 Hz, 1H, H-
7a, CH2CO2), 2.63 (dd, JH-7a = ꢀ15.2 and JH-2 = 5.7 Hz,
1H, H-7b, CH2CO2), 1.73 (m, 1H, H-3ax), 1.69–1.61 (br
m, 3H, H-3eq, H-4eq and H-5eq), 1.55 (m, 1H, H-4ax),
1.50 (m, 1H, H-5ax). 13C NMR (500 MHz, CDCl3) d
(ppm): 171.14 (CO2), 147.72 (CNO2), 133.86 (CSO2),
133.49 (CHCHCS), 131.89 (CHCHCNO2), 131.24
(CHCNO2), 124.35 (CHCS), 51.88 (CH3), 50.39 (C-2),
41.60 (C-6), 35.11 (CH2CO2), 28.10 (C-3), 25.18 (C-5),
18.28 (C-4). HRMS: MH+ = 343.0956, C14H18N2O6S
requires MH+ = 343.0964.
4.2.1.6. N-Fmoc-2-piperidylacetic acid methyl ester
1e. The synthetic procedures followed a known method.22
2-Piperidylacetic acid methyl ester hydrochloride 5 (0.10 g,
0.52 mmol) was dissolved in dry tetrahydrofuran (THF,
2 mL) followed by the addition of sodium hydride
(62 mg, 1.6 mmol, 60%) in dry THF. The reaction was
mixed for 3.5 h at room temperature, after which FmocCl
(207 mg, 0.78 mmol) in THF (3 mL) was added. After tak-
ing place overnight, the reaction was stopped with metha-
nol (5 mL). Evaporation yielded the crude product, which
was dissolved in water (10 mL) and extracted in DIPE
(1 · 10 mL). The organic phase was washed with water
(1 · 10 mL) and dried with MgSO4. Purification by column
chromatography with silica gel (acetone–petroleum ether
1:9) afforded 1e (102 mg, 0.27 mmol, 52%). 1H NMR
(500 MHz, CDCl3, ꢀ50 ꢁC) d (ppm): Major isomer
(56%): 7.80 (br, 2H, ArH), 7.63 (br, 2H, ArH), 7.44 (br,
2H, ArH), 7.36 (br, 2H, ArH), 4.79 (br, 1H, H-2), 4.46
(br m, 2H, NCO2CH2), 4.27 (br, 1H, NCO2CH2CH),
4.10 (br, 1H, H-6eq), 3.70 (s, 3H, CH3), 2.80 (br m, 1H,
H-6ax), 2.72 (br m, 1H, CH2CO2), 2.46 (br m, 1H,
CH2CO2), 1.75–1.35 (m, 6H, H-3ax, H-3eq, H-4ax, H-
4eq, H-5eq and H-5ax). Minor isomer (44%): 7.80 (br,
2H, H-50, ArH), 7.63 (br, 2H, H-20, ArH), 7.44 (br, 2H,
H-40, ArH), 7.36 (br, 2H, H-30, ArH), 4.85 (br, 1H, H-2),
4.36 (br m, 2H, NCO2CH2), 4.27 (br, 1H, NCO2CH2CH),
4.10 (br, 1H, H-6eq), 3.62 (s, 3H, CH3), 3.01 (br m, 1H, H-
6ax), 2.72 (br m, 1H, CH2CO2), 2.57 (br m, 1H, CH2CO2),
1.75–1.35 (m, 6H, H-3ax, H-3eq, H-4ax, H-4eq, H-5eq and
H-5ax). 13C NMR (500 MHz, CDCl3, ꢀ50 ꢁC) d (ppm):
Major isomer (56%): 171.78 (CO2CH3), 155.16
(NCO2CH2), 143.40 (2C, C-1a0, ArC), 141.05 (2C, C-5a0,
ArC), 127.62 (2C, C-40, ArCH), 126.96 (2C, C-30, ArCH),
4.2.1.4. N-Boc-2-piperidylacetic acid methyl ester
1c. The reaction was carried out as described for 1b,
except that acetic anhydride was replaced by di-tert-butyl-
dicarbonate. The reaction yielded 1c (1.2 g, 4.7 mmol,
48%). 1H NMR (500 MHz, CDCl3, ꢀ50 ꢁC) d (ppm):
Major isomer (57%): 4.62 (br m, 1H, H-2), 4.00 (br d,
JH-6ax = ꢀ12.9 Hz, 1H, H-6eq), 3.63 (s, 3H, CO2CH3),
2.70 (br dd, JH-6eq = ꢀ12.9 and JH-5ax = 12.6 Hz, 1H,
H-6ax), 2.58 (dd, JH-7b = ꢀ13.7 and JH-2 = 7.5 Hz, 1H,
H-7a, CH2CO2), 2.47 (dd, JH-7a = ꢀ13.7 and JH-2
=
7.5 Hz, 1H, H-7b, CH2CO2), 1.7–1.2 ppm (m, 6H, H-3ax,
H-3eq, H-4ax, H-4eq, H-5eq and H-5ax), 1.40 (s, 9H,
C(CH3)3). Minor isomer (43%): 4.71 (br m, 1H, H-2),
3.88 (br d, JH-6ax = ꢀ12.8 Hz, 1H, H-6eq), 3.60 (s, 3H,
CO2CH3), 2.78 (br dd, JH-6eq = ꢀ12.9 and JH-5ax
=
12.9 Hz, 1H, H-6ax), 2.62 (dd, JH-7b = ꢀ13.3 and
JH-2 = 7.5 Hz, 1H, H-7a, CH2CO2), 2.45 (dd, JH-7a
=
ꢀ13.3 and JH-2 = 7.5 Hz, 1H, H-7b, CH2CO2), 1.7–
1.2 ppm (m, 6H, H-3ax, H-3eq, H-4ax, H-4eq, H-5eq and
H-5ax), 1.39 (s, 9H, C(CH3)3). 13C NMR (500 MHz,
CDCl3, ꢀ50 ꢁC) d (ppm): Major isomer (57%): 172.21
(CO2CH3), 154.60 (NCO), 79.63 (CCH3), 52.07 (CO2CH3),
47.85 (C-2), 38.14 (C-6), 34.74 (CH2CO2), 28.14 (C(CH3)3),
28.11 (C-3), 24.88 (C-5), 18.52 (C-4). Minor isomer (43%):
172.39 (CO2CH3), 154.60 (NCO), 79.49 (CCH3), 52.18
(CO2CH3), 46.89 (C-2), 39.33 (C-6), 34.65 (CH2CO2),
28.14 (C(CH3)3), 28.11 (C-3), 25.08 (C-5), 18.58 (C-4).

A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
189
125.08 (2C, C-20, ArCH), 120.02 (2C, C-50, ArCH), 67.31
(NCO2CH2), 52.24 (CH3), 47.65 (C-2), 46.72 (C-10,
NCO2CH2CH), 39.32 (C-6), 34.54 (CH2CO2), 28.00 (C-
3), 24.91 (C-5), 18.38 (C-4). Minor isomer (44%): 172.17
(CO2CH3), 155.11 (NCO2CH2), 143.40 (2C, C-1a0, ArC),
141.05 (2C, C-5a0, ArC), 127.62 (2C, C-40, ArCH), 126.96
(2C, C-30, ArCH), 125.08 (2C, C-20, ArCH), 120.02 (2C,
C-50, ArCH), 67.20 (NCO2CH2), 52.24 (CH3), 47.65 (C-
2), 46.66 (C-10, NCO2CH2CH), 39.39 (C-6), 34.58
(CH2CO2), 27.58 (C-3), 25.16 (C-5), 18.58 (C-4). HRMS:
M+ = 379.1786, C23H25NO4 requires M+ = 379.1784.
1.96 (dddd, JH-4ax = ꢀ13.1, JH-5eq = 5.5, JH-3 = 4.3 and
JH-5ax = 3.4 Hz, 1H, H-4eq), 1.65 (ddddd, JH-5ax = ꢀ13.6,
JH-4eq = 5.5, JH-6eq = 4.8, JH-4ax = 3.9 and JH-6ax = 3.2 Hz,
1H, H-5eq), 1.65 (dddd, J
= ꢀ13.1, JH-5ax = 11.1,
H-4eq
JH-3 = 10.0 and JH-5eq = 3.9 Hz, 1H, H-4ax), 1.44 (ddddd,
JH-5eq = ꢀ13.6, JH-4ax = 11.1, JH-6ax = 10.3, JH-4eq = 3.4
and JH-6eq = 3.3 Hz, 1H, H-5ax). 13C NMR (500 MHz,
CDCl3) d (ppm): 174.85 (CO2), 51.59 (CH3), 48.58 (C-2),
46.41 (C-6), 42.41 (C-3), 27.38 (C-4), 25.51 (C-5).
4.2.2.2. N-Acetyl-3-piperidinecarboxylic acid methyl ester
1
4.2.1.7. N-Cbz-2-piperidylacetic acid methyl ester
1f. The reaction was carried out as described for 1b,
except that acetic anhydride was replaced by benzyloxy-
carbonyl chloride. The reaction yielded 1f (630 mg,
2b. Yield 1.35 g, 7.29 mmol, 66%. H NMR (500 MHz,
CDCl3)
d (ppm): Major isomer (53%): 3.98 (ddd,
JH-6ax = ꢀ13.2, JH-5eq = 4.8 and JH-5ax = 4.6 Hz, 1H, H-
6eq), 3.74 (br dd, JH-2ax = ꢀ13.5 and JH-3 = 3.5 Hz, 1H,
H-2eq), 3.69 (s, 3H, CO2CH3), 3.43 (dd, JH-2eq = ꢀ13.5
and JH-3 = 9.0 Hz, 1H, H-2ax), 3.06 (m, JH-6eq = ꢀ13.2
and JH-5eq = 3.2 Hz, 1H, H-6ax), 2.49 (dddd, JH-2ax = 9.2,
JH-4ax = 9.2, JH-2eq = 4.1 and JH-4eq = 4.1 Hz, 1H, H-3),
2.11 (s, 3H, NCOCH3), 2.01 (m, 1H, H-4eq), 1.79 (m,
1H, H-4ax), 1.65 (m, 1H, H-5eq), 1.47 (m, 1H, H-5ax).
Minor isomer (47%): 4.60 (m, JH-2ax = ꢀ13.2 Hz, 1H, H-
2eq), 3.70 (m, 1H, H-6eq), 3.66 (s, 3H, CO2CH3), 3.05
1
2.2 mmol, 84%). H NMR (500 MHz, CDCl3, ꢀ50 ꢁC) d
(ppm): Major isomer (55%): 7.41–7.31 (m, 5H, ArH),
5.11 (s, 2H, CO2CH2Ph), 4.79 (m, 1H, H-2), 4.11 (br d,
1H, H-6eq), 3.55 (s, 3H, CH3), 2.84 (br dd, JH-6eq
=
ꢀ13.4 and JH-5ax = 13.4 Hz, 1H, H-6ax), 2.67 (dd,
JH-7b = ꢀ14.3 and JH-2 = 7.9 Hz, 1H, H-7a, CH2CO2),
2.55 (dd, JH-7a = ꢀ14.3 and JH-2 = 7.4 Hz, 1H, H-7b,
CH2CO2), 1.72–1.55 (m, 4H, H-3ax, H-3eq, H-4eq and
H-5eq), 1.50 (m, 1H, H-4ax), 1.38 (m, 1H, H-5ax). Minor
isomer (45%): 7.41–7.31 (m, 5H, ArH), 5.09 (s, 2H,
CO2CH2Ph), 4.82 (m, 1H, H-2), 4.02 (br d, 1H, H-6eq),
3.60 (s, 3H, CH3), 2.87 (br dd, JH-6eq = ꢀ13.4 and
JH-5ax = 13.4 Hz, 1H, H-6ax), 2.66 (dd, JH-7b = ꢀ14.1
and JH-2 = 7.8 Hz, 1H, H-7a, CH2CO2), 2.56 (dd,
JH-7a = ꢀ14.1 and JH-2 = 7.4 Hz, 1H, H-7b, CH2CO2),
1.72–1.55 (m, 4H, H-3ax, H-3eq, H-4eq and H-5eq), 1.50
(m, 1H, H-4ax), 1.38 (m, 1H, H-5ax). 13C NMR
(500 MHz, CDCl3, ꢀ50 ꢁC) d (ppm): Major isomer
(55%): 171.95 (CO2CH3), 155.06 (NCO2CH2), 136.23
(C-i), 128.37 (2C, C-m), 127.87 (C-p), 127.55 (2C, C-o),
66.90 (CO2CH2Ph), 51.97 (CH3), 47.79 (C-2), 38.98 (C-
6), 34.79 (CH2CO2), 28.14 (C-3), 24.85 (C-5), 18.41
(C-4). Minor isomer (45%): 172.04 (CO2CH3), 155.04
(NCO2CH2), 136.21 (C-i), 128.37 (2C, C-m), 127.87 (C-
p), 127.55 (2C, C-o), 66.90 (CO2CH2Ph), 52.05 (CH3),
47.49 (C-2), 39.26 (C-6), 34.49 (CH2CO2), 27.82 (C-3),
25.03 (C-5), 18.46 (C-4). HRMS: M+ = 291.1469,
C16H21NO4 requires M+ = 291.1471.
(m, 1H, H-6ax), 2.83 (dd, JH-2eq = ꢀ13.2 and JH-3
=
=
10.7 Hz, 1H, H-2ax), 2.44 (dddd, JH-2ax = 10.7, JH-4ax
10.9, JH-2eq = 4.0 and JH-4eq = 4.0 Hz, 1H, H-3), 2.07
(s, 3H, NCOCH3), 2.07 (m, 1H, H-4eq), 1.80 (m, 1H,
H-5eq), 1.67 (m, 1H, H-4ax), 1.47 (m, 1H, H-5ax). 13C
NMR (500 MHz, CDCl3) d (ppm): Major isomer (53%):
173.67 (CO2CH3), 169.19 (NCOCH3), 51.98 (CO2CH3),
48.07 (C-2), 41.78 (C-6), 41.59 (C-3), 27.12 (C-4), 23.77
(C-5), 21.45 (NCOCH3). Minor isomer (47%): 173.31
(CO2CH3), 169.06 (NCOCH3), 51.81 (CO2CH3), 46.74
(C-6), 43.43 (C-2), 41.06 (C-3), 27.26 (C-4), 24.99 (C-5),
21.50 (NCOCH3). HRMS: M+ = 185.1051, C9H15NO3
requires M+ = 185.1052.
4.2.2.3. N-Boc-3-piperidinecarboxylic acid methyl ester
1
2c. Yield 1.54 g, 6.33 mmol, 77%. H NMR (500 MHz,
CDCl3, –50 ꢁC) d (ppm): Major isomer (71%): 4.09 (br,
1H, H-2eq), 3.96 (br, 1H, H-6eq), 3.67 (s, 3H, CO2CH3)
2.93 (br, 1H, H-2ax), 2.70 (br dd, JH-6eq = ꢀ11.3 and
JH-5ax = 11.3 Hz, 1H, H-6ax), 2.43 (br, 1H, H-3), 2.04 (br
d, JH-4ax = ꢀ11.5 Hz, 1H, H-4eq), 1.69 (br d, 1H, H-5eq),
1.52 (br, 1H, H-4ax), 1.43 (br, 1H, H-5ax), 1.41 (s, 9H,
C(CH3)3). Minor isomer (29%): 4.27 (br, 1H, H-2eq),
3.96 (br, 1H, H-6eq), 3.67 (s, 3H, CO2CH3) 2.77 (br, 1H,
H-2ax), 2.70 (br dd, JH-6eq = ꢀ11.3 and JH-5ax = 11.3 Hz,
4.2.2. 3-Piperidinecarboxylic acid derivatives. N-Protected
3-piperidinecarboxylic acid methyl esters were prepared by
starting from nicotinic acid 6 followed by hydrogenation of
the aromatic ring, esterification and N-protection (Scheme
2). The reactions were performed as described for the 2-
piperidylacetic acid derivatives 1a–c.
1H, H-6ax), 2.43 (br, 1H, H-3), 2.04 (br d, JH-4ax
=
ꢀ11.5 Hz, 1H, H-4eq), 1.69 (br d, 1H, H-5eq), 1.52
(br, 1H, H-4ax), 1.43 (br, 1H, H-5ax), 1.41 (s, 9H,
C(CH3)3). 13C NMR (500 MHz, CDCl3, ꢀ50 ꢁC) d
(ppm): Major isomer (71%): 174.33 (CO2CH3), 154.52
(NCO2), 79.79 (NCO2C(CH3)3), 52.15 (CO2CH3), 45.32
(C-2), 42.80 (C-6), 40.97 (C-3), 28.21 (C(CH3)3), 27.21
(C-4), 23.92 (C-5). Minor isomer (29%): 174.33 (CO2CH3),
154.57 (NCO2), 79.79 (NCO2C(CH3)3), 52.15 (CO2CH3),
44.52 (C-2), 44.03 (C-6), 40.97 (C-3), 28.21 (C(CH3)3),
27.21 (C-4), 24.24 (C-5). HRMS: M+ = 243.1468;
C12H21NO4 requires M+ = 243.1471.
4.2.2.1. 3-Piperidylacetic acid methyl ester 2a. Yield
530 mg, 3.70 mmol, 96%.
1H NMR (500 MHz, CDCl3) d (ppm): 3.66 (s, 3H, CH3),
3.14 (dd, JH-2ax = ꢀ12.4 and JH-3 = 3.5 Hz, 1H, H-2eq),
2.91 (ddd, JH-6ax = ꢀ12.3, JH-5eq = 4.8 and JH-5ax = 3.3 Hz,
1H, H-6eq), 2.80 (dd, J
= ꢀ12.4 and JH-3 = 9.3 Hz, 1H,
H-2eq
H-2ax), 2.62 (ddd, JH-6eq = ꢀ12.3, JH-5ax = 10.3 and
JH-5eq = 3.2 Hz, 1H, H-6ax), 2.44 (dddd, JH-4ax = 10.0,
JH-2ax = 9.3, JH-4eq = 4.3 and JH-2eq = 3.5 Hz, 1H, H-3),

190
A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
4.3. Enzymatic reactions
m, 2H, CH2CO2), 1.6–1.2 (m, 6H, H-3ax, H-3eq, H-4ax,
H-4eq, H-5eq and H-5ax), 1.48 (m, 2H, CO2CH2CH2),
1.34 (s, 9H, C(CH3)3), 1.23 (m, 2H, CO2CH2CH2CH2),
0.80 (m, 3H, CH2CH3). Minor isomer (40%): 4.66 (br m,
The conversion was mainly calculated according to the
equation c = eeS/(eeS + eeP) at conversions lower than
40%. Due to the presence of competitive nucleophiles,
MeOH and BuOH (and water), conversions were also deter-
mined from the peak areas of the substrate and the product
taking into account the different responses in the FID-
detector. This gave conversion values similar to those ob-
tained from ee-values. The determination of E is based on
the equation E = ln[(1 ꢀ c)(1 ꢀ eeS)]/ln[(1 ꢀ c)(1 + eeS)].23
Using linear regression E was achieved as a slope of a line.
E values may contain contributions from the competitive
nucleophiles and, for this reason, ‘E’ is used instead of E
throughout the work.
1H, H-2), 3.92 (br m, 2H, CO2CH2), 3.81 (br d, JH-6ax
=
ꢀ12.8 Hz, 1H, H-6eq), 2.71 (br dd, JH-6eq = ꢀ12.8
and JH-5ax = 13.1 Hz, 1H, H-6ax), 2.46 (br m, 2H,
CH2CO2), 1.6–1.2 (m, 6H, H-3ax, H-3eq, H-4ax, H-4eq,
H-5eq and H-5ax), 1.48 (m, 2H, CO2CH2CH2), 1.32 (s,
9H, C(CH3)3), 1.23 (m, 2H, CO2CH2CH2CH2), 0.80 (m,
3H, CH2CH3). 13C NMR (500 MHz, CDCl3, ꢀ50 ꢁC) d
(ppm): Major isomer (60%): 171.68 (CO2CH2), 154.47
(NCO), 79.43 (CCH3), 64.41 (CO2CH2), 47.78 (C-2),
38.06 (C-6), 34.84 (CH2CO2), 30.04 (CO2CH2CH2), 27.97
(C(CH3)3),
27.80
(C-3),
24.75
(C-5),
18.90
(CO2CH2CH2CH2), 18.37 (C-4), 13.73 (CH2CH3). Minor
isomer (40%): 171.68 (CO2CH2), 154.39 (NCO), 79.27
(CCH3), 64.41 (CO2CH2), 46.60 (C-2), 39.27 (C-6), 34.53
(CH2CO2), 30.04 (CO2CH2CH2), 28.01 (C(CH3)3), 27.65
(C-3), 25.02 (C-5), 18.90 (CO2CH2CH2CH2), 18.44 (C-4),
13.73 (CH2CH3). HRMS: M+ = 299.2095, C16H29NO4
requires M+ = 299.2097.
Volumes of the test reactions were 0.5–1.0 mL. In butano-
lyses, 1a or 1b (0.1 M) was dissolved in butanol or butanol/
DIPE mixture. Addition of the enzyme started the reac-
tion. The reactions were shaken at room temperature or
in an incubator (47 ꢁC). The enzyme was filtered off from
the samples (50 lL), which were analysed by GC equipped
with a chiral column (Varian permethylated b-cyclo-
dextrin). In interesterifications, 1a–f or 2b–c (0.1 M) were
dissolved in PrCO2Bu or PrCO2Bu–DIPE mixture (1 or
0.5 mL). Addition of the enzyme (75 mg mLꢀ1) started
the reaction. The reactions were shaken at room tempera-
ture or in an incubator (47 ꢁC). The samples (50 lL) were
taken at intervals and analysed as above or by HPLC
equipped with a chiralcel OD-column. Since the enantio-
mers of 2c were not separated by the chiral columns used,
the reaction products were analysed by separating sub-
strate 2c and product 11c with column chromatography
using acetone–petroleum ether 1:9 as solvent. Boc-protec-
tion was removed by HCl in methanol (0.83 M) and the
enantiomers of 2a produced were derivatised into 2b with
acetic anhydride. In aminolyses, 1b or 1c (0.1 M) was dis-
solved in TBME (1 or 0.5 mL). Addition of an amine
(iPrNH2 or BuNH2, 0.2 M) and the enzyme (75 or
80 mg mLꢀ1) started the reaction. The reactions were sha-
ken at room temperature. Samples (50 lL) were taken at
intervals and analysed as described above.
Acknowledgements
The authors thank Pa¨ivi Alanko for the work with 1a and
2a, Jukka Alilonttinen for measuring some of the NMR
spectra and TEKES for financial support.
References
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. Liljeblad, A.; Kiviniemi, A.; Kanerva, L. T. Tetrahedron:
Asymmetry 2004, 60, 671–677.
4. Li, X.-G.; Kanerva, L. T. Org. Lett. 2006, 8, 5593–5596.
5. Gedey, S.; Liljeblad, A.; Fulo¨p, F.; Kanerva, L. T. Tetra-
¨
hedron: Asymmetry 1999, 10, 2573–2581.
6. Liljeblad, A.; Kanerva, L. T. Tetrahedron 2006, 62, 5831–
5854.
7. Pousset, C.; Callens, R.; Haddad, M.; Larcheveˆque, M.
Tetrahedron: Asymmetry 2004, 15, 3407–3412.
8. Miyata, H.; Yamamoto, Y.; Konegawa, T.; Sakata, K. U.S.
Patent 0170473, 2005.
9. Toone, E. J.; Jones, J. B. Can. J. Chem. 1987, 65, 2722–
2726.
10. Kazlauskas, R. J. TIBTECH 1994, 12, 464–472.
11. Zaks, A.; Klibanov, A. M. Proc. Natl. Acad. Sci. U.S.A.
1985, 82, 3192–3196.
12. Alfonso, I.; Gotor, V. Chem. Soc. Rev. 2004, 33, 201–
209.
13. Jacobsen, E. E.; Hoff, B. H.; Moen, A. R.; Anthonsen, T. J.
Mol. Catal. B: Enzym. 2003, 21, 55–58.
14. Liljeblad, A.; Kanerva, L. T. Tetrahedron: Asymmetry 1999,
10, 4405–4415.
Gram-scale reaction was performed with N-Boc-2-piper-
idylacetic acid methyl ester 1c (750 mg, 2.92 mmol) by dis-
solving it into PrCO2Bu–DIPE -mixture (1:1, 29.2 mL) and
by adding butanol (690 lL, 0.2 M) and lipase PS-C II
(2.19 mg, 75 mg mLꢀ1). The reaction was shaken for
168 h at 47 ꢁC. The enzyme was filtered off and washed
with DIPE and acetone. (R)-1c and (S)-9c were separated
by column chromatography with silica gel by using ace-
tone–petroleum ether (1:99–5:95) as an eluent. Unreacted
20
(R)-1c [308 mg, 1.20 mmol, ees = 94%, ½aꢁD ¼ þ9:9 (c
4.54, CHCl3)] and the product (S)-9c [339 mg, 1.13 mmol,
20
eep = 99%, ½aꢁD ¼ ꢀ5:4 (c 1, CHCl3)] were separated. Lit-
20
erature value for (S)-1c has been reported ½aꢁD ¼ ꢀ8:3 (c
4.54, CHCl3).21
15. Li, X.-G.; Kanerva, L. T. Tetrahedron: Asymmetry 2005, 16,
1709–1714.
16. Sakai, T. Tetrahedron: Asymmetry 2004, 15, 2749–2756.
17. Martinelle, M.; Hult, K. Biochim. Biophys. Acta 1995, 1251,
191–197.
Compound (S)-9c: 1H NMR (500 MHz, CDCl3, ꢀ50 ꢁC) d
(ppm): Major isomer (60%): 4.55 (br m, 1H, H-2), 3.93 (br
d, 1H, H-6eq), 3.92 (br m, 2H, CO2CH2), 2.64 (br dd,
JH-6eq = ꢀ12.7 and JH-5ax = 12.7 Hz, 1H, H-6ax), 2.43 (br

A. Liljeblad et al. / Tetrahedron: Asymmetry 18 (2007) 181–191
191
´
18. Garcıa-Alles, L. F.; Gotor, V. Biotechnol. Bioeng. 1998, 59,
21. Morley, C.; Knight, D. W.; Share, A. C. J. Chem. Soc.,
Perkin Trans. 1 1994, 2903–2907.
163–170.
19. Laatikainen, R.; Niemitz, M.; Weber, U.; Sundelin, J.;
Hassinen, T.; Vepsa¨la¨inen, J. J. Magn. Reson. A 1996, 120,
1–10.
22. Helgen, C.; Bochet, C. G. J. Org. Chem. 2003, 68, 2483–
2486.
23. Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am.
Chem. Soc. 1982, 104, 7294–7299.
20. http://www.perchsolutions.com.