Synthesis of (S)-(+)-cericlamine through lipase-catalyzed aminolysis of azo acetates

Article abstract of DOI:10.1039/c2ob25247c

The kinetic enzymatic resolution of azo acetates via aminolysis with Candida antarctica lipase B has been investigated using benzylamine as amine component. The products obtained from this biotransformation in high enantiomeric purity can serve as valuable precursors for various amino alcohols, as exemplified by the synthesis of the serotonin reuptake inhibitor (S)-(+)-cericlamine.

Full text of DOI:10.1039/c2ob25247c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Organic &
Biomolecular
Chemistry
Cite this: Org. Biomol. Chem., 2012, 10, 3384
www.rsc.org/obc
COMMUNICATION
Synthesis of (S)-(+)-cericlamine through lipase-catalyzed aminolysis of azo
acetates†
Agnes Prechter,a Harald Gröger‡b and Markus R. Heinrich*a
Received 2nd February 2012, Accepted 8th March 2012
DOI: 10.1039/c2ob25247c
enriched form has been extremely limited due to the narrow sub-
strate scope, long reaction times and high enzyme loading
required for lipase-catalyzed kinetic resolutions.11 In this com-
munication, we now report recent advances leading to an
efficient access to enantiomerically enriched azo acetates through
lipase-catalyzed kinetic resolution.
Due to the low conversions observed for azo acetates in the
previously employed aqueous biphasic mixtures of methyl tert-
butyl ether (MTBE) and water,11 we first examined the hydroly-
sis of acetate 7a in a variety of non-dried organic solvents
including acetonitrile, MTBE, methanol, n-hexane, dimethylform-
amide (DMF), acetone and tetrahydrofuran (THF) (Table 1,
entries 1–8).
The only solvent leading to a basically useful conversion com-
bined with a moderate enantiomeric excess was found to be
acetonitrile (entry 1). In a second optimization step benzylamine
was added, which turns the overall reaction from a hydrolysis to
an aminolysis of the acetate.12 By this modification, the results
for THF, and especially acetonitrile (entries 12, 13), could be
further improved with regard to conversion and enantioselectiv-
ity. One reason for the positive effect of benzylamine might be
the decreased activity of CAL-B otherwise caused by liberated
acetic acid.13 Furthermore, advantages are likely to arise from
the ability of benzylamine to act as a better nucleophile than
water. Instead of acetic acid, N-benzylacetamide is now formed
in amounts equivalent to alcohol 8a.14 While fewer equivalents
of benzylamine (3 equiv) did not influence the overall outcome
(entry 14), larger quantities of benzylamine (>5 equiv) as well as
the addition of water led to remarkably lower conversions
The kinetic enzymatic resolution of azo acetates via aminoly-
sis with Candida antarctica lipase B has been investigated
using benzylamine as amine component. The products
obtained from this biotransformation in high enantiomeric
purity can serve as valuable precursors for various amino
alcohols, as exemplified by the synthesis of the serotonin
reuptake inhibitor (S)-(+)-cericlamine.
While lipase-catalyzed reactions of tertiary alcohols are usually
slow, enzymatic transformations remain fast when primary alco-
hols with more remote quaternary stereocenters are employed as
starting materials.1 Most primary alcohols that have so far been
studied belong to the group of diols 1,2 or derivatives in which
the unreactive tertiary alcohol is part of an ether,3 epoxide4 or
lactone.5 Remarkable results have also been achieved in desym-
metrizations of diols 2.6 Only very few studies have so far been
reported for enzymatic resolutions of primary alcohols 3 bearing
a nitrogen-substituted quaternary stereocenter in β-position. An
early example is the enantiomeric enrichment of racemic benza-
mide 4 through lipase AK-20 in the presence of vinyl acetate.7 A
comparable lipase-catalyzed transesterification has very recently
been used for the resolution of nitro compound 5 (Fig. 1).8
Our interest is focused on azo compounds 6, which are easily
accessible and can also be further converted to manifold pro-
ducts including amino alcohols, amino acids, indolines and iso-
quinolines.9,10 The broad substrate scope and substitution pattern
of 6 and O-acylated derivatives thereof makes an enantio-
selective approach to such molecules an attractive alternative to
currently developed desymmetrization strategies.6 So far, unfa-
vorably, the access to azo compounds 6 in enantiomerically
aDepartment für Chemie und Pharmazie, Pharmazeutische Chemie,
Friedrich-Alexander Universität Erlangen-Nürnberg, Schuhstraße 19,
91052 Erlangen, Germany. E-mail: Markus.Heinrich@medchem.uni-
erlangen.de; Fax: +49-9131-85-22585; Tel: +49-9131-85-24115
bDepartment für Chemie und Pharmazie, Organische Chemie,
Friedrich-Alexander-Universität Erlangen-Nürnberg, Henkestraße 42,
91054 Erlangen, Germany
†Electronic supplementary information (ESI) available: Experimental
procedures, characterization data and copies of 1H and 13C NMR spectra
of all new compounds. See DOI: 10.1039/c2ob25247c
‡Current address: Organische Chemie I, Fakultät für Chemie, Universi-
tät Bielefeld, 33501 Bielefeld, Germany. E-mail: harald.groeger@uni-
bielefeld.de.
Fig. 1 Substrates for kinetic enzymatic resolution.
3384 | Org. Biomol. Chem., 2012, 10, 3384–3387
This journal is © The Royal Society of Chemistry 2012

View Article Online
Table 1 Effects of non-polar solvents and benzylamine
Table 2 Variation of substrates
Entrya Solvent
Base (equiv) Conversionb % ee (8a)b Ec
Acetate (S)-7c,d
(%, ee)
Alcohol (R)-8c
(ee)
Entrya
R
Ee
1
2
3
4
5
6
CH3CN
CH3CNd
MTBE
MeOH
n-Hexane
DMF
—
—
—
—
—
—
—
—
32
<2
67
<1
21
<1
13
13
45
<1
16
28
45
46
35
21
<2
41
nde
4f
2.9
nde
1.1
nde
1.0
nde
2.8
2.2
1.6
nde
2.6
2.9
7.0
7.1
6.3
3.5
nde
10.2
3.3
6.1
1b
2
3
4
5
6
7
8
4-OMe
4-F
3,4-Cl2
4-Br
7a (84, 73% ee)f
7b (58, 93% ee)
7c (52, >99% ee)
7d (58, 96% ee)
7e (24, 84% ee)
7f (n.d.)h
8a (53% ee)
8b (40% ee)
8c (37% ee)
8d (35% ee)
8e (13% ee)
8f (18% ee)h
8g (32% ee)
8h (71% ee)h
8i (43% ee)
8j (29% ee)
6.8
7.1
9.6g
7.0g
2.7
1.5i
3.2
7.3i
6.0
4.8
nde
0
nde
45
35
18
nde
42
43
62
62
64
51
nde
76
46
68
H
7
8
Acetone
THF
2-OMe
3-OMe
3,4-OMe2
4-CN
4-NO2
7g (82, 56% ee)
9
MTBE
DMF
Acetone
THF
CH3CN
CH3CN
CH3CN
CH3CN
CH3CNg
CH3CN
CH3CN
CH3CN
BnNH2 (5)
BnNH2 (5)
BnNH2 (5)
BnNH2 (5)
BnNH2 (5)
BnNH2 (3)
BnNH2 (7)
BnNH2 (10)
BnNH2 (5)
7h (n.d.)h
10
11
12
13
14
15
16
17
18
19
20
9
7i (74, 83% ee)
7j (48, 90% ee)
10j
a Reaction conditions: rac-7 (0.30 mmol), CH3CN (12 mL), CAL-B
(120 mg), benzylamine (161 mg, 5 equiv), rt, 3 days. b This reaction was
performed on a 0.05 mmol scale for 6 days. c Enantiomeric excess
determined by chiral HPLC, yields are calculated based on the single
enantiomers. d Acetates were converted to their corresponding alcohols
for HPLC-measurement except in the case of 7j, which could be
measured directly. e E-value calculated from the ee data of the alcohol
and the acetate. f Enantiomeric excess calculated from conversion and
enantiomeric excess of the alcohol. g E-values are lower than those
reported in Table 3 due to continuation of the reaction beyond the
optimal conversion. h Enantiomeric excess of 7f and 7h not determined
due to low conversions (30% for 7f and 24% for 7h, each after 3 days).
i E-values calculated from the conversion and the ee data of the alcohol.
j This reaction was performed on a 0.06 mmol scale.
n-BuNH2 (5) 31
Et3N (5)
Et2NH (5)
32
19
a Reaction conditions: rac-7a (0.05 mmol), organic solvent (2 mL),
CAL-B (20 mg), rt, 3 days. b Conversion and enantiomeric excess of 8a
determined by chiral HPLC. c E-value calculated from the conversion
and the ee data of the alcohol 8a. d Water was added to the reaction
mixture: CH3CN–H2O 3 : 1(v/v). e nd
=
not determined. f (S)-
enantiomer. g 5% water were added to the reaction mixture.
(entries 15–17). Attempts with other bases instead of benzyl-
amine revealed that n-butylamine as an additive is able to give
even higher E-values, but at remarkably lower conversions
(entry 18). The reaction with triethylamine proceeded as if no
base had been added (entry 19, cf. entry 1), and diethylamine
led to an improved ee-value, but with insufficient conversion
(entry 20).
The newly found conditions (Table 1, entry 13) were then
applied to the fluorinated azo acetate 7b, a compound which had
shown the worst results in previous investigations.11 To get a
better insight in the conversion of each enantiomer, the reaction
course was followed by chiral HPLC over several days (see
ESI†).15,16 In this way, nearly total conversion of one enantiomer
was reached after 72 hours, leaving the remaining acetate (29%
of 50% theoretical yield) with necessarily high enantiomeric
excess.
Results from reactions with a broader spectrum of substrates
are summarized in Table 2. After 3 days, high ee-values were
observed for the remaining acetates 7 starting from racemic lipo-
philic halogenated azo acetates 7b–d (entries 2–4). While a
longer reaction time of 6 days was necessary to achieve an
acceptable enantiomeric excess for 7a (entry 1), the conversion
of unsubstituted azo acetate 7e proceeded rapidly, but with only
low selectivity (entry 5). In general, electron-donating substitu-
ents such as the methoxy group led to the lowest conversions
when situated in 2- or 4-position of the aromatic core (entries 1,
6, 8). A higher conversion, although at lower selectivity, was
reached with the 3-methoxy derivative 7g (entry 7). Electron-
withdrawing substituents like nitro or cyano were better tolerated
in the 4-position than the methoxy group, as shown by the
reactions of 7i and 7j (entries 9, 10). The fact that the desired
products (S)-7i and (S)-7j were still obtained with lower enantio-
meric excess than the halogenated derivatives demonstrated that
the latter group of substrates (entries 2–4) represents the most
suitable for this type of enzymatic resolution.17
The results obtained from the investigation of the substrate
scope (Table 2) raised the question whether high ee-values
(>95%) can especially be reached with halogenated azo acetates
at relatively low conversions. A survey within this group of com-
pounds is presented in Table 3. Interestingly, the conversion of
some azo acetates 7 required much shorter reaction times than
three days to yield the desired products with acceptably high
enantiomeric excess.
When gradually increasing the steric demand of the halogen
atom in the 4-position from fluorine to chlorine and bromine to
iodine, a constant decrease in reaction time was observed (entries
1–7). To reach roughly 90% ee with a nearly similar yield, the
4-fluoro derivative 7b required 48 hours, the 4-chloro derivative
7k 44 hours, the 4-bromo derivative 7d 33.5 hours and 4-iodo
derivative 7l only 24.5 hours (entries 1, 3, 5, 6).
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3384–3387 | 3385

View Article Online
Table 3 Variation of substrates and reaction times
Time Acetate (S)-7b,c
Alcohol (R)-8b
(ee)
Entrya
R
(h)
(%, ee)
Ed
1
2
3
4
5
6
7
8
4-F
4-F
4-Cl
4-Cl
4-Br
4-I
48
71
44
59
33.5
24.5
46.5
149
16
7b (70, 88% ee)e
7b (58, 96% ee)e
7k (70, 90% ee)
7k (64, 94% ee)
7d (72, 91% ee)
7l (76, 90% ee)
7l (66, >97% ee)
7m (66, 60% ee)
7n (60, 82% ee)
7n (22, 96% ee)
7o (66, 80% ee)f
7p (74, 89% ee)i
7c (80, 97% ee)
8b (48% ee)
8b (39% ee)
8k (50% ee)
8k (43% ee)
8d (51% ee)
8l (52% ee)
8l (48% ee)
8m (32% ee)
8n (37% ee)
8n (12% ee)
8o (nd)g
7.7
7.8
8.6
8.0
9.1
9.1
10.8
3.4
5.0
Scheme 1 Synthesis of (S)-(+)-cericlamine [(S)-10].
4-I
2-Br
3-Br
3-Br
2,4-Cl2 39
3,5-Cl2 12.5
3,4-Cl2 15
3,4-Cl2 19
(S)-2-amino-3-phenylpropanols. The results obtained by enzy-
matic kinetic resolution compare well to existing synthetic
strategies leading to structurally related amino alcohols.7,8
The tolerance of lipophilic substrates enabled an effective syn-
thesis of (S)-(+)-cericlamine, and E-values up to 18 were
observed in the enzymatic transformation leading to the required
(S)-(+)-cericlamine intermediate. Ongoing research is directed
towards dynamic kinetic resolutions22 as well as further process
optimization.
9
10
11
12
13
14
45
3.5
5.2h
8.6
8p (51% ee)i
8c (64% ee)
18.1
18.8
7c (74, >99% ee) 8c (59% ee)
a Reaction conditions: rac-7 (0.20 mmol), CH3CN (8 mL), CAL-B
(80 mg), benzylamine (107 mg, 5 equiv), rt. b Enantiomeric excess
determined by chiral HPLC, yields are calculated based on the single
enantiomers. c Unless otherwise stated, acetates were converted to their
corresponding alcohols for HPLC-measurement. d E-value calculated
from the ee data of the alcohol and the acetate. e Enantiomeric excess
calculated from conversion and enantiomeric excess of the alcohol.
f Determination of enantiomeric excess after conversion to phthalate
Acknowledgements
derivative. g nd
=
not determined. h E-value calculated from the
The authors would like to thank the Universität Bayern e.V. for a
“Bayerische Eliteförderung” fellowship awarded to A. Prechter.
We thank Michael Höfer for the experiments conducted during
his pharmaceutical chemistry internship.
conversion and the ee data of the alcohol. i Determination of
enantiomeric excess after conversion to amino alcohol.
Shifting the rather large bromo substituent into the 2- or 3-
position led to a sharp decrease (entry 8) or increase (entries 9,
10) in reaction rate, respectively. Both compounds 7m and 7n
were however converted with comparably lowered E-values. The
substituent effects observed for the three different dichloro
derivatives 7c, 7o and 7p are largely in agreement with the
observations made before. While a 2-substituent such as in 7o
(entry 11) is not well tolerated even at small size, 3-substitution
as in 7p (entry 12) allows for better results. As evidenced by the
experiments with 7c (entries 13, 14), two small and lipophilic
substituents in 3- and 4-position are likely to be one of the most
suitable patterns for the enzymatic reaction described herein.18
One possible application for azo acetates obtained from the
kinetic resolution is the synthesis of pharmaceutically relevant
amino alcohols. (S)-(+)-Cericlamine [(S)-10], an antidepress-
ant,19,20 is readily accessible from acetate (S)-7c through two
simple synthetic steps (Scheme 1).
Notes and references
1 (a) U. T. Bornscheuer and R. J. Kazlauskas, Hydrolases in Organic Syn-
thesis, Wiley-VCH, 2006, 2nd edn; (b) For a recent review discussing
lipase-catalyzed reactions see: A. Ghanem, Tetrahedron, 2007, 63, 1721.
2 (a) S. T. Chen and J. M. Fang, J. Org. Chem., 1997, 62, 4349; (b) R.
P. Hof and R. M. Kellog, J. Chem. Soc., Perkin Trans. 1, 1996, 2051;
(c) R. P. Hof and R. M. Kellog, Tetrahedron: Asymmetry, 1994, 5, 565;
(d) C. R. Johnson, Y. Xu, K. C. Nicolaou, Z. Yang, R. K. Guy, J.
G. Dong and N. Berova, Tetrahedron Lett., 1995, 36, 3291; (e) V.
P. Rocco, S. J. Danishefsky and G. K. Schulte, Tetrahedron Lett., 1991,
32, 6671–6674; (f) O. Jimenéz, M. P. Bosch and A. Guerrero, J. Org.
Chem., 1997, 62, 3496; (g) V. Khlebnikow, K. Mori, K. Terashima,
Y. Tanaka and M. Sato, Chem. Pharm. Bull., 1995, 43, 1659; (h) K.
E. Henegar, S. W. Ashford, T. A. Baughman, J. C. Sih and R. L. Gu, J.
Org. Chem., 1997, 62, 6588; (i) J. C. Sih, J. Am. Oil Chem. Soc., 1996,
73, 1377; ( j) R. G. Lovey, A. K. Saksena and V. M. Girijavallabhan, Tet-
rahedron Lett., 1994, 35, 6047.
3 (a) H. Estermann, K. Prasad, M. J. Shapiro, J. J. Bolsterli and M.
D. Walkinshaw, Tetrahedron Lett., 1990, 31, 445; (b) B. Wirz, R. Barner
and J. Huebscher, J. Org. Chem., 1993, 58, 3980; (c) R. P. Hof and R.
M. Kellog, J. Org. Chem., 1996, 61, 3423; (d) J. A. Hyatt and
C. Skelton, Tetrahedron: Asymmetry, 1997, 8, 523.
4 (a) P. Ferraboschi, D. Brembilla, P. Grisenti and E. Santaniello, J. Org.
Chem., 1991, 56, 5478; (b) P. Ferraboschi, S. Casati, P. Grisenti and
E. Santaniello, Tetrahedron: Asymmetry, 1993, 4, 9; (c) P. Ferraboschi,
S. Casati, P. Grisenti and E. Santaniello, Tetrahedron, 1994, 50, 3251;
(d) P. Ferraboschi, P. Grisenti, S. Casati and E. Santaniello, Synlett, 1994,
754; (e) P. Ferraboschi, P. Grisenti, A. Manzocchi and E. Santaniello,
Tetrahedron: Asymmetry, 1994, 5, 691; (f) Y. B. Seu and Y. H. Kho,
Tetrahedron Lett., 1992, 33, 7015; (g) T. Itoh, H. Ohara, Y. Takagi,
Reduction of azo acetate 7c to amino alcohol 9 by zinc in
hydrochloric acid was followed by reductive methylation.21 In
addition to the experiments reported in Table 3 (entries 13, 14),
the enzymatic resolution of 7c was carried out on a larger scale
to provide 0.84 g (76% yield based on single enantiomer, 97%
ee) of azo compound (S)-7c.
In summary, azo acetates have been shown to be valuable
starting materials for the enantioselective preparation of diverse
3386 | Org. Biomol. Chem., 2012, 10, 3384–3387
This journal is © The Royal Society of Chemistry 2012

View Article Online
N. Kanada and K. Uneyama, Tetrahedron Lett., 1993, 34, 4215; (h) See
ref. 2a.
5 T. Tauchi, H. Sakuma, T. Ohno, N. Mase, H. Yoda and K. Takabe, Tetra-
hedron: Asymmetry, 2006, 17, 2195.
6 (a) M. Ihara, M. Suzuki, K. Fukumoto and C. Kabuto, J. Am. Chem.
Soc., 1990, 112, 1164; (b) N. Watanabe, T. Sugai and H. Ohta, Chem.
Lett., 1992, 657; (c) S. Akai, T. Naka, Y. Takebe and Y. Kita, Tetrahedron
Lett., 1997, 38, 4243; (d) A. Fadel and P. Arzel, Tetrahedron: Asymmetry,
1997, 8, 283 and 371 (e) T. Ping, M.-H. Xu, Z.-Q. Wang, Z.-Y. Li and
G.-Q. Lin, Synlett, 2006, 1201; (f) S. Sano, M. Nakao, M. Takeyasu,
T. Honjo and Y. Nagao, Lett. Org. Chem., 2006, 3, 764.
(e) M. Weiß and H. Gröger, Synlett, 2009, 1251; (e) M. Weiß,
T. Brinkmann and H. Gröger, Green Chem., 2010, 12, 1580; (f) Review:.
V. Gotor, J. Biotechnol., 2002, 96, 35.
13 M. D. Romero, L. Calvo, C. Alba and A. Daneshfar, J. Biotechnol.,
2007, 127, 269.
14 For details, see ESI.†
15 Conversions determined by HPLC were verified by 1H-NMR spectra.
According to 1H-NMR, all reactions remained free of side-products.
16 Reaction conditions: rac-7b (0.20 mmol), CH3CN (8 mL), CAL-B
(80 mg), benzylamine (107 mg, 5 equiv), rt.
17 R/S-selectivity in agreement with previous studies, see ref. 11.
18 Experiments performed with two different enzyme batches led to nearly
identical results.
7 D. B. Berkowitz, J. A. Pumphrey and Q. Shen, Tetrahedron Lett., 1994,
35, 8743.
8 T. Nakano, Y. Yagi, M. Miyahara, A. Kamimura, M. Kawatsura and
T. Itoh, Molecules, 2011, 16, 6747.
19 (a) G. Aubard, J. Bure, A. P. Calvet, C. Gouret, A. G. Grouhel and J.-
L. Junien, Eur. Pat, 0237366, 1987; (b) J. G. Wettstein, C. J. Gouret and
J. L. Junien, Eur. J. Pharmacol., 1990, 183, 1477; (c) C. J. Gouret, J.
G. Wettstein, R. D. Porsolt, A. Puech and J.-L. Junien,
Eur. J. Pharmacol., 1990, 183, 1478; (d) C. J. Gouret, R. Porsolt, J.
G. Wettstein, A. Puech, X. Pascaud and J.-L. Junien, Arzneim.-Forsch.,
1990, 40, 633.
20 For other enantioselective synthetic approaches to cericlamine, see:
(a) B. Kaptein, H. M. Moody, Q. B. Broxterman and J. Kamphuis, J.
Chem. Soc., Perkin Trans. 1, 1994, 1495; (b) A. G. Steinig and D.
M. Spero, J. Org. Chem., 1999, 64, 2406.
9 For related review articles, see: (a) M. R. Heinrich, Chem.–Eur. J., 2009,
15, 820; (b) S. B. Höfling and M. R. Heinrich, Synthesis, 2011, 173.
10 (a) M. R. Heinrich, O. Blank and S. Wölfel, Org. Lett., 2006, 8, 3323;
(b) M. R. Heinrich, O. Blank and A. Wetzel, J. Org. Chem., 2007, 72,
476; (c) O. Blank, A. Wetzel, D. Ullrich and M. R. Heinrich, Eur. J. Org.
Chem., 2008, 3179.
11 F. R. Dietz, A. Prechter, H. Gröger and M. R. Heinrich, Tetrahedron Lett.,
2011, 52, 655.
12 For some enzymatic aminolysis reactions, see: (a) S. Puertas, R. Brieva,
F. Rebolledo and V. Gotor, Tetrahedron, 1993, 49, 4007; (b) A.
E. Sigmund, K. C. McNulty, D. Nguyen, C. E. Silverman, P. Ma, J.
A. Pesti and R. DiCosimo, Can. J. Chem., 2002, 80, 608; (c) M. López-
Garcia, I. Alfonso and V. Gotor, J. Org. Chem., 2003, 68, 648;
21 The configuration of the product was confirmed by comparison of the
optical rotation with reported values, see ref. 20a.
22 For possible light-induced racemization of quaternary azo compounds
see:N. A. Porter and L. J. Marnett, J. Am. Chem. Soc., 1973, 95, 4361.
This journal is © The Royal Society of Chemistry 2012
Org. Biomol. Chem., 2012, 10, 3384–3387 | 3387