Studies on N-activation for the lipase-catalyzed enantioselective preparation of β-amino esters from 4-phenylazetidin-2-one

Article abstract of DOI:10.1002/ejoc.201403467

The effect of N-substitution was examined for the enantio-selective lipase-catalyzed ring-opening reaction of racemic 4-phenylazetidin-2-one with methanol in dry organic solvents. Marked differences in the reactivity of various N-protected 4-phenylazetidin-2-ones were observed. Preparativescale reactions with Candida antarctica lipase B (Novozym 435 preparation) yielded N-acylated methyl (R)-3-amino-3- phenylpropanoates with enantiomeric excess (ee) values >99% in up to a 49% isolated yield, whereas Thermomyces lanuginosus lipase (Lipozyme TM IM) gave enantiomerically enriched methyl (S)-3-acetamido-3-phenylpropanoate. Candida antarctica lipase A catalyzed the cleavage of the N-chloroacetyl protective group, whereas all of the other examined lipases underwent the ring-opening reaction.

Full text of DOI:10.1002/ejoc.201403467

FULL PAPER
DOI: 10.1002/ejoc.201403467
Studies on N-Activation for the Lipase-Catalyzed Enantioselective Preparation
of β-Amino Esters from 4-Phenylazetidin-2-one
Riku Sundell[a] and Liisa T. Kanerva*[a]
Keywords: Biocatalysis / Enzymes / Enantioselectivity / Lactams / Kinetic resolution / Solvolysis
The effect of N-substitution was examined for the enantio-
selective lipase-catalyzed ring-opening reaction of racemic
4-phenylazetidin-2-one with methanol in dry organic sol-
vents. Marked differences in the reactivity of various N-pro-
tected 4-phenylazetidin-2-ones were observed. Preparative-
scale reactions with Candida antarctica lipase B (Novozym
435 preparation) yielded N-acylated methyl (R)-3-amino-3-
phenylpropanoates with enantiomeric excess (ee) values
Ͼ99% in up to a 49% isolated yield, whereas Thermomyces
lanuginosus lipase (Lipozyme TM IM) gave enantiomerically
enriched methyl (S)-3-acetamido-3-phenylpropanoate. Can-
dida antarctica lipase A catalyzed the cleavage of the N-
chloroacetyl protective group, whereas all of the other exam-
ined lipases underwent the ring-opening reaction.
β-amino acids and derivatives from the corresponding
nitriles.[4]
Introduction
β-Amino acids and their derivatives, including β-lactams
(azetidin-2-ones), are a group of pharmaceutically impor-
tant compounds and intermediates of synthetic products.[1]
Various chemo- and enzyme-catalyzed synthetic routes for
the preparation of these compounds in enantiomeric form
have been the target of intensive studies. For instance, the
transition-metal-catalyzed asymmetric hydrogenation of β-
dehydroamino acid derivatives and the Mannich reaction
with silyl enolates along with organocatalytic approaches
have been reported.[1a–1c,1e,1f] From a biocatalytic perspec-
tive, the lipase-catalyzed (EC 3.1.1.3) kinetic resolution of
racemic β-lactams by an enantioselective ring-opening reac-
tion represents a viable and extensively used approach for
the preparation of β-amino acid, β-amino ester, and β-di-
peptide enantiomers, leaving behind the less reactive β-
lactam enantiomer.[1d,1g,2] The formation of poly(β-alanine)
has even been described by the lipase-catalyzed ring-open-
ing of azetidin-2-one.[2d] The kinetic resolution of N-hy-
droxymethylated β-lactams by lipase-catalyzed O-acylation
followed by the removal of the N-methanol tail from the
resolved product has allowed the simultaneous preparation
of both β-lactam enantiomers.[1d,1g,3] The cascade reaction
that is catalyzed by nitrile hydratase and amidase enzymes
in Rhodococcus erythropolis AJ270 whole cells represents
more recent advances in the biocatalytic preparation of
Lipase enzymes, belonging to the serine hydrolase family,
hydrolyze lipids in nature. The ability of lipases to cleave an
amide bond is rare,[5] whereas serine proteases may undergo
a reaction at both the amide and ester bonds. Previous stud-
ies have explained the difference between the two types of
serine hydrolases by using the mechanistic details of amide
hydrolysis, that is, lipases lack the transition-state stabilizing
interaction between the active site and the amide nitrogen
atom, which is pivotal in the amide hydrolysis by serine pro-
teases.[6] Some lipases such as Candida antarctica lipase B
(CAL-B, the Novozym 435 preparation) and Burkholderia
cepacia lipase (lipase PS-D preparation) have been success-
fully used to cleave a β-lactam ring in a highly effective and
enantioselective manner,[1d,1g,2] which can be explained by
the fact that the amide bond in a β-lactam ring is less stable
than a normal resonance stabilized peptide bond. However,
the use of elevated temperatures or structural activation has
been a prerequisite with certain β-lactam structures. Thus,
in the first lipase-catalyzed β-lactam ring-opening case, N-
benzoyl protection was used to increase the reactivity.[7] The
CAL-B-catalyzed enantioselective ring-opening of many β-
lactams, such as that of 4-phenylazetidin-2-one (rac-1a),[8b]
with or without added water in organic solvents was pre-
viously reported to need the elevated temperature of
60 °C.[8] Fluorine-substitution allowed for the enantio- and
diastereoselective ring-opening of 3-trans-mono- and 3,3-di-
fluorinated rac-1a in lipase PS-D catalyzed transformations
to lead to the formation of the corresponding β-amino
esters and amides, β-dipeptides, and methyl α-d-glyco-
pyranoside conjugates.[2]
[a] Institute of Biomedicine/Department of Pharmacology,
Drug Development and Therapeutics/Laboratory of Synthetic
Drug Chemistry, University of Turku,
20014 Turku, Finland
E-mail: lkanerva@utu.fi
http://www.utu.fi/en/units/med/units/pharmacology/Pages/
home.aspx
In this current work, the effect of N-activation on the
ring-opening reaction of rac-1a at room temperature
(23 °C) has been studied (Scheme 1). N-Protecting groups
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201403467.
1500
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 1500–1506

Lipase-Catalyzed Enantioselective Preparation of β-Amino Esters
were selected on the basis of their ease of installation/re-
moval and by varying the electron-withdrawing properties
to affect the electrophilic nature of the lactam carbonyl
group toward a nucleophile and decrease the stability of the
ring C–N bond. Structurally the N-protected substrates can
be grouped as N-acyl (i.e., rac-1b–rac-1d) and N-alkyl (i.e.,
rac-1e–rac-1g) types. Some of the N-protected substrates
such as N-acetylated (R)-2b (a noncompetitive inhibitor of
human α-chymotrypsin)[9] are biologically active. In prac-
tice, the lipase-catalyzed reaction of rac-1a–rac-1g has been
examined by employing an alcoholysis reaction with meth-
anol (and 1-butanol) in dry organic solvents. The main fo-
cus of the studies involved the opening of the lactam ring
(Scheme 1, route a). However, the incorporation of groups
such as N-acetyl, N-chloroacetyl, and N-tert-butoxycarb-
onyl (N-Boc) might introduce the possibility for an enzy-
matic N-deprotection route (Scheme 1, route b) in addition
to a ring-opening route when the enzyme correctly binds
the substrate. On the basis of previous work, it was found
that the ring-opening reactions of rac-1a-type substrates are
enantioselective to give the (R) isomer by using lipases such
as CAL-B and lipase PS-D.[2,3,8b] If not otherwise stated,
the absolute configurations in the present work are in ac-
cordance with this, as confirmed by comparing the optical
rotations of products 2 to literature values and by chiral
GC analysis.
Scheme 2. Synthesis of rac-1a–rac-1g. Reagents and conditions:
(a) (1) chlorosulfonyl isocyanate (CSI), dichloromethane (DCM),
room temp., (2) K2CO3, Na2SO3, H2O, room temp.; (b) Ac2O, 4-
(N,N-dimethylamino)pyridine (DMAP), DCM, room temp.;
(c) (ClCH2CO)2O, DMAP, DCM, room temp.; (d) Boc2O, DMAP,
DCM, room temp.; (e) allyl bromide, K2CO3, acetonitrile, reflux;
(f) BrCH2CO2Et, trimethylsilyl chloride (TMSCl), Zn, toluene, re-
flux; (g) tert-butyldiphenylsilyl chloride (TBDPSCl), triethylamine
(TEA), DMAP, acetonitrile. See Supporting Information for fur-
ther details.
N-4-methoxybenzyl-protected (N-PMB-protected) deriva-
tive rac-1f was obtained by a Gilman–Speeter variation of
the Reformatsky reaction between N-(4-methoxybenzyl)-
benzaldimine (4) and ethyl bromoacetate (2 equiv.) in the
presence of zinc dust in refluxing toluene.[11] The syntheses
of rac-1a–rac-1g are presented in more detail in the Sup-
porting Information.
Enzymatic Reactions of rac-1a–rac-1g
As previously mentioned, the ring-opening reaction of
rac-1a with CAL-B and water has been successful at ele-
vated temperatures, with the (R) enantiomer of the β-lactam
being the more reactive one.[8b] Each of the substrates rac-
1a–rac-1g (50 mm) was first subjected to an alcoholysis re-
action with methanol (2 equiv.) in the presence of CAL-B
(30 mgmL–1) in dry diisopropyl ether (DIPE) at 23 °C. The
substrates can be grouped according to reactivity as those
that underwent a reaction with at least one of the screened
lipases (i.e., rac-1a–rac-1c) and those that did not undergo
Scheme 1. Kinetic resolution of rac-1: (a) by lipase-catalyzed ring-
opening of the β-lactam and (b) by lipase-catalyzed N-deprotection
through alcoholysis.
Results and Discussion
Synthesis of Racemic N-Protected β-Lactams
Racemic 4-phenylazetidin-2-one (rac-1a) was obtained any reaction (i.e., rac-1d–rac-1g). The fact that the ring-
by the 1,2-dipolar cycloaddition of chlorosulfonyl isocyan- opening reaction of N-Boc-protected rac-1d was not pos-
ate to styrene as described previously.[2c] The N-substituted sible with methanol and CAL-B (or other lipases) is in ac-
analogues rac-1b–rac-1d were obtained by common cordance with our previous work with amines as nucleo-
reactions with the corresponding anhydride Ac2O, philes.[2b] We propose that steric factors restricted the enzy-
(ClCH2CO)2O, and Boc2O (Scheme 2).
matic ring-openings of rac-1d–rac-1g. As further support,
The N-allyl derivative rac-1e was prepared by heating CAL-B has been shown to be ineffective in previous ring-
rac-1a at reflux with allyl bromide in the presence of K2CO3 opening experiments of N-para-methoxyphenyl-substituted
in acetonitrile, whereas tert-butyldiphenylsilyl-protected α-methylene-β-lactams with water, although the unprotec-
rac-1g was synthesized by the reaction of rac-1a with tert- ted β-lactams were good substrates for the enzyme.[12] Com-
butyldiphenylsilyl chloride, triethylamine, and DMAP in pounds rac-1d–rac-1f were omitted from further studies.
acetonitrile as an adaptation of a published method.[10] The The results for the successful cases are given in Table 1.
Eur. J. Org. Chem. 2015, 1500–1506
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1501

R. Sundell, L. T. Kanerva
FULL PAPER
Table 1. Alcoholysis of rac-1a–rac-1c (50 mm) with MeOH reached the theoretical 50% conversion much faster than
(2 equiv.) in the presence of a lipase preparation (30 mgmL–1) in
rac-1c (Figure 1), which demonstrates the importance of the
structure of the substrate over its expected chemical reactiv-
DIPE at 23 °C for 24 h (R2 = Me).
ity in enzymatic reactions. When methanol was replaced
with 1-butanol, the ring-opening reaction of rac-1a in DIPE
proceeded very slowly (conversion 5% in 24 h, Table 1, En-
try 2), which indicates the potential for side reactions rather
than the alcoholysis. In spite of reactivity differences, the
enantioselectivities of the reactions of rac-1a–rac-1c were
Entry Substrate Lipase
% Conv.[a] % ee1a–1c % ee2a–2c
E[b]
always excellent (E Ͼ 200). Although rac-1a was only reac-
tive with CAL-B, the use of Lipozyme TL IM (lipase from
Thermomyces lanuginosus immobilized on granulated silica)
as well as CAL-B led to the ring-opening of rac-1b (Table 1,
Entries 3 and 5). Interestingly, however, Lipozyme TL IM
favored the formation of (S)-2b with low enantioselectivity
(Table 1, Entry 5) rather than the formation of the expected
(R)-2b enantiomer. Finally, N-chloroacetyl-activated rac-1c
was reactive with all of the investigated lipase preparations
(Table 1, Entries 4 and 6–10), although the observed
enantioselectivity of the ring-opening for the (R) isomer
was negligible with enzymes other than CAL-B and the two
Burkholderia cepacia lipase preparations (Table 1, Entries 7
and 8). Moreover, the reaction of rac-1c with Candida ant-
arctica lipase A (CAL-A as the NZL-101-IMB preparation)
was uncommon, as the enzymatic cleavage of the N-
1
2
3
4
5
6
7
8
9
10
rac-1a
CAL-B
26 (25)
5 (18)
35
5
99
Ͼ99
7
Ͼ99
Ͼ99
Ͼ99
Ͼ99
50[d]
65
Ͼ200
Ͼ200
Ͼ200
Ͼ200
3
rac-1a[c] CAL-B
rac-1b
rac-1c
rac-1b
rac-1c
rac-1c
rac-1c
rac-1c
rac-1c
CAL-B
CAL-B
TL IM
TL IM
PS-C II
PS-D
50 (54)
50 (57)
13 (13)
11 (18)
32 (29)
21 (19)
46 (46)
4 (15)
8
5
47
27
34
3
Ͼ99
Ͼ99
39
Ͼ200
Ͼ200
3
RM IM
CAL-A
75[e]
7
[a] Conversion based on enantiomeric excess (ee) values (conver-
sion as the disappearance of 1 against an internal standard in pa-
renthesis). [b] E is the enantiomeric ratio value. [c] nBuOH
(2 equiv.) was employed instead of MeOH. [d] The product is (S)-
2b. [e] The product is (S)-1a through route b (Scheme 1).
In addition to the added alcohol, the supposed residual chloroacetyl group took place through route b (Scheme 1)
water in the seemingly dry lipase preparation in a dry or- rather than the ring-opening reaction (Table 1, Entry 10).
ganic solvent and, with rac-1a, the free amino group of the This observation further confirms the previously discovered
formed β-amino ester (and the acid if formed) can act as a exceptional properties of CAL-A, which can be employed,
competitive nucleophile for the ring-opening reaction of a for instance, as an N-acylation catalyst in the kinetic and
β-lactam.[2b,2d,3a] In such a case, the conversion, which is dynamic kinetic resolution of sterically hindered hetero-
the disappearance of a β-lactam, and the obtained enantio- cyclic proline as well as 2-piperidine- and 2-piperazine-sub-
meric excess values for the unreacted β-lactam should con- stituted carboxylic acid esters.[5c,14] This interesting slow de-
tain contributions from side reactions. Conversion values protection reaction was not studied further.
in Table 1 are based on ee1 (unreacted substrate) and ee2
(product) values through the equation conversion (c) = ee1/
(ee1 + ee2)[13] and on the disappearance of rac-1 against an
internal standard (in parenthesis), the latter method giving
the real total conversion. Differences in the conversion val-
ues can be expected when hydrolysis and/or aminolysis are
involved. Within the limits of experimental accuracy, the
methods gave very similar values, which indicate the lack of
or at least minimal possibility of side reactions in dry
DIPE. Moreover, HPLC analysis did not give signs of side
product formation. Significant differences were evident only
when the ring-opening reaction of rac-1a was performed
with 1-butanol (Table 1, Entry 2) and when CAL-A cleaved
the N-chloroacetyl group (Scheme 1, route b) from rac-1c
with methanol (Table 1, Entry 10).
The results of the employment of rac-1a–rac-1c as sub-
Figure 1. Alcoholysis of rac-1b (50 mm, squares) and rac-1c
strates in DIPE at 23 °C indicate that although rac-1a
(50 mm, circles) with MeOH (2 equiv.) in the presence of CAL-B
underwent the reaction slowly with CAL-B (30 mgmL–1)
and methanol (26% conversion after 24 h, 72 h was needed
to reach 49% conversion), both N-acetylated rac-1b and N-
chloroacetylated rac-1c reached 50% conversion in 24 h
(Table 1, Entries 1, 3, and 4, respectively). The results under
(10 mgmL–1) in TBME at 23 °C. Conversion (solid line) and eeS (S
= unreacted substrate, dashed line), eeP (P = formed product) al-
ways Ն99%.
The CAL-B catalyzed alcoholysis of rac-1b (50 mm) with
optimized conditions later revealed that the ring-opening methanol (2 equiv.) was used for further optimization stud-
reaction of rac-1b in tert-butyl methyl ether (TBME) ies. When the effects of the solvent were studied by carrying
1502
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 1500–1506

Lipase-Catalyzed Enantioselective Preparation of β-Amino Esters
out the reaction in TBME, toluene, and hexane in addition content was increased to 5 and 10 equiv. in the reaction of
to DIPE with 30 mgmL–1 of the enzyme, excellent enantio- rac-1b (50 mm) with 10 mgmL–1 of CAL-B, the result after
selectivities and reactivities were evident in all four solvents 24 h was the same as that with only 2 equiv. (Table 2, En-
(Table 2, Entries 1–4). These solvents are generally used in tries 6–8). Accordingly, we decided not to increase the
lipase-catalyzed transesterification reactions, and their ap- amount of methanol from 2 equiv. When rac-1b (50, 100,
plicability in many cases is substrate dependent. There are and 200 mm) was subjected to the reaction with CAL-B
often other aspects that affect the selection of a solvent be- (10 mgmL–1) and methanol (2 equiv.), the reactivity de-
sides seeking one in which the reaction proceeds with ap- creased somewhat with the increasing substrate concentra-
propriate reactivity and excellent enantioselectivity. For in- tion. However, all of the β-lactam concentrations yielded
stance, TBME can be regarded as a safer option than DIPE a 50% conversion in 24 h with excellent enantioselectivity
because of its lower potential for peroxide formation. Sub- (Figure 2). Accordingly, this method provides an opportu-
strate and product solubilities may also affect the selection nity to widely vary both the β-lactam and methanol con-
of a solvent. For instance, in the present work, the low solu- tents for synthetic purposes and still perform the reaction
bility of rac-1b in hexane is an obvious reason for the devia- with a low enzyme content without the need for elevated
tion between the conversion values after 24 h when deter- temperatures.
mined on the basis of ee values (50%) and on the use of an
internal standard (70%, Table 2, Entry 4).
Table 2. Effects of solvent and CAL-B content (5–50 mgmL–1) on
the alcoholysis of rac-1b (50 mm) with MeOH (2 equiv.) at 23 °C.
Entry Solvent
CAL-B
%
% ee1b % ee2b
E
[mgmL–1]
Conv.[a]
1
2
3
4
5
6
7
8
9
DIPE
30
30
30
30
5
10
10
10
50
50 (54)
50 (53)
46 (46)
50 (70)
47 (47)
50 (50)
50 (51)
50 (51)
50 (56)
99
Ͼ99
85
Ͼ99
90
Ͼ99
Ͼ99
Ͼ99
Ͼ99
Ͼ99
99
Ͼ99
99
Ͼ99
Ͼ99
Ͼ99
Ͼ99
99
Ͼ200
Ͼ200
Ͼ200
Ͼ200
Ͼ200
Ͼ200
Ͼ200
Ͼ200
Ͼ200
TBME
toluene
hexane[b]
TBME
TBME
TBME[c]
TBME[d]
TBME
Figure 2. Effect of substrate concentration on the alcoholysis of
rac-1b [50 mm (squares), 100 mm (spheres) and 200 mm (triangles)]
with MeOH (2 equiv.) and CAL-B (10 mgmL–1) in TBME at 23 °C
[conversion (solid line) and eeS (dashed line)].
[a] Conversion based on ee values (conversion as the disappearance
of 1 against an internal standard in parenthesis). [b] Poor initial
solubility of rac-1b. [c] With MeOH (5 equiv.). [d] With MeOH
(10 equiv.).
Preparative Scale Synthesis and Product Characterization
Preparative scale reactions (reaction volume 10–20 mL)
of rac-1a–rac-1c were conducted with methanol and CAL-
Finally, the optimization studies with rac-1b (50 mm) B (10–30 mgmL–1) in TBME under the conditions de-
were continued by investigating the effect of CAL-B content scribed in Table 3. When rac-1a (50 mm) was submitted to
(5–50 mgmL–1) on the reactivity and enantioselectivity of the reaction with methanol (2 equiv.) and CAL-B
the reaction in TBME (Table 2, Entries 2, 5, 6, and 9). The (30 mgmL–1), the reaction proceeded slowly and reached
enzyme content did not affect enantioselectivity. The theo- 47% conversion after 4 d (Table 3, Entry 1). In accordance
retical 50% conversion was reached in 24 h by using a with an excellent enantioselectivity (E Ͼ 200), the ester
CAL-B content of 10 mgmL–1 or higher (Table 2, Entries 2, product (R)-2a was isolated with an ee value of Ͼ99% and
6, and 9). However, when the enzyme content was in- the unreacted (S)-1a with 94%ee at the given conversion.
creased, the amount of the residual water in the reaction Although it is relatively common that preparative-scale re-
system also increased, which raised the possibility for a actions proceed slower than small-scale reactions because
hydrolysis side reaction. In accordance with this, the total of factors such as different mixing conditions, the large, un-
conversion values in parentheses gradually increased with a expected decrease in rate and long reaction time might in-
higher enzyme content. These values already differed by 6% crease the possibility of side reactions. The reaction be-
when 50 mgmL–1 compared to 10 mgmL–1 of CAL-B were tween rac-1a and methanol was repeated in the presence of
used (Table 2, Entry 9 vs. 6).
Boc2O (to protect the primary amino group in situ without
An increase in the methanol content from 2 equiv. with a catalyst)[2c,15] to rule out a possible enzymatic or chemical
respect to rac-1b might help to favor the desired alcoholysis peptide bond formation by aminolysis.[2b,2d] In spite of the
over the competing hydrolysis reaction. When the methanol small reduction in the conversion in the presence of Boc2O,
Eur. J. Org. Chem. 2015, 1500–1506
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1503

R. Sundell, L. T. Kanerva
FULL PAPER
the reaction proceeded in a similar manner (Table 3, En- lipases, whereas the N-chloroacetyl-protected rac-1c under-
try 1 vs. 2). This together with the fact that no peptide prod- went the ring-opening reaction with all of examined lipases
uct was isolated from the reaction mixture was taken as except CAL-A, which resulted in an exocyclic fission of the
proof of the lack of a competing aminolysis. When rac-1b C–N bond of the N-acyl moiety. Excellent enantioselectivity
(50 mm) was treated with methanol (2 equiv.) in the pres- for the (R) isomer (E Ͼ 200) was evident by using Candida
ence of CAL-B (30 mgmL–1), the reaction stopped at 50% antarctica lipase B (CAL-B as the Novozym 435 prepara-
conversion, and the corresponding methyl ester (S)-2b and tion) and Burkholderia cepacia lipase (lipase PS-D and PS-
unreacted (R)-1b were isolated in 41 and 40% yield, respec- C II preparations) as the catalysts. Preparative-scale ring-
tively, in enantiopure forms (Table 3, Entry 3). It was also opening reactions of rac-1a–rac-1c by employing CAL-B in
confirmed that a substrate concentration of 200 mm and TBME allowed for (R)-2a–(R)-2c to be separated in
only 10 mgmL–1 of CAL-B content yielded an excellent enantiopure forms with conversions of 41–50%. With this
outcome for a preparative-scale reaction (Table 3, Entry 4), method, wide variations of both β-lactam and methanol
consistent with the promise of the small-scale experiments. contents with a low enzyme concentration was possible
The production of (R)-1b and (S)-2b (Table 3, Entry 5) by without the need of elevated temperatures. We also showed
using Lipozyme TL IM on a preparative scale also pro- that the enantiodiscrimination of Lipozyme TL IM catalyst
ceeded with a similar efficiency as the screening experiment was exceptional, preferentially giving ring-opened (S)-2b
of Table 1, Entry 5). The reaction of rac-1c (50 mm) reached rather than the (R) enantiomer.
only 46% conversion in 24 h (Table 3, Entry 6) to allow for
the preparation of (R)-2c. We also confirmed the cleavage
Experimental Section
of the N-acyl group of rac-1c by using CAL-A (Scheme 1,
route b) and separating and characterizing enantiomerically
enriched (S)-1a (44%ee) and the unreacted 1c (Table 3, En-
try 7).
General Methods: All reagents and materials were purchased from
commercial sources and used as received, with the exception of the
solvents, which were dried over molecular sieves (3 Å) prior to use.
Powdered zinc was acid-washed prior to use.[16] Lipase preparations
from Amano (Burkholderia cepacia lipase as Lipase PS-D and
Lipase PS-C II) and Codexis [Candida antarctica lipase A (CAL-
A) as NZL-101-IMB, Candida antarctica lipase B (CAL-B) as No-
vozym 435, Rhizomucor miehei lipase as Lipozyme RM IM, and
Thermomyces lanuginosus lipase as Lipozyme TL IM] were used.
The lipase-catalyzed reactions were monitored by chiral HPLC or
GC analysis. HPLC analyses were performed with an HP 1090
HPLC/DAD that was equipped with a Daicel CHIRACEL-OD-H
(4.6 mmϫ250 mmϫ5 μm) column. GC analyses were performed
with a HP 6850 GC/FID that was equipped with a Chrompack CP-
Chiralsil-DEX CB (25 mϫ0.25 mmϫ0.25 μm) capillary column.
Retention times of studied compounds are presented in Table S1
(Supplementary Information). The enantiomeric ratio values E
were determined by E = ln[(1 – eeS)/(1 – eeS/eeP)]/ln[(1 + eeS)/(1 +
eeS/eeP)], which is obtained by substituting the conversion (c) =
(eeS)/(eeS + eeP) into the original equation of Chen and Sih (eeS
and eeP refer to the enantiomeric excess values of the unreacted
substrate and the formed product, respectively, at the point of con-
version).[13] Conversion values were calculated from the ee values
as well as from the disappearance of a β-lactam against an internal
standard. Analytical thin layer chromatography (TLC) was carried
out on Merck Kieselgel 60F254 sheets, and the spots were visualized
by UV (254 nm). Chromatographic separations were performed by
column chromatography on Kieselgel 60 (0.063–0.200 μm). The 1H
and 13C NMR spectroscopic data were recorded at 298 K in CDCl3
against an internal standard (TMS) with a Bruker Avance 500 spec-
trometer. Mass spectra were recorded in the positive mode with a
Bruker Daltonics micrOTOF-Q (ESI). Specific rotations were mea-
sured against the sodium D line (589 nm) with a Perkin–Elmer 341
Polarimeter, and the values are presented as 10–1 degcm–2 g–1. Melt-
ing points were measured with a Gallenkamp device.
Table 3. Preparative-scale kinetic resolution of rac-1a–rac-1c
(50 mm) with MeOH (2 equiv.) and CAL-B (30 mgmL–1) in TBME
at 23 °C.
Entry Substrate Lipase Time [h] % Conv.[a] % Yield 1[b] % Yield 2[b]
1
2
3
4
5
6
7
rac-1a
rac-1a[c] CAL-B
rac-1b CAL-B
rac-1b[d] CAL-B
CAL-B
96
96
24
24
48
24
72
47
41
50
50
57
46
5
38 (94, S)
33 (70, S)
40 (Ͼ99, S) 41 (Ͼ99, R)
47 (Ͼ99, S) 49 (Ͼ99, R)
40 (58, R) 34 (49, S)
36 (Ͼ99, R)
23 (Ͼ99, R)
rac-1b
rac-1c
rac-1c
TL IM
CAL-B
CAL-A
26 (91, S)
38 (5, R)
25 (Ͼ99, R)
6 (44, S)[e]
[a] Calculated from the ee values. [b] Isolated yield as calculated
from rac-1 (% ee value and configuration of enantiomer are in pa-
renthesis). [c] Kinetic resolution of rac-1a in the presence of Boc2O
(1 equiv.). Isolated product is (R)-2d. [d] rac-1b (200 mm), MeOH
(500 mm), and CAL-B (10 mgmL–1). [e] (S)-1a is the product in-
stead of 2.
Conclusions
The effect of six nitrogen protecting groups on the enan-
tioselective ring-opening reaction of 4-phenylazetidin-2-one
(rac-1a) was assessed with methanol and five lipases in im-
mobilized form in organic solvents. We were able to show
that the ring-opening reaction is sensitive to the structure
of the nitrogen protecting group. Either steric effects or
poor N-activation made the N-Boc-, N-allyl-, N-4-meth-
oxybenzyl-, and tert-butyldiphenylsilyl-substituted sub-
Small-Scale Enzymatic Reactions with Lactams rac-1a–rac-1g: In a
given small-scale experiment, rac-1a–rac-1g (50 mm) was dissolved
in a solvent [1 mL, including the internal standard (5 mm dihexyl
ether or methoxybenzene)], and then a lipase preparation (5–
50 mgmL–1) and alcohol (MeOH or 1-BuOH, 2–10 equiv.) were
strates (i.e., rac-1d– rac-1g) unreactive towards the studied added. The reaction mixture was shaken (170 rpm) at room tem-
1504 Eur. J. Org. Chem. 2015, 1500–1506
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Lipase-Catalyzed Enantioselective Preparation of β-Amino Esters
perature. Samples were taken at intervals from the reaction mixture,
and these were filtered and analyzed by the developed GC and/
temperature for 24 h before being filtered, and then the solvents
were evaporated. Column chromatography (ethyl acetate/hexane,
or HPLC methods (retention times for substrates, products, and 1:4; then pure acetone) yielded (S)-1b (40 mg, 0.21 mmol, 40%
standards are presented in the Supplementary Information).
yield, Ͼ99%ee) and (R)-2b (47 mg, 0.21 mmol, 41% yield,
Ͼ99%ee). Rf = 0.20 (ethyl acetate/hexane, 1:4 for 1b); Rf = 0 (ethyl
acetate/hexane, 1:4 for 2b). Data for (R)-1b: H NMR (500 MHz,
Preparative-Scale Enzymatic Reactions with Lactams rac-1a–rac-1c
1
Methyl (R)-3-Amino-3-phenylpropanoate [(R)-2a] and (S)-4-Phenyl-
azetidin-2-one [(S)-1a]: Lactam rac-1a (154 mg, 1.05 mmol) was dis-
solved in TBME (20 mL), and then MeOH (84 μL, 2.1 mmol,
2 equiv.) and CAL-B (30 mgmL–1, 600 mg) were added. The reac-
tion mixture was shaken (170 rpm) at room temperature for 72 h
before being filtered, and then the solvents were evaporated. Col-
umn chromatography (ethyl acetate/petroleum ether, 1:1) yielded
(S)-1a (59 mg, 0.40 mmol, 38% yield, 94%ee) as an off-white solid
and (R)-2a (67 mg, 0.37 mmol, 36% yield, Ͼ99%ee) as a thick oil.
Rf = 0.37 (ethyl acetate/petroleum ether, 1:1 for 1a); Rf = 0.05 (ethyl
acetate/petroleum ether, 1:1 for 2a). Data for (S)-1a: 1H NMR
(500 MHz, CDCl3, 298 K): δ = 2.87 (ddd, J = 1.0 Hz, J = 2.5 Hz,
J = 14.9 Hz, 1 H, CHaHb), 3.44 (ddd, J = 2.5 Hz, J = 5.3 Hz, J =
14.9 Hz, 1 H, CHaHb), 4.72 (dd, J = 2.5 Hz, J = 5.3 Hz, 1 H,
CHPh), 6.40 (br., 1 H, NH), 7.32 (m, 1 H, Ar), 7.38 (m, 4 H,
Ar) ppm. 13C NMR (126 MHz, CDCl3, 298 K): δ = 48.01 (CH2),
50.39 (CHPh), 125.65 (Ar), 128.24 (Ar), 128.86 (Ar), 140.20 (Ar),
168.13 (C=O) ppm. HRMS: calcd. for C9H9NONa [M + Na]+
170.05764; found 170.05986. [α]2D2 = –119.2 (c = 0.5, EtOH); ref.[3b]
[α]2D2 = –140.5 (c = 0.5, EtOH; 99%ee), m.p. 100–101 °C; ref.[17]
m.p. 116–117 °C. Data for (R)-2a: 1H NMR (500 MHz, CDCl3,
298 K): δ = 1.78 [br., 2 H, NH2], 2.67 (d, J = 6.3 Hz, 2 H, CH2),
3.69 (s, 3 H, CO2CH3), 4.42 (dd, J = 6.4 Hz, J = 7.3 Hz, 1 H,
CHPh), 7.26 (m, 1 H, Ar), 7.34 (m, 4 H, Ar) ppm. 13C NMR
(126 MHz, CDCl3, 298 K): δ = 43.95 (CH2), 51.69 (CO2CH3),
52.63 (CHPh), 126.17 (Ar), 127.46 (Ar), 128.68 (Ar), 141.61 (Ar),
172.51 (C=O) ppm. HRMS: calcd. for C10H13NO2Na [M + Na]+
202.08385; found 202.08423. [α]2D0 = +21.9 (c = 1.99, CHCl3); ref.[18]
[α]2D0 = +22.3 (c = 1.99, CHCl3; Ͼ98%ee).
CDCl3, 298 K): δ = 2.43 (s, 3 H, CH3), 2.97 (dd, J = 3.5 Hz, J =
16.4 Hz, 1 H, CHaHb), 3.51 (dd, J = 6.5 Hz, J = 16.4 Hz, 1 H,
CHaHb), 5.03 (dd, J = 3.4 Hz, J = 6.5 Hz, 1 H, CHPh), 7.32 (m, 3
H, Ar), 7.38 (m, 2 H, Ar) ppm. 13C NMR (126 MHz, CDCl3,
298 K): δ = 24.02 (COCH3), 45.69 (CH2), 52.52 (CHPh), 125.81
(Ar), 128.50 (Ar), 128.96 (Ar), 137.74 (Ar), 165.35 (COCH3),
167.51 (C=O) ppm. HRMS: calcd. for C11H11NO2Na [M + Na]+
212.06820; found 212.06721. [α]2D5 = –213.0 (c = 1.0, MeOH). Data
1
for (R)-2b: H NMR (500 MHz, CDCl3, 298 K): δ = 2.03 [s, 3 H,
NH(CO)CH3], 2.84 (dd, J = 5.9 Hz, J = 15.8 Hz, 1 H, CHaHb),
2.94 (dd, J = 5.8 Hz, J = 15.8 Hz, 1 H, CHaHb), 3.62 (s, 3 H,
CO2CH3), 5.43 (ddd, J = 5.8 Hz, J = 5.9 Hz, J = 8.4 Hz, 1 H,
CHNHAc), 6.58 (d, J = 7.6 Hz, 1 H, CHNHAc), 7.27 (m, 3 H,
Ar), 7.34 (m, 2 H, Ar) ppm. 13C NMR (126 MHz, CDCl3, 298 K):
δ = 23.44 (NHCOCH3), 39.69 (CH2), 49.48 (CHPh), 51.84
(CO2CH3), 126.26 (Ar), 127.67 (Ar), 128.74 (Ar), 140.48 (Ar),
169.29 [NH(CO)CH3], 171.77 (CO2CH3) ppm. HRMS: calcd. for
C12H15NO3Na [M + Na]+ 244.09441; found 244.09410. [α]2D5
=
+62.1 (c = 1.0, MeOH); ref.[8] [α]2D2 = +60.1 (c = 0.90, CHCl3), m.p.
99–100 °C; ref.[8] m.p. 94–95 °C.
Methyl (S)-3-Acetamido-3-phenylpropanoate [(S)-2b] and (R)-N-
Acetyl-4-phenylazetidin-2-one [(R)-1b]: As above, rac-1b (95 mg,
0.50 mmol) was treated with MeOH (43 μL, 1.0 mmol, 2 equiv.) in
the presence of Lipozyme TL IM (30 mgmL–1, 300 mg) in TBME
(10 mL) at room temp. for 48 h. Column chromatography (ethyl
acetate/hexane, 1:4; then pure acetone) afforded (R)-1b (38 mg,
0.20 mmol, 40% yield, 59%ee) and (S)-2b (39 mg, 0.17 mmol, 35%
yield, 49%ee). Rf = 0.23 (ethyl acetate/hexane, 1:4 for 1b). Rf = 0
(ethyl acetate/hexane, 1:4 for 2b). The 1H and 13C NMR spectra of
(R)-1b and (S)-2b together with the HRMS data were identical to
rac-1b and (R)-2b, respectively. Data for (R)-1b: [α]2D5 = +121.5 (c
= 1.0, MeOH). Data for (S)-2b: [α]2D5 = –36.8 (c = 1.0, MeOH);
ref.[20] [α]2D5 = –79.9 (c = 1.00, MeOH; Ͼ99%ee), m.p. 91–92 °C;
ref.[20] m.p. 99–101 °C.
Methyl (R)-3-(tert-Butoxycarbonyl)amino-3-phenylpropanoate [(R)-
2d] and (S)-4-Phenylazetidin-2-one [(S)-1a]: Lactam rac-1a (74 mg,
0.50 mmol) was dissolved in TBME (10 mL), and then Boc2O
(109 mg, 0.50 mmol, 1 equiv.), MeOH (43 μL, 1.06 mmol, 2 equiv.),
and CAL-B (30 mgmL–1, 300 mg) were added. The reaction mix-
ture was shaken (170 rpm) at room temperature for 72 h before
being filtered, and then the solvents were evaporated. Column
chromatography (ethyl acetate/petroleum ether, 1:9–1:1) yielded
(S)-1a (25 mg, 0.17 mmol. 33% yield, 70%ee) and (R)-2d (32 mg,
0.12 mmol, 23% yield, Ͼ99%ee). Rf = 0.37 (ethyl acetate/petro-
leum ether, 1:1 for 1a); Rf = 0.90 (ethyl acetate/petroleum ether, 1:1
for 2e). Data for (S)-1a: The 1H NMR, 13C NMR, and HRMS
data were identical to that above for (S)-1a. [α]2D2 = –93.6 (c = 0.5,
Methyl (R)-3-(Chloroacetamido)-3-phenylpropanoate [(R)-2c] and
(S)-N-Chloroacetyl-4-phenylazetidin-2-one [(S)-1c]: As above, rac-1c
(112 mg, 0.5 mmol) was treated with MeOH (43 μL, 1.06 mmol,
2 equiv.) in the presence of CAL-B (10 mgmL–1, 100 mg) in TBME
(10 mL) at room temp. for 24 h. Column chromatography (ethyl
acetate/hexane, 1:4; then pure acetone) afforded (S)-1c (29 mg,
0.13 mmol, 26% yield, 92%ee) and (R)-2c (32 mg, 0.12 mmol, 25%
yield, Ͼ99%ee). Rf = 0.19 (ethyl acetate/hexane, 1:4 for 1c). Rf
= 0 (ethyl acetate/hexane, 1:4 for 2c). Data for (S)-1c: 1H NMR
(500 MHz, CDCl3, 298 K): δ = 3.07 (dd, J = 3.6 Hz, J = 16.6 Hz,
1 H, CHaHb), 3.57 (dd, J = 6.6 Hz, J = 16.6 Hz, 1 H, CHaHb),
4.47 (s, 2 H, CH2Cl), 5.11 (dd, J = 3.6 Hz, J = 6.6 Hz, 1 H, CHPh),
7.35 (m, 3 H, Ar), 7.39 (m, 2 H, Ar) ppm. 13C NMR (126 MHz,
CDCl3, 298 K): δ = 42.99 (CH2Cl), 45.80 (CH2), 52.89 (CHPh),
125.92 (Ar), 128.84 (Ar), 129.07 (Ar), 136.78 (Ar), 163.41
(NHCOCH2Cl), 164.79 (C=O) ppm. HRMS: calcd. for
1
EtOH). Data for (R)-2d: H NMR (500 MHz, CDCl3, 298 K): δ =
1.42 [s, 9 H, C(CH3)3], 2.85 (m, 2 H, CH2), 3.62 (s, 3 H, CO2CH3),
5.11 (br., 1 H, CHPh), 5.46 (br., 1 H, NH), 7.30 (m, 5 H, Ar) ppm.
13C NMR (126 MHz, CDCl3, 298 K): δ = 28.35 [C(CH3)3], 40.80
(CH2), 51.19 (CHPh), 51.78 (CO2CH3), 79.68 [C(CH3)3], 126.12
(Ar), 127.53 (Ar), 128.67 (Ar), 141.17 (Ar), 155.05 (NCOtBu),
171.42 (C=O) ppm. HRMS: calcd. for C15H21NO4Na [M + Na]+
302.13628; found 302.13604. [α]2D0 = +29.2 (c = 1.4, CHCl3); ref.[19]
[α]2D0 = +29.9 (c = 1.4, CHCl3; Ͼ97%ee), m.p. 94–95 °C; ref.[19]
m.p. 92–93.5 °C.
C11H10NO2ClNa [M + Na]+ 246.02923; found 246.02951. [α]2D0
–145.9 (c = 1.0, CHCl3). Data for (R)-2c: 1H NMR (500 MHz,
=
Methyl (R)-3-Acetamido-3-phenylpropanoate [(R)-2b] and (S)-N- CDCl3, 298 K): δ = 2.88 (dd, J = 5.8 Hz, J = 15.9 Hz, 1 H,
Acetyl-4-phenylazetidin-2-one [(S)-1b]: Lactam rac-1b (100 mg, CHaHb), 2.97 (dd, J = 5.7 Hz, J = 15.9 Hz, 1 H, CHaHb), 3.64 (s,
0.52 mmol) was dissolved in DIPE (10 mL), and then MeOH
(43 μL, 1.06 mmol, 2 equiv.) and CAL-B (30 mgmL–1, 300 mg)
were added. The reaction mixture was shaken (170 rpm) at room
3 H, CH3), 4.08 (dd, J = 15.3 Hz, J = 19.5 Hz, 2 H, CH2Cl), 5.43
(dt, J = 5.8 Hz, J = 8.5 Hz, 1 H, CHPh), 7.29 (m, 3 H, Ar), 7.35
(m, 2 H, Ar), 7.76 (d, J = 7.6 Hz, 1 H, NHCOCH2Cl) ppm. 13C
Eur. J. Org. Chem. 2015, 1500–1506
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1505

R. Sundell, L. T. Kanerva
FULL PAPER
Wang, Z.-T. Huang, M.-X. Wang, Org. Biomol. Chem. 2010, 8,
4736–4743; c) Y.-F. Ao, D.-H. Leng, D.-X. Wang, L. Zhao, M.-
X. Wang, Tetrahedron 2014, 70, 4309–4316.
NMR (126 MHz, CDCl3, 298 K): δ = 39.63 (CH2), 42.65 (CH2Cl),
49.84 (CHPh), 51.98 (CO2CH3), 126.17 (Ar), 127.92 (Ar), 128.86
(Ar), 139.73 (Ar), 165.33 (COCH2Cl), 171.40 (CO2CH3) ppm.
HRMS: calcd. for C12H14NO3ClNa [M + Na]+ 278.05544; found
278.05495. [α]2D0 = +15.2 (c = 1.0, CHCl3).
[5] a) R. Fujii, Y. Nakagawa, J. Hiratake, A. Sogabe, K. Sakata,
Protein Eng. Des. Sel. 2005, 18, 93–101; b) A. Torres-Gavilán,
E. Castillo, A. López-Munguía, J. Mol. Catal. B 2006, 41, 136–
140; c) A. Liljeblad, P. Kallio, M. Vainio, J. Niemi, L. T.
Kanerva, Org. Biomol. Chem. 2010, 8, 886–895; d) P.-O. Syrén,
K. Hult, ChemCatChem 2011, 3, 853–860; e) J. Brem, L.-C.
Bencze, A. Liljeblad, M. C. Turcu, C. Paizs, F.-D. Irimie, L. T.
Kanerva, Eur. J. Org. Chem. 2012, 3288–3294.
(S)-4-Phenylazetidin-2-one [(S)-1a] and (R)-N-Chloroacetyl-4-phen-
ylazetidin-2-one [(R)-1c]: As above, rac-1c (169 mg, 0.76 mmol) was
treated with MeOH (62 μL, 1.52 mmol, 2 equiv.) in the presence of
CAL-A (30 mgmL–1, 450 mg) in TBME (15 mL) at room temp. for
72 h. Column chromatography (ethyl acetate/petroleum ether, 1:4–
3:1) afforded (S)-1a (6 mg, 0.04 mmol, 6% yield, 44%ee) and (R)-
1c (64 mg, 0.28 mmol, 38% yield, 5%ee). The 1H NMR, 13C
NMR, and HRMS data of (S)-1a and (R)-1c were identical to the
results above. Data for (S)-1a: [α]2D2 = –15.6 (c = 0.5, EtOH); m.p.
96–97 °C. Data for (R)-1c: [α]2D0 = +9.3 (c = 1.0, CHCl3).
[6] a) P.-O. Syrén, P. Hendil-Forssell, L. Aumailley, W. Besen-
matter, F. Gounine, A. Svendsen, M. Martinelle, K. Hult,
ChemBioChem 2012, 13, 645–648; b) P.-O. Syrén, FEBS J.
2013, 280, 3069–3083.
[7] R. Brieva, J. Z. Crich, C. J. Sih, J. Org. Chem. 1993, 58, 1068–
1075.
[8] a) E. Forró, F. Fülöp, Org. Lett. 2003, 5, 1209–1212; b) E.
Forró, T. Paál, G. Tasnádi, F. Fülöp, Adv. Synth. Catal. 2006,
348, 917–923.
[9] V. Peddie, M. Pietsch, K. M. Bromfield, R. N. Pike, P. J. Dug-
gan, A. D. Abell, Synthesis 2010, 1845–1859.
Supporting Information (see footnote on the first page of this arti-
cle): Retention times for GC and HPLC analysis, synthesis protocol
for rac-1a–rac-1g, 1H and 13C NMR spectra for rac-1a–rac-1g, (S)-
1a–(S)-1c, (R)-1b, and (R)-2a–(R)-2d.
[10] a) R. Singh, R. D. G. Cooper, Tetrahedron 1994, 50, 12049–
12064; b) M. Lee, S. S. Stahl, S. H. Gellman, Org. Lett. 2008,
10, 5317–5319.
[11] a) H. Gilman, M. Speeter, J. Am. Chem. Soc. 1943, 65, 2255–
2256; b) C. Palomo, F. P. Cossío, A. Arrieta, J. M. Odriozola,
M. Oiarbide, J. M. Ontoria, J. Org. Chem. 1989, 54, 5736–5745.
[12] W. Adam, P. Groer, H.-U. Humpf, C. R. Saha-Möller, J. Org.
Chem. 2000, 65, 4919–4922.
Acknowledgments
The Instrumentarium Science Foundation, Finland is gratefully ac-
knowledged (grant to R. S.).
[1] a) C. Palomo, J. M. Aizpurua, I. Ganboa, M. Oiarbide, in:
Enantioselective Synthesis of β-Amino Acids, 2nd ed. (Eds.: E.
Juaristi, V. Soloshonok), John Wiley & Sons, Hoboken, 2005,
p. 477–495; b) S. Kobayashi, H. Ishitani, Chem. Rev. 1999, 99,
1064–1094; c) D. Seebach, A. K. Beck, D. J. Bierbaum, Chem.
Biodiversity 2004, 1, 1111–1239; d) A. Liljeblad, L. T. Kanerva,
Tetrahedron 2006, 62, 5831–5854; e) B. Alcaide, P. Almendros,
[13] C.-S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am. Chem.
Soc. 1982, 104, 7294–7299.
[14] a) A. Liljeblad, J. Lindborg, A. Kanerva, J. Katajisto, L. T.
Kanerva, Tetrahedron Lett. 2002, 43, 2471–2474; b) A. Liljeb-
lad, A. Kiviniemi, L. T. Kanerva, Tetrahedron 2004, 60, 671–
677; c) A. Hietanen, K. Lundell, L. T. Kanerva, A. Liljeblad,
ARKIVOC 2012, v, 60–74.
C. Aragoncillo, Chem. Rev. 2007, 107, 4437–4492; f) B. Weiner, [15] Y. Basel, A. Hassner, J. Org. Chem. 2000, 65, 6368–6380.
W. Szyman´ski, D. B. Janssen, A. J. Minnaard, B. L. Feringa,
Chem. Soc. Rev. 2010, 39, 1656–1691; g) E. Forró, F. Fülöp,
Curr. Med. Chem. 2012, 19, 6178–6187.
[16] B. S. Furniss, A. J. Hannaford, P. W. G. Smith, A. R Tatchell,
Vogel’s Textbook of Practical Organic Chemistry, 5th ed., Pear-
son Education Ltd., Harlow, England, 1989, p. 467.
[17] E. Forró, F. Fülöp, Tetrahedron: Asymmetry 2001, 12, 2351–
2358.
[2] a) X.-G. Li, M. Lähitie, M. Päiviö, L. T. Kanerva, Tetrahedron:
Asymmetry 2007, 18, 1567–1573; b) X.-G. Li, M. Lähitie, L. T.
Kanerva, Tetrahedron: Asymmetry 2008, 19, 1857–1861; c) R. [18] J. Jiang, K. K. Schumacher, M. M. Joullié, Tetrahedron Lett.
Sundell, E. Siirola, L. T. Kanerva, Eur. J. Org. Chem. 2014, 1994, 35, 2121–2124.
6753–6760; d) L. W. Schwab, R. Kroon, A. J. Schouten, K. [19] F. A. Davis, J. M. Szewczyk, Tetrahedron Lett. 1998, 39, 5951–
Loos, Macromol. Rapid Commun. 2008, 29, 794–797. 5954.
[3] a) X.-G. Li, L. T. Kanerva, Adv. Synth. Catal. 2006, 348, 197– [20] J. You, H.-J. Drexler, S. Zhang, C. Fischer, D. Heller, Angew.
205; b) X.-G. Li, L. T. Kanerva, Tetrahedron: Asymmetry 2007,
18, 2468–2472.
Chem. Int. Ed. 2003, 42, 913–916; Angew. Chem. 2003, 115,
942–945.
[4] a) D.-Y. Ma, D.-X. Wang, J. Pan, Z.-T. Huang, M.-X. Wang,
J. Org. Chem. 2008, 73, 4087–4091; b) D.-H. Leng, D.-X.
Received: November 12, 2014
Published Online: January 27, 2015
1506
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 1500–1506