Communication
especially applicable to the concept, resulting in products of
high enantiopurity in high yields.
1H NMR spectroscopy and chiral HPLC. Unexpectedly, amide
products 2a–c formed under the same reaction conditions
using the two different lipase preparations, Novozyme 435 and
lipase PS-C I, provided the opposite absolute configuration, re-
spectively, in all cases (entries 1–6, Table 1). For the enzyme-
catalyzed reactions using Novozyme 435, amide products (À)-
2a, and (À)-2c were produced in very good to excellent yields
(89–94%) and very good ee values (83–89%). Amide product
(À)-2b was formed at lower yield (30%) under these condi-
tions, but with an excellent enantiopurity (97% ee). Nitrile 3
was, in this case, formed as a by-product, resulting from the re-
action between aminonitrile 1b and (4-methoxyphenyl)metha-
nimine, in turn formed from compound 1b during the enzy-
matic reaction.[37] For the somewhat larger amide product (À)-
2d, the contrary effect was instead recorded, quantitatively
formed from rac-1d but with lower resulting enantiomeric
excess (37%). This is likely due to an impaired racemization
step in this case due to the larger substrate structure.
The lipase-catalyzed racemization and asymmetric amidation
was first applied to the transformation of a-aminonitriles 1a–
d. 1-Amino-2-(4-fluorophenyl) acetonitrile 1c was selected as
a candidate for the initial lipase screening (entries 5–8, Table 1).
Four different lipase preparations were evaluated: Novo-
zyme 435 (immobilized lipase CAL-B from Candida antarctica)
lipase PS from Burkholderia cepacia, lipase PFL from Pseudomo-
nas fluorescens, and immobilized PS-C I from Burkholderia cepa-
cia. Of these, only reactions catalyzed by Novozyme 435 and
lipase PS-C I provided good conversions (95%) to amide prod-
ucts 2c at room temperature (entries 5–6), whereas no prod-
ucts were formed by using the PS and PFL preparations. Rais-
ing the temperature to 408C resulted in only marginally im-
proved conversions in the latter cases (11 and 4%).
Different acyl donors: phenyl acetate, 2,2,2-trifluoroethyl
acetate (TFEA),[36] and ethyl acetate, were also applied to the
enzyme-catalyzed reactions. The enantiomeric excess (ee) of
amide product 2c from the reactions using ethyl acetate was
slightly higher than using the other acyl donors. Moreover,
100 mg of lipase PS-C I and 50 mg of Novozyme 435 were suf-
ficient to catalyze the transformation of aminonitriles 1a–d to
the corresponding amide products 2a–d in similar reaction
times.
Using lipase preparation PS-C I, very similar yields/conver-
sions as for Novozyme 435 were obtained, and amide products
(+)-2a, (+)-2c, and (+)-2d were produced in 89–92% yield,
whereas product (+)-2b was formed at a lower rate (35%
yield). In the latter reaction from compound rac-1b, by-prod-
uct 3 was again formed. However, the enantiomeric excesses
of the products, of opposite configuration compared to the
products formed using Novozyme 435, were lower
After initial optimization of the dual lipase-catalyzed racemi-
zation and asymmetric amidation, the enzymatic reactions of
a-aminonitriles 1a–d were performed in tert-butyl methyl
ether (TBME) at room temperature by using ethyl acetate as
acyl donor. The results of the reactions were followed by
than for the CAL-B-catalyzed reactions (0–37% ee).
This effect is likely due to lower rates of the racemiza-
Table 1. Catalytic activities and stereoselectivities of lipase-catalyzed racemization and
asymmetric amidation of compounds 1a–d.[a]
tion step using PS-C I, compared with higher rates of
the asymmetric amidation step. Attempts to improve
the results by using silica gel as additive in the enzy-
matic reaction using lipase preparation PS-C I did not
result in any enhancement in these cases.[38]
To further evaluate the lipase catalytic activities in
racemization and asymmetric amidation, cyclic ami-
nonitrile structures were subsequently probed. Thus,
1-cyano-1,2,3,4-tetrahydroisoquinolines 4a and b,
representing important intermediates for isoquinoline
alkaloid syntheses,[39] were applied as aminonitrile
substrate candidates in the DKR process. Stereoselec-
tive acylation of this class of compounds, for exam-
ple, accomplished by using metal-based catalysts,[40]
organocatalysts,[41,42] and chiral auxiliaries,[43] results in
optically active isoquinoline Reissert-type products. In
the present case, Novozyme 435, PS-C I, PS, and PFL
were evaluated as racemizing and resolving agents,
and the enzymatic reactions were performed in TBME
at 408C by using phenyl acetate as acyl donor. The
Entry Product Lipase
Loading [mg] Time [days] Conversion [%][b] ee [%][c]
1
2
3[d]
4[e]
5
2a
2b
2c
Novozyme 435 50
PS-C I 100
Novozyme 435 50
PS-C I 100
Novozyme 435 50
9
10
12
12
10
10
4
98 (94)
97 (92)
33 (30)
38 (35)
95 (89)
95 (90)
11
(À) 83
(+) 15
(À) 97
(+) 52
(À) 89
(+) 37
(+) 56
(+) 62
(À) 37
0
6
PS-C I
PS
PFL
Novozyme 435 50
PS-C I 100
100
100
100
7[f]
8[f]
9
4
4
2d
10
10
quant. (93)
95 (89)
10
[a] Reactions carried out with compound 1 (0.05 mmol), ethyl acetate (3 equiv),
TMSCN (0.01 equiv), and lipase in TBME at RT. [b] Followed by chiral HPLC analysis
and 1H NMR spectroscopy. [c] Determined by chiral HPLC analysis on an OJ column;
see the Supporting information. [d] 63% a-Benylideneamino-a-phenylacetonitrile was
formed. [e] 59% a-Benylideneamino-a-phenylacetonitrile was formed. [f] Reactions
performed at 408C.
1
reactions were followed by H NMR spectroscopy and
chiral HPLC (Table 2).
To compare the catalytic activities, the different
enzyme preparations were first applied in equal
Chem. Eur. J. 2014, 20, 11322 – 11325
11323 ꢀ 2014 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim