Table 2 Recycling of the free lipase and nanoparticle-supported
lipase (Iron Oxide–Lipase) for kinetic resolution of racemic 2-bromo-
propionic acid
Reaction
cycle
Conversiona,d
(%)
Conversionb,d
(%)
Conversionc,d
(%)
Fig. 2 Structure of Candida rugosa lipase immobilized on solid-phase
microbeads (Resin–Lipase). Resin–Lipase was used for kinetic resolution
of racemic 2-bromopropionic acid (Table 2).
1
2
3
4
a
29
21
21
20
55
7
n/a
n/a
21
17
18
17
Catalyzed by Iron Oxide–Lipase. Yields were determined after
b
seven reaction days. Catalyzed by the free lipase. Yields were
of lipase protein per mg of nanoparticles was obtained through
this immobilization strategy.
determined after two reaction days. No products were detected in
the third and fourth reaction cycles even after two weeks.
Catalyzed by Resin–Lipase (Fig. 2). Conversion yields were
The enzymatic activity of the supported lipase was determined
by closely monitoring the kinetic resolution reactions of three
racemic carboxylates (Table 1). At fixed time intervals, small
samples were taken out from the reaction mixtures. Iron Oxide–
Lipase was magnetically concentrated and removed. The super-
natants were then subjected to GC analyses for the conversion
progress of carboxylates. The reactions were usually stopped when
about 45% of the carboxylates were transformed into esters. The
carboxylates and ester products were separated and examined by
1H NMR and Mass Spectrometry analyses. Optical rotation
measurements and GC analyses using a chiral GC column were
utilized for determining the absolute configurations and enantio-
meric excess (ee) values of resolution products. Average isolation
yields of 77% and 80% were obtained for n-butyl esters and
carboxylates, respectively (Table 1). The ee values of isolated
carboxylates and esters were usually higher than 99%, comparable
to those reported for the non-immobilized, free lipase.15
Apparently, immobilization of the lipase to nanoparticles did not
lead to significant loss of stereoselectivity of the enzyme.
c
d
determined by GC analyses.
conversion yield of 55% within two reaction days in its first round
reaction. However, the catalytic activity of the free enzyme
decreased rapidly from 55% to 7% in its second reaction cycle. It
completely lost its enzymatic activity in its third and fourth rounds
of reaction. On the other hand, the lipase immobilized on
micrometre-sized beads (Resin–Lipase) demonstrated long-term
stability, but its conversion yields were slightly lower than those of
Iron Oxide–Lipase (Table 2).
The enzymatic activity of the nanoparticle-supported lipase in its
initial round of reaction is lower than that of the free enzyme. This
is probably due to chemical bonding between the proteins and the
nanoparticle carriers. The -NH2 groups that were used for linking
to the glutaraldehyde bridge in Iron Oxide–Lipase might play an
important role in catalytic reactions. The decrease of enzymatic
activity was compensated by the long-term enzyme stability
provided by the nanoparticle supports. The nanoparticle-
supported lipase could be recovered and reused without significant
loss of its activity for 28 days. This is in contrast to the rapid loss of
activity of the free enzyme. Our observations support an early
report from Ulman on the improved long-term stability of a
nanoparticle-supported enzyme.3
The possibility of recycling the nanoparticle-supported lipase for
repeated uses has also been investigated. Racemic 2-bromopro-
pionic acid was chosen as a substrate for four rounds of reactions
(Table 2). In a typical experiment, 2-bromopropionic acid was
mixed with excessive n-butanol in the presence of a catalytic
amount of Iron Oxide–Lipase. After seven days, nanoparticles
were magnetically concentrated, isolated and washed with buffers.
The isolated lipase was then subjected to a new round of the same
reaction. Each time, the reaction was stopped after seven days.
Supernatants were analyzed by GC to determine the conversion
yields of 2-bromopropionic acid in each reaction cycle. Table 2
showed that the conversion percentages dropped from 29% to 20%
after four reaction cycles for Iron Oxide–Lipase. In comparison,
the free lipase demonstrated much higher enzymatic activity with a
In summary, maghemite nanoparticle-supported Candida rugosa
lipase was employed to probe kinetic resolution reactions of
racemic mixtures. Chemical immobilization of the lipase to the
surfaces of nanoparticles led to a decrease in the enzymatic activity,
but this process improved the long-term stability of the enzyme.
The nanoparticle-supported lipase could be recycled for subse-
quent new rounds of reactions, showing no significant drop in its
enzymatic activity in repeated reactions. The second advantage of
using nanoparticle supports is the facile recovery of the
immobilized enzyme. An external small permanent magnet that
is inexpensive and readily available from many commercial sources
can be utilized for magnetic concentration of the enzyme in high
efficiency. More importantly, our work here, for the first time,
demonstrated that the nanoparticle-supported lipase maintained
high stereoselectivity towards chiral resolution reactions. The ee
values of the nanoparticle-supported resolution products were
comparable to those of the free enzyme. This high stereoselectivity
in conjugation with the long-term stability and facile recovery of
magnetic nanocluster-supported enzymes will make economically
viable the use of magnetic nanoparticles for the industrial chiral
resolution processes. Such magnetic nanometre-sized carriers can
also be potentially utilized for supporting other types of large-scale
organic and biological transformations in the pharmaceutical and
Table 1 Resolution of racemic carboxylates using Iron Oxide–Lipase
I
II
Yielda
(%)
Eeb
(%)
Yielda
(%)
Eeb
(%)
Entry
R
X
1
2
3
a
CH3
CH3
(CH2)3CH3
Br
Cl
Br
76
77
80
b
99
99
99
80
78
82
99
99
92
Determined by GC analyses. Ee determined by chiral GC
analyses. Absolute configurations were determined by optical
rotation measurements.
This journal is ß The Royal Society of Chemistry 2005
Chem. Commun., 2005, 4432–4434 | 4433
Gardimalla, Hari M. R.
