Angewandte
Chemie
reaction rate was seen upon increasing amounts of gluta-
raldehyde (SI, Table S7).
Surprisingly, the KR activity of the hybrids seemed to be
negligibly affected by the amount of glutaraldehyde
employed in the first step, and all hybrids prepared from
1
7 wt% CALB gave essentially the same conversion and
enantiomeric excess (ee) after 1 h at 708C (SI, Table S8,
[
16]
entries 1, 3, and 4).
These observations suggest that
0
.1 equiv of glutaraldehyde is sufficient for immobilizing the
required amount of enzyme for an efficient KR. Lowering the
amount of CALB to 5.5 wt% on the other hand, resulted in
a significantly slower KR and only 18% of the acylated
product was obtained after 1 h (Table S8, entry 2). Interest-
ingly, the racemization rate did not improve when the enzyme
loading was lowered (SI, Figure S7, Table S6), which clearly
demonstrates that this co-immobilization strategy keeps the
two species sufficiently separated to avoid deactivating
interactions between them. Based on the results obtained
from the racemization and KR, the hybrid prepared from
Figure 1. A) Image of the hybrid catalyst containing 5 nm Au-tagged
CALB by HAADF-STEM, showing two sets of well-dispersed metal
nanoparticles. The larger and brighter set of nanoparticles belongs to
the Au tags that confirm a uniform distribution of the enzyme on the
support. The smaller set with darker contrasts belongs to the original
Pd nanoparticles, which are responsible for the racemization activity.
Scale bar: 0.1 mm. B) HAADF-STEM image of the corresponding hybrid
catalyst without Au nanoparticles tags for clearer a view of the Pd
nanoparticle distribution. The typical size range of the Pd nanoparti-
cles was found to be 1–2 nm. Scale bar: 20 nm.
0
.1 equiv glutaraldehyde, and 17 wt% CALB, hereafter
denoted as hybrid-GA -Ehigh, was used for characterization
0
.1
and DKR experiments.
From elemental analyses (ICP-OES), the loading of
enzyme and Pd on the hybrid-GA -E was determined to
STEM analyses of the corresponding hybrid catalyst in the
absence of the Au nanoparticle tags were also conducted as
a reference, which showed a well-dispersed Pd nanoparticle
distribution in the size-range of 1–2 nm (Figure 1B; SI,
Figure S10). Furthermore, this demonstrates that our
method of introducing glutaraldehyde and enzyme on the
0
.1
high
be 15.6 wt% and 4.80 wt%, respectively. These data confirm
that the glutaraldehyde-based method for immobilization of
CALB on the support was highly efficient, as 92% of the
loaded enzyme was successfully attached. Another piece of
desirable information is the distribution of CALB on the
support. However, direct observation of the enzyme by
transmission electron microscopy (TEM) is not possible.
Therefore, a method was developed that involved tagging of
CALB with gold nanoparticles to allow for the indirect
determination of the enzyme distribution (SI, Figure S1).
To provide a sufficient Au nanoparticle tagging, a solution
of CALB was first treated with tris(2-carboxyethyl)phosphine
0
Pd -AmP-MCF does not lead to any significant clustering or
agglomeration of the Pd nanoparticles present from the start.
The utilization of hybrid-GA -E as a catalyst for the
high
0
.1
DKR of 1-phenylethylamine was commenced by performing
the reaction at 808C, as both the kinetic resolution and the
racemization previously showed high efficiency at this
temperature. However, it was found that these conditions
were not ideal for the DKR, and only a moderate yield of the
desired product 2 was obtained as a result of partial enzyme
deactivation (Table 1, entry 1). The explanation for this
seemingly contradictory result originates from the difference
in reaction time between KR and DKR. In KR the reaction
time is short with negligible deactivation, whereas in DKR the
reaction time is long, leading to significant enzyme deactiva-
tion.
The problem of enzyme deactivation was circumvented by
lowering the temperature to 708C, which furnished 2 in 95%
yield and an excellent enantioselectivity (99% ee) after 16 h
(Table 1, entry 2). The DKR employing the hybrid could be
further improved by the addition of molecular sieves (4 ꢁ),
which afforded the desired product in quantitative yield and
99% ee after 16 h (Table 1, entry 3).
(
TCEP), which reductively cleaves a disulfide bridge that is
present on the enzyme surface into free thiols. Thiols are well-
known to efficiently coordinate to Au nanoparticles, and this
property was exploited when anchoring these nanoparticles to
CALB. TCEP-treated CALB was stirred overnight with
a colloidal solution containing either 2 or 5 nm Au nano-
particles, followed by concentration of the resulting reaction
solution in vacuo. The solution of the 2 nm Au reaction was
passed through a membrane centrifugation tube with a cutoff
of at 10 K prior to concentration. Elemental analysis (ICP-
OES) of the collected enzyme residue was for practical
reasons only possible for the 2 nm reaction, and it showed that
1
4% of the CALB had been successfully tagged with an Au
nanoparticle. This was considered to be sufficient for sub-
sequent TEM experiments. The tagged enzyme was then
To validate the utility of the concept of co-immobilizing
the Pd nanoparticles and the enzyme into the same cavities of
0
immobilized into the glutaraldehyde-functionalized Pd -
0
AmP-MCF as described above and analyzed by high-angle
annular dark-field scanning transmission electron microscopy
the support, a DKR was performed with separate Pd -AmP-
MCF and CALB-MCF (Table 1, entries 4,5). To make these
results fully comparable, the respective loadings of Pd and
enzyme used in the separate component reaction were the
same as those employed in the hybrid-GA -Ehigh. Moreover,
(
HAADF-STEM, Figure 1A; SI, Figures S8,9). To our
delight, a well-dispersed pattern of Au nanoparticles of
either 2 nm or 5 nm could be observed along with the Pd
nanoparticles in both cases. For better visualization of the Pd
nanoparticles with weaker contrast, additional TEM and
0
.1
0
the Pd -AmP-MCF used in this comparison was of identical
standard to that used for the hybrid catalyst preparation, and
Angew. Chem. Int. Ed. 2013, 52, 1 – 6
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!