Controlling performance of lipase immobilised on bioinspired silica

Article abstract of DOI:10.1039/c2tb00462c

Lipase (Candida antarctica lipase B) was immobilised on silica that was produced via a mild route using a recently developed biologically inspired silica formation method. This route of immobilisation was favourable compared to traditional methods due to

Full text of DOI:10.1039/c2tb00462c

Journal of
Materials Chemistry B
View Article Online
View Journal | View Issue
PAPER
Controlling performance of lipase immobilised on
bioinspired silica†
Cite this: J. Mater. Chem. B, 2013, 1,
1164
*
Claire Forsyth and Siddharth V. Patwardhan
Lipase (Candida antarctica lipase B) was immobilised on silica that was produced via a mild route using a
recently developed biologically inspired silica formation method. This route of immobilisation was
favourable compared to traditional methods due to its simplicity, mild conditions, low cost, one-step
procedure and short preparation time. Lipase was chosen as the enzyme due to its wide
implementation in industry, particularly in the food and pharmaceutical industries. Compared to other
methods using bioinspired silica entrapment of enzymes, we demonstrate several significant
improvements in the performance of the biocatalysts produced. Very high immobilisation efficiencies
(close to 100%) were achieved under optimised conditions. The immobilised enzymes also displayed
high levels of activity, which exceeded those reported in previous studies of lipase that was immobilised
on bioinspired silica. Furthermore, we demonstrate the control over enzyme activity which was achieved
through the entrapment conditions by regulating surface area, the average pore diameter and the
mean particle size of the silica support. The immobilised enzymes also had very good reuse potential,
and showed improved thermal and pH stability. The biocatalysts also performed well in turbulent
conditions, showed good storage properties and were successfully used in bench-top reactors, therefore
strongly supporting their use in commercial applications. The immobilised enzymes also performed
comparably to or better than the industrial benchmark, Novozymꢀ 435. The bioinspired method of
immobilisation used has significant benefits over current methods, and these factors could increase
potential applications in industry.
Received 27th August 2012
Accepted 21st December 2012
DOI: 10.1039/c2tb00462c
www.rsc.org/MaterialsB
potential, and can also lead to instability. Because of this,
expensive and energy intensive separations are oen necessary,
1 Introduction
In recent years, it has become necessary to investigate the
feasibility of more environmentally friendly, or ‘green’ process
routes, that make use of mild processing conditions. This is
largely due to increasingly stringent environmental regulations,
environmental concerns, and drives for greater efficiency.1 These
can greatly reduce operating costs (such as handling and waste
treatment), and by making an efficient use of resources, process
efficiency – and therefore economics – can be improved.2
A key area of green route investigation is in the application of
biocatalysts. Enzymes are examples of biocatalysts, and are
advantageous since they are able to catalyse reactions at near
ambient temperatures, they are highly specic and can there-
fore be used to reduce by-product formation, and in some cases,
enzymes can even make otherwise non-attainable reactions
feasible.2 Enzymes do, however, have several inherent problems
that restricts their use. One such problem is their soluble
nature; this hinders the ease of enzyme recovery and reusability
and product contamination can be problematic. The molecular
structure of enzymes is also prone to destruction through non-
benign conditions, such as high temperatures, extreme pH
conditions and the presence of organic solvents.2,3
To overcome such problems, considerable research is being
carried out in enzyme immobilisation techniques, whereby the
enzymes are effectively supported or conned for the duration
of the reaction.4,5 This allows their catalytic properties to be
preserved, while assisting with repeated and continuous use.
Effective immobilisation has also been shown to result in
greater enzyme stability over more intense process conditions,
therefore increasing potential enzyme applications.3,6 Recently,
additional advantages of enzyme immobilisation have been
reported and include possibility of using multi-enzyme systems,
improved activity and modication of substrate selectivity.4
There are various enzyme immobilisation techniques and these
include covalent attachment, adsorption,3,7 cross-linked
enzyme aggregates (CLEAs), entrapment and encapsulation.2
Many of the current immobilisation approaches suffer from
problems, such as long preparation times, tedious multi-step
procedures, loss of enzyme activity during immobilisation and
poor reusability. Within the entrapment/encapsulation
Department of Chemical and Process Engineering, University of Strathclyde, 75
Montrose Street, Glasgow G1 1XJ, UK. E-mail: Siddharth.Patwardhan@strath.ac.uk;
Tel: +44 (0)141-548-5786
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c2tb00462c
1164 | J. Mater. Chem. B, 2013, 1, 1164–1174
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry B
approaches of immobilisation, the sol–gel method is by far the Further consideration of these biomolecules has been used to
most common approach and this involves immobilising the establish synthetic “additives” which have proven to be effective
enzyme within a silica gel. Although maintaining high levels of for producing silica on a laboratory scale.2,24 Additives can greatly
enzyme activity, a major disadvantage of the sol–gel route is its affect the silica production process and oen act as catalysts,
long preparation time; it can take several hours to days and therefore allowing the process to be carried out rapidly, and
industrially this is impractical (Table S1†).6,8,9 Due to the mate- under mild conditions.9 We have reported previously that a key
rials and techniques involved, many could also not be classed as advantage of using such bioinspired methods is that the prop-
‘green’. These drawbacks restrict their industrial applications.3
The enzyme lipase – inparticular the strain Candida antarctica ne-tuned by carefully chosen synthetic conditions.24
lipase B (CALB, simply referred to as “lipase” herein) – forms the A comparison between bioinspired silica synthesis and
erties of silica (surface area, particle size, porosity, etc.) can be
cornerstone of this study, as it is a technologically important existing methods for enzyme immobilisation reveals that our
enzyme. Lipase is primarily associated with catalysing the method is rapid, easy and environmentally friendly (Table S1†).
hydrolysis of triglycerides (fats) into fatty acids and glycerol.7,10 Previous studies have reported immobilisation of a variety of
CALB is one of the more notable enzymes because of its high enzyme using bioinspired approaches,2 however, the success of
degree of selectivity, and importance in a range of industrial these for lipase immobilisation has been limited, with prob-
applications.10 Due to the ability to produce it on a large scale lems such as low activity being reported. Lipase immobilised on
through micro-organism proliferation, it has been widely bioinspired silica has been achieved by using poly-
implemented in industry, including the food processing indus- ethyleneimine (PEI),25 polyallylamine (PAH)26 or the R5
tries (particularly dairy), pharmaceuticals industries and chem- peptide27 as additives, however, these suffer from a few prob-
icals industries (with signicant use seen in producing synthetic lems discussed in Results and discussion section.
polymers and detergents),7 and more recently, in the renewable
In this work, dilute aqueous solutions of various additives
energy industry in biodiesel production.11,12 Its use on an (ethyleneamines) were used as additives along with sodium
industrial scale is, however, oen limited due to reasons metasilicate as a silica precursor, to immobilise the enzyme
mentioned, such as instability at higher temperatures and pH lipase (CALB) via an entrapment method. These additives were
extremes, and the difficulty to recover it. These drawbacks are chosen due to the rapid silica formation they induce28,29 and their
particularly important for the foods and pharmaceutical indus- low cost. There are no previous reports of using these combina-
tries, where the production of pure, non-contaminated (enzyme- tions for enzyme immobilisation, making the route highly novel.
free) products is a necessity. As a result, nding efficient and Furthermore, varying the chemistry of the additives allowed a
cost-effective means of immobilising lipase, in such a way that its systematic study to be performed and the properties of the bio-
inherent problems may be alleviated or overcome, is of interest. catalyst to be controlled. In this study, we aim to address
There are number of lipase immobilisation methods that following questions. (i) Is it possible to achieve signicant
have been tested. Perhaps the most common form of immobi- improvements in loading and activity of lipase using bioinspired
lised CALB is the commercially available Novozymꢀ 435, which silica entrapment method? (ii) Can we control the performance
is produced by Novozymes. It consists of CALB immobilised of biocatalyst through the synthetic conditions of entrapment?
onto a macroporous acrylic resin. Covalent binding of lipase (iii) Can this method produce stable biocatalysts that would be
onto commercial Eupergitꢀ supports and glyoxyl agarose beads suitable for use in industrially relevant continuous reactors?
have also been reported.3,13–15 Other examples of lipase immo-
bilisation include adsorption onto silica nanoparticles,10,16,17
2 Experimental
Chemical reagents
cross-linking using glutaraldehyde18,19 and sol–gel entrap-
ment.20,21 Many of these methods, however, suffer from tedious
and expensive preparation methods, poor stability, or the use of
toxic or hazardous chemicals.
All chemicals were purchased from Sigma-Aldrich and used as
received. Disodium hydrogen orthophosphate anhydrous
(Na2HPO4), sodium dihydrogen orthophosphate dihydrate
(NaH2PO4$2H2O), sodium metasilicate (Na2O3Si$5H2O),
ethanol (99.5% v/v), hydrochloric acid (1 M), lipase (CALB),
p-nitrophenol, p-nitrophenyl acetate, pentaethylenehexamine
(PEHA), tetraethylenepentamine (TEPA), triethylenetetramine
(TETA), diethylenetriamine (DETA), Novozymꢀ 435. All experi-
ments were performed at 20 ꢂC, unless stated otherwise. Unless
otherwise specied, the experiments were performed in a 0.1 M
pH 7 phosphate buffer. Experiments were carried out in tripli-
cate, and Minitab 16 was used to evaluate errors.
There appears to be considerable scope for improved enzyme
immobilisation techniques through the use of bioinspired silica.
Due to its mild, rapid, low cost and controlled protocol, biolog-
ically inspired silica has recently been demonstrated as a green
method for enzyme immobilisation via in situ entrapment of
enzymes.2 Bioinspired silica routes aim to mimic nature's in vivo
approach to producing silica which is termed ‘biosilicication’.
In nature ꢀ6.7 ꢁ 109 tonnes of silicon are estimated to be pro-
cessed per year by marine biological systems such as diatoms.22,23
The understanding of this biological silica formation has resul-
ted in silica being produced in vitro under mild conditions (near
ꢂ
neutral pH and ꢀ20 C), while still maintaining high levels of
Enzyme immobilisation
control, resulting in high quality silica formation.9 It is known
from studies carried out on silica producing species that bio- Silica was prepared using sodium metasilicate as the silica
logical molecules are involved in controlling biosilicication. precursor, and the amines PEHA, TEPA, TETA or DETA as the
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. B, 2013, 1, 1164–1174 | 1165

View Article Online
Journal of Materials Chemistry B
Paper
additives. Lipase (CALB) was added during silica condensation, product concentration was deduced using a calibration curve
producing immobilised enzyme. Separate solutions of sodium (Fig. S2†). This procedure was repeated for the various
metasilicate and amine were prepared using deionised water, and temperatures and pH conditions involved in the entirety of the
were mixed to obtain a 30 mM concentration of silicic acid, and experiment, since it was found that calibration did not remain
maintain a molar ratio of [Si] : [N] ¼ 1 : 1 in the nal solution. A unchanged for different temperature and pH conditions.
pre-determined volume of 1 M hydrochloric acid was added to this
Since the assay reaction can occur even without the presence
so that the pH was 7 ꢃ 0.10. A known mass of enzyme was added of lipase (although to a lesser extent), it was necessary to
immediately, and the resulting solution was given a gentle mix. determine non-enzymatic reaction rates in order to allow the
Thiswasthenleunstirredfor5minutes,beforebeingcentrifuged true increase in reaction rate, due the presence of lipase, to be
for 15 minutes at 8000 rpm, to allow the immobilised enzyme to be determined. When enzymes were present, the rate calculated
separated. The supernatant was removed and stored (for further was the sum of both the enzymatic and non-enzymatic reaction
testing), and deionised water was added to make up the volume rates. The non-enzymatic rate was then subtracted, giving the
again. This process of centrifugation and washing was repeated true enzymatic rate. A 1.0 M stock solution of p-nitrophenyl
another two times. The immobilised enzyme was then dried at 40 acetate was prepared by dissolving the necessary mass in
ꢂC in a vacuum oven for 5 hours. Control experiments were carried ethanol (99.5% v/v). This was diluted into 5, 10, 20, 40, 80 and
out where silica was prepared without enzyme being added.
100 mM solutions using ethanol. The assay mixture was then
It was necessary to deduce how much lipase had become prepared and consisted of: 3 mL 0.1 M, pH 7 phosphate buffer;
entrapped within the silica. This allowed the immobilisation 15 mL 0.1 M, pH 7 phosphate buffer (when free enzyme was
efficiency to be determined (and compared for the various present, 15 mL of enzyme solution was added, hence buffer of
amines), and was also necessary to ensure that the same this volume was added to maintain concentrations) and 30 mL
concentrations of enzyme were used for comparing the free and p-nitrophenyl acetate (in ethanol) substrate solution. This was
immobilised enzyme systems. This was achieved by measuring prepared for all p-nitrophenyl acetate concentrations, as well as
the protein content in the supernatant produced during enzyme a 0 mM solution as a control. The assays were kept gently stir-
immobilisation, using UV-spectroscopy (Shimadzu UVmini- red, and the absorbance was measured at 405 nm, every minute
1240) at 283 nm with a quartz cuvette. A lipase calibration curve for 10 minutes. The absorbance was then measured at various
(Fig. S1†) was produced to allow the relationship between lipase intervals aer, until there was no further change. This indicated
concentration and absorbance to be deduced.
that the reaction was complete. The p-nitrophenol calibration
curve equation was then used to convert the absorbance into
product concentrations. Linear equations were found to
describe the data for the rst 10 minutes, and these were used
to estimate the non-enzymatic reaction rate for each substrate
concentration. A rst order reaction model was then used to
estimate the rate constants.
The kinetic parameters and activity of free lipase were esti-
mated. This was necessary to allow a comparison between free
and immobilised systems, and hence allowed the impact of
immobilisation on lipase to be evaluated. The procedure was
similar to that used to deduce the non-enzymatic reaction rates,
however, 15 mL of 1 mg mLꢄ1 free lipase in buffer solution
(producing a 0.005 mg mLꢄ1 enzyme concentration overall in
the assay) was added, rather than the additional 15 mL of buffer
added previously.
The corresponding non-enzymatic reaction rates were sub-
tracted from the total reaction rates deduced, giving the
enzymatic reaction rates. The enzymatic reaction rates were
then plotted against substrate concentration, to produce a
Michaelis–Menten model. Through non-linear regression
techniques, the kinetic parameters were deduced; for this, the
Hill function in Origin 8.5 was used to describe the following
equation:
Materials characterisation
Lipase immobilised on silica were dried and mounted on SEM
sample holders with double sided sticky carbon tape. Upon gold
coating by sputtering, samples were analysed on a HITACHI SU-
6600 Field Emission Scanning Electron Microscope (FE-SEM) at
10 kV. For particle size measurements, sizes of a number of
particles from selected high resolution images were measured
manually. Fourier Transform Infra-Red (FTIR) spectroscopy was
used to analyse the dried immobilised enzyme samples, in order
to detect the formation of silica and the presence of lipase.
Attenuated Total Reectance-FTIR analyses (ABB Miracle MB
3000) of dried silica (without lipase) and dried immobilised
enzyme samples were performed, with 32 scans at a resolution of
4. Prior to centrifugation during enzyme immobilisation (i.e. aer
leaving the samples unstirred for 5 minutes), dynamic and static
light scattering (DLS/SLS) techniques were used to estimate
average particle/aggregate sizes. Measurements took place at 90ꢂ
(ALV/CG3 Compact Goniometer, ALV/LSE-5004 Electronic System
and 50 mW He–Ne laser (l ¼ 632 nm)). Immobilised enzyme
samples underwent N2 adsorption (Micromeritics ASAP 2420);
these gave details of surface area, pore volume and pore width.
Vmax½Sꢅ
Enzyme activity assay
v ¼
(1)
KM þ ½Sꢅ
The enzyme assay used involved the hydrolysis of p-nitrophenyl
acetate to p-nitrophenol (and acetate acid). At pH $ 7, the
p-nitrophenol mainly exists as a yellow anion. UV-vis spectros-
copy was therefore used to measure the absorbance of the
samples at 405 nm, and the corresponding p-nitrophenol
Where: v ¼ reaction rate; Vmax ¼ maximum rate; [S] ¼ substrate
concentration; KM ¼ Michaelis–Menten constant. The kinetic
parameter Vmax was used to estimate the specic enzyme activity
of the free enzyme. An enzyme activity unit (U) was dened as
1166 | J. Mater. Chem. B, 2013, 1, 1164–1174
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry B
the amount of enzyme activity that caused the disappearance of p-nitrophenol exists mainly as a yellow anion), only pH 7, 8, 9,
1 mmol of substrate, or formation of 1 mmol of product, per 10 and 11 tests were performed using this method (all at 20 ꢂC).
minute. The specic activity (U mgꢄ1) was dened as the activity
Attempts were also made to evaluate performance in acidic
of the enzyme per mg of enzyme.
conditions (pH 4), by adjusting the pH of the buffer. Since the
The kinetic parameters and activity of the immobilised p-nitrophenol was not in neutral or basic conditions, it remained
enzymes were also determined. This allowed the impact of colourless. Therefore, ata time ofzero and a time of10minutes, a
immobilisation and the effectiveness of the different amines to known volume of the assay solution was taken and added to the
be considered. The performance of Novozymꢀ 435, the industry pH 7 buffer, in order to produce the yellow anion. The absor-
standard immobilised CALB, was also evaluated. The procedure bance was measured and adjusted for the corresponding dilu-
was similar to that used for free enzyme, however, rather than tions; this allowed the product formation aer ten minutes to be
adding free enzyme, an equivalent volume of buffer was added, deduced, allowing a comparison of the systems.
and an appropriate mass of immobilised enzyme was added to
The effectiveness of the immobilised enzymes was tested
ensure a 0.005 mg mLꢄ1 enzyme concentration, as before. under extreme physical conditions, to simulate the turbulent
According to Novozyme, Novozymꢀ 435 is 21% w/w enzyme.30 ow conditions which may be encountered in an industrial
Therefore, since this range was similar to that achieved with the situation, rather than the benign, gently stirred assay conditions.
immobilised lipase produced herein, a fair comparison between It also allowed an insight into the mass transfer limitations of the
lipase supported on silica and that in Novozymꢀ 435 was systems. The procedure was similar to that used for the
possible.
temperature tests, however, all tests took place at 20 ꢂC, during
which the samples were subjected to shaking, vorticing or the
gentle stirring. Product formation aer at a short xed time was
measured, which produced initial rates of reaction. These rates
were compared for the same sample but when measured without
any mixing and reported as “relative activity”. The procedure was
performed without enzyme, with free enzyme, with the bio-
inspired immobilised enzymes and with Novozymꢀ 435. The
storage stability of the enzyme systems was also investigated.
Aer vacuum drying, the immobilised enzymes were stored at
ꢄ18 ꢂC, 4 ꢂC or 20 ꢂC, and their performance aer 5 weeks under
these conditions was evaluated.
Leaching and reuse
A known mass of immobilised enzyme (whose lipase content
had been determined) was placed in 5 mL of 0.1 M pH 7
phosphate buffer. The absorbance of the surrounding buffer at
283 nm (using a quartz cuvette) was measured at various time
intervals. The lipase calibration curve was then used to obtain
an enzyme concentration from this, and the loss of activity was
deduced. This procedure was carried out for lipase immobilised
using bioinspired silica, as well as Novozymꢀ 435.
Immobilised enzyme that had been used in a prior reaction
(and was present in the reaction medium for 1 hour) was
separated from the surrounding liquid by centrifugation, and
washed. The material was then reused in an assay, as described
for the immobilised enzyme. The product formation aer 10
minutes was measured, and compared to its initial use. This
procedure was repeated for additional cycles, and was carried
out for lipase immobilised using bioinspired silica, as well as
Novozymꢀ 435.
Bench-top reactors
The performance of supported lipase was evaluated in two
simple bench-top reactors – continuous stirred tank reactor
(CSTR) and plug ow reactor with a xed catalyst bed (PFR). As
given in Section 3.8, the reactors were designed and dried
powder of supported enzyme catalyst was loaded. Special care
was taken in the PFR by loading the immobilised enzyme into
the tubular reactor and keeping it in place with glass-wool and
lter paper to prevent any enzyme loss. The reactors were
operated for several hours while monitoring product concen-
tration and a steady state was attained.
Enzyme stability
The performance of the immobilised enzymes (including
Novozymꢀ 435) was tested under various temperatures to study
thermal stability. For all stability tests, it was necessary to
perform the assay without enzyme, with free enzyme, and with
immobilised enzyme, at each temperature chosen. The assay
compositions used were identical to those described previously,
with the p-nitrophenyl acetate substrate concentration at
10 mM. Thermal stability experiments were carried out between
20 ꢂC and 70 ꢂC in 10ꢂ increments; these temperatures were
established through the use of a water bath. The product
formation aer 10 minutes was measured, which allowed a
comparison of activity of the systems at various temperatures.
The procedure used for pH stability was similar to that used
in the temperature tests, however, rather than varying the
temperature, the pH of the assay was varied by altering the pH
3 Results and discussion
In order to address the questions given above, a complete
analysis of the immobilised enzymes produced was necessary. A
colorimetric enzyme assay scheme was used to assess the
activity of the immobilised enzymes compared to free enzyme;
the kinetic parameters were determined through non-linear
regression techniques. A comparison of the performance of the
immobilised enzymes and Novozymꢀ 435 (N-435), the industry
benchmark CALB, was also performed.
Improved loading and activity of immobilised lipase
of a 0.1 M phosphate buffer. Since the direct assay procedure An optimisation procedure took place by altering the mass of
only works well at pH $ 7 (it is only at these conditions that the enzyme added during immobilisation to see if a maximum
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. B, 2013, 1, 1164–1174 | 1167

View Article Online
Journal of Materials Chemistry B
Paper
immobilisation efficiency and w/w enzyme was reached: an
The enzymatic activities of various biocatalysts produced
example for PEHA is shown in Fig. 1a. There appeared to be a herein, free enzyme and Novozymꢀ-435 were investigated using
maximum enzyme capacity in the silica, with silica produced protocol detailed in the Experimental section. In each assay, the
using TETA having the greatest capacity, followed by DETA, total amount of enzyme used was kept constant irrespective
PEHA, and TEPA. The use of TETA and DETA were found to of their immobilisation method. The experimental data
produce the highest mass of immobilised enzyme, followed by conformed well to the Michaelis–Menten model (Fig. 3a). The
PEHA and TEPA (Fig. 1b).
kinetic parameters determined (Table 1 and Fig. 3b) showed
When low masses of enzyme were added, protein in the that Vmax was smaller and KM was larger for the bioinspired
supernatant was undeterminable, suggesting very high levels of immobilised systems, compared to free enzyme. Smaller Vmax
immobilisation. The p-nitrophenyl acetate assay was used to for silica immobilised enzymes, when compared with Novo-
conrm this and it was found that immobilisation efficiencies zymꢀ-435, suggests that silica bound enzyme is less active.
were close to 100%. The immobilisation efficiencies and w/w Since KM is effectively a dissociation constant, the larger value
enzyme of the samples compared well with many bioinspired suggested that the substrate had less of an affinity for the
enzyme immobilisation routes that used similar tech- enzyme in the immobilised systems and that a higher concen-
niques,2,25–27,31–33 and exceeded many too.26,27,34–39 Previously tration of the substrate is required to saturate the enzyme.
reported immobilisation efficiencies using lipase and bio- These trends have been observed in many studies of enzyme
inspired silica approaches included 12%,27 90%25 and 96%,26 immobilisation techniques.33,36,39,41 The immobilised enzymes
therefore suggesting that the use of PEHA, TEPA, TETA and produced showed very good potential. Almost 90% of free
DETA as additives was at least as efficient as these, or perhaps enzyme activity was preserved during immobilisation using
better.
DETA, and this was better than that achieved through alterna-
The morphologies and particle sizes of lipase–silica bio- tive bioinspired lipase immobilisation approaches, where
catalysts were investigated using Scanning Electron Microscopy 51%,26 10%25 and 14%27 of activity being maintained have been
(SEM) and the data is shown in Fig. 2a–i. From the SEM data it is reported. Even PEHA, the poorest performer herein, maintained
evident that the lipase–silica samples formed spherical or high levels of activity, exceeding previous approaches. The
pseudo-spherical nanoparticles, which were found to aggregate. results suggested that the amine used during immobilisation
Particle size analysis revealed that the particle were roughly ultimately affected the performance of the immobilised
around 250 nm in size with broad size distributions (Fig. 2i). enzyme, particularly in terms of reaction rates and activity. On
SEM results along with the gas adsorption and light scattering average, enzyme immobilised using DETA had the highest Vmax
,
data (discussed below) suggest that the lipase–silica particles followed by TETA, TEPA and PEHA. KM for PEHA and TETA were
exhibited features at three length-scales. The aggregation of the same, while those for DETA and TEPA were higher (Table 1
primary particles (not shown) allowed the formation of pores and Fig. 3). Novozymꢀ 435 compared very well to the free
(<20 nm, see discussion below for Fig. 5), the formation of enzyme. Although slightly lower on average, Vmax for Novozymꢀ
secondary particles (ꢀ250 nm, SEM data) and larger aggregates 435 and the free enzyme were very similar, and KM was almost
of these particles (1–2 mm, see discussion below for Fig. 5). identical. Novozymꢀ 435 also performed better than the lipase
Although the exact location of enzyme in silica is not clear at immobilised via bioinspired silica.
this point in time, it is reasonable to assume that the enzyme
It is known that immobilisation of enzymes can cause
may reside in the pores, as demonstrated for a related system in several effects leading to reduction in the enzyme activity. These
the literature.40 FTIR spectra (Fig. 2j) conrmed successful effects include enzyme distortion, blockage of enzyme active
immobilisation of the enzyme; this was evident from amide centre, diffusion limitations and hyperactivation.5 Since the
peaks at 1650 cmꢄ1 and 1450 cmꢄ1 (marked as dashed lines in activities of the biocatalysts reported herein are comparable
the Fig. 2j) that were not detected with control silica. A with the free enzyme activity, hyperactivation can be ruled out.
comparison with free lipase also conrmed successful enzyme The ndings suggested that immobilisation may have restricted
immobilisation. Strong bands in the region of 950–1150 cmꢄ1 the enzyme's overall activity. Active sites may have become
and 800 cmꢄ1 conrmed silica formation.
impaired during immobilisation, or the silica support may have
made the sites more difficult to access, possibly due to the
porous nature of the silica support, therefore reducing mass
transfer, and ultimately the reaction rate and activity of the
enzyme. Due to the wide pore sizes in silica (discussed below),
compared to the size of enzyme, it is also likely that enzyme
agglomerates were formed within these pores, attributing to
reduced enzyme activity.
In order to probe whether the reduced activity upon immo-
bilisation was caused by enzyme deactivation or diffusion
limitation, the enzyme activity was measured by inducing more
turbulent conditions. Although, extensive calculations and
analyses on diffusion can be performed by incorporating
internal and external diffusion models, the focus here was to
Fig. 1 Enzyme immobilisation efficiency and amount of enzyme immobilised
using PEHA (a) and all additives (b). In (b) , shows w/w enzyme (left axis) and
shows immobilisation efficiency (right axis).
1168 | J. Mater. Chem. B, 2013, 1, 1164–1174
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry B
Fig. 2 SEM images for lipase immobilised in silica using (a and b) DETA, (c and d) TETA, (e and f) TEPA and (g and h) PEHA. Scale bars are 1 mm for images on left and
5 mm for images on right. (i) Particle size measurements from SEM images. (j) ATR-FTIR spectra for free lipase, silica without lipase and lipase immobilised on silica.
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. B, 2013, 1, 1164–1174 | 1169

View Article Online
Journal of Materials Chemistry B
Paper
the immobilised enzymes, and the number of amines in a
molecular unit. DETA, which resulted in the best performance,
has fewest amines (3), while PEHA, which resulted in the
poorest performance, has most (6).
From porosity measurements, although surface areas and
pore volumes remained unchanged, the average pore diameter
was found to correlate with the enzyme activity (Fig. 5 and Table
S3†). All samples showed mesoporous structures, and TETA
immobilised enzyme was found to have the largest average pore
diameter (22 nm), while PEHA immobilised enzyme, the poor-
est performer, was found to have the smallest (13 nm). Smaller
pores were likely to have reduced mass transfer, resulting in the
lower rates and poorer performance which were observed. The
DLS/SLS results (Fig. 5, Table S2 and Fig. S3†) also suggested a
correlation between the number of amines and performance. In
the PEHA immobilised enzyme system, aggregates were shown
to have the largest mean radii (2400 nm) aer 5 minutes of
reaction and aggregation, while DETA immobilised enzyme
aggregates were the smallest (1200 nm). The larger particle sizes
suggested higher levels of aggregation, since it is known that for
these additives, aer 5 minutes of reaction, micron sized
aggregates are formed that are composed of 200–400 nm silica
particles.28,29 This may have inhibited mass transfer; larger
particle sizes would have made it more difficult for reactants to
penetrate further into the particle.42
Fig. 3 (a) Michaelis–Menten model of enzymatic rate data for immobilised
enzymes (including Novozymꢀ 435) pH 7, 20 ꢂC. (b) Enzymatic activity after
immobilisation compared with that of the free enzyme.
simply check if the immobilised enzyme was still active (i.e.
whether lower conversions were diffusion limited). It was
observed that signicant increases in reaction rate in the initial
stages of the reactions were seen for all immobilised enzyme
systems (Fig. 4). This was likely to have been due to the more
turbulent conditions increasing rates of mass transfer, since the
mass transfer coefficient is a function of the Reynolds number.
For example, enzyme immobilised using TETA had 25 times
faster initial reaction rates when shaken, compared to free
enzyme being just 3 times faster and Novozymꢀ 435 being 11
times faster. Although Novozymꢀ 435 showed increased rates of
product formation upon mixing, the synthesised immobilised
enzymes showed notably enhanced activities compared to
Novozymꢀ 435. Free enzyme, on the other hand, did not show
notable increase in activity under these harsher conditions.
This suggested that mass transfer during the assays was
limiting the immobilised enzymes' performance. It was also
possible that in turbulent conditions, shearing of the free
enzyme led to conformational deformation, resulting in deac-
tivation, while due to large particles of Novozymꢀ 435, mixing
did not inuence reaction rates. This may explain why the
activities of Novozymꢀ 435 and free enzyme were not as greatly
enhanced as the synthesised immobilised enzymes: the silica
supports may have been more effective at protecting the enzyme
from these conditions than Novozymꢀ 435, while free enzyme
had no protection at all.7
The effect of the additive chemistry on the improved enzyme
activity can be explained through the relation between mass
transfer and pore size of the support. Theory for liquid diffu-
sion, particularly in pores, is not as advanced as gaseous
Controlling enzyme performance through entrapment
conditions
Based on the structure of the amines, the data suggested that
there may have been some correlation between performance of
Fig. 4 Relative enzymatic activity compared to non-shaken equivalent. The
activity reported here is a measure of initial rate of the product formation.
Table 1 Comparison of enzyme performance (pH 7, 20 ꢂC)
Specic activity
[U mgꢄ1
% Activity
cf. free enzyme
Method
Vmax [mM minꢄ1
]
KM [mM]
]
PEHA
TEPA
TETA
DETA
Novozymꢀ 435
Free enzyme
5.4 ꢁ 10ꢄ4 ꢃ 5 ꢁ 10ꢄ5
6.9 ꢁ 10ꢄ4 ꢃ 1 ꢁ 10ꢄ4
7.7 ꢁ 10ꢄ4 ꢃ 1 ꢁ 10ꢄ4
8.3 ꢁ 10ꢄ4 ꢃ 1 ꢁ 10ꢄ4
9.0 ꢁ 10ꢄ4 ꢃ 1 ꢁ 10ꢄ4
9.5 ꢁ 10ꢄ4 ꢃ 5 ꢁ 10ꢄ5
0.57 ꢃ 0.1
0.77 ꢃ 0.2
0.57 ꢃ 0.1
0.61 ꢃ 0.1
0.32 ꢃ 0.02
0.33 ꢃ 0.02
0.11 ꢃ 0.01
0.14 ꢃ 0.02
0.16 ꢃ 0.02
0.17 ꢃ 0.01
0.18 ꢃ 0.01
0.19 ꢃ 0.02
57 ꢃ 7
73 ꢃ 12
81 ꢃ 10
87 ꢃ 5
95 ꢃ 3
100
1170 | J. Mater. Chem. B, 2013, 1, 1164–1174
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry B
the pore sizes of the silica from 5 nm to 24 nm. This highlighted
the effect of poor mass transfer on enzyme activity, and the
inuence of pore diameter.16
Improving biocatalyst stability and its use in continuous
reactors
For the free and immobilised enzyme systems, a clear reaction
optimum at 40 ꢂC was observed which compared well with
literature, where optimum conditions between 40 and 50 ꢂC
were generally reported.46–49 Importantly, at 60 ꢂC, the immo-
bilised enzymes (PEHA, TEPA, TETA, DETA and Novozymꢀ 435)
were still having an enzymatic effect, whereas the free enzyme
had become denatured (Fig. 6 and S4†). This strongly suggested
that immobilisation successfully improved the stability of the
enzyme, with the synthesised immobilised enzymes' stability
comparing very well to Novozymꢀ 435.
The pH for optimum activity appeared to be pH 8, which was
consistent with literature where pH values of 6–8 were typically
stated as optimums.48 As with the temperature tests, an increase
in stability of the immobilised enzymes was observed, and the
immobilised enzymes prepared using PEHA, TEPA, TETA and
DETA appeared to show better stability than Novozymꢀ 435
(Fig. 7 and S5†). At pH 10, the immobilised enzymes prepared
were still having an enzymatic effect, whereas the free enzyme
and Novozymꢀ 435 were no longer effective. The test in acidic
conditions also suggested an increased stability of the immo-
bilised enzymes; at pH 4, the immobilised enzymes were shown
to have more of an enzymatic effect than the free enzyme and
Novozymꢀ 435 (Fig. S5†). These results suggest improved
thermal and pH stability, which are consistent with advantages
identied from enzyme immobilisation.4
Fig. 5 Effect of number of nitrogen atoms per amine molecule on activity, pore
diameter (dp) and aggregate size in solution (Ragg).
diffusion. As a result, evaluation of porous liquid systems typi-
cally relies on approximations and empirical equations. A pore
can be approximated as a straight, cylindrical pipe. Darcy's law
(eqn (2)) and the Hagen–Poiseuille equation (eqn (3)) can be
used to show the relationship between diameter and pressure
drop. This is therefore analogous to the relationship between
the pore diameter and the ease of mass transfer through the
pore.43,44
l
DP ¼ 2 f rv2
(2)
d
32mlv
d2
DP ¼
(3)
where: DP ¼ pressure drop; f ¼ fanning friction factor; r ¼
density; v ¼ velocity; m ¼ viscosity; l ¼ pore length; d ¼ pore
diameter.
The improved thermal and pH stabilities upon entrapment
of CALB in silica, which is well documented, in particular for
silica supports,50 is attributed to the steric constraints the
support imposes on the enzyme structure, making denaturation
more difficult. Improved stability is also indicative of strong and
favourable enzyme–support interactions, which could arise
from the in situ enzyme entrapment that was only possible due
to the one-step green approach. In the case of pH stability, the
Both of these equations clearly show that bigger diameters
result in less resistance to ow. This suggests better mass
transfer with larger pore diameters, as suggested by the exper-
imental results. This idea is further reinforced with the
following adaption of Fick's law for liquid owing through
pores:45
ꢀ ꢁ
3
Ni ¼ Di
ðDCÞK
(4)
s
ꢀ
ꢁ
4
dm
dp
K ¼ 1 ꢄ
(5)
where: Ni ¼ ux of component i; Di ¼ diffusivity of solute i; 3 ¼
porosity of material; s ¼ tortuosity of material; DC ¼ concen-
tration driving force; K ¼ geometry factor to account for effect of
pore diameter; dm ¼ diffusing molecule diameter; dp ¼ pore
diameter.
The above interpretation strongly supports the experimental
results (Fig. 5) that as the pore diameter of the support is
increased, the geometry factor is larger, resulting in a higher
ux, hence better mass transfer (i.e. enzymatic activity). The
results presented herein and their interpretation is corrobo-
rated with a previous study immobilising lipase on SBA-15 silica
nanoparticles where it was shown that signicant improve-
ments in activity (three times more) were observed by increasing
Fig. 6 Thermal stability of biocatalysts reported as activity relative to that
at 20 ꢂC.
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. B, 2013, 1, 1164–1174 | 1171

View Article Online
Journal of Materials Chemistry B
Paper
immobilised enzyme being lost during centrifugation and
transfer. The leaching and reuse results suggest that there may
be a strong and benecial interaction between the enzyme and
silica. Furthermore, enzyme performance upon storage was
investigated. Storage at ꢄ18 ꢂC appeared to have the most
detrimental effect on the immobilised enzymes. Storage at 4 ꢂC
appeared most effective, and loss of activity due to storage at
ambient did not appear to be a signicant problem (Fig. 9).
Due to their improved stability and reuse potential, the
effectiveness of the immobilised enzymes on a larger scale was
tested by assembling a bench-top continuous stirred tank
reactor (CSTR) and a bench-top tubular plug ow reactor (PFR).
These reactor types were chosen due to their simplicity to
construct and operate, their compatibility with liquids and
suspended particles, and their popularity in industry.7 Material
balances were carried out across the systems to produce design
equations that allowed the volume of the reactors to be deter-
mined based on known parameters for CSTR (eqn (6)) and PFR
(eqn (7)).
Fig. 7 Relative activity at various pH conditions reported as activity compared to
that at pH 7.
interactions of CALB with the silica (which is known to vary
surface chemistry with changes in pH)51 as well as with amines
incorporated within silica may well be providing favourable
micro-environments for CALB and minimising detrimental
effects of high and low pH conditions.
F½Sinꢅ ꢄ F½Sꢅ
VCSTR
¼
(6)
rS
Sout
ð
d½Sꢅ
ꢄrS
VPFR ¼ F
(7)
The immobilised enzymes also retained very high residual
activity (>90%) aer leaching experiments, which highlights
stability of the enzyme over time as well as demonstrates
negligible levels of leaching (Fig. 8a). These results suggested
that the silica supports had strong entrapment of the enzyme,
while the small loss of activity is attributed to experimental
errors and loss of sample during recovery by centrifugation aer
the leaching experiments but prior to activity measurements.
Novozymꢀ 435 displayed no apparent leaching. On average,
DETA showed most loss in activity, still below 10%. This may be
for similar reasons to those mentioned previously such as pore
diameter; PEHA, with the smallest average pore diameter, dis-
played the lowest levels of loss of activity. When the immobi-
lised enzymes were recovered from the reaction mixtures, they
exhibited excellent reuse potential: this was comparable to
Novozymꢀ 435 (Fig. 8b). They showed only a very slight
reduction in activity from their primary use, with ꢀ5% reduc-
tion in activity being observed between any consecutive cycles.
The small drop in activity from both the leaching and reuse
experiments was likely to be due to silica nanoparticles with
Sin
where: F ¼ volumetric ow rate of inlet and outlet stream; Sin
¼
inlet substrate concentration; S ¼ reactor substrate concentra-
tion (N.B. this varies along the reactor length in a tubular
reactor); Sout ¼ outlet substrate concentration; P ¼ product
concentration; rs ¼ rate of substrate consumption. Despite
several assumptions made in eqn (6) and (7), and their imple-
mentation, these equations served as good rst
approximations.
The rate of substrate consumption was the sum of the non-
enzymatic and enzymatic reaction rates. Therefore, the
Michaelis–Menten rate equation was substituted in for the
enzymatic reaction rate using the kinetic parameters deduced
for the immobilised systems, and a simple rst order rate law
was used for the non-enzymatic rate, using the rate constant
Fig. 8 (a) Residual activity of immobilised enzymes after leaching experiments
and (b) reuse potential after up to five uses.
Fig. 9 Percentage of activity remaining after 5 weeks storage at various
conditions.
1172 | J. Mater. Chem. B, 2013, 1, 1164–1174
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry B
revealed that this loss was due to mass transfer limitations, and
not enzyme deactivation. The activities of all immobilised
enzymes produced exceeded prior studies of lipase immobilised
using bioinspired routes. The synthetic conditions were found
to control physical properties of the biocatalysts, which in turn
regulated the biocatalyst performance.
The immobilised enzymes showed excellent reuse potential
(comparable to Novozymꢀ 435), maintaining high levels of
activity aer repeated use. They also showed very low levels of
leaching, making them particularly suitable for the food and
pharmaceuticals industries. Importantly, the immobilised
enzymes were shown to perform better over a wider range of
temperatures and pH conditions than free enzyme; this was
indicative of improved enzyme stability. The immobilised
enzymes also showed better stability than Novozymꢀ 435,
particularly over the pH range studied. The immobilised
enzymes showed good storage potential, performed well in
turbulent conditions and were shown to perform well in bench-
top reactors, therefore taking a step forward in their commercial
utilisation.
The bioinspired method of immobilisation used has signif-
icant benets over current methods, largely due to its mild,
rapid, low-cost and simple preparation. These factors could
therefore increase potential applications in industry. Future
work, utilising the results presented herein, is directed towards
exploring other benets from the immobilisation method by
developing lipase based biocatalysts for applications in enan-
tioselective reactions, for example, and also demonstrating the
versatility of this method for other industrially relevant enzy-
matic processes.52
Fig. 10 Schematic representation of the CSTR (top) and the tubular reactor
set-up.
deduced for the non-enzymatic reactions as given in eqn (8) and
(9) below.
F½Sinꢅ ꢄ F½Sꢅ
VCSTR
¼
(8)
Vmax½Sꢅ
þ k½Sꢅ
KM þ ½Sꢅ
Sout
ð
d½Sꢅ
Vmax½Sꢅ
ꢀ
ꢁ
VPFR ¼ F
(9)
ꢄ
þ k½Sꢅ
Sin
KM þ ½Sꢅ
Suitable feed ow rates, substrate concentrations and
conversions (X) were chosen to ensure practical designs, and
this allowed the required reactor volumes to be estimated
(Fig. 10). The enzyme concentrations used were kept identical to
those used in the free and immobilised enzyme experiments. To
prevent enzyme being lost in the product, a lter was placed at
the entrance of the outlet pipes. This method of minimising
enzyme loss is common in industry.7
Aer leaving the systems to reach a steady-state, product
formation was measured, allowing the conversions achieved to
be calculated, and compared to the theoretical values. All
conversions achieved experimentally compared very well to the
theoretical values (within 15%). The reactors were operated for
several hours and no loss of enzyme activity was observed aer
the steady state was reached (Fig. S6†). The immobilised
enzymes produced therefore have the potential to be effective
on a larger scale.
Acknowledgements
We thank funding from Royal Society (Grant no. TG090299) for
the initiation of this project, the Department of Chemical and
Process Engineering and Carnegie Undergraduate Vacation
Scholarship. The authors also thank Professor Stephen Clarson
and Professor Peter Halling for various helpful discussions and
Carol Forsyth for help with the light scattering analysis.
References
1 P. T. Anastas and J. B. Zimmerman, Environ. Sci. Technol.,
2003, 37, 94A–101A.
2 L. Betancor and H. R. Luckari, Trends Biotechnol., 2008, 26,
566–572.
3 C. Mateo, J. M. Palomo, G. Fernandez-Lorente, J. M. Guisan
and R. Fernandez-Lafuente, Enzyme Microb. Technol., 2007,
40, 1451–1463.
4 D. Brady and J. Jordaan, Biotechnol. Lett., 2009, 31, 1639–
1650.
5 C. Garcia-Galan, A. Berenguer-Murcia, R. Fernandez-
Lafuente and R. C. Rodrigues, Adv. Synth. Catal., 2011, 353,
2885–2904.
6 U. Hanefeld, L. Gardossi and E. Magner, Chem. Soc. Rev.,
2009, 38, 453–468.
4 Conclusion
Lipase immobilised on silica that was produced via a bio-
inspired route clearly has signicant potential, with main
advantages being its mild processing conditions, the simplicity
at which it can be produced (one-step procedure), its low cost
and its short preparation time compared with current methods,
in addition to high activity and stability. Very high immobili-
sation efficiencies (close to 100%) were achieved through opti-
misation; these compared well with, or exceeded, those
achieved with many bioinspired enzyme immobilisation routes
which used similar techniques. Some loss of activity due to
immobilisation was observed as expected, however, it was
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. B, 2013, 1, 1164–1174 | 1173

View Article Online
Journal of Materials Chemistry B
Paper
7 V. M. Balcao, A. L. Paiva and F. X. Malcata, Enzyme Microb. 29 D. J. Belton, S. V. Patwardhan and C. C. Perry, J. Mater.
Technol., 1996, 18, 392–416. Chem., 2005, 15, 4629–4638.
8 J. Brinker and G. Scherer, Sol–Gel Science: The Physics and 30 Novozymes, private communication.
Chemistry of Sol–Gel Processing, Academic Press, 1990.
9 S. V. Patwardhan, Chem. Commun., 2011, 47, 7567–7582.
31 H. R. Luckari, J. C. Spain, R. R. Naik and M. O. Stone, Nat.
Biotechnol., 2004, 22, 211–213.
10 J. C. Cruz, P. H. Pfromm and M. E. Rezac, Process Biochem., 32 L. Li, Z. Jiang, H. Wu, Y. Feng and J. Li, Mater. Sci. Eng., C,
2009, 44, 62–69. 2009, 29, 2029–2035.
11 K. Ban, M. Kaieda, T. Matsumoto, A. Kondo and H. Fukuda, 33 S. A. Miller, E. D. Hong and D. Wright, Macromol. Biosci.,
Biochem. Eng. J., 2001, 8, 39–43. 2006, 6, 839–845.
12 U. Schoerken and P. Kempers, Eur. J. Lipid Sci. Technol., 34 J. K. Lai, T. H. Chuang, J. S. Jan and S. S. S. Wang, Colloids
2009, 111, 627–645. Surf., B, 2010, 80, 51–58.
13 E. Katchalski-Katzir and D. M. Kraemer, J. Mol. Catal. B: 35 L. Betancor, H. R. Luckari, J. H. Seo, O. Brand and
Enzym., 2000, 10, 157–176. J. C. Spain, Biotechnol. Bioeng., 2008, 99, 261–267.
14 O. Barbosa, C. Ortiz, R. Torres and R. Fernandez-Lafuente, 36 I. C. Kuan, J.-C. Wu, S.-L. Lee, C.-W. Tsai, C.-A. Chuang and
J. Mol. Catal. B: Enzym., 2011, 71, 124–132. C.-Y. Yu, Biochem. Eng. J., 2010, 49, 408–413.
15 D. S. Rodrigues, A. A. Mendes, W. S. Adriano, 37 M. Ramanathan, H. R. Luckari, A. Sarsenova, J. R. Wild,
L. R. B. Goncalves and R. L. C. Giordano, J. Mol. Catal. B:
Enzym., 2008, 51, 100–109.
E. K. Ramanculov, E. V. Olsen and A. L. Simonian, Colloids
Surf., B, 2009, 73, 58–64.
16 H. Jaladi, A. Katiyar, S. W. Thiel, V. V. Guliants and 38 S. V. Patwardhan and C. C. Perry, Silicon, 2010, 2, 33–39.
N. G. Pinto, Chem. Eng. Sci., 2009, 64, 1474–1479.
17 S. Zheng, Effect of Pore Curvature and Surface Chemistry of
Model Silica Hosts on Biocatalytic Activity of Immobilised
39 J. S. Edwards, A. Kumbhar, A. Roberts, A. C. Hemmert,
C. C. Edwards, P. M. Potter and M. R. Redinbo, Biotechnol.
Prog., 2011, 27, 863–869.
Lipase, Department of Chemical Engineering, University of 40 M. B. Cardoso, H. R. Luckari, V. S. Urban, H. O'Neill and
Cincinnati, 2005.
G. R. Johnson, Adv. Funct. Mater., 2010, 20, 3031–3038.
41 F. Neville, M. J. F. Broderick, T. Gibson and P. A. Millner,
Langmuir, 2011, 27, 279–285.
42 S. Fogler, Elements of Chemical Reaction Engineering, Pearson,
2004.
18 R. A. Sheldon, Biochem. Soc. Trans., 2007, 35, 1583–1587.
19 B. Devi, Z. Guo and X. B. Xu, J. Am. Oil Chem. Soc., 2009, 86,
637–642.
20 M. T. Reetz, A. Zonta and J. Simpelkamp, Biotechnol. Bioeng.,
1996, 49, 527–534.
43 E. M. Renkin, J. Gen. Physiol., 1954, 38, 225–243.
21 A. Tomin, D. Weiser, G. Hellner, Z. Bata, L. Corici, F. Peter, 44 U. Heinbuch and J. Fischer, Phys. Rev. A, 1989, 40, 1144–
B. Koczka and L. Poppe, Process Biochem., 2011, 46,
52–58.
1146.
45 M. Peters, T. Klaus and R. West, Plant Design and Economics
for Chemical Engineers, McGraw Hill, 2002.
46 J. Pfeffer, S. Richter, J. Nieveler, C.-E. Hansen, R. B. Rhlid,
R. D. Schmid and M. Rusnak, Appl. Microbiol. Biotechnol.,
2006, 72, 931–938.
22 P. Treguer, D. M. Nelson, A. J. Vanbennekom, D. J. Demaster,
A. Leynaert and B. Queguiner, Science, 1995, 268, 375–379.
23 R. Tacke, Angew. Chem., Int. Ed., 1999, 38, 3015–3018.
24 S. V. Patwardhan, S. J. Clarson and C. C. Perry, Chem.
Commun., 2005, 1113–1121.
47 A. P. A. Sujith, T. V. Hymavathi and P. Y. Devi, Int. J. Biol. Life
Sci., 2010, 6, 67–76.
25 J. C. McAuliffe, W. C. Smith, R. Bond, J. Zimmerman,
D. E. Ward, K. Sanford and T. H. Lane, Abstr. Pap. Am. 48 BRENDA Comprehesive Enzyme Information System.
Chem. Soc., 2005, 229, U1159.
26 G.-C. Chen, I. C. Kuan, J.-R. Hong, B.-H. Tsai, S.-L. Lee and
C.-Y. Yu, Biotechnol. Lett., 2011, 33, 525–529.
49 A. Proskova, Z. ZKopicova, J. Kucera and L. Skarkova, Res.
Agric. Eng., 2010, 56, 122–125.
50 D. Avnir, S. Braun, O. Lev and M. Ottolenghi, Chem. Mater.,
1994, 6, 1605–1614.
27 S. Emond, D. Guieysse, S. Lechevallier, J. Dexpert-Ghys,
P. Monsan and M. Remaud-Simeon, Chem. Commun., 2012, 51 S. V. Patwardhan, F. S. Emami, R. J. Berry, S. E. Jones,
48, 1314–1316.
R. R. Naik, O. Deschaume, H. Heinz and C. C. Perry, J. Am.
Chem. Soc., 2012, 134, 6244–6256.
28 D. J. Belton, S. V. Patwardhan, V. V. Annenkov,
E. N. Danilovtseva and C. C. Perry, Proc. Natl. Acad. Sci. 52 C. Forsyth, T. W. S. Yip and S. V. Patwardhan, Chem.
U. S. A., 2008, 105, 5963–5968.
Commun., 2013, DOI: 10.1039/c2cc38225c.
1174 | J. Mater. Chem. B, 2013, 1, 1164–1174
This journal is ª The Royal Society of Chemistry 2013