Enzyme-based hybrid macroporous foams as highly efficient biocatalysts obtained through integrative chemistry

Article abstract of DOI:10.1021/cm100823d

Integrative chemistry-based rational design has been used to synthesize the first lipase [C-CRl]@Glymo-Si(HIPE) and [C-TAl]@Glymo-Si(HIPE) hybrid macrocellular biocatalysts, where immobilization of crude enzymes is optimized, while circumventing the reactants' low kinetic diffusion, by the use of silica macroporous hosts. As a direct consequence, these new hybrid biocatalysts display unprecedented cycling catalysis performance, as demonstrated by the syntheses of butyloleate ester (used as biodiesel lubricant), hydrolysis of linoleic-glycero ester derivatives (end products used for detergent and soap generations), and trans-esterification (reaction involved in the synthesis of low viscosity biodiesel). Considering that the catalytic performances are given in terms of absolute conversion percentage and not just relative enzyme activity, the enzyme@Glymo-Si(HIPE) hybrid macrocellular biocatalysts presented in this study display unprecedented high yield cycling catalysis performances, where turnover numbers (TON) and turnover frequencies (TOF) show promise for real industrial applications. This study can be considered as a milestone for enzyme-based heterogeneous catalyzes, thereby enhancing their competitiveness with the supported-catalysts commonly used in industry, in total agreement with current sustainable development issues. Also, the new macrocellular biocatalysts are well-suited for large-scale industrial production because of their above-mentioned performance characteristics, further enhanced by their monolithic character, which eases the separation of the catalysts from other reaction components.

Full text of DOI:10.1021/cm100823d

Chem. Mater. 2010, 22, 4555–4562 4555
DOI:10.1021/cm100823d
Enzyme-Based Hybrid Macroporous Foams as Highly Efficient
Biocatalysts Obtained through Integrative Chemistry
†
,‡
†
N. Brun, A. Babeau Garcia, H. Deleuze, M.-F. Achard, C. Sanchez,
‡
†
§
†
†
F. Durand, V. Oestreicher, and R. Backov*
,†
†
Centre de Recherche Paul Pascal, UPR 8641-CNRS, Universit eꢀ Bordeaux 1, 115 Avenue Albert Schweitzer,
‡
3600 Pessac, France, Universit eꢀ Bordeaux I/CNRS, Institut des Sciences Mol eꢀ culaires (ISM), 351 Cours
3
§
de la Lib eꢀ ration, F 33405 Talence, France, and Laboratoire Chimie de la Mati eꢁ re Condens eꢀ e de Paris
(
LCMCP) UMR-7574 UPMC-CNRS, Coll eꢁ ge de France, 11, place Marcelin Berthelot,
7
5231 Paris Cedex 05, France
Received March 23, 2010. Revised Manuscript Received July 16, 2010
Integrative chemistry-based rational design has been used to synthesize the first lipase [C-CRl]@
Glymo-Si(HIPE) and [C-TAl]@Glymo-Si(HIPE) hybrid macrocellular biocatalysts, where immo-
bilization of crude enzymes is optimized, while circumventing the reactants’ low kinetic diffusion, by
the use of silica macroporous hosts. As a direct consequence, these new hybrid biocatalysts display
unprecedented cycling catalysis performance, as demonstrated by the syntheses of butyloleate ester
(used as biodiesel lubricant), hydrolysis of linoleic-glycero ester derivatives (end products used for
detergent and soap generations), and trans-esterification (reaction involved in the synthesis of low
viscosity biodiesel). Considering that the catalytic performances are given in terms of absolute
conversion percentage and not just relative enzyme activity, the enzyme@Glymo-Si(HIPE) hybrid
macrocellular biocatalysts presented in this study display unprecedented high yield cycling catalysis
performances, where turnover numbers (TON) and turnover frequencies (TOF) show promise for
real industrial applications. This study can be considered as a milestone for enzyme-based hetero-
geneous catalyzes, thereby enhancing their competitiveness with the supported-catalysts commonly
used in industry, in total agreement with current sustainable development issues. Also, the new
macrocellular biocatalysts are well-suited for large-scale industrial production because of their
above-mentioned performance characteristics, further enhanced by their monolithic character,
which eases the separation of the catalysts from other reaction components.
Introduction
methodology in use, but rather the enhanced function or
intended use, which will ultimately determine the best
overall synthetic pathway. From this way of thinking has
recently emerged the concept of integrative chemistry,
which is applied herein to the design of highly efficient
porous enzyme-based supported catalysts.
In chemistry, heterogeneous catalysis is certainly the
area where the strongest efforts are needed, because
catalyzed reactions are involved both in decontamination
processes and in synthesis of compounds dedicated either
to the pharmaceutical or petroleum industries. In this
regard, enzyme-based heterogeneous catalysts are con-
sidered as potentially outstanding candidates, but to date
enzyme immobilization/activation has been associated
with slow kinetics due to low diffusion rates. Coupled
with the high cost of these materials, their use on the
industrial scale has been severely restricted. To design
high-performing heterogeneous biocatalysts, all the key
parameters that determine the final activity must be
balanced. The reactive surface and catalyst accessibility
1
With the emergence of the green chemistry criteria,
modern chemical science research efforts, both academic
and industrial, are strongly directed toward novel syn-
thetic pathways, which decrease the use of solvents, are
safe for the environment, use bioresources as precursors,
and have some degree of reusability, thereby achieving
the notion of sustainable development. Simultaneously,
chemists are requested to envisage materials that are
increasingly complex in nature and structure, have more
than one functionality, and at the extreme are inspired by
living organisms. To construct such highly advanced
functional materials, chemical science cannot be re-
stricted to a single specific domain of chemistry. Rather,
a strong interdisciplinary approach is needed, crossing
chemical boundaries and beyond. In such a context, there
is a crucial need for “rational design” of functional
architectures, in which the primary concern is not only
the “chemistry of shapes” or even the competence and
2,3
*
Corresponding author. E-mail: backov@crpp-bordeaux.cnrs.fr.
(2) Backov, R. Soft Matter 2006, 2, 452.
(3) Prouzet, E.; Ravaine, S.; Sanchez, C.; Backov, R. New J. Chem.
2008, 32, 1284.
(
1) Anastas, P. T.; Warner, J. C. Green Chemistry Theory and Practice;
Oxford University Press: New York, 1998.
r 2010 American Chemical Society
Published on Web 07/30/2010
pubs.acs.org/cm

4
556 Chem. Mater., Vol. 22, No. 16, 2010
Brun et al.
1
7
Scheme 1. Structure of the Two Enzymes in Use in This Study:
(
myces lanuginosus Lipase (TLl), pdb-1dt3
involved in biodiesel production. Both structures are pro-
posed within Scheme 1. In this study, these two lipase
enzymes will be used as reactive guests stabilized by silica-
hybrid meso-macrocellular monolith hosts.
1
a) Candida Rugosa Lipase (CRl), pdb-1crl, and (b) Thermo-
8
1
9
Lipases have very complex catalytic mechanisms. In
aqueous homogeneous solution, lipases are in equilibri-
2
0,21
um between two conformational states:
a “closed”
inactive form, in which the catalytic triad (aspartate,
histidine, and serine) in the active site is covered by a
helical “lid” (flap), and an “open” (active) form, in which
the lid has been displaced, adopting a totally different
conformation and exposing the catalytic residues. This lid
opening leads to the formation of a short-lived acyl-
enzyme intermediate, which reacts with water or alcohol
to produce the acids or the esters, respectively. In parti-
cular, when hydrophobic substrates interact with the
lipase, the lid opens and thus exposes the active site by a
and stability must be optimized, while avoiding, as far as
4
possible, problems with slow diffusion. At this stage,
these are competitive goals, because design of high-
performance enzyme-based catalyst supports to improve
enzyme stability via confinement is not commensurate
with good accessibility. These features, along with the
relatively high cost of enzymes, are certainly the reasons
why biocatalyst industrial applications have not yet
reached a significant level.
2
2
process known as “facial activation”. The control of the
hydrophobic-hydrophilic balance of matrices has been a
way to optimize enzymatic efficiency.
8
,9
Currently, one area of research has been directed to
optimization of catalytic reactions in organic media,
because all pharmaceutical or bio fuel end products are
insoluble in water. Also, it is known that water molecules
Among heterogeneous catalysis strategies, immobiliza-
tion, and encapsulation of enzymes either within inorganic
2
5,25
act as enzyme lubricants,
work in organic solvents containing little or even no
water.
and that enzymes can also
4
,5
6-9
10
solids, sol-gel derived matrices, ormembranes have
been active research areas because of the great potential
2
6,27
Immobilization/activation of enzymes is then
11
12
for use as biocatalysts and biosensors, and also because
these enzyme-based hybrid functional materials intrinsi-
cally promote the use of biomass components. Lipases
the key to expanding the applications of these natural
catalysts, thereby leading to easy separation and purifica-
tion of the end products from reaction mixtures, efficient
recovery of the enzyme-based catalyst, versatility regard-
ing the hydrous/anhydrous character of the solvent in use,
and minimization as far as possible of slow diffusion that
occurs because of enzyme confinement.
(
enzymes with various biological functions, including en-
triacylglycerol acyl hydrolases, EC. 3.1.1.3) are ubiquitous
1
3
antioselective hydrolysis and esterification,
chiral
resolution, synthesis of enantiomeric monomers and
14
15
macromolecules for polymerization reactions,
etc.
Among the lipases from various sources, Candida rugosa
lipase (CRl) has received much attention because of its high
activity and specificity for a range of substrates, whereas
Thermomyces lanuginosus lipase (TLl) appears as an out-
standing candidate for transesterification catalysis reactions
Experimental Section
Materials. Tetraethyl-orthosilane (TEOS), tetradecyltrimethy-
lammonium bromide (TTAB), (3-Glycidyloxypropyl)trimethoxy-
silane, and dodecane were purchased from Fluka. Lipase from
Candida rugosa (E.C.3.1.1.3, Type VII, 700 U/mg) and lipase from
Thermomyces lanuginosus (solution, g100 000 U/g) were pur-
chased from Sigma Chemical (St. Louis, MO). Oleic acid, glyceryl
trilinoleate (98%), ethyl linoleate (g99%), linoleic acid (g99%),
n-heptane, ethanol, and 1-butanol were purchased from Sigma and
16
(
4) Mahmood, I.; Guo, C.; Xia, H.; Ma, J.; Jiang, Y.; Liu, H. Ind. Eng.
Chem. Res. 2008, 47, 6379.
(
5) Kang, Y.; He, J.; Guo, X.; Song, Z. Ind. Eng. Chem. Res. 2007, 46,
4474.
(
6) Audebert, P.; Demaille, C.; Sanchez, C. Chem. Mater. 1993, 5, 911.
(7) Avnir, D.; Braun, S.; Lev, O.; Ottolenghi, M. Chem. Mater. 1994, 6,
1
605.
(18) Grochulski, P.; Li, Y.; Schrag, J. D.; Bouthillier, F.; Smith, P.;
Harrison, D.; Rubin, B.; Cygler, M. J. Biol. Chem. 1993, 268,
12843.
(19) Brzozowski, A. M.; Savage, H.; Verma, C. S.; Turkenburg, J. P.;
Lawson, D. M.; Svendsen, A.; Patkar, S. Biochemistry 2000, 39,
15071.
(
8) Reetz, M. T.; Zonta, A.; Simpelkamp, J.; K o€ nen, W. Chem.
Commun. 1996, 1396.
9) Reetz, M. T.; Zonta, A.; Simpelkamp, J. Angew. Chem., Int. Ed.
(
1
995, 34, 301.
(
10) Li, N.; Sabakie, K. J. Membr. Sci. 2008, 314, 183.
(
11) Dzige, N.; Aydiner, C.; Imer, D. A.; Bayramoglu, M.; Tanriseven,
A.; Keskinler, B. Bioresour. Technol. 2009, 100, 1983.
(20) Van Tilbeurgh, H.; Sarda, L.; Verger, R.; Cambillau, C. Nature
1992, 369, 159.
(21) Van Tilbeurgh, H.; Egloff, M.-P.; Matinez, C.; Rugani, N.; Verger,
R.; Cambillau, C. Nature 1993, 362, 814.
(22) Brzozowski, A. M.; Derewenda, U.; Derewenda, Z. S.; Dodson,
G. G.; Lawson, D. M.; Turkenburg, J. P.; Bjorkling, F.;
Huge-Jensen, B.; Patkar, S. A.; Thim, L. Nature 1991, 351, 491.
(23) Okahata, Y.; Mori, T. Trends Biotechnol. 1997, 15, 50.
(24) Kuntz, I. D.; Kauzmann, W. Adv. Protein Chem. 1974, 28, 239.
(25) Rupley, J. A.; Careri, G. Adv. Protein Chem. 1991, 41, 37.
(26) Klibanov, A. M. Nature 2001, 409, 241.
(
(
(
12) Chung, K. E.; Lan, E. H.; Davidson, M. S.; Dunn, B. S.; Valentine,
J. S.; Zinc, J. I. Anal. Chem. 1995, 67, 1505.
13) Ong, A. L.; Kamarruddin, A. H.; Bhatia, S.; Long, W. S.; Lim,
S. T.; Kumari, T. Enzyme Microb. Technol. 2006, 39, 924.
14) Ghanen, A.; Enein, H. Y. A. Tetrahedron: Asymmetry 2004, 15,
3
331.
15) Karla, B.; Kumar, A.; Gross, R. Chem. Rev. 2001, 1001, 2097.
16) Marı
a, P. D.; Montero, J. M. S.; Sinisterra, J. V.; Alc ꢀa ntara, A. R.
(
(
´
Biotechnol. Adv. 2006, 24, 180.
17) Dizge, N.; Aydiner, C.; Imer, D. Y.; Bayramoglu, M.; Tanriseven,
(
(27) Koskinen, A. M.; Klibanov, A. M. Enzymatic Reactions in Organic
Media; Blackie-Pergamon: London, 1996.
A.; Keskinler, B. Bioresour. Technol. 2009, 100, 1983.

Article
Chem. Mater., Vol. 22, No. 16, 2010 4557
Aldrich (Paris, France). Other chemicals and solvents used in this
study were of analytical grade or HPLC grade.
catalytic run, the monolith was extracted from the vessel and
washed three times with the water-saturated heptane solution.
Trans-Esterification Catalysis Reactions. Reactions were per-
formed as batch processes in 4 mL of heptane in glass tubes
containing 100 μL of glyceryl trilinoleate previously dissolved in
methyl tert-butyl ether at a concentration of 100 mg/mL, 25 mg
of ethanol and a 387 mg piece of monolith [C-TLl]@Glymo-
Si(HIPE). The reaction mixture was incubated at 37 °C for 24 h
with slow stirring. Conversion of glyceryl trilinoleate was mon-
itored using HPLC.
Synthetic Pathways. Syntheses of Native Silicon High Inter-
nal Phase Emulsion Si-(HIPE). TEOS (5.02 g) was added to an
aqueous solution of tetradecyltrimethylammonium bromide
(
TTAB) (16.02 g). Then, concentrated hydrochloric acid solu-
tion (37%) was added (5.88 g). The aqueous phase was allowed
to stir for approximately 5 min for the TEOS to hydrolyze. Then
the emulsion was ground in a mortar with dropwise addition of
dodecane (40.05 g). The emulsion was transferred to a plastic
flask and left to age for 3 days at room temperature. The
resulting material was washed with THF, dried in air, and cured
at 650 °C for 6 h (heating rate of 2 °C/min with a first plateau at
Characterizations. Scanning electron microscopy (SEM) ob-
servations were performed with a Jeol JSM-840A scanning
electron microscope operated at 10 kV. Intrusion/extrusion
mercury measurements were performed using a Micromeritics
Autopore IV apparatus in order to assess the scaffolds’ macro-
cellular cell characteristics. Surface areas and pore characteristics at
the mesoscale were obtained by nitrogen adsorption-desorption
experiments using a Micromeritics ASAP 2010. Small-angle X-ray
scattering (SAXS) experiments were carried out on an 18 kW
rotating-anode X-ray source (Rigaku-200) using a Ge (111) crystal
as monochromator. The scattered radiation was collected on a two-
dimensional detector (Imaging Plate system from Mar Research,
Hamburg). The sample- detector distance was 500 mm. TEM was
performed using a HITACHI - H7650 TEM operated at 80 kV
equipped with an ORIUS GATAN (11MPX) camera.
2
00 °C for 2 h).
Synthesis of Grafted Glymo-Si(HIPE) Supports. A one-gram
piece of native Si-(HIPE was added to a solution of (3-Glycidylox-
ypropyl)trimethoxysilane organosilane (Glymo) (0.01 mol) in
chloroform (120 g). The suspension was placed under vacuum to
enhance impregnation. After 48 h at room temperature, the solu-
tion was filtrated, and the monoliths were then washed with
chloroform and acetone and dried in air.
Lipase Immobilization. Crude lipase from Candida rugosa
540 mg) was dispersed in distilled water (18 mL). The mixture
(
was first stirred for an hour. The immobilization of lipase was
carried out by introducing the Glymo-Si(HIPE) open cell monolith
HPLC Analytical System. The analytical system consisted of a
model 600 solvent delivery system with a manual injector (Waters,
Milford, MA). The compounds were separated on an Atlantis
DC18 (4.6 mm ꢀ 150 mm, 5 μm) column with an Atlantis DC18
guard column (Waters) using an isochratic mobile phase of HPLC-
grade acetonitrile. The column was operated at room temperature.
EMPOWER software (Waters) was used for data acquisition and
processing. Standards were dissolved in methyl tert-butyl ether
(250 mg) into the enzyme solution, and a vacuum (20 mbar) was
applied for 72 h at ambient temperature. Crude lipase from
Thermomyces lanuginosus (53 mg) was dissolved in distilled water
(
10 mL). The Glymo-Si(HIPE) (350 mg) was added to the enzyme
solution, and a vacuum (20mbar) was applied for 72 h at ambient
temperature. The immobilized lipases were then extracted from the
impregnation media and washed three times with distilled water in
order to remove the excess enzyme. The final [C-CRl]@Glymo-
Si(HIPE) and [C-TLl]@Glymo-Si(HIPE) heterogeneous catalysts
are then dried at room temperature for 12 h and stored in a
refrigerator at 4 °C. The stoichiometries of the Candida rugosa
and Thermomyces lanuginosus-based heterogeneous catalysts were
determined by combined C,H,N elemental analysis data and TGA
experiments.
(MTBE). All solutions were filtered through a 0.45 μm membrane
and degassed before use. The mobile phase flow-rate was 1 mL/min
and 20 μL samples were injected. A model 410 refractometer
(Waters, Milford, MA, USA) was used for detection of the
esterification reaction. For both the hydrolysis and trans-esterifica-
tion reactions the detector was an ultraviolet diode array 996
(Waters, Milford, MA). The absorbance maximum was at λ =
Set-up of the Catalytic Reactions. Esterification Catalysis
Reactions. The enzymatic esterification reaction consisted of
204 nm. A gradient elution was used: solvent A, acetonitrile; solvent
B, methyl tert-butyl ether; 4 min isochratic at 100% (v/v) A; 2 min
-
3
-3
1
-butanol (2.0 ꢀ 10 mol), oleic acid (1.0 ꢀ 10 mol), a piece of
monolith [C-CRl]@Glymo-Si(HIPE) (247 mg) and heptane
2 mL). The reaction mixture was incubated at 37 °C for 24 h
gradient to 70/30 (v/v) A/B; isocratic for 10 min at 70/30 (v/v) A/B;
0.5 min return to 100% (v/v) A; 10 min re-equilibration with 100% A.
(
under gentle stirring to avoid breaking the monoliths. Forma-
tion of the ester was monitored using HPLC to determine the
conversion rate.
Results and Discussion
Synthesis and Characterization of Biocatalyst Foams. In
the present study, the confinement media were organo-
silica-based hybrid macrocellular monolithic foams hav-
ing hierarchical porous structures. The monolith-type
material observed in Figure 1a was obtained through a
double template process making use of direct emulsion to
create the macroporosity and a lyotropic mesophase both
to stabilize the oil/water interface and to create micellar-
Homogeneous catalysis with free lipases was carried out in
-3
heptane (2 mL), with 1-butanol (2.0 ꢀ 10 mol), oleic acid (1.0 ꢀ
-
3
10
mol), and crude Candida rugosa lipase (17.5 mg). The reaction
was incubated at 37 °C for 40 h with vigorous stirring. Formation of
ester was monitored using HPLC to determine the conversion rate.
Triester Hydrolysis Catalysis Reactions. In this experiment, a
2
00 mg piece of monolith, [C-CRl]@Glymo-Si(HIPE), was
placed into 15 mL of water-saturated heptane. After reaching
the desired temperature between 37 and 38 °C, 100 μL of a
glyceryl trilinoleate (triester) dissolved in methyl-tert-butyl ether
2
8
induced mesoscopic porosity. The macroscopic cell sizes of
such architectures vary depending on the initial oil volume
fraction. We labeled this new series of materials “Si(HIPE)”
(
MTBE) was addedat a concentration of 100 mg/mL, leading
29
with the acronym HIPE for “high internal phase emulsion”.
finally to a triester solution with a concentration of 0.66 mg/mL.
The catalysis batch process was performed with slow stirring.
HPLC was used to check the disappearance of reagent
and appearance of both the final product and intermediates.
After 24 h of catalytic cycling, and prior to performing a new
(
28) Carn, F.; Colin, A.; Achard, M.-F.; Birot, M.; Deleuze, H.;
Backov, R. J. Mater. Chem. 2004, 14, 1370.
29) Cameron, N. R.; Sherrington, D. C. Adv. Polym. Sci. 1996, 126,
162.
(

4
558 Chem. Mater., Vol. 22, No. 16, 2010
Brun et al.
from the hollow sphere aggregation (previously called ex-
2
9
ternal junctions).
The distribution of the interconnected pore opening
sizes, as measured by mercury intrusion porosimetry
(Figure 1d), covers a broad range (50 to 5000 μm) with
a major contribution at 9000 μm. Specific morphological
characterizations were 94% porosity, 0.085 g/cm bulk
-
3
-
3
density, and 1.56 g/cm skeleton density. It is note-
worthy that these monolithic materials have the same
3
4
bulk density as common silica aerogel and can endure
the mercury impregnation process, thus indicating good
mechanical properties. At the mesoscopic length scale,
these hybrid foams have vermicular-type mesopores as-
˚
sociated with a wall-to-wall average distance of 31 A, as
shown in Figure 1c. Nevertheless, when considering the
nitrogen physisorption (Figure 1e,f) measurements,
which strongly suggest a relatively high microporosity
2
-1
2 -1
(BET=170m g ) and low mesoporosity (BJH = 45 m g ),
there is at first glance a certain collapse of mesoporosity
2
8
when compared with pure inorganic Si(HIPE) foams.
In fact, it is very well established that postgrafting of
silane tends to strongly block the mesoporous nodes,
minimizing nitrogen accessibility to the interior of the
3
5-38
mesopores.
This feature is certainly enhanced here
because of the enzyme adjunction. As a direct conse-
quence, all subsequent catalyzed reactions will occur
exclusively inside the macropores. The full surface char-
acterization both at the macroscopic and mesoscopic
length scales can be found in the Supporting Information.
Also at the microscopic length scale, it is important to
provide stoichiometries of the as-synthesized biocatalysts, in
order to both relate and normalize catalytic performances
with enzyme activity. The stoichiometries of the Candida
rugosa and Thermomyces lanuginosus based heterogeneous
catalysts were obtained through combined C, H, N elemental
analysis and thermal gravimetric analysis (TGA) and are,
Figure 1. Examples of support textural characteristics at three length
scales of the as-synthesized Enzyme(CRl)@Glymo-Si(HIPE) macrocel-
lular foam: (a) Photograph of the as-synthesized monoliths, (b) SEM
micrograph where the black arrow indicates an internal junction while the
white arrow indicates an external junction, (c) TEM pictures depicting
mesoporous void spaces (SAXS profile, embedded), (d) size distribution
of the macroscopic pore openings, (e) nitrogen adsorption (O)-
desorption (9) measurements, (f ) resulting distribution of the mesoscopic
pore sizes (obtained by the density functional theory). The full surface
characterization for all hybrid foams in use, both at the macroscopic and
mesoscopic length scales, can be found in the Supporting Information.
Recently, we developed a new generation of hybrid Organo-
3
0
Si(HIPE) series having organic functionalities, such as
mercapto, diamino, benzyl, dinitro, etc., derivatives at the
molecular level. These allow the functionalities of the final
porous materials to be directed toward either heterogeneous
-
5
respectively, [C. CR1]
9.10
@Si1.91(C O H )
6
0.09H O
0.06 H O. The
2
2
11 0.09 2
3
-
5
and [C-TLl]8.10 @SiO1.95(C O H )
6
2
11 0.05
3
crude Candida rugosa lipase had a purity of 3 wt %, whereas
the crude Thermomyces lanuginosus had a purity of 95 wt %
31
32
catalysis or photoluminescence. The present study is the
first use of (3-Glycidyloxypropyl)trimethoxysilane (Glymo)
as a functional agent to stabilize the embedded enzymes. This
methodology allows preparation of net-shaped monolithic
materials that can be easily manipulated, separated and
transferred (Figure 1a). As shown in Figure 1b, the texture
resembles an aggregate of hollow spheres because of the pH-
(see the Supporting Information for the employed Bradford
purification technique). Quantitatively, there was 2.3 mg of
active enzyme per gram of catalyst for [C-CRl]@Glymo-
Si(HIPE) and 31 mg of active enzyme per gram of catalyst for
[
C-TLl]@Glymo-Si(HIPE). Also the grafting of the Glymo
29
moieties through condensation was verified by Si NMR
spectroscopy (see the Supporting Information).
29,33
induced Euclidian character
of the inorganic walls. As a
Catalytic Activity of the Enzyme Biocatalyst Foams. Three
main reactions were chosen to test these enzyme-based
heterogeneous catalysts: a simple esterification reaction, a
reaction based on triester hydrolysis, and a trans-esterifica-
tion reaction associated with biodiesel production. All the
above-mentioned reactions can be found in Scheme 2.
direct consequence, two kinds of pore openings connecting
adjacent cells must be considered: the intrawall pore openings
(previously called internal junctions) and the pore openings
(
30) Ungureanu, S.; Birot, M.; Guillaumme, L.; Deleuze, H.; Babot, O.;
Julian, B.; Achard, M.-F.; Popa, M. I.; Sanchez, C.; Backov, R.
Chem. Mater. 2007, 19, 5786.
(
(
31) Ungureanu, S.; Deleuze, H.; Popa, M. I.; Sanchez, C.; Backov, R.
Chem. Mater. 2008, 20, 6494.
32) Brun, N.; Julian-Lopez, B.; Hesemann, P.; Guillaume, L.; Achard,
(34) H u€ sing, N.; Schubert, U. Angew. Chem., Int. Ed. 1998, 37, 22.
(35) Moller, K.; Bein, T. Chem. Mater. 1998, 10, 2950.
(36) Stein, A.; Melde, B. J.; Schroden, R. C. Adv. Mater. 2000, 12, 1403.
(37) Nicole, L.; Boissi ꢁe re, C.; Grosso, D.; Quach, A.; Sanchez, C.
J. Mater. Chem. 2005, 15, 3598.
(38) Hoffmann, F.; Cornelius, M.; Morel, J.; Fr o€ ba, M. Angew. Chem.,
Int. Ed. 2006, 45, 3216.
M.-F.; Deleuze, H.; Sanchez, C.; Backov, R. Chem. Mater. 2008,
20, 7117.
(33) Brinker, C. J.; Scherer, G. W. Sol-Gel Science: The Physics and
Chemistry of Sol-Gel Processing; Academic Press: London, 1990.

Article
Chem. Mater., Vol. 22, No. 16, 2010 4559
Scheme 2. Three Typical Reactions Chosen for Testing the As-Synthesized Enzyme-Based Catalyst Performances: Top, Esterification of
Oleic Acid (1) with Butanol (2) Providing Butyl Oleate Ester (3); Middle, Hydrolysis of Trilinolein (4) Providing Linoleic Acid (5)
0
00
000
and Glycerol (6) (Intermediate Products (4 ) 1,3-Dilinolein, (4 ) 1,2-Dilinolein, (4 ) 1-Monolinoleoyl-rac-glycerol); Bottom:
Trans-Esterification of Trilinolein (4) with Ethanol (7) Providing Glycerol (6) and Ethyl Linoleate Ester (8)
Enzyme-Based Heterogeneous Catalysis of Esterifica-
tion Reactions. The first chemical reaction chosen to
evaluate the as-synthesized enzyme-based catalyst per-
formance was the well-known esterification of oleic acid
with butanol to provide butyl-oleate ester (Scheme 2), this
ester being an essential biodiesel additive acting as lubri-
Figure 2a shows that the conversion kinetics are greatly
enhanced by heterogeneous catalysis compared to the
homogeneous system. The conversion rate (Figure 2b) for
homogeneous conversion rate is 0.9 μmol min- mg
1
-1
,
whereas the initial rate of the heterogeneous reaction is
-1
3.5 μmol min-1 mg . This result demonstrates without
2
3
9
cant during winter use. To assess enzyme heterogeneous
catalysis, it is also important to consider enzyme homo-
geneous catalysis, in order to appreciate the host confine-
ment effect. This first experiment is addressed in Figure 2.
ambiguity the enhancement of enzyme activity via con-
finement; in the case of Glymo-Si(HIPE) confinement,
the enzymatic activity for this esterification reaction is
increased by about 25-fold. Also the conversion by
homogeneous catalysis was not completed after 24 h,
but the heterogeneous system led to 100% conversion
(
39) Linko, Y.-Y.; L €a m €a , M.; Wu, X.; Uosukainen, E.; Sepp €a l €a , J.;
Linko, P. J. Biotechnologies 1998, 66, 41.

4
560 Chem. Mater., Vol. 22, No. 16, 2010
Brun et al.
Figure 3. Candida rugosa-based heterogeneous catalyst performance.
[C-CRl]@Glymo-Si(HIPE)-based esterification reactions: O, Glymo stabi-
lized enzyme; 4, without Glymo.
Thus, the results obtained for the heterogeneous catalysis
are due to the intrinsic biohybrid catalyst performance.
Also, when performing these cycles, the [C-CRl]@Glymo-
Si(HIPE) was stored at 4 °C for 2 months between cycles
Figure 2. Enzymatic catalytic performance in an esterification reaction.
a) O [C-CRl]@Glymo-Si(HIPE) heterogeneous catalysis; 9 [C-CRl]
1
0 and 11 and, as shown in Figure 3, without any loss of
(
catalytic activity. Thus, in addition to having high cata-
lytic activities, these enzyme-based catalysts can be stored
in a refrigerator at 4 °C without any special precautions.
When compared with the currently best enzyme-based
homogeneous catalysis (17.5 mg of commercial Candida rugosa crude
Lipase in heptane). (b) Zoom of the blue dashed area of a: O [C-CRl]@
Glymo-Si(HIPE) heterogeneous catalysis; 9 [C-CRl] homogeneous cat-
alysis (17.5 mg of commercial Candida rugosa crude Lipase in heptane).
The red dashed line corresponds to the slope of the conversion versus time
for the first 10 min of reaction.
40-43
heterogeneous catalysts,
[C-CRl]@Glymo-Si(HIPE)
shows significantly enhanced esterification catalytic perfor-
mance with excellent stability over time. The conversions
observed with other supports currently in use, e.g., polyur-
after about 12 h. The results in Figure 2 also show that the
enzyme concentration within the heterogeneous catalyst
was not excessive. This is the first reason why the same
catalytic experimental setup was chosen to evaluate the
cycling performance of the [C-CRl]@Glymo-Si(HIPE)
heterogeneous catalysts. Another reason is that this
experimental setup is nearly the same as that for the best
enzyme-based heterogeneous catalysts found in the
40
41
41
ethane foam, calcium carbonate, silica gel 60, Amber-
41
41
lite IRC-50, Celite 545, natural kaolin, and layered
42
43
double hydroxides, were always below 85% over at most 9
cycles with lipase life times of 12 days at best. Also, to
decrease the cost of this enzyme-based catalyst, we investi-
gated the need for Glymo entities. Catalytic tests with the
4
0-43
literarure.
[
C-CRl]@Si(HIPE) catalytic support free of the “Glymo”
As shown in Figure 3, the [C-CRl]@Glymo-Si(HIPE)
catalyst was able to cycle for 19 runs with 100% of
conversion in each cycle. There was a small decrease in
the catalytic activity for the 20th and 21st runs, in agree-
ment with the observation of partial collapse of the
monolith and consequent loss of enzyme into the reaction
batch. Despite final catalyst collapse it may be argued
that the conversion reached 100% for 19 runs because the
amount of active enzyme was very high (or oleic acid
concentration was very low). However, the oleic acid and
active enzyme concentrations were exactly the same as
those used for the homogeneous catalysis, which did not
achieve complete conversion even after 24 h (Figure 2).
stabilizing agent were performed, and the results are pro-
vided in Figure 3 (Δ symbols). The first conversion was
around 95% complete with a slower kinetics (double reac-
tion time). With cycling, the conversions decreased to 87%
with slower kinetics when compared to the catalyst support
having the Glymo entities acting as enzyme stabilizing
agents. This result confirms that even with crude enzyme,
Glymo molecules, which hydrolyze spontaneously in water
to generate diol functional groups, are indeed necessary to
improve enzyme stability and hydration and thereby the
44
catalytic activity.
Enzyme-Based Heterogeneous Catalysis of Hydrolysis
Reactions. The second reaction chosen to test the [C-CRl]
@Glymo-Si(HIPE) catalytic activity was triester hydro-
lysis (Scheme 2), specifically the hydrolysis of a trilinolein
ester which, with triglyceride esters of oleic and
palmitic acid, represents a major component of olive oil.
Moreover, the corresponding linoleic acid and linoleate
(
(
(
(
40) Awang, R.; Basri, M. R. G. M. Am. J. Biochem. Biotechnol. 2007, 3,
163.
41) Ghamgui, H.; Karra-Cha ^a bouni, M.; Gargour, Y. Enzyme Mi-
crob. Technol. 2004, 35, 355.
42) Rahman, M. B. A.; Tajudin, S. M.; Hussein, M. Z.; Rahman, R. N.
Z. R. A.; Sallehb, A. B.; Basri, M. Appl. Clay Sci. 2005, 29, 111.
43) Rahman, M. B. A.; Basri, M.; Hussein, M. Z.; Idris, M. N. H.;
Rahman, R. N. Z. R.; Salleh, A. B. Catal. Today 2004, 93-95, 405.
(44) Wheatleya, J. B.; Schmidt, D. E. J. Chromatogr., A 1999, 849, 1.

Article
Chem. Mater., Vol. 22, No. 16, 2010 4561
4
5-47
from the support, as claimed elsewhere.
However,
when the catalyst was rehydrated (Figure 4, dashed line),
the [C-CRl]@Glymo-Si(HIPE) regained the catalytic ac-
tivity of the enzyme. Water has to be added to ensure the
2
2-27
enzyme’s structural integrity and catalytic activity.
This means that the lack of catalyst hydration, even using
a water saturated heptane solution (0.6 wt %), is indeed a
major problem for the hydrolysis reaction itself. Also,
hydrolysis performances reached with the magnetic
4
7
microsphere bead catalysts were obtained with hydra-
tion between 25 and 65 wt % water (around 100 times
higher than that of the experiment described here), en-
hancing the intrinsic hydrolysis reaction, but also cer-
tainly leading to a triphasic system, i.e., water in oil
biphasic emulsion and the solid-state catalyst itself. This
overall scenario appears at first glance to make supported
catalyst use difficult on the industrial. This is the reason
why we chose to minimize, as far as possible, the water
content in the reaction medium when dealing with this
biocatalyzed hydrolysis reaction, working thereby in the
worst reaction conditions but optimizing the industrial
potential applications. The high performance of the
Figure 4. [C-CRl]@Glymo-Si(HIPE) linoleic acid production through
triester hydrolysis. The dashed line at the beginning of the 11th cycle
indicates that catalyst hydration has been performed prior this new cycle.
esters have two double bonds, which are easily by HPLC
(see Experimental Section ). Usually, enzyme-based catalysts
are tested by ester hydrolysis in water of either p-nitophenyl
[
C-CRl]@Glymo-Si(HIPE) catalyst, despite functioning
with very low amounts of water, is certainly due to the
intrinsic hydration of silica (see the catalysts’ stoichiom-
etry in the Experimental Section), which favors enzyme
stability. This inherent hydration characteristic is also
important in the following biocatalyzed reaction.
4
5
palmitate ( pNPP) or p-nitrophenyl acetate (pNPA), be-
cause the nitrophenyl group can be easily followed by UV-
vis experiments, and because water, being the solvent, is not a
limiting reagent. Real olive oil triester derivatives are very
seldom used because of several issues. First, the triesters are
mostly insoluble in water, so that the solution in use for the
catalysis can be at best only a “water-saturated” nonaqueous
medium. Second, both the triester, the intermediate diesters,
and the final oleic acid (Scheme 2) are not easily followed
using HPLC. Despite those problems, some successes have
been achieved using either polypropylene modified mem-
Enzyme-Based Heterogeneous Catalysis of Trans-Esterifi-
cation Reactions. The third reaction chosen for testing the
enzyme-based biocatalysts was the transesterification reac-
tion mainly involved in biodiesel production (Scheme 2). For
this particular catalysis reaction the C. Rugosa has been
shown to offer poor efficiency for biodiesel production,
49
50
whatever the support in use, for instance Celite or kaolite.
For this reason, we used Thermomyces lanuginosus lipase
Scheme 1) to generate a new [C-TLl]@Glymo-Si(HIPE)
4
6
47
branes, magnetic microspheres, or layered double hydro-
48
(
xides (LDHs) as supported media.
heterogeneous biocatalyst. This was also a direct way to
demonstrate that use of Glymo-Si(HIPE) as the host-
support is not restricted to the C. Rugosa lipase. In particular,
the lipase from T. Lanuginosus has demonstrated good
efficiency toward biodiesel production, when supported on
The results in Figure 4 show that [C-CRl]@Glymo-
Si(HIPE) keeps producing a conversion of at least 80%
for the first four cycles and above 65% for the remaining
cycles.
These conversion percentages are higher than those
obtained with enzyme stabilized in open structures: poly-
51
52
13
Celite, acrylic resin, polymeric macroporous foams,
5
3
4
propylene membranes (loss of around 6% of the initial
6
and silica-poly (vinyl alcohol) composite. As with C. Rugosa,
the T. Lanuginosus was employed as a crude material for
impregnation into the host Glymo-Si(HIPE), leading to the
catalyst [C-TLl]@Glymo-Si(HIPE). Figure 5 shows that
that the first cycle achieved 100% conversion within 24 h at
4
8
activity by the fourth cycle) and LDHs (conversion
lower than 80% for the first batch and loss of around
10% of the initial activity by the fourth cycle). Moreover,
reusability of the [C-CRl]@Glymo-Si(HIPE) is as good as
that reported for enzyme stabilized with magnetic porous
37 °C.
4
7
This result is much higher than the conversions ob-
tained previously for T. Lanuginosus lipase as guest for
microspheres. We first thought that the decrease of
conversion observed in Figure 4 was caused by the
inherent protein denaturation and substantial leakage
1
7,52,53
heterogeneous catalysts.
For instance, taking the
(
(
45) Ozyilmaz, G. J. Mol. Catal. B: Enzym. 2009, 56, 231.
46) Deng, H.-T.; Xu, Z.-K.; Dai, Z.-W.; Wu, J.; Seta, P. Enzyme and
Microbial Tech. 2005, 36, 996.
(49) Shah, S.; Sharma, S.; Gupta, M. N. Energy Fuels 2004, 18, 154.
(50) Iso, M.; Chen, B.; Eguchi, M.; Kudo, T.; Shrestha, S. J. Mol. Catal.
B: Enzymatic 2001, 16, 53.
(51) Soumanou, M. M.; Bornscheuer, U. T. Enzyme Microbial Technol.
2003, 33, 97.
(52) Hern ꢀa ndez-Matı
´
n, E.; Otero, C. Bioresour. Technol. 2008, 99, 277.
(53) Moreira, A. B. R.; Perez, V. H.; Zanin, G. M.; de Castro, H. F.
(
(
47) Yong, Y.; Bai, Y.-X.; Li, Y.-F.; Lin, L.; Cui, Y.-J.; Xia, C.-G.
Process Biochemistry 2008, 43, 1179.
48) Rahman, M. B. A.; Zaidan, U. H.; Basri, M.; Hussein, M. Z.;
Rahman, R. N. Z. R. A.; Salleh, A. B. J. Mol. Catal. B: Enzym.
2008, 50, 33.
Energy Fuels 2007, 21, 3689.

4
562 Chem. Mater., Vol. 22, No. 16, 2010
Brun et al.
performed in this study, no water was addedtothe reaction
medium (see Experimental Section), contrary to all pre-
vious mentioned T. Lanuginosus-based heterogeneous
1
7,52,53
catalysts.
of the [C-TLl]@Glymo-Si(HIPE) catalyst.
This feature is another major advantage
The catalytic performances of the enzyme based hybrid
catalysts in all three reactions are summarized in Table 1.
The turnover number (TON) and turnover frequency
(TOF) presented in Table 1 and the results reported in
Figures 3-5, which correspond to absolute conversion and
not to relative enzyme activity, indicate that the enzyme@
Glymo-Si(HIPE) hybrid macrocellular biocatalysts pre-
sented in this study have very high conversion efficiencies
in multiple catalytic cycles. For the first time, the TON
and TOF numbers indicate that these materials are
feasible for real industrial applications. We are currently
performing uniaxial flux catalysis, in order to better
optimize mass transport through the catalyst while mini-
mizing the use of solvent. Those investigations are in
progress and will be published in due time.
Figure 5. [C-TLl]@Glymo-Si(HIPE)-based trans-esterification re-
actions. The black line corresponds to the trilinoline ester conversion,
while the blue line corresponds to the ethyl linoleiate ester production.
For each experiment, the first conversions are equal to 100%; this is the
reason why the catalytic performances are presented in absolute con-
version performances and not using percentage relative activity.
Table 1. Catalytic Performance of Hybrid Foams
Conclusion
hybrid foam
catalytic
performance
In summary, we report herein the integrative chemistry-
based design of new enzyme-based macrocellular
heterogeneous monolith-type biocatalysts [C-CRl]@
Glymo-Si(HIPE) and [C-TLl]@Glymo-Si(HIPE), which
demonstrate several favorable characteristics: high con-
version percentages in multiple cycles, low steric hindrance
between enzymes and substrates, minimal problems with
slow diffusion rates, low cost of catalysts because of the use
of crude enzymes, efficient recovery of the monolith-type
catalysts allowing simple recovery and separation of end
products, versatility to work either in nonaqueous or water
saturated nonaqueous media, and versatility in choice of
the stabilized enzymes.
a
esterification hydrolysis transesterification
a
b
TON
TOF (h
105615
4400.63
23.5
3482.61
145.109
0.210
70.8749
2.95312
0.00483
-1
)
enzymatic activity
-1
-1
(
μmol min mg )
a
b
C-CRl]@Glymo-Si(HIPE) catalyst. [C-TLl]@Glymo-Si(HIPE)
[
catalyst. The TON and TOF numbers were calculated for a single cycle;
full TON and TOF can be obtained by multiplying by the number of
cycles for each catalyzed reaction.
best results to date, T. Lanuginosus stabilized within macro-
porous polyglutaraldehyde-modified polymer foams achiev-
ed 90.2% within 24 h when used as a monolith, at best
With these favorable characteristics, all associated with
current sustainable development issues, this study repre-
sents a significant step forward promoting robust re-
emergence of enzyme-based biocatalysts for industrial
applications.
9
7% when the monoliths were powdered, both systems at a
temperature of 65 °C. Thus, our results clearly demonstrate
the high efficiency of Glymo toward stabilizing T. Lanuginosus.
Also another factor may contribute to the rather high
conversion: namely, as described above for the second
reaction, the inherent hydration of silica matrices, allowing
optimum [C-TLl]@Glymo-Si(HIPE) catalytic performance
Acknowledgment. This work is partially supported by the
grand “Nano-cataHIPE” from the French national research
agency (ANR). The authors thank Dr. Kathryn Williams of
the University of Florida for her careful editing of the
manuscript.
(see above for stoichiometry). In fact, it is well-known that
water in small amounts is an important parameter for
optimization of transesterification catalytic performance
when enzymes are used as catalysts. We want to empha-
size that in all the cycling tests dedicated to transesterification
54
Supporting Information Available: Additional figures and
information (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
(54) Antczak, M. S.; Kubiak, A.; Antczak, T.; Bielecki, S. Renewable
Energy 2009, 34, 1185.