The chemo-enzymatic Baeyer-Villiger oxidation of cyclic ketones with an efficient silica-supported lipase as a biocatalyst

Article abstract of DOI:10.1016/j.apcata.2013.07.009

The synthesis and characterisation of new, silica-supported lipase biocatalysts and studies of their performance in Baeyer-Villiger oxidation (BVO) of cyclic ketones to lactones were described. Biocatalysts were obtained by immobilisation of Candida Antarctica lipase B onto siliceous materials with multimodal pore structure (MH) and also on typical SBA-15, chemically modified with organosilanes terminated with methyl, octyl and hexadecyl group. Biocatalysts structure was characterised by nitrogen adsorption and TEM/SEM imaging. The lipase presence was confirmed by FTIR spectroscopy, whereas protein loads were obtained from thermogravimetric data. They appeared to be notably larger for MH catalysts than SBA-15s and correlated with surface hydrophobicity of supports which decreased in the order Hd > Oc > Me, from 198 to 80 mg/g, for MH-Hd-L and MH-Me-L biocatalysts, respectively. Investigation of the biocatalysts performance (activity, stability, reusability) in BVO of cyclic ketones, using urea hydrogen peroxide as oxidant and ethyl acetate as both the peracid precursor and solvent, clearly demonstrated very large activity of MH biocatalysts. For the same lipase load activity decreased in the order MH-Me-L > MH-Oc-L > MH-Hd-L, with that of least active (MH-Hd-L) catalysts being even trice larger than of Novozyme-435 under equal mass basis. The biocatalysts showed good stability, even in the presence of 60% aq. hydrogen peroxide, and facile reuse. On this basis we proposed a general chemo-enzymatic BVO method for oxidation of cyclohexanones and cyclobutanone to obtain adequate lactones with high yields (80-98%).

Full text of DOI:10.1016/j.apcata.2013.07.009

Accepted Manuscript
Title: The chemo-enzymatic Baeyer–Villiger oxidation of
cyclic ketones with an efficient silica-supported lipase as a
biocatalyst
Authors: A. Dro z˙ d z˙ , A. Chrobok, S. Baj, K. Szyma n´ ska, J.
Mrowiec-Biało n´ , A.B. Jarz e˛ bski
PII:
S0926-860X(13)00407-9
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.apcata.2013.07.009
APCATA 14325
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
21-3-2013
8-6-2013
5-7-2013
Please cite this article as: A. Dro z˙ d z˙ , A. Chrobok, S. Baj, K. Szyma n´ ska, J. Mrowiec-
Biało n´ , A.B. Jarz e˛ bski, The chemo-enzymatic Baeyer–Villiger oxidation of cyclic
ketones with an efficient silica-supported lipase as a biocatalyst, Applied Catalysis
A, General (2013), http://dx.doi.org/10.1016/j.apcata.2013.07.009
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

The chemo-enzymatic Baeyer–Villiger oxidation of cyclic ketones with an efficient
silica-supported lipase as a biocatalyst
a
a,
a
b
b,c
b,c
A. Drożdż, A. Chrobok, * S. Baj, K. Szymańska, J. Mrowiec-Białoń, A. B. Jarzębski
a
Department of Chemical Organic Technology and Petrochemistry, Silesian University of
Technology, Krzywoustego 4, 44-100 Gliwice, Poland, Fax: + 48 32 237 10 32; e-mail:
Anna.Chrobok@polsl.pl
b
Department of Chemical and Process Engineering, Silesian University of Technology,
Strzody 7, 44-100 Gliwice, Poland
c
Institute of Chemical Engineering, Polish Academy of Sciences, Bałtycka 5, 44-100
Gliwice, Poland
Abstract
The synthesis and characterisation of new, silica-supported lipase biocatalysts and studies
of their performance in Baeyer–Villiger oxidation (BVO) of cyclic ketones to lactones
were described. Biocatalysts were obtained by immobilization of Candida Antarctica
lipase B onto siliceous materials with multimodal pore structure (MH) and also on typical
SBA-15, chemically modified with organosilanes terminated with methyl, octyl and
hexadecyl group. Biocatalysts structure was characterised by nitrogen adsorption and
TEM/SEM imaging. The lipase presence was confirmed by FTIR spectroscopy, whereas
protein loads were obtained from thermogravimetric data. They appeared to be notably
larger for MH catalysts than SBA-15s and correlated with surface hydrophobicity of
supports which decreased in the order Hd>Oc>Me, from 198 to 80 mg/g, for MH-Hd-L
and MH-Me-L biocatalysts, respectively.
Investigation of the biocatalysts performance (activity, stability, reusability) in BVO of
cyclic ketones, using urea hydrogen peroxide as oxidant and ethyl acetate as both the
peracid precursor and solvent, clearly demonstrated very large activity of MH biocatalysts.
For the same lipase load activity decreased in the order MH-Me-L> MH-Oc-L> MH-Hd-
L, with that of least active (MH-Hd-L) catalysts being even trice larger than of Novozyme-
1
Page 1 of 29

435 under equal mass basis. The biocatalysts showed good stability, even in the presence
of 60% aq. hydrogen peroxide, and facile reuse. On this basis we proposed a general
chemo-enzymatic BVO method for oxidation of cyclohexanones and cyclobutanone to
obtain adequate lactones with high yields (80-98%).
Keywords: chemo-enzymatic Baeyer-Villiger reaction, lactones, cyclic ketones oxidation,
silica-supported lipase B, enzyme carriers with hierarchical structure
Introduction
The Baeyer–Villiger oxidation (BVO) of ketones to lactones or esters is a reaction of
significant interest in organic chemistry owing to very wide range of possible applications,
e.g. in the synthesis of antibiotics, steroids, pheromones and monomers for polymerisation
[1, 2, 3, 4]. In the last decade catalysts based on molecular sieves modified with tin (Sn-
Beta) or Sn(IV) centers grafted onto mesoporous MCM-41 (Sn-MCM-41) were shown to
be excellent and highly selective catalysts for Baeyer-Villiger oxidation of cyclic ketones
2 2
with H O as oxidant [5, 6, 7, 8, 9]. The organic percarboxylic acids typically used as
oxidants in these reactions are fairly expensive, often poorly-stable and hazardous, and
this consequently limits their commercial application. Therefore, the chemo-enzymatic
approach appears to be a very attractive alternative. It involves oxidation of long- or
2 2
medium-chain carboxylic acids with H O or urea hydrogen peroxide (UHP) to generate in
situ peracid which is later used to oxidise ketones to lactones. The former reaction can
effectively be catalysed by lipases, and Candida Antarctica lipase B immobilized on
acrylic resin (commercially available as Novozyme-435), owing to its excellent stability in
organic solvents, has been frequently utilised for this purpose with a considerable success.
Authors of a pioneer work on lipase-mediated perhydrolysis demonstrated that oxidation
of cyclic ketones in toluene with stoichiometric amounts of myristic acid and a large (four-
fold) excess of 30% aq. hydrogen peroxide gave lactones in high yields at room
temperature after six days [10]. Shortly thereafter, another group employed an
autocatalytic concept and applied UHP as an anhydrous form of hydrogen peroxide for the
lipase-catalysed BVO [11]. Lactone synthesized on this way in the first step of the process
could undergo further lipase-mediated perhydrolysis to a peracid, which in turn was used
as oxidant in BVO of another molecule of the ketone. Using this strategy, 2-substituted
cyclohexanones were oxidised with ten-fold excess of UHP at room temperature to obtain
20-50% yields of lactones in two days [11]. A clear advantage of that method was the lack
2
Page 2 of 29

of both the carboxylic acid and the solvent. An interesting modification of the chemo-
enzymatic method that utilised ethyl acetate as solvent and a peracetic acid precursor was
proposed more recently [12, 13]; 2-methylcyclohexanone was oxidised with UHP and
Novozyme-435 as a catalyst at room temperature to obtain corresponding lactone with
95% yield in three days [12]. The most recent report on the chemo-enzymatic BVO
demonstrated that oxidation of ketones by peracid generated in situ in the reaction system
could be significantly accelerated by the presence of hydrogen-bond-donating ionic
liquids [14]. Utilizing this concept, cyclohexanone was oxidised to -caprolactone at 50 °C
with 62% yield in just 5 h [14].
To summarize, the chemo-enzymatic BVO is still under intensive studies for its usefulness
in biotransformations. However, the fairly long reaction times required, sometimes poor
stability of lipase in the presence of H
2 2
O and the modest yields resulting from this method
still remain major challenges to overcome.
Herein, we report results of studies of the chemo-enzymatic BVO of cyclic ketones using
Candida Antarctica lipase B. But contrary to Novozyme-435, the lipase was attached not
to a polymeric support but to porous silica materials that possess multi-modal hierarchical
pore structure (labelled MH), and the surface of which was additionally modified with
organosilanes to impart hydrophobic properties. We compared their performance with the
corresponding biocatalysts obtained using the most typical mesoporous silica (MPS)
materials of SBA-15 type, considered as a reference siliceous enzyme carrier, and also
with those obtained using Novozyme-435, applied as benchmark. These comparisons
should shed light on potentials of the new biocatalysts in BVO reaction.
An operational stability of native enzymes is rather limited and immobilization is very
often applied to ease this constraint. In consequence, the performance of heterogeneous
biocatalyst depends on both the enzyme carrier and the immobilization [15, 16].
Therefore, to obtain effective biocatalyst both factors need to be understood and taken into
account. A development of strategies that may improve stability of the enzymes in the
presence of hydrogen peroxide was extensively addressed in [15, 16, 17-23].
Lipases present their specific catalytic mechanisms of action in two structural forms. The
first called close features a polypeptide chain (lid or flat) which isolates an active centre
from the reaction medium and that hinders its chemical activity. In the second open form
of lipases the lid moves and an active centre is exposed to reactants [24]. Under the
favourable conditions, lipases can be immobilized via the hydrophobic surrounding of
3
Page 3 of 29

their active centre, fixing them in an open conformation. Interfacial activation on
hydrophobic supports at low ionic strength has been reported to be a simple and efficient
method to purify and immobilize lipases [24, 25, 26] on various supports: octadecyl–
Sepabeads [27], agarose (octyl-, butyl- and hexyl-agarose [24, 26]), nano-sized magnetite
particles [28], and more recently also on mesoporous organosilica, and methyl- and octyl-
modified MPS [25, 29].
On the whole, MPSs and those of SBA-15 and MCF type in particular, were more recently
reported to be excellent enzyme supports, to give biocatalysts of activity notably superior
to the most conventional polymeric enzyme carriers [30, 31]. That stems from their very
high mechanical, chemical and biological resistance, open structure facilitating access of
reactants to the active centres, large surface area that can easily be modified with various
chemical groups favourable for expressing enzyme activity, and also the ability to tune
pore sizes to those of enzyme molecules. More recently we demonstrated large potentials
of MH-type silica materials as ionic liquids carriers to afford very active catalysts [32].
Consequently, we deemed it important to test their potentials also as lipase/enzyme
carriers in the BVO reaction, considered as an exemplary enzyme catalysed reaction of
major importance.
2.
Materials and Methods
2.1
Materials
All cyclic ketones used in this study, UHP, n-decane and ethyl acetate were purchased
from Acros Organics, H (60% aqueous) was obtained by distillation of 30% H in
the presence of stabilisers Na , H PO , and 8-hydroxychinoline. Novozyme-435 and
2
O
2
2 2
O
4
P
2
O
7
3
4
native Candida Antarcticalipase B were a kind gift of Novozymes (Denmark). Silica with
multi-modal pore structure (MH) was synthesised according to the procedures described in
[32], and SBA-15 was obtained by the method proposed by Zhao et al. [33].
Tetraethoxysilane (98%), Pluronic P123, polyethylene glycol 35000 and
cetyltrimethylammonium bromide, used in the synthesis of SBA-15 and MH materials,
were purchased from Sigma-Aldrich, as well as hexadecyltrimethoxysilane.
Hexamethyldisilazane was purchased from Alfa Aesar, n-octyltriethoxysilane from ABCR
4
and HAuCl was from POCH (Poland).
4
Page 4 of 29

2.2
Methods
Chemical functionalisation of MH and SBA-15
Before grafting, both silica materials (SBA-15, MH) were exposed to water vapour for 5 h
and then dried at 200 ºC for 2 h. Organic hydrophobic groups: methyl (Me), octyl (Oc) and
hexadecyl (Hd) were grafted onto silica particles of SBA-15 and powdered MHs by
reacting suitable amounts of n-octyltriethoxysilane, hexamethyldisilazane and
hexadecyltrimethoxysilane, respectively, dissolved in toluene under reflux (24 h, 40 ºC)
with silanols present on the silica surface. Specifically, 50 mL of the solution containing
1
2
.5 mmol of the organic precursor was stirred under reflux with 1 g of SBA-15 or MH for
4 h, followed by solvent evaporation at 60 ºC.
Immobilization of lipase on functionalised MH and SBA-15
The functionalised materials (0.7 g) were rinsed by centrifugation (6500 rpm, 15 min) with
ethanol, distilled water and 0.1 M phosphate buffer (pH 7.0), then 6 mL of native Candida
Antarctica lipase B solution in 0.1 M phosphate buffer (pH 7.0) was added. The slurries
were mixed for 3 h and stored overnight at 4 ºC, whereupon the excess proteins were
washed off with 50 mL of 0.1 M phosphate buffer (pH 7.0). Finally the biocatalysts
obtained were dried at 4 ºC in a vacuum.
Deposition of gold nano-particles onto biocatalyst
One gram of the biocatalyst sample was dispersed in 100 mL of an aqueous solution of
-
1
HAuCl
NaBH
4
(0.2 g L ) and stirred in the dark for 20 min. Then, 2.0 mL of 0.1 M solution of
4
was added, and after 2 h mixing at room temperature, the suspension was filtered
and washed with water.
General procedures for BVO with the biocatalysts and activity assays
A solution of ketone (0.5 mmol) in ethyl acetate (1.5 mL) and n-decane (0.5 mmol,
internal standard) was introduced into the 25-mL round-bottom flask, whereupon the
biocatalyst (0.005 – 0.03 g) and UHP (or 60% aqueous H
2
O
2
) 1 mmol were added. The
o
flask was sealed with septum and mixed in a thermostat shaker (0.5 C) with orbital
stirring of 180 rpm at 25-70 °C for 29-50 h, depending of the reaction rate. Periodically,
during the reaction, 100 μl of the samples diluted with ethyl acetate were collected to
monitor progress of the reaction utilizing GC. When the reaction was completed, the
catalyst was filtered off and washed with ethyl acetate (3 x 5 mL). The filtrate was washed
3 4
with 5 mL of a 10% NaHCO solution in water, dried over anhydrous MgSO and
concentrated in vacuum. The yields of the lactones after purification by column
5
Page 5 of 29

chromatography (with hexane:ethyl acetate ratio of 4:1 as an eluent) were in the range of
5-99%.
8
Recycling of biocatalyst
In experiments with the biocatalyst recycle, the reactions were scaled up by factor of five.
After reaction completion, the biocatalysts were filtered off, washed with ethyl acetate and
dried under vacuum.
2.3
Instrumentation
Nitrogen adsorption/desorption isotherms from the silica materials were obtained using
a Micrometrics ASAP 2020 instrument at –196 °C to calculate the specific surface area
(SBET), pore volume (VpN2) and their average size (d
m
). The size of the pores was obtained
using Barret–Joyner–Halenda (BJH) method with Kruk-Jaroniec-Sayari correction. Prior
-3
to experiments the samples were outgassed at 150 °C and 1.33 x 10 Pa for 5 h.
The amounts of functional groups and enzymes attached into the silica matrices were
determined by a thermogravimetric method. The measurements were performed using
a Mettler Toledo STAR851 thermobalance. Samples of approximately 20 mg were heated
at a rate of 10 K/min in a standard 150 µl Pt crucibles under an air flow of 50 mL/min.
TG, DTG and SDTA curves were recorded.
A Thermo Scientific FTIR Nicolet 6700 spectrometer was used for the spectroscopic
studies of functionalised samples. The spectra of KBr pellets (diameter 13 mm) were
-1
-1
recorded in the range of 4000 – 600 cm with a resolution of 4 cm . The pellets were
prepared using 748 mg of KBr (spectroscopic grade) and 2 mg of sample.
To evaluate the surface hydrophobicity, MH samples were pressed into flat plates under
6
MPa pressure, followed by the contact angle measurement using a homemade device
equipped with a digital camera (MDA 1300).
Scanning electron microscopy (SEM) imaging was made on an JOEL JSM-6480. High-
resolution transmission electron microscopy (HRTEM) was performed using a FEI Titan
80-300 to identify gold nano-particles distribution on the biocatalyst surface.
1
13
H NMR and C NMR spectra of the lactones were recorded at 300 MHz in CDCl
3
(
Varian Unity Inova plus spectrometer with a TMS internal standard). All lactones were
characterised by comparing their NMR spectra with those of authentic samples. GC
analysis was performed using Perkin Elmer Clarus 500 chromatograph
a
6
Page 6 of 29

TM
SUPELCOWAX 10 column (30 m×0.2 mm×0.2μm) with n-decane as an external
standard.
3.
Results and discussion
3.1
Characterisation of the biocatalysts
Table 1 lists the properties of pore structure of pristine and functionalised materials as well
as of biocatalysts obtained from nitrogen adsorption experiments. It appeared that surface
area of MH samples was about half the value of the SBA-15 counterparts, but strikingly
enough, the values of mesopore volume and average pore diameter were roughly two
times larger in the functionalised MH samples than the corresponding SBA-15s. After
crushing the monoliths to obtain particles of ca. 10-30 μm size the continuous structure of
macropores (diameter 20 - 50 µm) seen in the original materials was effectively destroyed
and only small cavities on surface of the particles could be observed (cf. Supplementary
data). Both small sizes of MH particles (short diffusion pathway) and their extended
external surface area may facilitate transport of the reactants to and from the active sites.
The attachment of hydrophobic groups was confirmed by both FTIR and
thermogravimetric measurements. Figure 1 shows the FTIR spectra of SBA-15 material in
its pristine form and after hydrophobisation.
Fig. 1. FTIR spectra of the SBA-15 material and after its hydrophobisation with methyl,
octyl and hexadecyl groups (for the sake of clearness the spectrum of SBA-15 was
shifted up 60%).
-1
All spectra showed a broad adsorption band from 3100 to 3600 cm derived from
-1
different O-H groups. The bands in the range of 1250-1000 cm are ascribed to the Si-O-
-1
Si asymmetric stretching vibrations, while those at 803.3 and 960 cm indicate Si-O-Si
bending vibrations and Si-OH stretching vibrations, respectively. After hydrophobisation,
the characteristic bands of methyl and methylene groups were observed in the range of
-1
-1
2
850 to 3000 cm . For the SBA-Me sample, a band at 2964.7 cm is observed as a result
of asymmetric stretching vibrations of the methyl groups. Strong absorption asymmetric
-1
-1
stretching bands at 2926.5 cm and symmetric stretching bands at 2855.2 cm of
methylene groups were observed for the SBA-Hd sample. In the case of the SBA-Os
-1
-1
sample, these bands were present at 2927.7 cm and 2865.4 cm , respectively. In both
7
Page 7 of 29

-1
spectra, a weak band at 2959 cm was observed due to the asymmetric stretching
vibrations of terminal methyl groups. The intensities of absorption bands decreased in the
following order: Hd>Oc>Me, which correlates quite well with the wt% content of
functionalities determined by thermogravimetry (Table 1). Similar relationships were also
found for the corresponding MH materials.
The changes in sample masses recorded during TG analysis are shown in Figures 2 and 3.
The amount of functionalities was recalculated to obtain the values expressed in moles per
gram, and the highest ones of 1.2 and 2.0 were found for MH-Me and SBA-Me,
respectively. In the Hd and Oc series, the concentration was ca. two times lower.
However, for Hd materials, the amount was larger than for Oc series, owing to higher
activity of the organic group precursor containing more reactive methoxy groups. The
nominal content of precursors used for modification was the same in all cases, 1.5 mmol
per 1 g of silica, however, as 1 mol of methyl group precursors (hexamethyldisilazane)
contains six Me entities thus the hydrophobic efficiency (molar ratio of attached groups to
the nominal content of the precursors used for modification) was only 13% and 33% for
the SBA-Me and MH-Me materials, respectively (Table 1).
Compared to SBA-15s the MH samples possessed slightly less hydrophobic groups per 1
g of silica, due to smaller specific surface area and hence smaller specific number of
hydroxyls. The surface concentration of hydroxyl groups, determined as described in [34],
2
was similar in both materials and about 3 OH per nm .
FTIR spectroscopy and thermogravimetry analysis were also used to study proteins’
attachment, whereas their distribution on the silica supports was evaluated by inspecting
the TEM images of Au-NPs (acting as an electron-dense contrast agent [35]) distribution
in the modified biocatalyst.
Fig. 2. TG curves from the functionalised SBA-15s after lipase attachment and from
SBA-Me-L-R (recycled 5 times).
The major protein absorption bands typically seen in FTIR spectra are the amide I (1700 –
-1
-1
1
600 cm ), due to the C=O stretching vibrations, and amide II (1580 – 1510 cm ), due to
the N-H bending vibrations and C-N stretching vibration (see Supplementary data). Fig. 4
-1
shows the spectral range for the MH samples. The position of amides I (1650 cm ) and II
-1
(
1550 cm ) indicated that the immobilised lipases still maintained their native secondary
8
Page 8 of 29

-1
structure [36]. Additionally, the intensity of bands in the range of 3700 – 2800 cm was
increased due to the absorption of the N-H groups (stretching bands in the range of 3600 –
-1
-1
3
200 cm ) and methylene groups (2927 and 2857 cm ) (Fig. 4).
Fig. 3. TG curves from the functionalised MHs before and after lipase attachment.
Fig. 4. FTIR spectra from the functionalised MHs before and after lipase attachment.
The presence and uniform distribution of proteins over the biocatalysts surface were also
deduced from TEM images of Au-nanoparticles adsorbed onto amino groups of the
proteins (cf. Fig. 5).
As was expected, the enzyme load appeared to depend on pore structure of silica support
and the type of hydrophobic groups attached to their surfaces. Higher protein loads were
found for MHs (Table 1) and that could easily be attributed to their more open structure
and larger porosity (Table 1). For both silica supports a strong relationship between
hydrophobicity of functional groups and the amount of immobilized lipase was observed.
That amount decreased in the order Hd>Oc>Me, which very well correlates with decrease
in hydrophobicity of the functionalities and also with surface hydrophobicity of the
supports. The latter property could clearly be quantified by the wetting angles of water,
o
o
o
which for Hd-, Oc- and Me-modified MHs were found equal to 132 , 118 and 102 ,
respectively. The largest lipase load of about 198 mg/g of support was observed in MH-Hd
material.
Fig. 5. TEM image of Au-NPs on the SBA-Me-L biocatalyst
3.2 Performance of the biocatalysts in the chemo-enzymatic Baeyer-Villiger reaction
As can be seen from Scheme 1 the applied chemo-enzymatic method of lactone synthesis
involved immobilized lipase as the biocatalyst of peracid formation, urea hydrogen
peroxide (UHP) as oxidant and ethyl acetate as both the peracid precursor and solvent.
Making use of UHP and a heterogeneous biocatalyst we aimed to develop a clean BVO
process with a facile catalyst separation and reuse.
Already a superficial inspection of Scheme 1 indicates that a direct evaluation of the
biocatalysts’ activities from the rates of peracid formation could be somewhat difficult. To
tackle that problem we decided to assess their performance from the rates of lactone
9
Page 9 of 29

synthesis (BVO process), and more specifically, using those from the region where the
BVO kinetics is controlled by the enzyme-catalysed reaction.
Scheme 1.
Chemo-enzymatic Baeyer-Villiger oxidation of 2-methylcyclohexanone.
Therefore, the preliminary studies aimed at discriminating the regions of specific kinetic
control of the BVO process. That was made by varying amount of the most heavily
lipase-laden biocatalyst (MH-Hd-L, cf. Table 1) in the range of 5 – 30 mg, while keeping
constant (0.5 mmol) the amount of a model ketone (2-methylcyclohexanone). At a fixed
molar ratio of ketone : UHP (1:2) and an excess of ethyl acetate the rate of lactone
formation at RT appeared to be linearly dependent (pseudo-first order kinetics, correlation
coefficients: 0.997-0.999) on the biocatalyst content if it was in amount ≤ 10 mg, and it
remained constant for larger values. More specifically, the respective apparent rate
-1
constants, kobs, increased linearly from 0.06 to 0.12 h , with increase in the catalyst’s
-1
contents from 5 to 10 mg, respectively, and was ca. 0.12 h at its larger values. These
experiments clearly showed that in the system under study, synthesis of peracetic acid
with over 10 mg of MH-Hd-L had no effect on the BVO kinetics, in strong contrast to the
situation where its amount was less than 10 mg. Those findings clearly delineated the
region of effective control of the BVO process by the reaction of peracid synthesis and
hence by the biocatalyst itself.
For this reason, to compare activity of MH-Hd-L with that of Novozym-435 10 mg of
each catalysts per 0.5 mmol of the model ketone were used, whereas a commercial native
lipase preparation was applied in larger (0.5 g) amount. The experiments were performed
under the reaction conditions recommended for Novozym-435 [12]. Results of these
studies, displayed in Figure 6, clearly show that the native lipase preparation was the least
active, and it can be regarded as a strong, but not unexpected, argument for application of
the immobilized lipase preparations in the BVO processes.
Fig. 6. Comparative study of MH-Hd-L, Novozyme-435 and a native lipase performance
in the BV oxidation of 2-methylcyclohexanone (0.5 mmol) with UHP (1 mmol) in
ethyl acetate (1.5 mL) at RT.
1
0
Page 10 of 29

-1
But more importantly, the calculated value of the apparent rate constant (0.12 h ) for the
BVO reaction with MH-Hd-L appeared to be ca. three times larger then the value
-1
determined for Novozyme-435 (ca. 0.04 h ), under otherwise identical conditions. Thus in
the latter case the kinetics of BVO appeared to be severely constrained by the rate of
peracid formation, and that was not the case when the same amount of MH catalysts was
applied. In practical terms, the yield of lactone of about 99% was obtained within 29 h
using MH-Hd-L, whereas it took 50 h to achieve yield of slightly over 90% using
Novozyme-435 (cf. Figure 6). Those results also clearly demonstrate a superior efficiency
of the MH-bound lipase against the benchmark catalyst.
The comparative experiments performed also showed that an average reaction time of 29
h, necessary to achieve a complete conversion of ketone using UHP, could be reduced to
2 2
21 h when the same amount of 60% aq. H O was applied.
Next, using the commonly accepted methodology of studying the stability of solid
catalysts in liquid phase oxidations, we investigated the leaching of proteins from MH-Hd-
L catalyst. The experiments with fast catalyst filtration performed at the reaction
temperature demonstrated no further 2-methylcyclohexanone conversion in the filtrate
after catalyst’s removal (see Supplementary data). This provides a strong evidence that
MH-Hd-L operates as truly heterogeneous catalyst, even if minute amounts of the lipase
are leached from the solid matrix during the oxidation process.
At this stage we also checked whether a twofold molar ratio of oxidant (UHP) to ketone
was large enough to ensure an optimum kinetics. For that we increased the value of this
-1
ratio to 3:1 and it appeared to have no effect on the value of kobs (0.12 h ).
The comparative studies of SBA-15 and MH biocatalyst were carried by investigating
respective rates of the BVO reaction in the systems with different catalyst contents but the
same total mass of an active compound (lipase), equal to that present in 5 mg of MH-Me-L
i.e. 0.4 mg (cf. Table 1). Using that fairly low lipase content we aimed to carry out activity
studies deeply in a region of the biocatalyst-control of the BVO process to be able to show
even minor differences in the catalysts activities. From Figure 7 it can be seen that, on the
whole, MH biocatalysts appeared to be much more active (larger kobs) than those of SBA-
15 series and also that activation of silica surface with various hydrophobic groups had
appreciable effect on the catalytic properties, especially for MH biocatalysts. To this end,
specific activity of the biocatalysts appeared to decrease in the order Me>Oc>Hd, i.e.
reverse to hydrophobic properties of the surface reported earlier (vide supra) and also to
1
1
Page 11 of 29

protein load in the biocatalysts (cf. Table 1). It is noteworthy that specific activity of MH-
Me-L appeared to be even about 50% higher than of MH-Hd-L.
Fig. 7. The values of apparent rate constant kobs of BVO of 2-methylcyclohexanone (0.5
mmol) with UHP (1 mmol) in ethyl acetate (1.5 mL) at RT using MH (black) and
SBA-15 (grey) biocatalysts in amounts calculated to give 0.4 mg of lipase input.
A significant difference in efficacy of MHs and SBA-15s can easily be explained by large
differences in their pore structure, and ascribed to the presence of mesopores significantly
larger in MHs (cf. Table 1) than SBA-15s, and also in larger quantities (as can be deduced
by comparing the value of VpN2). Both these factors contribute to much more open
structure of MHs and hence to larger activities of the biocatalysts. Firstly by facilitating
lipase molecules penetration into pore structure during immobilization, that resulted in
larger protein loads (Table 1) and more homogeneous enzyme distributions, and secondly
by easing diffusion constraints for transport of reactants to the active sites of enzymes,
also directly boosting biocatalysts’ activities. A similar trend was also observed before
when mesporous silica of MCF type (with very open mesopore structure) and SBA-15
were applied as enzyme supports [37].
The reason for a minor difference in activity of biocatalysts of the same series (MH or
SBA), yet with different functionalities may be that Candida Antarctica lipase B does not
express “interfacial activation”; indeed, the putative “lid” above the active site cleft is
short and disordered and it is likely that it cannot cover the active site [38]. The highest
specific activity of the catalysts functionalised with methyl groups can be explained by
a ‘reasonable’ load of enzyme (of few wt%), i.e. less than in the catalysts with more
hydrophobic surfaces (Hd- and Oc-). In heavily loaded biocatalysts the enzyme molecules
often tend to aggregate and that impedes an access of reactants to their active sites, with
a negative effect on their specific activity. That has already been observed quite frequently
before.
The results of catalyst activity and reusability studies performed for 2-
2 2
methylcyclohexanone oxidation with 60% aq. H O over five successive runs showed very
high but slowly declining catalysts activity (cf. Figure 8), most likely caused by a slow
detachment and washing out of the enzymes. The latter supposition was confirmed by a
meticulous TGA analysis of the biocatalysts applied, which indicated a 6.5% loss in
1
2
Page 12 of 29

2 2
protein load after five cycles. It is noteworthy, that the idea of using 60% aq. H O , which
originated from two literature reports [39, 40], in fact had two objectives and aimed at
practical probing of the biocatalysts performance under harsh reaction conditions, and not
so much the BVO process itself. Indeed, according to one report [39], the exposure of
lipase to high concentration of aq. hydrogen peroxide resulted in its complete deactivation,
whereas according to the other one [40] activity of Novozyme-435 was lost after the
second cycle of epoxidation reaction with 50% aq. H
2
O
2
. Therefore, by demonstrating that
and can be reused at
new biocatalysts are highly active even in the presence of 60% H
2
O
2
least five times with relatively small loss in protein load and activity, we documented and
proved their quite exceptional catalytic properties.
Fig. 8. The effect of MH-Me-L (50 mg) reuse on rate constant (kobs - gray) and yield
(black) of BV oxidation of 2-methylcyclohexanone (2.5 mmol) with 60% aq.
2 2
H O
(5 mol) in ethyl acetate (7.5 mL) at RT.
Finally, the most active biocatalysts (MH-Me-L) was examined in the BVO synthesis of
various lactones from the corresponding ketones to determine its practical potentials. As
can be seen from Table 2 the BVO processes proceeded very efficiently; in fact, cyclic
ketones were readily oxidised to their corresponding lactones in high yields (35-98%)
under our mild conditions within reasonable reaction times. As model reactants, we chose
strained ketones, such as cyclobutanone and cyclopentanone, which have long been known
to form lactones in high yields, and the non-strained ketone cyclohexanone, which is much
more difficult to oxidise. The most reactive of these ketones appeared to be
cyclobutanone, which oxidised to -butyrolactone with 98% yield in 2 hours, while
cyclohexanone gave 80% -caprolactone after 75 hours. Cyclobutanone is very reactive
molecule because its ring strain is very high, and hence its very fast oxidation does not
come as surprise. The oxidation of 2-adamantanone and norcamphor yielded their
corresponding lactones also in high yields. Unfortunately, the oxidation of cyclopentanone
led to the formation of -valerolactone in the first step, which subsequently hydrolysed to
the corresponding hydroxyacid.
To sum up, the biocatalysts proposed in this study are attractive for application in the
chemo-enzymatic BVO reaction, comparing with Novozyme-435 as well as other standard
catalysts, e. g. Sn-Beta or Sn-MCM-41 (cf. comparison of the catalysts performance given
1
3
Page 13 of 29

in Supplementary data). Specific activity of the biocatalysts presented here is significantly
larger than that of Novozyme-435. On the other hand the application of these biocatalysts
require longer reaction times to achieve the same conversions of cyclohexanones,
compared with Sn-Beta catalyst. However, the biocatalytic reactions are carried out at
o
room temperatures, whereas standard catalysts are applied at 56-90 C.
Table 2
Oxidation of selected cyclic ketones (1 mmol) to lactones with UHP (2 mmol) in the
presence of 20 mg of MH-Me-L in ethyl acetate (3 mL) at RT.
4.
Conclusions
The devised biocatalysts, engineered by supporting the most conventional lipase on silica
materials exhibiting very open, bi-modal mesopore structure and purposely
hydrophobized surface, show most important advantages: very high activity, good stability
and reusability demonstrated in the chemo-enzymatic Baeyer-Villiger oxidation of various
cyclic ketons to the corresponding lactones.
Specific activity of biocatalysts functionalised with different groups decreased in the order
Me>Oc>Hd, and that of biocatalyst grafted with hexadecyl groups for the same mass of
biocatalyts appeared to be at least trice larger than of Novozyme-435 considered as
benchmark. Activities of all MH biocatalysts were also significantly larger than of the
corresponding biocatalysts fabricated using mesoporous silica of SBA-15 type, considered
as a reference silica enzyme support. That very high activity of the newly devised
biocatalysts could be attributed to the effective elimination of strong diffusion constraints
of the reaction kinetics, typically prevailing in liquid reaction systems.
The proposed environmentally benign chemo-enzymatic BVO process avoids the use of a
relatively unstable peracid, which in the proposed method is generated in situ in the
reaction system during the enzymatic stage, and the rate of its synthesis, affecting the rate
of the BVO process, could be effectively controlled by the applied biocatalysts.
Acknowledgments
This work was financed by the Ministry of Science and Higher Education (Grant no. N
N209 021739) and by the EU European Regional Development Fund supported project
POIG01.03.01-00-158/09-00.
1
4
Page 14 of 29

References
[
[
[
[
1]
2]
3]
4]
G. Krow, Org. React. 43 (1993) 251-798.
M. Renz, B. Meunier, Eur. J. Org. Chem. (1999) 737-750.
G. Brink, I. Arends, R. Sheldon, Chem. Rev. 104 (2004) 4105-4123.
D. E, Torres Pazmiño, R. Snajdrova, B-J. Baas, M. Ghobrial, M. D. Mihovilovic,
M. W. Fraaije, Angew. Chem. Int. Ed. Engl. 120 (2008) 2307-2310.
A. Corma, L. T. Nemeth, M. Renz, S. Valencia, Nature 412 (2001) 423-425.
M. Renz, T. Blasco, A. Corma, V. Fornes, R. Jensen, L. Nemeth, Chem. Eur. J. 8
[
5]
6]
[
(2002) 4708-4717.
[
[
[
[
7]
8]
9]
A. Corma, M. T. Navarro, L. Nemeth, M. Renz, Chem. Comm. (2001) 2190-2191
A. Corma, M. T. Navarro, M. Renz, A. Corma, J. Catal. 219 (2003) 242-246.
I. Nowak, A. Feliczak, I. Nekoksova, J. Cejka, Appl. Catal. A 321 (2007) 40-48.
10] S. C. Lemoult, P. F. Richardson, S. M. Roberts, J. Chem. Soc., Perkin Trans. 1
1995) 89-91.
11] B. K. Pchelka, M. Gelo-Pujic, E. Guibé-Jampel, J. Chem. Soc., Perkin Trans. 1
1998) 2625-2627.
(
[
(
[12] M. Y. Rios, E. Salazar, H. F. Olivo, Green Chem. 9 (2007) 459-462.
[13] M. Y. Rios, E. Salazar, H. F. Olivo, J. Mol. Catal. B 54 (2008) 61-66.
[14] A. J. Kotlewska, F. van Rantwijk, R. A. Sheldon, I. W. C. E. Arends, Green Chem.
1
3 (2011) 2154-2160.
[
15] C. T. Hou, Handbook of Industrial Biocatalysis, 2005 New York, NY:
Taylor&Francis/CRC.
[
16] K. Buchholz, V. Kasche, U. T. Bornscheuer, Biocatalysts and Enzyme
Technology, 2012 Weinheim, WILEY-VCH Verlag GmbH & Co. KGaA.
[
17] K. Hernandez, A. Berenguer-Murcia, R. C. Rodrigues, R. Fernandez-Lafuente,
Curr. Org. Chem. 16 (2012) 2687-2707.
[
[
[
18] U. Törnvall, M. Hedström, K. Schillén, R. Hatti-Kaul, Biochimie 92 (2010) 1867-
1
875.
19] U. Törnvall, C. Orellana-Coca, R. Hatti-Kaul, D. Adlercreutz, Enzyme Microb.
Technol. 40 (2007) 40, 447-451.
20] U. Törnvall, C. M. Fürst, R. Hatti-Kaul, M. Hedström, Rapid Commun. Mass
Spectrom. 23 (2009) 2959-2964.
1
5
Page 15 of 29

[21] A. A. Tzialla, I. V. Pavlidis, M. P. Felicissimo, P. Rudolf, D. Gournis, H. Stamatis,
Bioresour. Technol. 101 (2010) 1587-1594.
[
22] K. Hernandez, R. Fernandez-Lafuente, Process Biochem. 46 (2011) 873-878.
23] K. Hernandez, C. Garcia-Galan, R. Fernandez-Lafuente, Enzyme Microb. Technol.
[
4
9 (2011) 72-78.
[
[
[
[
[
[
[
24] G. Fernàndez-Lorente, Z. Cabrera, C. Godoy, R. Fernandez-Lafuente, J. M.
Palomo, J. M. Guisan, Proc. Biochem. 43 (2008) 1061-1067.
25] R. M. Blanco, P. Terreros, M. Fernández-Pérez, C. Otero, G. Díaz-González, J.
Mol. Catal. B 30 (2004) 83-93.
26] G. Fernàndez-Lorente, J. M. Palomo, Z. Cabrera, J. M. Guisàn, R. Fernàndez-
Lafuente, Enzyme Microb. Tech. 41 (2007) 565-569.
27] J. M. Palomo, G. Muñoz, G. Fernández-Lorente, C. Mateo, R. Fernández-Lafuente,
J. M. Guisán, J. Mol. Catal. B 19–20 (2002) 279-286.
28] D.G. Lee, K.M. Ponvel, M. Kim, S. Hwang, I.S. Ahn, C.H. Lee, J. Mol. Catal. B
5
7 (2009) 62-66.
29] E. Serra, E. Díez, I. Díaz, R. M. Blanco, Micropor. Mesopor. Mater. 132 (2010)
87-493.
4
20] K. Szymańska, J. Bryjak, J. Mrowiec-Białoń, A. B. Jarzębski, Micropor. Mesopor.
Mater. 99 (2007) 167-175.
[31] S. Hadson, J. Cooney, E. Magner, Angew. Chem. Int. Ed. 47 (2008) 8582-8594.
[32] A. Chrobok, S. Baj, W. Pudło, A. Jarzębski, Appl. Catal. A 366 (2009) 22-28.
[33] D. Zhao, Q. Huo, J. Feng, B. F. Chmelka, G. D. Stucky, J. Am. Chem. Soc. 120
(1998) 6024-6036.
[34] J. Mrowiec-Białoń, Thermochim. Acta 443 (2006) 49-52.
[
35] C. Lei, Y Shin, J. K. Magnuson, G. Fryxell, L. L. Lasure, D. C. Elliott, J. Liu, E. J.
Ackerman, Nanotechnology 17 (2006) 5531-5538.
[36] A. Natalello, D. Ami, S. Brocca, M. Lotti, S. M. Doglia, Biochem. J. 385 (2005)
5
11-517.
[37] K. Szymańska, J. Bryjak, A. B. Jarzębski, Top. Catal. 52 (2009) 1030-1036.
[38] Q. Jin, G. Jia, Y. Zhang, Q. Yang, C. Li, Langmuir 27 (2011) 12016-12024.
[39] F. Bjorkling, S. E. Godtfredsen, O. Kirk, J. Chem. Soc., Chem. Comm. (1990)
1
301–1303
[40] E. G. Ankudey, H. F. Olivo, T. L. Peeples, Green Chem. 8 (2006) 923–926
1
6
Page 16 of 29

Scheme 1.
Chemo-enzymatic Baeyer-Villiger oxidation of 2-methylcyclohexanone.
O
O
CH3
O
CH3
O
O
H
3
C
OOH
H
C
OH
3
O
H O
2
immobilized lipase
UHP
EtOH
H C
OEt
3
1
7
Page 17 of 29

Table 1
Properties of the materials.
Contents of
hydrophobic
groups wt%
Hydrophobic
efficiency
S
BET
-1
VpN2
d
m
Contents of
No. Support
2
3
-1
-1 a
[
m g ] [cm g ]
[nm]
lipase [mgg ]
[%]
-1
[
mmol g ]
1
4.0
MH
320
157
283
267
1.25
0.97
1.13
1.2
-
-
-
28.5
2
2.5
MH-Hd
14.2 (0.63)
6.1 (0.54)
1.8 (1.2)
42
36
13
198
116
80
28.3
3
3.0
MH-Oc
28.2
4
3.3
MH-Me
2
7.7
8.2
6.8
7.2
6.8
5
6
7
8
SBA-15
SBA-Hd
SBA-Oc
SBA-Me
750
310
450
396
0.9
-
-
-
0.52
0.66
0.62
15.8 (0.7)
5.9 (0.52)
3.1 (2.0)
47
35
22
142
99
75
a
in functionalized support
1
8
Page 18 of 29

Table 2
Oxidation of selected cyclic ketones (1 mmol) to lactones with UHP (2 mmol) in the
presence of 20 mg of MH-Me-L in ethyl acetate (3 mL) at RT.
a
Time
h]
Yield
[%]
Entry Ketone
Lactone
[
O
7
5
80
O
O
1
2
O
O
O
2
7
9
5
97
82
3
4
O
O
O
6
8
92
94
O
O
O
5
O
O
O
O
O
6
7
O
2
6
98
O
O
O
b
O
35
O
8
O
c
9
6
7
a
isolated yields after column chromatography
b
NMR analysis of post-reaction mixture confirmed the formation of corresponding
hydroxyacid which was formed from lactone hydrolysis reaction
c
yield determined by GC
1
9
Page 19 of 29

Fig. 1. FTIR spectra of the SBA-15 material and after its hydrophobisation with methyl,
octyl and hexadecyl groups (for the sake of clearness spectrum of the SBA-15 was
shifted up 60%).
1
1
1
40
20
00
SBA-15
SBA-Me
80
60
40
20
0
SBA-Oc
SBA-Hd
4
000
3500
3000
2500
2000
1500
1000
500
-1
Wavenumber, cm
Fig. 2. TG curves from the functionalised SBA-15s after lipase attachment and from
SBA-Me-L-R (recycled 5 times).
2
0
Page 20 of 29

1
1
1
00
SBA-Me
SBA-Me-L-R
90
80
70
SBA-Me-L
00
90
80
70
SBA-Hd
SBA-Hd-L
00
90
80
70
SBA-Oc
SBA-Oc-L
1
00 200 300 400 500 600 700 800
o
Temperature [ C]
Fig. 3. TG curves from the functionalised MHs before and after lipase attachment.
2
1
Page 21 of 29

100
MH-Me
95
90
85
MH-Me-L
100
MH-Oc
9
0
0
MH-Oc-L
8
100
90
80
70
MH-Hd
MH-Hd-L
100 200 300 400 500 600 700 800
o
Temperature [ C]
Fig. 4. FTIR spectra from the functionalised MHs before and after lipase attachment.
2
2
Page 22 of 29

1
1
1
00
MH-Me
8
6
4
2
0
0
0
0
0
MH-Me-L
MH-Oc
00
8
6
4
2
0
0
0
0
0
MH-Oc-L
00
MH-Hd
80
60
40
20
0
MH-Hd-L
4000 3800 3600 3400 3200 3000 2800
-
1
Wavenumber [cm ]
Fig. 5. TEM image of Au-NPs on the SBA-Me-L biocatalyst.
2
3
Page 23 of 29

nm
2
4
Page 24 of 29

Fig. 6. Comparative study of MH-Hd-L, Novozyme-435 and a native lipase performance
in the BV oxidation of 2-methylcyclohexanone (0.5 mmol) with UHP (1 mmol) in
ethyl acetate (1.5 mL) at RT.
100
-1
kobs = 0.12 [h ]
75
50
25
0
-1
kobs = 0.04 [h ]
-1
kobs = 0.02 [h ]
1
1
0
0 mg MH-Hd-L
0 mg Novozyme
.5 g native lipase
0
10
20
30
40
50
Time [h]
2
5
Page 25 of 29

Fig. 7. The values of apparent rate constant kobs of BVO of 2-methylcyclohexanone (0.5
mmol) with UHP (1 mmol) in ethyl acetate (1.5 mL) at RT using MH (black) and
SBA-15 (grey) biocatalysts in amounts calculated to give 0.4 mg of lipase input.
0
0
0
0
0
0
0
,07
,06
,05
,04
,03
,02
,01
0
MH
SBA
Me-L
Oc-L
Hd-L
2
6
Page 26 of 29

Fig. 8. The effect of MH-Me-L (50 mg) reuse on rate constant (kobs - gray) and yield
black) of BV oxidation of 2-methylcyclohexanone (2.5 mmol) with 60% aq.
(5 mol) in ethyl acetate (7.5 mL) at RT.
(
2 2
H O
2
7
Page 27 of 29

Highlights
A novel, very active lipase on silica biocatalyst is proposed.
Silica with multimodal pore structure appears superior to MPS as enzyme support.
The catalysts show very good activity in chemo-enzymatic Baeyer–Villiger oxidation.
Stability and recyclability of the biocatalysts has been demonstrated.
2
8
Page 28 of 29

Page 29 of 29