Lipase-catalyzed acylation in a continuous down-flow fixed-bed reactor

Article abstract of DOI:10.1134/S0023158412060092

The kinetic resolution of racemic 1-phenylethanol with ethyl acetate was investigated in a down-flow fixed-bed reactor operated in a continuous mode mainly at the molar ratio of 1 : 3 in 400 mL toluene at 70°C. The catalytic activity of the immobilized lipase was studied by: (i) changing the flow rates, (ii) utilizing different substrate concentrations, (iii) applying step changes using ethyl acetate, ethyl benzene, acetic acid, acetophenone etc., (iv) investigating the inhibitory effect of either the desired or the stoichiometric products (R)-1-phenylethyl acetate and ethanol, respectively), (v) elucidating the effect of water on the activity and stability of the immobilized lipase. The residence time distribution and the reactor hydrodynamics were also discussed along with kinetic modelling. The results were linked to the one-pot reactions. Pleiades Publishing, Ltd., 2012.

Full text of DOI:10.1134/S0023158412060092

ISSN 0023ꢀ1584, Kinetics and Catalysis, 2012, Vol. 53, No. 6, pp. 673–683. © Pleiades Publishing, Ltd., 2012.
Published in Russian in Kinetika i Kataliz, 2012, Vol. 53, No. 6, pp. 716–727.
LipaseꢀCatalyzed Acylation in a Continuous DownꢀFlow
FixedꢀBed Reactor1
S. Sahin, P. MäkiꢀArvela, M. Kangas, K. Eränen, J. Wärnå, T. Salmi, and D. Yu. Murzin
Process Chemistry Centre, Abo Akademi University, Turku, FIꢀ20500, Finland
eꢀmai: dmurzin@abo.fi
Received August 16, 2011
Abstract—The kinetic resolution of racemic 1ꢀphenylethanol with ethyl acetate was investigated in a downꢀ
flow fixedꢀbed reactor operated in a continuous mode mainly at the molar ratio of 1 : 3 in 400 mL toluene at
70
different substrate concentrations, (iii) applying step changes using ethyl acetate, ethyl benzene, acetic acid,
acetophenone etc., (iv) investigating the inhibitory effect of either the desired or the stoichiometric products
)ꢀ1ꢀphenylethyl acetate and ethanol, respectively), (v) elucidating the effect of water on the activity and
°C. The catalytic activity of the immobilized lipase was studied by: (i) changing the flow rates, (ii) utilizing
(R
stability of the immobilized lipase. The residence time distribution and the reactor hydrodynamics were also
discussed along with kinetic modelling. The results were linked to the oneꢀpot reactions.
DOI: 10.1134/S0023158412060092
INTRODUCTION
aqueous solvents catalyze numerous reactions such as
esterifications, transesterifications, aminolysis, while
on the contrary these processes are completely supꢀ
pressed by hydrolysis in water [12]. Different solvents
can strip off the essential water from the catalyst in difꢀ
ferent ways [13, 14]. Essential water, which is the water
around the enzyme to preserve the 3D structure of the
protein in a catalytically active form, acts as a molecꢀ
ular lubricant to add protein a greater flexibility [13].
Typically enzyme activity decreases with increasing
the polarity of the organic solvent [15]. Solvent hydroꢀ
The preparation of chiral drugs as single enantiꢀ
omers via enantioselective acylation catalyzed by
lipases is widely used in pharmaceutical synthesis [1–3].
Enantiopure compounds are efficiently produced by
lipase catalyzed kinetic resolution of enantiomers givꢀ
ing high regioꢀ and enantioselectivities in mild reacꢀ
tion conditions with multitude of substrates [4–8].
The two enantiomers of a racemic mixture react at difꢀ
ferent rates and one of the enantiomers is then transꢀ
formed to the desired product while the other remains
unchanged. After reaction, the mixture is enriched in
the slowerꢀreacting enantiomer.
phobicity, characterized by the log
the logarithm of the partition coefficient in an octanolꢀ
water system [16]. The solvents with a log < 2 are not
P value, indicates
P
The enantioselectivity is often given as enantioꢀ
favorable for enzymatic systems, because they strongly
distort the waterꢀenzyme interaction, which is essenꢀ
tial for the enzyme activity, whereas solvents with a
logP between 2 and 4 are weak water distorters and can
affect enzymatic activity in an unpredictable manner.
⎛xP − xP ⎞
R
S
meric excess of the products,
Р
(ee =
×
⎜
⎟
xP + xP
⎝
⎠
100) [9], which should be preferable closeRto 100%.
S
kS
R
Biocatalyst remains in active state in solvents with a
Additionally, the enantiomeric ratio
E
=
is given by
kS
log
P > 4, which are not able to distort the waterꢀ
S
the ratio of the rate constants for two enantiomers in
the substrates, S. A value of , which should preferable
enzyme interactions [17].
E
A few hydrolaseꢀcatalyzed enantioselective proꢀ
cesses have been carried out in continuous flow sysꢀ
tems [18–20]. Kinetic resolution of racemic alcohols
and vinyl acetate with the molar ratio of 1 : 2 over an
immobilized lipase was performed in a continuousꢀ
flow reactor at a 0.70 mL min–1 flow rate under
13 MPa of supercritical CO2 (scCO2) [21]. The ee of the
optically active acetates was 99% and the productivity of
the desired product was improved by over 400 times in a
continuousꢀflow reactor compared to the correspondꢀ
ing batch reactions using scCO2 [21, 22]. Continuous
be above 50, can also be calculated from the following
equation:
ln 1 − X 1 + ee(P)
[
(
(
)
)
]
(1)
E =
,
ln 1 − X 1 − ee(P)
[
(
(
)
)
]
in which
X is the conversion [10].
Several organic compounds are waterꢀunstable
and/or waterꢀinsoluble and thus their synthesis is carꢀ
ried out in organic solvents [11]. Enzymes in nonꢀ
1
The article is published in the original.
673

674
SAHIN et al.
L
L
L
immobilized lipase
acyl donor
M
O
M
M
+
( )
O
OH
OH
Fig. 1. Lipase catalyzed kinetic resolution following the Kazlauzkas' rule [6] of secondary alcohols.
HO
CH3
H3C
O
CH3
O
ethyl acetate
H3C
OH
+
immobilized lipase
(
R)ꢀ1ꢀphenylethanol
(
R
)ꢀ1ꢀphenylethyl acetate ethanol
CH3
HO
(S)ꢀ1ꢀphenylethanol
Fig. 2. Reaction scheme of enantioselective enzymatic kinetic resolution of racemic 1ꢀphenylethanol with ethyl acetate over imꢀ
mobilized lipase.
packedꢀbed reactors offer several advantages including
easy control and operation, reduction of labor costs,
etc. [23, 24]. Promising results have been obtained in
lipaseꢀcatalyzed kinetic resolution in continuousꢀflow
packedꢀbed bioreactors [25] showing that the producꢀ
tivity was considerably enhanced in a continuous flow
reactor compared to the corresponding batch mode.
The kinetic resolution of racemic 1ꢀphenylethanol
(Fig. 2) with ethyl acetate was performed in a downꢀ
flow continuous reactor (Fig. 3). The internal diameꢀ
ter and the length of the reactor are 1 cm and 12 cm,
respectively. The reactor was filled with 3.5 cm quartz
sand (>355 m) on the top of a 1 cm quartz wool layer,
µ
located at the bottom. Above this quarz sand layer, the
catalyst bed (2.25 cm), containing a mixture of 0.125 g
The aim of this work was to study transient kinetics
in the kinetic resolution of racemic 1ꢀphenylethanol immobilized lipase, Novozym 435, and 1.5 g quartz
with ethyl acetate over lipase catalysts in a downꢀflow
continuous reactor. The effect of acetophenone, ethyl
benzene, acetic acid, water and ethanol on the cataꢀ
lytic behavior was studied, since these compounds are
sand of a height of 2.9 cm, was placed. The uppermost
quartz layer which dimension is 3.5 cm in length disꢀ
tributed the liquid flow above the catalyst bed. In a
typical experiment, the kinetic resolution of racemic
1ꢀphenylethanol (racꢀ1) with ethyl acetate (EA) was
performed at the molar ratio of 1 : 3 in 400 mL toluene
present in oneꢀpot synthesis of Rꢀ1ꢀphenylethyl aceꢀ
tate [26]. The main studied parameters were flow rates
(and thus residence time), substrate concentrations.
The continuous flow reactor and the batch reactor
modes for the kinetic resolution of racemic 1ꢀpheꢀ
nylethanol were compared.
(J.T. Baker, 99%) at 70°C
with 3 mL min–1 flow rate at
atmospheric pressure. The liquid phase containing
substrates was saturated with argon prior pumping into
the reactor. The catalyst bed was washed with toluene
at room temperature for 15 min before and after each
experiment. The immobilized lipase was kept at +
overnight between the experiments.
4°C
EXPERIMENTAL
The SEM images of fresh and spent immobilized
lipase catalysts were taken with the scanning electron
microscope system Leo Gemini 1530 with a Thermo
Scientific UltraDry Silicon Drift Detector (SDD)
equipped with SE (secondary electron) and BSE
(backscattered electron) detectors including an Inꢀ
Lens detector.
The samples were analyzed with a gas chromatoꢀ
graph equipped with a FI detector and a chiral column
(ChirasilDex(250
ing temperature programme was applied: 100
0.30 C/min–130 C–15 C/min–200 (10 min). Injecꢀ
and the split
µ
m
×
0.250
µ
m
×
25 m) and the followꢀ
°C
(1 min)–
°
°
°
°C
tor and detector temperature was 280°C
KINETICS AND CATALYSIS Vol. 53
No. 6
2012

LIPASEꢀCATALYZED ACYLATION IN A CONTINUOUS DOWNꢀFLOW
675
ratio of 100 : 1 was applied. Acetophenone was used as
an internal standard for quantitative analysis.
Temperature
sensor
Residence time distribution was investigated by
applying step changes using either ethanol (Etax Aa,
99.5%) or hexane (J.T. Baker, 99%) or ethanol and
hexane in toluene. The experiments were performed at
Argon
flow
room temperature (25°C). In order to suppress possiꢀ
ble evaporaion of the corresponding compounds the
experiments with (i) 0.06 M ethanol, (ii) 0.06 M hexꢀ
ane or, (iii) 0.06 M ethanol and 0.06 M hexane mixꢀ
ture were performed. Toluene was pumped to the reacꢀ
tor for 10 min before and after the experiments. The
samples collected both from the inlet and outlet
streams of the reactor, were taken every 30 s to obtain
informative data. The effect of different flow rates was
investigated by applying flow rates ranging from 1.5 to
3 mL min–1.
Substrate
pump
Lipaseꢀfilled
continuous
flow reactor
Racemic 1ꢀphenylethanol
with ethyl acetate in toluene at 70°С
Chiral
product
Different concentrations of substrates were also
studied. The ratio of racꢀ1 (Acros, 97%) to ethyl aceꢀ
tate (SigmaꢀAldrich, >99.5%) was changed from 1 : 6
to 1 : 1 whereas the flow rates were kept constant. The
effect of different substrate concentrations was also
investigated at different volumetric flow rates. The
concentrations of racꢀ1 in the first and second experiꢀ
ments were 0.02 and 0.04 M whereas 0.06 and 0.12 M
EA were used, respectively.
Fig. 3. Simplified illustration of the experimental apparaꢀ
tus (~~: temperature control unit).
experimental data and only a limited amount of data
points was available a simple oneꢀparameter mixing
cell approximation (tanksꢀinꢀseries model, TIS) was
used to model the reactor setup. The residence time
distribution for the tanksꢀinꢀseries is given by
The effect of acetophenone (Acros, 99%), ethyl
n
i−1
⎡
⎢
⎤
⎥
benzene (Aldrich), acetic acid (J.T. Baker, 99
⎯
100%)
(2)
nt t
(
)
F t = = 1 − e−nt t
,
c
c0
( )
∑
i=1
and water on the activity and stability of the immobiꢀ
lized lipase was investigated. Moreover, the inhibitory
effects of either the desired or the stoichiometric prodꢀ
i − 1 !
(
)
⎢
⎣
⎥
⎦
and the corresponding washout function,
obtained from the usual relationship
W(t), is
ucts ((R)ꢀ1ꢀphenylethyl acetate (Acros, 96%) and ethꢀ
anol (Etax Aa, 99.5%), respectively) were studied as
well. The concentrations of racꢀ1 and ethyl acetate
were kept constant, being 0.02 and 0.06 M, respecꢀ
tively. After the start up of the reaction, the known
concentration of the compounds to be investigated
(0.02 M) was added to the known concentration of the
liquid phase at certain time intervals.
The activity of the immobilized lipase was deterꢀ
mined by comparing the activities obtained over fresh
and spent catalysts. The experiments were performed
using 0.02 M racꢀ1 and 0.06 M ethyl acetate in 400 mL
toluene with 3 mL min–1 flow rate. The very same experꢀ
iments were carried out 30 days after the first experiment.
SEM images of the fresh and spent catalysts were taken in
order to make visible morphological changes which
could be used to rationalize activity decrease.
(3)
W t = 1 − F t .
( )
( )
In order to get a satisfactory description (Fig. 4) of
the measured hydrodynamics six mixing cells were
used with a mean residence time of roughly 3 min,
which is more than three times longer than the resiꢀ
dence time calculated based on reactor volume and
volumetric flow. This is a strong indicator that the
pump and inlet lines account for the main total volꢀ
ume of the studied system. The suggestion was verified
in a separate experiment, in which it was found that
the mean residence time from feed vessel to reactor
inlet was approximately 2 min for the same experiꢀ
mental conditions. The degree of backꢀmixing in the
actual reactor bed was roughly the same as in the
upstream apparatus.
The number of mixing cells in the TIS model can
directly be related to the Peclet number, Pe, for the one
dimensional axial dispersion model:
RESULTS AND DISCUSSION
Pe2
The hydrodynamics of the reactor system was studꢀ
ied by following the responses to step changes in the
feed concentrations. The deviation from plug flow was
then evaluated by performing a rough time domain fit
(4)
n =
.
2 Pe −1 + e−Pe
(
)
Five to six mixing cells in the tanksꢀinꢀseries model
to measured concentration responses, such as that correspond to Peclet numbers of about 10 in the axial
shown in Fig. 4. Since there was some noise in the dispersion model. In other words, in the current reacꢀ
KINETICS AND CATALYSIS Vol. 53
No. 6
2012

676
Normalized concentration
SAHIN et al.
)ꢀ1ꢀphenylethyl acetate concentration, M
(R
1.2
1.1
0.8
0.6
0.4
0.2
0
0.004
0.003
0.002
0.001
0
·
t
= 45 s
3
Hexane
Ethanol
·
t
= 90 s
2
1
–0.2
0
50 100 150 200 250 300 350
Timeꢀonꢀstream, min
30
35
40
45
50
55
60
65 70
Time, min
Fig. 5. Kinetic resolution of racꢀ1 at 70
up of the first experiment using = 3 mL min
of the second experiment using =1.5 mL min was apꢀ
plied; 3, using = 1.5 mL min . The ratio between the conꢀ
V
centrations of the racꢀ1/EA is 1 : 3.
°
C. Notation:
1
, startꢀ
–1
;
2, startꢀup
V
–1
Fig. 4. Transient kinetic by applying step changes at 25
using =3 mL min . Notation:
toluene;
°
C
V
–1
–1
1
, 0.06 M of ethanol in
V
2
, 0.06 M of hexane in toluene.
tor a flow pattern is present in which convective transꢀ
port of mass (plug flow) in the system dominates over
dispersive transport by roughly one order of magniꢀ
tude. The deviations from ideal flow for the investiꢀ
gated system are small enough that the results shown
in Figs. 5–12 can be attributed to intrinsic kinetic and
possibly adsorption effects.
The effect of the residence time (
where V
τ = V V,
is the catalyst bed volume and is the volumetric flow
V
rate of the liquid, respectively) was investigated by
applying different volumetric flow rates. Initially, the
residence time was increased at 60 min timeꢀon stream
from 45 to 90 s. The results showed that doubling the
residence time without changing the concentrations of
substrates doubled (
as expected. The highest (
R
)ꢀ1ꢀphenylethanol conversion,
R)ꢀ1ꢀphenylethanol converꢀ
sion was 30% after 300 min timeꢀonꢀstream correꢀ
sponding to 100% selectivity of the desired acetate
((R)ꢀ1ꢀphenylethyl acetate). The (R)ꢀ1ꢀphenylethyl
acetate concentration was enhanced by 53% when the
residence time increased two fold (Fig. 5). No deactiꢀ
vation of the immobilized lipase was observed.
(R)ꢀ1ꢀphenylethyl acetate concentration, M
0.008
0.006
0.004
0.002
0
·
5
t
= 45 s
Different concentrations of the substrates were also
applied whereas the residence time was kept constant
(Fig. 6). The startꢀup of the reaction was carried out
with the ratio of racꢀ1 and ethyl acetate equal to 1 : 3
(0.02 and 0.06 M, respectively) for 180 min timeꢀonꢀ
stream. The ratio of the substrates was changed from
1 : 3 to 1 : 1 (0.02 and 0.02 M, respectively) in the secꢀ
ond step. Decreasing the concentration of ethyl aceꢀ
tate decreased the desired product concentration in
half (Fig. 6) meaning that the excess of ethyl acetate
which was used as an acyl donor in this reaction is necꢀ
essary to obtain high yields of the desired product In
the third step, racꢀ1 concentration was doubled, 0.04 M,
comparing to the startꢀup concentrations whereas
ethyl acetate concentration was kept constant (0.06 M).
Doubling racꢀ1 concentration enhanced the concenꢀ
4
2
3
1
0
200
400
600
800
1000
Timeꢀonꢀstream, min
Fig. 6. Kinetic resolution of racꢀ1 at 70
different concentrations of substrates. Notation:
°
C. The effect of
, startꢀ
, the raꢀ
tio between racemic 1ꢀphenylethanol (racꢀ1) to ethyl aceꢀ
tate (EA) (racꢀ1/EA) is 1 : 3; , the ratio between racꢀ1/EA
is 1 : 1; , the ratio between racꢀ1/EA is 2 : 3; , the ratio
between racꢀ1/EA is 1 : 6. The arrows ( ) show when the
different concentrations were applied at timeꢀonꢀstream.
1
; 2
–1
up of the first experiment using = 3 mL min
V
3
tration of (R)ꢀ1ꢀphenylethyl acetate by 50% (Fig. 6).
Compared to the second step, in the third step the
concentrations of racꢀ1 and ethyl acetate increased
4
5
→
KINETICS AND CATALYSIS Vol. 53
No. 6
2012

LIPASEꢀCATALYZED ACYLATION IN A CONTINUOUS DOWNꢀFLOW
)ꢀ1ꢀphenylethyl acetate concentration, M )ꢀ1ꢀphenylethyl acetate concentration, M
677
(R
(R
0.004
·
0.005
0.004
0.003
0.002
0.001
0
·
t
= 45 s
t
= 45 s
0.003
2
3
0.002
0.001
0
2
3
1
1
0
50
100
150
200
250
0
50
100 150 200 250 300
Timeꢀonꢀstream, min
Timeꢀonꢀstream, min
Fig. 8. Kinetic resolution of racꢀ1 at 70
°
C. Notation:
Fig. 7. Kinetic resolution of racꢀ1 at 70°C. Notation: 1, startꢀ
–1
–1
up of the experiment using = 3 mL min with racꢀ1/EA
1, startꢀup of the experiment using = 3 mL min with
V
V
is 1 : 3; 2–3, the effect of 0.02 M acetophenone addition to
racꢀ1/EA is 1 : 3; 2–3, the effect of 0.02 M ethyl benzene
addition to the reaction media between 60 to 120 min
timeꢀonꢀstream.
the reaction media between 90 to 180 min timeꢀonꢀstream;
3, acetophenone feed switchꢀoff.
from 0.02 and 0.02 M to 0.04 and 0.06 M, respectively centration decreased by 17% after adding 0.02 M aceꢀ
meaning that the ratio of racꢀ1 to ethyl acetate was
tophenone.
decreased from 1 to 0.666. Increased concentrations
of substrates increased (R)ꢀ1ꢀphenylethyl acetate conꢀ
The effect of ethyl benzene on the enzyme activity
and stability was also studied. Since ethyl benzene was
observed as the major side product in the oneꢀpot
reactions due to the hydrogenolysis of racemic 1ꢀpheꢀ
nylethanol, 0.02 M ethyl benzene addition to the reacꢀ
tion media at 60 until 120 min timeꢀonꢀstream was
studied in the continuous flow system. The results
indicated that presence of ethyl benzene did not retard
the acylation activity of lipase (Fig. 8).
centration by 200% (Fig. 6). Finally, the ratio of racꢀ1
to ethyl acetate was changed to 1 : 6 (0.02 and 0.12 M,
respectively). Very high excess of ethyl acetate gave the
highest desired product concentration (Fig. 6).
Generally, it can be stated that in experiments with
step changes the enzyme was the same in all experiꢀ
ments. However, enzyme activity can change with
time and furthermore, there are several effects, which
can influence the enzyme acitivity, such as water conꢀ
The highest inbitory effect was observed with 0.02 M
tent. Accordingly, the results presented in Figs. 7–12 acetic acid addition between 120 to 180 min timeꢀonꢀ
should be compared only within the course of a single
pulse experiment and not with each other. To illustrate
this point, we can compare the concentration of the
(R)ꢀ1ꢀphenylethyl acetate with the fresh catalyst in the
step 1 in Fig. 5 with the results obtained under similar
conditions in Figs. 7–12. It can be seen that the
stream (Fig. 9). Acetic acid was produced by hydrolyꢀ
sis of ethyl acetate with water generated in a side reacꢀ
tion of racemic 1ꢀphenylethanol and dehydration over
palladium supported on acidic supports in oneꢀpot
reactions [26]. Enzymes are very sensitive to pH
changes since pH significantly influences the secondꢀ
ary, tertiary and quaternary structure of proteins thus
affecting enzyme deactivation kinetics [27]. The
enzyme inactivation rate, as expected, in most cases is
significantly dependent upon the pH of the solution
[28]. However, the definition of pH measured in water
cannot be applied to aqueous mixture of organic solꢀ
vents since pH has no meaning in organic solvents
[28]. It is well known that the enzymatic activity in
organic solvents is related to the protonation state of
enzyme’s ionizable groups reflecting the pH of the last
aqueous solution where they were lyophilized [29].
former concentration is 0.002 M (Fig. 5, step 1),
whereas it was about 0.003 M or even more in the iniꢀ
tial step in experiments including addition of aceꢀ
tophenone (Fig. 7), ethyl benzene (Fig. 8), acetic acid
(Fig. 9), or on changing the ratio of racꢀ1 to ethylaceꢀ
tate (Fig. 10), and adding the product (Fig. 11) and
water (Fig. 12).
When 0.02 M of acetophenone was added to the
reaction media between 90 to 180 min timeꢀonꢀstream
at 70
results showed that acetophenone slightly inhibited
°C
passed with a flow rate of 3 mL min–1, the
the lipase activity (Fig. 7). The desired product conꢀ This phenomenon is called the pH memory. The
KINETICS AND CATALYSIS Vol. 53
No. 6
2012

678
)ꢀ1ꢀphenylethyl acetate concentration, M
0.004
SAHIN et al.
)ꢀ1ꢀphenylethyl acetate concentration, M
(R
(R
0.005
0.004
0.003
0.002
0.001
0
·
·
t
= 45 s
t
= 45 s
0.003
0.002
0.001
0
3
2
3
2
1
1
0
50
100
150
200
250
300
0
50
100 150 200 250 300
Timeꢀonꢀstream, min
Timeꢀonꢀstream, min
Fig. 9. Kinetic resolution of racꢀ1 at 70
°
C. Notation:
Fig. 10. Kinetic resolution of racꢀ1 at 70
°
C. Notation:
–1
–1
1
, startꢀup of the experiment using = 3 mL min with
V
1
, startꢀup of the experiment using = 3 mL min with
V
racꢀ1/EA is 1 : 3; 2–3, the effect of 0.02 M (R)ꢀ1ꢀphenylꢀ
racꢀ1/EA is 1 : 3; 2–3, the effect of 0.02 M acetic acid adꢀ
dition to the reaction media between 120 to 180 min timeꢀ
onꢀstream.
ethyl acetate addition to the reaction media between 60 to
120 min timeꢀonꢀstream.
one of the most hydrophilic organic solvents with—
results obtained from acetic acid addition (Fig. 9) are
in line with the influence of pH changes over enzymes.
One of the most often considered issues in the litꢀ
erature related to the activity and stability of enzymes
is the product inhibition. The influence of the desired
log
in principle occur when excess of ethanol is added, is
ethanolysis of ( )ꢀ1ꢀphenylethyl acetate. The extent
of ethanolysis is difficult to quantify, since the ee of
)ꢀ1ꢀphenylethyl acetate remained constant during
the reaction being about 98% and traces of ( )ꢀ1ꢀpheꢀ
nylethyl acetate were observed in the product mixture.
Furthermore, the corresponding value was ca. 100.
P = 0.24 value [30]. Another reaction, which could
R
(
R
product
product ethanol on the lipase activity was studied. A
portion of 0.02 M of )ꢀ1ꢀphenylethyl acetate was fed
(R)ꢀ1ꢀphenylethyl acetate and stoichiometric
S
(R
E
to the reaction media at 60 min timeꢀonꢀstream and
the feed continued for 60 min. The desired product
concentration formed in the reaction decreased by
30% indicating that enzyme activity decreased due to
the high product concentration (Fig. 10). This result
show that the product should be withdrawn from the
reaction media continuously to achieve high desired
product yields.
An increase in the desired product concentration after
the withdrawal of ethanol from the reaction media at
point
3 (Fig. 11) was observed. The activity of the
immobilized lipase enhanced because the liquid phase
was replaced by the fresh substrate mixture.
Essential water is needed to keep the enzyme active
even though the reaction media contains mainly
organic solvent and/or substrates. On the other hand,
water in the reaction mixture can cause unfavourable
equilibrium in reversed hydrolysis and lead to hydroꢀ
lytic side reactions in transferase catalyzed reactions.
It has been shown that the water content in the reacꢀ
tion mixture is best described in terms of water activity
[31]. A quick and more convenient way to determine
the water activity is to measure water concentration
[32]. There are several reports in the literature on the
effect of water activity on the catalytic behavior of
lipases in organic media [33], and it is clear that differꢀ
ent enzymes have quite different requirements for
water activity. Lipases can be active at water activities
very close to 0 [34].
Ethanol is stoichiometrically produced during the
kinetic resolution of (
acetate over lipase. When 0.02 M ethanol was added to
the reaction mixture (Fig. 11, step ), the concentraꢀ
R)ꢀ1ꢀphenylethanol with ethyl
2
tion of the desired product decreased by almost 100%
(Fig. 11). The inhibition of lipase due to high concenꢀ
tration of ethanol can be explained by the influence of
hydrophilic solvents on the enzyme activity in organic
solvents. The enzyme activity usually decreases with
increasing the hydrophilicity of the solvent [15]
because hydrophilic solvents apt to strip off essential
water which is needed by enzyme to preserve the 3D
structure of the protein in a catalytically active form
[8]. The active site of the enzyme intends to disperse
resulting in unfolding of the molecule [7] without the
The water effect was in the focus of interest of the
essential water around. The decrease in the activity of present study. In order to investigate the effect of
lipase is consistent with the literature since ethanol is excess water, water was added to the reaction media
KINETICS AND CATALYSIS Vol. 53
No. 6
2012

LIPASEꢀCATALYZED ACYLATION IN A CONTINUOUS DOWNꢀFLOW
)ꢀ1ꢀphenylethyl acetate concentration, M )ꢀ1ꢀphenylethyl acetate concentration, M
679
(R
(R
0.004
0.004
0.003
0.002
0.001
0
·
t
= 45 s
·
t
= 45 s
0.003
0.002
0.001
0
3
2
2
3
1
1
0
50 100 150 200 250 300 350
Timeꢀonꢀstream, min
0
50
100
150
200
Timeꢀonꢀstream, min
Fig. 12. Kinetic resolution of racꢀ1 at 70
°
C. Notation: 1,
Fig. 11. Kinetic resolution of racꢀ1 at 70
°
C. Notation:
–1
1
1
, startꢀup of the experiment using = 3 mL min with
racꢀ1/EA is 1 : 3; 2–3, the effect of 0.02 M ethanol addition
startꢀup of the experiment using = 3 mL min⎯ with
V
V
racꢀ1/EA is 1 : 3; 2–3, the effect of 0.02 M water addition to
to the reaction media between 60 to 120 min timeꢀonꢀstream.
the reaction media between 60 to 120 min timeꢀonꢀstream.
since it was formed as a result of a side reaction, dehyꢀ
dration of racemic 1ꢀphenylethanol, due to the acidic
support properties of palladium catalysts in batch
reactor [26]. The results showed that the excess water
(0.02 M) diminished enzyme activity significantly
(Fig. 12). The desired product concentration
decreased by 68% after the water addition. The
decrease can be explained by the limited diffusion of
the reactants to the active site of the enzyme. The reaꢀ
son for the gradual increase of the concentration of
The activity of the immobilized lipase recorded as
a function of time decreased by about 40% after
30 days (Fig. 13). The experiments performed with a
fresh catalyst gave around 0.004 M desired product
concentration, however, when the very same experiꢀ
ments were carried out over the 30 days spent catalyst, the
desired product concentration dropped from 0.004 to
0.0025 M. It is interesting to note that the activity in a
fixed bed reactor equal to 1.6 ×
10–6 mol/s is very close
to activity of an immobilized enzyme observed in a
batch reactor [35]. The decrease in the catalytic activꢀ
(R)ꢀ1ꢀphenylethyl acetate before 275 min is unclear.
After that time, the concentration of the product
increased suddenly to about the same level, as it was
before water addition. It should, however, be stated
that there can be several reasons for this kind of behavꢀ
ior, for example the conformation of the enzyme can
be changed as a function of water concentration. Since
the excess water generates an extra layer around the
enzyme and thus clogs the active site of the enzyme to
interact with organic phase including reactants, the
desired product formation decreased substantially.
The inhibitory effect of water disappeared after water
was removed from the inlet mixture (Fig. 12). The
Substrate concentrations, M
0.005
0.004
0.003
0.002
0.001
0
activity over fresh
catalyst
activity over
spent catalyst
hydrolysis of the product back to (R)ꢀ1ꢀphenylethanol
and acetic acid might also occur. Analogously, as it was
described above in the section describing ethanolysis,
also in the case of product hydrolysis, the ee of the
1, 2
product remained close to 98% and traces of (S)ꢀ1ꢀ
phenylethyl acetate were formed. Thus, it could be
concluded that the addition of water can decrease the
concentration of (R)ꢀ1ꢀphenylethylacetate via making
0
100 200 300 400 500 600
Timeꢀonꢀstream, min
the reaction mixture more polar as well as hydrolyzing
in some extent the product. The extent of hydrolysis
was not determined in this work.
Fig. 13. Kinetic resolution of racꢀ1 at 70°C. Notation:
1–2, startꢀup of the first and second experiment using
=
V
–1
3 mL min with racꢀ1/EA is 1 : 3.
KINETICS AND CATALYSIS Vol. 53
No. 6
2012

680
SAHIN et al.
(a)
(b)
Data 25Mar 2010
Data 25Mar 2010
Mag = 10x
WD = 11 mm
200
(c)
µ
m
200
(d)
µ
m
20
µ
m
20 µm
Fig. 14. images of the fresh immobilised lipase. (a) 30
×
, (b) 100
×
, (c) 250
×
, (d) 500×.
ity of the immobilized enzyme was also observed in the
As indicated above the same reaction was previꢀ
experiments conducted at different time intervals. The ously studied by the authors in a batch reactor [35].
SEM pictures from fresh and spent catalysts clearly Kinetics in the batch reactor was discussed in detail in
[35], thus the aim of the current work was to compare
batch and continuous operation and to reveal influꢀ
ence of substrates present during oneꢀpot synthesis on
acylation. In kinetic modelling performed for the
batch mode it was sufficient to describe kinetics withꢀ
out inclusion adsorption constants of reactants.
Experiments in the continuous mode clearly highꢀ
lighted influence of ethanol and the products, in particꢀ
ular they indicated that there is inhibition by the reactꢀ
ing phenvlalcohol, as well as the product and ethanol.
In [35] kinetic modelling revealed first order in ethꢀ
showed that the activity decrease was due to the morꢀ
phological changes of the spent catalyst. The SEM
pictures taken from the fresh catalyst (Fig. 14a) demꢀ
onstrate smooth surface of the catalyst. The same picꢀ
tures for the spent catalysts (Fig. 15a) give a clear
image of the destruction of the catalyst integrity. The
pictures taken with 200× and 500× magnififcation
(Figs. 14c, 14d) show that the surface of the fresh catꢀ
alyst is very smooth compared to the surface of the
spent catalyst (Figs. 15c, 15d) indicating that the
decrease of the enzyme activity might be a result of the
physical changes of the immobilised enzyme during the ylacetate thus in the present work also the first order
reactions caused by organic solvents and/or substrates. was utilized resulting in the following kinetic equation
k1CEACR−PE
(5)
r =
,
1 + KRꢀPEACCRꢀPEAC + KRꢀPECRꢀPE + KSꢀPECSꢀPE + KEtOHCEtOH
KINETICS AND CATALYSIS Vol. 53
No. 6
2012

LIPASEꢀCATALYZED ACYLATION IN A CONTINUOUS DOWNꢀFLOW
681
(a)
(b)
Data 25Mar 2010
Data 25Mar 2010
Mag = 10x
WD = 11 mm
200
µ
m
200 µm
(c)
(d)
20
µ
m
20 µm
Fig. 15. SEM images of the spent immobilised lipase. (a) 30
×
, (b) 100
×
, (c) 250
×
, (d) 500
×
.
where
respectively (
R
ꢀPEAC,
)ꢀphenylethylacetate ethylacetate, (
)ꢀ1ꢀphenylethanol as well as ethanol.
R
ꢀPE,
S
ꢀPE and EtOH denote
2
2
(6)
Q = xexp − xest
xexp,it − xest,it ,
= ∑∑(
t
)
R
R)ꢀ and
i
(
S
where xexp is the experimental value and xest denotes
the predictions given by the model, is the component
index and is the time value. The quality of the fit and
For the calculation of parameters, differential
i
equation (5) describing the changes in the concentraꢀ
tions profiles of the reagents and products was solved
by means of ModEst software [36]. Using Levenbergꢀ
Marquardt simplex method, the target function,
which was defined as incompliance between the
experimental and calculated values of concentrations
was used to find the parameters. The sum of the residꢀ
t
accuracy of the model description was defined by the
degree of explanation R2
;
2
⎛
⎞
xexp,i − xest,i
R2 = 100 1 −
,
(7)
⎜
⎜
⎝
⎟
⎟
2
xexp,i − xest
⎠
ual squares between the model and the experimental where
is the mean value of all the data points.
xest
data (Fig. 16) was minimized using the following
objective function;
The results of the simulations for experimental data
in Fig. 16 are given in table.
R2, %
k
1, L mol–1 min
12.30 0.58
KRꢀPE, L mol–1
KSꢀPE, L mol–1
KRꢀPEAC, L mol–1
K
EtOH, L mol–1
1180 405
86
490 340
negligible
154 59
The degree of explanation as high as 86% evidences mental data for the concentration profiles could be
a good compliance between the experimental data and probably still improved taking into account activity
the calculated values of the concentrations of the increase with time. Despite the relatively large number
reagents. The fit of the kinetic model to the experiꢀ of parameters in the proposed mechanism, the model
KINETICS AND CATALYSIS Vol. 53
No. 6
2012

682
SAHIN et al.
acetic acid is formed by ethylacetate hydrolysis.
R
0.007
ꢀPEAC concentration, 10–3
M
Therefore the effect of acetic acid on the activity of
enzyme was also investigated since the inhibitory
effect of acids on enzymes is very well known. The acid
changes catalytically active 3D structure of enzymes
causing the activity loss. The inhibition on the lipase
activity due to acetic acid was clearly observed in this
study. The experiments to elucidate the inhibitary
effect of either the desired or the stoichiometric prodꢀ
0.006
0.005
0.004
0.003
0.002
0.001
3
8
6
4
ucts ((R)ꢀ1ꢀphenylethyl acetate and ethanol, respecꢀ
tively) with the inhibitory effect of water on the activity
and stability of the immobilized lipase was also carried
out. The slight inhibition due to the desired product
was observed indicating that the product should be
withdrawn continuously from the reaction media to
increase the yield in the batch mode. The activity of
lipase was significantly diminished by excess ethanol.
The decreases of enzyme activity can be explained
with the increase in the hydrophilicity of the solvent.
Hydrophilic solvents apt to strip off essential water
which is needed by enzyme to preserve the 3D strucꢀ
ture of the protein in a catalyticaly active form. The
similar effect was observed due to the increase of water
concentration in the reaction media. The decrease in
the activity of enzyme was also followed as a function
of time. The desired product concentrations diminꢀ
ished over the catalysts used for 30 days. SEM pictures
demonstrated that the spent catalysts are much more
damaged whereas the fresh catalysts had smoother
surface than the spent samples.
2
7
5
1
0
10 20 30 40 50 60 70 80 90 100
Time, min
Fig. 16. Kinetic resolution of racꢀ1 at 70
different concentrations of reactants. Concentration of
)ꢀ1ꢀphenylethanol and ( )ꢀ1ꢀphenylethanol was always
0.01 M each. Concentrations of other reactants were
EA = 0.02 M, : EA = 0.06 M, : EA = 0.12 M; : EA =
0.06 M; : EtOH = 0.02; : EA = 0.06 M; : EA = 0.06 M;
ꢀPEAC = 0.02 M; : EA = 0.06 M.
°C. The effect of
(R
S
1:
2
3
4
5
6
7
R
8
was fitting the data with a good degree of accuracy. The
model was able to rationalize lipase catalyzed acylaꢀ
tion in a continuous down flow fixed bed reactor.
In kinetic modelling performed previously for the
batch mode it was sufficient to describe reaction kinetꢀ
ics without including adsorption constants of the reacꢀ
tants. Experiments in the continuous mode clearly
revealed influence of ethyl acetate, ethanol and the
reaction product. Kinetic analysis of the experimental
data obtained in the continuous mode was thus perꢀ
formed, taking into account a more complicated picꢀ
ture of adsorption, showing a good correspondence
between experimental and calculated values.
CONCLUSIONS
The utilization of the continuous flow reactor for
the kinetic resolution of racemic 1ꢀphenylethanol
with ethyl acetate in toluene over an immobilized
lipase was successfully demonstrated in this study. The
results were discussed in connection to the acylation of
racemic 1ꢀphenylethanol in toluene over immobilized
lipase utilizing oneꢀpot system operated in batch
mode. The effects of different residence times and
substrate concentrations were investigated by applying
different flow rates and ratios of racꢀ1to ethyl acetate,
respeLctively. The results showed the absence of the
external mass transfer.
The inhibitory effect of organic compounds used in
the reaction such as ethyl acetate, ace lie acid, aceꢀ
lophenone etc., on the activity and stability of the
immobilized lipase was also studied. No influence of
excess ethyl benzene on the enzyme activity was
observed whereas the enzyme activity was slightly
inhibited by acetophenone. Ethyl benzene is the major
side product formed in oneꢀpot reactions due to the
acidic character of the metal support. The formation
of ethyl benzene occurrs via hydrogenolysis reactions
of racemic 1ꢀphenylethanol under H2 flow. The water
is released during the hydrogenolysis reactions and
ACKNOWLEDGMENTS
This work is part of the activities at the Abo Akaꢀ
demi Process Chemistry Centre within the Finnish
Centre of Excellence Programmes (2000–2011) by
the Academy of Finland.
REFERENCES
1. Carrea, G. and Riva, S., Angew. Chem. Int. Ed., 2000,
vol. 39, p. 2226.
2. Chen, C.ꢀS. and Sih, C.J., Angew. Chem. Int. Ed., 1989,
vol. 28, p. 695.
3. GotorꢀFernandez, V., Brieva, R., and Gotor, V., J. Mol.
Catal., 2006, vol. 40, p. 111.
KINETICS AND CATALYSIS Vol. 53
No. 6
2012

LIPASEꢀCATALYZED ACYLATION IN A CONTINUOUS DOWNꢀFLOW
683
4. Machado, G.D.C., Paiva, L.M.C., Pinto, G.F., and 21. Lozano, P., de Diego, T., Carrie, D., Vaultier, M., and
Oestreicher, E.G., Enzyme Microb. Technol., 2001,
Iborra, J.L., Chem. Commun., 2002, p. 692.
vol. 28, p. 322.
22. Reetz, M.T., Wiesenhofer, W., Franci, G., and Leitner, W.,
5. Bornschauer, U.T. and Kazlauskas, R.J., Hydrolases in
Organic Synthesis: Regioꢀ and Stereoselective Biotransꢀ
formations, Weinheim: WileyꢀVCH, 2006.
Adv. Synth. Catal., 2003, vol. 345, p. 1221.
23. Jas, G. and Kirchning, A., Chem. Eur. J., 2003, vol. 9,
p. 5708.
6. Ghanem, A. and AboulꢀEnein, H.Y., Chirality, 2005,
24. Kirchning, A., Solodenko, W., and Mennecke, K.,
vol. 17, p. 1.
Chem. Eur. J., 2006, vol. 12, p. 5972.
7. Matsuda, T., Watanabe, K., Harada, T., Nakamura, K.,
Arita, Y., Misumi, Y., Ichikawa, S., and Ikariya, T.,
Chem. Commun., 2004, p. 2286.
8. Frings, K., Koch, M., and Hartmeier, W., Enzyme
Microb. Technol., 1999, vol. 25, p. 303.
25. Csajagi, Cs., Szatzker, G., Toke, E.R., Urge, L., Davas, F.,
and Poppe, L., Tetrahedron: Asymmetry, 2008, vol. 19,
p. 237.
26. MakiꢀArvela, P., Sahin, S., Kumar, N., Mikkola, J.ꢀP.,
Eranen, K., Salmi, T., and Murzin, D.Yu., Catal.
Today, 2009, vol. 140, p. 70.
9. Kazlauskas, R.J., Trends Biotechnol., 1994, vol. 12,
p. 464.
27. Sadana, A. and Henley, J.P., J. Biotechnol., 1988, vol. 7,
10. Chen, C.ꢀS., Fujimoto, Y., Girdaukas, G., and Sih, C.J.,
p. 95.
J. Am. Chem. Soc., 1982, vol. 104, p. 7294.
11. Ogino, H. and Ishikawa, H., J. Biosci. Bioeng., 2001,
28. Dixon, M. and Webb, E.C., Enzymes, New York: Acaꢀ
vol. 91, p. 109.
demic, 1979.
12. Colombo, G., Ottolina, G., and Carrea, G., Monatsh.
Chem., 2000, vol. 131, p. 527.
29. Fang, Y.W., Liu, S., Wang, S.J., and Lv, M., Biochem.
Eng. J., 2009, vol. 43, p. 212.
13. Zaks, A. and Klibanov, A.M., J. Biol. Chem., 1998,
30. Klibanov, A.M., Tibtech, 1997, vol. 15, p. 97.
vol. 26, p. 1.
31. Halling, P.J., Enzyme Microb. Technol., 1994, vol. 16,
14. Narayan, V.S. and Klibanov, A.M., Biotechnol. Bioeng.
1993, vol. 41, p. 390.
15. Yang, L., Dordick, J.S., and Garde, S., Biophys. J.
,
,
p. 178.
32. Wehtje, E., Kaur, J., Adlercreutz, P., Chand, S., and
Mattiasson, B., Enzyme Microb. Technol., 1997, vol. 21,
p. 502.
2004, vol. 87, p. 812.
16. Laane, C., Boeren, S., Vos, K., and Voeger, C., Biotechꢀ
nol. Bioeng., 1987, vol. 30, p. 81.
33. Ma, L., Persson, M., and Adlercreutz, P., Enzyme
Microb. Technol., 2002, vol. 31, p. 1024.
17. Faber, K., Biotransformations in Organic Chemistry
,
34. Valivety, R.H., Halling, J.P., and Macrae, A.R., FEBS
Lett., 1992, vol. 301, p. 258.
Berlin: Springer, 2004.
18. Patel, R.N., Banerjee, A., and Szarka, L.J., J. Am. Oil
Chem. Soc., 1996, vol. 73, p. 1363.
35. Kirilin, A., Sahin, S., MakiꢀArvela, P., Warna, J.,
Salmi, T., and Murzin, D.Yu., Int. J. Chem. Kinet.
2010, vol. 42, p. 629.
,
19. Chen, J.C. and Tsai, S.W., Biotechnol. Prog., 2000, vol. 16,
p. 986.
20. Ujang, Z., Husain, H.W., Seng, M.C., and Rashid, A.H.A., 36. Haario, H., ModEst User’s Guide 6.1, Helsinki: Profꢀ
Proc. Biochem., 2003, vol. 38, p. 1483.
Math Oy, 2002.
KINETICS AND CATALYSIS Vol. 53
No. 6
2012