S. Leuchs et al. / Journal of Molecular Catalysis B: Enzymatic 88 (2013) 52–59
57
and the aqueous phase has been observed. When these phases
were separated, the emulsion-like phase spontaneously split up
into two phases: The lower phase was clear, the upper phase was
emulsion-like. Further investigations revealed that the emulsion-
phase is only formed when LbADH is present in the aqueous phase.
On first view, emulsion-forming seems to be advantageous due to
the enhanced mass transfer. However, for the continuous synthesis
a good phase separation is required. Otherwise, parts of the aque-
ous, reactive phase can be carried out of the reactor. This leads to
leaching of the catalyst. For this reason, we determined the enzyme
activity in all three phases. The highest volumetric activity was
found in the aqueous phase, less in the interphase and no activity
was detectable in the organic phase (see Supporting information).
Conversion as a function of time of a typical experiment with
2-octanone as substrate is shown in Fig. 5. With the initial reac-
tion conditions, a conversion of 31% could be observed with an
−
1
apparent deactivation of 0.1% h . After 170 h on stream, the initial
−
1
enzyme concentration of 480 mg L (=2.4 mg lyophilised enzyme
−
1
preparation per 5 mL aqueous phase) was increased to 1480 mg L
(=7.4 mg lyophilised enzyme preparation per 5 mL aqueous phase)
which led to a doubling of the conversion from 26% to 48%. A dou-
bling of the residence time from 4 to 8 h showed less effect on the
conversion (from 48% to 56%) but led to a decrease in the space-
−
1
−1
−1 −1
time-yield (STY) from 142 mmol L
d
to 82 mmol L d . After
readjusting the residence time to 4 h, the conversion dropped to
42%, which, assuming a constant deactivation rate, is in accordance
with the steady-state before increasing the flow rate of the non-
2
-Propanol did not influence the phase behavior, but reduced the
3
volumetric activity as already indicated by the kinetic measure-
ments. These results show that emulsion forming needs monitoring
to prevent loss of the aqueous phase.
reactive phase. An overall turn-over number (TON) of 186 × 10
for the LbADH has been achieved. For the cofactor TON
+
NADP
3
was 26 × 10 , which is exceptionally high for an in vitro appli-
cation [38]. The conversion in the continuous synthesis is with
56% at its best not reaching the estimated equilibrium conver-
sion of 81%. In this experiment emulsification was observed to
a very low extent, so that continuous decantation was opera-
ble.
3.3. Continuous syntheses
The continuous synthesis offers advantages over batchwise syn-
thesis such as better catalyst utilisation, constant product quality
and easy automatisation [36]. As reactor, the biphasic mini-reactor
as described previously [24] was chosen. This can be considered
as a continuous stirred tank reactor (CSTR). For information gath-
ering, a CSTR offers more detailed information on the catalyst
system long term activity and stability than a plug flow reactor
Although the kinetic model is in line with the measured ini-
tial reaction rates, it was not possible to predict the course of the
biphasic batch-reaction or the continuous syntheses. Reasons for
this may lie in the complex reaction system. Partitioning of the
substrates and products is not included in the model and may be
not ideal. The prediction of a continuous experiment starting from
initial reaction rates is not straightforward, especially if a biphasic
experiment is described with data from monophasic initial rate
experiments. Conducting biphasic initial reaction rate experiments
might be helpful, but not straightforward to carry out.
(
PFR). A biphasic approach with pure substrate as second phase
was considered [37] but turned out to be challenging if carried out
continuously due to the aforementioned difficulties with the co-
substrate dosage and removal. Only the substrate coupled approach
with 2-propanol would be feasible. Then, a mixture of 2-propanol
and 2-octanone could be dosed into the reactor. Due to the miscibil-
ity of 2-propanol with water, the initial phase ratio would change
and removal of acetone and remaining 2-propanol is no longer
guaranteed. Thus, an approach where the 2-ketone and 2-propanol
are dissolved in MTBE was preferred. Both, the organic MTBE-phase
and the aqueous buffer-phase, were pre-saturated prior to use in
the continuous synthesis in order to prevent loss of the aqueous
phase due to solubility in MTBE. Initial LbADH concentration was
To show the potential and limitations of the approach, we
extended the experiments to homologous alkanone substrates with
higher and lower solubility in water. Subsequently, 2-heptanone,
2-nonanone, and 2-decanone were employed under the same reac-
tion conditions as for 2-octanone (Fig. 5 and Table 4). Kinetic data
for 2-heptanone hint towards lower reaction rates when compared
to 2-octanone experiments (Table 2). However, in the continuous
synthesis, with 2-heptanone as substrate, a conversion of 49% could
be realised. Under the same reaction conditions with 2-octanone as
substrate, a conversion of only 31% could be achieved. At a reaction
time of 140 h, the residence time was decreased from 4 h to 3 h
which led to a decrease in conversion to 29%. After increasing the
residence time again to 4 h, the conversion regained 45%. In total, a
−
1
+
4
80 mg L (=2.4 mg lyophilised enzyme preparation) and NADP -
−1
concentration was 0.1 mmol L if not stated otherwise. The outlet
concentrations were monitored by using a flow cell via online gas
chromatography.
2
-Octanone was chosen as substrate for comparison with pre-
3
3
vious studies, in which either a solubiliser was used in both batch
and continuous synthesis in a single phase [29,30,35], or biphasic
approaches with pure substrate as second phase [37] and a biphasic
approach with ionic liquid or MTBE as non-reactive phase in a
batch [23,35]. The reaction conditions were chosen based on the
results of the kinetic investigations and batch-experiments. All
TONLbADH of 478 × 10 and a TONNADP
+
of 22 × 10 were achieved.
It is noteworthy that the conversion with 2-heptanone as substrate
is higher than for 2-octanone, even though the kinetics with a lower
vmax,2-heptanone hintto a lowerratefor2-heptanone. Here, thehigher
availability of 2-heptanone seems to play the key role for the higher
reaction rate and thus a higher conversion. Still, the equilibrium
conversion of 85% with K = 0.536 [34,35] was not reached either. In
this experiment, emulsification was not significant.
continuous syntheses were carried out at R = 10 with a ketone
0
−1
concentration of 100 mmol L
and 2-propanol concentration of
−
1
1
.00 mol L
.
In contrast to the high conversion reached with 2-heptanone
as substrate, the maximum conversion achieved was 24% with 2-
nonanone as a substrate – under comparable reaction conditions.
Due to the formation of an interphase making decantation of the
organic phase impossible, the experiment was stopped after 72 h.
The influence of stirring speed in the aqueous phase and
cofactor-concentration were determined. Apparently, increasing
the stirring speed to more than 400 rpm led to no further increase
in the conversion. It also turned out that an initial stirring speed
of 200 rpm was beneficial for the long time enzyme performance.
So, the aqueous phase for all subsequent experiments was stirred
at 200 rpm for 2 h (except for the 2-decanone-experiment, here
the initial stirring time was 20 h) and increased to 500 rpm after-
4
3
Still, an overall TONLbADH of 6.4 × 10 and a TONNADP
+
of 3.0 × 10
were reached. Similar results were achieved with 2-decanone as
a substrate. The maximum conversion was 22% although the res-
idence time was 8 h instead of 4 h for 2-nonanone. Again, the
formation of an extended interphase caused the end of this exper-
−
1
wards. As increasing the cofactor concentration from 0.1 mmol L
−1
to 0.2 mmol L augmented the conversion from 25% only to 30%
data not shown), the lower concentration was fixed in view of a
better cofactor utilisation.
iment. Slightly higher TON were achieved when compared to the
4
(
2-nonanone experiment (TON
= 8.2 × 10 ; TON
+
= 3.8 ×
LbADH
NADP
3
1
0 ).