Ionic liquids as performance additives for electroenzymatic syntheses

Article abstract of DOI:10.1002/chem.200901046

Electroenzymatic syntheses combine oxidoreductase-catalysed reactions with electrochemical reactant supply. The use of ionic liquids as performance additives can contribute to overcoming existing limitations of these syntheses. Here, we report on the infl

Full text of DOI:10.1002/chem.200901046

DOI: 10.1002/chem.200901046
Ionic Liquids as Performance Additives for Electroenzymatic Syntheses
Christina Kohlmann,[a, b] Lasse Greiner,[b] Walter Leitner,[b, c] Christian Wandrey,[a] and
Stephan Lꢀtz*[a, d]
Abstract: Electroenzymatic syntheses
combine oxidoreductase-catalysed re-
actions with electrochemical reactant
supply. The use of ionic liquids as per-
formance additives can contribute to
overcoming existing limitations of
these syntheses. Here, we report on the
influence of different water-miscible
ionic liquids on critical parameters
such as conductivity, biocatalyst activity
and stability or substrate solubility for
three typical electroenzymatic synthe-
ses. In these investigations promising
ionic liquids were identified and have
been used as additives for batch elec-
trolyses on preparative scale for the
three electroenzymatic systems. It was
possible to improve the space-time-
yield for the electrochemical regenera-
tion of NADPH by a factor of three.
For an amino acid oxidase catalysed
resolution of a methionine racemate
with ferrocene-mediated electrochemi-
cal regeneration of the enzyme-bound
cofactor FAD a 50% increase in space
time yield and 140% increase in cata-
lyst utilisation (TTN) were achieved.
Furthermore, for the chloroperoxidase-
catalysed synthesis of (R)-phenylme-
thylsulfoxide with electrochemical gen-
eration of the required cosubstrate
H2O2 the space time yield and the cata-
lyst utilisation were improved by a
factor of up to 4.2 depending on the
ionic liquids used.
Keywords: biocatalysis
· biotech-
nology · electroenzymatic synthesis ·
ionic liquids · sustainable chemistry
Introduction
the reaction medium, biocompatible conducting salts could
contribute to the solution of this problem. Moreover, biocat-
alysts used in electroenzymatic synthesis suffer from deacti-
vation at the electrode surface in many cases, leading to in-
sufficient biocatalyst utilisation. Therefore, there is the need
for methods that stabilise the catalysts under process condi-
tions in order to make electroenzymatic syntheses competi-
tive. Furthermore, it is also difficult to use electroenzymatic
syntheses for the conversion of hardly water soluble sub-
strates, since classical methods such as the addition of an or-
ganic co-solvent lead to a decrease in conductivity of the re-
action medium, which will result in even lower productivi-
ties and biocatalyst utilisations, whereas the use of a second
organic phase will complicate the reaction setup tremen-
dously. Thus, co-solvents which do not reduce the conductiv-
ity of the reaction medium would be a great advantage for
electroenzymatic syntheses.
Ionic liquids (ILs) are promising candidates to overcome
all three limitations and improve electroenzymatic systems
synergistically. ILs have the following advantages: Firstly;
since ILs entirely consist of ions, they have outstanding elec-
trochemical properties.[4] Secondly, it is also demonstrated
that they can be advantageous for the performance of bio-
catalysts.[5] Finally, ILs can act as solubilisers for various sub-
stances.[6]
The combination of catalysed redox reactions with an elec-
trochemical reaction is state of the art for biosensor tech-
niques.[1] However, in synthetic reactions,[2] other methods
of supplying redox equivalents are more commonplace due
to their often superior productivities.[3] As the productivity
of electrochemical reactions is limited by the conductivity of
[a] Dr. C. Kohlmann, Prof. Dr. C. Wandrey, Dr. S. Lꢀtz
Institute of Biotechnology 2, Forschungszentrum Jꢀlich
52425 Jꢀlich (Germany)
Fax : (+49)2461-61-3870
E-mail: s.luetz@fz-juelich.de
[b] Dr. C. Kohlmann, Dr. L. Greiner, Prof. Dr. W. Leitner
Institut fꢀr Technische und Makromolekulare Chemie
RWTH Aachen University, 52074 Aachen (Germany)
[c] Prof. Dr. W. Leitner
Max-Planck-Institut fꢀr Kohlenforschung, Mꢀlheim a.d.R. (Germany)
[d] Dr. S. Lꢀtz
Present address: Head of Bioreactions
Novartis Institutes for BioMedical Research
4056 Basel (Switzerland)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901046.
11692
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 11692 – 11700

FULL PAPER
ILs had first been under investigation as electrolytes for
electrochemical applications.[7] At present they have caught
much attention as alternative reaction media for various cat-
alytic reactions, for example, in biocatalysis.[8] In our work
we describe the application of ILs as performance additives
in preparative scale electroenzymatic synthesis for the first
time.[9] The ILs are not used as pure reaction medium, but
rather as additives in volumetric contents of up to 10%.
Therefore, the electroenzymatic syntheses were carried out
in a homogenous water/IL mixture, which can also be con-
sidered as a salt solution containing organic ions. Within the
paper the amount of IL is given in volumetric contents
(vol%) as this unit is commonly used in biocatalysis. Addi-
tionally, diagrams were the results are plotted against the
concentration of the IL can be found in the Supporting In-
formation.
the chloroperoxidase (CPO) catalysed synthesis of (R)-phe-
nylmethylsulfoxide with electrochemical generation of the
required cosubstrate H2O2 (Scheme 3).[12] Depending on the
Scheme 3. CPO-catalysed synthesis of (R)-phenylmethylsulfoxide with
electrochemical generation of the required cosubstrate H2O2.
To generally demonstrate the potential of ionic liquids as
performance additives three representative electroenzymatic
synthesis reactions with different drawbacks were investigat-
ed. The first example is the electrochemical generation of
NADPH via a rhodium mediator, which can be applied for
different NADPH-consuming reactions (Scheme 1).[10] The
second example is an amino acid oxidase (d-AAO) cata-
lysed resolution of a methionine racemate with electrochem-
ical regeneration of the enzyme-bound cofactor FAD by fer-
rocene carboxylic acid (Scheme 2).[11] The third example is
reaction, different parameters such as the conductivity of
the reaction medium, the activity and the stability of the
biocatalyst or the substrate solubility, which can be influ-
enced by IL were investigated in presence of various IL (see
Table 1). From those experiments promising ILs were select-
ed and used as performance additive for the corresponding
electroenzymatic synthesis.
Results and Discussion
Electrochemical generation of
NADPH: The effective regen-
eration of NADPH represents
one of the major tasks for vari-
ous oxidoreductase-catalysed
reductions.
Beside
the
common enzyme- or substrate-
coupled approaches,[13] electro-
chemical regeneration via
a
suitable redox mediator pres-
ents an interesting alternative,
but is often significantly slower
than the other approaches.
Thus, for this reaction system
we tried to achieve higher pro-
ductivities by increasing the
conductivity of the reaction
media through addition of IL.
The application of other con-
ventional inorganic conducting
salts is not favoured, since high
concentrations of various salts
destabilise the cofactor.[14] For
this reason, not only the influ-
ence of different ILs on the
conductivity of the reaction
medium, but also on the stabil-
ity of the cofactors NADPH
and NADP+ were investigated.
Scheme 1. Electrochemical generation of NADPH via a rhodium mediator.
Scheme 2. d-AAO-catalysed resolution of a methionine racemate with electrochemical regeneration of the
enzyme-bound cofactor FAD by ferrocene carboxylic acid.
Chem. Eur. J. 2009, 15, 11692 – 11700
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11693

S. Lꢀtz et al.
Table 1. Investigated cations and anions.
[EMPY]
ACHTUNGTRENNUNG
leads to an almost tripled half life. [BMIM]ACHTUNGTRENNUNG
Cations
Anions
showed the lowest potential for cofactor stabilisation, but
with 10 vol% still an improvement by the factor of 1.5 was
possible. The stability of NADP+ is much higher than the
stability of NADPH under the chosen reaction conditions
and hence it is not limiting the reaction. Nevertheless, we
also investigated the influence of the same IL on the half
lives of the oxidised cofactor. Within a measuring time of
one week, only minor differences which are in the range of
experimental error have been detected. The ILs used seem
not to have an effect on the stability of the oxidised cofac-
tor.
methylsulfate [MeSO4]ꢀ
1,3-dimethylimidazolium [MMIM]+
ethylsulfate [EtSO4]ꢀ
1-ethyl-3-methylimidazolium
[EMIM]+
dimethylphosphate [Me2PO4]ꢀ
diethylphosphate [Et2PO4]ꢀ
1-butyl-3-methylimidazolium
[BMIM]+
1-hexyl-3-methylimidazolium
[HMIM]+
Figure 2 depicts the results of the conductivity measure-
ments for the same IL. As expected all ILs have a positive
influence on the conductivity of the reaction medium. The
conductivities achieved rise with increasing amount of IL.
2-(2-methoxyethoxy)ethysulfate
[MDEGSO4]ꢀ
1-methyl-3-octylimidazolium
[OMIM]+
The addition of [MMIM]ACHTNUGTRNEUGN[Me2PO4] led to the highest im-
provement. With 10 vol% an increase by a factor of 3.9 was
possible.
tetrafluoroborate [BF4]ꢀ
Furthermore, experiments concerning the activity and sta-
bility of two different alcohol dehydrogenase, which could
possibly be used in combination with the electrochemical re-
generation of NADPH were conducted. The results are
given in the Supporting Information. Since the coupling of
the enzymatic and the electrochemical reaction faces stabili-
ty issues, the process improvements from the IL additive
would be overcompensated by enzyme inactivation from the
mediator in this setup. Therefore, the preparative electro-
chemical generation of NADPH was not coupled with an
enzymatic reaction.[16]
1-ethyl-3-methyl-pyridinium
[EMPY]+
bromide [Br]ꢀ
1-ethyl-3-hydroxymethylpyridinium
[EM(OH)PY]+
Figure 1 summarises half lives of NADPH in presence of
2 to 10 vol% contents of different ILs. From these results it
is obvious, that in contrast to a lot of inorganic salts, ILs do
not have a destabilising effect on NADPH. On the contrary,
they are able to stabilise the reduced cofactor.[15]
As the most promising candidate [EMPY]ACTHNURGTNEUNG[EtSO4] was
chosen for the electrochemical reaction. Preparative scale
syntheses of NADPH were carried out with the addition of
different amounts of this IL as well as in pure buffer. The
results of these syntheses are summarised in Figure 3.
The addition of [EMPY]-
For each of the IL the stabilising effect increases with in-
creasing amount of IL. Best results were obtained with
A
to
the
reaction
medium significantly increases
the amount of NADP+ re-
duced per time and the turn-
over frequency (TOF=number
of regeneration cycles within a
defined time frame) of the
used rhodium mediator. Batch
experiments in pure buffer led
to a space time yield (STY=
amount of product per reaction
volume
and
time)
of
27 mmolLꢀ1 dꢀ1 and a TOFRhbpy
of 62 hꢀ1 while with 10 vol%
of [EMPY]ACHTUNRGTNE[UNG EtSO4] the STY
and the TOFRhbpy was nearly
tripled. The STY and the
TOFRhbpy increased linearly
with increasing amount of IL,
which is a clear evidence for
Figure 1. Half life of NADPH in the presence of different ILs. For each IL 2, 4, 6, 8 and 10 vol% of IL were
measured and grouped together (from left to right). The dashed line represents the half life in pure buffer
ACHTUNGTRENNUNG(~14.4 h).
11694
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 11692 – 11700

Ionic Liquids
FULL PAPER
enzyme (for the influence of
H2O2 on the used d-AAO, see
Supporting Information). Typi-
cally, catalase is added to de-
grade the peroxide produced.
However, in some cases the
stability of the flavo enzyme is
still limited.[17] Also, the reac-
tion medium has to meet the
conditions for both enzymes;
additionally, product separa-
tion or catalyst recycling might
be more difficult. Electrochem-
ical cofactor regeneration via a
ferrocene mediator represents
an anaerobic approach, which
avoids the formation of H2O2.
However, during the reaction
deactivation of the enzyme at
the electrode surface occurs.
Figure 2. Influence of different ILs on the conductivity of a 50 mmolLꢀ1 phosphate buffer pH 7. For each IL
an addition of 2, 4, 6, 8 and 10 vol% of IL was investigated and these results are grouped together (from left
to right). The dashed line represents the conductivity of the pure buffer (~7.3 mScmꢀ1).
Consequently, to improve this reaction system there is need
for a method to stabilise the biocatalyst, as well as an appro-
priate conducting salt to achieve higher productivities.
Again, these limitations were overcome by using ILs as per-
formance additives.
First, the influences of different ILs on the activity of d-
AAO were investigated. As indicated by Figure 4, the
enzyme was sensitive to these additives; only in three ILs an
activity higher than 75% was observed in the presence of
only 2 vol% IL. Best results were obtained with IL-contain-
ing phosphorylated anions. These results are in agreement
with investigations on the choice of the buffer medium; in
phosphate buffer the highest activities have been measured
(data not shown). In contrast to the results obtained by
Lutz–Wahl and co-workers,[18] who reported an increased ac-
tivity for isolated AAO in presence of 20 vol% of [MMIM]-
^
*
Figure 3. Influence of [EMPY]
of NADPH production.
ACTHUNGNRET[NUGN EtSO4] on the STY ( ) and TOFRhbpy ( )
the limitation of this reaction by the conductivity of the re-
action medium. Increasing the conductivity by using higher
concentration of the buffer salt
ACHTUGNRTNE[NUNG Me2PO4] no improvement was measured.
could result in reduced cofac-
tor stabilities. This underscores,
that ILs are synergistic perfor-
mance additives. However, the
obtained productivities are still
not high enough to compete
with other established cofactor
regeneration methods.[13]
d-AAO-catalysed resolution of
a
methionine racemate with
electrochemical regeneration
of FAD: FAD-dependent en-
zymes (e.g., oxidases) can be
regenerated
by
molecular
oxygen. However, H2O2 is
formed as a by-product of this
reaction, which deactivates the
Figure 4. Relative activity of d-AAO in presence of different ILs compared to the activity in pure buffer. For
each IL addition of 2, 4, 6, 8 and 10 vol% were measured and grouped together (from left to right). The
dashed line represents the activity in pure buffer (~9.9 Umgꢀ1).
Chem. Eur. J. 2009, 15, 11692 – 11700
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11695

S. Lꢀtz et al.
For those six ILs which showed highest remaining activi-
ties also stability investigations were carried out; results of
these experiments are presented in Figure 5. A clear desta-
bilisation was seen for both [MDEGSO4]ꢀ IL; even for only
4 vol% the half life was much lower than in pure buffer and
with higher amounts of IL the value decreased further. In
influence on the conductivity. With addition of 10 vol% the
conductivity was almost doubled.
Therefore it was obvious that [MMIM]ACTHNUTGRNEG[UN Me2PO4] showed
the best compromise in terms of enzyme performance and
conductivity. Thus, batch electrolyses were carried out in
presence of different amounts of this IL and compared with
the results obtained in pure buffer solution (see Figure 7).
presence of increased amounts of [BMIM]
bility decreased, too. However the half life reached was
higher than for the [MDEGSO4]ꢀ IL. For [MMIM]
[MeSO4],
[EMIM][Et2PO4] and [MMIM][Me2PO4] a contrary trend
ACHTUNGRTEN[NUNG MeSO4] the sta-
AHCTUNGTRENNUNG
A
ACHTUNGTRENNUNG
was observed; the stability increased with increased amount
of IL. Generally, the enzyme was more stable in these ILs
than in pure buffer.
^
*
Figure 7. Influence of [MMIM]ACHTUNGTENR[NUG Me2PO4] on STY ( ), TOFFcCOOH ( ) and
~
TTNd-AAO
(
) during d-AAO-catalysed resolution of a methionine race-
mate with electrochemical regeneration of FAD by ferrocene carboxylic
acid.
In pure buffer the STY was approximately 18 gLꢀ1 dꢀ1 and
the TOFFcCOOH for the ferrocene carboxylic acid was 8.6 hꢀ1,
whereas with 10 vol% [MMIM]ACHTNUGRTNEUNG[Me2PO4] a 1.5-fold increase
Figure 5. Half life of d-AAO in the presence of different ILs. For each IL
2, 4, 6, 8 and 10 vol% of IL were measured and grouped together (from
left to right). The dashed line represents the half life in pure buffer
of these numbers was found. The selectivity of the enzyme
was not influenced by the IL; d-methionine was converted
quantitatively resulting in an optical purity of ee >99.99%
for the l product. During the experiments improved TTNs
(TTN=produced amount of product in mol per mol biocat-
alyst) for the biocatalyst were observed with increasing
amounts of IL. With an addition of 10 vol% an increase of
TTNd-AAO by a factor of 2.4 was possible. These findings are
in line with the experiments showing that the enzyme was
also more stable during incubation in the presence of in-
ACHTUNGTRENNUNG(~45 h).
Since it was only possible to obtain a sufficient compro-
mise for activity and stability with [EMIM][Et2PO4] and
[MMIM][Me2PO4] conductivity measurements were carried
out for these two ILs. Figure 6 summarises the findings. It
can clearly be seen, that [MMIM][Me2PO4] has a stronger
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
creasing amount of this IL. Therefore, [MMIM]ACHTNUGTRNEG[UN Me2PO4]
does not only improve the storage stability of the enzyme,
but is also able to stabilise d-AAO under process conditions.
Thus, ILs can act as conducting salts and biocatalyst stabilis-
ers, although the electroenzymatic approach is not yet com-
petitive to the aerobic regeneration in terms of STY and
TTN.
CPO-catalysed synthesis of (R)-phenylmethylsulfoxide with
electrochemical generation of H2O2: Even if the CPO re-
quires H2O2 to be able to catalyse various reactions, high
concentrations of the co-substrate have to be avoided, since
CPO is both inhibited and deactivated by H2O2. The electro-
chemical reduction of O2 is a promising approach to gently
introduce the co-substrate, because H2O2 is dosed via a rela-
tively large surface, thus avoiding highly concentrated oxi-
dant.[12] However, the poor solubility of attractive substrates
Figure 6. Influence of two ILs on the conductivity of a 100 mmolLꢀ1
phosphate buffer pH 8. For each IL an addition of 2, 4, 6, 8 and 10 vol%
of IL were investigated and these results are grouped together (from left
to right). The dashed line represents the conductivity of the pure buffer
ACHTUNGTRENNUNG
(~16.2 mScmꢀ1).
11696
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 11692 – 11700

Ionic Liquids
FULL PAPER
limits the performance of the
synthesis. The application of
an organic co-solvent as solu-
biliser for the substrate
showed a drastic reduction of
the conductivity of the reac-
tion medium and was harmful
to the CPO at the same
time.[19] The solubility of the
substrate is crucial, since CPO
(compound I) is deactivated
rapidly in the absence of sub-
strate.[20] For these reasons,
there is need for a method to
stabilise the biocatalyst by
adding an appropriate solubil-
iser for the substrate and a
conducting salt to improve this
reaction system. Once more,
these limitations are overcome
by using ILs as performance
additives.
Figure 8. Relative activity of CPO in the presence of different ILs compared to the activity in pure buffer. For
each IL 2, 4, 6, 8 and 10 vol% of IL were measured and grouped together (from left to right). The dashed line
represents the activity in pure buffer (~191 Umgꢀ1).
In the beginning the influ-
ence of ten different ILs on
the activity of the CPO was in-
vestigated. The results of these
experiments are summarised in
Figure 8. As for the d-AAO all
ILs tested had a negative influ-
ence on the activity of the
CPO. With increasing amount
of IL the activity decreases.
Moreover, total inactivation of
the CPO occurred in presence
of pyridine cations as well as a
very drastic drop in activity for
[BMIM]ACHTUNGTRENNUNG[BF4]. Investigations
on the performance of CPO-
catalysing sulfoxidations in the
presence of different ILs have
also been carried out by
Figure 9. Half life of CPO in the presence of different ILs. For each IL 2, 4, 6, 8 and 10 vol% of IL were mea-
sured and grouped together (from left to right). The dashed line represents the half life in pure buffer (~45 d).
Chiappe et al.[21] In contrast to their work, we found that the
[EMIM]
ACHTUNGTRENNUNG
enzyme was active in presence of [MMIM]
G
theless even with small amounts of [EMIM]ACTHUNGTRENNUNG
the reasons explaining this fact can be that they investigated
higher contents of IL. However, Sanfilippo et al. reported a
tolerance of up to 30 vol% [MMIM]ACTHNUTRGNEUNG[MeSO4] for the CPO
tic increase in half life was measured. With more than
4 vol% of IL the half life of CPO was higher than in pure
buffer for all additives.
catalysing the oxidation of 1,2-dihydronaphthalene.[22] The
main reason might be that Chiappe et al.[23] did not always
adjust pH values of the IL/buffer mixture. In accordance to
their work we also found, that the enzyme tolerates higher
amounts of IL while catalysing the sulfoxidation of thioani-
sole than during chlorination reactions (data not shown).
Since seven of the tested ILs showed moderate CPO ac-
tivities, stability investigations were carried out for these
ILs. All results from these measurements are illustrated in
Figure 9. The stability of the CPO was increased with in-
creased amounts of IL for all investigated additives, only for
Since all ILs showed good results for CPO stability, also
conductivity and solubility measurements were carried out
for these ILs. Figure 10 summarises the results from the con-
ductivity measurements. As expected, all ILs had a positive
influence on the conductivity. With increasing amount of IL
the conductivity rises. Best results were obtained with
[MMIM]ACTHNURTGNENG[U MeSO4]; the addition of 10 vol% of this IL in-
creased the conductivity by the factor of 7.
Moreover, all ILs had a positive effect on the solubility of
the substrate thioanisole (see Figure 11). The addition of
merely 5 vol% IL led to an increase of at least 40% for
Chem. Eur. J. 2009, 15, 11692 – 11700
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11697

S. Lꢀtz et al.
Figure 10. Influence of different ILs on the conductivity of
a
100 mmolLꢀ1 acetate buffer pH 5. For each IL an addition of 2, 4, 6, 8
and 10 vol% of IL were investigated and these results are grouped to-
gether (from left to right). The dashed line represents the conductivity of
the pure buffer (~5.0 mScmꢀ1).
Figure 12. Influence of different ILs on the STY (light grey) and TTN
(dark grey) of CPO-catalysed syntheses of (R)-phenylmethylsulfoxide
with the electrochemical generation of H2O2.
and a maximum of 75 gLꢀ1 dꢀ1 for [EMIM]
ACTHNUGTRNE[UNG EtSO4] were ob-
served. In pure buffer, only a STY of 18 gLꢀ1 dꢀ1 could be
achieved. Also the catalyst utilisation was improved; in
buffer only a TTNCPO of 3300 was observed, while in the
presence of IL a minimum TTN of 78000 for [BMIM]-
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
TTN for CPO reported for other H2O2 supply systems.[24]
The influence on STY and TTN strongly depends on the
IL, since each IL leads to a different compromise between
enzyme performance, substrate solubility and conductivity.
While comparing the results from the syntheses with the in-
vestigations concerning enzyme performance, conductivity
and solubility, it becomes obvious, that especially the solu-
bility of thioanisole influences the reactions. For example,
these investigations showed relatively good results concern-
ing enzyme performance and conductivity, but only little
Figure 11. Influence of different ILs on the solubility of thioanisole in
100 mmolLꢀ1 acetate buffer pH 5. For each IL an addition
5 and
10 vol% of IL were investigated and these results are grouped together
(from left to right). The dashed line represents the solubility in pure
buffer (~3.8 mmolLꢀ1).
[MMIM]ACHTUNGTRENNUNG[Me2PO4]. Especially, [BMIM]ACHTUNTGERN[NUGN MeSO4] was able to
improve the solubility; 5 vol% of this IL led to an improve-
ment by a factor of 5.
effect on the substrate solubility for [MMIM]ACTHUNGTERNU[NG Me2PO4]. Ap-
plying this IL as performance additive for the electroenzy-
matic synthesis relatively high TTN were observed, but the
STY is still quite low, leading to the assumption, that the re-
action is not only limited by the electrochemical generation
rate of H2O2, but also by the solubility of the substrate. This
statement can be supported by the results obtained with
Taking all investigations into account, ILs are appropriate
performance additives for the CPO-catalysed oxidation of
thioanisole with electrochemical generation of the required
cosubstrate H2O2. Even if the chosen IL had a negative in-
fluence on the activity of the enzyme, some of them are able
to stabilise it tremendously. Furthermore, the IL can act as
solubilisers for thioanisole and have a positive influence on
the conductivity of the reaction medium at the same time.
Since the different ILs had different advantages, it was not
possible to select one IL which was superior. Therefore elec-
troenzymatic syntheses with an addition of 2 vol% IL have
been carried out for five of these ILs.
[BMIM]ACTHNURTGENNG[U MeSO4]. For this IL the highest improvement for
the substrate solubility was observed, leading to very good
TTN and STY. Taking all these results into account, ILs are
excellent performance additives for the reaction under in-
vestigation.
Figure 12 summarises the results for the syntheses in pure
buffer and with the different ILs. Depending on the IL an
improved STY of at least 35 gLꢀ1 dꢀ1 for [MMIM]
ACTHNUGTRNEUGN[Me2PO4]
11698
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 11692 – 11700

Ionic Liquids
FULL PAPER
Conclusion
We were able to successfully apply IL as performance addi-
tives for electroenzymatic syntheses for the first time. They
ideally combine the function of conducting salts, enzyme sta-
bilisers and co-solvents at the same time. The performance
of three completely different electroenzymatic reactions was
enhanced by small amounts of IL in the reaction medium.
Their addition led to higher conductivities of the reaction
media, stabilised biocatalysts and nicotine amide cofactors,
as well as to improved substrate solubility resulting in in-
creased productivities and improved catalyst utilisations. IL
as performance additives for electroenzymatic synthesis
seem to be able to overcome the main challenges of these
reactions and their application can be beneficial for various
syntheses. They define a new state of the art for electroenzy-
matic syntheses.
Scheme 4. Assay for the investigation of d-AAO activity.
Chloroperoxidase from Caldariomyces fumago: As activity assay the oxi-
dation of thioanisole to (R)-phenylmethylsulfoxide was investigated
(Scheme 3). Enzymatic activity was determined spectrophotometrically
measuring the difference in absorption at a wavelength of 284 nm during
one minute reaction time.[20] In the cuvette thioanisole solution
(2 mmolLꢀ1, 980 mL) and H2O2 solution (200 mmolLꢀ1
, 10 mL) were
mixed and incubated at 258C for 5 min; with the addition of enzyme so-
lution (10 mL) the reaction was started. As enzyme solution a combined
stock of three different commercial batches of CPO were used. The com-
bined enzyme stock had a protein content of 34.5 mgmLꢀ1 and an RZ
value of 0.92 (1.44 for pure CPO) corresponding to a total CPO content
Experimental Section
of 22 mgmLꢀ1
. In the standard assay without IL the activity was
Chemicals: All IL were supplied by Solvent Innovation (Cologne, Ger-
many). Chloroperoxidase from Caldariomyces fumago (E.C. 1.11.1.10)
and horseradish peroxidase (E.C. 1.11.1.7) were purchased from Fluka
(Taufkirchen, Germany). d-Specific amino acid oxidase from Trigonopsis
variabilis (E.C. 1.4.3.3), NADPH and NADP+ were provided by Jꢀlich
Chiral Solutions (Jꢀlich, Germany). All other reagents were purchased
from Sigma Aldrich (Schnelldorf, Germany) and were of analytical grade
or better.
191 Umgꢀ1
regarding thioanisole as substrate and 930 Umgꢀ1
re-
CPO
CPO
garding the MCD assay.[23]
Stability investigations
Cofactor stability: To measure the stability of NADPH, cofactor (1.5 mg)
was diluted in 2 mL of the corresponding IL/buffer mixture and stored at
258C. Samples were withdrawn in defined time periods and the remain-
ing absorption at 340 nm was measured. Additionally, for some of the
samples the remaining amount of cofactor was investigated by a photo-
metric enzyme coupled assay. For this assay, the absorption of a mixture
from acetophenone (30 mmolLꢀ1, 970 mL) and cofactor solution (20 mL)
was determined at 340 nm. Then alcohol dehydrogenase solution
(50 mgmLꢀ1, 10 mL) was added. After reaching a stable value for the ad-
sorption at 340 nm the amount of enzymatically active NADPH was cal-
culated from the difference of these two values.
To measure the stability of NADP+, cofactor (2 mg) was diluted in 2 mL
of the corresponding IL/buffer mixture and stored at 258C. Samples were
withdrawn in defined time periods and the remaining amount of enzy-
matically active cofactor was investigated by a photometric enzyme
assay. For that reason, the absorption of a mixture from sodium formate
(240 mmolLꢀ1, 970 mL) and cofactor solution (20 mL) was determined at
340 nm. Then formate dehydrogenase solution (10 mL) was added. After
reaching a stable value for the adsorption at 340 nm the amount of enzy-
matically active NADP+ was calculated from the difference of these two
values.
Methods: All results represent the average of at least two experiments.
The standard deviation was always less then 5%.
Preparation of IL/buffer mixtures: Depending on the reaction different
buffers were used: All investigations concerning the electrochemical gen-
eration of NADPH were carried out in phosphate buffer (50 mmolLꢀ1
,
pH 7). Investigations regarding the d-AAO-catalysed resolution of a me-
thionine racemate with electrochemical regeneration of FAD were car-
ried out in phosphate buffer (100 mmolLꢀ1, pH 8). In all experiments for
the CPO-catalysed synthesis of (R)-phenylmethylsulfoxide with electro-
chemical generation of H2O2 acetate buffer (100 mol lꢀ1, pH 5) was used.
The addition of some IL to the three buffers led to shifts in the pH of
ꢁ0.3 pH units. Therefore after addition of IL the pH was controlled and,
if necessary, fixed by addition of small amounts of phosphoric or acetic
acid respectively or the corresponding bases to maintain the desired pH.
Conductivity measurements: The conductivity was measured with a con-
ductivity meter (Cond 315i, WTW Weinheim, electrode TetraCon 325) at
a temperature of 298 K (258C).
Enzyme stability: The different enzymes were stored in the correspond-
ing IL/buffer mixtures at 258C; samples were withdrawn in defined time
periods and analysed for activity via the photometric assays described
previously (see above).
Enzyme activity investigations
d-Amino acid oxidase from Trigonopsis variabilis: As activity assay the
conversion of d-alanine to the corresponding imino acid 2-iminopropano-
ic acid, followed by the spontaneous hydrolysis to a-keto-g-(methylthio)-
Electroenzymatic syntheses
butyric acid was investigated with
a coupled peroxidase assay
(Scheme 4).[24] As H2O2 was formed during the amino acid oxidase cata-
lysed reaction, the H2O2 could be used to oxidise o-dianisidine by horse
radish peroxidase (HRP) leading to an auburn colouration. Enzymatic
activity was determined spectrophotometrically measuring the difference
in absorption at a wavelength of 436 nm during one minute reaction
time. Alanine and dianisidine solution were aerated with oxygen for
30 min before use and incubated at 258C for 5 min afterwards. One milli-
liter cuvette volume contained d,l-alanine (10 mmolLꢀ1), o-dianisidi-
ne·2HCl (0.2 mgmLꢀ1), horse radish peroxidase (~25 U) and d-amino
acid oxidase (~5 U). We also checked the activity of the HRP in pres-
ence of the used IL to ensure, that the HRP was more active than the d-
AAO.
Reaction setup for all electroenzymatic syntheses: Reactions where carried
out in a thermostated glass reaction vessel at 258C (workshops For-
schungszentrum Jꢀlich, Jꢀlich, Germany). A graphite fleece on a stainless
steal fixation was used as working electrode, a platinum net separated by
a dialysis sack functioned as counter electrode and as reference an Agj
AgCl electrode was used. The electrodes were connected to a potentio-
stat (263 A, Princeton Applied Research).
Electrochemical generation of NADPH: The reaction medium (x vol%
50 mm phosphate buffer pH 7 + y vol% [EMPY]ACTHNUTRGENUG[N EtSO4]; 200 mL) was
degassed with argon before use. NADP+ (2.5 mm) and rhodium mediator
(0.05 mm; prepared by the method of Kçlle and Grꢂtzel)[25] were added
and the reaction was started by applying a potential of ꢀ750 mV vs. Agj
Chem. Eur. J. 2009, 15, 11692 – 11700
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11699

S. Lꢀtz et al.
AgCl. Samples were withdrawn and analysed for concentration spectro-
photometrically at 340 nm.
Chem. 2002, 4, 147–151; c) F. van Rantwijk, R. A. Sheldon, Chem.
Rev. 2007, 107, 2757–2785; d) H. Y. Xiong, T. Chen, X. H. Zhang,
S. F. Wang, Electrochem. Commun. 2007, 9, 1648–1654; e) C.
Roosen, P. Mꢀller, L. Greiner, Appl. Microbiol. Biotechnol. 2008,
81, 607—614.
d-AAO-catalysed resolution of a methionine racemate with electrochemi-
cal regeneration of FAD: The reaction medium (x vol% 100 mm phos-
phate buffer pH 8 + y vol% [EMIM]ACHTNUGTRNEUNG[Me2PO4]; 300 mL) was degassed
with argon before use. Ferrocene carboxylic acid (0.5 mm), d,l-methio-
nine (20 mm) and d-AAO were added and the reaction was started by ap-
plying a potential of 350 mV vs. AgjAgCl. Samples were withdrawn and
analysed by HPLC (column: Crownpack CR+, perchloric acid pH 2,
0.9 mLminꢀ1, 408C).
[9] C. Kohlmann, L. Greiner, W. Leitner, C. Wandrey, S. Lꢀtz,
DE 102007044379A1, 2007.
[10] a) E. Steckhan, S. Herrmann, R. Ruppert, J. Thommes, C. Wandrey,
Angew. Chem. 1990, 102, 445–447; Angew. Chem. Int. Ed. Engl.
1990, 29, 388–390; b) E. Steckhan, S. Herrmann, R. Ruppert, E.
Dietz, M. Frede, E. Spika, Organometallics 1991, 10, 1568–1577;
c) E. Steckhan, M. Frede, S. Herrmann, R. Ruppert, E. Spika, E.
Dietz, DECHEMA Monogr. 1992, 125, 723–753; d) F. Hildebrand,
C. Kohlmann, A. Franz, S. Lꢀtz, Adv. Synth. Catal. 2008, 350, 909–
918.
[11] a) M. Frede, E. Steckhan, Tetrahedron Lett. 1991, 32, 5063–5066;
b) E. Steckhan, M. Frede, S. Herrmann, R. Ruppert, E. Spika, E.
Dietz, DECHEMA Monogr. 1992, 125, 723–753.
[12] a) S. Lꢀtz, E. Steckhan, A. Liese, Electrochem. Commun. 2004, 6,
583–587; b) C. Kohlmann, S. Lꢀtz, Eng. Life Sci. 2006, 6, 170–174;
c) S. Lꢀtz, K. Vuorilehto, A. Liese, Biotechnol. Bioeng. 2007, 98,
525–534.
CPO-catalysed synthesis of (R)-phenylmethylsulfoxide with electrochemi-
cal generation of H2O2: The reaction medium (98 vol% 100 mm acetate
buffer pH 8 + 2 vol% IL; 300 mL) was saturated with O2 before use
and aerated with O2 during the whole experiment. Thioanisole (20 mm)
and CPO were added and the reaction was started by applying a poten-
tial of ꢀ650 mV vs. AgjAgCl. Every 30 min thioanisole (6 mmol) was in-
jected. Samples were withdrawn and analysed by GC (column: Lipodex
G, hydrogen, 0.6 bar, 1608C (10 min) 208Cminꢀ1 to 1808C (7 min)).
[13] a) U. Kragl, Biotechnol. Bioeng. 1996, 52, 309–319; b) T. Stillger, M.
Bonitz, M. Villela, A. Liese, Chem. Ing. Tech. 2002, 74, 1035–1039;
c) M. Eckstein, T. Daussmann, U. Kragl, Biocatal. Biotransform.
2004, 22, 89–96; d) “Cofactor regeneration at the lab scale”: R.
Wichmann, D. Vasic-Racki in Technology Transfer in Biotechnology:
From Lab to Industry to Production, Vol. 02, Springer, Heidelberg,
2005, pp. 225–260; e) K. Goldberg, K. Schroer, S. Lꢀtz, A. Liese,
Appl. Microbiol. Biotechnol. 2007, 76, 237–248.
Acknowledgements
The authors thank Lilia Hꢂrter, Stephanie Corsten and Zꢀbeyda Dogan
for excellent work in the lab and the GK 1166 “BioNoCo” for fruitful
discussion and funding. L.G. thanks the cluster of excellence “Tailor-
Made Fuels from Biomass” (TMFB) for support.
[14] H. K. Chenault, G. M. Whitesides, Appl. Biochem. Biotechnol. 1987,
14, 147–197.
[15] C. Kohlmann, L. Greiner, S. Lꢀtz, Verwendung von ionischen Flꢀs-
sigkeiten sowie Sensor, DE 10 2008 061866 7, 2008.
[16] F. Hildebrand, S. Lꢀtz, Chem. Eur. J. 2009, 15, 4998–5001.
[17] E. Steckhan, M. Frede, S. Herrmann, R. Ruppert, E. Spika, E.
Dietz, DECHEMA Monogr. 1992, 125, 723–753.
[18] S. Lutz-Wahl, E. M. Trost, B. Wagner, Manns, A. Fischer, J. Biotech-
nol. 2006, 124, 163–171.
[19] S. Lꢀtz, K. Vuorilehto, A. Liese, Biotechnol. Bioeng. 2007, 98, 525–
534.
[20] M. P. J. van Deurzen, F. van Rantwijk, R. A. Sheldon, Tetrahedron
1997, 53, 13183–13220.
[21] C. Chiappe, L. Neri, D. Pieraccini, Tetrahedron Lett. 2006, 47, 5089–
5093.
[22] C. Sanfilippo, N. D’Antona, G. Nicolosi, Biotechnol. Lett. 2004, 26,
1815–1819.
[1] a) P. V. Bernhardt, Aust. J. Chem. 2006, 59, 233–256; b) S. W. May,
Curr. Opin. Biotechnol. 1999, 10, 370–375.
[2] a) “Electroenzymatic Synthesis”: E. Steckhan, Top. Curr. Chem.
1994, 170, 83–111; b) F. Hollmann, A. Schmid, Biocatal. Biotrans-
form. 2004, 22, 63–88; c) C. Kohlmann, W. Mꢂrkle, S. Lꢀtz, J. Mol.
Catal. B 2008, 51, 57–72.
[3] R. Ruinatscha, V. Hçllrigl, K. Otto, A. Schmid, Adv. Synth. Catal.
2006, 348, 2015–2026.
[4] a) F. Endres, Z. Phys. Chem. (Muenchen Ger.) 2004, 218, 255–283;
b) M. Galinski, A. Lewandowski, I. Stepniak, Electrochim. Acta
2006, 51, 5567–5580.
[5] a) S. H. Schçfer, N. Kaftzik, P. Wasserscheid, U. Kragl, Chem.
Commun. 2001, 425–426; b) P. Lozano, T. De Diego, D. Carrie, M.
Vaultier, J. L. Iborra, Biotechnol. Lett. 2001, 23, 1529–1533; c) N.
Kaftzik, P. Wasserscheid, U. Kragl, Org. Process Res. Dev. 2002, 6,
553–557; d) S. Lutz-Wahl, E. M. Trost, B. Wagner, A. Manns, L.
Fischer, J. Biotechnol. 2006, 124, 163–171.
[6] a) P. Wasserscheid, W. Keim, Angew. Chem. 2000, 112, 3926–3945;
Angew. Chem. Int. Ed. 2000, 39, 3772–3789; b) U. Kragl, M. Eck-
stein, N. Kaftzik, Curr. Opin. Biotechnol. 2002, 13, 565–571.
[7] P. Wasserscheid, T. Welton, Ionic Liquids in Synthesis; Wiley-VCH,
Weinheim, 2003.
[23] M. P. J. Van Deurzen, K. Seelbach, F. van Rantwijk, U. Kragl, R. A.
Sheldon, Biocatal. Biotransform. 1997, 15, 1–16.
[24] B. Geueke, W. Hummel, Enzyme Microb. Technol. 2002, 31, 77–87.
[25] U. Kçlle, M. Grꢂtzel, Angew. Chem. 1987, 99, 572–574; Angew.
Chem. Int. Ed. Engl. 1987, 26, 567–570.
[8] a) T. Welton, Chem. Rev. 1999, 99, 2071–2083; b) R. A. Sheldon,
R. M. Lau, M. J. Sorgedrager, F. van Rantwijk, K. Seddon, Green
Received: April 20, 2009
Published online: September 23, 2009
11700
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 11692 – 11700