Biocatalytic oxidation of benzyl alcohol to benzaldehyde via hydrogen transfer

Article abstract of DOI:10.1016/j.tet.2009.06.088

Various types of biocatalysts like oxidases, alcohol dehydrogenases, and microbial cells were tested for the oxidation of benzyl alcohol. Oxidases in combination with molecular oxygen led to low conversion. Alcohol dehydrogenases and microbial cells were tested in a hydrogen transfer reaction employing acetaldehyde, chloroacetone, and acetone as hydrogen acceptor. Excellent conversion (95%) could be achieved employing lyophilised cells of Janibacter terrae DSM 13953 at a substrate concentration of 97 mM.

Full text of DOI:10.1016/j.tet.2009.06.088

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Tetrahedron 65 (2009) 6805–6809
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Biocatalytic oxidation of benzyl alcohol to benzaldehyde via hydrogen transfer
a
b
a
b
c
´
Thomas Orbegozo , Ivan Lavandera , Walter M.F. Fabian , Barbara Mautner , Johannes G. de Vries ,
,
Wolfgang Kroutil a
*
a Department of Chemistry, Organic and Bioorganic Chemistry, University of Graz, Heinrichstr. 28, A-8010 Graz, Austria
b Research Centre Applied Biocatalysis, c/o Department of Chemistry, Organic and Bioorganic Chemistry, University of Graz, Heinrichstrasse 28, 8010 Graz, Austria
c DSM Pharmaceutical Products-Innovative Synthesis & Catalysis, PO Box 18, 6160 MD Geleen, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Various types of biocatalysts like oxidases, alcohol dehydrogenases, and microbial cells were tested for
the oxidation of benzyl alcohol. Oxidases in combination with molecular oxygen led to low conversion.
Alcohol dehydrogenases and microbial cells were tested in a hydrogen transfer reaction employing ac-
etaldehyde, chloroacetone, and acetone as hydrogen acceptor. Excellent conversion (95%) could be
achieved employing lyophilised cells of Janibacter terrae DSM 13953 at a substrate concentration of
97 mM.
Received 5 June 2009
Accepted 23 June 2009
Available online 26 June 2009
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
OH
O
oxidase
H2O2
catalase
Ph
Ph
H
Oxidation of alcohols to yield carbonyl compounds is one of the
most fundamental and important processes in synthetic organic
chemistry. In the search for alternatives driven by the immaturity of
many organic oxidation reactions1–5 a lot of emphasis has been put
on the development of ‘green’ chemical processes.6–11 For instance
for the biocatalytic oxidation of primary alcohols laccases in com-
bination with mediators12–14 as well as various redox enzymes have
been employed.15–21 Here we describe the chemoselective bio-ox-
idation of benzyl alcohol to the corresponding aldehyde avoiding
overoxidation to benzoic acid using oxidases, isolated alcohol de-
hydrogenases (ADHs), and microbial cells. For the last two options
various formal hydrogen acceptors were studied such as acetalde-
hyde, chloroacetone as well as acetone.
1
O2
2
H2O +1/2 O2
Scheme 1. Oxidation employing an oxidase and a catalase.
Out of the fourteen oxidases tested only four showed low con-
version of approximately 5%. The active oxidases originated from
Pichia pastoris, Candida boidinii, Hansenula sp. as well as the ga-
lactose oxidase from Dactylium dendroides. Since this was below
our expectations, we tested commercial alcohol dehydrogenases
(ADHs) for the oxidation of benzyl alcohol.
2. Results and discussion
2.1. Oxidases
2.2. Commercial alcohol dehydrogenases
A significant number (>200) of alcohol dehydrogenases are
commercially available.26 Out of these, a library of 33 selected ADHs
were tested in a biocatalytic hydrogen transfer reaction employing
acetaldehyde as hydrogen acceptor (Scheme 2). Acetaldehyde has
rarely been employed as hydrogen acceptor,27–29 probably due to
its supposed inhibition effect on enzymes. However, employing it
in excess in the testing allows identifying easily enzymes which are
stable in the presence of the reactive aldehyde moiety. Additionally
our aim was to identify a single biocatalyst/enzyme, which per-
formed the desired oxidation as well as the recycling of the cofactor,
which is in contrast to approaches where an additional enzyme is
used for the recycling.26
As a first approach we tested a library of fourteen commercial
alcohol oxidases for the oxidation of benzyl alcohol to benzalde-
hyde. Alcohol oxidases22–25 have been efficiently employed for the
oxidation of n-alkanols requiring just molecular oxygen as oxidant
(Scheme 1). Hydrogen peroxide is formed as side product, which
can be disproportioned by a catalase yielding water and molecular
oxygen.
* Corresponding author. Tel.: þ43 316 380 5350; fax: þ43 316 380 9840.
E-mail address: Wolfgang.Kroutil@uni-graz.at (W. Kroutil).
0040-4020/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2009.06.088

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
6806
T. Orbegozo et al. / Tetrahedron 65 (2009) 6805–6809
OH
The low conversion achieved employing acetone (3%) can be
attributed to the non-favoured reaction equilibrium. Calculating
the relative Gibbs free energies (MP2/cc-pVTZ//MP2/cc-pVDZ) for
a model-reactiondacetone plus ethanol leading to 2-propanol and
acetaldehydedshowed that the equilibrium is far on the left side,
thus the formation of a primary alcohol is highly favoured over the
O
Ph
Ph
H
1
2
NAD(P)+ Biocat.
NAD(P)H
O
OH
R
R
formation of a secondary alcohol (DG w4 kcal/mol). The related
R = H, CH3, CH2Cl
model reaction between ethanol and chloroacetone leading to 1-
chloro-2-propanol and acetaldehyde has a
D
G wꢀ1 kcal/mol,32
Scheme 2. Oxidation of benzyl alcohol via biocatalytic hydrogen transfer.
thus the equilibrium of the oxidation reaction employing chloro-
acetone is slightly on the side of the desired aldehyde. Therefore
chloroacetone should actually work very nicely as oxidant with
respect to thermodynamics, so either sterical hindrance (ketone
versus aldehyde) leads to the observed much slower reaction or
chloroacetone is too reactive leading to destruction/inhibition of
the enzyme.
2.2.1. Testing of ADHs
For preparative significance substrate 1 was used at a concen-
tration of 97 mM with an almost 10-fold excess of acetaldehyde
(951 mM) as hydrogen acceptor to ensure shifting the equilibrium
to benzaldehyde. Furthermore the elevated acetaldehyde concen-
tration represents an additional selection parameter, namely to
select for stability in the presence of the aldehyde. Although many
new ADHs have become commercially available recently, the ‘old’
and frequently employed horse liver ADH (HLADH)30 preparation
displayed the highest activity (Table 1). The ADH-A from Rhodo-
coccus ruber showed only low activity for primary alcohols which is
in accordance with previous reports where sec-alcohols are pref-
erentially transformed.31
2.2.3. Concentration of acetaldehyde
Since acetaldehyde was the best hydrogen acceptor the con-
version at varied concentrations of acetaldehyde was measured
(Fig. 2). Without acetaldehyde no oxidation was found, which
clearly indicated that it was required for the transformation.
100
90
80
70
60
50
40
30
20
10
0
Table 1
Oxidation of 1 via hydrogen transfer employing ADHs and acetaldehyde
ADHa
conv.b,c
Horse liver ADH
R. ruber ADH-A
PADH 101
PADH 102
PADH 103
KRED 117
KRED 118
KRED 123
KRED 126
þþþþ
þ
þþ
þ
þ
þþ
þ
þ
0
2
4
6
Acetaldehyde (eq.)
8
10
12
þ
a
PADHs and KREDs are commercially available from Codexis (PADH¼primary
ADH, KRED¼keto reductases). R. ruber ADH-A is described in literature29 and is also
commercially available from Codexis.
Figure 2. Biocatalytic oxidation of 1 at varied equivalents of hydrogen acceptor. Re-
action conditions: HLADH (3 mg), NADþ (3 mM), Pi buffer (600
(97 mM), acetaldehyde, 30 ꢁC, 120 rpm, 23 h.
mL, pH 7.5, 100 mM), 1
b
Reaction conditions: crude ADH preparation (7.5 mg), 1 mM NA(P)Dþ, Pi buffer
(600 m
L, pH 7.5, 50 mM), 1 (97 mM), acetaldehyde (951 mM), 30 ꢁC, 120 rpm, 23 h.
c
þþþþ indicates >50% conv., þþþ: 30–50%, þþ: 10–30%, þ: 5–10%.
The highest conversion after 23 h was achieved employing
2.5 equiv of acetaldehyde, increasing the amount of acetaldehyde
led to lower conversions. Although at higher concentrations of
acetaldehyde the equilibrium should be shifted to higher conver-
sions, the transformations slowed down most likely due to in-
hibition of the HLADH.
2.2.2. Alternative hydrogen acceptors
Having identified HLADH as suitable ADH various alternative
hydrogen acceptors were tested. From the three alternative ac-
ceptors (acetaldehyde, chloroacetone, acetone) studied (Fig. 1),
acetaldehyde clearly worked best leading to 50% conversion and
chloroacetone was second.
2.2.4. Oxidation employing molecular oxygen and an
NADPH-oxidase
60
50
40
30
20
10
0
Molecular oxygen as oxidant would be highly favourable due to
its high redox potential which would lead to improved conversion.
For this purpose we coupled the HLADH-catalysed oxidation with
the recycling of NADþ by the oxidase YcnD from Bacillus subtilis33
(Scheme 3). Although YcnD shows a preference for NADPH, sepa-
rate experiments showed that it also transforms NADH at slightly
reduced rate.
OH
O
ADH
Ph
Ph
H
1
2
NAD(P)+
NAD(P)H
Acetaldehyde
Chloroacetone
Cosubstrate
Acetone
O2
H2O2
NAD(P)-oxidase
YcnD
catalase
Figure 1. Alternative hydrogen acceptors employed in the hydrogen transfer process
for the oxidation of 1. Reaction conditions: HLADH (7 mg), NADþ (3 mM), Pi buffer
mL, pH 7.5, 100 mM), 1 (97 mM), acetaldehyde (951 mM), chloroacetone
(669 mM), or acetone (728 mM), 30 ꢁC, 120 rpm, 23 h.
1/2 O2 + H2O
(600
Scheme 3. Recycling of NAD(P)þ by an NAD(P)-oxidase.

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
T. Orbegozo et al. / Tetrahedron 65 (2009) 6805–6809
6807
Although the oxidation of primary alcohols to aldehydes and
further on to acids has been previously shown to work successfully
with NAD(P)-oxidases,19 the oxidation of 1 to benzaldehyde 2 led
only to low conversion (4%) employing HLADH (Table 2). Addition
of catalase had no measurable effect for HLADH; however, testing
the commercial primary ADH PADH103 the addition of catalase led
to a clear improvement of conversion (11%, 6% without catalase).
Another improvement was achieved by using larger reaction tubes
providing larger gas volume and therefore a higher excess of oxy-
gen. By doubling the gas volume an increase of conversion up to
21% for PADH103 was achieved, although the amount of oxygen
was not limiting in any of the experiments.
carboxylic acid is a rather difficult task in organic chemistry. Even
selected alcohol dehydrogenases have been shown to oxidise
aldehydes to carboxylic acids.34–36 Therefore, we tested the re-
action by J. terrae specifically for the formation of benzoic acid. We
could not detect any benzoic acid, which indicated us that J. terrae
acts as a highly chemoselective catalyst. Therefore this strain was
used further to investigate the optimum of the reaction conditions.
2.3.1. Optimisation of reaction conditions
As a first parameter the influence of the pH on conversion was
investigated. The lyophilised cell preparation of J. terrae could be
employed over a broad pH range with highest conversion from pH
7.5 even up to basic conditions (pH 10.5) (Fig. 3).
Table 2
Oxidation of 1 employing ADHs and NADPH-oxidase YcnD
100
80
60
40
20
0
ADH
cofactor
pH
tubea (mL)
catalase
c (%)b
HLADH
HLADH
NADþ
7.5
7.5
8.5
8.5
8.5
8.5
8.5
1.2
1.2
2
d
þ
þ
þ
d
þ
þ
4
4
3
5
6
NADþ
PADH101
PADH102
PADH103
PADH103
PADH103
NADPH
NADPH
NADPH
NADPH
NADPH
2
1.2
1.2
2
11
21
a
Size of eppendorf tube.
b
Reaction conditions: HLADH (3 mg)/PADH (1.8 mg), NADþ (3 mM), YcnD (10
M), Pi buffer (600
L, pH 7.5, 100 mM), 1 (97 mM), catalase (4.5 mg), 30 ꢁC,
120 rpm, 22 h.
mL,
13
m
m
5
6
7
8
pH
9
10
11
Since PADH103 was not sufficient active or stable (Table 1) and
since we experienced severe limitations with the commercial
supply of HLADH we started a screening of commercial microbial
strains to identify a better biocatalyst.
Figure 3. Biocatalytic oxidation of 1 at varied pH employing lyophilised cells of J. terrae
and acetaldehyde. Reaction conditions: Cells (24 mg), acetaldehyde (951 mM), Pi buffer
(600 m
L, 100 mM), 1 (97 mM), 30 ꢁC, 120 rpm, 23 h.
2.3. Microbial cells
Testing the conversion at varied equivalents of acetaldehyde
(Fig. 4) showed that the reaction reaches highest conversion
quickest already with five or even 2.5 equiv of acetaldehyde. Al-
though higher amounts of acetaldehyde should have led to a higher
conversion due to a shift of the equilibrium, the results suggested
that more than 5 equiv lead to inhibition or destruction of the
catalyst as already observed for HLADH (see Section 2.2.3). Without
acetaldehyde no oxidation was found, which clearly indicated that
it was required for the transformation.
A total of 218 micro-organisms (bacteria, yeasts stored in
lyophilised form) were tested for their ability to oxidise benzyl
alcohol 1 at the expense of acetaldehyde in a hydrogen transfer like
fashion. The library consists of pre-selected strains, which were
chosen for their known ability to stand organic chemicals or ca-
talyse chemical transformations of interest. Seventeen strains were
active showing a conversion above 5% (Table 3).
Table 3
Oxidation of 1 via hydrogen transfer employing lyophilised microbial strains and
acetaldehyde
100
80
60
40
20
0
Microbial straina
cb
Gordonia alkanivorans DSM 44369
J. terrae DSM 13953
þþ
þþþþ
þþþþ
þþþ
þþþþ
þ
Mycobacterium gilvum DSM 9487
Norcardia corynebacterioides DSM 20151
Norcardia nova DSM 43843
Pseudomonas cichorii DSM 50259
Pseudomonas elodea ATCC 31461
Pseudomonas sp. DSM 6978
Pseudomonas syringae DSM 1241
Ralstonia sp. DSM 6428
Ralstonia sp. DSM 9750
Arthrobacter sp. DSM 312
R. ruber DSM 44190
R. ruber DSM 44540
R. ruber DSM 43338
þþ
þþ
þþþ
þþ
0
2
4
6
Acetaldehyde (eq.)
8
10
12
þ
þþ
þþ
þþ
Figure 4. Variation of amount of hydrogen acceptor. Reaction conditions: Cells
(24 mg), acetaldehyde, Pi buffer (600
16 h.
m
l, pH 8.5, 100 mM), 1 (97 mM), 30 ꢁC, 120 rpm,
þþþ
þ
R. ruber DSM 44541
R. ruber DSM 44539
þþ
a
Strains are commercially available from the German culture collection (DSMZ) or
the American type culture collection (ATCC).
From the time course of the reaction it was concluded that the
oxidative hydrogen transfer of benzyl alcohol 1 already reached its
highest value after 5-h reaction time (Fig. 5).
b
þþþþ indicates >60% conv., þþþ: 30–60%, þþ: 10–30%, þ: 5–10%.
The most active strain was Janibacter terrae DSM 13953. The
chemo-selective oxidation of a primary alcohol to yield exclusively
the aldehyde without overoxidation to the corresponding
Therefore, to get a clearer picture for the optimum of the tem-
perature the transformation of 1 was stopped already after 2 h at
varied temperature (Fig. 6). The observed conversion increased

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
6808
T. Orbegozo et al. / Tetrahedron 65 (2009) 6805–6809
100
95
90
85
80
75
70
65
60
reaction time. The derivatives possessing a substituent in ortho
position were the poorest substrates probably due to sterical
hindrance.
In conclusion, we have identified J. terrae DSM 13953 as a suit-
able biocatalyst for the chemoselective oxidation of benzyl alcohol
derivatives to the corresponding aldehydes avoiding the formation
of the corresponding carboxylic acids via hydrogen transfer
employing acetaldehyde as hydrogen acceptor.
0
5
10
15
20
25
3. Experimental part
3.1. General
Time (h)
Figure 5. Time course of oxidation of 1 via hydrogen transfer. Reaction conditions:
Cells (800 mg), acetaldehyde (951 mM), Pi buffer (24 mL, pH 7.5, 100 mM), 1 (97 mM),
30 ꢁC, 120 rpm.
Acetaldehyde, benzyl alcohol, NADþ as well as HLADH were
purchased from Fluka (Sigma–Aldrich–Fluka, Vienna, Austria).
Substituted benzyl alcohols were purchased from Sigma–Aldrich
(Sigma–Aldrich–Fluka, Vienna, Austria). Enzymes named KRED and
PADH as well as NADPþ were obtained from Codexis (Redwood city,
USA). J. terrae DSM 13953 obtained from the Deutsche Sammlung
fu¨r Mikroorganismen und Zellkulturen (DMSZ, Braunschweig,
Germany, http://www.dmsz.de). ADH-A was employed as lyophi-
lised E. coli powder and prepared as previously described.30 Tryp-
ticase soy broth (Sigma T-8907) and yeast extract (Oxoid L21) were
purchased from Thermo Fisher (Vienna, Austria).
100
80
60
40
20
0
3.1.1. Preparation of lyophilised cells of J. terrae DSM 13953
Cultivation medium M92: Trypticase soy yeast extract medium:
30 g/L Trypticase soy broth (Sigma T-8907), 3 g/L yeast extract
(Oxoid L21), pH 7.0.
30
35
40
45
T (°C)
50
55
60
J. terrae was cultivated in M92 medium (330 mL) in baffled
shake flasks (1 L) at 120 rpm for three days at 28 ꢁC. The cells were
harvested by centrifugation (18,000 g) and washed twice with so-
dium phosphate buffer (50 mM, pH 7.5) before the cells were sus-
pended in a minimum amount of the above buffer, shock frozen
with liquid nitrogen and freeze dried.
Figure 6. Dependency of conversion on temperature for the oxidation of 1 via hy-
drogen transfer. Reaction conditions: Cells (20 mg), acetaldehyde (951 mM), Pi buffer
(600 mL, pH 7.5, 100 mM), 1 (97 mM), 2 h, 120 rpm.
from 71% at 30 ꢁC to reach its highest value at 37 ꢁC (90% conver-
sion). Even at 60 ꢁC the J. terrae preparation showed still reasonable
conversion (66% conv.).
Chloroacetone and propanal were tested as alternative hydro-
gen acceptors. Chloroacetone reacted slower than acetaldehyde
just leading to 65% conversion (95% for acetaldehyde). Propanal on
the other hand was equally suitable as acetaldehyde leading to the
same conversion (95%).
3.2. Experimental procedures
3.2.1. Biocatalytic oxidation employing HLADH and acetaldehyde
Alcohol dehydrogenase (3 mg) and NADþ (1.8
mmol) were sus-
pended in phosphate buffer (0.6 mL, 100 mM, pH 7.5) in eppendorf
tubes. The enzyme preparations were rehydrated by shaking at
The oxidation of benzyl alcohol was also demonstrated on
a 40-fold larger scale (250 mg) leading again to 95% conversion
within 6 h.
30 ꢁC, 120 rpm for 30 min. After addition of benzyl alcohol (6
m
L,
6.24 mg, 58
m
mol) the reaction mixture was shaken at 30 ꢁC and
120 rpm for 23 h. The reaction was stopped by extraction with ethyl
acetate (2 ꢂ 0.5 mL) and centrifugation (12,000 rpm, 2 min). The
organic phase was dried (Na2SO4) prior to determination of con-
version by GC.
2.3.2. Other benzyl alcohol derivatives
To test whether other benzyl alcohol derivatives were accepted
as well, various ortho-, meta- and para-substituted derivatives were
tested (Table 4). From the substrates tested the meta-derivatives
were transformed fastest leading to highest conversions within the
3.2.2. Biocatalytic oxidation employing HLADH and YcnD
Alcohol dehydrogenase (3 mg), the cofactor (1.8
mmol),
NAD(P)H-oxidase [YcnD from B. subtilis] (10 L, 13 M) were sus-
m
m
Table 4
pended in phosphate buffer (0.6 mL, 100 mM, pH 7.5) in eppendorf
tubes with or without catalase (4.5 mg). The enzyme preparations
were rehydrated by shaking at 30 ꢁC, 120 rpm for 30 min. After
Oxidation of derivatives of benzyl alcohol via hydrogen transfer employing lyophi-
lised cells of J. terrae DSM 13953 and acetaldehyde
Substituent
c (%)a
addition of benzyl alcohol (6 mL, 6.24 mg, 58 mmol) the reaction
d
95b
72
8
mixture was shaken at 30 ꢁC and 120 rpm for 22 h. The reaction was
stopped by extraction with ethyl acetate (2ꢂ0.5 mL) and centrifu-
gation (12,000 rpm, 2 min). The organic phase was dried (Na2SO4)
prior to determination of conversion by GC.
p-F
p-Cl
p-CH3
m-OCH3
m-CH3
m-I
o-CH3
o-F
16
84
96
88
4
3.2.3. Biocatalytic oxidation employing J. terrae
Typical optimised procedure: Lyophilised cells of J. terrae DSM
13953 (20 mg) were rehydrated in phosphate-buffer (0.6 mL,
100 mM, pH 7.5) in eppendorf tubes (2 mL) by shaking at 30 ꢁC,
23
a
Measured by GC–MS. Reaction conditions: Cells (100 mg), acetaldehyde
(951 mM), Pi buffer (3 mL, 100 mM), substrate (97 mM), 30 ꢁC, 120 rpm, 24 h.
b
Product contained 5% of cinnamaldehyde.
120 rpm for 30 min. Afterwards, acetaldehyde (8 mL, 6.3 mg,