Chemoenzymatic synthesis of (2S)-2-arylpropanols through a dynamic kinetic resolution of 2-arylpropanals with alcohol dehydrogenases

Article abstract of DOI:10.1039/c005098a

We applied Horse Liver Alcohol Dehydrogenase (HLADH) to the enantioselective synthesis of six (2S)-2-arylpropanols, useful intermediates in the synthesis of Profens. The influence of substrate structure and reaction conditions on yields and enantioselectivity were investigated. The high yields and high enantioselectivity towards the (S)-enantiomer obtained in the bioreduction of 2-arylpropionic aldehydes, clearly indicate the achievement of a DKR process through a combination of an enzyme-catalyzed kinetic reduction with a chemical base-catalyzed racemization of the unreacted aldehydes. The racemization step is represented by the keto-enol equilibrium of the aldehyde and can be controlled by modulating pH and reaction conditions.

Full text of DOI:10.1039/c005098a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Chemoenzymatic synthesis of (2S)-2-arylpropanols through a dynamic kinetic
resolution of 2-arylpropanals with alcohol dehydrogenases†
Paola Galletti,* Enrico Emer, Gabriele Gucciardo, Arianna Quintavalla, Matteo Pori and Daria Giacomini*
Received 14th April 2010, Accepted 16th June 2010
First published as an Advance Article on the web 13th July 2010
DOI: 10.1039/c005098a
We applied Horse Liver Alcohol Dehydrogenase (HLADH) to the enantioselective synthesis of six
(2S)-2-arylpropanols, useful intermediates in the synthesis of Profens. The influence of substrate
structure and reaction conditions on yields and enantioselectivity were investigated. The high yields
and high enantioselectivity towards the (S)-enantiomer obtained in the bioreduction of 2-arylpropionic
aldehydes, clearly indicate the achievement of a DKR process through a combination of an
enzyme-catalyzed kinetic reduction with a chemical base-catalyzed racemization of the unreacted
aldehydes. The racemization step is represented by the keto–enol equilibrium of the aldehyde and can
be controlled by modulating pH and reaction conditions.
cessfully reported.6 Among dehydrogenases, the most prominent
Introduction
enzymes out of this class are the widely employed yeast alcohol
The increasing attention dedicated to biocatalytic processes,
dehydrogenase (YADH, Baker’s Yeast) and the horse liver alcohol
which combine selectivity and environmental sustainability, in
dehydrogenase (HLADH). In particular, HLADH is a commer-
drug synthesis is not surprising.1 The main advantages of the
cially available, NAD(H)-dependent biocatalyst characterized by
a very broad substrate tolerance.7 Recently, a process to produce
this enzyme in bacteria, avoiding the use of animal sources, has
been established opening the way for large scale applications.
Following our interest in the use of enzymes for the synthesis
of bioactive molecules,8 we recently applied HLADH to the enan-
tioselective synthesis of (2S)-2-phenylpropanol and (2S)-2-(4-iso-
butylphenyl)propanol (S-Ibuprofenol) via an efficient dynamic
kinetic resolution (DKR) of the parent racemic aldehydes.9 (2S)-2-
Arylpropanols are useful intermediates in flavor manufacture and
can be oxidized10 to (2S)-2-arylpropionic acids, active ingredients
of the Profen class. Profens are a subclass of the non-steroidal
antiinflammatory class of drugs (NSAIDs) and in recent years,
have come to dominate this therapeutic area. Ibuprofen, for
example, is used in the treatment of a number of inflammatory
conditions such as arthritis, muscular strain, and cephalalgia.
Profens are chiral drugs and the (S)- and (R)-enantiomers differ
substantially in both their pharmacodynamics and pharmacoki-
netic properties. Prior to the early nineties, the (S)-enantiomer was
regarded as the eutomer of the Profens and the (R)-enantiomer as
the distomer. However, observations in the late eighties made this
distinction less clear and it is now recognised that the (S)-profens
are the enantiomers that inhibit prostaglandin synthetase.11 It
was also evidenced that the (R)-enantiomer of Ibuprofen can be
interconverted to the (S)-enantiomer in the liver and kidney of
pigs and rats. However this inversion requires CoA and ATP as
cofactors which was validated by showing that (R)-Ibuprofen-
CoA did not racemize in either buffer solution (pH 7.4) or human
plasma.12 Moreover, the (R)-Ibuprofen displays toxicity due to
its storage in fatty tissue as a hybrid glycerol ester, whose long-
term effects are not known.13 Effectively, most Profens are moving
towards single enantiomer administration as the rule. Enantiopure
Profens are usually obtained through a final kinetic resolution of
the esters by enzymatic hydrolysis,14 but performing a dynamic
kinetic resolution at an earlier stage may provide an alternative
use of enzymes as catalysts are represented by the high effi-
ciency in terms of yields, chemo-, regio-, and stereo-selectivity
along with their use in mild conditions.2 Enzymes have been
widely employed in the resolution of racemates (kinetic resolu-
tion, KR).3 However, the maximum yield of kinetic resolutions
is only 50 per cent, which is economically and ecologically
unattractive unless recycling of the undesired enantiomer is
easily achievable; the ultimate improvement in this context is
represented by dynamic kinetic resolution (DKR).4 A DKR
process combines a kinetic resolution and in situ racemization
of the unreacted enantiomer and can overcome the 50 per
cent yield limitation in classical kinetic resolution whenever the
racemization rate successfully competes with that of the resolution
reaction.
In DKR processes, enzymes have been successfully employed.
Typically, several lipases have been used in esterification or
hydrolysis in highly stereoselective DKR, with good yields and
under mild conditions.5 Dehydrogenases have been employed less
frequently than lipases in organic synthesis and in DKR because
of a number of drawbacks: 1) a limited number of commercially
available purified enzymes; 2) limited stability and activity in
non-aqueous solvents; and 3) requirement for a co-factor and its
associate recycling system that increase the complexity of the reac-
tion. However, recent applications of alcohol dehydrogenases and
carbonyl reductases show that the productivity of these systems,
particularly in the reduction mode, is constantly increasing, and
some seminal examples of DKR by dehydrogenases have been suc-
Department of Chemistry “G. Ciamician”, University of Bologna, Via
Selmi 2, Bologna, I-40126, Italy. E-mail: paola.galletti@unibo.it, daria.
giacomini@unibo.it; Fax: (+39)0512099456; Tel: (+39)0512099514
† Electronic supplementary information (ESI) available: Synthesis and
spectroscopic data of 2-arylpropanals 1b–f, HPLC analyses and procedure
for oxidation of (2S)-arylpropanols 2c and 2d to (2S)-2-arylpropionic
acids. See DOI: 10.1039/c005098a
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4117–4123 | 4117
©

View Article Online
Table 1 Enzymatic reduction of 2-phenylpropanal 1a to 2-
phenylpropanol 2a with excess NADH.a
route to their asymmetric synthesis thus making the process
simpler and more economically advantageous.
Here we present a detailed study of the enzymatic reduction
by dehydrogenases of arylpropionic aldehydes to obtain (2S)-
2-arylpropanols. In particular we investigated the racemization
step and the influence of pH on both the racemization of
the starting aldehyde and the enzymatic reduction. We then
applied the reaction to a wider series of racemic 2-arylpropanals
(Scheme 1) and even on a semi-preparative scale, and the influence
of substrate structures and reaction conditions on yields and
enantioselectivities was discussed in depth. As a further develop-
ment, we demonstrated the maintenance of the optical purities
of (2S)-flurbiprofenol and (2S)-fenoprofenol in a subsequent
oxidation reaction to the corresponding acids, thus realizing the
enantioselective synthesis of Flurbiprofen and Fenoprofen.
Enzyme, amount
Ent. (U/mmol), time
Co-solvent
(%)
2a (yield %)b S/R
1
2
3
4
5
6
7
8
YADH 500, 5 h
YADH 500, 24 h
YADH 500, 96 h
HLADH 25, 5 h
HLADH 25, 24 h
HLADH 5, 5 h
HLADH 5, 24 h
HLADH 25, 5 h
HLADH 25, 5 h
HLADH 500, 5 h
HLADH 500, 5 h
—
—
—
—
—
—
—
18
53
79
81
89
75
89
72
55
77
80
99/1
93/7
92/8
82/18
81/19
99/1
95/5
89/11
96/4
CH3CN (10)
THF (10)
Hexane (99)
Hexane (98)
9
10
11
63/37
62/38
a Reaction conditions: method A, 0.5 mM 2-phenylpropanal 1a, 1 mM
NADH, 0.1 M phosphate buffer pH 7.5, 27 ◦C, V = 5 mL, enzyme amount
expressed as U/mmol of aldehyde 1a; U refers to ethanol/acetaldehyde
activity as reported by suppliers. b Yields obtained by HPLC through
calibration plot.
Table 2 Enzymatic reduction of 2-phenylpropanal 1a to 2-
phenylpropanol 2a with NADH recycling.a
Enzyme, amount
(U/mmol), time
Co-solvent
(%)
2a
Ent.
pH
(yield %)b
S/R
1
2
3
4
5
6
7
8
YADH 500, 5 h
YADH 500, 24 h
YADH 50, 96 h
HLADH 25, 5 h
HLADH 25, 24 h
HLADH 5, 24 h
HLADH 25, 5 h
HLADH 5, 96 hc
HLADH 25, 24 h
HLADH 25, 5 h
HLADH 25, 24 h
HLADH 25, 24 h
HLADH 500, 5 h
HLADH 500, 5 h
—
—
—
—
—
—
7.5
7.5
7.5
7.5
7.5
8.0
7.5
7.5
8.0
7.5
8.0
8.0
7.5
7.5
—
traces
5
—
—
50/50
84/16
83/17
95/5
97/3
94/6
98/2
95/5
99/1
95/5
90/10
72/28
89
99
85
72
90
85
54
80
88
55
68
CH3CN (10)
CH3CN (16)
CH3CN (10)
THF (10)
THF (10)
THF (5)
Hexane (99)
Hexane (95)
9
10
11
12
13
14
Scheme 1 Enzymatic reduction of racemic 2-arylpropanals 1a–f to the
corresponding (2S)-2-arylpropanols 2a–f.
a Reaction conditions: method B, 0.5 mM 2-phenylpropanal, 0.01 mM
NADH, 0.5 M EtOH, 0.1 M phosphate buffer, 27 ◦C, V = 15 mL, enzyme
amount expressed as U/mmol of aldehyde 1a. b Yields obtained by HPLC
through calibration curve. c 1 mmol 1a (45 mM), 96 h, V = 24 mL, isolated
yield.
Results and discussion
Preliminary experiments were performed on racemic 2-
phenylpropanal. The enzymatic reduction was explored using the
two commercial alcohol dehydrogenases, YADH and HLADH in
phosphate buffer with an excess of NADH. Data are reported in
Table 1. With both enzymes (entries 3 and 5) the enantioselectivity
was good and the yield exceeded 50%, thus indicating that a
DKR process was in action but with YADH reaction times were
longer, reflecting the broader substrate specificity of HLADH. The
enantiomeric excess was always in favour of the (S)-enantiomer
and noticeably improved by lowering the enzyme amount (entries
6 and 7). To increase aldehyde solubility in the reaction mixture,
the reduction with HLADH was tested also in the presence of
an organic co-solvent (CH3CN, THF, and hexane). The reaction
proceeded in all cases but yields were lower and in hexane the
stereoselectivity was poor, probably because of the excess of
enzyme needed to catalyze the reaction.
results obtained at pH 7.5, in terms of yields and enantioselectivity,
were at least comparable with those obtained with NADH in
excess (cf. entry 4 Table 1 and entry 4 Table 2). On increasing
the pH or adding an organic co-solvent, the enantiomeric ratio
increased, even if the co-solvent addition lowered the enzyme
activity and extended the reaction time to obtain a complete
conversion.
The high yields obtained in the bioreduction of 2-
phenylpropionic aldehyde, clearly indicate the achievement of a
DKR process through a combination of an enzyme-catalyzed
kinetic resolution with a chemical base-catalyzed racemization.
In our case, the racemization step is represented by the keto–
enol equilibrium of the starting aldehyde. The stereoinversion of
arylpropionic aldehydes in aqueous medium is a feasible process
and supported by the reported racemization of R- and S-Ibuprofen
The same process was tested in the presence of ethanol as a co-
substrate to realize the in situ co-factor recycling (Table 2). With
YADH the reduction did not proceed whereas with HLADH the
4118 | Org. Biomol. Chem., 2010, 8, 4117–4123
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
(2-(4-isobutyl)phenyl-propionic acid)15 and 2-phenyl ketones.16 In
the reaction medium the unreacted (R)-aldehyde undergoes a
racemization process through the achiral enolic form (Scheme 2).
The stereoinversion was definitively proved by the reduction with
HLADH of the enantiopure (2R)-2-phenylpropanal; the (2S)-2-
phenylpropanol was obtained in 95% yield and an enantiomeric
ratio S/R of 95 : 5.9
Table 3 Enzymatic reduction of 2-arylpropanals 1b–e to 2-arylpropanols
2b–e with NADH recycling.a
Enzyme amount
Ent. Sub. (U/mmol)
10%
pH Co-solvent Yield (%)b S/R
1
2
3
4
5
6
7
8
1b
1b
1b
1c
1c
1c
1cd
1d
1d
1d
1d
1d
1d
1d
1dd
1e
1e
1e
1e
1e
1e
25
25
25
25
25
25
25
25
25
25
10c
25
25
10c
25
25
25
25
25
25
25
7.5
—
69
93
46
92
98
49
63
93
92
68
99
82
97
98
82
44
42
38
63
58
67
97/3
7.5 CH3CN
7.5 THF
7.5
7.5 CH3CN
7.5 THF
8
>99/1
>99/1
96/4
—
>99/1
>99/1
>99/1
66/34
82/18
>99/1
81/19
93/7
82/18
93/7
>99/1
92/8
CH3CN
—
7.5
9
7.5 CH3CN
7.5 THF
8
8
8
8
8
7.5
7.5 THF
7.5 CH3CN
8.0
8.0 THF
8.0 CH3CN
10
11
12
13
14
15
16
17
18
19
20
21
—
THF
CH3CN
CH3CN
CH3CN
—
Scheme 2 Resolution and racemization steps in reductive DKR of
2-phenylpropanal.
88/12
90/10
86/14
98/2
—
To extend the application scope, the enzymatic reduction with
HLADH was applied to 2-arylpropanals 1b–f suitable for use as
intermediates in the synthesis of Profens (Scheme 1, Table 3 and
4). The racemic aldehydes are not commercially available and, as
a shortcut for the purpose of this work, have been obtained from
the corresponding commercial 2-arylpropionic acids (see ESI†
for details). The bioreductions were performed using NADH in
catalytic amount, and in the presence of EtOH for the co-factor
recycling.
92/8
a Reaction conditions: method B, 0.1 M phosphate buffer, 0.5 mM
2-arylpropanals 1b–e, 0.01 mM NADH, 0.5 M EtOH, 24 h, enzyme
amount expressed as U/mmol of aldehyde. b Yields obtained by HPLC
through calibration curve. c Total amount, added in two aliquots at t = 0
and t = 5 h. d Reaction conditions: method C, 0.5 mmol aldehyde (5 mM),
96 h, isolated yield.
S-enantiopreference was definitively confirmed by comparison
with literature optical rotation data and HPLC analyses on chiral
columns (Scheme 3, see ESI† for details).
On changing substrates and reaction conditions the interplay
between the resolution and the racemization step becomes funda-
mental to get satisfactory yields and enantioselectivities (Table 3).
In the case of ibuprofenal 1b and flurbiprofenal 1c very good
yields and complete enantioselelectivity were achieved using a 10%
mixture of CH3CN in phosphate buffer (pH 7.5) (Table 3, entries 2
and 5). In the same conditions with fenoprofenal 1d the S/R ratio
was not satisfactory (entry 9) but it was improved by increasing pH
and lowering the enzyme amount (entry 14). To keep the enzyme
concentration low and obtain quantitative yields, the enzyme was
added in two aliquots at different reaction times (Table 3, cf. entries
13 and 14). In the case of naproxenal 1e, the enantiomeric ratio
was good (entry 20) but high yields could not be obtained even
if in some cases more than 50% was obtained thus indicating
the presence of a DKR (entry 21). In all cases we obtained the
(2S)-arylpropanol as the major enantiomer. For 2b and 2e the
configuration was established by comparison with samples of
enantiopure (S)-alcohols obtained by reduction with BH3·Me2S
of the commercial (S)-acids. For 2c and 2d (flurbiprofenol and
fenoprofenol), the enzymatic reduction was performed on a semi-
preparative scale (Table 3, entries 7 and 15), the (S)-alcohols
were oxidized to the corresponding (2S)-2-arylpropionic acid
(flurbiprofen and fenoprofen) using KMnO4 in acetone and the
Scheme 3
In both cases no racemization occurred in the oxidation process
and the enantiomeric ratio of the alcohols was conserved in
acids. With all substrates, when the reaction yields were low,
the unreacted aldehyde was not recovered but we detected the
corresponding 2-arylmethylketones as by-products. The formation
of a by-product is particularly evident in the case of naproxenal 1e
which never gives high yields and forms a significant amount of
arylmethylketone 3e (Scheme 4).
The conversion of 2-phenylpropanal into acetophenone is re-
ported in the literature through a metal catalysis,17 and it is known
also in the case of other 2-arylpropanals.18 In our case we identified
arylmethylketones in blank reactions with aldehydes in the same
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4117–4123 | 4119
©

View Article Online
Table 4 Enzymatic reduction of ketoprofenal 1f to ketoprofenol 2f with
the kinv becomes slower than kfast a decrease in the ee of the
preferred enantiomer at high conversions can occur but a good
DKR process can still be obtained if kinv is higher than the rate
NADH recycling.a
constant for reduction of the slower reacting enantiomer (kinv
>
k
slow). We observed with time a decrease of the S/R ratio on
increasing the reaction yields (entries 4–7, Table 1 and entries 4–5,
Table 2) and this could be due to a kinv slower than kS. Moreover,
both resolution and racemization steps are expected to be pH
dependent, particularly the keto–enol tautomerism that occurs
via proton transfer.20 Therefore we studied the bioreduction of 2-
phenylpropanal at different pH values in the range of pH 6–8.5
in consideration of the pH sensitivity of HLADH (Fig. 1). As
expected, yields are strongly influenced by pH value (Fig. 1a). At
pH 7.5 the yield was particularly good, but at pH 8.5 it was low.
This result could be due to a decreased enzyme activity or co-factor
binding21 at higher pH values. However, at the higher pH the best
enantioselectivity was obtained (Fig. 1b). Moreover a decreased
(S)-enantiomer % on increasing the reaction time and conversion
confirms that the racemization rate kinv is very close to or even
lower than kS, according to the kinetic analysis by Noyori.19
Enzyme amount
Entry (U/mmol)
10%
pH
Co-solvent 2f (yield %)b S/R
1
2
3
4
5
6
7
8
25
7.5
7.5
7.5
7.5
7.5
7.5
8.0
8.0
—
—
CH3CN
CH3CN
THF
THF
—
95
99
80
54
88
36
37
9
75/25
93/7
94/6
99/1
95/5
99/1
99/1
99/1
10c
25
10c
25
10 c
25
25
CH3CN
a Reaction conditions: method B, phosphate buffer, 0.5 mM 1f, 0.01 mM
NADH, 0.5 M EtOH, 24 h, enzyme amount expressed as U/mmol of
aldehyde (1f). b Yields obtained by HPLC through calibration curve. c Total
amount, added in two aliquots at t = 0 and t = 5 h.
Scheme 4
reaction conditions of the DKR process but in the absence of
the enzyme. The reduced products of the arylmethylketones, i.e
1-arylethanols, were never detected in the enzymatic reactions.
Ketoprofenal is a particularly interesting substrate for the
enzymatic reduction (Table 4) since 1f bears two carbonyl groups,
a diaryl ketone and an aldehyde. HLADH demonstrated a high
degree of chemoselectivity producing as a unique compound the
ketoalcohol 2f in high yield. No traces of the possible ketone reduc-
tion by-products (the diol or the hydroxy-aldehyde) were detected.
With this substrate both better yields and enantioselectivity were
obtained in the absence of a co-solvent (entry 2); by using 10% of
CH3CN and THF, S/R values were always good but yields were
significantly lower.
The high (S)-enantioselectivity of the enzymatic reduction, in
terms of the kinetic constants, means that the rate constant for
(S)-aldehydes is higher than that for (R)-aldehydes (kS ꢀ kR), but
the efficiency of the entire DKR process is also strictly related
to the rate of the racemization kinv. Noyori et al.19 developed a
quantitative treatment of a DKR process generated by a reduction
coupled with a stereoinversion and showed that quantitative
conversions and good enantioselectivity can be still obtained when
the racemization rate constant (kinv) and the rate constant for
reduction of the faster enantiomer are the same (Scheme 2). If
Fig. 1 Time course for pH effects on a) yields and b) enantioselectivity.
Reaction conditions: 0.5 mM 1a, 0.1 M phosphate buffers, 0.01 mM
NADH, 0.5 M EtOH.
As above mentioned, the racemization step is very important for
the efficiency of a DKR process, we thus studied in more detail the
racemization of 2-phenylpropanal, ibuprofenal and naproxenal
(1a,1b and 1e) in the absence of the enzymatic reduction. We
measured the decrease of optical rotation 22 with time of solutions
of enantiopure (2S)-2-phenylpropanal, (S)-ibuprofenal and (S)-
naproxenal (1a,1b and 1e) in phosphate buffer 0.1 M at pH 7.5
and 8, CH3CN was used as co-solvent to ensure complete substrate
4120 | Org. Biomol. Chem., 2010, 8, 4117–4123
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
dissolution. Data are plotted in Fig. 2. Considering racemization
as the formation of a racemate from a pure enantiomer in an
irreversible first-order reaction it follows that: 23
decomposition into the 2-arylmethylketone, in the case of 1b a
good DKR is obtained with longer reaction times.
-ln(at/ao) = kinvt
(1)
Conclusions
and
The enzymatic reduction with HLADH was successfully applied
to six racemic 2-arylpropanals affording (2S)-2-arylpropanols,
suitable for use as intermediates in Profen syntheses. In most cases
both excellent enantioselectivity and yields could be obtained by
modulating the reaction conditions, thus influencing the resolution
and the racemization step of the DKR process. In the case of
ketoprofenal complete chemoselectivity was also achieved.
The starting racemic 2-arylpropanals are in some cases already
used as intermediates in the synthesis of Profen24 or can be
obtained by hydroformylation25 or hydrovinylation26 of vinyl-
arenes. The enantiopure products, (2S)-2-arylpropanols, can in
turn be oxidized to (2S)-2-arylpropanoic acids, so that the process
here described could represent an interesting alternative for the
synthesis of enantiomerically pure Profens. Work is in progress
to study alternative more sustainable procedures for the final
oxidation step.
t1/2 = ln 2/kinv
(2)
at being the observed optical rotation at time t, a0 the initial
observed optical rotation, t1/2 the racemization half-life and kinv
the racemization constant.
Experimental
General
Fig. 2 Time course of optical rotation of aldehydes 1a (ꢀ), 1b (ꢀ) and 1e
(ꢁ) at pH 7.5 and 8. Conditions: aldehyde concentration 5 mM, phosphate
buffer 0.1 M, CH3CN 30%.
Chemicals were purchased from Sigma-Aldrich or TCI and used
without further purification. YADH: Sigma A3263. HLADH:
Sigma A9589 from Horse Liver, or evo-1.1.211 cloned isoenzyme
E from Evocatal, or cloned enzyme prepared according to existing
protocols27 in Dr Paradisi’s laboratory (UCD, Ireland) and kindly
given to us for the purpose of this research. TLC: Merck 60
F254. Column chromatography: Merck silica gel 20–300 mesh.
FT-IR: Nicolet 308 measured as films between NaCl plates, wave
numbers reported in cm-1. 1H and 13C NMR spectra: obtained on
a Varian GEMINI 200 and Varian INOVA 300 spectrometers
with a 5 mm probe. All chemical shifts have been quoted
relative to deuterated solvent signals, d in parts per million, J in
hertz. Elemental analysis: Perkin-Elmer 2400 Series II CHNS/O
analyzer. Reverse phase HPLC: Agilent Technology HP1100,
column ZORBAX-Eclipse XDB-C8 Agilent Technologies. The
compounds were eluted with CH3CN–H2O, gradient: from 30%
to 100% of CH3CN in 15 min, then 100% of CH3CN for 10 min,
T = 30 ◦C. Direct phase HPLC: Hewlett-Packard HP1090 Series
II, columns Daicel’s Chiralpack (0.46 cm Ø x 25 cm) Chiralcel
OD or OF. The compounds were eluted with Hexane/iPrOH, flux
0.5 mL min-1, T = 40 ◦C (see Table 5). 2-Phenylpropanal 1a and
2-phenylpropanol 2a are commercial and re-distilled before use.
Non-commercial racemic 2-arylpropanals 1b–f can be prepared
startingfromthecorrespondingcommercialcarboxylicacidsusing
two alternative procedures: method 1) preliminary esterification
and then iBu2AlH reduction,9 or method 2) over-reduction to
primary alcohols with BH3·Me2S28 and then Swern oxidation to
aldehydes.29 In both procedures 2-arylpropanol alcohols 2b–f can
be obtained as intermediates or by-products and can be used for
HPLC calibration plot. In the case of ketoprofenal 1f, to avoid
ketone protection/deprotection steps, method B was used. Full
details and data in ESI.†
According to eqn (1) a linear relationship with slope corre-
sponding to kinv values should be obtained. In our case, as can
be seen in Fig. 2, linearity was not observed during the entire
experimental course because the aldehydes form chiral hydrates
and by-products with time (HPLC monitoring). However over
the initial time range a linear behaviour (solid lines in Fig. 2)
was observed for all plots allowing calculation of the kinv values
reported in Table 5. Half-lives t1/2calc have been calculated from
kinv through eqn (2) and have been compared with the observed t1/2
(t1/2obs), the time corresponding to ao/2.
As expected, the racemization rate is highly controlled by
pH, for all aldehydes kinvs increase and half-lives decrease with
increasing pH (Table 5, cf. entries 1 and 2, 3 and 4, 5 and 6).
2-Phenylpropanal is the aldehyde with the fastest racemization
rate and this is also confirmed by the experimental observation
that enantiopure 2-phenylpropanal has a very short shelf-life.
Ibuprofenal 1b and 1e have slower kinv values but whereas for
naproxenal 1e the whole process is limited by the aldehyde
Table 5 Racemization constants kinv, half-lives t1/2calc and t1/2obs ob-
tained from data in Fig. 2
Entry Aldehyde pH
k
inv/10-6s-1
t1/2 obs/min
t1/2 calc/min
1
2
3
4
5
6
1a
1a
1b
1b
1e
1e
7.5
8
7.5
8
7.5
8
75
419 30
29
52
41
95
6
120
30
320
180
180
115
154
28
397
224
279
113
1
2
2
3
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4117–4123 | 4121
©

View Article Online
General procedure for enzymatic reduction.
3), 130.7 (d, J(C,F) = 4), 135.7, 145.5 (d, 1C, JC,F = 7), 159.9 (d,
1C, JC,F = 247); IR: n = 3373, 1023; elemental analysis calcd (%)
for C15H15FO: C 78.24, H 6.57; found: C 78.19, H 6.56. (S)-2c
obtained from semi-preparative enzymatic reaction: [a]2D0 = -11.5,
(c = 2.6, CHCl3).
Method A with excess NADH: into a vial equipped with a
magnetic stirrer (double spinfin magnetic stirring bar) all reagents
were added in the following order: 0.5 mL of a 5 mM solution
of the starting aldehyde in CH3CN or THF, 0.5 mL of a 10 mM
solution of NADH freshly prepared in 0.1M phosphate buffer
(pH 7.5), 0.1M phosphate buffer (pH 7.5) to reach a total final
volume of 5 mL and the chosen amount of enzyme (indicated in
Table 1–4) from a freshly prepared solution in phosphate buffer.
In the case of co-solvent absence, the aldehyde was directly added
to the reaction mixture. In the case of enzymatic reductions in
hexane, data were obtained by adding solid lyophilized cofactor
together with lyophilized enzyme into the reaction vessel before
adding the solvent mixture and the substrate.
2-(3-Phenoxy-phenyl)-propan-1ol (Fenoprofenol) 2d. Obtained
in 15% yield as a by-product in 1c preparation. 1H NMR
(300 MHz, CDCl3): d 1.30 (d, J = 7.2, 3H, CH3), 1.4 (bs, 1H,
OH), 2.97 (sextet, J = 7.2, 1H, CH), 3.72 (d, J = 7.2, 2H, CH2),
6.89–7.19 (m, 6H, arom), 7.30–7.42 (m, 3H, arom); 13C NMR
(75 MHz, CDCl3): d 17.5, 42.2, 68.4, 116.7, 117.9, 118.8, 122.3,
123.2, 129.7, 145.9, 157.0, 157.4. IR: n = 3350, 1584, 1489, 1250;
elemental analysis calcd (%) for C15H16O2: C 78.92, H 7.06; found:
C 79.02, H 7.05.
Method B with NADH recycling: into a vial equipped with a
magnetic stirrer (double spinfin magnetic stirring bar) all reagents
were added in the following order: 1.5 mL of a 5 mM solution of the
starting aldehyde in CH3CN or THF, 0.45 mL of EtOH, 1.5 mL of
a 0.1 mM solution of NADH freshly prepared in the appropriate
0.1 M buffer, 0.1M buffer to reach a total final volume of 15 mL
and the chosen amount of enzyme from a freshly prepared solution
in buffer. In the case of co-solvent absence, 2-arylpropanals were
diluted in EtOH.
Method C, semi-preparative scale: into a flask equipped with
a magnetic stirrer were added in the following order: 0.5 mmol
of the starting aldehyde 1a,c–d in 10 mL of CH3CN, 3.4 mL of
EtOH and phosphate buffer 0.1 M (pH = 8) to reach a final total
volume of 100 mL. After 15 min NADH (0.01 mmol, 7.1 mg)
and the enzyme (1 mg, 0.46 U/mg) were added to the solution.
After 96 h the reactions were worked up by adding 1 g of solid
NaCl to the aqueous phase and extracting 3 times with EtOAc
(3 ¥ 75 mL). The organic phases were dried over Na2SO4 and
concentrated in vacuo. Yields after flash chromatography (80/20
: cyclohexane/ethylacetate). Oxidation to 2-arypropionic acids is
reported in the ESI.†
(S)-2d obtained from semi-preparative enzymatic reaction:
[a]2D0 = -10.0 (c = 2.4, CHCl3).
2-(6-Methoxy-naphthalen-2-yl)-propan-1-ol (Naproxol) 2e.
Obtained in 10% yield as a by-product in 1e preparation. 1H
NMR (300 MHz, CDCl3): d 1.38 (d, J = 7.0, 3H, CH3CH),
1.4 (bs, 1H, OH), 3.11 (sextet, J = 7.0, 1H, CH), 3.70 (d, J =
7.0, 2H, CH2), 3.94 (s, 3H, OCH3), 7.14–7.28 (m, 2H, arom),
7.34–7.40 (m, 1H, arom), 7.63–7.76 (m, 3H, arom); 13C NMR
(50 MHz, CDCl3): d 17.6, 42.4, 55.3, 68.6, 105.6, 118.9, 125.9,
126.3, 126.2, 127.2, 129.1, 133.5, 138.7, 157.5; IR: n = 3300, 1604,
1028; elemental analysis calcd (%) for C14H16O2: C 77.75, H 7.46;
found: C 77.91, H 7.51.
[3-(2-Hydroxy-1-methyl-ethyl)-phenyl]-phenyl-methanone (Keto-
profenol) 2f. Obtained in 30% yield as a by-product in the
1
obtaining of 1f. H NMR (300 MHz, CDCl3): d 1.35 (d, J =
7.0, 3H, CH3CH), 1.4 (bs, 1H, OH), 3.07 (m, 1H, CH), 3.79 (d,
J = 6.6, 2H, CH2), 7.44–7.86 (m, 9H, arom); 13C NMR (75 MHz,
CDCl3): d 17.6, 42.3, 68.5, 128.3, 128.4, 128.6, 128.9, 130.0, 131.7,
132.4, 137.6, 137.8, 144.2, 196.8; elemental analysis calcd (%) for
C16H16O2: C 79.97, H 6.71; found: C 80.05, H 6.69.
Formation of arylpropanols was monitored by inverse phase
HPLC analysis: at different reaction times, aliquot samples
were filtered, diluted and directly injected. Calibration curves
obtained with pure arylpropanols (five dilutions, each in triplicate)
were used for quantitative analysis. Enantiomeric ratios were
determined by HPLC analysis on chiral columns.
Determination of enantiomeric ratio and configurational
assignment of major stereoisomers.
Enantiomeric ratios were determined by HPLC analysis on chiral
columns (see Table 6 in ESI† for full details). In the case of
alcohols 2a,2b,2e and 2f the (S)-configuration of the major
isomer was established by direct comparison with commercial
(S)-2-phenylpropanol or (S)-alcohols, obtained by reduction with
BH3·Me2S of the commercial (S)-acids. For 2c and 2d the (S)-
configuration of the major isomer was established by converting
the alcohols obtained in the semi-preparative procedure into acids
by oxidation with KMnO4 and comparing the optical rotation
with reported data (see ESI†).
2-(4-Isobutylphenyl)-propanol (Ibuprofenol) 2b. 1H NMR
(300 MHz, CDCl3): d 0.92 (d, J = 6.6, 6H, (CH3)2CH), 1.28 (d, J =
7.2, 3H, CH3CH), 1.4 (bs, 1H, OH), 1.86 (m, 1H, (CH3)2CHCH2),
2.46 (d, J = 7.2, 2H, CH2), 2.93 (sextet, J = 7.2, 1H, CH), 3.70
(d, J = 7.2, 2H, CH2OH), 7.10–7.17 (m, 4H, arom); 13C NMR
(75 MHz, CDCl3): d 17.5, 22.2, 30.0, 41.8, 44.8, 68.4, 127.0, 129.0,
139.6, 140.9; IR: n = 3372, 1684, 1513, 1465; elemental analysis
calcd (%) for C13H13O: C 81.20, H 10.48; found: C 81.15, H 10.52.
2-(2-Fluoro-biphenyl-4-yl)-propan-1-ol (Flurbiprofenol) 2c.
Obtained in 12% yield as a by-product in 1c preparation following
method A or in 95% yield following method B. 1H NMR
(200 MHz, CDCl3): d 1.34 (d, J = 7.0, 3H, CH3), 1.57 (bs, 1H,
OH), 3.02 (sextet, J = 7.0, 1H, CH), 3.77 (d, J = 7.0, 2H, CH2),
7.05–7.16 (m, 2H, arom), 7.28–7.61 (m, 6H, arom); 13C NMR
(50 MHz, CDCl3): d 17.4, 41.9, 68.4, 114.9 (d, J(C,F) = 23), 123.5
Acknowledgements
Financial support from MIUR, University of Bologna and the
Fondazione del Monte di Bologna e Ravenna. We are grateful to
Mrs Elena Benedetto for technical assistance. We appreciate Dr
Francesca Paradisi for the supply of cloned HLADH.
(d, JC,F = 3), 127.2 (d, 1C, JC,F = 18), 127.5, 128.4, 128.9 (d, JC,F
4122 | Org. Biomol. Chem., 2010, 8, 4117–4123
=
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
of ADH on 2-phenylpropanal see: J. Grunwald, B. Wirz, M. P. Scollar
and A. M. Klibanov, J. Am. Chem. Soc., 1986, 108, 6732–6134; (d) I.
Kelemen-Horva´th, N. Nemesto´thy, K. Be´lafi-Bako´ and L. Gubicza,
Chem. Papers, 2002, 56, 52–56; (e) G. Sello, F. Orsini, S. Bernasconi
and P. Di Gennaro, Molecules, 2006, 11, 365–369.
Notes and references
1 R. N. Patel, Coord. Chem. Rev., 2008, 252, 659–701.
2 K. Faber, Biotransformations in Organic Chemistry, Springer-Verlag,
Berlin, 2004..
3 (a) G. Kirchner, M. P. Scollar and A. M. Klibanov, Biosci. Biotechnol
Biochem, 1985, 107, 7072–7251; (b) A. Ghanem and H. Y. Aboul-Enein,
Chirality, 2005, 17, 1.
10 A. Basak, A. Nag, G. Bhattacharya, S. Mandal and S. Nag, Tetrahe-
dron: Asymmetry, 2000, 11, 2403–2408.
11 D. B. Campbell, Eur. J. Drug Metab.Pharmacokin., 1990, 5, 109; A. M.
Evans, Eu. J. Clin.Pharmacol., 1992, 42, 237–256.
12 (a) C. Reichel, H. Bang, K. Brune, G. Geisslinger and S. Menzel,
Biochem. Pharmacol., 1995, 50, 1803–1806; (b) R. D. Knihinicki, K. M.
Williams and R. O. Day, Biochem. Pharmacol., 1989, 38, 4389–4395.
13 K. Willams, R. O. Day, R. Knihinicki and A. Duffield, Biochem.
Pharmacol., 1986, 35, 3403–3405.
4 (a) H. Pellissier, Tetrahedron, 2008, 64, 1563–1601; (b) A. Ghanem,
Tetrahedron, 2007, 63, 1721–1754; (c) N. J. Turner, Curr. Opin. Chem.
Biol., 2004, 8, 114–119; (d) H. Pellissier, Tetrahedron, 2003, 59, 8291–
8327; (e) O. Pa`mies and J-E. Ba¨ckvall, Chem. Rev., 2003, 103, 3247–
3261; (f) F. F. Huerta, A. B. E. Minidis and J-E. Ba¨ckvall, Chem. Soc.
Rev., 2001, 30, 321–331; (g) For non-enzymatic DKR examples on 2-
arylpropanals see: X. Li and B. List, Chem. Commun., 2007, 1739–1741;
(h) J.-H. Xie, Z.-T. Zhou, W.-L. Kong and Q.-L. Zhou, J. Am. Chem.
Soc., 2007, 129, 1868–1869.
5 (a) R. Corbera´n and E. Peris, Organometallics, 2008, 27, 1954–1958;
(b) D. Stru¨bing, P. Krumlinde, J. Piera and J. E. Ba¨ckvall, Adv. Synth.
Catal., 2007, 349, 1577–1581; (c) Y. Liu, F. Wang and T. Tan, Chirality,
2009, 21, 349–353.
6 (a) P. K. Matzinger and H. G. W. Leuenberger, Appl. Microbiol.
Biotechnol., 1985, 22, 208–210; (b) A. R. Maguire and N. O’Riordan,
Tetrahedron Lett., 1999, 40, 9285–9288; (c) H. Miya, M. Kawada and
Y. Sugiyama, Biosci., Biotechnol., Biochem., 1996, 60, 95–98; (d) M. F.
Zu¨ger, F. Giovannini and D. Seebach, Angew. Chem., Int. Ed. Engl.,
1983, 22, 1012; (e) K. Nakamura, T. Miyai, K. Ushio, S. Oka and A.
Ohno, Bull. Chem. Soc. Jpn., 1988, 61, 2089–2093; (f) D. Seebach, M. F.
Zu¨ger, F. Giovannini, B. Sonnleitner and A. Fiechter, Angew. Chem.,
Int. Ed. Engl., 1984, 23, 151.
7 C. Hertweck and W. Boland, J. Prakt. Chem., 1997, 339, 754–757.
8 (a) G. Cainelli, D. Giacomini, P. Galletti and M. DaCol, Tetrahedron:
Asymmetry, 1997, 8, 3231–3235; (b) G. Cainelli, D. Giacomini and P.
Galletti, Synthesis, 2000, (2), 289–294; (c) G. Cainelli, V. De Matteis,
P. Galletti, D. Giacomini and P. Orioli, Chem. Commun., 2000, 2351–
2352; (d) G. Cainelli, P. Galletti, D. Giacomini, A. Gualandi and A.
Quintavalla, Helv. Chim. Acta, 2003, 86, 3548–3559; (e) G. Cainelli,
P. Engel, P. Galletti, D. Giacomini, A. Gualandi and F. Paradisi, Org.
Biomol. Chem., 2005, 3, 4316–4320.
14 H. Y. Zhang, X. Wang, C. B. Ching and J. C. Wu, Biotechnol. Appl.
Biochem., 2005, 42, 67–71.
15 X. Yuchun, L. Huizhou and C. Jiayong, Int. J. Pharm., 2000, 196,
21–26.
16 R. Cirilli, R. Costi, R. Di Santo, F. Gasparrini, F. La Torre, M. Pierini
and G. Siani, Chirality, 2009, 21, 24–34.
17 M. Tokunaga, H. Aoyama, Y. Shirogane, Y. Obora and Y. Tsuji, Catal.
Today, 2006, 117, 138–140.
18 D. P. Riley, D. P. Getman, G. R. Beck and R. M. Heintz, J. Org. Chem.,
1987, 52, 287–290.
19 (a) M. Kitamura, M. Tokunaga and R. Noyori, J. Am. Chem. Soc.,
1993, 115, 144–152; (b) J. Andraos, J. Phys. Chem. A, 2003, 107, 2374–
2387; (c) J. Steinreiber, K. Faber and H. Griengl, Chem.–Eur. J., 2008,
14, 8060–8072.
20 E. Iglesias, J. Org. Chem., 2003, 68, 2680–2688.
21 L. A. LeBrun and B. V. Plapp, Biochemistry, 1999, 38, 12387–
12393.
22 Performing the same study using data of ee vs time was not possible
because 2-phenylpropanal does not separate on chiral columns nor in
HPLC neither in GC.
23 E. J. Ebbers, G. J. A. Ariaans, J. P. M. Houbiers, A. Bruggink and B.
Zwanenburg, Tetrahedron, 1997, 53, 9417–9476.
24 N. J. Stuart, A. S. Sanders, US Patent 3,385,886.
25 J. J. Kim and H. Alper, Chem. Commun., 2005, 3059–3061.
26 C. R. Smith and T. V. RajanBabu, J. Org. Chem., 2009, 74, 3066–3072.
27 D. H. Park and B. V. Plapp, J. Biol. Chem., 1991, 266, 13296–13302.
28 A. Chen, L. Ren and C. M. Crudden, J. Org. Chem., 1999, 64, 9704–
9710.
9 (a) D. Giacomini, P. Galletti, A. Quintavalla, G. Gucciardo and F.
Paradisi, Chem. Commun., 2007, 4038; (b) D. Giacomini, G. Cainelli,
P. Galletti, G. Gucciardo, A. Quintavalla,(Alma Mater Studiorum,
University of Bologna), WO 2008074717, 2008.; (c) For other examples
29 A. Dondoni and D. Perrone, Organic Syntheses, 2004, coll. 10, 230.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 4117–4123 | 4123
©