Acid zeolites as alcohol racemization catalysts: Screening and application in biphasic dynamic kinetic resolution

Article abstract of DOI:10.1002/chem.200400713

Acid zeolites were screened as heterogeneous catalysts for racemization of benzylic alcohols. The most promising zeolites appeared to be H-Beta zeolites, for which the optimal reaction conditions were studied in further detail. The zeolite performance was

Full text of DOI:10.1002/chem.200400713

Acid Zeolites as Alcohol Racemization Catalysts: Screening and Application
in Biphasic Dynamic Kinetic Resolution
S. Wuyts, K. De Temmerman, D. E. De Vos, and P. A. Jacobs*[a]
Abstract: Acid zeolites were screened
as heterogeneous catalysts for racemi-
zation of benzylic alcohols. The most
promising zeolites appeared to be H-
Beta zeolites, for which the optimal re-
action conditions were studied in fur-
ther detail. The zeolite performance
was compared to that of homogeneous
acids and acid resins under similar re-
action conditions. In a second part of
the research, H-Beta zeolites were ap-
plied in dynamic kinetic resolution
(DKR) of 1-phenylethanol, which was
conducted by means of a two-phase ap-
proach and which resulted in yields
smoothly crossing the 50% border up
to 90%, with an enantiomeric excess
of >99%. To explore the applicability
of this biphasic methodology, several
other substrates were examined in the
standard racemization reaction and in
the biphasic dynamic kinetic resolu-
tion.
Keywords:
biphasic catalysis
·
heterogeneous catalysis
·
kinetic
resolution · racemization · zeolites
Introduction
many cases, racemization of a chiral substrate requires harsh
conditions that are incompatible with the milder reaction
conditions needed for successful resolution. Therefore, it is
important to catalyze racemization so that it can proceed
under milder conditions. For the racemization of chiral alco-
hols, transition-metal-catalyzed routes have been developed
based on hydrogen-transfer mechanisms.[14–20] Recently,
Maschmeyer et al. studied in detail the mechanisms of these
racemization reactions, which are closely related to Meer-
wein–Ponndorf–Verley reduction and Oppenauer oxidation
reactions (MPVO redox reactions).[21] Some racemization
One of the emerging methods to obtain enantiomerically
pure chiral compounds is the application of dynamic kinetic
resolution (DKR), in which a resolution step, often through
biotransformation, is combined in situ with a racemization
step.[1–6] A large number of resolution methods has been de-
scribed in the literature, comprising use of enzymes[7–10] and
other, purely chemical, tools.[11] The main drawback of clas-
sical kinetic resolution (KR) is the maximum yield of 50%
that can be obtained when starting from the racemic mix-
ture. When a racemization step is integrated in the reaction,
the main advantage of KR remains: one can start from the
cheap racemic mixture, but the yield of 50% can be sur-
passed and yields of up to 100% can be achieved in one
step. Moreover, depletion of the reacting enantiomer is
avoided in DKR, and in this way a decrease of the enantio-
meric purity of the product in the course of the reaction is
circumvented.[2,12] The mechanism of racemization and the
reaction conditions required depend on the stability of the
stereogenic center, which is itself influenced by the function-
al groups on the stereogenic atom or at the a-position.[13] In
catalysts were successfully combined with
a resolving
enzyme in DKR; however, the re-use of these catalysts is
not trivial, and in some resolution reactions 4-chlorophenyl
acetate is used[6] instead of more convenient acyl donors,
such as ethyl acetate[14b] or isopropenyl acetate.[17,20]
Alternatively, chiral alcohols, such as 1-phenyl-1-ethanol,
may be racemized by means of protonation, water loss, and
formation of an sp2 carbenium ion, which is planar and pro-
chiral.[22–24] Subsequent water addition is aselective and re-
sults in a racemic mixture (Scheme 1). If this Brçnsted acid
catalyzed racemization is performed in
a water-rich
medium, the elimination reaction to the alkene can be
avoided. As this racemization method is especially suitable
for alcohols, producing well-stabilized carbenium ions, for-
mation of an anti-Markovnikov product, for example, 2-phe-
nylethanol, is highly unlikely.
[a] S. Wuyts, K. De Temmerman, Prof. Dr. D. E. De Vos,
Prof. Dr. P. A. Jacobs
Centre for Surface Chemistry and Catalysis
KULeuven, Kasteelpark Arenberg 23, 3001Leuven (Belgium)
Fax : (+32)1632-1998
In line with our previous attempts to develop heterogene-
ous catalysts for alcohol racemization and potential applica-
E-mail: pierre.jacobs@agr.kuleuven.ac.be
386
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200400713
Chem. Eur. J. 2005, 11, 386 – 397

FULL PAPER
Figure 1. Racemization of 1-phenylethanol with homogeneous HCl (0.1
Scheme 1. Acid-catalyzed racemization of 1-phenyl-1-ethanol.
~
&
^
( ) and 0.5m ( )) and pTSA (0.1(”) and 0.5 m ( )) acids.
Table 1. Activity of resins in racemization of benzylic alcohols.[a]
tion in DKR,[25] our attention
was drawn to the use of
heterogeneous acids, like zeo-
lites, as racemization catalysts.
In the patent literature, the use
of Amberlystꢁ and Deloxanꢁ
acid resins for the racemization
of benzylic alcohols has been
described.[22,23] Zeolites may not
only be more stable than these
resins, but it is also known that
hardly any acidity leaks from
zeolites to an aqueous solvent;
this fact makes them favourable
for combination with the enzy-
matic reaction.
Entry
Resin
Support[b]
Acid
mmolH+
gꢀ1
eealcohol [%]
after 3 h
group
1Dowex
2
3
4
5
6
7
8
ꢁ MWC-1PA
COOH
9.5
3.8
10
4.8
4.8
4.8
4.3
4.8
0.8
0.15
100
100
100
86
65
48
63
23
56
30
ꢀ
Dowexꢁ CCR-1PA
Amberliteꢁ IRC-50
Dowexꢁ 50X2-100
Dowexꢁ 50X4-100
Dowexꢁ 50X8-100
Amberliteꢁ IR-118
COOH
ꢀ
ꢀ
PM
PS
PS
PS
PS
5
PF
PF–SiO2
COOH
ꢀ
SO3H
ꢀ
SO3H
ꢀ
SO3H
ꢀ
SO3H
SO3H
Amberlystꢁ
1
PS
ꢀ
9
10
Nafionꢁ NR-50
Nafionꢁ SAC-13
ꢀ
SO3H
ꢀ
SO3H
[a] Reaction conditions: 5 mL double-distilled water, 80 mg resin, 0.26 mmol (R)-1-phenylethanol at 608C
under vigorous stirring. [b] PA=polyacrylate; PM=polymethacrylate; PS=polystyrene resin; PF=perfluori-
nated resin; PF–SiO2 =perfluorinated resin dispersed on amorphous silica, 10–20%, >200 m2 gꢀ1
.
Herein, we investigate the
potential value of acid zeolites as heterogeneous alcohol
racemization catalysts. Relations are established between
the racemization activity of the zeolites and their structure
and composition. Re-use of the zeolite catalysts is also in-
vestigated. We then present a new reaction methodology in
which racemization, in seemingly “harsh, acid” conditions, is
combined with a resolving biotransformation using a lipase.
A preliminary account of this work has appeared previous-
ly.[24]
on either polystyrene (PS), polymethacrylate (PM), or poly-
acrylate (PA) and contained either weak carboxylic acid
groups (entries 1–3) or strong sulfonic acid groups (en-
tries 4–8). Some perfluorinated (PF) sulfonic acid resins
were also tested (entries 9 and 10).
The racemization of benzylic alcohols requires an acid of
sufficient strength; this explains why weak, immobilized car-
boxylic acids like Dowexꢁ MWC-1, Dowexꢁ CCR-1, and
Amberliteꢁ IRC-50 did not lead to any inversion of the alco-
hol configuration within 3 h (Table 1, entries 1–3), under the
previously defined reaction conditions. All strong acid resins
caused racemization, though a large range of activities was
seen. Within the Dowexꢁ 50 series, a clear trend of increas-
ing activity with degree of cross linking was observed (en-
tries 4–6, corresponding to 2, 4, and 8% cross linking, re-
spectively). This tendency can be explained in terms of the
gel character of the polymer matrix: a less cross-linked poly-
mer will have less accessible acid sites because of the gel
properties of the polymer matrix; a more cross-linked poly-
mer results in a more macroreticular network and makes
the acid sites far more accessible for the substrate molecules.
Of all the resins, macroreticular Amberlystꢁ 15, described
earlier as a racemization catalyst in the patent literature,[23]
gave the best results (entry 8). Perfluorinated acid resins
gave good activity, particularly if one takes into account
the small amount of acid groups per gram of material
Results and Discussion
Racemization
Homogeneous acids: Hydrochloric acid (HCl) and p-toluene-
sulfonic acid (pTSA) were both tested in 0.1and 0.5 m con-
centrations. As can be concluded from Figure 1, no evident
difference in racemization activity could be measured be-
tween these two acids: both 0.5m acid solutions were able to
racemize the substrate to an enantiomeric excess below 5%
within 3 h.
Heterogeneous resin catalysts: Different classes of commer-
cially available resins were screened in a standard racemiza-
tion of (R)-1-phenylethanol (Table 1). The resins were based
Chem. Eur. J. 2005, 11, 386 – 397
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
387

P. A. Jacobs et al.
(entry 9). The activity increased even more when the
Nafionꢁ polymer was dispersed on an amorphous silica sup-
port (entry 10).
Heterogeneous zeolite catalysts: The properties of the zeolite
samples are summarized in Table 2. Before use, all zeolites
were converted to the H+ form by repeated exchange with
NH4+, followed by calcination.
Table 2. Properties and origin of zeolite samples used.
Zeolite
Topology
Si/Al
ratio
Specific
Manufacturer
surface [m2 gꢀ1
]
CBV 600
CBV 712
CBV 720
CBV 760
CBV 780
ZM 101
CBV 21A
CBV 20A
ZM 510
CP 811BL-25
CP 814E-22
CP 806B-25
ZGO-52025
H-MCM-22
FAU
FAU
FAU
FAU
FAU
MOR
MOR
MOR
MOR
BEA
BEA
BEA
BEA
MWW
2.6
5.8
13
30
37
5.8
10
10
11
660
730
780
720
780
480
500
550
530
PQ zeolites
PQ zeolites
PQ zeolites
PQ zeolites
PQ zeolites
PQ zeolites
PQ zeolites
ZØocat
ZØocat
PQ zeolites
Zeolyst
PQ zeolites
Süd-Chemie
–
12.5
11
12.5
740
680
730
10–15
10–15
>400
n.d.[a]
[a] n.d.: not determined.
As is evident from Figure 2a, the activity of the H-Y zeo-
lites strongly depends on the Si/Al ratio. At low ratios (2.6,
5.8) hardly any activity can be discerned. The best results
are obtained with Si-rich zeolites.
The results for mordenites were, in general, better than
those for H-Y zeolites (Figure 2b). Sufficiently de-aluminat-
ed zeolites resulted in fast racemization. The fastest reac-
tions were observed with H-Beta zeolites from various sup-
pliers and synthesis batches (Figure 2c): the enantiomeric
excess (ee) of the substrate steeply decreased and fell below
10% after 2 h in all cases. All H-Beta zeolite samples tested
have a similar Si/Al ratio; consequently, activity differences
between these zeolites were only minor. Finally, H-MCM-22
was found to display considerable activity (Figure 2c). This
is not unexpected, as H-MCM-22 and H-Beta zeolites share
some of the same important properties, such as small crystal
size, relatively high Si/Al ratio, and affinity for hydrophobic
compounds.
Figure 2. Racemization of (R)-1-phenylethanol with acid zeolites. a) H-Y
zeolites, b) H-MOR zeolites, and c) H-Beta zeolites and H-MCM-22.
During racemization, no acetophenone was detected; ad-
dition of acetaldehyde to a reaction in a closed pressure
vessel did not influence its reaction rate. This confirms that
on acid zeolites, such as H-Beta, racemization is catalyzed
by Brçnsted acids and not by Lewis acids, as in the related
MPVO redox reactions.[21]
the apolar substrate takes place in an excess of water, it is
clear that the zeolite must be sufficiently hydrophobic so
that adsorption of the benzylic alcohol is preferred over
water adsorption.
Two phenomena might explain why only samples with
high Si/Al ratio have appreciable racemization activity.
Firstly, racemization is a strongly acid-catalyzed reaction,
and it is known that the acid strength in zeolites increases
with the Si/Al ratio. Secondly, the hydrophobicity of the lat-
tice increases with the Si/Al ratio. Because racemization of
Kinetics and temperature dependence of racemization: The
kinetics of the racemization reaction can be described either
as the rate of interconversion of enantiomers or as the rate
of formation of the racemic mixture.[26] The first approach
can be described by reversible first-order kinetics as shown
in Equation (1):
388
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 386 – 397

Acid Zeolite Catalysts
FULL PAPER
1
t = is the time required to obtain a mixture with an enantio-
2
meric excess of 50%.
If Equation (2) is valid for racemization, plots of the per-
centage S enantiomer formed versus time should be identi-
cal for reactions with 50 or 150% of the original amount of
substrate, but using the same amount of catalyst. This can
be clearly seen in Figure 3b, which proves that the assump-
tion of a first-order rate in alcohol substrate is correct.
Additionally, the temperature dependence of racemiza-
tion was investigated by screening the performance of the
different racemization catalysts (0.5m pTSA, Amberlystꢁ 15,
and H-Beta) at four different temperatures (Figure 4 and
Table 3). The highest activation energy (Ea =112 kJmolꢀ1)
d½RŠ
ð1Þ
ꢀ
¼ k1½RŠꢀkꢀ1½SŠ
dt
where [R] and [S] are the concentrations of the R and S
enantiomers, and k1 and kꢀ1 are the interconversion con-
stants for the R and S enantiomers, respectively. Initially,
when one of the enantiomers dominates the medium is
chiral, and therefore, k1 may differ from kꢀ1. In the case of a
diluted solution, however, one can assume that the differ-
ence is negligible and that k1 =kꢀ1 =k. If the racemization
reaction starts from the pure R enantiomer, the elaboration
of this first-order differential, with the condition [S]t =
[R]0ꢀ[R]t, leads to Equation (2):
ꢀ
ꢁ
½RŠ
0
ln
¼ 2kt
ð2Þ
2½RŠtꢀ½RŠ
0
where [R]0 and [R]t are the concentrations of R enantiomer
at time zero and at time t, respectively. Plotting the left-
hand term of Equation (2) versus time results in a straight
line for the data of the zeolite H-Beta catalyzed racemiza-
1
tion of (R)-1-phenylethanol (Figure 3a). The half-life t = can
2
1
easily be calculated from the equation t = =ln2/2k, in which
Figure 4. Arrhenius plots for the racemization of (R)-1-phenylethanol
2
over acid catalysts under standard conditions. Catalysts: Amberlystꢁ 15
~
&
^
(
), H-Beta ( ), and 0.5m pTSA ( ).
Table 3. Temperature dependence of racemization.
Catalyst
lnk=a(1/T)+b
R2
Ea
[kJmolꢀ1
]
Amberlystꢁ
H-Beta zeolite
pTSA (0.5m)
1
5
k=ꢀln8642(1/T)+20807
lnk=ꢀ11087(1/T)+28979
lnk=ꢀ13531(1/T)+36472
0.9937
0.9703
0.9881
71.9
92.1
112.5
was obtained for the homogeneous catalyst pTSA. The
values were lower for the H-Beta zeolite (92 kJmolꢀ1) and
especially for the Amberlystꢁ resin (72 kJmolꢀ1). This indi-
cates that for the heterogeneous catalysts, and particularly
for Amberlystꢁ, diffusion of the reactants might slow down
the overall conversion. However, the effect of diffusion on
the rate seems limited for H-Beta, and the reaction is as-
sumed to be practically under chemical control.
Substrate concentration: In order to investigate the effect of
the substrate concentration on racemization, reactions using
H-Beta were performed with different amounts of water
(Table 4). If the substrate concentration was too high
(>0.1m), the side reaction to the 1-phenylethyl ether
became important (Scheme 2, Table 4 entry 1). The dilution
Figure 3. a) Kinetics of H-Beta-catalyzed standard racemization of (R)-1-
phenylethanol. The straight line was fitted using Equation (2), assuming
reversible first-order kinetics. b) Percentage inverted alcohol as a func-
~
tion of time for H-Beta-catalyzed racemization reactions with 15 ( ), 30
&
^
( ), and 45 mg ( ) (R)-1-phenylethanol in 10 mL water, using 80 mg H-
Beta.
Chem. Eur. J. 2005, 11, 386 – 397
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
389

P. A. Jacobs et al.
Table 4. Effect of substrate concentration on racemization.[a]
showed no racemization after extended reaction times
(Table 5, entry 1). For (R)-3-chloro-1-phenyl-1-propanol, the
substrate solubility was low but sufficient to make racemiza-
tion possible (entry 2). The substrate was completely race-
mized within 48 h. The addition of different amounts of
water-miscible co-solvents, such as acetone, 2-butanone, or
acetic acid, did not result in any improvement of the reac-
tion rate. It is likely that the increased substrate solubility
was partially offset by the presence of an excess of organic
molecule, which possibly adsorbed competitively on the zeo-
lite. When ketones, such as acetone or 2-butanone, were
added as co-solvents to the racemization of (R)-1-phenyl-
ethanol, a decrease of the reaction rate was indeed ob-
served, which was proportional to the amount of co-solvent
added (data not shown). (S)-2-Butanol was also screened to
test the applicability of the method to simple aliphatic alco-
hols (entry 3); however, no racemization took place, even
after extended reaction times and at an increased tempera-
ture of 808C.
Volume of
water [mL]
Substrate
concentration [m]
Ether
eealcohol [%]
after 30 min
2
5
8
0.125
0.05
0.03
+
ꢀ
ꢀ
24
45
55
[a] Reactions in water at 608C, 0.26 mmol (R)-1-phenylethanol, 80 mg of
H-Beta.
In the case of benzylic alcohols with electron-donating
substituents, for example, 1-(4-methoxyphenyl)ethanol, or in
the case of (S)-1-indanol, the zeolite quickly turned purple
or pink, respectively (Table 5, entry 4). Similarly, formation
of colored products has been described upon adsorption of
4-vinylanisole on dry NaY, CaY, or NaX zeolites, or on
H-Y, H-Beta, and H-ZSM-5 zeolites.[27–29] As shown in
Figure 5, virtually identical UV/Vis DRS (DRS=diffuse re-
flectance spectroscopy) spectra can be observed when the
H-Beta zeolite is exposed to 4-vinylanisole instead of 1-(4-
methoxyphenyl)ethanol, and to indene instead of 1-indanol.
In the case of 1-indanol, the substrate disappeared during
the 24 h reaction time, and indene was formed in nearly
quantitative amounts. This contrasts with all former sub-
strates, which were not dehydrated. All of these factors indi-
cate that the condensation in the zeolite pores probably pro-
ceeds by means of cationic intermediates (Scheme 3). In the
spectrum of the zeolite reacted with 1-(4-methoxyphenyl)-
ethanol, the absorptions at 368 and 580 nm have previously
been ascribed to cationic intermediates.[28,29] Indene conden-
sation on zeolites has also been described[27] with similar
UV/Vis absorption maxima as in the present study. Thermo-
gravimetric analysis (TGA) of a H-Beta zeolite exposed to
1-(4-methoxyphenyl)ethanol showed that the sample lost ap-
proximately 20% of its weight between 300 and 5008C. This
proves that the pore volume is indeed filled to a large
extent with rather stable organic residues.
Scheme 2. Formation of 1-phenylethyl ether as a potential side reaction.
also affected the ee value obtained after a given time. It was
evident that the reaction rate increased with the substrate
and catalyst concentration.
Substrate scope: A number of commercially available chiral
alcohols were used as the racemization substrate with H-
Beta in the standard conditions developed for (R)-1-phenyl-
ethanol (Table 5). One of the drawbacks of the methodology
is the low solubility of apolar substrates in pure water. This
was definitely the case for (S)-1-(2-naphthyl)ethanol, which
Table 5. Racemization of chiral commercial benzylic alcohols.[a]
Entry
Substrate
t [h]
eealcohol [%]
1
48
>99
2
24
28
Dynamic kinetic resolution: As it is likely that racemization
occurs through a Brçnsted acid catalyzed mechanism, it
must be performed in pure water or in a water-rich environ-
ment to avoid undesired reactions of the carbenium ion. An
excess of water effectively suppresses dehydration to sty-
renes or ether formation (Scheme 2). On the other hand, a
successful kinetic resolution requires that the enantiomeri-
cally pure product is stable, not only towards racemization,
but also towards other chemical degradation reactions. The
most common enzymatic resolution of alcohols is the lipase-
48
48
1
3
4
>99
3
<5[b]
[a] Reactions in water at 608C, 0.26 mmol enantiopure alcohol, 80 mg of
H-Beta. [b] Less than 10% of the amount of 1-indanol remained.
390
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 386 – 397

Acid Zeolite Catalysts
FULL PAPER
In a biphasic reaction, the concentration of the substrate
in either phase affects the rates of racemization and esterifi-
cation. Therefore, the partitioning behavior of 1-phenyletha-
nol was quantified by comparison with the internal standard
tetradecane, which completely resides in the apolar layer.
Various parameters, such as presence of the zeolite catalyst
or the concentration of the acylating agent, were investigat-
ed (Table 6). The addition of H-Beta zeolite to the biphasic
Table 6. Distribution of 1-phenylethanol over the biphasic system.[a]
Octanoic acid
equiv
Zeolite[b]
Substrate in
organic phase[c] [%]
–
–
10
20
20
ꢀ
+
ꢀ
ꢀ
+
80
58
88
94
87
[a] Reaction system at 608C containing 50 mL water, 50 mL octane,
775 mg 1-phenylethanol, and internal standard. [b] 155 mg of H-Beta zeo-
lite. [c] Measurement after complete equilibration (>1h stirring at reac-
tion temperature).
system resulted in an increased residence of the substrate in
the aqueous layer, probably due to adsorption of the sub-
strate on the zeolite. Conversely, addition of 10 or 20 equiv-
alents, corresponding to approximately 2 or 4 mL, respec-
tively, of an apolar acid such as octanoic acid resulted in an
increased dissolution of 1-phenylethanol in the organic
layer. In all cases, the reaction products, namely, acetyl, oc-
tanoyl, and lauryl esters, were exclusively present in the or-
ganic layer due to their pronounced apolarity. Hence, an
easy product separation was possible.
Figure 5. DRS spectra of H-Beta zeolite exposed to: a) 1-(4-methoxyphe-
nyl)ethanol, b) 4-methoxystyrene, c) 1-indanol, and d) indene. H-Beta
zeolite (1g) in water (5 mL) was stirred overnight at 60 8C in the pres-
ence of alkene or alcohol (2 mmol).
The efficiency of the mass transfer of the substrate be-
tween the two layers was investigated in the following repre-
sentative experiment: a 0.37m HCl solution (50 mL) and
octane (50 mL), containing an internal standard, were intro-
duced into the reactor and brought to 608C. Subsequently,
1-phenylethanol (775 mg) was introduced into the upper
layer, and the amount of 1-phenylethanol in the upper layer
was monitored as a function of time under stirring at
300 rpm (Figure 6). An equilibrium concentration was
reached within one hour. An additional indication of effi-
cient mass transfer was the small difference between the ee
values of the alcohol substrate in the water and organic
Scheme 3. Cationic intermediates resulting from condensation reactions
within the H-Beta zeolite pores.
catalyzed (trans)acylation to enantiopure esters. While these
esters are not susceptible to acid-catalyzed racemization,
they are easily hydrolyzed in a water-rich environment, as
the lipase also accelerates the reverse-hydrolytic reaction.
To physically separate racemization in the aqueous medium
from the resolution reaction in the non-aqueous medium, a
biphasic water/apolar solvent system was used.
Figure 6. Representative distribution of 1-phenylethanol over the bi-
phasic system as a function of time. Conditions: 775 mg 1-phenylethanol,
50 mL of 0.37m HCl solution, 50 mL octane, 608C.
Chem. Eur. J. 2005, 11, 386 – 397
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
391

P. A. Jacobs et al.
layers during DKR. In the water layer the ee was lower due
to the ongoing racemization, but the difference between it
and the ee of the substrate in the organic layer was less than
5% for all dynamic kinetic resolution measurements per-
formed.
DKR was also considered. Of course, the reaction is then
limited by equilibrium, but this equilibrium can be shifted to
the right by using an appropriate excess of the acid.
The use of acetic acid (AA), which was attempted first,
was not successful (Table 8, entry 1). The high concentration
of acetic acid in the water probably facilitated the aselective
acid-catalyzed esterification, and thus led to low ee values
for the ester product. Moreover, the presence of a large
amount of acetic acid (20 equivalents equals 1.45 mL) also
increased the water content of the organic layer and thus fa-
vored the ester hydrolysis, with the direct consequence of a
low overall yield. When the C2 acid, with a logP=ꢀ0.44,
was replaced with a more apolar reagent such as octanoic
acid (OA, logP=2.68), the ee was very high (ꢁ98%), indi-
cating that the acylation was nearly exclusively catalyzed by
the enzyme in the upper layer. Between 2 and 20 equiva-
lents of the carboxylic acid were added at the start of the re-
action (Table 8, entries 2–5). With acid additions of five
equivalents and more, the eventual yields were well above
50%, proving that DKR, with
Effect of the acyl donor: Apart from the choice of the organ-
ic solvent, the most important parameter in biphasic dynam-
ic kinetic resolution is the choice of the acyl donor. One
may choose either (activated) esters or carboxylic acids as
the acylating agents; in these cases, the resolution is either a
transesterification or an esterification.
In enol esters, the leaving group is an unstable alcohol
that tautomerizes to a ketone or an aldehyde. These enol
esters are commonly used as acyl donors in kinetic resolu-
tion reactions to shift the equilibrium of the esterification to
completion. The use of 1-ethoxyvinyl esters, which has re-
cently been reported, is based on a similar principle, with
ethyl acetate as a byproduct.[30] In Table 7 the different vinyl
coupled racemization and reso-
lution, has indeed proceeded.
When comparing the activat-
ed esters to the acids as acyl
donors, it was seen that DKR
was faster with addition of the
Table 7. Dynamic kinetic resolution with enol esters as acylating agents.[a]
Entry
Acyl donor
logP[b]
Acyl
t
Enzyme
amount [mg]
Yield
[%]
eeester
[%]
equiv
[h]
1
2
3
4
5
vinyl acetate (VA)
0.58
0.94
3.59
5.80
3.59
2
2
2
2
10
2.5
1.5
4
5
24
11
11
11
11
110
18
10
28
22
74
80
73
8198
>99
>99
>99
>99
98
isopropenyl acetate (IPA)
vinyl octanoate (VO)
vinyl laurate (VL)
enol ester at
a
rate of
2 equivhꢀ1 than the reactions
with octanoic acid (Table 7,
entry 7 versus Table 8). Howev-
er, when the ester was com-
pletely added at the beginning
of the reaction, no profit con-
cerning yield, reaction time, or
vinyl octanoate (VO)
48
24
48
8
98
98
6
vinyl octanoate (VO)
3.59
20
110
7
vinyl octanoate (VO)
3.59
16[c]
1
1
0
>9990
[a] Reaction conditions: 1.27 mmol 1-phenylethanol, 50 mL water, 50 mL octane, air atmosphere, 608C,
413 mg H-Beta zeolite CP814E-22, Novozymꢁ 435, acyl donor added at the beginning of the reaction. [b] Cal-
culated values using IA_LogPꢁ predictor (available at http://www.logp.com). [c] 2 equivhꢀ1 VO.
Table 8. Dynamic kinetic resolution reactions with carboxylic acids as
acylating agents.[a]
esters that were applied in biphasic DKR are represented,
together with the maximal yield obtained when acyl donor
(two equivalents) was added at the start of the reaction (en-
tries 1–4). Clearly, in the biphasic system the enol ester was
susceptible to enzyme- or zeolite-catalyzed hydrolysis.
Apolar enol esters, such as vinyl octanoate (VO), prefer to
reside in the octane layer (approximately 85%, depending
on the reaction conditions) and are therefore less suscepti-
ble to zeolite-catalyzed hydrolysis. Consequently, yields of
the desired ester product were higher with VO than with
vinyl acetate (VA) or isopropenyl acetate (IPA).
Entry
Acid[b]
Acid
equiv
t
Yield
[%]
eeester
[%]
[h]
1[c]
2
AA
OA
OA
OA
OA
20
20
10
5
24
48
18
214
69
78
68
74
24
54
28
10
5.5
22
5.5
22
5.5
>99
98
3
>99
>99
>99
>99
>99
4
22
22
5
2
When the amount of VO was increased to 10 or 20 equiv-
alents (together with the amount of enzyme), yields of up to
73 or 81% could be achieved within 24 and 48 h, respective-
ly (Table 7, entries 5 and 6). Increasing the VO excess from
10 to 20 equivalents seemed to have a negligible effect;
moreover, the VO was completely hydrolyzed after 2 and
3 h, respectively. As the hydrolysis of these enol esters leads
to the formation of carboxylic acids, the ongoing resolution
reaction changes from a transesterification into an esterifica-
tion. Therefore, the use of these acids as acylating agents in
[a] Reaction conditions: 50 mL water, 50 mL octane, 1.27 mmol 1-phenyl-
ethanol, 413 mg H-Beta zeolite CP814E-22, 110 mg Novozymꢁ 435, 608C.
[b] AA=Acetic acid, OA=octanoic acid. [c] Reaction performed under
identical conditions, but with 6.35 mmol substrate.
enantioselectivity was gained with respect to the use of the
corresponding acid. In this context, the acid is preferred be-
cause it is much cheaper. By contrast, when VO was added
in portions of 2 equivhꢀ1, a fast and high-yielding DKR re-
392
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 386 – 397

Acid Zeolite Catalysts
FULL PAPER
action was achieved: a yield of 90% (>99% ee) could be
reached within 8 h after the addition of 16 equivalents of
VO in total (Table 7, entry 7).
donor hydrolysis, while the major part of the acyl donor was
lost through zeolite-catalyzed hydrolysis. Hence, an in-
creased amount of enzyme does not lead to substantial loss
of acyl donor, and results in a much faster overall resolution.
Increasing the amount of enzyme to 220 mg did not further
improve the yield (entry 3).
Effect of the organic solvent: As explained before, it is im-
perative to use a water-immiscible solvent in biphasic DKR,
in order to perform the enzyme-catalyzed resolution in the
absence of water. The solvents toluene, octane, and tetrade-
cane were investigated, using just two equivalents of vinyl
octanoate as the acyl donor. These solvents have different
logP values, and water and the alcohol substrate are expect-
ed to dissolve to a varying extent in them. As can be seen in
Table 9, the use of toluene resulted in a poor yield. This is
As the amount of zeolite affects the racemization rate, a
changing zeolite concentration also influences the evolution
of the substrate ee during DKR. After a few hours, the sub-
strate ee became stationary; this stationary ee value was
higher in reactions with less zeolite. The ester yield depends
on the zeolite concentration in a complex way. When IPA
was used, the zeolite evidently catalyzed the undesired hy-
drolysis of the acyl donor, and this effect was expected to
decrease the ester yield. On the other hand, when IPA was
used as the acyl donor in the presence of a large amount of
zeolite, a lot of acetic acid became available by hydrolysis of
the IPA, and this resulted in spontaneous formation of both
the (R)- and (S)-ester. This explains why the yield was more
or less constant, but the ester ee decreased when the
amount of zeolite was increased (Table 10, entries 4–7).
When VO was used as the acyl donor, aselective transes-
terification was reduced to a minimum. This allowed the use
of a larger amount of zeolite to keep the ee of the substrate
at a low level and to preserve a high product ee (Table 7,
entry 7).
Table 9. Effect of the solvent on DKR of 1-phenylethanol.[a]
Solvent
logP
Reaction
time [h]
Ester
yield [%]
eeester
[%]
toluene
octane
tetradecane
2.7
4.76
7.14
4.5
4
24
6
28
30
>99
>99
>99
[a] Reaction conditions: 1.27 mmol 1-phenylethanol, 50 mL water, 50 mL
organic solvent, air atmosphere, 608C, 413 mg H-Beta zeolite CP814E-
22, 11 mg Novozymꢁ 435, 2 equiv VO at t=0 h.
probably due to the relatively high water content of saturat-
ed toluene, and results in hydrolysis of both the acyl donor
and the ester product. Switching to the more apolar octane
or tetradecane resulted in substantially higher yields. How-
ever, in tetradecane the reaction was much slower while the
yield was not significantly increased (Table 9, entry 3).
Reaction temperature: As DKR comprises two different re-
actions with a different response to temperature, it is likely
that there exists an optimal temperature for the overall
DKR. According to the supplier, the lipase displays an opti-
mum activity in the range of 40–608C combined with a high
value for the enantiomeric ratio E. However, it is unclear
how fast the enzyme deactivation is. Literature indicates
that a reaction temperature of 608C should not result in
enzyme denaturation.[31] Furthermore, the temperature will
affect other phenomena, such as substrate distribution and
diffusion, and also the side reactions. Biphasic DKR reac-
tions were performed at 50, 60, and 808C, and the corre-
sponding yields and ee values were compared. As expected,
the racemization reaction was fastest at the highest tempera-
ture, resulting in a consistently low(er) ee value during the
whole reaction. On the other hand, the temperature depen-
dence of the yield was influenced by the resolution method.
In the case of transesterification
Amounts of enzyme and zeolite: For successful DKR, the
resolution rate should never exceed the racemization rate
too much, to avoid depletion of the resolved enantiomer;
this could result in an ee decrease of the ester reaction prod-
uct.[2,12] Therefore, the amounts of zeolite and enzyme must
be well adjusted while keeping the overall rate at an accept-
able level.
The impact of the amount of enzyme is shown in Table 10
(entries 1–3). As can be seen, an increase from 11 to 110 mg
resulted in a remarkable improvement of the ester yield and
ee (entries 1and 2). Control experiments showed that the
enzyme was only responsible for a minor fraction of the acyl
with VO (added at 2 equivhꢀ1),
the highest yield was obtained
Table 10. Effect of the amount of enzyme and zeolite.[a]
at 608C (Figure 7a and b). On
the other hand, with octanoic
acid as the acylating agent (ten
equivalents at t=0), the yield
was highest for T=808C (Fig-
ure 7c and d). In all reactions of
Figure 7, the ee of the 1-phenyl-
ethyl octanoate product was
>99%. In summary, the absence
of a clear pattern is probably
Entry
Enzyme
amount [mg]
H-Beta zeolite
amount [mg]
t
Acyl donor
Yield
[%]
eeester
[%]
eealcohol
[%]
[h]
1
2
3
4
5
6
7
11
110
220
110
110
110
110
413
413
413
413
207
104
52
3
3
3
8
8
8
8
VO
VO
VO
IPA
IPA
IPA
IPA
22
68
60
69
61
69
70
96
>99
>99
75
78
82
17
43
44
22
29
41
63
89
[a] Reaction conditions: 1.27 mmol 1-phenylethanol, 50 mL water, 50 mL octane, air atmosphere, 608C, H-
Beta zeolite CP814E-22, Novozymꢁ 435, 2 equivhꢀ1 acyl donor.
Chem. Eur. J. 2005, 11, 386 – 397
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
393

P. A. Jacobs et al.
Figure 7. Effect of temperature on DKR. Yield=yield of (R)-1-phenylethyl octanoate. Reaction conditions for a) and b): 1.27 mmol 1-phenylethanol,
50 mL water, 50 mL octane, air atmosphere, 413 mg H-Beta zeolite CP814E-22, 110 mg Novozymꢁ 435, 2 equivhꢀ1 vinyl octanoate. Temperatures in a):
~
&
^
50 ( ), 60 ( ), and 808C ( ). Reaction conditions for c) and d): 1.27 mmol 1-phenylethanol, 50 mL water, 50 mL octane, air atmosphere, 413 mg H-Beta
zeolite CP814E-22, 110 mg Novozymꢁ 435, 10 equiv octanoic acid. Temperatures in c): 50 ( ), 60 ( ), and 808C ( ). Data in b) and d) taken after 3 h.
~
&
^
due to the number and complexity of the reactions and phe-
nomena involved.
sion, and the basket containing the original enzyme was
mounted again in the reactor. As can be observed from
Figure 8, there is no significant difference in reaction prog-
ress at any time during the reaction; both curves follow a
similar course.
Reaction upscale, product isolation, and re-use: To evaluate
the usefulness of the system for synthesis on a gram scale,
the concentration of the substrate 1-phenylethanol was in-
creased fivefold, together with the amounts of zeolite and
enzyme. VO was added as an acylating agent at a rate of
2 equivhꢀ1. After 8 h, a maximum yield of 77% was ob-
tained (eeester =99%). In the next attempt, the substrate con-
centration was again increased by five times (0.127m,
0.775 g), but the amounts of enzyme and zeolite were kept
equal, and the acylating agent was octanoic acid. This result-
ed in a yield of 80% after 96 h (eeester =95%). This DKR
was repeated without the tetradecane internal standard, en-
abling isolation of the pure reaction product. After removal
of the zeolite-containing water layer, the organic layer was
washed three times with a 0.6m NaHCO3 aqueous solution
(50 mL) to remove the excess octanoic acid together with
the residual substrate. After drying with MgSO4 and solvent
evaporation, an isolated yield of 72% in 95% pure (R)-1-
phenylethyl octanoate was collected. This clearly shows that
the system is also useful for syntheses on a larger, gram
scale.
Figure 8. Re-use of the enzyme and zeolite suspension (after it was ex-
tracted to remove all residual substrate). Reaction conditions: 1.27 mmol
1-phenylethanol, 50 mL water, 50 mL octane, air atmosphere, 608C,
413 mg H-Beta zeolite CP814E-22, 110 mg Novozymꢁ 435, 2 equivhꢀ1
VO.
As the described protocol uses a heterogeneous zeolite
catalyst and an immobilized enzyme, the re-use of both cata-
lysts was attempted. After the first run, the organic layer
was removed from the reactor, and the water layer was
washed twice with pure octane to extract all residual sub-
strate. A fresh solution of the substrate (155 mg, 1.27 mmol)
in octane (50 mL) was added to the aqueous zeolite suspen-
Substrate scope: To evaluate the scope of the new DKR,
several benzylic alcohols were subjected to biphasic DKR,
and the results were compared to those with the standard
substrate (Table 11, entry 1). With 1-(4-methoxyphenyl)etha-
nol, the reaction was performed in toluene as the molecule
does not dissolve in octane. Under these conditions, the
yield was low and the ee was moderate, as might be expect-
394
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 386 – 397

Acid Zeolite Catalysts
FULL PAPER
Table 11. DKR of different benzylic alcohols.[a]
With 1-(4-bromophenyl)ethanol (Table 11, entries 6 and
7), a satisfactory yield of 74% 1-(4-bromophenyl)ethyl ester
in 98% ee was achieved within 96 h. Compared to 1-phenyl-
ethanol, the reaction was slower; this could be due to a
lower resolution rate or to a slower racemization rate, as a
result of the electron-withdrawing substituent in the para
position. The latter would translate into a continuously high
substrate ee during the reaction. Since a true depletion of
Entry
Substrate
t
eesubstrate
Yield
[%]
eeester
[%]
[h]
[%]
1
22
<5
78
1
98
2[b]
21–
3
[c]
the R enantiomer was not observed at any point during the
79
reaction, a slow resolution is probably the main reason for
the low overall reaction rate. Yields were again lower in tol-
uene than in octane.
Finally, 1-phenylpropanol was transformed into its octa-
noate by means of DKR: a yield of 79% with an enantio-
meric purity of 97% was achieved after 70 h (entry 8). In
order to get a smooth conversion, it was necessary to in-
crease the amount of enzyme to 330 mg.
3
17.5
<5
84
>99
4[b]
41
74
<5
66
39
94
5
62
>99
Conclusions
6
96
39
74
98
Use of zeolites as acid catalysts for alcohol racemization
were shown to be clearly superior to other heterogeneous or
homogeneous acids in the racemization of (R)-1-phenyletha-
nol. The best zeolite, H-Beta, was selected for an in-depth
investigation of reaction conditions, kinetics, and substrate
scope. Electron-rich benzylic alcohols were not suitable sub-
strates for racemization, because of the formation of stable
cationic dimers in the zeolite pores, but industrially relevant
substrates, like (R)-3-chloro-1-phenyl-1-propanol, could be
successfully racemized.
Biphasic dynamic kinetic resolution was studied in detail
and optimized. This reaction used two cheap, robust, and re-
usable catalysts. For a series of benzylic alcohols, the enan-
tiopure esters were obtained in yields well above 50% with
excellent enantioselectivities, proving that racemization and
resolution have been successfully coupled.
7[b]
8[d]
130
53
45
79
98
97
70
<5
[a] Reaction conditions: 1.27 mmol racemic alcohol, 50 mL water, 50 mL
octane, air atmosphere, 608C, 413 mg H-Beta zeolite CP814E-22, 110 mg
Novozymꢁ 435, 20 equiv octanoic acid. [b] Toluene was used as organic
solvent. [c] All substrate was reacted to other species (see above).
[d] Amount of enzyme increased to 330 mg.
ed, owing to the previously mentioned side reactions. After
21h no substrate was left in the reaction (entry 2). With 1-
(4-tolyl)ethanol, an enantiopure ester was obtained in yields
comparable to those for 1-phenylethanol. The low substrate
ee during the reaction indicated a smooth racemization
(entry 3). For the same substrate in toluene, both yield and
ee were decreased, in line with earlier data (entry 4). For 1-
(3-trifluoromethylphenyl)ethanol, the product yield slowly
approached 50% after an extended reaction time (entry 5).
The substrate ee increased during the reaction. Comparison
of ester yield and substrate ee indicated that the substrate
was not being racemized. In that case, kinetic resolution is
the only reaction going on, and the yield–substrate ee rela-
tion is given by yieldKR =100eeOH/(100+eeOH).
The lack of racemization, even at 908C, was not due to
zeolite deactivation, as zeolite isolated from this reaction
was white and was still able to racemize (R)-1-phenyletha-
nol. A more likely cause is that during DKR, more than
98% of the substrate resides in the organic layer, making it
practically unavailable for the zeolite. Apparently, the intro-
duction of a CF3 group strongly increases the logP value of
the substrate.
Experimental Section
Materials: All compounds were used as-received: 1-phenylethanol
(Fluka, >98%), (R)-1-phenylethanol (Acros, >99% ee), 1-phenyl-1-
propanol (Fluka, 98%), (R)-3-chloro-1-phenyl-1-propanol (Fluka,
98% ee), 1-(p-tolyl)ethanol (Fluka, 97%), 4-bromo-a-methylbenzyl alco-
hol (Acros, 98%), 4-methoxy-a-methylbenzyl alcohol (Aldrich, 99%), 3-
(trifluoromethyl)-a-methylbenzyl alcohol (Acros, 96%), (S)-1-indanol
(Aldrich, 99%), 1-(2-naphthyl)ethanol (Aldrich, 98%), styrene (Janssen
Chimica, 99%), indene (Acros, 90%), p-methoxystyrene (Acros, 96%),
octanoic acid (Fluka, 99.5%), vinyl laurate (Fluka, >99%), vinyl acetate
(Acros, >99%), isopropenyl acetate (Fluka, 99%), vinyl octanoate
(ABCR, 98%), n-octane (Acros, >99%), tetradecane (Acros, >99%),
toluene (Acros, >99%), acetic acid (UCB, >96%), toluene-4-sulfonic
acid monohydrate (Sigma, 99%), hydrochloric acid (VEL, 37%), Na-
fionꢁ NR-50 and SAC-13 (Aldrich), Novozymꢁ 435 Candida Antarctica
Lipase B immobilized on acrylic resin (Aldrich).
Instrumentation: Yields and enantiomeric purities of substrates and reac-
tion products were determined with GC (HP 6890) on a CP-CHIRASIL-
DEX CB chiral column (25 m) with FID detection. Yields were deter-
mined using tetradecane as an internal standard. The presence of the di-
Chem. Eur. J. 2005, 11, 386 – 397
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
395

P. A. Jacobs et al.
meric ether side product, for example, di(a-methylbenzyl) ether, was in-
vestigated by using GC (HP 5890A) on a 30 m CP-Sil-5 CB column with
FID, and a GC-MS 8000 (Fisons Instruments), equipped with a 30 m
BPX5 SGE column and an MD 800 mass-spectrometer.
resolution of 1-phenylethanol,[33] and because it is commercially available
as small particles (in the range of 0.3–0.9 mm). The basket was made of
inox gauze with a pore diameter of 0.15 mm; hence, the Novozymꢁ 435
particles were well retained inside the basket. The reactor was immersed
in an oil bath in such a way that the water layer was completely beneath
the oil surface. Resulting internal temperatures for the organic upper
layer are listed in Table 12. The slight temperature difference between
the two layers was favorable for the overall reaction, because high tem-
peratures favored the rate of racemization in the lower water layer, while
they decreased the stability of the biocatalyst in the upper layer.
UV/Vis DRS measurements were performed by using a UV/Vis light
source (Top Sensor Systems DH-2000 deuterium-halogen source) and a
photodiode array detector (Ocean Optics SD 2000), both with optical
fiber technology (Top Sensor Systems FCB-UV400-ME cable and FCG-
UV400-0.1-XHT probe). For TGA measurements, a Setaram TG-DTA
92 thermobalance was employed. The sample (approximately 25 mg) was
heated from 293 to 1073 K at 1 Kminꢀ1 in a flow of helium containing
20% oxygen.
Table 12. Temperatures of heating bath and upper layer in the biphasic
reactor.
Racemization reactions: Racemization activities of homogeneous or het-
erogeneous acids in water were compared by using (R)-1-phenylethanol
as a standard substrate.
Tbath [8C]
Torganic layer [8C]
50
44
60
55
80
73
Two dissolved acids, namely, HCl and pTSA, were tested in two different
concentrations (0.5m and 0.1m). (R)-1-Phenylethanol (31.7 mg,
0.26 mmol) was added to the acidic solution (5 mL) in a 10 mL glass reac-
tor and vigorously stirred (900 rpm) at 608C in air.
Acknowledgements
The solids tested as heterogeneous catalysts were either resins or zeolites.
All materials were commercially available samples, except the H-MCM-
22 zeolite, which was synthesized in-house following a standard synthesis
procedure.[32] For the standard racemization tests, the resin or zeolite
(80 mg) was suspended in doubly distilled water (5 mL) in a 10 mL glass
reactor, and (R)-1-phenylethanol (31.7 mg, 0.26 mmol) was added. Al-
though the substrate was only sparingly soluble in pure water at room
temperature, the solubility seemed to be sufficient under reaction condi-
tions. The reaction mixture was heated to 608C in air while being vigo-
rously stirred (900 rpm).
SW acknowledges KULeuven for a fellowship as Research Assistant.
This work was sponsored by the Belgian Federal IAP program (Supra-
molecular chemistry and supramolecular catalysis) and by the Flemish
government (GOA action). The authors also acknowledge Dr. Ir. Marijke
Groothaert for assistance with the UV/Vis DRS measurements.
Biphasic DKR reactions: The reaction set-up consisted of a 100 mL glass
batch reactor containing water (50 mL) and a water-immiscible solvent
(50 mL) with a density <1gcmꢀ3. A stirring rotor was installed in the
water layer and rotated at 300 rpm. This ensured dispersion of the zeolite
in the water layer and mass transfer of the substrate between the two
layers at a sufficient rate without turbulent mixing of the two phases;
hence, the water–organic interphase remained clearly observable. A trun-
cated-conical basket, containing the immobilized enzyme, was mounted
on the rotating shaft at an appropriate height to be totally immersed in
the upper organic layer without contact with the lower water layer
(Figure 9).
[1] R. S. Ward, Tetrahedron: Asymmetry 1995, 6, 1475.
[2] H. Stecher, K. Faber, Synthesis 1997, 1 .
[3] K. Faber, Chem. Eur. J. 2001, 7, 5004.
[4] H. Pellissier, Tetrahedron 2003, 59, 8291.
[5] O. Pamies, J.-E. Bꢂckvall, Curr. Opin. Biotechnol. 2003, 14, 407.
[6] O. Pamies, J.-E. Bꢂckvall, Chem. Rev. 2003, 103, 3247.
[7] N. J. Turner, Curr. Opin. Chem. Biol. 2004, 8, 114.
[8] U. T. Bornscheuer, R. J. Kazlauskas, Hydrolases in Organic Synthe-
sis. Regio- and Stereoselective Biotransformations, Wiley-VCH,
Weinheim, 1999.
[9] B. G. Davis, V. Borer, Nat. Prod. Rep. 2001, 18, 618.
[10] G. Carrea, S. Riva, Angew. Chem. 2000, 112, 2312; Angew. Chem.
Int. Ed. 2000, 39, 2226.
[11] D. E. J. E. Robinson, S. D. Bull, Tetrahedron: Asymmetry 2003, 14,
1407.
[12] C. S. Chen, Y. Fujimoto, G. Girdaukas, C. J. Sih, J. Am. Chem. Soc.
1982, 104, 7294.
[13] E. J. Ebbers, G. J. A. Ariaans, J. P. M. Houbiers, A. Bruggink, B.
Zwanenburg, Tetrahedron 1997, 53, 9417.
[14] a) J. H. Koh, H. M. Jung, M.-J. Kim, J. Park, Tetrahedron Lett. 1999,
40, 6281; b) H. M. Jung, J. H. Koh M.-J. Kim, J. Park, Org. Lett.
2000, 2, 2487.
[15] Q. B. Broxtermann, G. K. M. Verzijl, J. G. De Vries, WO 0190396,
2001.
[16] F. F. Huerta, A. B. E. Minidis, J.-E. Bꢂckvall, Chem. Soc. Rev. 2001,
30, 321.
[17] J. H. Choi, Y. H. Kim, S. H. Nam, S. T. Shin, M.-J. Kim, J. Park,
Angew. Chem. 2002, 114, 2479; Angew. Chem. Int. Ed. 2002, 41,
2373.
[18] A. Dijksman, J. M. Elzinga, Y.-X. Li, I. W. C. E. Arends, R. A. Shel-
don, Tetrahedron: Asymmetry 2002, 13, 879.
Figure 9. Biphasic reaction set-up.
[19] M. Ito, A. Osaku, S. Kitahara, M. Hirakawa, T. Ikariya, Tetrahedron
Lett. 2003, 44, 7521.
[20] J. H. Choi, Y. K. Choi, Y. H. Kim, E. S. Park, E. J. Kim, M.-J. Kim, J.
Park, J. Org. Chem. 2004, 69, 1972.
[21] D. Klomp, T. Maschmeyer, U. Hanefeld, J. A. Peters, Chem. Eur. J.
2004, 10, 2088.
[22] S. Y. Kalliney, M. V. Ruggeri, WO 91/08196, 1991.
[23] D. W. House, US 5476964, 1995.
Unless mentioned otherwise, typical reaction conditions were: 1-phenyl-
ethanol (155 mg, 1.27 mmol), doubly distilled water or aqueous acid
(50 mL), organic solvent (50 mL), 608C, H-Beta zeolite (413 mg), and
Novozymꢁ 435 lipase (11 or 110 mg, corresponding to 138 or 1375 propyl
laurate units, respectively). Products added during the reaction, for exam-
ple, the acyl donor, were introduced into the upper layer. Novozymꢁ 435
was chosen as it displays excellent activity and enantioselectivity in the
396
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2005, 11, 386 – 397

Acid Zeolite Catalysts
FULL PAPER
[24] S. Wuyts, K. De Temmerman, D. E. De Vos, P. A. Jacobs, Chem.
Commun. 2003, 1928.
[30] Y. Kita, Y. Takebe, K. Murata, T. Naka, S. Akai, J. Org. Chem. 2000,
65, 83.
[25] S. Wuyts, D. E. De Vos, F. Verpoort, D. Depla, R. De Gryse, P. A.
Jacobs, J. Catal. 2003, 219, 417.
[26] E. L. Eliel, S. H. Wilen, Stereochemistry of Organic Compounds,
Wiley, New York, 1994, p. 426.
[27] K. Pitchumani, V. Ramamurthy, Chem. Commun. 1996, 2763.
[28] V. J. Rao, N. Prevost, V. Ramamurthy, M. Kojima, L. J. Johnston,
Chem. Commun. 1997, 2209.
[31] E. E. Jacobsen, T. Anthonsen, Can. J. Chem. 2002, 80, 577.
[32] U. Müller, T. Hill, J. Henkelmann, A. Bçttcher, E. Zeller, DE
19951280, 2001.
[33] T. Ohtani, H. Nakatsukasa, M. Kamezawa, H. Tachibana, Y. Naoshi-
ma, J. Mol. Catal. B 1998, 4, 53.
[29] V. FornØs, H. Garcꢃa, V. Martꢃ, L. Fernꢄndez, Tetrahedron 1998, 54,
3827.
Received: July 13, 2004
Published online: November 24, 2004
Chem. Eur. J. 2005, 11, 386 – 397
www.chemeurj.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
397