Chemoenzymatic dynamic kinetic resolution of acyloins

Article abstract of DOI:10.1021/jo051661n

Acyloins (α-hydroxy ketones) are important building blocks in organic synthesis, e.g., for the total synthesis of epothilones. Optically pure acyloins can be obtained by lipase-catalyzed kinetic resolution (KR) of the racemate with, for example, Burkholde

Full text of DOI:10.1021/jo051661n

Chemoenzymatic Dynamic Kinetic Resolution of Acyloins
†
§
,†
Peter O¨ dman, Ludger A. Wessjohann, and Uwe T. Bornscheuer*
Department of Technical Chemistry and Biotechnology, Institute of Chemistry and Biochemistry,
Greifswald University, Soldmannstrasse 16, D-17487 Greifswald, Germany, and Department of
Bioorganic Chemistry, Leibniz Institute of Plant Biochemistry (IPB), Weinberg 3,
D-06120 Halle, Germany
uwe.bornscheuer@uni-greifswald.de
Received August 8, 2005
Acyloins (R-hydroxy ketones) are important building blocks in organic synthesis, e.g., for the total
synthesis of epothilones. Optically pure acyloins can be obtained by lipase-catalyzed kinetic
resolution (KR) of the racemate with, for example, Burkholderia cepacia lipase, but this process
suffers from a yield limitation of 50%. To devise a dynamic kinetic resolution (DKR), we studied
the racemization of two different acyloins and corresponding esters with various amine bases and
ion exchangers. No combination of base and solvent was found that could selectively racemize the
acyloin or corresponding ester under the conditions needed for a DKR. In contrast to bases, acidic
resins (ARs) were found to racemize the acyloins selectively in n-hexane and in water. Unfortunately,
the AR deactivated the lipase, preventing a one-pot DKR. Minor side reactions involving the AR,
the substrate acyloin, and the vinyl ester acyl donor were also observed. However, an efficient
DKR was made possible by the spatial separation of lipase and ion exchanger, with enzymatic
transesterification and AR-catalyzed racemization taking place simultaneously in two compartments
connected by a pump loop. The conversion of substrate alcohol was 91%, the selectivity toward the
product butyrate ester 90%, and the enantiomeric excess of the (S)-product 93% ee.
Introduction
Hydrolases are well suited as catalysts for resolution-
type reactions. The high regio- and enantioselectivity of
2
Today there is a great demand for optically pure
specialty chemicals in, for example, the pharmaceutical
industry, where the stereochemistry of the products must
be carefully controlled. In a kinetic resolution (KR), a
widely used method to separate the two enantiomers in
many lipases and esterases allows for the complete
separation of the enantiomers of many racemic alcohols,
esters, amides, and amines. However, combining enzy-
matic catalysis with in situ racemization in DKR can be
problematic, because many compounds only racemize
under harsh conditions that are incompatible with the
1
a racemic mixture, an enantioselective catalyst (e.g., an
enzyme) reacts faster with one of the enantiomers than
with the other. The main drawback of this method, which
has been extensively used for the resolution of a large
number of compounds, is that the product yield is limited
to 50%. This can be overcome by combining the kinetic
resolution with racemization of the unwanted enantiomer
by using a so-called dynamic kinetic resolution (DKR).
In DKR, the wanted enantiomer of the substrate is
continuously being replenished via the racemization, and
a (theoretical) 100% yield is possible.
3
enzyme activity. Common racemization methods include
4
5
3
metal-, base-, acid-, and enzyme-catalyzed methods,
along with thermal and reduction/oxidation methods.
Very promising results have been obtained using differ-
ent lipase-compatible Ru complexes for the racemization
of secondary alcohols via hydrogen transfer. Acid zeolites
(
2) Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in Organic
Synthesis - Regio- and Stereospecific Biotransformations, 2nd ed.;
Wiley-VCH: Weinheim, Germany, 2005.
(
3) Ebbers, E. J.; Ariaans, G. J. A.; Houbiers, J. P. M.; Bruggink,
A.; Zwanenburg, B. Tetrahedron 1997, 53, 9417-9476.
*
Tel.: +49 3834 864367. Fax: +49 3834 86 80066.
Greifswald University.
(4) Pami e` s, O.; B a¨ ckvall, J.-E. Trends Biotechnol. 2004, 22, 130-
†
135.
§
Leibniz Institute of Plant Biochemistry.
(5) Schnell, B.; Faber, K.; Kroutil, W. Adv. Synth. Catal. 2003, 345,
653-666.
(1) Faber, K. Chem.sEur. J. 2001, 7, 5004-5010.
1
0.1021/jo051661n CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/12/2005
J. Org. Chem. 2005, 70, 9551-9555
9551

O¨ dman et al.
SCHEME 1. Structures of Acyloinsa
TABLE 1. Results from Esterification of 8 with BCL or
CAL-B and Vinyl Butyrate in n-Hexane
lipase
T (°C)
time (h)
conversion (%)
E value
BCL
CAL-B
CAL-B
40
40
50
15
15
6
51
40
43
>300
48
65
TABLE 2. Deacylation of Acyloin Esters in n-Hexane/
2-Propanol (8:1) at 50 °C
a
1
general; 2 epothilone (R3 ) H or Me); 3 epothilone building
block (R4 ) H, acetate, or butyrate).
enantiomeric excessa (% ee)
acyloin
ester
conversiona
(%)
enzyme
ester
alcohol
have also recently been reported to successfully racemize
benzylic secondary alcohols under conditions compatible
with enzymatic catalysis.6
Acyloins are R-hydroxy ketones. The acyloin moiety 1
Scheme 1) is present in a variety of natural compounds,
such as antitumor agents and antibiotics. It has an
important function as a building block in organic syn-
thesis, e.g., the synthesis of epothilones. Epothilones are
a group of macrolide compounds, in nature synthesized
by the bacterium Sorangium cellulosum. In vitro, they
exhibit cytotoxic effects resembling those of paclitaxel
4
5
6
BCL
CAL-B
BCL
CAL-B
BCL
CAL-B
9
26
5
20
5
15
9
33
5
24
4
13
89
94
100
98
85
69
(
7
a
Conversion and enantiomeric excess were determined after
15 h.
antarctica lipase B (CAL-B), with vinyl butyrate as the
acyl donor, to see if this substrate was also suitable for
that kind of reaction. The results are shown in Table 1.
Both BCL and CAL-B showed good enantioselectivity
for this reaction. As can be seen, the 10 °C increase in
temperature doubled the reaction rate of CAL-B. From
Table 1, it also seems that this led to a moderate
improvement in enantioselectivity.
(
Taxol), also containing an acyloin moeity, but otherwise
structurally unrelated to this compound. Moreover, the
epothilones are active against multidrug-resistant cell
lines, which make them especially interesting.8
We have previously reported the efficient lipase- and
esterase-catalyzed kinetic resolution affording the enan-
tiomerically pure (S)-acyloin building block 3 used for the
synthesis of epothilones C and D.9
To increase the product yield in the resolution of the
optically pure acyloins, we developed a route for a
dynamic kinetic resolution as described in this paper.
To investigate the feasibility of a DKR according to
Scheme 3c, a kinetic resolution of 4-6 was performed in
this manner. CAL-B and BCL were used for deacylation
of racemic 4, 5, and 6 in 2-propanol. After 23 h, there
was no measurable conversion catalyzed by BCL, and
only ∼5% of each substrate had been converted by CAL-
B. All three reactions were (S)-selective with CAL-B. It
was suspected that the 2-propanol might inhibit the
enzymes at high concentrations, and deacylation was
therefore carried out in an n-hexane/2-propanol (8:1)
mixture. The results after 15 h are shown in Table 2.
The enzymes do seem to be inhibited or deactivated
by high concentrations of 2-propanol, since the reaction
worked much better in the n-hexane/2-propanol mixture.
Deactivation could be an effect of dehydration of the
enzyme, and inhibition could be possible, since the
substrate 2-propanol is present in high excess. E values
could not be calculated, because the equilibrium constant
for this reaction is unknown. An additional measurement
was done for the reaction with CAL-B and 4. After 22 h,
the conversion was 33%, and the enantiomeric excess of
alcohol was 94%, and for the remaining ester was 46%.
The 33% conversion after 22 h represents a 13% decrease
in reaction rate compared to the 26% conversion after
,10
Results
The acyloins 3-acetoxy-6-methyl-5-hepten-2-one (4),
-butyroxy-6-methyl-5-hepten-2-one (5), 3-butyroxy-4-
3
phenylbutan-2-one (6), 3-hydroxy-6-methyl-5-hepten-2-
one (7), and 3-hydroxy-4-phenylbutan-2-one (8) were used
in this investigation (Scheme 2).
For a dynamic kinetic resolution to be carried out, a
racemization method compatible with an enzymatic
reaction must be found. Three possible alternatives are
shown in Scheme 3.
Enzymatic hydrolysis of 4-6 (acyloin esters) and
esterification of acyloin 7 with vinyl acetate have previ-
ously been performed with excellent enantioselectivity.9
,10
Acyloin 8 was resolved via enzymatic transesterification
with Burkholderia cepacia lipase (BCL) or Candida
(
6) Wuyts, S.; De Temmerman, K.; De Vos, D. E.; Jacobs, P. A.
Chem.sEur. J. 2005, 11, 386-397.
(
7) (a) Davis, F. A.; Chen, B.-C. Chem. Rev. 1992, 92, 919-934. (b)
1
5 h. This could suggest that the enzyme is inhibited/
Wessjohann, L. A.; Scheid, G. Synthetic access to epothilones - natural
products with extraordinary anticancer activity. In Organic Synthesis
Highlights IV; Schmalz, H. G., Ed.; Wiley-VCH: Weinheim, Germany,
deactivated over time, e.g., by the 2-propanol.
Alternatively, organic bases were investigated for the
racemization efficiency in lipase-catalyzed hydrolysis
reactions according to Scheme 3a. First, racemic acyloin
acetate 4 was enantioselectively hydrolyzed to yield the
enantioenriched alcohol (S)-7 (94% ee) and ester (R)-4
(99% ee) to serve as starting materials for the racemiza-
tion studies. The enantioenriched acyloin and acyloin
ester were together subjected to racemization with the
organic bases 1,5,7-triazabicyclo[4.4.0]dec-5-ene (TBD),
2
3
000; pp 251-267. (c) Wessjohann, L. A. Angew. Chem., Int. Ed. 1997,
6, 715-718.
(
8) Kowalski, R. J.; Giannakakous, P.; Hamel, E. J. Biol. Chem.
1
997, 272, 2534-2541.
9) Scheid, G.; Ruijter, E.; Konarzycka-Bessler, M.; Bornscheuer, U.
(
T.; Wessjohann, L. A. Tetrahedron: Asymmetry 2004, 15, 2861-2869.
10) (a) Scheid, G.; Kuit, W.; Ruijter, E.; Orru, R. V. A.; Henke, E.;
Bornscheuer, U. T.; Wessjohann, L. A. Eur. J. Org. Chem. 2004, 1063-
074. (b) Wessjohann, L. A.; Scheid, G.; Bornscheuer, U.; Henke, E.;
Kuit, W.; Orru, R. WO Pat. Appl. 2002032844; Chem. Abstr. 2002, 136,
40534.
(
1
3
9552 J. Org. Chem., Vol. 70, No. 23, 2005

Chemoenzymatic Dynamic Kinetic Resolution of Acyloins
SCHEME 2. Acyloins Used in This Work
SCHEME 3. Possible Routes for a Dynamic
alcohol. Unfortunately, pure 2-propanol turned out to be
a bad environment for the resolution of acyloins (see
above), and a kinetic resolution is not feasible in this
solvent. In an n-hexane/2-propanol system however, TBD
and DBU are less selective and much slower than in both
pure n-hexane and 2-propanol, though still active.
The results from these experiments with organic bases
in different solvents are very interesting, and equally
confusing. The different racemization behavior regarding
reactivity and selectivity of the bases in different media
cannot be simply related to the polarity (dielectric
constant) or hydrophobicity (log P) of the solvent.
Next, we focused on the use of acidic resins as possible
racemization agents. Enantioenriched (S)-acyloins and
the corresponding (R)-esters were racemized at 50 °C
with different acidic resins in different solvents. The
results are presented in Table 4.
a
Kinetic Resolution
Table 4 shows that all four resins are able to selectively
2
racemize the acyloin in H O and n-hexane, without
a
(
a) Racemization of the acyloin ester in the presence of water,
affecting the ester. This means that it is possible to
racemize the residual substrate alcohol in an enzymatic
esterification in n-hexane. No hydrolysis of ester was
observed. Table 4 also shows that the vinyl acetate is not
inert in the presence of the resins, resulting in extensive
side reactions. To see if this would be a problem in a DKR
with vinyl butyrate present as acyl donor, a 10 µL/mL
to be combined with enzymatic hydrolysis of the acyloin ester. (b)
Racemization of the acyloin in organic solvent, to be combined with
enzymatic transesterification of the alcohol with, for example, a
vinyl ester as the acyl donor. (c) Racemization of the acyloin ester
in organic solvent, to be combined with deacylation of the acyloin
ester in the presence of an achiral alcohol as the acyl acceptor.
1
[
,8-diazabicyclo[5.4.0]undec-7-ene (DBU), 1,5-diazabicyclo-
4.3.0]non-5-ene (DBN), N,N,N′,N′-tetramethylethylene-
diamine (TMEDA), 1,4-diazabicyclo[2.2.2]octane (DAB-
CO), and triethylamine (Et N) in different solvents,
including n-hexane, methyl tert-butyl ether (MTBE),
THF, H O, THF with 5% v/v H O, and 2-propanol. The
2 2
vinyl butyrate solution in CH Cl was incubated with 35
TABLE 3. Changes in Optical Purity (∆ee) for 4 and 7
after Racemization with Different Bases in Different
Solvents at 50 °C
3
2
2
a
∆
ee after 6.5 h (%)
results are summarized in Table 3.
Some interesting solvent effects are shown in Table 3.
In n-hexane, TBD clearly racemizes the ester faster than
it does the alcohol. When 10% MeOH is added to the
n-hexane, the TBD effects complete methanolysis of the
ester. In pure THF and MTBE, like as in n-hexane, it
racemizes the ester preferably over the alcohol. When 5%
water is added to the THF, selectivity is inverted; the
TBD prefers the alcohol and leaves the ester untouched.
Water seems to be a good solvent for the racemization of
the alcohol form: TBD, DBU, and DBN all racemize the
alcohol faster than the ester in the presence of water. It
should be noted that in the case of hydrolysis of the (R)-
base
TBD
solvent
ester 4
alcohol 7
n-hexane
92
b
90
0
88
0
18
b
n-hexane with 10% MeOH
MTBE
16
23
16
15
19
4
c
H2O
THF
c
THF with 5% H2O
2-propanol
44
8
n-hexane with 13% 2-propanold
DBU n-hexane
2
2
0
MTBE
0
c
H2O
0
25
1
THF
2
c
THF with 5% H2O
0
8
2
ester in the presence of H O, a decrease of enantiopurity
2-propanol
n-hexane with 13% 2-propanol
20
10
2
7
d
of alcohol would be observed. A slight decrease in ester
concentration was observed in the presence of TBD and
9
DBN n-hexane
1
MTBE
0
0
2
DBU (but not with DBN) in H O, but the possibility of
H2O
0
34
0
ester hydrolysis was not further investigated. In TMEDA,
THF
0
DABCO, and Et
not shown).
3
N, no racemization was observed (data
2-propanol
4
0
1
d
n-hexane with 13% 2-propanol
2
Especially interesting are the racemization properties
of TBD and DBU in 2-propanol. Selective racemization
of the ester in the presence of an alcohol allows (in theory)
for a dynamic kinetic resolution, with enzymatic deacy-
lation of the acyloin ester yielding the ester of the achiral
a
The % ee of the starting material before racemization was 99%
ee (ester) and 94% ee (alcohol). The change in optical purity was
b
monitored by GC analysis to calculate ∆ee values. Ester com-
pletely methanolyzed. Racemization performed at 37 °C. d Mea-
c
sured after 22 h.
J. Org. Chem, Vol. 70, No. 23, 2005 9553

O¨ dman et al.
TABLE 4. Results from Racemization of Acyloins 7 (92 % ee) and 8 (87 % ee), and Their Corresponding Esters 4 (99 %
ee) and 6 (97 % ee) with Different Acidic Resins in Different Solvents
a
∆
ee after 22.5 h (%)
substrates
solvent
resin
ester
alcohol
4
4
4
4
6
6
6
6
6
6
6
6
6
6
and 7
and 7
and 7
and 7
and 8
and 8
and 8
and 8
and 8
and 8
and 8
and 8
and 8
and 8
n-hexane/water
n-hexane/water
n-hexane/water
n-hexane/water
n-hexane
Amberlyst 15
Amberlite IR-120
Amberlite 200
Dowex C
Amberlyst 15
Amberlite IR-120
Dowex C
Amberlite IR-120
Dowex C
Amberlyst 15
Amberlyst 15
Amberlyst 15
Amberlyst 15
Amberlyst 15
0
0
0
0
13
39
12
29
41
26
32
2
b
b
b
b
0
b
n-hexane
0
0
b
n-hexane
THF
0
0
0
0
0
0
0
THF
1
THF
1
n-hexane/CH2Cl2 (1:1)
MTBE
0
0
diethyl ether
vinyl acetate
3
side reactionsc
a
The % ee of the starting material before racemization was as stated above. The change in optical purity was monitored by GC analysis
b
c
to calculate ∆ee values. Measured after 6 h. Extensive side reactions with multiple products prohibited measurement of the respective
enantiomeric excesses of ester and alcohol.
TABLE 5. Loss of Alcohols 7 and 8 Due to Side
Reactions during Racemization with Acidic Resins
alcohol loss
after 7 h (%)
substrate
resin
solvent
8
Amberlyst 15
Amberlite IR-120
Dowex C
Amberlyst 15
Amberlite IR-120
Dowex C
Amberlyst 15
Amberlite IR-120
Dowex C
n-hexane
18
17
21
20
26
30
6
n-hexane
n-hexane
7
7
n-hexane
n-hexane
n-hexane
n-hexane/H2O
n-hexane/H2O
n-hexane/H2O
17
11
FIGURE 1. Reactor setup for a dynamic kinetic resolution of
with CAL-B and Amberlyst 15.
8
TABLE 6. Effect of Incubation with Amberlite IR-120
on Enzyme Activity
exchanger, is much more resilient than the powdered
BCL form, the deactivation is significant and will prohibit
high activity and stability and thus the efficiency of the
DKR and the reuse of the catalyst. The mechanism of
this deactivation is unknown, but it may have to do with
the earlier-mentioned side reaction between the acidic
resin and the acyloin. It is possible that there is a reactive
byproduct or intermediate present that deactivates the
enzyme or surface effects, and local pH effects influence
the activity.
To avoid this inactivation, we designed a simple reactor
setup with two 1.5 mL glass vials connected via a pump
to achieve a spatial separation between the acidic resin
and the enzyme (Figure 1). We investigated the DKR of
the racemic acyloin (R,S)-8 by transacylation with vinyl
butyrate and CAL-B with Amberlyst 15 as the racemi-
zation agent in n-hexane. Even though CAL-B is less
enantioselective than BCL in this reaction (see Table 1),
it was desirable to use the immobilized enzyme to
simplify retention of the catalyst in the vial. The reaction
started as a kinetic resolution at 40 °C for the first 15 h,
thereby reducing the initial concentration of alcohol in
contact with the acidic resin. After this KR, the temper-
ature was increased to 50 °C, and the reaction mixture
was pumped continuously through both vials.
remaining activity
enzyme
T [°C]
after 6 h [%]
BCL
CAL-B
37
50
26
51
mg of Amberlite IR-120 at 40 °C. After 63 h, the vinyl
butyrate concentration had decreased by 23%. Side
reactions were also observed during the racemization of
acyloins in the presence of the acidic resins. The extents
of these side reactions are shown in Table 5. 7 and 8 are
quickly consumed during racemization with acidic resins
in n-hexane. Such side reactions might be acid/base-
catalyzed aldol reactions or may be due to isomerization
of the acyloin (see below). In the biphasic system, the side
reaction is slower, but still significant.
Alternatively, three basic resins (Amberlite IRA-900,
Amberlite IRA-402, and Dowex A2) were used in n-
hexane at 50 °C starting from enantioenriched (R)-4 (98%
ee) and (S)-7 (90% ee). Unfortunately, after 6 h, complete
hydrolysis of the ester was achieved by all three resins.
Thus, a DKR with these agents is not possible.
To see if an acidic resin can be used for a one-pot DKR
with a lipase, two different enzymes were incubated with
Amberlite IR-120 for 6 h. The results are shown in Table
6
.
After 22 h, the resin had turned from grayish brown
to dark brown-black, and the resin and enzyme were
replaced by fresh material. The experiment was repeated
to estimate reproducibility. Also in this run, the resin
It is evident from Table 6 that both lipases lost
considerable activity upon incubation with the acidic
resin. Although CAL-B, which is immobilized on an ion
9554 J. Org. Chem., Vol. 70, No. 23, 2005

Chemoenzymatic Dynamic Kinetic Resolution of Acyloins
SCHEME 4. Possible Mechanism for Acid-Catalyzed Racemization of Acyloins with the Possibility for the
Formation of Isomer 9
had darkened after 22 h. Results from the two dynamic
kinetic resolution runs are presented in Table 7.
anone when determining 4 and 7, and 4-methyl acetophenone
when determining 5, 6, and 8). Determination of enzyme
activity was performed by incubating with 1 mM p-nitrophenyl
acetate at pH 7.5 (50 mM sodium phosphate buffer) and
measuring the amount of formed p-nitrophenol at 410 nm on
a UV spectrophotometer.
Enzymatic Transesterification of Racemic Acyloins.
Lipase (10 mg) and vinyl butyrate (75 µL) were added to 1
mL of acyloin (10 mg/mL) in n-hexane. The mixture was
shaken in a thermoshaker at 50 °C and 1000 rpm. Samples
were taken and analyzed directly by GC.
Enzymatic Deacylation of Racemic Acyloin Esters.
Acyloin ester (1 mL,10 mg/mL) in 2-propanol or n-hexane/2-
propanol (8:1) was shaken with 10 mg of lipase on a ther-
moshaker at 50 °C and 1000 rpm. Samples were taken and
analyzed directly by GC.
TABLE 7. Dynamic Kinetic Resolution of 8 with Vinyl
Butyrate, CAL-B, and Amberlyst 15 in n-Hexanea
enantiomeric
excess of ester
(% ee)
selectivityb
conversion
%)
(
(%)
T (h)
run 1
run 2
run 1
run 2
run 1
run 2
1
2
3
5c
38
80
98
42
75
91
100
88
100
100
81
90
92
95
91
94
94
93
d
2
0
d
a
Using the reactor setup shown in Figure 1. b Defined as the
c
percentage of converted alcohol that is turned into ester. Only
KR for 15 h at 40 °C. d DKR at 50 °C.
Racemization of Acyloins and Acyloin Esters with
Organic Bases. Enantioenriched acyloins and acyloin esters
were prepared by enzymatic hydrolysis as described previ-
It is obvious from Table 7 that a dynamic kinetic
resolution of acyloins is now possible by enzymatic
transesterification with vinyl butyrate combined with
chemical racemization of the substrate alcohol by an
acidic resin. It may seem surprising that the selectivity
increases between 22 and 30 h, but this can be explained
by the fact that the extent of the side reactions decreases
with decreasing alcohol concentration (increased degree
of conversion). It should be noted that the overall
selectivity could not increase from 88 to 100% (run 1),
which was attributed to a slight error in this measure-
ment.
In conclusion, we have shown for the first time that a
dynamic kinetic resolution of acyloin 8 can be simply and
efficiently achieved by combining lipase-catalyzed trans-
esterification with a racemization of the alcohol catalyzed
by an acidic resin (Amberlyst 15) in a two-compartment
reactor. In principle, this reaction should also work for
any other type of acyloin. Conversion and selectivity were
almost excellent (g91 and g88%, respectively), as was
the enantiomeric purity of the product (S)-acyloin bu-
tyrate ester (g91% ee).
9,10
ously.
The product was dissolved in 950 µL of solvent. Base
(50 µL, 5 mg/mL) in the same solvent was added, and the vials
were shaken on a thermoshaker at 50 °C and 900 rpm.
Samples were taken and analyzed by GC.
Racemization of Acyloins and Acyloin Esters with
Acidic/Basic Resins. Enantioenriched acyloins and acyloin
esters were prepared by enzymatic hydrolysis as described
9
,10
previously.
tioenriched product was dissolved in 500 µL of n-hexane, and
added to 35 mg of resin in 500 µL of H O. In a monophasic
Racemization in biphasic systems: The enan-
2
system, the enantioenriched product was dissolved in 1 mL of
solvent. The mixture was shaken at 50 °C and 1000 rpm on a
thermoshaker. Samples were taken and analyzed by GC. Basic
-
resins had been turned into their OH form by washing with
NaOH and water, followed by water removal with acetone and
2
drying with N .
Dynamic Kinetic Resolution of 8 in a Two-Compart-
ment Reactor. The reactor setup is depicted in Figure 1. The
reactor consisted of two 2 mL HPLC vials that were sealed
with PTFE/rubber septa, thermostated with an oil bath, and
stirred with magnetic stirrers. The reaction mixture was
continuously pumped at 1.2 mL/min between the two vials by
a peristaltic pump with Teflon PFA tubing (L ) 0.61 mm).
The tubes were connected to stainless steel needles that
penetrated the septum. The vials were ventilated with similar
needles to even out built-up pressure. Reaction: 1.5 mL of
substrate solution with a concentration of 10 mg of acyloin/
mL of n-hexane was stirred at 40 °C with 150 µL of vinyl
butyrate and 10 mg of CAL-B. After a certain period of time,
the pumps were started, and the mixture was allowed to move
to the other vial, which contained 50 mg of Amberlyst 15. The
temperature was increased to 50 °C. A total of 1.5 mL n-hexane
was added to the two flasks, to compensate for the extra
volume available when the connection to the Amberlyst 15 vial
was opened. As solvent evaporated during the course of the
reaction, n-hexane was added through the ventilation needles.
These needles were also used for sampling.
We suggest the following mechanism for the AR-
3
catalyzed racemization of the acyloins (Scheme 4). This
mechanism also explains a possible side product that
might contribute to the slightly reduced selectivity, as
the acyloin isomer 9 could have been formed during the
racemization. However, this byproduct is not yet identi-
fied.
Experimental Section
Racemic acyloin esters were synthesized and characterized
9,10
as described previously. Racemic acyloins were obtained by
9,10
saponification of the corresponding acyloin ester. Enzymatic
hydrolysis of acyloin esters was performed as reported previ-
9
,10
ously.
BCL (Amano PS) and CAL-B (Chirazyme L-2, c.-f.,
Acknowledgment. We thank Ms. Gisela Schmidt
(IPB) for synthetic support. L.A.W. is indebted to the
state of Sachsen-Anhalt for support as part of an
HWP-program.
C2), as well as all resins, were from commercial suppliers.
Analytical Methods. Acyloins and their respective esters
were analyzed by chiral gas chromatography on a Shimadzu
GC-14A with a Hydrodex-â-3P column and a FID detector.
Yields were determined with an internal standard (cyclohex-
JO051661N
J. Org. Chem, Vol. 70, No. 23, 2005 9555