Dynamic kinetic resolution of α-hydroxy acid Esters

Article abstract of DOI:10.1021/ol000014+

Enzymatic resolution in combination with ruthenium-catalyzed racemization of the substrate led to dynamic kinetic resolution of α-hydroxy esters in good yields and excellent ee's. Studies of different parameters showed that the best results were obtained

Full text of DOI:10.1021/ol000014+

ORGANIC
LETTERS
2000
Vol. 2, No. 8
1037-1040
Dynamic Kinetic Resolution of
r-Hydroxy Acid Esters
Fernando F. Huerta, Y. R. Santosh Laxmi, and Jan-E. Ba1ckvall*
Department of Organic Chemistry, Arrhenius Laboratory, Stockholm UniVersity,
SE-106 91 Stockholm, Sweden
jeb@organ.su.se
Received January 20, 2000
ABSTRACT
Enzymatic resolution in combination with ruthenium-catalyzed racemization of the substrate led to dynamic kinetic resolution of r-hydroxy
esters in good yields and excellent ee’s. Studies of different parameters showed that the best results were obtained using Pseudomonas
cepacia lipase, ruthenium catalyst 3, and 4-chlorophenyl acetate as acyl donor in cyclohexane.
Since 1987, “the enzymatic second-order asymmetric trans-
formation of different compounds” or dynamic kinetic
resolution (DKR) has been the focus of many studies.1 This
is due to the fact that, with this new methodology, the main
drawback of the traditional kinetic resolution (maximum 50%
yield) is overcome. With DKR the unreactive enantiomer is
continuously racemized and the product can be obtained
optically pure in 100% yield (Scheme 1).
of secondary alcohols3 and diols.4 In these reactions a
ruthenium catalyst racemizes a secondary alcohol during
enzymatic esterification. Here, we report our results on the
dynamic resolution of R-hydroxy esters, applying this
racemization-esterification methodology.
We first investigated the racemization5 of (S)-methyl
mandelate6 [(S)-1a] with different ruthenium catalysts, 3-7
(Scheme 2). The best results in the racemization were
Scheme 1. Dynamic Kinetic Resolution (DKR)
Scheme 2. Racemization of (S)-Methyl Mandelate
obtained with ruthenium complexes 3 and 4. Complex 3 has
been successfully used previously for the racemization of
In connection with our work on the combination of enzyme
and transition metal catalysis in organic synthesis,2 we
recently reported on the dynamic kinetic resolution (DKR)
(2) Ba¨ckvall, J. E.; Gatti, R.; Schink, H. E. Synthesis 1993, 343-348.
(b) Gatti, R. G. P.; Larsson, A. L. E.; Ba¨ckvall, J. E. J. Chem. Soc., Perkin
Trans. 1 1997, 577-584.
(3) Larsson, A. L. E.; Persson, B. A.; Ba¨ckvall, J. E. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1211-1212. (b) Persson, B. A.; Larsson, A. L.; LeRay,
M.; Ba¨ckvall, J. E. J. Am. Chem. Soc. 1999, 121, 1645-1650.
(4) Persson, B. A.; Huerta, F. F.; Ba¨ckvall, J. E. J. Org. Chem. 1999,
64, 5237-5249.
(1) Caddick, S.; Jenkis, K. Chem. Soc. ReV. 1996, 447-456. (b) Stecher,
H.; Faber, K. Synthesis 1997, 1-16. (c) Noyori, R.; Tokunaga, M.;
Kitamura, M. Bull. Chem. Soc. Jpn. 1995, 68, 36-56. (d) Stu¨rmer, R.
Angew. Chem., Int. Ed. Engl. 1997, 36, 1173-1174.
10.1021/ol000014+ CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/28/2000

pathway.7,9 This is confirmed by the fact that in the (S)-
methyl mandelate racemization the presence of the corre-
sponding ketone10 accelerates the reaction rate. However, in
some cases there was no significant effect by the use of added
ketone and therefore it was omitted (Table 1, entries 1, 3,
and 4). Further experiments revealed that for some R-hydroxy
esters the use of added ketone has a significant influence,
vide infra.
For the kinetic resolution of hydroxy esters,11 Pseudomo-
nas cepacia lipase (PS-C, type II from Amano) was
employed. 4-Chlorophenyl acetate was used as acyl donor
because, after the acyl transfer process, the resulting chlo-
rophenol does not interfere with the ruthenium catalyst.3 The
use of vinyl acetate as acyl donor would result in the
formation of acetaldehyde, which interferes with the metal
complex employed. Different conditions, such as tempera-
ture, solvent, substrate concentration, etc. were screened
under kinetic resolution conditions (Scheme 3, Table 2).
secondary alcohols in toluene at 70 °C. This procedure was
initially used for the racemization of (S)-1a. However, low
solubility of the catalyst in this solvent together with low
conversions made it necessary to change solvent. The use
of cyclohexane as the solvent and employing 20 mol % of
methyl 2-oxo-2-phenylacetate as ketone led to substantial
racemization. Thus, with enantiomerically pure (S)-1a, 82%
racemization had occurred after 24 h (Table 1, entry 2).
Scheme 3. Kinetic Resolution of Hydroxy Acid Esters
Table 1. Racemization of (S)-Methyl Mandelatea
entry Ru cat.b
basec
solvent temp (°C) eed (%)
1
2
3
4
5
6
7
3
PhCH3
C6H12
Et3N (4 mol %) PhCH3
Et3N (2 mol %) i-C8H18
C6H12
70
60
50
60
60
40
60
60
18
0
68
80
3e
4
4
5f
6
Et3N (2 equiv) CH2Cl2
Et3N (3 equiv) CCl4
100
100
7
a All the reactions were run for 24 h. b Unless otherwise noted, 2 mol %
of catalyst was used without added ketone. c Unless otherwise noted, 1
equivof base was used. d The ee was determined by HPLC on a Chiralcel
OD-H column. e Methyl 2-oxo-2-phenylacetate (20 mol %) was used as
ketone. f Methyl 2-oxo-2-phenylacetate (1 equiv) was used as ketone.
The choice of the proper solvent is critical, because
selectivity and especially reaction times may change sig-
nificantly (Table 2, entry 3 versus entries 4 and 5). According
to the results in Table 2, it seems to be a requirement that
the â-carbon of the R-hydroxy acid ester is a secondary
As was known from previous work in this field,3,4
ruthenium complex 4 showed a higher rate than 3, and
complete racemization was achieved after 24 h (Table 1,
entry 3). However, the use of complex 4 requires a base,
which might interfere with the enzyme. Ruthenium dihydride
complex 5 has recently been found to be responsible for the
reduction of ketones to secondary alcohols under hydrogen
transfer conditions.7 Unfortunately this complex did not work
in our system. Complexes 6 and 7 have been used recently
for the racemization of secondary alcohols,8 but with hydroxy
acid ester 1a we could not obtain any racemization.
Table 2. Kinetic Resolution of Compounds 1a
temp time yield
ee
entry enzyme substr solvent
(°C)
(h)
(%)b (%)c
1
2
3
4
5
6
7
8
9
N-435
PS-D
PS-C
PS-C
PS-C
PS-C
PS-C
PS-C
PS-C
PS-C
PS-C
1a
1a
1a
1a
1a
1b
1c
1d
1e
1f
PhCH3
PhCH3
PhCH3
i-C8H18
C6H12
C6H12
C6H12
C6H12
C6H12
C6H12
i-C8H18
70
50
50
60
60
60
60
60
60
60
40
48
48
100
24
24
28
24
24
24
24
24
0
40
40
49
44
42
46
39
49
48
47
92
99
99
96
99
99
99
99
18
62
The racemization mechanism has been studied extensively,
and it was shown to proceed via a hydrogen transfer
(5) For a review on racemization of optically active compounds, see:
Ebbers, E. J.; Ariaans, G. J. A.; Houbiers, J. P. M.; Bruggink, A.;
Zwanenburg, B. Tetrahedron 1997, 53, 9417-9476.
(6) Previous experiments showed that the presence of a carboxylic acid
functionality is not compatible with the ruthenium complexes for racem-
ization purposes.
(7) Aranyos, A.; Csjernyik, G.; Szabo´, K. J.; Ba¨ckvall, J. E. Chem.
Commun. 1999, 351-352.
(8) Koh, J. H.; Jung, H. M.; Kim, M.-J.; Park, J. Tetrahedron Lett. 1999,
40, 6281-6284. (b) Koh, J. H.; Jeong, H. M.; Park, J. Tetrahedron Lett.
1998, 39, 5545-5548.
10
11
1g
a Unless otherwise noted, all the reactions were performed on a 0.25
mmol scale with 15 mg of enzyme and 2 equiv of the acyl donor (4-Cl-
C6H4OAc) in 1.25 mL of the corresponding solvent. b Yield determined
1
by H NMR. c The ee was determined on the acetate 2 by HPLC using a
Chiralcel OD-H column.
1038
Org. Lett., Vol. 2, No. 8, 2000

Table 3. Dynamic Kinetic Resolution of R-Hydroxy Acids with PS-C and Ru Catalyst 3a
entry
substrate
product
time (h)
yield (%)b
ee (%)c
94
1d
48
80
2d
72
48
74
96
94
3
76 (73e)
4
72
69
98
5
48
48
24
80 (70e)
98
30
80
6
62
7f
60
a Unless otherwise noted, all the reactions were performed on a 0.25 mmol scale with 2 mol % of Ru catalyst 3, 15 mg of PS-C, and 2 equiv of acyl donor
1
(4-Cl-C6H4OAc) in 1.25 mL of cyclohexane under argon at 60 °C. b Unless otherwise noted, yield calculated by H NMR. c The ee was determined on the
acetate 2 by HPLC using a Chiralcel OD-H column. d 20 mol % of the corresponding ketone was used. e Isolated yield. f The reaction was performed in
isooctane at 40 °C.
carbon for a good kinetic resolution to occur. Substrates 1f
and 1g (Table 2, entries 10 and 11) with a primary â-carbon
atom show very poor ee’s. Since this study is aimed at
dynamic kinetic resolution, some of the parameters such as
solvents and temperature have to fit the optimal function of
the Ru catalyst (Table 1), where higher temperatures imply
faster racemization. The temperature in the kinetic resolution
was set at the high-temperature limit for the enzyme. Above
this temperature the selectivity of the enzyme drops dramati-
cally. For P. cepacia lipase (PS-C (type II) from Amano),
this limit was at 60 °C under the conditions described above.
In general, good results were obtained with the traditional
kinetic resolution of compounds 1a-e (Table 2, entries 3-9),
and conversions in the range of 40-50% with high ee’s were
obtained. The next step was to combine the racemization
and the enzymatic transformation and run the reaction in one
pot (Scheme 4).
Scheme 4. Dynamic Kinetic Resolution of Hydroxy Acid
Esters
In Table 3, our preliminary results are summarized. An
efficient dynamic kinetic resolution was achieved for sub-
strates 1a-e with goods yields and high ee’s. Mandelic acid
ester 1a gave an 80% yield of 2a in 94% ee and ester 1b
afforded 2b in 74% yield and 96% ee (Table 3, entries 1
and 2). Ring-substituted mandelic acid ester derivatives (1c
and 1d) showed high efficiency in these reactions, proving
(9) Haack, K. J.; Hashiguchi, S.; Fujii, A.; Takehera, J.; Ikariya, T.;
Noyori, R. Angew. Chem., Int. Ed. Engl. 1997, 36, 285. (b) Santosh Laxmi,
Y. R.; Ba¨ckvall, J. E. Chem. Commun. 2000, 611-612.
(10) Methyl 2-oxo-2-phenylacetate was used as ketone.
(11) Ebert, C.; Ferluga, G.; Gardossi, L.; Gianferrara, T.; Linda, P.
Tetrahedron: Asymmetry 1992, 3, 903-912.
Org. Lett., Vol. 2, No. 8, 2000
1039

that the presence of electron-withdrawing or -donating groups
in the ring does not affect the DKR process (Table 3, entries
3 and 4). The nonaromatic, cyclohexyl analogue (1e) worked
even better under the same reaction conditions (Table 3, entry
5), giving 2e in 80% yield and 98% ee. Thus, the scope of
this methodology is not restricted to aromatic hydroxy acids.
For substrates 1f and 1g, less efficient reactions were
obtained (Table 3, entries 6 and 7), which reflects the poor
results for these substrates in the kinetic resolution (Table
2, entries 10 and 11).
The reaction of 1g under DKR conditions (Table 3, entry
7) demonstrates the efficiency of the dynamic process
compared to the kinetic resolution of this compound.
Whereas kinetic resolution gave 2g in only 47% yield with
62% ee (Table 2, entry 11), DKR gave 2g in 60% yield with
80% ee. When this reaction was run longer (48 h instead of
24 h, Table 3, entry 7), the small gain in yield produced a
negative effect on the ee (66% yield, 60% ee). These results
indicate that for substrates 1f and 1g further experiments are
needed in order to find the proper reaction conditions for a
successful dynamic kinetic resolution.
subsequent biochemical racemization.12 When this one-pot
sequence was repeated four times, the R-acetoxy acid was
obtained in 80% yield and 98% ee.
In summary, we have shown that dynamic kinetic resolu-
tion of different R-hydroxy esters was carried out successfully
to give enanitomerically pure compounds in good yields and
in one single step. It was demonstrated that the high
enantioselectivity of the enzyme is compatible with the
ruthenium-catalyzed racemization of R-hydroxy esters. The
racemization process takes place via a hydrogen transfer
pathway involving dehydrogenation of the alcohol to ketone
and readdition of hydrogen. This methodology is applicable
to a variety of substrates, although substrates with a primary
carbon at the â-position of the R-hydroxy ester gave poor
results.
Supporting Information Available: Shvo catalyst prepa-
ration details and HPLC and GC data. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL000014+
Recently, Faber et al. developed a method for deracem-
ization of mandelic acid based on enzymatic resolution and
(12) Strauss, U. T.; Faber, K. Tetrahedron: Asymmetry 1999, 10, 4079-
4081.
1040
Org. Lett., Vol. 2, No. 8, 2000