Highly enantioselective preparation of multifunctionalized propargylic building blocks

Article abstract of DOI:10.1002/1521-3773(20020215)41:4<634::AID-ANIE634>3.0.CO;2-U

Halogenated propargylic alcohol 2a (X = Cl, R = Ph) was prepared by the enzymatic reduction of the corresponding ketone 1a with excellent enantioselectivity and chemical yield. Both enantiomers were synthesized with high total turnover numbers of the cata

Full text of DOI:10.1002/1521-3773(20020215)41:4<634::AID-ANIE634>3.0.CO;2-U

COMMUNICATIONS
Garcin, C. Piras, A. L. de Lacey, V. M. Fernandes, E. C. Hatchikian,
M. Frey, J. C. Fontecilla-Camps, J. Am. Chem. Soc. 1996, 118, 12989
12996.
[12] a) S. Niu, L. M. Thomson, M. B. Hall, J. Am. Chem. Soc. 1999, 121,
4000 4007; b) L. De Gioia, P. Fantucci, B. Guigliarelli, P. Bertrand,
Inorg. Chem. 1999, 38, 2658 2662; c) P. Amara, A. Volbeda, J. C.
Fontecilla-Camps, M. J. Field, J. Am. Chem. Soc. 1999, 121, 4468
4477.
[4] a) A. L. de Lacey, E. C. Hatchikian, A. Volbeda, M. Frey, J. C.
Fontecilla-Camps, V. M. Fernandez, J. Am. Chem. Soc. 1997, 119,
7181 7189; b) F. Dole, A. Fournel, V. Magro, E. C. Hatchikian, P.
Betrand, B. Guigliarelli, Biochemistry 1997, 36, 7847 7854.
[5] a) M. Y. Darensbourg, E. J. Lyon, J. J. Smee, Coord. Chem. Rev. 2000,
553 561; b) F. Osterloh, W. Saak, D. Haase, S. Pohl, Chem. Commun.
1997, 979 980; c) E. Bouwman, R. K. Henderson, A. L. Spek, J.
Reedijk, Eur. J. Inorg. Chem. 1999, 4, 217 219; d) S. C. Davies, D. J.
Evans, D. L. Hughes, S. Longhurst, J. R. Sanders, Chem. Commun.
1999, 1935 1936.
[6] D. Sellmann, F. Geipel, F. W. Heinemann, Chem. Eur. J. 2002, 8, in
press.
Highly Enantioselective Preparation of
Multifunctionalized Propargylic Building
Blocks**
[7] D. H‰u˚inger, Dissertation, Universit‰t Erlangen-N¸rnberg, 1996.
[8] [Fe(CO)(PMe3)2(−S3×)] (4): A THF suspension (30 mL) of [{Fe(CO)2-
(−S3×)}2] (240 mg, 0.33 mmol) and PMe3 (0.25 mL, 2.4 mmol) was
stirred for one day. Undissolved material was removed by filtration
and the red filtrate was evaporated to dryness. Purification of the
residue by dissolution in MeOH (20 mL), filtration, evaporation to
dryness, re-dissolution of the remaining residue in toluene (3 mL), and
precipitation by adding n-pentane (15 mL) yielded pure 4 as red-
brown powder. Yield 200 mg (62%) .Elemental analysis calcd (%) for
C19H26FeOP2S3 (483.39): C 47.11, H 5.41, S 19.86; found: C 47.19, H
5.59, S 20.06; IR (KBr): nÄ ˆ 1932 cmÀ1 (CO); MS (FD, CH3CN); m/z:
Thomas Schubert, Werner Hummel, and
Michael M¸ller*
Optically active propargylic alcohols are versatile building
blocks in the synthesis of a broad variety of natural
compounds and drugs.[1] Therefore, numerous synthetic
approaches have been developed to obtain propargylic
alcohols with high enantiomeric excesses. Chemical methods
offer straightforward access to these compounds, for example,
enantioselective reductions[2] or asymmetric zinc-mediated
additions to aldehydes,[3] as well as enzymatic methods with
oxidoreductases[4] or lipases[5]. Nevertheless, most published
procedures do not give enantiomerically pure a,b-alkynyl-
substituted methanols. Moreover, there has been only one
report of a chiral a,b-alkynyl a-chloro- or a-bromohydrin.
Corey and co-workers obtained (R)-4-triisopropylsilyl-1-
chloro-3-butyn-2-ol by oxazaborolidine reduction of the
corresponding ketone. However, the authors indicated that
the bulky triisopropylsilyl group was essential for an ee of
95%.[2c] In general, reduction methods applied to a-chloro- or
a-bromo ketones need to be mild as a result of possible side
reactions at the activated a position. Furthermore, zinc-
mediated additions to aldehydes cannot be used for the
preparation of a,b-alkynyl a-chloro- or a-bromohydrins.[6]
Nevertheless, propargylic a-chlorohydrins could be easily
transformed, for example, into epoxides, thus offering novel
comprehensive applications in organic syntheses.
‡
1
484 [Fe(CO)(PMe3)2(−S3×)] ; H NMR (399.7MHz, CD 2Cl2): d ˆ 7.68
(m, 1H; C6H4), 7.54 (m, 1H; C H4), 7.42 (m, 2H; C H4), 6.97(m, 2H;
6
6
C6H4), 6.83 (m, 2H; C6H4), 1.61 (d, 9H, 2JP, H ˆ 9.4 Hz; PC3H9), 1.36 (d,
2
9H, JP, H ˆ 8.9 Hz; PC3H9); 13C{1H} NMR (100.4 MHz, CD2Cl2): d ˆ
2
214.9 (dd, 2JP, C ˆ 20.1 Hz, JP, C ˆ 21.1 Hz, CO), 156.9 (d, JP,C ˆ 9.1 Hz),
154.7(d, JP, C ˆ 12.9 Hz), 139.5 (d, JP, C ˆ 17.4 Hz), 138.5, 130.9, 130.5,
130.1, 129.0, 128.9, 128.1, 121.7, 121.3, (C H4), 18.6 (d, 1JP,C ˆ 29.2 Hz),
6
1
18.0 (d, JP, C ˆ 26.6 Hz) (PCH3); 31P{1H} NMR (161.7MHz, CD 2Cl2):
2
d ˆ 21.05, 14.23 (d, JP, P ˆ 50.4 Hz PC3H9).
[9] X-ray structure analysis of 1: Dark red to black blocks of 1 were
obtained directly from a saturated THF/MeOH reaction solution.
Suitable single crystals were embedded in protecting perfluoropo-
lyether oil. Data were collected on a Siemens P4 diffractometer using
MoKa radiation (l ˆ 71.073 pm), a graphite monochromator, and w-
scan technique. Absorption correction was applied on the basis of y
scans (Tmin ˆ 0.123, Tmax ˆ 0.149). C25H30FeNiOP2S5, crystal size
0.09  0.56  0.52 mm, monoclinic, space group P21/n, a ˆ 1032.4(2),
b ˆ 1702.2(2), c ˆ 1712.6(3) pm, b ˆ 104.72(1), Vˆ 2.9109(7) nm3, Z ˆ
4, 1calcd ˆ 1.559 gcmÀ3, Tˆ 220(2) K, m ˆ 1.632 mmÀ1, 8715 measured
reflections (4.2 < 2q < 568), 7035 unique reflections, 5476 observed
reflections, 438 parameters, wR2 ˆ 0.0832, R1 ˆ 0.0379 (I > 2s(I)). The
structure was solved by direct methods and refined by full-matrix
least-squares calculations on F 2. The structure of 1 contains both
enantiomers of 1. A disorder is observed in which the two enantiomers
share a single site. The site occupation factors have been refined giving
93.4(2)% for the major and 6.6(2)% for the minor component. The
non-hydrogen atoms of the major component have been refined with
anisotropic displacement parameters while the corresponding atoms
of the minor component have been refined isotropically with group-
wise refined isotropic diplacement parameters for the carbon atoms.
The hydrogen atoms of the major component of 1 were taken from a
difference Fourier map and were kept fixed with a common isotropic
displacement parameter. No hydrogen atoms were taken into account
for the minor component. CCDC 172509 contains the supplementary
crystallographic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.ac.uk/conts/retrieving.html (or from the
Cambridge Crystallographic Data Centre, 12, Union Road, Cam-
bridge CB21EZ, UK; fax: (‡44)1223-336-033; or deposit@ccdc.cam.
ac.uk).
In preceding studies, we synthesized a broad variety of
enantiopure propargylic alcohols by the enzymatic reduction
of alkynones.[7] Thus, both enantiomers of a,b-alkynyl-sub-
stituted methanols are accessible. Our interest in a general
approach to obtain optically pure building blocks as inter-
mediates in organic syntheses encouraged us to continue our
[*] Dr. M. M¸ller, Dipl.-Chem. T. Schubert
Institut f¸r Biotechnologie 2, Forschungszentrum J¸lich GmbH
52425 J¸lich (Germany)
Fax : (‡49)2461-613870
E-mail: mi.mueller@fz-juelich.de
[10] a) D. Sellmann, D. H‰u˚inger, F. W. Heinemann, Eur. J. Inorg. Chem.
1999, 1715 1725; b) D. Sellmann, F. Geipel, F. W. Heinemann, Eur. J.
Inorg. Chem. 2000, 271 279; c) D. Sellmann, H. Binder, D. H‰u-
˚inger, F. W. Heinemann, Inorg. Chim. Acta 2000, 300 302, 829 836.
[11] a) Y. Higuchi, T. Yagi N. Yasuoka, Structure 1997, 5, 1671 1680; b) Y.
Higuchi, H. Ogata, K. Miki, N. Yasuoka, T. Yagi, Structure 1999, 7,
549 556.
Dr. W. Hummel
Institut f¸r Enzymtechnologie
Heinrich-Heine-Universit‰t D¸sseldorf (Germany)
[**] We gratefully acknowledge financial support from the Deutsche
Forschungsgemeinschaft (SFB 380). We thank Prof. Christian Wan-
drey for his continuous and generous support.
634
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4104-0634 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 4

COMMUNICATIONS
studies towards the reduction of a-halo propargylic ketones.
We report herein a straightforward approach to both enan-
tiomers of a,b-alkynyl a-chloro- or a-bromohydrins through
the enzymatic reduction of the corresponding ketones. By
subsequent treatment with a mild base, the alcohols were
converted into propargylic epoxides to give other highly
functionalized building blocks.
activity of TBADH and recLBADH decreases, probably as a
result of steric interactions. Interestingly, this characteristic
differs for HLADH. This enzyme generally prefers cyclic
ketones, whereas aliphatic compounds are known to be poor
substrates.[12] Applying propargylic ketones, the a-halogena-
tion increases the activity notably from nonhalogenated
through chlorinated to brominated derivatives, thus making
these compounds exceptional in the substrate range of
HLADH.
Our previous results provided evidence that in the reduc-
tion of propargylic ketones by oxidoreductases, one substitu-
ent of the substrate must not exceed a certain size.[7] Most
likely, this substituent has to be small enough to fit into a
cavity of the active site of the enzyme. Therefore, we
synthesized a number of alkynones with diverse steric
demands, and determined the ability of various alcohol
dehydrogenases to reduce these compounds. In a UV assay,
we identified three alcohol dehydrogenases (ADHs) that
accept propargylic substrates with substituents of the size of
chloro- and bromomethyl groups (data not shown). a-
Halogenated propargylic ketones 1, which were easily syn-
thesized in one step,[8] were reduced by horse liver alcohol
dehydrogenase (HLADH), Thermoanaerobium brockii ADH
(TBADH), and Lactobacillus brevis ADH (LBADH) (Ta-
ble 1). All these enzymes are commercially available.[9] Addi-
tionally, two are efficiently overproduced in a recombinant
Escheria coli strain (TEADH and recLBADH).[10, 11] All these
oxidoreductases reduce aromatic and aliphatic a-chloropro-
pargylic ketones with high activity. a-Bromopropargylic
ketones are also suitable substrates; however, the enzymatic
As a result of the low solubility of 1, about 25% of a short-
chain alcohol was added, surprisingly without any significant
loss of enzymatic activity of HLADH and recLBADH.[13]
Conveniently, as all the investigated ADHs recycle the
‡
cofactor NAD(P) by oxidation of the auxiliary alcohol, the
‡
addition of a second NAD(P) -reducing enzyme was not
necessary. The large excess of short-chain alcohol shifted the
substrate/product equilibrium towards the desired propargylic
alcohol 2, thus resulting in almost quantitative conversions
with high total turnover numbers (TTN) of the cofactor.[14]
Enzymatic reductions could be carried out on a preparative
scale by using deionized water instead of buffer[15] and were
easily scaled up by using fed-batch technique. Thus, mmol
quantities of substrate 1a were converted by using as little as
0.005 mol% of cofactor, which corresponds to a TTN of
20000, and small amounts of recLBADH (ꢀ100 units
enzyme gÀ1 substrate).[16] After a reaction time of 24 36 h,
the pure product (NMR and GC analysis) was easily isolated
by extraction without further purification in >95% yield.
HPLC analysis revealed >99% conversion and, additionally,
only one single enantiomer could be detected (ee >99%).[17]
Currently, investigations into the reduction of 1-(chloro or
bromo)-3-butyn-2-one are in progress. recLBADH converts
these substrates into enantiopure R alcohols 2 (X ˆ Cl, Br;
R ˆ H), thus resulting in an interesting switch of the
enantioselectivity of the enzymatic reduction. As the enan-
tiomers (S)-2 (X ˆ Cl, Br; R ˆ H) can be obtained by
recLBADH reduction of 1b d and subsequent removal of
the silyl protecting group, this enzyme offers unique access to
a pair of enantiomers.
Propargylic alcohols 2 can easily be converted in good yield
into the corresponding epoxides 3 without racemization.
Treatment with 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) in
EtOH/H2O (4:1) at ambient temperature led to ring closure
(Scheme 1). After workup by extraction, no further purifica-
tion was necessary since by-products could not be detected.
This general approach to enantiomerically pure terminal
propargylic epoxides[18] offers multifaceted applications in
organic chemistry as has been shown, for instance, by Hiyama
and co-workers in the synthesis of HMG-CoA reductase
inhibitor NK-104.[19] For this purpose, epoxide (R)-3b was
obtained by ex-chiral pool synthesis in six steps and with an
overall yield of 20%.[18a,b] Since terminal propargylic epoxides
Table 1. Relative enzymatic activities of HLADH, TBADH and
recLBADH.
Alcohol
HLADH[a]
TBADH[a]
recLBADH[a]
activity [%][b] (configuration)
2a
X ˆ Cl
32 (R)
24 (R)
15 (R)
35 (R)
6 (S)
53 (R)
51 (R)
7
S)
0 (
(
R ˆ Ph
X ˆ Cl
2b
2c
37 S)
4 ( 28 (S)
18 (S)[c]
R ˆ TBS
X ˆ Cl
7
R)
R ˆ TMS
X ˆ Br
R ˆ TMS
X ˆ H
2d
2e[d]
4 (R)
47 S)
(
142 (R)
R ˆ Ph
Assay conditions: substrate (2 mm), NAD(P)H (0.25 mm), MgCl2 (1 mm),
TEA/NaOH buffer (100 mm), pH ˆ 7.0. [a] All alcohols have an enantio-
meric excess of >99% as determined by HPLC (2a, e) or GC (Moshers
ester of 2b d) on chiral stationary phase. The configuration of 2a and b
was determined by comparison of the optical rotation of the corresponding
epoxide with literature data. The configuration of 2c, d was determined
with respect to mechanistic aspects of enzyme catalysis and by comparison
of the NMR data of the corresponding R and S Mosher×s ester. The
configuration of 2e was determined by comparison of the optical rotation
with literature data. [b] The enzyme activity was determined by the
decrease of the NAD(P)H extinction at 340 nm relative to the 100%
standard (HLADH: cyclohexanone; TBADH: 2-butanone; recLBADH:
ethyl 5-oxohexanoate). [c] ee 98.5%. [d] The change in configuration is
because of the different CIP priorities. TMS ˆ trimethylsilyl, TBS ˆ tert-
butyldimethylsilyl.
Scheme 1. Conversion of chlorohydrins 2 into epoxides 3.
Angew. Chem. Int. Ed. 2002, 41, No. 4
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4104-0635 $ 17.50+.50/0
635

COMMUNICATIONS
[3] For a review, see: a) L. Pu, H.-B. Yu, Chem. Rev. 2001, 101, 757; b) M.
Ishizaki, O. Hoshino, Tetrahedron: Asymmetry 1994, 5, 1901; c) Z. Li,
V. Upadhyay, A. E. DeCamp, L. DiMichele, P. J. Reider, Synthesis
1999, 1453; d) D. E. Frantz, R. F‰ssler, E. M. Carreira, J. Am. Chem.
Soc. 2000, 122, 1806, and references cited therein.
and chlorohydrins can be selectively modified at C1 or C2 at
their particular functionalities as well as at C3 or C4 at the
triple bond,[20] these compounds are of great use in synthetic
chemistry.
In conclusion, we have developed an efficient method to
obtain halogenated propargylic alcohols 2 as well as terminal
propargylic epoxides 3 with excellent enantiomeric excesses.
The commercial availability and the easy handling of all
components offers ready access to building blocks that were
hardly known previously. The multifunctionality of these
enantiopure C4 units allows highly flexible synthetic trans-
formations, thus making them interesting intermediates in the
synthesis of natural compounds and drugs.
[4] a) M. De Amici, C. De Micheli, G. Carrea, S. Spezia, J. Org. Chem.
1989, 54, 2646; b) C. W. Bradshaw, H. Fu, G.-J. Shen, C.-H. Wong, J.
Org. Chem. 1992, 57, 1526; c) C. W. Bradshaw, W. Hummel, C.-H.
Wong, J. Org. Chem. 1992, 57, 1532; d) C. Zheng, V. T. Pham, R. S.
Phillips, Catal. Today 1994, 22, 607; e) J.-M. Fang, C.-H. Lin, C. W.
Bradshaw, C.-H. Wong, J. Chem. Soc. Perkin Trans. 1 1995, 967; f) C.
Heiss, R. S. Phillips, J. Chem. Soc. Perkin Trans. 1 2000, 2821, and
references cited therein.
[5] For recent examples, see: a) C. Waldinger, M. Schneider, M. Botta, F.
Corelli, V. Summa, Tetrahedron: Asymmetry 1996, 7, 1485; b) K.
Nakamura, K. Takenaka, A. Ohno, Tetrahedron: Asymmetry 1998, 9,
4429; c) Sheng-Xue Xie, J. Ogawa, S. Shimizu, Biotechnol. Lett. 1998,
20, 935; d) M. Kurihara, K. Ishii, Y. Kasahara, N. Miyata, Tetrahedron
Lett. 1999, 40, 3183.
Experimental Section
[6] E. M. Carreira, personal communication.
[7] T. Schubert, W. Hummel, M.-R. Kula, M. M¸ller, Eur. J. Org. Chem.
2001, 4181.
(S)-2a: A solution of 1a (1.56 g, 8.76 mmol) in 2-propanol (30 mL) was
‡
added over 20 h (25 mL minÀ1) to a stirred solution of NADP (3.6 mg,
4.2 mmol; 0.05 mol%), 2-propanol (30 mL) and recLBADH (150 U) in
triethanolamine HCl buffer (150 mL; 100 mm; 1 mm MgCl2; pH 6.5) at
room temperature. After stirring for an additional 28 h, deionized water
(600 mL) was added, and the reaction mixture was extracted with CH2Cl2
(3 Â 200 mL). The combined organic layers were dried over MgSO4 and
concentrated in vacuo to give analytically pure (NMR, GC-MS) alcohol
[8] Ketones 1a, 1c, 1d were prepared according to: R. A. Earl, K. P. C.
Vollhardt, J. Org. Chem. 1984, 49, 4786; ketone 1b was synthesized by
acetylation with N-methoxy-N-methylchloroacetamide: S. Nahm,
S. M. Weinreb, Tetrahedron Lett. 1981, 22, 3815.
[9] HLADH and TBADH were purchased from Sigma Aldrich, Ger-
many; LBADH can be purchased from J¸lich Fine Chemicals,
Germany.
(S)-2a as a yellow oil (1.55 g, 8.60 mmol, 98% yield). >99% ee;[17] [a]D20
ˆ
‡25.4 (c ˆ 1.4, CHCl3); 1H NMR (300 MHz, CDCl3): d ˆ 2.56 (d, J ˆ
6.1 Hz, 1H; OH), 3.76 (dd, J ˆ 11.1, 6.5 Hz, 1H; CHCl), 3.83 (dd, J ˆ
11.1, 4.1 Hz, 1H; CHCl), 4.84 (m, 1H; CH), 7.36 (m, 3H; ArH), 7.47 (m,
2H; ArH); 13C NMR (75.5 MHz, CDCl3): d ˆ 49.3 (C1), 63.2 (C2), 86.1,
86.6 (C3, C4), 122.0, 128.6, 129.1, 132.0 (ArC); HR-MS (EI): calcd for
C10H9ClO: 180.0342, found: 180.0344.
[10] S-ADH from Thermoanaerobacter ethanolicus (TEADH) has been
proven to be extremely similar to TBADH (99.1% sequence identity)
and was cloned, sequenced, and expressed in E. coli: D. S. Burdette, C.
Vieille, J. G. Zeikus, Biochem. J. 1996, 316, 115. Wet E. coli cells
(150 g) that express TEADH yield approximately 60000 units of
TEADH: see ref. [4 f]. TEADH and TBADH showed identical
relative enzymatic activities on ketones 1a e (for TBADH, see
Table 1). In analytical batches, both enzymes converted substrate 1a
into enantiomerically pure (R)-2a (> 99% ee). We thank Prof.
Phillips for a kind donation of TEADH.
[11] B. Riebel, PhD thesis, University of D¸sseldorf, 1996. A 10-liter
fermentation of E. coli strain recLBADH HB101 ‡ yields approx-
imately 600000 units of recLBADH. One unit (U) of enzyme activity
is defined as the amount of recLBADH that catalyzes the oxidation of
NADPH (1 mmolminÀ1) when incubated with ethyl 5-oxohexanoate
(10 mm) and NADPH (0.25 mm) at 258C and pH 6.5 (100 mm
phosphate buffer, 1 mm MgCl2). Assay and batches were performed
by using crude cell extracts.
(R)-2a: A solution of 1a (1.56 g, 8.76 mmol) in ethanol (45 mL) was added
‡
over 2 h (15 mLhÀ1) to a stirred solution of NAD (6.0 mg, 8.42 mmol;
0.10 mol-%), ethanol (75 mL), and HLADH (250 U) in triethanolamine
HCl buffer (500 mL, 100 mm; pH 7.0) at room temperature. After stirring
for an additional 34 h, deionized water (1000 mL) was added and the
reaction mixture was extracted with CH2Cl2 (3 Â 300 mL). The combined
organic layers were dried over MgSO4 and concentrated in vacuo to yield
analytically pure (NMR, GC-MS) alcohol (R)-2a as a yellow oil (1.53 g,
8.50 mmol, 97%). >99% ee);[17] [a]D20 ˆ À25.2 (c ˆ 1.3, CHCl3);[17]
1H NMR and 13C NMR: as for (S)-2a.
(R)-3a: Alcohol (R)-2a (588 mg, 3.3 mmol) was added to a solution of
DBU (1.5 mL, 10.0 mmol) in EtOH/H2O (20 mL, 4:1). The mixture was
stirred for 90 min at room temperature, followed by addition of H2O
(100 mL) and extraction with ethyl acetate (3 Â 30 mL). The organic layers
were dried over Na2SO4 and concentrated in vacuo to yield analytically
pure (NMR, GC-MS) epoxide (R)-3a as an orange oil (441 mg, 3.1 mmol,
93%). Purification by flash chromatography (−Iso-hexane×/ethyl acetate
30:1; −Iso-hexane×: Fluka 34969, mixture of isomers) yielded (R)-3a as a
colorless oil (376 mg, 2.6 mmol, 80%) >99% ee;[21] [a]D20 ˆ À47.0 (c ˆ 1.3,
acetone);[18c] [a]2D0 ˆ À42 (c ˆ 2.0, acetone 78% ee). 1H NMR (300 MHz,
CDCl3): d ˆ 3.02 (d, J ˆ 3.3, 2H; CH2), 3.60 (t, J ˆ 3.3, 1H; CH), 7.34 (m,
3H; ArH), 7.47 (m, 2H; ArH); 13C NMR (75.5 MHz, CDCl3): d ˆ 40.4
(C2), 49.3 (C1), 83.6, 85.9 (C3, C4), 122.1, 128.5, 129.0, 132.1 (ArC).
[12] J. Peters in Biotechnology, 2nd ed., Vol. 8a (Eds.: H.-J. Rehm, G.
Reed), Wiley-VCH, Weinheim, 1998, pp. 437 439.
[13] An aqueous HLADH solution that contained 25% ethanol and an
aqueous recLBADH solution that contained 25% 2-propanol
showed >80% remaining activity after 15 h. In contrast, TBADH
displayed <40% remaining activity under these conditions (25%
2-propanol).
[14] In the case of the enzymatic reduction of 1b, only moderate
conversions were found. However, preliminary results suggest that
the use of enzyme-coupled cofactor regeneration could result in
quantitative conversion.
[15] The use of deionized water requires pH monitoring and adjustment
with aqueous NaOH. In contrast to HLADH and recLBADH, a
decrease in enzymatic activity of TBADH in deionized water was
observed. For simplification of the g-scale conversions with
recLBADH and HLADH, these reactions were performed in aqueous
buffer.
Received: July 24, 2001 [Z17588]
[1] For selected examples, see: a) W. R. Roush, R. J. Sciotti, J. Am. Chem.
Soc. 1994, 116, 6457; b) T. Kolasa, A. O. Stewart, C. D. W. Brooks,
Tetrahedron: Asymmetry 1996, 7, 729; c) D. Vourloumis, K. D. Kim,
J. L. Peterson, P. A. Magriotis, J. Org. Chem. 1996, 61, 4848.
[2] a) K. Matsumura, S. Hashiguchi, T. Ikariya, R. Noyori, J. Am. Chem.
Soc. 1997, 119, 8738; b) K. A. Parker, M. W. Ledeboer, J. Org. Chem.
1996, 61, 3214; c) C. J. H. Helal, P. A. Magriotis, E. J. Corey, J. Am.
Chem. Soc. 1996, 118, 10938; d) P. V. Ramachandran, A. V. Teodor-
[16] In a fed-batch optimized with respect to low amounts of cofactor,
ketone 1a (1.1 mmol) in deionized water (20 mL) and 2-propanol
(8 mL) was quantitatively reduced as described in the experimental
‡
section by using NADP (0.055 mmol) and recLBADH (20 U).
[17] Enantiomers of 2a were separated by means of HPLC on
a
Chiralcel OB column (250 Â 4 mm, equipped with
a precolumn,
¬
80 Â 4 mm, Daicel Chem. Ind., 208C, 0.5 mLminÀ1, −Iso-hexane×/
ovic, M. V. Rangaishenvi, H. C. Brown, J. Org. Chem. 1992, 57, 2379,
2-propanol 95:5), Rt ˆ 39.0 min ((S)-2a), 44.7min (( R)-2a).
and references cited therein.
636
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4104-0636 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 4

COMMUNICATIONS
these complexes are stable in aqueous media,[7] none could
perform alkene epoxidation with hydrogen peroxide. It is
tempting to ascribe this to the lack of a combination of the
active titanium site with a suitable hydrophobic environment.
The same phenomenon can acount for the lack of catalytic
activity in a study by Sherrington and Alder on TiIV-grafted
polysiloxane networks prepared from silanol-rich supports.
These materials, showed no activity in alkene epoxidation
with 30% aqueous H2O2.[8] This can be ascribed to either the
presence of residual silanol groups that make the materials
more hydrophilic thus rendering unsuitable catalytic ensem-
bles or to the presence of Ti centers with inappropriate
structures.
Herein, epoxidation catalysts are reported that, like TS-1,
epoxidize 1-octene but also substrates that are too large for
TS-1 such as cyclooctene and cyclododecene. In this ap-
proach, ensembles are made from robust titanium silsesqui-
oxane complexes. These are grafted on commercially avail-
able linear methylhydrosiloxane dimethylsiloxane copoly-
mers and then cross-linked by reaction with vinyl-terminated
polydimethylsiloxanes. The resulting titanium polysiloxane
materials are hydrophobic, three-dimensionally netted poly-
mers that enclose the titanium sites in cavities that can, in
principle, be varied according to the choice of the starting
materials. Since titanium polysiloxanes are found to epoxidize
alkenes with aqueous hydrogen peroxide, while titanium
silsesquioxane complexes alone do not have this ability, we
demonstrate the need for catalytic ensembles in this area of
science.
[18] Chiral nonracemic terminal propargylic epoxides have been described
in: a) M. Lopp, T. Kanger, A. M¸raus, T. Pehk, ‹. Lille, Tetrahedron:
Asymmetry 1991, 2, 943; b) T. Kanger, P. Niidas, A.-M. M¸¸risepp, T.
Pehk, M. Lopp, Tetrahedron: Asymmetry 1998, 9, 2499 (one example
¬
derived from tartaric acid derivative is given); c) R. Sanchez-
¬
Obregon, B. Ortiz, F. Walls, F. Yuste, J. L. GarcÌa Ruano, Tetrahedron:
Asymmetry 1999, 10, 947(3 examples with ee×s up to 88% are given);
synthetically useful enantioselectivities with terminal olefins are yet to
be achieved with chiral (salen)Mniii systems: E. N. Jacobsen in
Catalytic Asymmetric Synthesis (Ed.: I. Ojima), Wiley-VCH, New
York, 1993, p. 178.
[19] K. Takahashi, T. Minami, Y. Ohara, T. Hiyama, Bull. Chem. Soc. Jpn.
1995, 68, 2649.
[20] Alcohols (R)- and (S)-2a were converted into the corresponding E-
and Z allylic alcohols by using Red-Al or Lindlar catalyst, respec-
tively. The olefinic products were obtained in high yield, without
racemization (ee >99%) and excellent E/Z ratios (Red-Al:
E/Z 100:0; Lindlar catalyst: E/Z 5:95). Thus, another class of hardly
known highly functionalized synthetic intermediates is easily acces-
sible; T. Schubert, M. M¸ller, unpublished results.
[21] The ee values were determined by conversion of epoxide (S)-3a and
rac-3a into the terminal nitrile, by treatment with a threefold excess of
KCN in ethanol/water (4:1) at 508C for 4 h. The nitrile obtained was
separated by means of HPLC on a Chiralcel OB column (208C,
0.5 mLminÀ1, −Iso-hexane×/2-propanol 95:5), Rt ˆ 68.6 min ((R)-3a),
79.3 min ((S)-3a).
Titanium Silsesquioxanes Grafted on Three-
Dimensionally Netted Polysiloxanes: Catalytic
Ensembles for Epoxidation of Alkenes with
Aqueous Hydrogen Peroxide**
ˆ
Vinyl silsesquioxane trisilanol [(H2C CH)(c-C6H11)6-
Si7O9(OH)3] (1)[9] can be easily converted to new titanium
ˆ
Maria D. Skowronska-Ptasinska,
Martinus L. W. Vorstenbosch, Rutger A. van Santen,
and Hendrikus C. L. Abbenhuis*
silsesquioxane complexes [(H2C CH)(c-C6H11)6Si7O12TiX]
(2a, X ˆ h5-C5H5 (Cp), 2b, X ˆ OiPr) by reaction with
[Cl3TiCp] or [Ti(OiPr)4], respectively. These reactions are
Few catalysts have been truly efficient in alkene epoxida-
tion with aqueous hydrogen peroxide. Development of such
catalysts is an important goal since with regard to desirability,
this oxidant comes only second to oxygen itself.[1] Currently,
the best catalyst in this field is the synthetic titanium-
containing zeolite, titanium silicalite-1 (TS-1),[2] which is
active for a wide range of oxidation reactions, including
epoxidation.[3] For TS-1, activity seems to originate from a
combination of a robust active Ti(OSi )n site (n ˆ 3, 4),[4] and
ꢁ
its location in a hydrophobic channel or cavity in the MFI
(ZSM-5) structure.[5] The resulting catalytic ensemble pre-
vents poisoning of the active site by water as well as
unproductive decomposition of the oxidant. A series of
titanium silsesquioxane complexes with structural elements
that are very similar to the active TS-1 site have been reported
to function as homogeneous catalysts.[6] Although some of
similar to those previously described for the related, unfunc-
tionalized
silsesquioxane
trisilanol
[(c-C6H11)7Si7O9-
(OH)3].[6b, 10] For both new complexes, the 13C NMR spectra
(400 MHz, CDCl3) are particularly informative, showing four
peaks for the cyclohexyl methine carbon atoms (ratio 2:1:1:2
for 2a and 1:1:2:2 for 2b) characteristic for C2-symmetric,
monomeric silsesquioxane titanium species. Attempts to
obtain crystals of 2 suitable for X-ray analysis were unsuc-
cessful thus far.
[*] Dr. H. C. L. Abbenhuis, Dr. M. D. Skowronska-Ptasinska,
M. L. W. Vorstenbosch, Prof. Dr. R. A. van Santen
Schuit Institute of Catalysis
Eindhoven University of Technology
P. O. Box 513, 5600 MB Eindhoven (The Netherlands)
Fax : (‡31)40-24-55-05-4
E-mail: H.C.L.Abbenhuis@chem.tue.nl
Vinyl-bearing metallosilsesquioxanes readily undergo plat-
inum-catalyzed hydrosilylation,[11] thus for the immobilization
[**] We thank Degussa/Creavis GmbH for financial support.
Angew. Chem. Int. Ed. 2002, 41, No. 4
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4104-0637 $ 17.50+.50/0
637