Highly efficient catalytic asymmetric epoxidation of allylic alcohols by an oxovanadium-substituted polyoxometalate with a regenerative TADDOL-derived hydroperoxide

Article abstract of DOI:10.1021/ol027498p

(Matrix presented) The oxovanadium(IV) sandwich-type POM catalyzes the chemo-, regio-, and stereoselective epoxidation of allylic alcohols by chiral hydroperoxides with very high catalytic efficiency (up to 42 000 TON), a potentially valuable oxidation for the development of sustainable processes. By using the sterically demanding, TADDOL-derived hydroperoxide TADOOH as the chiral oxygen source, enantiomeric ratios (er) of up to 95:5 have been achieved.

Full text of DOI:10.1021/ol027498p

ORGANIC
LETTERS
2003
Vol. 5, No. 5
725-728
Highly Efficient Catalytic Asymmetric
Epoxidation of Allylic Alcohols by an
Oxovanadium-Substituted
Polyoxometalate with a Regenerative
TADDOL-Derived Hydroperoxide
Waldemar Adam,† Paul L. Alsters,‡ Ronny Neumann,§ Chantu R. Saha-Mo1ller,†
,†
Dieter Seebach, and Rui Zhang*
Institute of Organic Chemistry, UniVersity of Wu¨rzburg, Am Hubland,
D-97074 Wu¨rzburg, Germany, DSM Fine Chemicals, AdVanced Synthesis and
Catalysis, P.O. Box 18, 6160 MD Geleen, The Netherlands, Department of Organic
Chemistry, Weizmann Institute of Science, RehoVet 76100, Israel, and Laboratory of
Organic Chemistry, Eidgeno¨ssische Technische Hochschule, ETH Ho¨nggerberg, HCI,
Wolfgang Pauli Strasse 10, CH-8093 Zu¨rich, Switzerland
adam@chemie.uni-wuerzburg.de
Received December 19, 2002
ABSTRACT
The oxovanadium(IV) sandwich-type POM catalyzes the chemo-, regio-, and stereoselective epoxidation of allylic alcohols by chiral hydroperoxides
with very high catalytic efficiency (up to 42 000 TON), a potentially valuable oxidation for the development of sustainable processes. By using
the sterically demanding, TADDOL-derived hydroperoxide TADOOH as the chiral oxygen source, enantiomeric ratios (er) of up to 95:5 have
been achieved.
Despite the vast array of homogeneously catalyzed reactions
that has been developed in the past decades, industrial fine
chemical manufacture still heavily depends on classical
organic chemistry based on stoichiometric reagents. Large-
scale use of homogeneous catalysis not only requires high
selectiVity, but also sufficiently high catalyst actiVity to obtain
a process that is economically and environmentally viable.1
Development of highly active homogeneous catalysts that
meet industrial demands is a challenging task. At very low
catalyst concentrations and very high substrate-to-catalyst
ratios, the process becomes particularly sensitive to catalyst
deactivation by traces of impurities, which adversely affects
the robustness of the catalytic system. Unquestionably, there
is a need for highly persistent homogeneous catalysts.
Polyoxometalates (POMs), in particular their transition-
metal-substituted derivatives, have been extensively used as
oxidation catalysts in view of their resistance toward oxida-
tion, ease of preparation, and compatibility with various
oxygen sources.2,3 We found earlier the “sandwich-type”
POMs, viz. [ZnWM2(ZnW9O34)2]q-, may be used as oxida-
† University of Wu¨rzburg.
(2) For reviews, see: (a) Mizuno, N.; Misono, M. Chem. ReV. 1998, 98,
199-218. (b) Hill, C. L.; Prosser-McCartha, C. M. Coord. Chem. ReV. 1995,
143, 407-455. (c) Kozhevinikov, I. V. Chem. ReV. 1998, 98, 171-198.
(d) Neumann, R. Prog. Inorg. Chem. 1998, 47, 317-370. (e) Finke, R. G.
In Polyoxometalate Chemistry; Pope, M. T., Mu¨ller, A., Eds; Kluwer
Academic Publisher: Dordrecht, The Netherlands, 2001; pp 363-390.
‡ DSM Fine Chemicals.
§ Weizmann Institute of Science.
Eidgeno¨ssische Technische Hochschule.
(1) (a) Tucker, C. E.; De Vries, J. G. Top. Catal. 2002, 19, 111-118.
(b) Blaser, H. U. AdV. Synth. Catal. 2002, 344, 17-31.
10.1021/ol027498p CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/08/2003

tion catalysts in the presence of aqueous H2O2 because they
resist both oxidative and hydrolytic degradation.3c The good
performance of these POM catalysts with aqueous H2O2,3g,h
notably their tolerance of water, prompted us to use sandwich
POMs as epoxidation catalysts in combination with organic
hydroperoxides, which are often superior to those by H2O2
in regard to selectivity.4 Unfortunately, catalytic hydroper-
oxide oxidations are not well-suited for the manufacture of
fine chemicals in multipurpose plant equipment because of
the low process efficiency due to catalyst deactivation by
adventitious water.5 We report our unprecedented results on
the use of [ZnW(VO)2(ZnW9O34)2]12- as a highly active
catalyst for the selective epoxidation of allylic alcohols by
hydroperoxides.6 Stimulated by the high enantioselectivity
that we have obtained recently in the asymmetric Weitz-
Scheffer epoxidation of R,â enones by enantiopure hydro-
peroxides,7 notably the sterically demanding TADDOL-
derived hydroperoxide TADOOH,7b we also present here the
first POM-catalyzed asymmetric epoxidation of allylic al-
cohols by TADOOH.
catalyzed oxidations,8c and the racemic [1-(4-chlorophen-
yl)]ethyl hydroperoxide (4a) as oxygen donor (Table 1). This
Table 1. Catalytic Oxidation of Mesitylol (1a) by the Various
Transition-Metal-Substituted Polyoxometalates with the Racemic
Hydroperoxide 4a
selectivitya
tempb convna mba
diastereo
entry
M
(°C)
(%)
(%)
2a :3a
(threo:erythro)
1
2
3
4
5
6
7
8
OV(IV)
OV(IV)
Mn(II)
Ru(III)
Fe(III)
Zn(II)
20
50c
50
50
50
50
50
50
>95
>95
85
94
18
8
18
12
>95 >95:5
>95 >95:5
91:9
91:9
90:10
90:10
94:6
95:5
92:8
95:5
85
70
95
90
92
95
90:10
70:30
90:10
85:15
93:7
First an extensive screening of various transition-metal-
substituted sandwich-type POMs, namely [ZnWM2(ZnW9-
O34)2]q- or M-POM (M ) OVIV, MnII, RuIII, FeIII, ZnII, PdII,
and PtII, q ) 10-12), was conducted with racemic hydro-
peroxides. The reactions were carried out in a 1,2-dichlo-
roethane solution of the POMs, obtained by extraction from
an aqueous solution of their alkali metal derivatives, through
the addition of Aliquat 336 as a lipophilic quaternary
ammonium salt. No meticulous precautions were taken to
exclude water, other than brief drying over Na2SO4 (see
Supporting Information).
Pd(II)
Pt(II)
88:12
a Conversions (allylic alcohol), material balances, and product ratios were
determined by 1H NMR analysis of the crude reaction mixture, ca. 5% error
of the stated value. b For entries 3-8, 0.02 mol % of catalyst loading was
employed; no conversion was observed at 20 °C. c Reaction time was 6 h.
screening revealed that the oxovanadium(IV)-substituted
POM, namely [ZnW(VO)2(ZnW9O34)2]12- or OdV(IV)-
POM, was the most reactive and selective catalyst for the
epoxidation of allylic alcohols (Table 1). High yields were
obtained within 24 h at 20 °C (entry 1) or within 6 h at 50
°C (entry 2), with only 0.01 mol % of POM catalyst.
Although the sandwich-type POM [ZnW(VO)2(ZnW9-
O34)2]12- has been known for some time,9 apparently it had
hitherto not been used for catalytic epoxidations. Our results
on the epoxidation of a variety of substituted allylic alcohols
1 by the racemic (1-phenyl)ethyl hydroperoxide (4b) with
the OdV(IV)-POM catalyst are collected in Table 2. As is
evident, the secondary allylic alcohol 1a (entry 1) afforded
the epoxy alcohol in high diastereoselectivity (threo:erythro
91:09). The usual undesirable allylic oxidation (CH insertion)
was not observed, which manifests the high chemoselectivity
of this reaction. With tert-butyl hydroperoxide (Table 2, entry
2) instead of (1-phenyl)ethyl hydroperoxide (4b), a lower
conversion (65%) was obtained under identical reaction
conditions. 2-Cyclohexen-1-ol (1b) was found to be less
reactive than the acyclic allylic alcohols and gave selectively
the cis-epoxy alcohol 2b (entry 3), but a higher (50 °C)
temperature was necessary for complete conversion of the
allylic substrate. Significant in regard to regioselectivity is
1-methylgeraniol (1c) with two types of double bonds, which
gave only the 3,4 epoxide in excellent yield and high
diastereoselectivity (entry 4). Similarly, the primary allylic
Mesitylol (1a) was chosen as model substrate, which is
an established probe for the assessment of the chemoselec-
tivity (epoxidation versus allylic oxidation)8a,b and diaste-
reoselectivity (threo versus erythro epoxidation) in metal-
(3) For leading examples in catalysis, see: (a) Hill, C. L.; Brown, R. B.
J. Am. Chem. Soc. 1986, 108, 536-538. (b) Mansuy, D.; Bartoli, J. F.;
Battioni, P.; Lyon, D. K.; Finke, R. G. J. Am. Chem. Soc. 1991, 113, 7222-
7226. (c) Neumann, R.; Gara, M. J. Am. Chem. Soc. 1995, 117, 5066-
5074. (d) Bo¨sing, M.; No¨h, A.; Loose, I.; Krebs, B. J. Am. Chem. Soc.
1998, 120, 7252-7259. (e) Mizuno, N.; Nozaki, C.; Kiyoto, I.; Misono,
M. J. Am. Chem. Soc. 1998, 120, 9267-9272. (f) Weiner, H.; Finke, R. G.
J. Am. Chem. Soc. 1999, 121, 9831-9842. (g) Adam, W.; Alsters, P. L.;
Neumann, R.; Saha-Mo¨ller, C. R.; Sloboda-Rozner, D.; Zhang, R. Synlett
2002, 12, 2011-2014. (h) Adam, W.; Alsters, P. L.; Neumann, R.; Saha-
Mo¨ller, C. R.; Sloboda-Rozner, D.; Zhang, R. J. Org. Chem. In press.
(4) Sheldon, R. A.; Kochi, J. K. Metal-Catalyzed Oxidations of Organic
Compounds; Academic Press: New York, 1981; pp 48-62.
(5) For example, even in the presence of molecular sieves, the Sharpless-
Katsuki epoxidation still requires 5-10 mol % of Ti catalyst to protect the
latter from adventitious water; see: Johnson, R. A.; Sharpless, K. B. In
Catalytic Asymmetric Synthesis; Ojima, I., Ed.; Wiley: New York, 2000;
pp 231-280.
(6) Hydroperoxides such as tert-butyl hydroperoxide are known to serve
as the oxygen source for POM-catalyzed oxidations; however, these
epoxidations are of limited success in view of the low reactivity and, in
most cases, undesirable radical-type reactions, see: (a) Faraj, M.; Hill, C.
L. J. Chem. Soc., Chem. Commun. 1987, 1487-1489. (b) Neumann, R.;
Khenkin, A. M. Inorg. Chem. 1995, 34, 5753-5760.
(7) (a) Adam, W.; Rao, P. B.; Degen, H.-G.; Saha-Mo¨ller, C. R. J.
Am. Chem. Soc. 2000, 122, 5654-5655. (b) Aoki, M.; Seebach, D. HelV.
Chim. Acta 2001, 84, 187-207.
(8) (a) Adam, W.; Hajra, S.; Herderich M.; Saha-Mo¨ller, C. R. Org. Lett.
2000, 2, 2773-2776. (b) Adam, W.; Herold, M.; Hill, C. L.; Saha-Mo¨ller,
C. R. Eur. J. Org. Chem. 2002, 941-946. (c) Adam, W.; Wirth, T. Acc.
Chem. Res. 1999, 32, 703-710.
(9) The synthesis and X-ray structure of oxovanadium(IV) POM is
known; see: Tourne´, C. M.; Tourne´, G. F.; Zonnevijlle, F. J. Chem. Soc.,
Dalton Trans. 1991, 143-155.
726
Org. Lett., Vol. 5, No. 5, 2003

Table 2. Catalytic Epoxidationa of Allylic Alcohols by
[ZnW(VO)2(ZnW9O34)2]12- with the Racemic (1-Phenyl)ethyl
Hydroperoxide (4b)
Table 3. Catalytic Enantioselective Epoxidationa of the
Primary Allylic Alcohols 1 by [ZnW(VO)2(ZnW9O34)2]12- with
TADOOH
a All reactions were carried out in 1,2-dichloroethane at 20 °C (except
for entry 3) for 24 h with the particular allylic alcohol 1 (0.50 mmol), 1.1
equiv of (1-phenyl)ethyl hydroperoxide (used in all reactions unless
otherwise stated), and 0.01 mol % of [ZnW(VO)2(ZnW9O34)2]12-; in all
cases no enone (allylic oxidation) was detected. b Determined by 1H NMR
analysis of the crude reaction mixture, ca. 5% error of the stated values;
material balances >95%. c Isolated material after silica gel chromatography.
d tert-Butyl hydroperoxide was used as the oxygen source. e Based on 100%
conversion of allylic alcohol.
a Unless otherwise specified, all reactions were carried out in 1,2-
alcohol 1d (entry 5) was selectively oxidized to the corre-
sponding epoxy alcohol 2d in quantitative yield. In addition,
it was noted that under identical conditions, the unfunction-
alized cyclohexene (convn 13% within 24 h) was much less
reactive than the allylic alcohols (data not shown in Table
2).
dichloroethane at 20 °C with the particular allylic alcohol 1 (0.50 mmol),
1.1 equiv of TADOOH, and 0.01 mol % of [ZnW(VO)2 (ZnW9O34)2]12-
,
except for entry 3, for which 0.002 mol % of catalyst loading was used.
1
b Determined by H NMR analysis of the crude reaction mixture, ca. 5%
error of the stated values; material balances >95%. c Isolated material after
silica gel chromatography. d Enantiomeric ratio (er), determined by HPLC
analysis on a chiral column (Chiralcel OD), except entry 9, for which GC-
MS analysis on a chiral â-TG column was employed. e Assigned by direct
comparison of the specific rotation, determined on a polarimeter, with the
literature value. f Based on 100% conversion of allylic alcohol.
These successful results with the racemic hydroperoxide
prompted us to employ enantiomerically pure hydroperoxides
in the oxovanadium(IV)-POM-catalyzed epoxidation of al-
lylic alcohols, to assess the suitability of this catalyst system
for asymmetric oxygen transfer. A screening of diverse chiral
hydroperoxides under a variety of experimental conditions
disclosed that the sterically demanding TADDOL-derived
hydroperoxide TADOOH was especially effective for the
intended purpose, as representative results show in Table 3.10
The epoxidation of the primary allylic alcohol 1d by
TADOOH at 0 °C led to the (2R,3R) epoxy alcohol 2d in
high (er 95:5) enantioselectivity (Table 3, entry 1). Since
the reactivity was low (40% convn after 48 h) at 0 °C, the
reaction was carried out at 20 °C and complete consumption
of the allylic substrate was obtained within 30 h, to afford
the epoxide 2d quantitatively with an er value of 91:9 (entry
2).
TON!) of the oxovanadium(IV)-substituted POM in the
enantioselective epoxidation of allylic alcohols. The catalyst
loading in terms of vanadium concentration is 2-3 orders
of magnitude lower than those commonly employed in
vanadium-catalyzed epoxidations of allylic alcohols by
hydroperoxides.11,15 As expected, at elevated temperature (50
°C), the enhanced catalytic activity (>95% convn in 6 h)
was accompanied by a somewhat lower (er 85:15) enantio-
selectivity (entry 4).
In contrast to TADOOH, the enantiomerically pure (S)-
(-)-(1-phenyl)ethyl hydroperoxide12 gave substantially
lower (er 56:44) asymmetric induction with the substrate 1d,
catalyzed by OdV(IV)-POM (not shown in Table 3). An
(11) (a) Villa, A. L.; De Vos, D. E.; Verpoort, F.; Sels, B. F.; Jacobs, P.
A. J. Catal. 2001, 198, 223-231. (b) Haanepen, M. J.; Van Hooff, J. H. C.
Appl. Catal. A 1997, 152, 183-201. (c) Haanepen, M. J.; Elemans-Mehring,
A. M.; Van Hooff, J. H. C. Appl. Catal. A 1997, 152, 203-219.
(12) (S)-(-)-(1-Phenyl)ethyl hydroperoxide was prepared by enzymatic
kinetic resolution according to the literature procedure: Adam, W.; Hoch,
U.; Lazarus, M.; Saha-Mo¨ller, C. R.; Schreier, P. J. Am. Chem. Soc. 1995,
117, 11898-11901.
Significant for our aim to develop a robust catalytic
epoxidation system with hydroperoxides, similar catalytic
activity and identical enantioselectivity (entry 3) was achieved
even at the very low catalyst loading of only 0.002 mol %.
This illustrates unequivocally the high efficiency (42 000
(13) The resulting TADDOL was recovered in >95% yield by silica gel
chromatography without loss of optical purity, from which TADOOH may
be easily regenerated, see ref 7b.
(10) Only primary allylic alcohols were used because kinetic resolution
of racemic secondary allylic alcohols has been shown to give poor results
in metal-catalyzed asymmetric epoxidations, see: (a) Adam, W.; Humpf,
H.-U.; Roschmann, K. J.; Saha-Mo¨ller, C. R. J. Org. Chem. 2001, 66, 5796-
5800. (b) Adam, W.; Prikhodovski, S.; Roschmann, K. J.; Saha-Mo¨ller, C.
R. Tetrahedron: Asymmetry 2001, 12, 2677-2681.
(14) In contrast, our recent studies on the sandwich-type POM-catalyzed
epoxidation of allylic alcohols with H2O2 as oxygen source revealed that
the substituting transition metals are not directly involved in the oxygen-
transfer process, see refs 3g and 3h.
Org. Lett., Vol. 5, No. 5, 2003
727

On the basis of the present experimental facts, a catalytic
cycle is proposed, with the vanadium metal as reaction center
in the POM, which accounts adequately for the high
reactivity and selectivity displayed by the oxovanadium(IV)-
substituted POM (Scheme 1). In the proposed catalytic cycle,
the oxygen transfer takes place in a vanadium(V) template,
formed from the hydroperoxide and allylic alcohol. Clearly,
such a relatively rigid metal template plays a pivotal role in
steering the stereochemical course of the attacking chiral
hydroperoxide, since the restricted freedom of motion
through the POM ligand ensures the pertinent steric inter-
actions between the chiral hydroperoxide, the allylic alcohol,
and the POM ligand.17
Scheme 1. A Proposed Catalytic Cycle
additional advantage of the TADOOH, besides the high
enantioselectivity, is the fact that this chiral hydroperoxide
may be regenerated from the resulting TADDOL without
loss of optical purity.13
To substantiate this hypothesis, an unfunctionalized alkene,
namely 1-phenylcyclohexene, was epoxidized by TADOOH
with oxovanadium(IV) POM as catalyst under identical
reactions as for the allylic alcohols 1. Besides the poor
reactivity, the enantioselectivity was essentially negligible
(see Scheme S1 in the Supporting Information). Furthermore,
we found that related transition-metal-substituted POMs, for
example, Mn(II)-POM, which cannot react through a similar
peroxy-type template, afforded racemic epoxide in poor 2:3
product selectivity when the allylic alcohol 1g was oxidized
by TADOOH (see Table S1 in the Supporting Information).
In summary, the oxovanadium(IV)-substituted sandwich-
type POM [ZnW(VO)2(ZnW9O34]12- derives its remarkable
reactivity and selectivity in the catalytic epoxidation of allylic
alcohols from its resistance toward deterioration by hydro-
peroxides. For the first time, a POM-catalyzed asymmetric
epoxidation of allylic alcohols has been accomplished, and
that in high enantioselectivity (er values up to 95:5), by the
sterically demanding TADDOL-derived hydroperoxide
TADOOH as regenerative chiral oxygen source. The high
selectivities of this unprecedented oxygen-transfer process
have been rationalized in terms of a vanadium(V) template.
We anticipate that the efficiency (up to 42 000 TON) and
accessibility of the oxovanadium(IV) POM catalyst should
contribute to the development of sustainable oxidation
processes.
In view of these advantages, the TADOOH was used as
the chiral oxygen source for the asymmetric epoxidation of
a variety of allylic alcohols under the optimized conditions
(20 °C, 0.01 mol % of catalyst). Quite generally, the similarly
cis-disubstituted allylic alcohols 1e and 1f (entries 5 and 6)
were epoxidized in about as high enantioselectivities and
yields as substrate 1d (entry 2), but for the p-anisyl derivative
1e, the enantiomeric ratio dropped slightly to 85:15 (entry
6) from 92:8 (entry 5). The epoxidation of the monosubsti-
tuted allylic alcohols 1g and 1h afforded the corresponding
epoxides in moderate enantioselectivies (entries 7 and 8),
but in high yield. Geraniol (1i) was regioselectively epoxi-
dized exclusively to the 2,3 epoxide in excellent yield, albeit
in poor enantioselectivity (entry 9).
Of mechanistic relevance is the fact that the reactivity and
chemoselectivity (epoxidation versus allylic oxidation) of the
sandwich-type POM-catalyzed oxidation of the allylic alcohol
1a with the hydroperoxide 4a depend decisively on the
redox-type substituting transition metal in the central ring
of the sandwich POM (Table 1). This indicates that the
transition metal is directly involved in the oxygen-transfer
process.14 Moreover, the reactivity pattern of the oxovana-
dium(IV) POM in the present study parallels that of earlier
studies on the catalytic epoxidation of allylic alcohols by
oxovanadium complexes,15 i.e., allylic alcohols react faster
than unfunctionalized alkenes and geraniol shows a high
regioselective preference for the allylic double bond. In this
context, our more recent studies are of interest, in which for
the asymmetric epoxidation of allylic alcohols by TADOOH
er values of up to 86:14 have been achieved; instead of the
sandwich oxovanadium(IV) POM, the VO(OiPr)3 catalyst
was used in the presence of an achiral hydroxamic acid
ligand.16
Acknowledgment. We thank the European Commission
(SUSTOX grant, G1RD-CT-2000-00347) and the Fonds der
Chemischen Industrie for financial support, and A. K. Beck
(ETH, Zu¨rich) for the preparation of TADOOH. R.Z. is
grateful to the Alexander-von-Humboldt Foundation for a
postdoctoral fellowship (2001-2002).
Supporting Information Available: Experimental details
and data on the reactivity and selectivity of the Mn(II)-POM
(Table S1) and the epoxidation of 1-phenylcyclohexene
(Scheme S1). This material is available free of charge via
the Internet at http://pubs.acs.org.
(15) (a) Sharpless, K. B.; Michaelson, R. C. J. Am. Chem. Soc. 1973,
95, 6136-6137. (b) Michaelson, R. C.; Palermo, R. E.; Sharpless, K. B. J.
Am. Chem. Soc. 1977, 99, 1990-1992. (c) Sharpless, K. B.; Verhoeven, T.
R. Aldrichim. Acta 1979, 12, 63-73.
(16) Vogl, N. Diplomarbeit, University of Wu¨rzburg, Germany, 2002.
(17) It would be too speculative at this time to propose a definite structure
for the vanadium(V)-centered POM template to rationalize the stereochem-
ical course of the epoxidation.
OL027498P
728
Org. Lett., Vol. 5, No. 5, 2003