Vanadium-catalyzed asymmetric epoxidation: Proline-derived hydroxamic acids revisited

Article abstract of DOI:10.1055/s-2006-956494

Structural features of ligands and substrates affecting enantioselectivity in vanadium(V)-catalyzed epoxidation of allylic alcohols were investigated. For the hydroxamic acid ligands derived from arylhydroxylamines, 2-aryl-substituted allylic alcohols eme

Full text of DOI:10.1055/s-2006-956494

CLUSTER
3525
Vanadium-Catalyzed Asymmetric Epoxidation: Proline-Derived Hydroxamic
Acids Revisited
V
anadium-Cata
a
lyzed
A
sym
ï
metric
n
E
poxidationaba Bourhani, Andrei V. Malkov*
Department of Chemistry, WestChem, Joseph Black Building, University of Glasgow, Glasgow G12 8QQ, Scotland, UK
Fax +44(141)3304888; E-mail: amalkov@chem.gla.ac.uk
Received 1 September 2006
catalysis turns into a ligand-accelerated process.6 Al-
Abstract: Structural features of ligands and substrates affecting
though the detailed mechanism of enantiodifferentiation
enantioselectivity in vanadium(V)-catalyzed epoxidation of allylic
in vanadium-catalyzed epoxidation remains unclear, the
alcohols were investigated. For the hydroxamic acid ligands derived
best ligands developed by Yamamoto5 and others7 are
from arylhydroxylamines, 2-aryl-substituted allylic alcohols
emerged as the best substrates furnishing good level of enantio- likely to rely on the steric bulk of the substituents creating
selectivity.
an efficient chiral environment. In this context, high
selectivity exhibited by the catalytic system based on 3a
is rather conspicuous since the ligand does not offer any
appreciable steric bias.
Key words: asymmetric catalysis, epoxidations, homogeneous
catalysis, ligands, vanadium
In order to shed more light on the mechanism of enantio-
differentiation when 3a is used as a ligand, we synthesized
a series of analogues 3, 5–9 shown in Figure 1 where the
structure of the acid and the hydroxylamine moieties were
systematically varied. The results are collected in Table 1.
Epoxidation of prochiral allylic alcohols belongs to a priv-
ileged class of asymmetric transformation which is regu-
larly employed in organic synthesis for enantioselective
functionalizaton.1 Before the advent of the famous titani-
um–tartrate system,2 Sharpless reported3 that a complex
of vanadium(V) with proline-derived hydroxamic acid 3a
exhibited a good enantioselectivity in the epoxidation of
2-phenylcinnamyl alcohol (1 → 2, Scheme 1). However,
alcohol 1 turned out to be the only substrate giving appre-
ciable enantiomeric excess with ligand 3a, and more im-
portantly, vanadium systems were considered impractical
due to a pronounced ligand-decelerating effect.1a,4
R
O
O
O
N
N
R
OH
N
N
N
OH
N
OH
TFA
Ar
Ph
Ms
Ph
6a, Ar = 3,5-Me2C6H3
6b, Ar = 3,5-Cl2C6H3
6c, Ar = 1-naphthyl
3b, R = Ms
5a, R = Me
5b, R = i-Pr
5c, R = Bn
3c, R = CHO
O
O
t-BuOOH
O
O
O
1 mol% V(V),
3 mol% 3a
Ph
N
Ts
N
Ph
Ph
Ph
Ph
OH
OH
N
OH
N
OH
N
O
Ms
O
CHPh2
toluene, –20 °C
4 days
OH
Ph
MeO
O
7
9
8
2
1
90% (80% ee)
Figure 1 Chiral hydroxamic acids employed as ligands in vanadi-
um(V)-catalyzed asymmetric epoxidation of allylic alcohols.
O
O
O
N
N
N
Variation of the Carboxylic Acid Component
N
OH
R
R
OH
OH
F3C
O
Ph
First, the influence of N-substitution of proline was as-
sessed (3b–d). Sulfonamide 3b mirrored the performance
of the parent 3a (entry 2) while a more Lewis basic forma-
mide 3c capable of additional coordination to metal exhib-
ited poor reactivity and afforded racemic product (entry
3). Next, acyclic ligands 5a–c were examined to probe if
a conformational flexibility about the C–N bond benefits
the enantioselectivity.8 However, the effect was opposite;
the best in the series, alanine derivative 5a, showed only
34% ee (entry 4). Varying the nature of the 5-membered
ring also proved to be unproductive. Thus, hydroxamic
acids derived from 4-oxaproline (7) and di-O-isopropyl-
idenetartrate (8) gave rise to epoxyalcohol 2 in low ee
(entries 8, 9). It turned out that the rigid proline structure
with a non-coordinating trifluoroacetamide or sulfon-
amide group is crucial for attaining good selectivity.
4, R = CHPh2, CH2CPh3,
3a
CH(3,5-Me2C6H3)2
Scheme 1
It took another 20 years to bring vanadium back to
spotlight when Yamamoto5 introduced a new family of
ligands of which dihydroxamic acids 4, in particular,
challenged the dominance of the titanium systems in
terms of the scope and enantioselectivity. In a related in-
vestigation, we have demonstrated that in water vanadium
SYNLETT 2006, No. 20, pp 3525–3528
Advanced online publication: 08.12.2006
DOI: 10.1055/s-2006-956494; Art ID: Y01106ST
© Georg Thieme Verlag Stuttgart · New York
1
8
.1
2
.2
0
0
6

3526
Z. Bourhani, A. V. Malkov
CLUSTER
Variation of the Hydroxylamine Component
efits reactivity. In contrast, rates of epoxidation of the
fully conjugated cinnamyl alcohol 14 and its isomer 13
remained very low (entries 6 and 7).
Hydroxamic acids 6a–c and 9 were examined to assess
how the ligand efficacy depends on the nature of the hy-
droxylamine fragment. The results revealed that only aro-
matic hydroxylamines can provide an adequate level of
enantioselectivity (entries 5–7) while use of ligand 9 de-
rived from benzhydryl hydroxylamine furnished a nearly
racemic epoxide 2. Ligand 6a based on 3,5-dimethylphen-
ylhydroxylamine matched the enantioselectivity of the
parent 3a (entry 5). On the other hand, hydroxamic acids
6b and 6c derived from the respective 3,5-dichlorophenyl-
hydroxylamine and 1-naphthylhydroxylamine, produced
inferior results (entries 6 and 7).
OH
OH
Me
Ph
10
11
OH
OH
OH
Ph
Me
Ph
Ph
12
13
14
Figure
2
Substrates for vanadium(V)-catalyzed asymmetric
epoxidation.
Table 1 Vanadium(V)-Catalyzed Asymmetric Epoxidation of 2-
Phenylcinnamyl Alcohol (1) Employing Ligands 3, 5–9a
Table 2 Vanadium(V)-Catalyzed Asymmetric Epoxidation of
Allylic Alcohols 1, 10–13a
Entry
Ligand
3a
3b
3c
5a
6a
6b
6c
7
Yield (%)
eeb (%)
78
76
5
Entry
Ligand
3a
Alcohol
1
Yield (%) eeb (%), configc
1
2
79
70
28
72
75
36
66
90
81
59
1
2
3
4
5
6
7
79
40
89
91
52
13
<5
78 (S,S)
15 (S,S)
64 (+)
66 (+)
13 (S,S)
10 (S)
–
3a
10
3
3a
11
4
34
78
63
52
13
25
5
6a
11
5
3a
12
6
3a
13
7
3a
14
8
9
8
a The reactions were carried out in toluene at –20 °C for 48 h on a
1-mmol scale. The catalyst was generated in situ from (i-PrO)3VO (1
mol%) and the ligand (3 mol%).
10
9
b Determined by chiral GC or HPLC (ref. 6, 7f).
c The absolute configuration of the products deduced from their
optical rotation.
a The reactions were carried out in toluene at –20 °C for 48 h on a
1-mmol scale. The catalyst was generated in situ from (i-PrO)3VO (1
mol%) and the ligand (3 mol%).
b Determined by chiral HPLC (ref. 7f), the absolute configuration of
the product 2 deduced from its optical rotation was S,S.
To rationalize the observed pattern of reactivity, we sug-
gested that non-covalent interactions between the aromat-
ic group in position 2 of the substrates 1 and 11 and the
phenyl group of the hydroxylamine fragment of the ligand
might be involved (e.g. structure B¢, Figure 3).10 The
NMR investigation of the complexes of vanadium(V)
with hydroxamic acids revealed11 that the axial alkoxide
ligand is fairly labile and rapidly exchanges with neutral
alcohol added to the solution. It can be assumed that dia-
stereoisomeric complexes A–D, chiral at vanadium, are
the major components present in solution. With chiral
hydroxamic acid ligands, such as 3a, complexes A/B and
C/D form two sets of diastereoisomeric pairs (Figure 3),
whose relative reactivity would control the overall enan-
tioselectivity of the epoxidation. If proposed arene–arene
interactions do take place12 and exerts a considerable ef-
fect on the reaction rate, the diastereomeric structures A
and B should exercise the main influence on the reaction
outcome; while the relative reactivity and the relative pop-
ulation of A and B controlled by the chiral proline moiety
will be reflected in the enantioselectivity of the reaction.
Among solvents, toluene proved to be optimal with
dichloromethane and chloroform showing slightly re-
duced selectivities, while the use of polar solvents
(acetonitrile and methanol in particular) led to nearly
racemic products.
In order to establish the substrate scope, the most efficient
catalysts based on hydroxamic acids 3a and 6a were em-
ployed in epoxidation of a series of allylic alcohols 10–14
(Figure 2). Examination of the results (Table 2) revealed
an unusual trend. 2-Phenylcinnamyl alcohol (1, entry 1)
and its analogue 11 (entries 3, 4), which normally react
slower than geraniol (10, entry 2),6,7f exhibited faster con-
version rates.9 Presence of the aromatic group in position
2 of the substrate appears to be crucial since alcohol 12,
the positional isomer of 11, showed substantially reduced
reactivity and selectivity (entry 5). It is important to note
that the 2-aryl group in 1 and 11 for steric reasons is twist-
ed out of the conjugation with the double bond which ben-
Synlett 2006, No. 20, 3525–3528 © Thieme Stuttgart · New York

CLUSTER
Vanadium-Catalyzed Asymmetric Epoxidation
3527
alcohol (1 mmol) was then added in one portion and the mixture was
stirred for at r.t. for 10 min and then cooled to –20 °C. A 5 M solu-
tion of t-BuOOH in nonane (0.3 mL) was added and the mixture
was stirred at –20 °C for 48 h. Excess of the oxidant was decom-
posed by stirring with 3 mL of a sat. aq solution of Na2SO3 for 30
min. The resulting mixture was then diluted with H2O (10 mL) and
the aqueous phase was extracted with CH2Cl2 (2 × 20 mL). The
combined organic extracts were dried over MgSO4 and concentrat-
ed in vacuo. Purification of the products was accomplished by col-
umn chromatography on silica gel (15 × 3 cm) with an n-hexane–
EtOAc mixture (4:1). The absolute configuration of the epoxide
products was assigned by comparison of their optical rotations with
the literature data; the ee was determined using chiral GC or HPLC.
O
V
O
*R
Ar
R*
Ar
O
O
O
V
O
O
O
O
O
O
N
N
O
i-Pr
i-Pr
B
A
O
V
O
*R
Ar
R*
Ar
O
O
O
O
O
O
O
V
N
N
O
O
O
i-Pr
i-Pr
C
D
Epoxidation of 2-Phenylcinnamyl Alcohol (1)7f
Chiral HPLC: Chiracel OD-H; 0.5 mL/min; hexane–2-PrOH,
90:10, tR (R,R isomer) = 15.26 min, tR (S,S isomer) = 16.91 min.
TFA
N
N
O
V
O
O
O
Epoxidation of Geraniol (10)7f
Chiral GC: Supelco a-Dex 120 column, oven temp. 110 °C for 2
min, then 1.0 °C/min to 200 °C, tR (S,S isomer) = 24.38 min, tR (R,R
isomer) = 24.82 min.
O
O
R
B'
Epoxidation of Alcohol 11
Figure 3
(+)-(3-Methyl-2-phenyl-oxiranyl)-methanol was isolated as a clear
oil; [a]D20 +25.4 (c 0.5, CHCl3). Chiral GC: Supelco b-Dex 120 col-
umn; oven temp. 110 °C for 2 min, then 1 °C/min to 200 °C, tR (–
isomer) = 25.99 min, tR (+ isomer) = 27.69 min] showed 66% ee. 1H
NMR (400 MHz, CDCl3): d = 1.07 (d, J = 5.6 Hz, 3 H), 1.91–1.96
(m, 1 H) 3.55 (q, J = 5.6 Hz, 1 H) 3.91–4.02 (m, 2 H), 7.21–7.42 (m,
5 H). 13C NMR (100 MHz, CDCl3): d = 14.1 (CH3), 56.9 (CH), 64.6
(CH2), 127.0 (CH), 127.9 (CH), 128.4 (CH), 135.7 (C); IR (NaCl):
n = 3446, 2926, 1498, 1447, 1421, 1139, 1094, 1075, 1047, 917,
763, 704 cm–1. HRMS (CI): m/z calcd for C13H12O2: 165.0916;
found: 165.0918.
In conclusion, we have investigated structural features of
ligands and substrates affecting enantioselectivity in va-
nadium(V)-catalyzed epoxidation of allylic alcohols. For
the hydroxamic acid ligands derived from arylhydroxyl-
amines, the original architecture introduced by Sharpless
where proline served as the chiral controller turned out to
be optimal, while 2-aryl-substituted allylic alcohols
proved to be the best substrates achieving respectable
level of selectivity.
Epoxidation of 2-Methylcinnamyl Alcohol 127f
Chiral GC: Supelco b-Dex 120 column, oven temp. 110 °C for 2
min, then 1.5 °C/min to 200 °C, tR (R,R isomer) = 33.19 min, tR (S,S
isomer) = 33.71 min.
Ligand 3a
White solid, isolated as a 3:1 mixture of E- and Z-isomers; mp 104–
20
1
106 °C (hexane–EtOAc); [a]D +21.2 (c 1.00, CHCl3). H NMR
(400 MHz, CDCl3): d = 1.87–2.26 (m, 4 H), 3.67–3.79 (m, 2 H),
4.43 and 5.29 (br s, 1 H), 7.10–7.57 (m, 5 H), 8.55 and 9.04 (br s, 1
H). 13C NMR (100 MHz, CDCl3): d = 25.0 (CH2), 28.1 (CH2), 48.1
(CH2), 59.5 (CH), 116.2 (q, JCF = 284.7 Hz, CF3), 120.9 (CH), 125.8
(CH), 128.5 (CH), 140.5 (C), 156.5 (q, JCF = 39.1 Hz, COCF3),
168.6 (CO). IR (KBr): nmax = 3443, 1691, 1660, 1635, 1594, 1493,
1454, 1419, 1235, 1204, 1146, 758, 693 cm–1. HRMS (EI): m/z cal-
cd for C13H13O3N2F3: 302.0878; found: 302.0881.
Epoxidation of Alcohol 135a
Chiral HPLC: Chiracel OD-H; 0.5 mL/min; hexane–2-PrOH 95:5,
tR (R isomer) = 17.88 min, tR (S isomer) = 21.91 min.
Acknowledgment
We thank the EPSRC for the research grant GR/R86744 and the
University of Glasgow for an additional support. We thank project
students Précilia Jousselme and Jenny McGarvey for synthesis of
some intermediates.
Ligand 6a
Light yellow solid, isolated as a 1:1 E/Z mixture of isomers; mp 90–
92 °C (hexane–EtOAc); [a]D +39.0 (c 0.5, CHCl3). 1H NMR (400
MHz, CDCl3): d = 1.78–2.21 (m, 4 H), 2.21 (s, 3 H), 2.28 (s, 3 H),
3.66–3.74 (m, 2 H), 4.43 and 4.46 (m, 1 H), 6.73 and 6.92 (s, 1 H),
7.08 and 7.18 (s, 2 H). 13C NMR (100 MHz, CDCl3): d = 20.2 (CH3),
20.3 (CH3), 24.1 (CH2), 27.1 and 27.9 (CH2), 46.7 and 47.1 (CH2),
56.8 and 58.1 (CH), 117.8 and 124.1 (CH), 126.7 and 130.7 (CH),
136, 2 and 137.2 (C), 138.6 and 139.2 (C), 164.5 and 167.40 (CO),
References and Notes
(1) For reviews, see: (a) Berrisford, D. J.; Bolm, C.; Sharpless,
K. B. Angew. Chem., Int. Ed. Engl. 1995, 34, 1059.
(b) Katsuki, T. In Comprehensive Asymmetric Catalysis,
Vol. 2; Jacobsen, E. N.; Pfaltz, A.; Yamamoto, H., Eds.;
Springer: Heidelberg, 1999, 621. (c) Katsuki, T. Curr. Org.
Chem. 2001, 5, 663. (d) Keith, J. M.; Larrow, J. F.;
Jacobsen, E. N. Adv. Synth. Catal. 2001, 343, 5. (e) Adam,
W.; Malisch, W.; Roschmann, K. J.; Saha-Möller, C. R.;
Schenk, W. A. J. Organomet. Chem. 2002, 661, 3.
(2) Katsuki, T.; Sharpless, K. B. J. Am. Chem. Soc. 1980, 102,
5974.
the CF3 and F3CCO signals are too weak to be identified. IR: nmax
=
3155, 2961, 2924, 1704, 1614, 1594, 1455, 1408, 1382, 1322, 1147,
1018, 850, 704 cm–1. HRMS (EI): m/z calcd for C15H17O3N2F3:
330.1191; found: 330.1192.
General Procedure for Asymmetric Epoxidation7f
Ligand (3 mol%) and (i-PrO)3VO (2.5 mL, 10 mmol) were dis-
solved in dry toluene (3 mL) under nitrogen atmosphere and the re-
sulting deep brown solution was stirred at r.t. for 30 min. Allylic
Synlett 2006, No. 20, 3525–3528 © Thieme Stuttgart · New York

3528
Z. Bourhani, A. V. Malkov
CLUSTER
(8) (a) Malkov, A. V.; Hand, J. B.; Kočovský, P. Chem.
(3) (a) Sharpless, K. B.; Verhoeven, T. R. Aldrichimica Acta
1979, 12, 63. See also: (b) Michaelson, R. C.; Palermo, R.
E.; Sharpless, K. B. J. Am. Chem. Soc. 1977, 99, 1990.
(c) Takai, K.; Oshima, K.; Nozaki, H. Tetrahedron Lett.
1980, 21, 1657.
Commun. 2003, 1948. (b) Malkov, A. V.; Mariani, A.;
MacDougall, K. N.; Kočovský, P. Org. Lett. 2004, 6, 2253.
(c) Malkov, A. V.; Stončius, S.; MacDougall, K. N.;
Mariani, A.; McGeoch, G. D.; Kočovský, P. Tetrahedron
2006, 62, 264.
(4) Sharpless, K. B. Chem. Tech. 1985, 692.
(5) (a) Murase, N.; Hoshino, Y.; Oishi, M.; Yamamoto, H. J.
Org. Chem. 1999, 64, 338. (b) Hoshino, Y.; Murase, N.;
Oishi, M.; Yamamoto, H. Bull. Chem. Soc. Jpn. 2000, 73,
1653. (c) Hoshino, Y.; Yamamoto, H. J. Am. Chem. Soc.
2000, 122, 10452. (d) Zhang, W.; Basak, A.; Kosugi, Y.;
Hoshino, Y.; Yamamoto, H. Angew. Chem. Int. Ed. 2005,
44, 4389.
(9) Epoxidation of alcohol 1 carried out in H2O (0 °C, 48 h)
using ligand 6a produced epoxide 2 in 65% yield and 51%
ee. For the reaction conditions, see ref. 6.
(10) Note that 3,5-dimethylphenyl group is commonly employed
in design of chiral selectors in chiral chromatography:
Pirkle, W. H.; Koscho, M. E. J. Chromatogr. A 1999, 840,
151.
(6) Bourhani, Z.; Malkov, A. V. Chem. Commun. 2005, 4592.
(7) (a) Bolm, C.; Kühn, T. Synlett 2000, 899. (b) Bolm, C.;
Beckmann, O.; Kühn, T.; Palazzi, C.; Adam, W.; Rao, P. B.;
Saha-Möller, C. R. Tetrahedron: Asymmetry 2001, 12,
2441. (c) Traber, B.; Jung, Y.-G.; Park, T. K.; Hong, J.-I.
Bull. Korean Chem. Soc. 2001, 22, 547. (d) Wu, H.-L.;
Uang, B.-J. Tetrahedron: Asymmetry 2002, 13, 2625.
(e) Adam, W.; Beck, A. K.; Pichota, A.; Saha-Möller, C. R.;
Seebach, D.; Vogl, N.; Zhang, R. Tetrahedron: Asymmetry
2003, 14, 1355. (f) Malkov, A. V.; Bourhani, Z.; Kočovský,
P. Org. Biomol. Chem. 2005, 3194.
(11) (a) Bryliakov, K. P.; Talsi, E. P.; Kühn, T.; Bolm, C. New J.
Chem. 2003, 27, 609. (b) Bryliakov, K. P.; Talsi, E. P. Kinet.
Catal. 2003, 44, 334. (c) Bryliakov, K. P.; Talsi, E. P.;
Stas’ko, S. N.; Kholdeeva, O. A.; Popov, S. A.; Tkachev, A.
V. J. Mol. Catal. A: Chem. 2003, 194, 79.
(12) The nature of the arene–arene interactions is not clear at the
moment. For recent reviews on the p–p and CH–p
interactions of aromatic systems, see: (a) Hobza, P.; Havlas,
Z. Chem. Rev. 2000, 100, 4253. (b) Hunter, C. A.; Lawson,
K. R.; Perkins, J.; Urch, C. J. J. Chem. Soc., Perkin Trans. 1
2001, 651. (c) Cozzi, F.; Annunziata, R.; Benaglia, M.;
Cinquini, M.; Raimondi, L.; Baldridge, K. K.; Siegel, J. S.
Org. Biomol. Chem. 2003, 1, 157. (d) Meyer, E. A.;
Castellano, R. K.; Diederich, F. Angew. Chem. Int. Ed. 2003,
42, 1210.
Synlett 2006, No. 20, 3525–3528 © Thieme Stuttgart · New York