Construction of a new asymmetric reaction site: Asymmetric 1,4-addition of thiol using pentagonal bipyramidal Hf(salen) complex as catalyst

Article abstract of DOI:10.1016/j.tetlet.2004.01.095

Pentagonal bipyramidal Hf(salen) complex 1 was found to serve as a catalyst for 1,4-addition reaction of thiol to N-(2-alkenoyl)-2-oxazolidinones.

Full text of DOI:10.1016/j.tetlet.2004.01.095

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 2385–2388
Construction of a new asymmetric reaction site: asymmetric
1
,4-addition of thiol using pentagonal bipyramidal
Hf(salen) complex as catalyst
Kazuhiro Matsumoto, Akira Watanabe, Tatsuya Uchida, Kayoko Ogi and
Tsutomu Katsuki*
Department of Chemistry, Faculty of Science, Graduate School, Kyushu University 33, CREST, Japan Science and Technology (JST),
Hakozaki, Higashi-ku, Fukuoka 812-8581, Japan
Received 12 December 2003; revised 13 January 2004; accepted 16 January 2004
Abstract—Pentagonal bipyramidal Hf(salen) complex 1 was found to serve as a catalyst for 1,4-addition reaction of thiol to N-(2-
alkenoyl)-2-oxazolidinones.
Ó 2004 Elsevier Ltd. All rights reserved.
Optically active quadridentate salen ligands have been
widely used as chiral auxiliaries for asymmetric synthe-
sis, and various types of salen ligands have been syn-
thesized. Complexes of these salen ligands usually take
octahedral configuration and the ligands adopt square
planar (trans) geometry, placing ancillary ligands (X and
Y) at their apical positions (Fig. 1), but Hf- and Zr-
(salen) complexes have been reported to adopt penta-
gonal bipyramidal configuration (A), in which an
asymmetric catalysts, the latter complexes have not
drawn the attention of synthetic chemists and have not
been used as catalyst. On the other hand, Hf- and Zr-
complexes have been reported to adopt cis-b octahedral
configuration with loss of the neutral ligand upon their
2
heating, and we have disclosed that a chiral Zr(salen)
1
complex is an efficient catalyst for asymmetric Baeyer–
Villiger (B–V) oxidation, in which the Zr complex is
considered to adopt a cis-b octahedral configuration
upon complexation of the Criegee intermediate, a
additional neutral ligand (S) such as THF is coordinated
2
to the central metal ion in its equatorial position. These
3
dianionic bidentate ligand, with the zirconium ion.
octahedral and pentagonal bipyramidal salen complexes
are expected to show different asymmetric catalysis.
Different from the former complexes widely used as
These results suggest that coordination of a dianionic
ligand accelerates dissociation of the apical ligand and
urges the complex to take cis-b configuration. This
X
X
N
O
N
O
N
O
N
O
M
Y
M
S
3
3'
3
Y
3'
octahedral
pentagonal bipyramidal (M= Zr or Hf, S= THF)
O
M
X
X
X
∆
T
N
O
N
X
D-L
N
N
O
N
N
O
N
N
O
N
O
O
M
M
or
M
N
O
O
O
-
S
D
L
S
Y
Y
L
D
O
O
O
O
cis-β
A
B
C
s-cis
s-trans
Figure 1.
Keywords: Hf(salen) complex; Asymmetric catalysis; Michael addition; Thiol; Baeyer–Villiger oxidation.
*
Corresponding author. Tel.: +81-92-642-2610; fax: +81-92-642-2607; e-mail: katsuscc@mbox.nc.kyushu-u.ac.jp
0
040-4039/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.01.095

2386
K. Matsumoto et al. / Tetrahedron Letters 45 (2004) 2385–2388
complicates the use of pentagonal bipyramidal Hf- and
Zr-(salen) complexes as catalyst. However, the penta-
gonal bipyramidal configuration was expected to be
retained during a reaction, if the bidentate substrate is
of neutral or monoanionic nature. Thus, we examined
alyzed the desired reaction smoothly, but complex 1
showed much better enantioselectivity than complex 2,
though the ee values were somewhat scattered (entries 1
and 2, Method A). During this study, molecular sieves
were found to catalyze the addition of thiol, albeit
slowly. Thus, we examined the reaction in the absence of
molecular sieves but the enantioselectivity was reduced
considerably (entry 3, Method B). Therefore, we carried
out the reaction using the substrates and the solvent
dried with molecular sieves immediately before use
(entry 4, Method C). Under the conditions, however, the
enantioselectivity was poorly reproducible and was
found to be dependent on the catalyst used. This
strongly indicated that the complex 1 including some
ingredient(s) suffered diminished enantioselectivity.
1
a neutral bidentate ligand, in the presence of a Hf(salen)
,4-addition of thiol to N-(2-alkenoyl)-2-oxazolidinone,
4;5
complex in order to expand metallosalen chemistry.
Although two isomeric chelates (B and C) can be formed
by coordination of a bidentate substrate (D–L), we
expected that one of them would be prepared preferen-
tially, if the Hf(salen) complex possesses a bulky chiral
0
substituent at C3 and C3 and replacement of the neutral
ligand is slowed (Fig. 1). N-(2-Alkenoyl)-2-oxazolid-
inone can exist as an equilibrium of s-cis and s-trans
conformers, but the former should be more stable than
the latter due to the steric reason. The sterically less
crowded oxazolidinone-carbonyl group was expected to
first coordinate with the hafnium ion at the apical site
and then make a chelate, positioning the alkenoyl group
at the space between the chiral substituents. In addition,
(R)
N
(S)
Y
Hf
Y
Y
Hf
Y
N
O
N
O
N
the hafnium ion of the Hf(salen)(OR) complex serves as
2
O
Ph Ph
O
5
a Lewis acid and the alkoxide as a mild base. Thus, the
complex was expected to catalyze the
Ph Ph
Hf(salen)(OR)
desired 1,4-addition under mild conditions.
2
(R)
(R)
2
"
Y= PhO
1
2
Addition of benzenethiol to N-crotonoyl-2-oxazolid-
inone was first examined in dichloromethane with
Hf(salen) complex 1 or 2 as the catalyst in the presence
ꢀ
of molecular sieves 4 A (Table 1). Both complexes cat-
All the complexes 1 were prepared from the corre-
sponding Hf(salen)Cl by its treatment with lithium
Table 1. 1,4-Addition of benzenethiol to N-crotonoyl-2-oxazolidinone
using Hf(salen) 1 or 2 as catalyst
2
phenoxide in tetrahydrofuran and purified by recrys-
tallization from hexane–dichloromethane. We submitted
all the complexes 1 used for the above experiments to
elementary analysis and found that the complex giving
the satisfactory analysis (Found: C, 71.58; H, 4.76; N,
Hf(salen) (5mol %), C6H5SH
O
N
O
N
CH2Cl2, rt, 24 h
O
O
C6H5S
O
O
2
.33. Calcd for C72
H
54Hf
1
N
2
O
4
2
ÆH O: C, 71.60; H, 4.67;
a
b
Entry
Catalyst Method
Yield (%)
% Ee
Confign
N, 2.32) showed the best enantioselectivity (92–93% ee,
entry 5). This enantioselectivity was reproducible, as
long as the complex giving the satisfactory analysis was
c
1
2
3
4
5
1
2
1
1
1
A
A
72–81
66–90
36
80–88
5–10
58
S
S
S
S
S
c
d
B
6;7
used as the catalyst. In general, as the analytical value
e
C
C
72–90
75–81
27–93
92–93
of the complex became less satisfactory, enantioselec-
tivity became lower. This indicated contamination of a
metal salt that promotes non-enantioselective 1,4-addi-
tion. From the synthetic procedure of 1, contamination
e;f
a
Determined by chiral HPLC using Daicel Chiralcel OD-H (hexane/
i-PrOH ¼ 8/2 v/v).
b
c
4d
Absolute configuration was determined by chiroptical comparison.
ꢀ
8
of some lithium salt(s) was most suspicious. Thus, we
Reaction was carried out in the presence of MS 4 A.
d
e
ꢀ
Reaction was carried out in the absence of MS 4 A.
Solutions of N-crotonoyl-2-oxazolidinone and benzenethiol in
ꢀ
dichloromethane were dried with MS 4 A immediately before use,
examined the 1,4-addition with the complex inducing an
inferior enantioselectivity, in the presence of 12-crown-4
ether that can trap the lithium ion, and found that the
addition of the crown ether improved enantioselectivity,
though the chemical yield of the 1,4-adduct was reduced
(Scheme 1).
respectively.
Reaction was carried out with complex 1 that gave the satisfactory
elementary analysis.
f
Hf(salen) (5mol %), C6H5SH
O
N
O
N
CH2Cl2, rt, 24 h
O
O
C6H5S
1% ee, 72%
4% ee, 63% (1 eq. of 12-crown-4)
O
O
3
6
Elementary analysis of the used Hf(salen) complex, found: C, 71.06; H, 4.76; N, 2.25.
Scheme 1.

K. Matsumoto et al. / Tetrahedron Letters 45 (2004) 2385–2388
2387
a
Table 2. Effect of the b-substituent (R) on enantioselectivity
O
1
, UHP
Ph
O
Ph
1
(5mol %), C6H5SH
O
O
R
N
O
AcOEt, rt
UHP= urea• H2O2 1(C, 71.58; H, 4.76; N, 2.33): 82% ee, 62%
(C, 71.37; H, 4.72; N, 2.26): 82% ee, 60%
R
N
CH2Cl2, rt, 24 h
O
O
C6H5S
O
O
1
b
Entry
R
Yield (%)
% Ee
Scheme 3.
1
2
3
4
5
Et
77
73
31
10
87
84
59
59
5 11
n-C H
3 7
i-C H
Ph
t-Bu
structure, in which two phenoxy ligands occupied its
axial positions and one water molecule was equatorially
coordinated. However, the salen ligand took an unusual
umbrella structure, and the five-membered chelate ring
between the hafnium ion and the cyclohexanediamine
unit adopted an envelope conformation: the hafnium
ion, two nitrogen atoms and one of the methine carbons
existed in the same plane and the other methine carbon
existed below the plane. The OH-p interaction between
the water molecule and its left-side naphthalene ring and
c
N.R.
a
b
c
Solutions of N-crotonoyl-2-oxazolidinone and benzenethiol in
dichloromethane were dried with MS before use, respectively.
Determined by chiral HPLC using Daicel Chiralcel OD-H (hexane/
i-PrOH ¼ 8/2 v/v).
No reaction occurred.
With the pure complex 1, we next examined the reac-
tions of other N-[(E)-2-alkenoyl]-2-oxazolidinones
(
00
the CH-p interaction between the 2 -phenyl group and
Table 2). When the b-substituent was a primary alkyl
the phenyl group in the axial ligand seems to be
responsible for this unique structure. The space above
the hafnium ion is sterically crowded and highly asym-
metric, but the space below the ion is much less crow-
ded. Thus, it was expected that the bottom axial ligand
would be replaced with a donor ligand, the carbonyl
group of the oxazolidinone unit (vide supra), delivering
the alkenyl moiety to the asymmetric but sterically
crowded space by chelate formation with the coordina-
tion of the carbonyl group of the alkenoyl moiety.
Although the replacement of the water ligand eliminates
the OH-p interaction, the CH-p interaction should
remain and the right-side face of the alkenyl group is
group, the enantioselectivity of the reaction was greater
than 80% ee (entries 1 and 2). However, as the b-sub-
stituent became bulkier, the reaction rate and the
enantioselectivity were reduced (entries 3 and 4), and no
reaction occurred, when the substituent was a tert-butyl
group (entry 5). These results suggested that the b-sub-
stituent was placed in a sterically crowded space.
The reactions with other thiols were also examined by
using 1 as the catalyst (Scheme 2). The addition of
p-bromobenzenethiol proceeded smoothly with slightly
diminished enantioselectivity, while the addition of
butanethiol did not occur.
00
blocked by the 2 -phenyl group. Thus, thiol attacks the
si-face of the b-carbon of the alkenoyl group to give the
Recrystallization of 1 from CH Cl –heptane gave a
2
2
(
S)-adduct. This model also explains why good sub-
single crystal. Its X-ray diffraction analysis determined
9
the structure of 1 unambiguously (Fig. 2). In analogy
strates are limited to N-(2-alkenoyl)-2-oxazolidinones
bearing a primary E-substituent.
2a
with the report by Floriani and co-workers, complex 1
adopted a slightly distorted pentagonal bipyramidal
Considering the similarity in nature of the Zr and Hf
2a
ions, complex 1 was expected to be an efficient catalyst
for B–V oxidation. Indeed, the oxidation of 3-phenyl-
cyclobutanone with 1 proceeded with high enantio-
selectivity (Scheme 3), though the corresponding
Zr(salen) complex showed slightly better enantioselec-
tivity (87% ee). This oxidation is also considered to
proceed via a Criegee intermediate cis-b 1 adduct (vide
supra). It is noteworthy that enantioselectivity of the
oxidation was not much affected by the purity of 1,
probably because the lithium ion does not promote B–V
oxidation.
1
(5mol %), RSH
O
O
N
N
CH2Cl2, rt, 24 h
O
O
RS
O
O
R= p-BrC6H4: 82% ee, 100%
R= n-Bu: no reaction
Scheme 2.
In conclusion, we were able to demonstrate that a pen-
tagonal bipyramidal metallosalen complex can be used
as an asymmetric catalyst.
References and notes
1
. For the reviews, see: (a) Yamada, S. Coord. Chem. Rev.
1999, 190–192, 537–555; (b) Calligaris, M.; Nardin, G.;
Randaccio, L. Coord. Chem. Rev. 1972, 7, 385–403.
Figure 2. X-ray structure of complex 1 and the proposed 1-N-(2-
alkenoyl)-2-oxazolidinone adduct.

2388
K. Matsumoto et al. / Tetrahedron Letters 45 (2004) 2385–2388
2
. (a) Corazza, F.; Solari, E.; Floriani, C.; Chiesi-Villa, A.;
Guastini, C. J. Chem. Soc., Dalton Trans. 1990, 1335–1344;
b) Repo, T.; Klinga, M.; Pietik €a inen, P.; Leskel €a , M.;
Uustitalo, A.-M.; Pakkanen, T.; Hakala, K.; Aaltonen, P.;
L o€ fgren, B. Macromolecules 1997, 30, 171–175.
2 2
6. The complex used was recrystallized from CH Cl –heptane
until it gave a satisfactory elementary analysis, and dried at
70 °C in vacuum.
(
2 2
7. Typical experimental procedure was as follows: a CH Cl
(1 mL) solution of N-crotonoyl-2-oxazolidinone (31.0 mg,
0.20 mmol) and a CH Cl (1 mL) solution of benzenethiol
3
4
. Watanabe, A.; Uchida, T.; Ito, K.; Katsuki, T. Tetrahedron
Lett. 2002, 43, 4481–4485.
. For catalytic enantioselective 1,4-addition of thiol to
electron-deficient olefins, see: (a) Helder, R.; Arends, R.;
Bolt, W.; Hiemstra, H.; Wynberg, H. Tetrahedron Lett.
2
2
ꢀ
(26.4 mg, 0.24 mmol) were dried with MS 4 A, respectively.
Both the solutions (0.5 mL each) were placed in a flask. To
the solution, was added complex 1 (5.9 mg, 5 mol %) and
the whole solution was stirred at room temperature for
24 h. The mixture was directly passed through a short silica
gel column (hexane/AcOEt ¼ 4/1 v/v) and subjected to silica
1977, 25, 2181–2182; (b) Suzuki, K.; Ikegawa, A.; Mukaiy-
ama, T. Bull. Chem. Soc. Jpn. 1982, 55, 3277–3282; (c)
Emori, E.; Arai, T.; Sasai, H.; Shibasaki, M. J. Am. Chem.
Soc. 1998, 120, 4043–4044; (d) Kanemasa, S.; Oderaotoshi,
Y.; Wada, E. J. Am. Chem. Soc. 1999, 121, 8675–8676; (e)
Saito, M.; Nakajima, M.; Hashimoto, S. Chem. Commun.
2 2
gel TLC (neat CH Cl ) to give the addition product
(21.5 mg, 81%). The enantiomeric excess of the product
was determined to be 92% as described in the footnote of
Table 1.
2
000, 1851–1852; (f) Kobayashi, S.; Ogawa, C.; Kawamura,
8. Although lithium chloride did not catalyze addition of
benzenethiol to N-crotonoyl-2-oxazolidinone, addition of
lithium chloride to the present reaction medium reduced the
enantioselectivity.
M.; Sugiura, M. Synlett 2001, 983–985; (g) Sundararajan,
G.; Prabagaran, N. Org. Lett. 2001, 3, 389–392; (h)
Nishimura, K.; Tomioka, K. J. Org. Chem. 2002, 67,
4
31–434.
9. Crystallographic data for complex 1 have been deposited
with the Cambridge Crystallographic Data Centre, UK as
supplementary publication number CCDC 222762.
5
. For hafnium-mediated asymmetric thiol 1,4-addition, see
Ref. 4f.