A new dinuclear chiral salen complexes for asymmetric ring opening and closing reactions: Synthesis of valuable chiral intermediates

Article abstract of DOI:10.1016/j.jorganchem.2005.12.044

A new dinuclear chiral Co(salen) complexes bearing group 13 metals have been synthesized and characterized. The easily prepared complexes exhibited very high catalytic reactivity and enantioselectivity for the asymmetric ring opening of epoxides with H₂O, chloride ions and carboxylic acids and consequently provide enantiomerically enriched terminal epoxides (>99% ee). It also catalyzes the asymmetric cyclization of ring opened product, to prepare optically pure terminal epoxides in one step. The homogeneous dinuclear chiral Co(salen) have been covalently immobilized on MCM-41. The potential benefits of heterogenization include facilitation of catalyst separation and recyclability requiring very simple techniques. The system described is very efficient.

Full text of DOI:10.1016/j.jorganchem.2005.12.044

Journal of Organometallic Chemistry 691 (2006) 1862–1872
www.elsevier.com/locate/jorganchem
A new dinuclear chiral salen complexes for asymmetric ring opening
and closing reactions: Synthesis of valuable chiral intermediates
a
a
a
a
b
Santosh Singh Thakur , Shu-Wei Chen , Wenji Li , Chang-Kyo Shin , Seong-Jin Kim ,
c
a,b,
*
Yoon-Mo Koo , Geon-Joong Kim
a
Department of Chemical Engineering, Inha University, 253 Yonghyun-dong, Nam-gu, Incheon 402-751, Republic of Korea
b
RStech Corp. #305 Venture Town, 1688-5, Sinil-dong, Daedeok-gu, Daejeon 306-230, Republic of Korea
Department of Biological Engineering, ERC for Advanced Bioseparation Technology, Inha University, Incheon 402-751, Republic of Korea
c
Received 26 July 2005; received in revised form 8 December 2005; accepted 13 December 2005
Available online 28 February 2006
Abstract
A new dinuclear chiral Co(salen) complexes bearing group 13 metals have been synthesized and characterized. The easily prepared
complexes exhibited very high catalytic reactivity and enantioselectivity for the asymmetric ring opening of epoxides with H2O, chloride
ions and carboxylic acids and consequently provide enantiomerically enriched terminal epoxides (>99% ee). It also catalyzes the asym-
metric cyclization of ring opened product, to prepare optically pure terminal epoxides in one step. The homogeneous dinuclear chiral
Co(salen) have been covalently immobilized on MCM-41. The potential benefits of heterogenization include facilitation of catalyst sep-
aration and recyclability requiring very simple techniques. The system described is very efficient.
ꢀ 2006 Elsevier B.V. All rights reserved.
Keywords: Asymmetric catalysis; Chiral dinuclear catalyst; Chiral intermediates; Asymmetric ring opening reactions; Kinetic resolution
1. Introduction
faces or groups within the reagent through predetermina-
tion of the reaction trajectory. Very recently Kobayashi
et al. [5] reported dinuclear chiral niobium complex for
Lewis acid catalyzed enantioselective Mannich-type reac-
tion of imines with silicon enolates. Zhu et al. [6] used
heterodinuclear Ti–Ga salen in the enantioselective ring
opening of meso-epoxides with aryl selenols. The stereose-
lective synthesis of chiral terminal epoxide is of immense
academic and industrial interest due to their utility as ver-
satile starting materials as well as chiral intermediates for
the preparation of bioactive molecules [7–10]. Hydrolytic
kinetic resolution (HKR) technology [11] is the very
prominent way to prepare optically pure terminal epox-
ides among available methods [12].
Design of simple and efficient chiral catalyst for asym-
metric reactions is one of the most important tasks in
organic synthesis [1]. Quite recently, several groups made
significant progress towards the design and use of chiral
homo- and heterodinuclear complex for asymmetric catal-
ysis, with particularly important contributions reported
by Shibasaki et al. [2], Trost et al. [3] and Belokon and
Kagan et al. [4]. Homo- and heterodinuclear complexes
have enormous potential to revolutionize asymmetric
catalysis. They can activate both components of bimolec-
ular reaction simultaneously, overcome entropy barriers
associated with bringing the two reagents together, mini-
mize the energy barrier that arises from solvent shell rear-
rangements during the reaction, and recognize prochiral
In the asymmetric ring opening of epoxides catalyzed by
chiral [(salen)Co] complexes [13] the reaction involved two
separate chiral salen-metal species in the rate limiting step
of the reaction and subsequently, very efficient oligomeric
catalysts were developed based on this principle [14].
*
Corresponding author. Tel.: +82 32 860 7472; fax: +82 32 872 0959.
E-mail address: kimgj@inha.ac.kr (G.-J. Kim).
0022-328X/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.12.044

S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
1863
However, multistep and complicated synthesis method of
2. Experimental
oligomeric catalysts [14] and solubility problem during
HKR make it less attractive strategy.
2.1. General
Pursuant to our own efforts directed towards the design-
ing of the di- and multimeric chiral (salen)Co catalysts [15]
herein we report the catalytic activity of new easily synthe-
sized homogeneous and heterogeneous dinuclear chiral
(salen)Co–MX3 catalyst (Scheme 1) for the enantioselective
kinetic resolution of terminal epoxides with H2O and other
nucleophiles and asymmetric cyclization reaction. The
dinuclear catalysts 1b–9b show remarkable enhanced reac-
tivity and may be employed substantially lower loadings
than its monomeric analogues 1a–9a without suffering sol-
ubility problem and their deactivation.
All 1H NMR, 13C NMR and 69Ga NMR (I = 3/2)
were recorded using 400 MHz FT NMR spectrophotom-
eter (VARIANUNITYNOVA400) at ambient tempera-
ture. Optical rotation measurements were conducted
using a Jasco DIP 370 digital polarimeter. Gas Chroma-
tography analyses were performed on Hewlett-Packard
5890 Series II instruments equipped with FID detectors
using chiral column (CHIRALDEX G-TA and A-TA,
20 m · 0.25 mm i.d. (Astec) and HP 3396 integrators
with HP Chem Station software for data analysis. Chiral
Monomers (a)
Dimers (b)
tBu
tBu
MX or
3
Precatalyst
tBu
t-Bu
N
O
N
N
O
tBu
N
O
II
MX nH O
3
II
O
N
Co
2
N
O
Co
Co
Co
But
O
tBu
tBu
But
tBu
M
1:1 equiv.
tBu
MX3
X3
O
N
tBu
X
N
O
t
tBu
Precatalyst
Bu
1a-9a
tBu
tBu
M
1b-9b
Cl, Br, I
1a-3b Al
4a-6b Ga
Cl, I, NO3
Cl, I
Cl
7a-8b In
9a-9b Tl
Covalentally heterogenized chiral salen ligand on MCM-41
N
N
HO
t-Bu
OH
O
O
MCM-41
Si
S
t-Bu
t-Bu
H3CO
A
1. Co(OAc)2.4H2O, EtOH, Reflux
2. Anhydrous MX3, CH2Cl2, 2h
Heterogeneous chiral (salen)Co MX3 monomers
N
N
Co
MX
O
t-Bu
O
O
O
MCM-41
Si
S
3 t-Bu
t-Bu
H3CO
1A-6A
M
X
Cl, Br
Precatalyst Chiral Co (salen)
CH2Cl2,1 h
1A-2B Al
3A-4B Ga
Cl, NO3
Cl
Cl
5A-5B In
6A-6B Tl
N
N
Co
O
t-Bu
O
O
O
MCM-41
Si
S
t-Bu
t-Bu
H3CO
MX3
t-Bu
t-Bu
t-Bu
O
N
O
t-Bu
Co
N
B
1B-6B
Heterogeneous chiral (salen)Co MX3 dinuclear complex
Scheme 1.

1864
S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
HPLC analyses were performed on YOUNGLIN instru-
ment using a Chiralcelꢁ OD column (24 cm · 0.46 cm
i.d.; Chiral Technologies, Inc.) and (R,R)-Whelk-O1/
(S,S)-Whelk-O1 column (24 cm · 0.46 cm i.d.) (Regis) at
254 nm. UV spectra were recorded on UV–Vis spectro-
photometer (Optizen 2120 UV) interfaced with PC using
Optizen view 3.1 software for data analysis. Vibrational
circular dichroism (VCD) and IR were measured in
Chiralirꢂ ABB Bomem Inc using Bomem GRAMS-32
software. X-ray absorption spectroscopy (EXAFS) were
measured by Rigaku Model R-XAS (Rigaku, Japan)
applying Co K-edge energy (7708.9 eV), Ga K-edge
energy (10,367.1 eV) radiation and data were simulated
by using FEFF program. Solvents were used after distil-
lation. TBME used as such obtained from Aldrich. All
other reagents obtained from Aldrich, Fluka and TCI.
The general procedure for the kinetic resolution was
same as shown in published papers [11,15a].
3. Results and discussion
In a representative example, the detailed characteriza-
tion by 69Ga (I = 3/2) NMR taking [Ga(D2O)6]3+ as an
external reference at 0 ppm shows the formation between
Ga and Co salen unit for all chiral(salen)Co–GaCl3 mono-
mer and dimers. The monomeric chiral (salen)Co 4a shows
chemical shift on d = 78.2 ppm, while bimetallic chiral
(salen)Co complex 4b on d = 52.4 ppm. The chemical shift
difference confirms the presence of the catalyst 4a and 4b as
two distinct species. X-ray fine structure (EXAFS) data
support the formation of Ga–O (4a) and Ga–O–Ga (4b)
˚
complex. For complex 4a Ga–O bond length is 1.83 A
and Ga–O coordination number is 0.75 (ꢀ1) and for 4b
˚
1.75 A and 1.50 (ꢀ2), respectively (see Supporting Infor-
mation). The different vibrational circular dichroism
(VCD) spectra for the catalyst 4a and 4b provide the con-
figuration of monomeric and dinuclear forms (Fig. 1) as a
distinct two species than that of precatalyst chiral Co(sa-
len). The spectra of dimeric form shows that it is not a
physical mixture of Co(II) ligand and monomeric form.
Monomeric catalyst 4a and dinuclear catalyst 4b exhibit
the new characteristic UV–Vis absorption band at
365 nm while band at 420 nm corresponding to the precat-
alyst Co(II) salen has disappeared. The FAB-mass spectra
of monomer 4a and dinuclear 4b (Figs. 2 and 3) provide
2.2. Syntheses of dinuclear catalyst
To
a
solution of hydrated gallium nitrate
Ga(NO3)3 Æ nH2O (2.118 g, 8.28 mmol, 1.0 equiv.) in tetra-
hydrofuran (25 mL), precatalyst (R,R)-salenCo (5.0 g,
8.28 mmol, 1.0 equiv.) was added and stirred in at open
atmosphere at room temperature. As soon as the chiral
(salen)Co was added color of the solution changes from
brick red to dark olive green. The solution was stirred at
r.t. for 1 h. The resulting solution was concentrated under
reduced pressure. The crude solid was worked up with H2O
and CH2Cl2. Yield = 98–99% as a dark green solid powder.
69Ga NMR (solvent CDCl3; d = 78.2 ppm) with reference
to [Ga(D2O)6]3+ for the catalyst 1a. UV–Vis shows sharp
absorbance on 375 nm. The characteristic absorption band
of precatalyst Co(II) salen at 420 nm disappeared. In addi-
tion, anhydrous gallium (III) chloride, bromide or iodide
(8.28 mmol, 1.0 equiv.) was added to a stirred solution of
precatalyst chiral salen Co(II) in water and THF and syn-
thesized in a similar manner, mentioned above. The crude
solid was worked up with H2O and CH2Cl2. Yield = 98–
99% as a dark green solid powder. In case of anhydrous
salt source, monomers of Co–MX3 were prepared using
methylene chloride (MC) as a solvent in place of THF
and bimetallic complexes were prepared in a similar
method except taking 2:1 equiv. of precatalyst chiral
(salen)Co.
The heterogenized chiral salen A was prepared by ear-
lier reported method [16]. Cobalt insertion into the
MCM-41 bound salen ligand was accomplished by adding
a equimolar solution of Co(OAc)2 Æ 4H2O in EtOH solvent
and refluxed for 1 h. After 1 h the dark red complex was
rinsed sequentially with EtOH and CH2Cl2 and dried in
vacuo to yields a heterogenized Co(salen) complex. Subse-
quently heterogenized Co(salen)–MX3 complex 1A–6A
and 1B–6B were prepared by mixing stoichiometric
amount of reacting partners in CH2Cl2 stirred for 1 h
(Scheme 1).
(R)-Co-Salen-AlCl3 Monomer
(S)-Co-Salen-AlCl3 Monomer
(R)-Co-Salen-AlCl3 Dimer
(S)-Co-Salen-AlCl3 Dimer
(R)-Co-Salen-Ga(NO3)3 Monomer
(S)-Co-Salen-Ga(NO3)3 Monomer
(R)-Co-Salen-Ga(NO3)3 Dimer
(S)-Co-Salen-Ga(NO3)3 Dimer
(R)-Co-Salen-InCl3 Monomer
(S)-Co-Salen-InCl3 Monomer
(R)-Co-Salen-InCl3 Dimer
(S)-Co-Salen-InCl3 Dimer
(R)-Co-Salen
(S)-Co-Salen
1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000
Wavenumber (cm-1)
Fig. 1. Comparison of VCD spectra of mono- and dinuclear chiral
catalysts.

S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
1865
Fig. 2. FAB-mass of Ga monomer catalyst 4a.
Fig. 3. FAB-mass of Ga dinuclear catalyst 4b.
direct evidence of formation of dinuclear complex. Dinu-
3.1. Syntheses of optically pure terminal epoxides via
hydrolytic kinetic resolution (HKR)
clear complex 4b showed molecular mass over 1242 and
monomer complex 4a over 643. It has already been estab-
lished that oxygen atoms of the metal complexes of the
Schiff bases are able to coordinate to the transition and
group 13 metals to form bi- and trinuclear complex [17].
On these bases possible structure of chiral (salen)Co–
MX3 complexes is depicted in Scheme 1.
The catalytic activities of the dinuclear catalyst 4b for
HKR of the diverse and valuable racemic terminal epox-
ides are shown in Table 1.
Due to high selectivity factor (krel values) every epoxide
underwent resolutionin excellentyieldandhighee % employ-

1866
S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
Table 1
HKR of terminal epoxides catalyzed by the dinuclear catalyst
(R,R)- Cat.1b-9b
0.2-0.8 mol%
rt,2-7 h
OH
O
O
OH
ee%,42-50 % y
85
+
H2O
+
y
R
R
ee%
_
+
( )
R
2.0
equiv.
1.11
,
40-49 %
> 99
>
equiv.
Entry
1
Recovered epoxidesa
Catalyst
Catalyst loading (mol%)b
0.2
Time (h)
%Yield (ee)c
43 (99.3)
krel
O
4b
2
480
O
2
4b
4b
4b
4b
4b
4b
4b
0.2
0.5
0.2
0.4
0.4
0.5
0.8
3
7
3
3
4
2
6
45 (99.7)
43 (98.7)
45 (99.8)
49 (99.6)
40 (99.3)
42 (99.3)
40 (98.2)
510
110
205
167
173
145
135
3d
O
HO
4
O
Cl
O
5
H3CO
6
O
O
7
O
O
8d
O
O
9
4b
4b
4b
4b
4b
0.5
0.5
0.5
0.5
0.5
3
4
6
2
2
44 (99.4)
43 (99.8)
43 (96.8)
42 (99.3)
42 (99.3)
151
92
O
O
O
10
O
O
11e
12e
290
120
125
O
O
O
O
H3
C
13e
O
O
Cl
a
Isolated yield is based on racemic epoxides (theoretical maximum = 50%).
Loading on a per [Co] basis w.r.t. racemic epoxides.
ee % was determined by chiral GC or chiral HPLC.
THF was used as a solvent.
b
c
d
e
Solvents CH2Cl2:THF = 2:1. The selectivity factor (krel) was calculated using the equation = ln[1 ꢁ c(1 ꢁ ee)]/ln [1 ꢁ c(1 + ee)] where the conversion c
was set to equal the isolated yield of the recovered epoxide.
ing 0.2–0.8 mol% of catalyst in solvent free condition in most
of the cases. In a similar condition, catalyst 4a loading per
[Co] basis gives <50% ee and <30% isolated products (see
SupportingInformation).Theorderofreactivityfordinuclear
complexeswerefoundtobeCo–In>Co–Tl> Co–Ga> Co–Al
andforthecounterionsI > Cl > Br > NO3 (seeFigs. 4and5).

S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
1867
100
80
60
40
20
0
1
1 Precatalyst chiral (salen) Co
2 Catalyst 4b
3 Catalyst during reaction at t=0
4 Catalyst during reaction at t=1h
5 Catalyst during reaction at t=24h
6 Catalyst after 1'st cycle
1.0
Co-In dimer 7b
Co-Tl dimer 9b
Co-In monomer 7a
Co-Tl monomer 9a
Co-Ga dimer 4b
Co-Al dimer 1b
Co-Ga monomer 4a
Co-Al monomer 1a
2
3
4
0.5
0.0
5
6
0
2
4
300
400
500
600
700
Time (h)
Wave Length (in nm)
Fig. 4. Comparison of reactivity and enantioselectivity of catalysts Co–
MX3 for the HKR reaction of epichlorohydrin at r.t. using 0.2 mol % of
the catalyst per [Co] basis w.r.t. substrate.
Fig. 6. The UV–Vis spectral analysis of the catalyst during and after the
HKR of racemate methyl glycidyl ether at r.t.
100
90
Co-In dimer 7B
Co-In monomer 7A
Co-Ga dimer 4B
Co-Al dimer 1B
80
80
70
Co-Ga monomer 4A
Co-Al monomer 1A
60
60
50
40
30
20
10
0
Co-In dimer 7b
Co-In monomer 7a
Co-Tl dimer 9b
Co-Tl monomer 9a
Co-Ga dimer 4b
Co-Ga monomer 4a
Co-Al dimer 1b
40
20
0
Co-Al monomer 1a
0
2
4
6
8
5.0x102
1.0x103
1.5x103
2.0x103
2.5x103
Time (h)
Substrate/Catalyst mole ratio
Fig. 7. Comparison of reactivity and enantioselectivity of heterogeneous
catalysts Co–MX3/MCM-41 for the HKR reaction of epichlorohydrin at
r.t. using 0.2 mol% of the catalyst per [Co] basis w.r.t. substrate.
Fig. 5. The effect of substrate/catalyst mole ratio on the enantiosectivity
of the HKR of epichlorohydrin at r.t. in 1 h.
treated with
Work up
with CH2Cl2
Shows same reactivity
and selectivity
Catalysts 1b-9b
100
excess of H2O
90
80
70
60
50
1 st cycle
2 nd cycle
3 rd cycle
4 th cycle
5 th cycle
Scheme 2. Stability of the catalyst under control experiment.
In a control experiment, the dinuclear complexes trea-
ted with excess of water and worked up are found to have
the similar reactivity and selectivity (Scheme 2). The cat-
alyst prepared from hydrated MX3 Æ nH2O exhibited sim-
ilar activity and enantioselectivity for the present study.
EXAFS data of the water treated Co–Ga dimer reveal
that there is no dissociation of Co–Ga complex during
HKR reaction (see supporting information). However,
catalyst could be generated in situ by suspension of the
precatalyst in epoxide or epoxide/solvent and addition
of MX3. The continuous monitoring of UV–Vis spectra
of dinuclear catalyst 4b during the HKR reaction of
methyl glycidyl ether (Fig. 6) revealed that there is no
10
0
0
2
4
6
8
10
12
14
16
18
20
22
24
Time (h)
Fig. 8. Recyclability of the heterogeneous catalyst 3B in the asymmetric
HKR of ( ) ECH using 0.2 mol% catalyst at r.t.

1868
S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
Table 2
Kinetic data for the HKR of racemic ECH catalyzed by monomers and
dimers
Catalyst
No. of (salen)Co unit
kintra
(minꢁ1 · 10ꢁ2 a
kinter
(Mꢁ1 minꢁ1 a
)
)
1a
1b
4a
4b
7a
7b
9a
9b
a
1
2
1
2
1
2
1
2
1.0
10.2
5.07
11.4
15.2
22.0
12.1
21.0
44.4
47.8
49.8
66.0
49.3
61.3
Calculated from Fig. 10 using Eq. (1).
Vibrational circular dichroism (VCD) spectroscopy has
been applied to elucidate the stereochemistries of chiral
molecules, including the accurate estimation of enantio-
meric excess and their absolute configurations [19]. In the
present study VCD spectra was used to elucidate the ste-
reochemistries of chiral mono- and dinuclear catalyst and
optically pure chiral intermediates (Fig. 9), including the
accurate estimation of enantiomeric excess and their abso-
lute configurations. Optically pure samples as well as race-
mic samples were used as a reference to compare the VCD
spectra. The VCD spectra of opposite configuration, such
as R and S exhibited the reverse absorption peaks. The the-
oretical VCD spectra of optically pure epichlorohydrin and
styrene oxide were calculated using DFT/6-31G(d)/B3LYP
program. The observed VCD and IR spectra were obtained
for 1.0 M solution of (R)-(ꢁ) epichlorohydrin and (R)-(+)-
styrene oxide each in CDCl3 using Chiral IR FT VCD ABB
Bomem (Bio Tools) spectrometer, with a path length 94 lm
set at 8 cmꢁ1 resolution and a spectral collection time of
1 h. The comparison of observed and calculated spectra
shows very high level fidelity and clearly establish the abso-
lute configuration of these two chiral intermediates. The
Fig. 9. The calculated and observed VCD and IR spectra of optical
isomers of epichlorohydrin (A) and styrene oxide (B) obtained after the
HKR reaction.
dissociation and equilibrium of 4b to their monomeric
form. The dinuclear catalyst is very stable even after
HKR reaction.
R2=0.977 to 0.999
1.8
1a
5a
1b
1.6
The reactivity and enantioselectivity of hydrolytic
kinetic resolution of epichlorohydrin by covalently immo-
bilized chiral Co(salen) MX3 monomer and dimer over
MCM-41 is outlined in Fig. 7. The heterogenized catalysts
1A–6A and 1B–6B activity is found to be similar to the
homogeneous catalyst 1a–9a and 1b–9b, except taking
longer reaction time. The recylability of the heterogenized
catalyst is shown in Fig. 8.
The supported catalysts 1B–6B were removed by simple
filtration and repeated for recycling with no loss of reactiv-
ity and enantioselectivity up to three cycle. The potential
benefits of heterogenization include facilitation of catalyst
separation from reagents and reaction products, simplifica-
tion of methods for catalyst recycle, and possible adapta-
tion of immobilized catalyst to continuous-flow process
[18].
5b
1.4
9a
7a
9b
7b
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.01
0.02
0.03
0.04
0.05
Catalyst [M]
Fig. 10. Initial rate kinetics for the asymmetric HKR of the ECH
catalyzed by the monomer and dimmer catalysts.

S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
1869
shift in peak position between observed and calculated
VCD spectra might be due to solute–solvent interactions
[20]. The ee% determined by the VCD and those measured
by GC were found to be good agreement within 2% ee.
Kinetic studies of the HKR of epichlorohydrin (ECH)
were carried out studied as a model reaction (Table 2).
Considering the two-term rate equation involving both
intra- and intermolecular components (Eq. (1)) [15a,21,22].
of Co–In 7a and Co–Tl 9a show the intramolecular path-
way which is quite obvious from kinetic analysis. It seems,
Co–In 7a and Co–Tl 9a act as heterometallic complexes
exhibiting two different Lewis acid centers Co and In
[23a,23b] and Co and Tl [23c] and show strong synergistic
effect. However, dinuclear complex of Co–In 7b and Co–Tl
9b show two fold more reactive than corresponding mono-
meric analogy 7a and 9a. The central metal atom Co
appears to activate and control the orientation of epoxide
and stereoselectively bind only one enantiomer and the lat-
ter seems to activate and control the orientation of nucle-
ophlie H2O by enabling an enantioselective ring opening
of epoxides with nucleophiles. The proposed model for
HKR catalyzed by Co–In and Co–Tl complex is given in
Scheme 3 which may be a similar to enantioselective ring
opening of epoxides with 4-methoxyphenol catalyzed by
gallium heterobimetallic complexes as reported by Shiba-
saki et al. [2c]. The intra- and inter-molecular mechanism
for the 7a–9b may be a similar to earlier report [15a,21,22].
2
rate / kintra½catalystꢂ þ kinter½catalystꢂ
ð1Þ
Plots of rate/[catalyst] vs. [catalyst] should be linear with
slopes equal to kinter and y-intercepts corresponding to
kintra
.
Analysis of such plots with rate data obtained with dinu-
clear catalysts 1b–9b and monomeric catalysts 7a–9a
depicted linear correlations with positive slopes and non
zero y-intercepts, consistent with participation of both
inter- and intramolecular pathway for the HKR
(Fig. 10). Similar analysis of rate data obtained with mono-
meric catalysts 1a and 4a revealed y-intercepts of zero,
reflecting the absence of any first-order pathway for these
complexes. With dinuclear catalyst 7b, a maximum value
for intra- and intermolecular rate constants was obtained
showing highest reactivity and enantioselectivity. Thus,
the dinuclear catalyst provides appropriate relative proxim-
ity and orientation, which eventually reinforces the reactiv-
ity and selectivity relative to monomeric complex.
The HKR reactions follow the cooperative bimetallic
catalysis where epoxide and nucleophile activate simulta-
neously by two different (salen)Co–Ga catalyst molecules.
The active intermediate during HKR may be a CoIII–OH
complex [8a]. The linking of two (salen)Co unit through
the Al and Ga induces the cooperative mechanism, albeit
through a far less enantio-discriminating transition state
than that attained with the catalyst 1a and 4a. The reaction
mechanism for 4b is similar to 1b [15a]. Although the HKR
reaction is easily carried out with Co(II) complexes, but it
appears that the reactive species is in fact Co(III). Unlikely
to dinuclear complex of Co–Al and Co–Ga, the monomers
3.2. Kinetic resolution with other nucleophiles and
asymmetric ring closing reaction
3.2.1. Kinetic resolution with hydrochloride
Among the myriad of nucleophiles that have been
employed in epoxide ring openings, halide ions (which
afford the corresponding halohydrins) have been received
considerable attention [24,25]. Methods for the asymmetric
synthesis of chlorohydrins by enantioselective ring opening
of epoxides have relied upon the use of stoichiometric
amounts of chiral Lewis acid halides [26]. Garrett et al.
[27a], Denmark et al. [27b] and quite recently Wang et al.
[27c] disclosed the enantioselective ring opening of epox-
ides with TMSCl and SiCl4 to afford optically active chlo-
rohydrins. To our knowledge, we have observed first time
enantioselective ring opening of terminal epoxides with
HCl [15a]. We report herein one step synthesis of optically
pure chlorohydrins using new chiral catalyst 4b–9b dimer
via kinetic resolution of terminal epoxides with HCl at 0–
4 ꢃC in tert-butyl methyl ether (TBME) solvent. Before
starting the reaction the catalyst 4b–9b were treated
O
OH
N
N
O
Co
Cl
HO
R
_
+
(
)
O
R
(R,R)-Cat.4b-9b
M
OH
O
Cl
Cl
2-5 mol%
HCl
Cl
+
o
_
+
TBME
0-4 C,4-6 h
( )
R
R
N
O
N
N
N
O
2.22
1.00
Co
Cl
Co
,
equiv.
equiv.
ee%
19- 89
84-94 yield %
O
O
Cl
Cl
M
O
OH
Scheme 4. Asymmetric ring opening of terminal epoxides with HCl
H O
2
M
Cl
H
N
N
O
Cl
Cl
O
catalyzed by 4b and 9b (yield based on HCl).
R
Co
O
Cl
R
M
O
OH
OH
Cl
O
(R,R)-Cat.4b 2 mol%
(R,R)-Cat.4b 2 mol%
Cl
H
+ HCl
M=In , Tl
X =Cl, I
Cl
0-4 oC
1,4-dioxane
H
Cl
,
TBME ,
KOH
O
_
+
(
)
,
88 ee% 89% Yield
R
> 98 % ee 80% Yield
Scheme 3. Possible working model for the HKR of terminal epoxides
catalyzed by Co–In and Co–Tl heterometallic complex.
Scheme 5. Kinetic resolution/cyclization sequence in the presence of
catalyst 4b.

1870
S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
OH
OH
OH
OH
OH
O
Cl
O
Cl
O
Cl
Cl
Cl
89 % ee
93% yield
krel = 210
89 % ee
90 % yield
krel = 150
89 % ee
90 % yield
krel = 180
87% ee
89% yield
krel = 89
88 % ee
89 % yield
k
rel=103
OH
OH
OH
OH
Cl
O
Cl
O
Cl
O
Cl
O
O
O
O
85% ee
94% yield
krel = 81
84 % ee
93 % yield
krel = 78
80 % ee
92% yield
krel = 73
85 % ee
87 % yield
krel = 83
OH
OH
Cl
OH
OH
Cl
O
Cl
O
Cl
O
62 % ee
87 % yield
krel = 66
23 % ee
86 % yield
krel = 51
44 % ee
93 % yield
krel = 57
25 % ee
91 % yield
krel = 54
CH3
Cl
Fig. 11. Kinetic resolution products obtained using HCl and catalyst Co–In. Conditions are shown in Scheme 2. Yields correspond to the isolated product
based on HCl 5. The selectivity factor (krel) was calculated using the equation = ln [1 ꢁ c(1 + ee)]/ln[1 ꢁ c(1 ꢁ ee)] where the conversion c was set to equal
the isolated yield of the ring opened products.
in situ with HCl by taking 2:1 equiv. of HCl:4b in TBME
for 1 h at 0–4 ꢃC. The optimum condition for the kinetic
resolution of terminal epoxides with HCl is illustrated in
Scheme 4. TBME demonstrated the highest reactivity
amongst various solvents like CH2Cl2, CH3CN, THF and
n-hexane at 0–4 ꢃC. The reaction time and catalyst loading
amount depend upon the individual terminal epoxides.
In a representative example, rac-propylene oxide and
HCl were catalyzed by 2 mol% of 4b with major regioselec-
tivity to afford 1-chloro-propan-2-ol in 88% ee within 1 h at
0–4 ꢃC in TBME solvent. The ring opened 1-chloro-pro-
pan-2-ol in the presence of equimolar base and 2 mol%
of catalyst 4b (with respect to substrate) minor enantiomer
selectively ring closed to provide optically pure propylene
oxide along with higher optically pure 1-chloro-propan-2-
ol (Scheme 5) in 1,4-dioxane solvent at room temperature
within 4 h. Moreover, HKR technology is the direct way
to prepare enantiomerically enriched terminal epoxides
but this method has benefit in terms of easy separation.
Terminal epoxides containing halide, ether, ester and
aliphatic substituents underwent resolution to 89% ee of
the corresponding products with good yields (Fig. 11).
However, phenyl glycidyl ether and their 3-chloro, 3-
methyl derivatives, and benzyl glycidyl ether proved to be
problematic substrates and resolved with relatively low
ee%, even after high catalyst loading (3–10 mol%) and pro-
longed time (6 h). It appears that aromatic group might be
plagued by conflicting steric and electronic factors influenc-
ing regioselectivity in the epoxide ring opening with HCl.
In all present studied epoxides afforded with major regiose-
lectivity 2-ol chlorohydrins and very few amount (<2%)
regioisomer was experienced.
Tl > Co–Ga and for the counter ion Cl > I > NO3. The
enantioselective ring opening of epoxides with HCl cata-
lyzed by Co–Ga, Co–In and Co–Tl complex may occur
similar to Scheme 3 where Ga, In and Tl also play an
important role to activate the attacking reagent regio-
and stereoselectively.
3.2.2. Catalytic asymmetric ring closing reactions
In an important and interesting reaction the optically
pure epichlorohydrin and glycidol were obtained in good
ee with over 70% ee and 43% yield, respectively, catalyzed
by the catalyst 1b–4b in the presence of equivalent amount
base with respect to substrate (Scheme 6). The asymmetric
cyclization of 1,3-dichloro-2-propanol involves asymmetric
elimination of hydrogen chloride with base potassium
phosphate. The asymmetric cyclization of 1,3-dichloro-2-
propanol could proceed significantly only in the presence
of the catalyst.
OH
(R,R)-Cat. 1b-4b
5 mol%
O
Cl
OH
HO
K3PO4
,THF, rt, 7 h
(R)-
43 % Yield
_
+
( )
,
> 70 % ee
OH
(R,R)-Cat. 1b-4b
5 mol%
O
Cl
Cl
Cl
rt,
K3PO4,THF,
3 h
(S)-
Prochiral
,
70 % Yield
> 75 % ee
OH
O
(R,R)-Cat.4b 2 mol%
Cl
,
KOH
1,4-dioxane
_
+
(R)-
(
)
The reactivity and selectivity for kinetic resolution of
terminal epoxides with HCl catalyzed by monomers and
dimers of Co–Ga, Co–In, Co–Tl have not shown great dif-
ferences, however, the order of reactivity is Co–In > Co–
>78% ee,
41% Yield
Scheme 6. Asymmetric cyclization of chlorohydrin catalyzed by dinuclear
complex 1b–4b.

S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
1871
Table 3
ing route to 1,2-diol monoesters [29]. In an important and
model reaction, butyric acid and ( ) epichlorohydrin was
catalyzed by 4b with complete regioselectivity along with
75% ee and in a 43% yield (45% theoretical yield). The
resolved ring opened product followed by ring closing in
the presence of base and catalyst afforded glycidyl butyrate
in high 98% ee and quantitative yield (see Scheme 7).
(R)-glycidyl butyrate is very important compound and
has been used to introduce a stereogenic center in the syn-
thesis of Linezolid [30] which is currently marketed for the
treatment of multidrug resistant Gram-positive infections
such as nosocomial, community-acquired pneumonia,
and skin infections.
Asymmetric ring opening of terminal epoxides with carboxylic acids
catalyzed by 4b
OH
O
R'
O
(R,R)-Cat.4b, 2 mol%
rt
R'COOH
+
R
1,4-dioxane/TBME
_
+
( )
R
1.00
equiv.
O
50-75
2.22
equiv.
,
ee % 41-43 yield %
R'=CH3,C2H5 , n-C3H7, propiolic acid
CH Cl
2
R=
EntryA
1
R
R0
Catalyst Catalyst Time Yield ee
loadinga (h)
(%)b
(%)c
O
CH2Cl
4b
4b
2.0
2.0
3.0
4.0
42
61
OH
O
CH2Cl
2
41
63
OH
O
3
4
A
4b
4b
3.0
3.0
3.0
5.0
43
41
76
75
CH2Cl
CH2Cl
4. Conclusion
OH
O
In summary, we have synthesized homogeneous and
heterogeneous chiral dinuclear complex and demonstrated
their catalytic activity and selectivity for the asymmetric
ring opening (ARO) of terminal epoxides with variety of
nucleophiles and for asymmetric cyclization to prepare
optically pure terminal epoxides in one step. The ARO
reactions are summarized in Scheme 8. Yields and enanti-
oselectivities are good to excellent. The resolved ring
opened product combined with ring closing in the presence
of base and catalyst afforded terminal epoxides in high ee
and quantitative yield. The catalyst can be easily synthe-
sized by readily commercially available precatalyst Co(sa-
len) in both enantiomeric forms. Potentially, the catalyst
may be used on an industrial scale. Further studies con-
cerning the mechanism and scope of binuclear catalysts,
OH
For entry 1–2, 1,4-dioxane and 3–4 TBME was taken as solvent.
In mol% loading on a per [Co] basis w.r.t. racemic epoxide.
Isolated yield is based on racemic epoxides (theoretical
a
b
maximum = 45%).
c
ee% was determined by chiral GC or chiral HPLC.
O
O
OH
Cl
n-Pr
O
Cl
+
_
(
n-Pr
OH
(R)-
76 % ee
(R,R)Cat. 4b,TBME, rt
O
O
(R,R) Cat.4b
rt
2 mol %
K2CO3 ,CH2Cl2
n-Pr
O
(S) -> 98% ee
Glycidyl butyrate
O
HO
>99% ee
Cl
O
HO
Scheme 7.
R
O
OH
+
OH
>99% ee
R
R
86% ee
In the absence of catalyst reaction underwent slowly and
formed racemic mixture. Similarly optically active glycidol
was synthesized from asymmetric cyclization of 3-chloro-
propane-1,2-diol via kinetic resolution. The result shows
that (S)-3-chloro-propane-1,2-diol is preferentially cyclized
to give (S)-glycidol. The mechanism of asymmetric cycliza-
tion may follow the similar mechanism reported by Takei-
chi et al. [28].
O
O
_
O
>93% ee
R
( +
)
88% ee
R
R
+
R
HO
R
OH
O
Cl
R
O
O
R
OH
Cl
R'
90% ee
O
O
3.2.3. Asymmetric ring opening of terminal epoxides with
carboxylic acids
+
O
>95% ee
R'
O
R
The moderate to high enantioselectivity displayed by
these catalysts for asymmetric ring opening of terminal
epoxides with H2O and HCl led us to extend the possible
use of other nucleophiles. Carboxylic acids are interesting
candidates (Table 3) because of their low cost, ease of han-
dling and reaction of epoxides provides a direct and appeal-
Scheme 8. Coupled route for the synthesis of chiral intermediates
catalyzed by dinuclear salen complex.

1872
S.S. Thakur et al. / Journal of Organometallic Chemistry 691 (2006) 1862–1872
(d) T. Katsuki, V.S. Martin, Org. React. 48 (1996) 1.
[13] J.M. Ready, E.N. Jacobsen, J. Am. Chem. Soc. 123 (2001) 2687.
[14] (a) J.M. Ready, E.N. Jacobsen, Angew. Chem. Int. Ed. 41 (2002)
1374;
for asymmetric catalytic reactions are currently under
investigation for a broad applicability as a general catalyst.
Appendix A. Supplementary data
(b) D.E. White, E.N. Jacobsen, Tetrahedron: Asymmetry 14 (2003)
3633.
[15] (a) S.S. Thakur, W. Li, S.J. Kim, G.-J. Kim, Tetrahedron Lett. 46
(2005) 2263;
Supplementary data associated with this article can
be found, in the online version, at doi:10.1016/
j.jorganchem.2005.12.044.
(b) C.K. Shin, S.J. Kim, G.-J. Kim, Tetrahedron Lett. 45 (2004)
7429;
(c) G.-J. Kim, H. Lee, S.J. Kim, Tetrahedron Lett. 44 (2003) 5005;
(d) S.S. Thakur, W. Li, C.K. Shin, G.-J. Kim, Catal. Lett. 104 (2005)
151.
References
[16] D.W. Park, S.D. Choi, C.Y. Lee, G.-J. Kim, Catal. Lett. 78 (2002)
145.
[17] (a) D. Atwood, M.J. Harvey, Chem. Rev. 101 (2001) 37, and
references cited therein;
[1] (a) I. Ojima (Ed.), Catalytic Asymmetric Synthesis, second ed., Wiley-
VCH, New York, 2000;
(b) E.N. Jacobsen, A. Pfaltz, H. Yamamoto (Eds.), Comprehensive
Asymmetric Catalysis, vols. I–III, Springer, Heidelberg, 1999.
[2] (a) M. Shibasaki, M. Kanai, K. Funabashi, Chem. Commun. (2002)
1989;
(b) D. Atwood, P. Wei, Chem. Commun. (1997) 1427;
(c) E. Solari, F. Corazza, C. Floriani, A. Chiesi-Villa, C. Guastini, J.
Chem. Soc., Dalton Trans. (1990) 1345;
(b) M. Shibasaki, N. Yoshikawa, Chem. Rev. 102 (2002) 2187;
(c) T. Iida, N. Yamamoto, S. Matsunaga, H.-G Woo, M. Shibasaki,
Angew. Chem. Int. Ed. 37 (1998) 2223;
(d) S.J. Gruber, C.M. Harris, E. Sinn, Inorg. Chem. 2 (1968) 268;
(e) R.H. Holm, W. Everett, A. Chakravorty, Prog. Inorg. Chem. 7
(1966) 83;
(d) S. Matsunaga, T. Yoshida, H. Morimoto, N. Kumagai, M.
Shibasaki, J. Am. Chem. Soc. 126 (2004) 8777.
(f) H. Aoi, M. Ishimori, T. Tsuruta, Bull. Chem. Soc. Jpn. 48 (1975)
1897.
[3] (a) B.M. Trost, V.S.C. Yeh, Angew. Chem. 114 (2002) 889;
Angew. Chem. Int. Ed. 41 (2002) 861, and references therein;
(b) B.M. Trost, T. Mino, J. Am. Chem. Soc. 125 (2003) 2410;
(c) B.M. Trost, L.R. Terrell, J. Am. Chem. Soc. 125 (2003) 338.
[4] Yu.N. Belokon, M. North, V.I. Maleev, N.V. Voskoboev, M.A.
Moskalenko, A.S. Peregudov, A.V. Dmitriev, N.S. Ikonnikov, H.B.
Kagan, Angew. Chem. Int. Ed. 43 (2004) 4085.
[18] D.A. Annis, E.N. Jacobsen, J. Am. Chem. Soc. 121 (1999) 4147.
[19] K.M. Spencer, S.J. Cianciosi, J.E. Baldwin, T.B. Freedman, L.A.
Nafie, Appl. Spectrosc. 44 (1990) 235.
[20] For review see: T.B. Freedman, X. Cao, R.K. Dukor, L.A. Nafie,
Chirality 15 (2003) 743.
[21] R.G. Konsler, J. Karl, E.N. Jacobsen, J. Am. Chem. Soc. 120 (1998)
10780.
[5] S. Kobayashi, K. Arai, H. Shimizu, Y. Ihori, H. Ishitani, Y.
Yamashita, Angew. Chem. Int. Ed. 44 (2005) 761.
[6] M. Yang, C. Zhu, F. Yuan, Y. Huang, Y. Pan, Org. Lett. 7 (2005)
1927.
[22] R. Breinbauer, E.N. Jacobsen, Angew. Chem. Int. Ed. 39 (2000)
3604.
[23] (a) For examples using InCl3 and TlCl3 Æ 4H2O as Lewis acid, see:
B.C. Ranu, Eur. J. Org. Chem. (2000) 2347;
[7] H.C. Kolb, M.S. Van Nieuwenhze, K.B. Sharpless, Chem. Rev. 94
(1994) 2483.
[8] (a) L.P.C. Nielson, C.P. Stevenson, D.G. Backmond, E.N. Jacobsen,
J. Am. Chem. Soc. 126 (2004) 1360;
(b) S. Sengupta, S. Mondal, Tetrahedron Lett. 40 (1999) 8685;
(c) S. Uemura, K. Sohma, M. Okano, Bull. Chem. Soc. Jpn. 45
(1972) 860.
[24] C. Bonini, G. Righi, Synthesis (1994) 225.
[25] Halohydrins also serve as important synthetic intermediates in their
own right and are critical subunits for the synthesis of halogenated
marine natural products. For reviews see: W. Fenical, in: P.J. Scheuer
(Ed.), Marine Natural Products, vol. 2, Academic Press, New York,
1980, p. 174.
(b) S.K. Kim, E.N. Jacobsen, Angew. Chem. Int. Ed. 43 (2004) 3952;
(c) J.F. Larrow, E.N. Jacobsen, Top. Organomet. Chem. 6 (2004)
123, and references cited therein.
[9] S. Aertsa, A. Buekenhoudta, H. Weytena, I.F.J. Vankelecom, P.A.
Jacobs, Tetrahedron: Asymmetry 16 (2005) 657.
[10] (a) L. Aouni, K.E. Hemberger, S. Jasmin, H. Kabir, J.F. Larrow, I.
Le-Fur, P. Morel, T. Schlama, in: H.U. Blaser, E. Schidmt (Eds.),
Asymmetric Catalysis on Industrial Scale: Challenges, Approaches
and Solutions, Wiley-VCH Verlag Gmbh & Co., KgaA Weinheim,
Germany, 2004, p. 165;
(b) J.F. Larrow, K.E. Hemberger, S. Jasmin, H. Kabir, P. Morel,
Tetrahedron: Asymmetry 14 (2003) 3589.
[11] (a) M. Tokunaga, J.F. Larrow, F. Kakiuchi, E.N. Jacobsen, Science
277 (1997) 936;
[26] (a) N.N. Joshi, M. Srebnik, H.C. Brown, J. Am. Chem. Soc. 110
(1988) 6246;
(b) Y. Naruse, T. Esaki, H. Yamamoto, Tetrahedron 44 (1988)
4747.
[27] (a) C.E. Garrett, G.C. Fu, J. Org. Chem. 62 (1997) 4534;
(b) S.E. Denmark, P.A. Barsanti, J. Org. Chem. 63 (1988) 2428;
(c) L.S. Wang, T.K. Hollis, Org. Lett. 5 (2003) 2543.
[28] T. Takeichi, M. Arihara, M. Ishimori, T. Tsuruta, Tetrahedron 36
(1980) 3391.
(b) M.E. Furrow, S.E. Schaus, E.N. Jacobsen, J. Org. Chem. 63
(1998) 6776.
[29] (a) E.N. Jacobsen, F. Kakiuchi, R.G. Konsler, J.F. Larrow, M.
Tokunaga, Tetrahedron Lett. 38 (1997) 773;
[12] (a) T. Katsuki, in: E.N. Jacobsen, A. Pfaltz, H. Yamamoto (Eds.),
Comprehensive Asymmetric Catalysis, Springer, New York, 1999
(Chapters 18.1 and 35);
(b) M. Carron, K.B. Sharpless, J. Org. Chem. 50 (1985) 1560;
(c) G.H. Posner, Angew. Chem., Int. Ed. Engl. 17 (1978) 487.
[30] (a) S. Brickner, D. Hutchinson, M. Barbachyn, P. Manninen, D.
Ulanowicz, S. Garmon, K. Grega, S. Hendges, D. Toops, C. Ford,
G. Zurenko, J. Med. Chem. 39 (1996) 673;
(b) B.E. Rossiter, in: J.D. Morrison (Ed.), Asymmetric Synthesis, vol.
5, Academic Press, New York, 1985 (Chapter 7);
(c) R.A. Johnson, K.B. Sharpless, in: I. Ojima (Ed.), Catalytic
Asymmetric Synthesis, second ed., Wiley-VCH, New York, 2000
(Chapter 6A).
(b) M.A. Weidner-Wells, C.M. Boggs, B.D. Foleno, J. Melton, K.
Bush, R.M. Goldschmidt, D.J. Hlasta, Bioorg. Med. Chem. 10
(2002) 2345.