Highly active oligomeric Co(salen) catalysts for the asymmetric synthesis of α-aryloxy or α-alkoxy alcohols via kinetic resolution of terminal epoxides

Article abstract of DOI:10.1016/j.molcata.2010.06.015

A mixture of Co(salen) macrocycles, prepared via the ring expansion metathesis oligomerization of salen-functionalized cyclooctene monomers, among the most active soluble catalysts for the hydrolytic kinetic resolution (HKR) of terminal epoxides, is exploited as the catalyst in the ring-opening of epoxides using aliphatic alcohols or phenols as nucleophiles, leading to the direct synthesis of optically active α-aryloxy alcohols or α-alkoxy alcohols. The catalyst is compared to other dimeric, oligomeric and monomeric Co(salen) complexes including a pimelate-linked macrocyclic Co(salen) catalyst and a dimeric Co(salen) catalyst referred to as a bisalen. The catalysts that contain multiple Co(salen) units within a single molecular framework allow for substantial decreases in catalyst loading compared with the monomeric catalyst. The cyclooctene-based Co(salen) macrocycle catalyst allows for good activity and enantioselectivity in the ring-opening of terminal epoxides with phenols as nucleophiles, giving enhanced turnover frequencies relative to many literature catalysts. The cyclooctene-based Co(salen) macrocycle catalyst and the bisalen catalysts are shown to be the most active in the asymmetric ring-opening of (±)1,2-epoxyhexane with methanol, out-performing the other catalysts tested. The Co(salen) macrocycle catalyst is recycled 3 times in this reaction with some loss in activity but no noteworthy change in selectivity.

Full text of DOI:10.1016/j.molcata.2010.06.015

Journal of Molecular Catalysis A: Chemical 329 (2010) 1–6
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Editor’s choice paper
Highly active oligomeric Co(salen) catalysts for the asymmetric synthesis of
␣-aryloxy or ␣-alkoxy alcohols via kinetic resolution of terminal epoxides
Xunjin Zhua, Krishnan Venkatasubbaiaha, Marcus Weckb, Christopher W. Jonesa,∗
a School of Chemical & Biomolecular Engineering, Georgia Institute of Technology, 311 Ferst Dr., Atlanta, GA 30332, USA
b Department of Chemistry and Molecular Design Institute, New York University, New York, NY 10003, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 May 2010
Received in revised form 8 June 2010
Accepted 13 June 2010
Available online 19 June 2010
A mixture of Co(salen) macrocycles, prepared via the ring expansion metathesis oligomerization of
salen-functionalized cyclooctene monomers, among the most active soluble catalysts for the hydrolytic
kinetic resolution (HKR) of terminal epoxides, is exploited as the catalyst in the ring-opening of epox-
ides using aliphatic alcohols or phenols as nucleophiles, leading to the direct synthesis of optically
active ␣-aryloxy alcohols or ␣-alkoxy alcohols. The catalyst is compared to other dimeric, oligomeric
and monomeric Co(salen) complexes including a pimelate-linked macrocyclic Co(salen) catalyst and a
dimeric Co(salen) catalyst referred to as a bisalen. The catalysts that contain multiple Co(salen) units
within a single molecular framework allow for substantial decreases in catalyst loading compared with
the monomeric catalyst. The cyclooctene-based Co(salen) macrocycle catalyst allows for good activity and
enantioselectivity in the ring-opening of terminal epoxides with phenols as nucleophiles, giving enhanced
turnover frequencies relative to many literature catalysts. The cyclooctene-based Co(salen) macrocycle
catalyst and the bisalen catalysts are shown to be the most active in the asymmetric ring-opening of
Keywords:
Ring-opening
Epoxide
Phenol
Enantioselectivity
Cooperativity
(
)1,2-epoxyhexane with methanol, out-performing the other catalysts tested. The Co(salen) macrocy-
cle catalyst is recycled 3 times in this reaction with some loss in activity but no noteworthy change in
selectivity.
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
mon to many epoxide ring-opening reactions [20,21]. Our groups
reported the preparation of mixtures of Co(salen) macrocycles
Studies on asymmetric reactions using optically active metal-
losalen complexes as catalysts [1–7] continue to increase, owing
to the versatility of the salen ligand, being a “privileged cat-
alyst,” as designated by Jacobsen [8]. Salen complexes can be
synthesized easily and exhibit distinctive catalytic performance
in a wide array of enantioselective reactions [9–12]. Using chi-
ral cobalt (salen) complexes as catalysts, Jacobsen and co-workers
reported the hydrolytic kinetic resolution (HKR) of terminal epox-
ides using water as the nucleophile and proposed a dual activation
pathway, namely, the activation of two substrates (e.g., an elec-
trophile and a nucleophile) simultaneously by two catalytic species
in the transition state of the rate-determining step [13–16]. Based
on this mechanistic context, complexes that contain multiple
metal centers in appropriate relative proximity and orientation
can provide improved reactivity relative to monometallic catalysts
[17–19]. Jacobsen and co-workers reported in 2001 the prepara-
tion of mixtures of cyclic oligomeric Co(salen) complexes that were
designed to enforce the cooperative bimetallic mechanism com-
(1) (Fig. 1) by the ring-expanding olefin metathesis of the Co(II)
complex of a monocyclooct-4-en-1-yl-functionalized salen ligand,
with the oligomeric catalysts exhibiting extremely high reactiv-
ity and selectivity (using the Co(III) complex) in the HKR of a
variety of racemic terminal epoxides under neat conditions with
very low catalyst loadings [22]. Recently a Co(bisalen) complex (2)
(Fig. 1), with high reactivity and selectivity in the HKR of epox-
ides, was also reported by us [23]. Based on the hypothesis that
strong performance in the HKR reaction may correlate with high
activity in other epoxide ring-opening reactions such as aryloxy-
lation [20,21,24–35] and alkoxylation [20,21], we extend here the
scope of our Co(salen) macrocycle 1(OTs) catalyzed reactions to
the direct synthesis of optically active ␣-aryloxy alcohols and ␣-
alkoxy alcohols that are valuable targets for asymmetric synthesis
and key intermediates in a variety of pharmaceutically impor-
tant compounds [36]. We report that 1, upon aerobic oxidation
under acidic conditions, exhibits excellent catalytic properties in
the ring-opening of epoxides with phenols and aliphatic alcohols,
allowing for substantial decreases in catalyst loading compared
to reactions using the monomeric complex or dinuclear chiral
(salen) Co-MX3 (M = Al, Ga) [24–26], and excellent enantioselec-
tivity.
∗
Corresponding author. Tel.: +1 404 385 1683; fax: +1 404 894 2866.
E-mail address: cjones@chbe.gatech.edu (C.W. Jones).
1381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.06.015

2
X. Zhu et al. / Journal of Molecular Catalysis A: Chemical 329 (2010) 1–6
Fig. 1. Structures of Co(salen) macrocycles (1), bi-Co(salen) (2), oligomeric Co(salen) (3) and monomeric Co(salen) (4).
2. Results and discussion
intramolecular cooperative action of two cobalt sites present in a
single catalyst framework, which has been described in previous
reports [6].
2.1. Preparation of catalysts
A mixture of the Co(salen) macrocycles (1) was prepared via the
ring expansion olefin metathesis reaction of salen-functionalized
cyclooctene monomers. The mixture of oligomeric salen species
obtained had a stoichiometric cobalt content after metallation
based on elemental analysis (EA) (calcd: 8.42%, found: 8.25%) and
exhibited a mass spectrum essentially identical to that reported
previously [22]. For comparison, two additional cooperative cata-
lysts were prepared, following literature reports. These included
the bisalen monomer (2) previously reported by us [23] and the
pimelate-linked macrocyclic catalyst (3) described by Jacobsen
[21]. The elemental analyses show that 2 had a stoichiometric
cobalt content (calcd: 8.53%, found: 8.56%) after metallation while 3
gave a somewhat lower metalation efficiency (calcd: 9.10%, found:
8.00%).
2.3. Catalytic activity in the alcoholic and phenolic kinetic
resolutions
Based on the results of the HKR studies, we selected the most
active catalyst 1(OTs), for use in the ring-opening of epoxides with
phenols as well as aliphatic alcohols. The kinetic resolution of ter-
minal epoxides with phenols has been widely studied recently with
supported and unsupported bimetallic Co(salen)-MX3 catalysts
[25–35], following initial studies using supported and unsupported
cooperative Co(salen) catalysts by Jacobsen [20,21,24]. Here, the
asymmetric ring-opening of racemic terminal epoxides with phe-
nol and its derivatives was selected to evaluate the reactivity and
enantioselectivity of 1(OTs). The addition of 0.45 equivalents of
phenol derivative to racemic epichlorohydrin or 1,2-epoxyhexane
in tert-butyl methyl ether (TBME) solution was examined with
catalyst loadings as low as 0.02% on a per cobalt basis relative
All catalysts were oxidized prior to use by stirring in THF under
air in the presence of toluene sulfonic acid (HOTs), to provide the
corresponding activated catalysts in Co(III) form with TsO− as the
counter ion.
2.2. Catalytic activity in the hydrolytic kinetic resolution
The pre catalysts 1–4 have previously been used in the HKR of
terminal epoxides [8,20,22,23], but they have been run under dif-
ferent conditions. To set the stage for our epoxide ring-opening
studies using aliphatic alcohols and phenols as nucleophiles, we
first compared the catalytic activity of catalysts 1–4 in the HKR of
racemic epichlorohydrin under identical conditions. The reactions
were completed in 20, 30 and 120 min at ambient temperature (RT)
with 0.1 mol% cobalt loading of catalysts 1(OTs), 3(OTs) and 2(OTs),
respectively, affording the (R)-enantiomer of the epoxide in >99%
ee. In comparison, the standard monometallic Co(salen) catalyst
4(OTs) required a prolonged reaction time (15 h) to reach >99% ee
in this reaction. The aforementioned data clearly demonstrate that
1(OTs) and 3(OTs) possess similar outstanding catalytic activities in
the HKR of epoxides (Fig. 2). Compared with 4(OTs), the enhanced
reactivity with 1(OTs), 2(OTs) and 3(OTs) at the same metal load-
ing indicates that the HKR of terminal epoxides takes place via the
Fig. 2. HKR of racemic epichlorohydrin catalyzed by 1(OTs), 2(OTs), 3(OTs) and
4(OTs) at ambient temperature with 0.1 mol% cobalt catalyst loading.

X. Zhu et al. / Journal of Molecular Catalysis A: Chemical 329 (2010) 1–6
3
Table 1
Preparation of ␣-aryloxy alcohols by ring-opening of epoxides with phenol catalyzed by 1(OTs).
Co (mol%)b
T (◦C)
Reaction time (h)
Yield (%)c
ee (%)d
a
Entry
R1
R2
1
2
3
4
5
6
H
CH2Cl
CH2Cl
CH2Cl
n-Butyl
n-Butyl
n-Butyl
0.02
0.02
0.02
0.02
0.02
0.05
RT
RT/4
RT
RT/4
RT
4
9
91
89/85
96
95/97
91
87
97
92/99
95
94/99
97
99
3-CH3
3-Cl
H
3-CH3
3-Cl
21/48
24
12/24
19
24
a
Reactions were carried out with [epoxide]0 = 5 M in TBME.
Catalyst loading on a per Co basis relative to epoxides.
Isolated yields based on phenol.
b
c
d
ee of product, as determined by chiral GC or HPLC analysis.
Table 2
The initial TOF in asymmetric ring-opening of racemic epoxides with phenols cat-
alyzed by 1(OTs).
Entry
Reactants
Co loading (%)
Initial TOF (min−1
)
1
2
3
4
ECH, phenol
0.02
0.02
0.02
0.02
31
10
8.3
2.5
ECH, 3-cresol
ECH, 3-chlorophenol
EH, 3-cresol
Scheme 1. Preparation of ␣-aryloxy alcohols by the ring-opening of epoxides with
phenol catalyzed by 1(OTs).
Note: ECH = epichlorohydrin; EH = 1,2-epoxyhexane.
to epoxide (Scheme 1). The results are summarized in Table 1.
At a catalyst loading of 0.02–0.05 mol%, both electron-poor and
electron-rich phenols reacted at room temperature with a small
series of epoxides to provide the corresponding chiral ␣-aryloxy
alcohols in good to excellent yields and high ee’s. The kinetic reso-
lutions were accomplished with only 0.02 mol% of catalyst 1(OTs) in
less than 12 h using phenol, whereas more than 24 h were required
when using the bulkier molecules 3-chlorophenol or 3-cresol as
nucleophiles. Use of these nucleophiles also resulted in lower
enantioselectivities compared to the phenol case. For comparison,
with 3-chlorophenol as the nucleophile in the kinetic resolution
of epichlorohydrin (ECH), the ␣-aryloxy alcohol was synthesized
in high yield and optical purity after 24 h using only 0.05 mol%
of 1(OTs). In contrast, under otherwise identical conditions, 12 h
were required to obtain a similar result using 0.8 mol% of cata-
lyst 3(OTs) [21]. Higher enantioselectivities could be obtained at
lower reaction temperatures and with longer reaction times using
catalyst 1(OTs). For example, when the reaction was carried out
at 4 ◦C, improved enantioselectivities (99%) were obtained (entry
2, 4 and 6, Table 1) in 24–48 h. Overall, it is evident that 1(OTs)
exhibits excellent activity and selectivity in the kinetic resolutions
of racemic epichlorohydrin and 1,2-epoxyhexane with a variety of
phenol derivatives.
Scheme 2. Preparation of ␣-alkoxy alcohols by ring-opening of epoxides with alco-
hols catalyzed by 1(OTs).
than the estimated initial TOF with several previously reported
unsupported or supported Co(salen) catalysts, which ranged from
∼0.05–1 min−1 [25,29,31,32,34].
The catalytic asymmetric ring-opening of epoxides using
aliphatic alcohols as nucleophiles is a related reaction that has
proven particularly difficult to promote, with few effective cata-
lysts reported to date [20,21]. In the context of kinetic resolution
of terminal epoxides, alcoholic ring-opening provides an attrac-
tive strategy for the direct preparation of optically active ␣-alkoxy
alcohols. A variety of alcohols were found to effectively ring-
open epoxides when using 1(OTs) (Table 3). In all cases, ␣-alkoxy
alcohols were synthesized in high yields and enantioselectiv-
ities. The ring-opening reaction of racemic epichlorohydrin or
1,2-epoxyhexane with methanol was chosen as a model reac-
tion with a Co loading of 0.1% (Scheme 2 and entries 1 and 3 in
Table 3). Catalyst 1(OTs) promoted the kinetic resolution of termi-
nal epoxides effectively to afford 1-chloro-3-methoxy-2-propanol
and 1-methoxy-2-hexanol, respectively, in highly enantioenriched
form (entry1 and 3, Table 3). As for the less nucleophilic reagent
1-hexanol or the bulky 2-bromobenzyl alcohol, the reaction was
completed effectively at room temperature in less than 12 h in these
cases using a higher Co loading (1% Co), while low temperatures
The data in Table 1 give times required to fully resolve epoxides
using different phenols as nucleophiles, an important parameter
needed by synthetic chemists. For catalysis scientists, comparing
the reactivity of various catalysts is more easily achieved by com-
paring initial reaction rates. To this end, the data in Table 2 give
the initial TOFs of catalyst 1(OTs) for the reactions with various
epoxides and phenols. Using 1(OTs) as catalyst, the initial TOF
for the reactions between epoxides and phenols is much higher
Table 3
Preparation of ␣-alkoxy alcohols by ring-opening of epoxides with alcohol catalyzed by 1(OTs).
a
Entry
R1
R2
Co (mol%)b
T (◦C)
Reaction time (h)
Yield (%)c
ee (%)d
1
2
3
4
5
CH3
n-Hexyl
CH3
2-BrPhCH2
2-BrPhCH2
CH2Cl
0.1
1
0.1
1
RT
RT
RT
RT
4
2
4
3
8
24
97
99
96
99
97
99
99
99
98
99
n-Butyl
n-Butyl
CH2Cl
n-Butyl
1
a
Reactions were carried out with [epoxide]0 = 5 M in CH3CN for 2-24 h unless indicated otherwise.
Catalyst loading on a per Co basis relative to epoxides.
Isolated yields based on alcohol.
b
c
d
ee of product, as determined by chiral GC or HPLC analysis.

4
X. Zhu et al. / Journal of Molecular Catalysis A: Chemical 329 (2010) 1–6
Table 4
Compared with the monomeric Co(salen) 4(OTs), the enhanced
reactivity with all the multi-center catalysts 1(OTs), 2(OTs) and
3(OTs) indicates that the kinetic resolution of terminal epoxides
with alcohols also takes place via the intramolecular cooperative
action of two cobalt sites present in a single catalyst framework,
when using these cooperative catalysts.
The initial TOF in asymmetric ring-opening of racemic 1,2-
epoxyhexane with methanol catalyzed by 1(OTs), 2(OTs),
3(OTs) and 4(OTs) with 0.1 mol% Co loading.
Entry
Initial rate (min−1
)
1(OTs)
2(OTs)
3(OTs)
4(OTs)
6.7
3.3
2.5
0.7
To further demonstrate the cooperativity of the catalysts, we
assessed the order of the reactions with respect to catalyst con-
centration during the kinetic resolution of racemic epichlorohydrin
with phenol using activated 1(OTs). The plot of initial rate vs. [cat-
alyst] gives insight into the catalyst reaction order. In this case, the
plot is linear within the reported range of catalyst loadings, consis-
tent with a first-order kinetic dependence on catalyst concentration
(Fig. 4) [13]. This stands in contrast to the second-order depen-
dence observed with monomeric catalysts [15,16,20] and indicates
that epoxide ring-opening with activated 1(OTs) takes place via the
cooperative reactivity of two (or possibly more) metal sites within
a single cyclic framework under the diluted conditions used here.
As noted in Table 4, the initial TOF using 1(OTs) for ring-opening of
epoxide with methanol as well as phenol is about ten times higher
than using monomeric Co(salen) 4(OTs). Thus, active, bimetallic
transition states are more effectively created intramolecularly with
cooperative catalysts at high dilution, whereas conversion arising
from inter-molecular cooperativity contributes very little to the
observed rate under the conditions used here.
were required to produce optically pure 1-(2-bromobenzyloxy)-2-
hexanol.
The catalytic results support the hypothesis that strong perfor-
mance in the HKR reaction can correlate with efficient reactivity
in other epoxide ring-opening reactions such as aryloxylation and
alkoxylation. In the case of 1(OTs), it is suggested that the same
cooperative macrocyclic structure that allowed for extremely effi-
cient HKR catalysis also allows for conversion of more demanding
substrates such as alcohols and phenols.
As noted above, the literature contains very few examples of
effective catalysts for ring-opening of epoxides with aliphatic alco-
hols as nucleophiles. The best example is Jacobsen’s work using
3(OTs) [21]. In that system, the reactions were conducted under
different conditions than those used here. To better compare the
catalytic activity of catalysts 1(OTs)–4(OTs) in the kinetic res-
olution of terminal epoxides with alcohols, the ring-opening of
1,2-epoxyhexane with 0.6 equivalents of methanol was examined
under identical conditions with all four catalysts, using CH3CN as
the solvent at 0.1 mol% catalyst loading relative to epoxide. While
complexes 1(OTs) and 3(OTs) displayed similar activities in the HKR
(vide supra), in our hands, 1(OTs) proved to be superior for the
kinetic resolution of terminal epoxides with alcohols. The initial
TOFs using the four catalysts are given in Table 4, again show-
ing that 1(OTs) is the most active catalyst. In this case, 2(OTs),
which was inferior to 1(OTs) and 3(OTs) in the HKR, unexpectedly
showed similar activities to 1(OTs), completing the reaction within
3 h, affording the (R)-enantiomer of the epoxide in >99% ee (50–53%
GC conversion) at low catalyst loadings (Fig. 3). The cause of this
different behavior using the two nucleophiles (water vs. methanol)
is not yet clear. It should be noted that the somewhat lower meta-
lation efficiency of 3 might be one reason for the lower reactivity of
3(OTs) compared with 1(OTs). Although the reactions were com-
paredatthe samecobalt loading, lowermetalationefficiency means
a larger fraction of the available salen ligands are not metalated in
3(OTs) compared to 1(OTs), which may affect the intramolecular
site pairing probabilities.
2.4. Catalyst recovery and recycle
The recovery and reuse of the catalyst was assessed. For these
studies, the ring-opening reactions of ( )-1,2-epoxyhexane using
0.6 equivalents of methanol in the presence of 1(OTs) (0.2 mol%
Co) was chosen. After each cycle, the catalyst was recovered via
removal of the product, residual reactant, and solvent under vac-
uum, followed by washing with a mixture of water and ethanol
(1:1, v/v) to further clean the liquid residue, followed by drying
under vacuum prior to reuse. The recycle studies show that the
selectivity of the recycled catalyst remained about the same after
three runs, but the reactivity clearly decreased (Fig. 5). The time
required to achieve ≥99% ee of the epoxide increased from 45 to
90 min over three runs. It should be noted that no regeneration of
the catalyst with p-toluenesulfonic acid was carried out between
individual runs. Future studies may focus on elucidating the mech-
anisms that lead to loss of activity upon recycle, with past work on
HKR reactions suggesting that ligand hydrolysis, decomposition,
metal leaching, metal reduction, and counter ion exchange may be
important [37–40].
Fig. 3. Asymmetric ring-opening of racemic 1,2-epoxyhexane with methanol cat-
alyzed by 1(OTs), 2(OTs), 3(OTs) and 4(OTs) with 0.1 mol% Co loading.
Fig. 4. Initial reaction rates of the phenolic ring-opening of ECH catalyzed by 1(OTs).

X. Zhu et al. / Journal of Molecular Catalysis A: Chemical 329 (2010) 1–6
5
(40 m × 25 mm × 0.25 ␮m). An Shimadzu HPLC with a UV detector
and a CHIRALCEL OD (4.6 × 250 mm, 10 mic) column for separation
was also utilized. Elemental analyses were performed by Desert
Analytics Lab (Tucson, AZ, USA).
4.1. Preparation of catalysts
The Co(salen) macrocycle (1) and the bi-Co(salen) (2) complexes
were prepared according to published literature methods [22,23].
The 1H NMR spectra and matrix assisted laser desorption ionization
time-of-flight (MALDI-TOF) mass spectra obtained for 1 as well as
2 were similar to literature reports.
Elemental analysis for Co(salen) macrocycles (1): calcd (%) C
70.37, H 8.07, N 4.00, Co 8.42; found: C 70.63, H 8.04, N 3,56, Co
8.25. Elemental analysis for bi-Co(salen) (2): calcd (%) C 70.42, H
7.73, N 4.06, Co 8.53; found: C 69.70, H 7.61, N 3.92, Co 8.56. The
monomeric Co(salen) (4) was used as received from Sigma–Aldrich.
The pimelate-linked oligomeric Co(salen) (3) catalyst was pre-
pared in close analogy to the method reported by Jacobsen and
co-workers [21]. A solution of (1R,2R)-1,2-diaminocyclohexane
mono-(+)-tartrate salt (1.75 g, 6.59 mmol) in THF was mixed with
K2CO3 in H2O (8.2 mL) and the resulting solution was refluxed for
30 min. Next, 0.5 equivalent of dialdehyde, which was prepared
by condensation of 3-tert-butyl-2,5-dihydroxy benzaldehyde and
pimelic acid, was added as a solution in THF (22 mL). The reac-
tion was stirred at reflux for 2 h, cooled to room temperature and
diluted with ethyl acetate (100 mL). After separation, a yellow solid
Fig. 5. Asymmetric ring-opening of racemic 1,2-epoxyhexane with methanol cat-
alyzed by 1(OTs) with 0.2 mol% Co loading.
3. Conclusions
Upon aerobic oxidation under acidic conditions, the mixture of
Co(salen) macrocycles 1(OTs) exhibited excellent catalytic prop-
erties in the ring-opening of epoxides with phenols and alcohols,
showing substantial decreases in catalyst loading compared to
the monomeric salen catalyst 4(OTs), and displaying excellent
enantioselectivity. The high reactivity and enantioselectivity of
1(OTs) may be associated with the incorporation of salen units
into an extremely flexible cyclic framework allowing for improved
bimetallic cooperativity. Kinetic studies indicated that epoxide
ring-opening with 1(OTs) followed a first-order kinetic depen-
dence on catalyst, which was consistent with an intramolecular
bimetallic cooperative reaction of multiple centers within the
cyclic oligomeric framework. While macrocycles 1(OTs) and 3(OTs)
displayed similar activity in the HKR reaction, complex 1(OTs)
proved superior for the kinetic resolution of terminal epoxides
with aliphatic alcohols under the conditions used. Catalyst 1(OTs)
also proved to be substantially more active than other cooper-
ative Co(salen) catalysts reported in the literature for epoxide
ring-opening with phenols. Catalyst 2(OTs), which was infe-
rior in the HKR, unexpectedly showed similar activity to 1(OTs)
in ring-opening of 1,2-epoxyhexane with methanol. The recy-
cle studies showed that over three runs in the ring-opening of
1,2-epoxyhexane with methanol, the activity of 1(OTs) clearly
decreased but that catalytic selectivity remained excellent.
was obtained quantitatively, and characterized by 1H NMR and 13
C
NMR spectroscopies and MS, with the results consistent with the
literature. Metalation using cobalt(II) actetate tetrahydrate with
the (salen) macrocycles in a dichloromethane/methanol mixture
afforded the oligomeric Co(salen) (3) pre-catalyst. Elemental anal-
ysis for Co(salen) macrocycles (3): calcd (%) C 64.91, H 6.85, N 4.33,
Co 9.10; found: C 64.49, H 6.78, N 4.55, Co 8.00.
4.2. General procedure for the catalytic synthesis of ˛-alkoxy
alcohols or ˛-aryloxy alcohols
The alcohol or phenol of choice (45 mmol), epoxide
(10.00 mmol), internal standard chlorobenzene (100 ␮L) and
CH3CN or tert-butyl methyl ether (TBME) (0.2 mL) were mixed
with activated 1 (7.2 mg, 0.01 mmol Co containing for Table 2
entry 3) at room temperature or 4 ◦C, and the solution was stirred
until GC analysis indicated complete conversion of the alcohol.
The reaction was then diluted with 5 mL Et2O and filtered through
a plug of silica gel to remove the catalyst. The plug was washed
with 20 mL Et2O. The filtrates were combined and concentrated
under reduced pressure to provide the crude product. Further
purification included flushing through a silica gel column and
distillation under vacuum for recovery of individual compounds.
The ee of the product was determined by chiral GC or HPLC.
4. Experimental section
Reagents were used as received unless otherwise noted.
Dichloromethane (DCM) was dried by passing through columns of
activated alumina. Toluene and tetrahydrofuran (THF) were dried
by passing through columns of activated copper oxide and alu-
mina successively. 1H and 13C NMR spectra were acquired with
a Varian Mercury 400 MHz spectrometer, and chemical shifts are
reported in ppm with reference to the corresponding residual
nuclei of the deuterated solvents. Mass spectra were analyzed
using a VG 7070 EQ-HF hybrid tandem mass spectrometer. Gel
permeation chromatography (GPC) analyses were performed on
American Polymer Standards columns equipped with a Waters 510
pump and UV detector, using poly(styrene) standards for calibra-
tion and THF as the mobile phase at a flow rate of 1.0 mL/min.
Enantiomeric excesses were determined by capillary gas-phase
chromatography (GC) analysis on a Shimadzu GC 14A instru-
ment equipped with a FID detector and a Chiraldex ␥-TA column
Acknowledgement
We are thankful to the Department of Energy Office of
Basic Energy Sciences through Catalysis Contract No. DEFG02-
03ER15459 for financial support of this work.
References
[1] P.G. Cozzi, Chem. Soc. Rev. 33 (2004) 410.
[2] T. Katsuki, Chem. Soc. Rev. 33 (2004) 437.
[3] L. Canali, D.C. Sherrington, Chem. Soc. Rev. 28 (1999) 85.
[4] N.E. Leadbeater, M. Marco, Chem. Rev. 102 (2002) 3217.
[5] D.J. Darensbourg, Chem. Rev. 107 (2007) 2388.
[6] N. Madhavan, C.W. Jones, M. Weck, Acc. Chem. Res. 41 (2008) 1153.
[7] R.M. Haak, S.J. Wezenberga, A.W. Kleij, Chem. Commun. 46 (2010) 2713.

6
X. Zhu et al. / Journal of Molecular Catalysis A: Chemical 329 (2010) 1–6
[8] T.P. Yoon, E.N. Jacobsen, Science 299 (2003) 1691.
[25] K.-Y. Lee, C.-Y. Lee, G.-J. Kim, React. Kinet. Catal. Lett. 93 (2008) 75.
[26] R.B. Kawthekar, C.-H. Ahn, G.-J. Kim, Catal. Lett. 115 (2007) 62.
[27] Y.-S. Kim, X. Guo, G.-J. Kim, Chem. Commun. (2009) 4296.
[28] R.B. Kawthekar, Y.H. Lee, G.J. Kim, J. Porous Mater. 16 (2009) 367.
[29] K.Y. Lee, C.Y. Lee, G.J. Kim, Bull. Korean Chem. Soc. 30 (2009) 389.
[30] X. Guo, Y.S. Kim, G.J. Kim, Top. Catal. 52 (2009) 153.
[31] Y.S. Kim, X. Guo, G.J. Kim, Top. Catal. 52 (2009) 197.
[32] R.B. Kawthekar, W. Bi, G.J. Kim, Appl. Organometal. Chem. 22 (2008) 583.
[33] K.Y. Lee, R.B. Kawthekar, G.J. Kim, Bull. Korean Chem. Soc. 28 (2007) 1553.
[34] Y.S. Kim, C.Y. Lee, G.J. Kim, Bull. Korean Chem. Soc. 30 (2009) 1771.
[35] R.B. Kawthekar, W.T. Bi, G.J. Kim, Bull. Korean Chem. Soc. 29 (2008) 313.
[36] M.P. Kirkup, R. Rizvi, B.B. Shankar, S. Dugar, J. Clader, S.W. McCombie, S. Lin, N.
Yumibe, K. Huie, M. Heek, D.S. Compton, H.R. Davis, A.T. McPhail, Bioorg. Med.
Chem. Lett. 6 (1996) 2069.
[9] C. Mazet, E.N. Jacobsen, Angew. Chem., Int. Ed. 47 (2008) 1762.
[10] N. Madhavan, M. Weck, Adv. Synth. Catal. 350 (2008) 419.
[11] L. Yang, D. Wang, Z. Huang, M. Wang, J. Am. Chem. Soc. 131 (2009) 10390.
[12] H. Suematus, S. Kanchiku, U. Shigefumi, T. Katsuki, J. Am. Chem. Soc. 130 (2008)
10327.
[13] M.T. Kunaga, J.F. Larrow, F. Kakiuchi, E.N. Jacobsen, Science 277 (1997) 936.
[14] S.E. Schaus, B.D. Brandes, J.F. Larrow, M. Tokunaga, K.B. Hansen, A.E. Gould, M.E.
Furrow, E.N. Jacobsen, J. Am. Chem. Soc. 124 (2002) 1307.
[15] E.N. Jacobsen, Acc. Chem. Res. 33 (2000) 421.
[16] E.N. Jacobsen, M.H. Wu, in: E.N. Jacobsen, A. Pfaltz, H. Yamamoto (Eds.), Com-
prehensive Asymmetric Catalysis, vol. 2, Springer, Berlin, 1999, p. 1309.
[17] R. Breinbauer, E.N. Jacobsen, Angew. Chem., Int. Ed. 39 (2000) 3604.
[18] D.E. White, E.N. Jacobsen, Tetrahedron: Asymmetry 14 (2003) 3633.
[19] R.I. Kureshy, S. Singh, N.-U.H. Khan, S.H.R. Abri, I. Ahmad, A. Bhatt, R.V. Jasra,
Chirality 17 (2005) 590.
[37] C.S. Gill, K. Venkatasubbaiah, N.T.S. Phan, M. Weck, C.W. Jones, Chem. Eur. J. 14
(2008) 7306.
[20] J.M. Ready, E.N. Jacobsen, J. Am. Chem. Soc. 123 (2001) 2687.
[21] J.M. Ready, E.N. Jacobsen, Angew. Chem., Int. Ed. 41 (2002) 1374.
[22] X. Zheng, C.W. Jones, M. Weck, J. Am. Chem. Soc. 129 (2007) 1105.
[23] K. Venkatasubbaiah, C.S. Gill, T. Takatani, D. Sherrill, C.W. Jones, Chem. Eur. J.
15 (2009) 3951.
[38] L.P.C. Nielsen, C.P. Stevenson, D.G. Blackmond, E.N. Jacobsen, J. Am. Chem. Soc.
126 (2004) 1360.
[39] S. Jain, K. Venkatasubbaiah, C.W. Jones, R.J. Davis, J. Mol. Catal. A: Chem. 315
(2010) 8.
[40] S. Jain, X. Zheng, C.W. Jones, M. Weck, R.J. Davis, Inorg. Chem. 46 (2007)
8887.
[24] D.A. Annis, E.N. Jacobsen, J. Am. Chem. Soc. 121 (1999) 4147.