Olefin metathesis as a tool for multinuclear Co(III)salen catalyst construction: Access to cooperative catalysts

Article abstract of DOI:10.1039/b911856j

The construction of novel (macrocyclic) multinuclear Co(iii)salen catalysts is reported. Olefin metathesis has been used as a key construction tool for the multimetallic structures starting from versatile allyl-substituted salen scaffolds. The Co(iii) complexes were tested in the hydrolytic kinetic resolution of (rac)-1,2-epoxyhexane and epoxide ring opening reactions using methanol as the nucleophile. The preliminary results suggest a cooperative mode of catalysis in the case of bis-Co(iii)salen macrocycle 10. The Royal Society of Chemistry.

Full text of DOI:10.1039/b911856j

View Article Online / Journal Homepage / Table of Contents for this issue
This article is published as part of the Dalton Transactions themed issue entitled:
New Talent
Showcasing the strength of research being carried out by
tomorrow's leaders in the field of inorganic chemistry
Guest Editor Polly Arnold
University of Edinburgh, UK
Published in issue 2, 2010 of Dalton Transactions
Image reproduced with permission of Mark Muldoon
Articles in the issue include:
PERSPECTIVES:
Modern multiphase catalysis: new developments in the separation of homogeneous catalysts
Mark J. Muldoon, Dalton Trans., 2010, DOI: 10.1039/b916861n
Probing bioinorganic chemistry processes in the bloodstream to gain new insights into the origin of
human diseases
Elham Zeini Jahromi and Jürgen Gailer, Dalton Trans., 2010, DOI: 10.1039/b912941n
COMMUNICATIONS:
Facile entry to 3d late transition metal boryl complexes
Ba L. Tran, Debashis Adhikari, Hongjun Fan, Maren Pink and Daniel J. Mindiola,
Dalton Trans., 2010, DOI: 10.1039/b912040h
Probing the kinetics of ligand exchange on colloidal gold nanoparticles by surface-enhanced raman
scattering
Yuhua Feng, Shuangxi Xing, Jun Xu, Hong Wang, Jun Wei Lim and Hongyu Chen,
Dalton Trans., 2010, DOI: 10.1039/b912317b
Visit the Dalton Transactions website for more cutting-edge inorganic and organometallic research
www.rsc.org/dalton

PAPER
www.rsc.org/dalton | Dalton Transactions
Olefin metathesis as a tool for multinuclear Co(III)salen catalyst construction:
access to cooperative catalysts†
Robert M. Haak,a Marta Mart´ınez Belmonte,a Eduardo C. Escudero-Ada´n,a Jordi Benet-Buchholza and
Arjan W. Kleij*a,b
Received 17th June 2009, Accepted 30th October 2009
First published as an Advance Article on the web 12th November 2009
DOI: 10.1039/b911856j
The construction of novel (macrocyclic) multinuclear Co(III)salen catalysts is reported. Olefin
metathesis has been used as a key construction tool for the multimetallic structures starting from
versatile allyl-substituted salen scaffolds. The Co(III) complexes were tested in the hydrolytic kinetic
resolution of (rac)-1,2-epoxyhexane and epoxide ring opening reactions using methanol as the
nucleophile. The preliminary results suggest a cooperative mode of catalysis in the case of
bis-Co(III)salen macrocycle 10.
electronic/steric features have also been prepared to serve as
references in the catalytic studies.
Introduction
Complexes based on salen ligands are considered privileged
catalysts owing to the remarkably wide variety of reactions they are
Results and discussion
able to catalyze.1 Notable examples (among many others) include
asymmetric epoxidation2 and enantioselective ring opening and
hydrolytic kinetic resolution of epoxides.3 In these reactions, high
catalytic activity is accompanied by excellent enantioselectivity.
Apart from their catalytic versatility, salen complexes are also
characterized by their facile, modular synthesis that allows for
the straightforward introduction of various peripheral groups,
metal ions and chiral modules.4 Because of these latter features,
metallosalens have received recognition as useful and versatile
supramolecular building blocks for materials applications.5
At the end of the last century, Jacobsen and coworkers discov-
ered that metallosalen catalysts perform a dual role of activating
both the epoxide and the attacking nucleophile in the ring opening
of epoxides.3 In subsequent reports by a number of research
groups, it was shown that cooperativity between metal centers
may be enhanced by positioning two or more salen units in close
mutual proximity. Strategies that have been used so far include
covalent attachment of two building blocks to create a dinuclear
catalyst,6 supramolecular assembly of multimetallic catalysts7 or
the synthesis of oligomers and polymers containing pendant
metallosalen units via ring opening metathesis polymerization.8
In this paper, we report the design and construction of
new multinuclear Co(III)salen catalysts for epoxide ring opening
reactions, using olefin metathesis as a key coupling strategy. Fur-
thermore, a small set of mononuclear analogues with comparable
Cobalt-salen complexes 4 and 5a–5c (the monomeric analogues
of the multinuclear compounds, vide infra) were prepared by
condensation of chiral diamine (R,R)-2 with two equivalents of
the commercially available allyl-functionalized salicylaldehydes
1a and 1b (Scheme 1).9 The resulting ligands 3 were allowed to
react with Co(OAc)2·4H2O and one equiv. of lutidinium p-toluene
sulfonic acid (LPTS) or acetic acid, respectively, either in situ or
after isolation and characterization of the ligand precursor (in the
case of 4).
aInstitute of Chemical Research of Catalonia (ICIQ), Av. Pa¨ısos Catalans
16, 43007, Tarragona, Spain. E-mail: akleij@iciq.es; Fax: +34 977920224;
Tel: +34 977920247
bInstitucio´ Catalana de Recerca i Estudis Avanc¸ats (ICREA), Pg. Llu´ıs
Companys 23, 08010, Barcelona, Spain
† Electronic supplementary information (ESI) available: Copies of relevant
spectra and crystallographic comments and GC/HPLC methods. CCDC
reference numbers 735901 for 4b, 735902 for dialdehyde 7 and 735903 for
dialdehyde 6. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/b911856j
Scheme 1 Synthesis of Co(III)salen complexes 4 and 5a–c. Complex 5c
was obtained by crystallization of 5b from acetone.
Alternatively, we found that condensation of 1a and (R,R)-2 in
refluxing ethanol, followed by addition of Co(OAc)2 and LPTS
at room temperature, cleanly yielded complex 4, thus avoiding
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 593–602 | 593
©

precautions such as working under an inert atmosphere or using
degassed solvents.10
Interestingly, in the case of 5b we observed a doubling of
1
practically all signals in the 13C{ H} NMR spectrum, indicative of
the loss of C2-symmetry when a sufficiently strong axial ligand
such as acetate coordinates to the cobalt center. For tosylate
complexes 4 and 5a similar signal duplication was not observed,
probably due to the more labile binding of the tosylate anion,
which leads to a fast dynamic equilibrium on the NMR timescale.
Initially, we attempted to couple Co(III)salen complexes 4 and
5a (Scheme 1) functionalized with pendant allyl groups at the
3,3¢ or 5,5¢ positions using ruthenium-catalyzed olefin metathesis
in order to create, in a direct fashion, oligomeric metallosalen
species (Scheme 2). However, no conversion was observed, despite
the fact that the corresponding nickel centered complexes gave a
clean and complete conversion.11
Scheme 4 Preparation of oligomeric and macrocyclic multinuclear
Co(III)salens 9 and 10, respectively.
Scheme
2
Attempted cross-metathesis of allyl-functionalized
Co(III)salens. [Ru] stands for Grubbs I catalyst.
Therefore we decided to use another strategy towards the con-
struction of multinuclear Co(III)salen complexes. This approach
starts with the synthesis of the dialdehydes 6 and 7 (Scheme 3,
both isolated in 61% yield) via metathesis using Grubbs I catalyst.
Although the crude products are obtained with a trans/cis
ratio of 4 : 1, pure trans-products could be obtained by a single
recrystallization from DCM–MeOH.
Scheme 5 Preparation of dinuclear complex 14.
LPTS. Oligomers 8 and 9 are only sparingly soluble and therefore
we also probed the same reaction using dialdehyde 6. Interestingly,
in the latter synthesis macrocyclic, dinuclear Co(III)salen 10 (67%
isolated yield) was selectively formed. A one-pot procedure proved
to be successful and comprises the condensation reaction of
bis(salicylaldehyde) 6 and (R,R)-2 in refluxing ethanol, followed
by insertion of cobalt and in situ oxidation in the presence of LPTS
(Scheme 4).
Scheme 3 Preparation of bifunctional building blocks 6 and 7.
Building blocks 6 and 7 were subsequently used for the construc-
tion of multinuclear Co(III)salen scaffolds 9, 10 and 14 as outlined
in Schemes 4 and 5. For the synthesis of oligomeric Co(III) complex
9, bis(salicylaldehyde) 7 was reacted with an equimolar amount
of (R,R)-2 in MeOH, yielding oligomeric ligand 8 (Scheme 4).
To obtain the Co(III) complex 9, the procedure of Jacobsen and
Ready was followed.12 Thus, 8 was reacted with cobalt(II) acetate
under argon, followed by oxidation with air in the presence of
The cyclic nature of 10 was unambiguously demonstrated by
NMR and MALDI-TOF mass spectrometry (ESI†). Only small
594 | Dalton Trans., 2010, 39, 593–602
This journal is
The Royal Society of Chemistry 2010
©

◦
˚
amounts of insoluble oligomeric side products were formed during
the reaction which could be conveniently removed by filtration
over Celite. This method thus constitutes an attractive route
to the selective construction of chiral macrocyclic, dinuclear
metallosalen scaffolds.
Asymmetrically substituted (within each salen unit) complex
14 was prepared using the procedure of Weck and coworkers
(Scheme 5).13 Initially, 1c was reacted with mono-protected
diamine 1114,15 yielding monoprotected “half-salen” 12, which was
reacted in situ with half an equivalent of 7 to give asymmetrically
substituted bis-salen 13. The corresponding Co(III)-tosylate com-
plex 14 was then obtained analogously to 9.
All new compounds were characterized by NMR spec-
troscopy, MALDI-TOF mass spectrometry, IR spectroscopy and
HRMS/elemental analyses.16 For dialdehydes 6 and 7 and for
racemic Co(III) complex 5c (Scheme 1) the identity was also
established through single crystal X-ray diffraction (Fig. 1 and
2, Tables 1 and 2).
Table 1 Selected inter-atomic distances (A) and bond/dihedral angles ( )
of dialdehydes 6 and 7 and Co(III)salen complex 5c
6
C(7)-C(8)
C(9)-C(9A)
7
1.5129(11)
1.3303(15)
C(8)-C(9)-C(9A)-C(8A)
C(7)-C(8)-C(9)-C(9A)
180.00(8)
-120.92(9)
C(7)-C(8)
C(8)-C(8A)
5c
Co(1)-O(1W)
Co(1)-O(2)
Co(1)-O(3)
Co(1)-N(2)
Co(1)-N(3)
Co(1)-C(2A)
1.498(3)
1.327(4)
C(7)-C(8)-C(8A)-C(7A)
C(4)-C(7)-C(8)-C(8a)
-180.00(17)
127.5 (2)
2.086(11)
N(2)-Co(1)-O(3)
N(3)-Co(1)-O(2)
C(2A)-Co(1)-O(1W)
N(2)-Co(1)-N(3)
O(2)-Co(1)-C(2A)
O(3)-Co(1)-O(1W)
176.68(5)
178.04(5)
175.61(6)
85.92(6)
90.11(6)
89.61(5)
1.9064(10)
1.9013(11)
1.8806(13)
1.8874(12)
1.9907(15)
diametrically opposed mutual orientation is remarkable in view
of the fact that in the synthesis of 10 these moieties clearly need
to point in the same direction for selective formation of the
metallomacrocycle.
In the crystal structure of 5c, the Co(III) center is in a slightly
distorted octahedral geometry (Table 1) with the salen ligand
donor occupying the equatorial positions. The axial ligands are
water and remarkably an acetonyl ligand instead of the expected
acetate anion. Moreover, the acetone is bound to the metal center
via a Co–C bond. This interesting feature hints at possibilities for
the mild activation of C–H bonds in the a-position to carbonyl
functionalities. At this stage we believe that the acetonyl ligand
probably originates from an (intermolecular) proton exchange
with the acetate base followed by preferred crystallization of 5c
instead of 5b.
Although the C-bound acetone complex 5c represents a rather
remarkable structure, there are some literature precedents for
Co(salen) and/or similar structures bearing stable Co–C bonded
axial ligands.17 These complexes have for instance been used as
model compounds for vitamin B12-containing enzymes.
In view of our interest in the potential of multinuclear
Co(III)salens as cooperative catalysts we tested the catalytic
activity of complexes 4, 5a, 9, 10 and 14 in hydrolytic kinetic
resolution (HKR) using (rac)-1,2-epoxyhexane [(rac)-15a] as a
benchmark substrate. The results of these studies are summarized
in Table 3.
Fig. 1 Molecular structures determined for dialdehydes 6 (top) and 7
(bottom). H-atoms and co-crystallized solvent molecules are omitted for
clarity.
First, the performance of complexes 4, 5a, 9, 10 and 14 in
the HKR under neat conditions was assessed. Unfortunately,
complexes 4, 5a, 9 and 10 were completely inactive under these
conditions. Only dinuclear catalyst 14 (the only well-soluble cata-
lyst under these conditions) showed good activity (Table 3, entry
1) together with high enantioselectivity for both the remaining
epoxide as well as diol product. We attribute the low activity, in
particular that of multinuclear catalysts 9, to their low solubility in
the epoxide reagent and therefore subsequent HKR experiments
were carried out using MeCN as co-solvent.
In acetonitrile complex 14 catalyzes, as expected, the HKR
of (rac)-15 more slowly than under neat conditions (cf. entries
1 and 6) probably due to competition of the solvent for axial
coordination to 14. Remarkably, macrocyclic Co(III)salen 10
shows the highest catalytic activity under these conditions with
moderate enantio-discrimination (entry 5). Complexes 4, 5a and 9
are virtually inactive in acetonitrile. In the case of 9, this is ascribed
to its low solubility even in the presence of a co-solvent. The
Fig. 2 Molecular structure determined for 5c. H-atoms, co-crystallized
solvent molecules and partial disorder of the c-hexyl group are omitted for
clarity.
The structures of 6 and 7 clearly reveal the rigid nature of
these bifunctional building blocks with trans configuration of the
olefinic part, with 7 in particular showing an almost perfectly
planar geometry. Interestingly, in 6, the salicylaldehyde units
on the two phenyl rings point in opposite directions. This
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 593–602 | 595
©

Table 2 Crystal and structure determination data for dialdehydes 6 and 7 and complex 5c
6
7
5c
Empirical formula
Formula weight
T/K
Crystal system
Space group
C18H16O4
296.31
293(2)
C20H20O6
356.36
100(2)
Monoclinic
P21/c
C34H45N2O7Co
652.65
100(2)
Triclinic
¯
P1
a = 11.3743(4)
b = 11.5252(4)
c = 14.8043(6)
a = 78.151(2)
b = 68.412(2)
g = 62.374(2)
1597.42(10)
2
Monoclinic
P21/n (no.14)
a = 4.4309(4)
b = 9.7243(8)
c = 16.3855(11)
a = 90
˚
Unit cell dimensions/A
a = 4.031(2)
b = 12.305(7)
c = 16.759(10)
a = 90
Unit cell angles/◦
b = 94.510(5)
g = 90
b = 93.249(16)
g = 90
3
˚
V/A
703.82(10)
2
830.0(8)
2
Z
F(000)
312
1.398
0.80 ¥ 0.30 ¥ 0.20
0.71073
0.099
4.29, 36.4
4891
2851
0.0223
376
1.426
0.3 ¥ 0.03 ¥ 0.02
0.71073
0.105
2.43, 31.61
10529
2765
0.1316
692
1.357
0.40 ¥ 0.20 ¥ 0.15
0.71073
0.588
2.00, 29.16
16551
8399
0.0222
rc/g cm-3
Crystal size/mm
˚
l/A
m/mm-1
q
min, qmax (◦)
Total reflections
Unique reflections
Rint
Data [I > 2.0 s(I)]
2460
2851, 101
1.04
1601
2765, 109
1.02
7273
8399, 447
1.04
N
ref, Npar
GOF on F2
R1, wR2 (all data)
R1, wR2 [I > 2.0 s(I)]
˚
0.0533, 0.1384
0.0467, 0.1324b
0.573, -0.274
0.1244, 0.1876
0.0687, 0.1573a
0.378, -0.332
0.0437, 0.0976
0.0363, 0.0928c
0.657, -0.335
3
Larg. peak/hole (e/A )
2
2
2
2
2
a w = 1/[s (Fo2) + (0.0674P)2] where P = (Fo + 2Fc2)/3. b w = 1/[s (Fo2) + (0.0747P)2 + 0.1875P] where P = (Fo + 2Fc2)/3. c w = 1/[s (Fo2) +
(0.0500P)2 + 0.7166P] where P = (Fo2+2Fc2)/3.
Table 3 HKR of (rac)-epoxides catalyzed by 4, 5, 9, 10 and 14a
c
c
Entry
Substrate
Solvent
Catalyst
t/h
conv. (%)b
eeepox
eediol
1
2
3
4
5
6
7
8
15a
15a
15a
15a
15a
15a
15b
15b
—
14d
4
5a
9
10
14
10
14
21
54
54
54
54
54
54
54
50
6
0
0
29
8
80
89
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
MeCN
n.d.
n.d.
n.d.
25
12
14
n.d.
n.d.
n.d.
43
n.d.
31
18
27
25
74
a See the Experimental part for details. Conditions: 2.0 mmol (rac)-epoxide, 1.1 mmol demineralized H2O, 0.1 mol% Co, 0.2 mL solvent. b Please note: in
the HKR of (rac)-epoxides the maximum yield of enantiomerically pure epoxide/diol is 50%. c N.d. = not determined. d 0.2 mol% Co was used.
combined results nevertheless point out that bimetallic macrocylic
10 is the most efficient catalyst and we suggest some degree
of cooperativity between both Co(III) centres in this species.
Besides 15a, the epoxides 1,2-epoxy-5-hexene, styrene oxide,
epichlorohydrin, and cyclohexene oxide were tested in HKR,
however no conversion was observed using either 10 or 14 as
catalyst. On the other hand, benzyl glycidyl ether (15b) was found
to be an effective substrate, leading to the corresponding diol 16a
with moderate conversion and, when catalyst 14 was used, high ee
(74%, Table 3, entries 7 and 8).
Using methanol as co-solvent, we observed that the HKR of
(rac)-15 is significantly suppressed (Table 4, entries 1–5). More-
over, the main product in this case arises from epoxide ring opening
by methanol, leading to synthetically valuable mono-protected
diols 17a–e.12,18 Again, dinuclear macrocyclic catalyst 10 is more
effective in this transformation than mononuclear catalysts 4 and
5a, and dinuclear 14, emphasizing the importance of the presence
of two proximal metal centres to induce cooperative effects in
this Co(salen)-mediated methanolysis. A similar observation was
done by Ready and Jacobsen who described a cyclic oligomeric
596 | Dalton Trans., 2010, 39, 593–602
This journal is
The Royal Society of Chemistry 2010
©

Table 4 Methanolysis of (rac)-epoxides catalyzed by 4, 5a, 9, 10, 14, and 20a
d
Entry
Substrate
H2O (equiv.)
Catalyst
t/h
15 bee (%)
17 cconv./% (ee/%)
17¢ (conv./%)
16 e(conv./%)
1
2
3
4
5
6
15a
15a
15a
15a
15a
15a
0.55
0.55
0.55
0.55
0.55
—
4
5a
9
10
14
4
40
55
55
55
55
1.5
5
21
1.5
5
21
1.5
5
21
54
54
21
54
54
54
21
54
54
54
120
54
120
54
120
54
120
16
7
32 (20)
27 (7)
0
41 (3)
0
1
9
27 (28)
1
4
10 (10)
5
15
31 (13)
70 (9)
57 (0)
66 (48)
68 (2)
46 (14)
51 (1)
89 (17)
92 (16)
2 (n.d.)
12 (33)
18 (30)
14 (21)
39 (8)
7 (19)
24 (n.d.)
24 (18)
40 (20)
8
14
0
9
0
trace
1
5
trace
1
4
1
3
6
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
2
1
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
8
8
0
0
0
0
0
0
0
0
0
n.d.
7
n.d.
n.d.
n.d.
8
n.d.
n.d.
4
n.d.
n.d.
11
17
30
45
12
n.d.
n.d.
n/af
n/af
n.d.
5
7
8
15a
15a
—
—
5a
10
9
15b
15b
15c
15c
15d
15d
15e
15e
15a
15b
—
—
—
—
—
—
—
—
—
—
10
14
10
14
10
14
10
14
20
20
10
11
12
13
14
15
16
17
18
8
5
7
4
n.d.
n/af
n/af
19
20
21
15c
15d
15e
—
—
—
—
—
20
20
20
a See the Experimental part for details. Quantities used: 2.0 mmol (rac)-epoxide, 0.1 mol% Co, 0.2 mL MeOH. b Enantiomeric excess of the remaining
epoxide 15; n.d. = not determined. c Please note: in the enantioselective ring opening of (rac)-epoxides the maximum yield of enantiomerically pure
epoxide/product is 50%. d Only the conversion is reported since separation of the enantiomers of 17¢ was not achieved with chiral GC. e Because of the
generally low conversion to 16 in these reactions, the ee was not determined. f Not applicable because the substrate is a meso-compound.
catalyst displaying high activity in the ring opening of epoxides
by alcohols, whereas a mononuclear analogue displayed only low
activity.12 Previous publications have, however, mainly focused
on the scope of the nucleophile, employing primarily phenols
and other activated alcohols12,18 or making use of intramolecular
processes.19
When water is present in significant amounts, the main reaction
product 17a is produced with low ee, regardless of the catalyst
employed (Table 4, entries 1, 2 and 4). Hence, we performed the
reaction in the absence of water. In this case, methanolysis was the
only reaction observed (Entries 6–8). In contrast with the results
reported by Aerts et al.,20 we found that methanol attack takes
place selectively, but not exclusively, on the terminal carbon atom
of the epoxide. Macrocyclic catalyst 10 gives the best results with
substrate 15a, with a total of 37% conversion after 21 h (17a:17¢a =
84 : 16, entry 8),21 again suggesting a cooperative mode of action of
the two metal centres. A number of other epoxides were tested in
this procedure. Benzyl glycidyl ether (15b) shows good conversion,
however product 17b is obtained with low ee (Table 4, entries 9
and 10). Styrene oxide (15c), on the other hand, was converted to
17c with much higher ee (48%) when 10 was employed as catalyst,
whereas the use of 14 led to virtually racemic product (Table 4,
entries 11 and 12). Epichlorohydrin (15d) shows good results in
terms of conversion (about 50% using either 10 or 14, Table 4,
entries 13 and 14), however giving the product with low ee. The
secondary epoxide cyclohexene oxide (15e) was rapidly converted
to 2-methoxycyclohexanol (17e), with almost full conversion (89%
to 17e, 8% to 16e) being obtained after only 21 h using catalyst
10 (Table 4, entry 15) as opposed to a 54 h reaction time needed
with catalyst 14. However, the enantioselectivity for the formation
of 17e was low using either 10 or 14 as catalyst (Table 4, entries
15 and 16). Finally, 1,2-epoxy-5-hexene was tested in both HKR
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 593–602 | 597
©

and methanolysis, but no conversion was obtained, despite its
structural similarity to 15a.
to involve a degree of cooperativity, since corresponding mono-
metallosalens 4 and 5a and binuclear 14 show a lower conversion
rate for a number of substrates. The enantioselectivity obtained
in methanolysis with 10 is generally higher than with its parent
derivative 20 or bimetallic 14. Our future investigations will focus
on the optimization of the macrocyclic scaffold presented herein,
to increase both the reactivity as well as enantioselectivity for
various types of epoxide substrates.
The low to moderate enantioselectivity shown in the methanol-
ysis of 15a–e may be explained by the rigidity of the macrocycle,
which could prevent the two cobalt centers from approaching
each other in a conformation favorable for enantioselective
induction.6 With the objective of improving the catalytic activity
and selectivity, we investigated a macrocyclic scaffold similar to
10 but with more flexible linking units. Therefore, backbone 6 was
hydrogenated using Wilkinson’s catalyst (Scheme 6). The resulting
bis-salicylaldehyde 18 was then reacted with an equimolar amount
of (R,R)-2 to give the corresponding salen ligand 19 in high yield.
As in the case of complex 10, selective formation of the cyclic
dimer was observed. Ligand 19 was subsequently converted into
the bis-Co(III)salen macrocycle 20 as reported for 10.
Experimental
General remarks
All starting materials were purchased from commercial sources
and used without further purification unless stated otherwise.
When needed, dry solvents were obtained using a solvent pu-
rification system (SPS) from Innovative Technology. LPTS was
prepared according to a previously reported method.12 Hydro-
genation experiments were carried out using standard Schlenk
techniques. NMR measurements were carried out on a Bruker
400 MHz spectrometer at ambient temperature, and chemical
shifts are given in ppm with respect to the residual solvent peak.
Elemental analyses were performed by the Unidad de Ana´lisis
Elemental, Universidad de Santiago de Compostela (Spain).
Catalytic experiments were monitored by GC using an Agilent
Technologies 6890 N Network GC System equipped with an HP-
5 column (60.0 m ¥ 320 mm ¥ 0.25 mm). Mass spectrometric data
and chiral GC data were obtained from the Research Support
Unit of the ICIQ. FT-IR spectra (solid state) were recorded on a
Bruker Tensor 27 apparatus.
Scheme 6 Preparation of the more flexible macrocycle 20 from bis-
aldehyde 6.
Synthesis of salen ligand 3a
Unfortunately, when tested in catalysis complex 20 gave rise
to inferior results compared to 10 (Table 4, entries 17–21). In
the HKR, very low to no conversion was noted for all studied
substrates, while reasonable conversions in methanolysis were only
obtained by prolonged reaction times of up to 120 h (Table 4,
entries 18–21). Surprisingly, however, when using substrate 15b
in combination with catalyst 20, product 17b was obtained with
a significantly higher ee than using either 10 or 14 as catalysts
(Table 4, cf. entries 9, 10 and 18). These results demonstrate that
further optimization of the catalyst structure is needed to improve
the catalytic efficiency. We are currently investigating macrocyclic
scaffolds similar to 10 and 20 focusing on larger ring systems
comprising more Co(III)salen units.
Ligand 3a was obtained as a yellow oil (110 mg, 0.27 mmol, 90%)
by condensation of 3-allylsalicylaldehyde (99.3 mg, 0.61 mmol)
and (R,R)-2 (34.5 mg, 0.30 mmol) in MeOH at reflux (3 h),
followed by evaporation of the solvent and subsequent drying
1
at the oil pump. H NMR (CDCl3, 400 MHz) d 13.61 (br, 2H,
=
OH), 8.27 (s, 2H, N CH), 7.13 (dd, J = 7.4, 1.6 Hz, 2H, ArH),
7.04 (dd, J = 7.7, 1.6 Hz, 2H, ArH), 6.75 (t, J = 7.5 Hz, 2H,
ArH), 6.07–5.94 (m, 2H, allyl), 5.09–5.06 (m, 2H, allyl), 5.04 (s,
2H, allyl), 3.43–3.38 (m, 4H, allyl), 3.32 (dd, J = 5.7, 4.0 Hz, 2H,
c-hexyl CH–N), 1.98–1.83 (m, 4H, c-hexyl), 1.79–1.66 (m, 2H, c-
hexyl), 1.54–1.40 (m, 2H, c-hexyl); 13C NMR (CDCl3, 100 MHz):
d 164.9, 158.9, 136.6, 132.4, 129.6, 127.6, 118.2, 118.1, 115.5, 72.5,
33.5, 33.1, 24.2; MS (ES+): m/z = 425.2 (M + Na)+, 403.2 (M +
H)+; FT-IR (neat, cm-1): 2933, 2858, 2361, 1628, 1491, 1450, 1272,
1241, 1144, 1097, 995, 913, 845, 754, 669.
Conclusions
In conclusion, new Co(III)salen complexes 4, 5a, 9, 10, 14 and
20 have been prepared. These complexes were obtained via
intermediates having useful allyl groups that can be used in
olefin metathesis providing multinuclear metallosalen complexes.
Especially remarkable is the selective formation of cyclic dimeric
structures when bis-salicylaldehydes 6 and 18 are reacted with
(R,R)-1,2-cyclohexanediamine. These Co(III)salen complexes were
tested in the HKR and methanolysis of a variety of (rac)-epoxides.
Of particular interest is the macrocyclic Co(III)salen based catalyst
10, which proved to be the most efficient catalyst species for
some of the investigated substrates. Its mode of action is likely
Synthesis of Co(III)salen complex 4
This compound was prepared from 3a (79.0 mg, 0.20 mmol),
Co(OAc)2·4H2O (55.6 mg, 0.22 mmol) and LPTS (55.0 mg,
0.20 mmol) using a previously reported procedure.12 Yield: 86.0 mg
(0.14 mmol, 70%). 1H NMR (DMSO-d6, 400 MHz): d 8.02 (s, 2H,
=
CH N), 7.46–7.57 (m, 4H, ArH), 7.29 (d, J = 5.7 Hz, 2H, ArH),
7.11 (d, J = 7.8 Hz, 2H, ArH), 7.02 (d, J = 7.6 Hz, 2H, ArH), 6.63
(t, J = 7.4 Hz, 2H ArH), 6.29–6.41 (m, 2H, allyl), 5.26 (dd, J =
17.0, 1.9 Hz, 2H, allyl), 5.09 (dt, J = 10.0, 1.0 Hz, 2H, allyl), 3.95
(dd, J = 14.5, 7.0 Hz, 2H, allyl), 3.81 (dd, J = 14.6, 6.6 Hz, 2H,
598 | Dalton Trans., 2010, 39, 593–602
This journal is
The Royal Society of Chemistry 2010
©

allyl), 3.61 (br d, J = 8.2 Hz, 2H, c-hexyl), 3.05 (br d, J = 11.5 Hz,
2H, c-hexyl), 2.28 (s, 3H, CH3 tosylate), 2.00 (br d, J = 8.2 Hz, 2H,
c-hexyl), 1.79–1.93 (m, 2H, c-hexyl), 1.52–1.62 (m, 2H, c-hexyl);
13C NMR (DMSO-d6, 100 MHz): d 164.2, 162.8, 138.2, 136.6,
134.0, 133.3, 133.1, 128.1, 125.5, 120.0, 118.5, 115.4, 114.8, 69.6,
35.0, 29.4, 24.2, 21.3; MS (MALDI+): m/z = 1089.2 (2M - OTs)+,
935.2 (2M - 2OTs + OH)+, 459.1 (M - OTs)+; FT-IR (neat, cm-1):
2928, 2858, 2360, 1707, 1633, 1596, 1545, 1498, 1464, 1433, 1387,
1324, 1224, 1159, 1120, 1032, 1008, 905, 866, 815, 751, 679; HRMS
(MALDI+) m/z calcd. for [C26H28CoN2O2]+: 459.1477, measured:
459.1598.
162.7, 162.6, 156.2, 153.1, 152.9, 138.5, 138.3, 125.5, 125.3, 123.5,
122.6, 118.7, 118.5, 116.8, 115.9, 115.2, 115.1, 69.8, 69.6, 56.4,
56.1, 30.7, 29.4, 29.1, 24.5, 24.4, 24.1; MS (MALDI+): m/z =
1055.3 (2M - 2OAc + OH)+, 542.1 (M - AcO + Na)+, 519.1 (M -
AcO)+; FT-IR (neat, cm-1): 3215, 2932, 1635, 1581, 1551, 1462,
1372, 1322, 1255, 1164, 1123, 1101, 1044, 999, 967, 915, 828, 760,
686; HRMS (MALDI+) m/z calcd. for C30H35CoN2O6: 578.1822,
measured: 578.1833.
Synthesis of dialdehyde 6
A flame dried Schlenk vessel under argon was charged with 3-allyl-
salicylaldehyde (1a, 814.9 mg, 5.02 mmol), dry DCM (2.5 mL) and
Grubbs I catalyst (43.8 mg, 0.053 mmol, 1.0 mol%), respectively.
This mixture was allowed to stir at room temperature for 21 h,
after which a pale yellow solid had precipitated, which was filtered
and washed with MeOH (cis/trans 1 : 4). Recrystallization from
DCM–MeOH afforded pure trans-product (487.7 mg, 1.65 mmol,
66% of theoretical maximum).
Synthesis of Co(III)salen complex 5a
A mixture of 5-allyl-2-hydroxy-3-methoxybenzaldehyde (95.8 mg,
0.50 mmol) and (R,R)-2 (25.8 mg, 0.23 mmol) in EtOH (7 mL)
was refluxed for 30 min. Subsequently, the reaction mixture was
cooled down and Co(OAc)2·4H2O (65 mg, 0.26 mmol) was added,
immediately followed by addition of LPTS (70 mg, 0.25 mg).
The resulting brown solution was stirred at rt for an additional
30 min, after which it was concentrated in vacuo. The brown
suspension thus obtained was diluted with DCM and the resulting
solution was filtered over Celite, after which the filtrate was
partly concentrated. Subsequent addition of hexane led to a
brown precipitate, which was filtered, air-dried and subsequently
1H NMR (CDCl3, 400 MHz): d 11.29 (s, 2H, OH), 9.88 (s, 2H,
CHO), 7.39–7.43 (m, 4H, ArH), 6.96 (t, J = 7.6 Hz, 2H, ArH),
5.69–5.71 (m, 2H, CH CH), 3.41 (d, J = 4.8 Hz, 4H, CH); 13C
=
NMR (CDCl3, 100 MHz): d 196.7, 159.5, 137.0, 131.7, 129.4,
129.2, 120.2, 119.5, 31.7; MS (EI+): m/z = 295.1 (M - H)+, 296.1
(M+); FT-IR (neat, cm-1): 3038 (OH), 2852 (CH2), 1649 (CHO),
=
=
1618 (RCH CHR), 1442 (CH2), 981 (RCH CHR), 752 (CHAr);
Anal. calcd. for C18H16O4: C, 72.96; H, 5.44; found: C, 72.35; H,
5.19.
1
identified as 5a (159 mg, 0.23 mmol, 100%). H NMR (DMSO-
=
d6, 400 MHz): d 7.89 (s, 2H, CH N), 7.46 (d, J = 7.8 Hz, 2H,
Ar–H), 7.10 (d, J = 7.7 Hz, 2H, Ar–H), 7.04 (s, 2H, Ar–H), 6.87
(s, 2H, Ar–H), 5.93–6.04 (m, 2H, allyl), 5.11 (d, J = 17.2 Hz,
2H, allyl), 5.07 (d, J = 10.1 Hz, 2H, allyl), 4.00 (s, 6H, MeO),
3.61 (br, 2H, c-hexyl), 3.33 (br, 4H, allyl-H overlap with solvent
residual peak), 3.03 (br, 2H, c-hexyl), 2.28 (s, 3H, ArMe), 1.95–
2.05 (m, 2H, c-hexyl), 1.75–1.95 (m, 2H, c-hexyl), 1.50–1.60 (m,
2H, c-hexyl); 13C NMR (DMSO-d6, 100 MHz): d 163.7, 154.4,
153.4, 138.1, 137.4, 127.9, 125.5, 125.4, 125.3, 118.3, 117.6, 115.5,
69.4, 56.8, 38.6, 29.3, 24.1, 20.7; MS (MALDI+): m/z = 1209.4
(2M - OTs)+, 1055.4 (2M - 2OTs + OH)+, 713.2 (M + Na)+,
519.1 (M - OTs)+; FT-IR (neat, cm-1): 2935, 1634, 1459, 1324,
1259, 1162, 1112, 1012, 683; HRMS (MALDI+) m/z calcd. for
[C35H39CoN2NaO7S]+: 713.1702, measured: 713.1706.
Synthesis of dialdehyde 7
The synthesis and purification was performed as reported for 6
using 3-methoxy-5-allyl-salicylaldehyde (1b, 970.0 mg, 5.05 mmol)
and Grubbs I catalyst (111 mg, 0.152 mmol, 3 mol%). The catalyst
was added in two portions with a time interval of two hours. The
isolated yield after recrystallization was 549 mg (1.54 mmol, 61%).
1H NMR (CDCl3, 400 MHz): d 10.92 (s, 2H, OH), 9.88 (s, 2H,
CHO), 6.98 (s, 2H, ArH), 6.93 (s, 2H, ArH), 5.65–5.69 (m, 2H,
=
CH CH), 3.89 (s, 6H, OMe), 3.37 (br d, J = 4.5 Hz, 4H, CH2);
13C NMR (CDCl3, 100 MHz): d 196.4, 150.1, 148.2, 131.8, 130.5,
123.5, 120.5, 118.6, 56.3, 38.1; MS (EI+): m/z = 355.1 (M-H)+,
356.1 (M+); FT-IR (neat, cm-1): 2360, 1645, 1446, 1428, 1323,
1
4
1292, 1222, 750; Anal. calcd. for C20H20O6· H2O: C, 66.56; H,
Synthesis of Co(III)salen complex 5b
5.73; found: C, 66.68; H, 5.56.
A degassed solution of Co(OAc)2·4H2O (260 mg, 1.05 mmol) in
MeOH (3 mL) was added to a degassed solution of 5-allyl-2-
hydroxy-3-methoxybenzaldehyde (385 mg, 2.00 mmol) in MeOH
(3 mL) under argon, after which (rac)-2 (112 mg, 0.98 mmol) was
added to the mixture. The resulting solution was allowed to stir at
rt under an atmosphere of argon for 30 min. Subsequently, HOAc
(0.7 mL, 728 mg, 12.1 mmol) was added and the argon atmosphere
was replaced by air. The resulting brown solid was filtered, washed
Synthesis of oligomeric Co(III)salen complex 9
Dialdehyde 7 (38.0 mg, 0.11 mmol) was reacted with (R,R)-2
(11.5 mg, 0.10 mmol) in MeOH at rt. The solvent was slowly
evaporated until a yellow solid precipitated, which was found to
be poorly soluble in common organic solvents. This solid was then
reacted with Co(OAc)2·4H2O (25.0 mg, 0.10 mmol) in EtOH at
reflux for 2 h. After cooling to rt, LPTS (28.1 mg, 0.10 mmol)
was added to this mixture. The brown solid that resulted was
filtered and dried in vacuo to yield 69.0 mg of product. Due to
its low solubility in organic solvents, it could only be analyzed
by 1H NMR and FT-IR. Since the product is oligomeric, the 1H
NMR spectrum is generally characterized by broad peaks, except
with MeOH and dried under air to yield 4 (146 mg, 0.252 mmol,
1
=
26%). H NMR (DMSO-d6, 400 MHz): d 7.77 (s, 2H, CH N),
6.92 (s, 1H, ArH), 6.91 (s, 1H, ArH), 6.76 (s, 1H, ArH), 6.72 (s, 1H,
ArH), 5.97–6.06 (m, 2H, allyl), 5.14 (d, J = 16.9 Hz, 2H, allyl), 5.08
(d, J = 10.0 Hz, 2H, allyl), 3.98 (s, 3H, OMe), 3.93 (s, 3H, OMe),
3.31 (d, J = 6.6 Hz, 4H, CH2 allyl), 3.00 (br, 2H, c-hexyl), 2.00 (br
d, J = 7.52 Hz, 2H, c-hexyl), 1.65–1.90 (m, 4H, c-hexyl), 1.50–1.65
(m, 2H, c-hexyl), 1.31 (s, 3H, OAc); 13C (DMSO-d6, 100 MHz): d
1
for the signals of the tosylate groups which are sharp. H NMR
=
(DMSO-d6, 400 MHz): d 7.78–8.43 (br, CH N region), 7.46 (d,
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 593–602 | 599
©

J = 7.7 Hz, tosylate ArH), 6.41–7.33 (br, Ar), 7.10 (d, J = 7.7 Hz,
Ts), 5.37–5.81 (br, allyl), 3.64–3.85 (br, allyl), 2.28 (s, tosylate CH3),
0.80–2.00 (c-hexyl); FT-IR (neat, cm-1): 3222, 2933, 2860, 1633,
1496, 1461, 1396, 1318, 1182, 1123, 1049, 1034, 1010, 815, 761,
709, 682.
1440, 1391, 1360, 1266, 1202, 1172, 1097, 1038, 987, 938, 860, 827,
804, 772, 730, 712; Anal. calcd. for C62H84N4O6·Et3NHCl·3H2O:
C, 69.62; H, 9.11; N, 5.97; found: C, 69.51; H, 9.36; N, 6.16; the
presence of NEt3·HCl was confirmed by 1H NMR and could not
be removed.
Synthesis of bis-Co(III)salen complex 14
Synthesis of macrocyclic Co(III)salen complex 10
A degassed solution of Co(OAc)2·4H2O (13.3 mg, 0.053 mmol) in
MeOH (1 mL) was added to a degassed solution of 13 (23.7 mg,
0.024 mmol) in DCM (1 mL) under Ar. The resulting solution was
stirred for 30 min, after which LPTS (19.0 mg, 0.068 mmol) was
added and the reaction was allowed to proceed in air. After 18 h,
the reaction mixture was concentrated in vacuo, stripped with Et2O
Dialdehyde 6 (29.6 mg, 0.10 mmol) was dissolved in EtOH, after
which (R,R)-2 (11.4 mg, 0.10 mmol) was added. This mixture was
heated to reflux for 1 h and subsequently cooled down, followed
by addition of Co(OAc)2·4H2O (25.0 mg, 0.1 mmol). The mixture
was then heated at reflux for another 0.5 h. After cooling the
resulting brownish red suspension to rt, LPTS (28.2 mg, 0.1 mmol)
was added, after which the reaction was stirred at rt for 1 h.
The reaction mixture was then evaporated, redissolved in a small
quantity of DCM and filtered over Celite. The desired product was
precipitated from the filtrate by addition of hexane, isolated by
filtration and dried in vacuo (39 mg, 0.032 mmol, 67%). 1H NMR
and dried at the oil pump yielding a dark brown solid (34.5 mg,
1
=
100%). H NMR (DMSO-d6, 400 MHz): d 7.93 (s, 2H, CH N),
=
7.84 (s, 2H, CH N), 7.46 (d, J = 7.8 Hz, 4H, ArH tosylate), 7.41
(s, 2H, ArH), 7.10 (d, J = 7.7 Hz, 4H, ArH tosylate), 7.02 (s, 2H,
=
ArH), 7.00 (s, 2H, ArH), 6.94 (s, 2H, ArH), 5.53 (br, 2H, CH CH),
4.13 (s, 6H, OMe), 3.62 (br, 4H, allyl), 2.92 (br, 4H), 2.39 (s, 6H,
Me), 2.27 (s, 18H, tBu), 1.95–2.04 (m, 4H, c-hexyl), 1.79–1.95 (m,
4H, c-hexyl), 1.51–1.63 (m, 4H, c-hexyl). Note that part of the c-
hexyl protons are buried under the H2O peak; 13C NMR (DMSO-
d6, 100 MHz): d 164.2, 163.8, 161.9, 155.3, 153.3, 145.7, 141.6,
137.3, 136.4, 135.6, 130.2, 128.8, 128.5, 126.6, 126.0, 120.8, 119.8,
119.0, 118.3, 69.3, 69.1, 58.3, 37.1, 35.4, 33.4, 31.3, 30.8, 30.1, 29.2,
24.0, 21.9, 20.6; MS (MALDI+): m/z = 1265.5 (M - OTs)+, 1129.5
(M - 2OTs + Cl)+, 1094.5 (M - 2OTs)+; FT-IR (neat, cm-1): 3397,
2948, 2360, 1633, 1597, 1541, 1460, 1391, 1318, 1257, 1168, 1121,
1024, 1009, 911, 818, 764, 681. HRMS (MALDI+): m/z calcd. for
[C69H87Co2N4O9S]+: 1265.4852, measured: 1265.4777.
=
(DMSO-d6, 400 MHz): d 8.00 (s, 4H, CH N), 7.50 (dd, J = 7.9,
1.4 Hz, 4H, ArH), 7.46 (d, J = 8.0 Hz, 4H, ArH tosylate), 7.28
(d, J = 5.8 Hz, 4H, ArH), 7.10 (d, J = 7.8 Hz, 4H, ArH tosylate),
6.71 (m, 2H, ArH), 6.61 (t, J = 7.4 Hz, 4H, ArH), 4.07 (d, J =
=
14.6 Hz, 4H, CH CH); 3.79 (d, J = 11.8 Hz, 4H, CH2), 3.61 (d,
J = 8.1 Hz, 4H, CH2), 3.03 (d, J = 11.2 Hz, 4H, c-hexyl, CH2),
2.28 (s, 6H, CH3 tosylate), 1.94–2.06 (m, 4H, c-hexyl), 1.79–1.94
(m, 4H, c-hexyl), 1.46–1.66 (m, 8H, c-hexyl); 13C NMR (DMSO-
d6, 100 MHz): d 164.1, 163.0, 145.9, 137.4, 134.1, 133.0, 132.9,
130.0, 127.9, 125.4, 118.2, 114.6, 69.5, 30.6, 24.1, 21.0, 20.7; MS
(MALDI+): m/z = 1033.3 (M - OTs)+, 879.3 (M - 2 OTs + OH)+,
862.3 (M - 2 OTs)+; FT-IR (neat, cm-1): 2930, 2860, 2360, 1632,
1597, 1546, 1435, 1388, 1323, 1224, 1154, 1120, 1104, 1032, 1009,
981, 924, 889, 857, 815, 751, 710, 679; HRMS (MALDI+) m/z
calcd. for [C55H55Co2N4O7S]+: 1033.2450, measured: 1033.2597.
Bis-aldehyde 18
In an oven-dried two-necked flask equipped with a magnetic
stirring bar under an atmosphere of argon, dialdehyde 6 (33.7 mg,
0.11 mmol) and Rh(PPh3)3Cl (4.7 mg, 0.005 mmol) were dissolved
in 5 mL of dry THF. Subsequently, the reaction vessel was
equipped with a hydrogen balloon and the reaction was left stirring
at rt under ambient H2 pressure. After 24 h, conversion was
Synthesis of bis-salen ligand 13
(R,R)-2 was reacted with HCl gas (generated in situ from H2SO4
and NaCl) in dry Et2O at 0 C. The resulting mono-ammonium
◦
salt 11 was filtered, washed with dry Et2O, dried in vacuo and
subsequently reacted with 1c (116.2 mg, 0.50 mmol) as described
in the literature.13 The resulting ammonium-protected half-salen
(12) (0.5 mmol) was reacted in situ with dialdehyde 7 (85 mg,
0.25 mmol) in MeOH–DCM 1 : 1 in the presence of Et3N (140 mL,
0.5 mmol; added dropwise). The reaction mixture was stirred for
18 h and subsequently filtered. The filtrate was evaporated and
stripped with Et2O until a yellow foam appeared, which was dried
at the oil pump. Yield: 192 mg (0.195 mmol, 82%). H NMR
(CDCl3, 400 MHz): d 13.69 (s, 2H, OH), 13.62 (s, 2H, OH), 8.24
(s, 2H, CH N), 8.22 (s, 2H, CH N), 7.29 (s, 2H, ArH), 6.97 (s,
2H, ArH), 6.64 (s, 2H, ArH), 6.56 (s, 2H, ArH), 5.50–5.57 (br,
1
checked by H NMR (67%), after which another 0.005 mmol of
Rh(PPh3)3Cl was added to the reaction. After 48 h, the reaction
was stopped, the mixture was filtered over SiO2 (eluent: DCM)
and the filtrate was evaporated, yielding 18 (32.7 mg, 0.11 mmol,
1
96%). H NMR (CDCl3, 400 MHz) d 11.26 (s, 2H, OH), 9.87
=
(s, 2H, CH O), 7.36–7.42 (m, 4H, ArH), 6.94 (t, J = 7.5 Hz,
2H, ArH), 2.67–2.75 (m, 4H, CH2), 1.64–1.73 (m, 4H, CH2); 13
C
NMR (CDCl3, 100 MHz) d 196.8, 159.7, 137.2, 131.5, 131.1, 120.2,
119.4, 29.1, 28.9; MS (ES-) m/z 297.1 ([M - H]-); Anal. calcd.
for C18H18O4: C, 72.47; H, 6.08; found: C, 72.73; H, 6.73; FT-IR
(neat, cm-1): 2920, 2843, 1643, 1617, 1451, 1387, 1330, 1290, 1264,
1247, 1225, 1126, 1079, 1010, 970, 892, 848, 796, 758, 701.
1
=
=
=
2H, CH CH), 3.81 (s, 6H, OMe), 3.21–3.38 (m, 4H, c-hexyl),
3.19 (br, 4H, allyl), 1.79–1.98 (m, 8H, c-hexyl), 1.63–1.78 (m, 4H,
c-hexyl), 1.40–1.54 (m, 4H, c-hexyl), 1.39 (s, 18H, tBu), 1.23 (s,
18H, tBu); 13C NMR (CDCl3, 100 MHz): d 165.8, 165.7, 157.9,
149.9, 148.1, 139.9, 136.2, 130.4, 130.1, 118.1, 117.7, 126.8, 126.1,
122.3, 114.4, 72.7, 72.4, 56.0, 45.8, 38.3, 34.9, 34.0, 33.2, 33.0,
31.4, 29.4, 24.2; MS (ES-): m/z = 1015.6 (M + Cl)-, 979.6 (M -
H)-, 980.6 (M-); FT-IR (neat, cm-1): 2932, 2859, 1627, 1465, 1465,
Salen ligand 19
A mixture of 18 (11.0 mg, 0.037 mmol) and (R,R)-2 (4.2 mg,
0.037 mmol) in methanol (5 mL) was heated to reflux for 2 h. The
reaction mixture was then cooled down to rt and concentrated
in vacuo, after which the resulting suspension was filtered. The
residue, macrocyclic salen 19, was washed with a small quantity
600 | Dalton Trans., 2010, 39, 593–602
This journal is
The Royal Society of Chemistry 2010
©

◦
of cold MeOH and dried in air. Yield: 13.6 mg (0.018 mmol,
50–150 ◦C at 12.5 C min-1). Enantioselectivities were obtained
98%).1H NMR (CDCl3, 400 MHz) d 13.58 (br s, 4H, OH), 8.24
from chiral GC or HPLC studies.
=
(s, 4H, CH N), 7.09 (d, J = 7.1 Hz, 4H, ArH), 6.97 (d, J =
The regioselectivity of the products was assigned by correlation
with product mixtures synthesized using NaOH in MeOH or
H2SO4 in MeOH. The first method results in a 17a/17a¢ ratio
of 99 : 1, whereas the latter method gives 17a/17a¢ = 55 : 45, as
determined using a combination of GC and NMR techniques.22
7.3 Hz, 4H, ArH), 6.69 (t, J = 7.5 Hz, 4H, ArH), 3.25–3.35 (br,
4H, alkyl), 2.59–2.73 (br, 8H, alkyl), 1.80–1.99 (m, 8H, alkyl),
1.56–1.78 (m, 12H, alkyl), 1.40–1.54 (m, 4H, alkyl); 13C NMR
(CDCl3, 100 MHz)) d 165.0, 159.0, 132.4, 130.2, 129.2, 118.1,
118.0, 72.5, 33.1, 29.6, 29.4, 24.2; MS (ES-): m/z 751.4 (M - H)-,
773.4 (M - 2H + Na)-; HRMS (ES-) calcd. for [C48H55N4O4]-:
751.4223, found: 751.4239; FT-IR (neat, cm-1): 3853, 3734, 3675,
2971, 2360, 2341, 1735, 1622, 1559, 1446, 1377, 1262, 1076, 845,
747, 668.
Acknowledgements
Financial support from ICIQ, ICREA, Consolider Ingenio 2010
(project CSD2006-0003) and the Spanish Ministereo de Educacio´n
y Ciencias (MEC, project no. CTQ2008-02050/BQU) is acknowl-
edged. We thank Dr Noem´ı Cabello and Alba Gonzalez Arandilla
for the mass spectrometric studies. Simona Curreli and Enrique
Cequier are thanked for the chiral GC and HPLC studies.
Macrocyclic cobalt-salen 20
A degassed solution of Co(OAc)2·4H2O (15.6 mg, 0.063 mmol)
in MeOH was added to a degassed solution of ligand 19 (22 mg,
0.029 mmol) in DCM under argon. This mixture was stirred at rt
for 15 min, during which a red precipitate appeared. Subsequently,
LPTS (17.0 mg, 0.061 mmol) was added and the reaction mixture
was allowed to stir for another 30 min under air. The resulting
brown solution was concentrated in vacuo, redissolved in dry DCM
and filtered over Celite. The product was precipitated from the
filtrate by addition of hexane, filtered and dried in air, yielding
Notes and references
1 T. P. Yoon and E. N. Jacobsen, Science, 2003, 299, 1691–
1693.
2 E. M. McGarrigle and D. G. Gilheany, Chem. Rev., 2005, 105, 1563–
1602; T. Katsuki, Adv. Synth. Catal., 2002, 344, 131–147; P. G. Cozzi,
Chem. Soc. Rev., 2004, 33, 410–421.
3 E. N. Jacobsen, Acc. Chem. Res., 2000, 33, 421–431.
4 A. W. Kleij, Eur. J. Inorg. Chem., 2009, 193–205.
5 For reviews, see: A. W. Kleij, Chem.–Eur. J., 2008, 14, 10520–10529;
S. J. Wezenberg and A. W. Kleij, Angew. Chem., Int. Ed., 2008, 47,
2354–2364.
1
25.8 mg (0.021 mmol, 74%) of 20. H NMR (CDCl3, 400 MHz)
=
d 7.90–8.07 (m, 4H, CH N), 7.45 (d, J = 7.8 Hz, 4H, ArH
tosylate), 7.36–7.51 (m, 4H, ArH, partly overlaps with tosylate
peak at d 7.45), 7.18–7.37 (m, 4H, ArH), 7.09 (d, J = 7.8 Hz,
4H, ArH tosylate), 6.43–6.66 (m, 4H, ArH), 3.18 (d, J = 5.1 Hz,
4H, c-hexyl), 2.94–3.12 (m, 8H, alkyl), 2.27 (s, 6H, CH3 tosylate),
1.91–2.06 (m, 8H, alkyl), 1.73–1.91 (m, 8H, alkyl), 1.46–1.66 (m,
8H, alkyl); 13C NMR (CDCl3, 100 MHz)) d 156.1, 145.6, 136.7,
135.7, 127.1, 124.6, 119.1, 113.7, 68.6, 28.7, 28.6, 23.3, 23.1, 19.9.
Two aromatic signals are missing probably due to overlap; MS
(MALDI+): 1037.3 (M - 2OTs)+, 911.3 (M - 2OTs + OEt)+,
897.6 (M - 2OTs + OMe)+, 883.3 (M - 2OTs + OH)+, 866.3
(M - 2OTs)+; HRMS (MALDI+) calcd. for (C55H59Co2N4O7S)+:
1037.2763, found: 1037.2765; FT-IR (neat, cm-1): 3397, 2925,
2361, 1733, 1635, 1596, 1545, 1434, 1184, 1125, 1038, 1011, 950,
813, 680.
6 W. Hirahata, R. M. Thomas, E. B. Lobkovsky and G. W. Coates, J. Am.
Chem. Soc., 2008, 130, 17658–17659; C. Mazet and E. N. Jacobsen,
Angew. Chem., Int. Ed., 2008, 47, 1762–1765; R. G. Konsler, J. Karl
and E. N. Jacobsen, J. Am. Chem. Soc., 1998, 120, 10780–10781.
7 J. Park, K. Lang, K. A. Abboud and S. Hong, J. Am. Chem. Soc., 2008,
130, 16484–16485; B. M. Rossbach, K. Leopold and R. Weberskirch,
Angew. Chem., Int. Ed., 2006, 45, 1309–1312; S. J. Wezenberg, E. C.
Escudero-Ada´n, J. Benet-Buchholz and A. W. Kleij, Inorg. Chem., 2008,
47, 2925–2927.
8 X. Zheng, C. W. Jones and M. Weck, J. Am. Chem. Soc., 2007, 129,
1105–1112; M. Holbach and M. Weck, J. Org. Chem., 2006, 71, 1825–
1836; N. Madhavan and M. Weck, Adv. Synth. Catal., 2008, 350, 419–
425.
9 Please note that for the synthesis of 4 (rac)-1,2-diaminocyclohexane
was used. Racemic compound rac-4 was subjected to X-ray diffraction
studies.
10 See also: Z. Chen, M. Furutachi, Y. Kato, S. Matsunaga and M.
Shibasaki, Angew. Chem., Int. Ed., 2009, 48, 2218–2220.
11 R. M. Haak and A. W. Kleij, manuscript in preparation.
12 J. M. Ready and E. N. Jacobsen, J. Am. Chem. Soc., 2001, 123, 2687–
2688.
Catalytic experiments
General procedure for HKR reactions. Co(III)-salen complex
(0.002 mmol) was weighed in a small vial, after which a stirring
bar, rac-15 (2.0 mmol), MeCN (0.2 mL) and demineralized
water (0.55 mmol) were added. The resulting mixture was stirred
at rt, while aliquots were periodically taken from the reaction
mixture, filtered over SiO2 (eluent: MeCN) and analyzed by
13 M. Holbach, X. Zheng, C. Burd, C. W. Jones and M. Weck, J. Org.
Chem., 2006, 71, 2903–2906.
14 E. J. Campbell and S. T. Nguyen, Tetrahedron Lett., 2001, 42, 1221–
1225.
15 The procedure given in reference 14 uses 1.0 equiv. of 1.0 M HCl
in Et2O. We, found the reaction to proceed rapidly (<10 min) and
selectively when HCl gas, generated in situ by dropwise addition of
conc. H2SO4 to solid NaCl, is bubbled through a cooled solution of 9
in Et2O at rt.
◦
GC (HP-5 column, gradient from 50–150 ◦C at 12.5 C min-1).
Enantioselectivities were obtained from chiral GC or HPLC
studies.
16 Note that for all Co-complexes it was difficult to obtain satisfactory
C-elemental analyses. This is a feature that has recently also been
encountered with other oligomeric metallosalens. See: A. C. W. Leung,
J. K.-H. Hui, J. H. Chong and M. J. MacLachlan, Dalton Trans., 2009,
5199. For all Co(III)salen complexes and other new compounds, copies
General procedure for methanolysis reactions. Co(III)-salen
complex (0.002 mmol) was weighed in a small vial, after which
a stirring bar, rac-15 (2.0 mmol) and MeOH (0.2 mL) were
added. The resulting mixture was stirred at rt, while aliquots were
periodically taken from the reaction mixture, filtered over SiO2
(eluent: MeCN) and analyzed by GC (HP-5 column, gradient from
1
of H NMR and MALDI-TOF mass spectra have been deposited in
the ESI†.
17 C. Huilan, H. Deyan, L. Tian, Y. Hong, T. Wenxia, J. Chen, P. Zheng
and C. Chen, Inorg. Chem., 1996, 35, 1502–1508; C. Huilan, H. Deyan,
Y. Hong, T. Wenxia, Y. Yao and Y. Wenjuan, Acta Crystallogr., Sect.
C: Cryst. Struct. Commun., 1996, 52, 2204–2206; M. F. Summers, L. G.
This journal is
The Royal Society of Chemistry 2010
Dalton Trans., 2010, 39, 593–602 | 601
©

Marzilli, N. Bresciani-Pahor and L. Randaccio, J. Am. Chem. Soc.,
1984, 106, 4478–4485; M. Cesari, C. Neri, G. Perego, E. Perrotti and
A. Zazzetta, J. Chem. Soc. D, 1970, 276–277.
20 S. Aerts, A. Buekenhoudt, H. Weyten, I. F. J. Vankelecom and P. A.
Jacobs, Tetrahedron: Asymmetry, 2005, 16, 657–660.
21 Please note that macrocyclic 10 also shows the highest initial activity.
22 For reference data on 17a see: C. L. Francis, N. M. Williamson and
A. D. Ward, Synthesis, 2004, 2685–2691; For reference data on 17a¢
see: A. J. Bloodworth and D. J. Lapham, J. Chem. Soc., Perkin Trans.
1, 1981, 3265–3271.
18 D. A. Annis and E. N. Jacobsen, J. Am. Chem. Soc., 1999, 121, 4147–
4154.
19 M. H. Wu, K. B. Hansen and E. N. Jacobsen, Angew. Chem., Int. Ed.,
1999, 38, 2012–2014.
602 | Dalton Trans., 2010, 39, 593–602
This journal is
The Royal Society of Chemistry 2010
©