A novel type of catalysts for the asymmetric C-C bond formation based on chiral stereochemically inert cationic Co iii complexes

Article abstract of DOI:10.1007/s11172-012-0320-2

Schiff bases derived from (1R,2R)-1,2-diaminocyclohexane and 1 eq. of salycylic (or substituted salycylic) aldehyde form stereochemically inert positively charged chiral octahedral Co ⁱⁱⁱ complexes of Δ-configuration with the stereoselectivity approaching 100%. To evaluate the calatylic activity and stereoinduction of the resulting complexes with various counteranions in the outer sphere, a model reaction of trimethylsilyl cyanide addition to benzaldehyde was used. O-trimethylsilylmandelonitrile formed in the process had an enantiomeric purity up to 27%. Complexes with F ^- counterion showed high catalytic activity.

Full text of DOI:10.1007/s11172-012-0320-2

Russian Chemical Bulletin, International Edition, Vol. 61, No. 12, pp. 2252—2260, December, 2012
2252
A novel type of catalysts for the asymmetric С—С bond formation based on
chiral stereochemically inert cationic СoIII complexes
Yu. N. Belokon,a V. A. Larionov,a А. F. Mkrtchyan,b V. N. Khrustalev,a А. Nijland,c
А. S. Saghyan,b I. А. Godovikov,a А. S. Peregudov,a К. К. Babievsky,a N. S. Ikonnikov,a and V. I. Maleeva
aA. N. Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences,
28 ul. Vavilova, 119991 Moscow, Russian Federation.
Fax: +7 (499) 135 5085. Eꢀmail: yubel@ineos.ac.ru
bScientific and Production Center Armbiotechnology, National Academy of Sciences of the Republic of Armenia,
14 Gyurjyan Str., 375056 Erevan, Republic of Armenia.
Eꢀmail: bioreg@anrb.ru
cStratingh Institute for Chemistry, University of Groningen,
Nijenborgh 4, 9747 AG Groningen, The Netherlands.
Eꢀmail: aike_n@hotmail.com
Schiff bases derived from (1R,2R)ꢀ1,2ꢀdiaminocyclohexane and 1 eq. of salycylic (or substiꢀ
tuted salycylic) aldehyde form stereochemically inert positively charged chiral octahedral CoIII
complexes of ꢀconfiguration with the stereoselectivity approaching 100%. To evaluate the
calatylic activity and stereoinduction of the resulting complexes with various counteranions in
the outer sphere, a model reaction of trimethylsilyl cyanide addition to benzaldehyde was used.
Oꢀtrimethylsilylmandelonitrile formed in the process had an enantiomeric purity up to 27%.
Complexes with F– counterion showed high catalytic activity.
Key words: asymmetric catalysis, metal complex catalysis, reaction of the trimethylsilylꢀ
cyanation of aldehydes.
Synthesis of enantiomerically pure organic compounds,
particularly pharmaceutical forms, is a problem of exꢀ
treme importance in chemical technology (see Refs 1 and 2).
From the economical viewpoint, catalytic methods of
asymmetric synthesis are the most attractive (see Refs 1
and 2). For this reason, the formation of С—С bond
using asymmetric catalysts, is still one of the most popular
fields of investigations both in fundamental and in applied
chemistry. Particular attention of scientists is directed to
chiral metal complexes (Lewis acids), as asymmetric catalysts
of these reactions (see Ref. 3). A conceptional framework of
this catalysis is very simple. A strong coordinative interacꢀ
tion between a metal ion and a substrate provides the enꢀ
ergy sufficient for the further orientation of the substrate
in the position which is optimal for an asymmetric reacꢀ
tion. This orientation results from the repulsive force (nonꢀ
bonding interactions) of the chiral ligand groups. Howevꢀ
er, it is the strong interaction between a metal and an
intermediate product, that sufficiently decreases the total
rate of the reaction rendering the catalytic effect either
impossible or hindered. In this connection, during the
last decade, considerable advances have been made in the
novel area of asymmetric catalysis — an asymmetric orꢀ
ganic catalyis (see Refs 4 and 5). Catalysis of this type is
based on the use of chiral organic molecules as catalysts,
providing an orientation of a substrate in the space due to
the few bonding weak interactions with the catalyst groups.
Such catalysts are represented by positively charged chiral
quaternary ammonium salts (see Refs 6 and 7), which
have originally been obtained from Cinchona type alkaꢀ
loids (see Ref. 7). Brønsted acids (see Refs 8 and 9) and
their negatively charged conjugated bases obtained from
1,1´ꢀbiꢀ2ꢀnaphtol (BINOL) derivatives also belong to such
catalysts (see Ref. 10). Unfortunately, with rare excepꢀ
tion, the effectiveness of organic asymmetric catalysis
does not reach the average level of catalysts based on
Lewis acids, particularly, because of the limited set of
chiral frameworks of organic catalyts. Thus, the organic
framework of the Cinchona type alkaloids is difficult to
modify (see Ref. 7), and the environment of the acid group
of phosphorus acids based on BINOL needs to be stericalꢀ
ly overloaded to reach a high asymmetric induction (see
Ref. 10), which results naturally in the loss of their cataꢀ
lytic activity. Obviously, a new approach to construct
asymmetric catalysts is required that serves to develop
asymmetric catalysts devoid of disadvantages of the curꢀ
rently available catalysts. An attempt to develop this apꢀ
proach is described in the present work.
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 12, pp. 2231—2239, December, 2012.
1066ꢀ5285/12/6112ꢀ2252 © 2012 Springer Science+Business Media, Inc.

Catalysts based on cationic CoIII complexes
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
2253
Results and Discussion
A novel catalyst suggested for the asymmetric syntheꢀ
sis are positively charged СоIII complexes (Fig. 1) with
octahedral surroundings that are stereochemically inert
and possess both central and meridional chirality.
Positively charged hydrophobic metal complexes 1—3
have been obtained according to the specially developed
procedure using the assembly from building blocks
(Scheme 1). Modules constituting the chiral metal comꢀ
plex cation represent substituted salycylic aldehydes,
(1R,2R)ꢀ1,2ꢀdiaminocyclohexane and the cobalt salt.
The synthesis of compounds 1—3 comprises several
steps (see Scheme 1). At the first step the Schiff base
derived from (1R,2R)ꢀ1,2ꢀdiaminocyclohexane and salyꢀ
cylic (or substituted salycylic) aldehyde is obtained, and at
the second step, two molecules of the Schiff base and the
cobalt salt yield the target complex. At the final step, the
counterion may be changed readily for any other one by
ionꢀexchange chromatography; in our case Cl–, F– and
–
BF4 counterions were used. Complexes may be formed
as a mixture of two diastereomers, (R,R) and (R,R) (see
Fig. 1). However, the stereoselectivity of the resulting comꢀ
plexes is so high that the ꢀisomer is formed in a minor
amount, and attempts to isolate it as an individual comꢀ
pound failed.
Complexes 1—3 obtained are darkꢀbrown crystal comꢀ
pounds fairly soluble in ethyl alcohol and polar aprotic
solvents.
The structure of compounds 1—3 has been confirmed
by the elemental analysis data. Since complexes 1—3 are
diamagnetic (the configuration of the CoIII ion is 4s23d6),
they were also characterized by the NMR 1Н spectroscopy
data. Spectra of the compounds 1—3 show very similar set
of signals corresponding to the expected one for the comꢀ
plexes having С2ꢀsymmetry, and the significant differꢀ
Fig. 1. Stereochemically inert, chiral, octahedral, positively
charged СоIII complexes.
ences are observed only among the chemical shifts of
the NH2ꢀgroup protons. In the case of chlorine as a counꢀ
terion, the signal of a proton of the amino group is locatꢀ
ed in the range of 6.40—6.75 ppm, and that of the secꢀ
ond proton — in the range of 1.45—2.09 ppm. When
a fluoride anion presents a counterion in the outer comꢀ
plex sphere, the signal of the first proton of amino group
shifts to the range of 2.53—2.90 ppm, and the signal of
the other proton remains in the same interval as
for the chloride anion, i.e. between 1.45—2.09 ppm.
This implies the hydrogen bonding of fluoride anion
Scheme 1
R = H (1), OCH3 (2), CH2—CH=CH2 (3); X = HCO3– (a), Cl– (b), F (c), BF4 (d)
–
Reagents and conditions: i. CHCl3, 0 C; ii. Na3[Co(CO3)3]•3H2O, EtOH, ; iii. ionꢀexchange chromatography.

2254
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
Belokon et al.
with the complex to be more strong than that of the chloꢀ
ride anion.
tion to the metal complex cation coordinated by two proꢀ
tons of NHꢀgroups of cyclohexanediaminic moieties, and
the NH F bond lengths are in the range from 2.65 to
...
To ascribe the configuration of complexes based on
the substituted salycylic aldehydes, a NMR experiment
using NOESY procedure for the compound [ꢀ3]+Cl–
(Fig. 2) derived from (1R,2R)ꢀ1,2ꢀdiaminocyclohexane
and 3ꢀallylꢀ2ꢀhydroxybenzaldehyde has been performed.
Figure 2 shows the cross peak observed between the
proꢀRꢀproton of the amino group of the cyclohexane diꢀ
aminic moiety and the proton of the >CH—N= moiety,
which evidences their spatial proximity. Such a mutual
disposition of these moieties for the complex based on
(1R,2R)ꢀ1,2ꢀdiaminocyclohexane may occur only in the
isomer of ꢀconfiguration.
2.7 Å. The cobalt complexes in the crystal are assembled
in the infinite coupled chains due to the cationꢀanion inꢀ
teractions.
Basing on the analysis of hydrogen bond lengths in the
complexes [ꢀ1]+F– and [ꢀ3]+Cl–, one may state that
these bonds are sufficiently strong (see Refs 11—14) and,
correspondingly, the complexes per se may catalyze a wide
range of asymmetric chemical reactions, like chiral thioꢀ
ureas (see Ref. 15).
The presence of strong hydrogen bonds between the
counterions [ꢀ3]+Cl–, [ꢀ1]+F– and protons of the coꢀ
ordinated amino groups suggests that the latter have
a sufficient Brønsted acidity and the resulting cationic
complexes are metalꢀactivated organic Brønsted acids (see
Ref. 16). Correspondingly, they may be considered as a new
type of chiral organocatalysts of modular structure. To
support this proposal, these complexes were used as cataꢀ
lysts of the asymmetric С—С bond formation reactions.
Condensation of trimethylsilyl cyanide with benzaldeꢀ
hyde was chosen as a model reaction (Scheme 2). Accordꢀ
ing to the current ideas, this reaction is catalyzed both by
Bronsted and Lewis acids and bases. The acids activate
the aldehyde carbonyl group, whereas the bases activate
Absolute configurations of complexes [ꢀ3]+Cl– and
[ꢀ1]+F– were confirmed by the Xꢀray diffraction method
with their structures shown in Figures 3 and 4, correꢀ
spondingly.
Referring to Figure 3, one chlorine anion forms four
hydrogen bonds with protons of NH2ꢀgroups of both
...
complexes, the NH Cl bond lengths ranging from
3.24—3.26 Å.
In the structure of [ꢀ1]+F– the fluorine anion is bondꢀ
ed to a cation (see Fig. 4), where the coordination sphere
of the anion includes two chloroform molecules forming
additional hydrogen bonds with the fluorine ion in addiꢀ

1
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5

2
Fig. 2. Twoꢀdimensional NMR experiment using NOESY procedure for complexes [ꢀ3]+Cl–. Cross peaks between the proꢀRꢀ
protones of the amino group of the cyclohexane diamine moiety and the protone of >CH—N= moiety are encircled with oval.

Catalysts based on cationic CoIII complexes
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
2255
C(42)
C(10)
C(11)
C(12)
C(9)
C(8)
C(43)
C(44)
C(31)
C(30)
C(41)
C(40)
C(32)
C(7)
N(1)
C(64)
C(5)
C(6)
C(62)
C(39)
N(5)
C(37)
C(13)
N(2)
C(38)
C(36)
C(49)
C(4)
C(18)
C(19)
C(3)
C(45)
N(6)
O(2)
C(17)
C(63)
C(50)
O(4)
Cl(1)
C(1)
C(2)
Co(1)
N(3)
C(33)
Co(2)
C(51)
C(52)
O(1)
C(23)
C(20)
C(14)
C(35)
C(34)
N(4)
O(3)
N(8)
C(54)
C(46)
C(55)
C(22)
C(21)
C(29)
C(61)
N(7)
C(24)
C(25)
C(53)
C(48)
C(47)
C(56)
C(57)
C(15)
C(60)
C(59)
C(16)
C(28)
C(26)
C(65)
C(58)
Cl(2)
O(5)
C(27)
Fig. 3. Structure of the complex [ꢀ3]+Cl– according to the Xꢀray diffraction data.
C(5)
C(10)
C(9)
C(8)
Cl(2)
C(4)
Cl(1)
C(6)
C(3)
N(1)
C(7)
C(11)
C(27)
C(12)
Cl(3)
C(2)
C(1)
O(1)
O(2)
C(13)
N(2)
N(4)
Co(1)
F(1)
C(15)
C(26)
C(14)
N(3)
Cl(4)
C(25)
C(19)
C(16)
C(17)
C(28)
C(21)
Cl(6)
C(24)
C(20)
C(22)
C(18)
Cl(5)
C(23)
Fig. 4. Structure of the complex [ꢀ1]+F– according to the Xꢀray diffraction data.
trimethyl silyl cyanide (TMSCN) molecule, forming
pentacoordinated silicon compound wherein the cyanideꢀ
ion have an additional mobility (see Ref. 17). New comꢀ
plexes were expected to be effective catalysts for the reacꢀ
tion as Brønsted acids, and an outerꢀsphere anion would
be a base, activating TMSCN. Thus, the complexes could
be bifunctional catalysts comprising both basic and acidic
groups in a molecule.
Reaction of trimethylsilylcyanation of benzaldehyde
was performed with 2.5 mol.% of the catalyst at room
temperature. The regularities of the solvent influence on
the chemical and enantioselective yield was studied using

2256
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
Belokon et al.
Scheme 2
Table 1. The effect of solvent on the enantiomeric purity and
chemical yield of the product of reaction between TMSCN and
PhCHO, catalyzed by the cationic complex [ꢀ3]+Cl– а
Solvent
Yield (%)b
Product ее (%)c,d
CH2Cl2
CHCl3
MeCN
Toluene
THF
160
140
186
100
100
100
27
25
16
13
16
21
Reagents and conditions: catalysts are compounds 1—3 (2 mol.%),
CH2Cl2, 25 C, Ar.
Hexane
а Reaction conditions: PhCHO (26 mg, 0.246 mmol), TMSCN
(31.5 mg, 0.17 mmol), catalyst (2 mol.%, 0.00492 mmol), solꢀ
vent (1 mL), Ar atmosphere, stirring for 3 h.
b Yield according to NMR.
c Enantiomeric excess (ее) was determined using chiral GLC
analysis.
d The product of Sꢀconfiguration was formed.
–
–
X = Cl , F–, BF4
The influence of temperature on the process enantioꢀ
selectivity is insignificant, since ee varies from 23 to 27%
in the temperature range of –20 to +28 С. However the
influence of substituents in the salycylic aldehyde moiety
is highly significant (Table 2). The highest asymmetric
induction of 27% has been observed for catalysis by the
complex 3, whereas for the compounds 1 and 2 this value
was in the range of 5—13%.
Nearly all complexes obtained were highly active in
catalysis, since (S)ꢀOꢀtrimethylsilyl mandelonitrile was
formed in high and, in some cases, in quantitative yield.
The absolute configuration was determined from the comꢀ
parison of the retention times of the complex obtained
and a known enantiomer on a chiral column. It was found
R = H (1), OCH3 (2), CH2—CH=CH2 (3)
the complex [ꢀ3]+Cl– as a catalyst. A comparison of CD
spectra of [ꢀ3]+X– (Х– = HCO3–, Cl– и F–) complex
showed that the spectra did not change with the anion
(Fig. 5). This indicates that the counteranion has essenꢀ
tially no influence on the conformations of the chelate
rings in complexes in solution.
In all solvents except for chloroform, the complex has
proved to be an effective catalyst for the addition reaction of
TMSCN and PhCHO (Table 1). However, as one could
expect, in basic solvents capable of forming hydrogen bonds
with protons of the complex amino groups (MeCN and
THF), the asymmetric induction was significantly less than
in aprotic solvents. In terms of yield — enantiomeric purity,
the best solvent to perform this reaction proved to be
CH2Cl2, which was used in all the following experiments.
Table 2. The effect of substituents in the salycylic aldehyde moiꢀ
ety of complexes 1—3 and structure of counterion on the rate
and enantioselectivity of reaction of the trimethylsilylation of
benzaldehydeа
•10–8
Compꢀ
lex
Complex
cation
Counterꢀ
anion
t/h
Yield
(%)
ee
(%)b
400
[ꢀ1]+
[ꢀ1]+
[ꢀ1]+
[ꢀ2]+
[ꢀ2]+
[ꢀ3]+
[ꢀ3]+
[ꢀ3]+
[ꢀ3]+
Cl–
F–
2
2
15
3
3
3
3
3
3
57
100
61
81
100
60
100
100
6
9
6
13
7
300
200
1
3
1b
1c
1d
2b
2c
3b
3c
3c
3d
2
–
BF4
Cl–
F–
100
5
0
Cl–
F–
27
20
20
21
–100
–200
–300
F– с
–
BF4
а Reaction conditions: PhCHO (26 mg, 0.246 mmol), TMSCN
(31.5 mg, 0.17 mmol), catalyst (2 mol.%, 0.00492 mmol), solꢀ
vent CH2Cl2 (1 mL), Ar atmosphere, stirring for 3 h.
b In all cases mandelonitrile of (S)ꢀconfiguration was formed.
с Substrate:catalyst ratio of 500 : 1 was used.
365 415 465 515 565 615 665 715 /nm
Fig. 5. CD spectra for complexes [ꢀ3]+Cl– (1), [ꢀ3]+F– (2)
and [ꢀ3]+НCО3–(3).

Catalysts based on cationic CoIII complexes
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
2257
that the catalyst activity depends mainly on the anion
structure, decreasing in a series F > Cl > BF4. Such
a dependence corresponds well to the decrease in the counꢀ
teranion basicity. For example, conjugated acids of these
anions, HF и HCl, have рКа values in dimethylsulfoxide
equal to 15 and 1.8, correspondingly (see Refs 18 and 19).
Figure 6 compares kinetics of mandelonitrile accumulaꢀ
tion in a catalytic reaction in the presence of [ꢀ3]+F–
and [ꢀ3]+Cl– complexes (0.5 mol.%) in CH2Cl2 at room
temperature. Obviously, fluoride complex is several orꢀ
ders higher in catalytic activity than chloride complex.
In this case catalytic activity is observed even with a subꢀ
strate : catalyst ratio of 500 : 1. This shows that the reacꢀ
tion can proceed with a lesser amount of catalyst.
ness of the resulting complexes is on a level with the most
active catalysts of this reaction known in literature. Comꢀ
pounds 1—3 are readily modified by a series of diamines
and salycylic aldehydes. It can be expected that modest
values of asymmetric inductions observed could be further
improved, and the complexes 1—3 could be successively
used as catalysts for other reactions involving С—С bond
formation. Thus, the synthesis of chiral bifunctional cataꢀ
lysts of a fundamentally new class was successful. Accordꢀ
ingly, the problem of development of a novel class of chiral
bifunctional catalysts formulated in the introduction was
successfully solved.
Experimental
An additional activation of silicon atom by the fluoride
anion is well known (see Ref. 20). Figure 7 illustrates the
structure of a hypothetical intermediate state of the reaction
that takes into account the experimental facts observed.
Thus, positively charged СоIII complexes, formed by
Schiff monobases of salycylic aldehydes and chiral diꢀ
amines (1—3), synthesized in our group, proved to be
highly effective activators of the trimethylsilylcyanation
reaction. They act as bifunctional chiral catalysts having
both acidic and basic groups in a molecule. The effectiveꢀ
NMR spectra were recorded on «Bruker Avance 300» inꢀ
strument, working at resonance frequencies of 300.13 MHz
(1H) and 282.38 MHz (19F), NOESY experiment was performed
using a «Bruker Avance 600» instrument, working at resonance
frequencies of 600.22 MHz (1H) and 282.38 MHz (19F). Proton
chemical shifts (, ppm) were measured relative to the residual
solvent peak of a nonꢀdeutered solvent. As solvents CDCl3,
CD3ОD, D2О and (CD3)2CО were used. Optical rotations were
measured with a «Perkin—Elmer 341» polarimeter using a 5 cm
thermostated cuvette at 25 С. In the study, sorbents such as silica
gel 60 («Merck») and Sephadex LНꢀ20 («Supelco») were used.
Enantiomeric analysis of trimethylsilyl cyanohydrine ethers
was performed on a gas chromatograph 3700, equipped with
a plazma ionization detector, using a chiral stationaly phase
DPꢀTFAꢀꢀcD (32 м×0.20 мм), with hydrogen as a carrier gas.
A racemic form of each compound was used as a standard.
Elemental analysis for all the compounds obtained was perꢀ
formed in the Laboratory of Elemental Analysis of INEOS RAS.
All the solvents in use were purified according to convenient
procedures.
C (%)
100
80
[ꢀ3]+F–
60
40
Na3[Co(CO3)3] was synthesized according to the previously
reported procedure (see Ref. 21).
[ꢀ3]+Cl–
20
Synthesis of (1R,2R)ꢀ1,2ꢀdiaminocyclohexane. Resolution
of the commercially available 1,2ꢀtransꢀcyclohexanediamine
(«Aldrich») was performed according to the known procedure
(see Ref. 22). Enantiomeric purity of the compound, determined
for its bisꢀtrifluoroacetyl derivative using chiral gas chromatoꢀ
graphy, exceeded 99%.
50
100
150
200
250
300 t/nm
Fig. 6. Relative rate of Oꢀtrimethylsilylmandelonitrile formation
in catalytic reaction between TMSCN and PhCHO with comꢀ
plexes [ꢀ3]+F– and [ꢀ3]+Cl– (0.5 мол.%) in CH2Cl2 at room
temperature.
Synthesis of cationic complexes of the CoIII ion
(general procedure)
Synthesis of the Schiff monobase derived from (1R,2R)ꢀ1,2ꢀ
diaminocyclohexane and salycylic and substituted salycylic aldeꢀ
hydes. To the cooled on an ice bath solution of (1R,2R)ꢀ1,2ꢀ
diaminocyclohexane (0.5 mL, 0.475 g, 4.16 mmol) in CHCl3
(50 mL) a solution of salycylic (or substituted salycylic) aldehyde
(2 mmol) in CHCl3 (50 mL) was slowly added dropwise (over 3 h).
Then the reaction mixture was stirred for 5 h without interꢀ
rupting, and evaporated to dryness in vacuo. The resulting yelꢀ
low oily substance, presenting a mixture of monoꢀ and bisꢀSchiff
bases in a ~6 : 1 ratio, was used further without any purification.
The yield of the target mono Schiff base exceeded 80%.
Fig. 7. Structure of a hypothetic transition state for the reaction
of the trimethylsilylcyanation of benzaldehyde.

2258
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
Belokon et al.
Synthesis of chiral positively charged octahedral cobalt(III)
complexes. To the Schiff base (2 mmol) dissolved in ethanol
(12 mL), Na3[Co(CO3)3]•3H2O (0.4 g, 1.1 mmol) was added.
The reaction mixture was refluxed for 3 h, then the undisꢀ
solved residue of Na3[Co(CO3)3]•3H2O was filtered, and the
filtrate was evaporated to dryness in vacuo. ꢀIsomer of comꢀ
plex was separated from the corresponding concomitant byꢀprodꢀ
ucts and a minor amount of ꢀisomer by column chromatograꢀ
phy (SiO2, 2×10 cm, CHCl3/(CH3)2CO in a 5 : 1 ratio), with
a preliminary change of the counterion by chlorine, since in this
case the separation is more effective. The second darkꢀbrown
fraction from the column comprises the target ꢀisomer*.
Then the complex with Cl– counterion was purified by gel
chromatography on «Sephadex LHꢀ20», eluting with EtOH/
C6H6 mixture in a 1 : 1 ratio. To change counterions, 0.16 mmol
of the complex obtained was dissolved in 50% aqueous EtOH
(10 mL) and passed through a column (20×100 mm) packed with
1H NMR (300 MHz, CDCl3), : 1.25 (m, 1 H, —CH2—); 1.49
(d, 1 H, —CH2—, J = 12.7 Hz); 1.72—2.10 (m, 5 H, —CH2—,
—NH2); 2.33 (m, 1 H, —CH2—), 2.55—2.72 (m, 1 H, —CH2—);
2.78 (d, 1 H, >CH—N, J = 8.9 Hz); 3.89 (br. t, 1 H, >CH—N,
J = 9.4 Hz); 6.52 (m, 2 H, Ar, —NH2); 6.64 (d, 1 H, Ar,
J = 8.4 Hz); 6.99—7.11 (m, 1 H, Ar); 7.23 (dd, 1 H, Ar, J = 7.9
Hz, J = 1.7 Hz); 7.98 (s, 1 H, —CH=N—). 19F NMR (CDCl3),
: –69.91.
ꢀBis[2ꢀmethoxyꢀ6ꢀ{[(1R,2R)ꢀ2ꢀaminocyclohexyl]iminoꢀ
methyl}phenolate]cobalt(III) chloride [ꢀ2]+Cl–. Yield 253 mg
(43%). 1H NMR (300 MHz, (CD3)2CO), : 1.18—1.56 (m, 3 H,
—CH2—); 1.61—1.98 (m, 4 H, —CH2—); 2.22—2.40 (m, 1 H,
—CH2—); 2.70 (dd, 1 H, >CH—N, J = 11.0 Hz); 3.28 (s, 3 H,
OCH3); 3.33 (br. s, 1 H, —NH2); 3.94 (br. t, 1 H, >СH—N,
J = 10.8 Hz); 6.30 (t, 1 H, Ar, J = 7.7 Hz); 6.63 (dd, 1 H, Ar,
J = 7.7, J = 1.6 Hz); 6.75 (br. t, 1 H, —NH2, J = 9.4 Hz); 7.06
(dd, 1 H, —CH=N—, J = 7.9 Hz, J = 1.6 Hz).
anionꢀexchange resin that contains the required anion (F–, BF4
ꢀBis[2ꢀallylꢀ6ꢀ{[(1R,2R)ꢀ2ꢀaminocyclohexyl]iminomethyl}ꢀ
–
or any other) as a counterion. Removal of the solvent in vacuo
and drying in a dessicator over P2O5 gave the target product.
Catalytic trimethylsilylcyanation of benzaldehyde (general proꢀ
cedure). A solution of the catalyst (2.0 mol.%, 0.005 mmol) in
CH2Cl2 (1 mL) was placed in a Schlenk flask under argon flow,
followed by a consequtive addition of benzaldehyde (0.025 mL,
0.0261 g, 0.25 mmol) and trimethylsilyl cyanide (0.04 mL, 0.0317 g,
0.32 mmol). The reaction mixture was stirred under argon at
25 С for a certain time, then the catalyst was removed by flash
chromatography on SiO2 (ø 2 mm×10 m, eluent — CH2Cl2).
Enantiomeric excess of the product obtained was determined by
the gas chromatography on a chiral column.
phenolate]cobalt(III) chloride [ꢀ3]+Cl–. Yield 317 mg (47%).
M.p. 160—162 С, []25 –2470 (c 0.06, MeOH). Found (%):
D
C, 57.66; H, 6.56; Cl, 11.44; N, 7.89. C32H42ClCoN4O2•
•0.4CHCl3•H2O. Calculated (%): C, 57.66; H, 6.63; N, 8.3; Cl,
11.56. 1H NMR (300 MHz, CDCl3), : 1.16—1.34 (m, 1 H,
—CH2—); 1.45—2.12 (m, 6 H, —CH2—, —NH2); 2.35 (dt, 1 H,
—CH2—, J = 21.9 Hz, J = 11.0 Hz); 2.51—2.72 (m, 1 H,
—CH2—); 2.72—2.95 (m, 3 H, >CH—N, —СH2— (All)); 3.92
(t, 1 H, >CH—N, J = 9.6 Hz); 4.69 (dd, 13.5, 2 H, =CH2 (All),
J = 20.9 Hz); 5.34 (m, 1 H, =СH— (All)); 6.39—6.66 (m, 2 H,
Ar, —NH2); 6.92 (d, 1 H, Ar, J = 7.0 Hz); 7.05—7.19 (m, 1 H,
Ar); 7.96 (s, 1 H, —CH=N).
ꢀBis[2ꢀ{[(1R,2R)ꢀ2ꢀaminocyclohexyl]iminomethyl}phenolꢀ
ꢀBis[2ꢀallylꢀ6ꢀ{[(1R,2R)ꢀ2ꢀaminocyclohexyl]iminomethyl}ꢀ
phenolate]cobalt(III) fluoride [ꢀ3]+F–. Yield 267 mg (42%),
[]25D –2621.1 (c 0.038, MeOH). Found (%): C, 60.72; H, 7.03;
F, 2.34; N, 8.59. C32H42FCoN4O2•0.2CHCl3•H2O. Calculatꢀ
ed (%): C, 60.95; H, 7.02; F, 2.99; N, 8.83. 1H NMR (300 MHz,
CDCl3), : 1.20 (m, 2 H, —CH2—); 1.38—2.19 (m, 7 H,
—CH2—, —NH2); 2.53 (s, 1 H, —NH2); 2.67—2.97 (m, 3 H,
>CH—N, —СH2— (All)); 3.66 (m, 1 H, >CH—N); 4.68 (dd, 2 H,
=СH2— (All), J = 23.6 Hz, J = 13.5 Hz); 5.23—5.47 (m, 1 H,
=СH— (All)); 6.42 (t, 1 H, Ar, J = 7.4 Hz); 6.89 (d, 1 H, Ar,
J = 6.8 Hz); 7.08 (d, 1 H, Ar, J = 7.5 Hz); 7.93 (s, 1 H, —CH=N).
ꢀBis[2ꢀallylꢀ6ꢀ{[(1R,2R)ꢀ2ꢀaminocyclohexyl]iminomethyl}ꢀ
phenolate]cobalt(III) tetrafluoroboride [ꢀ3]+BF4–. Yield 270 mg
ate]cobalt(III) chloride [ꢀ1]+Cl–. Yield 278 mg (45%),
[]25 –1610 (c 0.056, MeOH). Found (%): C, 51.75; H, 5.80;
D
Cl, 16.46; Co, 9.4; N, 8.9. C26H34ClCoN4O2•0.6CHCl3•H2O.
Calculated (%): C, 51.65; H, 5.96; N, 9.06; Cl, 16.05; Co, 9.53.
1H NMR (300 MHz, CDCl3), : 1.17—1.36 (m, 1 H, —CH2—);
1.53 (dd, 1 H, —CH2—, J = 26.0 Hz, J = 12.9 Hz); 1.72—2.09
(m, 5 H, —CH2—, —NH2); 2.31 (dd, 1 H, —CH2—, J = 22.0 Hz,
J = 12.4 Hz); 2.63 (d, 1 H, —CH2—, J = 11.0 Hz); 2.78 (d, 1 H,
>СН—N, J = 9.3 Hz); 3.90 (br. t, 1 H, >CH—N, J = 9.9 Hz);
6.40—6.57 (m, 2 H, Ar,—NH2); 6.63 (d, 1 H, Ar, J = 8.4 Hz);
7.04 (ddd, 1 H, Ar, J = 8.6 Hz, J = 7.0 Hz, J = 1.7 Hz); 7.23 (dd,
1 H, Ar, J = 7.80 Hz, J = 1.60 Hz); 7.97 (s, 1 H, —CH=N).
ꢀBis[2ꢀ{[(1R,2R)ꢀ2ꢀaminocyclohexyl]iminomethyl}phenolꢀ
ate]cobalt(III) fluoride [ꢀ1]+F–. Yield 245 mg (40%),
1
(41%), []25 –2321.2 (c 0.066, MeOH). H NMR (300 MHz,
D
CDCl3) : 1.21 (m, 1 H, —CH2—); 1.48 (dd, 1 H, —CH2—,
J = 25.5 Hz, J = 12.7 Hz); 1.59—2.07 (m, 5 H, —CH2—, —NH2);
2.27 (dd, 1 H, —CH2—, J = 22.4 Hz, J = 12.0 Hz); 2.57 (d, 1 H,
—CH2—, J = 9.9 Hz); 2.68—2.91 (m, 3 H, >CH—N, —СH2—
(All)); 3.86 (br. t, 1 H, >CH—N, J = 9.4 Hz); 4.63 (m, 2 H,
=СH2— (All)); 5.19—5.40 (m, 1 H, =СH— (All)); 6.24—6.49
(m, 2 H, Ar, —NH2); 6.85 (d, 1 H, Ar, J = 6.9 Hz); 7.07 (d, 1 H,
Ar, J = 6.8 Hz); 7.92 (s, 1 H, —CH=N—). 19F NMR (CDCl3),
: –69.842.
[]25 –1850 (c 0.076, MeOH). Found (%): C, 52.38; H, 5.98;
D
F, 2.34; Co, 9.4; N, 8.92. C26H34FCoN4O2•0.7CHCl3•H2O.
Calculated (%): C, 52.22; H, 6.02; N, 9.12; F, 3.09; Co, 9.6.
1H NMR (300 MHz, CDCl3), : 1.11 (m, 1 H, —CH2—);
1.15—2.06 (m, 8 H, —CH2—, —NH2); 2.70 (d, 1 H, >CH—N,
J = 10.2 Hz); 2.90 (s, 1 H, —NН2); 3.55 (dd, 1 Н, >CH—N,
J = 14.1 Hz, J = 7.1 Hz); 6.61 (t, 1 H, Ar, J = 7.3 Hz); 6.96
(d, 1 H, Ar, J = 8.4 Hz); 7.16 (t, 1 H, Ar, J = 7.3 Hz); 7.22
(d, 1 H, Ar, J = 7.8 Hz); 7.88 (s, 1 H, —CH=N). 19F NMR
(CDCl3), : –72.405.
Xꢀray diffraction analysis of complexes [ꢀ1]+F– and
[ꢀ3]+Cl–. Unit cell parameters and reflexion intensities for the
compounds [ꢀ1]+F– and [ꢀ3]+Cl– were measured with an
automatic diffractometer «Bruker SMART APEX II CCD»
(T = 100 K, MoꢀKꢀradiation, graphite monochromator, ꢀ и
ꢀscanning). For the data obtained, the calculation of Xꢀray
absorbance was performed using a computer programme
SADABS23. Main crystal structural data are presented in Table 3.
ꢀBis[2ꢀ{((1R,2R)ꢀ2ꢀaminocyclohexyl)iminomethyl}phenolꢀ
ate]cobalt(III) tetrafluoroboride [ꢀ1]+BF4–. Yield 244 mg (42%).
* Minor ꢀisomer remains on the start and is not eluted by
the system; and the first fraction (yellow one) is an unreacted
Schiff base.

Catalysts based on cationic CoIII complexes
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
2259
Table 3. Main crystallographic data and refining parameters for the compounds [ꢀ1]+F– and
[ꢀ3]+Cl–
Parameter
[ꢀ1]+F–•2CHCl3
[ꢀ3]+Cl–•0.25CH3OH
Empirical formula
Molecular weight
T/K
C28H36N4O2FCl6Co
751.24
100
C32.25H43N4O2.25ClCo
617.09
100
Syngony
Space group
Monoclinic
P21
Monoclinic
P21
a/Å
b/Å
c/Å
/deg
12.8489(14)
8.4379(9)
15.4329(16)
90
12.4079(8)
20.2137(13)
13.0439(8)
90
/deg
91.263(2)
90
1672.8(3)
2
108.829(1)
90
3096.5(3)
4
/deg
V/Å3
Z
dc/g cm3
1.491
1.324
F(000)
/mm–1
772
1.030
1306
0.677
2max/deg
55.6
52
Number of reflexions measured
Number of independent reflections
Number of reflections with I > 2(I)
Number of refined parameters
R1 (I > 2(I))
18273
7862
7143
379
0.0627
0.1596
1.003
29676
12186
8550
734
0.0551
0.1196
1.003
wR2 (all data)
GOOF
Flack parameter
Tmin/Tmax
0.00(2)
0.747/0.820
0.02(2)
0.849/0.977
Structures of both compounds were determined by a direct methꢀ
od and refined by fullꢀmatrix least squares, with anisotropic disꢀ
placement parameters for the nonꢀhydrogen atoms. Single crysꢀ
tal of the compound [ꢀ1]+F– comprises two solvate chloroꢀ
form molecules in the independent part of the elemental cell,
one of which is disordered in two positions with the occupancies
0.65:0.35. The independent part of the unit cell of the compound
[ꢀ3]+Cl– contains 1/2 of the solvate methanol molecule. Hyꢀ
drogen atoms of amino groups and the solvate methanol moleꢀ
cule were detected objectively in the substractive Fourieꢀsyntheꢀ
sis and included in a refinement with fixed positional (model
«Rider») and thermal (Uiso(H) = 1.5Ueq(C) for CH3ꢀgroups,
Uiso(H) = 1.5Ueq(O) for OHꢀgroup and Uiso(H) = 1.2Ueq(N) for
NH2ꢀgroups) parameters. Positions of the other hydrogen atoms
were calculated geometrically and refined in the isotropic apꢀ
proximation with fixed positional (model «Rider») and thermal
(Uiso(H) = 1.5Ueq(C) for CH3ꢀgroups and Uiso(H) = 1.2Ueq(C)
for all the other groups) parameters. All calculations were perꢀ
formed using the complex of programmes SHELXTL (see Ref. 24).
Tables of atomic coordinates, bond lengths and anisotropic
temperature parameters for the compounds [ꢀ1]+F–•2CHCl3
and [ꢀ3]+Cl–•1/4CH3OH have been deposited with the Camꢀ
bridge Crystallographic Data Centre (CCDC 889589 и 820740,
correspondingly).
chemical substances and production of novel materials",
school "Development of organic synthesis methodology
and production of compounds with valuable applied propꢀ
erties" Pꢀ8).
References
1. H. J. Federsel, K. M. J. Brands, Org. Process Res. Dev., 2007,
11, 494.
2. H. Hugl, Asymmetric Synthesis: The Essentials, 2nd ed., Eds
M. Christmann, S. Bräse, WileyꢀVCH Verlag GmbH & Co.
KgaA, Weinheim, 2008, p. 307.
3. H. Yamamoto, C. H. Cheon, Chiral Lewis Acids and Brønstꢀ
ed Acids in Asymmetric Synthesis, in Catalytic Asymmetric Synꢀ
thesis, Ed. I. Ojima, John Wiley & Sons, Inc. 2010, p. 119.
4. P. M. Pihko, H. Rahaman, in Enantioselective Organocataꢀ
lyzed Reactions I, Ed. R. Mahrwald, Springer, Heidelberg,
2011, 185—207.
5. J. G. Hernandez, E. Juaristi, Chem. Commun., 2012, 48, 5396.
6. S. E. Denmark, N. D. Gould, L. M. Wolf, J. Org. Chem.,
2011, 76, 4337.
7. T. Marcelli, J. H. van Maarseveen, H. Hiemstra, Angew.
Chem., Int. Ed. (Engl.), 2006, 45, 7496.
8. A. G. Doyl, E. N. Jacobsen, Chem. Rev., 2007, 107, 5713.
9. T. Akiyama, Chem. Rev., 2007, 107, 5744.
10. M. Terada, Synthesis, 2010, 1929, as well as articles cited
therein.
The work was financially supported by the Presidium
of Russian Academy of Sciences (Programme of Fundaꢀ
mental Research "Development of synthetic routes to

2260
Russ.Chem.Bull., Int.Ed., Vol. 61, No. 12, December, 2012
Belokon et al.
11. U. Koch, P. L. A. Popelier, J. Phys. Chem., 1995, 99, 9747.
12. G. A. Jeffrey, An introduction to Hydrogen Bonding, Oxford
University Press, 1997.
13. E. Espinosa, E. Molins, C. Lecomte, Chem. Phys. Lett., 1998,
285, 170.
14. J.ꢀA. van den Berg, K. R. Seddon, Cryst. Growth Des., 2003,
3, 643—661.
15. H. Miyabe, Y. Takemoto, Bull. Chem. Soc. Jpn, 2008, 81, 785.
16. H. Yamamoto, K. Futatsugi, Angew. Chem., Int. Ed. Engl.,
2005, 44, 1924; C. H. Choen, H. Yamamoto, Chem. Comꢀ
mun., 2011, 47, 3043.
19. R. Goumont, E. Magnier, E. Kizilian, F. Terrier, J. Org.
Chem., 2003, 68, 6566, as well as articles cited therein.
20. A. D. Dilman, V. V. Levin, M. Karni, Y. Apeloig, J. Org.
Chem., 2006, 71, 7214, as well as articles cited therein.
21. H. F. Bauer, W. C. Drinkard, J. Am. Chem. Soc., 1960,
82, 5031.
22. H.ꢀJ. Schanz, M. A. Linseis, D. G. Gilheany, Tetrahedron
Asymmetry, 2003, 14, 2763.
23. G. M. Sheldrick, SADABS, v. 2.03, Bruker/Siemens Area
Detector Absorption Correction Program; Bruker AXS Inc.,
Madison, WI, 2003.
17. Y. N. Belokon, B. Green, N. S. Ikonnikov, V. S. Lariꢀ
chev, B. V. Lokshin, M. A. Moscalenko, M. North, C. Oriꢀ
zu, A. S. Peregudov, G. I. Timofeeva, Eur. J. Org. Chem.,
2000, 2655.
24. G. M. Sheldrick, Acta Crystallogr. Sect. A, 2008, 64, 112.
Received August 20, 2012;
18. F. G. Bordwell, Acc. Chem. Res., 1988, 21, 456.
in revised form Octover 25, 2012