Synthesis and physicochemical study of NiII complexes with tetradentate acyclic and macrocyclic N2S2 ligands as thiosalen analogs

Article abstract of DOI:10.1007/s11172-005-0234-3

N,N′-Polymethylenebis(thiosalicylidene)iminate and macrocyclic dithiadiazadibenzocycloalkadiene complexes of nickel(II) were synthesized and their electrochemical and spectroscopic properties were studied. Dithiadiazadibenzocycloalkadiene complexes contai

Full text of DOI:10.1007/s11172-005-0234-3

Russian Chemical Bulletin, International Edition, Vol. 54, No. 1, pp. 173—188, January, 2005
173
II
Synthesis and physicochemical study of Ni complexes
with tetradentate acyclic and macrocyclic
N S ligands as thiosalen analogs
2
2
K. P. Butin,aꢀ A. A. Moiseeva, E. K. Beloglazkina, Yu. B. Chudinov, A. A. Chizhevskii, A. V. Mironov,
a
a
a
a
a
a
b
a
B. N. Tarasevich, A. V. Lalov, and N. V. Zyk
aDepartment of Chemistry, M. V. Lomonosov Moscow State University,
Leninskie Gory, 119992 Moscow, Russian Federation.
1
Fax: +7 (095) 939 5546. Eꢀmail: butin@org.chem.msu.su
N. D. Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences,
b
4
7 Leninsky prosp., 119991 Moscow, Russian Federation
N,N´ꢀPolymethylenebis(thiosalicylidene)iminate and macrocyclic dithiadiazadibenzoꢀ
cycloalkadiene complexes of nickel(II) were synthesized and their electrochemical and specꢀ
troscopic properties were studied. Dithiadiazadibenzocycloalkadiene complexes containing
two DMSO molecules coordinated to Ni2 and two outerꢀsphere ClO4– anions were syntheꢀ
sized by the reaction of the corresponding macrocyclic ligands with Ni(ClO ) •6H O. The
+
4
2
2
structure of 3,6ꢀdithiaꢀ10,14ꢀdiazadibenzo[a,g]cyclopentadecaꢀ9,14ꢀdienylnickel(II)[bis(diꢀ
methyl sulfoxide) bisꢀperchlorate] was established by Xꢀray diffraction. The UVꢀVis spectroꢀ
scopic data are consistent with octahedral structures of diiminobis(sulfide) complexes, a squareꢀ
planar structure of the thiosalen complex, and distorted tetrahedral structures of other
diiminodithiolate complexes. The reaction of Sꢀtertꢀbutylthiosalicylaldehyde with hydrazine
hydrate afforded di(orthoꢀtertꢀbutylthiobenzal)azine. The reaction of the latter with anhydrous
NiCl produced a colored complex with the simplest molecular formula Ni(C H N S ) in
2
16 12
2 2
1
1
5% yield. Semiempirical PM3(tm) calculations and the results of UVꢀVis, ESR, and H NMR
spectroscopy demonstrate that this complex has most probably a dimeric structure, in which
two Ni centers adopt a nearly squareꢀplanar configuration. The complexes are clearly divided
into two types according to their electrochemical behavior in DMF solutions. The type 1 is
characterized by reversibility of the first reduction steps. The type 2 is characterized by irreversꢀ
ible twoꢀelectron reduction as the first step accompanied by deposition of Ni metal on the
electrode surface. Rapid electrochemically initiated alkylation occurs in the presence of variꢀ
n
n
ous alkylating agents (Bu I, Bu Br, (DmgH) CoCH ) in a solution of complex 1 in DMF.
2
3
Key words: nickel(II), thiosalen, homoleptic thiosalen complexes, diiminobis(sulfide) comꢀ
plexes, macrocyclic S,Sꢀalkylidenethiosalicylimines, electrochemistry, UVꢀVis spectroscopy,
3
,6ꢀdithiaꢀ10,14ꢀdiazadibenzo[a,g]cyclopentadecaꢀ9,14ꢀdienylnickel(II) bisꢀdimethyl sulfoxꢀ
ide bisꢀperchlorate, crystal and molecular structures.
II
II
Transition metal complexes with the tetradentate N X
Salen complexes of Co and Ni are extensively studꢀ
ied as potential catalysts for electrochemical reduction of
2
2
ligands, which are Schiff bases, viz., alkylenediamine
–
19
derivatives (N is imine nitrogen; X = O (salen),
organic halides (see, for example, the study and referꢀ
–
S (thiosalen), OR, SR, PR , etc.), have catalytic activity
ences cited therein). In recent years, electrochemical acꢀ
tivation of freons has attracted attention in connection
not only with problems of ozone layer destruction but
also with a search for ways of converting facilities from
freon production to new freon alternative products. Reꢀ
cently, Ni (Salen) has been proposed as an efficient
electrocatalyst for reduction of polyfluoroalkyl chlorides.
Salen complexes are used also as model compounds to
elucidate the mechanisms of action of particular enzymes.
For example, the Mn (Salen) complex simulates the acꢀ
tion of Mnꢀcontaining forms of superoxide dismutase,
2
in many reactions. Salen (N,N´ꢀethylenebis(salicylꢀ
ideneiminate)) and its derivatives are the most well known
tetradentate Schiff bases, which form rigid planar coordiꢀ
1
20
nation about the central metal ion. Achiral and chiral
2
1
II
salen complexes of middle and late transition metals
catalyze alkene epoxidation, including asymmetric
2
1
2
—4
5
6
epoxidation,
cyclopropanation, aziridination, sulfide
oxidation,7 the Diels—Alder reaction, activation
,8
9
1
0—12
III
of C—H bonds,
opening.1
and asymmetric epoxide ring
3—18
22
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 1, pp. 169—183, January, 2005.
066ꢀ5285/05/5401ꢀ0173 © 2005 Springer Science+Business Media, Inc.
1

1
74
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
Butin et al.
and the FeIII(Salen) complex simulates the action of enꢀ
2
3
zymes responsible for sulfoxidation of organic sulfides.
II
In the present study, we synthesized Ni complexes
with tetradentate acyclic and macrocyclic N S ligands
2
2
(
thiosalen analogs) and investigated them by electroꢀ
chemical methods. Sulfurꢀcontaining ligands have atꢀ
tracted our attention due primarily to the fact that metal
atoms in natural metalloenzymes are often coordinated
by the sulfur atom of cysteine or methionine. However,
there is no simple and unambiguous answer to the quesꢀ
tion of why metal atoms in some natural enzymes are
bound to sulfur, whereas metal atoms in other enzymes
are bound to oxygen. The present study was aimed at
II
synthesizing Ni complexes with different coordination
modes and geometry and examining the effect of the elecꢀ
tronic structure and geometry of the ligand on the redox
potential and reactivity of the reduced forms of the comꢀ
I
plexes (with metals in low oxidation states, viz., Ni
and Ni ) in alkylation at the metal atoms.
0
Complexes of thiosalen and its homologs and analogs
with metals have been studied in much less detail comꢀ
pared to salen complexes (see, for example, the studꢀ
2
4,25
ies
and references cited therein). The replacement of
two phenoxide oxygen atoms in salen by two thiolate
sulfur atoms giving rise to thiosalen should, in principle,
lead to changes in the redox properties of the correspondꢀ
ing complexes. Since sulfur is a mediumꢀstrength soft
1
2
—
2
3
—
3
4
—
4
6
—
5
8
—
6, 10 7, 11 8, 12 9, 13
n
m
2
2
2
3
3
2
3
3
base, whereas oxygen is a strong hard base, it is often
expected (see the studies2
6,27
and references cited therein)
that coordination by sulfur will stabilize metal in low
I
oxidation states (for example, Ni ), i.e., lead to a decrease
II
I
in the potential of the Ni /Ni transition. However, even
attempts to explain the known facts in terms of the simple
principle of hard and soft acids and bases are complicated
by the fact that nickel can be considered as either hard or
soft depending not only on its oxidation state but also on
the nature of the ligands.26
type ligands. Compounds 6—9 are complexes with unꢀ
charged diiminobis(sulfide) ligands and, hence, they conꢀ
tain counterions (perchlorate ions). The structure of nickel
di(orthoꢀthiobenzal)azine complex 14 is discussed below.
Results and Discussion
2
8
29
Complexes 1 (nickel thiosalen complex) and 2
II
were synthesized according to procedures developed earꢀ
lier. The previously unknown nickel diiminodithiolate
complexes containing more than three methylene groups
between the amino groups were prepared according to a
Synthesis of Ni thiosalicylimine complexes
We studied NiII thiosalicylimine complexes. Comꢀ
pounds 1—5 are complexes with charged iminothiolateꢀ
Scheme 1
Compound
n
3
4
4
6
5
8
15
4
16
6
17
8
Yield (%)
72
60
35
99
99
99

Nickel complexes, analogs of Ni(thiosalen)
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
175
Table 1. Spectroscopic data and results of elemental analysis for ligands and intermediate products of their synthesis
Compound
Found
Calculated
1H NMR (CDCl3)
IR
/cm
(
%)
–
1
δ*
C
H
N
3
,6ꢀDithiaꢀ10,13ꢀdiazadibenzo[a,g]ꢀ
65.91 5.67 9.02
66.22 5.56 8.58
2.83 (s, 4 H, CH S); 3.97 (s, 4 H, CH N);
1690
1695
2
2
cyclotetradecaꢀ9,13ꢀdiene (10)
7.20 (t, 2 H, Harom); 7.43 (d, 2 H, Harom);
.74 (d, 2 H, Harom); 8.80 (s, 2 H, HC=N)
1.84 (m, 2 H, CH ); 2.83 (t, 4 H, CH S);
7
3
,7ꢀDithiaꢀ11,14ꢀdiazadibenzo[a,h]ꢀ
66.85 5.70 8.46
67.02 5.92 8.23
2
2
cyclopentadecaꢀ10,14ꢀdiene (11)
3.70 (t, 4 H, CH N); 7.29 (t, 2 H, Harom);
2
7
7
.39 (t, 2 H, Harom); 7.50 (d, 2 H, Harom);
.68 (d, 2 H, Harom); 8.64 (s, 2 H, HC=N)
3
,6ꢀDithiaꢀ10,14ꢀdiazadibenzo[a,g]ꢀ
67.05 5.97 8.46
67.02 5.92 8.23
1.85—1.90 (t, 2 H, CH ); 2.95 (s, 4 H, CH S);
1695
1700
1700
2
2
cyclopentadecaꢀ9,14ꢀdiene (12)
3.82 (t, 4 H, CH N); 7.22 (t, 2 H, Harom);
2
7
7
.41 (t, 2 H, Harom); 7.80 (d, 2 H, Harom);
.81 (d, 2 H, Harom); 8.80 (s, 2 H, HC=N)
3
,7ꢀDithiaꢀ11,15ꢀdiazadibenzo[a,h]ꢀ
66.98 6.24 8.41
67.76 6.25 7.90
2.08 (m, 2 H, CH ); 2.95 (m+t, 2 H,
2
cyclohexaꢀ10,15ꢀdiene (13)
CH +4 H, CH S); 3.68 (t, 4 H, CH N);
2
2
2
7
7
.22 (m, 2 H, Harom); 7.30 (t, 2 H, Harom);
.95 (d, 2 H, H ); 8.80 (s, 2 H, HC=N)
arom
t
N,N´ꢀTetramethylenebis(orthoꢀtertꢀ
butylthiobenzal)diimine (15)
70.78 7.95 7.91
70.86 8.23 6.36
1.06 (s, 18 H, Bu ); 1.74 (m, 4 H, CH );
2
3.65 (m, 4 Н, CH N); 7.22 (m, 2 Н, Harom);
2
7
.42, 7.54 (both t, 2 H each, Harom); 7.51, 7.87
(
both d, 2 H each, Harom); 7.96, 8.01 (both d,
2
1
H each, Harom); 8.62, 9.01 (both s, 2 H each,
: 1, HC=N)
t
N,N´ꢀHexamethylenebis(orthoꢀtertꢀ
butylthiobenzal)diimine (16)
71.49 8.80
71.74 8.60
—
—
1.04 (s, 18 H, Bu ); 0.90—1.86 (m, 8 H, CH );
1700
2
3.70 (m, 4 Н, CH N); 7.15 (m, 2 Н, Harom);
2
7
.08, 7.27 (both t, 2 H each, Harom); 7.44, 7.68
(
both d, 2 H each, Harom); 7.82, 8.02 (both d,
2
1
H each, Harom); 8.78, 8.89 (both s, 2 H each,
: 1, HC=N)
N,N´ꢀOctamethylenebis(orthoꢀtertꢀ
butylthiobenzal)diimine (17)
Di(orthoꢀtertꢀbutylthiobenzal)ꢀ
azine (20)
72.48 8.88
72.53 8.93
68.12 7.22 7.33
68.75 7.29 7.29
—
1700
1700
—
*
All spinꢀspin coupling constants of the aromatic fragments and CH groups are 8.7—8.8 and 6.6—6.8 Hz, respectively.
2
procedure, which has been used earlier30 to synthesize
nickel thiosalen complex 1. This procedure involves the
preparation of di(orthoꢀtertꢀbutylthiobenzal)diimine folꢀ
lowed by its reaction with nickel chloride, which is acꢀ
companied by thioether bond cleavage to form a metal
thiolate complex (Scheme 1).
1
Diimines 15—17 were identified by IR and H NMR
spectroscopy and their compositions were confirmed by
elemental analysis (Table 1).
the stretching band of the imino group is shifted to longer
wavelengths (1635—1625 cm ) compared to that of the
starting ligand (to 1700 cm–1).
Ligands 15—17 can form complexes of two structural
types and can be coordinated as either tetradentate ligands
(A) (resulting complexes have a nickelꢀcontaining ring of
size n+3) or bisꢀbidentate ligands (B) (polymeric comꢀ
plexes).
1
–1
It should be noted that the H NMR spectra of comꢀ
pounds 15—17 show two signals of the HC=N group (in a
ratio of 1 : 1) and two sets of signals for aromatic protons.
This may be evidence that the bulky tertꢀbutyl group hinꢀ
ders free rotation about the Carom—C(=N) bond, so that
the H —H and H —H protons become nonequivalent
a
b
c
d
and have different chemical shifts (see Table 1).
Complexes 3—5 are poorly soluble in all solvents used
in the present study. In the IR spectra of these complexes,
In the subsequent discussion, we arbitrarily assign the
structure A to the complexes synthesized. However, an

1
76
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
Butin et al.
chlorate anions as counterions have been synthesized.
II
However, examples of complexation of Ni with macroꢀ
cyclic S,Sꢀdialkylthiosalicylimine ligands, which contain
polymethylene bridges between both the nitrogen and sulꢀ
fur atoms,* are lacking in the literature. The macrocyclic
ligands and their complexes with nickel(II) were syntheꢀ
sized according to Scheme 2.
In the first step of the synthesis, bis(orthoꢀformylꢀ
phenyl)ꢀ1,4ꢀdithiabutane (18) and bis(orthoꢀformylꢀ
phenyl)ꢀ1,5ꢀdithiapentane (19) were prepared by nucleoꢀ
philic displacement of the nitro group in orthoꢀnitroꢀ
benzaldehyde with 1,2ꢀethanedithiol or 1,3ꢀpropaneꢀ
dithiol in DMF in the presence of K CO . Then diꢀ
2
3
formylbis(sulfides) 18 and 19 were used in the reactions
with 1,2ꢀdiaminobutane or 1,3ꢀdiaminopropane. The reꢀ
actions were carried out in MeOH or EtOH at room temꢀ
perature, and anhydrous MgSO was added to bind water
4
that was eliminated. The reactions produced the correꢀ
sponding macrocyclic diiminobis(sulfides) (10—13). In
the final step of the synthesis, macrocyclic ligands (10—13)
were used for complexation with Ni(ClO ) •6H O to preꢀ
4
2
2
pare complexes 6—9 containing two DMSO molecules
(
elemental analysis data).
unambiguous choice between the structures A and B can
be made only based on the results of Xꢀray diffraction
study.
We also examined the possibility of preparing nickel
di(orthoꢀthiobenzal)azine complex 14 containing a single
bond between two nitrogen atoms, which is an analog of
We prepared diiminobis(sulfide) complexes 6—9 for
the first time by the reactions of the corresponding macꢀ
Only the CuII complex with the thiosalenꢀtype macrocyclic
diiminobis(sulfido) ligand, in which two S atoms are linked by
*
2
5
rocyclic ligands 10—13 with nickel perchlorate. Earlier,
II
Ni complexes with neutral ligands of the di(Sꢀalkylthioꢀ
the —(CH ) — bridge and two N atoms are linked by the o,o´ꢀbiꢀ
2
2
3
0
salicyl)diimine series containing iodide, bromide, or perꢀ
phenylene bridge, has been described in the literature.
Scheme 2
Compound
Yield (%)
n
m
6
13
2
7
30
2
8
57
3
9
36
3
18
38
—
2
19
56
—
3
2
3
2
3
i. K CO , DMF. ii. EtOH or MeOH, MgSO anhydrous. iii. EtOH/CH CN/DMSO.
2
3
4
3

Nickel complexes, analogs of Ni(thiosalen)
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
177
diiminodithiolate complexes 1—5. The synthesis was carꢀ
ried out using a sequence of reactions analogous to that
employed for the preparation of complexes 3—5. In the
first step, the reaction of hydrazine with tertꢀbutylthioꢀ
salicylaldehyde afforded azine 20, which was then used in
C(21)
C(20)
S(3)
a
O(9)
C(5)
N(2)
C(4)
C(3)
N(1)
the reaction with anhydrous NiCl (Scheme 3).
2
C(13)
C(14)
C(15)
C(16)
C(6)
Scheme 3
Ni
S(2)
C(19)
S(1)
C(1)
C(7)
C(8)
C(2)
S(4)
C(12)
O(10)
C(18)
C(9)
C(11)
C(17)
C(23)
C(22)
C(10)
b
Elemental analysis demonstrated that the composiꢀ
tion of complex 21 corresponds to the simplest molecular
formula Ni(C H N S ). The fact that the complex conꢀ
1
6
12
2 2
tains the metal ion coordinated by the nitrogen atom is
also confirmed by IR spectroscopy (vibrational band of
the C=N group, which is observed at 1700 cm–1 in the
spectrum of di(orthoꢀtertꢀbutylthiobenzal)azine 20, is
–
1
shifted to 1615 cm in the spectrum of complex 21; see
Tables 1 and 2)).
Crystal and molecular structure of
Fig. 1. Crystallographic molecular structure of the
[
complex 8
12•Ni(DMSO)2]2 dication (a) and the structure calculated by
+
the PM3(tm) method (b).
The structure of complex 8 was established by Xꢀray
3
1
diffraction. Earlier, a squareꢀplanar coordination of
metal atoms has been revealed for uncharged dithiolateꢀ
type thiosalen complexes of divalent metals, including
nickel complex 1. A squareꢀplanar configuration was
found also in compounds (analogous to N,Nꢀtrimethylꢀ
eneꢀbridged complex 2) containing substituents in the
benzene rings.2 However, the planes of the benzene rings
in these compounds deviate from the S N plane by ~35°.
pound 8 are given in Table 3. Selected bond lengths and
bond angle are listed in Table 4. The structure of the
2
+
[12•Ni(DMSO)2] dication (8) is shown in Fig. 1. The
central nickel atom has an octahedral coordination with
two DMSO molecules as axial ligands. The Ni atom lies
almost strictly in a plane passing through the N S atoms
4
2
2
of the macrocyclic ligand (deviation is 0.050(13) Å).
2
2
II
II
In the bisꢀsulfideꢀtype anionic Ni bis(hexylthio)salen
complex, the metal atom has an octahedral coordination
Analysis of the data on the octahedral Ni complexes
retrieved from the Cambridge Structural Database3 demꢀ
onstrated that the Ni—N (2.031(12) and 027(11) Å) and
Ni—S (2.398(4) and 2.381(4) Å) distances vary in ranges
typical of Ni—NC (2.006—2.212 Å, 41 records) and
Ni—S(Alkyl)Ar (2.344—2.575 Å, 48 records) bonds, reꢀ
spectively. The S—Ni—N, S—Ni—O, and N—Ni—O
angles are close to 90°. The S(1)—Ni—S(2) angle is slightly
smaller than 90°, whereas the N(1)—Ni—N(2) angle is,
on the contrary, slightly larger than 90°, which is, apparꢀ
ently, associated with the difference in the number of
carbon atoms in the bridges between the N atoms and the
S atoms in the chelating rings. The planes of the benzene
2
–
25
involving two Br ions as axial ligands. The Ni—N and
Ni—S bonds in this complex are longer than those in
planar complex 1, i.e., the Ni atom in the highꢀspin comꢀ
plex forms a weaker bond with N S compared to the
2
2
corresponding bond in lowꢀspin complex 1. An elongaꢀ
tion of the Ni—ligand bonds may be attributable either to
the fact that the d
x –y
2 2
orbital in the octahedral complex,
unlike that in the squareꢀplanar complex, is occupied or
to distortion of the ligand geometry in the former complex.
The crystallographic data, characteristics of Xꢀray difꢀ
fraction study, and results of structure refinement of comꢀ

1
78
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
Butin et al.
Table 2. Physicochemical and spectroscopic data for complexes 3—9
Compound
Crystal
color
M.p.
/°С
Found
Calculated
1H NMR
IR
/cm
(
%)
N
a
–1
(DMSOꢀd ), δ
6
C
H
N,N´ꢀTetramethyleneꢀ
bis(thiosalicylideneiminato)ꢀ
nickel(II) (3)
N,N´ꢀHexamethyleneꢀ
bis(thiosalicylideneiminato)ꢀ
nickel(II) (4)
N,N´ꢀOctamethyleneꢀ
bis(thiosalicylideneiminato)ꢀ
nickel(II) (5)
Paleꢀ
green
102—107 56.12 4.24 7.19 Very poorly soluble
1625
1630
1635
1625
b
(разл.)
56.13 4.24 7.27
—
Paleꢀ
green
78—83
(разл.)
57.67 4.93 6.79 Very poorly soluble
b
58.13 5.37 6.78
—
Paleꢀ
green
70—74
(разл.)
59.36 5.34 6.09 Very poorly soluble
b
59.88 5.94 6.35
—
3
,6ꢀDithiaꢀ10,13ꢀdiazadiꢀ
Red
270c
35.90 3.74 3.94 3.02 (s, 4 H, CH S);
2
benzo[a,g]cyclotetradecaꢀ9,13ꢀ
dienylnickel(II)[bis(dimethyl
sulfoxide) bisꢀperchlorate] (6)
35.69 4.08 4.08 3.86 (s, 4 H, CH N);
2
7.28 (t, 2 H, Harom);
7.46 (d, 2 H, Harom);
7
7
8
.60 (d, 2 H, Harom);
.90 (t, 2 H, Harom);
.80 (s, 2 H, HC=N)
—
3
,7ꢀDithiaꢀ11,14ꢀdiazadiꢀ
Красноꢀ
коричꢀ
невый
165c
205c
36.95 4.21 4.28
36.62 4.28 3.71
1630
1630
benzo[a,h]cyclopentadecaꢀ10,14ꢀ
dienylnickel(II)[bis(dimethyl
sulfoxide) bisꢀperchlorate] (7)
3
,6ꢀDithiaꢀ10,14ꢀdiazadiꢀ
Оливковоꢀ
green
36.95 3.99 3.54 2.15 (m, 2 H, CH );
2
benzo[a,g]cyclopentadecaꢀ9,14ꢀ
dienylnickel(II)[bis(dimethyl
sulfoxide)bisꢀperchlorate] (8)
36.62 4.28 3.71 3.05 (s, 4 H, CH S);
2
4.00 (t, 4 H, CH N);
2
7.30 (t, 2 H, Harom);
7
7
7
8
.38 (d, 2 H, Harom);
.50 (d, 2 H, Harom);
.77 (t, 2 H, Harom);
.69 (s, 2 H, HC=N)
3
,7ꢀDithiaꢀ11,15ꢀdiazadiꢀ
Оливковоꢀ
green
162c
36.89 3.66 4.81 2.20 (m, 2 H, CH );
1635
1615
2
benzo[a,h]cyclohexadecaꢀ10,15ꢀ
dienylnickel(II)[bis(dimethyl
sulfoxide)bisꢀperchlorate] (9)
37.52 3.65 4.46 2.90 (m+t, 2 H, CH +4 H,
2
CH S); 3.78 (t, 4 H,
2
CH N); 7.25 (m, 4 H,
2
Harom); 7.42 (t, 2 H, Harom);
7
8
.93 (t, 2 H, Harom);
.70 (s, 2 H, HC=N)
d
Body color 130—135c 51.42 3.00 8.53 7.90 (t, 2 H, Harom);
Ni(C H N S ) (21)
16
12
2 2
(
разл.)
51.10 3.06 8.51 7.98 (t, 2 H, Harom);
8
8
9
.32 (d, 2 H, Harom);
.24 (d, 4 H, Harom);
.12 (s, 2 H, HC=N)
a
b
c
All spinꢀspin coupling constants of the aromatic fragments and CH groups are 8.7—8.8 and 6.6—6.8 Hz, respectively.
In DMFꢀd7.
For perchlorates, the decomposition temperatures are given (without premelting).
2
d
The mass spectrum (direct probe, electron impact, 70 eV), m/z (I (%)): 299 (25%, Mꢀ29, M – N H, or M – CH NH), 135 (75%,
2
2
C H CHNS), 103 (20%, C H CN), 91 (85%, C H ), 77 (100%, C H ), 64 (40%, C H ), 51 (50%, C H ). The fragments with peak
6
4
6
5
7
7
6
5
5
4
4
3
intensities of no lower than 20% are given. No peaks at m/z > 299 are observed.
rings are inclined to the basal plane of the complex deterꢀ
mined by the N, N, S, and S atoms (43.6(5)° and 55.3(4)°)
and are located on the same side of this plane, whereas
three carbon atoms of the trimethylene group between the
nitrogen atoms are located on the opposite side of the
basal plane. This configuration is, apparently, associated
with the chair conformation of the sixꢀmembered NiN C
2 3
ring, and the difference in the inclination angles of the
benzene rings is attributable to nonlinearity of the carbon
chain between the sulfur atoms. The distances from the
C(1) and C(2) atoms to the nearest atoms of the benzene
rings (C(12) and C(19), respectively) are virtually equal

Nickel complexes, analogs of Ni(thiosalen)
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
179
Table 3. Crystallographic data, details of Xꢀray diffraction study, and characteristics of
structure refinement for compound 8
Molecular formula
C H Cl N NiO S
23 32 2 2 10 4
Molecular weight
Space group
694.03
P2 /a
1
Unit cell parameters:
a/Å
b/Å
c/Å
β/deg
V/Å3
Number of formula units
Calculated density/g cm–3
Linear absorption coefficient/cm–
Absorption correction
Tmin/Tmax
15.600(4)
11.985(1)
17.302(3)
103.72(2)
3143(2)
4
1.594
11.08
ψ scanning technique
1
0.933
F(000)
1560
θ range/deg
2—25
Scanning mode
Ranges of reflection indices
Number of measured reflections
Number of independent reflections
Rint = 0.014
ω/1.33θ
–4 < h < 18, 0 < k < 14, –20 < l < 20
2848
2179
Number of reflections with I > 3σ(I )
Number of refinement variables
Leastꢀsquares refinement against F
Goodnessꢀofꢀfit
2074
379
1.360
R factors based on reflections with I > 3σ(I )
R
0.068
0.095
0.53/–0.94
wR
Residual density (emax/emin), e/Å–3
to each other (2.87 Å), whereas the deviations of the C(1)
Xꢀray diffraction analysis (see Fig. 1). The planes of the
benzene rings are located on the same side of the basal
plane of the complex, whereas three carbon atoms of the
polymethylene fragment between the nitrogen atoms are
located on the opposite side. However, the sixꢀmembered
NiN C ring in the calculated structure, unlike that in the
and C(2) atoms from the S N plane are substantially
2
2
different (0.31 and 0.97 Å, respectively).
The crystal structure of complex 8 is shown in Fig. 2.
The [12•Ni(DMSO)2]2+ dications form layers parallel to
the (010) plane alternating with the layers formed by the
2
3
–
ClO4 anions.
The molecular geometry of the [12•Ni(DMSO)2]2+
dication calculated by the semiempirical PM3(tm)
method3 is, in principle, identical to that determined by
4
Table 4. Selected interatomic distances (d) and bond angles (ω)
for compound 8
b
a
c
Bond
d/Å
Angle
ω/deg
Ni—N(1)
Ni—N(2)
Ni—S(1)
Ni—S(2)
Ni—O(9)
Ni—O(10)
S(3)—O(9)
2.035(10)
2.029(10)
2.394(5)
2.383(4)
2.057(11)
2.130(9)
1.455(10)
N(1)—Ni—N(2)
S(1)—Ni(1)—S(2) 85.3(3)
N(2)—Ni(1)—S(1) 175.0(3)
N(2)—Ni—S(2)
S(1)—Ni—O(10)
S(2)—Ni—O(10)
N(1)—Ni—O(10)
N(1)—Ni—O(9)
97.6(4)
90.4(3)
90.0(3)
88.6(3)
91.1(4)
90.1(4)
S(4)—O(10) 1.519(9)
Fig. 2. Crystal structure of complex 8.

1
80
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
Butin et al.
experimental structure, adopts a twistꢀboat conformation
rather than an ideal chairꢀlike conformation.
A
Since crystallographic data for complexes 3—5 are
lacking, we carried out semiempirical PM3(tm) calꢀ
culations for a series of complexes analogous to
2
.5
2.0
1.5
1.0
0.5
0
II
Ni (Thiosalen), in which the number of the bridging CH
2
groups varied from 1 to 11. The calculations demonstrated
that the complexes with n = 1 or 2 are planar. However,
the improper NSSN dihedral angle gradually increases
from 0° to 23° as n increases from 1 to 6 and then remains
constant (21—22°) for larger n. Consequently, the coorꢀ
dination of the isolated molecule tends to change from
lowꢀspin squareꢀplanar to highꢀspin pseudotetrahedral
with increasing n.
2
1
300
400
500
600
700
800
900 λ/nm
Fig. 3. UVꢀVis spectra of complexes 1 (1) and 21 (2) (DMF,
UVꢀVis spectroscopic study
–3
–1
1
0
mol L ).
The spectra of macrocyclic ligand 13 in DMF soluꢀ
ε 18600 dm mol cm–1) belongs, apparently, to the
3
–1
tions show a very intense absorption maximum at 316 nm
intraligand transfer (Fig. 3).
In the spectra of complexes 3—5, no intense bands
were found at λ > 256 nm. In the spectrum of complex 2,
3
–1
–1
(
31645 cm–1) (ε 36500 dm mol cm ) associated with
intraligand charge transfer. In the spectrum of comꢀ
plex 9 containing ligand 13, this maximum is slightly
–
1
a transition at 304 nm (32894 cm ) was observed, but
this band disappeared within 5 min after dissolution in
DMF and the initially brown solution turned almost colꢀ
orless. In solutions, complexes 3—5 and, probably, comꢀ
plex 2, unlike complex 1, have, presumably, pseudoꢀ
tetrahedral or (due to coordination by the solvent)
pseudooctahedral or squareꢀpyramidal structures rather
than squareꢀplanar structures. For a pseudotetrahedral
–
1
hypsochromically shifted to 308 nm (32467 cm ) and
the extinction coefficient sharply decreases to
3
–1
1
895 dm mol cm–1 due, presumably, to additional coꢀ
ordination by the solvent (DMF) molecules. Theoretiꢀ
3
–1
–1
cally, three lowꢀintensity (ε 1—10 dm mol cm
)
3
3
spinꢀallowed d,d transitions, viz., A → T (at
2
g
2g
–1
3
3
–1
8
000—10000 cm ), A → T (13000—15000 cm ),
2g 2g
and A → T (2400—27000 cm ), and even less inꢀ
3
3
–1
2g
2g
–
1
coordination, three bands at 4000—7000 cm
1
1
tense spinꢀforbidden transitions to the D and G states
should be observed for octahedral or pseudooctahedral
highꢀspin Ni complexes. However, we failed to unamꢀ
biguously reveal absorption bands in these spectral reꢀ
gions because of very poor solubility of the complexes in
all solvents. If these bands are present, their intensity is at
most 5 dm mol cm . For complexes 6 and 7, a chargeꢀ
transfer band was observed at 311 nm (32154 cm
ε 3050 dm mol cm ) and 319 nm (31348 cm–1,
3
–1
cm–1), 7000—11000 cm
–1
–1
(
(
(
ε 10—50 dm mol
ε 100—200 dm mol cm–1), and 15000—20000 cm
3
–1
II
1
ε 250—500 dm mol–1 cm
3
–1 1
)
should be observed,
–
1
whereas bands at wavelengths shorter than 20000 cm
should be absent (except for transitions associated with
the organic ligand). These compounds should be weakly
colored compared to squareꢀplanar complexes, and this is
actually the case (see Table 2).
3
–1
–1
–
1
,
3
–1
–1
3
–1
–1
ε 1600 dm mol cm ), respectively. The octahedral
Ni(H O) (ClO4–) complex is much more readily
2
+
Electrochemical study
2
6
2
soluble, and its spectrum in DMF solutions at a concenꢀ
tration of 4•10–3 mol L shows pronounced bands at
–1
Electrochemical study of complexes 1—9 and macroꢀ
cyclic ligands 10—13 was carried out by cyclic
voltammetry (CV) and rotating disk electrode (RDE)
voltammetry in DMF on a glassyꢀcarbon electrode
(GCE), a Pt electrode and, in some cases, an Au elecꢀ
trodes in DMF solutions in the presence of 0.05 M
Bu NClO as the supporting electrolyte. The electroꢀ
–
1
3
–1
–1
6
92 nm (14450 cm , ε 7.5 dm mol cm ), 618 nm
–
1
–1
(
(
16181 cm , ε 6.25 dm3 mol cm–1), and 412 nm
24272 cm–1, ε 12.5 dm mol cm ) and a shoulder at
3
–1
–1
3
–1
–1
2
83 nm (35336 cm–1, ε 7200 dm mol cm ) correꢀ
sponding to charge transfer.
The spectrum of thiosalen complex 1 corresponds to
4
4
II
that expected for a lowꢀspin squareꢀplanar Ni comꢀ
chemical oxidation and reduction potentials are given in
Table 5.
–
1
plex. It has a singlet at 454 nm (22026 cm
,
3
–1
–1
1
(
640 dm mol cm ) and the second band at 388 nm
It should be noted that the first oxidation potential
(0.78 V relative to Ag|AgCl|KCl(satur.)) and the first reꢀ
duction potential (–1.64 V) of thiosalen N S complex 1
25773 cm–1, ε 7200 dm3 mol cm ). The third,
–1
–1
–
1
even more intense, band at 295 nm (33898 cm
,
2
2

Nickel complexes, analogs of Ni(thiosalen)
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
181
Table 5. Electrochemical oxidation potentials (EOx) and reduction potentials (ERed) of the complexes and ligands in DMF (concenꢀ
a
tration was 3•10–4 mol L ) in the presence of 0.05 M Bu NClO measured at glassyꢀcarbon (GC) and Pt electrodes relative to
–1
4
4
Ag|AgCl|KCl(satur.)
Comꢀ
pound
EOx/V
–ERed/V
GC
Pt
GC
Pt
1b
2
3
0.78/0.72; 1.23/1.16
1.04; 1.30
1.16
0.78/0.72; 1.22/1.14
1.00; 1.18
1.16
1.64/1.58; 2.21; 2.41
1.34(2e)/–0.01
1.28(2e)/0.04
1.64/1.57; 2.20; 2.40
0.91; 1.24/–0.02
4
5
1.16
1.15
1.15
1.39(2e)/0.04
1.63; 2.06
1.26
c
6
7
8
9
1
1
1
1
2
Is not oxidized
1.30(2e)/–0.05; 2.19; 2.40; 2.75
1.51(2e)/0.02; 2.08; 2.39; 2.76
1.57(2e)/0.03; 2.07; 2.34; 2.76
1.47(2e)/–0.21; 2.11; 2.39; 2.70
2.14; 2.44; 2.70
1.30(2e)/0.00
1.60(2e)/–0.15; 2.10
1.56(2e)/–0.08; 2.08
1.42(2e)/–0.32; 2.11
2.06 d
2.09
2.12
2.02
Is not oxidized
Is not oxidized
Is not oxidized
0
1.40
1.45
1.32
1.35
1.36
1.42
1.20
1.16
1
2
3
1
2.20; 2.44; 2.94
e
2.07; 2.31 ; 2.71
e
1.18; 1.40
1.15
2.10; 2.40 ; 2.79
f
0.94/0.82(1:1); 1.70; 2.47; 2.73
1.24(2e)/–0.16
0.96/0.82(1:1); 1.64
1.26(2e)/–0.16
Ni(ClO ) •
4
2
6
H O
2
a
b
c
d
e
f
The reverse peak potential is given after the slash; 2е is a twoꢀelectron process.
In MeCN; 0.05 M Bu NBF .
4
4
Since the compound is very poorly soluble, a saturated solution was used.
At an Au electrode; no pronounced peaks are observed at a Pt electrode.
The low peak.
The compound very slowly dissolves in DMF, and dissolution at 60 °C is completed in approximately 40 min. At room temperature,
the concentration of 5•10–4 mol L is achieved.
–1
are close to the oxidation potential (~0.80 V relative to
fact, the reactions of even the latter reduced form with
electrophiles or radicals occur at the nickel atom and are
accompanied by charge transfer from the ligand to the
metal. The complexes under consideration are clearly diꢀ
vided into two types according to the electrochemical
behavior in the cathodic region in DMF solutions. Comꢀ
pounds 1 and 21 (Figs 4 and 5) belong to the type 1
34
Ag|AgCl|NaCl (0.1 M) ) and reduction potential (–1.65 V
2
1
relative to a saturated calomel reference electrode ) of
the salen N O complex, i.e., the replacement of the coꢀ
2
2
ordinated phenoxide ligands with the thiolate ligands has
virtually no effect on these redox potentials. However,
electrooxidation of the S N and S O complexes occurs
2
2
2
2
by different mechanisms. Earlier, it has been demonꢀ
strated3 that in solvents having a weak or moderate coorꢀ
dination ability (for example, in MeCN), Ni(Salen) is
electropolymerized to give polyꢀNi(Salen) through the
formation of orthoꢀ and paraꢀbonds between the Ph rings
of the ligands. In a strong donor solvent DMSO, a sixꢀ
coordinate solvate complex is formed and the metal atom
4
C
1
B
A1
F
II
III
rather than the ligand is oxidized (Ni ꢀ Ni ). We have
found that electrooxidation of Ni(Thiosalen) 1 in MeCN
10 µA
(
0.1 M Bu NClO ) at a Pt electrode did not give electroꢀ
4 4
active polymeric films, i.e., oxidation of complex 1 in
MeCN also occurs at the metal atom, as opposed to the
oxygen complex.
2
0
–2
E/V
A2
G
2
D
Complexes with N ꢀ and N O ꢀtype salicylideneꢀ
4
2
2
iminate and diaminodiiminate ligands are reduced to give
E
I
II •–
either the corresponding Ni complexes or Ni (L ), i.e.,
either the metal atom or the ligand are reduced (see the
studies1 and the data considered below). In spite of this
Fig. 4. Cyclic voltammograms of complex 1 (CH CN, 0.05 M
Bu NPF , the concentration was 5•10 mol L ) (1) and comꢀ
plex 1 in the presence of an excess (5•10 mol L–1) of Bu I (2).
3
–
3
–1
4
6
,35
–2
n

1
82
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
Butin et al.
1
0 µA
These orbitals are lower in energy than the b1g orbitals
in the planar complex and than the t orbitals in the
2
pseudotetrahedral complex and are the lowestꢀenergy unꢀ
occupied πꢀtype orbitals. This interaction is stronger in
the planar configuration than in the tetrahedral one with
the result that:
1
) the average Ni—ligand bond length in the planar
complex is smaller than that in the tetrahedral complex
2
[
(
Ni—N(sp ), 1.68 (D ) and 1.96 Å (T ); Ni—S, 2.15
4h d
1
0
–1
–2
E/V
37
D ) and 2.28 Å (T )];
4h
d
2
) the reduction potential of the planar complex should
be less negative than those of the pseudotetrahedral comꢀ
plexes, which is actually the case, as can be seen from a
comparison of complex 1 with complexes 2—5.
Fig. 5. Cyclic voltammogram of complex 21 (DMF, 0.05 M
–
4
–1
Bu NClO , the concentration was 5•10 mol L with respect
4
4
to the monomeric formula).
It is more likely that reduction of the conformationꢀ
ally rigid squareꢀplanar complex is accompanied by charge
transfer from lowꢀvalent nickel to the ligand. The balance
between the interactions involving the lone electron pairs
of the ligand and the π* orbitals, which stabilize the tetraꢀ
hedral and planar coordination, respectively, can be such
that lowꢀvalent nickel in a conformationally rigid system
will be more stable than that in flexible complexes, which
acquire two electrons, resulting in deposition of nickel
metal on the electrode surface.
characterized by reversibility of the first reduction steps.
Complexes 2—9 belong to the type 2, which is characterꢀ
ized by irreversible twoꢀelectron reduction as the first
step accompanied by deposition of Ni metal on the elecꢀ
trode surface (Fig. 6) (this is the reason why the cyclic
voltammogram shows a peak of oxidative desorption of
nickel with a characteristic triangular shape in the reverse
anodic scan).
The electrochemical behavior of complexes 6—9 with
macrocyclic ligands is supposedly described by a dissociaꢀ
Structural studies of lowꢀvalent Ni complexes (NiI
0
II
and Ni ) demonstrated that nickel(0) coordinated by "soft"
tive mechanism, because Ni complexes with acyclic
1
donor ligands always has a tetrahedral configuration
tetradentate S,S´ꢀdialkylthiosalenꢀtype ligands are unꢀ
35
25
(
[
see also the study and references cited therein). The
Ni (PMe ) ] complex also has a tetrahedral structure.
stable in DMSO solutions and dissociate to give the
I
2+
diiminodisulfide ligand and Ni .
3
4
However, reduction of the structurally rigid squareꢀplaꢀ
The CV curves for reduction of Ni(ClO ) •6H O and
free ligand 13 are shown in Fig. 6, d. The reduction of
4 2 2
II
I
nar Ni complex with N ꢀcyclam to Ni leads only to a
4
slight tetrahedral distortion accompanied by elongation
of the Ni—N bonds by 0.12 Å.
Ni(ClO ) •6H O was accompanied by the appearance of
4
2
2
3
5
a twoꢀelectron peak in the reverse scan to the anodic
region, viz., an anodic peak at +0.16 V. This peak has a
characteristic sharp triangular shape and corresponds to
oxidative desorption of Ni metal from the electrode surꢀ
face. The curves shown in Fig. 6, c are very similar to the
sum of the curves of the free ligand and Ni(ClO4)2
(cf. Fig. 6, d). However, the macrocyclic S,S´ꢀpolyꢀ
methylenethiosalenꢀtype ligands do not, apparently, disꢀ
sociate in DMF solutions to any noticeable degree. This
conclusion can be drawn from the following experimenꢀ
tal data:
A qualitative correlation diagram of the energy levels
for the tetrahedral and squareꢀplanar nickel complexes,
in which binding occurs only through the lone electron
36
pairs of the ligand, is shown in Fig. 7. The squareꢀplanar
complex is characterized by the highꢀlying unoccupied
antibonding b1g orbital (d
x –y
2 2
orbital of the metal atom).
Upon reduction, this orbital becomes occupied and,
hence, the tetrahedral configuration containing additional
electrons on the t orbitals becomes more stable. Howꢀ
2
ever, the imino groups stabilize lowꢀvalence states of the
metal due to back donation (interaction between the
nonbonding levels of the metal atom (e , b ) and the π*
(1) reduction of complexes 6—9 occurs at more
negative potentials than the reduction potential of
Ni(ClO ) •6H O;
g
2g
orbital of the C=N group):
4
2
2

Nickel complexes, analogs of Ni(thiosalen)
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
183
a
b
2
µA
5
µA
1
0
–1
E/V
1
0
–1
–2
E/V
c
d
5
µA
2
1
5
µA
1
0
–1
–2
E/V
1
0
–1
–2
E/V
Fig. 6. Cyclic voltammograms (DMF, 0.05 M Bu NClO ) of complexes 3 (a), 8 (b), 9 (c) (CV curves, which were measured after the
4
4
direction of the potential scan was reversed, are indicated by dashed lines), ligand 13 (d, curve 1), and NiClO •6H O (d, curve 2). The
4
2
concentrations, mol L–1: 5•10 (3), 3•10 (8, 9), 10 (13, NiClO •6H O).
–4
–4
–3
4
2
0
(
2) in the region of oxidation of complexes 6—9, anꢀ
process, the formation of Ni during electrochemical reꢀ
duction of undissociated complexes plays, apparently, the
main role.
odic peaks are either absent or strongly shifted to positive
potentials compared to the peaks observed for oxidation
of the free ligands (see Table 5);
It is known that nickel complexes can be divided into
two groups according to the electrochemical behavior.
The first group includes complexes, which are stepwise
reduced through the intermediate formation of Ni , which
is transformed into Ni at substantially higher cathodic
(
3) dissolution of the crystals of complexes 6—9
in DMF was not accompanied by the appearance of
the green color characteristic of Ni(ClO ) •6H O soluꢀ
tions.
I
4
2
2
0
Therefore, dissolution of compounds 6—9 in DMF
leads, apparently, only to elimination of the coordinated
DMSO molecules, resulting in the transformation of comꢀ
pounds 6—9 into squareꢀplanar lowꢀspin complexes, due
to which their NMR spectra can be measured. Although
dissociation of the diiminodisulfide complexes into the
starting ligand and Ni2 in DMF cannot be completely
ruled out and this process can contribute to the overall
potentials. The second group includes complexes, which
are electrochemically reduced immediately to the correꢀ
sponding zeroꢀvalent particles in one reversible twoꢀelecꢀ
II
tron step. The first group includes Ni complexes with
II
cyclams, dmgH, etc., whereas Ni complexes with the
bpy, phen, PPh , or dppe ligands belong to the second
3
+
38
group (see the study and references cited therein). Reꢀ
0
duction affords the 18ꢀelectron bisꢀchelate Ni complex,

1
84
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
z
Butin et al.
processes occur at the ligand, whereas metal remains in
the oxidation state of 2+. Then the initial addition of
x
y
II
two electrons gives rise to an electroneutral Ni comꢀ
plex with the diradicalꢀdianionic ligand in which the
charge and spin are delocalized on two independent
Td
D4h
b1g
•
–
(
N=CH—C H —S) systems (Scheme 4). Let us then
6 4
assume that charge transfer from the ligand to metal ocꢀ
curs in the reduced complex giving rise to the correꢀ
sponding Ni0 complex with the neutral ligand. This
18ꢀelectron complex should have a tetrahedral structure,
the transformation to which is substantially hindered by
the presence of short bridges between the nitrogen and
sulfur atoms. If the coordination unit remains planar or
nearly planar, the metal—ligand bonds should be subꢀ
stantially elongated (to decrease destabilization of the
antibonding occupied b1g orbital; see Fig. 6), which is
also hindered by the bridges. Consequently, twoꢀelecꢀ
tronꢀreduced complexes with the macrocyclic N2S2
ligands would be unstable and rapidly undergo decompoꢀ
sition with metal elimination.
NiII NiI Ni0
t2
a1g
Ni0
NiI NiII
e
b2g
b1g
t2
eg
At high negative potentials, fourꢀelectron reduction of
a1g
the complexes occurs. Transfer of all electrons to the
II
"
ligand" gives rise to the Ni complex with the tetraanionic
ligand (see Scheme 4). In this complex, the acceptor abiliꢀ
ties of the π* orbitals of the C=N groups and the d orbitals
of the S atoms are substantially lower and, hence, the
e
b2g
0
•
• 2–
Ni [(L —L ) ] state (L—L is the macrocyclic ligand)
cannot be sufficiently stabilized. However, the squareꢀ
Fig. 7. Correlation diagram of the molecular orbitals of tetraꢀ
hedral and squareꢀplanar complexes containing Ni in different
oxidation states (+2, +1, and 0) in the absence of πꢀbonding.
II
4–
planar 16ꢀelectron Ni [(L—L) ] state is stabilized beꢀ
cause the initial ligand is transformed to a strongerꢀfield
ligand upon reduction. Therefore, instability of the twoꢀ
electronꢀreduced complex and stability of the fourꢀelecꢀ
tronꢀreduced complex can be explained assuming that all
electrons are initially transferred to the ligand, but the
twoꢀelectronꢀreduced complex is more readily transꢀ
which loses one chelating ligand to give a 14ꢀelectron
complex, for example:
(bpy) Ni ] + 2e– → (bpy)Ni + bpy.
II +
0
[
2
0
For macrocyclic ligands 10—13, an analogous process
cannot occur; instead, Ni metal is eliminated.
formed into the Ni state (which eliminates nickel) comꢀ
pared to the fourꢀelectron reduced complex.
For complexes 2—5, attempts to isolate the correꢀ
sponding free ligands failed. Hence, these complexes were
assigned to the electrochemical type 2 based on the apꢀ
In our laboratory, we used the voltammetric method
for the detection of fast alkylation of chelate transition
metal complexes based on electrochemical activation of
the complexes by oneꢀ or twoꢀelectron reduction due to
0
pearance of a peak of oxidative desorption of Ni followꢀ
ing the first irreversible twoꢀelectron peak in the reverse
scan (see Fig. 6, b).
which the latter become active in S 2 reactions with alkyꢀ
N
lating agents. The criteria for the occurrence of the reacꢀ
tion are as follows: (1) the appearance of a new peak in
the cathodic region corresponding to the alkylation prodꢀ
uct, (2) a strong decrease in reversibility of reduction of
the complex, and (3) the appearance of oxidation peaks
We observed an unusual electrochemical situation for
the macrocyclic complexes. In the reverse scan to the
anodic region, the peak associated with oxidative desorpꢀ
tion of nickel metal appeared in the anodic scan after the
potential of the first reduction step was achieved. Howꢀ
–
–
–
of I , Br , or (dmgH) Co(I) (for the alkylating agents
2
0
n
n
ever, the oxidation peak of Ni disappeared, when the
Bu I, Bu Br, and (dmgH) CoCH , respectively). Using
2 3
reverse scan was carried out at more negative potentials
complex 1 as an example, we demonstrated that all three
agents can be used for alkylation.
(
Fig. 6, b, c). Therefore, more reduced complexes are
0
more stable and do not eliminate Ni .
In MeCN (0.05 M Bu NPF ), complex 1 is reduced in
4
6
In the present investigation, only tentative concluꢀ
sions about the factors responsible for this phenomenon
can be drawn. Let us assume that all electroreduction
three oneꢀelectron steps at E –1.64, –2.21, and 2.41 V,
p
respectively, and gives two oneꢀelectron peaks at E 0.76
p
and 1.26 V in the anodic region (see Fig. 4, curve 1).

Nickel complexes, analogs of Ni(thiosalen)
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
185
Scheme 4
Reversibility of the first reduction peak provides evidence
in thiosalen complex 1, are directed away from rather
than toward the nickel atom:
I
that the Ni radical anion is stable and can be subjected to
n
alkylation. Upon the addition of, for example, Bu I (see
Fig. 4, curve 2), a new peak (peak F; E = –1.87 V) is
p
observed already in the first scan. The height of this peak
increases with increasing time of electrolysis at the potenꢀ
tial of the first peak of compound 1. In this case, the
height of the reverse peak A strongly decreases and the
1
peak corresponding to oxidation of the iodide ion (peak
Other possible structures of complex 21, which are
more reasonable from the point of view of the spatial
arrangement of the orbitals, can be proposed: two regions
of twoꢀcenter bonding are present in cisꢀ14, whereas threeꢀ
center bonding can occur in transꢀ14. Then, in addition
G; E = +0.5 V) appears in the anodic scan. These changes
p
are indicative of fast electrochemically activated alkyꢀ
lation.38
Study of the structure of complex 21
We failed to grow crystals of complex 21 suitable for
Xꢀray diffraction. Below we discuss the possible molecuꢀ
lar structures of 21 based on theoretical predictions, specꢀ
troscopic characteristics, and electrochemical data. Comꢀ
plex 21 can be a priori described by several structural
formulas, each corresponding to the simplest molecular
formula Ni(C H N S ).
1
6
12
2 2
Structure 21A characterized by the cisoid conformation
of the CH=N—N=CH bond system is, apparently, not
fully reasonable because the coordinated lone electron
pairs of the nitrogen atoms in this structure, unlike those

1
86
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
Butin et al.
a
b
occurs ~0.7 V more easily than that of complex 1, for
which the first reduction step is reversible. Of all the comꢀ
plexes under consideration, complex 21 is most similar in
the electrochemical behavior to compound 1. The UVꢀVis
spectrum of compound 21 (see Fig. 3) is analogous to the
spectrum of thiosalen complex 1 and differs from the
spectra of all other compounds under consideration (in
addition, complexes 1 and 21 in the solid state and in
solution are similar in color). The ESR spectrum of a
solution of complex 21 in DMF has no signals, and no
noticeable broadening of the lines is observed in the
N
N
N
Ni
N
S
S
Ni
S
S
S
c
S
Ni
Ni
N
N
N
1
H NMR spectrum. These facts confirm the presence of
II
lowꢀspin diamagnetic Ni in molecule 21. The aboveꢀ
N
mentioned calculated, electrochemical, and spectroscopic
data provide evidence that the structure of compound 21
is most likely described by a dimeric structure (21B), in
which both nickel atoms are in a squareꢀplanar coordinaꢀ
tion environment. It should be emphasized that an unamꢀ
biguous conclusion about the structure of compound 21
can be drawn only based on the results of additional inꢀ
vestigation.
However, it should be noted that the mass spectrum of
compound 21 (direct probe, electron impact, 70 eV) has
no peaks corresponding to molecular ions of the dimer or
monomer (m/z 656 or 328), and the largest mass of the
fragment ion is 299.
S
S
Fig. 8. Structures of model complexes 22A (a), 22B (b), and
2C (c) calculated by the PM3(tm) method.
2
to the polymeric structure analogous to that shown for
complexes 3—5, the following two structures can be preꢀ
sented: dimeric (21B) and monomeric (21C) structures
containing sixꢀ and sevenꢀmembered chelate rings, reꢀ
spectively, in which compound 14 serves as either the
tetraꢀ or tridentate ligand.
Semiempirical PM3(tm) calculations of model comꢀ
pounds 22A—22C, which differ from molecules 21—23 in
that they have no Ph groups, demonstrated that the N—N
bond order in model structure 22A (Fig. 8, a) analogous
to 21A is as low as 0.3, and the conjugated C=C—C=N
bond system in the sevenꢀmembered ring in model strucꢀ
ture 22C analogous to 21C is strongly twisted (dihedral
angle is 67°; Fig. 8, b). Only dinuclear structure 22B
analogous to dimer 21B (Fig. 8, c) has an "adequate"
geometry. In calculated structure 22B, both nickel atoms
have an almost planar coordination (improper SNNS diꢀ
hedral angle is 13°; N—Ni, 1.86 Å; S—Ni, 2.21 Å).
Experimental
Materials, electrodes, apparatus, and methods. Dithiosalicylꢀ
aldehyde was synthesized according to a known procedure.
Thiosalicylaldehyde was prepared according to a procedure pubꢀ
39
4
0
lished earlier. SꢀtertꢀButylthiosalicylaldehyde was synthesized
from orthoꢀnitrobenzaldehyde and tertꢀbutanethiol according to
a known procedure.3 Complexes 1 and 2 were prepared as
1
28
29
described earlier.
Dimethylformamide of highꢀpurity grade was purified by
successive refluxing and vacuum distillation over anhydrous
CuSO and P O .
The H NMR spectra were recorded on a VarianꢀVXRꢀ400
instrument at 400 MHz.
4
2
5
1
The IR spectra were measured on a URꢀ20 instrument in a
thin film or Nujol mulls.
The UVꢀVis spectra were recorded on SpecordꢀM40
(
200—900 nm) and Heliosꢀα Nicolet (200—1100 nm) instruꢀ
ments in a 0.1ꢀcm quartz cell at 20—22 °C.
The mass spectra were measured on a Varian MAT 202
instrument (direct probe, electron impact, 70 eV).
Study of compound 21 by ESR spectroscopy in a DMF
solution (concentration was ~5•10–4 mol L ) was carried out
on a Bruker EMX 6ꢀ1 radiospectrometer at room temperature
and at 77 К.
–1
Xꢀray diffraction study of compound 8 was performed on an
Enraf Nonius CADꢀ4 diffractometer (graphite monochromaꢀ
tor, λ(MoKα) = 0.71069) equipped with a scintillation detector
at room temperature. The structure was solved by direct methods
Electrochemical study (see Table 5 and Figs 4 and 5)
demonstrated that quasireversible reduction of complex 21
4
1
(CSD program ) and refined by the fullꢀmatrix leastꢀsquares

Nickel complexes, analogs of Ni(thiosalen)
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
187
method against F with anisotropic displacement parameters for
all nonhydrogen atoms using the JANA2000 program package.
The positions of the hydrogen atoms were not revealed.
ethanol and diethyl ether to prepare compound 13 as a white
powder in a yield of 1.15 g (38%), m.p. 131 °C (lit. data :
133 °C).
4
2
43
Electrochemical studies were carried out on a PIꢀ50ꢀ1.1
potentiostat. Glassyꢀcarbon (φ = 1.8 mm) or platinum (φ =
.5 mm) disks were used as the working electrodes; a 0.05 M
1,5ꢀBis(2ꢀformylphenyl)ꢀ1,5ꢀdithiapentane (19). orthoꢀNitroꢀ
benzaldehyde (3 g, 0.02 mol) and propaneꢀ1,3ꢀdithiol (1.2 g,
0.011 mol) were dissolved in DMF (25 mL) and then finely
dispersed potassium carbonate (3 g) was added. The reaction
mixture was stirred at 60 °C for two days, filtered, and concenꢀ
trated to 5 mL. Then ethanol (20 mL) was added. Cooling of the
mixture afforded a white precipitate, which was filtered off and
recrystallized from ethanol to prepare compound 19 as a white
3
Bu NClO solution in DMF or a 0.05 M Bu NPF solution in
4
4
4
6
MeCN served as the supporting electrolyte; Ag/AgCl/KCl(satur.)
was used as the reference electrode. All measurements were
carried out on stationary or rotating electrodes at a potential
scan rate varying from 20 to 500 mV s–1 under argon.
4
3
Quantumꢀchemical calculations were carried out using the
powder in a yield of 1.77 g (56%), m.p. 51—53 °C (lit. data :
54—56 °C).
33
semiempirical SCF PM3 method, which was extended by inꢀ
cluding the parameters for all firstꢀrow transition metals and
selected secondꢀ and thirdꢀrow transition metals. This extended
method (PM3(tm)) is implemented in the HyperChem program
package (HyperCube Inc., FL, USA). Geometry optimization
of the molecules was carried out with a gradient of no higher
than 10 kal (Å mol)–1 as the convergence criterion.
Di(orthoꢀtertꢀbutylthiobenzal)azine (20). orthoꢀtertꢀButylꢀ
thiobenzaldehyde (0.77 g, 4 mmol) and hydrazine hydrate
(0.03 g, 2 mmol) in benzene (60 mL) were stirred at room
temperature for 0.5 h. Then the mixture was slowly brought to
boiling and refluxed for 3 h using a Dean—Stark trap. The
solvent was removed in vacuo and the oily product was used
without additional purification.
The spectroscopic data and results of elemental analysis for
the free ligands and intermediates of their synthesis are given in
Table 1. The corresponding data for the nickel complexes are
listed in Table 2. The yields of the products are given in the
schemes presented in the text.
Synthesis of complexes
Complexes of N,N´ꢀpolymethylenebis(orthoꢀthiobenzal)diꢀ
II
imines and di(orthoꢀthiobenzal)azine with Ni (general proceꢀ
Synthesis of ligands
dure). The corresponding N,N´ꢀpolymethylenebis(orthoꢀthioꢀ
benzal)diimine 15—17 or di(orthoꢀtertꢀbutylthiobenzal)azine 20
(
1 mmol) in anhydrous ethanol (3 mL) was mixed with anꢀ
Macrocyclic Sꢀalkylthiosalicylimines 10—13 (general proceꢀ
dure). α,ωꢀBis(2ꢀformylphenyl)ꢀα,ωꢀdithiaalkane 13 or 14
hydrous nickel chloride (0.13 g, 1 mmol). The reaction mixture
was refluxed for 4 h and cooled to ~20 °C. The finely dispersed
paleꢀgreen precipitate was filtered off, washed successively with
anhydrous ethanol and anhydrous diethyl ether, and dried in air.
All complexes are poorly soluble in DMF and insoluble in other
organic solvents and water.
SꢀAlkylthiosalicylimine complexes of NiII (general procedure).
Macrocyclic diiminobis(sulfides) 10—13 were dissolved in a
minimum volume of chloroform, 1—2 drops of DMSO were
added, and the reaction mixture was mixed with an equimolar
amount of Ni(ClO ) •6H O dissolved in a minimum volume of
(
4 mmol) was suspended in methanol or ethanol (20 mL). Then
diamine (1,2ꢀdiaminoethane or 1,3ꢀdiaminopropane) (4 mmol)
was added and the reaction mixture was stirred at room temꢀ
perature for 30 min until α,ωꢀbis(2ꢀformylphenyl)ꢀα,ωꢀdiꢀ
thiaalkane was completely dissolved. Then anhydrous MgSO4
(
2 g) was added to the solution and the reaction mixture was
stirred for 4 h and filtered off. The drying agent was washed with
EtOH and the combined solutions were concentrated in vacuo
to obtain a yellow oil with a small amount of white crystals. The
mixture was dissolved in chloroform (5 mL), the crystals that
remained undissolved were filtered off, and the solution was
concentrated. The target diimine as a yellow oil was used in the
synthesis of the complexes.
4
2
2
ethanol. The reaction mixture was refluxed for 20 min. After
cooling to 0 °C, the crystals that precipitated were filtered off,
washed with a small amount of ethanol, and dried in air.
N,N´ꢀPolymethylenebis(orthoꢀtertꢀbutylthiobenzal)diimines
5—17 (general procedure). orthoꢀtertꢀButylthiobenzaldehyde
References
1
(
0.77 g, 4 mmol) was dissolved in benzene (60 mL) and the
corresponding diamine (2 mmol) was added. The reaction mixꢀ
ture was refluxed using a Dean—Stark trap for 3 h. The residual
benzene was removed in vacuo. The yellow oil, which crystalꢀ
lized with time, was used in the synthesis of the complexes
without additional purification. The complexes were prepared
in ~100% yields.
1. M. Calligaris and L. Randaccio, in Comprehensive Coordinaꢀ
tion Chemistry, Eds G. Wilkinson, and R. D. Gillard,
Pergamon, Oxford, 1987, 2, Ch. 20.
2. A. K. Jorgensen, Chem. Rev., 1989, 89, 431.
3. P. J. Pospisil, D. H. Carsten, and E. N. Jacobsen, Chem.
Eur. J., 1996, 2, 744 (and references cited therein).
4. K. P. Bryliakov and E. P. Talsi, Inorg. Chem., 2003, 42, 7258.
5. T. Fukuda and L. Katsuki, Tetrahedron, 1997, 53, 7201.
6. J. Du Bois, C. S. Tomooka, J. Hong, and E. M. Carreira,
Acc. Chem. Res., 1997, 30, 364.
7. C. Bolm and F. Bienewald, Angew. Chem., Int. Ed. Engl.,
1995, 34, 2640.
8. T. Fukuda and L. Katsuki, Tetrahedron Lett., 1997, 38, 3435.
9. S. E. Schaus, J. Branalt, and E. N. Jacobsen, J. Org. Chem.,
1998, 63, 403.
1
,4ꢀBis(2ꢀformylphenyl)ꢀ1,4ꢀdithiabutane (18). orthoꢀNitroꢀ
benzaldehyde (3 g, 0.02 mol) and ethaneꢀ1,2ꢀdithiol (1 g,
.011 mol) were dissolved in DMF (25 mL) and then finely
0
dispersed potassium carbonate (3 g, 0.02 mol) was added. The
reaction mixture was stirred at 60 °C for one day, filtered, and
concentrated to 5 mL, after which ethanol (20 mL) was added.
A yellow flocculent precipitate that formed upon cooling (m.p.
1
19 °C) was filtered off and dissolved in a minimum amount of
DMF. The product was precipitated with a 1 : 1 mixture of

1
88
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 1, January, 2005
Butin et al.
1
1
1
1
1
1
1
1
1
1
2
0. M. D. Kaufman, P. A. Greico, and D. W. Bougie, J. Am.
Chem. Soc., 1993, 115, 11648.
1. J. F. Larrow and E. N. Jacobsen, J. Am. Chem. Soc., 1994,
28. N. Goswami and D. Eichhorn, Inorg. Chem., 1999, 38, 4329.
29. M. F. Corrigan and B. O. West, Aust. J. Chem., 1976,
29, 1413.
30. B. Xie, L. Wilson, and D. Stanbury, Inorg. Chem., 2001,
40, 3606.
31. T. Yamamura and M. Tadokoro, Chem. Lett., 1989, 7, 1245.
32. F. H. Allen and O. Kennard, Chem. Des. Autom. News,
1993, 8, 1.
33. J. J. P. Stewart, J. Comput. Chem., 1989, 10, 209.
34. M. VilasꢀBoas, C. Freire, B. de Castro, P. A. Christensen,
and A. R. Hillman, Inorg. Chem., 1997, 36, 4919.
35. G. Music, J. H. Riebenspies, and M. Y. Darensbourg, Inorg.
Chem., 1998, 37, 302.
36. K. P. Butin, R. D. Rakhimov, and I. G. Il´ina, Izv. Akad.
Nauk, Ser. Khim., 1999, 71 [Russ. Chem. Bull., Int. Ed.,
1999, 49, 71].
37. K. W. Muir, Molecular Structure by Diffraction Methods,
Vol. 1, Chemical Society, London, 1973, p. 580.
38. E. K. Beloglazkina, A. A. Moiseeva, A. V. Churakov, I. S.
Orlov, N. V. Zyk, J. A. K. Howard, and K. P. Butin, Izv.
Akad. Nauk, Ser. Khim., 2002, 436 [Russ. Chem. Bull., Int.
Ed., 2002, 51, 467].
1
16, 12129.
2. K. Hamachi, R. Irie, and L. Katsuki, Tetrahedron Lett.,
996, 37, 4979.
1
3. E. N. Jacobsen, F. Kakiuchi, R. G. Konsler, J. F. Larrow,
and M. Tokunaga, Tetrahedron Lett., 1997, 38, 773.
4. M. Tokunaga, J. F. Larrow, F. Kakiuchi, and E. N. Jacobsen,
Science, 1997, 277, 936.
5. J. L. Leighton and E. N. Jacobsen, J. Org. Chem., 1996,
6
1, 389.
6. L. E. Martinez, J. L. Leighton, D. H. Carsten, and E. N.
Jacobsen, J. Am. Chem. Soc., 1995, 117, 5897.
7. R. L. Puddock and S. T. Nguyen, J. Am. Chem. Soc., 2001,
1
23, 11498.
8. J. M. Ready and E. N. Jacobsen, J. Am. Chem. Soc., 2001,
23, 403.
1
9. A. A. Isse, A. Gennaro, and E. Vianello, J. Electroanal.
Chem., 1998, 444, 241.
0. V. G. Koshechko and V. D. Pokhodenko, Izv. Akad. Nauk,
Ser. Khim., 2001, 1843 [Russ. Chem. Bull., Int. Ed., 2001,
5
0, 1929].
39. H. S. Kasmai and S. G. Tadokoro, Synthesis, 1989, 763.
40. D. Leaver, J. Smolicz, and W. H. Stafford, J. Chem. Soc.,
1962, 740.
41. L. G. Akselrud, Yu. N. Gryn, P. V. Zavalij, V. K. Pecharsky,
and V. S. Fundamensky, Abstrs of the 12th European Crystalꢀ
lographic Meeting, Moscow, 1989, p. 155.
41. V. Petricek and M. Dusek, Jana2000. The Crystallographic
Computing System, Institute of Physics, Praha, Czech Reꢀ
public, 2000.
2
1. A. A. Stepanov, M. G. Peterleitner, S. M. Peregudova, and
L. I. Denisovich, Elektrokhimiya, 2002, 38, 630 [Russ.
J. Electrochem., 2002, 38 (Engl. Transl.)].
2. D. P. Riley, Chem. Rev., 1999, 99, 2573.
3. V. K. Sivasubramanian, M. Ganesan, S. Rajagopal, and
R. Ramaraj, J. Org. Chem., 2002, 67, 1506.
2
2
2
2
2
2
4. A. Cristensen, H. S. Jensen, V. McKee, C. J. McKenzie, and
M. Munch, Inorg. Chem., 1997, 36, 6080.
5. D. S. Bohle, A. Zafar, P. A. Goodson, and D. A. Jaeger,
Inorg. Chem., 2000, 39, 712.
6. M. H. Chisholm, E. R. Davidson, J. C. Huffman, and K. B.
Quinian, J. Am. Chem. Soc., 2001, 123, 9652.
7. S. V. Kryatov, B. S. Mohanraj, V. V. Tarasov, O. P. Kryatova,
and E. V. RybakꢀAkimova, Inorg. Chem., 2002, 41, 923.
42. L. Lindoy and R.Smith, Inorg. Chem., 1981, 20, 1314.
Received March 5, 2004;
in revised form November 1, 2004