Influence of the nature of organic components in dinuclear copper(II) pivalates on the composition of thermal decomposition products

Article abstract of DOI:10.1016/j.poly.2010.02.021

Tetrabridged dinuclear complexes ((2-NH₂)C₅H₄N)₂Cu ₂(μ₂-OOCCMe₃)₄ (2·S{cyrillic}₆N{cyrillic}₆) and ((3-NH₂)C₅H₄N

Full text of DOI:10.1016/j.poly.2010.02.021

Polyhedron 29 (2010) 1734–1746
Contents lists available at ScienceDirect
Polyhedron
journal homepage: www.elsevier.com/locate/poly
Influence of the nature of organic components in dinuclear copper(II) pivalates
on the composition of thermal decomposition products
a
a
Irina Fomina a, , Zhanna Dobrokhotova , Grygory Aleksandrov , Artem Bogomyakov b,1, Matvey Fedin b,1
,
*
Alexander Dolganov c,2, Tatyana Magdesieva c,2, Vladimir Novotortsev a, Igor Eremenko a
a N.S. Kurnakov Institute of General and Inorganic Chemistry, Russian Academy of Sciences, Leninsky prosp. 31, 119991 Moscow, GSP-1, Russian Federation
b International Tomography Center, Siberian Branch, Russian Academy of Sciences, Institutskaya str. 3a, 630090 Novosibirsk, Russian Federation
c Department of Chemistry, M. V. Lomonosov Moscow State University, Leninskie Gory 1/3, 119991 Moscow, GSP-1, Russian Federation
a r t i c l e i n f o
a b s t r a c t
Article history:
Tetrabridged dinuclear complexes ((2-NH2)C5H4N)2Cu2(
l
2-OOCCMe3)4 (2ꢀC6H6) and ((3-NH2)C5H4N)2Cu2
(
l
2-OOCCMe3)4 (3) and the cocrystallization product ((4-NMe2)C5H4N)2Cu(
g
2–OOCCMe3)2ꢀ2((4-
Received 8 October 2009
Accepted 11 February 2010
Available online 17 February 2010
NMe2)C5H4N)2Cu2(l2-OOCCMe3)4 (4) were synthesized by the reaction of the polymer [Cu(OOCCMe3)2]n
(1) with aminopyridine ligands (L) of different nature (Cu: L = 1:1) in C6H6. The solid-state thermal
decomposition of these compounds was studied by differential scanning calorimetry and thermogravi-
metry, and their electrochemical behavior was investigated by cyclic voltammetry. All newly synthesized
complexes were studied by X-ray diffraction, the magnetic properties of the complexes were investi-
gated, and ESR measurements were performed.
Keywords:
Copper pivalates
Aminopyridine ligands
Synthesis
X-ray diffraction study
Magnetic properties
Electrochemical behavior
Solid-state thermal decomposition
Ó 2010 Elsevier Ltd. All rights reserved.
1. Introduction
A rise of the temperature leads to a gradual elimination of organic
fragments and the clustering of the metal core up to the formation
In recent years, the use of metal complexes as molecular precur-
sors for the preparation of nanosized metal particles [1,2] has at-
tracted great interest because of the unique physical and catalytic
properties [3–5] of such metallic materials. The solid-phase ther-
molysis of solutions of complex molecular precursors in suitable
matrices [6–10] is used for the synthesis of metals as often as the
gas-phase decomposition [1,11] or, for example, the sol–gel tech-
nologies [2]. However, the solid-state thermal decomposition of
complexes containing metal cations in the absence of external
reducing agents differs in that the organic components of precur-
sors play an important role in this process. For instance, it is known
that metals can easily be generated in the final step of thermal
decomposition using complexes with ligands having strong reduc-
ing properties. This approach has recently been applied to the ther-
molysis of methylhydrazine 3d-metal complexes, which afforded
metals or even alloys (in the case of the co-thermolysis of the com-
plexes) [12] at rather low temperatures (below 400 °C). Presum-
ably, these reactions proceed via the intramolecular reduction of
metals with the involvement of ligands having reducing properties.
of metal particles. Evidently, these processes essentially depend on
the nature of metal ions in the precursors, as well as on the struc-
tural and electronic features of both the ligands having a reducing
ability and the molecules as a whole.
As for the known type of carboxylate complexes, viz., dinuclear
tetrabridged d-block metal pivalates LM(l-Piv)4ML, where
M = Mn(II) [13,14], Fe(II) [15], Co(II) [16–18], Ni(II) [19,20], or
Cu(II) [21,22], it can be noted that the pivalate ligand (Piv = (CH3)3
CCO2ꢁ) is a potential reducing agent with respect to metal ions be-
cause it contains a large number of C–H bonds. However, it ap-
peared that the nature of the metal ions, as well as the axial
ligand, also play an important role. For example, the solid-phase
thermolysis of dinuclear manganese and iron pivalates (M = Mn
and Fe) did not afford metals in spite of the presence of strong
reducing agents (aminopyridine ligands) as the ligands L [23]. On
the other hand, the thermolysis of tetrabridged dinuclear com-
plexes with nickel atoms sometimes afforded a metallic phase [24].
In the present study, we focused on the influence of the nature
of the ligands L on the phase composition of thermolysis products
and investigated the electrochemical behavior of a series of dinu-
clear copper pivalates LCu(l-Piv)4CuL (L is substituted pyridine).
* Corresponding author. Tel.: +7 (495) 955 4835; fax: +7 (495) 952 1279.
E-mail address: fomina@igic.ras.ru (I. Fomina).
The results of this study allow the evaluation of the pathways of
the transformations of these molecules in the course of thermal re-
dox processes.
1
Fax: +7 (383) 333 1399.
2
Tel.: +7 (495) 939 3065; fax: +7 (495) 932 8846.
0277-5387/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2010.02.021

I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
1735
Crystal suitable for X-ray diffraction was chosen from those pre-
cipitated from the benzene mother liquor.
2. Experimental
2.1. Synthesis
2.2. Methods
The synthetic operations were carried out under pure argon in
oxygen-free solvents using the standard Schlenk technique. Start-
ing copper pivalate 1 was synthesized according to a procedure de-
scribed earlier [21]. The synthesis of new compounds was carried
out with the use of trimethylacetic acid (Acros Organics), 2-amino-
Microanalyses were carried out on a Euro Vector Element CHN
analyzer (Model EA 3000). The IR spectra were recorded on a Spe-
cord M80 spectrometer in KBr pellets (4000–300 cmꢁ1). Magnetic
measurements were performed on a Quantum Design MPMSXL
SQUID magnetometer in the 2–300 K temperature range at a mag-
netic field strength of 5 kOe. The paramagnetic components of the
pyridine,
(Aldrich).
3-aminopyridine,
and
4-dimethylaminopyridine
magnetic susceptibility
v were calculated taking into account the
diamagnetic contribution estimated from the Pascal constants.
2.1.1. ((2-NH2)C5H4N)2Cu2(
l
-OOCCMe3)4ꢀC6H6 (2ꢀC6H6)
The effective magnetic moment was calculated by the equation
1=2
3k
N
TÞ1=2where NA, b, and k are Avogadro’s
Benzene (40 mL) was added to
a
mixture of polymer
1
,leff ¼ ð
ꢀ vTÞ ꢃ ð8v
Ab2
[Cu(OOCCMe3)2]n (0.331 g, 1.169 mmol; per formula unit
Cu(OOCCMe3)2) and 2-aminopyridine (0.109 g, 1.169 mmol), and
the reaction mixture was stirred at 80 °C until the reagents were
completely dissolved. The solution was concentrated to 15 mL
and kept at room temperature for 24 h. Green crystals suitable for
X-ray diffraction were gathered, washed with cold C6H6, and dried
under a stream of argon. The solvate of compound 2 with one C6H6
molecule was obtained in a yield of 0.477 g (96%). Anal Calc. for
C36H54Cu2N4O8: C, 54.25; H, 6.82; N, 7.02. Found: C, 54.29; H,
number, the Bohr magneton, and the Boltzmann constant, respec-
tively. The Q-band ESR spectra were recorded on a Bruker Elexsys
E580 spectrometer (34 GHz) equipped with a standard EN 5107D2
resonator and an Oxford Instruments cryostat for cooling. All
experiments were carried out with the use of polycrystalline sam-
ples, which were pre-dried with argon to remove the mother liquor
(for approximately 1 min) and sealed in quartz tubes in vacuo. The
calculations were performed using the EasySpin program package
[25]. The thermal decomposition of compounds was studied by dif-
ferential scanning calorimetry (DSC) and thermogravimetric analy-
sis (TGA) on DSC-20 and TG-50 units of a Mettler TA-3000
thermoanalyzer. In all experiments, samples of the compounds
were heated under dry argon at a constant rate of 5 deg/min. For
each compound, two DSC experiments and three TG experiments
were performed. The weight loss upon thermal degradation was
determined directly on a TG-50 unit; the accuracy of weighing
was 2*10ꢁ3 mg. The thermal decomposition was studied in steps
by differential scanning calorimetry, which involved the division of
the total temperature range into intervals. The size and number of
these intervals were determined based on the overall patterns of
weight loss upon decomposition. This approach allowed us to
determine the weight loss in each temperature range and compare
the results of DSC with the TG data. The results of these methods
were in satisfactory agreement, which confirms the reliability of
the experimental data. The accuracy of the determination of anom-
alous points and thermal effects in the DSC curves was 1° and
0.5%, respectively. The X-ray powder diffraction analysis of the
decomposition products was carried out on a FR-552 monochro-
6.91; N, 7.09%. IR (KBr) m
/cmꢁ1: 3480 m, 3360 m, 3336 s, 3200 s,
2968 s, 2952 m, 2908 m, 2868 m, 1644 s, 1620 s, 1580 s, 1572 s,
1536 s, 1496 s, 1480 s, 1452 s, 1408 s, 1372 m, 1356 m, 1336 m,
1268 m, 1228 m, 1160 m, 1080 w, 1052 w, 1016 m, 952 w, 936
w, 896 m, 848 m, 804 m, 792 m, 776 s, 744 w, 680 m, 652 m, 620
m, 548 m, 516 m, 456 m, 400 w, 364 w, 356 w, 320 w.
2.1.2. ((3-NH2)C5H4N)2Cu2(
l
-OOCCMe3)4. (3)
Benzene (30 mL) was added to
a
mixture of polymer
1
[Cu(OOCCMe3)2]n (0.3 g, 1.129 mmol; per formula unit
Cu(OOCCMe3)2) and 3-aminopyridine (0.105 g, 1.129 mmol), and
the reaction mixture was stirred at 80 °C until the reagents were
completely dissolved. The solution was concentrated to 15 mL
and kept at room temperature for 24 h. Green crystals suitable
for X-ray diffraction were gathered, washed with cold C6H6, and
dried under a stream of argon. The yield of compound 3 was
0.390 g (96%). Anal. Calc. for C30H48Cu2N4O8: C, 50.11; H, 6.72; N,
7.78. Found: C, 50.16; H, 6.80; N, 7.82%. IR (KBr) m
/cmꢁ1: 3432 m,
3364 m, 3268 m, 2968 s, 2960 s, 2928 m, 2868 m, 1684 s, 1600
s, 1484 s, 1448 s, 1416 s, 1376 m, 1360 m, 1316 m, 1272 m,
1224 m, 1196 m, 1132 w, 1076 w, 1052 w, 1020 m, 936 w, 896
m, 856 m, 796 m, 788 s, 696 m, 644 m, 620 m, 548 m, 520 m,
440 m, 416 w, 392 w, 352 w, 316 w.
mator chamber (CuK 1 radiation) using germanium as the internal
a
standard (X-ray diffraction patterns were processed on an IZA-2
comparator with an accuracy of 0.01 mm) and with the use of
the STOE Powder Diffraction System. The electrochemical oxida-
tion and reduction potentials were measured with a digital IPC_-
Win potentiostat/galvanostat connected to a personal computer.
The voltammograms were recorded in acetonitrile with 0.05 M n-
Bu4NBF4 as a supporting electrolyte at 20 °C in 8 mL-electrochem-
ical cell. Oxygen was removed from the solutions by purging with
dry argon. Working electrode was a stationary platinum disk with
a surface area of 0.049 cm2. A platinum plate was used as an aux-
iliary electrode, and a saturated silver chloride electrode was used
as the reference electrode (the potential with respect to Fc/
Fc+ = ꢁ0.46 V in CH3CN). The measured potentials were corrected
for ohmic losses. The number of electrons involved in each step
of the electrochemical process was determined by the ferrocene
reference method.
2.1.3. Cocrystallization product [((4-NMe2)C5H4N)2Cu2(
0.5Cu(
2–OOCCMe3)2((4-NMe2)C5H4N)2] (4)
Benzene (40 mL) was added to mixture of polymer
[Cu(OOCCMe3)2]n (0.331 g, 1.169 mmol; per formula unit
Cu(OOCCMe3)2) and 4-dimethylaminopyridine (0.143 g,
l
-OOCCMe3)4ꢀ
g
a
1
1.169 mmol), and the reaction mixture was stirred at 80 °C until
the reagents were completely dissolved. The solution was concen-
trated to 15 mL and kept at ꢂ5 °C for 24 h. Green crystals were
gathered, washed with cold C6H6, and dried under a stream of ar-
gon. The yield of compound 4 was 0.493 g (96%). Anal. Calc. for
C92H150Cu5N12O20: C, 53.59; H, 7.33; N, 8.15. Found: C, 53.62; H,
7.36; N, 8.14%. IR (KBr) m
/cmꢁ1: 3482 m, 3364 m, 3332 s, 2968 s,
2980 s, 2956 s, 2924 s, 2868 m, 2816 m, 1684 m, 1620 s, 1604 s,
1572 s, 1540 s, 1536 s, 1444 s, 1416 s, 1392 s, 1372 m, 1364 m,
1360 m, 1340 m, 1228 m, 1160 w, 1116 m, 1068 m, 1008 s, 980
w, 952 m, 936 w, 896 m, 888 m, 804 s, 788 m, 756 m, 680 m, 676
m, 616 m, 544 w, 524 m, 488 w, 432 m, 388 w, 352 w, 312 w.
2.3. X-ray data collection
The X-ray data sets for complexes 2ꢀC6H6, 3, and 4 were col-
lected according to a standard procedure [26] on a Bruker AXS
SMART 1000 APEX II diffractometer equipped with a CCD detector

1736
I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
(graphite monochromator,
x
scanning technique, scan step 0.3°,
quantitatively to give green crystals of dinuclear copper carboxyl-
ate complexes with aminopyridines having the composition ((2-
exposure time per frame 10 s). For all complexes, semiempirical
absorption corrections were applied (min/max transmission was
0.7088/0.9181 for 2ꢀC6H6, 0.7374/0.9100 for 3, and 0.7010/0.9008
for 4) [27]. The structures of all complexes were solved by direct
methods and using Fourier techniques and were refined by the
full-matrix least squares against F2 with anisotropic thermal
parameters for all non-hydrogen atoms (except for the carbon
atoms of the solvent molecules in structures 2ꢀC6H6 and 4 and
the carbon atoms of one disordered tert-butyl group (each C atom
of the methyl groups in the tert-butyl substituent group was lo-
cated in two positions with occupancy factors of ꢂ0.5)). The hydro-
NH2)C5H4N)2Cu2(
a solvate with one benzene molecule (2ꢀC6H6)) and ((3-NH2)
C5H4N)2Cu2( -OOCCMe3)4 (3). In spite of the ratio Cu(OOCCMe3)2:
L3 = 1:1, only crystals of the cocrystallization product Cu(g2
-OOCCMe3)4
l-OOCCMe3)4 (2) (the complex was isolated as
l
-
OOCCMe3)2((4-NMe2)C5H4N)2ꢀ2((4-NMe2)C5H4N)2Cu2(
l
(4) consisting of mono- and dinuclear complexes were isolated by
crystallization after the reaction of L3 with polymeric copper piva-
late 1.
According to X-ray diffraction data, dinuclear complexes 2 and
3 with the ligands L1 and L2 are structurally similar in spite of
the fact that the pyridine rings in these ligands contain the amino
group in different positions (Fig. 1, Table 2).
gen atoms in complexes 2ꢀC6H6, 3, and
4 were positioned
geometrically and refined using the riding model. In all structures,
there is a rotational disorder of the tert-butyl groups of the pivalate
ligands (formally due to rotation about the single C–CMe3 bond),
which is strongly manifested in the anisotropic temperature fac-
tors of the C atoms of the tert-butyl groups. In all crystal structures,
all bond lengths in the tert-butyl groups were constrained to be
equal to the bond lengths between the corresponding atoms (the
apparent standard deviation was 0.03 A). The benzene molecules
of solvation in the refinement were fitted to a regular hexagon
(with d = 1.39 A) in structure 2ꢀC6H6. In addition, there are disor-
dered solvent molecules (apparently, benzene) in structure 4. The
calculations were carried out with the use of the SHELX97 program
package [28]. The crystallographic parameters and the X-ray data
collection and refinement statistics for the structures of 2ꢀC6H6,
3, and 4 are given in Table 1.
It should be noted that both structures are characterized by the
presence of hydrogen bonding between the protons of the amino
groups and the oxygen atoms of the carboxylate bridges (intramo-
lecular in 2, N–HꢀꢀꢀO(OOCR) 2.929–3.002 Å; the N–HꢀꢀꢀO angles are
136.2–155.4°; intermolecular in 3, N–HꢀꢀꢀO(OOCR) 3.028–3.110 Å;
the N–HꢀꢀꢀO angles are 116.5–128.4).
The crystal of the cocrystallization product Cu(
Me3)2((4-NMe2)C5H4N)2ꢀ2((4-NMe2)C5H4N)2Cu2( -OOCCMe3)4 (4)
(Fig. 2, Table 2) consists of two different complex molecules, viz.,
the dinuclear molecule ((4-NMe2)C5H4N)2Cu2( -OOCCMe3)4 (4a)
located on a crystallographic center of symmetry and the mononu-
clear molecule Cu(
2-OCCMe3)2((4-NMe2)C5H4N)2 (4b) having the
g
2–OOCC
l
l
g
crystallographic symmetry C2h. Although the overall structure of
the dinuclear molecule (4a) differs only slightly from the struc-
tures of dinuclear complexes 2 and 3, the distance between the
copper atoms in molecule 4a (CuꢀꢀꢀCu, 2.664(3) Å) is somewhat
longer than those in complexes 2 and 3 (CuꢀꢀꢀCu, 2.6307(9) and
2.6360(18) Å, respectively). It is noteworthy that the Cu–N dis-
tance between the copper atom and the nitrogen atoms of the pyr-
idine ligands in the dinuclear molecule (4a) (2.138(10) Å; N–
CuꢀꢀꢀCu, 179.3(3)°) is essentially different from the Cu–N distance
in mononuclear complex 4b containing the trans-coordinated pyr-
idine ligands (Cu–N, 2.00(2) Å), whereas this distance is approxi-
mately equal to the corresponding distances in 2 and 3 (2.163(5),
2.177(5) Å and 2.155(10), 2.161(10) Å, respectively).
3. Results and discussion
3.1. Synthesis and X-ray diffraction study
It is known that aminopyridines act as reducing agents [29] and
readily undergo condensation with carbonyl-containing organic
moieties [29]. Since pivalate complexes containing carbonyl frag-
ments served as the starting compounds for the thermolysis, we
introduced aminopyridine ligands with different structures into
dinuclear copper pivalates by the reactions of the known polymer
[Cu(OOCCMe3)2]n (1) [21,22] with 2-amino- (L1), 3-amino-(L2), and
4-dimethylaminopyridine (L3) in the ratio Cu:L = 1:1 in nonpolar
benzene. In the first two cases, the reaction proceeds virtually
As opposed to complexes 2 and 3, the hydrogen bonding in
compound 4 is absent.
All newly synthesized copper complexes 2–4 contain strong do-
nor aminopyridine ligands in the coordination sphere. However, it
Table 1
Crystallographic data and details of the structure refinement for complexes 2ꢀC6H6, 3 and 4.
Complex/parameter
2ꢀC6H6
3
4
Formula
Formula weight
T (K)
C36H54Cu2N4O8
797.91
200(2)
C30H48Cu2N4O8
719.80
296(2)
C92H150Cu5N12O20
2061.94
296(2)
Crystal system
Space group
a (Å)
b (Å)
c (Å)
monoclinic
Cc
monoclinic
P2(1)/n
monoclinic
C2/m
35.539(2)
11.7266(6)
29.472(1)
100.020(1)
12095.4(11)
12
9.6551(10)
20.554(2)
18.9913(19)
101.977(2)
3686.9(7)
4
20.4591(14)
30.797(2)
10.9921(8)
114.8860(10)
6282.8(8)
2
b (°)
V (Å3)
Z
Dcalc (g cmꢁ3
)
1.315
1.087
1.297
1.201
1.090
0.889
l
(mmꢁ1
)
Total number of reflections/unique
Rint
35 134/23 135
0.0427
24 315/7598
0.0984
22 373/6647
0.1626
hmax (°)
26.73
26.73
26.73
S
0.996
0.969
1.000
R1 [I > 2
r(I)]
0.0673
0.0618
0.0832
wR2 [I > 2
r
(I)]
0.1675
0.1014
0.1569
R1 (all data)
0.0915
0.1302
0.1411
wR2 (all data)
0.1872
0.1792
0.1749

I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
1737
Fig. 1. Structures of the dinuclear complexes ((2-NH2)C5H4N)2Cu2(
l
-OOCCMe3)4 (2) (a) and ((3-NH2)C5H4N)2Cu2(l-OOCCMe3)4 (3) (b) (hydrogen atoms are omitted for
clarity).
Table 2
Main structural and magnetic characteristics for the dinuclear copper pivalate complexes.
Compound
CuꢀꢀꢀCu (Å)
2.6307(9)
Cu–N (Å)
Cu–O (Å)
leff
(
l
B) 300–2 K.
Exchange parameters-2Jc (cmꢁ1
)
((2-NH2)C5H4N)2Cu2(
((3-NH2)C5H4N)2Cu2(
l
l
-OOCCMe3)4 (2)a
2.163(5), 2.177(5)
1.942(5) ꢁ 1.990(5)
2.084 ꢁ 0.090
1.993 ꢁ 0.085
2.519 ꢁ 1.254
214(7)
244(5)
-OOCCMe3)4 (3)
2.6360(18)
2.155(10), 2.161(10) 1.937(8) ꢁ 1.989(8)
2-OOCCMe3)2((4-NMe2)C5H4N)2
2.00(2)
1.965(16)
b
Cu(
((4-NMe2)C5H4N)2Cu2(
(C9H7N)2Cu2( -OOCCMe3)4 (5)
(2,3-Me2C5H3N)2Cu2(OOCCMe3)4 (6) [24]
((2-NH2)(6-CH3)C5H3N)2 Cu2(OOCCMe3)4 (7) 2.730(1) [23]
g
b
l
-OOCCMe3)4
2.664(3)
2.6543(9) [24] 2.223(3)
2.138(10)
1.961(10) ꢁ 1.999(9)
214(5)
l
1.956(3) ꢁ 1.985(2)
1.889 ꢁ 0.105 [24] 264(7)
1.883 ꢁ 0.088 [24] 261(1)
2.296(3)
2.243(3)
1.963(2) ꢁ 1.974(2)
1.922(6) ꢁ 2.019(5)
1.976 ꢁ 0.232
249(0)
((NH2)2C5H3N)2Cu2(OOCCMe3)4ꢀC6H6 (8)
2.762(1) [23]
1.920 ꢁ 0.171 [23] 256(3)
a
b
c
Three independent molecules in unit cell.
In the co-crystallizate 4.
Exchange-coupled dimmers model [35].
Fig. 2. Structures of the di- and mononuclear molecules (4a and 4b) in the cocrystallization product Cu(
OOCCMe3)4 (4) (hydrogen atoms are omitted for clarity).
g
2-OOCCMe3)2((4-NMe2)C5H4N)2ꢀ2((4-NMe2)C5H4N)2Cu2(
l-
should be noted that there are hydrogen bond networks between
the protons of the amino groups and the oxygen atoms of the
carboxylate ligands in the crystal structures of dinuclear molecules
2 and 3, whereas the co-crystallizate consist of discrete dinuclear

1738
I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
and mononuclear complexes and not form hydrogen bonds be-
tween the carboxylate and dimethylaminopyridine ligands. In fur-
ther investigations of the magnetic properties, ESR measurements,
and electro- and thermochemical studies, we used not only the
above-mentioned complexes but also other dinuclear copper piva-
lates LCu(Piv)4CuL (L are pyridine derivatives: quinoline (C9H7N)
(5), 2,3-dimethylpyridine (2,3-(CH3)2C5H3N) (6), 2-amino-6-meth-
ylpyridine (2-NH2)(6-CH3)C5H3N) (7), and 2,6-diaminopyridine
2,6-(NH2)2C5H3N) (8)) and the cocrystallization product Cu2(-
Piv)4(2,6-(NH2)2C5H3N)2ꢀCu2(Piv)4(CH3CN)2 (9), whose synthesis
and structures have been described in our recent publications
[23,24].
3.2. Magnetic properties
The magnetic properties of dinuclear complexes 2 and 3 differ
only slightly from those of the related dinuclear pivalate-bridged
copper complexes studied earlier [23,24,30–34,16]. In dinuclear
complexes 2, 3, and 5–8, the exchange spin-spin interactions are
antiferromagnetic, as clearly seen from the temperature depen-
dences of
crease in value leff
eff = 1.994 lB (300 K) ꢁ 0.085 lB (2 R) for 3;
for [24];
eff = 1.976 lB (300 K) ꢁ 0.232
l
eff (Fig. 3). A lowering of the temperature leads to a de-
eff = 2.084 lB (300 K) ꢁ 0.090 lB (2 R) for 2;
eff = 1.889 lB
eff = 1.883 lB
(l
l
l
(300 K) ꢁ 0.105 lB (5 R)
5
l
l
(300 K) ꢁ 0.088 lB (5 R) for 6 [24];
2.1
a
b
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
1.8
1.5
1.2
0.9
0.6
0.3
0.0
0
50
100
150
200
250
300
0
50
100
150
200
250
300
T (K)
T (K)
2.0
c
e
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
d
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0
50
100
150
200
250
300
0
50
100
150
T (K)
200
250
300
T (K)
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
f
0
50
100
150
200
250
300
0
50
100
150
T (K)
200
250
300
T (K)
Fig. 3. Temperature dependences of the effective magnetic moment for the complexes ((2-NH2)C5H4N)2Cu2(
(C9H7N)2Cu2(
-OOCCMe3)4 (5) (c), (2,3-Me2C5H3N)2Cu2(OOCCMe3)4 (6) (d), ((2-NH2)(6-CH3)C5H3N)2Cu2(OOCCMe3)4 (7) (e), and ((NH2)2C5H3N)2Cu2(OOCCMe3)4ꢀC6H6 (8) (f),
(d are experimental values, is the theoretical curve).
l-OOCCMe3)4 (2) (a), ((3-NH2)C5H4N)2Cu2(l-OOCCMe3)4 (3) (b),
l

I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
1739
lB (5 R) for 7;
l
l
eff = 1.920 lB (260 K) ꢁ 0.171 lB (5 R) for 8 [23];
3.3. ESR spectra
and eff = 2.693
l
B (300 K) ꢁ 0.120
l
B (2 R) for 9 [23]). The param-
eters of exchange interactions were estimated using a model of ex-
change-coupled dimers [35]. The values of exchange parameters
obtained are given in Table 2.
Compound 4 was obtained as a result of cocrystallization of the
dinuclear and mononuclear complexes in a ratio of 2:1 {(Cu–
Cu) + 1/2 Cu}. For cocrystallization product 4, the temperature
dependence of leff is shown in Fig. 4. At 300 K, the value of leff is
The electron spin resonance (ESR) is widely used to study mag-
netic interactions in exchange-coupled paramagnetic clusters [36].
Generally, the information obtained by ESR refers to the determi-
nation of g-tensors parameters and dipole interaction parameters
(zero-field splitting tensor D); however, in some cases, the ex-
change interactions and their signs were determined as well
[37,38]. The information on symmetry of the environment of para-
magnetic ions and the contribution of the anisotropic exchange
interactions can be obtained by analyzing the components of the
g and D tensors. Hence, we performed systematic studies and a
comparative analysis of a series of dinuclear tetrabridged copper
2.52
lB. A lowering of the temperature leads to a decrease in the
value of
leff, which reaches the value 1.35 lB at 70 K, after which
leff remains virtually unchanged down to helium temperatures
(at 5 K, the value of leff is 1.25 lB (Table 2)).
The paramagnetism of compound 4 is associated with the
presence of unpaired electrons of CuII ions in both dinuclear
and mononuclear complexes. As in the case of compounds 2, 3,
pivalate complexes 2–7 and 9 by Q-band ESR spectroscopy
(34 GHz).
The ESR spectra of compounds 2–7 and 9 are typical of the state
with the total electron spin S = 1 and a large zero-field splitting.
The intensity of the observed signals of the triplet (S = 1) state,
which are unambiguously assigned to the dinuclear copper com-
plexes under study, decreases with decreasing temperature, and
these signals are difficult to observe at temperatures below ꢃ70–
100 K. This fact is in complete agreement with the results of mag-
netic susceptibility measurements and is indicative of the singlet
ground state of the copper dimers (unobservable by ESR) and the
antiferromagnetic exchange. Hence, all compounds were studied
at room temperature (293 K) and at T = 150 K. The low-tempera-
ture experiments, in which the spectral lines are substantially nar-
rower, allowed us to more precisely determine the magnetic
resonance parameters of the systems. However, the ESR spectra
are very difficult to record at T < 150 K due to depletion of the ex-
cited triplet state. In addition, the experiments performed at two
temperatures showed that the parameters of the copper dimers
in the complexes under study depend only slightly on the
temperature.
and 5–8 (Fig. 3), the dependence
leff(T) in the 300–70 K temper-
ature range for 4 is determined by antiferromagnetic spin–spin
exchange interactions in 4a, resulting in the spin pairing in 4a
at temperatures below 70 K. The residual leff value (ꢂ1.3 lB
)
in the 70–5 K temperature range is determined primarily by
the unpaired electrons of CuII in complex 4b. In the solid-phase
of 4, the intermolecular spin-spin exchange interactions between
complexes 4a and 4b can be ignored, and the contributions from
the dinuclear and mononuclear complexes to
v can be estimated
as (Cu–Cu)+ (Cu). The component (Cu) was evaluated from
v
v
v
the Curie–Weiss law using temperature dependence of the in-
verse magnetic susceptibility in the 5–25 K temperature range,
where the contribution of the dinuclear complex to the magnetic
susceptibility can be ignored. We analyzed the dependence 1/
v(T) (based on the first eight values) and obtained the values
C = 0.20771 K cm3/mol and h = ꢁ0.37381 K (Fig. 4b). The C value
is in good agreement with the theoretical value (0.2057 K cm3/
mol) determined for one-half mole of the noninteracting para-
magnetic centers with the spins S = 1/2 and the g factor of
2.10. The parameter of exchange interaction in 4a was estimated
with the use of the model of exchange-coupled dinuclear com-
plexes taking into account the contribution from the mononu-
clear complex.
Fig. 5 shows the experimental and theoretically approximated
ESR spectra of complexes 2–7 and 9 at T = 293 and 150 K. All
spectra are essentially similar, except for the spectrum of cocrys-
tallization product 4, which shows not only the signal of the
triplet state (S = 1) but also the typical signal of the magnetically
isolated copper ion. The very similar ESR spectra of the complex
[Cu(Sup)2H2O]2 were interpreted earlier [39]. Let us mention the
The obtained values for parameter of exchange interaction J
(ꢁ214(5) cmꢁ1) has the same order of magnitude as the corre-
sponding parameters for complexes 2, 3, and 5–8 (Table 2).
For cocrystallization product 9, the average parameter of ex-
change interaction was estimated, both exchange clusters being
assumed to be equal. The obtained values for parameter of ex-
following main features: (1) the intense signals in
B0 ꢃ 950 and 1350 mT correspond to the allowed transitions
with M = 1, and the splitting between these lines is caused
a field
D
by the electron dipole–dipole interaction (zero-field splitting);
(2) the signals in a field B0 ꢃ 500 mT correspond to the forbidden
change interaction J is ꢁ268(4) cmꢁ1
.
400
a
b
2.6
350
300
250
200
150
100
50
2.4
2.2
2.0
1.8
1.6
1.4
1.2
0
0
50
100
150
200
250
300
0
50
100
150
T(K)
200
250
300
T (K)
Fig. 4. Temperature dependences of leff (a) and the inverse magnetic susceptibility 1/
v
(b) for compound 4 (d are experimental values,
is the theoretical curve).

1740
I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
Fig. 5. Experimental ( ) and calculated ( ) ESR spectra of compounds 2–7 and 9 at 293 K (on the left side, mmw ꢃ 34.2 GHz) and 150 K (on the right side, mmw ꢃ 34.4 GHz).
double-quantum transitions with
D
M = 2 typical of triplet states;
As a rule, in copper ions, the hyperfine coupling constants for the z
component of g tensor (Azz) are substantially larger than the con-
stants for the x,y components (Axx,yy). As a result, only the Azz val-
ues were reliably determined in the calculations from the well-
(3) the low-intensity signals in a field B0 ꢃ 1100–1200 mT (2, 3,
5–7, and 9) also correspond to double-quantum transitions;
however, the contribution of insignificant impurities of free cop-
per ions, which can give signals at the same magnetic field
strength, cannot be ruled out.
In the calculations, the model took into account the total elec-
tron spin S = 1, which is coupled with two equivalent 63Cu nuclei
with the corresponding hyperfine coupling tensors and has a
zero-field splitting commonly described by the scalar parameters
D and E. All tensors were assumed to be collinear, which was suf-
ficient for obtaining good agreement with the experimental data,
and the g tensor was assumed to be nearly axially symmetrical.
However, in some cases, it was necessary to take into account
the rhombic distortion to reach agreement with experimental data
and, consequently, to take into account the non-zero parameter E.
resolved structure of the signals with
DM = 2 at T = 150 K. In all
cases, the spectral lines have a Lorentzian line shape, which is
indicative of a relaxation broadening.
The calculated magnetic resonance parameters of complexes 2–
7 and 9 are given in Table 3. As can be seen from this table, the dif-
ferences between compounds 2–7 and 9 are, on the whole, small.
This result was expected because the magnetic interactions are lo-
cated in the dinuclear fragment, and the influence of substituents
more distant than those involved in the direct coordination envi-
ronment of the copper ion should not have a substantial effect. This
is also confirmed by very similar magnetic susceptibility curves.
However, as it follows from the ESR spectra and the components
Table 3
Calculation parameters for the EPR spectraa of the compounds 2–7, 9.
Compound
g-Tensor
D (cmꢁ1
)
E (cmꢁ1
)
Azz (MHz)
2
3
4
[2.0702.0622.375]
[2.078 2.062 2.390]
[2.075 2.075 2.390]
[2.055 2.055 2.253]b
[2.070 2.062 2.380]
[2.071 2.067 2.380]
[2.074 2.066 2.385]
[2.080 2.060 2.355]
0.387
0.385
0.404
ꢁ0.003
ꢁ0.002
0
210
205
200
580b
210
210
202
205
5
6
7
9
0.383
0.391
0.408
0.390
ꢁ0.004
ꢁ0.001
ꢁ0.002
ꢁ0.006
a
The accuracy of the parameters determination is estimated to be: 0.003 for the g-tensor components, 0.002 cm-1 for D and E, 2 MHz for Azz. The signs of the D and E
parameters should be understood as the relative ones, i.e. [ D E].
b
Parameters of the magnetically isolated copper(II) ion in the compound 4.

I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
1741
of the g and D tensors, the degree of rhombic distortion (i.e., the
deviation of the coordination environment of the copper ions from
the axial symmetry) slightly varies from compound to compound.
copper center is the first step giving rise to an unstable mixed-va-
lence complex.
LCuIIð
l
-OOCCMe3Þ CuIIL þ e ! LCuIIð
l
-OOCCMe3Þ CuIL
4
4
The formation of the mixed-valence complexes [CuIICuI] has
been observed earlier in the reduction of dinuclear Cu(II) tetraace-
tate complexes [43,45]. The peak potential values for dinuclear
complexes 1–9 (Table 4) are shifted in the cathodic direction, thus
reflecting the stronger electron-donating properties of the pivalate
ligands compared to the acetate ligands but the shapes of CV
curves for complexes with pivalate and acetate [43] bridging li-
gands are similar. A broadening of the reduction peak and the ab-
sence of the reoxidation peak in the reverse scan of the cyclic
voltammogram indicate that the electron transfer is followed by
the structural reorganization of the mixed-valence complex. The
structural reorganization can involve both a change in the type
of the coordination polyhedron of the copper ions and, for example,
the elimination of one of the bridging pivalate ligands. In addition,
the complexes [CuIICuI] are prone to disproportionation to form
metallic Cu [39,43–45]. However, a weak oxidation peak at a po-
tential of 0.1 V (corresponding to oxidation of Cu0) was observed
for dinuclear complexes 1–9 only at a slow potential scan rate
(650 mV/s), if the scan was reversed after the first reduction peak.
This indicates that process (1)
3.4. Cyclic voltammetric studies
Apparently, the solid-state thermal decomposition of molecular
precursors, for example, transition metal complexes, including
dinuclear metal pivalates, is the redox process, which, in some
cases, can lead to reduction of metal ions (up to the metallic phase)
and oxidation of organic ligands [22–24,40–42]. The mechanism of
these transformations is very complex and is determined primarily
by the electronic nature of the precursor components providing the
possibility of electron density redistribution in the molecules upon
thermal decomposition. Although, the structural parameters of the
molecules and the crystal lattice structure often play an important
role as well.
It is reasonable to suggest that the electrochemical reduction
and oxidation potentials of precursor complexes are among possi-
ble diagnostic criteria for the evaluation of the impact of intramo-
lecular redox processes. Hence, we performed the voltammetric
study of the redox behavior (solutions in MeCN) of the dinuclear
complexes LCu(l-Piv)4CuL (2-9), where L is the substituted
pyridine.
2½CuII—CuIꢄ ! ½CuII—CuIIꢄ þ ½CuI—CuIꢄ
½CuI—CuIꢄ ! 1=2½CuII—CuIIꢄ þ Cu0
ð1Þ
ð2Þ
ð3Þ
The potentials of the observed oxidation and reduction peak
potentials of the complexes are given in Table 4. The reduction of
complexes 1–9 proceeds in two successive one-electron steps,
whereas both one- and two-electron waves are observed during
oxidation. The oxidation and reduction of complexes 1-9 are irre-
versible processes, as evidenced by the absence of pronounced
peaks in the reverse scans of CV- curves. Consequently, the values
presented in Table 4 are not thermodynamic redox potentials. This
makes it somewhat difficult to interpret their relation to possible
structural rearrangements of the molecules of complexes in solu-
tion. However, for a series of compounds containing the structur-
ally and compositionally similar metal carboxylate core
Cu(Piv)4Cu, certain tendencies in the electron density redistribu-
tion in the molecules can be revealed. All peaks present in the vol-
tammograms correspond to the diffusion-limited processes.
or
½CuI—CuIꢄ ! CuII þ Cu0
is slow on the CV time scale. However, if the potential scan is re-
versed after the second reduction peak (corresponding to the for-
mation of the complex [CuICuI]), a very intense oxidation peak of
Cu0 can be observed in the reverse scan (Fig. 6). This is evidence that
the disproportionation according to Eq. (2) or (3) goes virtually to
completion during the cyclic voltammetry experiments.
It should be noted that the reduction peaks in the voltammo-
grams of complexes 2–9 at ꢁ1.2 – ꢁ1.5 V potential range assigned
to the second electron transfer are strongly broadened, and these
peaks are shifted in the cathodic direction by 100–200 mV after re-
peated scanning. This makes it difficult to precisely determine the
redox potential values and is indicative of the occurrence of chem-
ical and adsorption processes accompanying the electron transfer.
In this context, Table 4 gives the potential values only of the first
reduction peaks observed in the voltammograms.
3.4.1. Reduction
The reduction peaks observed in CV’s are diffusion-controlled
(as follows from Ip–v1/2 linear dependence (v = scan rate) and,
according to the estimation of peak current values, correspond to
consecutive one-electron processes. A comparative analysis of
the reduction potentials (Ered) of the dinuclear copper complexes
suggests that, in the case of complexes 1–9, the reduction of the
Since the electron density back-donation from the copper cen-
ters to the axial ligand is unlikely to occur in these systems, the li-
gand acts only as an electron-pair donor. Hence, the reduction
Table 4
Redox properties of compounds 1–9 (C = 7ꢅ10ꢁ4 M, CH3CN, Pt, 100 mV/s, Bu4NBF4, vs. Ag/AgCl/KCl).
Complex
ꢁEred (V)
0.68
Eox (V)
Decomposition product
Cu, Cu2O, CuO
Cu2piv4 (1) [21,22]
1.27
1.80
1.42
1.39
1.32
1.39
1.13
1.51
0.97
1.64
0.85
1.31
0.91
1.31
Cu2piv4(2-(NH2)C5H4N)2 (2)
Cu2piv4(3-(NH2)C5H4N)2 (3)
Co-crystallization product Cu(
Cu2piv4(C9H7N)2 (5) [24]
Cu2(piv)4(2,3-Me2C5H3N)2 (6) [24]
0.65
0.69
0.74
0.35
0.42
Cu
Cu
Cu
CuO
g
2–OOCCMe3)2((4-NMe2)C5H4N)2ꢀ2((4-NMe2)C5H4N)2Cu2(
l
-OOCCMe3)4 (4)
CuO, Cu2O, Cu
Cu2(piv)4((2-NH2)(6-CH3)C5H3N)2 (7) [23]
0.77
0.95
0.89
Cu
Cu2(piv)4(2,6-(NH2)2C5H3N)2 (8) [23]
Cu
Co-crystallization product Cu2(piv)4(2,6-(NH2)2C5H3N)2ꢀCu2(piv)4(CH3CN)2 (9) [23]
Cu, Cu2O

1742
I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
oxidation potentials of the dinuclear copper complexes increase
with decreasing donating ability of the axial ligands. Model piva-
late polymer 1, which is transformed into the dinuclear complex
(MeCN)Cu(Piv)4Cu(MeCN) in MeCN solution, is oxidized in one
two-electron wave. The structure of the dinuclear complex has
been established earlier in cocrystallization product 9 [23]. Since
redox-active axial substituents are formally absent in compound
1, both copper centers apparently undergo oxidation.
0.02
0.01
Scheme 1 shows five main possible pathways of oxidation and
subsequent fragmentation of dinuclear copper pivalate complexes
2–9. For simplicity, all ligands containing amino groups in the
complexes are designated as L and the pivalate bridges are omit-
ted. In the case of the path a, a reversible one-electron oxidation
peak corresponding to oxidation of the copper center (because
the structure of the complex remains unchanged) followed by oxi-
dation of the axial ligand should be observed in the voltammo-
gram. The path b involves the irreversible one-electron oxidation
of the ligand followed by its elimination and the replacement by
molecules of the solvent (CH3CN), which is present in a large ex-
cess. In the second step, the copper center is oxidized. If the pro-
cess proceeds according to the mechanism c, the first oxidation
peak should be two-electron because the ligand that is eliminated
(amino-substituted pyridine) is redox-unstable at this potential
and, consequently, it should immediately give an electron to the
electrode (ECE process: electrochemical step-chemical step-elec-
trochemical step). This means that the oxidized axial ligand is actu-
ally eliminated. The driving force for the replacement of the ligand
is that the metal in higher oxidation states (metal becomes a hard-
er Lewis acid) is preferably coordinated by the harder base CH3CN,
which is present in a large excess.
0.00
-0.01
-0.02
-2000
-1500
-1000
-500
0
500
E,mV
Fig. 6. Typical CV curve for complex
(s = 0.049 cm2), Bu4NBF4, vs. Ag/AgCl/KCl).
3
(7 ꢆ 10ꢁ4 M, CH3CN, 50 mV/s, Pt
potential of dinuclear complexes should be shifted to negative val-
ues with increasing donating ability of the ligand, which is actually
observed in cyclic voltammetry experiments (Table 4). Based on
the observed sequence in the reduction potential values of the
complexes, it can be concluded that thermal decomposition of
more electron-rich complexes (which are reduced at more cathodic
potentials) affords products of higher degrees of reduction of Cu(II)
ions in pivalate complexes 2–9. It should be noted that in our case
the evaluation of the electron-donating properties of the ligands by
analyzing the reduction potentials of the complexes is more cor-
rect than a comparison of the oxidation or reduction potentials
of the free ligands. This is associated with the fact that pyridine
is not oxidized in the potential range under study, and the oxida-
tion of various aminopyridines occurs at the amino group in a
rather narrow potential range (0.8–0.9 V, with respect to Ag/
AgCl/KCl). According to the literature data [46], the oxidation of
amines involves a wide range of consecutive chemical reactions
and it impedes the direct comparison of the measured potentials.
The reduction of pyridine derivatives containing electron-donating
substituents in the ring (which are considered in the present study)
occurs at high cathodic potentials (>2.4 V), which also makes it dif-
ficult to perform a comparative analysis.
It is worth mentioning that in some cases [38,44] for binuclear
copper complexes with carboxylate bridging ligands two sequen-
tial two-electron reductions were observed. Such behavior is often
the case when binuclear motif is destroyed in solution yielding
mononuclear species or if there is no interaction between metallic
centers. As follows from the estimation of magnetic properties of
complexes 2–9 (see above), it is not our case. Besides, UV–Vis
investigation of freshly prepared solutions of the complexes
showed that they are stable at least on the timescale of CV
measurements.
The mechanism f should involve the irreversible one-electron
oxidation of the ligand giving rise to the radical–cation followed
by the fast
a-proton abstraction (typical of amino- and methyl-
substituted pyridines). In the second step, the further one-electron
oxidation of the metal center or the ligand can occur.
The path d is, in essence, similar to the above-considered path f.
In the first step, the copper center is oxidized to form a mixed-va-
lence complex followed by the fast intramolecular electron trans-
fer from the axial ligand to Cu(III) to give the ligand radical-
cation. The latter rapidly releases the proton, hydrogen-bonded
to the oxygen atom of the pivalate bridge. This process can be
the driving force for the intramolecular electron transfer. In the
second step, the newly formed Cu(II) center in the complex con-
taining the oxidized axial ligand undergoes one-electron oxidation.
Owing to complexity of the observed electrochemical behavior,
it is difficult to decide with certainty only from the CV data which
of the above-considered mechanisms is true for each of complexes
2–9. Potential-controlled electrolysis with the identification of
products will be more helpful. However, some hypotheses can be
proposed. Since one of the aims of the present study was to eluci-
date the mechanism of intramolecular redox processes occurring
during thermal decomposition of pivalate complexes, the mecha-
nisms d and f (Scheme 1), which involve the intramolecular elec-
tron transfer and the
a-proton abstraction from the axial ligand,
are of the most interest.
Let us compare the oxidation of complexes 5 and 6 containing
the apical ligands, which are not oxidized at potentials lower than
2 V (Table 4). For these complexes, the metal orbitals (Cu(II)/
Cu(III)) should be involved in the oxidation. As opposed to complex
5, the cyclic voltammogram of complex 6 shows two one-electron
peaks at E = 1.13 and 1.51 V. It can be speculated that the methyl
substituents of complex 6 play a key role in the difference in the
electrochemical behavior of complexes 5 and 6. Apparently, in
the case of complex 6, Cu(II) is initially oxidized to Cu(III), resulting
in an even more substantial donation of electron density from the
axial ligand. This, in turn, leads to an increase in the acidity of the
3.4.2. Oxidation
Although the mechanism of reduction of dinuclear complexes
2–9 seems to be more or less evident, the mechanism of oxidation
remains unclear. First, it is worthy of note that both the copper
centers and the amino groups of the axial ligands can be oxidized
(Scheme 1). The irreversibility of all oxidation peaks indicates that
the initially formed electron transfer product is subjected to subse-
quent fast chemical reactions. As can be seen from Table 4, the

I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
1743
Scheme 1. Possible oxidative transformations of dinuclear copper pivalates.
protons of the methyl group at the
a
position and a strengthening
The thermogravimetric analysis showed that the solid-state
decomposition of compound 1 under argon does not afford stable
intermediates (Fig. 7a), proceeds in a rather narrow temperature
range (230–300 °C), and is endothermic (Fig. 7b). In the 300–
365 °C temperature range, highly exothermic processes occur,
which may be associated with the formation of new structures of
solid decomposition product. The weight loss upon thermolysis is
73.3 1.5%. The phase composition of the final solid decomposition
product determined by X-ray powder diffraction shows the pres-
ence of a mixture of metallic copper and copper oxides: Cu,
Cu2O, and CuO. Although it is impossible to precisely determine
the ratio of the phases in the decomposition products, the thermo-
gravimetric analysis showed that the weight of the solid product
after heating of 1 under an inert atmosphere is 26.7 2.0% of the
initial weighed sample, which is equal, within experimental error,
to the value of 27.20% calculated from the empirical formula on the
assumption that the decomposition affords a mixture of the phases
Cu, Cu2O, and CuO in a ratio of 1:1:1.
of the interaction between the oxygen atom of the pivalate bridge
and the proton of the methyl group via hydrogen bonding. This
promotes the intramolecular oxidation of the ligand by the copper
center to form the ligand radical–cation accompanied by the fast
proton transfer from the ligand to the oxygen atom of the pivalate
ligand. The newly formed Cu(II) center again undergoes one-elec-
tron oxidation at E = 1.51 V. The replacement of the axial ligand
by the solvent molecules does not apparently take place because
the oxidation potential of the complex is not equal to the oxidation
potential of model pivalate complex 1, which does not contain het-
erocyclic apical ligands.
The oxidation of complex 3 proceeds in one two-electron step.
This suggests that the oxidation of complex 3 follows the path c.
For compounds 2, 4, 7, 8, and 9 containing the easily oxidizable ax-
ial aminopyridine ligands, it is difficult to draw a definite conclu-
sion. Both the mechanisms d and c are possible. The fact that the
thermal decomposition of these compounds affords metallic cop-
per agrees well with both the strong electron-donating ability of
the ligands (see the reduction potentials of the complexes in Table
Our results differ from the results of the earlier study, in which
the thermolysis of polymeric copper pivalate was investigated un-
der a stream of nitrogen by the simultaneous TGA–DTGA–DSC
4) and the presence of a-H atoms that facilitate the intramolecular
electron transfer from the ligand to the metal (see above).
Therefore, the electrochemical study of the properties of com-
plexes 1–9 led to the following conclusions. First, a comparative
analysis of the first reduction peak potential values of the com-
plexes shows that they reflect the electron-donating properties of
the axial ligands and allow the prediction of the impact of intramo-
lecular redox processes during the thermal decomposition of this
type of compounds. Second, the analysis of the possible mecha-
nisms of electrochemical oxidation of complexes 1–9 shows that
3.5
a
3
2.5
2
1.5
1
0.5
0
the presence of highly acidic
a-H atoms in the axial ligand can
facilitate the intramolecular electron transfer from the ligand to
the metal, thus increasing the degree of reduction of Cu ions (up
to Cu0).
0
100
200
t,C
300
400
15
10
5
b
3.5. Thermal decomposition
As follows from the results of the electrochemical study, copper
pivalate complexes with strong donor axial aminopyridine ligands
2–4 can serve as promising molecular precursors for the prepara-
tion of metallic copper by the thermolysis of these complexes via
intramolecular redox transformations. However, in the first step
we performed thermal decomposition of the starting pivalate coor-
dination polymer 1, which consists of the dinuclear complex units
Cu(Piv)4Cu linked to each other by intermolecular Cu–OPiv bonds
[22] and which contain no additional ligands with reducing prop-
erties, as the model compound.
0
-5
-10
0
100
200
t,C
300
400
Fig. 7. Temperature dependences of the weight loss (a) and the heat flow (b) for the
polymer [Cu2(Piv)4]n (1).

1744
I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
(thermogravimetric analysis – differential thermogravimetric anal-
ysis – differential scanning calorimetry) method on a Q-1500 D
instrument; the heating rate was 2.5 deg/min (the final tempera-
ture of the investigation was not reported). Under the above-de-
scribed thermolysis conditions, metallic copper was not detected,
and the final mixture consisted only of oxides Cu2O and CuO [47].
The present study showed that even the thermolysis of poly-
meric copper(II) pivalate (1) is accompanied by intramolecular re-
dox processes. It can be expected that the presence of pyridine
ligands containing easily oxidizable amino groups, which can act
as inner-sphere reducing agents, in the dinuclear system Cu(Piv)4-
Cu would lead to deeper intramolecular redox transformations
during the thermal decomposition.
It appeared that complexes 2 and 3 containing aminopyridine
ligands decomposed in the same manner. It is practically impossi-
ble to isolate individual decomposition steps (Fig. 8). It should be
noted that an endothermic effect is observed in the curves of heat
flow (37–90 °C) (Fig. 8c). The endothermic effect in this tempera-
ture range is not associated with decomposition of the samples,
as evidenced by a comparison of the TGA (thermogravimetric anal-
ysis) curves and DSC (differential scanning calorimetry) curves.
Apparently, this effect is attributed to the breaking of hydrogen
bonds that are present in the crystals of 2 and 3. The decomposi-
tion of the complexes started at temperatures above 170 °C (2)
and 208 °C (3) (Fig. 8), the onset of decomposition being accompa-
nied by melting (the first endothermic effect (Fig. 8b). The further
deep decomposition proceeds in a rather broad temperature range
(170–350 °C and 208–350 °C for 2 and 3, respectively) and is loss in
this temperature range is 80.7 2.0% and 81.2 2.0% for 2 and 3,
respectively. The phase composition of the final solid decomposi-
tion product determined by X-ray powder diffraction shows that
metallic copper was obtained as the only decomposition product.
The thermogravimetric analysis showed that the weight of the so-
lid product after heating in an inert atmosphere is 19.3 2.0% and
18.8 2.0% of the initial weighed sample, which is equal, within
experimental error, to the value of 17.66% calculated from the
empirical formula on the assumption that the decomposition affor-
ded metallic copper.
14
12
10
8
6
4
a
2
0
0
100
200
300
400
500
t,C
10
5
0
-5
-10
-15
-20
-25
-30
b
0
50
100 150 200 250 300 350 400 450 500
t,C
Fig. 9. Temperature dependences of the weight loss (a) and the heat flow (b) for
copper complex 4.
2 and 3, it is practically impossible to isolate individual decompo-
sition steps. The deep decomposition proceeds in a rather broad
temperature range (210–350 °C) and is accompanied by the
appearance of a complex endothermic effect (Fig. 9b). The weight
loss in this temperature range is 85.1 1.5%. The phase composi-
tion of the final solid decomposition product determined by X-
ray powder diffraction shows that metallic copper was obtained
as the only decomposition product.
Therefore, the study of the thermolysis of complexes 1–4
showed that the introduction of aminopyridine ligands containing
easily oxidizable substituents into the coordination sphere of the
metal atoms leads to a higher degree of intramolecular redox
processes.
An analysis of the results of the present study showed that the
temperature of the onset of decomposition of complexes 2–9 is
lower than the boiling point of the free ligand (substituted pyri-
dine), the decomposition of all complexes, except for 6, being
accompanied by melting (Table 5). The thermal stability of com-
plexes 2–9 depends on the position and nature of the substituents
in pyridine (Table 5). The complexes can be classified into two
groups according to this feature. The complexes belonging to the
first group (2, 3, 5, and 6) contains substituents at positions 2
and 3 (ortho and meta). There is a correlation between the boiling
point of the ligand in these complexes and the temperature of the
onset of decomposition of the complexes (Fig. 10). The decrease in
the temperature of the onset of decomposition of the complexes
The decomposition of cocrystallization product 4 (Fig. 9) upon
heating starts at temperatures above 186 °C, the onset of decompo-
sition being accompanied by melting. As in the case of compounds
a
14
12
10
8
6
4
belonging to this group [
is 37.2 °C for 2, 40 °C for 3; 31 °C for 5, and 10.7 °C for 6.
The second group of complexes ( 7–9) is characterized by the
presence of substituents at positions 2 or 6 (ortho or ortho0). The
presence of two substituents in the ortho and ortho0 positions leads
to a substantial decrease in the temperature of the onset of decom-
position of the complexes [
(80 °C for 7, 146 °C for 8, 145 °C for ((NH2)2C5H3N2), and 34 °C
(CH3CN) for 9).
It was shown that the phase composition of the solid decompo-
sition products of complexes 2–9 is determined by the nature of
the substituents in pyridine [23,24]. As demonstrated previously,
only copper oxide is generated from complex 5 upon decomposi-
tion. The presence of active alkyl substituents in the apical pyridine
molecules in complex 6 substantially influences the character of
the thermolysis products. The decomposition of 6 affords a mixture
of metallic copper and its oxides. The introduction of aminopyri-
dine ligands into the coordination sphere of copper (in complexes
D
= (Tb.p.
ligand ꢁ Tonset
decomposition.)]
of
of
2
0
0
100
200
300
400
500
t,C
1.5
c
b
D
= (Tb.p. of ligand ꢁ Tonset of decomposition.)]
1
0.5
0
0
-10
-20
-30
-40
-0.5
-1
-1.5
-2
-2.5
20
70
t,C
120
0
100 200 300 400 500
t,C
Fig. 8. Temperature dependences of the weight loss (a) and the heat flow (b) for the
copper complexes (s – 2, d – 3).

I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
1745
Table 5
be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/
retrieving.html, or from the Cambridge Crystallographic Data Cen-
ter, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336
033; or e-mail: deposit@ccdc.cam.ac.uk.
Characteristics of the thermolysis processes for compound 1–9.
Complex
Tb.p.
Tonset
Decomposition
product
of N-
of
(°C)
ligand
decomposition
(°C)
Cu2piv4 (1) [21,22]
Cu, Cu2O, CuO
Acknowledgments
Cu2piv4(2-(NH2)C5H4N)2 (2)
Cu2piv4(3-(NH2)C5H4N)2 (3)
Co-crystallization
207.2;
248
> 191.2; 186
170
208
Cu
Cu
Cu
This study was financially supported by the Russian Foundation
for Basic Research (Project No. 07-03-00408, 07-03-00707, 08-03-
00365, 08-03-00343, 08-03-00142, and 08-03-00326), the Council
on Grants of the President of the Russian Federation (NSh-
3672.2010.3) the Target Programs for Basic Research of the Presid-
ium of the Russian Academy of Sciences and the Division of Chem-
istry and Materials Science of the Russian Academy of Sciences,
Federal agency for science and innovation (State Contract
02.513.12.3098).
product Cu(g2
-
OOCCMe3)2((4-
NMe2)C5H4N)2.2((4-
NMe2)C5H4N)2Cu2(l-
OOCCMe3)4 (4)
Cu2piv4(C9H7N)2 (5) [24]
Cu2(piv)4 (2,3-Me2C5H3N)2 (6)
[24]
Cu2(piv)4((2-NH2)(6-
CH3)C5H3N)2 (7) [23]
Cu2(piv)4(2,6-(NH2)2C5H3N)2
(8) [23]
238;
163.7;
207
153
CuO
CuO, Cu2O, Cu
208;
128
139
Cu
285
Cu
Co-crystallization product
Cu2(piv)4(2,6-(NH2)2C5H3N)2
ꢀCu2(piv)4(CH3CN)2 (9) [23]
285 81
140 (for
(NH2)2C5H3N2)
115 (for
Cu, Cu2O
References
[1] C. Brèchignac, P. Houdy, M. Lahmani, Nanomaterials and Nanotechnology,
EMRS, 424, Springer-Verlag, Berlin Heidelberg, 2007.
CH3CN)
[2] C.B. Murray, C.R. Kagan, M.G. Bawendi, Annu. Rev. Mater. Sci. 30 (2000) 545.
[3] N. Herron, D.L. Thorn, Adv. Mater. 10 (1998) 1173.
[4] C. Desvaux, C. Amiens, P. Fejes, P. Renaud, M. Respaud, P. Lecante, E. Snoeck, B.
Chaudret, Nat. Mater. 4 (2005) 750.
[5] D. Cahen, G. Hodes, Adv. Mater. 14 (2002) 789.
First group
250
200
150
100
50
3-(NH2)C5H3N
[6] R. Padiyath, J. Seth, S.V. Babu, L.J. Matienzo, J. Appl. Phys. 73 (1993) 2326.
[7] R. Dlecohier, H. Van den Bergh, Appl. Surf. Sci. 43 (1989) 61.
[8] S.A. Gromilov, S.V. Korenev, I.A. Baidina, Russ. J. Struct. Chem. 3 (2002) 527.
[9] S.A. Gromilov, S.V. Korenev, I.V. Korolkov, K.V. Yusenko, I.A. Baidina, Russ. J.
Struct. Chem. 3 (2004) 508.
[10] B.P. Tolochco, A.Yu. Gaponov, B.B. Bohonov, V.P. Isupov, K.A. Tarasov, M.R.
Sharaphutdinov, S.S. Shatskaya, J. Surf. Invest. X-ray Synchrotron and Neutron
Tech. 3 (2001) 20.
C9H7N
2-(NH2)C5H3N
2,3-(CH3)2C5H3N
[11] O.Yu. Gorbbenko, S.I. Troyanov, A. Meets, A.A. Bosak, Polyhedron 16 (1997)
1999.
[12] B. Takayuki, O. Yutaka, T. Yasutaka, Mater. Chem. Phys. 78 (2003) 645.
[13] M.A. Kiskin, I.G. Fomina, G.G. Aleksandrov, A.A. Sidorov, V.M. Novotortsev,
Yu.V. Rakitin, Zh.V. Dobrokhotova, V.N. Ikorskii, Yu.G. Shvedenkov, I.L.
Eremenko, I.I. Moiseev, Inorg. Chem. Commun. 8 (2004) 89.
[14] I.L. Eremenko, M.A. Kiskin, I.G. Fomina, A.A. Sidorov, G.G. Aleksandrov, V.N.
Ikorskii, Yu.G. Shvedenkov, Yu.V. Rakitin, V.M. Novotortsev, J. Cluster Sci. 16
(2005) 331.
0
140
160
180
200
220
240
260
Т (b.p. of N-ligand), °С
Fig. 10. Plot of the temperature of the onset of decomposition vs. the boiling point
of the ligand for complexes 2, 3, 5, and 6.
[15] A.A. Sidorov, I.G. Fomina, G.G. Aleksandrov, Yu.V. Rakitin, V.M. Novotortsev,
V.N. Ikorskii, M.A. Kiskin, I.L. Eremenko, Russ. Chem. Bull. 53 (2004) 483.
[16] Y. Cui, D.L. Long, X.Y. Huang, F.K. Zheng, W.D. Chen, J.S. Huang, Chinese J.
Struct. Chem. 18 (1999) 107.
[17] E.V. Pakhmutova, A.E. Malkov, T.B. Mikhailova, A.A. Sidorov, I.G. Fomina, G.G.
Aleksandrov, V.M. Novotortsev, V.N. Ikorskii, M.A. Kiskin, I.L. Eremenko, Russ.
Chem. Bull. 52 (2003) 2105.
[18] I.G. Fomina, G.G. Aleksandrov, Zh.V. Dobrokhotova, O.Yu. Proshenkina, M.A.
Kiskin, Yu.A. Velikodny, V.N. Ikorskii, V.M. Novotortsev, I.L. Eremenko, Russ.
Chem. Bull. 55 (2006) 1909.
[19] I.L. Eremenko, S.E. Nefedov, A.A. Sidorov, I.I. Moiseev, Russ. Chem. Bull. 48
(1999) 405.
[20] N.I. Kirilova, Yu.T. Struchkov, M.A. Porai-Koshits, A.A. Pasynskii, A.S.
Antsyshkina, L.Kh. Minacheva, G.G. Sadikov, T.Ch. Idrisov, V.T. Kalinnikov,
Inorg. Chim. Acta 42 (1980) 115.
[21] T.O. Denisova, E.V. Amelrchenkova, I.V. Pruss, Zh.V. Dobrokhotova, O.P.
Fialkovskii, S.E. Nefedov, Russ. J. Inorg. Chem. 7 (2006) 1020.
[22] E.G. Ilrina, S.I. Troyanov, K.M. Dunaeva, Russ. J. Coord. Chem. 18 (1992) 882.
[23] I.G. Fomina, Zh.V. Dobrokhotova, G.G. Aleksandrov, M.A. Kiskin, M.A. Bikov,
V.N. Ikorskii, V.M. Novotortsev, I.L. Eremenko, Russ. Chem. Bull. 56 (2007)
1722.
[24] I.G. Fomina, Zh.V. Dobrokhotova, M.A. Kiskin, G.G. Aleksandrov, O.Yu.
Proshenkina, A.L. Emelina, V.N. Ikorskii, V.M. Novotortsev, I.L. Eremenko,
Russ. Chem. Bull. 56 (2007) 1712.
[25] S. Stoll, A. Schweiger, J. Magn. Reson. 178 (2006) 42.
[26] SMART (Control) and SAINT (Integration) Software, Version 5.0, Bruker AXS Inc.,
Madison, WI, 1997.
2–4, 7, and 8) results in a higher degree of intramolecular redox
processes and the formation of metallic copper as the only solid
decomposition product. The appearance of Cu2O among the ther-
molysis products of complex 9 is apparently attributed to the pres-
ence of the acetonitrile complex in cocrystallization product 9. At
the same time, the presence of the complex containing the diami-
nopyridine ligand in the cocrystallization product stimulates the
formation of metallic copper and is, apparently, responsible for
the absence of copper(II) oxide, which is formed upon thermolysis
of Cu(II) pivalates with quinoline 5 and lutidine 6 or upon decom-
position of 1.
To sum up, we showed that dinuclear pivalate copper(II) com-
plexes 2–9 containing pyridine derivatives as the axial ligands
decompose upon heating, their thermolysis proceeding as the
intramolecular redox process. In some cases, this results in reduc-
tion of Cu(II) ions to Cu2O or metallic copper and oxidation of or-
ganic ligands. The voltammetric study of the redox behavior of
the dinuclear complexes LCu(l-OOCCMe3)4CuL (2–9), where L is
substituted pyridine, confirmed the intramolecular redox mecha-
nism of the solid-phase thermolysis.
[27] G.M. Sheldrick, SADABS, Program for Scanning and Correction of Area Detector
Data, Göttingen University, Göttingen, Germany, 1997.
[28] G.M. Sheldrick, SHELX-97: Program for the Solution of Crystal Structures,
Götinngen, Germany, 1997.
[29] M.B. Smith, J. March, March’s Advanced Organic Chemistry: Reactions,
Mechanisms and Structure, vol. 6, Wiley, Interscience, JohnWiley & Sons,
Inc., Publication, 2007. and references cited therein.
[30] Y. Kani, M. Tsuchimoto, S. Ohba, H. Matsushima, T. Tokii, Acta Crystallogr.,
Sect. C: Crystallogr., Struct. Commun. 46 (2000) 923.
4. Supplementary data
CCDC 763453, 763454 and 763455 contain the supplementary
crystallographic data for compounds 2, 3, and 4. These data can

1746
I. Fomina et al. / Polyhedron 29 (2010) 1734–1746
[31] I.L. Eremenko, M.A. Golubnichaya, S.E. Nefedov, I.B. Baranovsky, I.A.
Olshnitskaya, O.G. Ellert, V.M. Novotortsev, L.T. Eremenko, D.A. Nesterenko,
Russ. J. Inorg. Chem. 41 (1996) 2029.
[32] A.A. Pasynskii, T.Ch. Idrisov, K.M. Suvorova, V.T. Kalinnikov, J. Coord. Chem. 2
(1976) 1060.
[33] Yu.V. Rakitin, V.T. Kalinnikov, Modern Magnetochemistry, Science,
St.Peterburg, 1994, pp. 220.
[34] Y. Zhang, L. Jianmin, Z. Min, Q. Wang, X. Wu, Chem. Lett. (1998) 1051.
[35] B. Bleaney, K.D. Bowers, Proc. Roy. Soc. A 214 (1952) 451.
[36] A. Bencini, D. Gatteschi, EPR of Exchange Coupled Systems, Springer-Verlag,
Berlin, Germany, 1990.
[37] M. Fedin, S. Veber, I. Gromov, K. Maryunina, S. Fokin, G. Romanenko, R.
Sagdeev, V. Ovcharenko, E. Bagryanskaya, Inorg. Chem. 46 (2007) 11405.
[38] S.L. Veber, M.V. Fedin, A.I. Potapov, K.Yu. Maryunina, G.V. Romanenko, R.Z.
Sagdeev, V.I. Ovcharenko, D. Goldfarb, E.G. Bagryanskaya, J. Amer. Chem. Soc.
130 (2008) 2444.
[39] P. Koegerler, P.A.M. Williams, B.S. Parajon-Costa, E.J. Baran, L. Lezama, T. Rojo,
A. Mueller, Inorg. Chim. Acta 268 (1998) 239.
[40] M.A. Kiskin, G.G. Aleksandrov, Z.V. Dobrokhotova, V.M. Novotortsev, Yu.G.
Shvedenkov, I.L. Eremenko, Russ. Chem. Bull. 5 (2006) 779.
[41] N. Yu, S.E. Kozitsyna, I.A. Nefedov, Z.V. Yakushev, M.N. Vargaftik, I.I. Moiseev,
Mendeleev Commun. 5 (2007) 261.
[42] Z.V. Dobrokhotova, I.G. Fomina, M.A. Kiskin, A.A. Sidorov, V.M. Novotortsev, I.L.
Eremenko, Russ. Chem. Bull. 1 (2006) 250.
[43] I. Toledo, M. Arancibia, C. Andrade, I. Crivelli, Polyhedron 17 (1998) 173.
[44] S. Perlepes, E. Libby, W.E. Streib, K. Folting, G. Christou, Polyhedron 11 (1992)923.
[45] C. Dendrinou-Samara, P.D. Jannakoudakis, D.P. Kessisssoglou, G.E.
Manoussakis, D. Mentrzafos, A. Terzis, J. Chem. Soc., Dalton Trans. (1992) 3259.
[46] M.M. Baizer, H. Lund (Eds.), Organic Electrochemistry, Marcel Dekker Inc.,,
New York and Basel, 1983 (Chapter. 15 and 21.1.3.).
[47] E.G. Ilrina, N.A. Santalova, K.M. Dunaeva, Russ. J. Inorg. Chem.
2301.
9 (1991)