Unexpected behaviour of copper(I) towards a tridentate Schiff base: Synthesis, structure and properties of new Cu(I)-Cu(II) and Cu(II) complexes

Article abstract of DOI:10.1016/S0020-1693(01)00663-6

The reaction of CuBr with 2,6-bis[1-(2,6-diisopropylphenylimino)ethyl]pyridine L afforded a new Cu(I)-Cu(II) derivative [CuBrL]₂[Cu₂Br₄] (1), while the reaction of [Cu(CH₃CN)₄]PF₆ with L in THF yielded the new Cu(I) compound CuL(THF)(CH₃CN)PF₆ (2). Derivative 2 further reacted with halogenated solvents to yield halogeno-Cu(II) salts, [CuClL]PF₆ (3) using CHCl₃ and [CuBrL]Br₃ (4) using CHBr₃. Compounds 1, 3 and 4 have been fully characterised by X-ray crystallography; they contain essentially similar [CuXL]⁺ cations with a square planar copper(II) co-ordination. However, the structure of compound 1 must be viewed as built of tetranuclear units since two [Cu^IIBrL]⁺ cations are bridged by a [Cu₂^IBr₄]^2- anion with a rather strong Cu^II···Br (2.689(2) A?) secondary interaction. These secondary interactions (Cu^II cation···FPF₅ or Br anion) are weaker in 3 and 4. The electrochemical properties of compounds 1-4 are discussed, too.

Full text of DOI:10.1016/S0020-1693(01)00663-6

www.elsevier.com/locate/ica
Inorganica Chimica Acta 324 (2001) 300–308
Unexpected behaviour of copper(I) towards a tridentate Schiff
base: synthesis, structure and properties of new Cu(I)ꢀCu(II) and
Cu(II) complexes
Benoˆıt Le Gall a, Franc¸oise Conan a,*, Nathalie Cosquer a, Jean-Michel Kerbaol a,
Marek M. Kubicki b, Estelle Vigier b, Yves Le Mest a, Jean Sala Pala a
a Laboratoire de Chimie, Electrochimie Mole´culaires et Chimie Analytique, UMR CNRS 6521, Uni6ersite´ de Bretagne Occidentale, BP 809,
29285 Brest Ce´dex, France
b Laboratoire de Synthe`se et d’Electrosynthe`se Organome´talliques, UMR CNRS 5632, Uni6ersite´ de Bourgogne, 21000 Dijon, France
Received 26 April 2001; accepted 3 September 2001
Abstract
The reaction of CuBr with 2,6-bis[1-(2,6-diisopropylphenylimino)ethyl]pyridine L afforded a new Cu(I)ꢀCu(II) derivative
[CuBrL]2[Cu2Br4] (1), while the reaction of [Cu(CH3CN)4]PF6 with
L in THF yielded the new Cu(I) compound
CuL(THF)(CH3CN)PF6 (2). Derivative 2 further reacted with halogenated solvents to yield halogeno-Cu(II) salts, [CuClL]PF6 (3)
using CHCl3 and [CuBrL]Br3 (4) using CHBr3. Compounds 1, 3 and 4 have been fully characterised by X-ray crystallography;
they contain essentially similar [CuXL]+ cations with a square planar copper(II) co-ordination. However, the structure of
compound 1 must be viewed as built of tetranuclear units since two [CuIIBrL]+ cations are bridged by a [CuI2Br4]2− anion with
II
II
,
a rather strong Cu ···Br (2.689(2) A) secondary interaction. These secondary interactions (Cu cation···FPF5 or Br anion) are
weaker in 3 and 4. The electrochemical properties of compounds 1–4 are discussed, too. © 2001 Elsevier Science B.V. All rights
reserved.
Keywords: Copper complexes; Disubstituted pyridine ligand; Crystal structures; Electrochemistry
active catalysts for olefin polymerisation render them
more attractive for chemists [2]; (ii) in addition, metal
complexes derived from N3 terdendate ligands have
been shown to form a large variety of molecular archi-
tectures, ranging from macrocyclic helicates to infinite
coordination polymers [3].
1. Introduction
The co-ordinating
ability
of
the
2,6-di-
acetylpyridinebis(imine) ligands (Scheme 1), which may
be viewed as terpyridine analogues, i.e. planar NNN
terdentate ligands, has been studied for long; numerous
wernerian complexes of first row transition metals are
well known. To date most of these complexes are either
five or six co-ordinate derivatives containing one or two
neutral tridentate meridionally co-ordinated ligands [1].
This family of ligands has recently found a renewal
of interest since: (i) the unexpected and recent discovery
that such complexes, in particular the pyridine-2,6-di-
imine iron(II) and cobalt(II) complexes, may act as very
* Corresponding author. Tel.: +33-2-9801 6708; fax: +33-2-9801
7001.
E-mail address: francoise.conan@univ-brest.fr (F. Conan).
Scheme 1.
0020-1693/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0020-1693(01)00663-6

B. Le Gall et al. / Inorganica Chimica Acta 324 (2001) 300–308
301
In regard to copper chemistry, only few examples
2.2. Preparations
have been fully characterised till today, e.g. with cop-
per(II) [CuL%(NO3)2] and very recently [CuL¦](PF6)2 [4]
2.2.1. [CuBr(C33H43N3)]2[Cu2Br4] (1)
2
(Scheme 1) and with copper(I) a thermotropic liquid-
crystalline complex with a L-type ligand, non-meso-
morphic itself, although bearing long chains [5].
A filtered THF solution of CuBr (0.6 g, 4.2 mmol)
was transferred in a Schlenk flask containing the ligand
L (1.0 g, 2.1 mmol). The mixture was heated under
reflux for 3 h. A solution colour change from light
yellow to orange occurred immediately. After cooling
to r.t., the orange solution was concentrated to dryness
under vacuum and then the solid extracted with about
30 mL of CH2Cl2. After filtration, the resulting solution
was layered with 30 mL of hexane; after 3 days, or-
ange–red monocrystals were obtained (yield 60%).
Anal. Found: C, 46.3; H, 5.0; N, 4.7. Calc. for
C66H86Br6Cu4N6: C, 46.7; H, 5.1; N, 4.9%. IR (KBr
pellets, w¯max cm−1): 2965s, 2927s and 2868m (wCH);
1615m and 1585s (wCN imine); 1466s, 1442m, 1382m,
1367s, 1321m, 1265s, 1216m, 1180vw, 1103w, 1057m,
1037m, 986vw, 938w, 835vw, 816m, 799s, 779m, 760w,
740w, 694vw, 651vw, 572vw. v298 K=2.57 mB. UV–Vis
(umax/nm (CH3CN)(log o)): 234(4.66), 312(4.31),
482(3.64).
Considering the great potentiality of such Schiff base
type species, we became thus interested in studying their
co-ordination especially to copper(I) and to other
metals. As we were starting this work, Halcrow pub-
lished results related to the chemistry of 2,6-bis(imi-
nomethyl)pyridine copper derivatives and described the
X-ray structure of a copper(I) complex [CuL§]BF4
2
(Scheme 1) [6].
In this paper, we present some unexpected results
related to the reactions of copper(I) derivatives to-
wards the 2,6-bis[1-(2,6-diisopropylphenylimino)ethyl]-
pyridine L.
2. Experimental
2.2.2. Cu(C33H43N3)(CH3CN)(THF)PF6 (2)
2.1. General procedures
A stoichiometric mixture of [Cu(CH3CN)4]PF6 (1.0 g,
2.68 mmol) and L (1.29 g, 2.68 mmol) was transferred
in a round Schlenk flask. THF (100 mL) was poured
via canulation and the mixture was refluxed for 1.5 h.
Immediately the solution colour turned from light yel-
low to dark brown. After cooling to r.t., the brown
solution was filtered off and concentrated to dryness
under vacuum. The dark crude product was washed
with pentane and then recrystallised in THF and left at
−20 °C for 2 days before filtration (crude yield 80%).
Anal. Found: C, 57.8; H, 6.2; N, 6.4. Calc. for
C39H54CuF6N4OP: C, 58.3; H, 6.8; N, 7.0%. IR (KBr
pellets, w¯max cm−1): 2962s, 2927s and 2869m (wCH);
2270vw (wCN,CH CN), 1710m, 1635m and 1589s (wCN
imine); 1464s, 14430s, 1370s, 1327m, 1310m, 1255s, 1242s,
1210m, 1144w, 1102m, 1059m, 1043m, 1017m, 936w,
842s, vb (wPF), 815s, 775m, 756m, 740w, 716w, 602vw,
All reactions were performed in Schlenk tubes under
a dry dioxygen-free dinitrogen atmosphere. Solvents
were distilled using standard techniques and were thor-
oughly deoxygenated before use. Elemental analyses
were performed by the ‘Service Central d’Analyses du
CNRS’, Vernaison, France. IR spectra were obtained
with a Nicolet Nexus spectrometer (KBr pellets). ESR
spectra were run on a Bruker Elaxys spectrometer
(X-band). NMR spectra were recorded with a Bruker
AMX 400 MHz. UV–Vis spectra were recorded on an
Anthe´lie Secoman spectrometer. The solid state mag-
netic susceptibility measurements were carried out on
powder samples using either a Gouy balance Johnson
Matthey at room temperature (r.t.) or a commercial
SQUID magnetometer from Quantum Design, from 2
to 250 K. The susceptibilities were corrected for the
intrinsic diamagnetism of the sample container.
The electrochemical measurements were performed, un-
der Ar, using a microAutolab apparatus from Eco-
1
558s (dPF), 447vw, 390vw. NMR H(THF-d8, d ppm):
1.16 (d, 24, CH(CH3)2), 2.35 (s, 3, CH3CN), 2.42 (s, 6,
NꢁCCH3), 2.81 (m, 4, CH(CH3)2), 7.18 (m, 4, Haryl),
7.26 (d, 2, Haryl), 8.33 (d, 2, HPy), 8.52 (t, 1, HPy).
chemie (Roucaire), and
a
three compartment
micro-electrochemical cell equipped with a rotating disk
electrode (Metrohm) as a working electrode (Pt disk,
¥=0.3 cm). The reference potential was that of the
ferrocenium/ferrocene (Fc+/Fc) couple. The solvents
were commercial acetonitrile and THF; tetrabutylam-
monium hexafluorophosphate was the supporting elec-
2.2.3. [CuCl(C33H43N3)]PF6 (3)
Compound 2 (0.50 g, 0.68 mmol) was dissolved in
chloroform (30 mL) at r.t. The mixture was left under
stirring for 1 h. After filtration and concentration under
vacuum, the resulting solution was kept at r.t. for 3
days. Compound 3 slowly precipitated as dark-orange
microcrystals, which were filtered and dried (yield 85%).
Anal. Found: C, 53.6; H, 5.8; N, 5.7. Calc. for
C33H43CuF6N3P·0.1CHCl3: C, 53.9; H, 5.9; N, 5.7%.
(KBr pellets, w¯max/cm−1): 2961s, 2929m and 2870m
trolyte
purified
by
standard
techniques.
[Cu(CH3CN)4]PF6 [7] and the ligand L [8] were pre-
pared as described in the literature. CuBr was pur-
chased from Aldrich.

302
B. Le Gall et al. / Inorganica Chimica Acta 324 (2001) 300–308
(wCH); 1621m and 1592s (wCN imine); 1463m, 1441w,
1383w, 1375m, 1323w, 1266s, 1219w, 1181w, 1104m,
1094m, 1058m, 1042m, 989w, 938m, 830vs–vb, 814s,
800s, 778s, 760m, 741m, 558s, 456vw, 435w, 405m,
370w, 327vw, 266w. v298 K=1.82 mB.
bon atoms of the solvent. Because the molecular ge-
ometry found for the cation in 4 closely resembles those
observed in 1 and 3 and because the formation of 4 is
interesting from a chemical point of view, we decided to
introduce this structure in the actual paper, in spite of
rather bad final statistics. Crystallographic data and
final discrepancy factors are gathered in Table 1.
Single crystals of 3 were obtained in the NMR tube
1
used to run the H NMR spectra of 2 in CDCl3.
2.2.4. [CuBr(C33H43N3)]Br3 (4)
Compound 2 (1.05 g, 1.43 mmol) was dissolved at r.t.
with stirring in bromoform (30 mL). After filtration
and concentration under vacuum, the resulting solution
was kept at r.t. Compound 4 slowly precipitated as
orange brown crystals, which were filtered and dried
(yield 85%). Although the crystal structure determina-
tion showed a formula 4·2CHBr3, the elemental analy-
sis is consistent with a lower amount of bromoform.
Anal. Found: C, 33.7; H, 3.5; N, 3.4. Calc. for
C33H43Br4CuN3·1.5CHBr3: C, 33.3; H, 3.6; N, 3.4%. IR
(KBr pellets, w¯max cm−1): 2967s, 2925m and 2867m
(wCH), 1616m and 1586s (wCN imine), 1463m, 1443m,
1369s, 1321m, 1265s, 1216m, 1150m, 1092m, 1057m,
1040m, 985w, 938w, 839w, 801s, 778m, 760w,736w,
692w, 654s, 570w, 537w, 450w, 438w, 356w, 314w.
3. Results and discussion
3.1. Syntheses, spectroscopic and magnetic studies
Reaction of the copper(I) bromide CuBr with 2,6-
bis[1-(2,6-diisopropyl phenylimino)ethyl]pyridine (L) in
THF at reflux afforded a new copper salt 1 for which
the elemental analysis is in good agreement with the
formula Cu4Br6L2. If we consider in this complex the
presence of an usual L ligand in its neutral state, the
unexpected value for the Cu/Br ratio (4/6) indicates the
presence of copper in at least two oxidation states,
presumably Cu(I) and Cu(II). The presence of Cu(II)
was confirmed by magnetic studies since magnetic sus-
ceptibility measurements run in the range 2–250 K
reveal 1 to be paramagnetic and that it follows a Curie
2.3. X-ray structure analyses
Law with the Curie constant C of 0.82 emu K mol−1
,
Crystals of 1, 3 and 4·2CHBr3 suitable for X-ray
studies were obtained from CH2Cl2, CDCl3 and CHBr3
solutions, respectively, and were mounted on a Nonius
Kappa CCD diffractometer. The unit cell determina-
tions and data collections were carried out with Mo Ka
a value typical of two magnetically diluted copper(II)
centres (S=1/2). The EPR powder spectrum of com-
pound 1 recorded at r.t. displays a signal centred at
g=2.096, besides, the EPR spectrum of a frozen solu-
tion of 1 (CH2Cl2, 150 K) exhibits the usual patterns of
a Cu(II) atom in an elongated axial environment (half
,
radiation (0.71073 A) at low temperature (110 K). The
measured intensities were reduced with DENZO program
[9]. The structures were solved via direct methods (1, 3)
or Patterson (4) with SHELXS-97. All models were fur-
ther refined with full-matrix least-squares methods
filled dx − 2 level) i.e. g//=2.129, gÞ=2.063, A//=170
2
y
G. Since the X-ray studies indicate the asymmetric unit
consists of [CuBrL] and [Cu2Br4] moieties (see below),
these data indeed agree with the formula 2[CuIIBrL]+
[CuI2Br4]2− (Scheme 2). Co-ordination of the tridentate
ligand to a copper centre was clearly indicated by an
important shift of the wC=imine from 1644 to 1585 cm−1
on the infrared spectrum (KBr pellets).
(
SHELXL-97) based on ꢀF2ꢀ [10a]. All non-hydrogen
atoms were refined with anisotropic thermal parame-
ters. The hydrogen atoms were included in calculated
positions and refined with a riding model. While the
determinations of the crystal structures of 1 and 3 have
been carried without particular problems (for 3, the
absolute-structure parameter [10b] of 0.003(7) confirms
the absolute structure reported here), that of 4 merits
some comments. The crystals of 4 were of poor quality
(large reflections) and very thin (0.6×0.5×0.04 mm).
Attempts to cut smaller (more regular) samples failed.
Moreover, the measured crystal has been accidentally
lost before the treatment of data, and consequently, no
absorption correction could be applied. The structure
of 4·2CHBr3 has been deduced from Patterson synthesis
which allowed to locate the six bromine atoms of two
solvent bromoform molecules. Further difference
Fourier analyses clearly revealed at first the Br3− anion
and subsequently all atoms of the cation and the car-
With the aim to better understand the behaviour of
copper(I) complexes towards the ligand L, we planned
to study the reaction of
L
with the salt
[Cu(CH3CN)4]PF6. This reaction, carried out by reflux-
ing stoichiometric amounts of the two compounds in
THF, afforded a new derivative for which elemental
analysis is in good agreement with the following for-
mula CuL(CH3CN)(THF)PF6 (2) (Scheme 2). In addi-
1
tion to the usual tridentate ligand features [9], the H
NMR spectrum of 2, recorded in THF-d8, exhibits a
supplementary resonance at 2.35 ppm indicative of the
presence of a co-ordinated acetonitrile molecule. This
observation is corroborated by the very weak band
pointed out at 2270 cm−1 on the infrared spectrum
(KBr pellets). Furthermore, a strong band at 1589

B. Le Gall et al. / Inorganica Chimica Acta 324 (2001) 300–308
303
Table 1
Crystal data and structure refinement for compounds [CuBrL]2[Cu2Br4] (1), [CuClL]PF6 (3) and [CuBrL]Br3 (4)
Compound
1
3
4·2CHBr3
a
Empirical formula
M
C33H43Br3Cu2N3
848.51
C33H43ClCuF6N3P
725.66
C35H45Br10CuN3
1370.38
Crystal system
Space group
Crystal colour, habit
monoclinic
P21/a
red, prism
18.721(1)
8.854(1)
monoclinic
P21
red, irregular
8.7956(2)
11.9903(2)
16.8802(4)
triclinic
(
P1
green, plate
10.1152(3)
10.4189(2)
21.8128(7)
102.808(1)
92.071(1)
97.678(1)
2216.4(2)
2
,
a (A)
,
b (A)
,
c (A)
21.103(1)
h (°)
i (°)
k (°)
108.143(5)
101.53(2)
3
,
V (A )
3324.0(4)
4
1.696
4.913
1700
1744.30(6)
2
1.382
0.807
Z
Dcalc (mg m−3
)
2.053
9.535
1310
0.6×0.5×0.04
2.03–29.15
Absorption coefficient (mm−1
F(000)
)
754
Crystal size (mm−3
)
0.2×0.15×0.1
2.29–26.68
−23+22, −11+8, −26926
9518/6281/4154 [R(int)=0.0537] 6659/6659/6535
6281/0/367
0.45×0.25×0.15
2.88–27.54
911, −15+14, 921 −11+13, −9+13, 929
11350/10126/8143 [R(int)=0.0611]
q Range (°) for coll.
hkl/ranges
Reflections collected/unique/gr[I\2|(I)]
Data/restraints/parameters
Final R indices
6659/1/406
10126/0/442
R1=0.1029,
wR2=0.2685
R1=0.1215,
wR2=0.2931
0.2, 0.0
R1 b=0.0561,
wR2 c=0.0962
R1 b=0.1092,
wR2 c=0.1143
0.0195, 17.6244
1.053
R1=0.0256,
wR2=0.0672
R1=0.0264,
wR2=0.0681
0.0308, 0.4815
1.083
[I\2|(I)]
R indices (all data)
W
d(a, b)
Goodness-of-fit on F2
1.217
2.832, −4.076
−3
,
zmax, zmin (e A
)
0.736, −0.674
0.303, −0.410
a The empirical formula corresponds to the asymmetric unit which is consistent to the half of the chemical formula.
b R1=ꢁ(ꢀFoꢀ−ꢀFcꢀ)/ꢁꢀFoꢀ.
c wR2=[ꢁw(Fo2−Fc2)2/ꢁ[w(Fo2)2]1/2
.
d w=1/[|2(Fo2)+(aP)2+bP] with P=(Fo2+2Fc2)/3.
Scheme 2.
cm−1 attributable to the CN imine bond reveals the
co-ordination of the tridentate ligand to the copper
centre. Any effort to obtain suitable single crystals for
an X-ray analysis was unsuccessful, the structure for
this compound remains unresolved so far it might be
polynuclear since the FAB mass spectrum, not clearly
elucidated, displayed peaks at m/z values higher than
those expected for mononuclear species.
spectrum of complex 2 from a CDCl3 solution failed
because it reacted with the solvent affording a para-
magnetic derivative [CuClL]PF6 (3) for which single
crystals came out from the deuterated solvent (Scheme
2). Basically on its IR spectrum we observed the disap-
pearance of the band at 1710 cm−1 and a slight shift of
the band characteristic of the imine function from 1589
to 1592 cm−1. The complex 3 can be also prepared
from CHCl3 solutions of 2.
It is noteworthy that our attempts to obtain a NMR

304
B. Le Gall et al. / Inorganica Chimica Acta 324 (2001) 300–308
Similar experiments conducted at r.t. in CHBr3 af-
presence of two cationic entities [CuBrL]+ related to each
other by an inversion centre and one centrosymmetric
dimer [Cu2Br4]2− (Fig. 1). For compound 3 (monoclinic
space group P21), the asymmetric unit consists of a
[CuClL]+ cation and a PF6− anion (Fig. 2), and for
forded a new copper(II) salt 4 which analysed consistently
forCuLBr4·1.5CHBr3; its X-ray analysis actually revealed
the presence of a tribromide anion Br3− (see below).
(
3.2. Crystal structures
compound 4·2CHBr3 (triclinic space group P1), it con-
sists of a [CuBrL]+ cation, a Br3− anion and two CHBr3
Compound 1 crystallises in the monoclinic space group
P21/a and its crystal structure is consistent with the
molecules of crystallisation, all in general positions (Fig.
3).
Fig. 1. Perspective drawing of the compound [CuBrL]2[Cu2Br4] (1) showing the atom numbering. Thermal ellipsoids are drawn at the 50%
probability level.
Fig. 2. ORTEP representation of the asymmetric unit for compound [CuClL]PF6 (3). Thermal ellipsoids are drawn at the 50% probability level.

B. Le Gall et al. / Inorganica Chimica Acta 324 (2001) 300–308
305
Fig. 3. ORTEP representation of the asymmetric unit for compound [CuBrL]Br3 (4). Thermal ellipsoids are drawn at the 50% probability level.
usual s CꢀC bond distances. This indicates a high p
electronic delocalisation within the terdentate ligand.
In compound 1, the centrosymmetric [Cu2Br4]2− an-
ion is alike to those of the [Me3PhN]2[Cu2Br4] and
[Et4N]2[Cu2Br4] salts previously described by Jagner et
al. [11]. The copper(I) atoms co-ordination corresponds
to a strongly distorted planar triangle characterised by
The principal bond lengths and angles for the
[CuXL]+ cations of 1, 3 and 4 are listed in Table 2.
Obviously, these three cations present similar features
with an essentially planar CuN3X co-ordination polyhe-
dron of the copper(II) atom (Table 2). Note, however,
that if in 1 the N3X tetra-atomic system is exactly
planar, it exhibits a slight butterfly-like deformation in
3 and 4. Although the ML fragment is not bisected by
a crystallographic mirror plane, the two halves of the
co-ordinated ligand exhibit almost similar metric fea-
tures and therefore the [CuXL]+ cation may reach the
Cs or even higher C2v molecular symmetry (Table 2).
The bond angle values around the metal atom, in the
range 78–102°, are indicative of strong distortions of
the co-ordination polyhedron, with respect to the
square, arising from the bite of the terdentate ligand.
The metalꢀligand bond lengths, which include
CuꢀNimine bond distances somewhat longer than the
CuꢀNpy bond length, are usual and in good agreement
whith those previously reported [4]. For instance, values
observed for the CuꢀNpy bond distances, labelled a in
,
CuꢀBr distances varying from 2.325(2) to 2.400(2) A
,
(vs. 2.310(1)–2.417(1) A in Me3PhN salt, 2.319(2)–
,
,
2.454(2) A in Et4N salt and 2.337(2)–2.454(1) A in the
MePh3P salt) and BrꢀCuꢀBr angles in the range
114.26–123.62° (Table 3). However, the Cu(I)···Cu(I)
,
distance 2.601(2) A is clearly shorter than the corre-
sponding values in these salts (2.937(3), 2.738(9) and
,
2.697(2) A).
In compound 1, the short contact observed between
the copper atom of the cation and one of the bromo
,
ligand of the anion (Cu(1)···Br(2) 2.689(2) A) corre-
sponds to the well known tendency of square planar
copper(II) to extend its co-ordination number up to five
or even six [12]. As a result, the whole structure must be
viewed as built of ‘tetranuclear’ units (Fig. 1).
,
Table 2, fall in the range 1.913(7)–1.938(5) A; they are
close to those observed for other Cu(II) complexes with
almost similar organic ligands L% and L¦ such as
In compound 3, the PF6− anion displays a regular
octahedral geometry characterised by PꢀF distances in
,
,
[CuL%(NO3)2] (1.913(3) A) [4a] and [CuL¦](PF6)2
the range 1.590(2)–1.608(2) A and FꢀPꢀF angles vary-
2
,
(1.9419(11) and 1.9561(11) A) [4b] (Scheme 1). As in
ing from 89.42(7) to 90.77(8)°. A weak interaction
these compounds, in complexes 1, 2 and 4, the
CimineꢀNimine f bond lengths are significantly shorter
(approximately 0.05 A) than the CpyꢀNpy d distances,
arises between the anion and the copper(II) atom as
,
shown by the Cu···F(2) distance (2.645(2) A).
In compound 4, the Br3− anion is almost linear
(BrBrBr=175.34(6)°) and symmetrical (BrBr 2.250(2)
,
consistent with the double-bond character of the former
(Table 2). The CpyꢀCimine e bond lengths, and even
those of CimineꢀCH3 g, are significantly shorter than
,
and 2.255(2) A). It is well known that this anion, like
the I3− anion, may be symmetrical or unsymmetrical

306
B. Le Gall et al. / Inorganica Chimica Acta 324 (2001) 300–308
depending on its environment [13]. As in derivative 3 a
weak interaction is observed between the anion and the
peak at −0.70 V [14]; furthermore, in the case of
compound 4, the reduction peak of CHBr3 is combined
with the copper deposit peak. The redox processes
located at potentials higher than 0.0 V correspond to
the oxidation of Br− or Br3−, and in the case of
compound 1 also correspond to that of the [CuI2Br4]2−
anionic species. This has been verified by independent
CV but the exact assignation of theses processes has not
been scrutinised. Therefore, the redox process com-
prised in both cases between −0.5 and 0.0 V corre-
sponds to the Cu(II)/Cu(I) exchange for the [CuBrL]+
moiety (peaks A/A%) (Fig. 4). The Cu(II) redox state in
this moiety was emphasized by the reduction wave
observed by rotating disk electrode voltammetry (not
shown). The cyclic voltammograms (Figs. 4 and 5) were
recorded starting from a potential more cathodic than
the A/A% system (−0.5 V), thus the Cu(II) moiety was
converted into the Cu(I) complex in the diffusion layer
before the run. As exemplified in Fig. 5, at low scan
,
copper(II) atom (Cu···Br(2) 3.141(2) A).
3.3. Electrochemical studies
For
complexes
[CuBrL]2[Cu2Br4]
(1),
CuL(CH3CN)(THF)PF6 (2) and [CuBrL]Br3 (4), cyclic
voltammetry (CV) and rotating disk electrode (RDE)
voltammetry studies were performed in CH3CN or
THF in order to evaluate their redox behaviour; every
attempt of characterisation of compound [CuClL]PF6
(3) failed since decomposition occurred in any solvent
tested.
For complexes 1 and 4, a typical voltammogram
displays several redox processes as shown in Fig. 4. On
the cathodic scan (EB −0.5 V), the electrode re-
sponses are ascribed to the reduction of Cu(I) to Cu(0)
with copper deposit, as indicated by the redissolution
Table 2
a
[CuLX]+ cation in 1, 3 and 4: selected bond lengths a (A), bond angles (°) and planarity b
,
1
3
4
Cu co-ordination
a
b
c
1.938(5)
2.071(5)
2.328(2) c
78.4(2)
99.6(2)
1.922(2)
2.070(2)
1.913(7)
2.067(6)
2.325(2) c
79.7(3)
2.133(5)
2.062(2)
2.079(6)
2.1450(5) d
79.14(6)
h
78.1(2)
99.6(2)
79.40(6)
99.64(5)
78.8(3)
99.8(2)
i
101.74(5)
101.0(2)
b
Planarity and de6iations
N1
N2
N3
Br1
Cu
Br2
0.001(2)
N1
N2
N3
Cl
Cu
F2
0.0594(9)
−0.0480(7)
−0.0476(7)
0.0361(5)
0.0609(7)
2.674(2)
N1
N2
N3
Br1
Cu
Br2
0.073(4)
−0.058(3)
−0.057(3)
0.041(2)
0.130(3)
3.182(3)
−0.001(3)
−0.001(2)
0.001(2)
0.292(2)
2.956(3)
L ligand
d
e
f
g
h

l
m
1.316(8)
1.460(9)
1.287(8)
1.486(9)
1.449(8)
118.1(4)
113.2(6)
115.9(6)
113.2(5)
125.6(4)
1.336(8)
1.467(9)
1.273(8)
1.483(9)
1.452(8)
118.6(4)
114.1(6)
115.5(6)
113.7(4)
126.7(4)
1.335(2)
1.495(2)
1.285(3)
1.492(3)
1.446(2)
118.7(2)
112.8(2)
115.3(2)
113.8(2)
125.8(2)
1.333(2)
1.483(2)
1.294(2)
1.484(3)
1.447(2)
118.8(2)
112.4(2)
114.7(2)
114.7(2)
122.8(2)
1.325(9)
1.36(1)
1.48(1)
1.30(2)
1.50(2)
1.28(1)
1.49(1)
1.439(9)
1.51(2)
1.431(9)
120.1(5)
111.6(6)
116.9(6)
112.9(5)
126.7(5)
118.7(5)
110.9(6)
115.0(6)
114.1(5)
125.8(5)
k
p
a Since the ML fragment is not bisected by a crystallographic mirror plane, two values corresponding to the two halves of the ligand are given
when appropriate.
b Distance of the atom from the mean plane N1N2N3X.
c X=Cl.
d X=Br.

B. Le Gall et al. / Inorganica Chimica Acta 324 (2001) 300–308
307
nated species [CuIL]+; while at high scan rate the
oxidation of both species in equilibrium can be seen
(systems A/A% and B/B%), at lower scan rate only the
easiest oxidation of [CuIBrL] into [CuIIBrL]+ at E%°(2)
can be observed via the CE process. This interpretation
is corroborated by the observation that the peak of the
second process disappears in the presence of an excess
of Br− anion in the same conditions for CV. A higher
stability of the [CuIIBrL]+ derivative as compared to
[CuIBrL] can be easily rationalised both in terms of
charge stabilisation and Cu co-ordination (square vs.
tetrahedral).
The Cu(I) compound CuL(CH3CN)(THF)PF6 (2) is
quite unstable in acetonitrile in which decomplexation
of the L ligand occurs very quickly; however, this
complex could be electrochemically characterised in
THF. The corresponding CV displays two quite close
redox processes, one pseudo-reversible for the oxidation
of Cu(I) (E%°= +0.23 V; DEp=190 mV), the other
irreversible at Ep (red)= −0.07 V to the reduction of
Cu(I) into Cu(0) with deposit.
Table 3
,
Selected bond lengths (A) and bond angles (°) associated to the
[Cu2Br4]2− anion and the anion–cation interactions in 1
Bond lengths
Br(2)ꢀCu(2)
Br(3)ꢀCu(2)
Cu(2)ꢀCu(2)i
2.325(2)
2.391(2)
2.601(2)
Br(2)ꢀCu(1)
Cu(2)ꢀBr(3)i
2.689(2)
2.400(2)
Bond angles
Br(2)ꢀCu(2)ꢀBr(3)
122.09(5) Br(2)ꢀCu(2)ꢀBr(3)i
123.62(5)
65.74(4)
78.1(2)
164.2(2)
99.6(2)
Cu(2)ꢀBr(2)ꢀCu(1) 124.59(4) Cu(2)ꢀBr(3)ꢀCu(2)i
N(1)ꢀCu(1)ꢀN(3)
N(3)ꢀCu(1)ꢀN(2)
N(3)ꢀCu(1)ꢀBr(1)
N(1)ꢀCu(1)ꢀBr(2)
N(2)ꢀCu(1)ꢀBr(2)
78.4(2)
153.2(2)
99.6(2)
91.7(2)
93.1(2)
N(1)ꢀCu(1)ꢀN(2)
N(1)ꢀCu(1)ꢀBr(1)
N(2)ꢀCu(1)ꢀBr(1)
N(3)ꢀCu(1)ꢀBr(2)
Br(1)ꢀCu(1)ꢀBr(2)
100.1(2)
104.15(4)
Br(3)ꢀCu(2)ꢀBr(3)i 114.26(4) Br(2)ꢀCu(2)ꢀCu(2)i 178.13(6)
Br(3)ꢀCu(2)ꢀCu(2)i
57.30(4) Br(3)iꢀCu(2)ꢀCu(2)i
56.95(4)
Symmetry transformation used to generate equivalent atoms: (i):
−x, −y+1, −z+1.
3.4. Discussion
As reported above, the three new Cu(II) complexes
1, 3 and 4, which have been fully characterised by
X-ray crystal structure determinations, are prepared
from Cu(I) starting materials.
Reproductive formation of 1 in a rather good yield
by reaction of CuBr with the ligand in THF, thus by
concomitant oxidation of the copper(I), is not ex-
plained away since (i) the starting THF solution of
CuBr is pure as regard to copper(II) (ESR) with only
traces of paramagnetism; (ii) the synthesis was per-
formed under oxygen free atmosphere; (iii) the ligand
and the solvent do not possess significant oxidant prop-
erties; and (iv) no trace of metallic copper indicative of
a dismutation process was found.
Fig. 4. Cyclic voltammogram of compound 1 (approximately 10−3 M
in MeCN, 0.2 M Bu4NPF6, platinum electrode, scan rate 0.1 V s−1).
rate, 6, (up to 6=0.1 V s−1), a single reversible redox
process is seen (peaks A/A%: E%°(1)= −0.24 V; DEp=
140 mV; ipa/ipc:1) for both compounds 1 and 4; at
higher scan rates, (0.2B6B1.0 V s−1), the peaks inten-
sity of process A/A% decreases giving rise to another
reversible redox process (peaks B/B%: E%°(2)= −0.08 V;
DEp=120 mV). This scan rate dependent behaviour
appears typical of a CE process, for which we propose
the following square scheme involving equilibria be-
tween bromo and non-bromo CuIL and CuIIL species
(Scheme 3):
Scheme 3.
Along with this scheme, at the starting potential,
[CuIBrL] would be in equilibrium with the unbromi-
Fig. 5. Cyclic voltammograms of compound 1 at different scan rates
(a: 6=0.1, b: 6=0.2, c: 6=0.5, d: 6=1.0 V s−1).

308
B. Le Gall et al. / Inorganica Chimica Acta 324 (2001) 300–308
(b) E.C. Alyea, G. Ferguson, R.J. Restivo, P.H. Merrell, J.
For both compounds 3 and 4, the formation of a
Chem. Soc., Chem. Commun. (1975) 269;
(c) E.C. Alyea, G. Ferguson, R.J. Restivo, Inorg. Chem. 14
(1975) 2491;
(d) E.C. Alyea, P.H. Merrell, Inorg. Chim. Acta 28 (1978) 91;
(e) P.H. Merrell, E.C. Alyea, L. Ecott, Inorg. Chim. Acta 59
(1982) 25;
(f) E.C. Alyea, Inorg. Chim. Acta 76 (1983) L239;
(g) D.A. Edwards, M.F. Mahon, W.R. Martin, K.C. Molloy,
P.E. Fanwick, R.A. Walton, J. Chem. Soc., Dalton Trans. (1990)
3161.
copper(II) complex results from an halogen atom trans-
fer from an halogenated solvent to a copper(I) species.
It seems likely that these reactions proceed via a radical
mechanism [15] considering the previously reported for-
mation of the copper(II) derivative [Cu(AA)2Br]PF6)
(AA=4,4%-(5-nonyl)-2,2%-bipyridine) by reaction at r.t.
in acetonitrile between the copper(I) complex
[Cu(AA)2]PF6 and one equivalent of CHBr3. On the
other hand, formation of the tribromide Br3− anion in
4 with bromoform as the only source of bromine is
quite unexpected.
The crystal structure of 1 (despite the anion–cation
short contact via a Br bridge), its magnetic properties
and colour clearly show that, as a mixed valence
Cu(I)ꢀCu(II) complex, this compound belongs to the
class I of the Robin and Day classification [16].
Previous studies have shown that the [Cu2Br4]2−
anion is capable of considerable flexibility with respect
to its geometry under different steric and electrostatic
constraints in the solid state [11]. As far as we know,
[2] (a) G.J.P. Britovsek, V.C. Gibson, B.S. Kimberley, P.J. Maddox,
S.J. McTavish, G.A. Solan, A.J.P. White, D.J. Williams, Chem.
Commun. (1998) 849;
(b) B.L. Small, M. Brookhart, A.M.A. Bennet, J. Am. Chem.
Soc. 120 (1998) 4049;
(c) D. Reardon, F. Conan, S. Gambarotta, G. Yap, Q. Wang, J.
Am. Chem. Soc. 121 (1999) 9318.
[3] See for instance: (a) E.C. Constable, T. Kulke, M. Neuburger,
M. Zehnder, Chem. Commun. (1997) 489;
(b) R.F. Carina, G. Bernardelli, A.F. Williams, Angew. Chem.,
Int. Ed. Engl. 32 (1993) 1463.
[4] (a) R.J. Restivo, G. Ferguson, J. Chem. Soc., Dalton Trans.
(1976) 518;
(b) B. de Bruin, E. Bill, E. Bothe, T. Weyhermu¨ller, K.
Wieghardt, Inorg. Chem. 39 (2000) 2936.
[5] L. Douce, A. El-Ghayoury, A. Skoulios, R. Ziessel, Chem.
Commun. (1999) 2033.
,
the Cu(I)···Cu(I) distance in 1 (2.601(2) A) is the short-
est ever seen for this anion. However, much shorter
Cu(I)···Cu(I) distances have been previously reported
for instance in the trinuclear complex [Cu3(RN5R)3] (R:
[6] J.M. Holland, X. Liu, J.P. Zhao, F.E. Mabbs, C.A. Kilner, M.
Thornton-Pett, M.A. Halcrow, J. Chem. Soc., Dalton Trans.
(2000) 3316.
,
PhMe) (2.348(2) and 2.358(2) A) [17].
[7] G.J. Kubas, Inorg. Synth. 19 (1979) 90.
[8] See for example: (a) B. C¸ etinkaya, E. C¸ etinkaya, M. Brookhart,
P.S. White, J. Mol. Cat. A: Chem. 142 (1999) 101;
(b) B.L. Small, M. Brookhart, Macromolecules 32 (1999) 2120.
[9] Z. Otwinowski, W. Minor, Methods Enzymol. 276 (1997) 307.
[10] (a) G.M. Sheldrick, SHELXS and SHELXL97, University of Go¨ttin-
gen, Go¨ttingen, Germany;
(b) G. Bernardinelli, H.D. Flack, Acta Crystallogr. Sect. A 43
(1987) 75.
[11] (a) S. Andersson, S. Jagner, Acta Chem. Scand. Ser. A 39 (1985)
423;
4. Supplementary material
Full crystallographic data for reported complexes
have been deposited with the Cambridge Crystallo-
graphic Data Centre CCDC Nos. 162614–162616 for
compounds [CuBrL]2[Cu2Br4] (1), [CuClL]Pf6 (3), [Cu-
ClL]Br3 (4) respectively. Copies of this information may
be obtained free of charge from the Director, CCDC,
12 Union Road, Cambridge CB2 1EZ, UK (fax: +44-
1223-336-033; e-mail: deposit@ccdc.cam.ac.uk or www:
http://www.ccdc.cam.ac.uk).
(b) M. Asplund, S. Jagner, Acta Chem. Scand. Ser. A 38 (1984)
135;
(c) S. Andersson, S. Jagner, Acta Crystallogr. Sect. C 31 (1987)
1089.
[12] (a) T. Ha, A.M. Larsonneur-Gallibert, P. Castan, J. Jaud, J.
Chem. Crystallogr. 29 (1999) 565;
(b) J. Costa, R. Delgado, G.B. Drew, V. Fe´lix, A. Saint-Maurice,
J. Chem. Soc. Dalton Trans. (2000) 1907.
[13] (a) F.A. Cotton, G.E. Lewis, W. Schwotzer, Inorg. Chem. 25
(1986) 3528;
(b) J.J. Mayerle, G. Wolmersha¨user, G.B. Street, Inorg. Chem.
18 (1979) 1161.
[14] S. Torelli, C. Belle, I. Gautheir-Luneau, J.L. Pierre, E. Saint-
Aman, J.M. Latour, L. Le Pape, D. Luneau, Inorg. Chem. 39
(2000) 3526.
Acknowledgements
CNRS and Universite´ de Bretagne Occidentale are
acknowledged for financial support and the ‘MENRT’
is acknowledged for a doctoral fellowship (BLG). We
thank Dr P. Molinie´ (IM Nantes, France) for magnetic
measurements and Dr S. Triki (Brest) for stimulating
and encouraging discussions.
[15] K. Matyjaszewski, T.E. Patten, J. Xia, J. Am. Chem. Soc. 117
(1997) 674.
[16] M.B. Robin, P. Day, Adv. Inorg. Chem. Radiochem. 10 (1967)
248.
References
[17] (a) K.M. Merz, R. Hoffmann, Inorg. Chem. 27 (1988) 2120;
(b) J. Beck, J. Strahle, Angew. Chem. Int. Ed. Eng. 24 (1985)
409.
[1] (a) E.C. Alyea, P.H. Merrell, Syn. React. Inorg. Metal-Org.
Chem. 4 (1974) 535;