Copper complexes of 2,6-bis(iminomethyl)pyridine derivatives and of l,3-bis(pyridin-2-yl)pyrazole. Effects of ligand bulk and conformational strain on the ground state of a six-co-ordinate copper(n) ion

Article abstract of DOI:10.1039/b005105p

Complexation of hydrated Cu(BF4)2 by 2 molar equivalents of L2R (L2Me = 2,6-bis{methyliminomethyl}pyridine, L2Cy = 2,6-bis{cyclohexyliminomethyl}pyridine, L2Bu' = 2,6-bis{to7-butyliminomethyl}pyridine) afforded [Cu(L2R)2][BF4]2 (R = Me, 1; Cy, 2; or Bu', 3) in moderate yields. EPR spectroscopy in solution and the solid state demonstrates that 1 and 2 adopt the expected {dy ji}' electronic ground state, and that the pseudoJahn-Teller elongation axis is fluxional in solid 1 and static in solid 2. In contrast, 3 exhibits a {dzi}' ground state by EPR. The crystal structures of 2-MeNO2 and 3-|Me2CO contain rhombic six-co-ordinate copper(n) ions, which differ principally in the length of the Cu-N bonds close to the molecular y axis. Treatment of [Cu(NCMe)4]BF4 with varying ratios of L2R yields [Cu(L2R)2]BF4 (R = Me, 4; Mes, 5; or Cy, 6: L2Mes = 2,6-bis{2,4,6-trimethylphenyliminomethyljpyridine) or [{Cu(L2Bu')}2][BF4]2 7. The single crystal structure of 5-MeCN shows a flattened tetrahedral copper(i) centre with two bidentate L2Mes ligands. Reaction of hydrated Cu(ClO4)2 with 2 molar equivalents of L3 (l,3-bis{pyridin-2-yl}pyrazole) yields [Cu(L3)2][ClO4]2 8, whose crystal structure demonstrates a six-co-ordinate copper(n) ion with long Cu-N{pyridine} bonds. Powder EPR spectroscopy demonstrates a {dj}1 ground state for 8 in the solid. The Royal Society of Chemistry 2000.

Full text of DOI:10.1039/b005105p

View Online / Journal Homepage / Table of Contents for this issue
Copper complexes of 2,6-bis(iminomethyl)pyridine derivatives
and of 1,3-bis(pyridin-2-yl)pyrazole. Effects of ligand bulk and
conformational strain on the ground state of a six-co-ordinate
copper(II) ion
Joanne M. Holland,a Xiaoming Liu,a Jing P. Zhao,b Frank E. Mabbs,b Colin A. Kilner,a
Mark Thornton-Petta and Malcolm A. Halcrow*a
a School of Chemistry, University of Leeds, Woodhouse Lane, Leeds, UK LS2 9JT.
E-mail: M.A.Halcrow@chem.leeds.ac.uk
b CW EPR Service Centre, Department of Chemistry, University of Manchester, Oxford Road,
Manchester, UK M13 9PL
Received 26th June 2000, Accepted 2nd August 2000
First published as an Advance Article on the web 18th September 2000
Complexation of hydrated Cu(BF4)2 by 2 molar equivalents of L2R (L2Me = 2,6-bis{methyliminomethyl}-
pyridine, L2Cy = 2,6-bis{cyclohexyliminomethyl}pyridine, L2But = 2,6-bis{tert-butyliminomethyl}pyridine)
afforded [Cu(L2R)2][BF4]2 (R = Me, 1; Cy, 2; or But, 3) in moderate yields. EPR spectroscopy in solution and the
1
2
2
solid state demonstrates that 1 and 2 adopt the expected {dy Ϫ z } electronic ground state, and that the pseudo-
1
2
Jahn–Teller elongation axis is fluxional in solid 1 and static in solid 2. In contrast, 3 exhibits a {dz } ground state
by EPR. The crystal structures of 2ؒMeNO2 and 3ؒ¹Me2CO contain rhombic six-co-ordinate copper() ions,
¯
²
which differ principally in the length of the Cu–N bonds close to the molecular y axis. Treatment of [Cu(NCMe)4]-
BF4 with varying ratios of L2R yields [Cu(L2R)2]BF4 (R = Me, 4; Mes, 5; or Cy, 6: L2Mes = 2,6-bis{2,4,6-trimethyl-
phenyliminomethyl}pyridine) or [{Cu(L2But)}2][BF4]2 7. The single crystal structure of 5ؒMeCN shows a flattened
tetrahedral copper() centre with two bidentate L2Mes ligands. Reaction of hydrated Cu(ClO4)2 with 2 molar
equivalents of L3 (1,3-bis{pyridin-2-yl}pyrazole) yields [Cu(L3)2][ClO4]2 8, whose crystal structure demonstrates a
1
2
six-co-ordinate copper() ion with long Cu–N{pyridine} bonds. Powder EPR spectroscopy demonstrates a {dz }
ground state for 8 in the solid.
Introduction
The Jahn–Teller effect1 in [CuL2]2ϩ (L = meridional tridentate
ligand) complexes manifests itself as an elongation along the
1
2
2
molecular x axis (Scheme 1), affording a {dy Ϫ z } electronic
ground state. In the solid state this elongation is often dynamic-
ally disordered over the molecular x and y axes, which leads to
unusual, temperature-dependent EPR behaviour.2–6 We have
discovered that a unique change in ground state can be enforced
at the complexes [Cu(L1R)2]2ϩ by introduction of bulky ‘R’ sub-
stituents, which sterically inhibit this Jahn–Teller fluxionality
through steric interactions between the ‘R’ groups on one lig-
and and the backbone of the other perpendicular one (Scheme
1
1).5,7 Hence, [Cu(L1H)2]2ϩ shows the usual {dy Ϫ z } ground
2
2
state by EPR spectroscopy, while [Cu(L1Mes)2]2ϩ adopts a
1
2
{dz } ground state, which should correspond structurally to a
identical ligand donors that exhibit different electronic ground
states. The copper() complex chemistry of these and other L2R
ligands is also presented. Finally, we report the properties of
compression along the molecular z axis (Scheme 1, A).7 A
meaningful structural comparison of the two compounds was
impossible, however, because the solid state fluxionality shown
by [Cu(L1H)2]2ϩ 5,6 prevents its true Cu–N distances from being
determined.8
[Cu(L3)2]2ϩ, which was synthesized with a view to obtaining a
1
2
6-co-ordinate {dz } copper() complex without the use of
bulky ligand substituents. The molecular axis convention
shown in Scheme 1 is used throughout the following discussion;
in those crystal structures that show a rhombic stereochemistry,
the long, medium and short Cu–N bonds are assumed to lie
close to the molecular x, y and z axes, respectively.
There is only one previous example of a 6-co-ordinate, tetra-
gonally compressed copper() centre with a homoleptic donor
set, in the co-ordination polymer KAlCuF6.9 We therefore
wished to examine the generality of our result, by preparing
copper() complexes of other meridional tridentate ligands
whose distal steric bulk could readily be modified. We describe
here the syntheses, structures and spectroscopy of a series of
complexes [Cu(L2R)2]2ϩ (R = Me, Cy or But). Included in this
work are the crystal structures of [Cu(L2Cy)2][BF4]2 and
[Cu(L2But)2][BF4]2, which for the first time allow a genuine
structural comparison between copper() complexes with
Results and discussion
Copper(II) complexes of L2R
Surprisingly few [M(L2R)2]nϩ (Mnϩ = transition ion) complexes
have been reported previously,10–13 and no [Cu(L2R)2]2ϩ
3316
J. Chem. Soc., Dalton Trans., 2000, 3316–3324
This journal is © The Royal Society of Chemistry 2000
DOI: 10.1039/b005105p

View Online
Table 1 Selected Q-band EPR data for the copper() complexes in this study. Solution spectra were run in 10:1 MeCN–toluene. Spectra with two
identical g values have axial symmetry. All hyperfine couplings are to 63,65Cu and are in G. Estimated errors in these parameters are g 0.002, A 2 G
Phase
T/K
g1
g2
g3
A1
A2
A3
1 [Cu(L2Me)2][BF4]2
2 [Cu(L2Cy)2][BF4]2
3 [Cu(L2But)2][BF4]2
Powder
292
110
5b
2.179
2.176
2.182
2.230
2.252
2.252
2.256
2.252
2.231
2.231
2.219
2.220
2.220
2.220
2.221
2.153
2.133
2.090
2.230
2.083
2.081
2.073
2.081
2.202
2.202
2.219
2.215
2.196
2.200
2.221
2.068
2.068
2.060
2.090
2.043
2.042
2.040
2.042
2.005
2.000
1.998
2.001
2.011
2.004
2.001
—
—
—
—
135
144
154
144
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
155
155
161
165
141
155
170
a
Powder
292
110
10
110
292
110
10
110
292
117
10
MeCN
Powder
MeCn
Powder
—
—
—
—
8 [Cu(L3)2][ClO4]2
a
a The spectrum of this compound in MeCN at 110 K is poorly resolved, and could not be simulated accurately. See text for details. b Two signals
present in the spectrum, in approximately equal proportions. See text for details.
aldehyde with an excess of the appropriate amine at room tem-
perature or in refluxing methanol. Complexation of Cu(BF4)2ؒ
6H2O by 2 equivalents of L2Me, L2Cy or L2But in MeCN yielded
green solutions, from which green microcrystalline products
could be obtained in moderate yields by addition of diethyl
ether. These microcrystals analysed as the desired complexes
[Cu(L2R)2][BF4]2 (R = Me, 1; Cy, 2; or But, 3), while FAB mass
spectrometry showed a highest molecular ion corresponding
to the fragment [63Cu(L2R)2]ϩ (R = Me, m/z = 385; Cy, 657; But,
553). While 2 and 3 could be crystallised from MeNO2, MeCN
or acetone by vapour diffusion with Et2O, 1 could only be
obtained as a solid from MeCN–Et2O. Analogous complex-
ations employing L2Ph or L2Mes afforded deep brown solu-
tions, from which no clean solid products could be isolated (see
below).
In addition to π → π* absorptions in the UV region, two
d–d peaks are apparent in the UV/vis spectra of complexes
1 and 2 in MeCN at 298 K,18 at λmax = 709 (εmax = 48 {1} or 61
{2} MϪ1 cmϪ1) and ca. 480 nm (sh). These are very close
to those shown by [Cu(L4R)2]2ϩ complexes in solution.14 The
higher wavelength band lacks a high-wavelength shoulder,
which is typical for six-co-ordinate copper() centres.18 That
1 and 2 exhibit identical d–d maxima suggests the cyclohexyl
substituents in the latter compound are not sufficiently bulky
to cause inter-ligand steric repulsions of the type shown by
[Cu(L1R)2]2ϩ (R = Pri, Ph or Mes). However, for 3 a single
higher wavelength d–d absorption was observed, at λmax = 746
nm (εmax = 63 MϪ1 cmϪ1), showing that the ligand tert-butyl
substituents do perturb the copper ion. For all three complexes
these absorptions have essentially identical wavelengths and
intensities in MeCN and in MeNO2, which demonstrates that
minimal ligand dissociation occurs in these solvents.
All EPR spectra in this study were run at Q-band. In 10:1
MeCN–toluene solution at 120 K, complex 1 affords a broad
Scheme 1 Molecular axes for [Cu(L1R)2]2ϩ (A) and [Cu(L2R)2]2ϩ (B).
spectrum that can be simulated with g1 = 2.22, g2 = 2.10,
g3 = 2.07, with peak-to-peak linewidths of 400 G on g1 and g2.
This g1 > g2 ≈ g3 > ge pattern is characteristic for a pseudo-
1
2 2
complex has previously been characterised by EPR spectro-
scopy. In addition, we are only aware of two literature reports
of [Cu(L4R)2]2ϩ complexes, whose EPR spectra were also not
Jahn–Teller elongated copper() complex with a {dy
}
Ϫ z
ground state.19 The powder EPR spectrum of 1 is more com-
plex, however. At 290 K an ‘inverse’ g1 ≈ g2 > g3 > ge spectrum
is observed (Table 1). As the temperature is lowered this spec-
trum becomes more rhombic, which mostly reflects a lowering
of g2. No hyperfine coupling is apparent at any temperature.
measured.14 However, we noted that the d9 species [Ni(L4Bz)2]ϩ
1
2
2
(Bz = benzyl) shows a {dy Ϫ z } ground state by EPR spectro-
scopy.15 Hence, it seemed that R substituents more bulky than
benzyl would be required in [Cu(L2R)2]2ϩ to cause inter-ligand
repulsions of the type shown by [Cu(L1R)2]2ϩ (Scheme 1). As a
2
This behaviour is similar to that exhibited by [Cu(terpy)2][PF6]2
5
control, the complex [Cu(L2Me)2]2ϩ was also prepared.
and [Cu(L1H)2][BF4]2 in the powder, and is diagnostic of
17
1
The ligands L2Ph10,16 and L2But
have been reported
a {dy
}
copper() complex, whose pseudo-Jahn–Teller
2
2
Ϫ z
previously. Other L2R ligands (R = Me, Mes {mesityl} or Cy
elongation axis is fluxional in the solid state.1,7 Below 20 K the
spectrum cannot be simulated by a single copper() spin; a
{cyclohexyl}) were prepared by reaction of pyridine-2,6-dicarb-
J. Chem. Soc., Dalton Trans., 2000, 3316–3324
3317

View Online
Table 3 Selected bond lengths (Å) and angles (Њ) at copper in the single
Table 2 Selected bond lengths (Å) and angles (Њ) at copper in the single
crystal structure of [Cu(L2Cy)2][BF4]2ؒMeNO2 2ؒMeNO2
2
t
1
1
–
–
crystal structure of [Cu(L Bu )2][BF4]2ؒ2Me2CO 3ؒ2Me2CO
Cu(1)–N(1)
Cu(1)–N(2)
Cu(1)–N(3)
1.936(2)
2.085(2)
2.091(2)
Cu(1)–N(4)
Cu(1)–N(5)
Cu(1)–N(6)
2.004(2)
2.412(2)
2.443(2)
Molecule 1
Molecule 2
Cu(1)–N(1)
Cu(1)–N(2)
Cu(1)–N(3)
Cu(1)–N(4)
Cu(1)–N(5)
Cu(1)–N(6)
1.932(2)
2.295(2)
2.304(2)
1.958(2)
2.312(2)
2.450(2)
Cu(2)–N(7)
Cu(2)–N(8)
Cu(2)–N(9)
Cu(2)–N(10)
Cu(2)–N(11)
Cu(2)–N(12)
1.927(2)
2.278(2)
2.313(2)
1.934(2)
2.352(2)
2.349(2)
N(1)–Cu(1)–N(2)
N(1)–Cu(1)–N(3)
N(1)–Cu(1)–N(4)
N(1)–Cu(1)–N(5)
N(1)–Cu(1)–N(6)
N(2)–Cu(1)–N(3)
N(2)–Cu(1)–N(4)
N(2)–Cu(1)–N(5)
79.02(9)
78.70(9)
N(2)–Cu(1)–N(6)
N(3)–Cu(1)–N(4)
N(3)–Cu(1)–N(5)
N(3)–Cu(1)–N(6)
N(4)–Cu(1)–N(5)
N(4)–Cu(1)–N(6)
N(5)–Cu(1)–N(6)
96.73(8)
101.79(8)
98.09(8)
87.27(8)
75.28(8)
74.45(8)
149.72(8)
178.94(9)
105.61(8)
104.67(9)
157.65(9)
100.46(9)
89.51(8)
N(1)–Cu(1)–N(2)
N(1)–Cu(1)–N(3)
N(1)–Cu(1)–N(4)
N(1)–Cu(1)–N(5)
N(1)–Cu(1)–N(6)
N(2)–Cu(1)–N(3)
N(2)–Cu(1)–N(4)
N(2)–Cu(1)–N(5)
N(2)–Cu(1)–N(6)
N(3)–Cu(1)–N(4)
N(3)–Cu(1)–N(5)
N(3)–Cu(1)–N(6)
N(4)–Cu(1)–N(5)
N(4)–Cu(1)–N(6)
N(5)–Cu(1)–N(6)
77.68(6)
77.48(6)
176.09(7)
106.18(6)
100.81(6)
155.16(6)
103.26(6)
92.97(6)
93.84(6)
101.56(6)
93.32(6)
91.38(6)
77.62(7)
75.38(6)
152.99(6)
N(7)–Cu(2)–N(8)
N(7)–Cu(2)–N(9)
N(7)–Cu(2)–N(10)
N(7)–Cu(2)–N(11)
N(7)–Cu(2)–N(12)
N(8)–Cu(2)–N(9)
N(8)–Cu(2)–N(10)
N(8)–Cu(2)–N(11)
N(8)–Cu(2)–N(12)
N(9)–Cu(2)–N(10)
N(9)–Cu(2)–N(11)
N(9)–Cu(2)–N(12)
N(10)–Cu(2)–N(11)
N(10)–Cu(2)–N(12)
78.33(7)
77.29(7)
178.84(7)
102.67(6)
102.92(6)
155.62(6)
102.82(7)
90.49(6)
92.75(6)
101.56(6)
94.42(6)
93.06(6)
77.23(6)
77.22(6)
N(11)–Cu(2)–N(12) 154.33(6)
Fig. 1 Q Band powder EPR spectra at 10 K of: (a) complex 2; (b) 3.
better simulation of the spectrum profile at 5 K is obtained
assuming two species in approximately equal proportions. One
of these signals affords g values that resemble those of 1 in
frozen solution, while the other signal has an apparently axial
‘inverse’ g pattern (Table 1). Hence, we suggest that at T ≤ 20 K
we observe a combination of exchange-averaged crystal and
resolved molecular spectra from this sample.
The EPR spectra of complexes 2 and 3, while differing
between the two, are virtually identical in the powder and in
frozen MeCN, and show only small variations with temperature
in the powder (Table 1). This shows that these complexes have
static structures in the solid, and that their molecular and elec-
tronic structures in the solid state and in solution are the same.
For 2, the EPR spectra in both phases are rhombic and exhibit
Fig. 2 View of the [Cu(L2Cy)2]2ϩ dication in the crystal of 2ؒMeNO2,
showing the atom numbering scheme employed. For clarity, all H atoms
have been omitted. Thermal ellipsoids are drawn at the 35% probability
level.
a g1 > g2 ≈ g3 > ge pattern (Fig. 1a). This shows that, like 1, 2
1
2
2
possesses a tetragonally elongated {dy Ϫ z } copper() centre.
the Jahn–Teller distortion in solid 2 is static rather than
dynamic, in keeping with the EPR data. The Cu(1)–N(4) dis-
tance [2.004(2) Å] is longer than Cu(1)–N(1) [1.936(2) Å], which
presumably reflects increased conformational strain within the
former ligand backbone caused by the lengthened Cu–N(5)
and Cu–N(6) bonds. The ligand imine functions exhibit indis-
However, for 3 the EPR spectra are almost axial with
1
2
g1 ≈ g2 > g3 ≈ ge, which is consistent with a {dz } ground state
(Fig. 1b).19 The alternative description of solid 3, as a fluxional
1
2
2
{dy Ϫ z } species akin to 1, is ruled out because g3 ≈ ge, and
because an almost identical spectrum is obtained in frozen solu-
tion. Therefore, in contrast to 1 and 2, 3 must be formulated
either as a tetragonally compressed 6-co-ordinate complex, or
as a trigonal bipyramidal copper() centre with one non-co-
ordinated imine donor. The former scenario was confirmed by
the single crystal structure analysis described below.
tinguishable C᎐N bond lengths between 1.268(4) and 1.280(3)
᎐
Å, despite the wide variation in Cu–N{imine} distances of
2.085(2)–2.443(2) Å. This suggests that there is only weak
Cu→π*{N᎐C} back donation in this complex.
᎐
Comparison of the two structures is hampered by the fact
Structural analyses were accomplished on single crystals of
composition 2ؒMeNO2 and 3ؒ¹Me2CO. Both complexes
that complex 3ؒ¹Me2CO contains two crystallographically
¯
²
independent cations, whose Cu–N bond lengths show some dif-
ferences, particularly along the molecular x axis (Fig. 3, Table
3). This suggests that the co-ordination sphere in 3 is rather
plastic in solution. However, the near invariance of the solid
state EPR spectrum of 3 with temperature,2–5 and the almost
identical and near-spherical anisotropic thermal parameters
exhibited by all twelve N atoms in the structure (Fig. 3),1 both
¯
²
exhibit rhombically distorted octahedral stereochemistries
(Figs. 2 and 3, Tables 2 and 3). The [Cu(L2Cy)2]2ϩ dication in 2
shows a clear axis of elongation along the N(5)–Cu(1)–N(6)
vector, while the Cu–N bonds along the molecular y axis [N(2)–
Cu(1)–N(3)] are only slightly longer than the z axis [N(1)–
Cu(1)–N(4)] (Fig. 2, Table 2). This supports the conclusion that
3318
J. Chem. Soc., Dalton Trans., 2000, 3316–3324

View Online
Fig. 3 Views of the crystallographically independent [Cu(L2But)2]2ϩ dications in the asymmetric unit of 3ؒ Me2CO. Details as for Fig. 2.
1
–
2
imply that 3 is not fluxional in the solid state. Therefore,
comparison of the metric parameters at copper in 2 and 3 is
justified. It is clear that, in 3, the Cu–N bond lengths along the
molecular x and y axes have converged compared to 2; this has
mostly been achieved through an elongation of the y-axis
Cu–N bonds by 0.193(3)–0.222(3) Å. In contrast, the z-axis
Cu–N bonds in 3 are only 0.004(3)–0.070(3) Å shorter than in
2. Hence, the structural distortion in 3 might better be
described as an elongation of the co-ordination sphere within
the xy plane, rather than as a z-axis compression.
Copper(I) complexes of L2R
As described above, our attempted syntheses of [Cu(L2R)2]-
[BF4]2 (R = Ph or Mes) were unsuccessful, yielding brown solu-
tions from which no pure solid products could be obtained. We
suggested that this may reflect spontaneous reduction of the
hydrated Cu(BF4)2 starting material to CuI in the reaction mix-
Scheme 2 Possible molecular structures for [Cu(L2R)2]ϩ (R = Me, Mes
ture, in the presence of the L2R ligands, which should have
enhanced π-acceptor capability with aryl ‘R’ substituents. In
order to test this theory, the copper() complex chemistry of the
L2R ligands was examined.
or Cy), assuming 4-co-ordination at copper.
Assuming 4-co-ordination at copper, there are three possible
structural isomers for [Cu(L2R)2]ϩ (Scheme 2), which in
principle could be distinguished by NMR spectroscopy. At
Complexation of [Cu(NCMe)4]BF4 by L2R (R = Me, Mes or
Cy) in MeCN or MeNO2 in all cases yielded deep brown solu-
tions, from which brown microcrystalline solids could be
isolated upon addition of Et2O. On the basis of microanalytical
data, these complexes were formulated as [Cu(L2R)2]BF4
(R = Me, 4; Mes, 5; or Cy, 6). This was supported by IR
spectroscopy, which demonstrated the presence of L2R and
1
room temperature, 4–6 (and 7) each afford sharp H NMR
spectra in CDCl3, showing only one C2 or m-symmetric L2R
ligand environment, which is not consistent with any of the
three possible isomers. No decoalescence of any of these reson-
ances was observed between 298 and 220 K. Therefore, in order
to elucidate the mode of ligand binding in 4-6, a single crystal
structure determination of one of these complexes was under-
taken (see below).
Ϫ
BF4 only, and by FAB mass spectrometry which showed
highest molecular ions corresponding to [63Cu(L2R)2]ϩ (4, m/z =
385; 5, 801; 6, 657). An analogous reaction employing L2Ph
also yielded a brown solid, that gave a highest mass peak at
m/z 633, corresponding to [63Cu(L2Ph)2]ϩ. However, despite
the clean mass spectrum, the microanalysis was consistently low
in C, H and N, suggesting contamination by a copper() species.
In contrast to the above results, treatment of [Cu(NCMe)4]-
BF4 with L2But under the above conditions affords a red
microcrystalline powder (7) which analysed consistently for
‘[Cu(L2But)]BF4ؒH2O’. Importantly, the FAB mass spectrum of
7 displayed a highest mass peak at m/z = 616, corresponding to
[{63Cu(L2But)}2]ϩ. Therefore, 7 was formulated as a dimeric
complex containing the [{Cu(L2But)}2]2ϩ dication, which prob-
ably adopts a helical structure similar to those of [{Cu-
(L1R)}2]2ϩ,20 [{Cu(terpy)}2]2ϩ 21 or related compounds.22,23
Presumably, the steric bulk of the But substituents prevents the
formation of [Cu(L2But)2]ϩ. Preparations of 4–7 employing
varying Cu:L2R ratios of between 1 and 2:1 affected the yields,
but not the identities, of the products obtained.
Crystals of formula 5ؒMeCN were grown from MeCN–Et2O.
The structure shows that the complex adopts a flattened tetra-
hedral geometry, with each ligand binding in a bidentate fash-
ion through the pyridine N-donor and one imine group
(Scheme 2, A; Fig. 4). While the two Cu–N bonds to one chelate
ligand [Cu(1)–N(4) and Cu(1)–N(6)] are crystallographically
identical (Table 4), the other ligand is bonded slightly asym-
metrically, with Cu(1)–N(1) 2.0186(15) and Cu(1)–N(3)
2.0442(15) Å. The dihedral angle ‘θ’ between the planes of the
two chelating ligands [Cu(1), N(1), N(3)] and [Cu(1), N(4),
N(6)], which should be 90Њ for an ‘ideal’ tetrahedral geometry, is
71.2(2)Њ. This is within the range previously observed for other
CuI/bis-diimine complexes, which often show distortions
towards planarity with θ as low as 49Њ.24 Although the pyridine
ring and both imine groups of each ligand are essentially
coplanar, the non-bonded imine groups of each ligand are
rotated by ca. 180Њ about the C{ipso}–C{aldimine} single bond,
J. Chem. Soc., Dalton Trans., 2000, 3316–3324
3319

View Online
Table 4 Selected bond lengths (Å) and angles (Њ) for [Cu(L2Mes)2]-
BF4ؒMeCN 5ؒMeCN
Cyclic voltammograms of complexes 4–7 were measured in
MeCN–0.1 M NBun BF4 at 298 K. All these compounds
4
showed a chemically reversible CuII–CuI couple, at E =
₁
₂
Ϫ0.43 (4), ϩ0.19 (5), Ϫ0.31 (6) and Ϫ0.11 V (7) vs. ferrocene–
Cu(1)–N(1)
Cu(1)–N(3)
2.0186(15)
2.0442(15)
Cu(1)–N(4)
Cu(1)–N(6)
2.0292(15)
2.0344(15)
ferrocenium. The potentials for 4, 6 and 7 are broadly compar-
able to the reductions shown by related complexes such as
[Cu(L1R)2]2ϩ [e.g. Epa{CuII–CuI)} = –0.06 V for R = H and
–0.45 V for R = Mes]27 or [Cu(terpy)2]2ϩ 21a,28 derivatives under
similar conditions. However, the substantially more positive
potential shown by 5 is consistent with the improved π-acceptor
capability anticipated for L2Mes compared to the alkyl-
substituted L2R ligands. Given the rather positive potential of
the [Cu(L2Mes)2]2ϩ/ϩ couple, and since the brown colours
obtained upon mixing copper() salts with L2R (R = Ph or
Mes) are visually identical to those of 4–7, it is almost certain
that our inability to prepare [Cu(L2R)2]2ϩ with aryl ‘R’ groups
reflects reduction of the copper() content of the reaction mix-
tures during complexation. In addition to the CuII–CuI couple,
N(1)–Cu(1)–N(3)
N(1)–Cu(1)–N(4)
N(1)–Cu(1)–N(6)
81.85(6)
123.13(6)
139.88(6)
N(3)–Cu(1)–N(4)
N(3)–Cu(1)–N(6)
N(4)–Cu(1)–N(6)
128.45(6)
106.98(6)
82.52(6)
4–7 also exhibit an irreversible CuI–Cu0 reduction at Epc
=
Ϫ1.79 (4), Ϫ1.28 (5), Ϫ1.85 (6) and Ϫ1.90 V (7), with an
associated oxidative desorption spike near –0.6 V reflecting
deposition of copper metal at the electrode surface.
A copper(II) complex of L3
Fig. 4 View of the [Cu(L2Mes)2]ϩ cation in the crystal of 5ؒMeCN.
1
2
We were keen to investigate whether a {dz } 6-co-ordinate
Details as for Fig. 2.
copper() complex could be obtained without the use of bulky
ligand substituents. We noted that the crystal structure of
[Fe(L3)2][PF6]2 had been published previously, which showed
that L3 enforces extremely long Fe–N{pyridine} bonds in this
complex.29 This is a result of the inclusion of a 5-membered
ring at the centre of the tridentate L3 framework, rather than a
6-membered ring as in L1R. We therefore suggested that con-
formational strain may induce correspondingly long Cu–
so that the N atom lone pairs are oriented away from the copper
ion. This ligand conformation enforces contacts between Cu(1)
and the non-bonded aldimino C–H moieties of Cu(1) ؒ ؒ ؒ H(6)
2.87 and Cu(1) ؒ ؒ ؒ H(31) 2.90 Å.
There is an intramolecular stacking interaction between the
two mesityl substituents C(17)–C(25) and C(42)–C(50) which
are separated by 3.4 Å, the dihedral angle between the least
squares planes of the two groups being 2.3(2)Њ and the centroids
of the two phenyl rings being offset by 2.9 Å.25 The relative
orientations of the two groups places a methyl group above the
centre of each phenyl ring, so that the distances between these
C atoms C(25) and C(48) and the centroids of the neighbouring
rings are 3.4–3.5 Å. This is suggestive of C-H-π interactions
between these two groups, although confirmation of this by
location of the methyl group H atoms in the difference map was
not possible. There is an intermolecular π–π stacking inter-
action between the mesityl group C(32)–C(40) and C(32Ј)–
C(40Ј) related by 1 Ϫ x, 1 Ϫ y, 1 Ϫ z, which are strictly
coplanar by symmetry, lie 3.4 Å apart and whose centroids are
offset by 3.6 Å. The molecular structure of complex 5 is similar
to that of [Cu(L2{C6H4(OMe)-4})2]BF4, which was published
during the course of this work.13 This literature compound does
not possess the intramolecular π–π interaction exhibited by 5 in
the crystal, however, which leads it to adopt a more regular
tetrahedral geometry with essentially perpendicular chelating
ligands.
The deep colours of complexes 4–7 originate from one or two
Cu→L2R charge transfer absorptions.23,26 For 4 this appears as
a single peak λmax = 465 nm (εmax = 2,200 MϪ1 cmϪ1), while for 6
and 7 it is split into two bands, at λmax = 455 (sh) and 476 nm
(εmax ≈ 2,600 MϪ1 cmϪ1 per L2R ligand). For 5, which bears aryl
imine substituents, these peaks are weaker and at higher wave-
lengths, at λmax = 495 (εmax = 881) and 581 nm (367 MϪ1 cmϪ1).
The sensitivity of these MLCT absorptions to the identity of
the L2R substituent suggest that they represent transitions pre-
dominantly into the π* orbitals of the ligand aldimine moieties.
The spectra of 4–7 also contain a second MLCT transition,23 at
λmax = 354–369 nm (εmax = 1,600–10,500 MϪ1 cmϪ1), and lower
wavelength π → π* transitions. The latter are substantially
more complex than the π → π* region for 1–3, consistent
with the reduced symmetry of the asymmetrically co-ordinated
L2R ligands in 4–6.
N{pyridine bonds} in [Cu(L3)2]2ϩ, and that this would result in
1
2
a {dz } ground state for this complex.
Complexation of Cu(ClO4)2ؒ6H2O by 2 mol of L3 in MeNO2
yields a bright green solution, from which green microcrystals
analysing as [Cu(L3)2][ClO4]2 8 were obtained following concen-
tration of the solution and addition of diethyl ether. Import-
antly, the IR spectrum of this compound was indistinguishable
from those of its iron() and nickel() congeners,29 which
is consistent with the [Cu(L3)2]2ϩ dication adopting a 6-co-
ordinate structure in the solid, rather than an equally plausible
4-co-ordinate geometry with each ligand binding bidentately.30
This proposal was confirmed by a crystal structure analysis.
Crystals of complex 8 give highly mosaic diffraction patterns,
whose mosaicity increases upon cooling (see Experimental
section). However, an analysis at room temperature was
achieved of a poorly diffracting crystal of 8, which is iso-
morphous with the high temperature form of [Cu(L1H)2]-
[BF4]2.5,7 Crystals of [Cu(L3)2](BF4)2 are also apparently
isomorphous with these two salts but suffer from twinning,
presenting an apparently tetragonal unit cell with a = 8.7295(6)
and c = 37.461(2) Å at 150 K. The structure of 8 shows the
complex to adopt the expected distorted octahedral stereo-
chemistry, with two very short Cu–N{pyrazole} bonds of
Cu(1)–N(2) 1.824(8) and Cu(1)–N(19) 1.838(9) Å (Fig. 5, Table
5). While the Cu(1)–N{pyridine} bonds to one L3 ligand [N(2)–
C(18)] are crystallographically indistinguishable (Table 5), the
other ligand is bound rather asymmetrically with Cu(1)–N(25)
2.353(9) and Cu(1)–N(31) 2.233(9) Å. This asymmetric co-
ordination is reflected in the chelate ring bite angles for this
ligand, which are N(19)–Cu(1)–N(25) 69.8(3) and N(19)–
Cu(1)–N(31) 76.5(3)Њ. All other metric parameters within the
structure are unexceptional.
The powder EPR spectrum of complex 8 is very similar to
that of 3, with an almost axial g⊥ > g|| ≈ ge pattern whose
parameters vary only slightly between 290 and 10 K (Table 1).
As for 3, this demonstrates that 8 has the desired {dz } ground
1
2
3320
J. Chem. Soc., Dalton Trans., 2000, 3316–3324

View Online
Table 5 Selected bond lengths (Å) and angles (Њ) for [Cu(L3)2][ClO4]2 8
This would open new opportunities for the design of molecular
magnetic switches or sensors. We are actively investigating these
ideas.
Cu(1)–N(2)
Cu(1)–N(8)
Cu(1)–N(14)
1.824(8)
2.297(8)
2.326(8)
Cu(1)–N(19)
Cu(1)–N(25)
Cu(1)–N(31)
1.838(9)
2.353(9)
2.233(9)
Experimental
N(2)–Cu(1)–N(8)
N(2)–Cu(1)–N(14)
N(2)–Cu(1)–N(19)
N(2)–Cu(1)–N(25)
N(2)–Cu(1)–N(31)
N(8)–Cu(1)–N(14)
N(8)–Cu(1)–N(19)
N(8)–Cu(1)–N(25)
75.0(4)
71.4(4)
N(8)–Cu(1)–N(31)
N(14)–Cu(1)–N(19)
N(14)–Cu(1)–N(25)
N(14)–Cu(1)–N(31)
N(19)–Cu(1)–N(25)
N(19)–Cu(1)–N(31)
N(25)–Cu(1)–N(31)
93.9(3)
107.3(4)
95.2(3)
95.2(3)
69.8(3)
76.5(3)
146.3(3)
Unless stated otherwise, all manipulations were performed in
air using commercial grade solvents. Pyridine-2,6-dicarb-
178.5(5)
110.9(3)
102.7(3)
146.4(3)
106.3(4)
94.9(3)
aldehyde,31 L2Ph,16 L2But,17 L3 29 and [Cu(NCMe)4]BF4
32
were prepared by the literature procedures. Cu(BF4)2ؒ6H2O
(Avocado), KH (Aldrich) and all solvents were used as
supplied.
Syntheses
2,6-Bis(methyliminomethyl)pyridine (L2Me). Pyridine-2,6-
dicarbaldehyde (0.20 g, 1.5 × 10Ϫ3 mol) and methylamine (1.5
cm3 of a 2 M solution in MeOH, 3.0 × 10Ϫ3 mol) were stirred
in MeOH (25 cm3) at room temperature for 1 h. The near-
colourless solution was evaporated to dryness, leaving a yellow
oil which formed a cream solid upon storage at Ϫ30 ЊC. Yield
0.21 g, 89% (Found: C, 66.9; H, 6.8; N, 26.3. Calc. for C9H11N3:
C, 67.1; H, 6.9; N, 26.1%). mp ca. 0 ЊC. EI mass spectrum: m/z
160, [M Ϫ H]ϩ; 146, [M Ϫ CH3]ϩ; and 119, [M Ϫ CHNCH3]ϩ.
NMR spectra (CDCl , 293 K): 1H; δ 8.37 (s, 2 H, HC᎐N), 7.91
᎐
3
(d, J 7.5, 2 H, Py H3/5), 7.73 (t, J 7.5 Hz, 1H, Py H4) and 3.53 (s,
6 H, CH3); 13C, δ 163.1 (HC᎐N), 154.3 (Py C2/6), 137.1 (Py C4),
᎐
122.1 (Py C3/5) and 48.1 (CH3).
2,6-Bis(2,4,6-trimethylphenyliminomethyl)pyridine (L2Mes).
Method as for L2Me, using 2,4,6-trimethylphenylamine (0.40 g,
3.0 × 10Ϫ3 mol) but with a 1 h reflux. A yellow solid precipitated
upon cooling the reaction mixture, which was filtered off,
washed with MeOH and dried. Yield 0.48 g, 87% (Found: C,
81.0; H, 7.3; N, 11.5. Calc. for C25H27N3: C, 81.3; H, 7.4; N,
11.4%). mp 156–158 ЊC. EI mass spectrum: m/z 369, [M]ϩ; 354,
[M Ϫ CH3]ϩ; 250, [M Ϫ C9H11]ϩ; and 236, [M Ϫ NC9H11]ϩ.
NMR spectra (CDCl , 293 K): 1H, δ 8.38 (s, 2 H, HC᎐N), 8.37
Fig. 5 View of the [Cu(L3)2]2ϩ dication in the crystal of complex 8.
Details as for Fig. 2. The pyrrolic N atoms of the L3 pyrazole rings are
disordered over the C(3), C(6), C(20) and C(23) sites (see Experimental
section).
state in the solid.19 Interestingly, the parallel region of the spec-
trum becomes less well resolved as the temperature decreases,
which is the opposite of the usual trend. The reason for this is
uncertain, but may be related to the reduced crystallinity of 8
at lower temperatures (see above). In frozen MeCN or MeNO2
solution, 8 affords a very different, poorly resolved EPR spec-
trum, which could not be simulated meaningfully but whose
shape is suggestive of a tetragonal copper() species with
g|| > g⊥ > ge.19 Since 8 shows identical d–d maxima in MeCN
and MeNO2, the copper() ion does not undergo significant
solvolysis in these solvents. Therefore, we tentatively propose
that the solution EPR spectrum of 8 reflects partial isomeris-
ation of the dissolved complex, to a [4 ϩ 2] co-ordination
geometry analogous to that shown by [Cu(L5)2][BF4]2 (L5 =
3-{2,5-dimethoxyphenyl}-1-{pyridin-2-yl}pyrazole).30 We note
᎐
3
(d, J 7.5, 2 H, Py H3/5), 7.94 (t, J 7.5 Hz, 1H, Py H4), 6.89 (s, 4 H,
Ph H3/5), 2.28 (s, 6 H, 4-CH3) and 2.14 (s, 12 H, 2,6-CH3); 13C,
δ 163.1 (HC᎐N), 151.6 (Py C2/6), 147.8 (Ph C1), 137.7 (Py C4),
᎐
133.5 (Ph C2/6), 128.8 (Ph C3/5), 126.8 (Ph C4), 122.6 (Py C3/5),
20.8 (4-CH3) and 18.2 (2,6-CH3).
2,6-Bis(cyclohexyliminomethyl)pyridine (L2Cy). Method as
for L2Me, using cyclohexylamine (0.30 g, 3.0 × 10Ϫ3 mol). Fol-
lowing a 1 h reflux, the colourless solution was evaporated to
dryness, leaving a tan oil which solidified upon standing at
Ϫ30 ЊC. Yield 0.38 g, 87% (Found: C, 76.8; H, 9.4; N, 14.4.
Calc. for C19H27N3: C, 76.7; H, 9.2; N, 14.1%). mp ca. 25 ЊC. EI
1
29
that the H NMR spectrum of [Ni(L3)2][BF4]2 is much more
complex than would be expected from its solid state structure,
and demonstrates the presence of several paramagnetic species
in solution.27 It is likely that the solution structure of 8 is
similarly complicated.
mass spectrum: m/z 297, [M]ϩ; and 214, [M Ϫ C6H11]ϩ. NMR
1
spectra (CDCl , 293 K): H; δ 8.46 (s, 2 H, HC᎐N), 8.05 (d,
᎐
3
J 7.5, 2 H, Py H3/5), 7.92 (t, 7.5 Hz, 1H, Py H4), 3.35 (m, 2H, Cy
CH) and 1.59 (m, 20H, Cy CH2); 13C, δ 159.4 (HC᎐N), 154.7
᎐
(Py C2/6), 136.9 (Py C4), 122.0 (Py C3/5), 69.6 (Cy C1), 34.2 (Cy
C2/6), 25.6 (Cy C3/5) and 24.7 (Cy C4).
Concluding remarks
This study has established that our observation of sterically
induced ground-state variability in [Cu(L1R)2]2ϩ is not confined
to this class of complexes, and appears to be general to other
[CuL2]2ϩ systems where L is a meridional tris-imine ligand. This
demonstrates an apparently general, but hitherto unreported,
method by which the magnetic properties of a co-ordinated
metal ion may be modulated by careful design of its steric
environment. While there is little obvious applicability for these
copper() systems, appropriate distal substitution of analogous
[FeL2]2ϩ or [CoL2]2ϩ (L = meridional tris-imine ligand) may
afford a new degree of control over the spin-crossover
behaviour that is often exhibited by this class of compound.
Bis[2,6-bis(methyliminomethyl)pyridine]copper(II) bis(tetra-
fluoroborate) 1. A mixture of L2Me (0.12 g, 7.70 × 10Ϫ4 mol)
and Cu(BF4)2ؒ6H2O (0.13 g, 3.85 × 10Ϫ4 mol) was stirred in
MeCN (20 cm3) at room temperature for 30 min. The resultant
dark brown-green solution was filtered and reduced to ca. 5 cm3
volume. Vapour diffusion of Et2O into this solution yielded oily
green microcrystals. The supernatant was removed by decanta-
tion, and the solid washed twice with Et2O and dried in vacuo to
afford a green powder. Yield 0.090 g, 42% (Found: C, 38.5; H,
4.0; N, 14.9. Calc. for C18H22B2CuF8N6: C, 38.6; H, 4.0; N,
15.0%). FAB mass spectrum: m/z 385, [63Cu(L2Me)2]ϩ; and 224,
[63Cu(L2Me)]ϩ. UV/vis spectrum: λmax/nm (εmax/MϪ1 cmϪ1)
J. Chem. Soc., Dalton Trans., 2000, 3316–3324
3321

View Online
(MeCN) 226 (33,700), 309 (8,100), 481 (sh) and 709 (48);
(MeNO2) 481 (sh) and 709 (50).
(0.19 g, 7.75 × 10Ϫ4 mol). The product formed deep red needles
from MeNO2–Et2O. Yield 0.19 g, 63% (Found: C, 43.5; H, 5.7;
N, 10.2. Calc. for C15H23BCuF4N3ؒH2O: C, 43.7; H, 5.9; N,
10.2%). FAB mass spectrum: m/z 616, [63Cu2(L2But)2]ϩ; 395,
[63Cu(L2But)11BF4]ϩ; and 308, [63Cu(L2But)]ϩ. UV/vis spectrum
(MeCN): λmax/nm (εmax/MϪ1 cmϪ1) 217 (41,000), 240 (24,700),
273 (sh), 283 (15,443), 369 (2,400), 455 (sh) and 476 (2,700). 1H
Bis[2,6-bis(cyclohexyliminomethyl)pyridine]copper(II)
bis(tetrafluoroborate) 2. A solution of Cu(BF4)2ؒ6H2O (0.13 g,
3.85 × 10Ϫ4 mol) and L2Cy (0.23 g, 7.70 × 10Ϫ4 mol) in MeCN
was stirred at room temperature for 30 min. The deep green
solution was filtered and concentrated to ca. 5 cm3 volume.
Vapour diffusion of Et2O into this solution formed green
microcrystals, which were filtered off and dried. The product
was recrystallised from MeNO2–Et2O. Yield 0.18 g, 55%
(Found: C, 53.8; H, 6.8; N, 9.8. Calc. for C38H54B2CuF8N6ؒH2O:
C, 53.7; H, 6.6; N, 9.9%). FAB mass spectrum: m/z 657,
[63Cu(L2Cy)2]ϩ; 379, [63Cu(L2Cy)F]ϩ; and 360 [63Cu(L2Cy)]ϩ.
UV/vis spectrum: λmax/nm (εmax/MϪ1 cmϪ1) (MeCN) 226
(45,200), 308 (12,000), 478 (sh) and 709 (61); (MeNO2): 478 (sh)
and 706 (70).
NMR spectrum (CDCl , 293 K): δ 8.71 (s, 2H, HC᎐N), 8.37 (m,
᎐
3
3H, Py H3/4/5) and 1.21 (s, 18H, CH3).
Bis(1,3-bis{pyridin-2-yl}pyrazole)copper(II) diperchlorate 8. A
solution of L3 (0.50 g, 2.25 × 10Ϫ3 mol) and Cu(ClO4)2ؒ6H2O
(0.42 g, 1.13 × 10Ϫ3 mol) in MeNO2 (20 cm3) was stirred at
room temperature for 30 min. The intensely green solution was
filtered and reduced to ca. 5 cm3 volume. Vapour diffusion of
Et2O into this solution yielded deep green microcrystals. Yield
0.54 g, 68% (Found: C, 44.1; H, 2.9; N, 16.2. Calc. for
C26H20Cl2CuN8O8: C, 44.2; H, 2.9; N, 15.9%). FAB mass spec-
trum: m/z 507, [63Cu(L3)2]ϩ; 285, [63Cu(L3)]ϩ; and 223, [L3H]ϩ.
UV/vis spectrum: λmax/nm (εmax/MϪ1 cmϪ1) (MeCN) 220 (sh),
227 (sh), 249 (23,000), 255 (sh), 270 (20,700), 277 (20,000), 307
(35,700), 320 (sh), 368 (sh), 415 (sh) and 761 (119); (MeNO2):
415 (sh) and 761 (129). CAUTION: although we have experi-
enced no difficulties in handling complex 8, metal–organic per-
chlorates are potentially explosive and should be treated with
due care.
Bis[2,6-bis(tert-butyliminomethyl)pyridine]copper(II)
bis(tetrafluoroborate) 3. Method as for complex 2, using L2But
(0.19 g, 7.70 × 10Ϫ4 mol). The product was recrystallised from
MeNO2–Et2O. Yield 0.15 g, 52% (Found: C, 48.9; H, 6.4; N,
11.3. Calc. for C30H46B2CuF8N6: C, 49.5; H, 6.4; N, 11.6%).
FAB mass spectrum: m/z 553, [63Cu(L2But)2]ϩ; and 308,
[63Cu(L2But)]ϩ. UV/vis spectrum: λmax/nm (εmax/MϪ1 cmϪ1)
(MeCN) 243 (56,700), 274 (51,100) and 746 (63); (MeNO2) 746
(62).
Bis(1,3-bis{pyridin-2-yl}pyrazole)copper(II) bis(tetrafluorobo-
rate). Method as for complex 8, using Cu(BF4)2ؒ6H2O (0.39 g,
1.13 × 10Ϫ3 mol). Yield 0.55 g, 71% (Found: C, 46.0; H, 3.0; N,
16.4. Calc. for C26H20B2CuF8N8: C, 45.8; H, 3.0; N, 16.4%).
Bis[2,6-bis(methyliminomethyl)pyridine]copper(I) tetrafluor-
oborate 4. A solution of L2Me (0.25 g, 1.55 × 10Ϫ3 mol) and
[Cu(NCMe)4]BF4 (0.24 g, 7.75 × 10Ϫ4 mol) was stirred in
MeCN (20 cm3) at room temperature for 15 min. The dark
brown solution was filtered and reduced to ca. 5 cm3 volume.
Vapour diffusion of Et2O into this solution yielded dark brown
microcrystals. Yield 0.26 g, 71% (Found: C, 45.1; H, 4.7; N,
17.2. Calc. for C18H22BCuF4N6: C, 45.7; H, 4.7; N, 17.8%). FAB
mass spectrum: m/z, 385 [63Cu(L2Me)2]ϩ; and 224, [63Cu-
(L2Me)]ϩ. UV/vis spectrum (MeCN): λmax/nm (εmax/MϪ1 cmϪ1)
214 (47,100), 219 (41,600), 239 (13,700), 274 (sh), 285 (4,700)
Single crystal structure determinations
Single crystals of X-ray quality of [Cu(L2Cy)2][BF4]2ؒMeNO2
2ؒMeNO2, [Cu(L2But)2][BF4]2ؒ¹Me2CO 3ؒ¹Me2CO and [Cu-
¯
²
¯
²
(L2Mes)2]BF4ؒMeCN 5ؒMeCN were grown by diffusion of
ether vapour into solutions of the complexes in the appropriate
solvents. Crystals of [Cu(L3)2][ClO4]2 8 were similarly grown
from MeNO2–Et2O. Experimental details from the structure
determinations are given in Table 6. All structures were solved
by direct methods (SHELXS 8633) and refined by full matrix
least squares on F2 (SHELXL 9734), with H atoms placed in
calculated positions.
1
294 (4,100), 362 (1,600) and 465 (2,200). H NMR spectrum
(CDCl , 293 K): δ 8.54 (s, 2 H, HC᎐N), 8.33 (t, J 6.6 Hz, 1 H, Py
᎐
3
H4), 8.21 (d, J 6.6 Hz, 2 H, Py H3/5) and 3.43 (s, 6 H, CH3).
Bis[2,6-bis(2,4,6-trimethylphenyliminomethyl)pyridine]-
copper(I) tetrafluoroborate 5. Method as for complex 4, using
L2Mes (0.57 g, 1.55 × 10Ϫ3 mol). The product formed brown
needles from MeCN–Et2O. Yield 0.54 g, 78% (Found: C, 66.4;
H, 6.2; N, 10.4. Calc. for C50H54BCuF4N6ؒCH3CN: C, 67.1; H,
6.2; N, 10.5%). FAB mass spectrum: m/z 801, [63Cu(L2Mes)2]ϩ;
and 432, [63Cu(L2Mes)]ϩ. UV/vis spectrum (MeCN): λmax/nm
(εmax/MϪ1 cmϪ1) 225 (65,400), 234 (68,100), 283 (24,900), 354
(10,500), 495 (881) and 581 (367). 1H NMR spectrum (CDCl3,
[Cu(L2Cy)2][BF4]2ؒMeNO2 2ؒMeNO2. The F atoms of one
Ϫ
BF4 anion were disordered, and were modelled using two
orientations in a 60:40 occupancy ratio with the B–F distances
restrained to 1.36(2) Å. A badly disordered region of solvent
lying across a crystallographic 2-fold axis, which could not be
identified unambiguously, was modelled using 4 half-occupied
carbon environments. All non-H atoms with occupancy >0.5
were refined anisotropically.
293 K): δ 8.53 (s, 2 H, HC᎐N), 8.48 (d, J 6.4, 2 H, Py H3/5), 8.42
᎐
(t, J 6.4 Hz, 1 H, Py H4), 6.79 (s, 4 H, Ph H3/5), 2.21 (s, 6 H,
2
t
1
1
–
–
[Cu(L Bu )2][BF4]2ؒ2Me2CO 3ؒ2Me2CO. The asymmetric
4-CH3) and 1.74 (s, 12 H, 2,6-CH3).
Ϫ
unit contains 2 complex cations, 4 BF4 anions and one mole-
cule of acetone. All non-H atoms were refined anisotropically,
and no disorder was detected.
Bis[2,6-bis(cyclohexyliminomethyl)pyridine]copper(I)
tetrafluoroborate 6. Method as for complex 4, using L2Cy (0.46
g, 1.55 × 10Ϫ3 mol). The product formed a dark brown solid
from MeCN–EtO2. Yield 0.45 g, 78% (Found: C, 61.0; H, 7.0;
N, 11.2. Calc. for C38H54BCuF4N6: C, 61.2; H, 6.8; N, 11.3%).
FAB mass spectrum: m/z 657, [63Cu(L2Cy)2]ϩ; and 360,
[63Cu(L2Cy)]ϩ. UV/vis spectrum (MeCN): λmax/nm (εmax/MϪ1
cmϪ1) 220 (43,200), 246 (20,800), 279 (sh), 285 (17,900), 368
[Cu(L2Mes)2]BF4ؒMeCN 5ؒMeCN. During refinement, the
Ϫ
BF4 anion was found to be disordered over three sites, which
were modelled with a 50:30:20 occupancy ratio with all B–F
distances restrained to 1.38(2) Å, and F ؒ ؒ ؒ F distances to
2.25(2) Å. All non-H atoms with occupancy ≥0.5 were refined
anisotropically.
1
(1,900), 455 (sh) and 476 (4,600). H NMR spectrum (CDCl3,
293 K): δ 8.39 (s, 2 H, HC᎐N), 8.14 (m, 3 H, Py H3/4/5), 3.27 (m,
᎐
[Cu(L3)2][ClO4]2 8. All crystals of this compound that were
examined had a high mosaicity, which increased upon cooling.
The crystal used for data collection had a mosaicity of 0.914(3)
at 293 K, compared to a more typical value of 0.4–0.5. Both
perchlorate anions were disordered over two distinct orienta-
2 H, Cy CH), 1.56 (m, 10 H, Cy CH2) and 1.13 (m, 10 H, Cy
CH2).
Bis[2,6-bis(tert-butyliminomethyl)pyridine]dicopper(I)
bis(tetrafluoroborate) 7. Method as for complex 4, using L2But
3322
J. Chem. Soc., Dalton Trans., 2000, 3316–3324

View Online
Table 6 Experimental details for the single crystal structure determinations
2
t
[Cu(L2Cy)2][BF4]2ؒMeNO2
2ؒMeNO2
[Cu(L Bu )2][BF4]2ؒ2Me2CO
[Cu(L2Mes)2]BF4ؒMeCN
5ؒMeCN
[Cu(L3)2][ClO4]2 8
1
–
1
–
3ؒ2Me2CO
Formula
Mr
Crystal class
Space group
C39H57B2CuF8N7O2
917.10
Monoclinic
P21/c
C31.5H49B2CuF8N6O0.5
756.93
C52H57BCuF4N7
930.40
Monoclinic
P21/c
C26H20Cl2CuN8O8
706.94
Monoclinic
P21
Triclinic
¯
P1
a/Å
b/Å
c/Å
α/Њ
13.6323(1)
17.8182(1)
19.2083(1)
—
104.5180(6)
—
4516.77(5)
4
0.559
11.1235(1)
17.0639(2)
19.9979(3)
81.6900(8)
75.7200(8)
89.0990(8)
3639.12(8)
4
11.6577(2)
26.3758(4)
16.2105(2)
—
105.1990(10)
—
4809.96(13)
4
0.513
8.753(2)
8.789(2)
19.381(5)
—
92.334(14)
—
1489.7(6)
2
0.975
β/Њ
γ/Њ
U/Å3
Z
µ(Mo-Kα)/mmϪ1
0.674
T/K
150(2)
70847
8844
150(2)
114065
16691
150(2)
114573
10969
293(2)
7778
4711
Measured reflections
Independent reflections
Rint
0.050
0.056
0.159
0.064
0.043
0.127
0.098
0.042
0.111
0.061
0.089
0.260
R(F)
wR(F2)
S
1.043
—
1.027
—
1.024
—
1.066
0.04(4)
Flack parameter
tions, which were modelled with occupancy ratios of 60:40
(anion 1) and 80:20 (anion 2). All Cl–O distances were
restrained to 1.39(2) Å and non-bonded O ؒ ؒ ؒ O distances to
2.27(2) Å, while individual Uij parameters for each O atom were
restrained to be the same with an esd of 0.04 Å2; rigid bond
restraints were also applied to these thermal parameters. The
resolution of the structure was too low to distinguish between
the pyrazole N1 and C3 sites on the L3 ligands, which could in
any case be disordered.29 Hence these sites in the structure,
namely C(3), C(6), C(20) and C(23), were all refined as C atoms.
All non-H atoms except the minor disorder orientation of
anion 2 were refined anisotropically.
ledged from the Royal Society (research fellowship to MAH),
the EPSRC and the University of Leeds.
References
1 L. R. Falvello, J. Chem. Soc., Dalton Trans., 1997, 4463.
2 J.-V. Folgado, W. Henke, R. Allmann, H. Stratemeier, D. Beltrán-
Porter, T. Rojo and D. Reinen, Inorg. Chem., 1990, 29, 2035.
3 M. Duggan, B. J. Hathaway and J. Mullane, J. Chem. Soc., Dalton
Trans., 1980, 690.
4 J. H. Ammeter, H. B. Bürgi, E. Gamp, V. Meyer-Sandrin and
W. P. Jensen, Inorg. Chem., 1979, 18, 733.
5 N. K. Solanki, E. J. L. McInnes, F. E. Mabbs, S. Radojevic,
M. McPartlin, N. Feeder, J. E. Davies and M. A. Halcrow, Angew.
Chem., Int. Ed., 1998, 37, 2221.
6 M. A. Leech, N. K. Solanki, M. A. Halcrow, J. A. K. Howard and
S. Dahaoui, Chem. Commun., 1999, 2245.
CCDC reference number 186/2129.
See http://www.rsc.org/suppdata/dt/b0/b005105p/ for crystal-
lographic files in .cif format.
7 A. J. Bridgeman, M. A. Halcrow, M. Jones, E. Krausz and
N. K. Solanki, Chem. Phys. Lett., 1999, 314, 176.
8 C. J. Simmons, New J. Chem., 1993, 17, 77.
Other measurements
9 G. Wingefeld and R. Hoppe, Z. Anorg. Allg. Chem., 1984, 516, 223;
K. Finnie, L. Dubicki, E. R. Krausz and M. J. Riley, Inorg. Chem.,
1990, 29, 3908; M. Atanasov, M. A. Hitchman, R. Hoppe,
K. S. Murray, B. Moubaraki, D. Reinen and H. Stratemeier, Inorg.
Chem., 1993, 32, 3397; V. M. Masters, M. J. Riley and M. A.
Hitchman, J. Synchrotron Rad., 1999, 6, 242.
10 F. Lions and K. V. Martin, J. Am. Chem. Soc., 1957, 79, 2733.
11 G. J. P. Britovsek, V. C. Gibson, B. S. Kimberley, P. J. Maddox,
S. J. McTavish, G. A. Solan, A. J. P. White and D. J. Williams, Chem.
Commun., 1998, 849.
12 A. L. Vance, N. W. Alcock, J. A. Heppert and D. H. Busch, Inorg.
Chem., 1998, 37, 6912.
13 L. Douce, A. El-ghayoury, A. Skoulios and R. Zeissel, Chem.
Commun., 1999, 2033.
14 P. H. Merrell, E. C. Alyea and L. Ecott, Inorg. Chim. Acta, 1982, 59,
25; M. S. Nelson, A. J. Lavery and M. G. B. Drew, J. Chem. Soc.,
Dalton Trans., 1986, 911.
15 A. J. Blake, A. J. Lavery, T. I. Hyde and M. Schröder, J. Chem. Soc.,
Dalton Trans., 1989, 965.
16 T. W. Bell, F. Guzzo and M. G. B. Drew, J. Am. Chem. Soc., 1991,
113, 3115.
17 J. Mlochowski, E. Kubicz, K. Kloc, M. Mordarski, W. Peczynska
and L. Syper, Liebigs Ann. Chem., 1988, 455.
Infrared spectra were obtained as Nujol mulls pressed between
KBr windows between 400 and 4000 cmϪ1 using a Nicolet
Avatar 360 spectrophotometer, UV/visible spectra with a
Perkin-Elmer Lambda 900 spectrophotometer in 1 cm quartz
cells, NMR spectra on a Bruker ARX250 spectrometer
operating at 250.1 (1H) and 62.9 MHz (13C), electron impact
and positive ion fast atom bombardment mass spectra on
a VG AutoSpec spectrometer, the FAB spectra employing
a 3-nitrobenzyl alcohol matrix. CHN microanalyses were
performed by the University of Leeds School of Chemistry
microanalytical service. Melting points are uncorrected. EPR
spectra were obtained using a Bruker ESP300E spectrometer,
fitted with an ER5106QT Q-band resonator and an ER4118VT
cryostat. Spectral simulations were performed using in-house
software which has been described elsewhere.35
All electrochemical measurements were carried out using an
Autolab PGSTAT30 voltammetric analyser, in MeCN contain-
ing 0.1 M NBun BF4 (prepared from NBun OH and HBF4)
4
4
as supporting electrolyte. Cyclic voltammetric experiments
involved the use of platinum working and counter electrodes
and an Ag–AgCl reference electrode; all potentials quoted are
referenced to an internal ferrocene–ferrocenium standard and
were obtained at a scan rate of 100 mV sϪ1.
18 A. B. P. Lever, Inorganic Electronic Spectroscopy, 2nd edn., Elsevier,
Amsterdam, 1984, pp. 554–572.
19 B. A. Goodman and J. B. Raynor, Adv. Inorg. Chem., 1970, 13, 135.
20 N. K. Solanki, A. E. H. Wheatley, S. Radojevic, M. McPartlin and
M. A. Halcrow, J. Chem. Soc., Dalton Trans., 1999, 521.
21 (a) K. T. Potts, M. Kesharvaz-K, F. S. Tham, H. D. Abruña and
C. Arana, Inorg. Chem., 1993, 32, 4450; (b) E. C. Constable,
A. J. Edwards, M. J. Hannon and P. R. Raithby, J. Chem. Soc.,
Chem. Commun., 1994, 1991; (c) E. C. Constable, T. Kulke,
M. Neuberger and M. Zehnder, Chem. Commun., 1997, 489.
Acknowledgements
The authors thank Dr Michael Hitchman (University of
Tasmania) for useful discussions. Financial support is acknow-
J. Chem. Soc., Dalton Trans., 2000, 3316–3324
3323

View Online
M. Kesharvaz-K, F. S. Tham, H. D. Abruña and C. R. Arana, Inorg.
Chem., 1993, 32, 4436.
22 C. Piguet, G. Bernardinelli and A. F. Williams, Inorg. Chem., 1989,
28, 2920; S. Rüttiman, C. Piguet, G. Bernardinelli, B. Bocquet and
A. F. Williams, J. Am. Chem. Soc., 1992, 114, 4230; R. F. Carina,
G. Bernardinelli and A. F. Williams, Angew. Chem., Int. Ed. Engl.,
1993, 32, 1463; R. E. Carina, A. F. Williams and C. Piguet, Helv.
Chim. Acta, 1998, 81, 548.
27 N. K. Solanki and M. A. Halcrow, unpublished data.
28 G. D. Storrier, S. B. Colbran and D. C. Craig, J. Chem. Soc., Dalton
Trans., 1998, 1351; M. A. Halcrow, E. K. Brechin, E. J. L. McInnes,
F. E. Mabbs and J. E. Davies, J. Chem. Soc., Dalton Trans., 1998,
2477.
29 A. T. Baker, D. C. Craig, G. Dong and A. D. Rae, Aust. J. Chem.,
1995, 48, 1071.
30 M. A. Halcrow, E. J. L. McInnes, F. E. Mabbs, I. J. Scowen,
M. McPartlin, H. R. Powell and J. E. Davies, J. Chem. Soc., Dalton
Trans., 1997, 4025.
31 N. W. Alcock, R. G. Kingston, P. Moore and C. Pierpoint, J. Chem.
Soc., Dalton Trans., 1984, 1937.
23 M. G. B. Drew, A. Lavery, V. McKee and S. M. Nelson, J. Chem.
Soc., Dalton Trans., 1985, 1771.
24 For recent examples see e.g. S. M. Scott, K. C. Gordon and
A. K. Burrell, Inorg. Chem. 1996, 35, 2452; M. K. Eggleston,
P. E. Fanwick, A. J. Pallenberg and D. R. McMillin, Inorg. Chem.,
1997, 36, 4007; M. A. Halcrow, N. L. Cromhout and P. R. Raithby,
Polyhedron, 1997, 16, 4257; M. T. Miller, P. K. Gantzel and T. B.
Karpishin, Inorg. Chem., 1998, 37, 2285; M. T. Miller, P. K. Gantzel
and T. B. Karpishin, Inorg. Chem., 1999, 38, 3414; C. L. Foster,
C. A. Kilner, M. Thornton-Pett and M. A. Halcrow, Acta
Crystallogr., Sect. C, 2000, 56, 319.
25 C. A. Hunter and J. K. M. Sanders, J. Am. Chem. Soc., 1990, 112,
5525.
26 K. T. Potts, M. Kesharvaz-K, F. S. Tham, H. D. Abruña and
C. R. Arana, Inorg. Chem., 1993, 32, 4422; K. T. Potts,
32 G. J. Kubas, Inorg. Synth., 1990, 28, 68.
33 G. M. Sheldrick, Acta Crystallogr., Sect. A, 1990, 46, 467.
34 G. M. Sheldrick, SHELXL 97, Program for the refinement of crystal
structures, University of Göttingen, 1997.
35 F. E. Mabbs and D. Collison, Electron Paramagnetic Resonance
of d Transition Metal Compounds, Elsevier, Amsterdam, 1992,
ch. 7.
3324
J. Chem. Soc., Dalton Trans., 2000, 3316–3324