Structural, magnetic, and electronic properties of phenolic oxime complexes of Cu and Ni

Article abstract of DOI:10.1021/ic2020644

Square planar complexes of the type Ni(L¹)₂, Ni(L²)₂, Cu(L¹)₂, and Cu(L ²)₂, where L¹H = 2-hydroxy-5-t- octylacetophenone oxime and L²H = 2-hydroxy-5-n-propylacetophenone oxime, have been prepared and characterized by single-crystal X-ray diffraction, cyclic voltammetry, UV/vis spectroscopy, field-effect-transistor measurements, density functional theory (DFT) and time-dependent DFT (TDDFT) calculations, and, in the case of the paramagnetic species, electron paramagnetic resonance (EPR) and magnetic susceptibility. Variation of alkyl groups on the ligand from t-octyl to n-propyl enabled electronic isolation of the complexes in the crystal structures of M(L¹)₂ contrasting with π-stacking interactions for M(L²)₂ (M = Ni, Cu). This was evidenced by a one-dimensional antiferromagnetic chain for Cu(L²)₂ but ideal paramagnetic behavior for Cu(L¹)₂ down to 1.8 K. Despite isostructural single crystal structures for M(L²) ₂, thin-film X-ray diffraction and scanning electron microscopy (SEM) revealed different morphologies depending on the metal and the deposition method (vapor or solution). The Cu complexes displayed limited electronic interaction between the central metal and the delocalized ligands, with more mixing in the case of Ni(II), as shown by electrochemistry and UV/vis spectroscopy. The complexes M(L²)₂ showed poor charge transport in a field-effect transistor (FET) device despite the ability to form π-stacking structures, and this provides design insights for metal complexes to be used in conductive thin-film devices.

Full text of DOI:10.1021/ic2020644

Article
pubs.acs.org/IC
Structural, Magnetic, and Electronic Properties of Phenolic Oxime
Complexes of Cu and Ni
Alexander M. Whyte,† Benjamin Roach,† David K. Henderson,† Peter A. Tasker,† Michio M. Matsushita,‡
Kunio Awaga,‡ Fraser J. White,† Patricia Richardson,† and Neil Robertson*,†
†School of Chemistry and EaStChem, University of Edinburgh, Joseph Black Building, West Mains Road, Edinburgh, Scotland EH9
3JJ, U.K.
‡Department of Chemistry, Graduate School of Science, and Research Center of Materials Science, Nagoya University, Chikusa-ku,
Nagoya 464-8602, Japan
S
* Supporting Information
ABSTRACT: Square planar complexes of the type Ni(L1)2, Ni(L2)2,
Cu(L1)2, and Cu(L2)2, where L1H = 2-hydroxy-5-t-octylacetophenone
oxime and L2H = 2-hydroxy-5-n-propylacetophenone oxime, have been
prepared and characterized by single-crystal X-ray diffraction, cyclic
voltammetry, UV/vis spectroscopy, field-effect-transistor measurements,
density functional theory (DFT) and time-dependent DFT (TDDFT)
calculations, and, in the case of the paramagnetic species, electron
paramagnetic resonance (EPR) and magnetic susceptibility. Variation of
alkyl groups on the ligand from t-octyl to n-propyl enabled electronic
isolation of the complexes in the crystal structures of M(L1)2 contrasting
with π-stacking interactions for M(L2)2 (M = Ni, Cu). This was
evidenced by a one-dimensional antiferromagnetic chain for Cu(L2)2 but
ideal paramagnetic behavior for Cu(L1)2 down to 1.8 K. Despite
isostructural single crystal structures for M(L2)2, thin-film X-ray diffraction and scanning electron microscopy (SEM) revealed
different morphologies depending on the metal and the deposition method (vapor or solution). The Cu complexes displayed
limited electronic interaction between the central metal and the delocalized ligands, with more mixing in the case of Ni(II), as
shown by electrochemistry and UV/vis spectroscopy. The complexes M(L2)2 showed poor charge transport in a field-effect
transistor (FET) device despite the ability to form π-stacking structures, and this provides design insights for metal complexes to
be used in conductive thin-film devices.
bis(o-diiminobenzosemiquinonate) nickel(II),5,23 with much
less attention given to the development of a wider variety of
examples.
1. INTRODUCTION
Because of their diverse redox, spectroscopic, and magnetic
properties, transition metal complexes are of widespread
importance in the study of functional molecular materials,1−3
In this context it is important to investigate new classes of
planar, electronically delocalized metal complexes or to re-
with key properties including accessible oxidations/reduc-
tions,4,5 and tunable intermolecular interactions through
evaluate appropriate existing materials previously used for a
modification of the ligand6 or by exchange of the transition
different purpose. This requires an extensive study of different
metal ion.7 The presence of unpaired electrons in some cases
classes of planar metal complexes to assess their electronic
structure; their processability into thin films by evaporation or
has played a key role in the extensive field of molecular
solution methods; their molecular packing in the solid state; the
magnetism and, when coupled with charge transport properties,
can lead to novel magneto-conductive effects.8−10 Metal
resulting intermolecular interactions, their magnetic properties,
and the possibility of charge-transport properties. Phenolic
complexes with a planar, electronically delocalized structure
oxime ligands are used extensively in extractive hydro-
have proven particularly attractive for development of
metallurgy,24−26 but the cooperative electronic properties in
cooperative electronic properties because of the strong
molecule−molecule interactions that can arise from π-stacking
materials formed by transition metal complexes of these ligands
of the planar units.11−13 This motif has led to materials showing
has to date been overlooked. We have prepared homoleptic
ferromagnetism,14 one-dimensional magnetic chains,15 semi-
nickel(II) and copper(II) complexes using two structurally
conductive behavior,16,17 and a variety of structural or magnetic
related phenolic oxime ligands, 2-hydroxy-5-t-octylacetophe-
none oxime (L1H) and 2-hydroxy-5-n-propylacetophenone
phase-transitions.18 The majority of research in this field,
however, has focused on well-established families of planar
metal complexes such as metal-bis-1,2-dithiolenes,19−21 metal
phthalocyanines,22 and structurally related analogues such as
Received: September 21, 2011
Published: November 18, 2011
© 2011 American Chemical Society
12867
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876

Inorganic Chemistry
Article
oxime (L2H). The size of the alkyl group at the 5-position on
the aromatic ring has been varied, which affects both the
solubility of the resulting complexes formed and thus the
processing methods available, and also intermolecular inter-
actions in the solid state, which determine the packing adopted
by the molecules and the corresponding cooperative electronic
properties that arise. We report a systematic study of the
electronic, magnetic, structural, and thin-film properties of
these materials to provide a thorough assessment of their
potential for use in functional molecular materials and devices.
precipitate was collected by filtration, washed with ethanol (10 mL),
then n-hexane (10 mL) and dried in a vacuum desiccator overnight.
Cu(L1)2. MS (ESI): m/z (%) = 588.15 (76.41%) [M+]. Calculated
for C32H48CuN2O4, C 65.33, H 8.22, N 4.76; found C 65.34, H 8.24,
N 4.76.
Cu(L2)2. MS (ESI): m/z (%) = 447.4 (100%) [M+]. Calculated for
C32H28CuN2O4, C 58.98, H 6.30, N 6.25; found C 59.04, H 6.16, N
6.13.
Ni(L1)2. 1H NMR (CDCl3): 0.64 (s, 9H), 1.25 (s, 6H), 1.49 (s,
2H), 2.42 (s, 3H), 6.62 (d, 1H), 7.07 (d, 1H), 7.19 (s, 1H), 10.90 (s,
1H). MS (ESI): m/z (%) = 583.11 (89.08%) [M+]. Calculated for
C32H48NiN2O4, C 65.88, H 8.29, N 4.80; found C 65.78, H 8.95, N
4.95.
Ni(L2)2. 1H NMR (CDCl3): 0.96 (t, 3H), 1.64 (m, 2H), 2.40 (s,
3H), 2.57 (t, 2H), 6.94 (d, 1H), 7.12 (d, 1H), 7.24 (s, 1H). MS (EI):
m/z (%) = 442.1 (100%) [M+]. Calculated for C32H28NiN2O4, C
59.62, H 6.37, N 6.32; found C 59.72, H 6.30, N 6.21.
2. EXPERIMENTAL SECTION
2.1. Synthesis of Materials. All chemicals were purchased from
Sigma Aldrich and used as bought without further purification.
2-Hydroxy-5-t-octylacetophenone Oxime (L1H). Acetyl chloride
(86.3 g, 1.1 mol) was added dropwise to a solution of 4-tert-
octylphenol (206.3 g, 1 mol) in toluene (600 mL), and the resulting
solution refluxed for 4 h and then stirred overnight. Aluminum
trichloride (133.3 g, 1.0 mol) was added in portions to the stirred
solution over 4 h, and the reaction mixture was then heated to reflux
for 3 h and stirred overnight at room temperature (RT). The reaction
was quenched with excess HCl (20%, 500 mL) which was added
dropwise over 1 h. The organic phase was separated from the aqueous,
washed with distilled water (2 × 250 mL), and filtered using phase
separation paper. The solvent was removed in vacuo yielding a viscous
yellow oil (230.2 g) 67% pure by GC. Hydroxylamine sulfate (221.5 g,
1.35 mol) and sodium acetate (221.4 g, 2.7 mol) were added to a
solution of crude phenolic ketone (230.2 g, 0.9 mol) in ethanol (600
mL). The resulting suspension was refluxed for 2 h, allowed to cool,
and poured onto toluene (500 mL) and distilled water (300 mL). The
organic layer was separated, washed with distilled water (2 × 100 mL)
and brine (100 mL), and filtered through phase separation paper. The
solvent was removed in vacuo, and the resulting yellow solid was
purified via complexation with copper, using the same general
procedure as detailed below. The complex solution was stripped
using sulphuric acid solution (0.1 M, 3 × 200 mL), washed with
distilled water (200 mL), and filtered through phase separation paper.
The solvent was removed in vacuo, and the resulting off-yellow solid
was recrystallized from hexane yielding an off-white solid (135.2 g,
2.2. Experimental Measurements. Crystallographic data were
routinely collected at 100 K. Single crystals suitable for X-ray
diffraction (XRD) were prepared by various methods: Cu(L1)2 by slow
evaporation of a MeOH/DCM/THF solution; Cu(L1)2·DMSO by
recrystallization from a concentrated dimethylsulfoxide (DMSO)
solution; Cu(L2)2 by vapor diffusion of EtOH into a CHCl3 solution;
Ni(L1)2 by slow evaporation of a (CH3O)2CO solution and Ni(L2)2
by vapor diffusion of MeCN into a CHCl3 solution. A table of
crystallographic data is included in the Supporting Information. Thin
film XRD was carried out on a Rigaku ultraX-18HB at RT. Data were
collected from 2θ angle of 5−40° at a rate of 2° per minute. Powder
XRD was carried out using a Bruker AXS D8 diffractometer. Thin
films were grown in a vacuum chamber on various substrates: Si
(1,0,0), quartz, indium tin oxide (ITO), glass, field-effect transistor
(FET) (bottom contact configuration) substrates and polyethylene
terephthalate (PET). Different substrates were required for the
different types of characterization: Si was used for thin film XRD
and infrared spectroscopy (IR); quartz for UV/vis and scanning
electron microscopy (SEM); glass for UV/vis and PET for electron
paramagnetic resonance (EPR). All the substrates, except PET and the
FETs, were cleaned in individual solutions of IPA, acetone, and then
chloroform prior to use. Deposition was carried out via vacuum
sublimation in a temperature range between 135 and 210 °C at a
pressure of 2.8 × 10−4 Pa. This resulted in a growth rate of 0.1−0.3 Å/
s which was monitored using a quartz crystal microbalance (QCM).
The material to be sublimed was heated inside an inert crucible by
applying a current. Films of approximately 100 nm thickness were
produced according to the QCM. All cyclic voltammetry measure-
ments were carried out in dry DCM using 0.3 M TBABF4 electrolyte
in a three electrode system, with each solution being purged with N2
prior to measurement. The working electrode was a 0.2 mm2 Pt wire
sealed in glass. The reference electrode was Ag/AgCl calibrated against
Ferrocene/Ferrocenium in the background electrolyte, and the
counter electrode was a Pt rod. All measurements were made at RT
using an μAUTOLAB Type III potentiostat, driven by the electro-
chemical software GPES. Cyclic voltammetry (CV) measurements
used scan rates of 0.1 V/s, and DPV was carried out at a step potential
of 0.01005 V, modulation amplitude of 0.10005 V, modulation time of
0.05 s and an interval time of 0.5 s. Solution UV/vis spectra were
recorded in solution in DCM using a quartz cell of path length 1 cm
on a Perkin-Elmer Lambda 9 spectrophotometer, controlled by a
datalink PC, running UV/Winlab software and in thin films on a Jasco
V-570 UV/vis/NIR spectrophotometer. Magnetic susceptibility
measurements were performed on powder samples from 1.8 to 300
K using a Quantum Design MPMS-XL SQUID magnetometer with
MPMS MultiVu Application software to process the data. The
magnetic field used was 0.1 T. Diamagnetic corrections were applied
to the observed paramagnetic susceptibilities by using Pascal’s
constants. EPR spectra were measured on a Bruker ER200D X-band
spectrometer with simulations using the Bruker EPR simulation
package SimFonia.27 EPR spectra of dry DCM solutions were
recorded at RT. Thin film EPR spectra were obtained on a JEOL
JES-FA200 ESR Spectrometer using 100 nm films on a PET substrate
1
0.51 mol, 51% yield). H NMR (CDCl3): δ (ppm) = 0.78 (s, 9H),
1.43 (s, 6H), 1.78 (s, 2H), 2.44 (s, 3H), 7.32 (d), 7.43 (s), 7.46 (d).
MS (ESI): m/z (%) = 264.20 (100%) [M+]
2-Hydroxy-5-n-propylacetophenone Oxime (L2H). Acetyl chloride
(11.77 g, 0.15 mol) was added dropwise to a stirred solution of 4-n-
propylphenol (13.62 g, 0.1 mol) in toluene (250 mL), and the
resulting mixture was refluxed for 4 h and stirred overnight at RT.
Aluminum trichloride (16.00 g, 0.12 mol) was added in 1 g portions to
the stirred solution over 4 h at 10 min intervals, and the reaction was
then refluxed for 3 h before being quenched with aqueous
hydrochloric acid (18%, 100 mL) added dropwise over an hour.
The organic layer was separated from the aqueous, washed with
distilled water (2 × 50 mL), and then dried using excess magnesium
sulfate. Evaporation of the solvent under reduced pressure gave an
orange oil which was dissolved in ethanol (200 mL). A filtered
solution of hydroxylamine hydrochloride (10.42 g, 0.15 mol) and
potassium hydroxide (8.42 g, 0.15 mol) was added, and the resulting
reaction mixture heated at reflux for 3 h, allowed to cool, and then
poured into a toluene/water (100 mL:100 mL) solution. The organic
layer was washed (50 mL water, 50 mL brine) and dried in vacuo to
yield an orange oil. The salicylketoxime ligand was then recrystallized
as an off-white solid (12.52 g, 65%) by cooling a concentrated n-
1
heptane (50 mL) solution. H NMR (CDCl3): δ = 0.96 (t, 3H), 1.64
(m, 2H), 2.58 (t, 2H), 2.61 (s, 3H), 7.26 (d, 1H), 7.32 (s, 1H), 7.42
(d, 1H). MS (ESI): m/z (%) = 194.31 (100%) [M+]
The same general procedure was adopted for the synthesis of all of
the metal complexes. The ligand (0.02 mol) was added to a stirred
solution of Cu(OAc)2·H2O or Ni(OAc)2·4H2O (0.01 mol) in ethanol
(50 mL). The resulting solution was stirred for 24 h at RT, and the
12868
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876

Inorganic Chemistry
Article
Cu(II) atom in the DMSO adduct, [Cu(L1)2DMSO] (Figure
2), which has a five coordinate square pyramidal structure. The
with a center field of 312 mT and a sweep width ±100 mT. Geometry
optimizations of the isolated complexes, Cu(L2)2 and Ni(L2)2, were
carried out at the B3LYP/6-31G(d,p) level of theory,28−30 using
Gaussian 03.31 The X-ray crystallographic coordinates were used as the
starting structures, and minima on the potential energy surface were
confirmed by the absence of any imaginary frequencies. The molecular
orbital isosurfaces were visualized using ArgusLab 4.0.32 FET
measurements were carried out using bottom contact devices
consisting of interdigitated platinum source and drain electrodes
with a silicon dioxide dielectric layer. The source and drain electrodes
were 8 μm thick with a gap between the interdigitated electrodes of 2
μm. The dielectric layer was 300 nm thick, and the gate electrode was
made from an n-doped silicon wafer. FET testing was done in darkness
under vacuum. Imaging of thin films was carried out using a Hitachi S-
4300 Scanning Electron Microscope.
3. RESULTS AND DISCUSSION
The salicylketoxime ligands discussed in this work are planar,
bidentate, and bind through two heteroatoms making them
interesting candidates for investigation to assess their potential
in magnetic or conducting materials. Synthesis of both Ni and
Cu complexes of two different phenolic oxime ligands was
readily achieved by direct reaction of the metal acetate salt with
the protonated ligand in ethanol (Figure 1), enabling
Figure 2. Unit cell of Cu(L1)2.DMSO. Hydrogen atoms have been
omitted for clarity and only the higher occupancy site of the
disordered (CH3)2S unit of the DMSO is shown.
bite angle is larger in the Ni(II) complexes, which is a
consequence of the shorter Ni−O and Ni−N bonds requiring
the metal atom to move toward the center of the N···O bite.
The crystal packing of Cu(L1)2 and Ni(L1)2 are compared in
Figure 3. Because of the bulky t-octyl chains, π-overlap of the
Figure 1. Molecular structure of the four complexes, M = Ni, Cu; R =
t-octyl (L1), n-propyl (L2).
comparison of the role of the central metal in the electronic
properties, and the role of the alkyl chain in the packing
properties of the molecules. The complexes achieve good
stability through the pseudomacrocylic arrangement arising
from hydrogen bonding between the oximic hydrogen and the
phenolic oxygen of the opposing ligands.
3.1. Crystal Structures. In the crystal structures of the
four unsolvated complexes, Cu(L1)2, Ni(L1)2, Cu(L2)2, and
Ni(L2)2, the metal atom lies on a crystallographic inversion
2−
center which results in perfectly planar N2O2 donor sets.
Figure 3. Unit cells of (a) Cu(L1)2 viewed along the a-axis and (b)
Ni(L1)2 viewed along the b-axis. Hydrogen atoms have been omitted
for clarity, and only the higher occupancy site of one of the octyl
groups, which is disordered in Cu(L1)2, is shown.
Bond lengths and angles and the chelate bite distances in the
coordination sphere, together with the O···O contact distances
in the outer sphere which are associated with oxime OH to
phenolate oxygen hydrogen bond, are shown below (Table 1).
As noted previously,24 the bond lengths to the Ni(II) atom are
significantly shorter than those to the Cu(II) atom. As might be
expected,33 the N2O2 donor set forms longer bonds to the
2−
Table 1. Selected Bond Length and Angles
a
[Cu(L1)2·DMSO]
part a
part b
Cu(L1)2
Ni(L1)2
Cu(L2)2
Ni(L2)2
́
M−O (Å)
1.919(2)
1.957(4)
89.7(1)
1.901(3)
1.964(4)
90.5(1)
1.870(1)
1.962(1)
91.05(5)
89.95(5)
2.735(2)
2.583(2)
1.819(1)
1.883(2)
92.12(6)
87.88(6)
2.666(2)
2.482(2)
1.883(1)
1.948(2)
91.84(7)
88.16(7)
2.753(2)
2.551(2)
1.827(1)
1.883(1)
92.40(5)
87.60(5)
2.678(2)
2.471(1)
́
M−N (Å)
O−M−N (deg)
O−M−N′ (deg)
N···O bite distance (Å)
MO···OH
89.5(1)
88.9(1)
́
2.745(5)
2.598(4)
2.733(5)
2.591(4)
a
This complex has no crystallographically imposed symmetry and two lengths and angles associated with the chelate units a and b in Supporting
́
Information, Figure S3 are shown. The DMSO oxygen atom forms a bond to Cu of 2.229(3) Å and defines angles of 93.8(1), 98.3(1), 95.2(1), and
91.9(1)o with atoms O2, O45, N6, and N81.
12869
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876

Inorganic Chemistry
Article
planar portions of the molecules is not favored and the
interplanar distance between overlapping aromatic rings is 6.97
1
1
1
́
́
́
Å in Cu(L )2·DMSO, 7.58 Å in Cu(L )2, and 7.12 Å in Ni(L )2.
Ni(L2)2 and Cu(L2)2 are isomorphous and, in contrast to
M(L1)2, display a herringbone packing motif with the π-
stacking of the molecules extending along the b-axis in a regular
stack (Figure 4). The interplanar packing distance is slightly
longer, 3.22 Å, in Ni(L2)2 than in Cu(L2)2, 3.17 Å.
Figure 5. Cyclic voltammetry of complexes Cu(L1)2 [open squares],
Ni(L1)2 [blue dashed lines], Cu(L2)2 [green lines] and Ni(L2)2 [red
solid circles], at a scan rate of 0.1 V/s between 0 and 2 V.
Each complex exhibits an irreversible oxidation occurring at
approximately the same potential, indicating that the highest
occupied molecular orbital (HOMO) of each complex is
substantially localized on the ligand. Replacement of the
phenolic protons by Ni2+ has little effect on the redox processes
compared with the free ligands ,and no other redox processes
have been generated through the redox activity of the metal.
None of the samples displayed a reduction process between 0
and −2 V.
3.3. UV/vis Spectroscopy. The UV−vis spectra of the
ligands, L1H and L2H, in the range 250−800 nm (Supporting
Information, Figure S11) are very similar. Both spectra are
dominated by a strong absorption band occurring at 315 nm,
assigned as a π−π* transition. The bis-ligand Ni(II) and Cu(II)
complexes formed from L1H and L2H have also been studied
by UV−vis absorption spectroscopy (Figure 6). The absorption
Figure 4. Crystal packing of Ni(L2)2. The mean planes (red and blue)
have been calculated using both aromatic rings on each molecule using
the program Mercury 2.3.34−37 Hydrogen atoms have been omitted for
clarity.
t
n
The steric effects of the octyl or propyl substituents clearly
have an impact on the solid state packing of the complexes. The
complexes of L2 form a stack with short intermolecular
distances suggesting strong π-stacking interactions, whereas
those formed from L1H have larger interplanar distances, too
great for effective π-overlap. This presents an opportunity to
investigate the effect of intermolecular π-stacking interactions
on magnetic and conduction properties, as the L1H/L2H pairs
share a common core electronic structure. It provides a
comparison between materials composed of essentially
electronically identical molecules that interact in the solid
state with those that are effectively isolated.
3.2. Electrochemistry. Both the ligands were studied in
the range of −2 to 2 V at various scan rates and both display
similar behavior. The cyclic voltammetry (CV) measurements
(Table 2) indicate that L1H and L2H are oxidized irreversibly at
a
Table 2. Electrochemical Data Measured by CV and DPV
differential pulse
cyclic voltammetry
pc (V)
voltammetry
sample
E
Epc (V)
L1H
1.40
1.42
1.43
1.29
1.40
1.35
1.80
1.78
1.80
1.82
1.32
1.75
1.71
1.72
1.74
1.82
1.73
L2H
1.34
1.22
1.24
1.27
1.20
Cu(L1)2
Ni(L1)2
Cu(L2)2
Ni(L2)2
Figure 6. UV/vis spectra of DCM solutions of the complexes overlaid
with thin film absorption spectra of M(L2)2 complexes on SiO2
substrates (arbitrary absorbance units).
a
Supporting Information, Figures S5−S10. The second oxidation
spectra for the Ni and Cu complexes are similar in overall
structure; however, there is a marked red shift in the Ni spectra
relative to that of Cu. The spectra for both Ni complexes are
almost identical, and are dominated by a strong absorption
band at 305 nm, with a weaker band observed at 383 nm
whereas the Cu spectra for both complexes show a prominent
UV band around 260 nm, with a weaker band appearing around
346 nm. At high concentrations all of the complexes display
Laporte forbidden d-d transitions in the region 606−654 nm.
process was not well-defined in complexes Cu(L2)2 and Ni(L2)2 by
CV.
1.41 and 1.42 V. L2H displays another irreversible oxidation at
1.78 V which is not as well-defined in the CV voltammogram of
L1H.
The voltammograms of the complexes (Figure 5) are very
similar to those of the uncomplexed ligands, which implies that
the redox processes involved are predominantly ligand based.
12870
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876

Inorganic Chemistry
Article
The more intense bands centered between 346 and 383 nm
(Figure 6) are assumed to be predominantly intraligand in
character, but their energies are dependent on the nature of the
complexed metal. The difference in energy of the UV
absorptions when comparing the Cu and Ni complexes
indicates a metal orbital contribution to the transition, while
the differing alkyl groups exert only a minimal influence. These
transitions are in the same general range as that witnessed in
the free ligand, indicating metal and ligand orbital mixing is
limited, although noticeably greater for Ni than for Cu.
The electronic absorption spectra of 100 nm thin films of the
M(L2)2 complexes on quartz are shown in Figure 6. The thin
film spectrum of the Cu(II) complex closely resembles its
solution spectrum but for the Ni(II) complex there is a
noticeable degree of red shifting. The lack of red shifting in the
Cu(II) could infer that the intermolecular interactions in the
solid state are weaker, attributed to the different metal ions
affecting intermolecular interactions. For example, with the
Ni(II) complexes, the larger involvement of the metal in the
frontier orbitals of the individual molecules may facilitate
stronger orbital interaction between molecules.
mixed metal(dxz) and ligand character, but the HOMO-3 is
solely based on the dz orbital.
2
3.5. SOMO Wrestling. The calculation of the electronic
structure of copper containing complexes continues to be a
challenge for electronic structure methods. Recent work on Cu-
phthalocyanines has shown that the predicted electronic
structure, particularly the ordering of the occupied valence
orbitals, is sensitive to the level of theory applied, and in the
case of DFT, sensitive to the nature of the functional
employed.38−41 While in most cases the unpaired electron is
2
2
calculated to reside in the Cu-based dx −y orbital, in agreement
with experimental evidence from EPR measurements,42,43
contradictory predictions place it variously above and below
the doubly occupied HOMO orbital.
In the course of this work, similar effects were also observed
for the copper complexes, where not just the electronic
structure, but the geometrical structure was seen to be sensitive
to the nature of the functional chosen. Relatively large basis sets
6-31+g(d,p) and cc-pVTZ were used in combination with a
number of functionals to investigate the sensitivity of the
electronic structure to the level of theory (Supporting
Information, Table S3). Of the functionals under study, the
“pure” functionals (OLYP, OPBE, PBE, SVWN, and BLYP)
tend to place the SOMO within the HOMO−LUMO gap,
predicting that the lower energy HOMO is a doubly occupied
ligand-based π-orbital. HF and the hybrid and long-range
corrected functionals (B3LYP, M06-HF, PBE1PBE, BH and
HLYP, CAM-B3LYP), in contrast, tend to place the SOMO at
an energy below the doubly occupied ligand-based π HOMO.
In the case of OLYP, OPBE, PBE and B3LYP(VWN5) the
geometry of the copper complexes was predicted to be
significantly nonplanar around the Cu center.
3.4. Computational Procedures. To estimate the en-
ergies of the HOMO and the lowest unoccupied molecular
orbital (LUMO) in our complexes, single molecule gas phase
calculations were carried out at the B3LYP/6-31G(d,p) level of
theory for complexes Cu(L2)2 and Ni(L2)2. The extended alkyl
chain on the L1H complexes is not expected to alter the
electronic properties significantly from that of the L2H
complexes. The results (inset Figure 7) indicate that the
Recent work by Kronick et al.44 has suggested the self-
interaction error, inherent in the majority of DFT functionals, is
responsible for the difference in the electronic structures
predicted by local and hybrid functionals. The self-interaction
error, which is more pronounced in the nonhybrid functionals,
tends to destabilize more localized orbitals, in this case the
2
2
singly occupied Cu-based dx −y orbital, raising it in energy
above the doubly occupied ligand-based HOMO. When a
fraction of HF exchange is included in the hybrid functionals
2
2
the self-interaction error is reduced, and the Cu-based dx −y
orbital is calculated to lie lower in energy than the HOMO
(Supporting Information, Figure S12, Table S3).
Figure 7. Time dependent DFT generated UV−vis spectrum (red
dashed lines) overlaid with the experimental spectrum (blue solid
lines) of Ni(L2).
Irrespective of the functional employed, all computational
methods consistently predict that the unpaired electron does
2
2
indeed reside in the mainly Cu-based dx −y orbital, consistent
with the experimental EPR results (vide infra). The insensitivity
of the electrochemical redox peaks to the choice of central
metal atom and the lack of an observable oxidation from the
Cu(II) SOMO also supports a doubly occupied ligand based
orbital as the HOMO. Figure 10 (inset) depicts the calculated
spin density of the unpaired electron on the Cu(II) complex
using the B3LYP/6-31G(d,p) level of theory. As expected the
HOMO of Ni(L2)2 is of mixed metal/ligand character and is
delocalized over both the aromatic part of the molecule and the
Ni(II) center. They also show that the LUMO is localized
solely on the ligand. The energy of the HOMO has been
calculated as −5.18 eV and the LUMO is −1.33 eV. The
energies of the frontier orbitals and the large HOMO−LUMO
gap of 3.85 eV is consistent with the observation of accessible
oxidation but no accessible reduction in the electrochemistry.
The electronic transitions witnessed in solution have been
assigned using time dependent DFT, with the simulated spectra
in good agreement with that obtained experimentally (Figure
7). The transition at 383 nm has been assigned as being from
the HOMO to LUMO (92%) whereas the transition at 305 nm
is from HOMO-2 to LUMO (62%) and HOMO-3 to LUMO
(24%). The HOMO is composed of Ni(dyz) and Lπ character
with the LUMO assigned as a Lπ* orbital. The HOMO-2 is of
2
2
unpaired electron appears to be based in the dx −y orbital of the
metal, with some delocalization onto the donor atoms of the
ligand.
3.6. Thin Film XRD. Complexes of the type M(L2)2 were
found to be volatile, and thin films on various substrates have
therefore been prepared by sublimation under reduced
pressure. XRD measurements were carried out at RT between
angles of 5° and 40° (2θ). Infrared (IR) spectroscopy carried
out on these films indicates that the thin films are structurally
12871
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876

Inorganic Chemistry
Article
the same as the powder sample prior to deposition (Supporting
Information, Figure S1, S2). This also reassures us that sample
decomposition has not taken place during volatilization.
XRD measurements carried out on powder samples of
Cu(L2)2 matches the calculated powder pattern generated from
the single crystal data indicating the bulk powder is in the same
crystallographic phase as that which was isolated from solution
as a single crystal. However, XRD measurements on powdered
Ni(L2)2 appears to show additional peaks that do not
correspond to the structure obtained from the single crystal
data. This indicates that the sample contains a mixture of
crystallographic phases or a trace amount of a crystalline
impurity.
Shown in Figure 8 is the thin film XRD of Cu(L2)2 on a Si
wafer. Contrasting it with the predicted powder pattern appears
Figure 9. Out of plane thin film XRD of Ni(L2)2 on a Si substrate
(blue) overlaid with the calculated powder pattern from the single
crystal data (red) and the experimental powder pattern (black).
3.7. Magnetic Susceptibility. Measurements carried out
on a powder sample of Cu(L1)2 are shown in Supporting
Information, Figures S16 and S17, which clearly indicates
paramagnetic behavior exhibited by the sample with no
indication of any significant magnetic ordering or interactions
at low temperature. This result is expected because of the steric
bulk of the tertiary octyl chain present on the ligand, L1H. The
steric effect of the ligand on the crystal packing results in a large
intermolecular distance between Cu(II) centers, approximately
7 Å, such that the molecule behaves as an isolated paramagnet.
Fitting to the Curie−Weiss law in the range 1.8−300 K
(Supporting Information, Figure S16) the Curie constant has
been determined as 0.3794 cm3 K mol−1, leading to a g factor of
2.012, and the Weiss constant as 0.0118 K. Because of the spin
orbit coupling present in Cu(II), the g-value is often anisotropic
and typically lies between 2.0 and 2.30.45 As the g factor is close
to the free electron value of 2.002, this would suggest the
unpaired electron has significant ligand character.
Figure 8. Out of plane thin film XRD of Cu(L2)2 on a Si substrate
(blue) overlaid with the calculated powder pattern from the single
crystal data (red) and the experimental powder pattern (black).
to suggest a different crystallographic phase in the thin film
from that present in the single crystal. Note that the intense
peak at 33° (2θ) is due to the substrate. The large peaks at
6.18° and 8.38° (2θ) in the thin film XRD correspond to a d-
spacing of 14.28 and 10.54 Å, respectively, but have not been
indexed because of the lack of distinct peaks available. From the
single crystal data, the intense peaks at 8.13, 9.00, and 12.45°
correspond to a d-spacing of 10.86, 9.81, and 7.10 Å,
respectively. These peaks are due to reflection from the
(100), (002), and (102) planes.
As witnessed with the Cu(II) complex, the analogous Ni(II)
sample also appears to exhibit a different structure when
thermally grown on a Si thin film, as can be observed in Figure
9. The prominent peaks at 8.04 and 10.47° (2θ) correspond to
an intermolecular d-spacing of 10.98 and 8.23 Å, respectively.
Contrasting this with the single crystal data, the dominant
peaks are at 8.13, 8.92, and 12.30°, similar to the thin film
obtained from the volatile Cu(II) complex, with the same
assignment of planes.
At low temperature the χT vs T plot for Cu(L2)2 shows a
downward trend indicating the metal centers are interacting
antiferromagnetically (Figure 10). Treating the sample as a 1-D
chain of interacting S = 1/2 molecules, the experimental data
were fit to the Bonner−Fisher expression46 modified by Estes
et al.47 This calculates the isotropic g-value as 2.024 and J as
0.49 cm−1. The small J value indicates a weak antiferromagnetic
coupling, giving a susceptibility peak at about 6 K. The g-value
is rather low; however this is consistent with the Curie−Weiss
fit and with the behavior of Cu(L1)2.
The outcome of modifying the alkyl chain on the chelating
ligand can be directly seen by comparing the magnetic data; the
small npropyl chain has caused the interplanar distance to
shorten to allow π−π stacking between molecules and as a
result the Cu(II) centers can now interact magnetically. The
change from paramagnetic behavior in Cu(L1)2 to weak
antiferromagnetic coupling in Cu(L2)2 can be attributed to
the interplanar distance shortening to ∼3 Å in Cu(L2)2 from
∼7 Å in Cu(L1)2.
From the single crystal XRD data the Cu(L2)2 and Ni(L2)2
complexes appear isomorphous but on a thin film these
materials are clearly exhibiting a different phase to each other.
The single crystal data have all come from solution grown
samples but to explore the possibility of polymorphism from
vapor grown samples, attempts to isolate single crystals by
sublimation methods have been made. However, crystals grown
thus far have not been of a suitable quality for single crystal
analysis.
3.8. EPR. Because of experimental limitations in magnetic
susceptibility measurements, a more accurate estimation of the
g-value is obtained from EPR. Measurements have been carried
out on solution samples of both Cu(II) complexes at RT. In the
case of Cu(L1)2, the EPR signal is split into a quartet of triplets
owing to the Cu nuclear spin of 3/2 (quartet) and 14N spin of 1
12872
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876

Inorganic Chemistry
Article
Figure 10. χT vs T plot of Cu(L2)2 from 1.8−300 K. The Bonner−
Fisher model curve (red line) displays a good fit to the experimental
data (black squares). The curve has been corrected for a temperature
independent paramagnetic parameter of 3.15 × 10−4 cm3 mol−1.
Powder EPR of Cu(L2)2 showing the change in g-factor with
temperature (hollow circles) has been overlaid. Inset is the calculated
spin-density at the B3LYP/6-31G(d,p) level of theory.
Figure 11. Solution EPR results for Cu(L2)2 (upper) and simulated
spectrum (lower). Measurement carried out in DCM at 25 °C.
(triplet). Note that both isotopes of Cu have the same nuclear
spin. From the EPR simulation, the following values have been
determined: coupling to the 63Cu nuclei of 90.5 G, coupling to
the 65Cu nuclei of 94.5 G, coupling to both the 14N nuclei of
14.3 G, a line width of 10.0 G and giso of 2.11. At lower
magnetic field the resolution of the hyperfine coupling is not
well resolved and thus simulation is only possible for the
resolved part of the spectrum at higher magnetic field; a higher
frequency EPR experiment may improve the resolution. The
coupling of the unpaired electron to the 14N nucleus indicates
that in the complex it is not completely localized on the Cu(II)
center but some of the electron density is on the N donor
atoms of the ligand, which implies that the SOMO of the
complex is delocalized onto part of the ligand. This is
consistent with the computational results above. The simulated
g factor is slightly larger than that determined by fit to the
susceptibility data although broadly consistent.
resembles the χT vs T data from magnetic susceptibility
measurements. This reaffirms that using a sterically smaller
ligand is having the desired effect on the intermolecular
interactions in the solid state.
3.9. SEM Imaging. The performance of a thin film device,
such as a transistor, can be influenced by crystallinity and grain
boundaries48,49 so SEM has been carried out to investigate the
thin film morphology. The Ni(II) and Cu(II) complexes appear
to exhibit different morphologies to each other on both ITO
and Si. The Cu(II) crystals appear needle like in shape while
the Ni(II) film consists of crystallites of various shapes with less
even coverage of the substrate. SEM images of the thin films on
Si, ITO, and FET substrates at different magnifications (Figure
12, Supporting Information, Figure S19), highlight that the
Ni(II) complex has not formed homogeneous films. The
source-drain gap of our FET substrates (vide infra) is only 2
μm so the films, although not ideal, should still be suitable for
FET studies. Both films were formed under similar conditions,
but a higher temperature was reached in the deposition of
Cu(L1)2 and this may be responsible for the pronounced
change in morphology. Both films were deposited at a rate of
0.1−0.3 Å per second but in the deposition of Cu(L2)2 a
maximum temperature of 213 °C was reached as opposed to
157 °C in the deposition of Ni(L2)2.
The solution deposited materials, in contrast to the vacuum
grown films, form large needle-like crystallites much bigger than
those which are formed by sublimation. The solution grown
crystallites effectively bridge the source-drain gap of 2 μm
whereas not all the vacuum deposited particles are this large.
3.10. FET Measurements. A series of FET measurements
was carried out using the less sterically hindered, M(L2)2,
complexes to form thin films by both vapor processing and
solution coating. From the cyclic voltammetry it appears that
these materials are weak electron donors therefore p-type
conduction may be expected. However, none of these
complexes produced gate voltage effects when measurements
were carried out using the FET configuration described earlier
(see for example Supporting Information, Figure S20). Instead
A similar spectrum was obtained for Cu(L2)2, (Figure 11),
indicating as expected that the paramagnetic systems are closely
related. From the EPR simulation, the following values have
been determined: coupling to the 63Cu nuclei of 90.5 G,
coupling to the 65Cu nuclei of 96.5 G, coupling to both the 14
N
nuclei of 15.8 G, a line width of 10.0 G and giso of 2.10. Again
we see a delocalization of the SOMO onto the N donor on the
ligand, the extent of which is almost identical to Cu(L1)2 by
estimation from the simulation parameters. Complexes such as
Cu(II) tetraphenyl porphyrin (TTP) have been previously
studied by EPR42 and reported to have an average metal
coupling of 95 G, 14N coupling of 14 G and a g-factor of 2.117.
2
2
In this Cu(TTP) example where the spin is based in the dx −y
orbital, the simulation parameters closely match that of our
system implying the unpaired electrons are in similar
environments with similar delocalization onto ligand orbitals.
The powder EPR results of Cu(L2)2, (Figure 10), display
how the g-factor changes as the material magnetically orders at
low temperature. In this case the g-factor indicates a greater
degree of electron delocalization than was estimated by
simulation from the solution measurement. The value falls
from approximately 2.060 to almost 2.057 as the antiferro-
magnetic interactions become more significant at lower
temperature. Reassuringly, the change in g-factor broadly
12873
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876

Inorganic Chemistry
Article
From the electrochemical studies, absorption spectroscopy,
and the computational work, the materials appear to have a
large energy gap between the HOMO and LUMO, and the
electrochemical processes, within the solvent window, are
irreversible. FET results indicate that these materials are poor
conductors and likely not suitable for this application. In
keeping with this, computational, EPR, and magnetic measure-
ments on the Cu(II) complexes have indicated that
intermolecular interactions are weak, which is affirmed by
thin film absorption spectroscopy. The poor FET performance
appears, in part, attributable to the minimal mixing of metal d-
orbitals with the ligands frontier orbitals, and consequently the
lack of facile redox processes. This lack of strong orbital overlap
within the molecule, evidenced by the electrochemical data,
may also play a role in the weak intermolecular interactions.
To maximize intermolecular interactions between frontier
orbitals and to ensure accessible oxidation and reduction
processes, significant delocalization across the molecule is
desirable. In metal-phthalocyanine complexes for example,
extensive delocalization is already inherent in the free-base
compound, and addition of the metal center is not required to
play a specific role in enhancing delocalization. This contrasts
with metal-bis-1,2-dithiolene complexes where the metal plays a
key role in mediating orbital mixing of the two dithiolene
ligands across the complex to provide an extended, delocalized
system. Hence, in dithiolene complexes, the identity of the
central metal and resulting orbital energies is crucial in
determining molecular electronic properties. The phenolic
oxime ligands studied here fall into the latter category where
the metal is required to play an electronic role in mediating
delocalization across the complex. However, in the case of
Ni(II) and especially Cu(II), the metal does not fulfill that role.
EPR spectroscopy indicated a largely Cu(II) centered radical,
with a similar degree of metal−ligand spin density distribution
to that of a Cu-porphyrin complex with some spin density
delocalized onto the donor nitrogen atoms. In such Cu(II)
complexes, we noted the computational complexity that has
previously been observed regarding the relative energy of the
SOMO compared with the other orbitals. Experimentally, the
electrochemical observation of ligand-based oxidation rather
than Cu-centered oxidation suggests a doubly occupied ligand-
based HOMO as the highest orbital. Our extensive survey of
different levels of theory has given guidance on the ability of
different methods to generate theoretical results consistent with
this interpretation. This assessment of a range of functionals
will aid future computational study of other similar Cu(II)
coordination complexes.
Figure 12. SEM images of the volatile complexes on Si and ITO
substrates (upper four images). Also shown at the bottom are drop
coated samples deposited onto FET interdigitated electrodes with a 2
μm gap. Samples were drop coated from saturated solutions of DCM.
Scale shown inset.
each material exhibited insulating behavior under a gate
potential in an FET device, despite the films being highly
crystalline and the π-stacking capability designed into the
complexes. The poor performance may be due to an
unfavorable arrangement of the molecules in the solid state;
however, a likely contributing factor is the lack of significant
mixing of metal d-orbitals with the ligands' frontier orbitals,
particularly for the Cu(II) complexes. The lack of mixing will
limit the delocalization of any introduced holes over the
molecule shown by the little difference between the oxidation
potential of the metal complexes and the uncomplexed ligand.
In addition, the localized ligand orbitals which result will limit
the intermolecular interactions required for charge transport.
3.11. Discussion. Although phenolic oxime complexes
have been extensively studied in metal extraction technology,
systematic study of their electronic properties has been
overlooked, despite their similarity to planar metal complexes
under investigation in molecular magnetic and semiconducting
materials. Variation of the alkyl substituent para to the hydroxyl
group in 2-hydroxy-acetophenone oxime has allowed us to
control the extent of intermolecular interactions in the Ni(II)
and Cu(II) complexes. A bulky alkyl substituent electronically
isolates the molecules in the solid state, while a smaller alkyl
chain allows short intermolecular distances, as witnessed by
comparisons of their single crystal X-ray structures, while still
retaining sufficient solubility to enable solution processing. We
have also shown these materials to be volatile and processable
via vacuum sublimation methods, consistent with the stability
imposed through the pseudomacrocyclic structure.
Still of interest is the characterization and further study of the
thin film crystal packing, in particular how the magnetic
properties alter with crystal packing. The SEM results, coupled
with thin-film XRD, have shown visually how deposition
technique and substrate can have a major impact on the film
morphology and crystal size, but further exploration of how the
deposition conditions affects thin film morphology is required
to attempt to understand the thin film polymorphism.
4. CONCLUSIONS
We have prepared a series of metal phenolic oxime complexes
designed to probe the role of (i) the central metal and (ii) the
peripheral alkyl in determining both the intrinsic electronic
properties of the molecules and the structural packing that leads
to intermolecular interactions and resulting materials proper-
ties. Through extensive electronic, structural, thin-film, and
12874
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876

Inorganic Chemistry
Article
(11) Shimazaki, Y.; Yajima, T.; Takani, M.; Yamauchi, O. Coord.
Chem. Rev. 2009, 253, 479−492.
materials characterization, a thorough assessment has been
possible of the potential of this class of molecules in functional
devices and by implication insight has been gained into the
design of such metal complexes in general. We successfully
showed control of intermolecular π-stacking interactions by
alteration of the peripheral alkyl chain. Despite this processing
control however, intermolecular magnetic interactions re-
mained weak and no evidence of charge transport between
molecules was observed in an FET. We attribute this mainly to
the limited mixing of the metal and ligand orbitals.
Design of new families of metal complexes suitable for
functional materials should therefore focus on a good energetic
match between orbitals of the metal and the delocalized ligand,
or should select polydentate or macrocyclic ligands that already
possess extensive delocalization analogous to the properties of
phthalocyanines, leading to more redox activity and more
delocalized frontier orbitals. Application of the structural and
processing control through alkyl-chain variation, along with the
electronic influence of the central metal achieved in this work,
to molecules where greater metal−ligand orbital mixing is
evident will lead to systems where the unique properties of
metal complexes in semiconducting devices can be fully
realized.
(12) Jia, J.; Hubberstey, P.; Champness, N.; Schroder, M.; Hosseini,
̈
M. W. Supramolecular Chemistry of 4,4′-Bipyridine-N,N′-dioxide in
Transition Metal Complexes: A Rich Diversity of Co-ordinate, Hydrogen-
Bond and Aromatic Stacking Interactions; Springer: Berlin/Heidelberg,
Germany, 2009; Vol. 132, pp 135−161.
(13) Ye, B.-H.; Tong, M.-L.; Chen, X.-M. Coord. Chem. Rev. 2005,
249, 545−565.
(14) Coomber, A. T.; Beljonne, D.; Friend, R. H.; Bredas, J. L.;
Charlton, A.; Robertson, N.; Underbill, A. E.; Kurmoo, M.; Day, P.
Nature 1996, 380, 144−146.
(15) Staniland, S. S.; Fujita, W.; Umezono, Y.; Awaga, K.; Camp, P.
J.; Clark, S. J.; Robertson, N. Inorg. Chem. 2005, 44, 546−551.
(16) Dalgleish, S.; Yoshikawa, H.; Matsushita, M. M.; Awaga, K.;
Robertson, N. Chem. Sci. 2011, 2, 316−320.
(17) Navor, N. L.; Robertson, N.; Weyland, T.; Kilburn, J. D.;
Underhill, A. E.; Webster, M.; Svenstrup, N.; Becher, J. J. Chem. Soc.,
Chem. Commun. 1996, 1363.
(18) Pei, W.-B.; Wu, J.-S.; Tian, Z.-F.; Ren, X.-M.; Song, Y. Inorg.
Chem. 2011, 50, 3970−3980.
(19) Mercuri, M. L.; Deplano, P.; Pilia, L.; Serpe, A.; Artizzu, F.
Coord. Chem. Rev. 2010, 254, 1419−1433.
(20) Cho, J.-Y.; Domercq, B.; Jones, S. C.; Yu, J.; Zhang, X.; An, Z.;
Bishop, M.; Barlow, S.; Marder, S. R.; Kippelen, B. J. Mater. Chem.
2007, 17, 2642−2647.
ASSOCIATED CONTENT
* Supporting Information
(21) Deplano, P.; Pilia, L.; Espa, D.; Mercuri, M. L.; Serpe, A. Coord.
Chem. Rev. 2010, 254, 1434−1447.
■
S
(22) Bao, Z.; Lovinger, A. J.; Dodabalapur, A. Appl. Phys. Lett. 1996,
69, 3066−3068.
Crystallographic data in CIF format. Further details are given in
Tables S1−S3 and Figures S1−S20. This material is available
free of charge via the Internet at http://pubs.acs.org.
(23) Noro, S.-i.; Chang, H.-C.; Takenobu, T.; Murayama, Y.;
Kanbara, T.; Aoyama, T.; Sassa, T.; Wada, T.; Tanaka, D.; Kitagawa,
S.; Iwasa, Y.; Akutagawa, T.; Nakamura, T. J. Am. Chem. Soc. 2005,
127, 10012−10013.
AUTHOR INFORMATION
Corresponding Author
*Fax: +44 (0)131 650 4743. Phone: +44 (0)131 650 4755. E-
mail: neil.robertson@ed.ac.uk.
■
(24) Smith, A. G.; Tasker, P. A.; White, D. J. Coord. Chem. Rev. 2003,
241, 61−85.
(25) Tasker, P.; Gasperov, V. Macrocyclic Chem. 2005, 365−382.
(26) Szymanowski, J. Hydroxyoximes and Copper Hydrometallurgy;
CRC Press: Boca Raton, FL, 1993.
(27) WIN-EPR Simfonia, version 1.25; Bruker Analytische
Messtechnik GmbH: Rheinstetten, Germany, 1996.
(28) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
(29) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789.
(30) Ditchfield, R.; Hehre, W. J.; Pople, J. A. J. Chem. Phys. 1971, 54,
724−728.
(31) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; J. A. Montgomery, J.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J.
J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; ;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; M. A. Al-Laham.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision
E.01; Gaussian Inc.: Wallingford, CT, 2004.
ACKNOWLEDGMENTS
■
We thank the EPSRC and the Japanese Science and
Technology agency (JST) for funding. This work has made
use of the resources provided by the EaStChem Research
Computing Facility (http://www.eastchem.ac.uk/rcf). This
facility is partially supported by the eDIKT initiative (heep://
www.edikt.org).
REFERENCES
■
(1) McKeown, N. B. Phthalocyanine Materials: Synthesis, Structure and
Function; Cambridge University Press: Cambridge, U.K., 1998.
(2) Dithiolene Chemistry: Synthesis, Properties, and Applications; Stiefel,
E. I., Ed.; John Wiley & Sons: Hoboken, NJ, 2004; Vol. 52, p 1- 738.
(3) Long, N. J. Metallocenes: An introduction to sandwich complexes.
Blackwell Science Ltd: Oxford, U.K., 1998.
(4) Wada, H.; Taguchi, T.; Noda, B.; Kambayashi, T.; Mori, T.;
Ishikawa, K.; Takezoe, H. Org. Electron. 2007, 8, 759−766.
(5) Shin-ichiro, N.; Taishi, T.; Yoshihiro, I.; Ho-Chol, C.; Susumu,
K.; Tomoyuki, A.; Takayoshi, N. Adv. Mater. 2008, 20, 3399−3403.
(6) Robertson, N.; Cronin, L. Coord. Chem. Rev. 2002, 227, 93−127.
(7) Bao, Z.; Lovinger, A. J.; Brown, J. J. Am. Chem. Soc. 1998, 120,
207−208.
(8) Uji, S.; Shinagawa, H.; Terashima, T.; Yakabe, T.; Terai, Y.;
Tokumoto, M.; Kobayashi, A.; Tanaka, H.; Kobayashi, H. Nature
2001, 410, 908−910.
(32) Thompson, M. A. ArgusLab 4.0.1; Planaria Software LLC:
Seattle, WA.
(33) Ramakrishnan, C.; Geetha, Y. J. Chem. Sci. 1990, 102, 481−496.
(34) Taylor, R.; Macrae, C. F. Acta Crystallogr., Sect. B 2001, 57,
815−827.
(9) Fujiwara, H.; Kobayashi, H.; Fujiwara, E.; Kobayashi, A. J. Am.
Chem. Soc. 2002, 124, 6816−6817.
(10) Kobayashi, H.; Sato, A.; Arai, E.; Akutsu, H.; Kobayashi, A.;
Cassoux, P. J. Am. Chem. Soc. 1997, 119, 12392−12393.
12875
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876

Inorganic Chemistry
Article
(35) Bruno, I. J.; Cole, J. C.; Edgington, P. R.; Kessler, M.; Macrae, C.
F.; McCabe, P.; Pearson, J.; Taylor, R. Acta Crystallogr., Sect. B 2002,
58, 389−397.
(36) Macrae, C. F.; Edgington, P. R.; McCabe, P.; Pidcock, E.;
Shields, G. P.; Taylor, R.; Towler, M.; van de Streek, J. J. Appl.
Crystallogr. 2006, 39, 453−457.
(37) Macrae, C. F.; Bruno, I. J.; Chisholm, J. A.; Edgington, P. R.;
McCabe, P.; Pidcock, E.; Rodriguez-Monge, L.; Taylor, R.; van de
Streek, J.; Wood, P. A. J. Appl. Crystallogr. 2008, 41, 466−470.
(38) Marom, N.; Hod, O.; Scuseria, G. E.; Kronik, L. J. Chem. Phys.
2008, 128, 164107.
(39) Aristov, V. Y.; Molodtsova, O. V.; Maslyuk, V. V.; Vyalikh, D. V.;
Zhilin, V. M.; Ossipyan, Y. A.; Bredow, T.; Mertig, I.; Knupfer, M. J.
Chem. Phys. 2008, 128, 034703.
(40) King, E. R.; Betley, T. A. J. Am. Chem. Soc. 2009, 131, 14374−
14380.
(41) Westcott, B. L.; Gruhn, N. E.; Michelsen, L. J.; Lichtenberger,
D. L. J. Am. Chem. Soc. 2000, 122, 8083−8084.
(42) Basu, P. J. Chem. Educ. 2001, 78, 666.
(43) Calle, C.; Schweiger, A.; Mitrikas, G. Inorg. Chem. 2007, 46,
1847−1855.
(44) Marom, N.; Kronik, L. Appl. Phys. A: Mater. Sci. Process. 2009,
95, 159−163.
(45) Carlin, R. L. Magnetochemistry; Springer-Verlag : Berlin,
Germany, 1986.
(46) Bonner, J. C.; Fisher, M. E. Phys. Rev. 1964, 135, A640.
(47) Estes, W. E.; Gavel, D. P.; Hatfield, W. E.; Hodgson, D. J. Inorg.
Chem. 1978, 17, 1415−1421.
(48) Sung Hoon, K.; Sun Hee, L.; Jin, J. IEEE Electron Device Lett.
2010, 31, 1044−1046.
(49) Tello, M.; Chiesa, M.; Duffy, C. M.; Sirringhaus, H. Adv. Funct.
Mater. 2008, 18, 3907−3913.
12876
dx.doi.org/10.1021/ic2020644 | Inorg. Chem. 2011, 50, 12867−12876