The trivalent copper complex of a conjugated bis-dithiocarbazate schiff base: Stabilization of Cu in three different oxidation states

Article abstract of DOI:10.1021/ic302596h

The new tribasic N₂S₂ ligand H₃ttfasbz has been synthesized by condensation of 4-thenoyl 2,2,2-trifluoroacetone and S-benzyl dithiocarbazate. On complexation with copper(II) acetate, spontaneous oxidation to the Cu^III oxidation state is observed, and the complex [Cu(ttfasbz)] has been isolated and characterized structurally. Reduction to the EPR active Cu^II analogue has been achieved chemically and also electrochemically, and in both cases, the process is totally reversible. The Cu^III/II redox potential of the complex is remarkably low and similar to that of the ferrocenium/ferrocene couple. Further reduction to the formally monovalent (d¹⁰) dianion [Cu^I(ttfasbz)]^2- may be achieved electrochemically. Computational chemistry demonstrates that the three redox states [Cu(ttfasbz)], [Cu(ttfasbz)]^-, and [Cu(ttfasbz)]^2- are truly Cu^III, Cu^II, and Cu^I complexes, respectively, and the potentially noninnocent ligand does not undergo any redox reactions.

Full text of DOI:10.1021/ic302596h

Article
pubs.acs.org/IC
The Trivalent Copper Complex of a Conjugated Bis-dithiocarbazate
Schiff Base: Stabilization of Cu in Three Different Oxidation States
Mohammad Akbar Ali,† Paul V. Bernhardt,*,‡ Mathilde A. H. Brax,‡ Jason England,§ Anthony J. Farlow,‡
Graeme R. Hanson,∥ Lee Len Yeng,† Aminul Huq Mirza,† and Karl Wieghardt§
†Chemistry Programme, Faculty of Science, Universiti Brunei Darussalam, BE 1410, Brunei Darussalam
‡School of Chemistry and Molecular Biosciences, The University of Queensland, Brisbane, 4072, Australia
§Max Planck Institute for Chemical Energy Conversion, Stiftstrasse 34-36, 45470 Mulheim an der Ruhr, Germany
̈
∥Centre for Advanced Imaging, The University of Queensland, Brisbane, 4072, Australia
S
* Supporting Information
ABSTRACT: The new tribasic N2S2 ligand H3ttfasbz has
been synthesized by condensation of 4-thenoyl 2,2,2-
trifluoroacetone and S-benzyl dithiocarbazate. On complex-
ation with copper(II) acetate, spontaneous oxidation to the
CuIII oxidation state is observed, and the complex [Cu-
(ttfasbz)] has been isolated and characterized structurally.
Reduction to the EPR active CuII analogue has been achieved
chemically and also electrochemically, and in both cases, the process is totally reversible. The CuIII/II redox potential of the
complex is remarkably low and similar to that of the ferrocenium/ferrocene couple. Further reduction to the formally
monovalent (d10) dianion [CuI(ttfasbz)]2− may be achieved electrochemically. Computational chemistry demonstrates that the
three redox states [Cu(ttfasbz)], [Cu(ttfasbz)]−, and [Cu(ttfasbz)]2− are truly CuIII, CuII, and CuI complexes, respectively, and
the potentially noninnocent ligand does not undergo any redox reactions.
Chart 1
INTRODUCTION
■
Although still an unusually high oxidation, a number of trivalent
copper complexes are now well characterized. Typically, two or
more ligand donor atoms in the CuIII complex are formally
anionic, which compensates for the high positive charge of the
metal. The types of ligands that are known to stabilize trivalent
copper include N-deprotonated oligopeptides,1−4 N-confused
porphyrins,5 corroles,6,7 oxamates,8 thiolates,9 and also various
N-deprotonated amido10 and mixed donor amido/thiolate
ligands.8,11−13 In all cases, the ligands are weak acids and are
deprotonated (as di-, tri-, or tetraanions) upon coordination. As
all of these ligands also possess unsaturated and delocalized
electronic structures, a clear definition of the metal oxidation
state is not straightforward. This is generally termed ligand
noninnocence,14,15 where redox reactions of the organic ligand
may be coupled with reactions on the metal center.
Computational chemistry has proven effective in defining the
exact nature of the electronic ground state and thus providing
an accurate picture of the d electronic configuration.15−17
A particularly unusual class of ligands is based on
acetylacetone (pentane-2,4-dione) bis-S-alkyl-isothiosemicarba-
zone (Chart 1). Ligands from this family, for example, the S-
methyl analogue acasme3−, bind as planar, trianionic,
tetradentate N-donors and have been found to stabilize CuIII
as well as FeIV complexes.11,18 However, the trianions (in their
coordinated form) were shown to be noninnocent and could
also exist in a two-electron oxidized (monoanionic) form based
on structural and spectroscopic evidence.19 A structurally
related compound is the S-benzyldithiocarbazate derivative of
the β-diketone 4-thenoyl-2,2,2-trifluoroacetone, which, in its
trianionic form (ttfasbz3−), shares many of the features of
acasme3− but bears an N2S2 set of donor atoms. Herein, we
report the Cu coordination chemistry of this new ligand.
EXPERIMENTAL SECTION
Syntheses. S-Benzyl dithiocarbazate was synthesized as previously
described.20 All other reagents were AR grade and were obtained
commercially.
■
H3ttfasbz. Solid 4-thenoyl-2,2,2-trifluoroacetone (0.64 g, 2.88
mmol) was added to a solution of S-benzyl dithiocarbazate (1.13 g,
5.70 mmol) in hot EtOH (25 mL). The mixture was refluxed for 24 h.
The mixture was cooled to room temperature, and an off-white
Received: November 27, 2012
Published: January 16, 2013
© 2013 American Chemical Society
1650
dx.doi.org/10.1021/ic302596h | Inorg. Chem. 2013, 52, 1650−1657

Inorganic Chemistry
Article
precipitate was filtered off and washed with a small volume of cold
EtOH (0.17 g, 10%); mp 180−184 °C. Anal. calcd for C24H21F3N4S5:
C, 49.47; H, 3.63; N, 9.61. Found: C, 49.5; H, 3.6; N, 9.6%. 1H NMR
(DMSO-d6) δ: 3.55 and 4.18 (AB q, 2H); 4.37 (q, 2H); 4.42 (q, 2H);
7.2−7.5 (m, 11H); 7.74 (d, 1H); 7.88 (d, 1H); 8.02 (s, 1H), 11.64 (s,
1H). 13C NMR (DMSO-d6) δ: 37.8, 38.3, 88.0 (q, CF3), 121.1, 124.9,
127.3, 127.4, 128.5, 128.6, 128.7, 129.2, 129.5, 131.4, 132.3, 133.6,
135.6, 136.7, 149.9, 194.2, 204.5.
Table 1. Crystal Data for [Cu(ttfasbz)]
formula
Fw
C24H18CuF3N4S5
643.26
cryst syst
triclinic
space group
a (Å)
b (Å)
c (Å)
α (°)
β (°)
γ (°)
V (Å3)
Z
P1 (no. 2)
̅
5.0358(3)
13.267(1)
20.367(1)
87.617(5)
85.132(4)
79.557(5)
1332.9(1)
2
[CuIII(ttfasbz)]. The ligand H3ttfasbz (50 mg; 120 μmol) was
dissolved in boiling abs. ethanol (15 mL) and combined with a hot
solution of copper(II) acetate hydrate (30 mg; 120 μmol) in EtOH
(50 mL). The mixture was heated on a water bath for ca. 10 min. The
reaction mixture was left to stand for about 3 days whereupon the
product that had formed was filtered off, washed with abs. ethanol, and
dried in a desiccator over anhydrous silica gel. Yield, 3.6 mg (4%).
Anal. calcd for C24H24CuF3N4O2.5N4S5: C, 41.82, H, 3.51; N, 8.12%.
Found: C, 41.5; H, 2.6; N, 8.0. IR (cm−1): 1662s, 1654s, 1560s, 1508s,
1414s, 1350s, 1314s, 1129s, 993s. Crystals suitable for X-ray work were
obtained by slow evaporation of a MeCN solution of the complex.
Physical Methods. Cyclic voltammetry was performed on a
BAS100B/W potentiostat employing a glassy carbon working
electrode, Pt auxiliary electrode, and Ag/AgCl reference electrode.
The supporting electrolyte was 0.1 NaNO3, and all solutions were
T (K)
293(2)
0.71073
1.254
λ (Å)
μ (cm−1
)
ρcalcd
1.603
a
R (obs data)
0.0377
b
wR2 (all data)
0.0601
a
b
2
2
2
R(Fo) = Σ||Fo| − |Fc||/Σ|Fo|. Rw(Fo ) = [Σ w(Fo − Fc2)/ΣwFo ]1/2
.
1
purged with Ar before measurement. IR spectra were measured. H
and 13C NMR spectra were recorded on a Varian Unity-300 (300
expand the electron density in the calculations were chosen to match
the orbital basis.35,36 The RIJCOSX37,38 approximation was used to
accelerate the calculations. The authenticity of each converged
structure was confirmed by the absence of imaginary vibrational
frequencies. Use of the Conductor-like Screening Model (COSMO),39
with DMF as the solvent, and the empirical dispersion forces
correction vdw0640 had little impact of the electronic structures of the
various calculated states relative to their gas phase counterparts.
Hence, all discussions herein are based on the latter.
The self-consistent field calculations were tightly converged (1 ×
10−8 Eh in energy, 1 × 10−7 Eh in the density change, and 1 × 10−7 in
maximum element of the DIIS error vector). The geometry
optimizations for all complexes were carried out in redundant internal
coordinates without imposing symmetry constraints. In all cases, the
geometries were considered converged after the energy change was
less than 5 × 10−6 Eh, the gradient norm and maximum gradient
element were smaller than 1 × 10−4 Eh Bohr−1 and 3 × 10−4 Eh
Bohr−1, respectively, and the root-mean square and maximum
displacements of all atoms were smaller than 2 × 10−3 and 4 × 10−3
Bohr, respectively.
In this paper, we describe our computational results by using the BS
approach pioneered by Noodleman et al.41 Given that several broken-
symmetry solutions to the spin-unrestricted Kohn−Sham equations
may be obtained, the general notation BS(m,n)42 has been adopted,
where m(n) denotes the number of spin-up (spin-down) electrons at
the two interacting fragments. Canonical and corresponding orbitals43
as well as spin density plots were generated with the program
Chimera.44
1
MHz H, 75 MHz 13C) spectrometer in DMSO-d6.
EPR Spectroscopy. Multifrequency (X- and S-band) electron
paramagnetic resonance (EPR) spectra were measured with a Bruker
Elexsys E500 EPR spectrometer equipped with an EIP microwave
frequency counter and a Bruker ER034 M Teslameter for calibration
of the microwave frequency and magnetic field, respectively. Low
temperatures were achieved at the sample position in the super high
Q-cavity using a Bruker nitrogen flow through system in conjunction
with a Eurotherm temperature controller. Room temperature X-band
EPR spectra employed the same cavity but utilized an AquaX cell to
minimize dielectric loss from the DMF solvent. At S-band frequencies,
normal 3 mm ID quartz EPR tubes were inserted into a flexline split
ring resonator. Samples for EPR measurements at both room and low
temperatures were prepared as 1 mM DMF solutions. Spin
Hamiltonian and line width parameters were determined by spectral
simulation with the program XSophe-Sophe-XeprView running on a
personal computer with Mandriva Linux 2010.2 as the operating
system.21 A Kivelson line width model22−25 (σ = α + βMI + γMI2
+
δMI3) was used for the simulation of the isotropic room temperature
spectra measured at both X-band (α = 9.04, β = 4.51, γ = 1.20, and δ =
−0.086) and S-band (α = 8.70, β = 2.51, γ = 2.23, and δ = −0.0364)
microwave frequencies. For the X-band anisotropic frozen solution
spectrum, a Hyde−Froncisz distribution of g and A values26,27 [σν
=
2
Σi=x,y,z (σRi2 + (σgi2/gi2 +σAiMI)2)] was used to simulate the variation
of line width (σRi = 1.24, 0.0, 2.31; σgi/gi = 0.000611, 0.00171,
0.00155; σAi = −2.033, 2.632, 0.240 × 10−4 cm−1; and i = x, y, z) as a
function of the nuclear spin quantum number (MI) and resonant
fields.
Crystallography. X-ray data were collected on an Oxford
Diffraction Gemini CCD diffractometer employing graphite-mono-
chromated Mo Kα radiation (0.71073 Å) and operating within the
range 2 < 2θ < 50 Å. Data reduction and empirical absorption
corrections (multiscan) were performed with Oxford Diffraction
CrysAlisPro software (Oxford Diffraction, vers. 171.36.21). The
structure was solved by direct methods with SHELXS and refined
by full-matrix least-squares analysis with SHELXL-97.28 All non-H
atoms were refined with anisotropic thermal parameters. The
molecular structure diagram was produced with ORTEP3.29 Crystal
data are given in Table 1.
DFT Calculations. All DFT calculations were performed using the
ORCA software package.30 The geometry optimizations of all
complexes and single-point calculations using the atomic coordinates
of the optimized geometries and X-ray crystal structures were carried
out using the B3LYP functional.31−33 In all calculations, def2-TZVP
basis sets were applied to all atoms.34 Auxiliary basis sets used to
RESULTS AND DISCUSSION
■
Ligand Synthesis. The Schiff base condensation reaction
(Scheme 1) between 2-thenoyltrifluoroactone (A, which exists
primarily as the enol tautomer in solution) and 2 equiv of S-
benzyl-dithiocarbazate does not lead to the open chain di-imine
C, but instead, the asymmetric pyrazoline (E) was obtained.
Intramolecular nucleophilic attack by the (deprotonated) NH
group occurs exclusively at the more electrophilic C-atom
adjacent to the −CF3 group (D, in the E-isomeric form). The
yield of this product is rather poor relative to similar
condensation reactions45−47 despite many attempts using
alternative solvents and reaction conditions. The monoimine
(B), which exists as the keto tautomer in DMSO and in the
solid state,48 was commonly a major product. The progress of
the reaction was followed by TLC, and a mixture of compounds
1651
dx.doi.org/10.1021/ic302596h | Inorg. Chem. 2013, 52, 1650−1657

Inorganic Chemistry
Article
Scheme 1. Formation of the Cyclic Product H3ttfasbz
Figure 1. ORTEP molecular structure of [CuIII(ttfasbz)]. Selected
bond lengths (Å) and angles (°): N2−Cu, 1.897(2); N3−Cu,
1.909(2); S1−Cu, 2.1802(9); S3−Cu, 2.1769(9); N2−Cu−N3,
96.7(1); N2−Cu−S3, 166.41(8); N2−Cu−S1, 87.75(8); and S3−
Cu−S1, 89.49(3).
Figure 3. X-band (9.449 GHz) frozen solution (1 mM in DMF) EPR
spectra (130 K) of (a) [CuIII(ttfasbz)] as prepared (partically reduced
by the solvent), (b) [CuIII(ttfasbz)] reduced with cobaltocene to its
monoanion [CuII(ttfasbz)]−, and (c) fully reoxidized with FeCl3
(bottom). The resonances labeled with a “*” arise from MnII
impurities in the FeCl3 oxidant.
was present in solution even after 24 h of reflux. The pyrazoline
H3ttfasbz precipitates from EtOH at room temperature,
enabling its separation from the other components.
1H and 13C NMR clearly discriminate the cyclic pyrazoline E
from its open chain isomer C. The methylene protons in
isomer C are necessarily equivalent due to the planar symmetry
1
of this molecule, so yield a single H NMR resonance. In the
pyrazoline E, the rigidity of the five-membered ring results in an
AB quartet from the now inequivalent methylene protons. This
is evident in the 1H NMR spectrum (AB doublets at 3.55, 4.18
ppm). The remaining peaks are due to the nonequivalent S-
benzyl methylenes and the three thiophene protons. A similar
cyclization reaction has been reported for analogous S-methyl
dithiocarbazate in reaction with other β-diketones.45,47
Copper Complexation. Reaction of H3ttfasbz with
(divalent) [Cu(OAc)2] in EtOH led to the neutral complex
[Cu(ttfasbz)]. Given that the ligand is a tribasic acid, this
formally corresponds to trivalent copper. Crystals suitable for
X-ray studies were obtained, and the crystal structure of the
Figure 2. Cyclic voltammetry of [Cu(ttfasbz)] (1 mM in DMF and 0.1
M Et4NClO4). Working electrode glassy carbon; sweep rate 100 mV/s.
Upper trace confined to the window −600 < E < +1000 mV, while the
lower trace is within the window −2000 < E < +500 mV vs Fc+/0
.
Arrows indicate the initial direction of sweep.
1652
dx.doi.org/10.1021/ic302596h | Inorg. Chem. 2013, 52, 1650−1657

Inorganic Chemistry
Article
Chart 2
Figure 4. Experimental (a) and simulated (b) X-band (9.343 GHz)
EPR spectra (1 mM in DMF, 130 K) of [CuII(ttfasbz)]− prepared by
reduction of the parent CuIII complex with cobaltocene.
complex is shown in Figure 1. The coordination geometry is
distorted square planar with no other contacts to the metal ion
perpendicular to the CuN2S2 plane closer than 3.55 Å. The
Cu−N (∼1.90 Å) and Cu−S (∼2.18 Å) bonds in this structure
are significantly shorter than typical CuII N2S2 complexes of
related ligands (CuII−N ∼1.94−1.98 Å and CuII−S ∼2.24
Å)49−51 and are consistent with the few known comparable
CuIII complexes.11,12 These two features (square planar
geometry and short coordinate bonds) both suggest a low
spin d8 ground state akin to isoelectronic and more common
square planar NiII complexes. The three chelate rings defined
by the tetradentate coordinated ligand are close to planar. The
least-squares plane defined by the ligand backbone (comprising
the continuous chain of non-H atoms from S2 to S3 inclusive)
reveals a modest tetrahedral distortion of the CuN2S2 moiety
with trans pairs of donor atoms lying on the same side of the
best plane as each other: S1 (+0.326 Å) and N3 (+0.121 Å)
and S3 (−0.272 Å) and N2 (−0.215 Å). The two benzyl groups
adopt different orientations (syn/anti) with respect to their
adjacent coordinated S-donor.
Figure 5. Experimental and simulated EPR spectra of [CuII(ttfasbz)]−.
(a) X-band (ν = 9.449 GHz) spectra (1 mM in DMF, 298 K) of
[CuII(ttfasbz)]− prepared by reduction of the parent CuIII complex
with cobaltocene. Note the presence of a three line signal (*) due to a
transient organic radical. (b) Composite computer simulation of the
EPR spectra of [CuII(ttfasbz)]− plus the radical in a ratio of 1
(radical):1.8 (CuII).
point) were unambiguously defined as the two noncoordinated
N-atoms (N1 and N4) and the apical methylene bridging the
two imine functional groups. Deprotonation of both thioamide
N-atoms (N1 and N4) is typical in dithiocarbazate Schiff base
coordination chemistry as well as related thiosemicarbazone
Most importantly, all H-atoms were clearly resolved in
difference electron density maps, and the three sites of
deprotonation (taking structure C in Scheme 1 as a starting
Table 2. Spin Hamiltonian Parameters for [CuII(ttfasbz)] and Related Compounds in Chart 2 (ND, Not Determined)
[CuII(ttfasbz)]−
[CuII(phmi)]2‑
[CuII(cdMenappd)]
[CuII(salen)]
gx
2.0261
2.0263
2.1056
37.32
2.028
2.035
2.119
36.8
2.036
2.053
2.168
ND
2.054
2.054
2.212
30.7
gy
gz
Ax (63Cu) (10−4cm−1
)
Ay (63Cu) (10−4cm−1
)
27.90
35.2
25.0
30.7
Az (63Cu) (10−4cm−1
Ax (14N) (10−4cm−1
)
186.53
15.43
195.0
12.6
189.0
ND
203.7
ND
)
Ay (14N) (10−4cm−1
Az (14N) (10−4cm−1
giso
Aiso (63Cu) (10−4cm−1
Aiso (14N) (10−4cm−1
)
)
16.54
15.1
12.7
ND
14.17
12.3
ND
ND
2.055
2.06
2.084
77.0
2.0999
86.0
)
84.11
88.5
)
14.70
12.6
12.0
ND
this work
ref 12
ref 53
ref 54
1653
dx.doi.org/10.1021/ic302596h | Inorg. Chem. 2013, 52, 1650−1657

Inorganic Chemistry
Article
Electrochemistry. Cyclic voltammetry of [CuIII(ttfasbz)]
revealed a variety of metal- and ligand-centered redox
processes. Two quasi-reversible single-electron reductions
were found at −126 mV vs Fc+/0 and −1669 mV vs Fc+/0
(Figure 2) when cycling within the potential window −2000 <
E < +500 mV vs Fc+/0. These are assigned to the
[CuIII/II(ttfasbz)]0/− and [CuII/I(ttfasbz)]−/2− couples on the
basis of their potentials and stoichiometry. On sweeping to
higher potentials, a multielectron irreversible anodic wave due
to thiophene oxidation was seen. The multielectron nature of
this process was not investigated further, but electropolymeri-
zation of thiophene monomers at similar potentials is a facile
reaction, and similar chemistry is likely here. On the reverse
sweep, an additional Cu-based response emerges at slightly
lower potential (Figure 2, upper trace), which we tentatively
assign to oligomeric complexes bridged by thiophene residues.
These results demonstrate that [CuIII(ttfasbz)] exhibits a
CuIII/II redox potential similar to the ferrocenium/ferrocene
couple so the trivalent form of the complex is only a mild
oxidant. Cobaltocene under the same conditions yielded a
Table 3. Selected Bond Distances (Å) from the DFT
(B3LYP) Geometry Optimized Structures of [Cu(ttfasbz)],
[Cu(ttfasbz)]−, and [Cu(ttfasbz)]2−, Where S Is the Total
Electron Spin Quantum Number and the Terms UKS and
RKS Refer to the Use of Unrestricted and Restricted Kohn−
Sham Approaches, Respectively
[Cu(ttfasbz)]
[Cu(ttfasbz)]−
[Cu(ttfasbz)]2−
S
0
0.5
0
method
RKS
UKS
RKS
Cu−S(1)
Cu−S(3)
2.22200
2.22551
1.93612
1.95803
1.74203
1.29564
1.35162
1.34378
1.38520
1.41401
1.34471
1.34149
1.30275
1.73319
2.32066
2.29366
1.97010
2.00254
1.73524
1.28753
1.37017
1.33097
1.39018
1.41490
1.33365
1.36335
1.29437
1.73239
2.41480
2.36313
2.09817
2.15985
1.71822
1.29222
1.37403
1.33240
1.39878
1.42096
1.33117
1.35973
1.29791
1.71870
Cu−N(2)
Cu−N(3)
S(1)−C(1)
C(1)−N(1)
N(1)−N(2)
N(2)−C(3)
C(3)−C(5)
C(5)−C(6)
C(6)−N(3)
N(3)−N(4)
N(4)−C(7)
C(7)−S(3)
reversible [Co(cp)2]+/0 redox response at −1400 mV vs Fc+/0
.
This compound was thus an ideal reductant to generate the
monoanion [CuII(ttfasbz)]− by chemical reduction (see
below), while the monovalent complex was not accessible by
chemical reduction (CuII/I −1669 mV vs Fc+/0). For
comparison, the related Cu complexes of N2S2 bis-thiosemi-
carbazones of glyoxal (and analogs) reported by Donnelly et
al.52 exhibit CuII/I couples at higher potentials (approximately
in the range −1000 to −800 mV vs Fc+/0 in DMSO). So, the
extensive delocalization and extra negative charge of the ligand
ttfasbz3− stabilize the CuIII redox state by more than 600 mV at
complexes where a change in bond order of the −HN−CS
moiety of the free ligand to an ene-thiolate −NC−S− occurs
upon coordination. The loss of the proton from the thenoyl-
trifluoroacetone moiety is reminiscent of Schiff base complexes
of ligands derived from acetylacetone, such as (acacsme)3− in
Chart 1, where the central methylene group is acidic and takes
on partial aromatic character once deprotonated.11
Figure 6. Qualitative frontier molecular orbital diagrams displaying the HOMO and orbitals of high Cu-character for [Cu(ttfasbz)] (left) and
[Cu(ttfasbz)]− (right), plus a spin density plot and Mulliken spin population analysis for the latter. Note that the doubly occupied orbitals of
dominant Cu character have been localized.
1654
dx.doi.org/10.1021/ic302596h | Inorg. Chem. 2013, 52, 1650−1657

Inorganic Chemistry
Article
Scheme 2. Formal Oxidation States of [Cu(ttfasbz)] and Its One-Electron and Two-Electron Reduced Forms
the expense of the CuI form. Although the low potential CuII/I
couple is totally reversible, we did not attempt to generate this
monovalent complex either in bulk solution or in the solid
state. On the basis of its redox potential, which is more than
250 mV lower than cobaltocene, it would be very oxygen
sensitive.
an N2O2 coordination sphere such as [Cu(salen)].54 This
reflects a greater spin density on the ligating nitrogen nuclei as
a consequence of enhanced covalent bonding. The S-donor
atoms are evidently important in this greater degree of
coordinate bond covalency.
At room temperature, a simpler EPR spectrum was obtained
(Figure 5a) comprising four resonances (from isotropic
hyperfine coupling with the Cu nucleus) whose line widths
vary as a function of the nuclear spin quantum number, a
consequence of incomplete motional averaging of the g and A
matrices (see the Experimental Section). Superposed on each
of these resonances is a five-line isotropic pattern (relative
intensities 1:2:3:2:1) arising from hyperfine coupling to two
magnetically equivalent nitrogen nuclei coordinated to the CuII
ion. An additional three resonances with narrow linewidths
(highlighted with * symbols, Figure 5a) are present and
characteristic of an isolated organic radical bearing an unpaired
EPR Spectroscopy. A freshly prepared solution of
[CuIII(ttfasbz)] in DMF yielded a very weak frozen solution
(140 K) EPR spectrum (Figure 3a, blue trace). In principle, the
low spin trivalent (d8) complex should be diamagnetic, but
partial reduction by the solvent had occurred to generate a
small amount of the paramagnetic [CuII(ttfasbz)]−. This was
confirmed by addition of an equivalent amount of [Co(cp)2],
which resulted in significant enhancement of the CuII EPR
spectrum (Figure 3b, black trace). It should be noted that both
[Co(cp)2] and [Co(cp)2]+ are EPR silent. Quantitative
reoxidation of the sample was achieved by addition of excess
solid ferric chloride, which removed all CuII EPR signals
(Figure 3c, red trace). The only detectable EPR signals were
from trace MnII (S = 5/2, I = 5/2) impurities in the FeCl3
oxidant, and these are seen as a weak sextet arising from the I
1/2⟩ Kramers doublet.
electron coupled to a single N nucleus (giso = 2.0062, Aiso
=
13.00 × 10−4 cm−1). Addition of the simulated spectra of the
isolated radical and the CuII species in a ratio of 1 (radical):1.8
(Cu) yields a composite spectrum (Figure 5b) that is in very
close agreement to the experimental spectrum.
The presence of this organic radical is almost certainly
coupled with the redox reaction that led to the partially reduced
“as prepared” species in Figure 3. In the absence of any other
compounds, the radical must result from one-electron DMF
oxidation by the EPR silent trivalent [CuIII(ttfasbz)]. However,
the EPR spectrum of this species is not that of the DMF radical
cation, which is more complicated,55 but must be from a DMF
radical decomposition product. This radical decays over periods
of days at room temperature to leave only the signal of the CuII
complex (Figure S1 in the Supporting Information), so its
identity remains unknown. This is consistent with the (stable)
CuII EPR signal always being dominant of the radical, which
eventually decays completely (Figure S1 in the Supporting
Information).
Computational Studies. Although all evidence points to
the isolated neutral complex having a formal trivalent Cu
oxidation state, the ligand ttfasbz3− is potentially redox active
(noninnocent). As a consequence, an alternative formulation of
the neutral complex as containing a CuII ion coordinated to a
ligand radical dianion, which strongly antiferromagnetically
couple to one another to yield the observed S = 0 ground state,
is also feasible. In an effort to differentiate between these two
possible electronic structures, we undertook a DFT (B3LYP)
computational analysis of the closed-shell S = 0 (CuIII), open-
shell S = 0 (CuII, plus ligand radical dianion), and S = 1 states of
[Cu(ttfasbz)] using the RKS, BS(1,1), and UKS formalisms,
respectively. The structural parameters of the RKS geometry
optimized structure (Table 3), which is ca. 2.5 kcal lower in
energy than the corresponding S = 1 solution, display good
The anisotropic frozen solution EPR spectrum of
[CuII(ttfasbz)]− is consistent with a pseudosquare planar Cu
2
2
complex with a dx −y ground state. The spectrum is highly
resolved, revealing hyperfine coupling to both 63Cu and 65Cu
isotopes (Figure 4, low field “z” resonance) and superhyperfine
coupling to the two N-donor atoms (14N, I = 1, 99.632%; 15N, I
= 1/2, 0.368%), resulting in each Cu resonance being split into
five dominant resonances from the 14N nuclei.
Computer simulation of the anisotropic frozen solution EPR
spectrum of [CuII(ttfasbz)]− with a rhombic spin Hamiltonian
(eq 1)
H = β·B·g·S + ∑ (S·A·I − g β B·I)
n
n
63,65Cu
2
+ ∑ ∑ (S·Ai·I − g β B·I)
n
n
14,15N i=1
(1)
was undertaken with the XSophe-Sophe-XeprView computer
simulation suite21 to determine the spin Hamiltonian
parameters listed in Table 2. We employed the quadratic
version of the Hooke and Jeeves optimization algorithm to
optimize the spin Hamiltonian parameters through a
comparison of the experimental and simulated spectra, which
are shown in Figure 4. Interestingly, these spectra are highly
resolved, revealing the 14N superhyperfine coupling along the x,
y, and z principle axes, which is also apparent in the anisotropic
EPR spectra of [CuII(phmi)]2+ (N2S2, Chart 2)12 and
[Cu(cdMenappd)]53 (N2SO) but not CuII complexes having
1655
dx.doi.org/10.1021/ic302596h | Inorg. Chem. 2013, 52, 1650−1657

Inorganic Chemistry
Article
agreement with those of the X-ray crystal structure.
Interestingly, the orientations of the thiophene and benzyl
groups were well reproduced in the gas phase calculations,
indicating they are not merely a consequence of crystal packing
forces.
manner. The stabilization of Cu in three different oxidation
states is quite remarkable, and our computational and
spectroscopic results indicate that the three oxidation state
changes are truly metal centered, and this potentially
noninnocent ligand does not undergo any redox reactions.
The role of the −CF3 and thiophene substituents in stabilizing
this unusual oxidation state remains unclear, and studies on
analogues derived from acetylacetone are warranted.
Efforts to geometry optimize using the BS(1,1) formalism
yielded only RKS solutions. However, a BS(1,1) solution was
obtained by performing a single point calculation using the
atomic coordinates of the UKS S = 1 geometry optimized
structure, but this was even higher in energy (ca. 3.5 kcal above
the RKS solution). Furthermore, efforts to obtain a BS(1,1)
solution using the true atomic coordinates (the X-ray crystal
structure) were unsuccessful, and the singlet closed shell state
was in this case calculated to be more than 10 kcal lower in
energy than the triplet (S = 1) state. Hence, we can confidently
rule out an electronic structure comprising a CuII ion
coordinated to a ligand radical dianion and conclude that the
closed shell solution containing CuIII (see the four doubly
occupied d orbitals in Figure 6) is the electronic ground state.
In addition, we have calculated the geometry-optimized and
electronic structures of the mono- and dianions [Cu-
(ttfasbz)]1−/2−. As expected, the C−C, C−N, N−N, and C−
S bond distances for the respective S = 1/2 and S = 0 ground
states (broken symmetry solutions could not be found in either
case) are very similar to those determined and calculated for
the neutral species. In contrast, the corresponding Cu−N and
Cu−S bond lengths increase (Table 3) with increasing negative
charge of the four coordinate species. This is a clear indication
that reduction of the diamagnetic neutral complex to the mono-
and dianion (Scheme 2) is predominantly a metal-centered
process, so involve sequential reduction of CuIII to CuII and
then CuII to CuI. This is reflected in the qualitative frontier
molecular orbital schemes for [CuI I I (ttfasbz)],
[CuII(ttfasbz)]1−, and [CuI(ttfasbz)]2−. High levels of cova-
lency were seen in all frontier orbitals, and for the sake of
formal oxidation state assignment, it was necessary to localize
the occupied orbitals containing Cu d character. The latter
contains five (see the Supporting Information) and the former
two complexes (Figure 6) four doubly occupied orbitals with
predominant Cu character. In addition, the monanion has a
SOMO with only 52% Cu character, and a Mulliken spin
density population analysis shows that the unpaired electron is
evenly distributed over the metal ion and the ligand, both of
which reflect the highly covalent bonding situation in this
complex and also is consistent with the strong 14N super-
hyperfine coupling apparent in the EPR spectra (Figures 3−5)
between the N donor nuclei and the unpaired electron.
ASSOCIATED CONTENT
* Supporting Information
■
S
Figures of EPR spectra, DFT calculated spin density plots and
Mulliken populations, and qualitative frontier molecular orbital
diagram; tables of the energies of the calculated [Cu(ttfasbz)]n
species and DFT geometry optimized atomic xyz coordinates;
and .cif file. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: p.bernhardt@uq.edu.au.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge financial support from the
University of Queensland and the Max Planck Society for
granting a fellowship to J.E.
REFERENCES
■
(1) Diaddario, L. L.; Robinson, W. R.; Margerum, D. W. Inorg. Chem.
1983, 22, 1021−1025.
(2) McDonald, M. R.; Fredericks, F. C.; Margerum, D. W. Inorg.
Chem. 1997, 36, 3119−3124.
(3) Pratesi, A.; Zanello, P.; Giorgi, G.; Messori, L.; Laschi, F.; Casini,
A.; Corsini, M.; Gabbiani, C.; Orfei, M.; Rosani, C.; Ginanneschi, M.
Inorg. Chem. 2007, 46, 10038−10040.
(4) Bossu, F. P.; Chellappa, K. L.; Margerum, D. W. J. Am. Chem. Soc.
1977, 99, 2195−2203.
(5) Maeda, H.; Ishikawa, Y.; Matsuda, T.; Osuka, A.; Furuta, H. J.
Am. Chem. Soc. 2003, 125, 11822−11823.
(6) Will, S.; Lex, J.; Vogel, E.; Schmickler, H.; Gisselbrecht, J.-P.;
Haubtmann, C.; Bernard, M.; Gorss, M. Angew. Chem., Int. Ed. Engl.
1997, 36, 357−361.
(7) Bruckner, C.; Brinas, R. P.; Krause Bauer, J. A. Inorg. Chem. 2003,
̈
̃
42, 4495−4497.
(8) Cervera, B.; L. Sanz, J.; J. Ibanez, M.; Vila, G.; Lloret, F.; Julve,
M.; Ruiz, R.; Ottenwaelder, X.; Aukauloo, A.; Poussereau, S.; Journaux,
Y.; Carmen Munoz, M. J. Chem. Soc., Dalton Trans. 1998, 781−790.
CONCLUSIONS
■
(9) Krebs, C.; Glaser, T.; Bill, E.; Weyhermuller, T.; Meyer-Klaucke,
̈
The highly conjugated trianion (ttfasbz)3− when coordinated as
an N2S2 ligand stabilizes trivalent Cu. The CuIII/II redox
potential of the complex is remarkably low (slightly lower than
the ferrocenium/ferrocene couple), which has enabled its
isolation and structural characterization. In fact, no oxidant was
employed in the synthesis of [CuIII(ttfasbz)] from
[CuII(OAc)2], so dioxygen appears to be the oxidant.
Reduction of [CuIII(ttfasbz)] with cobaltocene generates
[CuII(ttfasbz)]−, which exhibits a characteristic axial EPR
spectrum of a d9 complex and further reduction to a formally
monovalent (d10) dianion [CuI(ttfasbz)]2−, may be achieved
electrochemically. The cyclization reaction leading to the
pyrazoline H3ttfasbz is evidently reversible as complexation
reactions starting with this cyclic compound progress in a facile
W.; Wieghardt, K. Angew. Chem., Int. Ed. 1999, 38, 359−361.
(10) Oliver, K. J.; Waters, T. N. J. Chem. Soc., Chem. Commun. 1982,
1111−1112.
(11) Knof, U.; Weyhermuller, T.; Wolter, T.; Wieghardt, K. J. Chem.
̈
Soc., Chem. Commun. 1993, 726−728.
(12) Hanss, J.; Kruger, H.-J. Angew. Chem., Int. Ed. Engl. 1996, 35,
̈
2827−2830.
(13) Desbenoit, N.; Galardon, E.; Roussel, P.; Artaud, I.; Tomas, A. J.
Coord. Chem. 2009, 62, 2472−2479.
(14) Butin, K. P.; Beloglazkina, E. K.; Zyk, N. V. Russ. Chem. Rev.
2005, 74, 531−553.
(15) Kaim, W. Inorg. Chem. 2011, 50, 9752−9765.
(16) Ghosh, M.; Weyhermuller, T.; Wieghardt, K. Dalton Trans.
̈
2010, 39, 1996−2007.
(17) Brown, S. N. Inorg. Chem. 2012, 51, 1251−1260.
1656
dx.doi.org/10.1021/ic302596h | Inorg. Chem. 2013, 52, 1650−1657

Inorganic Chemistry
Article
(18) Gerbeleu, N. V.; Simonov, Y. A.; Arion, V. B.; Leovac, V. M.;
Turta, K. I.; Indrichan, K. M.; Gradinaru, D. I.; Zavodnik, V. E.;
Malinovskii, T. I. Inorg. Chem. 1992, 31, 3264−3268.
(19) Knof, U.; Weyhermuller, T.; Wolter, T.; Wieghardt, K.; Bill, E.;
̈
Butzlaff, C.; Trautwein, A. X. Angew. Chem., Int. Ed. Engl. 1993, 32,
1635−1638.
(20) Akbar Ali, M.; Tarafdar, M. T. H. J. Inorg. Nucl. Chem. 1977, 39,
1785−1791.
(21) Hanson, G. R.; Gates, K. E.; Noble, C. J.; Griffin, M.; Mitchell,
A.; Benson, S. J. Inorg. Biochem. 2004, 98, 903−916.
(22) Kivelson, D. J. Chem. Phys. 1960, 33, 1094−1106.
(23) Atkins, P. W.; Kivelson, D. J. Chem. Phys. 1966, 44, 169−174.
(24) Wilson, R.; Kivelson, D. J. Chem. Phys. 1966, 44, 154−168.
(25) Wilson, R.; Kivelson, D. J. Chem. Phys. 1966, 44, 4440−4444.
(26) Froncisz, W.; Hyde, J. S. J. Chem. Phys. 1980, 73, 3123−3131.
(27) Hyde, J. S.; Froncisz, W. Annu. Rev. Biophys. Bioeng. 1982, 11,
391−417.
(28) Sheldrick, G. M. Acta Crystallogr., Sect. A 2008, A64, 112−122.
(29) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30, 565.
(30) Neese, F. ORCA, an Ab Initio, Density Functional and
Semiempirical Electronic Structure Package, vers. 2.8; Universitat
̈
Bonn: Bonn, Germany, 2010.
(31) Becke, A. D. J. Chem. Phys. 1986, 84, 4524−4529.
(32) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652.
(33) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. Rev. B: Condens. Matter
Mater. Phys. 1988, 37, 785−789.
(34) Weigend, F.; Alrichs, R. Phys. Chem. Chem. Phys. 2005, 7, 3297−
3305.
(35) Eichkorn, K.; Weigend, F.; Treutler, O.; Ahlrichs, R. Theor.
Chem. Acc. 1997, 97, 119−124.
̈
(36) Eichkorn, K.; Treutler, O.; Ohm, H.; Haser, M.; Ahlrichs, R.
̈
Chem. Phys. Lett. 1995, 240, 283−290.
(37) Kossmann, A. S.; Neese, F. Chem. Phys. Lett. 2009, 481, 240−
243.
(38) Neese, F.; Wennmohs, F.; Hansen, A.; Becker, U. Chem. Phys.
2009, 356, 98−109.
(39) Klamt, A.; Schuurmann, G. J. Chem. Soc. Perkin Trans. 2 1993,
̈
̈
799−805.
(40) Grimme, A. S. J. Comput. Chem. 2006, 27, 1787−1799.
(41) Noodleman, L.; Peng, C. Y.; Case, D. A.; Mouesca, J. M. Coord.
Chem. Rev. 1995, 144, 199−244.
(42) Kirchner, B.; Wennmohs, F.; Ye, S.; Neese, F. Curr. Opin. Chem.
Biol. 2007, 11, 134−141.
(43) Neese, F. J. Phys. Chem. Solids 2004, 65, 781−785.
(44) Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.;
Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. J. Comput. Chem. 2004,
25, 1605−12.
(45) Iskander, M. F.; El-Sayed, L.; El-Toukhy, A.; Tawflk, M.
Transition Met. Chem. 1982, 7, 135−140.
(46) Du Preez, J. G. H.; Gerber, T. I. A.; Knoesen, O. Inorg. Chim.
Acta 1987, 133, 3−5.
(47) Gerber, T. I. A.; Hosten, E.; Knoesen, O.; Mayer, P. J. Coord.
Chem. 2007, 60, 2369−2375.
(48) Akbar Ali, M.; Mirza, A.; Butcher, R.; Chowdhury, A. Transition
Met. Chem. 2011, 36, 471−479.
(49) Castineiras, A.; Bermejo, E.; West, D. X.; El-Sawaf, A. K.;
̃
Swearingen, J. K. Polyhedron 1998, 17, 2751−2757.
(50) Cowley, A. R.; Dilworth, J. R.; Donnelly, P. S.; Gee, A. D.;
Heslop, J. M. Dalton Trans. 2004, 2404−2412.
́
(51) Calatayud, D. G.; Lopez-Torres, E.; Mendiola, M. A.; Pastor, C.
J.; Procopio, J. R. Eur. J. Inorg. Chem. 2005, 2005, 4401−4409.
(52) Xiao, Z.; Donnelly, P. S.; Zimmermann, M.; Wedd, A. G. Inorg.
Chem. 2008, 47, 4338−4347.
(53) Pereira, E.; Gomes, L. R.; Low, J. N.; de Castro, B. Polyhedron
2008, 27, 335−343.
(54) Bhadbhade, M. M.; Srinivas, D. Inorg. Chem. 1993, 32, 6122−
6130.
(55) Eastland, G. W.; Rao, D. N. R.; Symons, M. C. R. J. Chem. Soc.,
Faraday Trans. 1986, 82, 2833−2842.
1657
dx.doi.org/10.1021/ic302596h | Inorg. Chem. 2013, 52, 1650−1657