Copper(II)-coordinated α-Azophenols: Effect of the metal-ion geometry on phenoxyl/phenolate oxidation potential and reactivity

Article abstract of DOI:10.1002/ejic.201402312

Two copper(II) complexes were synthesized from tridentate ligands involving a (N_quinoline,N_azo,O_phenol) donor set. The copper(II) ion is tetracoordinate with a chloride ion occupying the fourth position of the coordination sphere. Both X-ray diffraction and electron paramagnetic resonance (EPR) spectroscopy reveal that the copper(II) ion geometry is square planar in 1. In contrast, significant tetrahedral distortions are observed in 2, as a result of the steric clash between the hydrogen atoms of the methyl substituent of the quinoline group and the chloride ion. Cyclic voltammetry curves of 1 and 2 in CH₂Cl₂ display a reversible oxidation wave at E_1/2 = 0.59 V and 0.56 V versus ferrocenium/ferrocene, respectively, which was assigned to the phenoxyl/phenolate redox couple. Compounds 1⁺ and 2⁺ were generated and characterized by UV/Vis and EPR spectroscopy. Their reactivity with benzyl alcohol was investigated by kinetic measurements. Tetrahedral distortion at the metal ion in a copper(II) complex containing tridentate ligands with a (N_quinoline,N_azo,O_phenol) donor set is tuned by the α-quinoline substituent. The Cu^II-phenoxyl radical species oxidizes benzyl alcohol at rate constants that are essentially independent of the distortion at the metal center.

Full text of DOI:10.1002/ejic.201402312

SHORT COMMUNICATION
DOI:10.1002/ejic.201402312
Copper(II)-Coordinated α-Azophenols: Effect of the
Metal-Ion Geometry on Phenoxyl/Phenolate Oxidation
Potential and Reactivity
Amélie Kochem,*[a,b] Alexandre Carrillo,[b] Christian Philouze,[b]
Maurice van Gastel,[a] Amaury du Moulinet d’Hardemare,[b] and
Fabrice Thomas*[b]
Keywords: Metalloenzymes / Oxidation / Copper / Ligand design / Tridentate ligands / Radicals
Two copper(II) complexes were synthesized from tridentate
ligands involving a (Nquinoline,Nazo,Ophenol) donor set. The
copper(II) ion is tetracoordinate with a chloride ion occupy-
ing the fourth position of the coordination sphere. Both X-
ray diffraction and electron paramagnetic resonance (EPR)
spectroscopy reveal that the copper(II) ion geometry is
square planar in 1. In contrast, significant tetrahedral distor-
tions are observed in 2, as a result of the steric clash between
the hydrogen atoms of the methyl substituent of the quinol-
ine group and the chloride ion. Cyclic voltammetry curves
of 1 and 2 in CH2Cl2 display a reversible oxidation wave at
E1/2 0.59 V and 0.56 V versus ferrocenium/ferrocene,
respectively, which was assigned to the phenoxyl/phenolate
redox couple. Compounds 1+ and 2+ were generated and
characterized by UV/Vis and EPR spectroscopy. Their reac-
tivity with benzyl alcohol was investigated by kinetic mea-
surements.
=
with a value of 0.15 V versus the normal hydrogen electrode
(NHE) at pH 7.5,[3] whereas the Y272·/Y272 redox potential
ranges between 0.50 (pH 5.5) and 0.38 V (pH 8.5), with a
value of 0.40 V vs. NHE at pH 7.5.[4] This pH dependency
is ascribed to the protonation equilibrium of the axial
tyrosine Y495, which becomes a weaker donor at low pH
values.
Introduction
Galactose oxidase (GO) is a mononuclear copper en-
zyme that catalyzes the aerobic oxidation of a broad range
of primary alcohols. The X-ray crystal structure of inactive
GO (known as GOsemi) revealed an active site comprising a
square pyramidal copper(II) ion coordinated to two
histidine (H496 and H581) and two tyrosine (Y495 and Y272
)
The influence of both the electronics and coordination
geometries imposed by the ligand on the CuII/CuI redox
potential are widely understood in inorganic complexes and
can be easily extended to GO. Many efforts have been made
over the last decade to quantify the effect of the phenolate
substituent on the phenoxyl/phenolate redox potential in
coordination compounds,[5] which have provided a rational
residues from the peptide chain, as well as an exogenous
water (or acetate) molecule that replaces the substrate.[1]
During turnovers, GO shuttles between an oxidized
·
CuII–Y272 radical form (GOox) and a reduced CuI–Y272
form (GOred).[2] The former is potent for the two-electron
oxidation of alcohols, whereas the latter reduces dioxygen
into hydrogen peroxide. A third form, the inactive and crys-
tallized one, is GOsemi. It corresponds to the intermediate
redox state and contains the CuII–Y272 entity. The redox
potentials for the interconversion between these three forms
were shown to be pH dependent. The CuII/CuI redox poten-
tial ranges between 0.16 (pH 5.5) and 0.12 V (pH 10.5),
·
explanation for the very low Y272 /Y272 redox potential. In
comparison, much less is known about the influence of the
coordination geometry on the phenoxyl/phenolate oxi-
dation potential. In a seminal report, Shimazaki et al. inves-
tigated the oxidative chemistry of copper(II) tripodal li-
gands involving a N3O donor set (two pyridines, one ter-
tiary amine, and one phenolate donor).[6] Upon progressive
replacement of the pyridines by methylpyridines (which in-
creases the distortion at the metal center), they evidenced
an increase in the oxidation potential. In closely related
complexes, we observed a similar trend.[7] However, we no-
ticed that steric hindrance at the α position of the iminic
nitrogen atom affects the positioning of the phenolate,
which could occupy either the axial or equatorial site in the
square-pyramidal coordination polyhedron.[8] Given that
[a] Université Grenoble Alpes, Département de Chimie
Moléculaire, UMR CNRS 5250,
BP 53, 38041 Grenoble Cedex 9, France
E-mail: Fabrice.Thomas@ujf-grenoble.fr
http://dcm.ujf-grenoble.fr
[b] Max Planck Institute for Chemical Energy Conversion,
Stiftstrasse 34-36, 45470 Mülheim an der Ruhr, Germany
E-mail: amelie.kochem@cec.mpg.de
http://www.cec.mpg.de
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejic.201402312.
Eur. J. Inorg. Chem. 2014, 4263–4267
4263
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
SHORT COMMUNICATION
dramatic differences in Cu–O bond lengths and orbital is located 1.587 Å above this plane. The Cu–Cl and Cu–N1
overlaps exist between the two isomers, it is not possible to bonds are significantly longer in 2 than in 1: 2.263(2) versus
draw clear-cut conclusions about the detailed origin of the 2.240(2) Å for Cu–Cl, 2.030(3) versus 2.006(4) Å for Cu–
increase in phenolate oxidation potential.
N1. In addition, in 2 the N2–Cu–Cl and N1–Cu–O1 angles
We herein describe two copper complexes, 1 {[CuII(L1)- differ significantly from 180° (153 and 163°, respectively).
Cl]} and 2 {[CuII(L2)Cl]}, prepared from the tridentate Also noticeable is the steric clash between the hydrogen
ligands L1H and L2H involving a (Nquinoline,Nazo,Ophenol
)
atoms of the methyl substituent of the quinoline group and
coordination sphere (Scheme 1). In both ligands, the α-azo- the chloride ion that forces the copper and chloride ions
phenol donor is incorporated into a rigid ligand framework out of the aromatic ligand plane. The result is tetrahedral
to prevent any isomerization, whereas the metal ion geom- distortion of the square planar copper center, while main-
etry is tuned by the hindrance provided by the α-substituent taining the Cu–O bond length. The final interesting feature
of the quinoline.
is the nonplanarity of the quinoline ring; its pyridinic part
is pushed away by the chloride ion.
Figure 1. Crystal structures of (a) 1 and (b) 2 shown with 30% ther-
mal ellipsoids. The numbering is similar in both (a) and (b) but is
indicated only for (a). Hydrogen atoms are omitted for clarity.
Scheme 1. Ligands and complexes.
The UV/Vis spectra of the complexes are characterized
by two intense absorption bands at 384 (12360 m–1 cm–1)
and 568 nm (12220 m–1 cm–1) for
1
and at 373
Results and Discussion
(13920 m–1 cm–1) and 575 nm (12390 m–1 cm–1) for 2. These
The new azo ligand L2H [2-tert-butyl-4-methoxy-6-(2- bands are comparable to those of the free L1H ligand and
methylquinolin-8-ylazo)phenol] was prepared by using a its zinc complex,[9] which shows that they involve predomi-
classical method consisting in the diazotization of the ap- nantly ligand-centered orbitals (see the Supporting Infor-
propriate 8-aminoquinoline with NaNO2 in concentrated mation for the time-dependent DFT assignment). The pow-
hydrochloric acid and by coupling the obtained diazonium der X-band electron paramagnetic resonance (EPR) spectra
salts with 2-tert-butyl-4-methoxyphenol in a way similar to of 1 and 2 are rather uninformative as a result of line broad-
that used to prepare L1H.[9] To lower the amount of by- ening, but measurements at the Q-band frequency show
product, we used fresh 8-amino-2-methylquinoline, which slightly rhombic (S = 1/2) spectra with resonances at g1 =
was readily synthesized by reduction of commercially avail- 2.193, g2 = 2.049, and g3 = 2.034 for 1 and at g1 = 2.208,
able 2-methyl-8-nitroquinoline with dihydrogen over g2 = 2.052, and g3 = 2.034 for 2 (Figure 2). The frozen-
Pearlman’s catalyst. Mixing equimolar equivalents of CuCl2 solution X-band EPR spectrum of 1 in CH2Cl2 remains
with L1H or L2H in the presence of triethylamine afforded broad (giso = 2.06), but it is partly resolved in the presence
complex 1 or 2; both were recrystallized by slow evapora- of tetra-n-butylammonium perchlorate (TBAP, 0.1 m) [g1 =
tion of methanol. Complex 1 crystallizes within the Pnma 2.193, g2 ≈ g3 = 2.042 (A1 = 19.4 mT, A2 = A3 = 3.0 mT for
space group, whereas 2 crystallizes in the less symmetrical one Cu nucleus)], whereas that of 2 displays satisfactorily
¯
P1 group.
resolved resonances at g1 = 2.213, g2 ≈ g3 = 2.050 (A1 =
The X-ray crystal structures of 1 and 2 are depicted in 15.8 mT, A2 = A3 = 3.0 mT for Cu) even in the absence
Figure 1. In both complexes, the copper(II) ion is tetracoor- of TBAP. The good agreement between the g-tensor values
dinate to one quinoline nitrogen atom (N1), one azo nitro- measured in the solid state and in CH2Cl2 solution confirms
gen atom (N2), one phenolato oxygen atom (O1), and a that the solvent does not affect the structure of the com-
chloride atom (Cl). Both the Cu–O1 and Cu–N2 bond plexes. Importantly, the higher g1 values obtained for 2 is
lengths are almost identical in 1 and 2 [1.900(3) and clear-cut evidence for substantial mixing of the dz2 orbital
1.896(3) Å for Cu–O1, 1.952(4) and 1.947(3) Å for Cu–N2]. into the dx2–y2, dxz, and dyz orbitals, in agreement with the
As anticipated, the copper(II) ion lies within a square- tetrahedral distortions present in its solid-state struc-
planar geometry for 1, with a mirror plane running through ture.[10,11] Accordingly, DFT calculations show that the
the aromatic ligand plane. In the case of 2, the metal ion metallic SOMO is a pure dx2–y2 orbital in 1 (the d orbitals
geometry is significantly distorted from square planar: the account for 58% of the SOMO, with 57% belonging to the
copper ion is displaced 0.277 Å above the mean plane de- dx2–y2 orbital), whereas the contribution of the metal to the
fined by the O1, N1, and N2 atoms, whereas the Cl atom SOMO of 2 is 56% and distributed throughout the dx2–y2
Eur. J. Inorg. Chem. 2014, 4263–4267
4264
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
SHORT COMMUNICATION
(46.9%), dz2 (0.3%), dxz (5.5%), and dyz (2.8%) orbitals. It ronment around the copper atom in 2, which both force a
is also significant that the calculations reproduce the shift twist between the quinoline and the phenolate rings and
in g1 from 1 to 2, with a calculated Δg1 of 0.012.
enlarge the coordination bonds on the quinoline side, does
not significantly influence the phenoxyl/phenolate redox
potential.
Cations 1+ and 2+ were generated chemically by addition
of one molar equivalent of the one-electron oxidant tris(p-
bromophenyl)aminium hexachloroantimonate at 233 K in
CH2Cl2. Oxidation was monitored by UV/Vis spectroscopy:
the main visible band is redshifted in the cation (relative to
the main visible band in the neutral precursor), as it is ob-
served at 572 (11100 m–1 cm–1) and 583 nm (11760 m–1 cm–1)
for 1+ and 2+, respectively. Shoulders are observed at
430 nm for 1+ and 2+ (7200 and 7000 m–1 cm–1, respectively)
and at 520 nm (9020 m–1 cm–1) for 2+. In addition, a broad
band appears at approximately 1000 nm in the spectra of
the cations. This evolution closely matches that observed
during oxidation of the zinc complex of L1H {[ZnII(L1)2]},
which produces a phenoxyl radical species,[9] and this sup-
ports the assignment of the reversible oxidation wave of 1
and 2 as the phenoxyl/phenolate redox couple. Time-
dependent DFT calculations, if considering a triplet state
for 1+ (see below), correctly predict a band at 904 nm (ƒ =
0.05), which is assigned to an azoquinolineǞphenoxyl
ligand-to-ligand charge transfer (LLCT) transition, and an
intense band at 578 nm (ƒ = 0.123), which is assigned to a
charge-transfer (CT) transition involving the azophenoxyl
moiety. The analogous LLCT transition is predicted for the
broken symmetry singlet 2+ at 886 nm (ƒ = 0.05), as well as
CT transitions at higher energies. Consistent with the
UV/Vis data, EPR spectroscopy shows that the copper(II)
signal is quenched by the addition of one molar equivalent
of tris(p-bromophenyl)aminium to 1 and 2, as expected for
the formation of magnetically interacting copper(II) phen-
oxyl radical pairs.[12] This hypothesis is further confirmed
by the observation of a broad ΔMS = 2 resonance at around
g ≈ 4 for 1+ and 2+ in parallel mode at 4.5 K. Accurate
integration of this resonance was difficult owing to the line-
width; therefore, we refrained to fit the temperature depen-
dence of the intensity to extract J values. However, the fact
that the resonance becomes barely observable above 20 K
is consistent with a coupling that is either ferromagnetic of
very weakly antiferromagnetic. The electronic structure of
the cations was investigated by theoretical calculations. The
triplet, singlet diamagnetic, and broken symmetry spin
states were considered for each compound. In either case,
the CuIII–phenolate singlet spin state was the highest in en-
ergy, consistent with the experimental data. It is located
6.35 kcalmol–1 above the CuII–phenoxyl triplet, which cor-
responds to the ground spin state for 1+. The antiferromag-
netically exchange-coupled CuII–phenoxyl singlet spin state
is located 0.67 kcalmol–1 above the triplet, a situation that
resembles that observed in CuII–salen radical complexes.
The dx2–y2 and phenoxyl π orbitals in 1+ are orthogonal to
each other; therefore, a direct overlap between the ligand-
centered and metal-centered SOMOs is not possible and the
coupling is ferromagnetic. The absence of the ΔMS = 1
transition in perpendicular mode for 1+ likely results from
Figure 2. Q-Band EPR spectra of powdered samples of (a) 1 and
(b) 2 (solid lines) and simulations (dotted lines) by using the param-
eters given in the text. Microwave frequency: 33.9 GHz, power:
0.5 mW, mod. frequency: 100 kHz, amp.: 0.4 mT, T = 298 K.
The cyclic voltammetry (CV) curves of 1 and 2 in
CH2Cl2 (Figure 3) display reversible oxidation waves at
E1/2 = 0.59 and 0.56 V versus ferrocenium/ferrocene
(Fc+/Fc), respectively, which are assigned to the phenoxyl/
phenolate redox couple (see below). An irreversible system
a
is observed at higher potentials (Ep = 1.00 and 1.02 V for
1 and 2, respectively), which has not been further investi-
gated. In previous series of copper(II) complexes synthe-
sized from the tripodal ligand, the incorporation of a
methyl substituent at the α position of the pyridine results
in an increase in the phenoxyl/phenolate redox potential by
about 0.07 V.[6,7] It is remarkable that the phenoxyl/phenol-
ate redox couple herein differs by only 0.03 V (2.9 kJmol–1)
and that it is lower for 2 than for 1. To assign this slight
difference, we recorded the CV curve of the free L1H and
L2H ligands in the presence of nBu4N+OH–. Under such
conditions, the phenol unit is deprotonated [to afford (L1)–
and (L2)–], and the redox process corresponds to a simple
electron transfer, as with the complex. The phenoxyl/phe-
a
nolate redox wave is irreversible and is observed at Ep
=
–0.50 and –0.52 V versus Fc+/Fc for (L1)– and (L2)–, respec-
tively. This difference between (L1)– and (L2)– is very similar
to that observed for the complexes. Thus, the distorted envi-
Figure 3. Cyclic voltammetry curves of (a) 1 and (b) 2 in 0.5 mm
CH2Cl2 solutions (+ 0.1 m TBAP). Scan rate = 0.1 Vs–1, T = 298 K.
The potentials are referenced to the Fc+/Fc couple. Inset: Spin-
density plot of 1+ (triplet state).
Eur. J. Inorg. Chem. 2014, 4263–4267
4265
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
SHORT COMMUNICATION
both zero-field spitting parameters that are larger than the was proposed by Fukuzumi et al. to explain benzyl alcohol
energy quantum of the EPR[12] and a large linewidth of the oxidation by the zinc(II)–phenoxyl complex of a tripodal
resonance. The CuIII–phenolate singlet remains the highest ligand.[17] They were also identified by Wieghardt et al. over
energy spin state for 2+. Although the energy difference is the course of benzyl alcohol oxidation by copper(II) com-
rather small between the triplet and broken symmetry plexes involving planar tridentate iminosemiquinonate radi-
singlet (0.17 kcalmol–1, within the error of the calculation), cal ligands.[18] Thus, the mechanism for benzyl alcohol oxi-
the calculated ground spin state is the antiferromagnetically dation by 1+ and 2+ clearly differs from that reported for
exchange-coupled CuII–phenoxyl singlet. The tetrahedral copper(II)–phenoxyl complexes involving tetradentate salcn
distortion at the copper center indeed favors mixing of the ligands[14,15] and rather compares with complexes involving
dx2–y2 orbital with the dxz and dyz orbitals. This allows for tridentate iminosemiquinonate ligands.
a non-negligible overlap between the copper and the radical
π orbital and subsequent antiferromagnetic interactions.
Not surprisingly, the geometry-optimized structures of the
CuII–phenoxyl complexes display a quinoidal distribution
of the bond lengths within the phenoxyl ring, with fairly
similar C–C and C–O bond lengths in 1+ and 2+ (see the
Supporting Information). Comparison of the bond lengths
within the ligand framework before and after oxidation
shows that the Cu–O bond length increases by 0.022 and
by 0.015 Å upon oxidation (for 1+ and 2+, respectively),
which is consistent with the phenoxyl radical becoming a
weaker donor than the phenolate ion. In contrast, the for-
mation of the phenoxyl radical is accompanied by a de-
crease in the Cu–Cl bond length (by 0.034 and 0.036 Å for
1+ and 2+, respectively).
Figure 4. Plot of kobs versus the benzyl alcohol concentration for
the oxidation of benzyl alcohol by 1+ (open circles) and 2+ (filled
The chemically generated radical complexes exhibit mod-
squares). [1+] = [2+] = 7ϫ10–5 m, T = 298 K.
erate stability with similar decomposition rate constants
[kdec = 0.033(3) and 0.030(3) min–1 for 1+ and 2+, respec-
tively] at 298 K in CH2Cl2. Addition of benzyl alcohol
(10% final concentration) to CH2Cl2 solutions of 1+ and 2+
Conclusions
induces a fast recovery of the copper(II) EPR signal, which
We have shown in this work that the methyl substituent
is identical to that of the neutral complexes in the same
included in the rigid conjugated ligand framework of L1H
or L2H induces a twist between the quinoline and phenolate
rings in addition to tetrahedral distortions at the copper
center. Such distortions neither influence the oxidation
potential nor the chemical stability of the radical species.
In contrast, the magnetic interaction between the spins of
the copper(II) center and the ligand radical is significantly
affected. This suggests that the strength of the metal–
medium.[13] Thus, the radical complexes oxidize benzyl
alcohol to regenerate 1 and 2. It has been shown that the
copper radical species [CuII(salcn)]·+ (salcn = N,NЈ-bis(3,5-
di-tert-butylsalicylidene)-1,2-cyclohexanediamine)
also
oxidizes benzyl alcohol, but it evolves toward its proton-
ated – not its initial bis(phenolate) – form over the course
of the reaction.[14] This implies that protons are released in
solution in the cases of 1+ and 2+, likely because of the
lower pKa of the phenolate groups embedded in the conju-
gated ligand framework. Such behavior has been recently
reported for some copper radical salophen complexes.[15]
GC analysis of the oxidation products at the plateau re-
vealed the formation of 0.8Ϯ0.2 equiv. of benzaldehyde per
copper, whereas we did not detect any diol that may have
resulted from oxidative C–C coupling of transient ketyl rad-
ical intermediates.[16] It is striking that the rate constants
kobs do not show a linear dependence with benzyl alcohol
concentration, as reported for [CuII(salcn)]·+,[14] but a para-
bolic shape (Figure 4). In addition, the plots of kobs as a
function of [PhCH2OH] are similar for 1+ and 2+; this con-
firms that the distortion present in 2+ is not a crucial factor
O
phenolate bond is the main determinant in the redox poten-
tial.[19] This also shows that the approach of the substrate
is not perturbed by distortion (at least at this extent) in the
present ligand framework. In addition, the denticity of the
ligand appears to be a major determinant in the mechanism
for benzyl alcohol oxidation by 1+ and 2+ versus that by
tetradentate salcn ligands. This implies the formation of a
dimer as a key intermediate, presumably favored by the
lower denticity of L1H and L2H.
Experimental Section
General: Full experimental details for the synthesis of L1H, L2H,
for the kinetics of benzyl alcohol oxidation. The square de- and their copper complexes are described in the Supporting Infor-
mation.
pendence of the rate constant kobs with benzyl alcohol con-
centration is evidence for the formation of a dimeric com-
plex (in which two benzyl alcohol molecules bind to the
L2H: 1H NMR (500 MHz, [D6]acetone): δ = 1.51 [s, 9 H, C(CH3)3],
2.86 (s, 3 H, CH3 quinoline), 3.90 (s, 3 H, O-CH3), 7.03–7.07 (2 d
metal) as the active species. The occurrence of such dimers AAЈ system, 4J = 3.10 Hz, 2 ϫ 1 H, CH phenolic), 7.60 (d, 3J =
Eur. J. Inorg. Chem. 2014, 4263–4267
4266
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org
SHORT COMMUNICATION
C. T. Lyons, E. C. Wasinger, T. D. P. Stack, J. Am. Chem. Soc.
2012, 134, 7367.
[6] Y. Shimazaki, S. Huth, S. Hirota, O. Yamauchi, Bull. Chem.
Soc. Jpn. 2000, 73, 1187.
8.50 Hz, 1 H, CH quinoline), 7.71 (t, J = 7.85 Hz, 1 H, CH quinoline),
3
7.99 (d, 3J = 8.05 Hz, 1 H, CH quinoline), 8.21 (d, J = 7.6 Hz, 1 H),
8.36 (d, 3J = 8.50 Hz, 1 H, CHquinoline), 15.87 (s, 1 H, O-Hphenol
)
ppm. MS (ESI): m/z = 350 (100) [M + H]+. C21H23N3O2 (349.4):
[7] N. Berthet, V. Martel-Frachet, F. Michel, C. Philouze, S. Ham-
man, X. Ronot, F. Thomas, Dalton Trans. 2013, 42, 8468.
[8] F. Michel, S. Hamman, C. Philouze, C. Perez Del Valle, E.
Saint-Aman, F. Thomas, Dalton Trans. 2009, 832.
[9] A. Kochem, M. Orio, C. Philouze, H. Jamet, A. du Mou-
linet d’Hardemare, F. Thomas, Eur. J. Inorg. Chem. 2011, 45.
[10] R. J. M. Klein Gebbink, M. Watanabe, R. C. Pratt, T. D. P.
Stack, Chem. Commun. 2003, 630.
[11] H. Yokoi, A. W. Addison, Inorg. Chem. 1977, 16, 1341.
[12] M. Orio, O. Jarjayes, H. Kanso, C. Philouze, F. Neese, F.
Thomas, Angew. Chem. Int. Ed. 2010, 49, 4989; Angew. Chem.
2010, 122, 5109.
calcd. C 72.18, H 6.63, N 12.03; found C 71.82, H 6.69, N 11.94.
1: MS (ESI/CH3CN): m/z = 397.2 [M – Cl]+, 829.1 [2M – Cl]+,
467.0 [M + Cl]– C20H20ClCuN3O2 (433.4): calcd. C 55.43, H 4.65,
N 9.70; found C 55.17, H 4.70, N 9.92.
2: MS (ESI/CH3CN): m/z = 411.1 [M – Cl]+, 857.1 [2M – Cl]+,
481.1 [M + Cl]–. C21H22ClCuN3O2 (447.4): calcd. C 56.37, H 4.96,
N 9.39; found C 56.22, H 5.02, N 9.48.
CCDC-975643 (for 1) and -975644 (for 2) contain the supplemen-
tary crystallographic data for this paper. These data can be ob-
tained free of charge from the Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[13] The distortion at the copper center is still present in the benzyl
alcohol adduct: The X-band EPR spectra of 1 and 2 in the
presence of 10% (in volume) benzyl alcohol were simulated by
Supporting Information (see footnote on the first page of this arti-
cle): Materials and methods, synthesis, crystal structures, EPR
spectra, and DFT and time-dependent DFT calculations.
ugust 6, 2014
using the following parameters: For 1: g1 = 2.213, g2 ≈ g3
=
2.033 (A1 = 17.7 mT, A2 = A3 = 2.8 mT for Cu and A2 = A3 =
1.7 mT and 0.7 for two distinct populations of nitrogen atoms);
for 2: g1 = 2.230, g2 ≈ g3 = 2.040 (A1 = 15.3 mT, A2 = A3
=
2.0 mT for Cu and A2 = A3 = 2.0 mT for two equivalents nitro-
gen atoms).
Acknowledgments
[14] K. Asami, K. Tsukidate, S. Iwatsuki, F. Tani, S. Karasawa, L.
Chiang, T. Storr, F. Thomas, Y. Shimazaki, Inorg. Chem. 2012,
51, 12450.
This work was supported by the Department of Molecular Chemis-
try (DCM) (doctoral grant to A. K.) and by Labex ARCANE
(grant number ANR-11-LABX-0003-01).
[15] K. Asami, A. Takashina, M. Kobayashi, S. Iwatsuki, T. Ya-
jima, A. Kochem, M. van Gastel, F. Tani, T. Kohzuma, F.
Thomas, Y. Shimazaki, Dalton Trans. 2014, 43, 2283.
[16] The radical species (1+ and 2+) are hardly regenerated, even not
in air, likely because of the high phenoxyl/phenolate oxidation
potential and acidification of the medium. This behavior con-
trasts that observed for galactose oxidase, but it has been re-
cently reported for copper(II)–salophen radical complexes (see
ref.[14]).
[17] S. Itoh, M. Taki, S. Takayama, S. Nagatomo, T. Kitagawa, N.
Sakurada, R. Arakawa, S. Fukuzumi, Angew. Chem. Int. Ed.
1999, 38, 2774; Angew. Chem. 1999, 111, 2944.
[18] P. Chaudhuri, M. Hess, U. Florke, K. Wieghardt, Angew.
Chem. Int. Ed. 1998, 37, 2217; Angew. Chem. 1998, 110, 2340.
[19] B. Adam, E. Bill, E. Bothe, B. Goerdt, G. Haselhorst, K.
Hildenbrand, A. Sokolowski, S. Steenken, T. Weyhermüller, K.
Wieghardt, Chem. Eur. J. 1997, 3, 308.
[1] N. Ito, S. E. V. Phillips, C. Stevens, Z. B. Ogel, M. J. McPher-
son, J. N. Keen, K. D. S. Yadav, P. J. Knowles, Nature 1991,
350, 87.
[2] J. W. Whittaker, Chem. Rev. 2003, 103, 2347.
[3] C. G. Saysell, T. Barna, C. D. Borman, A. J. Baron, M. J. Mc-
Pherson, A. G. Sykes, J. Biol. Inorg. Chem. 1997, 2, 702.
[4] C. G. Saysell, C. D. Borman, A. J. Baron, M. J. McPherson,
A. G. Sykes, Inorg. Chem. 1997, 36, 4520; P. M. Wood, Trends
Biochem. Sci. 1987, 12, 250.
[5] A. Philibert, F. Thomas, C. Philouze, S. Hamman, E. Saint-
Aman, J.-L. Pierre, Chem. Eur. J. 2003, 9, 3803; A. Kochem,
M. Orio, O. Jarjayes, F. Neese, F. Thomas, Chem. Commun.
2010, 46, 6765; L. Chiang, A. Kochem, O. Jarjayes, T. J. Dunn,
H. Vezin, M. Sakaguchi, T. Ogura, M. Orio, Y. Shimazaki, F.
Thomas, T. Storr, Chem. Eur. J. 2012, 18, 14117; R. C. Pratt,
Received: April 14, 2014
Published Online: August 6, 2014
Eur. J. Inorg. Chem. 2014, 4263–4267
4267
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim