A series of metal complexes with the non-innocent N,N′- bis(pentafluorophenyl)-o-phenylenediamido ligand: Twisted geometry for tuning the electronic structure

Article abstract of DOI:10.1039/b714974c

A series of homoleptic complexes with non-innocent ligands derived from N,N′-bis(pentafluorophenyl)-o-phenylenediamine (H₂ ^Fpda) are reported. [Ni^II(^Fsbqdi)₂] (1), [Pd^II(^Fsbqdi)₂] (2), [Co ^II(^Fsbqdi)₂] (3), and [Cu^II( ^Fsbqdi)₂] (4) were synthesized, where ( ^Fsbqdi)^1- represents a radical anion formed by one-electron oxidation of the doubly deprotonated H₂^Fpda. The oxidation states of ligands and metals in complexes 1-4 were assigned by single crystal X-ray crystallography performed at low temperatures. Complex 4 is the first Cu^II complex where both o-phenylenediamine derived ligands are monoanionic radicals. The bulky N-C₆F₅ substituents force the complexes 1, 3, and 4 to adopt a twisted geometry (intermediate between square-planar and tetrahedral). The electronic structures of the neutral compounds 1-4 and of some of their cationic and/or anionic neighboring redox states were probed using EPR and UV-VIS-NIR spectroelectrochemistry. The twisted geometry of the complexes results in considerable changes in their electronic structures compared to the well known square-planar complexes while the strongly electron withdrawing N-C₆F₅ groups have a great influence on redox properties. The Royal Society of Chemistry.

Full text of DOI:10.1039/b714974c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
A series of metal complexes with the non-innocent
N,Nꢀ-bis(pentafluorophenyl)-o-phenylenediamido ligand:
twisted geometry for tuning the electronic structure†‡
Marat M. Khusniyarov,a Klaus Harms,a Olaf Burghaus,a Jo¨rg Sundermeyer,*a Biprajit Sarkar,b
Wolfgang Kaim,b Joris van Slageren,c Carole Dubocd and Jan Fiedlere
Received 28th September 2007, Accepted 4th December 2007
First published as an Advance Article on the web 15th January 2008
DOI: 10.1039/b714974c
A series of homoleptic complexes with non-innocent ligands derived from
N,Nꢀ-bis(pentafluorophenyl)-o-phenylenediamine (H2 pda) are reported. [NiII(Fsbqdi)2] (1),
F
[PdII(Fsbqdi)2] (2), [CoII(Fsbqdi)2] (3), and [CuII(Fsbqdi)2] (4) were synthesized, where (Fsbqdi)1−
F
represents a radical anion formed by one-electron oxidation of the doubly deprotonated H2 pda. The
oxidation states of ligands and metals in complexes 1–4 were assigned by single crystal X-ray
crystallography performed at low temperatures. Complex 4 is the first CuII complex where both
o-phenylenediamine derived ligands are monoanionic radicals. The bulky N–C6F5 substituents force the
complexes 1, 3, and 4 to adopt a twisted geometry (intermediate between square-planar and
tetrahedral). The electronic structures of the neutral compounds 1–4 and of some of their cationic
and/or anionic neighboring redox states were probed using EPR and UV-VIS-NIR
spectroelectrochemistry. The twisted geometry of the complexes results in considerable changes in their
electronic structures compared to the well known square-planar complexes while the strongly electron
withdrawing N-C6F5 groups have a great influence on redox properties.
Introduction
ortho-Phenylenediamine, H2pda, and its derivatives are classical
examples of non-innocent N-donor ligands.1 They can serve as
monodentate or bidentate ligands2 and they can form complexes in
the doubly deprotonated dianionic form (pda)2−, which in turn can
be oxidized to o-benzosemiquinonediiminato radical monoanions
(sbqdi)1− and to neutral o-benzoquinonediimine (bqdi)0 ligands.
These different forms and oxidation states are summarized in
Scheme 1.
Complexes containing derivatives of H2pda reveal interesting
electronic properties and can undergo a series of electron transfer
reactions.3 Many of these electron transfer processes take place
not at the metal but at the ligand(s), demonstrating the non-
innocent nature of the latter.4 Recently such complexes received
considerable attention with respect to revision and reinterpretation
of their crystal5 and electronic6 structures, to the investigation
Scheme 1 Different forms and oxidation states of o-phenylenediamine
derived ligands.
of their reactivity,7 and to detailed theoretical studies aimed at
understanding the nature of these potential multispin systems.6,8
Determination of metal and ligand oxidation states in late
transition metal complexes with non-innocent ligands is difficult
when the oxidation state of the metal is highly variable. For
example, a neutral complex [Co(1L)2]9 may be formulated as
[CoII(1sbqdi)2] – CoII complex with two monoanionic radical
ligands, [CoIII(1sbqdi)(1pda)] – CoIII complex with one monoan-
ionic and one dianionic ligand, or even as a Co0, CoI or CoIV
complex with the appropriate set of oxidized or reduced ligands
(ligands abbreviated according to Fig. 1 and Scheme 1, “L” is
used when the oxidation state of the ligand is not assigned or
doubtful). Wieghardt et al.10 have shown that in such a case the
oxidation state of the metal and of the non-innocent ligand may
be assigned on the basis of single-crystal X-ray structure analyses,
because (pda)2−, (sbqdi)1−, and (bqdi)0 have distinctly different
aFachbereich Chemie der Philipps-Universita¨t, Hans-Meerwein-Straße,
D-35032, Marburg, Germany. E-mail: jsu@chemie.uni-marburg.de
bInstitut fu¨r Anorganische Chemie der Universita¨t Stuttgart, D-70550,
Stuttgart, Germany
c1. Physikalisches Institut der Universita¨t Stuttgart, D-70550, Stuttgart,
Germany
dGrenoble High Magnetic Field Laboratory, CNRS, 25 Avenue des Martyrs,
BP 166, F-38042, Grenoble Cedex 9, France
eJ. Heyrovsky-Institute of Physical Chemistry, CZ-18223, Prague, Czech
Republic
† CCDC reference numbers 662503–662506. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/b714974c
‡ Electronic supplementary information (ESI) available: Synthesis of
H2Fpda. See DOI: 10.1039/b714974c
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1355–1365 | 1355
©

View Article Online
Scheme 2 Synthesis strategy.
[CoIII(1sbqdi)2Cl],9 and [CoIII(5sbqdi)2I].6 The neutral (bqdi)0 is
well represented by X-ray crystallographic structure determination
of [FeII(1bqdi)(CN)4](Me2-DABCO),17 [CuI(6bqdi)(PPh3)2](BF4),18
[RuII(1bqdi)(PPh3)2(MeCN)2](PF6)2,19 [RuII(1bqdi)(PPh3)2Cl2],20
[RuII(1bqdi)(L)(H2O)](PF6)2,21 [RuII(1bqdi)I(L)](PF6)21 (L
=
1,4,7-trimethyl-1,4,7-triazacyclononane), and [RuII(5bqdi)(bpy)2]-
(PF6)2.14 These studies clearly demonstrated how intraligand
bond lengths vary in a systematic manner if (pda)2−, (sbqdi)1− and
neutral (bqdi)0 are compared.
In 1966 Balch and Holm3 reported the synthesis and properties
of a series of [M(1L)2] complexes, where M = Ni, Pd, Pt, Co. The
structures of the nickel22 and cobalt9 complexes were determined
later to show a planar geometry. Herebian et al.5 reported that
the diamagnetic planar complexes [M(4L)2] and [M(5L)2] (M =
Ni, Pd, Pt) are best described as containing a divalent d8 metal
ion and two antiferromagnetically coupled radical anions (singlet
diradical species). This interpretation was confirmed by density
functional theory (DFT) calculations.8 The electronic structure of
planar [CoL2] complexes is still a matter of debate with respect to
the presence of CoII versus CoIII.6 All structurally characterized
complexes of the general formula [ML2] were found to be
strictly planar, except for two copper complexes [CuI(7bqdi)2]Cl
and (HNEt3)2[Cu(8L)2], which adopt tetrahedral geometry, but
no data about their electronic structures were reported.23 Very
recently Wieghardt and coworkers reported tetrahedral complexes
with bulky o-phenylenediamine derived ligand.24 These complexes
adopt almost perfect tetrahedral geometry that change the spin-
state of the metal and in turn totally change the coupling pattern.
Herein we present a series of the complexes [MII(Fsbqdi)2],
where M is Ni, Pd, Co or Cu, based on the new N,Nꢀ-
Fig. 1 Ligands and complexes discussed in this paper.
bond distance patterns (Fig. 2). Since the relevant differences of
bond lengths are in the order of a few pm, accurate crystal structure
analyses at low temperatures are required to assign oxidation states
of the metal and of the ligand with sufficient confidence.
F
bis(pentafluorophenyl)-o-phenylenediamine (H2 pda, Fig. 1).
As a consequence of intramolecular interactions present in
[MII(Fsbqdi)2], the complexes have rarely observed twisted ge-
ometry, almost exactly intermediate between square-planar and
tetrahedral. An investigation of electronic structures of these
unusual twisted complexes by means of X-ray crystallogra-
phy, electrochemistry, magnetism and spectroscopy is reported
here.
Fig. 2 Characteristic bond distances of o-phenylenediamine-derived
ligands in different oxidation states: (Et4N)2[W0(CO)3(1pda)] (left),13
[NiII(5sbqdi)2] (center),5 [RuII(1bqdi)(PPh3)2(MeCN)2](PF6)2 (right).19
As shown in Scheme 1, the parent H2pda can be doubly depro-
tonated to form two different dianions: o-phenyleneaminoimido
(pai)2− and o-phenylenediamido (pda)2−. Monodentate (pai)2− has
been observed in [ReV(2pai)Cl3(PPh3)2]11 and bidentate (pda)2−
has been found in [TiIV(3pda)Cl(CpSiMe3)],12 [TiIV(3pda)ClCp],12
and (Et4N)2[W0(CO)3(1pda)]·{H2(1pda)}.13 The oxidation of
(pda)2− results in the formation of two new species: One-electron
oxidation of (pda)2− gives the open-shell o-benzosemiquinonedi-
iminato radical monoanion (sbqdi)1−, and two-electron oxidation
gives the closed-shell neutral o-benzoquinonediimine (bqdi)0.
The radical anion (sbqdi)1− can be found in [NiII(4sbqdi)2],5
[MII(5sbqdi)2]5,14 (M = Ni, Pd, Pt), [CoIII(1sbqdi)2(py)]-Cl,15
Experimental
The precursor N,Nꢀ-bis(pentafluorophenyl)-o-phenylenediamine
F
(H2 pda) was prepared from o-phenylenediamine and hexafluo-
robenzene according to a known procedure25 (included in ESI‡).
n-Butyllithium (1.6 M solution in n-hexane) was purchased from
27
Aldrich. The anhydrous salts NiCl2(DME),26 and PdCl2(SMe2)2
were prepared according to literature procedures. All syntheses
were performed under a dry, oxygen-free argon atmosphere using
standard Schlenk techniques and dried solvents.
[CoIII(1sbqdi)2(PPh3)](ClO4),16
[CoIII(5sbqdi)2(py)](OAc),14
1356 | Dalton Trans., 2008, 1355–1365
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Preparation of complexes
887 w, 928 m, 995 s, 1010 s, 1034 m, 1117 m, 1148 m, 1169 s, 1288 s,
1318 m, 1346 m, 1445 s, 1505 s, 1514 s, 1533 m, 1584 w, 1631 w,
1647 w, 3058 w, 3086 w; dH(200.1 MHz, CDCl3) 6.18–6.23 (4H, m),
6.71–6.76 (4H, m); dF(188.3 MHz, CDCl3) −162.5 (8F, pst, 3J =
[NiII(Fsbqdi)2] 1. H2 pda (2.64 g, 6.00 mmol) was dissolved
F
in diethyl ether (50 mL) and the resulting solution was cooled
to −50 ◦C. A solution of 1.6 M n-BuLi in n-hexane (7.5 mL,
12.00 mmol) was added slowly, dropwise at the end, with vigorous
stirring. After the mixture was stirred for 15 min, a suspension
of anhydrous NiCl2(DME) (0.66 g, 3.00 mmol) in diethyl ether
(15 mL) was added via a cannula. The yellow color of the reaction
mixture changed immediately to red. The mixture was allowed
to warm to room temperature and stirred overnight under an
argon atmosphere. The oxidation was accomplished by passing a
stream of pure oxygen through the solution for 5 min and stirring
the resulting dark-purple suspension for 30 min. Solvents were
removed in vacuo, the crude product was repeatedly extracted with
boiling n-hexane from a glass frit to give dark-purple microcrystals
in n-hexane solution. The Schlenk flask with crystals was cooled
to −80 ◦C, the n-hexane solution was decanted, the crystals were
washed with cold n-pentane and dried in vacuo (1.86 g, 66%).
The yield can be increased by adding 5 g of silica gel (Merck
60) to the residue suspended in toluene (50 mL), stirring for about
20 min, filtering, evaporating the toluene, and washing of the dark-
purple microcrystals with cold n-pentane. An additional amount
of 0.51 g analytically pure complex was thus obtained. Single
crystals of 1 suitable for X-ray crystallography were obtained from
a THF/n-hexane mixture (1 : 6) by cooling. Dark-purple crystals
(2.37 g, 85%); mp 241 ◦C (decomp.) (Found: C, 45.52; H, 0.71;
N, 6.83. Calc. for C36H8F20N4Ni: C, 46.24; H, 0.86; N, 5.99%);
mmax/cm−1 516 m, 600 m, 737 s, 771 m, 802 m, 831 w, 876 m, 1000 s,
1115 m, 1165 s, 1307 m, 1326 s, 1349 m, 1447 s, 1505 s, 1586 w,
1631 w, 1647 w, 3061 w; dH(200.1 MHz, CDCl3) 6.52–6.57 (4H, m),
8.12–8.20 (4H, m); dF(188.3 MHz, CDCl3) −162.7 (8F, pst, 3J =
3
3
20, Fmeta), −155.6 (4F, t, J = 20, Fpara), −144.1 (8F, d, J = 20,
Fortho); HRMS: m/z (EI) 981.9475 (M+. Calc. for C36H8F20N4Pd:
981.9464).
F
[CoII(Fsbqdi)2] 3. H2 pda (4.48 g, 10.18 mmol) was dissolved
in diethyl ether (60 mL) and the resulting solution was cooled
to −50 ◦C. A solution of 1.6 M n-BuLi in n-hexane (12.7 mL,
20.32 mmol) was added slowly, dropwise at the end, with vigorous
stirring. After the mixture was stirred for 15 min, a solution of
anhydrous CoCl2 (0.6 g, 4.62 mmol) in THF (100 mL) was added
through a cannula. The mixture was allowed to warm to room
temperature and stirred overnight under an argon atmosphere. A
color change from yellow to orange was observed. The oxidation
was accomplished by passing a stream of pure oxygen through the
solution for 5 min and stirring the resulting dark-green suspension
for 30 min. Solvents were removed in vacuo, the raw product was
repeatedly extracted with boiling n-hexane (100 mL) from a glass
frit to give dark microcrystals suspended in n-hexane solution.
The Schlenk flask with crystals was cooled to −80 ◦C, the n-
hexane solution was decanted, the crystals were washed with
cold n-hexane and dried in vacuo. Single crystals of 3 suitable
for X-ray crystallography were obtained from a THF/n-hexane
mixture (1 : 1) by cooling. Dark-green crystals (2.91 g, 67%); mp
191 ◦C (decomp.) (Found: C, 46.06; H, 0.80; N, 6.60. Calc. for
C36H8F20N4Co: C, 46.23; H, 0.86; N, 5.99%); mmax/cm−1 516 m,
575 w, 597 m, 638 w, 693 m, 725 s, 739 s, 772 m, 802 m, 834 w, 875
m, 995 s, 1007 s, 1116 m, 1141 m, 1165 m, 1278 m, 1341 m, 1507 s,
1587 w, 1629 w, 1646 w, 3063 w; HRMS: m/z (EI) 934.9771 (M+.
Calc. for C36H8F20N4Co: 934.9762).
3
3
20, Fmeta), −156.8 (4F, t, J = 20, Fpara), −146.6 (8F, d, J = 20,
Fortho); HRMS: m/z (EI) 933.9765 (M+. Calc. for C36H8F20N4Ni:
933.9783)
[CuII(Fsbqdi)2] 4. H2 pda (1.76 g, 4.00 mmol) was dissolved
F
in diethyl ether (40 mL) and the resulting solution was cooled
to −50 ◦C. A solution of 1.6 M n-BuLi in n-hexane (5.0 mL,
8.00 mmol) was added slowly, dropwise at the end, with vigorous
stirring. After the mixture was stirred for 15 min, a suspension of
anhydrous well-ground CuCl2 (0.27 g, 2.00 mmol) in THF (15 mL)
was added through a cannula. The mixture was allowed to warm to
room temperature and stirred for 2 d under an argon atmosphere.
The color of the reaction mixture changed gradually from yellow
to green. The oxidation was accomplished by passing a stream
of pure oxygen through the solution for 5 min and stirring the
resulting dark-blue suspension for 30 min. After solvents were
removed in vacuo, the mixture was extracted with toluene and the
extract was concentrated by a rotary evaporator. The raw product
was dissolved in wet ethyl acetate and the resulting solution
was quickly filtered through a layer of neutral Al2O3 (Fluka).
The filtrate was concentrated under reduced pressure to yield a
dark-blue solid. Recrystallization from a THF/n-hexane mixture
(1 : 1) gave a microcrystalline precipitate that was washed with
cold n-pentane and dried in vacuo. Single crystals of 4 suitable
for X-ray crystallography were obtained from a THF/n-hexane
mixture (2:1) by cooling. Dark-blue solid (0.80 g, 42%); mp
190 ◦C (decomp.) (Found: C, 45.71; H, 0.74; N, 6.32. Calc. for
C36H8F20N4Cu: C, 46.00; H, 0.86; N, 5.96%); mmax/cm−1 482 m, 591
m, 639 w, 736 s, 749 m, 803 m, 833 m, 856 m, 872 m, 995 s, 1008 s,
1025 m, 1119 m, 1141 m, 1165 m, 1298 m, 1315 m, 1353 m, 1420 m,
[PdII(Fsbqdi)2] 2. H2 pda (500 mg, 1.136 mmol) was dissolved
F
in diethyl ether (20 mL) and the resulting solution was cooled
to −50 ◦C. A solution of 1.6 M n-BuLi in n-hexane (1.4 mL,
2.240 mmol) was added slowly, dropwise at the end, with vigorous
stirring. After the mixture was stirred for 15 min, a suspension
of PdCl2(SMe2)2 (0.172 g, 0.570 mmol) in THF (10 mL) was
added through a cannula. The mixture was allowed to warm
to room temperature and stirred overnight under an argon
atmosphere. The oxidation was accomplished by passing a stream
of pure oxygen through the solution for 5 min and stirring
the resulting grey–blue suspension for 1 h. After solvents were
removed in vacuo, the mixture was extracted with toluene. The
toluene extract was filtered through a layer of SiO2 (Merck 60)
and concentrated by a rotary evaporator. The raw product was
recrystallized from a toluene/THF mixture (5 : 4) to give a sky-
blue precipitate that was washed with cold n-pentane and dried
in vacuo. When PdCl2(COD) or PdCl2(PhCN)2 was used instead
of PdCl2(SMe2)2 the yield was lower. Single crystals of 2 suitable
for X-ray crystallography were obtained from a THF/n-hexane
mixture (1 : 2) by cooling. Sky-blue solid (188 mg, 34%); mp 303–
306 ◦C (decomp.) (Found: C, 44.88; H, 0.86; N, 5.95. Calc. for
C36H8F20N4Pd: C, 43.99; H, 0.82; N, 5.70%); mmax/cm−1 498 m,
597 s, 643 w, 699 m, 713 w, 728 s, 738 m, 771 w, 800 m, 823 m,
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1355–1365 | 1357
©

View Article Online
1499 s, 1515 s, 1585 w, 1621 w, 1643 w; HRMS: m/z (EI) 938.9722
using a Nicolet 510 FT-IR spectrometer. NMR spectra were
recorded at room temperature on a Bruker ARX 200 spectrometer
at 200.1 MHz for 1H and at 188.3 MHz for 19F. 1H NMR
spectra were referenced (in ppm) to the residual proton signal
of CDCl3 (7.26 ppm), 19F NMR spectra were referenced to the
external standard CFCl3. CDCl3 was used as solvent in all cases.
Abbreviations used are: s = singlet, d = doublet, t = triplet,
pst = pseudo-triplet; J values are given in Hz. EPR spectra
in the X band were recorded with a Bruker System ESP 300
equipped with a Bruker ER035M gaussmeter and a HP 5350B
microwave counter. EPR spectra at 285 GHz were recorded using
a multifrequency spectrometer. A Gunn diode operating at 95 GHz
and equipped with a third harmonic generator has been used as
a radiation source. An InSb bolometer (QMC Instruments) was
used for detection. The main magnetic field was provided by a
superconducting magnet (Cryogenics Consultant) which generates
fields up to 12 T. Owing to different field sweep conditions, the
absolute values of the g components were obtained by calibrating
the precisely measured g anisotropy data with the isotropic g
value from X-band measurements. While this procedure does not
account for the temperature dependence of g, the values extracted
are identical with those obtained using an added standard. The
accuracy of the g values is estimated at 0.0003. UV-VIS-NIR
absorption spectra were recorded on J & M TIDAS and Bruins
Instruments Omega 10 spectrophotometers. Cyclic voltammetry
was carried out in 0.1 M Bu4NPF6 solutions using a three-electrode
configuration (glassy carbon working electrode, Pt counter elec-
trode, Ag/AgCl reference) and a PAR 273 potentiostat and
function generator. The ferrocene/ferrocenium (Fc/Fc+) couple
served as internal reference. A two-electrode capillary was used in
radical complex generation for X band EPR studies. The magnetic
susceptibility measurements were performed on powder samples
using a Quantum Design MPMS XL7 SQUID magnetometer.
The data were corrected for the diamagnetic contributions to
the molar magnetic susceptibility using Pascal’s constants and for
temperature independent paramagnetism (TIP).
(M+. Calc. for C36H8F20N4Cu: 938.9726).
X-Ray crystallographic data collection and refinement
of the structures
Dark single crystals of 1–4 were coated with inert oil, picked up
with a glass fiber and mounted in the cold nitrogen stream of
the STOE IPDS 2 diffractometer. Intensity data were collected
at 193 K using graphite monochromated Mo-Ka radiation (k =
˚
0.71073 A). Data collection was performed by hemisphere runs
taking frames at 1.0◦ in x. Final cell constants were obtained from
a least squares fit of a subset of several thousand strong reflections.
Intensity data of 2, 3, and 4 were corrected for absorption using
the semiempirical correction routine MulScanAbs,28 the data of 1
“analytical” using the indexed faces of the crystal.28 Disordered
THF is present in crystal structures of 1, 3, and 4 and has
been modeled with restraints for both geometry and anisotropic
displacement factors. They fill two independent “infinite” channels
parallel to the z axis. The refinement showed that one channel
contains 1, the other 1/2 THF with respect to one metal complex.
In the crystal structure of 2 the complex molecule adopts the
crystallographic site symmetry “222”. One molecule of disordered
THF is present in the asymmetric unit. All structures have
been solved by the direct methods in SHELXS-9729 (1–4) and
refined using the full matrix least squares refinement procedure of
SHELXL97,29 all non-hydrogen atoms anisotropically, H-atoms
have been placed at calculated positions and refined with Uiso(H) =
1.2 Ueq(C). Crystallographic data of the compounds are listed in
Table 1. Further information on structure refinement can be found
in ESI‡.
Physical measurements
High-resolution mass spectra were obtained on a Finnigan MAT
95S mass spectrometer (70 eV, EI). IR spectra were obtained
Table 1 Crystallographic data for 1·1.5(C4H8O), 2·4(C4H8O), 3·1.5(C4H8O), and 4·1.5(C4H8O)
1·1.5(C4H8O)
2·4(C4H8O)
3·1.5(C4H8O)
4·1.5(C4H8O)
Chem. formula
Crystal size/mm
Fw
C36H8F20N4Ni, 1.5(C4H8O)
0.22 × 0.15 × 0.14
1043.33
C52H40F20N4O4Pd
0.27 × 0.15 × 0.02
1271.28
Orthorhombic
Fddd (No. 70)
15.0923(14)
C36H8CoF20N4, 1.5(C4H8O)
0.34 × 0.18 × 0.07
1043.55
C36H8CuF20N4, 1.5(C4H8O)
0.14 × 0.11 × 0.05
1048.16
Crystal system
Space group
Monoclinic
C2/c (No. 15)
58.619(3)
15.3293(6)
8.6322(4)
97.288(3)
7694.1(6)
8
Monoclinic
C2/c (No. 15)
59.292(3)
15.3580(4)
8.6688(3)
97.952(9)
7818.0(5)
8
Monoclinic
C2/c (No. 15)
59.219(3)
15.3696(8)
8.5864(5)
97.394(4)
7750.1(8)
8
˚
a/A
˚
b/A
19.2020(14)
35.566(3)
˚
c/A
b/◦
3
˚
V/A
Z
10307.1(15)
8
193
1.638
4.83
T/K
193
1.801
6.47
193
1.773
5.79
193
1.797
7.05
q calc./g cm−3
l(Mo-Ka)/cm−1
Refl. collected/hmax
Unique refl./I > 2r(I)
No. params/restraints
R1a/goodness of fitb
wR2c (unique refl.)
42093/25
6746/4771
687/326
0.0590/1.031
0.1545
16959/25
2200/1643
204/7
0.0458/0.990
0.1064
48108/25
6890/4714
686/258
0.0521/0.976
0.1478
39236/25
6763/4464
685/262
0.0392/0.924
0.0979
ꢀ
ꢀ
ꢀ
ꢀ
2
2
a Observation criterion: I > 2r(I). R1 = ꢁFo| − |Fcꢁ/ |Fo|. b Goodness of fit (GoF) = [ {w(Fo − Fc2)2}/(n − p)]1/2
.
c wR2 = [ {w(Fo
−
ꢀ
2
2
2
Fc2)2}/ {w(Fo } ]1/2, where w = 1/r2(Fo2) + (aP)2 + bP and P = (Fo + 2Fc2)/3.
1358 | Dalton Trans., 2008, 1355–1365
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Results and discussion
Syntheses and characterization
The common synthesis of [ML2] complexes is performed by aero-
bic oxidation of the H2pda ligand in the presence of metal salt and
aqueous ammonia3 or triethylamine5 as bases. In our experiments,
the use of triethylamine for the synthesis of 1 led to low yields
and to the necessity to purify the product in the presence of
large amounts of unreacted ligand. Therefore, n-butyllithium was
employed as a base to give the best results. The general synthetic
procedure for complexes 1–4 involves a double deprotonation
of the ligand with n-butyllithium under anaerobic conditions,
the addition of the appropriate anhydrous metal salt, and slow
complexation, followed by oxidation with pure oxygen (Scheme 2).
Complexes 1 and 2 are diamagnetic at room temperature
showing sharp resonance signals in the proton and 19F NMR
spectra. As solids they are stable in contact with air, and toluene
solutions can be kept for at least 2 d without inert gas protection.
Complexes 3 and 4 are paramagnetic at room temperature (see
below). As solids, complexes 3 and 4 are stable in contact with
air, but after 1 d in non-dried solvents they start to decompose,
showing a set of colored spots on the TLC plate.
Our attempts to oxidize 3 with chloroform, tetrachloromethane
or iodine to generate a five-coordinated CoIII species failed. When
the strong oxidizing agent PhICl2 was used, ligand oxidation
to (Fbqdi)0, decomplexation, and isolation of a new phenazine
derivative—an isomer of (Fbqdi)0—was observed.30
Crystal structures
Table 1 contains the crystallographic data and Table 2 summarizes
important bond lengths of the complexes. Crystal structures of 1,
3, and 4 are isomorphous. The structure of the nickel complex 1
consists of two doubly deprotonated one-electron oxidized radical
anion ligands attached to the four-coordinated NiII (Fig. 3).
From a comparison of the C–C and C–N bond distances of
the ligands in 1 with corresponding bonds of the complexes
Fig. 3 Molecular structures of heavily twisted 1 (top) and less twisted
2 (bottom). Thermal ellipsoids are drawn at the 30% probability level,
hydrogen atoms are not shown.
reported previously (Fig. 2), it is evident that both ligands in 1
are in the radical monoanion (Fsbqdi)1− form. The C–N bond
˚
Table 2 Selected bond lengths [A] of complexes 1–4
[PdII(Fsbqdi)2] (2)
[NiII(Fsbqdi)2] (1)
[CoII(Fsbqdi)2] (3)
[CuII(Fsbqdi)2] (4)
N1–Pd1
1.993(3)
M1–N1
M1–N2
M1–N3
M1–N4
C1–C2
C3–C4
C5–C6
1.909(3)
1.902(4)
1.895(3)
1.889(4)
1.429(6)
1.421(7)
1.409(6)
1.418(6)
1.426(6)
1.427(8)
1.414(6)
1.435(6)
1.350(6)
1.352(7)
1.338(7)
1.366(7)
1.356(5)
1.365(5)
1.355(6)
1.362(5)
1.913(3)
1.903(3)
1.905(3)
1.900(3)
1.423(5)
1.400(6)
1.399(5)
1.428(5)
1.417(5)
1.400(6)
1.415(5)
1.429(5)
1.362(5)
1.364(5)
1.358(6)
1.369(6)
1.357(4)
1.368(4)
1.359(5)
1.370(4)
1.946(2)
1.947(3)
1.941(3)
1.942(3)
1.428(4)
1.415(5)
1.412(5)
1.437(4)
1.429(4)
1.402(6)
1.420(5)
1.432(5)
1.355(5)
1.355(5)
1.347(5)
1.362(5)
1.351(4)
1.351(4)
1.356(4)
1.360(4)
C7–C8
C9–C9#1
1.402(5)
1.428(8)
C7–C7#1
1.436(7)
C1–C6
C19–C20
C21–C22
C23–C24
C19–C24
C2–C3
C8–C9
C7–N1
1.365(6)
1.360(5)
C4–C5
C20–C21
C22–C23
C1–N1
C6–N2
C19–N3
C24–N4
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1355–1365 | 1359
©

View Article Online
length at average (av.) 1.360(5) A along with a quinonoid type
larger angles of 74.3(2)◦ with the central ligand planes. The crystal
˚
˚
=
distortion of the six-membered ring with two short C C bonds
structure of 2 shows C–N bond lengths of 1.360(5) A, along with a
˚
˚
at av. 1.352(7) A and four longer C–C bonds at av. 1.422(6) A are
in good agreement with the presence of two radical monoanions
(Fsbqdi)1− in 1. Note that observed bond lengths are accurate
characteristic quinonoid type distortion of the six-membered ring
with two short C C bonds at 1.365(6) A and four longer C–C
bonds at av. 1.417(6) A. Taking into account the accuracy of bond
˚
=
˚
˚
˚
˚
˚
within 0.018 A (3r) for C–C and 0.015 A for C–N bonds, and
the p/p effect of the perfluorinated phenyl rings on the ligand
bond pattern must be small since the C6F5 rings form dihedral
angles in the range of 51.0(2)–55.4(2)◦ with the corresponding
NCCN planes. This allows us to determine the oxidation state
of the ligand. Since the whole complex possesses zero charge,
the physical (or spectroscopic)31 oxidation state of Ni is +2
with d8 electronic configuration. High-spin (HS) NiII is favored
in tetrahedral complexes that could be expected in the twisted
complex 1 as well. Indeed, the bond distances Ni–N found in 1 are
considerably longer than those reported previously for square-
planar [NiII(sbqdi)2] complexes containing low-spin (LS) NiII.
Diamagnetism of 1 then demands the two radical ligands to be
antiferromagnetically coupled to the HS NiII (vide infra).
length determination ( 0.018 A for C–C and 0.015 A for C–N
bonds) we identify both ligands as (Fsbqdi)1− radical monoanions
and the metal as PdII (electronic configuration d8). Pd–N bonds
in slightly distorted 2 are close to such distances found in square-
planar LS PdII complexes reported previously.5 The ground state
singlet can then be easily achieved by antiferromagnetic coupling
of the two radical ligands via a superexchange mechanism.
The crystal structure of the cobalt complex 3 closely resembles
that of 1 (Fig. 4). The two nearly planar NCCN entities form an
interplanar angle of 54.1(1)◦ whereas the pairs of the perfluori-
nated phenyl rings are almost coplanar with dihedral angles of
2.0(2)◦ and 5.2(1)◦ with respect to each other, and in the range
of 50.8(1)–55.8(1)◦ with respect to the NCCN planes. The ligand
bond pattern in 3 closely resembles the bond patterns observed
in complexes 1, 2 and 4 (vide infra). The C–N bonds show partly
The two essentially planar NCCN-chelate ligands32 in 1 are
twisted with respect to each other forming a dihedral an-
gle of 53.9(1)◦, whereas almost all structurally characterized
complexes of the type [NiII(sbqdi)2] are known to adopt a
nearly planar geometry. A similar twisted geometry as for 1
was reported by tom Dieck et al. for a related nickel com-
plex [Ni(DAD)2] bearing two sterically demanding N,Nꢀ-bis(2,6-
dimethylphenyl)diazadiene ligands.33 This complex, doubtfully
formulated as a Ni(0) compound with two closed-shell DAD
ligands, reveals a dihedral angle between the two diimine planes
of 44.5◦.
˚
double bond character with an average length of 1.364(4) A while
the six-membered rings reveal a quinonoid type distortion with
two short C C bonds at av. 1.363(6) A and four longer C–C
bonds at av. 1.414(5) A. This bond pattern suggests that the two
˚
=
˚
We attribute the considerable distortion from planarity of 1 to
the repulsive interaction between the bulky C6F5 groups engaged
in intramolecular stacking (Fig. 3). Two pairs of the perfluorinated
phenyl rings in syn position are nearly coplanar. The ring to ring
˚
distance is in the range 3.20–3.28 A. The hypothetical planar
complex 1 would require a ring–ring distance of approximately
34
˚
2.87 A. According to DFT calculations of Lorenzo et al. the
optimum interplanar distance for the (C6F6)2 dimer is 3.31–
˚
3.37 A. Assuming that optimum ring–ring distances for two
pentafluorophenyl rings in 1 and for (C6F6)2 are similar, strong
repulsion between two pairs of C6F5 rings in a hypothetical planar
complex 1 is anticipated. Distortion of the complex towards a
tetrahedral species is accompanied by a synchronous rotation of
the phenyl rings and provides a release in steric strain by increase
of the C6F5/C6F5 distances. Similar intramolecular C6F5/C6F5
35,36
˚
distances of 3.2–3.3 A were found in other X-ray structures.
The molecular structure of the palladium complex 2 differs
significantly from that of 1 (Fig. 3). The molecular structure of
2 possesses D2 symmetry in the crystal and shows less deviation
from planarity in comparison with 1. The dihedral angle formed
by the two NCCN planes is only 22.4(2)◦. We ascribe this trend
towards planarity to the longer bond distances Pd–N at 1.993(3)
˚
˚
A in comparison with Ni–N (av. 1.899(4) A). This contributes
to holding the perfluorinated phenyl rings at a sufficient distance
even without great deformation from the planar configuration.
The rings in syn position are nearly coplanar with a dihedral angle
◦
˚
of 3.4(2) . The interplanar distance C6F5/C6F5 is 3.21 A, closely
resembling the interplanar distance in 1. As a consequence of
increased planarity at the metal the perfluorophenyl rings form
Fig. 4 Molecular structures of 3 (top) and 4 (bottom). Thermal ellipsoids
are drawn at the 30% probability level, hydrogen atoms are not shown.
1360 | Dalton Trans., 2008, 1355–1365
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
ligands are in the radical monoanion form (Fsbqdi)1−. Taking into
account the non-charged nature of the complex, the oxidation state
of the cobalt ion is +2 (d7 electronic configuration) with a possible
HS/LS alternative. The HS configuration would be favored for a
distorted tetrahedral structure. The long Co–N bond distances in
II
˚
3 (av. 1.905(3) A) support our hypothesis of HS Co ion.
Complex 4 represents the first structurally characterized CuII
complex containing two o-phenylenediamine derived ligands.
Complexes of o-semiquinones and o-semiquinonemonoimines are
known, however.10,37 The oxidation states of the two ligands
along with the oxidation state of the metal (+2, d9 electronic
configuration) were unambiguously assigned on the basis of the
bond length pattern. The crystal structure of 4 resembles that of 1
or 3, showing the presence of two almost planar ligand backbones
and a twisted geometry around the metal center with a dihedral
angle of 54.0(1)◦. A structurally characterized CuI (d10) complex
containing two o-benzoquinonediimine ligands [CuI(7bqdi)2]Cl23
showed tetrahedral geometry around the metal with an interplanar
angle of 84.6(4)◦.
Summarizing the structural results, the complexes 1–4 contain
F
pairs of H2 pda derived ligands which are largely equivalent
within each complex. In all complexes 1–4 the oxidation state
of both ligands was unambiguously determined to be −1, which
renders these complexes to be described as [NiII(Fsbqdi)2] (1),
[PdII(Fsbqdi)2] (2), [CoII(Fsbqdi)2] (3), and [CuII(Fsbqdi)2] (4). The
distortion from square-planar geometry is obviously the result of
repulsion between the C6F5 moieties favoring HS configurations
in 1 and 3. The palladium complex 2 experiences less distortion
from planarity as a result of longer M–N distances in comparison
with those involving the first row transition elements.
Spectroelectrochemistry, EPR and magnetism
To confirm and specify the oxidation state assignments based
on the aforementioned detailed structural results we used ad-
ditional information from electrochemistry, spectroscopy (UV-
VIS-NIR, EPR) and magnetism. The combination, referred to as
spectroelectrochemistry, allowed us to study certain neighboring
oxidation states of the complexes 1–4. Representative EPR and
UV-VIS-NIR spectra from spectroelectrochemical measurements
are shown in Fig. 5–10, while Tables 3 and 4 summarize pertinent
redox potentials and absorption data, respectively.
Fig. 6 UV-VIS-NIR spectroelectrochemical response of the transitions
2 → 2+ (top), 2 → 2− (center), 2− → 22− (bottom) in CH3CN/0.1 M
Bu4NPF6.
Table 3 Redox potentials from cyclic voltammetrya
Comp.
E
1/2(red2)b
E
1/2(red1)b
E1/2(ox1)
1
2
3
4
−1.00
−1.01
−1.06
−0.89
−0.48
−0.42
−0.37
−0.40
0.56c
0.61
0.27c
0.12c
a Potentials in V vs FeCp2+/0; measurements in acetonitrile/0.1 M
Bu4NPF6. b Reversible reduction processes (DE = 60–80 mV). c Anodic
peak potentials for irreversible oxidation.
tentials are remarkably similar (Table 3) which is a first indication
for ligand-based processes. Much less negative reduction poten-
tials for 1–4 are observed compared to the previously reported
[MII(4sbqdi)2] and [MII(5sbqdi)2].5 In a first approximation this
can be ascribed to the stabilization of the LUMO by electron
withdrawing N-C6F5 groups.38 The irreversibility of the oxidation
processes is in accord with our unsuccessful attempts to generate
cationic complexes chemically.30
Fig. 5 UV-VIS-NIR spectroelectrochemical response of the transition 1
→ 1− in CH3CN/0.1 M Bu4NPF6.
All compounds 1–4 could be reversibly reduced in two one-
electron steps under cyclic voltammetry conditions whereas the
oxidation processes were not always reversible. The reduction po-
Starting with the diamagnetic nickel compound 1, it exhibits a
typical intense absorption in the near infrared which is attributed
to a ligand-to-ligand charge transfer (LLCT) transition. Such
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1355–1365 | 1361
©

View Article Online
Table 4 Absorption dataa from spectroelectrochemistryb
Compound kmax(e)a
1
824(15400), 660(5000), 515(14100), 378(7600), 303 sh,
246(26200)
1060(5500), 830 sh, 660 sh, 524(4900), 387(8800),
318(16500), 244(26900)
895(14200), 688 sh, 572 sh, 335(4900), 265(14600),
232(14400)
1−
2
2−
22−
2+
3
1483(10500), 895(1100), 666(1700), 390(2800), 254 sh,
238(18800)
1486(5100), 667(1000), 613 sh, 558 sh, 370(4300), 253 sh,
240(19900)
1830(7700), 1332 sh, 907(1000), 485(4800), 295(11100),
229(13800)
1030(3900), 778(10700), 468 sh, 425(10800), 310(17800),
254 sh, 225(33500)
2050(1900), 755 sh, 667(5700), 478 sh, 320(21900), 251 sh,
232(36700)
680(1300), 340(27500), 233(36200)
1150(5500), 905(3600), 680 sh, 600(8300), 515 sh,
369(17400), 263 sh, 237(21400)
1975(3900), 1151(1900), 705(8100), 570(6100), 373(18500),
268 sh, 237(20400)
3−
32−
4
4−
4+
997(9200), 713(9500), 587(10100), 510(9000), 407(11200),
236(26800)
a Wavelengths in nm, molar extinction coefficients in dm3 mol−1 cm−1 b In
.
Fig. 7 EPR spectra of in situ generated 2− in CH3CN/0.1 M Bu4NPF6 at
298 K (top) and 110 K (bottom).
CH3CN/0.1 M Bu4NPF6, except for 4n (CH2Cl2/0.1 M Bu4NPF6).
Fig. 8 Temperature dependence of the effective magnetic moment of 3
measured at the field of 1 T.
Fig. 10 UV-VIS-NIR spectroelectrochemical response of the transitions
3 → 3− (top) and 3− → 32− (bottom) in CH3CN/0.1 M Bu4NPF6.
transitions have been discussed in detail for planar singlet
diradical compounds5,8 where the metal d8 center ensures
interaction between redox-active o-semiquinone ligands. Appa-
rently, this type of NIR band occurs also in 1 which exhibits
considerable twisting between the semiquinonediimine backbones.
It was possible to study the reversible one-electron reduction to
1− under the conditions of OTTLE spectroelectrochemistry, i.e.
requiring stability for at least 1 min. This reduction produced a
Fig. 9 X-Band EPR spectrum of 3 in frozen THF/toluene measured at
4 K was simulated with g1 = 4.328, g2 = 1.583, g3 = 1.462, A1 = 213.8,
A2 = 63.3, A3 = 69.5 G. The weak signal at g ∼ 2.0 is likely due to radical
impurities.
1362 | Dalton Trans., 2008, 1355–1365
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
broad band system at lower energies (Fig. 5, Table 4), suggesting
the occurrence of a ligand-to-ligand intervalence charge transfer
(LLIVCT) transition between the dianionic (Fpda)2− and the
monoanionic (Fsbqdi)− forms of the ligand.5,8 As could be
expected, LLCT band observed for 1 is absent in the spectrum
of reduced 1−. Although the structure of [NiII(Fpda)(Fsbqdi)]− is
not known, it may be safely assumed that, for steric reasons, it
cannot become planar. EPR experiments showed only a single
line centered at g = 1.97 which broadens but does not show
individual g components at 110 K. The deviation from g = 2.0 to
lower values indicates mixing in of low-lying excited states with
non-zero angular momentum.39
The palladium analogue 2 with its reduced twist also exhibits
diamagnetism and an LLCT band at about 900 nm (Fig. 6). One-
electron reduction creates a radical species (g = 2.020) which shows
significant g anisotropy g1 − g3 = 0.103 in frozen solution (Fig. 7).
This g splitting is attributed to the much higher spin–orbit coupling
(SOC) constant of PdII (1600 cm−1) vs NiII (630 cm−1)40 and has
been observed similarly for o-semiquinonemonoimine compounds
of Pd.41 In agreement with a more planar structure and thus better
p/p overlap than in the Ni analogue the LLIVCT transition for
[PdII(Fpda)(Fsbqdi)]− occurs as an intense band at about 1500 nm,
a feature which diminishes on further reduction (Fig. 6).
It is instructive to note that reduced species 1− and 2− produce
EPR spectra with significantly different g values, it is lower than
2.0 for 1−, but larger than 2.0 for 2−. If the electronic structure of 1−
and 2− would be similar, the direction of Dg would be the same and
only the magnitude would be stronger for the palladium complex.
The different trend of Dg indicates that there are different excited
states in 1− and 2− mixed into the ground state by SOC. Since PdII
in LS is close to planar 2 and presumably in its monoanion 2−, NiII
is expected to be HS in 1− as well as in its neutral counterpart 1 if
geometry of the complex preserves upon reduction. This difference
in the electronic structure is attributed to the different distortion
from square-planar towards tetrahedral geometry in 1 vs 2.
The oxidation occurs reversibly in the spectroelectrochemical
experiment and leads to one intense long-wavelength band
assigned to an LLIVCT process involving the ligands as the
monoanionic (Fsbqdi)1− donor and the neutral (Fbqdi)0 acceptor
in [PdII(Fsbqdi)(Fbqdi)]+. Such NIR bands and their assignment
were described recently in a similar way for corresponding o-
iminophenolate complexes of PdII.41
antiferromagnetically to two radical ligands (↓/↑↑↑/↓), a LS
CoII with two radicals coupled antiferromagnetically with each
other (↑/↑/↓), or a LS CoII antiferromagnetically coupled to
the both radical ligands (↑/↓/↑). The situation (↑/↑/↓) can be
safely discarded, since it would require strong antiferromagnetic
coupling between the two remote ligands which seems to be
unlikely in the highly twisted arrangement of 3. Furthermore,
a HS CoII (SCo = 3/2) rather than LS alternative (SCo = 1/2) is
expected in such a highly twisted arrangement that overall give
the only reasonable description (↓/↑↑↑/↓). The ground state of
3 can be effectively probed by EPR spectroscopy at cryogenic
temperatures. The highly anisotropic g tensor and a well resolved
hyperfine structure at g1 (59Co, I = 7/2, A1 = 213.8 G) indicate,
that the SOMO has predominantly metal character (Fig. 9). The
effective (total) g and A tensors obtained for the ground doublet
can be converted into the local tensors using the spin projection
techniques.42 The following equations were used:
geff (St = 1/2) = −2/3gR + 5/3gCo
Aeff (St = 1/2) = +5/3ACo,
where gR is the local g tensor for the radical ligands (SR = S1
+ S2; SR = 1), gCo—the local g tensor for cobalt ion (SCo = 3/2),
and ACo—the local hyperfine coupling tensor for cobalt; geff and
Aeff are the “observable” tensors. Assuming that gR is isotropic
with all diagonal components equal to 2.00 the intrinsic cobalt
g values were calculated as g1Co = 3.297, g2Co = 1.749, g3Co
=
1.677 and intrinsic cobalt hyperfine coupling components were
A1Co = 128.3, A2Co = 38.0, A3Co = 41.7 G. These results suggest
that the already highly anisotropic gCo and ACo tensors became
even more anisotropic upon coupling with the ligands. The highly
anisotropic gCo tensor is believed to be the result of admixture of
low-lying excited states into the ground state. A very similar EPR
spectrum was reported for a twisted [CoII(FDAD)2]36 containing
two radical ligands. The averaged g value can be calculated from
the EPR data according to the following equation:
ꢂ
1
gav
=
(g21 + g22 + g23),
3
where gi (i = 1, 2, 3) are components of the effective g tensor
for the ground state doublet. Calculated gav is then 2.79 that is
in a good agreement with the value obtained form the SQUID
measurements.
The electronic spectrum of 3 features a strong band at 778 nm
assigned to an LLCT transition. The reduction of 3 produces a
long-wavelength absorption at 2050 nm in the first step, attributed
to the LLIVCT transition. On second reduction this feature
disappears completely (Fig. 10).
According to the bond lengths analysis the copper complex
4 comprises a CuII ion with two radical ligands. The effective
magnetic moment of 1.7 lB is virtually temperature-independent in
the temperature range of 50–300 K (Fig. 11). This value is in good
agreement with the spin-only value of 1.73 lB for one unpaired
electron. It indicates strong antiferromagnetic coupling within the
system, whereby the excited quartet state is not populated even
at 300 K. According to its EPR spectrum, the spin density is
predominantly located on the ligand and not on the metal (vide
infra).
The cobalt complex
3 shows a virtually temperature-
independent effective magnetic moment (leff) of 2.54 lB. At
temperatures below 6 K an abrupt decrease of leff is observed
that could be ascribed to the presence of intermolecular anti-
ferromagnetic interactions or/and saturation effect (Fig. 8). The
values of leff observed for 3 point to strong antiferromagnetic
coupling within the system resulting in a doublet ground state
(St = 1/2) populated at the whole temperature range (2–300 K).
The strong deviation of leff from the spin-only value of 1.73 lB is
due to the excited states mixed into the ground state by SOC. The
corresponding gav = 2.93 was calculated using the equation:
leff
gav = ꢁ
S(S + 1)
According to X-ray analysis 3 consists of a CoII ion and
two radical ligands. The following electronic structures for the
ground doublet could then be proposed: a HS CoII coupled
Thus, the strong antiferromagnetic coupling between the
magnetic orbital of CuII and those of the radical ligands is
expected with |JCu–R| ꢂ 200 cm−1. Because of the twisted
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1355–1365 | 1363
©

View Article Online
Fig. 11 Temperature dependence of the effective magnetic moment of 4
measured at the field of 1 T.
geometry, the remote ligands couple weakly as compared to
the strong antiferromagnetic coupling between the metal and
the radical ligands (|JR–R| ꢃ |JCu–R|). This result differs
from that obtained for CuII complexes with N,O-donor o-
benzosemiquinonemonoiminato(1−) ligands in square-planar10
or twisted37 geometries, where the antiferromagnetic coupling
between the remote ligands (JR–R) is dominant10 or comparable37
to JCu–R, leading to the population of excited states at room
temperature. Below 50 K leff decreases gradually and reaches
the value of 1.45 lB at 1.8 K. Such behavior can be ascribed
to antiferromagnetic intermolecular interactions in solid 4. In
fact, the magnetic susceptibility data can be readily simulated via
the classical Curie–Weiss equation with a Weiss temperature of
−0.74 K.
Fig. 13 UV-VIS-NIR spectroelectrochemical response of the transitions
4 → 4− (top) and 4 → 4+ (bottom) in CH3CN/0.1 M Bu4NPF6.
of a higher-energy NIR band at 1000 nm, attributed to LLIVCT
transition [CuII(Fsbqdi)(Fbqdi)]1+ → [CuII(Fbqdi)(Fsbqdi)]1+.
Conclusion
The alternative arrangements of the three spins as ↑/↓/↑ or
↑/↑/↓ to effect an St = 1/2 ground state can be best probed via
EPR spectroscopy. One would expect predominantly ligand-based
SOMO in the former and predominantly metal-based SOMO
in the latter case. The high-frequency EPR spectrum in Fig. 12
clearly shows that the ↑/↓/↑ situation with the ligand-centered
SOMO is realized. This might be expected for a highly twisted
system where the two semiquinones cannot efficiently couple
antiferromagnetically.37 The fact that g < 2 reflects the presence of
low-lying excited states in agreement with previous observations.37
With the employment of the new bulky ligand N,Nꢀ-
bis(pentafluorophenyl)-o-phenylenediamine we succeeded in the
synthesis of the four homoleptic complexes [MII(Fsbqdi)2] (M =
Co, Ni, Pd, Cu) which show strongly (Co, Ni, Cu) or mod-
erately (Pd) twisted geometry around the metal centers. The
twisted geometry of the complexes seems to be the result of the
intramolecular p–p interaction between the two pairs of C6F5-
rings. All reported complexes 1–4 comprise divalent metal ions
with a pair of radical monoanions (Fsbqdi)−. The twisted ligands
arrangement leads to a HS NiII and CoII ions in complexes 1 and 3,
whereas LS metal ions were previously reported for square-planar
complexes. Superexchange interactions between the remote radical
ligands are not as effective as in planar complexes. This leads to
predominant antiferromagnetic coupling between the metal ions
and the radical ligands in complexes 1, 3 and 4, that is in contrast to
the strong antiferromagnetic coupling between two radical ligands
in square-planar complexes reported previously. The four electron
withdrawing N-C6F5 groups considerably stabilize reduced species
and destabilize oxidized ones. The much less negative reduction
potentials found for complexes 1–4 when compared to the
known complexes afford an opportunity for the isolation and
crystallographic characterization of the doubly reduced species
[MII(pda)2]2− that has never been realized before.
Fig. 12 EPR spectrum of 4 at 285 GHz in CH2Cl2/toluene (4/1) at 5 K;
g⊥ = 1.989, gꢁ = 1.959.
Akin to a twisted bis(o-semiquinonemonoimine)copper(II)
system,37 the reduction of 4 to 4− produces a long-wavelength
band at about 2000 nm (Fig. 13), which is assigned as LLIVCT
transition [CuII(Fpda)(Fsbqdi)]1− → [CuII(Fsbqdi)(Fpda)]1−. On the
other hand, one-electron oxidation to 4+ causes the emergence
Acknowledgements
We thank the Deutsche Forschungsgemeinschaft and Fonds der
Chemischen Industrie for financial support. M. M. K. appreciates
1364 | Dalton Trans., 2008, 1355–1365
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Dr E. Bill (MPI for Bioinorganic Chemistry, Germany) for very
fruitful discussions.
18 J. Rall, A. F. Stange, K. Hu¨bler and W. Kaim, Angew. Chem., Int. Ed.,
1998, 37, 2681.
19 D. Venegas-Yazigi, H. Mizra, A. B. P. Lever, A. J. Lough, J. Costamagna
and R. Latorre, Acta Crystallogr., Sect. C, 2000, C56, e281.
20 D. Venegas-Yazigi, H. Mizra, A. B. P. Lever, A. J. Lough, J. Costam-
agna, A. Vega and R. Latorre, Acta Crystallogr., Sect. C, 2000, C56,
e245.
21 T. Ju¨stel, J. Bendix, N. Metzler-Nolte, T. Weyhermu¨ller, B. Nuber and
K. Wieghardt, Inorg. Chem., 1998, 37, 35.
22 G. S. Hall and R. H. Soderberg, Inorg. Chem., 1968, 7, 2300.
23 H.-Y. Cheng, B.-C. Tzeng and S.-M. Peng, Bull. Inst. Chem. Acad. Sin.,
1994, 41, 51; H.-Y. Cheng, P.-H. Cheng, C.-F. Lee and S.-M. Peng,
Inorg. Chim. Acta, 1991, 181, 145.
References
1 M. D. Ward and J. A. McCleverty, J. Chem. Soc., Dalton Trans., 2002,
275; A. Vlcˇek, Jr., Coord. Chem. Rev., 2002, 230, 225; P. Chaudhuri and
K. Wieghardt, Prog. Inorg. Chem., 2001, 50, 151; A. B. P. Lever and
S. I. Gorelsky, in Struct. Bonding, Springer, Berlin, Heidelberg, 2004,
vol. 107, ch. 7, pp. 77-114; C. G. Pierpont, Coord. Chem. Rev., 2001,
216–217, 99; W. Kaim and B. Schwederski, Pure Appl. Chem., 2004, 76,
351; K. Ray, T. Petrenko, K. Wieghardt and F. Neese, Dalton Trans.,
2007, 16, 1552.
2 A. Mederos, S. Dom´ınguez, R. Herna´ndez-Molina, J. Sanchiz and F.
Brito, Coord. Chem. Rev., 1999, 193–195, 913.
3 A. L. Balch and R. H. Holm, J. Am. Chem. Soc., 1966, 88, 5201.
4 E. I. Stiefel, J. H. Waters, E. Billig and H. B. Gray, J. Am. Chem. Soc.,
1965, 87, 3016.
24 K. Chlopek, E. Bothe, F. Neese, T. Weyhermu¨ller and K. Wieghardt,
Inorg. Chem., 2006, 45, 6298.
25 D. Sorokin, PhD Thesis, Philipps-Universita¨t Marburg, 2006.
26 L. G. L. Ward and D. M. Paul, Inorg. Synth., 1971, 13, 154.
27 P. K. Byers, A. J. Canty, H. Jin, D. Kruis, B. A. Markies, J. Boersma
and G. van Koten, Inorg. Synth., 1998, 32, 162.
5 D. Herebian, E. Bothe, F. Neese, T. Weyhermu¨ller and K. Wieghardt,
J. Am. Chem. Soc., 2003, 125, 9116.
28 PLATON-A Multipurpose Crystallographic Tool, A. L. Spek, Utrecht
University, Utrecht, The Netherlands, 2005.
6 E. Bill, E. Bothe, P. Chaudhuri, K. Chlopek, D. Herebian, S. Kokatam,
K. Ray, T. Weyhermu¨ller, F. Neese and K. Wieghardt, Chem.–Eur. J.,
2005, 11, 204.
29 SHELX97-Programs for Crystal Structure Analysis (Release 97–2),
G. M. Sheldrick, Institu¨t fu¨r Anorganische Chemie der Universita¨t,
Go¨ttingen, Germany, 1998.
7 A. V. Reshetnikov, M. O. Talismanova, A. A. Sidorov, S. E. Nefedov,
I. L. Eremenko and I. I. Moiseev, Russ. Chem. Bull., Int. Ed., 2001,
50, 142; M. O. Talismanova, I. G. Fomina, A. A. Sidorov, G. G.
Aleksandrov, N. T. Berberova, A. O. Okhlobystin, E. V. Shinkarꢀ, I. F.
Golovaneva, I. L. Eremenko and I. I. Moiseev, Russ. Chem. Bull., Int.
Ed., 2003, 52, 2701; I. G. Fomina, A. A. Sidorov, G. G. Aleksandrov,
S. E. Nefedov, I. L. Eremenko and I. I. Moiseev, J. Organomet. Chem.,
2001, 636, 157.
30 M. M. Khusniyarov, K. Harms and J. Sundermeyer, J. Fluorine Chem.,
2006, 127, 200.
31 C. K. Jo¨rgensen, in Oxidation Numbers and Oxidation States, Springer,
Heidelberg, Germany, 1969.
32 The NCCN plane is defined by the six carbon atoms of the o-
phenylenediamine ring, the two attached nitrogen atoms, and the metal.
33 H. tom Dieck, M. Svoboda and T. Greiser, Z. Naturforsch., B: Anorg.
Chem., Org. Chem., 1981, 36, 823.
8 V. Bachler, G. Olbrich, F. Neese and K. Wieghardt, Inorg. Chem., 2002,
41, 4179; D. Herebian, K. E. Wieghardt and F. Neese, J. Am. Chem.
Soc., 2003, 125, 10997.
9 S.-M. Peng, C.-T. Chen, D.-S. Liaw, C.-I. Chen and Y. Wang, Inorg.
Chim. Acta, 1985, 101, L31.
34 S. Lorenzo, G. R. Lewis and I. Dance, New J. Chem., 2000, 24, 295.
35 M. P. Thornberry, C. Slebodnick, P. A. Deck and F. R. Fronczek,
Organometallics, 2000, 19, 5352.
36 M. M. Khusniyarov, K. Harms, O. Burghaus and J. Sundermeyer,
Eur. J. Inorg. Chem., 2006, 2985.
10 P. Chaudhuri, C. N. Verani, E. Bill, E. Bothe, T. Weyhermu¨ller and K.
Wieghardt, J. Am. Chem. Soc., 2001, 123, 2213.
11 G. Bandoli, T. I. A. Gerber, J. Perils and J. G. H. du Preez, Inorg. Chim.
Acta, 1998, 278, 96.
37 S. Ye, B. Sarkar, F. Lissner, T. Schleid, J. van Slageren, J. Fiedler and
W. Kaim, Angew. Chem., Int. Ed., 2005, 44, 2103; S. Ye, W. Kaim, M.
Niemeyer, F. Lissner, T. Schleid, J. van Slageren, J. Fiedler, B. Sarkar
and C. Duboc, unpublished work.
12 V. Tabernero, T. Cuenca and E. Herdtweck, Eur. J. Inorg. Chem., 2004,
3154.
38 A. M. Bond, in Comprehensive Coordination Chemistry II, ed. J. A. Mc-
Cleverty and T. J. Meyer, Elsevier, Pergamon, Oxford, 2004, vol. 2,
197.
13 D. J. Darensbourg, K. K. Klausmeyer and J. H. Reibenspies, Inorg.
Chem., 1996, 35, 1535.
39 W. Kaim, Coord. Chem. Rev., 1987, 76, 187.
14 H.-Y. Cheng, C.-C. Lin, B.-C. Tzeng and S.-M. Peng, J. Chin. Chem.
Soc., 1994, 41, 775.
15 P.-H. Cheng, H.-Y. Cheng, C.-C. Lin and S.-M. Peng, Inorg. Chim.
Acta, 1990, 169, 19.
40 J. A. Weil, J. R. Bolton and J. E. Wertz, in Electron Paramagnetic
Resonance, Wiley, New York, 1994, pp. 532–533.
41 S. Kokatam, T. Weyhermu¨ller, E. Bothe, P. Chaudhuri and K.
Wieghardt, Inorg. Chem., 2005, 44, 3709.
16 S. Ne`meth, L. I. Simandi, G. Argay and A. Ka´lman, Inorg. Chim. Acta,
1989, 166, 31.
17 G. G. Christoph and V. L. Goedken, J. Am. Chem. Soc., 1973, 95, 3869.
42 A. Bencini and D. Gatteschi, in Electron Paramagnetic Resonance
of Exchange Coupled Systems, Springer, Heidelberg, Germany,
1990.
This journal is
The Royal Society of Chemistry 2008
Dalton Trans., 2008, 1355–1365 | 1365
©