Amphiphilic paramagnetic neutral gold dithiolene complexes

Article abstract of DOI:10.1039/b820819k

The sulfiding of benzils with P₄S₁₀ in 1,3-dimethyl-2-imidazolidinone (DMI) as solvent allows for a direct synthesis of neutral radical, gold dithiolene complexes based on 1,2-bis-(4-alkoxy-phenyl) ethylene-1,2-dithiolate ligands wit

Full text of DOI:10.1039/b820819k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/dalton | Dalton Transactions
Amphiphilic paramagnetic neutral gold dithiolene complexes†
Romain Perochon,a Lydia Piekara-Sady,a,b Witold Jurga,b Rodolphe Cle´racc,d and Marc Fourmigue´*a
Received 21st November 2008, Accepted 6th February 2009
First published as an Advance Article on the web 3rd March 2009
DOI: 10.1039/b820819k
The sulfiding of benzils with P4S10 in 1,3-dimethyl-2-imidazolidinone (DMI) as solvent allows for a
direct synthesis of neutral radical, gold dithiolene complexes based on 1,2-bis-(4-alkoxy-
phenyl)ethylene-1,2-dithiolate ligands with n-butyl, n-octyl and n-dodecyl chains. The three neutral and
soluble complexes Au-OC4, Au-OC8 and Au-OC12 exhibit a near infrared (NIR) absorption band
around 1.5 mm and EPR characteristics which confirm a strong delocalization of the spin density on the
electron-rich dithiolene ligands. X-Ray crystal structures of Au-OC4 and Au-OC12 are compared with
those of the corresponding nickel complexes. They are characterised by segregation of the alkyl chains
into layered structures with a stacking of the radical complexes into alternated spin chains, confirmed
by the temperature dependence of the magnetic susceptibility which attests for antiferromagnetic
interactions and a singlet ground state. Observations under polarising microscope and DSC
experiments do not reveal a thermotropic behaviour for Au-OC12.
Introduction
Neutral nickel dithiolene complexes1 are currently intensively
investigated for several applications such as laser dyes,2,3 dyes for
liquid crystal devices,4 light compensation filters for near infrared
radiation and optical recording disks.5 The essential molecular
characteristic of the complexes used in these devices is their
strong NIR absorption, found at wavelengths ranging from 0.8–
1.5 mm. Furthermore, these complexes possess excellent thermal
and photochemical stability and their neutral character ensures a
good solubility in non-polar solvents or in liquid crystal hosts.4 No
other dithiolene complexes than those derived from the Ni–Pd–Pt
triad have been used for these applications so far, since they are
the only ones (commercially) available in their neutral soluble form
Scheme 1
with such a strong NIR absorption. In that respect, gold dithiolene
complexes offer a very attractive perspective as they can also be
isolated in their neutral form, from the one-electron oxidation
of the precursor [AuIII(dt)2]- complexes.6 In some cases however,
these neutral complexes are not stable and in most cases, they are
obtained in a poorly soluble microcrystalline form. Only a few
complexes were isolated by electro-crystallisation from the mono-
anionic salts as single crystals, that is [Au(dddt)2],7 [Au(bdt)2],8
[Au(F2pdt)2]9 and [Au(tmdt)2] (Scheme 1).10
tallise) on the anode. Very recently, Wieghardt et al. described
one example of a stable, soluble, neutral radical gold complex Au-
tBu (Scheme 2) derived from 1,2-di-(4-tert-butylphenyl)ethylene-
1,2-dithiolate, together with a complete analysis of its electronic
structure.12 Another gold complex [Au(tBu2bdt)2] (Scheme 1)
derived from 3,5-di-tert-butyl-benzene-1,2-dithiolate was reported
to decompose upon attempted recrystallisation.13
This electro-crystallisation procedure11 is very attractive in the
search for conductive materials but can not be easily expanded
to larger quantities of insulating material. Furthermore, it is
essentially restricted to insoluble species which precipitate (crys-
aSciences Chimiques de Rennes, UMR 6226 CNRS-Universite´ de Rennes I,
Campus de Beaulieu, F-35042, Rennes, France. E-mail: marc.fourmigue@
univ-rennes1.fr
Scheme 2
bInstitute of Molecular Physics, Polish Academy of Science, M. Smolu-
chowskiego 17, 60-179, Poznan´, Poland
Besides the tert-butyl groups, long alkyl chains are also an
efficient way to solubilise neutral molecules. Furthermore, when
long enough, segregation patterns can take place which can
favour the formation of layered structures.14 This strategy has
been earlier developed for the preparation of thermotropic nickel
dithiolene complexes, allowing for the isolation of smectic and
cCNRS, UPR 8641, Centre de Recherche Paul Pascal (CRPP), Equipe
“Mate´riaux Mole´culaires Magne´tiques”, 115 avenue du Dr. Albert
Schweitzer, Pessac, F-33600, France
dUniversite´ de Bordeaux, UPR 8641, Pessac, F-33600, France
† CCDC reference numbers 710443–710445. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/b820819k
3052 | Dalton Trans., 2009, 3052–3061
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
nematic15 as well as discotic mesophases.16 Considering that the
analogous Pd or Pt complexes exhibit different thermotropic
behaviour attributed to a tendency to dimerise15b and that the
radical [Au(dddt)2] complex crystallises into dimers,7 we wanted to
investigate the behaviour of the analogous neutral gold dithiolene
complexes, in order to evaluate if the tendency of radical species
to dimerise to form a 2-electron s bond, was also controlling
the structural features of these paramagnetic gold complexes,
in competition with the other intermolecular forces at work,
particularly when the complexes are functionalized with long alkyl
chains.
Scheme 3 Syntheses of the different Ni and Au complexes, with the yields
given for the reactions performed in dioxane or in DMI.
In this article, we described original syntheses of paramagnetic
neutral gold dithiolene complexes Au-OC4, Au-OC8 and Au-OC12,
substituted with four alkoxy chains of variable length, and we
investigate their electronic and structural properties, in solution
as well as in the solid state, with a special emphasis on the
similarities and differences they exhibit with the analogous dia-
magnetic nickel complexes Ni-OC4, Ni-OC8 and Ni-OC12 reported
earlier.16
Only very few examples of neutral radical gold dithiolene
complexes have been reported in the literature.7–10,12,13 They are
always based on the chemical or electrochemical oxidation of
the corresponding anionic and diamagnetic Au(III) dithiolene
complexes [Au(dt)2]-. The latter are easily obtained from the
corresponding dithiolate and a Au(III) salt such as AuCl4 .6 In
-
the present case, the phenyl-substituted dithiolate ligands are
not easily isolable and the sulfiding procedure described16 for
the nickel derivatives from benzils and P4S10 has been reported12
to provide the gold complex Au-tBu, but as monoanion, in
13% yield. It subsequent oxidation to the neutral species was
performed with ferricinium hexafluorophosphate in 70%, that is
a combined yield of 9% from the starting benzil derivative. Using
the same conditions (P4S10–dioxane–KAuCl4) with the alkoxy-
substituted benzils 1a–c, we were able to isolate directly the
oxidized neutral complexes Au-OC4, Au-OC8 and Au-OC12 (rather
than the mono anions) in comparable yields (10%). Furthermore,
using DMI as solvent allowed us to strongly increase the yields
up to 25–35% (Scheme 3), providing the three neutral radical
gold complexes in quantities large enough for their complete
characterization.
Results and discussion
Syntheses
The preparation of phenyl-substituted nickel dithiolene complexes
is most often based on the sulfiding of benzils with P4S10 or
Lawesson reagent in 1,4-dioxane, followed by hydrolysis of the
intermediate phosphorous thioesters in the presence of the nickel
salt such as NiCl2·6H2O to afford directly the oxidized, neutral
nickel complex (Scheme 3). This strategy has been successfully
engaged for the preparation of benzils derivatized with long alkoxy
chains such as 1a–c to afford Ni-OC4, Ni-OC8 or Ni-OC12 in ª 50%
yield.16 A recent patent description claimed that the yield of these
sulfiding reactions was much improved by the use of 1,3-dimethyl-
2-imidazolidinone (DMI) as solvent rather than dioxane.17 We
performed accordingly the sulfiding of the alkoxy-substituted
benzils 1a–c to find that much better results were indeed obtained
as the yields of Ni complexes were increased up to 85% in
DMI.
Electrochemistry
Electrochemical data for the six Ni and Au complexes were inves-
tigated by cyclic voltammetry (Table 1) and cyclic voltammogram
of Au-OC12 is shown in Fig. 1. While the chain length in the Ni
Table 1 Electrochemical data (in V vs. SCE). The electrochemical gap for the neutral compounds, DE, value refers to the difference E1/2 (+1/0) - E1/2
(0/-1). Values in parentheses for the Ni complexes are those reported in ref. 16c
Ered (-1/-2)
E1/2 (0/-1)
E1/2 (+1/0)
DE
Ref.
Ni complexes
Ni-OC4
Ni-OC8
Ni-OC12
Ni-H
-1.05 (irr)
-1.15 (irr)
-1.33 (irr)
-0.87
-0.101 (-0.06)
-0.089 (-0.05)
-0.107 (-0.06)
-0.045
0.877
0.904
0.915
—
0.98
0.99
1.02
—
This work (16c)
This work (16c)
This work (16c)
18
Au complexes
Au-OC4
—
—
—
0.251
0.240
0.244
0.32a
0.28a
0.464
0.505
0.753
0.765
0.758
0.84a
0.84a
—
0.50
0.51
0.52
0.52
0.56
—
This work
This work
This work
12
12
13
9
Au-OC8
Au-OC12
Au-Ph
-1.64a
-1.70a
-1.89
-1.42
Au-tBu
[Au(tBu2bdt)2]
[Au(F2pdt)2]
—
—
a For comparison purposes, data reported vs. Fc+/Fc have been transformed vs. SCE by adding 0.39 V.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 3052–3061 | 3053
©

View Article Online
Fig. 1 Cyclic voltammogram of Au-OC12, 1.5 10-3 M in CH2Cl2 at
100 mV s-1.
Fig. 2 UV-vis–NIR spectrum of Au-OC12 in CH2Cl2.
and Au complexes does not induce significant modifications, the
-1/0 redox process, observed around -0.10 V in the Ni complexes
is shifted systematically to higher potentials in the corresponding
gold complexes by ª 0.25 V. Furthermore, the electron-releasing
effect of the alkoxy chain is clearly seen when one compares the
potentials of the -1/0 and 0/+1 redox processes, shifted to more
cathodic potentials when the para substituent on the phenyl ring
goes from hydrogen to tert-butyl or biphenyl to the alkoxy chain.
This cathodic shift affects more strongly the 0/+1 than the -1/0
redox processes. As a consequence, the neutral gold complexes
described here have a smaller electrochemical gap DE [defined as
can be observed for example when comparing the spectroscopic
properties of Ni–H and [Ni–H]- (Table 2), with an associated
decrease of intensity, also observed when going from the neutral
nickel to the neutral gold complexes.
Structural properties
The good solubility brought by the long alkyl chains allowed for
the recrystallisation of the six different complexes. Good X-ray
quality structures could be determined for Ni-OC4, Au-OC4 and
Au-OC12 while the n-octyl derivatives Ni-OC8 and Au-OC8 could
not afford tractable crystals. The very thin crystals obtained with
Ni-OC12 did not allow us to complete the structure refinements.
Only the heavy atoms (Ni, S) could be refined anisotropically,
giving however a tentative idea of the structural arrangement.
Ni-OC4 and Au-OC4 are closely related even if not perfectly
isomorphous; they exhibit slightly different disorder models on
two out of the four butyl chains (Fig. 3). They both crystallize in
E
1/2(+1/0) - E1/2(0/-1)] than the tert-butyl derivative Au-tBu.
Optical spectroscopy
The three nickel complexes Ni-OC4, Ni-OC8 or Ni-OC12 exhibit a
strong absorption band in the NIR (Table 2), at lmax = 933–934 nm,
at slightly lower energy than that observed in the unsubstituted
phenyl derivative Ni-H.18 On the other hand, the corresponding
neutral gold complexes exhibit NIR absorption bands at much
lower energy, with wavelengths above 1.5 mm (Fig. 2). Even in
the gold series, this NIR absorption band is shifted toward even
lower energies, in the presence of the strongly electron-releasing
alkoxy groups. This result is in accordance with the decreased
electrochemical gap noted for these complexes (Table 1) and shows
that NIR bands at very low energy can be found when moving
from the nickel to the analogous, neutral, soluble, gold complexes.
Note that this absorption band at low energy corresponds to that
observed in the corresponding, isoelectronic monoanionic nickel
complexes. This shift toward lower energy in the radical species
Table 2 NIR absorption band characteristics
l/nm
e/M-1 cm-1
Ref.
Ni-OC4
933
934
933
855
943
1565
1565
1566
1495
1452
1110
52 800
41 900
31 300
29 800
10 600
21 400
25 600
33 400
21 200
27 000
10 000
This work
This work
This work
19
Ni-OC8
Ni-OC12
Ni-H
[Ni-H]-
20
Au-OC4
Au-OC8
Au-OC12
Au-tBu
This work
This work
This work
12
13
12
[Au(tBu2bdt)2]
[Au(mnt)2]
Fig. 3 Projection view perpendicular to the MS4 mean plane for Au-OC4
(top) and Ni-OC4 (bottom).
3054 | Dalton Trans., 2009, 3052–3061
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 3 Selected average bond distances, together with the tetrahedral distortion angle (angle between the two MS2 planes of each metallacycle) and the
angles between the Ph and MS2C2 mean planes (one line for each of the two crystallographically independent metallacycles)
Tetrahedral
distortion
Ph vs. MS2C2
plane angles
=
C C
Au–S
C–S
Ref.
Au-OC4
Au-OC12
Au-tBu
2.306(5)
1.750(7)
1.374(8)
1.383(10)
1.36(2)
3.54(5)
1.42(6)
2.1(1)
35.5(1), 52.5(1)
41.6(1),
49.9(1)
39.8(1), 40.8(1)
43.3(1),
44.5(1)
This work
2.290(5)
2.291(5)
1.736(10)
1.749(20)
This work
12
40.4(4), 55.4(5)
47.5(4),
55.4(4)
Ni-OC4
2.137(1)
1.725(3)
1.405(6)
5.63(4)
28.76(9), 51.15(8)
42.03(9),
This work
42.19(8)
41, 43
41, 43
31.04(4), 54.36(4)
42.64(4),
Ni-OC12
Ni-OMe
—
—
—
5–6
This work
21
2.1241(6)
1.712(2)
1.392(3)
5.00(2)
42.81(5)
38.25(9), 52.65(6)
47.12(7),
Ni-H
2.124(1)
1.704(4)
1.414(6)
4.51(5)
22
49.92(6)
¯
the triclinic system, space group P1, with the complex in general
position in the unit cell. In both complexes, the metal is in a square
planar environment with a very limited tetrahedral distortion as
the angles between the two MS2 planes do not exceed 5◦ (Table 3).
The phenyl rings are not co-planar with the metallacycle but make
dihedral angles, from 35–50◦ in Au-OC4, from 29–51◦ in Ni-OC4.
The strong distortions from planarity for the whole complexes,
which are a consequence of this rotation of the phenyl rings out
of the metallacycle planes, are comparable to those reported in
analogous neutral Ni and Au complexes. As we will see below, they
are expected to limit in some way the stacking of the complexes
on top of each other.
The Ni-OC12 complex afforded only weakly diffracting crystals
and the structure could not be fully determined as only the Ni
and S atoms were refined anisotropically. We therefore report,
in Table 3, only the unit cell parameters and give in Fig. 4 a
view of the complex, essentially for comparison with the Au-
OC12 derivative. Both complexes are not isostructural (Fig. 4),
the Au-OC12 one crystallises in the triclinic system, space group
Fig. 4 Projection views perpendicular to the MS4 mean plane of Ni-OC12
and Au-OC12, showing the different structural organisations of the dodecyl
chains around the central core.
¯
P1 with a complex in general position in the unit cell. The
dihedral angles between the phenyl rings and the metallacycles
are now all almost identical, between 40 and 44◦. Furthermore,
the long n-dodecyl chains are not disordered, with two of them
in a fully extended conformation, two other kinked by almost
90◦. In a whole, the dodecyl derivatives appear less disordered
than the butyl ones, a probable effect of their amphiphilic
character and the segregation of the long alkyl chains (see
below).
X-Band EPR spectroscopy
The X-band EPR spectra of the neutral gold complexes dissolved
in CH2Cl2 solution display at room temperature one single
Lorentzian line at g = 2.011 with a line width of 23, 25 and 31 G
for Au-OC4, Au-OC8 and Au-OC12, respectively. This remarkable
evolution of the line width with the chain length is attributed here
to a smaller tumbling rate in solution with the longer chains. In
frozen solution, anisotropic signals around g ª 2 typical of an S =
The intramolecular bond distances within the metallacy-
cles (Table 3) are analogous to those described in reference
complexes, with a shortening of the C–S bonds and con-
=
comitant lengthening of the C C double bond when com-
1
2
system with rhombic g factors are observed (Fig. 5 and Table 4).
pared with the corresponding anionic species, an indication
that the oxidation essentially affects the dithiolene ligands
rather than the central metal centre, as demonstrated by
Wieghardt.12,13
The g tensor principal values collected in Table 4 compare very well
with those reported for Au-tBu or [Au(mnt)2],12 indicating a very
common electronic structure for all these neutral gold complexes.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 3052–3061 | 3055
©

View Article Online
Table 4 EPR parameters
Compound
gmin
gint
gmax
<gav >
gmax - gmin
104|A0|/cm-1
Ref.
Au-OC4
Au-OC8
Au-OC12
Au-tBu
1.945
1.952
1.950
1.944
1.928
2.026
2.028
2.026
2.030
2.039
2.059
2.057
2.058
2.065
2.075
2.010
2.012
2.011
2.013
2.014
0.114
0.105
0.108
0.121
0.147
<5
<5
<5
6.6
7.8
This work
This work
This work
12
[Au(mnt)2]
12
Fig. 6 X-Band EPR spectrum of Au-OC4 powder sample at room
temperature. Simulation with S = 1, g = 2.0, D = 0.03 cm-1, line width
550 G and Lorentzian line shape. Part of the impurity lines comes from
isolated Au-OC4 radical defects.
Fig. 5 X-Band EPR spectrum of Au-OC12 in frozen CH2Cl2 solution at
77 K. The solid line represents simulation with the parameters reported in
Table 4 using the line shape function Lorentzian/Gaussian = 0.5.
As detailed in the following, we note however a clear evolution
within the nature of the substituents on the metallacycles. A hyper-
fine splitting from the 197Au nucleus (I = 3/2, 100% abundance)
was reported in the spectra of Au-tBu and [Au(mnt)2]. In Au-
OC4, Au-OC8 and Au-OC12, no such hyperfine coupling could be
observed, giving a maximum value for |A0| of 5 10-4 cm-1. As the
theoretical A0 value for 6 s gold is 1026 10-4 cm-1, the very low
values obtained in these gold dithiolene complexes (5–8 10-4 cm-1)
was convincingly attributed to a very small metal character, with
the complexes best described as Au(III) species with a ligand
radical.12 This analysis is further substantiated here when we
observe that smaller anisotropies (<gav >, gmax - gmin) and smaller
A0 values are found indeed in those complexes characterized
by more “electron-rich” dithiolene ligands such as Au-OC4, Au-
OC8 and Au-OC12. In other words, the electron-releasing alkoxy
chains most probably lead to an enhanced ligand radical character
of the magnetic orbital. This very strong delocalisation of the
spin density on the two dithiolene ligands might lead to sizeable
intermolecular magnetic interactions in the solid state, as shown
below.
are essentially temperature independent, in contrast with highly
conducting single component molecular metals based on neutral
gold dithiolene complexes.6c,23 The powder spectrum of Au-OC4 is
shown in Fig. 6. A very broad line for a powder sample suggests
a dimeric structure of radicals in a solid.24 The simulation of the
broad line, assuming that dipole–dipole interaction within pairs of
radicals contributes to the linewidth, gives an estimation of zero
field splitting parameter, D ≤ 0.03 cm-1. Such low value of D is
consistent with a rather large separation between radical pairs (see
below).
Solid state structural and magnetic properties of Au-OC4
The X-ray crystal structure resolution of the two new neutral
radical gold dithiolene complexes Au-OC4 and Au-OC12 allows for
an in-depth analysis of the relationship between their structure and
their magnetic behaviour, considering that the SOMO is essentially
localized on the two metallacycles, as already calculated for the
model compound Au-H.12
Indeed, while frozen solution spectra are due to the single radical
species, the EPR on powder samples are significantly different.
They consist of a single line at g ª 2 with a peak to peak line width
at room temperature of 700, 550 and 900 G for Au-OC4, Au-
OC8 and Au-OC12, respectively. Narrow impurity lines are visible
on each broad line and therefore a Curie contribution should be
taken into account in susceptibility analysis. The line widths are
much broader than for other neutral gold radical compounds and
A view of the unit cell of Au-OC4 (Fig. 7) shows that the radical
complexes are organised into stacks running along a, without any
possibility for inter-stack magnetic interaction. A side view of one
of these chains (Fig. 8) shows alternating Au ◊ ◊ ◊ Au distances of
5.016(13) and 5.910(9) A, with two types of intermolecular overlap
noted A and B in Fig. 8. Interaction A is characterized by much
˚
˚
˚
shorter S ◊ ◊ ◊ S [3.942(7) A] and Au ◊ ◊ ◊ S [3.878(9) A] intermolecular
˚
distances than interaction B [S ◊ ◊ ◊ S 4.211(9), Au ◊ ◊ ◊ S 4.586(4) A].
3056 | Dalton Trans., 2009, 3052–3061
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Fig. 9 Temperature dependence of the cT product for Au-OC4. The solid
line is a fit to the alternating spin chain model (see text) with a = 0.3.
Fig. 7 Projection view along a of the unit cell of Au-OC4.
Ng2 b2
kT
c = c0 + r
+ (1 - r)calt .chain
(1)
It afforded c0 = -1.9 10-5 emu mol-1, r = 2.4%, J/k =
-246(1) K (J = -171 cm-1) and a = 0.3, indicating a sizeable
antiferromagnetic interaction and a strong dimerization within
the chains. Data obtained from the integration of the powder
EPR signal have been superimposed on the SQUID measurements
(Fig. 8), showing a very similar temperature dependence and
confirming the intrinsic character of the EPR signal.
Solid state structural and magnetic properties of Au-OC8. The
small quantities obtained for Au-OC8 and its reluctance to afford
crystalline material amenable to X-ray crystal structure resolution
did not allowed us to fully characterise this compound. However,
the temperature dependence of its magnetic susceptibility could be
evaluated from the integration of its powder EPR signal between
77 K and room temperature. As shown in Fig. 10, the cT vs. T plot
exhibits a behaviour very close to that observed above for Au-OC4
(Fig. 9), with antiferromagnetic interactions and a singlet ground
state, indicating most probably very common structural features,
that is a sizeable dimerization of the radical species.
Fig. 8 Left: view of the alternated chain running along a in Au-OC4.
Right: detail of the two different overlap interactions between radical
species. The n-butyloxyphenyl groups have been removed for clarity.
From
a magnetic point of view, these chains can be
best described as alternated spin chains.25 Such magnetic
systems are described with two exchange couplings J and
aJ with 0 < a < 1 and the following spin Hamiltonian:
. In the case of an antiferromag-
S2i-1S2i + aS2iS2i+1
H = -2J
(
)
Â
i=1..n
netic interaction (J < 0), they are characterized by a singlet
ground state, in contrast to the uniform spin chain system (a =
1, Bonner–Fisher model), which exhibits a finite susceptibility at
low temperatures. The other limit (a = 0) is simply the dimer
Singlet–triplet system (Bleaney–Bower model). The temperature
dependence of the magnetic susceptibility determined from mag-
netic measurements on a polycrystalline sample of Au-OC4 is
shown in Fig. 9 as the cT product. It demonstrates the presence
of antiferromagnetic interactions and a singlet ground state and
can be best fitted within the alternated spin chain model according
to eqn (1) where r is the fraction of Curie-type magnetic defects
and calt.chain follows analytical expressions reported for alternated
chains.26
Fig. 10 Temperature dependence of the cT product for Au-OC8,
determined from the integration of the powder EPR data (c in arbitrary
units).
Solid state structural and magnetic properties of Ni-OC12 and Au-
OC12. The solid state arrangement for Ni-OC12 is characterised
by a strong segregation of the alkyl chains, giving rise to a layered
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 3052–3061 | 3057
©

View Article Online
structure (Fig. 11, left) with inter-digitated chains. A different
inter-digitation pattern is observed in the structure of Au-OC12
(Fig. 11, right)
Fig. 13 Temperature dependence of the cT product for Au-OC12. The
solid red line is a fit to the singlet–triplet model (see text).
chain model with J/kB = -163(5) K, g = 2.0(1), r = 0.007(3) and
a = 0.04(2).
The dimerisation observed here in the two gold complexes
can be compared with the structures of the few other neutral
radical gold dithiolene complexes reported so far. Indeed, while
[Au(bdt)2] and [Au(F2pdt)2] crystallise into uniform spin chains,8,9
[Au(dddt)2] was isolated into strongly associated dimers,7 with
structural characteristics very different from those of Au-OC4 and
Fig. 11 Projection views of the unit cells of Ni-OC12 (left) and Au-OC12
(right).
Because of the paramagnetic nature of Au-OC12, we also inves-
tigated the temperature dependence of its magnetic susceptibility,
in order to correlate it with its solid state arrangement. As for the
Au-OC4 complex, the Au-OC12 molecules organize into alternated
chains running along a (Fig. 12). Two types of overlap can be iden-
tified, noted A and B in Fig. 12 with two intermolecular Au ◊ ◊ ◊ Au
˚
Au-OC12 since the shortest intermolecular S ◊ ◊ ◊ S [3.59(1) A] are
much shorter than in the two gold complexes described here. As
a consequence, [Au(dddt)2] was reported to be diamagnetic, the
consequence of a strong pairing. In the complexes described here,
the chains are indeed dimerised, but to a much smaller extend than
in [Au(dddt)2]. It might be attributed to the combined effects of the
alkyl chain segregation and the twist of the phenyl rings relative
to the metallacycles which impedes closer interactions, since the
nickel complexes exhibit very similar structures.
˚
distances at 5.16(1) and 5.81(1) A, respectively. Interaction A
˚
is characterized by shorter Au ◊ ◊ ◊ S [4.20(1) A] intermolecular
˚
distances than interaction B [Au ◊ ◊ ◊ S 4.525(9) A] but slightly
˚
˚
longer S ◊ ◊ ◊ S distances [A: 4.35(2), 4.38(2) A; B: 4.20(2) A]. The
evolution of the cT product vs. T (Fig. 13) was properly fitted
here within the Bleaney–Bower model of an isolated singlet–triplet
system (using the following spin Hamiltonian: H = -2JS1S2)27
with J/kB = -165(5) K (-115 cm-1), g = 2.0(1) and r = 0.017(5).
Note, that an equally satisfactory fit was obtained by introducing
a small inter dimer interaction, that is, with an alternated spin
Thermal behaviour
It should be recalled here that the Ni-OC12 complex was prepared
already in 1986, concomitantly by Veber28 and Ohta,16 in the hope
to find a mesomorphic behaviour. First claimed by the two authors,
the existence of mesomorphism in Ni-OC12 was then denied on the
basis of X-ray diffraction experiments on powder samples.29 The
two compounds have been examined under cross-polarisers and no
mesophase could be observed, neither in Ni-OC12, nor in Au-OC12.
Furthermore, to compare with DSC experiments reported for Ni-
OC12,16 we also investigated the thermal behaviour of Au-OC12
(Fig. 14). The two compounds exhibit a closely related behaviour,
◦
◦
with a first transition at 73 C (Ni-OC12) or 81 C (Au-OC12), a
second weak complex transition at 122–131 ◦C (Ni-OC12) or 124–
138 C (Au-OC12) followed by melting at 170.8 C (Ni-OC12) or
171 ◦C (Au-OC12).
◦
◦
Of particular note are also the large enthalpies of transi-
tion associated with the melting process, comparable to those
observed for the low-temperature transition. Indeed, for most
liquid crystals either calamitic or discotic, the enthalpy for the
crystal-to-mesophase transition is normally much bigger than the
enthalpy for the mesophase-to-liquid transition, in sharp contrast
to the behaviour of Ni-OC12 or Au-OC12, indicating that the
first transition from the crystalline state is probably a crystal-to-
crystal solid state transition in both compounds. In conclusion,
it appears that the Au-OC12 gold complex exhibits a thermal
Fig. 12 Left: view of the alternated chain running along a in Au-OC12.
Right: detail of the two different overlap interactions between rad-
ical species. The n-dodecyloxyphenyl groups have been removed for
clarity.
3058 | Dalton Trans., 2009, 3052–3061
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
General synthesis of bis[1,2-di(4-n-alkoxyphenyl)ethane-
1,2-dithiolene]nickel
The general synthesis consists, under argon atmosphere, of heating
at 110 ◦C a mixture of 4,4¢-bis(alkoxy)benzil 1a–c (5 mmol)
and phosphorus pentasulfide (10 mmol) in 5 mL of dioxane
or 1,3-dimethyl-2-imidazolidinone for 2 h. The reaction mixture
was cooled to 60 ◦C. A solution of nickel chloride hexahydrate
(2.5 mmol) in water (1 mL) was added to the reaction mixture and
reacted to 90 ◦C for 2 h in air. To the resulting mixture was added
ethanol and the precipitate was filtered, washed with ethanol and
dried under vacuum. The purification of products was carried out
by chromatography on silica gel, eluting with petroleum ether–
CH2Cl2 (7 : 3). Crystals were obtained by slow diffusion of MeOH
on a solution of the product in CH2Cl2.
Fig. 14 DSC trace of Au-OC12.
Bis[1,2-di(4-n-butyloxyphenyl)ethane-1,2-dithiolene]nickel, Ni-
OC4. Needle black crystals; 56% yield (dioxane), 85% yield
◦
1
behaviour essentially identical to the corresponding diamagnetic
nickel complex Ni-OC12.
(DMI), mp 245 C; H NMR (300 MHz, CDCl3) d 0.99 (t, J =
6 Hz, 12H, CH3), d 1.50 (st, J = 6 Hz, 8H, CH2), d 1.78 (qt,
J = 6 Hz, 8H, CH2), d 3.97 (t, J = 6 Hz, 8H, CH2), d 6.79 (d,
J = 9 Hz, 8H, CH), d 7.32 (d, J = 9 Hz, 8H, CH); 13C NMR
(75 MHz, CDCl3) d 13.85 (CH3, 4C), d 19.22 (CH2, 4C), d 31.25
(CH2, 4C), d 67.76 (CH2, 4C), d 114.30 (CH, 8C), d 130.35 (CH,
8C), d 134.20 (Csp2, 4C), d 159.84 (Csp2, 4C), d 180.38 (Csp2, 4C).
Anal. calcd for C44H52NiO4S4: C 63.53, H 6.30. Found: C 63.52, H
6.09. UV-vis–NIR (CH2Cl2): lmax. = 933 nm (e = 52 833 M-1 cm-1).
Conclusions
We have shown that oxidised neutral radical gold complexes can
be obtained directly from the sulfiding of benzils. The electron-
releasing alkoxy groups bring not only solubility but also allows
for a small electrochemical gap, also confirmed by the NIR
absorption at very low energy (1.5 mm) when compared with
analogous Ni complexes. The comparison of the X-ray crystal
structures of Ni-OC4 and Au-OC4 in one hand, Ni-OC12 and Au-
OC12 on the other hand, shows that the paramagnetic nature of
the gold complexes is probably not at the origin of the observed
dimerisation of the stacks since very similar dimerisation is also
found in the closed-shell diamagnetic neutral nickel complexes.
This important observation lets us infer that other neutral nickel
complexes which exhibit unambiguous discotic mesophases,30 can
be used as models to prepare the corresponding paramagnetic gold
complexes.
Bis[1,2-di(4-n-octyloxyphenyl)ethane-1,2-dithiolene]nickel, Ni-
OC8. Brown precipitate, 48% yield (dioxane), 68% yield (DMI),
mp 198 ◦C; 1H NMR (300 MHz, CDCl3) d 0.90 (t, J = 6 Hz, 12H,
CH3), d 1.26–1.46 (m, 48H, CH2), d 1.79 (qt, J = 6 Hz, 8H, CH2),
d 3.96 (t, J = 6 Hz, 8H, CH2), d 6.79 (d, J = 9 Hz, 8H, CH), d 7.32
(d, J = 9 Hz, 8H, CH); 13C NMR (75 MHz, CDCl3) d 14.10 (CH3,
4C), d 22.65 (CH2, 4C), d 26.03 (CH2, 4C), d 29.23 (CH2, 8C), d
29.36 (CH2, 4C), d 31.81 (CH2, 4C), d 68.10 (CH2, 4C), d 114.31
(CH, 8C), d 130.35 (CH, 8C), d 134.19 (Csp2, 4C), d 159.84 (Csp2,
4C), d 180.41 (Csp2, 4C). Anal. calcd for C60H84NiO4S4: C 68.23,
H 8.02. Found: C 68.27, H 7.91. UV-vis–NIR (CH2Cl2): lmax.
=
934 nm (e = 41 923 M-1 cm-1).
Bis[1,2-di(4-n-dodecyloxyphenyl)ethane-1,2-dithiolene]nickel,
Ni-OC12. Need◦le brown crystals; 48% yield (dioxane), 58% yield
Experimental
1
(DMI), mp 164 C; H NMR (300 MHz, CDCl3) d 0.89 (t, J =
Procedures and methods
6 Hz, 12H, CH3), d 1.27–1.47 (m, 72H, CH2), d 1.79 (qt, J = 6 Hz,
8H, CH2), d 3.96 (t, J = 6 Hz, 8H, CH2), d 6.79 (d, J = 9 Hz,
8H, CH), d 7.32 (d, J = 9 Hz, 8H, CH); 13C NMR (75 MHz,
CDCl3) d 14.11 (CH3, 4C), d 22.68 (CH2, 4C), d 26.03 (CH2, 4C),
d 29.21 (CH2, 4C), d 29.34 (CH2, 4C), d 29.39 (CH2, 4C), d 29.56
(CH2, 4C), d 29.59 (CH2, 4C), d 29.63 (CH2, 4C), d 29.65 (CH2,
4C), d 31.91 (CH2, 4C), d 68.10 (CH2, 4C), d 114.31 (CH, 8C),
d 130.34 (CH, 8C), d 134.19 (Csp2, 4C), d 159.85 (Csp2, 4C), d
180.43 (Csp2, 4C). Anal. calcd for C76H116NiO4S4: C 71.28, H 9.13.
Found: C 70.84, H 9.12. UV-vis–NIR (CH2Cl2): lmax. = 933 nm
(e = 31 300 M-1 cm-1).
All reagents were commercially available and used without fur-
ther purification. 1H and 13C NMR spectra were recorded on
Bruker Avance 300 spectrometer, chemical shifts are reported
in ppm referenced to TMS for 1H and 13C NMR. UV-vis–
NIR spectra were recorded on a Cary 5 spectrophotometer in
CH2Cl2. Cyclic voltammetry was carried out on a 1.5 ¥ 10-3
M
solution of metal complex in CH2Cl2 (anhydrous grade) containing
0.2 M n-Bu4NPF6 as supporting electrolyte. Voltammograms were
recorded at 0.1 V s-1 on a platinum disk electrode (1 mm2).
Potentials were measured vs. Saturated Calomel Electrode (SCE).
Elemental analyses were performed at the CNRS Service de
Microanalyse, Gif sur Yvette, France.
General synthesis of bis[1,2-di(4-n-alkoxyphenyl)ethane-
1,2-dithiolene]gold
Synthesis of 4,4¢-bis(alkoxy)benzyl 1a–c. These compounds
In dioxane: under argon atmosphere, a mixture of 4,4¢-
bis(alkoxy)benzil 1a–c (5 mmol) and phosphorus pentasulfide
were prepared according to literature methods.16c
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 3052–3061 | 3059
©

View Article Online
(10 mmol) in 1,4-dioxane (5 mL) was refluxed for 3 h. Then the
mixture was cooled, filtered to remove the unreacted phosphorus
pentasulfide and washed with a small portion of dioxane. A
solution of potassium tetrachloroaurate (2.5 mmol) in water
(1 mL) was added and refluxed for 2 h. The resulting precipitate
was filtered, washed with ethanol and water. The purification of
products was carried out by chromatography on silica gel, eluting
with petroleum ether–CH2Cl2 (7 : 3).
Bis[1,2-di(4-n-dodecyloxyphenyl)ethane-1,2-dithiolene]gold, Au-
OC12. Needle brown crystals; 21% yield (dioxane), 34% yield
(DMI), mp 161 ◦C. Anal. calcd for C76H116AuO4S4: C 64.33, H
8.24. Found: C 64.33, H 8.43. UV-vis–NIR (CH2Cl2): lmax.
=
1566 nm (e = 33 400 M-1 cm-1).
Crystallography
In DMI: after heating under an argon atmosphere at 110 ◦C, a
mixture of 4,4¢-bis(alkoxy)benzil 1a–c (5 mmol) and phosphorus
pentasulfide (10 mmol) in 1,3-dimethyl-2-imidazolidinone (5 mL)
for 2 h, the reaction mixture was cooled to 60 ◦C. A solution
of potassium tetrachloroaurate (2.5 mmol) in water (1 mL) was
added to the reaction mixture which was stirred at 90 ◦C for
2 h in air. To the resulting mixture was added ethanol and the
precipitate was filtered, washed with ethanol and dried under
vacuum. The purification was carried out by chromatography on
silica gel, eluting with petroleum ether–CH2Cl2 (7 : 3). Crystals
were obtained by slow diffusion of MeOH on a solution of
products in CH2Cl2.
Experimental data are collected in Table 5. Crystals were quickly
R
transferred for the mother liquor to a drop of Paratoneꢀ, a crystal
was picked up with a cryoloop and then frozen at 100 K under a
stream of dry N2 on a APEX II Brucker AXS diffractometer for X-
˚
ray data collection (MoKa radiation, l = 0.71073 A). Structures
were solved by direct methods (SHELXS-9731 or SIR-92)32 and
refined (SHELXL-97)31 by full-matrix least-squares methods as
implemented in the WinGX software package.33 An empirical
absorption (multi-scan) correction was applied. Hydrogen atoms
were introduced at calculated positions (riding model) included
in structure factor calculation but not refined. The structures of
Ni-OC4 and Au-OC4 are strongly affected by disorder (Fig. 3).
In Au-OC4, the disorder affecting the butyl chain on O4 was
described with two chains. The disorder affecting the complete
parabutyloxyphenyl group (on O2/O2A) was described with two
components. The geometry of the C13-C18/C13A-C18A phenyl
rings was constrained to a regular hexagon with the AFIX66
instruction and the C13-C17, C13A-C17A atoms were refined
isotropically only. Distances were also constrained at the end group
of the butyl chain with DFIX for C32-C33 and C32A-C33A. In Ni-
OC4, disorder was found to affect two butyl chains linked to O1
and O3. Furthermore, one complete parabutyloxyphenyl group
based on O2 was also affected by disorder (the phenyl ring C13-
C18 with the chain O2-C19-C22). Geometrical constraints on the
phenyl rings were applied (AFIX66) and atoms C13A, C15A,
Bis[1,2-di(4-n-butyloxyphenyl)ethane-1,2-dithiolene]gold, Au-
OC4. Black needle crystals; 10% yield (dioxane), 29% yield
(DMI), mp 213 ◦C. Anal. calcd for C44H52AuO4S4: C 54.48, H
5.40. Found: C 54.28, H 5.47. UV-vis–NIR (CH2Cl2): lmax.
=
1565 nm (e = 21 391 M-1 cm-1).
Bis[1,2-di(4-n-octyloxyphenyl)ethane-1,2-dithiolene]gold, Au-
OC8. Brown precipitate, 10% yield (dioxane), 23% yield (DMI),
mp 179 ◦C. Anal. calcd for C60H84AuO4S4: C 60.33, H 7.09.
Found: C 58.69, H 7.01. UV-vis–NIR (CH2Cl2): lmax. = 1565 nm
(e = 25 615 M-1 cm-1).
Table 5 Crystallographic data
Phase
Ni-OC4
Au-OC4
Ni-OC12
Au-OC12
Formula
FW
C44H52NiO4S4
831.81
C44H52AuO4S4
970.06
C76H116NiO4S4
1278.71
C76H116AuO4S4
1418.89
Crystal system
Triclinic
P1
Triclinic
P1
Triclinic
P1
9.254(7)
16.303(13)
25.290(2)
94.19(1)
103.68(2)
102.80(2)
3583.40(7)
2
1.187
100
0.42
—
—
—
—
—
Triclinic
P1
¯
¯
¯
¯
Space group
˚
a/A
9.2955(16)
13.891(3)
18.494(3)
106.023(9)
99.822(9)
107.817(8)
2098.6(7)
2
1.316
100
0.702
1.19–27.54
32 423
9508
0.0651
6871
Multi-scan
1.0, 0.812
556
0.0638
0.1094
9.3768(5)
13.8350(9)
18.8220(12)
110.519(4)
96.237(4)
107.257(3)
2121.3(2)
2
1.519
100
3.705
1.19–27.41
27 248
9513
0.0571
7868
Multi-scan
1.0, 0.615
523
0.0370
0.716
10.2099(9)
13.5720(10)
26.967(3)
86.099(11
84.742(8)
78.986(7)
3647.6(6)
2
1.292
100
2.177
2.9–27.06
68 356
15 757
0.153
8699
Multi-scan
1.0, 0.768
765
0.0642
0.0789
˚
b/A
˚
c/A
a/◦
b/◦
g /◦
3
˚
V/A
Z
D
calcd/Mg m-3
T/K
m/mm-1
q range/◦
Measured reflections
Independent reflections
Rint
I > 2s(I) reflections
Absorption correction
—
—
—
—
—
—
T
max, Tmin
Refined parameters
R(F), I > 2s(I)
wR(F2), all data
-3
˚
Dr (e A )
-0.51, 0.46
-1.24,1.09,
-0.70, 0.80
3060 | Dalton Trans., 2009, 3052–3061
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
C18A and C17B were refined isotropically only. In Au-OC12, a
disorder model was used for four terminal carbon atoms (C75A
& B to C78A & B) of one dodecyl chain and all C–C distances
had to be constrained within the disordered group. C77A & B and
C78A & B were refined isotropically only. The very thin crystals
obtained for Ni-OC12 did not allowed a complete resolution. All
atoms were identified in successive Fourier difference maps but
only heavy atoms (Ni, S) could be refined anisotropically.
8 (a) G. Rindorf, N. Thorup, T. Bjørnholm and K. Bechgaard, Acta
Crystallogr., Sect. C: Cryst. Struct. Commun., 1990, 46, 1437; (b) N. C.
Schiødt, T. Bjørnholm, K. Bechgaard, J. J. Neumeier, C. Allgeier, C. S.
Jacobsen and N. Thorup, Phys. Rev. B, 1996, 53, 1773.
9 O. J. Dautel, M. Fourmigue´, E. Canadell and P. Auban-Senzier, Adv.
Funct. Mater., 2002, 12, 693.
10 W. Susuki, E. Fujiwara, A. Kobayashi, Y. Fujishiro, E. Nishibori, M.
Takata, M. Sakata and H. H. Kobayashi, J. Am. Chem. Soc., 2003, 125,
1486.
11 P. Batail, K. Boubekeur, M. Fourmigue´ and J.-C. P. Gabriel, Chem.
Mater., 1998, 10, 3005.
12 S. Kokatam, K. Ray, J. Pap, E. Bill, W. E. Geiger, R. J. LeSuer, P. H.
Rieger, T. Weyhermu¨ller, F. Neese and K. Wieghardt, Inorg. Chem.,
2007, 46, 1100.
13 K. Ray, T. Weyhermu¨ller, A. Goossens, M. W. J. Craje and K.
Wieghardt, Inorg. Chem., 2003, 42, 4082.
14 M. Cotrait, J. Gaultier, C. Polycarpe, A. M. Giroud and U. T. Mueller-
Westerhoff, Acta Crystallogr., Sect. C: Cryst. Struct. Commun., 1983,
39, 833.
15 (a) A. M. Giroud, A. Nazzal and U. T. Mueller-Westerhoff, Mol. Cryst.
Liquid Cryst., 1980, 56, 225; (b) U. T. Mueller-Westerhoff, A. Nazzal,
R. J. Cox and A. M. Giroud, Mol. Cryst. Liquid Cryst., 1980, 56, 249;
(c) A. M. Giroud and U. T. Mueller-Westerhoff, Mol. Cryst. Liquid
Cryst., 1977, 41, 11.
16 (a) K. Ohta, A. Takagi, H. Muroki, I. Yamamoto, K. Matsuzaki,
T. Inabe and Y. Maruyama, J. Chem. Soc., Chem. Commun., 1986,
883; (b) K. Ohta, A. Takagi, H. Muroki, I. Yamamoto and K.
Matsuzaki, Mol. Cryst. Liquid Cryst., 1987, 147, 15; (c) H. Horie, A.
Takagi, H. Hasebe, T. Ozawa and K. Ohta, J. Mater. Chem., 2001, 11,
1063.
Magnetic measurements
The magnetic susceptibility measurements were obtained with the
use of a Quantum Design SQUID magnetometer MPMS-XL. This
magnetometer works between 1.8–400 K for dc applied fields
ranging from -7–7 T. Measurements were performed on poly-
crystalline samples of Au-OC4 (25.2 mg) and Au-OC12 (15.28 mg).
The magnetic data were corrected for the ferromagnetic impurity,
the sample holder and the diamagnetic contributions. EPR spectra
were obtained either on a X-band Bruker EMX-8/2.7 spectrom-
eter in Rennes or on a Radiopan SE/X spectrometer equipped
with liquid nitrogen cooling system in Poznan. Simulations were
performed with Bruker WinEPR Symphonia.
Acknowledgements
17 K. Takuma, Y. Irizato, K. Katho, PCT Int. Appl.1990, 22 pp. WO
9012019.
Financial support from the Re´gion Bretagne through a PhD grant
(to R. P.), the Re´gion Aquitaine, the European network MAG-
MANet (NMP3-CT-2005-515767), the University of Bordeaux
and the CNRS (particularly for an Associated Researcher position
to L. P.-S. in Rennes) are gratefully acknowledged. We warmly
thank Dr P. Davidson (Laboratoire de Physique des Solides,
Orsay, France) for the polarising microscope observations and
Dr T. Roisnel (CDIFX, Rennes, France) for the X-ray data
collections.
18 K.-M. Sung and R. H. Holm, J. Am. Chem. Soc., 2002, 124, 4312.
19 A. Davison, N. Edelstein, R. H. Holm and A. H. Maki, J. Am. Chem.
Soc., 1963, 85, 2029.
20 A. Davison, N. Edelstein, R. H. Holm and A. H. Maki, Inorg. Chem.,
1963, 2, 1227.
21 K. Arumugam, J. E. Bollinger, M. Fink and J. P. Donahue, Inorg.
Chem., 2007, 46, 3283.
22 M. Megnamisi-Belombe and B. Nuber, Bull. Chem. Soc. Jpn., 1989,
4092.
23 (a) J. P. M. Nunes, M. J. Figueira, D. Bello, I. C. Santos, B. Ribeiro,
E. B. Lopes, T. Henriques, J. Vidal-Gancedo, J. Veciana, C. Rovira
and M. Almeida, Chem.–Eur. J., 2007, 13, 9841; (b) W. Suzuki, E.
Fujiwara, A. Kobayashi, Y. Fujishiro, E. Nishibori, M. Takata, M.
Sakata, H. Fujiwara and H. Kobayashi, J. Am. Chem. Soc., 2003, 125,
1486.
24 H. Yokoi and M. Chikira, J. Am. Chem. Soc., 1975, 97, 3975.
25 O. Kahn, in Molecular Magnetism, VCH, Weinheim, 1993.
26 (a) W. Duffy and K. P. Barr, Phys. Rev., 1968, 165, 647; (b) J. W. Hall,
W. E. Marsh, R. R. Weller and W. E. Hatfield, Inorg. Chem., 1981, 20,
1033.
References
1 For a recent book on dithiolene complexes see: Dithiolene Chemistry,
Prog. Inorg. Chem., ed. E. I. Stiefel, vol. 52.
2 K. H. Drexhage and U. T. Mueller-Westerhoff, IEEE J. Quantum
Electron., 1972, QE–8, 759.
3 U. T. Mueller-Westerhoff, B. Vance, in Compr. Coord. Chem., ed. Sir
G. Wilkinson, FRS, Pergamon Press, Oxford, 1987, ch. 16.5 pp. 595–
631.
27 B. Bleaney and K. D. Bowers, Proc. R. Soc. London, Ser. A, 1952, 214,
4 K. L. Marshall, S. Schudel and I. A. Lippa, Proceedings SPIE. The Int.
Soc. Opt. Eng., 2004, 52B, 201.
451.
28 M. Veber, R. Fugnitto and H. Strzelecka, Mol. Cryst. Liquid Cryst.,
1983, 96, 221.
29 M. Veber, P. Davidson, C. Jallabert, A. M. Levelut and H. Strzelecka,
Mol. Cryst. Liquid Cryst. Lett., 1987, 5, 1.
30 K. Ohta, H. Hasebe, H. Ema, M. Moriya, T. Fujimoto and I.
Yamamoto, Mol. Cryst. Liquid Cryst., 1991, 208, 21.
31 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008,
64, 112.
32 A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M. C. Burla,
G. Polidori and M. Camalli, J. Appl. Crystallogr., 1994, 27, 435.
33 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837.
5 S. Kuwabara, U.S. Pat. Appl. Publ., 2005, 26, pp. US 2005035336.
6 See for example: (a) J. M. Tunney, A. J. Blake, E. S. Davies, J. McMaster,
C. Wilson and C. D. Garner, Polyhedron, 2006, 25, 591; (b) N. C.
Schiødt, P. Sommer-Larsen, T. Bjørnholm, M. Folmer, Nielsen, J.
Larsen and K. Bechgaard, Inorg. Chem., 1995, 34, 3688; (c) D. Belo,
H. Alves, E. B. Lopes, M. T. Duarte, V. Gama, R. T. Henriques, M.
Almeida, A. Perez-Benitez, C. Rovira and J. Veciana, Chem.–Eur. J.,
2001, 7, 511.
7 A. J. Schultz, H. H. Wang, L. C. Soderholm, T. L. Sifter, J. M. Williams,
K. Bechgaard and M.-H. Whangbo, Inorg. Chem., 1987, 26, 3757.
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 3052–3061 | 3061
©