Metallomesogens by ligand design

Article abstract of DOI:10.1039/b212610a

In this Perspective, some of the basic principles of ligand design will be described as they apply to the synthesis of metal complexes with liquid crystal properties. Liquid crystal properties of metal complexes are a delicate function of both ligand and metal and examples will be chosen to illustrate this statement. In particular, the effect of the metal complex moiety on the mesomorphism of a series of imine and diimine liquid crystals is described. The Royal Society of Chemistry 2003.

Full text of DOI:10.1039/b212610a

View Article Online / Journal Homepage / Table of Contents for this issue
Metallomesogens by ligand design†
Richard W. Date, Eva Fernandez Iglesias, Kathryn E. Rowe, James M. Elliott and
Duncan W. Bruce*
School of Chemistry, University of Exeter, Stocker Road, Exeter, UK EX4 4QD.
E-mail: d.bruce@exeter.ac.uk
Received 18th December 2002, Accepted 29th January 2003
First published as an Advance Article on the web 23rd April 2003
In this Perspective, some of the basic principles of ligand
design will be described as they apply to the synthesis of
metal complexes with liquid crystal properties. Liquid crys-
tal properties of metal complexes are a delicate function of
both ligand and metal and examples will be chosen to illus-
trate this statement. In particular, the effect of the metal
complex moiety on the mesomorphism of a series of imine
and diimine liquid crystals is described.
of some Ni() dithiolenes (Fig. 1). In the context of the current
article, this is most appropriate as they had designed their
metal–ligand combination in a rather specific manner. Thus,
they proposed that the central metal-dithiolene unit might be
considered equivalent to a phenyl ring and thus the complex
could be considered as being a structural analogue of a 4,4Љ-
dialkylterphenyl (which has liquid crystal properties). Since
then, the subject has grown hugely and has been the subject of
several reviews2 and a book.3
Introduction
Liquid crystals based on metal complexes (metallomesogens)
have been known since the early part of the 20th century, but
they were clearly regarded as curiosities as reports of their syn-
thesis and properties were sporadic until the early 1980s. The
renaissance of the subject is often and properly ascribed to a
publication by Giroud and Mueller-Westerhoff in 1977,1 in
which they reported the synthesis and liquid crystal properties
Fig. 1 Mesomorphic nickel() dithiolene complexes.
This article seeks to take a number of examples from the
metallomesogen literature to try to exemplify how the demands
† Based on the presentation given at Dalton Discussion No. 5, 10–12th
April 2003, Noordwijkerhout, The Netherlands.
Richard Date was born in 1963 and studied Chemistry with
Physics at the University of Southampton. After completing
an M.Sc. in spectroscopy from the University of East Anglia
he then studied for a Ph.D. in the synthesis of thermotropic
liquid crystals in the research
Eva Fernandez Iglesias was
born in Oviedo, Spain. She
graduated in Chemistry at the
University of Oviedo in 1999
after completing her final year
at the University of Exeter
under an ERASMUS agree-
ment. She stayed in Exeter to
complete her Ph.D. with
Professor Duncan Bruce,
graduating in 2002. She now
works in the chemical industry
in the south-west of England.
group of Professor G. R.
Luckhurst at the University of
Southampton. After working
for six years as a post-doctoral
research assistant for Dr Alan
Fawcett at The Queen’s
University of Belfast on the
synthesis of liquid-crystalline
polymers, he moved in 1998
to the research group of
Eva Fernandez Iglesias
Professor Duncan Bruce at the
University of Exeter where
he has been working on the
synthesis of metallomesogens.
Richard Date
Duncan Bruce graduated from the University of Liverpool in 1981 and remained there for his Ph.D.
under the supervision of David Cole-Hamilton, during which time he studied phosphine complexes of
Pt(II) and Rh(I) as potential photocatalysts for the decomposition of water. In 1984, he took up a
Temporary Lectureship in Inorganic Chemistry at the University of Sheffield and, in 1986, was
awarded a Royal Society Warren Research Fellowship, which he held there until 1991. He was then
appointed Lecturer in Chemistry and was promoted to Senior Lecturer in 1994, in which year he
became co-director of the Sheffield Centre for Molecular Materials. In 1995, he was appointed
Professor of Inorganic Chemistry at the University of Exeter, the position he now holds. Professor
Bruce has research interests in different aspects of materials chemistry, central to which is an abiding
passion for the materials chemistry of coordination compounds and, in particular, their liquid crystal
properties. He was awarded the first Young Scientist prize by the British Liquid Crystal Society in
1990, was RSC Sir Edward Frankland Fellow in 1994/95 and was awarded one of the Corday-Morgan
Medals and Prizes in 1996. He was a founding member of the RSC’s Materials Chemistry Forum and
is a former member of the Dalton Council.
Duncan Bruce
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
1914

View Article Online
of liquid crystal design may be met, and also challenged, by
metal–ligand combinations. In order to make this article self-
contained, a brief introduction to liquid crystals will be given
here.
ily have to have liquid crystal properties. However, something
which is mesomorphic actually shows liquid crystal behaviour.
Nematic mesophases
The nematic phase (abbreviated, N) is the most disordered
type of liquid crystal mesophase, and in the context of
thermotropic liquid crystals, we will identify three types of
nematic phase, namely that derived from rod-like molecules
and two derived from disk-like molecules. In the first
(Fig. 2a), the phase comprises rod-like molecules which have
long-range, orientational order in one direction, but no
positional order, the orientational order being about the
unique, long axis. This is the phase used in the vast majority of
liquid crystal displays.
All nematic phases share the feature that they are organised
through a one-dimensional orientational order of molecules
about their unique axis, and so it is not surprising, therefore,
that there exists a nematic phase of disk-like molecules (Fig. 2b)
which is, unfortunately, known commonly as the nematic
discotic phase (ND); unfortunate because phases ought to be
characterised by their symmetry and not by the shape of the
molecules of which they are comprised. The final example is the
so-called Ncol phase—a nematic phase of columns (Fig. 2c)
which is occasionally found in the mesomorphism of disk-like
mesogens.‡
A brief introduction to liquid crystals
The liquid crystal state (known as a mesophase) represents a
discrete state of matter which exists between the solid and
liquid states; like a liquid it is fluid and yet like a solid, it has
order. This combination of order and fluidity results in meso-
phases with anisotropic physical properties and it is these
which have led to their widespread application in, for example,
low-power-consumption displays. Liquid crystals are first
classified according to the method used to destroy the order of
the solid state, of which there are two possibilities. In lyotropic
liquid crystals, this order is destroyed by the action of solvent,
perhaps in conjunction with heat, and typical examples of
mesomorphic materials would include amphiphilic materials,
certain polymers or disk-like materials with highly polar
peripheries.4 Such materials will not be considered further.
In thermotropic liquid crystals, however, transitions between
the solid, liquid and liquid crystal states are effected by the
action of temperature and, on heating, a typical material might
show transitions from solid to liquid crystal to isotropic liquid
(the adjective ‘isotropic’ is added in recognition of the fact that
the liquid crystal is an anisotropic liquid). Transitions between
the solid state and mesophases are typically strongly first order
from a thermodynamic point of view and often supercool,
while transitions between a mesophase and the isotropic liquid
are weakly first order. Transitions between mesophases are
either weakly first order or, occasionally, second order. The sin-
gle most important criterion for the formation of liquid crystal
mesophases is that the molecules should be structurally aniso-
tropic for this introduces additional, anisotropic dispersion
forces between the molecules which are strong enough to stabil-
ise the mesophase. In the context of this article, this anisotropy
can be a delicate interplay of metal and ligand.
Smectic mesophases
Smectic phases are layered phases which, in addition to the
orientational order found in nematic mesophases, offer a degree
of translational order, too. While there are several such phases,
only two are described here, namely the smectic A (SmA) and
smectic C (SmC) phases (note that the labels, A, C etc. are
nothing more than labels and do not carry implicit structural
information). Thus, the SmA phase is the most disordered type
of smectic phase and in it, the molecules are organised per-
pendicular to the layer direction as indicated in Fig. 3a. In the
SmC phase (Fig. 3b), the organisation is similar except that the
molecules are now tilted at some angle, θ, to the layer normal;
chiral versions of this phase have C2 symmetry in the layers
and as such, are ferroelectric—a property which allows their
exploitation in display devices. It is important to note that in
both the SmA and SmC phase, the ‘layers’ as shown in the
figure are not well-defined and in fact, the organisation is best
described as a sinusoidal molecular density wave. These phases
Within thermotropic liquid crystals, there are many classifi-
cations according to shape and also, occasionally, the meso-
phases that they form. For many years, only two classifications
were necessary, namely calamitic (rod-like) and discotic (disk-
like), as this effectively described most, if not all, of the known
mesogens. However, in addition we now have to take into
account other structural motifs such as bowlic liquid crystals,5
dendrimeric mesogens,6 polycatenar mesogens7 and bent-core
mesogens.8 To help the first-time reader in the subject, it may be
helpful also to know that the term mesogenic refers to some-
thing which is ‘liquid-crystal like’ but which does not necessar-
‡ Interestingly, this phase exists in both a thermotropic and lyotropic
form.
Kathryn Rowe obtained her
Ph.D. on the synthesis of metal
containing liquid crystals from
Sheffield University in 1997
under the supervision of
Professor Duncan Bruce. She
then joined Unilever Research
where she became a Global
Project Leader developing new
antiperspirant product forms.
In 2003 she joined the New
Product Development group at
GSK as a senior formulation
James Elliott’s first job was
with Omega Diagnostics in
Scotland, where he was part of
the team that won the Queen’s
export award in 1994. He
graduated
from
Glasgow
Caledonian University in 1995
and completed his Ph.D. thesis
in inorganic liquid crystals at
the University of Hull, under
Dr John R. Chipperfield. His
first postdoctoral position was
with Professor Duncan Bruce
at the University of Exeter,
and he is currently working
with Professor Ekk Sinn at the
University of Missouri-Rolla
in the USA.
scientist
responsible
for
managing multiple projects
that deliver cosmetic and OTC
products to the consumer.
Kathryn Rowe
James Elliott
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1915

View Article Online
Fig. 2 Structure of three nematic phases: (a) nematic phase of rod-like molecules, N, (b) the ND phase, (c) the Ncol phase.
phases were commonly, but incorrectly, referred to as ‘dis-
cotic’—D, for several years. However, we now appreciate that
the columns may be formed by a variety of different molecular
types, and it is the arrangement of these columns in two
dimensions which described the structure of the phases.
The most common types of columnar phase are the hex-
agonal (Colh), the rectangular (Colr) of which there are more
than one type according to symmetry (e.g. P2gg and C2mm
)
and the oblique (Colo). Fig. 4 shows (a-i) a side-on view of
the columnar hexagonal phase and then ‘aerial’ views of the
hexagonal (a-ii), oblique (b) and rectangular (c and d) lattices.
In this figure, circles represent disks which are orthogonal
within the columns, whereas ellipses represent disks which are
tilted.
Fig. 3 Schematic structure of the SmA (a) and SmC (b) phases.
are also very fluid and there is a very high probability of
molecules moving between layers.
Columnar mesophases
Disk-like metallomesogens forming columnar phases
As the name implies, the basic building block for this type of
mesophase is the column and in very many cases, the columns
are formed from disk-like molecules—one reason why the
The aim here is simply to design a metal complex whose shape
will approximate to a disk and then surround it with a sufficient
number of peripheral alkyl(oxy) chains so that a columnar
Fig. 4 Structure of the common columnar phases.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1916

View Article Online
Fig. 5 Mesomorphic porphyrins and phthalocyanines.
mesophase will be stabilised. Thus, for example, there are
species forming columnar phases appear to require six or more
peripheral chains.
very many examples of mesomorphic metallophthalocyanines
known where the basic phthalocyanine core has been elabor-
ated either with flexible chains (normally eight) or with more
complicated organic groups which in their turn carry flexible
chains. Examples are given as Fig. 5 and it should be noted that
while β-substituted porphyrins (Fig. 5(d)) do form columnar
phases, meso-tetraphenylporphyrins do not, forming either
crystalline or ‘discotic lamellar’ (DL) phases—the latter being a
disordered, layer-like arrangement of disk-like molecules. This
raises a design question which requires some understanding of
the rôle of the peripheral chains.
This is rather well exemplified by considering a series of
β-diketonato complexes of Cu() (Fig. 7) where the four-
chained species is not mesomorphic, rather forming a crystal-
line lamellar phase,10 while the eight-chain complex does indeed
form liquid crystal columnar phases. This is also true for
Thus, the presence of the chains serves to enhance the aniso-
tropy of the system and the fact that they are flexible and,
therefore, often do not crystallise easily leads to a lowering of
the otherwise high melting point of the disk core. However, the
number and position of the chains is not unimportant and there
is a somewhat unquantified notion that the chains should
somehow properly cover the periphery of the molecule which
means that with the exception of the pseudo-polymeric colum-
nar phases of tetra(µ-alkanoato)dimetal() complexes (Fig. 6),9
Fig. 6 Structure of the repeat unit in the mesophase of copper()
alkanoate mesogens.
Fig. 7 Mesomorphic β-diketonates.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1917

View Article Online
by Ringsdorf and Lattermann.16 Thus, while Ringsdorf had
found that the use of 4-alkoxybenzoyl, 3,5-dialkoxybenzoyl and
3,4,5-trialkoxybenzoyl group (Fig. 10, R = I, II, IV) did not lead
to mesomorphic materials, Lattermann found that the use of
a 3,4-dialkoxybenzoyl group (Fig. 10, R = III) did. These
materials also illustrate the fact that the central unit of a
discotic mesogen does not have to be planar, and there are
examples of mesomorphic hexaazacrown ethers which also
testify to this.17
The interplay between ligand and metal is also a major factor
in determining mesomorphism and this is illustrated with two
examples. Thus, phthalocyanines have been described where the
central metal moiety is a dihydroxymetal() unit based on
either Si or Sn.18 The materials are mesomorphic, but it is also
possible to effect a polymerisation of the materials via a con-
densation reaction to realise materials which are linked spinally
by –O–M–O–M– units, changing the nature of the liquid
crystal behaviour from columnar to lamellar.
Fig. 8 Mesogenic dithiolene mesogens.
structurally similar Ni() dithiolene complexes (Fig. 8).
Thus, while the four-chained derivatives (X = H) were claimed
initially to show columnar mesophases,11 this was shown sub-
sequently to be a mis-assignment12 and mesomorphic deriv-
atives were not realised until four, further chains were added
(Fig. 8, X = OCnH2n ϩ 1O).13
Another feature of phthalocyanines is, of course, the cavity
size at the centre of the ring and inclusion of large metals leads
to non-planar structures where the metal sits out of the plane.
This is the case found in phthalocyanine derivatives of Pb()
and the effect was to modify the arrangement in the mesophase
to give an antiferroelectric arrangement (Fig. 11) which was the
most efficient way for the complexes to organise and fill space.19
Just before leaving this small section describing β-diketon-
ates, it is worth contemplating the two complexes shown as
Fig. 9. Curiously, the more rod-like complexes (A) show colum-
nar phases,14 while the more disk-like complexes B show only
mesophases typical of rod-like materials!15
However, there is a subtlety to the notion of ‘covering the
periphery’ as exemplified by derivatives of [14]aneN4 described
Fig. 11 Antiferroelectric arrangement of Pb() phthalocyanines in the
mesophase.
Finally here, this discussion of discotic systems would not be
complete without mention of some rather beautiful complexes
in which the disk is assembled by forming multinuclear
complexes. Thus, Barberá et al. described20 some trinuclear
complexes of Au() which formed what was described as a
‘metallacrown’ using disubstituted pyrazole ligands (Fig. 12).
Both the complexes with X = H and X = OC10H21 were meso-
morphic showing columnar hexagonal phases close to room
temperature, although the latter complex had lower transition
temperatures on account of the fact that it exists as isomers
(relative orientation of the pyrazoles around the gold triangle)
and that the extra chains cannot be accommodated in the
molecular plane and so must reduce the structural anisotropy.
Another polynuclear system was described by Praefcke and
co-workers based on ortho-metallated complexes of Pd().21
The formation of liquid crystals based on ortho-metallation is
well known in calamitic systems (vide infra), but in these
Fig. 9 Rod- and disk-like β-diketonate complexes.
Fig. 10 Mesomorphic [14]aneN4 compounds.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1918

View Article Online
the dithiolenes shown in Fig. 1 illustrate this well. These
complexes, along with the dithiobenzoates shown in Fig. 14,22
illustrate an important effect that complex formation can have,
for while the complexes do form liquid crystal phases (and at
elevated temperatures for the dithiobenzoates), the constituent
ligands are not mesomorphic.
Fig. 14 Mesomorphic metal() dithiobenzoates.
However, this need not be the case and, for example,
complexes of the cyanobiphenyl mesogens (used in commercial
displays) have been complexed to Pd() and Pt() to give
mesomorphic materials (Fig. 15).23 An interesting aspect of this
work was that while the ligands show only N and SmA phases,
some complexes showed, in addition, a SmC phase demon-
strating that coordination to form a new species can change the
mesomorphism of the parent ligand.
Fig. 12 Mesogens made of ‘golden crowns’.
materials, the approach was cleverly extended by using larger
ligands bearing two sites for ortho-metallation, allowing the
formation of large, macrocyclic systems as shown in Fig. 13.
In these complexes, there was a marked difference between
the melting points (temperature at which the solid melts either
to mesophase or isotropic) and the clearing points (temperature
at which the liquid crystal phase becomes isotropic) when
comparing Pd and Pt. This is not expected from theory and
was attributed24 to weaker, perhaps more labile interactions
between the Pd and the organonitrile. Takahashi25 later studied
the directly analogous isonitrile materials for Pd and Pt with
bromide and iodide ligands and found that while iodo com-
plexes were mesomorphic for both metals, bromo complexes
were mesomorphic only for Pt. However, here the melting
and clearing points were very similar for the same combin-
ation of metal, halogen and ligand, and in these systems,
the metal–ligand (isonitrile) interaction will be very much
stronger than in the case of the nitriles, supporting the
proposal of weak Pd–nitrile interactions in complexes of
cyanobiphenyls.
There is here a curious comparison to be made between the
dithiobenzoate and the cyanobiphenyl complexes, for in the
former case, linking two, one-ring ligands (which are not
mesomorphic) together about the metal centre leads to meso-
morphic complexes and, therefore, it is not surprising that
linking the mesomorphic cyanobiphenyls about a metal has a
similar effect. By analogy then, one might expect that it would
be possible to use, for example, 4-alkoxybenzonitriles as ligands
to Pd() as the resulting complexes (Fig. 16) would be suffi-
ciently anisotropic to show liquid crystal behaviour. In fact, this
is not the case, yet if more chains are added to the benzonitrile
(Fig. 16), disk-like materials result which show columnar
phases;26 related isonitriles behave similarly.27
Fig. 13 Large, tetrametal macrocycles.
These complexes have a wide and varied mesomorphism. For
example, thermotropically many show columnar hexagonal
phases and yet in those materials which do not, it is often the
case that addition of the electron acceptor trinitrofluorenone
can induce a mesophase. Use of the chiral electron acceptor
2Ј-(2,4,5,7-tetranitro-9-fluorenylideneaminoxy)propionic acid
(fortunately know as TAPA!) resulted in chiral phases. Further,
in collaboration with Usol’tseva, it was found that the materials
would show lyotropic mesophases in solvents such as alkanes.
Simple rod-like metallomesogens from simple
rod-like ligands
However, it is also observed that simple, linear complexes of
d10 metal centres can show liquid crystal properties if acetylene
ligands are used28 or if complexes are formed from acetylene
and isonitrile ligands (Fig. 17),29 which suggests that in the
simple, calamitic alkoxybenzonitrile complexes (Fig. 16), the
two chloride ligands are sufficient large to reduce the aniso-
tropy and suppress mesomorphism. This supposition finds
some support in the X-ray single crystal structure30 of the
hexylcyanobiphenyl complex of Pt() (Fig. 18) which shows the
steric extent of the chloride ligands.
In principle, the design of calamitic metallomesogens should be
no more difficult than that of discotic systems given that it is
know that a rod-like motif is required. However, experience
shows that rod-like systems are a little more subtle, something
which it is hoped will emerge from the following discussion.
Many rod-like systems have been, and continue to be, based
on metals in Groups 9, 10 and 11 which are capable of forming
complexes with a square-planar or linear geometry, and indeed
Fig. 15 M() complexes of cyanobiphenyls.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1919

View Article Online
coordination chemistry. The ligand studied most extensively
is the 4-alkoxystilbazole (Fig. 20)32,33 and its mesomorphic
complexes will be reviewed briefly. It is important to note
that the ligand itself shows a crystal smectic E phase (an
orthorhombic phase where rotation about the long axes is con-
certed within layers and where there is correlation between the
layers—not a true liquid crystal phase) between around 80 to
90 ЊC and despite our initial reports of a SmB phase, probably
no other mesophase.
Fig. 20 4-Alkoxy-4Ј-stilbazole.
Pyridines can bind readily to M() (M = Pd, Pt) and so com-
plexes analogous to those of the cyanobiphenyls were syn-
thesised (Fig. 21A). The complexes of both metals melted at
high temperatures (T > 200 ЊC), but for palladium, this led
simply to decomposition (perhaps with transient phase form-
ation), while for platinum, SmC phases could be observed prior
to decomposition. Two strategies were then identified to reduce
the transition temperatures of these materials, both of which
involved employing different co-ligands. Thus, it was possible
to replace the chloride ligands with alkanoate ligands (in fact,
the materials were synthesised by first obtaining the desired
metal() alkanoate and then reacting with stilbazole) which
led34 to complexes (Fig. 21B) in which flexible, lateral chains
extended alongside the rigid core of the mesogen. In purely
organic systems,35 such a design invariably leads to materials
which show a nematic phase as the lateral chain(s) prevent
association into smectic layers and these complexes were no
different showing nematic phases at temperatures of around
150 to 160 ЊC.
Fig. 16 Alkoxy-substituted benzonitriles and benzoisonitriles.
Fig. 17 Small, two-ring complexes of d10 metals.
Fig. 18 Molecular structure of trans-[PtCl2(NC–Ph–Ph–C6H13)2].
In the second strategy, the anisotropy and symmetry of the
system were lowered by making an unsymmetric complex con-
taining both a stilbazole and a coordinated alkene (Fig. 21C).
These materials invariably showed a SmA phase at temper-
atures normally below 100 ЊC.34,36
Of course, other effects can come into play and there is some
evidence for the action of aurophilic interactions in promoting
mesomorphism in certain gold() and gold() complexes. Thus,
Fig. 19 shows some gold complexes31 which have liquid crystal
properties and which are very much ‘shorter’ than one might
expect a compound to be to form a mesophase.
Monoligand complexes were also available with Rh()
and Ir(), and here the materials were obtained by the reaction
of the ligand with [M2(µ-Cl)2(COD)2] in CH2Cl2 under CO
(Fig. 22).37 The complexation had a significant effect on the
ligand and now the complexes showed N and SmA phases at
temperatures between around 85 to 135 ЊC (Fig. 23); such
behaviour is typical of small, dipolar mesogens and it is clear
that the cis-[MCl(CO)2] unit has had a very positive effect on
the mesomorphic behaviour. This is attributed to the increased
linear polarisability of the complex compared to the ligand and
also the enhanced anisotropy of that polarisability.37–39 Thus,
for example, the linear polarisability of the ligand nearly
doubles on complexation.
Fig. 19 Small, mesomorphic complexes of gold.
While nitriles will ligate a number of metal centres, pyridines
often bind more strongly and so there was interest in their
Fig. 21 Stilbazole complexes of Group 10 metals.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1920

View Article Online
Fig. 22 Group 8 complexes of 4-alkoxystilbazoles.
Fig. 23 Phase diagram for the Ir() complexes of alkoxystilbazoles.
Complexes of ortho-metallated ligands
Fig. 25 Pd() imine complexes.
In one of the earlier contributions to the study of metallo-
mesogens, Ghedini and co-workers adopted a systematic
approach and complexed a series of azobenzene liquid crystals
to Pd() via ortho-metallation reactions, forming a series of
dimeric and monomeric complexes with the general structure
shown in Fig. 24.40
fragments might promote additional ordering. This assumption
was found to be true in the solid state from a single crystal
X-ray structure and 2H NMR studies in the nematic phase
showed that while an Nb phase was not found, the data sug-
gested that the design criteria adopted were valid and that the
phase behaviour was not that of a ‘conventional’ nematic
phase, but showed evidence of the behaviour predicted for the
Nb phase.42
Azine ligands also proved amenable to this chemistry
(Fig. 26).43 In the complexes where X = Cl and Br, only the trans
Fig. 24 Ortho-metallated Pd() complexes of azobenzene mesogens.
The dimeric complexes in particular are interesting for
examination of the ‘core’ of the molecule suggests something
disk-like, yet the distribution of chains is clearly insufficient to
stabilise a columnar phase. Further, the mesomorphism is un-
equivocally that associated with rod-like molecules with N and
SmA phases being seen and this has caused some interest in the
physical properties of these materials since their initial discovery.
Such a general motif suggests scope for structural modifi-
cation and some of these are described briefly now. Thus, azo
compounds can be replaced by imines which can also ortho-
metallate and, of course, monovalent, bidentate ligands can be
used to cleave the dihalogen bridge to give a range of com-
plexes, exemplified by those shown in Fig. 25.41 The complex
shown as A showed a SmC phase and so by making the ligand
chains chiral, it was possible to realise materials which showed a
chiral SmC phase (labeled SmC*). As described above, such
phases can exhibit ferroelectric response and ferroelectric
switching was observed in the millisecond time domain. The
complex shown as B was designed in attempts to realise a
material with the biaxial nematic phase, Nb. This phase is like
that shown in Fig. 2(a), but has, in addition, long-range corre-
lations perpendicular to the main director, . The complex
possess a hydrocarbon CH3 group and a fluorocarbon CF3
group and it was envisaged that the incompatibility of these two
Fig. 26 Ortho-metallated Pd() azine complexes.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1921

View Article Online
isomer was observed (by 1H NMR spectroscopy), while for
X = SCN, two isomers were observed in a 60 : 40 ratio and it
was assumed that these were cis and trans. In each of these
examples, the complex was assumed to be planar by com-
parison with related structures in the literature.44 However,
where X = OAc, the situation was more complex and trans and
Table 1 Thermal behaviour of metal salicylaldimine complexes (2)
M
n = m
Transition
T /ЊC
∆H/kJ molϪ1
Pd
Pd
10
16
Cr–Colh
Colh–I
Cr–Colh
Colh–I
99
104
52
128
5.6
134
3.6
1
cis isomers in the ratio 3 : 1 were produced consistently. H
111
NMR studies went on to show that the trans isomer was optic-
ally active and hence, the structure had to be that of an ‘open
book’ (Fig. 26, R = Me), although in the synthesis a racemic
mixture was produced.
b
Cu
Cu
10
16
Cr–I
Cr–Colh
Colh–I
93
57
102
135
4.5
For the non-acetate-bridged dimers, the only mesophase seen
was SmC which was typically in the range 102 to 249 ЊC (for
R = C9H19) while for the acetate-bridged complexes, a nematic
phase was seen for C6H13 ≤ R ≤ C8H17 and for R ≥ C7H15, a SmC
phase was seen in addition. Mesophase ranges were very much
larger in the planar materials.
b
b
Ni
Ni
10
16
Cr–I
Cr–I
77
90
b
Fe–Cl
Fe–Cl
10
16
Cr–Colh
Colh–I
68
172
160
15
16
Colh–Ia
a Melting point not seen as the material is obtained as a glass and forms
a glass on cooling. b Measurement not possible due to compound
cooling into glassy state.
Using the fact that the trans-isomer was chiral, a derivative
was synthesised where the bridging carboxylate group was the
optically pure (R)-2-chloropropionate.45 Synthesised from the
µ-Cl2 species by reaction with the sodium salt of the acid, a
1
mixture was produced which was shown by H NMR to have
the following composition: trans-ΛR,R (34%), trans-∆R,R
(34%), cis-R,R (32%). Thus, while the trans-components
described a pair of diastereoisomers, the cis isomer was optic-
ally pure by virtue of the chiral acid groups. The mixture so
produced had the phase sequence:
It is not proposed to discuss the rod-like derivatives in any
detail, but one or two points are pertinent. First is that in con-
sidering the structures of some derivatives, particularly those
with N-aryl groups, the position is similar to that found in the
basic ortho-metallated palladium complexes as the ‘molecular
core’ is almost disk-like, yet the phases shown are typically N,
SmA and SmC. Second is the observation that these ligands
gave the first examples of calamitic, paramagnetic materials
through the work of Galyametdinov, and several EPR studies
have subsequently been carried out. Third is that derivatives
showing columnar phases53 appear effectively non-existent,
save for two isolated reports (Fig. 29).54
Cr ؒ 102 ؒ SmC* ؒ 119 ؒ SmA ؒ 149 ؒ I
and physical measurements showed that the SmC* phase was
ferroelectric, with a rise time of around 330 ms at a square wave
voltage of 17 V and 0.5 Hz and a cell thickness of 11 µm. Such
response times were some three orders of magnitude longer
than those found in SmC* phases of more conventional, calam-
itic, organic materials, something probably due to the greater
viscosity of the palladium complexes which in turn resulted
from the molecular shape.
However, examination of these ligands shows that they
are readily functionalised as the ‘synthetic deconstruction’ in
Fig. 30 illustrates. Therefore, we undertook the synthesis of
some more heavily substituted derivatives in order to realise
columnar phases, details of which are presented in Scheme 1.
The material targeted used a combination of 4-alkyloxy-
benzoate functions attached to the aniline ring and 3,4,5-
trialkyloxybenzoate functions attached to the salicylaldehyde
ring. In the scheme, the reduction of the nitrobenzene to the
aniline may also be accomplished using hydrogen over Pd/C, or
using hydrazine over Pd/C. Complexation of the ligands was
achieved by reaction with the metal acetate (M = Pd, Cu, Ni) or
chloride (Fe). All new materials were obtained pure as evi-
Complexes of ortho-metallated ligands and of
salicylaldimines
If, arguably, the cation [Ru(bipy)3]2ϩ is the drosophila fly of
coordination chemistry (at least in terms of photochemistry),
then metal complexes of salicylaldimines perform a similar
function in the study of metallomesogens. The first reports of
mesomorphic derivatives were made by Galyametdinov who
was followed by Ghedini, Serrano, Hoshino and then many
others. A generic structure is shown as Fig. 27 and the literature
up to about 1998 is well described in a review by Hoshino.46
1
denced by CHN analysis and H and 13C NMR spectroscopy
(except for those containing paramagnetic metal centres).
For all ligands synthesised (1), the chain lengths of the two
parts of the molecule were the same (n = m) and it was found
that none of the ligands was mesomorphic, melting to the iso-
tropic state at temperatures between 58 ЊC (n = m = 16) and 168
ЊC (n = m = 1). In considering the metal complexes (2), chain
length was clearly a major factor in destabilising the crystal
phase to reveal the liquid crystal phases, for none of the com-
plexes with n = m = 1 or 4 was mesomorphic and melting points
were between ca. 160 and 290 ЊC. It was only at much longer
chain lengths that any mesophases were seen and, then, not for
each metal. The thermal behaviour of derivatives with n = m =
10 and 16 is summarised in Table 1.
All of the mesomorphic complexes showed a columnar hex-
agonal phase which was readily identified by its optical texture
(Fig. 31). In comparing the behaviour of the different com-
plexes, several points are worthy of further consideration. Thus,
none of the Ni() complexes was mesomorphic but the very
broad nature of the NMR spectra indicated that they were
paramagnetic and hence, did not exist in a planar geometry.
The most likely alternative would be tetrahedral or some
twisted state on the way from square planar to tetrahedral,
Fig. 27 Basic structure of metal complexes of salicylaldimines.
The beauty and attractiveness of these systems is that they
are readily and widely variable and that they can form stable
complexes with a range of metals. For example, Ni, Cu, V᎐O,
᎐
Fe–Cl and Pd have all been incorporated and, more recently,
work by Galyametdinov47 followed by ourselves48 and then
Binnemans,49 has shown that they can act as ligands to the full
range of lanthanide elements. A range of representative struc-
tures for these materials is given in Fig. 28,46,50,51 and certain
Pd() derivatives have recently been shown to have good
photorefractive properties.52
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1922

View Article Online
Fig. 28 Illustrative metal salicylaldimine complexes.
Fig. 29 Salicylaldimine complexes showing columnar phases.
leading to bulky structures with little or no anisotropy, sup-
That only one copper complex is mesomorphic is not so easy to
pressing totally liquid crystal phase formation. This is consist-
ent with the observed melting points which are lower than
might have been expected by comparison with, for example, the
clearing points of the Cu() complexes (for related materials
where one is mesomorphic and the other not, it is appropriate
to compare the melting point of the non-mesomorphic deriv-
ative with the clearing point of its mesomorphic counterpart).
explain as the thermal properties are similar to those of the
palladium complexes, suggesting that the decyloxy derivative
might have been expected to be mesomorphic.
The palladium complexes were square planar (good NMR
spectra were obtained) and here, both complexes showed a
columnar hexagonal phase. Interestingly, lengthening the chain
from decyloxy to hexadecyloxy significantly destabilised the
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1923

View Article Online
Fig. 30 ‘Synthetic deconstruction’ of mesomorphic metal salicylaldimate complexes.
Scheme 1 Synthesis of the salicylaldimine ligands. i) CnH2n 1Br/K2CO3/DMF/∆; ii) KOH/EtOH/∆; iii) DCC/DMAP/molecular sieves/THF;
ϩ
iv) SnCl2ؒ2H2O/iPrOH/∆; v) EtOH/Hϩ(cat.)/∆; vi) [M(OAc)2] (M = Pd, Cu, Ni) or FeCl3/iPrOH/∆.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1924

View Article Online
ation number metal centres might be incorporated into liquid
crystals while retaining the structural anisotropy necessary for
mesophase formation. In order to pursue this discussion, it is
necessary to introduce one more aspect of liquid crystal
behaviour, namely the mesomorphism of polycatenar liquid
crystals.
Polycatenar liquid crystals
Polycatenar mesogens54b,57 are based on a calamitic motif
(Fig. 33), having a rather extended, central core which termin-
ates in several alkoxy chains, hence giving rise to the name
polycatenar (literally many-tailed). The nomenclature of poly-
catenar systems depends on the number of terminal chains, so
that those with four chains are termed tetracatenar, those with
six chains hexacatenar and so on.
Fig. 31 Optical texture of the columnar hexagonal phase of 2
(M = Pd).
crystal phase (as it did with the copper derivatives), yet stabil-
ised the mesophase. Finally here, we note that the chloroiron()
complexes show the most stable mesophases of all, with the
Colh phase clearing above 150 ЊC. Of course, these complexes
are not planar, rather being based on a square pyramidal geom-
etry. As such, the complexes might be considered as ‘bowlic’ in
nature, implying a stacking as indicated in Fig. 32 which would
provide additional sterically driven interactions between the
molecules, further stabilising their organisational state. Such
arrangements have been proposed before in calixarene com-
plexes of W()55 and in oxovanadium() mesogens,56 for
example.
Fig. 33 Schematic representation of polycatenar mesogens showing
different potential terminal ring configurations.
The number and arrangement of the chains has a profound
effect on the mesomorphism and emphasis will be given here to
tetracatenar mesogens in which the terminal chains are in the
3,4-positions of the terminal rings.
For pentacatenar and hexacatenar mesogens, the dominant
phase is columnar and this mesomorphism arises as the
extended and numerous terminal chains occupy a very large
volume, preventing the molecules existing within a lamellar
phase as they would be unable to fill the space (Fig. 34).
Fig. 32 Schematic drawing of the packing in bowlic mesogens.
Thus, with a versatile ligand base, it is possible to realise a
variety of complexes not only from the point of view of includ-
ing several metals, but also in terms of the chemical modifi-
cation which is possible allowing properties to be tuned from
rods to disks.
In contemplating the mesomorphism of the Fe() complexes
above, we might begin to consider how, if at all, high coordin-
Fig. 34 ‘Cartoon’ to show the poor space-filling ability of hexacatenar mesogens.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1925

View Article Online
However, experimental observations are consistent with the
idea that perhaps three of these molecules might come together
to describe a circular cross-section (Fig. 35), many of which
could stack up to give columns which would organise into a
mesophase. The orientation of the molecules from one ‘slice’ of
column to the next is random.
Fig. 36 High-coordinate Pd–Cp complexes.
mesophase formation, for the introduction of the ligand(s)
which increase the coordination number (Cp for Pd and two
CO ligands for Mn/Re) reduces this anisotropy greatly. The
only way that the effect of this reduction can be overcome is to
make the ligands more anisotropic so that these ligands have a
proportionately smaller, negative effect.63
Another important point in relation to all these materials is
that they show only nematic phases and this bears comment.
Perhaps the fullest comparison62a is for the Re() complexes
with four rings in the ligand (Fig. 37A; M = Re) where the
mesomorphism was determined for a large number of ligands
as both n and m were varied. Typically the ligands showed the
following phase sequence:§
Fig. 35 Schematic representation of hexacatenar mesogens in the
columnar phase.
The reason for the interest in the tetracatenar systems men-
tioned is that short-chain-length homologues show nematic
and/or SmC phases, while longer-chain homologues show
columnar phases—that is to say the mesomorphism of rods and
disks can be combined into a single, homologous series. Thus,
when the chain length is short, the materials resemble calamitic
materials and so N or smectic phases are expected. However, as
the chains lengthen or as the temperature increases, the volume
the chains occupy increases and so, as for the hexacatenar
systems above, they cannot organise into lamellar phases and
so columnar phases result. The nature of this transition is
discussed elsewhere.58
Cr ؒ J ؒ SmI ؒ SmC ؒ N ؒ I
in the overall range of about 90 to 300 ЊC, whereas all of the
complexes showed only the nematic phase and normally in the
range 120 to 180 ЊC. The suppression of the smectic phases (in
these and the ligands of the Pd complexes) and the reduction
in transition temperatures reflect the introduction of a group
which perturbs the anisotropy of the system. Thus, mesophase
stability in less anisotropic systems is reduced and the fact that
the perturbation is located as a kind of lateral group prevents
smectic phase formation, leaving the nematic as the only one
seen.
In order to see how general such arguments might be, it is
necessary to consider a small collection of high-coordination-
number systems and these are now discussed. From our own
work with the Mn() and Re() imines, we were keen to establish
whether the use of more extended ligands would amount to a
general approach to realising mesomorphic complexes where
the metal was six coordinate. Thus, a series of 1,4-diazabuta-
diene complexes was synthesised and then complexed to Re()
as shown in Fig. 38.64
High coordination number systems—interplay
between metal and ligand
Calamitic systems
This discussion commences in considering the palladium-imine
complexes described above and here, ortho-metallation was
used to generate metallomesogens in which the metal was
formally high-coordinate due to the inclusion of a Cp ring
(Fig. 36).59,60
Ortho-metallation is also known for Mn() and Re() and
imine ligands were employed to make mesomorphic complexes
of the octahedral, d6 ions; representative structures are shown
in Fig. 37.61,62
The mesomorphism of ligand and complex are exemplified in
the following example:
One thing which characterises both the Pd–Cp complexes
and these metal tetracarbonyls is that the ligand systems are
somewhat more extended than the other ortho-metallated sys-
tems described, and groups that reported the high-coordin-
ation-number systems observed that for smaller, two-ring
ligands, mesomorphic complexes did not result. The reason
for this lies in the structural anisotropy required to ensure
§ The J phase is a crystal smectic rather than a true smectic phase. The
SmI phase is a fluid, tilted hexagonal phase.
Fig. 37 Mesomorphic Group 6 imine complexes.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1926

View Article Online
Fig. 38 Mesomorphic Re() diazabutadienes.
Fig. 39 Mesomorphic Re() complexes of 2,2Ј-bipyridines.
Thus, the transition temperatures are not reduced on complex-
ation and no strongly polar function associated with the ligated
‘Re(CO)4’ unit.
ation and, indeed, in some longer-chain homologues, it was
observed that the SmC phase of the ligand was retained on
complexation.
Next, we can consider another family of α-diimines and their
complexes, this time based on 5,5Ј-disubstituted-2,2Ј-bipyrid-
ines (Fig. 39).63a
Therefore, we sought some other systems which might act as
comparators in order that we might delineate the effect of the
ligand conformational change and any dipoles associated with
the metal fragment. In making this comparison, we turned to
tetracatenar complexes of a series of diimine complexes which
are now described.
Here, the mesomorphism is exemplified by the following
example:
Polycatenar systems
In addition to the simple, calamitic bipyridines described above,
the synthesis was undertaken of some tetra- and hexa-catenar
derivatives, too.65 Of these, it is the tetracatenar derivatives
(Fig. 40) which are of interest here (ligands are compound 7),
and their complexes (8) with Re().
The thermal behaviour of the ligands has been reported in
detail63 and, typically for tetracatenar materials, they show N
and SmC phases at short chain length and Colh phases at long
chain length; certain intermediate chain lengths also show a
cubic phase, although a discussion of the nature of this fascin-
ating, 3-D phase is beyond the scope of this manuscript. These
ligands may then be reacted with [ReBr(CO)5] to give bromo-
rheniumtricarbonyl complexes (8) and a comparison of the
mesomorphism for certain ligands and complexes is given in
Fig. 41 (labeling on the x-axis corresponds to the compound
number with the carbon chain length in parentheses).
What is found is that the clearing point of the complexes is
reduced over that of the ligands, although while there is a gentle
decrease in the clearing point of the ligands, it increases steadily
in the complexes. The ligands show only Colh phases for the
homologues under consideration and while the complexes also
Here, there is some reduction in transition temperatures, but no
loss of the smectic phase upon complexation.
What the above results show is that in the two α-diimine
systems, there is no major reduction in transition temperatures
as was found with the Re() imines, and that the smectic phases
found in the ligands can be retained in the complexes. The dif-
ference in the diimine systems is that, as indicated in the respect-
ive figures, the free ligands exist in a transoid arrangement
which becomes cisoid on complexation. Thus, from a ligand
with no dipole moment associated with the diimine function,
a complex is generated where an appreciable dipole moment
exists (e.g. calculated gas-phase dipole moment for 2,2Ј-bipyr-
idine is 3.7 D) and there is, in addition, an appreciable dipole
moment associated with the Re–Br bond. By contrast, in the
imine systems, there is no conformational change on complex-
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1927

View Article Online
Fig. 40 Tetracatenar bipyridine complexes of Re().
a single methylene unit to the terminal chains, so that the
dodecyloxy homologue shows only a SmC phase while the tri-
decyloxy homologues shows only a Colh phase. This is rather
unexpected and this progression from lamellar to columnar
phases in tetracatenar mesogens has been discussed else-
where.58b,68 The mesomorphism of the ligands (9) and their
Re() complexes (10) is then compared in Fig. 43.
Fig. 41 Comparative mesomorphism of bipyridine ligands 7 and their
Re() complexes 8.
show this phase, the two longer-chain homologues both show a
66
¯
cubic phase (symmetry Ia3d); the longest-chain homologue
shows a Colr phase in addition, although its symmetry has not
been determined.
In order to take the change in ligand dipole into account, we
then proceeded to synthesise some tetracatenar 3,8-disubsti-
tuted phenanthrolines (Fig. 42) as a permanent dipole moment
of 4.1 D67 exists due to the fixing of the ring nitrogens by the
exocyclic double bond.
These ligands are not direct analogues of the bipyridines as
the ‘arms’ are bound to the core by acetylene functions rather
than esters, but nevertheless are comparable in most other ways.
Once more, the ligands (9) show N and SmC phases at shorter
chain lengths and the Colh phase for longer-chain homologues
and a curious feature of these compounds is that the change
from SmC to Colh mesomorphism happens with the addition of
Fig. 43 Comparative mesomorphism of the phenanthroline ligands 9
and their Re() complexes 10.
The first thing that the plot shows is that the clearing points
decrease in both ligand and complex with increasing chain
length, in contrast to the behaviour of the bipyridines. The lig-
and decyloxy homologue shows a SmC phase which disappears
on complexation, but the tetradecyloxy and hexadecyloxy
homologues show the Colh phase in both ligand and complex.
However here, the clearing point of the complexes is higher
than that of the ligands and the mesomorphic range is wider in
the complexes due to some destabilisation of the crystal phase.
Fig. 42 Tetracatenar phenanthrolines and their Re() complexes.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1928

View Article Online
Scheme 2 Synthesis of the mesomorphic amidines (11) and their complexes of Re() (12): i) 3,4-dihydroxybenzaldehyde/piperidine/butanol/∆; ii)
CnH2n ϩ 1Br/butanone/NaI/∆; iii) hydrazine/‘Pd/C’/EtOH; iv) triethylorthoformate/toluene/AcOH/∆; v) [ReMe(CO)5]/toluene/∆.
Thus, there is a very different pattern of behaviour in the
phenanthroline materials.
The final class of ligands and complexes to be considered are
based on an amidine function which has never previously been
employed in liquid crystal compounds to the best of our know-
ledge. The structure of the ligands and their complexes is shown
in Scheme 2. These ligands were chosen for, while they possess
a permanent dipole moment as drawn in the scheme, the
‘Re(CO)4’ group does not contribute a significant dipole on
complexation in contrast to the ‘ReBr(CO)3’ unit.
The ligands themselves (11) were indeed mesomorphic and
formed columnar phases over a relatively narrow range of
temperature, not far above 100 ЊC. These ligands are quite
different from the bipyridine and phenanthroline materials as
they bend at the amidine carbon and, it is assumed, adopt
the motif suggested in the scheme. On complexation, it is
assumed that the ligand opens out a little, but that the com-
plexes are still bent about the Re centre. The mesomorphism
of ligand (11) and complex (12) are presented in Fig. 44,
plotted on the same scale as those discussed previously for
ease of reference.
In this series, fewer comparisons are available but neverthe-
less, trends are apparent. Thus for both ligand and complex,
clearing point and melting point decrease with increasing ter-
minal chain length and this time it is noted that the transition
temperatures and mesomorphic range in the complexes are
appreciably lower than in the free ligands.
Fig. 44 Comparative mesomorphism of the amidine ligands 11 and
their Re() complexes 12.
led to a reduction in transition temperatures, although in the
complexes themselves, these transition temperatures increased
with chain length, contrary to observations made for the ami-
dines. However, the phenanthrolines were different again and in
these systems, the clearing point of the complexes were higher
than those of the ligands.
There is clearly no obvious set of factors which will explain
the above behaviour consistently, yet there are lessons appar-
ent in controlling both transition temperatures and, to a
degree, mesomorphism. It is clear that all of these properties
are a complex and delicate function of the chemical com-
position of the system and that surprises await those who
would seek them.
Conclusions
So what might be learned from the three series of materials
above? The results from the amidines are consistent with
those discussed earlier with regard to calamitic imines in which
the ‘Re(CO)4’ unit acted to reduce transition temperatures,
although the effect is not so great. With the bipyridine systems,
complexation (this time of an ‘ReBr(CO)3’ group) once more
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1929

View Article Online
20 J. Barberá, A. Elduque, R. Gimenez, L. A. Oro and J.-L. Serrano,
Angew. Chem., Int. Ed. Engl., 1996, 35, 2832.
Acknowledgements
21 K. Praefcke, D. Singer and B. Gündogan, Mol. Cryst., Liq. Cryst.,
1991, 223, 191; D. Singer, A. Liebmann, K. Praefcke and J. H.
Wendorff, Liq. Cryst., 1993, 14, 785; B. Gündogan and K. Praefcke,
Chem. Ber., 1993, 126, 1253; K. Praefcke, B. Bilgin, J. Pickardt and
M. Borowski, Chem. Ber., 1994, 127, 1543; K. Praefcke, B. Bilgin,
N. Usol’tseva, B. Heinrich and D. Guillon, J. Mater. Chem., 1995, 5,
2257.
22 (a) H. Adams, A. C. Albeniz, N. A. Bailey, D. W. Bruce, A. S.
Cherodian, R. Dhillon, D. A. Dunmur, P. Espinet, J. L. Feijoo,
E. Lalinde, P. M. Maitlis, R. M. Richardson and G. Ungar, J. Mater.
Chem., 1991, 1, 843; (b) D. W. Bruce, D. A. Dunmur, S. E. Hunt,
P. M. Maitlis and R. Orr, J. Mater. Chem., 1991, 1, 857; (c) D. W.
Bruce, R. Dhillon, D. A. Dunmur and P. M. Maitlis, J. Mater.
Chem., 1992, 2, 65.
23 D. W. Bruce, E. Lalinde, P. Styring, D. A. Dunmur and P. M. Maitlis,
J. Chem. Soc., Chem. Commun., 1986, 581; H. Adams, N. A. Bailey,
D. W. Bruce, D. A. Dunmur, E. Lalinde, M. Marcos, C. Ridgway,
A. J. Smith, P. Styring and P. M. Maitlis, Liq. Cryst., 1987, 2, 381.
24 D. W. Bruce, S. Fan and G. R. Luckhurst, Liq. Cryst., 1994, 16,
1093.
25 T. Kaharu, T. Tanaka, M. Sawada and S. Takahashi, J. Mater.
Chem., 1994, 4, 859.
26 M. Lee, Y. S. Yoo and M. Choi, Bull. Korean Chem. Soc., 1997, 18,
1067.
27 S. Coco, F. Díez-Expósito, P. Espinet, C. Fernández-Mayordomo,
J. M. Martín-Alvarez and A.-M. Levelut, Chem. Mater., 1998, 10,
3666.
28 Y. G. Yang and J. X. Wen, Liq. Cryst., 1998, 25, 765.
29 P. Alejos, S. Coco and P. Espinet, New J. Chem., 1995, 19, 799.
30 P. Styring, H. Adams N. A. Bailey and D. W. Bruce, unpublished
work.
31 S. Coco, P. Espinet, S. Falagán and J. M. Martin-Alvarez, New J.
Chem., 1995, 19, 959.
32 D. W. Bruce, D. A. Dunmur, E. Lalinde, P. M. Maitlis and P. Styring,
Liq. Cryst., 1988, 3, 385.
33 D. W. Bruce, Adv. Inorg. Chem., 2001, 52, 151.
34 J. P. Rourke, F. P. Fanizzi, N. J. S. Salt, D. W. Bruce, D. A. Dunmur
and P. M. Maitlis, J. Chem. Soc., Chem. Commun., 1990, 229.
35 W. Weissflog, in Handbook of Liquid Crystals, eds. D. Demus,
J. Goodby, G. W. Gray, H.-W. Spiess and V. Vill, Wiley-VCH,
Weinheim, 1998, vol. 2B.
36 J. P. Rourke, F. P. Fanizzi, D. W. Bruce, D. A. Dunmur and
P. M. Maitlis, J. Chem. Soc., Dalton Trans., 1992, 3009.
37 D. W. Bruce, D. A. Dunmur, M. A. Esteruelas, S. E. Hunt, R. Le
Lagadec, P. M. Maitlis, J. R. Marsden, E. Sola and J. M. Stacey,
J. Mater. Chem., 1991, 1, 251.
We thank the EPSRC and the University of Exeter for support,
Johnson Matthey for generous loans of precious metal salts and
H. C. Stark for gifts of ammonium perrhenate.
References
1 A.-M. Giroud and U. T. Mueller-Westerhoff, Mol. Cryst., Liq.
Cryst., 1977, 41, 11.
2 A. M. Giroud-Godquin and P. M. Maitlis, Angew. Chem., Int. Ed.
Engl., 1991, 30, 375; P. Espinet, M. A. Esteruelas, L. A. Oro,
J. L. Serrano and E. Sola, Coord. Chem. Rev., 1992, 117, 215;
S. A. Hudson and P. M. Maitlis, Chem. Rev., 1993, 93, 861;
A. Polishchuk and T. V. Timofeeva, Russ. Chem. Rev., 1993,
62, 291; D. W. Bruce, J. Chem. Soc., Dalton Trans., 1993, 2983;
R. Deschenaux and J. W. Goodby, in Ferrocene: Homogeneous
Catalysis, Organic Synthesis, Materials Science, eds. A. Togni and
T. Hayashi, VCH Weinheim, 1995, ch. 9, p. 471; F. Neve, Adv.
Mater., 1996, 8, 277; B. Donnio and D. W. Bruce, Struct. Bonding
(Berlin), 1999, 95, 193; S. R. Collinson and D. W. Bruce, in
Transition Metals in Supramolecular Chemistry, ed. J. P. Sauvage,
Wiley, New York, 1999, ch. 7, pp. 285–369; D. W. Bruce, in Inorganic
Materials, eds. D. W. Bruce and D. O’Hare, Wiley, New York, First
Edition, 1992, ch. 8, pp. 405–490; Second Edition, ch. 8, pp. 429–
522; A. M. Giroud-Godquin, in Handbook of Liquid Crystals, eds.
D. Demus, J. Goodby, G. W. Gray, H. W. Spiess and V. Vill, Wiley-
VCH, Weinheim, 1998, ch. XIV, p. 901; B. Donnio, D. Guillon,
R. Deschenaux and D. W. Bruce in Comprehensive Coordination
Chemistry II, eds. J. A. McCleverty and T. J. Meyer, to be published
in 2003.
3 Metallomesogens: Synthesis, Properties and Applications, J. L.
Serrano, ed., VCH, Weinheim, 1996.
4 For a discussion of the different types of material forming lyotropic
liquid crystal mesophases, see: T. K. Attwood, J. E. Lydon, C. Hall
and G. J. T. Tiddy, Liq. Cryst., 1990, 7, 657.
5 A. M. Levelut, J. Malthête and A. Collet, J. Phys. (France), 1986,
47, 351; H. Zimmerman, R. Poupko, Z. Luz and J. Billard,
Z. Naturforsch., Teil A, 1985, 40, 149.
6 See e.g. (a) S. A. Ponomarenko, N. I. Boiko, V. P. Shibaev, R. M.
Richardson, I. J. Whitehouse, E. A. Rebrov and A. M. Muzafarov,
Macromolecules, 2000, 33, 5549; (b) J. Barberá, M. Marcos and
J.-L. Serrano, Chem. Eur. J., 1999, 5, 1834; M. Marcos, R. Giménez,
J.-L. Serrano, B. Donnio, B. Heinrich and D. Guillon, Chem. Eur. J.,
2001, 7, 1006; (c) J. H. Cameron, A. Facher, G. Lattermann and
S. Diele, Adv. Mater., 1997, 9, 398; (d ) V. Percec, W. D. Cho,
G. Ungar and D. J. P. Yeardly, Angew. Chem., Int. Ed., 2000, 39,
1597.
7 J. Malthête, H.-T. Nguyen and C. Destrade, Liq. Cryst., 1993, 13,
171; H.-T. Nguyen, C. Destrade and J. Malthête, Adv. Mater., 1997,
9, 375.
8 G. Pelzl, S. Diele and W. Weissflog, Adv. Mater., 1999, 1, 707.
9 A.-M. Giroud-Godquin, J.-C. Marchon, D. Guillon and
A. Skoulios, J. Phys. Lett. (Paris), 1984, 45, 681; H. Abied,
D. Guillon, A. Skoulios, A.-M. Giroud-Godquin and J.-C.
Marchon, Liq. Cryst., 1987, 2, 269.
38 D. W. Bruce, D. A. Dunmur, M. R. Manterfield, P. M. Maitlis and
R. Orr, J. Mater. Chem., 1991, 1, 255.
39 C. Bertram, D. W. Bruce, D. A. Dunmur, S. E. Hunt, P. M. Maitlis
and M. McCann, J. Chem. Soc., Chem. Commun., 1991, 69.
40 See e.g. M. Ghedini, M. Longeri and R. Bartolino, Mol. Cryst., Liq.
Cryst., 1982, 84, 207; M. Ghedini, S. Licoccia, S. Armentano and
R. Bartolino, Mol. Cryst., Liq. Cryst., 1984, 108, 269; M. Ghedini,
S. Armentano and F. Neve, Inorg. Chim. Acta, 1987, 134, 23;
M. Ghedini, S. Licoccia, S. Armentano and R. Bartolino,
Mol. Cryst., Liq. Cryst., 1984, 108, 269.
41 M. J. Baeña, J. Barbera, P. Espinet, A. Ezcurra, M. B. Ros and
J.-L. Serrano, J. Am. Chem. Soc., 1994, 116, 1899.
42 L. Omnès, B. A. Timimi, T. Gelbrich, M. B. Hursthouse,
G. R. Luckhurst and D. W. Bruce, Chem. Commun., 2001, 2248.
43 P. Espinet, E. Lalinde, M. Marcos, J. Pérez and J.-L. Serrano,
Organometallics, 1990, 9, 555.
10 A.-M. Giroud-Godquin and J. Billard, Mol. Cryst., Liq. Cryst.,
1981, 66, 147; A.-M. Levelut, J. Chim. Phys., 1983, 80, 149.
11 M. Veber, R. Fugnitto and H. Strzelecka, Mol. Cryst., Liq. Cryst.,
1983, 96, 221; K. Ohta, A. Tkagi, H. Muroki, I. Yamamoto,
K. Matsuzaki, T. Inabe and Y. Maruyama, J. Chem. Soc., Chem.
Commun., 1986, 883; K. Ohta, A. Takagi, H. Muroki, I. Yamamoto,
K. Matsuzaki, T. Inabe and Y. Maruyama, Mol. Cryst., Liq. Cryst.,
1987, 147, 15.
12 M. Veber, P. Davidson, C. Jallabert, A.-M. Levelut and
H. Strzelecka, Mol. Cryst., Liq. Cryst. Lett., 1987, 5, 1.
13 K. Ohta, H. Hasebe, H. Ema, T. Fujimoto and I. Yamamoto,
J. Chem. Soc., Chem. Commun., 1989, 1610.
44 R. C. Elder, R. D. P. Cruea and R. F. Morrison, Inorg. Chem., 1976,
15, 1623.
45 P. Espinet, J. Etxebarria, M. Marcos, J. Pérez, A. Rémon and
J.-L. Serrano, Angew. Chem., Int. Ed. Engl., 1989, 28, 1065.
46 N. Hoshino, Coord. Chem. Rev., 1998, 174, 77.
14 K. Ohta, O. Takenaka, H. Hasebe, Y. Morizumi, T. Fujimoto and
I. Yamamoto, Mol. Cryst., Liq. Cryst., 1991, 195, 135.
15 N. J. Thompson, G. W. Gray, J. W. Goodby and K. J. Toyne,
Mol. Cryst., Liq. Cryst., 1991, 200, 109.
16 C. Mertesdorf and H. Ringsdorf, Liq. Cryst., 1989, 5, 1757;
G. Lattermann, Liq. Cryst., 1989, 6, 619.
47 Yu. G. Galyametdinov, G. I. Ivanova and I. V. Ovchinnikov,
Bull. Acad. Sci. USSR Div. Chem. Sci., 1991, 40, 1109.
48 See e.g. F. Martin, S. R. Collinson and D. W. Bruce, Liq. Cryst.,
2000, 27, 859; K. Binnemans, D. W. Bruce, S. R. Collinson, R. Van
Deun, Y. G. Galyametdinov and F. Martin, Proc. R. Soc. London A,
1999, 357, 3063.
17 J. Malthête, A.-M. Levelut and J.-M. Lehn, J. Chem. Soc., Chem.
Commun., 1985, 1794; J. Malthête, A.-M. Levelut and J.-M. Lehn,
J. Chem. Soc., Chem. Commun., 1992, 1434.
18 C. Sirlin, L. Bosio and J. Simon, Mol. Cryst., Liq. Cryst., 1988, 155,
231; C. Sirlin, L. Bosio and J. Simon, J. Chem. Soc., Chem.
Commun., 1987, 379.
49 For an excellent and comprehensive review of lanthanide mesogens,
see: K. Binnemans and C. Görller-Walrand, Chem. Rev., 2002, 102,
2303.
50 M. Lehmann, T. Sierra, J. Barberá, J.-L. Serrano and R. Parker,
J. Mater. Chem., 2002, 1342.
51 J. Barberá, M. Marcos, A. Omenat, J.-L. Serrano, J. I. Martínez and
P. J. Alonso, Liq. Cryst., 2000, 27, 255.
19 P. Weber, D. Guillon and A. Skoulios, J. Phys. Chem., 1987, 91, 2242.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1930

View Article Online
52 I. Aiello, D. Dattilo, M. Ghedini, A. Bruno, R. Termine and
A. Golemme, Adv. Mater., 2002, 14, 1233; I. Aiello, D. Dattilo,
M. Ghedini and A. Golemme, J. Am. Chem. Soc., 2001, 123,
5598.
53 In parallel with this work, there have been conference reports
(International Liquid Crystal Conference, Edinburgh, July 2002) of
columnar derivatives by the Zaragoza group.
54 (a) C. K. Lai, C-H. Chang and C-H. Tsai, J. Mater. Chem., 1998, 8,
599; (b) H.-T. Nguyen, C. Destrade and J. Malthête, Adv. Mater.,
1997, 9, 375.
55 B. Xu and T. M. Swager, J. Am. Chem. Soc., 1993, 115, 1159.
56 A. G. Serrette, P. J. Carroll and T. M. Swager, J. Am. Chem. Soc.,
1992, 114, 1887; A. G. Serrette and T. M. Swager, J. Am. Chem. Soc.,
1993, 115, 8879.
60 D. P. Lydon, G. W. V. Cave and J. P. Rourke, J. Mater. Chem., 1997,
7, 403.
61 See e.g. D. W. Bruce and X.-H. Liu, J. Chem. Soc., Chem. Commun.,
1994, 729; X.-H. Liu, M. N. Abser and D. W. Bruce, J. Organomet.
Chem., 1998, 551, 271; (Publisher’s Erratum, 1999, 577, 150).
62 (a) X.-H. Liu, B. Heinrich, I. Manners, D. Guillon and D. W. Bruce,
J. Mater. Chem., 2000, 10, 637; (b) M.-A. Guillevic, M. E. Light,
S. J. Coles, T. Gelbrich, M. B. Hursthouse and D. W. Bruce, J. Chem.
Soc., Dalton Trans., 2000, 1437.
63 (a) K. E. Rowe and D. W. Bruce, J. Chem. Soc., Dalton Trans., 1996,
3913; (b) D. W. Bruce, Adv. Mater., 1994, 6, 699.
64 S. Morrone, G. Harrison and D. W. Bruce, Adv. Mater., 1995, 7, 665;
S. Morrone, D. Guillon and D. W. Bruce, Inorg. Chem., 1996, 35,
7041.
57 J. Malthête, H.-T. Nguyen and C. Destrade, Liq. Cryst., 1993, 13,
171.
65 K. E. Rowe and D. W. Bruce, J. Mater. Chem., 1998, 8, 331.
66 K. E. Willis, K. E. Rowe, D. W. Bruce and G. Ungar, unpublished
work.
67 P. E. Fielding and R. J. W. LeFèvre, J. Chem. Soc., 1951, 1811.
68 A. I. Smirnova, D. Fazio, E. Fernandez Iglesias, C. G. Hall,
D. Guillon, B. Donnio and D. W. Bruce, Mol. Cryst, Liq. Cryst.,
in press.
58 (a) D. Guillon, B. Heinrich, A. C. Ribeiro, C. Cruz and H. T.
Nguyen, Mol. Cryst. Liq. Cryst., 1998, 317, 51; (b) D. Fazio,
C. Mongin, B. Donnio, Y. Galerne, D. Guillon and D. W. Bruce,
J. Mater. Chem., 2001, 11, 2852.
59 M. Ghedini, D. Pucci and F. Neve, Chem.Commun., 1996, 996.
D a l t o n T r a n s . , 2 0 0 3 , 1 9 1 4 – 1 9 3 1
1931