Discotic liquid crystals of transition metal complexes, 31: Establishment of mesomorphism and thermochromism of bis[1,2-bis(4-n-alkoxyphenyl)ethane-1,2-dithiolene]nickel complexes

Article abstract of DOI:10.1039/b007135h

Two series of bis[1,2-bis(4-n-alkylphenyl)ethane-1,2-dithiolene]nickel, C_n-Ni (n= 1-12), and bis[1,2-bis(4-n-alkoxyphenyl)ethane-1,2-dithiolene]nickel, C_nO-Ni(n = 1-12, 14, 16, 18), have been synthesized. Their mesomorphism, thermochromism, supramolecular structures and π-acceptor property have been investigated by using different scanning calorimetry, polarizing microscopy, temperature-dependent X-ray diffraction technique, electronic spectroscopy and cyclic voltammetry. From the X-ray diffraction and electronic spectral results, it was established that the C_nO-Ni complexes for n ≤ 10 exhibit two differently colored discotic lamellar (D_L) mesophases whereas none of the C_n-Ni complexes has a mesophase, and that the thermochromism (brown→green) is attributable to a slow transformation from the Ni-Ni bonded dimers to the Ni-S bonded dimers.

Full text of DOI:10.1039/b007135h

View Article Online / Journal Homepage / Table of Contents for this issue
Discotic liquid crystals of transition metal complexes, 31:{
establishment of mesomorphism and thermochromism of bis[1,2-
bis(4-n-alkoxyphenyl)ethane-1,2-dithiolene]nickel complexes{
Hiroko Horie, Akira Takagi, Hiroshi Hasebe, Takumi Ozawa and Kazuchika Ohta*
Department of Functional Polymer Science, Faculty of Textile Science and Technology,
Shinshu University, 386-8567 Ueda, Japan. E-mail: ko52517@giptc.shinshu-u.ac.jp
Received 4th September 2000, Accepted 9th January 2001
First published as an Advance Article on the web 12th February 2001
Two series of bis[1,2-bis(4-n-alkylphenyl)ethane-1,2-dithiolene]nickel, Cn–Ni (n~1–12), and bis[1,2-bis(4-n-
alkoxyphenyl)ethane-1,2-dithiolene]nickel, CnO–Ni(n~1–12, 14, 16, 18), have been synthesized. Their
mesomorphism, thermochromism, supramolecular structures and p-acceptor property have been investigated by
using different scanning calorimetry, polarizing microscopy, temperature-dependent X-ray diffraction technique,
electronic spectroscopy and cyclic voltammetry. From the X-ray diffraction and electronic spectral results, it
was established that the CnO–Ni complexes for n¡10 exhibit two differently colored discotic lamellar (DL)
mesophases whereas none of the Cn–Ni complexes has a mesophase, and that the thermochromism
(brownAgreen) is attributable to a slow transformation from the Ni–Ni bonded dimers to the Ni–S bonded
dimers.
1. Introduction
In 1977, the first calamitic liquid crystalline dithiolene metal
complex was reported by Giroud and Mueller-Westerhoff.2,3
Since then, two long chain-substituted rod-like dithiolene
transition metal (Ni, Pt) complexes which show smectic and
nematic liquid crystal phases have been studied.3–7 These
dithiolene metal complexes have unusual electronic structures
to make them superior electron acceptors.3,5–7
In 1983, it was reported by Veber et al. that four long chain-
substituted dithiolene nickel complex bis[1,2-bis(4-n-dodecyloxy-
phenyl)ethane-1,2-dithiolene]nickel (abbreviated as C12O–Ni)
shows a mesophase.8 In 1986, we also synthesized the
homologous complexes (n~9, 11), using a different synthetic
route (Scheme 1) and reported that these complexes exhibit a
discotic mesophase.9,10 However, in 1987, Veber et al. denied
the existence of mesomorphism in these four long chain-
substituted complexes and they claimed that the phase is not
mesomorphic but crystalline.11 Prior to this paper, Takagi in
our group had already synthesized a series of the homologous
CnO–Ni and Cn–Ni complexes for n~1–12, using synthetic
routes illustrated in Schemes 1 and 2. He found a mesophase in
CnO–Ni for n~9, 11, two mesophases in CnO–Ni for n~10, 12,
and no mesophase in all the Cn–Ni complexes.12 He also
noticed two differently colored mesophases for C12O–Ni that
Veber et al. had overlooked.
Hence, in order to establish the mesomorphism and
thermochromism of CnO–Ni, we have synthesized additional
CnO–Ni homologues for n~14, 16, 18 by a novel synthetic
route (Scheme 3) different from the previous two routes, and
reinvestigated the physical properties in detail for all the
complexes, CnO–Ni (n~1–12, 14, 16, 18) and Cn–Ni (n~1–12).
Scheme 1 Previous synthetic route for bis[1,2-bis(4-n-alkoxyphenyl)-
{For part 30, see ref. 1.
ethane-1,2-dithiolene]nickel, 1. EtOH~ethanol, THF~tetrahydro-
furan, NiCl2(dppe)~dichloro[1,3-bis(diphenylphosphino)ethane]nick-
el(II), and AcOH~acetic acid.
{Elemental analysis data and yields for compounds Cn–Ni and CnO–Ni
are available as supplementary data. For direct electronic access see
http://www.rsc.org/suppdata/jm/b0/b007135h/
DOI: 10.1039/b007135h
J. Mater. Chem., 2001, 11, 1063–1071
This journal is # The Royal Society of Chemistry 2001
1063

View Article Online
Scheme 2 Synthetic route for bis[1,2-bis(4-n-alkylphenyl)ethane-1,2-
dithiolene]nickel, 2. EtOH~ethanol, THF~tetrahydrofuran,
NiCl2(dppe)~dichloro[1,3-bis(diphenylphosphino)ethane]nickel(II), and
AcOH~acetic acid.
2. Experimental
Scheme 3 New synthetic route for bis[1,2-bis(4-n-alkoxyphenyl)ethane-
1,2-dithiolene]nickel, 1. AcOH~acetic acid, DMA~N,N-dimethyl-
acetamide.
2.1 Synthesis
Synthesis A. The synthetic route to the CnO–Ni complexes
for n~1–12 is shown in Scheme 1.9,10,12 In this synthetic route
4-bromophenol, the starting material, was converted to 4-n-
alkoxybromobenzene 2. The reaction of the Grignard reagent
prepared from 4-n-alkoxybromobenzene with (Z)-dichlor-
oethylene in the presence of dichloro[1,3-bis(diphenylpho-
Table 1. In the electronic supplementary data,{ the elemental
analysis data, yields and the crystalline shapes obtained from
recrystallization are summarized. The detailed procedures were
described in a previous paper.10
sphino)propane]nickel(II) gave 4,4’-di-n-alkoxystilbene
3
(Tamao–Kumada reaction13). 4,4’-Di-n-alkoxybenzil 4 was
obtained by oxidation of the stilbene derivative 3 by selenium
dioxide.14 The preparation of the ligand and the formation of
the complex was carried out in one pot; 4,4’-di-n-alkoxybenzil 4
was treated with phosphorus pentasulfide and then the
complexation was carried out in situ with nickel chloride
hexahydrate. All the 4,4’-di-n-alkoxybenzil 4 and the corre-
sponding CnO–Ni complexes 1 were recrystallized from n-
hexane and ethyl acetate, respectively. These CnO–Ni com-
plexes recrystallized from ethyl acetate are black for n~1–5,
dark brown for n~6 and brown for n~7–18, as summarized in
Synthesis B. The synthetic route to the Cn–Ni complexes for
n~1–12 is shown in Scheme 2. The alkyl-substituted nickel
complexes, Cn–Ni (2: n~1–12), could be prepared by almost
the same synthetic route illustrated in Scheme 1. 4-n-Alkyl-
benzenes 5 were prepared by Tamao–Kumada reaction.13 All
the complexes except for C1–Ni and C2–Ni were recrystallized
from ethyl acetate. The recrystallization solvent for C1–Ni and
C2–Ni was dioxane. These Cn–Ni complexes obtained from
recrystallization are black for n~1–5, dark green for n~6 and
green for n~7–12, as summarized in Table 1.
1064
J. Mater. Chem., 2001, 11, 1063–1071

View Article Online
Table 1 The crystalline shapesa and colors obtained from recrystalliza-
tion for the complexes Cn–Ni and CnO–Ni
microscope, Olympas BH-2, equipped with a heating plate
controlled by a thermoregulator, Mettler FP80 and 82, and
measured with differential scanning calorimeters, a Rigaku
Thermoflex TG-DSC. To establish the mesophases, powder X-
ray patterns were measured with Cu-Ka radiation using a
Rigaku Geigerflex equipped with a hand-made heating plate
controlled by thermoregulator.15 Reduction potentials of these
complexes were measured with cyclic voltammetry, Yanagi-
moto Polarographic Analyzer P-1100, in methylene chloride
solutions containing 0.1 M tetrabutylammonium perchlorate
as supporting electrolyte. The measurements were made at
glassy-carbon working electrode vs. a saturated calomel
n
Cn–Ni
n
CnO–Ni
1
2
Black needles
Black needles
1
2
Black needles
Black needles
3
4
Black needles
Black needles
3
4
Black needles
Black needles
5
6
7
8
Black needles (K1)
Dark green needles
Green needles
Green powder (K1)
Green powder (K1)
Green powder (K1)
Green powder (K1)
Green powder (K1)
5
6
7
8
Black needles (K1)
Dark brown needles (K1)
Brown needles (K1)
Brown needles (K1)
Brown needles (K2)
Brown needles (K1)
Brown needles (K1)zPlates(K2)
Brown needles (K2)
Brown needles (K2)
Brown needles (K2)
Brown needles (K3)
9
9
10
11
12
10
11
12
14
16
18
electrode (SCE); sweep rates 10 mV sec21
.
3. Results and discussion
aAll the complexes except for C1–Ni and C2–Ni were recrystallized
from ethyl acetate. The recrystallization solvent for C1–Ni and C2–
Ni was dioxane.
3.1 Synthesis
The problem of the synthetic routes in Schemes 1 and 2 is that
the length of the chains must be determined in the early stage.
Moreover, disposal of the unreacted selenium dioxide after the
oxidation reaction is very difficult and expensive. Therefore, we
have synthesized the CnO–Ni complexes for n~14, 16, and 18
by a new synthetic route as shown in Scheme 3. The benzil
derivatives 4 were synthesized from 4-methoxybenzaldeyde by
the method of Wenz.16 This synthetic route is advantageous for
the synthesis of a series of the homologues with different length
of the lateral chains, because the alkylation in the late stage
permits us to use a common precursor, 11, for all the
complexes.
Synthesis C. A novel synthetic route to the CnO–Ni
complexes for n~14, 16 and 18 is shown in Scheme 3. These
CnO–Ni complexes recrystallized from ethyl acetate are brown,
as summarized in Table 1. The detailed procedures are
described only for the representative complex C14O–Ni in the
following:
4,4’-Bis(tetradecyloxy)benzil 4. In an atmosphere of dry
nitrogen, anhydrous potassium carbonate, 0.74 g (5.4 mmol),
and n-tetradecyl bromide, 1.5 g (5.4 mmol), were added to a
solution of 4,4’-dihydroxybenzil, 0.52 g (2.2 mmol), in 20 ml of
N,N-dimethylacetamide and the mixture was stirred for
17 hours at 90 ‡C. The reaction mixture was extracted with
chloroform. The organic layer was washed with water, dried
over sodium sulfate and the solvent was evaporated. The
purification was carried out by recrystallization from ethyl
acetate to give 1.0 g of white crystals 4. Yield 73%; mp 73.7 ‡C;
IR (KBr pellet, cm21); 2940, 2850, 1660, 1600, 1260. 1H-NMR
(CDCl3, TMS): d/ppm~0.90(m, 6H, -CH3), 1.30(m, 48H,
-(CH2)-), 4.02(t, J~6.3 Hz, 4H, CH2), 6.89(d, J~7.8 Hz, 4H,
arom. H), 7.87(d, J~7.8 Hz, 4H, arom.-H).
Table 2 Phase transition temperatures (T) and enthalpy changes (DH)
of Cn–Ni
Bis[1,2-bis(4-n-tetradecyloxyphenyl)ethane-1,2-dithiolene]-
nickel 1. A mixture of 1.0 g (1.6 mmol) of 4,4’-bis(tetradecy-
loxy)benzil 4, 1.4 g (3.2 mmol) of phosphorus pentasulfide, and
40 ml of dioxane was refluxed for 5 hours. The hot reaction
mixture was filtered to remove the unreacted phosphorus
pentasulfide and washed with a small portion of hot dioxane.
And a solution of nickel(II) dichloride hexahydrate, 0.19 g
(0.79 mmol), in 8 ml of ethanol was added to the filtrate and the
reaction mixture was refluxed for 2 hours. After it had been
cooled by immersion in an ice–water bath, a brown powder was
formed and was collected by filtration to give the crude
complex 1. Purification was performed by recrystallization
from ethyl acetate to afford 0.62 g of brown needle like crystals.
Yield 57%; cp 163.2 ‡C; IR (KBr pellet, cm21): 2930, 2860,
1600, 1510, 1470, 1360, 1300, 1250, 1170, 1140, 880; 1H-NMR
(CDCl3, TMS) d/ppm~0.88(t, J~4 Hz, 12H, -CH3), 1.26–
1.49(m, 96H, -(CH2)-), 3.9(t, J~6 Hz, 8H, -OCH2-), 6.50(d,
J~4.2 Hz, 8H, arom.-H), 6.96(d, J~4.2 Hz, 8H, arom.-H);
Anal. Found (Calcd. for C84H132O4S4Ni): C 72.36% (72.43%),
H 9.54% (9.55%).
ð0Þ
2.2 Measurements
The complexes 1 and 2 were identified by elemental analysis
using a Perkin–Elmer Elemental Analyzer 240B. The phase
transition behaviors of 1 and 2 were observed with a polarizing
aPhase nomenclature: K~crystal and I.L.~isotropic liquid.
J. Mater. Chem., 2001, 11, 1063–1071
1065

View Article Online
1066
J. Mater. Chem., 2001, 11, 1063–1071

View Article Online
Fig. 1 Phase transition temperature versus number of the carbon atoms in the alkoxy chain (n). Open circles: clearing points. Filled circles and
squares: melting points of brown crystals and green crystals, respectively. Triangles: phase transitions from brown D1 to green D2. Crosses: crystal–
crystal phase transition.
3.2 Thermal behavior of Cn–Ni (n~1–12)
Table 2 summarizes the phase transition temperatures and the
phase transition enthalpy changes (DH) which were determined
by the DSC measurements and polarizing microscopic
observations. None of the Cn–Ni complexes show a mesophase.
It is noteworthy that these Cn–Ni complexes obtained from
recrystallization are green for n~7–12, whereas the CnO–Ni
complexes recrystallized from ethyl acetate are brown for n~7–
18, as summarized in Table 1. This is very suggestive of a color
change from brown D1 mesophase to green D2 mesophase in
CnO–Ni complexes for n¢10, as will be described in the latter
part of this paper.
Fig. 2 Miscibility diagram between C9O–Ni and C11O–Ni. Points in
the miscibility diagram were onset temperatures observed in the DSC
measurements as the peaks or shoulders. The open and filled circles
could be observed as clear and big peaks, but most of the crosses were
observed as complicated small peaks and shoulders. Curves a and b:
freezing depression curves. Curve c: clearing point curve. Point e:
eutectic point.
Fig. 3 A schematic miscibility diagram between two pure compounds,
A and B. Filled and open circles represent melting points (TmA and
TmB) and clearing points (TcA and TcB) of the pure compounds,
respectively. Curves a and b: freezing depression curves. Curve c:
clearing point curve. Point e: eutectic point.
J. Mater. Chem., 2001, 11, 1063–1071
1067

View Article Online
Fig. 4 Temperature-dependent X-ray diffraction patterns of C16O–Ni
at rt, 107 and 150 ‡C. The peaks denoted D1 (b) and D2 (g) in this figure
are reflections due to the brown D1 and green D2 phases, respectively.
Fig. 5 Temperature-dependent X-ray diffraction patterns of C18O–Ni
at 90 and 150 ‡C. The peaks denoted D1 (b) and D2 (g) in this figure are
reflections due to the brown D1 and green D2 phases, respectively.
3.3 Thermal behavior of CnO–Ni (n~1–18)
observed for a transition from brown crystal to green crystal.
For n~9, it was observed for a transition from brown crystal
K2 to green mesophase D (g).
In Fig. 1, phase transition temperatures of the virgin samples
and the non-virgin samples are separately plotted against the
carbon number (n) in the alkoxy-side-chain. As can be seen
from this figure, each of the virgin samples n¢10 gives both
brown and green mesophases, whereas the non-virgin sample
does not show the brown mesophase but only the green
mesophase.
Although we reported the C11O–Ni complex gives only one
mesophase in our previous papers,9,10 the present detailed
reinvestigation revealed that it has two differently colored
mesophases like the other complexes for n~10, 12–18.
Therefore, we correct the phase transition sequence of the
C11O–Ni complex as shown in Table 2. Although Veber et al.
Table 3 summarizes the phase transition temperatures and the
phase transition enthalpy changes (DH). The CnO–Ni com-
plexes for n~1–4 were black needle-like crystals and the
crystals directly melt into a green isotropic liquid (I.L.) with
rapid decomposition. The CnO–Ni complexes for n~5–8 show
not a mesophase but crystal–crystal phase transitions. C9O–Ni
shows a green mesophase, and each of the complexes for n¢10
shows two differently colored mesophases. The results of the
polarizing microscopic observation are described for the
representative C12O–Ni complex in the following.
When the virgin sample of the brown K2 crystals was heated,
it melted into a brown D1 mesophase at 73.0 ‡C. On further
heating, a slow transition from the brown D1 phase to a green
D2 phase could be noticed at around 125 ‡C. The green D2
phase cleared into a green I.L. at 170.8 ‡C. On cooling down to
around 167 ‡C, the green I.L. changed into green smooth plates
surrounded by lustrous rings. On further cooling, the green
plates and rings were cracked and changed into green K1
crystals at around 56 ‡C. When the green K1 phase was heated
again, it melted into the green D2 phase at 56.2 ‡C and then the
green D2 phase cleared into the green I.L. at 170.8 ‡C. Thus, for
the non-virgin sample, neither the brown K2 nor D1 phases
appeared. Hence, only virgin sample of the complex can give
the brown K2 crystalline phase and the brown D1 mesophase.
However, it should be emphasized that the brown virgin form
could return on recystallization of the green non-virgin sample
from the solvent, ethyl acetate.
In DSC measurements, the peak due to the transition from
brown D1 phase to green D2 phase showed a heating rate
dependence. The faster the heating rate was, the higher the
transition temperature became (122–131 ‡C). This phenom-
enon was observed for each of the CnO–Ni complexes for
n¢10. Thus, the color change from D1 (b) to D2 (g) is
significantly slow. For n~7–8, the color change could be
˚
Fig. 6 Interlayer distance (A) versus number of carbon atoms in the
alkoxy chain (n). d1 (b) and d2 (g) represent the interlayer distances due
to the brown D1 and green D2 phases, respectively.
1068
J. Mater. Chem., 2001, 11, 1063–1071

View Article Online
Fig. 8 Tentative transformation from the nickel–nickel bonded dimer
D1 mesophase to the nickel–sulfur bonded dimer D2 mesophase.
mentioned above, the non-virgin sample of the C11O–Ni
complex gave a short curve corresponding to the transition
from the brown D1 phase to green D2. The curve obviously
starts from the transition temperature from the brown D1
phase to green D2 at 107 ‡C of the C11O–Ni complex and
disappears at ca. 80 wt%. The reason for the appearance of this
curve is not clear at the present time.
As can be seen from this figure, an area of green D phase
exists over two freezing depression curves (a and b) and under a
clearing point curve (c). Generally, a pure compound can be
thermodymanically divided into two phases between the
crystalline phase (K) and liquid (L) at the melting point
(Tm), as follows:
Fig. 7 Solid absorption spectra of the green K1 and brown K2 phases of
C12O–Ni. Fine powders of these solids were dispersed in transparent
silicone oil and placed between two P-102 glass plates. This dispersion
avoided the dissolution of these solids into their green solution, so that
the brown colored K2 crystals remained without changing the color.
Tm
K ? I
reported only one mesophase for the C12O–Ni complex,8 it has
two mesophases of brown D1 and green D2 as described above.
Hence, the mp (Tm) is the boundary between K and L. For a
mixture of two components, A and B, as schematically
illustrated in Fig. 3, both of the freezing point depression
curves (a and b) start from their melting points of the pure A
and B components (TmA, TmB). Therefore, these freezing
depression curves, a and b, also show the boundary between
crystalline phase and liquid. Hence, the area (M) over these two
freezing depression curves (a and b) with a eutectic point (e)
must have the characteristics of a liquid, thermodynamically.
Generally, liquid crystalline phases of a mixture of two
compounds can be seen in areas over two freezing depression
curves (a and b) with a eutectic point (e) and under a clearing
point curve (c). As can be seen from Fig. 2, the area denoted as
D (green) phase could be observed over two freezing depression
curves (a and b) with the eutectic point (e: 51 wt% 43.8 ‡C) and
under the clearing point curve (c). Therefore, we could judge
that this area of D (green) thermodynamically shows a
mesophase.
3.4 Miscibility diagram between C11O–Ni and C9O–Ni
Previously, we carried out miscibility tests between the C11O–
Ni and C9O–Ni complexes9,10 in order to establish mesomorph-
ism of these CnO–Ni complexes. However, at that time we
noticed only a green mesophase for the C11O–Ni complex.
Since we found another brown mesophase for this complex in
this work, we reexamined the miscibility between the C11O–Ni
and C9O–Ni complexes. As a result, a miscibility diagram was
obtained as shown in Fig. 2. These tests were carried out for the
non-virgin samples. Although the non-virgin sample cannot
give the phase transitions from brown D1 to green D2 as
Table 4 Main features of the experimental electronic spectra for the
C12O–Ni and Ni(S2C2H2)2 complexes
a
C
12O–Ni
Ni(S2C2H2)2
3.5 Temperature-dependent X-ray diffraction studies on the
CnO–Ni complexes
Green K1 solid
film lmax/nm
Brown K2
solid film lmax/nm
In CHCl3
max/nm
In n-hexane
lmax/nm
l
Fig. 4 shows X-ray diffraction patterns of C16O–Ni at rt, 107
and 150 ‡C. The pattern at rt shows many fine sharp reflections,
which is typical for crystalline phases. The pattern at 107 ‡C
show a halo at around 2H~20, which corresponds to the
molten alkoxy chains. In the low angle region, two sets of
periodic reflections corresponding to two lamellar structures
could be observed, but the longer- and shorter-spacing of
reflections became bigger and smaller, respectively, with time,
holding at 107 ‡C. This gradual change could be detected over
860
719
550 B1u (z)
444
370 B2u (y)
290
270
ca. 600
ca. 500
ca. 420
625
468
337
302
280
ca. 500
ca. 420
210
aRef. 19.
J. Mater. Chem., 2001, 11, 1063–1071
1069

View Article Online
Table 5 The reduction potential of the complexes, Cn–Ni and CnO–Ni
complexes consist of a disk-like core and four peripheral
chains. Hence, we concluded that both the D1 and D2 phases
are discotic lamellar mesophases (DL). The brown D1
mesophase did not give a natural texture from the green I.L.,
whereas the green D2 mesophase gave a natural plane texture
surrounded by a lustrous ring as reported previously.10 This
texture resembled very well the texture of the discotic lamellar
mesophase D1 in C8–Cu(II)18 which had been well established
as DL2 by miscibility phase diagram and X-ray diffraction
studies.19 These facts also support our present conclusion. It is
very interesting that the complexes which consist of the same
disk-like core and eight peripheral chains show a hexagonal
columnar mesophase (Colh), as previously reported.20
a
E1/2 of Cn–Ni
n
Liquid crystallinity
Non-mesogen
0b
1
2
3
4
5
6
7
8
9
10
11
12
0.03
20.01
20.02
20.01
20.01
20.01
20.03
20.03
20.03
20.03
20.03
20.03
20.02
The layer distances of these brown D1 and green D2
mesophases [d1 (b), d2 (g)] are plotted against number of the
carbon atoms in the alkoxy chain (n), in Fig. 6. As can be seen
a
E1/2 of CnO–Ni
n
Liquid crystallinity
Non-mesogen
˚
from this figure, the d2 (g) distance is about 2 A longer than the
d1 (b) distance for all the complexes.
1
2
3
4
5
6
7
8
20.05
20.06
20.06
20.06
20.05
20.06
20.05
20.05
3.6 Thermochromism of the CnO–Ni complexes
The CnO–Ni complexes for 7¢n show a color change from
brown to green for their virgin samples. For 10¢n, the
transition from the D1 phase to the D2 phase is accompanied by
a color change from brown to green, as described above. In
order to establish this color change for C12O–Ni, we carried out
measurements of the solid absorption spectra of the visible
region for the brown K2 crystalline phase of the virgin sample
and the green K1 crystalline phase of the non-virgin sample.
These solid fine powder samples were dispersed in transparent
silicone oil and placed between two glass plates (Glass techno,
P-102 glass) which are transparent at 2100–300 nm.
9
20.06
20.06
20.07
20.06
Mesogen
10
11
12
aVolts vs. SCE in CH2Cl2. bBis(diphenyldithiolene)nickel.
one hour. These longer-spacings of reflections correspond to
reflections of the higher temperature mesophase D2. For the
DSC measurements at 10 ‡C min21 heating rate, the transition
from D1 to D2 could be observed as a broad endothermic peak
at 111–118 ‡C. However, this X-ray diffraction pattern was
recorded, holding the temperature at 107 ‡C. This means that
the transition is so slow that it was detected at higher
temperature for the short-time DSC measurements, compared
with the long-hour X-ray diffraction measurements. Hence, it is
obvious that the brown D1 mesophase is not so stable even
under the D1–D2 phase transition temperature observed by
DSC. This instability made it very difficult to record an X-ray
diffraction pattern of the pure brown D1 mesophase, but the D1
phase could be assigned as a lamellar type of mesophase. The
pattern at 150 ‡C showed a set of periodic reflections, (001),
(002), (003), (004) and (005), at low angle region corresponding
to a lamellar structure, and a very broad halo due to the molten
alkoxy chains in the medium angle region. Hence, this D2 phase
could also be assigned as a lamellar type of mesophase. The
difference between these two lamellar mesophases is color: the
D1 and D2 mesophases are brown and green, respectively.
Fig. 5 shows X-ray diffraction patterns of C18O–Ni at 90 and
150 ‡C, corresponding to the brown D1 and green D2
mesophases, respectively. Even at 90 ‡C, which is low enough
from the D1 (b)–D2 (g) phase transition temperature (103–
111 ‡C) observed by DSC, the X-diffraction pattern still
contains some reflections due to the D2 (g) mesophase
(Fig. 5). From these X-ray patterns, both the brown D1 and
green D2 mesophases could be assigned as lamellar types of
mesophases.
As shown in Fig. 7, the spectrum of the brown K2 crystalline
phase gave three peaks at ca. 420, ca. 500, and ca. 600 nm. On
the other hand, that of the green K1 crystalline phase gave two
peaks at ca. 420 and ca. 500 nm. Thus, the color change from
brown to green originates from the lack of the peak at ca.
600 nm. It can be thought that these spectra of the brown K2
and green K1 crystalline phases correspond to those of the
brown D1 and green D2 mesophases, respectively. It was
revealed from X-ray analysis that each of the D1 and D2
mesophases is a DL mesophase. Since the spectral difference
between the D1 and D2 mesophases is regarded as the same as
that between the K2 and K1 crystalline phases, the transition
from the brown D1 phase to the green D2 phase corresponds to
the lack of the peak at ca. 600 nm.
These peaks were assigned by using electronic spectral
analysis for the core complex, bis(ethylene-1,2-dithiolato)-
nickel, Ni(S2C2H2)2, reported by Herman et al.21 In Table 4 are
collected the absorption spectral data of the green K1
crystalline phase, the brown K2 crystalline phase, and the
chloroform solution (green) of the C12O–Ni complex, and that
of the n-hexane solution of Ni(S2C2H2)2 reported by Herman et
al. The peaks at 550 and 370 nm for Ni(S2C2H2)2 in n-hexane
have been assigned to B1u and B2u symmetry, respectively. The
B1u symmetry contains LpA(Ls*z0.21dz220.84Ni s) transi-
tion, which is the transition in the z-axis direction and suggests
the existence of the Ni–Ni bond. On the other hand, the B2u
symmetry contains LpALp* transition in the y-axis direction.
These peaks at 550 and 370 nm of the core complex correspond
to ca. 600 and ca. 420 nm peaks of the solid absorption spectra
of the present C12O–Ni complex. Therefore, the phase
transition from the brown D1 phase to the green D2 phase is
accompanied by a lack of the peak at ca. 600 nm, which
suggests that the intermolecular Ni–Ni bonds in the z-axis
direction are broken or weakened. The tentative molecular
models in these two DL mesophases are proposed as shown in
Fig. 8. In the brown D1 phase, two molecules form a dimer
through the Ni–Ni bond between them. When it is heated, the
Ni–Ni bond is gradually broken and a new Ni–S bond between
Generally, disk-like molecules containing more than six side-
chains pile up one-dimensionally to give columnar mesophases.
On the other hand, four-chain-substituted disk-like molecules
tend to show discotic lamellar mesophases. Although many
columnar liquid crystalline compounds have been reported,
very few discotic lamellar liquid crystals have been reported.
Only four long chain-substituted disk-like metal complexes
show discotic lamellar mesophases in most cases.17 The present
1070
J. Mater. Chem., 2001, 11, 1063–1071

View Article Online
these two molecules slowly forms instead because the Ni–Ni
bond is at a higher energy level than a Ni–S bond, that is, the
Ni–S bond is more stable than the Ni–Ni bond. Hence, two
molecules slip over each other.22 As shown in Fig. 8, the layer
distance of the green D2 phase (d2 (g)) should be longer than
that of the brown D1 phase (d1 (b)). As can be seen from Fig. 6,
the d2 (g) distance is actually longer than the d1 (b) distance for
all the complexes for n¢10.
Acknowledgements
This work was partially supported by Grant-in Aid for
Research (12129205) by the Ministry of Education, Science,
Sports and Culture of Japan.
References
Thus, the thermochromism (brownAgreen) is attributable to
the slow transformation from the Ni–Ni bonded dimers to the
Ni–S bonded dimers.
1
Part 30: K. Hatsusaka, K. Ohta, I. Yamamoto and H. Shirai,
J. Mater. Chem., 2001, 11, 423.
2
A. M. Giroud-Godquin and P. M. Maitlis, Angew. Chem., Int. Ed.
Engl., 1991, 30, 375; K. Ohta and I. Yamamoto, J. Synth. Org.
Chem. Jpn., 1991, 49, 486.
3.7 Electrochemistry
3
4
5
6
7
A. M. Giroud and U. T. Mueller-Westerhoff, Mol. Cryst. Liq.
Cryst., 1977, 41, 11.
A. M. Giroud and U. T. Mueller-Westerhoff, Mol. Cryst. Liq.
Cryst., 1980, 56, 225.
U. T. Mueller-Westerhoff, A. Nazzal, R. J. Cox and A. M. Giroud,
Mol. Cryst. Liq. Cryst., 1980, 56, 249.
U. T. Mueller-Westerhoff, A. Nazzal, R. J. Cox and A. M. Giroud,
J. Chem. Soc., Chem. Commun., 1980, 497.
P. M. Cotrait, J. Gaultier, C. Polycarpe, A. M. Giroud and
U. T. Mueller-Westerhoff, Acta Crystallogr., Sect. C, 1980, C39,
833.
M. Veber, R. Fugnitto and H. Strzelecka, Mol. Cryst. Liq. Cryst.,
1983, 96, 221.
K. Ohta, A. Takagi, H. Muroki, I. Yamamoto, K. Matsuzaki,
T. Inabe and Y. Maruyama, J. Chem. Soc., Chem. Commun., 1986,
884.
The half-wave potentials E1/2 for the first reduction of these
complexes, Cn–Ni (n~1–12) and CnO–Ni (n~1–12), are listed
in Table 5 with their liquid crystallinity in the right-hand
column. In this table, n shows the number of the carbon atoms
in peripheral chains and n~0 corresponds to the core complex,
bis(diphenyldithiolene)nickel. From this table, it is apparent
that the reduction potentials of these complexes are nearly
constant irrespective of the alkyl or alkoxy chain length. The
values, 20.01 to 20.03 vs. SCE, observed for Cn–Ni (n~1–12)
are somewhat less positive than that of the core complex
(z0.03 V vs. SCE), whereas they are more positive than those
of the analogous complexes, CnO–Ni. The values, 20.05 to
20.07 V vs. SCE, observed for CnO–Ni (n~1–12) are much less
positive than that of the core complex. Thus, the reduction
potentials mainly depend on the types of the substituents. It is
attributable to the more electron donating property of the
alkoxy substituents than that of the alkyl substituents. From
Table 5, we can derive a very interesting relationship between
the mesomorphic property and reduction potential for
changing the alkoxy chain length. The mesomorphic property
of the complexes depends on the alkoxy chain length, whereas
the reduction potential does not. That is to say, we can change
the peripheral chain length to obtain the mesomorphic
properties without changing the electrochemical properties.
This seems to be very useful for the functionalization of
mesogenic compounds.
8
9
10 K. Ohta, A. Takagi, H. Muroki, I. Yamamoto, K. Matsuzaki,
T. Inabe and Y. Maruyama, Mol. Cryst. Liq. Cryst., 1987, 147, 15.
11 M. Veber, P. Davidson, C. Jallabert, A. M. Levelut and
H. Strzelecka, Mol. Cryst. Liq. Cryst., Lett., 1987, 5, 1.
12 A. Takagi, MSc Thesis, Shinshu University, Ueda, 1987, Ch. 1, 2
13 K. Tamao, K. Sumitani and M. Kumada, J. Am. Chem. Soc.,
1972, 94, 4374; M. Kumada, K. Tamao and K. Sumitani, Org.
Synth., 1978, 58, 127.
14 N. Sonoda, Y. Yamamoto, S. Murai and S. Tsutsumi, Chem. Lett.,
1972, 229.
15 H. Ema, MSc Thesis, Shinshu University, Ueda, 1988, Ch. 7
16 G. Wenz, Makromol. Chem., Rapid Commun., 1985, 6, 577.
17 K. Ohta, R. Higashi, M. Ikejima, I. Yamamoto and N. Kobayashi,
J. Mater. Chem., 1998, 8, 1979; and references cited therein.
18 K. Ohta, H. Murloki, A. Takagi, I. Yamamoto and K. Matsuzaki,
Mol. Cryst. Liq. Cryst., 1986, 135, 247.
19 K. Ohta, H. Muroki, A. Takagi, K. Hatada, H. Ema, I. Yamamoto
and K. Matsuzaki, Mol. Cryst. Liq. Cryst., 1986, 140, 131.
20 K. Ohta, Y. Inagaki-Oka, H. Hasebe and I. Yamamoto,
Polyhedron, 2000, 19, 276.
21 Z. S. Herman, R. F. Kirchner, G. H. Loew, U. T. Mueller-
Westerhoff, A. Nazzel and M. C. Zerner, Inorg. Chem., 1982, 21,
46.
22 S. Alvarez, R. Vicente and R. Hoffmann, J. Am. Chem. Soc., 1985,
107, 6253; I. Okura, N. Kaji, S. Aono, T. Kita and A. Yamada,
Inorg. Chem., 1985, 24, 453.
4. Conclusion
Two series of Cn–Ni (n~1–12) and CnO–Ni (n~1–12, 14, 16,
18), have been synthesized. It was established that the CnO–Ni
complexes exhibit two differently colored discotic lamellar (DL)
mesophases for n¢10, and that the thermochromism (brown
Agreen) is attributable to the slow transformation from the
Ni–Ni bonded dimers to the Ni–S bonded dimers. The
reduction potential of these complexes does not depend on
the chain length, whereas the mesomorphic property does.
J. Mater. Chem., 2001, 11, 1063–1071
1071