Chemical engineering of donor-acceptor liquid crystalline dyads and triads for the controlled nanostructuration of organic semiconductors

Article abstract of DOI:10.1039/c6ce00365f

Multi-segregated columnar structures provide a geometrically ideal scheme for ambipolar organic semiconductors, but are not easy to design. A set of novel materials with dyad and triad architectures based on 2 different discotic cores is reported and the conditions of emergence of such complex structures are investigated. The designed molecules associate together electron-donor triphenylene cores (D) and perylene or naphthalene diimides as acceptor moieties (A), both entities being linked via alkyl chain spacers. The evaluation in solution of their HOMO/LUMO energy levels by cyclic voltammetry demonstrates the preservation of the individual features of the D and A units. Their thermal and self-organization behaviors were studied by polarized-light optical microscopy, differential scanning calorimetry, temperature-dependent small-angle X-ray scattering and dilatometry, which permitted detailed investigation of the self-organization behaviour. These D-A compounds turned out to spontaneously self-organize into columnar mesophases at room temperature, with the D and A moieties segregated into either alternated stacks within mixed columns or in distinct columns, the latter providing an ideal configuration for 1D hole and electron transport pathways. In view of potential applications of the triad/dyad template, thin films of these self-organized materials were also probed by atomic force microscopy and grazing incidence X-ray scattering.

Full text of DOI:10.1039/c6ce00365f

CrystEngComm
View Article Online
View Journal
PAPER
Chemical engineering of donor–acceptor liquid
crystalline dyads and triads for the controlled
nanostructuration of organic semiconductors†
Cite this: DOI: 10.1039/c6ce00365f
Yiming Xiao,a Xiaolu Su,a Lydia Sosa-Vargas,a Emmanuelle Lacaze,b
c
c
Benoît Heinrich, Bertrand Donnio, David Kreher,a
*
a
a
*
*
Fabrice Mathevet and André-Jean Attias
Multi-segregated columnar structures provide a geometrically ideal scheme for ambipolar organic semi-
conductors, but are not easy to design. A set of novel materials with dyad and triad architectures based on
2 different discotic cores is reported and the conditions of emergence of such complex structures are in-
vestigated. The designed molecules associate together electron-donor triphenylene cores (D) and perylene
or naphthalene diimides as acceptor moieties (A), both entities being linked via alkyl chain spacers. The
evaluation in solution of their HOMO/LUMO energy levels by cyclic voltammetry demonstrates the preser-
vation of the individual features of the D and A units. Their thermal and self-organization behaviors were
studied by polarized-light optical microscopy, differential scanning calorimetry, temperature-dependent
small-angle X-ray scattering and dilatometry, which permitted detailed investigation of the self-organization
behaviour. These D–A compounds turned out to spontaneously self-organize into columnar mesophases at
room temperature, with the D and A moieties segregated into either alternated stacks within mixed col-
umns or in distinct columns, the latter providing an ideal configuration for 1D hole and electron transport
pathways. In view of potential applications of the triad/dyad template, thin films of these self-organized
materials were also probed by atomic force microscopy and grazing incidence X-ray scattering.
Received 12th February 2016,
Accepted 29th April 2016
DOI: 10.1039/c6ce00365f
www.rsc.org/crystengcomm
Indeed, these nanostructured morphologies offer the possibil-
ity to spontaneously form separate conduction pathways for
Introduction
The control of supramolecular self-assemblies and the nano-
scale morphologies of π-conjugated architectures in organic
semiconducting thin films can be seen as a critical issue for
the improvement of device performance in the field of or-
ganic optoelectronics.1 More particularly, the realization of
nanostructured 3D morphologies of different types of electro-
active constituents such as electron donor (D) or electron ac-
ceptor (A) building blocks is of great interest for organic solar
cell (OPV) and for ambipolar organic field effect transistor
(OFET) applications based on bulk heterojunction thin films.2
hole and electron charge carriers, consequent to the antago-
nistic chemical nature of each building block. This condition
is, however, not sufficient and optimal charge transport prop-
erties also imply the formation of appropriate long-range cor-
related structures with defined orientations and the ultimate
control of the regular pathway alternation.
A common strategy is based on the elaboration of conju-
gated D–A block copolymers, which show the ability to un-
dergo phase separation into well-ordered cylindrical or lamel-
lar morphologies with periodicities ranging from the tens to
hundreds of nanometers scale.3 However, these morphol-
ogies turn out to be difficult to realize and require fine
tuning of the block structures and lengths.
a Institut Parisien de Chimie Moléculaire (IPCM), Chimie des Polymères, Sorbonne
Université, UPMC Université Paris 06, UMR 8232, F-75005 Paris, France.
E-mail: fabrice.mathevet@upmc.fr, andre-jean.attias@upmc.fr
As an alternative route toward D–A nanostructured 3D
morphologies, the semiconducting architectures can be fitted
with liquid crystalline properties.4 Liquid crystals are indeed
outstanding examples of soft self-assemblies that combine or-
der and fluidity, within components self-organizing into a
wide diversity of long-range ordered and periodic structures.5
In addition, liquid crystal assemblies can form thin films
which are intrinsically self-healing since structural defects are
corrected automatically during the process of self-assembly.
b Institut des NanoSciences de Paris (INSP), CNRS-Université Pierre et Marie
Curie, UMR 7588, 4 Place Jussieu, 75005 Paris, France
c Institut de Physique et Chimie des Matériaux de Strasbourg (IPCMS),
CNRS-Université de Strasbourg, UMR 7504, F-67034 Strasbourg, France.
E-mail: bertrand.donnio@ipcms.unistra.fr
† Electronic supplementary information (ESI) available: Synthetic schemes, ex-
perimental procedures and chemical characterization of compounds, detailed
polymorphic data, DSC, SAXS in the pristine state and at higher temperature, in-
dexation tables of SAXS patterns of TPT, dilatometric curve of T, and GIXS in
the as prepared state of TP. See DOI: 10.1039/c6ce00365f
This journal is © The Royal Society of Chemistry 2016
CrystEngComm

View Article Online
Paper
CrystEngComm
Last but not least, the large variety of available π-conjugated
mesogens allows the accurate tuning of the electronic and
transport properties of the material through the control of the
segregation at the molecular level.6
Among the large diversity of such π-conjugated mesomor-
phic building blocks, the disk-shaped ones involved in
discotic liquid crystals (DLCs) are of particular interest.7
Consisting of a large π-conjugated polycyclic aromatic core
peripherally equipped with several alkyl chains, DLCs give
rise to phases in which the cores stack into columns produc-
ing an efficient 1D conductive pathway, whereas the sur-
rounding flexible alkyl chains play the role of soft insulators
between adjacent columns.
lumnar self-organization might even represent a more favor-
able alternative to lamello-columnar phases, since both types
of stacked species would then alternate in all directions of
the lattice plane. With this highest degree of intermingling
of conducting channels, the number of excitons photo-
generated close to the large inter-column donor–acceptor
interface and the conduction along the 1D hole and electron
transport pathways would be optimum. To the best of our
knowledge, such an ideal configuration for efficient forma-
tion of excited charge transfer states and good ambipolar
charge carrier mobilities has never been realized so far, as in-
deed the emergence of such multi-segregated columnar struc-
tures is not trivial.
Triphenylene and rylene diimide derivatives are archetypal
p-type (D) and n-type (A) DLCs, respectively, and each one
has extensively been studied for its charge transport and
photovoltaic properties.8,9 Nevertheless, only a few examples
of D–A mesomorphic molecular architectures based on the
combination of these entities have been reported until now,
and none of them give rise to self-organized structures with
long-range segregation between the D and A pi-fragments.4c,e
Zhao et al. in that regard very recently reported combined
triphenylene and perylenediimide or coronenediimide moie-
ties into D–A–D triad-type systems, which were found to self-
organize in a lamello-columnar mesophase giving distinct p
and n conducting channels through regular alternation of
two disrupted rows of triphenylene columns and one contin-
uous row of diimide species. The same favorable lamello-
columnar self-organization was also found for the three
triads, in relation to the side-connection through the short
spacers of the rylenes in the D–A–D architecture.10
In this context, novel dyad and triad D–A mesomorphic
materials based on discotic triphenylene and rylene diimide
fragments were synthesized and investigated for the possible
emergence of multi-segregated columnar structures. The fam-
ily of molecules presented hereafter associates electron-donor
triphenylene cores and rylene diimides as acceptor moieties,
both entities being linked via alkyl chain spacers
(Scheme 1a). The four compounds involve either a perylene
or a naphthalene fragment, since both are recognized n-type
DLCs with similar semiconducting properties but with differ-
ent geometries, which determine the mesomorphism. We
demonstrated that three of these D–A compounds turned out
to spontaneously self-organize into columnar mesophases at
room temperature, with the D and A moieties segregated into
either distinct columns or in alternated stacks within mixed
columns.
All the D–A systems were fully characterized by NMR and
standard techniques. Their electronic properties, and more
particularly their HOMO/LUMO energy levels, were evaluated
by cyclic voltammetry to show the potential of these materials
as ambipolar materials. Their thermal and self-organization
behaviors were further investigated by polarized-light optical
In the same line, end-connection by the tip through a lon-
ger spacer would split the continuous rylene row and possibly
yield true columnar mesophases, although the preservation
of the mesomorphic properties is not straightforward.11 Co-
Scheme 1 Chemical structures of a) the various targeted D–A mesomorphic dyads and triads and b) the individual model compounds:
triphenylene T, naphthalene N and perylene P.
CrystEngComm
This journal is © The Royal Society of Chemistry 2016

View Article Online
CrystEngComm
Paper
microscopy (POM), differential scanning calorimetry (DSC),
temperature-dependent small-angle X-ray scattering (SAXS)
and dilatometry, which permitted deep investigation of the
self-organization behavior. In view of potential applications
of the triad/dyad template, thin films of these self-organized
materials were moreover probed by atomic force microscopy
(AFM) and grazing incidence X-ray scattering (GIXS) to ad-
dress alignments on the substrate surface.
chain of the symmetrical diimide derivative P by partial
saponification under strong basic conditions. Finally, the
coupling between 5 and 8 by imidification reaction in melted
imidazole, catalyzed by zinc acetate (Lewis acid) according
to the literature procedure14 leads to the dyad in 75% yield.
Similarly, the triad TPT is obtained by coupling the triphenylene
amine 5 with the commercially available 3,4,9,10-perylene tetra-
carboxylic dianhydride 6.
Concerning the D–A dyad TN and D–A–D triad TNT homo-
logues based on the naphthalene acceptor unit, the previous
strategy developed for the perylene-based compounds could
not be applied. Indeed, our attempts to prepare the mono-
anhydride mono-imide naphthalene derivative failed. The
mono-imidification from 1,4,5,8-naphthalenetetracarboxylic
dianhydride (NTCDA) did not work, and the mono-cleavage
from the symmetric diimide naphthalene was also not possi-
ble. Thus, we developed a strategy based on a statistical ap-
proach, as described in Scheme 3. The imidification of
NTCDA was eventually performed in one-pot by adding the
two amine derivatives 5 and 12 in a 1 : 1 ratio. This way, the
dyad TN and triad TNT were obtained in acceptable yields af-
ter easy purifications.
Results and discussion
Synthesis
The D–A dyad TP and D–A–D triad TPT were prepared as
depicted in Scheme 2. Whatever the target molecule is, the
key intermediate is the triphenylene core bearing a single
alkyloxyamine side chain 5 prepared from the monohydroxy-
pentaalkoxytriphenylene 3 (see the ESI†).12 The unsymmetri-
cal triphenylene 3 was alkylated with the building block 10
bearing a “pre-Gabriel” functional group to form the interme-
diate 4 which was deprotected with an ethanolic hydrazine
solution to give the expected triphenylene amine 5. For the
synthesis of the dyad TP, the asymmetrically substituted
monoimide perylene 8 was prepared according to Thelakkat's
procedure13 based on the cleavage of one terminal branched
The model compounds T,15 N16 and P13 (Scheme 1b) were
also prepared.
Scheme
2 Synthesis of D–A dyad TP and D–A–D triad TPT based on triphenylene and perylene cores. Reagents and conditions: i′)
1-bromopentane (feeding ratio 2 : 1 to catechol), K2CO3, KI, tetrabutylammonium bromide, EtOH, reflux, 24 h, 83%; i) under the same conditions as
i′ with a feeding ratio of 1 : 1 with catechol, 79%; ii) FeCl3, DCM/MeCN, 0 °C, 2 h, 22%; iii) 10, K2CO3, DMF, 120 °C, 12 h, 82%; iv) aqueous hydrazine
hydrate, EtOH, reflux, 1.5 h, quantitative; v) 12, Zn(OAc)2, imidazole, 160 °C, 2 h, 81%; vi) KOH, t-BuOH, reflux, 1.5 h; AcOH; HCl, 43%; vii) Zn(OAc)2,
imidazole, 160 °C, 2 h, 75%; viii) Zn(OAc)2, imidazole, 160 °C, 2 h, 45%; ix) 1,10-dibromodecane, DMF, 100 °C, 72%; x) NH2OH, EtOH, pyridine, 100
°C, 2 h; RedAl, toluene, reflux, 2 h, quantitative.
This journal is © The Royal Society of Chemistry 2016
CrystEngComm

View Article Online
Paper
CrystEngComm
Scheme 3 Synthesis of D–A dyad TN and triad TNT based on triphenylene and naphthalene cores. Reagents and conditions: ZnIJOAc)2, imidazole,
160 °C, 2 h, TN 33%, TNT 20%.
The full synthetic and characterization details of TP, TPT,
TN, TNT, T, N and P are reported in the ESI.†
present similar electrochemical behaviors (see their CVs in
Fig. S1†), which allowed us to determine their redox potential
values. They are listed in Table S0 (ESI†).
Based on their redox potentials, the HOMO/LUMO energy
levels of the four molecules can be calculated by using the
following classical semi-empiric equations: estimated
Electrochemical behavior and HOMO/LUMO energies
The electrochemical behaviors of TP, TPT, TN and TNT were
investigated by cyclic voltammetry, the optimized experimen-
tal conditions being obtained with chloroform solutions (10−3
M) and by using 0.5 M n-Bu4NPF6 as the electrolyte. As an ex-
ample, the cyclic voltammogram (CV) of TP is reported in
Fig. 1.
In reduction, the voltammogram of TP shows two well-
separated and reversible one-electron reduction waves at
−0.67 and −0.86 V (vs. Ag/AgCl) related to the formation in
the perylene acceptor moiety of the radical anion and
dianion, respectively. In oxidation, a unique one-electron oxi-
dation wave corresponding to the signature of the
triphenylene donor moiety radical cation at 1.03 V (vs. Ag/
AgCl) is evidenced. The other derivatives TPT, TN and TNT
electron affinity (EA) or LUMO energy level, ELUMO = −(4.8 +
red1
E1/2
(vs. Fc)) and estimated ionized potential (IP) or
HOMO energy level, EHOMO = −(4.8 + E1/2 (vs. Fc)).17 Their
values are summarized in Table 1. The estimated electron af-
finity (EA) or ELUMO values are in the range of −3.77 to −3.87
eV for TP and TPT and −3.73 to −3.75 eV for TN and TNT, in-
dicating that for a perylene dyad/triad, the LUMO energies
are determined mainly by the LUMO energy of the perylene
moiety (−4.1 eV for the P model compound), and for the
naphthalene dyad/triad derivatives, the LUMO energies are
governed by the LUMO energy value of the naphthalene moi-
ety (−3.41 eV for the N model compound). Moreover, the esti-
mated ionization potential (IP) or EHOMO values are in the
range of −5.43 to −5.47 eV for the four derivatives, indicating
that these HOMO values are mainly the HOMO value
reported for the triphenylene moiety T (−5.4 eV), under the
same experimental conditions.
ox1
As a consequence, through comparison of all the reported
HOMO–LUMO values in Table 1, we can conclude that the
donor and acceptor units in this series of dyads and triads
do not interact with each other and that the connection of
the segments through the aliphatic spacers is a way to force
their juxtaposition and preserve their genuine electronic
properties.
Table 1 HOMO/LUMO energy levels of TP, TPT, TN and TNT
TP
TPT
TN
TNT
Ta
Na
Pa
E
E
HOMO (eV)
LUMO (eV)
−5.47
−3.77
1.7
−5.43
−3.87
1.56
−5.46
−3.75
1.71
−5.45
−3.73
1.72
−5.4
−1.7
3.7
−7.04
−3.41
3.63
−6.3
−4.1
2.2
ECgV (eV)
Fig.
1
Cyclic voltammetry (deconvoluted signal) of dyad TP in
n-Bu4NPF6. The
s−1 at room
a
chloroform solution (10−3 M) including 0.5
M
a
The HOMO/LUMO values of model compounds T,18 N,19 and P9d
experiment was carried out at
temperature.
a rate of 0.05 V
are published in the previous reports.
CrystEngComm
This journal is © The Royal Society of Chemistry 2016

View Article Online
CrystEngComm
Paper
Characterization of the self-organization behavior
(ca. −10 °C), by undergoing a sudden transition even visible
at a low rate. The stiffening occurs in this case within a few
degrees, as classically observed for organic liquids. For the
mesomorphic analogues, the process is anticipated at higher
temperatures and spread over tens of degrees, which means
that the collective motions responsible for the fluidity are
hindered by the molecular organization.
HexakisIJpentyloxy)-triphenylene, which is a representative
of the “T” unit in triads and dyads, is crystalline at room tem-
perature and melts at 65 °C to a Colhex phase formed by the
stacking of the triphenylene rings into columns and by the
melting of peripheral chains in a continuum.20 This structure
is demonstrated in the SAXS patterns by the presence of the
scattering signals hπ ≈ 3.5 Å, arising from the face-to-face
stacked rings, and hch ≈ 4.5 Å, associated with the lateral dis-
tances between molten chains, in addition to the series of
sharp reflections of the hexagonal lattice, namely the very
strong first order (d10 = 17.6 Å) and a few weak higher order
reflections (see Fig. S3†).
Although the triad/dyad architectures follow the electronic
and molecular properties of the constitutive acceptor/donor
segments, the separation of the latter in juxtaposed domains
is also required for the expression of the D–A couple proper-
ties in the final material. The different natures of the seg-
ments should intrinsically initiate distinct self-aggregation
processes, while the appropriate design of the interconnec-
tion and of a peripheral chain would ultimately lead to the
coveted juxtaposition of different domains isolated from each
other by a continuum of aliphatic chains. This molecular de-
sign naturally leads to self-organization in mesophases,
which were first investigated by POM and DSC (results sum-
marized in Table S1†), prior to deeper structural characteriza-
tion by SAXS.
As evidenced by POM, except the dyad TN which is an iso-
tropic liquid at room temperature, the other triads/dyad,
TNT, TP and TPT, are soft solids at room temperature and
gradually liquefy at higher temperature leading to birefrin-
The SAXS patterns of TNT, TP and TPT already contain
these diffuse signals hπ, hch and the sharp reflection (10) in
the pristine state at similar positions, revealing the develop-
ment of the columnar structure from the beginning (see Fig.
S4†). Consistently with the absence of mesophase observed
by POM, the patterns of TN only show a blurred shoulder
and a broad peak at the hπ and (10) locations, coming from
facing mesogens and from loose piles separated by chains.
Heating TP in the fluid state of the mesophase and subse-
quent cooling (see Fig. S4† and 3) sharpen and increase the
number of visible small-angle reflections, impressively rising
to six: (10), (11), (20), (21), (30), (22) (with q-ratios
1 : √3 : √4 : √7 : √9 : √12). The thermal treatment moreover
leaves the location and shape of the hπ scattering unmodified
(correlation length at 20 °C: ξ ≈ 100 Å, corresponding to ca.
30 mesogens) and thus restricts structural changes to a lat-
eral rearrangement of columns consecutive to the fluidization
of the aliphatic periphery. In TNT, the sharpening of the
gent fluid textures typical of
a mesophase, which are
maintained until the transition to the isotropic liquid, above
170 °C (TNT) or at 145–150 °C (TP and TPT). On cooling of
the isotropic liquid, TNT and TP display pseudo focal-conic
and dendritic textures, typical of columnar mesophases,
while unconventional dendritic textures indicative of low
symmetric structures are observed for TPT (Fig. 2). No tex-
tural changes occur on further cooling and the samples
therefore exhibit a single mesophase, as also confirmed by
the DSC traces that only display the isotropization peak
(Fig. 2 and S2†). In particular, crystalline states are not ob-
served for these compounds at low temperature, but the
mesophases gradually freeze in an anisotropic glassy state be-
tween ca. 80 °C and room temperature, the transitional range
being detectable as a slope change in DSC runs at a high
heating rate. In contrast, the non-mesomorphic TN rigidifies
into amorphous glass at a substantially lower temperature
Fig. 2 POM textures (left) of TN in the isotropic liquid state at 20 °C, and of TNT, TP and TPT in the mesophase on cooling, at 100 °C, 135 °C and
150 °C, respectively; DSC traces (right) recorded during the second heating run at 5 °C min−1 of T (hexakisIJpentyloxy)-triphenylene), with melting
and isotropization transitions at 65 °C and 121 °C, of TN, with a glass transition at −10 °C, of TNT, TP and TPT, with transitions to isotropic liquid at
173 °C, 146 °C and 159 °C, respectively. Inset view: DSC traces of the second heating run at 20 °C min−1 of TNT, TP and TPT, with a glass-like tran-
sition between frozen and fluid states of the mesophase.
This journal is © The Royal Society of Chemistry 2016
CrystEngComm

View Article Online
Paper
CrystEngComm
umn types in the pristine state is revealed by the doubling of
the main peak around 18 Å and by the appearance of an ad-
ditional reflection at double periodicity (see Fig. S4†). As for
the TP dyad, the pristine TPT triad structure correlates to
intermediate range and irreversibly transforms on first
heating above the glass transition, in a long-correlated orga-
nization characterized by the appearance of numerous sharp
reflections (see Fig. S5†). The π-stacking and the average sec-
tion of individual T or P columns are hardly affected by the
transformation, but the molecular stacks are rearranged in
the lattice plane, resulting in a symmetry break to an oblique
lattice (Colobl) (see Table S2†). The Colobl phase then is held
up to the isotropization, directly recovered on cooling from
the isotropic liquid and maintained on cooling in the frozen
room temperature state, with only little variation of the lat-
tice parameters with temperature. Small changes are never-
theless noticed in the frozen state, namely a slight re-
broadening of the small-angle reflections (corresponding to
roughly ξ = 400 Å) and the appearance of a weak wide-angle
scattering maximum D (see Fig. 3). Such features most likely
originate from the setting up of 3-dimensional correlations
between mesogens from neighboring columns. All the meso-
phase structural data are summarized in Table 2.
Fig. 3 SAXS patterns of triads/dyads TN, TNT, TP and TPT at room
temperature, in the isotropic liquid (TN) or in frozen mesomorphic
states (TNT, TP and TPT); curves on the top are zoomed to view weak
signals; the inset view in the TPT graph is an additional pattern with a
lower-angle SAXS line.
reflection (10) also proves the extension of the 2-dimensional
lattice to a longer range, but at the cost of the widening of
the hπ scattering (ξ ≈ 50 Å) in combination with the appear-
ance of only one higher order reflection. Hence the columnar
stacking of TNT is less regular in the extended 2-dimensional
lattice than the TP one. Finally, it should be stated that the
2-dimensional lattices of TNT and TP contain one single col-
umn; therefore the columns are composed of a priori
undifferentiated triphenylene/naphthalene and triphenylene/
perylene mesogenic stacks (Tmeso/Nmeso and Tmeso/Pmeso).
In contrast, the elemental T and P units of TPT turned out
to form distinct columns arranged in large lattices containing
several molecular stacks of two T and one P columns sepa-
rated by the aliphatic continuum. Differentiation of the col-
Description of bulk structures
Rylene diimide and triphenylene belong to the class of meso-
gens that self-organize by π-stacking and even show similar
stacking distances.22,23 The different chemical natures and
shapes of both types of mesogens make however their inti-
mate mixing in stacks highly unlikely and in contrast, pro-
mote self-association in distinct columns. Consistently, struc-
tures with differentiated rylene diimide and triphenylene
columns have recently been reported for other D–A triads;10
a
new one has been observed for TPT which will be described
now (see below).
Table 2 Mesomorphic structures and geometrical parameters of reference T (N = 1 mesogen per molecule), of dyads (N = 2) TN and TP, and of triads
(N = 3) TNT and TPT, at temperature T (20 °C on cooling)
T [°C]
Phasea
Parametersb
T
105
65
20
140
20
140
20
Colhex
Colhex
Iso
Colhex
Colhex
Colhex
Colhex
Colobl
a = 20.3 Å; A = 357 Å2 (Z = 1); hπ = 3.62 Å (ξπ = 60 Å) Scol = 357 Å2; Vmol = 1276 Å3; hmol = 3.57 Å
a = 20.3 Å; A = 356 Å2 (Z = 1); hπ = 3.49 Å (ξπ = 120 Å) Scol = 356 Å2; Vmol = 1244 Å3; hmol = 3.49 Å
D = 16.6 Å (ξ = 180 Å); σT/N = 290 Å2; Vmol = 1990 Å3; Hmol = 3.4 Å
TN
TNT
a = 18.5 Å; A = 297 Å2 (Z = 1/3); hπ = 3.53 Å (ξπ = 25 Å) Scol = 297 Å2; Vmol = 3180 Å3; hmol = 3.58 Å
a = 18.2 Å; A = 288 Å2 (Z = 1/3); hπ = 3.45 Å (ξπ = 50 Å) Scol = 288 Å2; Vmol = 2920 Å3; hmol = 3.38 Å
a = 19.6 Å; A = 334 Å2 (Z = 1/2); hπ = 3.52 Å (ξπ = 40 Å) Scol = 334 Å2; Vmol = 2300 Å3; hmol = 3.45 Å
a = 19.1 Å; A = 316 Å2 (Z = 1/2); hπ = 3.44 Å (ξπ = 100 Å) Scol = 316 Å2; Vmol = 2110 Å3; hmol = 3.35 Å
a = 79.3 Å; b = 39.5 Å; γ = 92.9°; A = 3120 Å2 (Z = 4); hπ = 3.61 Å (ξπ = 20 Å) a/b = 2.01; Scol = 260 Å2;
Vmol = 3350 Å3; hmol = 4.29 Å
TP
TPT
140
20
Colobl
a = 75.3 Å; b = 40.5 Å; γ = 93.7°; A = 3050 Å2 (Z = 4); hπ = 3.48 Å (ξπ = 50 Å)a/b = 1.86; Scol = 254 Å2;
Vmol = 3050 Å3; hmol = 4.01 Å
a
b
Colhex, Colobl: hexagonal and oblique columnar mesophases; Iso: isotropic liquid. Colhex, Colobl: a, b, γ, A (Z), hπ (ξ): 2-dimensional lattice
parameters, lattice area and number of molecular stacks per lattice, π-stacking distance from the scattering maximum and correlation length
(Scherrer); Iso: D (ξ), σ: lateral distance between mesogens from the peak maximum and correlation length, section of piled T/N units,
according to σT/N = (D/0.9763)2;21 Scol = A/(NZ): columnar cross-section; Vmol: molecular volume, directly measured for T or calculated from refer-
ence measurements for triads/dyads; hmol = (1/N)Vmol/Scol: molecular slice thickness; Hmol = (1/N)Vmol/σT/N: piling distance.
CrystEngComm
This journal is © The Royal Society of Chemistry 2016

View Article Online
CrystEngComm
Paper
The loss of differentiation in TP and TNT therefore ex-
cludes intimate mixing and is merely due to the permutation
of entire groups of identical mesogens. Exchanges of entire
columns between lattice nodes were reported for a previous
system,24 but this mechanism preserves the contrast between
neighboring columns at local range and thus combines with
the appearance of a diffuse superstructure signal not ob-
served here. Hence different groups of stacked mesogens of
the same nature alternate along the columnar axes. A lower
limit for the mesogen group size can be derived from the cor-
relation length ξπ of the stacking (see Table 2), which basi-
cally lies in the few tens of mesogens range in the columnar
phases and appears only slightly reduced with respect to the
reference compound T: ξπ/hπ ≈ 35 mesogens for T at 65 °C,
ξπ/hπ ≈ 30 mesogens for TP at 20 °C, ξπ/hπ ≈ 15 mesogens for
TNT at 20 °C. Whatever the real average group size, their per-
mutations between neighboring columns level out contrasts
and shape anisotropies in the plane of the lattice without al-
tering their long-range correlation, which ultimately leads to
the observed hexagonal P6mm arrangements of apparent
undifferentiated columns (see Fig. 4). Both mesogens are
therefore averaged to the same disc-shaped units of dimen-
sions defined by the columnar section Scol and by the molec-
ular slice thickness hmol, obtained as the ratio of the half-
dyad or third-triad volume and Scol. The accordance between
hmol and the π-stacking distances hπ confirmed the non-tilted
stacking inside the columns inherited from the Colhex phase
of the reference compound T (see Table 2 and Fig. 5).
The differences mainly come down to the shortened correla-
tion lengths: the lateral packing of the T/N units only corre-
lates over ca. ten units, the long-range orientational order
has vanished and the π-stacking signal has flattened to a
shoulder, revealing a very disordered piling.
The homogeneous aliphatic crown and the true disc-like
shape of Tmeso mesogens naturally promote hexagonal ar-
rangements which persist in the TP dyad, i.e. after the inter-
mingling of a stoichiometric amount of Pmeso between T
units. One might naively have expected that with only half
the amount of Pmeso, the Colhex phase would also be pre-
served, but instead the TPT triad organizes in a Colobl lattice
including twelve columns. The reason is obviously that a spe-
cific structural feature (discussed hereafter) restores the dif-
ferentiation between T and P columns and that the lattice
therefore involves an integer number of triad stacks. The dif-
ferent cross-sectional shapes are restored at the same time
and loom large constraints in the close packing of the lath-
shaped Pmeso (aspect ratio AR ≈ 1.5) with disc-shaped Tmeso
(AR = 1), while the areas are, in contrast, close (see next part).
These constraints certainly explain the appearance of a tilt-
angle of mesogens within columns whereas T and PT form
untilted mesophases. This angle of ca. 30° is directly deter-
mined from the ratio of the common stacking distance of
mesogens, hπ, and of the slice thickness of the structure, hmol
(see Fig. 5), and mechanically leads to 15% shrinking of the
lattice periodicities in the tilt directions.
The aspect ratio differences then get largely reduced when
both mesogens tilt in different directions, i.e. toward long
sides for Pmeso and in the transverse direction for Tmeso
(Fig. 6). The tilt induction can, however, not compensate for
the cross-sectional shapes and has no effect on other features
determining the organization such as the end-on connected
The local-range packing in the isotropic liquid phase of
the non-mesomorphic dyad TN resembles a disordered modi-
fication of this Colhex structure, within particular cross-
sectional areas of piled T/N units σT/N and piling distances
Hmol close to Scol and hmol in TNT (see Table 2 and Fig. 5).
Fig. 4 Schematic view of the self-organization within the Colhex mesophases of the TNT triad and TP dyad: left: entire groups of naphthalene or
perylene diimide mesogens, Nmeso or Pmeso (black), and of triphenylene mesogens, Tmeso (red), randomly alternate along and between columns;
middle: mesogen groups superpose into unique undifferentiated Nmeso/Tmeso or Pmeso/Tmeso columns of average cylindrical cross-section; right:
hexagonal p6mm lattice (green lozenge) formed by undifferentiated columns.
This journal is © The Royal Society of Chemistry 2016
CrystEngComm

View Article Online
Paper
CrystEngComm
which the orientation of the columnar cross-sections is only
correlated over a few first neighbors.
In the fluid state of the Colhex and Colobl phases, the col-
umns are largely free to slide along their axes relative to each
other and the mesophases are therefore true 2-dimensional
mesophases. This dynamics freezes in room temperature in
the solid state, and short-range correlations appear between
mesogen registries from neighboring columns in the Colobl
phase, resulting in a modified wide-angle scattering profile.
In contrast, the mixed T/P columns of the Colhex phase basi-
cally exclude such correlations and only freeze at the lattice
nodes in the solid state, which improves the regularity of the
2-dimensional arrangement.
Discussion of the self-organization process
The self-organization type observed with these compounds
basically differs from the lamello-columnar mesomorphism
recently reported for other triads based on the same T-units
linked to rylene cores.10 The connection through short
spacers in the previous triads (5 methylenes compared to 10
here) was found to promote the lamellar character of the mo-
lecular organization, which was moreover reinforced by the
side-on connection and the comparable sizes of both moie-
ties. In contrast, end-on connected rylene cores self-
assembled in layers would associate laterally through their
short sides and therefore need to tilt by very large angles (60–
65°) to equal the area imposed by the alternating rows of T
columns. Such large tilt angles would strongly constrain the
arrangement and moreover realize lesser efficient segregation
with the aliphatic moieties, so that the entire surrounding of
the rylene columns by the aliphatic continuum and the for-
mation of true columnar phases are finally preferable to the
lateral association in layers.
Fig. 5 Columnar cross-section Scol (top) and molecular slice thickness
hmol (bottom, symbols and solid lines) compared to the π-stacking dis-
tance from the scattering maximum hπ (dotted lines), as a function of
temperature T, for triads/dyads TN (red open circles, σT/N and Hmol in-
stead of Scol and hmol (see text)), TNT (blue lines and solid circles), TP
(magenta lines and triangles), TPT (green lines and triangles), and for
reference T (black lines and rectangles); lines are guides for the eyes.
spacers of the Pmeso compared to the crown-like aliphatic pe-
riphery of the Tmeso. During annealing in the fluid state, the
twelve columns of the Colobl lattice then develop as the opti-
mal lateral arrangement of these unequal columns, for which
schematic views based on symmetry, observed reflection se-
ries, geometrical parameters and molecular distances can be
proposed (see Fig. 6). Conversely, the pristine state can be
understood as an early stage of the same arrangement, in
Fig. 6 Schematic view of the self-organization within the Colobl mesophase of the TPT triad; middle: the Pmeso and Tmeso mesogens stack into dif-
ferentiated columns with a tilt angle of ca. 30° reducing their projection in the lattice plane; right: possible arrangements within the plane of the
oblique lattice (green frame); the blue dotted line defines a distorted honeycomb lattice formed by T units and surrounding the P units at the cen-
ters; the continuum of molten aliphatic chains fills the spaces between columns.
CrystEngComm
This journal is © The Royal Society of Chemistry 2016

View Article Online
CrystEngComm
Paper
Various structures according to this template arise in the
triad/dyad series, depending upon the ability of the individ-
ual compounds to preserve large groups of stacked mesogens
and differentiated columns. Before deepening these aspects,
the influence of the assembling process in the triad/dyad ar-
chitecture and of the restricted molecular diffusion should
be considered, for which the comparison of T and its dimer
TT might be a pertinent reference. The isotropization temper-
ature of the dimer was actually shown to lie 14 °C above that
of the monomer (e.g. T), but the mesophase structure was
identical and the melting transition did not move signifi-
cantly.20 In analogy with a previous work, a portion of the iso-
tropization temperature increase noticed for TNT, TPT and
TP might originate from the general architecture, but the
other self-organization features are likely due to the associa-
tion/connection of the different mesogens.
The most remarkable aspect is maybe the different effects
of the rylene proportion on both cores: the TNT triad shows a
Colhex phase as the reference T and with even the highest iso-
tropization temperature of the series, but the mesomorphism
vanishes with the increase of the Nmeso proportion to 1 : 1
(TN dyad). In contrast, the Colhex phase reappears for the
highest Pmeso proportion (TP dyad), while the intermediate
TPT triad exhibits a Colobl phase. Yet Nmeso and Pmeso are
clearly similar considering their chemical nature, lateral in-
teractions and segregation power with respect to the alkyl
chains and to other mesogens. Their principal difference is
actually their size, which requires to be compared to Tmeso to
understand the potential role of this characteristic on the
self-organization. For this purpose, T was compared to two
analogous reference compounds, P and N, with the corre-
sponding rylene mesogen and aliphatic peripheries equiva-
lent to the one of T. These new references are non-mesomor-
phic, but the volumes can accurately be compared in the
isotropic liquid phase, from which the volume in the meso-
phases deviate by less than one percent (see Fig. 7). After sub-
traction of the alkyl chain contribution, it appears that meso-
gens Nmeso and Pmeso have volumes 30% below and 5% above
Tmeso, respectively. Considering the similar stacking dis-
tances hπ found for all compounds at the same temperatures,
(see Fig. 5), the mesogen sections lie in the same ratios as
the volumes, thus for hπ ≈ 3.45 Å at room temperature:
SNmeso = 84, SPmeso = 121, STmeso = 116 Å2.
Fig.
(hexakisIJpentyloxy)-triphenylene),
naphthalene-diimide) and
diimide) measured versus temperature by dilatometry; Cr: crystal
phase; Colhex: hexagonal columnar mesophase; Iso: isotropic liquid;
dotted line: extrapolation; for the complete experimental curve of T,
extending to the isotropic liquid phase, see Fig. S6.†
7
Molecular
volumes
of
reference
(N,N′-diIJpentadecan-8-yl)-
(N,N′-diIJpentadecan-8-yl)-perylene-
compounds
T
N
P
aliphatic chains and differentiation of columns limited to lo-
cal range. Instead, it turned out preferable to maintain the
long-range correlated lattice and lose the differentiation.
Hence the distribution of groups of identical mesogens over
all columns yielded the single-column Colhex phase.
The three columns per elementary hexagonal cell (two for
the distorted honeycomb and one for the centers) prevent
any separation of column types under the conditions of 1 : 1
stoichiometry. Long-range differentiation should therefore
not occur in the dyads, but still could be preserved at the lo-
cal range, which is indeed observed for TP. In this way, Tmeso
and Pmeso remain separated in different groups over the cor-
relation length of the stacking, but at a larger distance, the
equivalence of the lattice nodes imposed by the stoichiometry
just results in the random distribution of groups in the lat-
tice and in the related symmetry elevation to Colhex. The
same organization appears with Nmeso already for the triad,
but in relation to its small size hampering the insertion be-
tween six neighboring T columns. The 2 : 1 stoichiometry pro-
motes long-range honeycomb lattices and relaxes these steric
constraints through frequent alternation of the mesogen na-
ture at the lattice nodes. This averaging effect however van-
ishes in the dyad and the irregular environment of the Nmeso
groups looms large, which ends up with a strongly disordered
piling and the disappearance of the mesomorphism, likely in
relation to a more intimate mixing of both mesogens.
As discussed above, the 2 : 1 stoichiometry of the triads is
intrinsically favorable to differentiation, since the major com-
ponent can occupy the nodes of the (regular or distorted)
honeycomb lattice and the minor component, the centers of
the hexagonal cells. The circumscribed area at the centers is
however identical to the area at the honeycomb lattice nodes
and the formation of a regular structure therefore requires
that both components have approximately the same cross-
sectional area. This is obviously the case for the Tmeso/Pmeso
couple and differentiation therefore spontaneously appears
for TPT. The smaller size of Nmeso in contrast prevents the
formation of a regular honeycomb lattice in TNT. This could
have resulted in an amorphous state with segregation from
Spontaneous alignment in thin films
Applications of organic semiconducting materials often re-
quire thin films with controlled morphology.7c When these
materials exhibit mesophases, the film preparation is facili-
tated by the long-range orientational order combined with
the fluidity and the self-healing ability against structural de-
fects inherent to mesomorphism.6b,c In addition, the associa-
tion of different mesogens in the triad/dyad architectures
This journal is © The Royal Society of Chemistry 2016
CrystEngComm

View Article Online
Paper
CrystEngComm
strongly hampers the crystallization and the structures
forming at higher temperature just freeze on cooling at room
temperature.10 Our triad/dyads meet these requirements: they
link recognized semiconducting units, most show meso-
morphism and none does crystallize, which opens up possi-
bilities to freeze morphologies having been obtained in the
fluid state of mesophases and limiting the grain boundary
formation. The results depend on the architecture but also
on the type of mesophase, which is addressed hereafter for
the mesomorphous triad/dyad pair TPT/TP.
For this purpose, tens of nanometer scale films of the TP
dyad were deposited by spin-coating on top of a silicon wafer
from chloroform solutions. The GIXS patterns of the as pre-
pared films prove the self-organization behaviour, which is
identical to the pristine bulk structure, namely a Colhex phase
of undifferentiated T/P columns (see Fig. S7†). As revealed by
the unique broadened signal, the organization involves a lot
of disorder, in relation to the frozen state at the deposition
temperature (room temperature) and with the fast deposition
technique. Surprisingly, the domains turned out to be spon-
taneously oriented at this stage of the preparation. Consistent
with previous findings for classical discotic systems,25 this
spontaneous orientation consists of planar lying column
rows, as demonstrated by the splitting of the first order re-
flection in spots at 0° and 60° from the meridian and by the
hπ scattering located at the equator. The annealing step in
the fluid state of the mesophase logically does not modify
this already optimal orientation, but develops the long-range
order similarly to that in the bulk. The appearance of numer-
ous sharp dots corresponding to higher order reflections
then attests to the regularity of the structure and to the
nearly perfect planar orientation (see Fig. 8, top left). Films
prepared by the same procedure and studied using tapping
mode atomic force microscopy (AFM) appear relatively uni-
form before thermal treatment, with no specific nanostruc-
tures and with roughness less than 1 nm (see Fig. S9†). After
their annealing in the fluid state of the mesophase, the films
become clearly nanostructured, with terraces of 1.7 nm or of
multiples of 1.7 nm height (see Fig. 8 top right). This height
coincides with the inter-row spacing d10 from the SAXS pat-
terns, in agreement with the planar orientation of the col-
umns evidenced in the GIXS patterns.
In the highly symmetric Colhex structure of TP, the surface
alignment of the {10} planes turned out to be preferred over
the alternative alignment of the lattice plane, which means that
the interactions at the interfaces favor an edge-on orientation
of the mesogens over face-on. TPT is constituted by the same
mesogens and should thus also lead to lying column align-
ments in the mesophase. This was indeed verified in some
patterns showing an intense hπ scattering on the equator
Fig. 8 Top. GIXS pattern (left) compared to topography AFM image in tapping mode (right) of thin films of dyad TP, spin-coated on silicon wafer
and annealed at 120 °C for 2 hours in vacuum; inset views: alignment of mixed T/P columns (red & black discs) on top of the substrate (lying col-
umns), orientation of the crystallographic axes (blue) and height profile (the green arrow in the AFM image shows the profile direction). Bottom:
GIXS pattern of the TPT triad (left), dip-coated on silicon wafer and annealed at 120 °C for 2 hours in vacuum; inset views: alignment of tilted T
and P columns (yellow-red and grey shaded) in the Colobl phase on top of the substrate (lying columns), corresponding orientation of the crystallo-
graphic axes (blue) and orientation of the crystallographic axes for crystal-like domains (green); profiles of GIXS pattern (right) within sectors ψ
3°, for azimuthal angles ψ = 0°, 25°, and 87° (origin on the meridian).
CrystEngComm
This journal is © The Royal Society of Chemistry 2016

View Article Online
CrystEngComm
Paper
beside the molten chain ring (see Fig. 8, bottom). The small-
angle spots related to the same mesophase moreover confirm
the arrangement in the Colobl lattice, with the alignment of
the a*-axis on the meridian and a somewhat smaller elonga-
tion with respect to the bulk structure (a ≈ 70 Å, b ≈ 45 Å; γ
= 98 3°). In coexistence with the major Colobl fraction, the
film contains domains of another phase for which the stack-
ing direction flips to the surface normal and the correspond-
ing scattering signal splits in semi-diffuse bows lying close to
eventually shows that by relevant chemical engineering it is
possible to control the nanostructuration of D–A discotic
multiads to give rise to differentiated D and A columnar ar-
rangements. Since such molecular organizations provide the
ideal configuration for 1D hole and electron transport and
are highly desirable for potential semiconducting applica-
tions, this work opens up promising opportunities for the
elaboration of ambipolar devices and, more widely, for opto-
electronic applications.
¯
the meridian. This indicates the appearance of (00l)/(hkl)/(hkl)
reflections from a 3-dimensional phase, with crystalline order
and still molten alkyl chains. Depending on the conditions of
film preparation, the Colobl domains even nearly vanished
and were replaced by further crystalline-like states, whose in-
vestigation would have required additional measurements
and is beyond the scope of the contribution (see example in
Fig. S8†).
The occurrence of these crystal-like states in the films
complicates obtaining of the uniform lying column align-
ments desired for the investigation of semiconducting prop-
erties. The achievement of this goal might however only be a
matter of process optimization since the films already con-
tain variable amounts of aligned Colobl domains and since
the organization in the bulk was shown to freeze above room
temperature. Further trials and the resort to more sophisti-
cated techniques combining in-plane alignment and surface-
induced alignment are indeed justified by the differentiation
of the P and T columns in the mesophase and the possible
ambipolar properties of the final material.
Acknowledgements
This work has been supported by the French National Agency
(ANR) in the frame of its program in Nanosciences and Nano-
technologies (TRAMBIPOLY project ANR-08-NANO-051). Y.
Xiao and X. Su thank the China Scholarship Council, Ministry
of Education, People's Republic of China. We thank the
Pohang Accelerator Laboratory (PAL) for giving us the oppor-
tunity to perform the GIXS measurements, MEST and
POSTECH for supporting these experiments, Drs. Tae Joo
Shin and Hyungju Ahn for adjustments and help, and other
colleagues from the 9A USAXS beamline for assistance. This
research was supported by the Leading Foreign Research In-
stitute Recruitment Program through the National Research
Foundation of Korea (NRF) funded by the Ministry of Science,
ICT & Future Planning (NRF-2010-00453). The authors also ac-
knowledge the International Research Network (GDRI, CNRS)
on “Functional Materials for Organic Optics, Electronics and
Devices” (FUNMOOD).
Conclusion
References
A new series of liquid crystalline D–A dyads and D–A–D triads
based on triphenylene and rylene diimide cores as donor (D)
and acceptor units (A), respectively, both entities being linked
by alkyl chain spacers, have been successfully designed, syn-
thesized and characterized. In solution, the HOMO/LUMO
values of the individual D/A cores turn out to be preserved in
the multiads. In the solid state, the linker authorizes the
stacking in individual columns, isolated from each other by
molten chains. One of these molecules effectively shows self-
organization in a genuine columnar mesophase with the opti-
mally intermingled arrangement of differentiated D and A
columns. In contrast, the three other molecules give rise to
mixed D/A columns or even to amorphous states, and geo-
metrical assembling rules for the appropriate molecular de-
sign could subsequently be derived. The coveted structure of
distinct columns appears indeed in conjunction with the 2 : 1
stoichiometry of the triad, compatible with the honeycomb
arrangements, and with the size-compatibility of mesogens,
realized for the perylene diimide/triphenylene couple. The
unfavorable 1 : 1 stoichiometry of the dyad architecture or the
substitution of perylene diimide by the smaller naphthalene
diimide leads to a columnar phase with mixed columns or
even to the vanishing of the mesomorphism when both unfa-
vorable features were combined. This comprehensive study
1 (a) S. R. Forrest, Nature, 2004, 428, 911–918; (b) C. Wang, H.
Dong, W. Hu, Y. Liu and D. Zhu, Chem. Rev., 2012, 112,
2208–2267.
2 (a) L. Dong, W. Li and W.-S. Li, Nanoscale, 2011, 3,
3447–3461; (b) M. Wang and F. Wudl, J. Mater. Chem.,
2012, 22, 2429–24314; (c) L. F. Dössel, V. Kamm, I. A. Howard,
F. Laquai, W. Pisula, X. Feng, C. Li, M. Takase, T. Kudernac,
S. de Feyter and K. Müllen, J. Am. Chem. Soc., 2012, 134,
5876–5886; (d) J. M. Mativetsky, M. Kastler, R. C. Savage, D.
Gentilini, M. Palma, W. Pisula, K. Müllen and P. Samori, Adv.
Funct. Mater., 2009, 19, 2486–2494; (e) R. Charvet, Y.
Yamamoto, T. Sasaki, J. Kim, K. Kato, M. Takata, A. Saeki, S.
Seki and T. Aida, J. Am. Chem. Soc., 2012, 134, 2524–2527; ( f )
D. Zeng, I. Tahar-Djebbar, Y. Xiao, F. Kameche, N. Kayunkid,
M. Brinkmann, D. Guillon, B. Heinrich, B. Donnio, D. A.
Ivanov, E. Lacaze, D. Kreher, F. Mathevet and A.-J. Attias,
Macromolecules, 2014, 47, 1715–1731; (g) R. A. Segalman, B.
McCulloch, S. Kirmayer and J. J. Urban, Macromolecules,
2009, 42, 9205–9216; (h) S. Huettner, M. Sommer, J.
Hodgkiss, P. Kohn, T. Thurn-Albrecht, R. H. Friend, U.
Steiner and M. Thelakkat, ACS Nano, 2011, 5, 3506–3515.
3 (a) M. Sommer, A. S. Lang and M. Thelakkat, Angew. Chem.,
Int. Ed., 2008, 47, 7901–7904; (b) S.-Y. Ku, M. A. Brady, N. D.
This journal is © The Royal Society of Chemistry 2016
CrystEngComm

View Article Online
Paper
Treat, J. E. Cochran, M. J. Robb, E. J. Kramer, M. L. Chabinyc
CrystEngComm
2009, 159, 875–879; (e) N. Mizoshita, H. Monobe, M. Inoue,
M. Ukon, T. Watanabe, Y. Shimizu, K. Hanabusa and T. Kato,
Chem. Commun., 2002, 428–429; ( f ) P. Staffeld, M. Kaller, S. J.
Beardsworth, K. Tremel, S. Ludwigs, S. Laschat and F.
Giesselmann, J. Mater. Chem. C, 2013, 1, 892–901.
and C. J. Hawker, J. Am. Chem. Soc., 2012, 134, 16040–16046.
4 (a) H. Hayashi, W. Nihashi, T. Umeyama, Y. Matano, S. Seki,
Y. Shimizu and H. Imahori, J. Am. Chem. Soc., 2011, 133,
10736–10739; (b) L. Bu, X. Guo, B. Yu, Y. Qu, Z. Xie, D. Yan,
Y. Geng and F. Wang, J. Am. Chem. Soc., 2009, 131,
13242–13243; (c) X. Kong, Z. He, Y. Zhang, L. Mu, C. Liang, B.
Chen, X. Jing and A. N. Cammidge, Org. Lett., 2011, 13,
764–767; (d) P.-O. Schwartz, L. Biniek, E. Zaborova, B.
Heinrich, M. Brinkmann, N. Leclerc and S. Mery, J. Am.
Chem. Soc., 2014, 136, 5981–5992; (e) S. K. Gupta, S. Setia, S.
Sidiq, M. Gupta, S. Kumar and S. K. Pal, RSC Adv., 2013, 3,
12060–12065; ( f ) E. Beltran, M. Garzoni, B. Feringan, A.
Vancheri, J. Barbera, J. L. Serrano, G. M. Pavan, R. Gimenez
and T. Sierra, Chem. Commun., 2015, 51, 1811–1814; (g) K. J.
Lee, J. H. Woo, E. Kim, Y. Xiao, X. Su, L. M. Mazur, A.-J.
Attias, F. Fages, O. Cregut, A. Barsella, F. Mathevet, L. Mager,
J. W. Wu, A. D'Aléo and J.-C. Ribierre, Phys. Chem. Chem.
Phys., 2016, 18, 7875–7887.
9 (a) X. Zhan, A. Facchetti, S. Barlow, T. J. Marks, M. A. Ratner,
M. R. Wasielewski and S. R. Marder, Adv. Mater., 2011, 23,
268–284; (b) S. Alibert-Fouet, I. Seguy, J.-F. Bobo, P. Destruel
and H. Bock, Chem. – Eur. J., 2007, 13, 1746–1753; (c) C. Li
and H. Wonneberger, Adv. Mater., 2012, 24, 613–636; (d) Z.
Anl, J. Yu, S. C. Jones, S. Barlow, S. Yoo, B. Domercq, P. Prins,
L. D. A. Siebbeles, B. Kippelen and S. R. Marder, Adv. Mater.,
2005, 17, 2580–2583; (e) M. Funahashi and A. Sonoda,
J. Mater. Chem., 2012, 22, 25190–25197; ( f ) F. Würthner, C. R.
Saha-Möller, B. Fimmel, S. Ogi, P. Leowanawat and D.
Schmidt, Chem. Rev., 2016, 116, 962–1052.
10 K.-Q. Zhao, L.-L. An, X.-B. Zhang, W.-H. Yu, P. Hu, B.-Q.
Wang, J. Xu, Q.-D. Zeng, H. Monobe, Y. Shimizu, B. Heinrich
and B. Donnio, Chem. – Eur. J., 2015, 21, 10379–10390.
11 W. Kreuder, H. Ringsdorf, O. Herrmann-Schönherr and J. H.
Wendorff, Angew. Chem., Int. Ed. Engl., 1987, 26, 1249–1252.
12 X. Kong, Z. He, H. Gopee, X. Jing and A. N. Cammidge,
Tetrahedron Lett., 2011, 52, 77–79.
5 Handbook of Liquid Crystals, ed. J. W. Goodby, P. J. Collings,
T. Kato, C. Tschierske, H. F. Gleeson and P. Raynes, Wiley-
VCH, Weinheim, Germany, 2nd edn, 2014.
6 (a) M. O'Neill and S. M. Kelly, Adv. Mater., 2011, 23, 566–584;
(b) F. Lincker, A.-J. Attias, F. Mathevet, B. Heinrich, B.
Donnio, J.-L. Fave, P. Rannou and R. Demadrille, Chem.
Commun., 2012, 48, 3209–3211; (c) L. Mazur, A. Castiglione, K.
Ocytko, F. Kameche, R. Macabies, A. Ainsebaa, D. Kreher, B.
Heinrich, B. Donnio, S. Sanaur, E. Lacaze, J.-L. Fave, K.
Matczyszyn, M. Samoc, J. W. Wu, A.-J. Attias, J.-C. Ribierre
and F. Mathevet, Org. Electron., 2014, 15, 943–953; (d) W.
Pisula, M. Zorn, J. C. Young, K. Müllen and R. Zentel,
Macromol. Rapid Commun., 2009, 30, 1179–1202; (e) Y.
Shimizu, K. Oikawa, K. Nakayama and D. Guillon, J. Mater.
Chem., 2007, 17, 4223–4229; ( f ) M. Funahashi, J. Mater.
Chem. C, 2014, 2, 7451–7459; (g) Y. Geng, S. Chang, K. Zhao,
Q. Zeng and C. Wang, J. Phys. Chem. C, 2015, 119,
18216–18220; (h) H. Iino, T. Usui and J. Hanna, Nat.
Commun., 2015, 6, 6828.
7 (a) S. Chandrasekhar, B. K. Sadashiva and K. A. Suresh,
Pramana, 1977, 9, 471–480; (b) S. Kumar, Chem. Soc. Rev.,
2006, 35, 83–109; (c) S. Sergeyev, W. Pisula and Y. H. Geerts,
Chem. Soc. Rev., 2007, 36(12), 1902–1929; (d) B. R. Kaafarani,
Chem. Mater., 2011, 23, 378–396; (e) T. Wöhrle, I. Wurzbach,
J. Kirres, A. Kostidou, N. Kapernaum, J. Litterscheidt, J. C.
Haenle, P. Staffeld, A. Baro, F. Giesselmann and S. Laschat,
Chem. Rev., 2016, 116, 1139–1241.
13 A. Wicklein, A. Lang, M. Muth and M. Thelakkat, J. Am.
Chem. Soc., 2009, 131, 14442–14453.
14 F. Würthner, C. Thalacker, S. Diele and C. Tschierske, Chem.
– Eur. J., 2001, 7, 2245–2253.
15 N. Boden, R. C. Borner, J. Bushby, A. N. Cammidge and M. V.
Jesudason, Liq. Cryst., 1993, 15(6), 851–858.
16 H. Langhals and S. Kinzel, J. Org. Chem., 2010, 75, 7781–7784.
17 (a) J. L. Bredas, R. Silbey, D. S. Boudreaux and R. R. Chance,
J. Am. Chem. Soc., 1983, 105, 6555–6559; (b) M. Thelakkat and
H. Schmidt, Adv. Mater., 1998, 10, 219–223; (c) P. Bauer, H.
Wietasch, S. M. Lindner and M. Thelakkat, Chem. Mater.,
2007, 19, 88–94.
18 I. Seguy, P. Jolinat, P. Destruel, J. Farenc, R. Mamy, H. Bock,
J. Ip and T. P. Nguyen, J. Appl. Phys., 2001, 89, 5442–5448.
19 L. E. Polander, S. P. Tiwari, L. Pandey, B. M. Seifried, Q.
Zhang, S. Barlow, C. Risko, J.-L. Bredas, B. Kippelen and S. R.
Marder, Chem. Mater., 2011, 23, 3408–3410.
20 S. Zamir, R. Poupko, Z. Luz, J. B. Hiker, C. Boeffel and H.
Zimmermann, J. Am. Chem. Soc., 1994, 116, 1973–1980.
21 M. Marcos, R. Giménez, J. L. Serrano, B. Donnio, B. Heinrich
and D. Guillon, Chem. – Eur. J., 2001, 7, 1006–1013.
22 C. W. Struijk, A. B. Sieval, J. E. J. Dakhorst, M. van Dijk, P.
Kimkes, R. B. M. Koehorst, H. Donker, T. J. Schaafsma, S. J.
Picken, A. M. van de Craats, J. M. Warman, H. Zuilhof and
E. J. R. Sudholter, J. Am. Chem. Soc., 2000, 122, 11057–11066.
23 S. Kumar, Liq. Cryst., 2004, 31, 1037–1059.
8 (a) D. Adam, P. Schumacher, J. Simmerer, L. Hausling, K.
Siemensmeyer, K. H. Etzbach, H. Ringsdorf and D. Haarer,
Nature, 1994, 371, 141; (b) H. Iino, Y. Takayashiki, J. Hanna,
R. J. Bushby and D. Haarer, Appl. Phys. Lett., 2005, 87,
192105; (c) H. Iino, J. Hanna, R. J. Bushby, B. Movaghar and
B. J. Whitaker, Appl. Phys. Lett., 2006, 100, 043716; (d) Yasuo
Miyakea, B. A. Fujiib, M. Ozakib and Y. Shimizu, Synth. Met.,
24 I. Bury, B. Heinrich, C. Bourgogne, G. H. Mehl, D. Guillon
and B. Donnio, New J. Chem., 2012, 36, 452–468.
25 A. Tracz, J. K. Jeszka, M. D. Watson, W. Pisula, K. Müllen and
T. Pakula, J. Am. Chem. Soc., 2003, 125, 1682–1683.
CrystEngComm
This journal is © The Royal Society of Chemistry 2016