Controlling spontaneous mirror symmetry breaking in cubic liquid crystalline phases by the cycloaliphatic ring size

Article abstract of DOI:10.1039/c9cc09206d

Rod-like molecules combining a fork-like triple chain end and a cycloaliphatic apex are introduced as a new design concept for materials with broad ranges of bicontinuous cubic (Cub_bi) phases. By ring expansion from n = 4 to 12 a sequence of three Cub_bi phases is observed; the achiral double gyroid la3d phase, a chiral Im3m phase and an achiral re-entrant la3d phase. The chiral Im3m phase is formed if the helical twist between the molecules along the networks is in the range of 8.6°-9.5°, either for the individual compounds or their mixtures.

Full text of DOI:10.1039/c9cc09206d

ChemComm
View Article Online
View Journal
COMMUNICATION
Controlling spontaneous mirror symmetry
breaking in cubic liquid crystalline phases by the
Cite this:DOI: 10.1039/c9cc09206d
cycloaliphatic ring size†
Received 26th November 2019,
Accepted 9th December 2019
Tino Reppe, Christian Dressel, Silvio Poppe and Carsten Tschierske
*
DOI: 10.1039/c9cc09206d
rsc.li/chemcomm
Rod-like molecules combining a fork-like triple chain end and a and on average perpendicular to the networks. The most preferred
cycloaliphatic apex are introduced as a new design concept for Cub phase is the double gyroid phase (space group la3%d) involving
bi
8,9
materials with broad ranges of bicontinuous cubic (Cubbi) phases. two networks with three-way junctions (see Fig. 1a). There is a
By ring expansion from n = 4 to 12 a sequence of three Cubbi phases second common, but more complex type of Cubbi phases with a
14
%
is observed; the achiral double gyroid la3d phase, a chiral ‘‘Im3m’’ different space group and a significantly larger lattice parameter,
%
phase and an achiral re-entrant la3d phase. The chiral ‘‘Im3m’’ whose precise structure is still under discussion. Its X-ray diffrac-
phase is formed if the helical twist between the molecules along tion pattern was indexed to the space group lm3%m and a triple
1
5
the networks is in the range of 8.68–9.58, either for the individual network structure (Fig. 1b) or a double network involving
1
6
compounds or their mixtures.
additional spherical aggregates were suggested.
This Cubbi phase was recently found to be inherently chiral
4
,5
Mirror symmetry breaking and chirality synchronisation of and the discovery of this supramolecular chirality provided an
achiral soft and liquid crystalline (LC) materials provides a improved general understanding of the cubic phases formed by
1
2,4a
new non-ex-chiral-pool route to chiral functional materials. It rod-like compounds.
Accordingly, the clashing of the bulky
has been observed in nematic, smectic and columnar LC molecular peripheries leads to a helical twist (Fig. 1c) which is
2
,3
phases, but recently it was also found in bicontinuous cubic transmitted to the macroscopic scale by the network structure of
4
,5
[ ] 6
(
Cubbi) LCs and even in isotropic liquids (Iso
1
* ). The Cubbi the Cubbi phases. The la3%d phase involves two enantiomorphic
phases have attracted much attention by virtue of their ubiquitous networks with opposite handedness (red, blue) and therefore it
presence in nature and their significant interest for conducting and is achiral (Fig. 1a). However, in the lm3%m phase this degeneracy
photonic materials, as matrices for protein crystallization and is broken either by a third network (Fig. 1b), that chirality cannot
7,8
templates for sol–gel syntheses. Cubbi phases are formed by
different types of amphiphilic compounds either in aqueous
8
systems (lyotropic phases) or as pure compounds (thermotropic
9
phases). Amphiphilicity can be due to a polar–apolar molecular
structure or a combination of rod-like polyaromatic cores
with flexible chains, both leading to nano-segregation of the
1
0
incompatible units. Among the rod-like molecules forming
0
0
Cubbi phases the 4 -n-alkyloxy-3 -nitrobiphenyl-4-carboxylic
1
1
0
acids (ANBCs), the 1,2-bis(4 -n-alkanoyloxybenzoyl hydrazines)
BABHs) and the so-called polycatenar compounds, having
more than only two terminal chains at their ends are the most
common. In the Cubbi phases the polyaromatic cores form the
networks and the end chains fill the continuum around them. In
these networks the rods are aligned (anti)parallel to each other
1
2
(
1
3
Fig. 1 Schematics showing (a) the achiral la 3% d (double gyroid) and (b) the
[
]
chiral Im3m * triple network structure of the Cubbi phases formed by
polycatenar mesogens and (c) clashing of the molecular peripheries leads
to helical twist; (d) shows the continuous transmission of helicity across
the junctions of the gyroid; (a) and (b) were reproduced with permission
from ref. 4a, copyright, 2014, Wiley-VCH.
Institute of Chemistry, Martin Luther University Halle-Wittenberg, Kurt-Mothes-Straße 2,
06120 Halle, Germany. E-mail: carsten.tschierske@chemie.uni-halle.de
†
Electronic supplementary information (ESI) available: Methods, syntheses,
analytical data, additional XRD and structural data. See DOI: 10.1039/c9cc09206d
This journal is © The Royal Society of Chemistry 2019
Chem. Commun.

View Article Online
Communication
ChemComm
Fig. 2 Investigation of compound 2/c8. (a) DSC heating and cooling traces
ꢀ1
[ ]
(
10 K min ); (b and c) optically active domains in the Cubbi/Im3m * phase at
40 1C as observed between slightly uncrossed polarizers after rotation in
b) clockwise and (c) anticlockwise direction (contrast enhanced).
1
(
Scheme 1 Chemical structures of the compounds 1–6 under investigation.
The lattice type is additionally confirmed by the distinct positions
be extinguished, or by the synchronization of the helix sense of the small angle X-ray diffraction peaks in these two phase types,
17
between the networks (as shown for the SmQ phase ). The highest with the most intense reflections being either O6 (211) and O8
[
]
symmetry chiral space group derived from the achiral lm3%m space (220) for la3%d or O14 (321), O16 (400) and O20 (422) for Im3m *
4
4a
group would be I432. However, it is not sure if indeed the highest (see Fig. S2a, S3a and Tables S2, S3, ESI†). In the la3%d phases the
symmetry lattice is chosen and, in addition, it is questionable if the lattice parameter acub has a ratio of O3acub/4 to the molecular
proposed structures, either involving some 5-way junctions (see red length, indicating that the distance between the middle of the
15
16
net in Fig. 1b) or spherical aggregates, could be stable for these networks corresponds to approximately one (0.86–1.09) molecular
rod-like molecules, and probably an alternative model has to be length (Table S5, ESI†). The deviation from being exactly one
[
]
developed. Though, ‘‘Im3m * ’’ is still used to indicate this type of could be due to a chang ´ı ng degree of intercalation of rods and
1
2b
spontaneously chiral Cubbi phase, it is in this case not considered apexes (Fig. S7, ESI†). The LC character of the cubic phases is
as a crystallographic space group. obvious from the diffuse wide angle scattering with maxima at
The availability of materials with Cubbi phases is limited, 0.45–0.46 nm (see Fig. S2b and S3b, ESI†) confirming the
and it is usually observed for only few compounds with specific absence of any fixed positions for the individual molecules.
8
,9
alkyl chain lengths in narrow temperature ranges. There are
Compound 1/ph with a benzene ring at the apex forms an
only few systems which are known to be less sensitive and show enantiotropic la3%d phase between 114 and 162 1C.4 Replace-
a
Cubbi phases over wider ranges, namely the ANBCs and ment of the benzene ring by a cyclohexane ring (2/c6) increases
9
,11,12
BABHs.
Cubbi phases over wide temperature ranges is still a challenging la3%d to Im3m * , while there is no significant change of the
demand. Cubbi–Iso transition temperature (Table 1). The additional axial
Therefore, the design of new materials forming the cubic phase range and changes the Cubbi phase type from
[
]
Herein such a new class of Cubbi phase forming compounds H-atoms make cylohexane an effectively bulkier group com-
is provided and the focus is on the elucidation of the molecular pared to the relatively flat benzene ring of 1/ph.
structural parameters which favour Cubbi phase formation and
[
]
determine the specific type of cubic phase, either la3%d or Im3m*. To
this end compounds 2/cn (Scheme 1) having a cycloaliphatic unit Table 1 Phase transitions, lattice parameters (acub) and twist angles
0
0
between the molecules in neighbouring rafts (F/1) in the Cub phases of
compounds 1–6
attached to the apex of a three-forked 2,5 -diphenyl-2,2 -dithiophene
core were synthesized and compared with related non-cyclic (3–5)
and a polycyclic adamantane substituted compound (6/ada).
The compounds of series 2–6 were synthesized from the
bi
a
a
ꢀ1
Compd
T /1C [DH kJ mol
]
a
cub/nm (T/1C)
F/1
8.2
8.6
8.5
—
—
9.5
—
—
—
4a
1
2
/ph
/c4
Cr 114 [39.7] la3%d 162 [2.3] Iso
11.0 (116)
10.4 (130)
10.5 (135)
15.6 (125)
15.4 (100)
9.4 (108)
14.8 (90)
—
4a,6a
previously reported phenol 1/OH
by acylation using standard
procedures, see ESI.† Though the isotropic liquid and the cubic 2/c5
phases both appear completely dark between crossed polarizers,
the transition to the cubic phases can be detected by a strong
increase of the viscosity at the Iso - Cubbi transition which is 3/br5
associated with a DSC peak. There is a hysteresis of this transition,
being 6 to 14 K lower on cooling than on heating (see Table S1,
ESI†), which is another typical feature of cubic LC phases. The 6/ada
%
Cr 92 [17.7] la3d 148 [3.3] Iso
%d 150 [3.0] Iso
Cr 88 [16.3] la
3
[
]
2
2
2
/c6
/c8
/c12
Cr 93 [33.9] Im3m * 160 [4.3] Iso
Cr 59 [33.1] Im3m * 148 [4.5] Iso
Cr 103 [36.6] la3d 113 [2.9] Iso
Cr 75 [29.8] Im3m * 109 [3.4] Iso
[ ]
%
[
]
3
4
5
/br8
/lin8
Cr 57 [43.7] Iso [
]
Cr 90 [5.7] Im3m * 122 [1.9] Iso
16.0 (110)
8.85 (70)
15.1 (120)
/Et
3
Cr 49 [8.9] la3d 87 [2.7] Iso
Cr 84 [11.2] Im3m * 156 [3.6] Iso
%
10.1
—
[ ]
type of cubic phase is identified by investigation between
a
ꢀ1
DSC peak temperatures on heating at 10 K min ; for transitions on
[
]
slightly uncrossed polarizers where only the Im3m * phase cooling, see Table S1, ESI; for molecular structures, see Scheme 1;
abbreviations: Cr = crystalline solid, la3%d = bicontinuous cubic phase
shows a conglomerate of optically active dark and bright chiral
domains (Fig. 2b, c and Fig. S5, ESI†), whereas the la3%d phase
[
]
with double gyroid structure; Im3m * = mirror symmetry broken
bicontinuous cubic phase; Iso = isotropic liquid; F(la3%d) = 70.51/
4
a
is achiral and does not show such domains (Fig. S4, ESI†). [0.354acub/0.45]; for DSCs, see Fig. S1, ESI.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2019

View Article Online
ChemComm
Communication
In a next step the effect of the ring size was investigated in the aliphatic part (n = 8). The branched compound 3/br8 does
systematically. All compounds 2/c4–2/c12 with cycloaliphatic end not form any LC phase, it has a melting point at 57 1C and on
groups form cubic phases which do not crystallize on cooling (see cooling it crystallizes at 4 1C (Fig. S1g, ESI†). The linear compound
[
]
DSC in Fig. 2a and Fig. S1, ESI†). Only after storage at 20 1C for 4/n8 forms an Im3m * cubic phase, too, just like 2/c8, but the
several days partial crystallization takes place. The Cub –Iso transi- Cub –Iso transition is reduced by 26 K. This shows that cyclization
bi
bi
tion temperatures first increase with growing ring size from 148 1C is a powerful tool to stabilize cubic LC phases.
for 2/c4 to 160 1C for 2/c6 and then decrease again to 113 1C for 2/c12 The distorting effect of chain branching becomes smaller for
Table 1). In addition, there is a significant effect of the ring size on compounds 3/br5 with two shorter ethyl branches, but never-
the cubic phase type. The cyclobutane and cyclopentane carboxylates theless, there is still a large stabilizing effect of cyclization,
(
[
]
2
/c4 and 2/c5 show the la3%d phase, whereas the Im3m* phase is enhancing the cubic phase stability by 40 K and changing the
[
]
observed for the compounds 2/c6–2/c8 with larger ring sizes, and cubic phase type from Im3m * for 3/br5 to la3%d for 2/c5,
surprisingly, the largest one, the cyclododecane carboxylate 2/c12 reflecting the smaller effective size of the cyclic apex.
[
]
forms a la3%d phase again. Thus, the sequence la3%d–Im3m*–la3%d
with an re-entrance of the la3%d phase is observed by increasing the group to the apex of 3/br5 retains the Cubbi phase (5/Et
Increasing the volume of the apex by adding one more ethyl
), but
the phase type is changed from Im3m * to la3%d. Miscibility
of compound 2/c4 with a smaller cyclobutane is 1 nm longer than experiments with the la3%d phase of 2/c4 or 2/c5 indicate the
3
[
]
ring size.‡ Remarkably, the lattice parameter of the la3%d phase (acub
)
[
]
that of compound 2/c12 with the larger cyclododecane ring induction of a chiral ribbon of the Im3m * phase between them
Table 1), meaning that the number of molecules per unit cell (see Fig. S6, ESI†), thus confirming that the two la3%d phases are
(
[
]
(
ncell) decreases by ring expansion. In the Im3m * phase, occurring of a different type, in line with the different size of their apexes.
between the two la3%d phases, the lattice parameter is about 50%
The adamantane-1-carboxylate 6/ada is related to 5/Et by
larger and the number of molecules per unit cell is about three to having a triply branched aliphatic units at the apex, but in this case
3
[
]
five times the number in the la3%d phases (see Table S4, ESI†).
incorporated in a polycyclic entity. 6/ada forms a Im3m * phase,
The chiral Im3m * phase is also found in the contact region though the total number of aliphatic C-atoms at the apex is
between the achiral la3%d phases of the two compounds 2/c4 and increased from 7 in 5/Et
to 10 in 6/ada (Table 1). This observation
/c12. In the contact area between them a concentration is in line with the smaller effective size of the adamantyl group due
[
]
3
2
gradient develops and a ribbon composed of dark and bright chiral to the reduced conformational freedom of the polycyclic hydro-
domains can be observed by optical microscopy between not exactly carbon compared to the flexible diethylbutanoate unit of 5/Et3,
901 crossed polarizers (Fig. 3). Rotating the analyzer into the opposite capable of assuming several bulky conformations.
direction with respect to the polarizer reverses the brightness of the Thus, the effect of the apex substituent could be traced back
domains, whereas rotation of the sample between crossed polarizers to a series of competing volume effects. The first determines the
does not change the domains. The maximum of the Cub–Iso degree of interfacial curvature, i.e. if a lamellar, Cubbi or a columnar
[
]
10
transition of the induced Im3m* phase on heating is at 143 1C. phase is formed. It is a remarkable feature of the investigated
This means that mixing the la3%d phases of two compounds with compounds that exclusively Cubbi phases were observed. This
different ring size causes the same effect as adjusting the ring size by volume effect is only expressed by the change of the Cubbi phase
synthesis of discrete single compounds. This is of practical interest type and the Cub –Iso transition temperature (Table 1).
bi
as it allows the preparation of eutectic multi-component mixtures
As shown in Table 1, the lattice parameter acub decreases with
[
]
forming the chiral Im3m* phase over broad temperature ranges growing effective apex size. The number of molecules calculated
down to ambient temperature and below. It also shows that simple from acub for each column segment between the junctions in the two
mixing of two achiral compounds can lead to a spontaneously gyroid networks of the la3%d phase is about 29 for 2/c4 and only about
18
mirror symmetry broken system.‡
19 for 2/c12. Simultaneously, the distance between the junctions
In order to investigate the effect of the aliphatic ring structure decreases from 3.7 nm for 2/c4 to 3.3 nm for 2/c12 (see Table S4,
on self-assembly two compounds related to 2/c8 were synthesized, ESI†). Hence, the number of molecules organized side-by-side in
the linear n-nonanoate 4/n8 and the 2-propylhexanoate 3/br8 with a each stratum (raft) of the columns with an assumed height of
branched apex, both with the same total number of eight C-atoms 0.45 nm is 3.5 for 2/c4 and only about 2.6 for 2/c12 (see Table S4,
ESI†). This means that with growing apex size there is a decreasing
number of molecules in each of the rafts and also the number of
rafts along the column segments between the junctions decreases.
Most importantly, the apex volume affects the twist between
the rods organized along the networks (Fig. 1c). This twist is
caused by the clashing of the bulky molecular ends and is
transmitted by the network junctions throughout the whole
network. The la3%d phase has trigonal junctions with a fixed
[
]
Fig. 3 Chiral domains of the Im3m * phase induced between the achiral
la 3% d phases of compounds 2/c4 (top) and 2/c12 (bottom) at T = 129 1C
(
dihedral angle of 70.51 between them (Fig. 1a). A continuous
twist of the molecules between the junctions by this angle is
required for a smooth merging of the rods without defects at
the junctions (Fig. 1d). With growing apex volume the distance
[
]
contrast enhanced); on cooling the Im3m * phase appears at 130 1C and
on heating it disappears at 143 1C; for the contact region between 2/c4
and 5/Et , see Fig. S6, ESI.†
3
This journal is © The Royal Society of Chemistry 2019
Chem. Commun.

View Article Online
Communication
ChemComm
between the junctions decreases (see above) and consequently Notes and references
the twist angle F between adjacent rod-like cores along the
column segments increases (Table 1 and Fig. S4, ESI†). A twist
angle between the molecules in adjacent rafts of 8.2–8.61 is the Im3m * phase was also induced between the la
[ ]
‡
A phase sequence la3%d–Im3m * –la3%d with increasing terminal alkyl
11 12
chains length was observed for the ANBCs and BABHs. For the latter
[
]
3
%d phases of
18
different homologues.
found in the la3%d phases. If the twist angle exceeds this value
[
]
the gyroid is replaced by the Im3m * structure. However, the
la3%d lattice appears again upon further chain volume enlarge-
1 H. K. Bisoyi, T. J. Bunning and Q. Li, Adv. Mater., 2018, 30, 1706512;
B. Chang, X. Li and T. Sun, Eur. Polym. J., 2019, 118, 365; Y. Sang,
J. Han, T. Zhao, P. Duan and M. Liu, Adv. Mater., 2019, 1900110.
(a) C. Tschierske, Liq. Cryst., 2018, 45, 2221; (b) C. Tschierske and
G. Ungar, ChemPhysChem, 2016, 17, 9.
ment, as for example observed in the series of cyclic com-
pounds 2/cn. The calculated twist angle is 9.51 in the re-entrant
la3%d phase of 2/c12 and an even larger value of 10.11 is calculated
2
3
K. V. Le, H. Takezoe and F. Araoka, Adv. Mater., 2017, 1602737.
[
]
4 (a) C. Dressel, F. Liu, M. Prehm, X. B. Zeng, G. Ungar and C. Tschierske,
Angew. Chem., Int. Ed., 2014, 53, 13115; (b) M. Alaasar, S. Poppe, Q. Dong,
F. Liu and C. Tschierske, Chem. Commun., 2016, 52, 13869.
(a) T. Kajitani, S. Kohmoto, M. Yamamoto and K. Kishikawa, Chem.
Mater., 2005, 17, 3812; J. M. Wolska, J. Wilk, D. Pociecha,
J. Mieczkowski and E. Gorecka, Chem. – Eur. J., 2017, 23, 6853;
(b) H. R. Brand and H. Pleiner, Eur. Phys. J. E: Soft Matter Biol. Phys.,
for the la3%d phase of 5/Et (Table 1). Hence, the chiral Im3m *
3
phase becomes the thermodynamically more stable alternative
mode of self-assembly replacing the achiral double gyroid if the
intermolecular twist exceeds 8.61. However, this structure
appears to be even more sensitive to the twist angle change
and becomes the less stable structure as the twist in the
competing la3%d phase reaches 9.51. This is a generally observed
5
2019, 42, 142.
6
(a) C. Dressel, T. Reppe, M. Prehm, M. Brautzsch and C. Tschierske,
Nat. Chem., 2014, 6, 971; (b) C. Dressel, W. Weissflog and
C. Tschierske, Chem. Commun., 2015, 51, 15850; (c) M. Alaasar,
M. Prehm, Y. Cao, F. Liu and C. Tschierske, Angew. Chem., Int. Ed.,
016, 55, 312; M. Alaasar, S. Poppe, Q. Dong, F. Liu and
C. Tschierske, Angew. Chem., Int. Ed., 2017, 56, 10801.
7 T. Kato, J. Uchida, T. Ichikawa and T. Sakamoto, Angew. Chem., Int.
Ed., 2018, 57, 4355.
Bicontinuous Liquid Crystals, ed. M. L. Lynch and P. T. Spicer, CRC
Press, Taylor & Francis Group, Boca Raton, FL, 2005; L. van‘t Hag,
S. L. Gras, C. E. Conn and C. J. Drummond, Chem. Soc. Rev., 2017,
[
]
feature‡ and any new or modified model of the Im3m * phase
has to give a convincing explanation of this observation.
2
Overall, a new design concept of robust thermotropic Cubbi
phase forming materials, based on tricatenar compounds with
a cycloaliphatic apex is introduced, providing a tool for stabiliz-
ing Cubbi phases over broad temperature ranges. In this respect
these compounds are advantageous over related compounds
with linear or branched alkyl chains at the apex. It also allows
8
46, 2705.
[
]
9 (a) G. Ungar, F. Liu and X. B. Zeng, in Handbook of Liquid Crystals,
ed. J. W. Goodby, P. J. Collings, T. Kato, C. Tschierske, H. Gleeson
and P. Raynes, Wiley-VCH, 2nd edn, 2014, vol. 5; (b) S. Kutsumizu,
Isr. J. Chem., 2012, 52, 844; (c) M. Yoneya, Chem. Rec., 2011, 11, 66.
0 C. Tschierske, J. Mater. Chem., 2001, 11, 2647.
1 G. W. Gray, B. Jones and F. Marson, J. Chem. Soc., 1957, 393;
S. Kutsumizu, K. Morita, T. Ichikawa, S. Yano, S. Nojima and
T. Yamaguchi, Liq. Cryst., 2002, 29, 1447.
the engineering of the cubic phase type (la3%d vs. Im3m * ) and
hence the control of emergence of supermolecular chirality.
The remarkable re-entrance of the la3%d phase is not yet fully
1
1
understood. However, it appears that for a twist angle in the
[
]
range between 8.6 and 9.51 the chiral Im3m * phase becomes
the thermodynamically more stable alternative structure with
respect to the achiral double gyroid (la3%d), irrespectively if it is
1
2 (a) H. Schubert, J. Hauschild, D. Demus and S. Hoffmann, Z. Chem.,
1
978, 18, 256; (b) Y. Yamamura, Y. Nakazawa, S. Kutsumizu and
formed by a single compound or in mixtures of compounds
with different apex size. This observation is difficult to explain
with the presently existing models of the cubic ‘‘Im3m’’
K. Saito, Phys. Chem. Chem. Phys., 2019, 21, 32705.
3 H. T. Nguyen, C. Destrade and J. Malthete, Adv. Mater., 1997, 9, 375;
D. W. Bruce, Acc. Chem. Res., 2000, 33, 831.
4 A. M. Levelut and M. Clerc, Liq. Cryst., 1998, 24, 105.
15 X. B. Zeng, G. Ungar and M. Imperor-Clerc, Nat. Mater., 2005, 4, 562.
1
1
1
5,16
phase
and requires a further refinement of these models.
1
1
1
6 K. Saito, Y. Yamamura, Y. Miwa and S. Kutsumizu, Phys. Chem.
Chem. Phys., 2016, 18, 3280.
7 H. J. Lu, X. B. Zeng, G. Ungar, C. Dressel and C. Tschierske, Angew.
Chem., Int. Ed., 2018, 57, 2835.
8 S. Kutsumizu, S. Miisako, Y. Miwa, M. Kitagawa, Y. Yamamura and
K. Saito, Phys. Chem. Chem. Phys., 2016, 18, 17341.
Conflicts of interest
There are no conflicts to declare.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2019