Systematic exploitation of thermotropic bicontinuous cubic phase families from 1,2-bis(aryloyl)hydrazine-based molecules

Shoichi Kutsumizu *, Yutaro Yamada , Tadashi Sugimoto§ , Nina Yamada , Taro Udagawa and Yohei Miwa
Department of Chemistry and Biomolecular Science, Faculty of Engineering, Gifu University, Yanagido, Gifu 501-1193, Japan. E-mail:; Fax: +81 58 293 2794; Tel: +81 58 293 2573

Received 13th December 2017 , Accepted 29th January 2018

First published on 2nd February 2018

Rational design of molecules that exhibit a thermotropic bicontinuous cubic (Cub) phase has been earnestly desired. In this work, we describe the suitable selection of a molecular motif that has enabled the systematic exploitation of eight new series of Cub-phase molecules with symmetric molecular cores, N-n (1), PB-n (2), S-n (3), and PEB-n (4), and unsymmetric cores, B-N-n (5), B-PB-n (6), B-S-n (7), and B-PEB-n (8). These eight series all originate from achiral chain-core-chain type rod-like molecules that exhibit two types of Cub phases, an achiral Ia3d phase, and a chiral phase. All the Ia3d phases formed were found to be isomorphous structures, with their cell dimensions being proportional to the core size, and the same was true for the latter chiral phase. We demonstrated that the formation is mainly governed by the segregation between core and alkyl moieties of the molecules, and thus, by the weight fraction of the core portion fcore. This work also demonstrates that the central dicarbonylhydrazine linkage bearing intermolecular hydrogen bonding ability exhibits a pinning effect that prevents slippage of π-stacks of molecules, which is critical for the formation of the two Cub phases that are composed of chiral networks with twisted molecular arrangements. In each series, the emergence of spontaneous chirality formation that occurred in the chiral phase was limited to between 0.36 and 0.50 in the range of fcore. An interesting insight was that the introduced unsymmetry of the molecular core strongly influenced the phase behavior, which lowered the temperature range of Cub phases to around that of the smallest core series B-n, while the high temperature limit (Tc) was roughly proportional to the core size, as determined by the strength of intermolecular π–π interactions.


Self-assembled superstructures of bicontinuous cubic (Cub) liquid crystalline (LC) phases have long attracted much attention because of their unique three-dimensionally extended network structures and the numerous application concepts that they have inspired.1–11 Analogous structures are also evident in other soft and hard materials and also in living systems. Because of their ubiquitous presence in nature, several mechanisms have been proposed for the formation of the Cub phase, including the possibility of the simultaneous cooperation of multiple mechanisms.12 In mathematics, the concept of triply periodic minimal surfaces has been established as a complementary description of the network structures.13 Furthermore, a recently renewed understanding of the formation based on the networks with a twisted molecular arrangement14 has provided a new recognition that this formation is closely connected to the spontaneous chirality formation of achiral LC molecules, which is a more general subject in stereochemistry.15–19 In the Cub LC phases, two typical representatives are an achiral Ia3d phase and a chiral phase. The latter phase has been previously recognized as having an Im3m symmetry.20–25 In the two phases, the twisted arrangement of the molecules is continuously transferred through the network junctions across entire networks. In the Ia3d phase, the opposing chiralities of the two networks cancel each other, which alters the system to being achiral, but this cannot take place in the chiral Cub (Cub[*]) phase, whose most probable symmetry is I432. The factors that determine the exclusive formation of one of the two phases is a new and important issue. For this purpose, structural modification is demanded because of the molecular factors that affect the cancelation or emergence on the macroscopic scale of the local chirality of molecular aggregation.

Three well-known Cub phase-forming molecules are presented in Scheme 1. They are all chain-core-chain type molecules, possessing apparently a rod shape but are rendered into an anti-spindle-like shape at high temperatures.12,26 This is a structural motif that is required for the Cub-phase formation. The side-by-side close packing of such molecules would produce a twist along the lateral direction to reduce the steric hindrance of the thermally activated terminal tails, forming a three-dimensional network instead of a layered structure that is usually favorable in the rod-shaped molecules.12,26–29 This idea was not enough for the exploitation of Cub-phase molecules, because even a slight chemical modification of these molecules, especially with respect to the core structure, very often removes the phase formation. For example, a series a with M = Sr forms a Cub phase (with Ia3d symmetry),30 but the analogous Mg and Ca salts form no Cub phase.31 In series b, the 3′-nitro32–35 and 3′-cyano series1,36 (X = NO2 or CN) exhibit both types of Cub phases but other molecular series having X = F, Cl, Br, I do not.1 The third series c is denoted as B-n in this report, where n represents the number of carbon atoms in the alkoxy group. These molecules form two types of Cub phases in a broad range where n = 6–22 (and more),26,37,38 but replacement of the central dicarbonylhydrazine [–C(O)NHNHC(O)–] group with an aldazine linkage [–CH[double bond, length as m-dash]N–N[double bond, length as m-dash]CH–] canceled the phase formation.39 Therefore, information on the rational design of molecules that exhibit the Cub phases is still incomplete, although more than 100 examples have been reported to date. Poor systematic studies have also hindered a quantitative understanding of the phase behavior.

image file: c7cp08345a-s1.tif
Scheme 1 Three representative Cub-phase forming compounds.

Our challenge in this reported work is chemical modification into the molecular core moiety. This work revealed that the selection of a suitable molecular motif can avoid removal of the Cub phases. The motif is schematically shown in Scheme 2, which, beginning with the structure of B-n, maintains the central linkage, thereby, exerting its intermolecular hydrogen bonding. In contrast to the previous failure of molecular exploitation using core modification, we have successfully exploited eight thermotropic Cub-phase families. Two design strategies are presented, the first of which is an extension of the aromatic portions of the molecules, which maintains its central symmetry so that the two rings Ar and Ar′ are the same. These compounds are denoted N-n (1), PB-n (2), S-n (3), and PEB-n (4), where both Ar and Ar′ are naphthalene, phenylbenzene (biphenyl), stilbene, and phenyethynybenzene, respectively. The second strategy concerns an extension of only the Ar′ part while maintaining the benzene ring at the Ar part of the molecule. The second series is denoted as B-N-n (5), B-PB-n (6), B-S-n (7), and B-PEB-n (8). In this case, the molecular core is unsymmetric, and the effect of the change in symmetry on the phase behavior was of great interest. For comparison, we also synthesized a molecule containing a bis(phenylbenzylidene)hydrazine core PBID-n (9), and it was shown that even for the molecules having an extended π-system of Ar and Ar′, the central dicarbonylhydrazine group is essential for the formation of the Cub phases. The effect of structural modification on the phase behavior is discussed in detail.

image file: c7cp08345a-s2.tif
Scheme 2 Structural motif useful for the Cub-phase formation. Ar and Ar′ are aromatic ring parts.

Results and discussion

Molecular structures of the Cub phase-forming molecules synthesized along with the design strategies in Scheme 2

In this reported work, totally 31 molecules were synthesized and their molecular structures are listed in Scheme 3. The detailed synthetic procedures, analytical data, and experimental procedures are given in the ESI. The phase transition temperatures with associated enthalpy changes obtained using DSC for all molecules discussed herein are summarized in Table 1. Our exploitation is systematic, and the effect of the size of the molecular core (Lcore) in this series was evaluated using the density functional theory (DFT) M06/6-31G* (see the ESI for computational details) based on the assumption of Ci symmetry as tabulated in Table 2.
image file: c7cp08345a-s3.tif
Scheme 3 Molecular structures of the Cub phase-forming molecules 1–8 synthesized along with the design strategies in Scheme 2 and the molecule 9 for comparison.
Table 1 Mesophase types, phase transition temperatures (T/K), transition enthalpies (ΔH/kJ mol−1 in parenthesis), and cubic cell dimensions (a/m) of compounds examined in this worka
Compd Phase sequence a Ia3d/nm a I432/nm
a Peak temperatures as determined from first heating and first cooling DSC scans with 5 K min−1 unless otherwise indicated. Abbreviations: Cr = crystalline solid; M = un-identified mesophase; Sm = higher order smectic phase; SmC = smectic C phase; Ia3d = bicontinuous cubic phase with Ia3d symmetry; I432 = chiral cubic phase probably having I432 symmetry; Colh = hexagonal columnar phase; Iso = isotropic liquid.
N series
N-12 Heating: Cr 467 (47.9) SmC 496 (9.5) Iso
Cooling: Cr 452 (−45.9) SmC 495 (−8.2) Iso
N-16 Heating: Cr 456 (52.8) Ia3d 478 (5.6) Iso 9.47
Cooling: Cr 439 (−40.3) Ia3d 476 (−5.0) Iso
N-18 Heating: Cr 415 (87.1) Sm 451 (51.0) I432 475 (6.5) Iso 14.80
Cooling: Cr 384 (−88.6) M 413 (−5.5) Sm 437 (−39.7) I432 472 (−4.9) Iso
N-20 Heating: Cr 418 (110.1) Sm 444 (44.6) Ia3d 475 (5.9) Iso 9.80
Cooling: Cr 385 (−105.5) Sm 429 (−38.9) Ia3d 471 (−5.3) Iso
N-22 Heating: Cr 418 (115.8) Sm 435 (38.3) Ia3d 470 (6.7) Iso 10.21
Cooling: Cr 387 (−102.9) Sm 428 (−34.6) M 441 (−1.2) Ia3d 464 (−5.1) Iso
PB series
PB-10 Heating: Cr 494 (43.2) SmC
Cooling: Cr 490 (−42.5) SmC Turned back from 523 K
2nd heating: Cr 494 (42.8) SmC 598 (8.2) Iso/decomp
PB-12 Heating: Cr 490 (46.5) SmC Ia3d 588 (7.5) Iso/decomp 9.51
Cooling: Cr 484 (−30.2) SmC Turned back from 523 K
PB-14 Heating: Cr 487 (44.1) SmC 548 (0.6) Ia3d 581 (6.2) Iso/decomp 10.08
Cooling: Cr 482 (−40.2) SmC Turned back from 523 K
PB-16 Heating: Cr 484 (45.6) Ia3d 574 (5.3) Iso/decomp 10.75
Cooling: Cr 474 (−42.9) Ia3d Turned back from 523 K
PB-18 Heating: Cr 480 (44.8) Ia3d 503 (0.6) I432 570 (6.4) Iso/decomp 10.92 16.21
Cooling: Cr 470 (−42.7) Ia3d Turned back from 523 K
PB-19 Heating: Cr 474 (42.2) I432 570 (6.7) Iso/decomp 17.03
Cooling: Cr 464 (−37.9) I432 Turned back from 523 K
PB-20 Heating: Cr 476 (45.2) I432 538 (0.3) Ia3d 568 (5.8) Iso/decomp 11.24 17.25
Cooling: Cr 465 (−41.9) I432 Turned back from 523 K
PB-21 Heating: Cr 472 (44.8) I432 525 (0.4) Ia3d 563 (1.3) Colh 566 (5.0) Iso/decomp 11.35 17.73
Cooling: Cr 460 (−43.4) I432 Turned back from 523 K
PB-22 Heating: Cr 472 (45.9) I432 Ia3d 558 (0.7) Colh 565 (3.2) Iso/decomp 11.41 18.07
Cooling: Cr 457 (−39.0) I432 /Ia3d Turned back from 523 K
S series
S-16 Heating: Cr 505 (49.1) SmC
Cooling: Cr 503 (−43.8) SmC Turned back from 523 K
3rd heating: Cr 505 (48.7) SmC 563 (1.4) Ia3d 602 (4.8) Iso/decomp 11.30
S-18 Heating: Cr 503 (47.5) SmC
Cooling: Cr 501 (−44.9) SmC Turned back from 523 K
3rd heating: Cr 503 (43.8) SmC 542 (0.6) Ia3d 591 (0.4) I432 600 (4.9) Iso/decomp 11.90 17.85
S-20 Heating: Cr 500 (48.9) I432
Cooling: Cr 488 (−47.9) I432 Turned back from 523 K
3rd heating: Cr 500 (46.5) I432 596 (5.0) Iso/decomp 19.12
S-22 Heating: Cr 495 (49.4) I432
Cooling: Cr 482 (−47.4) I432 Turned back from 523 K
3rd heating: Cr 494 (49.0) I432 568 (0.3) Ia3d 587 (0.9) Colh 594 (4.1) Iso/decomp 12.04 18.99
PEB series
PEB-18 Heating: Cr 485 (46.0) SmC 536 (1.7) I432 564 (4.5) Iso/decomp 17.72
Cooling: Cr 483 (−37.9) SmC Turned back from 523 K
PEB-22 Heating: Cr 479 (47.7) I432 543 (0.6) Colh 557 (5.7) Iso/decomp 19.55
Cooling: Cr 474 (−42.9) I432 Turned back from 543 K
B-N series
B-N-12 Heating: Cr 394 (50.9) Sm 414 (24.0) SmC 466 (10.1) Iso
Cooling: Cr 398 (−28.0) SmC 465 (−9.0) Iso
B-N-22 Heating: Cr 403 (155.0) Ia3d 450 (11.6) Iso 9.99
Cooling: Cr 388 (132.5) Ia3d 448 (−11.8) Iso
B-PB series
B-PB-10 Heating: Cr 425 (23.4) Sm 435 (24.0) SmC
Cooling: Cr 421 (−20.8) SmC Turned back from 480 K
3rd heating: Cr 425 (24.1) Sm 434 (20.2) SmC 526 (12.5) Iso/decomp
B-PB-14 Heating: Cr 406 (66.0) Sm 423 (13.9) Sm 428 (21.4) SmC
Cooling: Cr 362 (−22.8) Sm 378 (−14.0) Sm 417 (−35.1) SmC Turned back from 480 K
3rd heating: Cr 401 (25.6) Sm 424 (14.5) Sm 428 (21.3) SmC 511 (13.3) Iso/decomp
B-PB-16 Heating: Cr 405 (79.2) Sm 418 (13.5) Sm 427 (22.7) SmC
Cooling: Cr 378 (−52.3) Sm 399 (−14.8) Sm 417 (−21.8) SmC Turned back from 460 K
3rd heating: Cr 404 (33.6) Sm 419 (13.6) Sm 427 (22.9) SmC 474 (1.2) Ia3d 505 (7.6) Iso/decomp 9.71
B-PB-18 Heating: Cr 406 (93.0) Sm 414 (13.0) Sm 426 (23.0) SmC 440 (1.7) I432
Cooling: Cr 383 (−53.7) Sm 395 (−11.0) Sm 412 (−25.4) SmC I432 Turned back from 450 K
3rd heating: Cr 405 (38.5) Sm 414 (12.8) Sm 426 (22.7) SmC 441 (1.5) I432 501 (7.5) Iso/decomp 15.38
B-PB-20 Heating: Cr 406 (91.6) Sm 411 (8.5) Sm 425 (26.2) I432 472 (0.2) Ia3d 505 (6.4) Iso
Cooling: Cr 386 (−76.3) Sm 408 (−23.5) I432 475 (0.2) Ia3d 504 (−5.6) Iso
3rd heating: Cr 399 (60.4) Sm 416 (13.7) Sm 422 (17.7) I432 474 (0.3) Ia3d 503 (9.1) Iso 10.18 16.40
B-PB-22 Heating: Cr 407 (141.0) Sm 421 (28.5) I432 440 (0.3) Ia3d 495 (7.1) Iso
Cooling: Cr 386 (−103.3) Sm 407 (−27.2) I432 Ia3d 493 (−6.5) Iso
3rd heating: Cr 408 (74.7) Sm 420 (20.7) I432 443 (0.4) Ia3d 494 (15.5) Iso 10.58 17.17
B-S-22 Heating: Cr 413 (98.5) Sm 420 (29.4) SmC 439 (2.6) I432 486 (0.3) Ia3d 515 (7.3) Iso 11.67 18.16
Cooling: Cr 409 (−111.0) SmC I432 Ia3d 512 (−6.0) Iso
B-PEB-22 Heating: Cr 404 (130.6) Sm 416 (34.5) I432 463 (0.3) Ia3d 490 (5.9) Iso 11.05 18.11
Cooling: Cr 386 (−70.6) Sm 407 (−34.4) I432 Ia3d 488 (−2.6) Iso
PBID-22 Heating: Cr 412 (58.5) Sm 449 (1.0) Sm
Cooling: Cr 417 (−50.1) Sm 448 (−1.2) Sm Turned back from 455 K
3rd heating: Cr 420 (48.5) Sm 448 (1.2) Sm 464 (3.5) Sm 518 (8.8) Iso

Table 2 Estimated core sizes (Lcore) of Cub phase-forming series developed in this work
Core L core/nm Core L core/nm
B 1.41
N 1.87 B-N 1.64
PB 2.27 B-PB 1.87
S 2.74 B-S 2.07
PEB 2.77 B-PEB 2.09

Phase behavior of symmetric series B-n, N-n, PB-n, S-n, and PEB-n.

The type of LC phase that each molecule showed was determined by the combined methods of polarizing optical microscopy (POM) and X-ray diffraction (XRD) studies, and the chirality of the so-called “Im3m20–25 or I432 phase15–19 was confirmed by the presence of a conglomerate of chiral domains under POM observations (see the ESI for the detailed data). Fig. 1 shows typical XRD patterns of S-22 in the three LC phases. Fig. 2 compares the phase diagrams of five families of molecules as a function of temperature and the weight fraction of the core portion (fcore). The data for B-n were cited from the literature.26 In this case, it is better to compare the phase behavior on the scale of fcore rather than the alkyl chain length n. This is because for a given n, for example, n = 22, the role of the alkyl tail would be less when a larger molecular core is incorporated into the compound.
image file: c7cp08345a-f1.tif
Fig. 1 XRD patterns of S-22 in the I432, Ia3d and Colh phases together with their schematic nanostructures; the green portion represents the molecular core regions.

image file: c7cp08345a-f2.tif
Fig. 2 Phase diagrams of five Cub-phase series as a function of temperature and weight fraction of the core part (fcore).

The most important insight gained from Fig. 2 is that although the last PEB-n series exhibits a slight difference in phase, for example no Ia3d phase, all five series with benzene (B-n), naphthalene (N-n), biphenyl (PB-n), stilbene (S-n), and phenylethnylbenzene (PEB-n) show very similar phase behaviors, in that the dependence of the type of phase on the fcore value is very similar to each other. In particular, the range of the I432 phase is very narrow, from about 0.36 to 0.5. However, the larger core size slightly shifts the range to the right side. The second point about these series is that in the upper four series of materials, a common feature of the SmC → Ia3dI432 → Ia3d sequence is visible when going from the right side to left. Therefore, the fcore value is an important factor in controlling the type of the phase formed.

Finally, in the case of a symmetric core without the ability to form hydrogen bonds, PBID-22, no Cub phase was observed. It was found that this compound only forms three kinds of smectic phases (see Table 1 and Fig. S39, ESI). This fact will be discussed later.

Phase behaviors of unsymmetric series B-N-n, B-PB-n, B-S-n, and B-PEB-n

In the case of compounds with unsymmetric cores, it is examined as to what factors determine the type of phase formed. Fig. 3 compares the phase behavior of two series of compounds, the symmetric PB-n series and the unsymmetric B-PB-n series. In the case of the symmetric series, the phase diagram is drawn as a function of temperature and the fcore value, and different phases are indicated with different colors, and thinner colors below the line of melting temperature, Tm, represent virtual phase regions that would appear when the crystalline state is destabilized. On the other hand, the bar graphs represent the phase behavior of the unsymmetric series and beginning from the right, the bars correspond to n = 10, 14, 18, and 22, respectively. Clearly, the phase regions of each bar graph for the unsymmetric series almost match the phase regions for the symmetric series. This indicates that even for unsymmetric core series, such as the B-PB-n series, the formation is mainly governed by the fraction fcore, and thus core-alkyl segregation is a main factor in the control of the formation of the Cub phase. Here, an important point to consider is that the bar graphs for the unsymmetric series extend to more than 50 K below the Tm line of the symmetric PB-n series. This decrease should be ascribed to the effect of the unsymmetry of the molecular core.
image file: c7cp08345a-f3.tif
Fig. 3 Comparison of phase behaviors of two series, symmetric PB-n series (open circles) and unsymmetric B-PB-n series (bar graphs), in the diagram of phase temperature vs. weight fraction of the core part (fcore).

To look more close at the effect of the unsymmetry of the molecular core, the phase behavior of nine compounds with different cores but the same chain length (n = 22) are compared in Fig. 4. Clearly, the extension of the molecular core produces an extension of the Cub-phase temperature range, the temperature interval of the Cub phase for B-22 is 24 K (Ia3d), whereas the interval for the symmetric S-22 is 96 K and it is 75 K for the unsymmetric B-S-22. Both of these are more than three times wider than that for B-22 although the latter two regions include the Ia3d and I432 phases. However, in the case of the symmetric cores, the extension of the core also raises the high temperature limit of the mesophases (Tc), which is in most cases beyond 550 K (277 °C). By contrast, very interesting observations are the results for the unsymmetric cores, where the extension of the core elevates Tc, but this increase is suppressed in comparison with the case of the symmetric cores. In fact, Tc is almost at the midpoint of Tc of its two symmetric counterparts. In contrast, Tm is almost independent of the core extension of the aromatic ring at one end and lies almost at the same temperature as that of B-22. Therefore it appears that the extension of the core extends the Cub phase region in a manner similar to as if the extended region of the symmetric case had moved to a lower temperature.

image file: c7cp08345a-f4.tif
Fig. 4 Thermotropic phase behaviors of nine 1,2-bis(aryloyl)hydrazine-based compounds having C22 chains. It is clear that the compounds having unsymmetric cores show low Tms but much wider Cub phase temperature regions as compared with their parent compound having benzene as its aromatic ring (B-22).

Core size Lcore and cell dimension a

Fig. 5 summarizes the cell dimensions (a's) for nine compounds having different core sizes (Lcore's) but with the same chain length (n = 22) attached to both sides. The cell dimensions were obtained from the XRD patterns of the compounds after indexing the diffraction peaks. Regardless of the core symmetry, a good linear relationship is visible between the cell dimensions and core sizes for both Ia3d- and I432-Cub phases. This suggests that all Ia3d phases possess essentially an isomorphous structure and the same argument is true for all I432 phases. Therefore it is realized that control of the core size can tune the cell dimensions of both Cub phases.
image file: c7cp08345a-f5.tif
Fig. 5 Plots of cell dimension (a) vs. core size (Lcore) for both symmetric and unsymmetric 1,2-bis(aryloyl)hydrazine-based compounds having C22 chains. The a values are extrapolated values at 500 K.

In the case of the unsymmetric cores with the total core size of Lcore, larger and smaller aromatic cores are actually attached at each side of the central dicarbonylhydrazine group. Irrespective of this fact, the cell dimensions of both the symmetric and unsymmetric cores are on the same a vs. Lcore lines for the Ia3d- and I432-Cub phases. This indicates that the unsymmetric cores with Lcore act virtually as a symmetric core extended by Lcore/2 at both sides. This suggests a time variant, random distribution of larger and smaller aromatic cores along the network. The suggestion is also consistent with the aforementioned experimental result of Tc of an unsymmetric core being at the midpoint of Tcs of its two symmetric counterparts. The similar observation was previously obtained for the compounds with a symmetric core but having two different terminal chains Cn and Cm. The compounds actually acted like they have C(n+m)/2.19

As an additional insight, if one extrapolates two as to Lcore = 0, it will produce 5.70 and 8.55 nm for the Ia3d- and I432-Cub phases, respectively. This clarifies the size relationship between the two Cub phases, i.e., the hypothetical volume ratio of the alkyl regions for the two phases is estimated to be ≈3.4, which corresponds to the ratio of the numbers of the molecules contained in one unit cell under the assumption that the chains are in an identical state.

Information from Fourier transform infrared (FT-IR) analysis of intermolecular core–core interactions

To obtain insight into the intermolecular core–core interactions of the compounds, the FT-IR aromatic ring C[double bond, length as m-dash]C stretching (ν(C[double bond, length as m-dash]C)) band frequencies in the Cub phase region (except for the PBID-22 case, which is the Sm phase region) were plotted versus temperature as shown in Fig. 6. This band may be a marker of the intermolecular π–π interactions, and a lower ν(C[double bond, length as m-dash]C) frequency would correspond to a stronger interaction. Nevertheless, a decrease in the band frequency with an increase in temperature would itself reflect a change in the intramolecular electron delocalization. Thus, to eliminate the temperature effect, the values were extrapolated at a given temperature of 500 K and compared as shown in Fig. 6(b). The plot shows a clear tendency for a decrease in the frequency of the ν(C[double bond, length as m-dash]C) band with the increase of the core size. This result is reasonable if one considers that the larger core size would produce a stronger intermolecular π–π interaction. An important point here is that a common relationship exists irrespective of the core symmetry (i.e., symmetric or unsymmetric).
image file: c7cp08345a-f6.tif
Fig. 6 (a) Temperature variation of FT-IR aromatic ring C[double bond, length as m-dash]C stretching (ν(C[double bond, length as m-dash]C)) band frequencies in the Cub phase region, and (b) the plot of the extrapolated value at 500 K vs. the core size, for nine compounds.

Information on the intermolecular hydrogen bonding interactions was obtained from the FT-IR N–H stretching (ν(N–H)) band frequencies. Unexpectedly, the eight series of the compounds showed almost the same ν(N–H) frequency values that ranged between 3270 and 3320 cm−1 in the Cub phase region, with no marked differences between the compounds in the series (see Fig. S48, ESI).

Fig. 7 is a plot of Tm (lower-limit of LC phases) and Tc (higher-limit of LC phases) vs. the molar mass of the molecular core. Roughly speaking, all Tcs are near the same straight line, regardless of the core symmetry. This is a reasonable result since the core–core interaction is a major factor that determines Tc. This result is related to the fact that LC aggregation such as a smectic, cubic, or hexagonal columnar phase can be maintained if the core–core cohesive interaction is operative and prohibits independent free motions of the molecules as in a normal liquid. This is also consistent with the analysis of the above IR results that a stronger intermolecular π–π interaction is obtained in the larger core (Fig. 6b). In this case, it can be said that the core–core interaction originates primarily from intermolecular π–π interactions, because Tc of non-hydrogen-bonding PBID-22 is also present on the same line.

image file: c7cp08345a-f7.tif
Fig. 7 Plots of Tm (violet) and Tc (black) against the molar mass of the molecular core for nine 1,2-bis(aryloyl)hydrazine-based compounds and non-hydrogen-bonding PBID-22, both of which having C22 chains.

In contrast, the results for Tm of the symmetric and unsymmetric core cases are different. The slope of the unsymmetric core case is nearly zero, whereas the slope of the symmetric core case is positive. The origin of the independence of Tm from the core size in the former case is not clear, but is probably related to the unsymmetry of the core, which would provide a profound effect on the side-by-side molecular arrangement and would introduce a disorder upon crystallization. This would lower Tm to almost the same temperatures as that of the smallest, symmetric core of B-22.

Molecular design requirements

The molecular design requirements revealed by this work are now worth discussing. The results of this work have proven that the structural motif presented in Scheme 2 is quite useful in the Cub-phase formation. This motif is composed of a central molecular core and a single aliphatic chain attached at each terminal. Extension of the molecular core from B-n to PEB-n with a doubling of Lcore maintained the Cub-phase formation and increased the thermal stability. This is in marked contrast to the previous failure of exploitation of the Cub phase-forming molecules that were based on core modifications, as mentioned in the Introduction.

The dicarbonylhydrazine [–C(O)NHNHC(O)–] group in the central part of the molecule, which links two aromatic rings, is a prerequisite for the phase formation, even for the extended π-core series with stronger core–core cohesive interactions. This is because the absence of this portion allows a molecular shift along the long molecular axis that will reduce the steric hindrance of the thermally activated terminal chains and brings about a slipped π-stack of molecules. In fact, it was found that the layer thickness of non-hydrogen-bonding PBID-22 gradually decreases with an increase in temperature (see Fig. S39, ESI). Therefore, the dicarbonylhydrazine group produces a pinning effect against the shift, where, as another way of reducing the steric hindrance, a helical network with long molecular axes gradually twisted along the network direction is employed. This also supports our general understanding of the Cub-phase formation.12,26 In the preceding section, it was mentioned that the major core–core cohesive interaction is an intermolecular π–π interaction, which governs the high temperature limit of LC phases (Tc). However, in the case of the Cub-phase formation, intermolecular hydrogen bonding plays a critical role.

In contrast, a symmetric arrangement of two identical aromatic rings at both sides of the central linkage as the case of the initial B-n series is not important in the phase formation. Attachment of different sizes of rings was not found to disturb the Cub-phase formation, although it seems unfavorable for the nearly side-by-side arrangement of molecules with their long axes perpendicular to the two-fold axis of the network direction. By contrast, interestingly, this was found to shift the temperature range of phase formation toward lower temperatures. Here, what is important is that the attachment of different sizes of rings never implies an unsymmetric arrangement of the rings in the Cub phase but instead suggests a random distribution along the networks in a time-varying structure. As a result, the clearing temperature Tc to an isotropic liquid is a mid-value for two symmetric molecules that have either smaller or larger rings, and the low temperature limit Tm is as low as that of the symmetric counterpart with a smaller ring. These are two important findings for development of molecular strategies for producing the desired temperature range of the Cub phases.

Our systematic exploitation of the Cub phases has also revealed that a factor that determines the phase formation is fcore, and the Ia3d phase forms in the range between fcore = 0.3–0.4 and fcore = 0.45–0.6. In addition, the I432 phase forms in the range of fcore = 0.36–0.5, although the larger core size seems to slightly shift the range toward larger fcore values. Note that the importance of fcore does not imply that the operating mechanism is not the same as that operating in block polymers. This is readily understood if one notices that the range of fcore = 0.36–0.5 induces the chiral I432 phase in the present system instead of a lamellar structure that is found in polymeric systems. With respect to the introduction of core unsymmetry, a larger aspect ratio of the core seems to favor greater stability in the I432 phase but the origin is not clear at present. The relationship between the cell dimension (a) and the core size (Lcore) indicates that all obtained Ia3d phases possess an isostructure, and the same argument is true for all I432 phases. Thus, the present core extension could possibly tune the periodicity of the Cub phase structure.


This work is a systematic exploitation of thermotropic Cub-phase families, and it includes core modifications beginning with the well-known B-n series to develop eight new series, four symmetric molecular core series, N-n (1), PB-n (2), S-n (3), and PEB-n (4), and four unsymmetric core series, B-N-n (5), B-PB-n (6), B-S-n (7), and B-PEB-n (8). Examination of their thermotropic phase behaviors has revealed that they are all Cub phase-forming materials, and they proved that suitable selection of a molecular motif as presented in Scheme 2 avoids removal of the Cub phases, in contrast to the previous failures of exploitation. In this case, it was found that the presence of the central dicarbonylhydrazine linkage with intermolecular hydrogen bonding capability is critical for the phase formation. This work also supports our general understanding of the Cub-phase formation. The steric hindrance of thermally activated terminal chains will cause packing frustration in the parallel arrangement of chain-core-chain type molecules, while the cohesive molecular interaction between these linkages works to place the centers of the molecules along the network lines. To avoid the frustration, a helical network is formed with long molecular axes gradually twisted along the network direction. As a result, two Cub phases are formed. The existing relationship between the cell dimension (a) and the core size (Lcore) indicated that all Ia3d-Cub phases obtained possess an isostructure, and the same argument was true for all I432-Cub[*] phases. Thus, the core extension could tune the periodicity of the Cub phase structure.

Furthermore, two important findings that lead to molecular design strategies were revealed. First, the clearing temperature (Tc) to isotropic liquid is mainly determined by the intermolecular π–π interaction, and the larger core produces a stronger π–π interaction. Second, the core symmetry has a profound effect on the melting temperature (Tm) to LC phases, such as the smectic or Cub phase, and introduction of the unsymmetry lowers Tm to that of the symmetric, smallest core B-n. As mentioned in the introduction, the Cub phase structure has great potential for use in many attractive applications. In this reported work, we have improved the basic knowledge required for designing new Cub phase-forming molecules, which we believe will inspire striking innovations based on the Cub phase structure in the near future.


The preparations for symmetric and unsymmetric Cub-phase molecules 1–8 and non-hydrogen bonding PBID-22 (9) were carried out according to the preparation routes described in Scheme S1, ESI. The detailed synthetic procedures and characterization data are compiled in the ESI. The size of the molecular core (Lcore) examined in this reported work was evaluated using the density functional theory (DFT) M06/6-31G* based on the assumption of Ci symmetry. Identification of the mesophases was done using polarizing optical microscopy (POM, Olympus BX53P equipped with a Mettler FP82HT hotstage and a Mettler FP90 controller), differential scanning calorimetry (DSC, SII Nanotechnologies DSC7020), and X-ray diffraction (XRD, Rigaku NANO-Viewer Imaging Plate system using CuKα radiation (λ = 0.154 nm)). To confirm the phase behaviors, time-resolved XRD measurements were performed using synchrotron radiation at the Photon Factory (PF) in the High-energy Accelerator Research Organization (KEK). All the detailed experimental and computational procedures and analytical data are compiled in the ESI.

Conflicts of interest

The authors declare no competing financial interest.


The authors thank Prof. Dr Kazuya Saito and Prof. Dr Yasuhisa Yamamura at the University of Tsukuba for helpful discussions, and Prof. Dr Kaori Ando at Gifu University for help with the synthesis of PBID-22. The authors also thank Akane Kawafuchi, Sho Nakamura, Masaya Kuratsubo, and Kenya Yamauchi at Gifu University for their various experimental aids. Financial support for this work was partly provided by a Grant-in-Aid for Scientific Research (C) 25410091 and 16K05748 from the Japan Society for the Promotion of Science (JSPS) (for SK) and by a Grant-in-Aid for Young Scientists (B) 25810072 from JSPS (for YM). Beam time at PF-KEK provided by Programs 2012G673, 2013G687, 2014G591, 2015G514, 2016G627, and 2017G566 is also acknowledged. For help with the synchrotron experiments we thank Dr Nobutaka Shimizu and Dr Noriyuki Igarashi at PF-KEK. The computations were partly performed at the Research Center for Computational Science (RCCS), Okazaki. The authors are grateful to the two reviewers for their careful reading of the manuscript and for constructive advice.


  1. G. W. Gray and J. W. Goodby, Smectic Liquid Crystals, Textures and Structure, Leonard Hill, Glasgow, 1984, pp. 68–81 Search PubMed.
  2. Handbook of Liquid Crystals, ed. D. Demus, J. W. Goodby, G. W. Gray, H.-W. Spiess and V. Vill, Wiley-VCH, Weinheim, 1998 Search PubMed.
  3. I. W. Hamley, Introduction to Soft Matter, Polymer, Colloids, Amphiphiles and Liquid Crystals, John-Wiley, Chichester, 2000 Search PubMed.
  4. Bicontinuous Liquid Crystals, Surfactant Science Series, ed. M. L. Lynch and P. T. Spicer, CRC Press, Taylor & Francis Group, Boca Raton, 2005, vol. 127 Search PubMed.
  5. C. Tschierske, Annu. Rep. Prog. Chem., Sect. C: Phys. Chem., 2001, 97, 191–267 RSC and references therein.
  6. S. Diele, Curr. Opin. Colloid Interface Sci., 2002, 7, 333–342 CrossRef CAS.
  7. S. Kutsumizu, Curr. Opin. Solid State Mater. Sci., 2002, 6, 537–543 CrossRef CAS.
  8. M. Impéror-Clerc, Curr. Opin. Colloid Interface Sci., 2005, 9, 370–376 CrossRef.
  9. S. T. Hyde, M. O’Keeffe and D. M. Proserpio, Angew. Chem., Int. Ed., 2008, 47, 7996–8000 CrossRef CAS PubMed.
  10. S. Kutsumizu, Isr. J. Chem., 2012, 52, 844–853 CrossRef CAS.
  11. C. Tschierske, Angew. Chem., Int. Ed., 2013, 52, 8828–8878 CrossRef CAS PubMed.
  12. K. Saito, M. Hishida and Y. Yamamura, J. Phys. Soc. Jpn., 2017, 86, 084602 CrossRef.
  13. S. Hyde, S. Andersson, K. Larsson, Z. Blum, T. Landh, S. Lidin and B. W. Ninham, The Language of Shape, The Role of Curvature in Condensed Matter, Physics, Chemistry and Biology, Elsevier, Amsterdam, 1997 Search PubMed.
  14. Y. Nakazawa, Y. Yamamura, S. Kutsumizu and K. Saito, J. Phys. Soc. Jpn., 2012, 81, 094601 CrossRef.
  15. C. Dressel, F. Liu, M. Prehm, X. Zeng, G. Ungar and C. Tschierske, Angew. Chem., Int. Ed., 2014, 53, 13115–13120 CrossRef CAS PubMed.
  16. C. Dressel, T. Reppe, M. Prehm, M. Brautzsch and C. Tschierske, Nat. Chem., 2014, 6, 971–977 CrossRef CAS PubMed.
  17. C. Tschierske and G. Ungar, ChemPhysChem, 2016, 17, 9–26 CrossRef CAS PubMed.
  18. K. Saito, Y. Yamamura, Y. Miwa and S. Kutsumizu, Phys. Chem. Chem. Phys., 2016, 18, 3280–3284 RSC.
  19. S. Kutsumizu, S. Miisako, Y. Miwa, M. Kitagawa, Y. Yamamura and K. Saito, Phys. Chem. Chem. Phys., 2016, 18, 17341–17344 RSC.
  20. A.-M. Levelut and M. Clerc, Liq. Cryst., 1998, 24, 105–115 CrossRef CAS.
  21. S. Kutsumizu, K. Morita, S. Yano and S. Nojima, Liq. Cryst., 2002, 29, 1459–1468 CrossRef CAS.
  22. K. Saito and M. Sorai, Chem. Phys. Lett., 2002, 366, 56–61 CrossRef CAS.
  23. X. Zeng, G. Ungar and M. Impéror-Clerc, Nat. Mater., 2005, 4, 562–567 CrossRef CAS PubMed.
  24. K. Ozawa, Y. Yamamura, S. Yasuzuka, H. Mori, S. Kutsumizu and K. Saito, J. Phys. Chem. B, 2008, 112, 12179–12181 CrossRef CAS PubMed.
  25. M. Vogrin, N. Vaupotič, M. M. Wojcik, J. Mieczkowski, K. Madrak, D. Pociecha and E. Gorecka, Phys. Chem. Chem. Phys., 2014, 16, 16067–16074 RSC.
  26. S. Kutsumizu, H. Mori, M. Fukatami, S. Naito, K. Sakajiri and K. Saito, Chem. Mater., 2008, 20, 3675–3687 CrossRef CAS.
  27. D. Guillon and A. Skoulios, Europhys. Lett., 1987, 3, 79–85 CrossRef CAS.
  28. H. Chung and M. Caffrey, Biophys. J., 1994, 66, 377–381 CrossRef CAS PubMed.
  29. D. Fazio, C. Mongin, B. Donnio, Y. Galerne, D. Guillon and D. W. Bruce, J. Mater. Chem., 2001, 11, 2852–2863 RSC.
  30. V. Luzzati and P. A. Spegt, Nature, 1967, 215, 701–704 CrossRef CAS.
  31. V. Luzzati, A. Tardieu and T. Gulik-Krzywick, Nature, 1968, 217, 1028–1030 CrossRef CAS PubMed.
  32. G. W. Gray, B. Jones and F. Marson, J. Chem. Soc., 1957, 393–401 RSC.
  33. D. Demus, G. Kunicke, J. Neelsen and H. Sackmann, Z. Naturforsch., A: Astrophys., Phys. Phys. Chem., 1968, 23, 84–90 CAS.
  34. S. Kutsumizu, M. Yamada and S. Yano, Liq. Cryst., 1994, 16, 1109–1113 CrossRef CAS.
  35. S. Kutsumizu, K. Morita, T. Ichikawa, S. Yano, S. Nojima and T. Yamaguchi, Liq. Cryst., 2002, 29, 1447–1458 CrossRef CAS.
  36. S. Kutsumizu, K. Saito, S. Nojima, M. Sorai, Y. G. Galyametdinov, I. Galyametdinova, R. Eidenschink and W. Haase, Liq. Cryst., 2006, 33, 75–84 CrossRef CAS.
  37. H. Schubert, J. Hauschild, D. Demus and S. Hoffmann, Z. Chem., 1978, 18, 256 CrossRef CAS.
  38. D. Demus, A. Gloza, H. Hartung, A. Hauser, I. Rapthel and A. Wiegeleben, Cryst. Res. Technol., 1981, 16, 1445–1451 CAS.
  39. W. G. Shaw and G. H. Brown, J. Am. Chem. Soc., 1959, 81, 2532–2537 CrossRef CAS.


Electronic supplementary information (ESI) available: Detailed synthesis procedures and characterization data, and supporting data such as DSC, XRD, POM, DFT, and IR data. See DOI: 10.1039/c7cp08345a
Present address: Sumitomo Riko Company Ltd, 3-1 Higashi, Komaki-City, Aichi 485-8550, Japan.
§ Present address: Teva Takeda Pharma Ltd., 1040-22 Matsunoki-cho, Takayama-City, Gifu 506-0802, Japan.

This journal is © the Owner Societies 2018