Introduction

Liquid crystal (LC) is among the most representative electroresponsive soft materials [1, 2]. For example, typical LC display devices utilize rod-shaped nematic LC molecules that tend to align parallel to each other and switch macroscopic orientations in response to an applied electric field (E-field) to control the transmission of light [3, 4]. This response is realized by the coupling between the E-field and the LC molecules. The LC molecules need a preinstalled E-field-responsive molecular unit. However, designing new E-field responsive LC molecules is not easy. This is because the emergence of the LC property is largely governed by the “shape” of the molecule, and the E-field-responsive molecular unit is not always compatible with this principle. In our group, we employed a different strategy to create new E-field responsive soft materials. We focused on a “dopant” (Fig. 1). A dopant in LC materials is a compound that can largely change the properties of the LC material when a small amount is added. One of the most representative examples is cholesteric LCs [5, 6]. When a chiral dopant is added to a nematic LC, macroscopic helical molecular ordering of several hundred to thousands of nanometers emerges while maintaining the nanoscale parallel molecular arrangement. As a result, a wavelength of light that corresponds to the helical pitch is reflected, and bright reflection color becomes visible. The amount of the dopant is generally smaller than that of the host LC molecules; thus, the dopant can be called a “catalyst” that changes the properties of LC materials. Our strategy for developing novel electroresponsive materials is to couple the E-field to the dopant rather than the host LC molecules. As the amount of dopant necessary is only “catalytic”, the only requirement for the dopant is the minimum solubility to the host LC material. Generally, there is no strong obligation for the molecular shape. Therefore, a variety of molecular designs and E-field responsiveness have become possible. In this Focus Review, we introduce several electroresponsive materials that were developed in our group based on this concept.

Fig. 1
figure 1

Schematic representations of electroresponsive functional soft materials by electroresponsive dopants, together with the molecular structures utilized for the realization of these materials

Full size image

Electroresponsive chiral dopant for color-modulable cholesteric LC

The first LC compound discovered by Friedrich Reinitzer in 1888 was a cholesterol derivative [7, 8]. These molecules organize themselves into a macroscopic helical geometry in their LC phase, which is called the cholesteric phase or chiral nematic phase. The most characteristic feature of cholesteric LCs is selective reflection due to internal helical ordering. The reflection wavelength (λ) is determined by the helical pitch (P) as follows:

$$\lambda =nP$$
(1)

where n is the average refractive index of the LC. When the wavelength and the corresponding helical pitch fall into the range of the visible region, a bright reflection color can be observed. Meanwhile, the cholesteric phase is formed not only from the cholesterol derivative but also from a nematic LC doped with a chiral molecule (dopant). In such a case, the reflection wavelength of the cholesteric LCs is determined as follows:

$$\lambda =nP=n/{\beta }_{M\,\cdot }C$$
(2)

where βM is the helical twisting power based on a molar fraction of the dopant (HTP; μm–1) and C is the molar concentration of the dopant.

For the application of cholesteric LCs to display devices, we need to endow the LCs with electrical responsiveness. Cholesteric LCs can change the direction of the internal helical axis in response to an applied electric field (E-field) [9, 10]. When the helical axis is perpendicular to the substrate surface (planer state), a specific wavelength of light (λ) will be reflected as described above (selective reflection). When the helical axis is randomly oriented (focal conic state), the incident light will be scattered, and the device will look opaque. These two states can be reversibly switched using an E-field, and each state can be maintained even after the removal of the E-field. This bistability is applicable to electronic paper (E-paper) [11, 12]. However, this system can only switch the reflection on and off but cannot change the color of the LC device.

To change the reflection color of the cholesteric LC devices, the helical pitch must be changed by electrical stimuli. However, this method is known to be difficult. In the earliest example, a dedicated electrical cell was utilized for the display device, which can apply an E-field perpendicular to the helical axis [13]. In this way, the constituent polar rod-shaped molecule is forced to align parallel to the E-field. Accordingly, the helical pitch will be expanded to shift the reflection wavelength to a longer wavelength region. This method not only requires a dedicated cell but also a voltage above 100 V. Additionally, this device cannot realize a color shift in the visible wavelength region. The other method is to utilize a molecule where the molecular dipole is perpendicular to the long axis of the molecule (negative dielectric anisotropy) [10, 14, 15]. Cholesteric LCs obtained using such molecules can tilt the helical axis in response to the applied voltage (>40 V) [10]. However, because each domain has a different tilting angle, the tilting orientation can be different by the LC domains; thus, the interference of light scattering with the reflection color is likely unavoidable. The use of a command surface is also reported to change the reflection color of cholesteric LCs [16, 17].

Before our work, nearly all previous approaches for the electrical modulation of cholesteric reflection color were based on a design principle that involves perturbing the arrangement of host LC molecules by directly controlling the LC molecules. On the other hand, we focused on electrically modulating the properties of the chiral dopant to change the cholesteric reflection color (Figs. 2 and 3). The only example of this approach is the use of ferroelectric LC as a dopant for cholesteric LCs [18]. To date, we have succeeded in developing two kinds of electroresponsive chiral dopants. One can change its concentration within the host LC molecules (Generation 1: IonD, Fig. 2) [19]. Due to this dopant, the electrically modulated reflection color can be memorized, which can recover its original color simply by heating. This bistable color modulation was realized for the first time. This character can be a good candidate for full-color electronic paper using a single LC cell. The other type of electroresponsive dopant is the redox-responsive dopant, which can change its HTP upon oxidation/reduction (Generation 2: FcD, Fig. 3) [20]. This dopant can realize quick color modulation (<1 s) by the application of a voltage as low as 1.5 V.

Fig. 2
figure 2

a Schematic representation of the mechanism behind the reflection color change by EPD of IonD. Pictures I and II show a center part of the device before and after the E-field application, respectively. The final color was induced by heating the device to an isotropic temperature followed by cooling to generate a cholesteric phase. b Deposition of particles composed of ionic chiral dopant IonD at the electrode surfaces. (i–iii) SEM images of the electrode surfaces derived from the devices, (i) in the initial state, (ii) after E-field application (40 Vpp, 100 Hz for 360 s), and (iii) followed by heating (80 °C for 24 h). (iv) Enlarged SEM image of the deposited material in the red square in (ii). (v) SEM-EDX mapping of phosphorous in (iv). c Memory of E-field-modulated reflection colors. The initial blue color (picture I and transmittance spectrum I) changed to green (picture II and transmittance spectrum II) upon application of an E-field (40 Vpp, 100 Hz) for 0, 60, 120, 240, and 360 s. The E-field was turned off whenever each spectrum was measured. The color remained unchanged for 1 week at 45 °C after the E-field was removed (picture III and transmittance spectrum III). After the device was subsequently heated at 80 °C for 24 h, the color reverted to its initial state (picture IV and transmittance spectrum IV). Pictures of a cholesteric LC device were taken at 45 °C. d Time-dependent transmittance spectral changes at the center wavelength of the reflection band before and after turning off the E-field application. e A cholesteric LC display device utilizing a simple 8 × 8 matrix cell. Reproduced from [19] with permission (Copyright: 2016 Wiley-VCH)

Full size image
Fig. 3
figure 3

Chemical structures of a ferrocene-appended chiral dopant FcD and b its oxidized form FcD+, together with schematic representations of FcD- and FcD+-doped cholesteric LCs comprising 5OCB sandwiched between two glass substrates. c Schematic representation and photograph (inset) of the FcD-doped cholesteric LC device containing EMIm-OTf in 5OCB. The illustration represents the mechanism of electrochemical modulation of the reflection color. d Photographs of FcD- and FcD+-doped cholesteric LCs in a sandwich-type glass cell at 37 °C at varying concentrations of FcD and FcD+. e Photographs and f transmittance spectra of the FcD-doped LC device containing 3.0 mol% EMIm-OTf in 5OCB in its initial state (I) after the application of 1.5 V for 4 s (II), and subsequent application of 0 V for 8 s (III) at 37 °C. g Changes in the transmittance of the LC device at 510 nm upon switching the applied voltage between +1.5 and 0 V. h Details of the transmittance change of the LC device at 510 nm. Reproduced from [20] with permission (Copyright: 2018 American Chemical Society)

Full size image

The concept of dopant Generation 1 is to utilize electrophoretic deposition to change the concentration of the dopant by electrical stimuli. According to Eq. 2, the change in the concentration of the chiral dopant leads to a change in the reflection color of the cholesteric LC material. Thus, we designed the molecule IonD (Fig. 1), which has an imidazolium-based ionic moiety and binaphthyl moiety; this is a well-known framework often employed as a core structure of chiral dopants [6]. This dopant was revealed to exhibit an HTP of βM = 18 μm–1. The LC material based on IonD (1.1 mol%) contains 4′-pentyloxy-4-cyanobiphenyl (5OCB, Fig. 1), which acts as the host LC, a neutral chiral dopant (NeuD, βM = 83 μm–1, 3.4 mol%, Fig. 1) to compensate for the low HTP of IonD for realizing reflection in the visible range. We also added a small amount of acetic acid (1.1 mol%), which was found to enhance the reproducibility and uniformity of the E-field-induced color change. When this LC mixture was introduced to a sandwich-type electrical cell coated with a polyimide orientation film, the LC device exhibited a bright blue reflection color at 45 °C (Fig. 2a, I). The shape of the transmittance spectrum of the LC device (Fig. 2c, I in the right top spectrum) was typical for a cholesteric LC whose bandwidth (Δλ) was 47 nm. This value matched well with the calculated value of 55 nm based on the HTP values of the chiral dopants together with the ordinary (no = 1.53) and extraordinary (ne = 1.71) refractive indices of 5OCB at 45 °C. Upon application of an alternating current (AC) E-field at a voltage of 40 V peak-to-peak (Vpp) at 100 Hz, the reflection color gradually changed to a longer wavelength region and reached a plateau in 360 s, resulting in the emergence of a green reflection color (Fig. 2a, II). The green color remained unchanged for at least one week (Fig. 2c, d). After the cell was heated at 80 °C for 24 h, the device recovered its original blue reflection color (Fig. 2c, d). Among several previous works on the E-field modulation of cholesteric reflection color, our device is the first to realize rewritable color memory. Our design of the mechanism for this color change was an electrophoretic deposition (EPD) of the ionic chiral dopant IonD to the electrode surface (Fig. 2a). Once IonD is deposited onto the electrode surface, the effective concentration of IonD is significantly reduced, thereby elongating the reflection wavelength according to Eq. 2. In addition, the deposited dopant remains on the electrode surface unless it is redissolved into the host LC again. The mechanism of this color modulation/memorization was clarified by time-of-flight secondary-ion mass spectrometry (TOF-SIMS) and scanning electron microscopy (SEM) with energy dispersive X-ray microanalyzer (EDX) of the electrode surface of the cell, in which the deposited dopant was clearly observed (Fig. 2b). By using a matrix cell, we can write and erase different numbers repeatedly (Fig. 2e).

The difference between chiral dopant Generation 1 (IonD) and Generation 2 (FcD) is the background mechanism of the color change. According to Eq. 2, the reflection color can be changed by the change in concentration of chiral dopant and change in the HTP (βM). Although several chiral dopants can change their HTP values upon application of external stimuli [21], mostly using light [22,23,24,25,26,27,28,29,30], electrical stimuli have never been utilized to change the HTP. During the investigation of chiral dopant Generation 1, we noticed that ionic chiral dopants tend to exhibit lower HTP values than those of neutral dopants. From this observation, we decided to introduce redox-responsive units as a part of the chemical structure of chiral dopants, as redox-responsive units can change their ionicity upon reduction or oxidation. Thus, we designed FcD (Fig. 1), in which the redox-responsive ferrocene (Fc) unit was introduced to the binaphthyl unit. When FcD (3.0 mol%, Fig. 1) was mixed with 5OCB (Fig. 1) and introduced into a sandwich cell, the material exhibited a blue reflection color centered at 479 nm (Fig. 3d, top, 3.0 mol%). When we chemically oxidized FcD with NOBF4 to generate FcD+, the reflection color changed to green (λ = 530 nm) (Fig. 3d, bottom, 3.0 mol%). By measuring the reflection wavelength at various concentrations of FcD (Fig. 3d), we determined the HTP values (βM) of FcD and FcD+ to be 116 and 101 μm–1, respectively (Fig. 3a, b). These results strongly support our hypothesis that ionic chiral dopants show lower HTP values than those of their neutral counterpart. To realize the redox-responsive color change, we need to consider several factors. The first step is to introduce an electrolyte to enhance the redox reaction within the LC mixture, as electrochemical reactions are promoted in the presence of an electrolyte [31]. Thus, we employed 1-ethyl-3-methylimidazolium trifluoromethanesulfonate (EMIm-OTf) (3.0 mol%). The second point is the counter redox couple. When oxidation of FcD occurred at one side of the electrode, another component had to be reduced at the other electrode. For this purpose, we fabricated a PEDOT film on one side of the electrode (Fig. 3c) [32]. After 1.5 V was applied to the LC device using 3.1 mol% FcD in 5OCB, color modulation from blue to green occurred quickly (Fig. 3e–h). When we measured the time for 90% color change, it was only 0.4 s, which is the fastest among any E-field responsive color-modulable cholesteric LC system. In addition, the color modulation occurred with a 1.5 V dry-cell battery.

Currently, the electroresponsive dopant Generation 1 can achieve color memory, but the color modulation speed is very slow (forward: 6 min (45 °C), backward: 24 h (at 80 °C)), and the necessary voltage is very high (40 V). On the other hand, Generation 2 can realize a quick color change (forward: 0.4 (37 °C), backward: 2.7 s (37 °C)). If we want to achieve a full-color E-paper, both color memory and quick response, which are usually mutually exclusive, must be realized simultaneously. A chiral dopant with multiredox properties could be a solution to realize this function.

Electroresponsive helical supramolecular dopant for dual-responsive columnar LC material

Nematic LCs are known to be quickly alignable by an external E-field, and this character is elegantly utilized in LC display devices, as described above. However, columnar LCs cannot be easily aligned by the E-field. This is due to their two-dimensional molecular ordering, which is higher than that of one-dimensionally ordered nematic LCs. High molecular ordering leads to high structural integrity and thus leads to high viscosity and low fluidity of the LC material. Although columnar LCs have promising applications for organic electronics, separation membranes, and high-density memory devices due to their capabilities for electron transport, mass transport, and polarization along the columnar axis, respectively, low alignability has always been an issue for developing practical devices out of columnar LCs. In 2011, Aida and his coworkers reported a simple but general approach to this issue [33]. They found that introducing an amide group at the side chain unit of the disk-shaped core is an effective method to confer E-field responsiveness to the resultant columnar LCs. This paper demonstrated the effectiveness of the amide groups to realize a homeotropic alignment of various 1D-stacked π-conjugated discs parallel to the applied E-field [34,35,36]. However, the responsiveness is not satisfactory; the orientation is achieved by the slow cooling of their isotropic melt to a temperature that shows the LC phase under the application of an E-field. Nevertheless, achieving in situ isothermal orientation of columnar LCs is still difficult.

In 2019, we reported a new columnar LC system that is mainly composed of rod-shaped nematic LC molecules doped with disk-shaped molecules capable of H-bonding-induced supramolecular polymerization [37] (Fig. 4c). Considering that rod-shaped molecules tend to align parallel to each other and disk-shaped molecules tend to stack one-dimensionally, it is difficult to coassemble these two types of molecules. When the polymer-LC affinity is large, it results in the dispersion of the polymer in the LC matrix (Fig. 4b). On the other hand, when the polymer LC affinity is small, macroscopic phase separation occurs (Fig. 4d). In the latter case, the cross-linked network, formed by the insoluble polymer bundles, transformed the LC material into a gel to exhibit interesting physical properties [38]. In our case (Fig. 4c), introducing the mesogenic moiety, similar to that of rod-shaped nematic LC molecules, at the termini of the side chain of the disk-shaped molecules (Fig. 1) was the key to realizing the order-increasing mesophase transition to a columnar phase; the mesogenic moiety functions as a compatibilizing unit for mixing these two molecules. This unusual coassembly system was not discovered through de novo design but by chance. As a continuation of the above-described E-field responsive cholesteric LC devices, we investigated a helical supramolecular polymer as a new chiral dopant. However, when we utilized a chiral benzenetricarboxiamide (BTA) analog appended with a long aliphatic side chain, one of the most commonly utilized chiral monomers for helical supramolecular polymerization through H-bonding interactions, we observed bundled fibers that were phase-separated from the host nematic LC 5OCB (Fig. 4d). To prevent phase separation, we introduced an oxy-cyanobiphenyl (OCB) moiety at the termini of the aliphatic side chain (OCBBTA) and mixed it with 5OCB (OCBBTA/5OCB = 1/6). This time, the two components were completely miscible. To our surprise, the whole system turned into a columnar rectangular phase upon cooling from its isotropic melt to 100 °C (Fig. 5a, b). This is counterintuitive, as 86 mol% of the LC mixture is composed of rod-shaped 5OCB, which preferentially aligns parallel to each other to generate a nematic LC phase. After being mixed with OCBBTA, 5OCB underwent an order-increasing mesophase transition to a columnar phase (Fig. 4c).

Fig. 4
figure 4

Schematic representations of possible modes of supramolecular polymerization in LC media. A monomer can form its one-dimensional supramolecular polymer in a nematic LC medium. Depending on the affinity between the monomer or the resultant supramolecular polymer and the LC medium, four different modes are expected: a monomer dispersion, b polymer dispersion, c order-increasing mesophase transition and d phase transition. Reproduced from [37] with permission (Copyright: 2019 The American Association for the Advancement of Science)

Full size image
Fig. 5
figure 5

a, d, g, j XRD profiles, b, e, h, k POM images, and c, f, i, l CD spectra of a–c OCBBTA*-5OCB (molar ratio = 1/6) at 100 °C, d–f OCBBTA*-BMAB (molar ratio = 1/6) at 140 °C, g–i NOBTA*-5OCB (molar ratio = 1/3) at 110 °C, and j–l OCBTPP-5CB* (molar ratio = 1/12) at 40 °C. a–f, g–i, and j–l are reproduced from [37] (Copyright: 2019 American Association for the Advancement of Science) [40], (Copyright: 2019 American Chemical Society) and [47] (Copyright: 2022 Wiley-VCH), respectively, with permission

Full size image

Columnar LC material is composed of columns that are two-dimensionally packed. In our new rod-disk hybrid columnar LC, the individual column has a core-shell structure in which a helically assembled OCBBTA is surrounded by a helically assembled 5OCB (Fig. 4c). When looking at the orientation of the functional groups, the amide group is parallel to the columnar axis, while the cyano group is oriented perpendicularly and radially to the columnar axis (Fig. 6c, d). This unique orientation of the polar functional groups led to the emergence of the dual frequency response of the LC material, the macroscopic orientation of which can be controlled in different directions at two different frequencies (Fig. 6b). This dual frequency response function is often reported for nematic LC materials [39] but has not been reported for columnar LCs before our study. We introduced OCBBTA-5OCB (molar ratio = 1/6) into an electrical cell with a comb-type electrode and applied a direct current (DC) E-field (40 V µm–1) at 100 °C (Fig. 6a, b). The synchrotron 2D X-ray diffraction profile showed two clear diffraction arcs only in the equatorial direction [37]. This indicates that the LC columns are oriented parallel to the applied DC E-field. Polar plots of the polarized infrared (IR) absorption intensities arising from the hydrogen-bonded (H-bonded) amide N–H (3239 cm–1) and C ≡ N (2224 cm–1) groups showed that the N–H and C ≡ N groups preferentially align parallel and perpendicular to the applied E-field. On the other hand, when the applied E-field was switched from DC to AC (10 kHz, 24 V µm–1) at 100 °C, the orientation of the LC columns changed from parallel to perpendicular relative to the applied E-field. The orientation of the N–H and C ≡ N groups also changed accordingly (Fig. 6c, d). Importantly, orientation switching can be achieved isothermally through electrical reconstruction, in which electrical disassembly and subsequent assembly of the core-shell columns occur. Generally, isothermal orientation switching of the columns in columnar LCs is difficult due to their high structural integrity, which leads to a high energy barrier for reorientation (Fig. 6j). However, our hybrid columnar LC system can entropically lower the energy barrier by the disassembly of OCBBTA and 5OCB (Fig. 6i). Thus, our hybrid columnar LCs exhibit high levels of structural integrity and stimuli responsiveness, which are mutually exclusive.

Fig. 6
figure 6

a Schematic representation of a comb-type electrical cell comprising two parallel glass plates (20 μm of separation) for in-plane E-field application. b, e, f Schematic representations and c, d, g, h polar plots of the polarized FT-IR absorption intensities of b–d OCBBTA*-5OCB (molar ratio = 1/6) at 100 °C and e–h OCBBTA*-BMAB (molar ratio = 1/6) at 110 °C. b–d The sample was introduced into the comb-type electrical cell and oriented parallel and perpendicular to the directions of the applied DC E-field (40 V µm–1) and AC E-field (square wave, 10 kHz, 24 V µm–1), respectively. e, g The sample was introduced into the comb-type electrical cell and subjected to a DC E-field (40 V µm–1) upon cooling at 1 °C min–1 from its isotropic phase. f, h The sample was sandwiched by CaF2 plates and placed in a 9 T magnetic field upon cooling at 1 °C min–1 from its isotropic phase. c, d, g, h Polarized FT-IR absorption intensities due to H-bonded N − H (3239 cm−1; purple), C ≡ N (2224 cm−1; orange), N − O• (1363 cm−1; red), and aryl C − O (1249 cm−1; light green) were plotted at every 5° interval and were normalized relative to each corresponding minimum value. Arel is the relative absorbance. i, j Energy diagrams for the electrical reconstruction of columnar LC materials. i The noncovalent core-shell column bears an intermediate state featuring disassembled core and shell modules, which j the covalent column lacks. a–d, i, j, e–h are reproduced from [37] (Copyright: 2019 The American Association for the Advancement of Science) and [40] (Copyright: 2019 American Chemical Society), respectively, with permission

Full size image

The hybrid system exhibits several merits, as it confers a stimuli responsiveness to structurally rigid columnar LCs; in addition, various functions can easily be accommodated with the columnar LCs just by mixing the components. For example, the azobenzene-based rod-shaped nematic LC molecule 4-butyl-4’-methoxy azobenzene (BMAB) can be used instead of 5OCB to give a core-shell columnar LC (Fig. 5d, e). When OCBBTA and BMAB were mixed in a molar ratio of 1/6, they showed a columnar rectangular LC phase at 140 °C. Interestingly, even though the molecular structure of BMAB does not match with the terminal unit of OCBBTA, they can be integrated to form a columnar LC phase without macroscopic phase separation, demonstrating the flexibility of the hybrid columnar LC system. Since BMAB undergoes photoisomerization and OCBBTA has E-field responsiveness, this hybrid system OCBBTA-BMAB exhibits both E-field and photoresponsiveness. For the investigation of the E-field and photoresponsiveness, we introduced the LC sample OCBBTA-BMAB (molar ratio = 1/6) into a sandwich-type electrical cell partly covered with ITO (Fig. 7a). Under a polarized optical microscope (POM), the LC material exhibited a fan-shaped texture typical of columnar LCs (Fig. 5e). When a voltage of 20 V µm–1 was applied to the sample at 140 °C, the fan-shaped texture in the E-field-exposed area gradually disappeared and became entirely nonbirefringent [37]. This indicates that the columns are aligned perpendicular to the substrate (homeotropic alignment). On the other hand, when a 365 nm ultraviolet (UV) light was irradiated on the sample in the cell at 140 °C, a phase transition from a columnar LC phase to an isotropic melt occurred and generated a dark image under POM [37]. By taking advantage of this dual responsiveness (E-field and light), we fabricated an LC-based rewritable “AND” logic gate (Fig. 7c). For this purpose, an E-field of 5 V µm–1 was utilized, which is too weak to align the LC columns isothermally but can assist the alignment of the columns during the reassembly process from the isotropic melt. When this voltage was applied to the sample, no change in the texture was observed under POM (Fig. 7c, ii, iii). When a 365 nm UV light is applied to a selected area of the sample, they once turned dark but immediately became birefringent once the light was turned off (Fig. 7c, vi, vii). Then, when an E-field and light were applied, the selected area remained dark even after the input signal was removed (Fig. 7c, iv, v). This operation is called the “AND” logic gate. Importantly, the material can be initialized by irradiating the whole area with UV light (Fig. 7c, i).

Fig. 7
figure 7

a Schematic representation of a sandwich-type glass cell comprising two parallel glass plates (5 μm of separation) for vertical E-field application and UV light irradiation. b Schematic representations of the photoinduced depolymerization and repolymerization of OCBBTA*-BMAB (molar ratio = 1/6) between the columnar LC and its isotropic liquid upon UV light irradiation (365 nm) and backward thermal relaxation. c POM images of OCBBTA*-BMAB (molar ratio = 1/6) upon application of a power-tuned DC E-field (5 V μm–1) and/or upon irradiation with 365 nm UV light with a lattice-patterned photomask. Dashed squares represent areas exposed to UV light, whereas the entire area was exposed to the E-field. Scale bars represent 200 μm. Reproduced from [37] with permission (Copyright: 2019 The American Association for the Advancement of Science)

Full size image

To investigate the flexibility of the disk-shaped molecule, we synthesized NOBTA bearing a bulky organic radical, 2,2,6,6-tetramethylpiperidine 1-oxyl (TEMPO) (Fig. 1) [40]. Usually, such bulky spherical molecular units are incompatible with disk- or rod-shaped molecules due to the mismatch in their assembling preference. However, our hybrid columnar LC system is flexible enough to accommodate such bulky units. Although NOBTA was a viscous liquid by itself, a mixture of NOBTA and 5OCB in a molar ratio of 1/3 exhibited a hexagonal columnar LC phase at 110 °C (Fig. 5g, h). Interestingly, the structural order of the NOBTA-5OCB system (p6mm symmetry) is higher than that of the OCBBTA-5OCB or OCBBTA-BMAB system (c2mm symmetry). Therefore, the NOBTA-5OCB system is less prone to be oriented by the E-field. When the sample was subjected to an in-plane E-field of 40 V µm–1 in its mesophase temperature range at 110 °C using a comb-type electrode, the texture under POM remained unchanged [40]. This is in sharp contrast to the previous OCBBTA-5OCB system, which can be oriented isothermally [37]. However, when the E-field was applied during the cooling process (1 °C min–1) from its isotropic melt, the columnar mesophase was unidirectionally ordered, as confirmed by the POM texture, which showed bright- and dark-field images alternately every 45° [40], as well as by the polarized FT-IR spectral profile (Fig. 6g). Notably, NOBTA-5OCB (molar ratio = 1/3) could be aligned by a magnetic field. When the sample was sandwiched by calcium fluoride plates and cooled from its isotropic melt to 110 °C at the rate of 1 °C min–1 under a 9 T magnetic field, the resultant columnar LC material exhibited a dichroic feature in POM [40] and polarized FT-IR (Fig. 6h). In contrast, the OCBBTA-5OCB system was not aligned by a magnetic field. This indicates that the magnetic-field response was conferred by the TEMPO moiety. However, based on electron paramagnetic resonance (EPR) spectral measurements and magnetic susceptibility measurements, NOBTA-5OCB (molar ratio = 1/3) did not show a sign [40] of a magneto-LC effect [41,42,43,44,45,46]. The observed magnetic-field alignment might be due to the effect of the bulkiness of the TEMPO group, which makes the LC more mobile. This is the first example of a columnar LC material that is responsive to an E-field and magnetic field.

One important aspect of this hybrid columnar LC system is the molecular-level mixing of the shape-incompatible molecules. By means of circular dichroism (CD) spectroscopy, we clarified that the helicity of the supramolecular polymer is transferred to the chiral arrangement of the rod-shaped molecules and vice versa. In the case of OCBBTA-5OCB (molar ratio = 1/6), the CD spectra of the disk-shaped molecule and its helical polymer mostly overlap with that of 5OCB, and thus, the relationship between the helical ordering of the two components was difficult to investigate (Fig. 5c). On the other hand, for OCBBTA-BMAB (molar ratio = 1/6), the chromophores of OCBBTA and BMAB have different absorption bands, and thus, the chiral ordering of each component can be investigated (Fig. 5f). The CD spectra of OCBBTA-BMAB (molar ratio = 1/6) at 140 °C showed distinct bands originating not only from the OCB moiety (320 nm) but also from BMAB (370 nm) (Fig. 5f). This indicates that the core and shell parts are helically assembled within the column. As for BMAB, NOBTA-5OCB (molar ratio = 1/3) was optically active in the absorption range of the OCB moiety (320 nm) and TEMPO (~500 nm) (Fig. 5i). This indicates that the radical moiety is in close vicinity to the helical core. These observations clearly demonstrate that the helical arrangement of the disk-shaped molecules is successfully transferred to the arrangement of the rod-shaped molecules.

Here one question arose; can the chirality of the shell part consisting of rod-shaped molecules be transferred to the helicity of the core part consisting of disk-shaped molecules? To explore this possibility, we designed a new achiral H-bonding disk-shaped molecule, OCBTPP, which has a tetraphenyl porphyrin (TPP) core appended with four OCB-terminated achiral side chains (Fig. 1) [47]. In the previous design using BTA, the electronic absorption and CD spectral bands for the OCB unit and BTA unit overlapped with each other, and thus, it was difficult to investigate the chiral transfer from the shell to the core. Through the new molecule OCBTPP, we could observe the spectral bands separately between the OCB unit and TPP unit due to the large difference in the absorption range. For a rod-shaped molecule, we employed 4-cyano-4’-(2-methylbutyl)biphenyl (CB*), which has a chiral center on its C5 side chain (Fig. 1). This molecule was reported to show a smectic A phase at > –54 °C and a cholesteric phase at –54~ –30 °C [48]. As for the above examples, OCBTPP-CB* (molar ratio = 1/12) exhibited a columnar phase (p2 oblique lattice) below 48 °C (Fig. 5j, k). Interestingly, the CD spectral profile clearly exhibited a bisignate Cotton effect at 400–450 nm, which corresponds to the Soret band due to the porphyrin core (Fig. 5l). This clearly indicates that chiral transfer from the shell to the core took place. Therefore, the helical core and the shell exhibit a strong interaction to affect the molecular arrangement. More interestingly, we found an anomalous relationship between the degree of chirality of the rod-shaped molecule and the CD intensity. When the difference between the peak top CD intensity values of positive and negative Cotton effects ( | A | ) was calculated and plotted as a function of the %-enantiomeric excess (%ee) of CB*, the plot showed a sigmoidal curve [47]. Such a nonlinear relationship is known as the “majority rule”, which is often observed in the supramolecular polymerization of chiral monomers in an achiral solvent [49, 50]. In supramolecular polymerization, a similar relationship called the “sergeant and soldier principle” is also known, in which the incorporation of a chiral monomer into the supramolecular polymer of an achiral monomer can sigmoidally increase the CD intensity [5, 51,52,53,54]. However, when we utilized a mixture of chiral CB* and achiral 5-cyanobiphenyl CB as a rod-shaped component and varied the %-mole fraction of CB* in CB*/CB (χ), an anomalous bell-shaped relationship was observed (Fig. 8a). The maximum CD intensity reached 1800 mdeg μm–1 at χ = 50%. This corresponds to the anisotropic dissymmetry factor gabs of 0.049 (λ = 441 nm), which is the highest value among all reported helical supramolecular polymers [5, 53, 54] as well as columnar LCs [55, 56]. Interestingly, the vibrational circular dichroism (VCD) spectroscopy of the same sample set gave an ordinary sigmoidal relationship between the χ value and the intensity of the VCD band (Δabs) (Fig. 8b,c). This is similar to the typical “sergeant-soldier principle” [5, 51,52,53,54]. These results indicate that the anomalous nonlinear relationship observed for CD intensity is not due to the enhanced helical induction but the change in the spatial relationship between the chromophores. The detailed investigation between the intercolumnar distance and χ revealed that the intercolumnar distance decreased upon increasing the achiral component (Fig. 8d). Considering that the shape and intensity of the CD spectrum are highly sensitive to the stacking manner of chromophores [57, 58], the anomalous trend in the CD intensity is likely due to the change in the stacking manner of the porphyrin units caused by the change in the intercolumnar interaction. Namely, the enhancement in the CD intensity from χ = 100 to 50% is due to the change in the stacking manner of the chromophores by the increased intercolumnar interaction, and the subsequent CD intensity drop is due to the decrease in the chiral component in the material.

Fig. 8
figure 8

a ΔA values of OCBTPP (molar ratio = 1/12) in CBR100/CB0, CBR75/CB25, CBR50/CB50, CBR25/CB75 (red circles), CB (black circle), CBS25/CB75, CBS50/CB50, CBS75/CB25, CBS100/CB0 (blue circles) at 40 °C. b VCD spectra of OCBTPP in CBR/CB (molar ratio = 1/12) (red trace) and OCBTPP in CBS/CB (molar ratio = 1/12) (blue trace) at 30 °C. c VCD peak top signal intensities that correspond to the C = O stretching vibration (~1638 cm–1) of OCBTPP (molar ratio = 1/12) in CBR100/CB0, CBR75/CB25, CBR50/CB50, CBR25/CB75 (red circles), CB (black circle), CBS25/CB75, CBS50/CB50, CBS75/CB25, and CBS100/CB0 (blue circles). d Intercolumnar distances of OCBTPP in CB*/CB (molar ratio = 1/12) at various % mole fractions of CB* at 40 °C. Reproduced from [47] with permission (Copyright: 2022 Wiley-VCH)

Full size image

Summary and perspective

In this Focus Review, we summarized our new concept of electroresponsive dopants to construct novel electroresponsive soft materials. The dopants we designed respond to electrical stimuli in different ways. The one used for the cholesteric display bearing a rewritable color memory function (IonD) responds to an E-field to undergo electrophoretic deposition and change its concentration in the LC material. On the other hand, the one used for the cholesteric display with a quick color modulation function (FcD) undergoes a redox reaction to change its ionicity and eventually changes its HTP value. The disk-shaped supramolecularly polymerizable dopants (OCBBTA, NOBTA) not only function as a property-changing molecule but also as a structure-transforming agent to a higher-order molecular ordering of the system. The response of those dopants is not the response of the individual molecules but rather their polymerized 1D architecture, in which they can change their orientation depending on the direction and the frequency of the E-field. The dopant concentration we employed was below 15 mol%. This value could be lowered more by an appropriate molecular design. Our study demonstrates that modifying the constituent molecule of such a small amount can drastically change the properties of the host materials. This is an efficient way to confer stimuli-responsiveness to a material, allowing facile material design and energy-efficient function. One interesting perspective is to design a dopant that can operate within a lyotropic liquid crystal phase. By doing so, a material could be designed that can operate in a biological system, such as a cell membrane, with a minimum amount of dopant molecules.

Thus far, LC materials have created an era through their application to display devices. The exploration of new possibilities for LC materials is currently active. We hope that our work will help reveal a new direction for next-generation LC materials.