Hale
Ocak
*ab,
Marco
Poppe
a,
Belkız
Bilgin-Eran
b,
Gürkan
Karanlık
b,
Marko
Prehm
a and
Carsten
Tschierske
*a
aInstitute of Chemistry, Organic Chemistry, Martin Luther University Halle-Wittenberg, Kurt-Mothes-Str. 2, D-06120 Halle, Germany. E-mail: ocak_hale@hotmail.com; carsten.tschierske@chemie.uni-halle.de
bDepartment of Chemistry, Yildiz Technical University, Davutpasa Yerlesim Birimi, TR-34220 Esenler, Istanbul, Turkey
First published on 29th July 2016
A bent-core compound derived from a 4-cyanoresorcinol core unit with two terephthalate based rod-like wings and carrying chiral 3,7-dimethyloctyloxy side chains has been synthesized in racemic and enantiomerically pure form and characterized by polarizing microscopy, differential scanning calorimetry, X-ray diffraction and electro-optical investigations to study the influence of molecular chirality on the superstructural chirality and polar order in lamellar liquid crystalline phases. Herein we demonstrate that the coupling of molecular chirality with superstructural layer chirality in SmCsPF domain phases (forming energetically distinct diastereomeric pairs) can fix the tilt direction and thus stabilize synpolar order, leading to bistable ferroelectric switching in the SmC* phases of the (S)-enantiomer, whereas tristable modes determine the switching of the racemate. Moreover, the mechanism of electric field induced molecular reorganization changes from a rotation around the molecular long axis in the racemate to a rotation on the tilt-cone for the (S)-enantiomer. At high temperature the enantiomer behaves like a rod-like molecule with a chirality induced ferroelectric SmC* phase and an electroclinic effect in the SmA′* phase. At reduced temperature sterically induced polarization, due to the bent molecular shape, becomes dominating, leading to much higher polarization values, thus providing access to high polarization ferroelectric materials with weakly bent compounds having only “weakly chiral” stereogenic units. Moreover, the field induced alignment of the SmCsPF(*) domains gives rise to a special kind of electroclinic effect appearing even in the absence of molecular chirality. Comparison with related compounds indicates that the strongest effects of chirality appear for weakly bent molecules with a relatively short coherence length of polar order, whereas for smectic phases with long range polar order the effects of the interlayer interfaces can override the chirality effects.
There are presently two distinct types of soft ferroelectrics, representing smectic liquid crystals (LCs).3 One type is formed by chiral molecules organized in layers and assuming a uniform tilt (SmC* phases).4 The other one is provided by so-called bent-core molecules which are also organized in layers.5–8 For these molecules the bent shape restricts the rotation around the long axis, thus producing sterically induced polar order (SmCPF and SmAPF phases) due to the directed packing of the bent molecules with a uniform bending direction.9,10 In this case neither chirality nor tilt is required for the occurrence of polar order and ferroelectricity. For the tilted LC phases of bent-core liquid crystals (BCLCs) there are four distinct structures depending on the tilt direction and the polar direction, designated as SmCsPF, SmCsPA, SmCaPF and SmCaPA and shown in Fig. 1a.11
Fig. 1 (a) The four states of the SmCP phases of bent core molecules11 and the contribution of core packing (Pb) and a combination of molecular chirality and the tilt-direction (Pc, right) to the overall polarization P;15 the molecules are shown in a side view along the polar axis and the color indicates the sign of the superstructural layer chirality; (b) the effect of the distinct switching and reorganization modes of chiral BCLCs on Pb (Pc is fixed by the molecular chirality) and superstructural layer chirality; (A) rotation around the long axis; (B) rotation on the tilt cone and (C) flipping of the tilt direction.29 |
An interesting question concerns the combination of the two distinct approaches to ferroelectricity by introducing a bend into chiral rod-like mesogens or by introducing chirality into bent-core mesogens by molecular design,12–14 or by mixing achiral bent-core mesogens with rod-like chiral compounds.13 For these combined systems the coexistence of two kinds of polarization, one due to the packing of the bent-cores (bent-core polarization: Pb) and the other one resulting from molecular chirality, leading to reduced local symmetry (C2) and coupling with the tilt direction, as known in SmC* phases (chirality polarization: Pc), is in principle possible (see Fig. 1a).15
Besides the molecular chirality, chirality at the superstructural level must also be considered in the polar SmC phases of bent-core mesogens. As the combination of tilt and polar order gives rise to a reduced C2 symmetry, these layers are inherently chiral and their chirality is reversed either by tilt reversal or by reversal of the polar direction, whereas simultaneously reversing both tilt and polar directions, retains the chirality (Fig. 1b).11 Depending on the mode of combination of the polar direction and the tilt direction in adjacent layers the four different phase structures can be divided into racemic structures with alternating chirality (red + blue) in adjacent layers (SmCsPA and SmCaPF) and uniformly chiral structures (SmCsPF and SmCaPA) with identical layer chirality which can exist in two enantiomeric forms (blue or red, only one enantiomer of the chiral structures is shown in Fig. 1a).11 In line with the basic concepts of stereochemistry it was shown that in the chiral synclinic ferroelectric phases or states (SmCsPF*) the sign of the tilt is coupled with the molecular chirality, as only one combination of molecular and superstructural chirality represents the low energy diastereomer (matched case) and the other diastereomer is disfavored (mismatched case).15 However, surprisingly it was also observed that enantiomerically enriched chiral bent-core molecules preferably form synclinic antiferroelectic (SmCsPA*) and anticlinic ferroelectric (SmCaPF*) structures, which represent racemic structures with alternating right and left handed layers (Fig. 1a).16 This was explained by a dominance of the effects of interlayer interfaces9,17 on tilt correlation over the effects of molecular chirality. However, these investigations were based on the comparison of homogeneously chiral molecules, involving branched alkyl chains with achiral molecules comprising linear n-alkyl chains, representing different molecular structures.18–27 As there are significant effects of chain branching on interlayer coupling, mesophase type and mode of switching,9,16,20,22,23 this investigation cannot be fully conclusive and direct comparisons of uniformly chiral compounds with their racemic mixtures would be required. There is only a handful of reports about this kind of comparison28–31 and some of them have only appeared in conference proceedings15 or were only partly published.32 Moreover, previous work in this field was mostly focused on strongly bent molecules (with about 120° angles),16,19,21,23,26,28 forming the typical macroscopic polar “banana” phases composed of polar layers (see Fig. 1a). However, there are also bent-core molecules with reduced bend which, depending on temperature, can change their shape from nearly rod-like to strongly bent and therefore show transitions between non-polar LC phases and polar “banana” phases.33–37 Typical compounds are provided by benzoylated 4-cyanoresorcinols,36–42 having a reduced molecular bend due to the effect of the 4-CN substituent on the conformation of the adjacent COO linking unit.38
An especially interesting case is provided by compound 1.29 In the SmCsPA* phase of the enantiomer (S)-1 (see Fig. 2a) the switching of the field induced SmCsPF* state by rotation around the long axis (process (A) in Fig. 1b) inverts the sign of the layer chirality, whereas the molecular chirality cannot change; this leads to a diastereomeric state which after switching no longer represents the energetic minimum (mismatched case). Under the applied field the energetically preferred diastereomeric state (matched case) is re-established in a relaxation process by tilt-flipping without inversion of the polar direction (see (C) in Fig. 1b and 2b).29
Fig. 2 (a) Chemical structure, LC phases and transition temperatures (T/°C) of the pure enantiomer and the racemic mixture of compound 1;29 values in parenthesis are monotropic phases which can only be observed on cooling; abbreviations: Cr = crystalline solid state; Iso = isotropic liquid state; SmA = non-tilted and nonpolar smectic phase; SmCs = non-polar and synclinic tilted smectic phase; SmCsPA = synclinic tilted and antiferroelectric switching polar SmC phase (see Fig. 1a); the asterisk (*) indicates LC phases formed by uniform chiral molecules, being affected by the molecular chirality (favoring the layer chirality indicated in blue); (b) molecular reorganization during the antiferroelectric switching by rotation around the long axis (A) and relaxation to the thermodynamically more stable (matched) state by tilt flipping (C) under the applied field (see Fig. 1b), as observed for (S)-1.29 |
Here we report a new bent-core molecule 2 based on the same 4-cyanoresorcinol core, but with terephthalate based rod-like wings,43–45 replacing the biphenyls in compound 1, and carrying the same chiral 3,7-dimethyloctyloxy groups46 in the racemic (rac-2) and the uniformly chiral (S,S)-form ((S)-2, see Table 1). Just like compound 1, compound 2 shows a non-polar SmA phase and tilted smectic phases (SmC), but these SmC phases are very distinct from those of the related compound 1 where only one type of antiferroelectric switching SmCsPA(*) phase was observed for the racemate and the enantiomer. In contrast, for compound 2 a sequence of paraelectric and superparaelectric SmC phase ranges is found for the racemate which is replaced by SmC* phases showing distinct ferroelectric modes for the enantiomer. This shows that coupling of molecular chirality with the superstructural layer chirality in SmCsPF* domains stabilizes synpolar correlation which provides access to high polarization ferroelectric materials with weakly bent compounds having only “weakly chiral” stereogenic units. In addition, classical and non-classical electroclinic effects4 were found for enantiomers and racemates which are associated with the field induced alignment of the local domain structure of the smectic phases of 2. Conclusions are drawn concerning the interplay between chirality induced and sterically induced polar order.
Compd | T/°C [ΔH kJ mol−1] |
---|---|
a Transition temperatures as determined by the polarizing microscope under quasi equilibrium conditions and enthalpy values in italics in brackets taken from the 1st heating and cooling scans at a rate of 10 K min−1, see the Experimental section; abbreviations: SmA = uniaxial smectic phase, the prime in SmA′ indicates a de Vries-like SmC phase composed of tilt randomized SmC domains; SmC = paraelectric SmC phase; SmCPR*, SmCPAR = tilted smectic phases composed of SmCsPF domains (in the domains the tilt direction and the polar direction are uniform), with preferred synpolar (SmCPR*) or antipolar (SmCPAR) Ising-like correlation of the domains;57 SmCs′PF* = ferroelectric switching polar smectic phase with a uniform (synclinic) tilt; SmCs′PA = high permittivity paraelectric smectic phase with a synclinic tilt; SmCa′PA = high permittivity paraelectric smectic phase with an anticlinic tilt; SmCa′PA* = ferroelectric switching polar smectic phase with slow relaxation to a SmCa′PA ground state structure; the prime in the phase designation indicates a special domain-like structure of these polar smectic phases which are distinct from the typical B2 phases; for other abbreviations, see Fig. 2. b Melting point of the highest temperature crystalline phase is given and the enthalpy value is the sum of all Cr–Cr transitions and melting events (see Fig. 3). | |
rac-2 | Cr 98 [33.8]b (SmCa′PA 73 [—] SmCs′PA 81 [2.4] SmCPAR 87 [—] SmC 93 [—]) SmA′ 105 [—] SmA 159 [6.8] Iso |
(S)-2 | Cr 106 [46.0]b (SmCa′PA* 73 [—] SmCs′PF* 81 [2.2] SmCPR* 87 [—] SmC* 93 [—] SmA′* 105 [—]) SmA* 159 [6.8] Iso |
Scheme 1 Synthesis of the bent-core compound (S)-2 and its racemic mixture rac-2. Reagents and conditions: (i) (1) NH2OH·HCl, (2) Ac2O, (3) KOH;38 (ii) DCC, DMAP, CH2Cl2. |
Throughout all steps of the synthesis of the enantiomer (S)-2 the stereogenic center of the (S)-3,7-dimethyloctyloxy group was not touched and hence it can be assumed that the final product has approximately equal optical purity as (S)-(−)-β-citronellol (≥99%), used as a starting material.29,31,51rac-2 represents an optically inactive racemic mixture of the (R,R), (S,S), (R,S) and (S,R) stereoisomers. Despite being a complex mixture, the LC phase transition temperatures and the sharpness of the phase transitions are the same as for the enantiomeric pure single component compound (S)-2 (see Fig. 3).
In the cooling scans there are two exotherms for (S)-2 and rac-2, occurring at almost the same temperature, a larger one (ΔH = −6.7 kJ mol−1) for the iso–LC transition at T = 159 °C and a smaller one (ΔH ∼ −2.8 kJ mol−1) for an additional phase transition in the LC range at T = 81 °C. Besides these phase transitions with enthalpy changes, there are four additional continuous transitions in the LC range without visible enthalpy change, occurring at T = 105, 93, 87 and 73 °C. So, in total, six distinct LC ranges can be distinguished. The high temperature phase (SmA(*), for rac-2 also a certain range of SmA′) is enantiotropic, whereas the tilted LC phases (SmC(*) and the randomly tilted SmA′* phases) are monotropic (metastable) and therefore could only be observed on cooling from the SmA(*) phase. Nevertheless, the monotropic phases are long time persistent and thus can be investigated by different methods. In the following sections the discussion is focused on the four smectic phases above T = 73 °C, whereas the low temperature phase (SmCa′PA(*)) below this temperature will be discussed separately in Section 2.3.4.
Fig. 4 Textures of the LC phases of (a–c) rac-2 (left) and (d–f) (S)-2 (right) as observed in a 10 μm polyimide (PI) coated ITO cell (planar alignment) between crossed polarizers (arrows) on cooling at the indicated temperatures; the insets show the homeotropic samples at the same temperatures, as obtained between non-treated glass plates; for additional textures showing the SmA(*)–SmA′(*)–SmC(*) transitions, see Fig. S5 (ESI†). |
XRD investigations of surface aligned samples of rac-2 at T = 130 °C in the SmA phase indicate a sharp small angle reflection at d = 4.34 nm with the corresponding second order reflection having the maximum positioned on the meridian and a diffuse wide angle scattering (d = 0.48 nm) having its maxima on the equator, in line with the proposed SmA structure (Fig. 5a). The observed d-value is only a bit smaller than the molecular length (Lmol = 5.0 nm; d/Lmol = 0.86–0.87), thus indicating a monolayer smectic phase, allowing a possible maximum tilt angle of 30°.
There is no change in the small- and wide-angle scattering until the phase transitions at T = 81 °C, where the layer reflection adopts a crescent-like shape (Fig. 5b and c). This would be in line with a small tilt of about 7°, though from the 2D-XRD patterns the presence of a tilt cannot be proven unambiguously. However, electro-optical investigations of planar samples (see Section 2.3.1) indicate a randomized or uniformly tilted organization in all LC phases below T = 105 °C.
The plot of the d-values of the small angle reflection depending on temperature indicates an increase of the layer spacing in the SmA range between T = 157 and 105 °C, in line with a growing packing density, leading to alkyl chain stretching. An increase of the packing density with decreasing T is supported by the observed increase of the birefringence of the planar fan texture, indicating a growing orientational order (Fig. 4a–c), and by the plot of the maximum of the wide angle scattering against temperature (see Fig. 5d, red dots), showing an almost linear reduction of the lateral intermolecular distances. In the SmA range between 179 and 130 °C the increase of the d-value of the layer reflection (black dots) is nearly linear and the slope starts decreasing at about 130 °C, the layer spacing reaches a maximum at T = ∼105 °C and then decreases. The decrease of the layer spacing could be caused by different reasons, among them an increase of the molecular bend38 and a developing uniform or random tilt of the molecules with respect to the layer normal. There is a continuous and nearly linear decrease of d between 105 and 87 °C across the SmA–SmC transition temperature at T = 93 °C. This could mean that already in the SmA phase a random tilt of the molecules develops and therefore the phase range below T ∼ 105 °C, designated as SmA′, can be considered as a de Vries type SmC phase52 composed of small SmC domains which have a randomized tilt direction and grow in size with decreasing temperature. The tilt correlation remains short range until the continuous transition at T = 93 °C, where the tilt domains reach a critical coherence length, giving rise to phase biaxiality in the SmC phase. The de Vries character of the SmA′ phase range is further supported by the absence of a clear reduction of birefringence in the planar texture across the SmA′–SmC transition (see Fig. S5b and c, ESI†), confirming that the tilt should already be present in the SmA′ phase range, and by the observation of a field induced tilt in the SmA′(*) range (see Section 2.3.2).
Fig. 6 Switching current responses obtained for (a–d) rac-2 and (e–h) (S)-2 under a triangular wave field (10 Hz, 5 kΩ) in a 6 μm ITO cell in the distinct phases at the indicated temperatures; please consider that different scales were used for the P-axis; additional switching curves are collated in Fig. S6 (ESI†) (rac-2, burst and field dependence in SmCs′PA), and Fig. S7–S9 (ESI†) ((S)-2, transition SmC*–SmCPR* and burst in SmCs′PF*) and Fig. S10 (ESI†) ((S)-2 and rac-2 in SmCa′PA*) phases. |
Fig. 8 Development of the spontaneous polarization in the smectic phases of rac-2 and (S)-2 depending on the temperature, for numerical data, see Table S1 (ESI†). |
At this discontinuous phase transition a relatively broad and non-symmetric “single” peak is formed (Fig. 6d) which is positioned close to zero voltage crossing and splits into two if a modified triangular wave voltage with an additional relaxation time at 0 V is applied (Fig. S6b, ESI†).59 Therefore, the single peak does not indicate ferroelectric switching. The switching appears to represent a kind of superparaelectric switching60 where relatively large SmCsPF domains merge by formation of a field-induced macroscopic polar SmCsPF structure and relax back to an apolar polydomain structure around E = 0 (similar to Fig. 7j–l). The polarization reaches a value of Ps = 270 nC cm−2 in the SmCs′PF* range and 300 nC cm−2 in the SmCa′PA* phase before crystallization (Fig. 8), which is only about one third to one half of the typical values observed for B2-like SmCPA phases composed of polar layers (∼600–800 nC cm−2).61 This suggests that none of the SmC phases of rac-2 are formed by uniformly polar layers as known for the B2 phases of the strongly bent type of bent-core mesogens, thus supporting a polar domain structure of these LC phases. For this residual domain structure a prime is added in the phase designation (SmCs′PA).
Optical investigations in planar cells (Fig. 7) provide information about the alignment and field-induced reorganization of the molecules in the smectic phases. In the SmCPAR and SmCs′PA phases the direction of the dark extinctions is inclined with respect to the polarizer direction under an applied E-field and this orientation is retained after removing the applied field (Fig. 7g–i and m–o), indicating that a structure with a long range synclinic tilted organization is present under the electric field and that it is retained even after removal of the E-field. That the position of the extinctions does not change after field reversal means that the reorganization of the molecules takes place by inversion of only the polar direction, i.e. with inversion of the chirality of the field induced SmCsPF states.
Though the field direction does not change the tilt direction, there is a significant effect of the electric field on the tilt angle in the SmA′, SmC and SmCPAR phases. The strongest effect is found in the SmC phase. Despite the phase biaxiality (Fig. 9), the extinctions in planar samples of the SmC phase are parallel to the polarizers in the absence of an applied field (Fig. 9e), which is attributed to the presence of anticlinic packing of the SmC domains. Inclination of the extinction crosses is induced under an applied E-field, rising with growing field strength until a saturation angle of ∼23° is reached (Fig. 9a and e–g), indicating a field induced tilt alignment. We attribute the steep increase of β at low E-field to the removal of any anticlinic packing and the further growth is attributed to the synclinic alignment of the tilt director of the SmCsPF domains in the SmCsPF state developing under the E-field, due to the inherent coupling between the polarization direction and the tilt direction (Fig. 9a). The coexistence of areas with extinction crosses of the opposite direction (Fig. 7 and 9g) indicates that a macroscopic mixture of oppositely tilted, and hence oppositely chiral, SmCsPF areas is formed under the applied field.
Fig. 9 (a) Field strength dependence of the optical tilt angle β of rac-2 in the distinct smectic phases; □ indicates the roughly estimated value from (d); in the SmCs′PA phase the tilt is field-independent and amounts to ∼30° (see Fig. 7m–o); the accuracy of the tilt angle measurements is in the range of ±2°; (b–d) circular domains in the SmA′ phase at (b) E = 0, (c) E = 12 V μm−1 and (d) E = 20 V μm−1 and (e–g) in the SmC phase at (e) E = 0, (f) E = 2 V μm−1 and (g) E = 12 V μm−1; white arrows show the orientation of the polarizers. |
In the SmCPAR phase a synclinic tilt of 12–13° is found already at E = 0 (Fig. 7h), indicating a preference for synclinic SmCsPF domain correlation even in the absence of an E-field and the tilt increases with growing field strength to about 22°, which is approximately the same value as that found in the SmC phase (Fig. 9a). In the SmCs′PA phase the tilt is significantly larger (∼30°) and no field dependence is observed (Fig. 7m–o). No tilt is found in the SmA phase at the highest temperature, though below T = 105 °C in the SmA′ range a tilt of ∼10° is induced at a relatively high field strength of ≥20 V μm−1 (Fig. 9a and d).62 This supports the proposed de Vries like SmC (and field induced SmCsPF) domain structure of the uniaxial smectic phase in the SmA′ range.
Fig. 11 Field strength dependence of the optical tilt of (S)-2 in the distinct smectic phases (average of the amounts of right and left tilt of two measurements); in the SmCs′PF phase the tilt is field-independent and amounts to ∼30° (see Fig. 10p–r); the accuracy of the tilt angle measurements is in the range of ±2°. |
A large difference in birefringence between the field-on and field-off states in the SmCPR* and SmCs′PF* phases of (S)-2 (see Fig. 12) is observed in 10 μm PI-coated ITO cells. This is not found in 6 μm cells (Fig. 10) and also not for the SmCPAR and SmCs′PA phases of the achiral racemate rac-2 (see Fig. 7). Therefore, it is assigned to the development of a chirality induced helical modulation of the tilt direction after removal of the alignment field. In the SmCPR* phase of (S)-2 the position of the extinctions is inclined with respect to the direction of the polarizers in the field induced SmCsPF* state and this is retained at 0 V, though a significant decrease of birefringence is observed (color change from orange to green, see Fig. 12a and b). The birefringence slowly decreases further (pink/blue), and the extinctions become aligned parallel to the polarizers after a few (∼10–20) seconds (see Fig. 12c inset). It appears that immediately after switching off the applied field only a local modulation of the tilt direction takes place, meaning that helix formation in the SmCPR* phase is obviously a slow process. In the SmCs′PF* phase at lower temperature relaxation almost immediately leads to a flipping of the extinctions to positions parallel to the direction of the polarizers (Fig. 12f) and a smooth texture with reduced birefringence (yellow) is formed quickly (Fig. 12g),70 indicating that the driving force for helix formation is stronger.
For the racemate the same transition is found, but in this case it takes place immediately after switching off the field without any delay (Fig. 13f and g). As the same textural changes were found for the enantiomer as well as for the racemate, it is unlikely that helix formation could be mainly responsible. A transition from a synclinic to an anticlinic phase structure is the likely reason for this textural change.71 Accordingly, a relaxation to a non-polar SmCaPA(*) state takes place by rotation on the tilt cone for the racemate as well as for the enantiomer (see Fig. 14d and h), thus retaining a homogeneous layer chirality. Though the tilt correlation in the ground state changes at T = 73 °C, there is no change in the number and shape of the current peaks (Fig. S10, ESI†) and no jump in the Ps values (see Fig. 8), i.e. there is obviously no effect of the changing tilt correlation on the type of switching, being superparaelectric for rac-2 and ferroelectric for (S)-2. Therefore we conclude that this low temperature SmC phase still retains some residual domain character (indicated by the prime in the phase designation), but the polar domains have exceeded a critical size at which the emerging antipolar interlayer interfaces become dominating for the optical properties (see discussion in the Summary and conclusions section). Whereas the transition from the field induced SmCsPF(*) state to the non-polar SmCaPA(*) structure is fast for the racemate, it is slow for (S)-2. Thus it appears that in the case of (S)-2 the polar SmCsPF* state is still stabilized by the chirality–tilt-coupling, supporting synpolar correlation of SmCsPF* domains and making this relaxation slow. On the other hand, in anticlinic phases the chirality–tilt-coupling stabilizes the antipolar order as known for the SmCA* phases of chiral rod-like LCs.4 The SmCa′PA* phase of (S)-2 appears to be just at the cross-over from synpolar to antipolar organization. Chirality–tilt-coupling is still stabilizing the field induced SmCsPF* state, thus retaining ferroelectric switching and delaying the relaxation process, whereas at E = 0 the SmCaPA* structure, governed by the emerging anticlinic interlayer interfaces, becomes dominating. Despite the ferroelectric switching of the field-induced SmCsPF* states we choose the phase designation SmCa′PA*, describing the most stable ground state structure.
There are significant differences for the polar low temperature SmC(*) phases following the SmC(*) phase. The low temperature phases of rac-1 and (S)-1 both show antiferroelectric switching indicated by two sharp and well-separated polarization current peaks per half period of the applied triangular voltage and having a high Ps value (Ps = 700–800 nC cm−2) as typically observed for bent-core mesogens and there is no effect of molecular chirality on the switching mechanism (Fig. 15b and d). Optical investigations indicated that the field induced polar SmCsPF(*) states relax to a racemic apolar SmCsPA(*) structure by rotation around the long axis (A), which reverses the superstructural chirality (Fig. 15b and d). This means that the low temperature smectic phase behaves like a typical B2 phase of a bent-core mesogen. For (S)-1 where the molecular chirality is uniform, this switching leads to two energetically different diastereomeric field-induced SmCsPF* states, depending on the direction of the applied electric field. The less stable SmCsPF* state (mismatched case) slowly relaxes under the applied field, where tilt flipping with constant polar direction (process (C) in Fig. 1b) leads to the formation of the more stable diastereomeric state (matched case, see Fig. 2b). This kind of slow reorganization process cannot be observed for (S)-2, because the switching takes place by rotation on a cone which retains the superstructural layer chirality once developed, and thus the field induced state always represents the energy minimum diastereomer. This indicates that compared to compound 2 the coupling between molecular and superstructural chirality is much weaker for compound 1 and thus in this case chirality can neither induce ferroelectric switching in the polar smectic phases nor it can change the mode of switching from a rotation around the long axis to the chirality retaining rotation on a tilt-cone. As another difference, there is an abrupt transition from the behavior as a typical rod-like molecule to typical bent-core behavior for compound 1 whereas for compound 2 this transition is smooth and a true B2-like phase is never achieved. The very distinct behavior of compounds 1 and 2 must be due to the change in the structure and the length of the rod-like wings by introduction of the COO linking unit. The rigid biphenyl wings in compound 1 obviously lead to a pronounced bent molecular shape at reduced temperatures, being responsible for polar packing, hence, in this case polar order is dominated by Pb. The terephthalate wings in compound 2, being more flexible, provide improved attractive intermolecular interactions and, due to the increased molecular length, probably also an increased nematic order parameter, which do not lead to an abrupt change from non-polar to polar smectic phases. For this compound the balance between the molecular shape and the intermolecular interactions obviously contributes to a continuous development of polar order with decreasing temperature. In this case diastereomeric relationships have a significant impact on the polar packing of the bent cores, i.e. molecular chirality leads to ferroelectricity and rotation on a cone.
Fig. 15 Summary of phase structures, shapes of polarization curves and modes of electric field induced switching in the distinct SmC phases of compound 1.29 In principle, the same scheme can be applied for the SmC phases of compound 3.28 |
The bent-core compound 3, reported by Sadashiva et al.,28 behaves very similar to compound 1 (see Table 2). The high temperature smectic phases (SmCs, SmCa) of the racemate rac-3 are non-polar and the enantiomer (S)-3 shows chirality induced ferroelectric switching on a cone in the SmCγ* and SmC* phases (see Fig. 15a and c). The low temperature phase has a racemic SmCsPA(*) structure, in this case with additional in-plane modulation (SmsPA(*)). There is Pb-based antiferroelectric switching around the long axis with two well separated peaks and high polarization values (see Fig. 15b and d). Only the relaxation process by flipping of the layer chirality (process (C) in Fig. 1b, see also Fig. 2b), found for (S)-1, was not reported for (S)-3, probably it is inhibited by the layer modulations in this case.
T/°C [ΔH kJ mol−1] | |
---|---|
a SmCγ* = ferrielectric switching tilted smectic phase; SmsPA(*) = modulated synclinic tilted and antiferroelectric switching smectic phase (ColobPA(*) phase); X(*) = non-specified biaxial smectic phase. | |
rac-3 | Cr 114 SmsPA 118 SmCs 142 SmCa 148 X 148.5 Iso |
(R)-3 | Cr 115 SmsPA* 118 SmCγ* 141 SmC* 147 X* 148 Iso |
Finally we would like to include the bent-core compound (S)-4,16 with a strongly bent core involving Schiff base units, reported by Takezoe et al., in the discussion. This compound and related compounds were only synthesized as uniformly chiral (S,S)-enantiomers and therefore no direct comparison with the racemate is possible in this case. For (S)-4 different field-induced and ground state structures coexist and ferroelectric switching takes place by rotation on a cone with high polarization values. The racemic SmCaPF* state was identified as one of the stable ground state structures18 and under an applied field SmCsPF* states with either handedness coexist with the racemic SmCaPF* state.16 It appears that in this case there is no coupling between molecular chirality and layer chirality. If ferroelectric switching is induced by the molecular chirality or by the modification of the interlayer interfaces due to the presence of chain branching cannot be decided, as the behavior of the related racemate is unknown in this case.
Overall it appears that the coupling between molecular chirality and supramolecular chirality in polar SmC(*) phases increases in the order 4 < 3 < 1 < 2. In the same order the molecular shape changes from a typical bent-core mesogen (compound 4) to a weakly bent molecule (compound 2) and the contribution of Pb decreases. This molecular structural change is associated with a transition form long range polar order (polar layers) to local polar order in domains, a reduction of the tilt, a reduction of the importance of anticlinic tilt correlation and a change in the switching from rotation on a tilt cone to the rotation around the molecular long axis.72
Below 105 °C there is a tilt in all LC phases, only the coherence length of the tilt is different, being short range in the SmA′(*) phase and long range in the distinct SmC(*) phases. At high temperature this compound behaves like a rod-like molecule with its properties dominated by the polarization induced by molecular chirality (Pc), leading to ferroelectric switching in the SmC* phase and a significant electroclinic effect in the SmA′* phase of enantiomer (S)-2. At lower temperature the rotation around the molecular long axis becomes restricted, giving rise to the simultaneous development of sterically induced polar order (Pb) and superstructural layer chirality. These smectic phases are considered as being composed of small SmCsPF domains preferring an antipolar correlation which leads to the macroscopically polarization randomized SmCPAR and SmCs′PA phases for rac-2.57 The field induced alignment of these SmCsPF domains leads to nonclassical electroclinic effects in the SmA′, SmC(*) and SmCP(A)R(*) phases of rac-2 and (S)-2.63 For rac-2 the tilt direction is retained by reversing the applied field which is attributed to surface alignment effects (SmC) and rotation around the long axis (SmCs′PA and SmCPAR).
In the LC phases of the uniformly chiral compound (S)-2 additional diastereomeric relationships between the fixed molecular and the transiential superstructural chirality of the SmCsPF domains lead to energetically different pairs with the more stable being preferred, thus fixing just one specific combination of the tilt direction and the polar direction. For this reason the tilt is uniform under an applied E-field and it is inverted by field reversal, thus retaining the once developed energetically favored diastereomeric pair by reorganization via rotation on the tilt cone. Because a uniform synclinic tilt is present, a uniform direction of polarization is inevitably stabilized due to the coupling of the tilt direction and the polar direction, thus leading to ferroelectric and ferroelectric-like switching in the SmCs′PF* and SmCPR* phases of (S)-2, respectively. This coupling of Pb with molecular chirality provides new access to high polarization ferroelectric LC materials by utilizing molecules combining a weak molecular chirality with a weakly bent core structure. The polarization is almost the same in rac-2 and (S)-2 which shows that polarization is determined by Pb and means that the domain size and the magnitude of polarization are not affected by molecular chirality, it only removes the antipolar domain correlation. Investigation of the low temperature SmCa′PA* phase of compound 2 and a comparison of the phase sequence of this compound with related chiral bent-core molecules (compounds 1, 3 and 4) indicate that a further increase in the molecular bend and growing polar order (increasing contribution of Pb) obviously favour anticlinic and antipolar SmC phases and reduce the influence of molecular chirality. This means that only for weakly bent molecules can the molecular chirality support synpolar order. The reason might be that for rod-like or weakly bent molecules synclinic tilt is inherently preferred by interlayer fluctuations (providing an entropic gain). However, these fluctuations favor an antipolar packing of the molecules in adjacent layers if a molecular bend emerges and increases (Fig. 16).71 For strongly bent molecules the effects of the competing interlayer interfaces become dominating, obviously overriding the effects of molecular chirality on tilt and polarization and thus leading to a preference for antiferroelectric phases, even for the enantiomers.
Fig. 16 Effects of out-of-plane fluctuations (green arrows) on layer coupling depending on the molecular bend. |
Overall this work provides clues concerning the relationships between molecular chirality induced polar order and sterically induced polar order, as well as leading to a new approach to high polarization ferroelectric LCs. Moreover, it supports the SmCsPF domain model of various paraelectric and polarization randomized LC phases occurring at the cross-over between rod-like and bent molecular shapes.
Differential scanning calorimetry (Perkin-Elmer DSC-7) was performed in 30 μl-pans for 2–5 mg samples with heating and cooling rates of 10 K min−1; peak temperatures for the first heating and cooling scans are given in Table 1. The assignment of the mesophases was made on the basis of the combined results of optical textures, DSC and X-ray diffraction (XRD).
XRD measurements were done at Cu-Kα line (μ = 1.54 Å) using a standard Coolidge tube source with a Ni-filter. Investigations of oriented samples were performed using a 2D-detector (HI-Star, Siemens or Vantec 500, Bruker Siemens AG). Alignment was achieved by slow cooling (0.1 K min−1) of a small droplet on a glass surface; the beam was applied parallel to the surface. The sample to detector distance was 8.8 cm and 26.9 cm for the wide angle and small angle measurements, respectively, and the exposure time was 30 min. Powder samples were investigated in thin capillaries (∅ = 1 mm).
Switching experiments and electro-optical investigations were performed in a 10 μm polyimide (PI) coated ITO cell or 6 μm non-coated ITO cell (EHC, Japan) with a measuring area of 1 cm2. The cells were filled in the isotropic state. Switching experiments were carried out with the triangular-wave method73 using a combination of a function synthesizer (Agilent, model 33220A, load was set to 10 kΩ), an amplifier (FLC electronics, model A400), and the current response traces were recorded using an oscilloscope (Tektronix, model TDS2014) across a 5 kΩ resistance.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6sm00960c |
This journal is © The Royal Society of Chemistry 2016 |