Remarkable stabilisation of the intercalated smectic phases of nonsymmetric dimers by tert-butyl groups

ABSTRACT

The synthesis and characterisation of two groups of nonsymmetric dimers, the 1-(4-cyanobiphenyl-4‘−yloxy)-ω-(4-butyl, 4-(1-methylpropyl) or 4-t-butylanilinebenzylidene-4’−oxy)alkanes, and the 1-(4-cyanobiphenyl-4‘−yl)-ω-(4-butyl, 4-(1-methylpropyl) or 4-t-butylanilinebenzylidene-4’−oxy)alkanes, are reported. The length and parity of the flexible spacer is varied. The tert-butyl homologues show higher melting points than the corresponding sec-butyl or n-butyl substituted dimers, suggesting that chain branching improves packing efficiency within the crystalline structure. The branched chain homologues have a stronger tendency to exhibit smectic phases than the n-butyl-substituted dimers, and for longer spacers are exclusively smectic. A comparison of the nematic-isotropic transition temperatures (T_NI) for dimers containing the different terminal chains is possible for one set of materials, and reveals a large reduction in T_NI on passing from the n-butyl to sec-butyl-substituted, but a much smaller decrease on changing sec-butyl for tert-butyl. A different trend is observed for the smectic A-isotropic transition temperatures for which the tert-butyl substituted dimers show a higher value than the corresponding sec-butyl homologue, and only marginally lower than that of the n-butyl-substituted dimer. This surprising behaviour is interpreted in terms of the ability of the tert-butyl group to pack more efficiently into the intercalated smectic A phase as the spacer length increases.

GRAPHICAL ABSTRACT

Keywords:

• Intercalation
• branched chain
• smectic phases
• twist-bend

Introduction

Liquid crystalline dimers consist of molecules containing two mesogenic units linked via a flexible spacer, most commonly an alkyl chain [1,2]. In recent years, there has been a dramatic resurgence of interest in this class of materials arising first from the observation of the twist-bend nematic phase for dimers containing an odd-membered spacer (see, e.g. [3–25]), and even more recently from the discovery of the twist-bend smectic phases [26–28]. Prior to this, liquid crystal dimers had been the focus of considerable research interest following the discovery of the intercalated smectic phases for nonsymmetric dimers, in which the two mesogenic units differ in structure [29–31]. The intercalated smectic phases were first observed for members of the 1-(4-cyanobiphenyl-4’−yloxy)-ω-(4-alkylanilinebenzylidene-4’-oxy)alkanes [29].

The acronym used to refer to this family of dimers is CBOnO.m where CB stands for the cyanobiphenyl unit, O for the ether linkages, n for the number of methylene units in the spacer, “.‘’. for the benzylideneaniline moiety and m is the number of carbon atoms in the terminal alkyl chain. The discovery of the intercalated smectic phases arose from the observation of unprecedented smectic behaviour on varying the terminal chain length in the CBO6O.m series [29]. Specifically, the first eleven members (m = 0–10) were found to show an enantiotropic nematic phase, and in addition, smectic phases were observed for m = 0–6 and 10. The smectic A – nematic transition temperature, T_SmAN, increased over the first three members and subsequently fell for m = 3–6. No smectic behaviour was observed for m = 7–9 but reappeared for m = 10, which showed the highest value of T_SmAN of the series. This behaviour was hugely surprising, and in marked contrast to that seen for conventional low molar mass liquid crystals comprising molecules containing a single mesogenic unit for which T_SmAN normally increases on increasing the length of a terminal alkyl chain. This novel behaviour was accounted for in terms of how the structure of the SmA phase changed as m was varied for the CBO6O.m series. For early members of the series the ratio of the layer thickness, d, to the molecular length, l, d/l was about 0.5 whereas for m = 10, d/l = 1.8. The latter value indicated an interdigitated arrangement of the molecules stabilised by the electrostatic interaction between the polar and polarizable cyanobiphenyl units, and the smectic phase resulted from the molecular inhomogeneity arising from the long terminal alkyl chains, see Figure 1. For the early members of the series, d/l ≈ 0.5 was interpreted in terms of the formation of an intercalated smectic A phase, in which differing parts of the molecule overlap, see Figure 1. The possibility that the dimers formed horseshoe conformations and hence, d/l ≈ 0.5 was ruled out as being energetically unfavourable. The driving force for the formation of the intercalated arrangement has been attributed to a specific favourable interaction between the unlike mesogenic units, suggested to be an electrostatic quadrupolar interaction between groups having quadrupole moments of opposite signs [32]. The ability to accommodate the terminal chain in the intercalated arrangement is determined by the length of the spacer. The absence of smectic behaviour for m = 7–9 reflects a competition between two incompatible structures, neither of which wins, and hence, the nematic phase is observed rather than the monolayer smectic A phase (Figure 1). The study of the CBOnO.m family of dimers not only revealed the intercalated smectic A phase but also the intercalated smectic C and I phases, and the intercalated crystal B and J phases [31]. Moreover, the intercalated smectic phases have been observed not only for dimers [33] but also for higher oligomers including trimers [34,35] and tetramers [36]. It should also be noted that the tendency of nonsymmetric dimers to form intercalated packing arrangements may play an important role in the formation of the twist-bend nematic phase [37–47].

Figure 1. Sketches of the molecular organisation within (a) the interdigitated, (b) the intercalated and (c) the monolayer smectic A phases composed of nonsymmetric liquid crystal dimers: d denotes the layer spacing, and l the molecular length.

The proposed structure of the intercalated smectic phase in which unlike molecular fragments overlap has now been widely accepted in terms of accounting for d/l ratios of much less than one for smectic phases exhibited by dimers and higher oligomers [48]. There remain, however, a number of unresolved issues with the intercalated structure and the nature of the driving force for its formation. A notable example is CBO12O.2, which exhibits an intercalated smectic A phase although the significant mismatch in length between the spacer and terminal chain would suggest that the structure will contain a considerable concentration of voids [31]. It has also been shown that mixtures of symmetric dimers such as 2.O2O.2 and CBO10OCB exhibit induced intercalated smectic A phases [31]. It is very difficult to visualise how these symmetric dimers containing very different spacer lengths can pack efficiently into an intercalated arrangement while maximising the interactions between the unlike mesogenic units. In order to investigate the nature of these intercalated phases further, and in particular to address the question of how do dissimilar fragments pack within the intercalated structure, here we report the synthesis and characterisation of two groups of nonsymmetric dimers in which the length of the spacer and steric bulk of the terminal chain are varied, see Figure 2. In both groups of materials, the terminal substituent is either a butyl chain (X = 4), or the isomeric sec-butyl (X=s4) and tert-butyl groups (X=t4). In the CBOnO.X dimers, the spacer has been varied to consider both the effects of length and parity on phase behaviour (Figure 2), and these are directly comparable to the CBOnO.m dimers published previously [29,31].

Figure 2. The nonsymmetric dimers reported here and the acronyms used to refer to them.

In the second set of materials discussed, CBnO.X, the spacer is now linked to the cyanobiphenyl unit using a methylene group and for even values of n, this endows the molecular curvature required for twist-bend liquid crystal phases to be observed as shown by the CB6O.m [47] and CB6O.Om [45] series. The branched materials will allow us to study how the steric bulk of the terminal chain affects the stability of such phases as well as the intercalated smectic phases.

Experimental

he synthesis of the diether-linked CBOnO.X dimer series followed the steps outlined in Scheme 1. The 4-{4-[(ω-bromoalkyl)oxy]phenyl}benzonitriles (1) were prepared according to a procedure described by Crivello et al.,[49] and the aldehyde (2) was formed by a Williamson ether synthesis. This was subsequently combined with the relevant n, sec- or tert-butyl aniline to form the desired Schiff’s base (3).

Scheme 1. Synthesis of the CBOnO.X dimers, where n = 4, 5, 6, 11 or 12 and xX = n-, sec- or tert-butyl.

The synthesis of the CBnO.X series followed the procedure outlined in Scheme 2. Noncommercial acid chlorides (4) were prepared by reacting the corresponding bromoalkanoic acid with thionyl chloride. Compounds (5)–(8) were prepared as reported previously [49]. The 1-(4-cyanobiphenyl-4’−yl)-ω-(4-formylphenyl-4’−oxy)alkanes (n = 5, 6, 8, 10) (8) were then combined with n-, sec- or tert-butylaniline to form the desired Schiff base (9).

Scheme 2. Synthesis of the CBnO.X dimers, where n = 5, 6, 8 or 10, and X = n-,sec- or tert-butyl.

For full synthetic procedures and analytical data, see ESI.

Results

CBOnO.X series

The transitional properties of the CBOnO.X dimers are listed in Table 1. Conventional nematic phases were assigned on the basis of the observation of characteristic schlieren textures when viewed through the polarised light optical microscope containing two- and four-point brush defects and which flashed on the application of mechanical stress, see Figure 3. CBO4O.4 shows an enantiotropic nematic phase and on extensive supercooling focal conic fans developed in coexistence with regions of homeotropic alignment indicative of a smectic A phase, see Figure 3. The transition temperatures for CBO4O.4 are in excellent agreement with those published previously [31]. The X-ray diffraction pattern of the nematic phase shown by CBO4O.4 shows diffuse wide- and small-angle signals with the latter centred at about 19.1 Å. This corresponds to approximately half the molecular length and indicates a locally intercalated arrangement of the molecules. The monotropic nature of the SmA phase precluded its characterisation using X-ray diffraction, although it seems wholly reasonable to assume that this will have an intercalated structure. CBO4O.s4 exhibited only an enantiotropic nematic phase whereas CBO4O.t4 did not exhibit liquid crystalline behaviour.

Figure 3. (Colour online) The textures obtained using the polarised light optical microscope for CBO4O.4 in (left) the nematic, and (right) the smectic A phase.

Table 1. The transition temperatures and associated scaled entropy changes, ∆S/R, for the CBOnO.X series.

Dimer	TCr-/°C (∆S/R)	TEHexB/°C TJSmI/°C¶ (∆S/R)	THexBSmA/°C TSmISmC/°C§ (∆S/R)	TSmCSmA/°C (∆S/R)	TSmAN/°C TSmX2N/°C† (∆S/R)	TNTBN/°C (∆S/R)	TNI/°C TSmAI/°C‡ (∆S/R)
CBO4O.4	148 (10.09)				98^a-		224 (1.45)
CBO4O.s4	181 (10.27)						188 (0.74)
CBO4O.t4	227 (13.98)
CBO5O.4	108 (11.37)					73^b (0.004)	150 (0.50)
CBO5O.s4	136 (13.45)				72^†,a-		94 ^b (0.11)
CBO5O.t4	168 (11.64)
CBO6O.4	111 (11.11)	89 ^b (1.19)	99 ^b (0.92)		139 (0.22)		199 (1.63)
CBO6O.s4	158 (9.05)				144 ^b (1.03)		166 (1.05)
CBO6O.t4	189 (12.90)						172^‡,a
CBO11O.4	95 (13.84)	66¶ (0.62)	98§ (0.91)	112 (0.03)	142 (0.79)		152 (0.88)
CBO11O.s4	91 (11.21)	63^¶,b (0.92)	78^§,b (0.66)	109^a-			120^‡,b (2.70)
CBO11O.t4	141 (12.40)	73^¶,a-	80^§,a-	120^b (0.02)			132^‡,b (3.32)
CBO12O.4	112 (10.43)		112 (1.03)				159‡ (3.48)
CBO12O.s4	136 (14.34)						142‡ (4.23)
CBO12O.t4	159 (15.15)						148‡ (4.46)

^aTemperature obtained using POM.

^bData obtained from DSC cooling trace.

CBO5O.4 also showed an enantiotropic nematic phase and on cooling a twist-bend nematic, N_TB, phase [47], and the transition temperatures are in excellent agreement with literature values, although the N_TB phase had previously been assigned as a SmA phase [31]. CBO5O.s4 exhibited a monotropic nematic phase. On cooling this, a jagged focal conic fan texture was seen fleetingly prior to crystallisation, and this unknown smectic phase has been denoted SmX₂, see Figure 4. The monotropic nature of this phase precluded its further study using X-ray diffraction. CBO5O.t4 did not exhibit liquid crystalline behaviour.

Figure 4. (Colour online) The textures obtained using the polarised optical microscope for CBO5O.s4 in (left) the nematic, and (right) the smectic X₂ phase.

CBO6O.4 shows a sequence of four liquid crystal phases: nematic, smectic A, hexatic smectic B (HexB) and soft crystal E. The smectic A phase was identified on the basis of the observation of coexisting regions of focal-conic fans and homeotropic alignment. On cooling into the HexB phase, the fans became smoother at the transition and less well-defined, and at the transition to the E phase a scratch-like pattern developed across their backs, see Figure 5. In the nematic phase, the X-ray pattern consists of only diffuse wide- and small-angle signals reflecting the liquid-like ordering of the molecules. The small-angle signal in the nematic phase is very diffuse and centered at around 18 Å. The small-angle signal becomes sharp in the SmA phase while the wide-angle peak remains broad indicating a liquid-like arrangement of the molecules within the layers, see Figure 5. The layer spacing, d, in the smectic A phase is 18.0 Å and approximately half the molecular length, l, of CBO6O.4 estimated to be 33.4 Å, indicating an intercalated arrangement of the molecules. The increased in-plane order of the HexB phase is evident in the narrowing of the wide-angle signal in the X-ray diffraction pattern; the presence of several sharp signals in the X-ray pattern of the E phase indicates a highly ordered structure consistent with the phase assignment, see Figure 5. The layer spacings measured in the HexB and E phases are 18.0 Å and 18.3 Å, respectively, with associated d/l ratios of 0.54 and 0.55, respectively, showing that the intercalated arrangement of the molecules persists in these higher ordered phases. The monotropic E and HexB phases were not identified in the previous study, but otherwise the transition temperatures are in excellent agreement [29]. CBO6O.s4 exhibits only N and SmA phases, while CBO6O.t4 cools directly from the isotropic phase into the SmA phase.

Figure 5. (Colour online) The textures obtained using the polarised light optical microscope and the corresponding X-ray diffraction patterns for CBO6O.4 in (from left to right) the nematic, the smectic A, hexatic B and soft crystal E phases.

CBO11O.4 shows a rich smectic polymorphism, having the phase sequence on cooling: I-N-SmA-SmC-SmI-J-Cr. A characteristic schlieren texture was observed in the nematic phase, which on cooling became largely homeotropic with small regions of focal conic fans indicating a SmA phase, see Figure 6. The X-ray pattern of the smectic A phase contained a sharp low angle signal and a broad wide-angle signal. On cooling, the homeotropic regions developed a weakly birefringent schlieren texture characteristic of a SmC phase, and the X-ray diffraction pattern consisted of a sharp low angle and diffuse wide-angle signal reflecting the liquid-like arrangement of the molecules within the smectic layers. On cooling the smectic C phase, the schlieren texture becomes notably brighter and somewhat less well-defined, see Figure 6. The wide-angle reflection in the X-ray diffraction pattern narrows and splits, indicating a tilted hexatic phase. For an aligned sample, the pattern shows one of the wide-angle peaks is in an equatorial position with respect to the small-angle signals, indicating that the molecules are tilted towards nearest neighbours in the local in-plane hexagons, and hence the phase has been assigned as a SmI phase. On cooling the SmI phase, scratch-like defects appear on the texture and the birefringence changes. These changes are accompanied by a splitting of the wide-angle signal into a number of sharp peaks in the X-ray diffraction pattern suggesting this to be a crystal J phase. The layer spacings and the d/l ratios measured in each of the phases are listed in Table 2, and in each an intercalated arrangement of the molecules is evident. We also note that the position of the diffuse signal in the low angle region for the nematic phase is centred at around 21 Å suggesting a locally intercalated arrangement of the molecules. CBO11O.s4 and CBO11O.t4 show the same sequence of smectic phases but do not exhibit a nematic phase and instead show SmA-I isotropic transitions. The phase assignments were based on the observation of similar optical textures and X-ray diffraction patterns as described for CBO11O.4. The layer thicknesses measured in the SmA phase shown by CBO11.s4 and CBO11O.t4 are also listed in Table 2 and these reveal intercalated arrangements.

Figure 6. (Colour online) The textures obtained using the polarised optical microscope and the corresponding X-ray diffraction patterns for CBO11O.4 in (a) the nematic, (b) the smectic A, (c) smectic C, (d) smectic I and (e) soft crystal J phases.

Table 2. Layer spacings, d, and ratio of layer spacing to molecular length, l, d/l, for the CBO11O.X and CBO12O.X series.

Dimer	Phase	d/Å	d/l
CBO11O.4	SmA	21.0	0.58
CBO11O.4	SmC	20.8	0.58
CBO11O.4	SmI	21.2	0.59
CBO11O.4	J	21.1	0.59
CBO11O.s4	SmA	20.6	0.60
CBO11O.t4	SmA	20.7	0.63
CBO12O.4	SmA	21.9	0.55
CBO12O.4	HexB	22.2	0.56
CBO12O.s4	SmA	21.5	0.56
CBO12O.t4	SmA	21.5	0.59

CBO12O.4, CBO12O.s4 and CBO12O.t4 exhibit SmA-I transitions, and CBO12O.4 also shows the hexatic B phase. These phases were identified on the basis of optical textures and the associated X-ray diffraction patterns, see Figure S1, and as described earlier for other homologues. The layer spacings for each CBO12O.X dimer measured using X-ray diffraction are listed in Table 2, and again reveal in each case an intercalated arrangement of the molecules.

CBnO.X series

The transitional properties of the CBnO.X series are listed in Table 3. CB5O.4 shows an enantiotropic nematic phase and two monotropic smectic phases, see Figure 7. The X-ray pattern of the nematic phase consisted of diffuse peaks in the wide- and small-angle regions. On cooling the uniform texture of the nematic phase, a focal conic fan texture appears and the small-angle diffraction signal narrows whereas the wide-angle peak remains diffuse. This is indicative of a transition to the smectic A phase, and the diffraction patterns suggest an intercalated structure. On further cooling, the fans become rather blocky and truncated, and the texture is smoother. This is not a characteristic texture for the purposes of phase identification, and crystallisation precluded the study of the phase using X-ray diffraction. The phase has been arbitrarily designated the SmX₁ phase. CB5O.s4 also exhibits an enantiotropic nematic phase and on cooling, forms the SmX₁ phase seen for CB5O.4. This was confirmed on the basis of a phase diagram constructed using binary mixtures of the two dimers that revealed complete miscibility across the complete composition range, see Figure S2. CB5O.t4 exhibits only a monotropic nematic phase.

Figure 7. (Colour online) The textures obtained using the polarised optical microscope and the corresponding X-ray diffraction patterns for CB5O.4 in (a) the nematic, (b) the smectic A and (c) smectic X₁ phases.

Table 3. The transition temperatures and associated scaled entropy changes, ∆S/R, for the CBnO.X series.

Dimer	TCr-/°C (∆S/R)	TX4-/°C¶ (∆S/R)	TSmX5- /°C	TSmX3-/°C (∆S/R)	TSmC-/°C† (∆S/R)	TSmX1SmA /°C (∆S/R)	TSmAN/°C (∆S/R)	TNTBN/°C (∆S/R)	TNI/°C TSmCI/°C‡ (∆S/R)
CB5O.4	123 (4.95)					80^a-	99^a-		204 (1.74)
CB5O.s4	143 (6.37)					109^a-			167 (1.30)
CB5O.t4	204 (12.90)								160^a
CB6O.4	103 (9.11)							75^b (0.14)	115 (0.32)
CB6O.s4	104 (29.08)
CB6O.t4	154 (23.96)
CB8O.4	84 (15.47)	25^b (0.56)		70^b (0.73)	85^b (1.28)			87^a-	127 (0.39)
CB8O.s4	102 (15.52)	58^b (1.46)		64^b (1.01)					82^‡,b (2.80)
CB8O.t4	130 (8.99)
CB10O.4	86 (14.09)	28^b (0.58)		75^b (0.99)	101^b (1.28)				129 (0.70)
CB10O.s4	83 (5.72)	36^b (1.41)^c	63^b	67^b (0.92)					94‡ (2.64)
CB10O.t4	128 (6.17)								105^‡,a-

^aTemperature obtained using POM.

^bData obtained from DSC cooling trace.

^cCombined value for SmX₄-SmX₅-SmX₃ transition.

CB6O.4 exhibits an enantiotropic nematic phase and a monotropic twist-bend nematic phase, and the local periodicity in both, as determined from the position of the small-angle diffuse signal in their X-ray diffraction patterns, suggested a locally intercalated arrangement of the molecules [47]. Neither CB6O.s4 or CB6O.t4 exhibited liquid crystalline behaviour.

CB8O.4 shows five liquid crystalline phases, see Figure 8. On cooling from the isotropic liquid, a nematic phase develops which transforms on further cooling into a twist-bend nematic phase. At this transition there is a cessation of the optical flickering associated with director fluctuations in the nematic phase, and the schlieren texture becomes blocky, see Figure 8. This texture changes to give a focal conic fan texture, which may be sheared to give a schlieren texture containing predominantly two-brush point singularities suggesting that this an anticlinic SmC_A phase. On cooling, this texture becomes brighter and smoother with the defect lines blurring, and at a lower temperature, scratches develop across the texture, see Figure 8. Similar changes were observed for members of the CT6O.m series [36], and by analogy we have designated the higher temperature phase as a hexatic SmX₃ phase and the lower temperature phase as a soft-crystal X₄ phase. Nematic behaviour is extinguished for CB8O.s4 and instead this compound undergoes a direct SmC_A-I transition. The schlieren texture is particularly temperature dependent suggesting strong changes of the tilt angle. These cease abruptly at the transition to the lower temperature phase, the schlieren texture blurs and regions of mosaic-like pattern develop. On further cooling, fracture-like defects develop and there is a subtle change in birefringence, see Figure S3. We have tentatively assigned these two lower temperature smectic phases to be the same as those shown by CB8O.4, i.e. SmX₃ and X₄, respectively. CB8O.t4 does not exhibit liquid crystallinity.

Figure 8. (Colour online) The textures obtained using the polarised optical microscope for CB8O.4 in (from left to right) the nematic, the twist-bend nematic, the smectic C, the smectic X₃ and the soft crystal X₄ phases.

CB10O.4 shows very similar phase behaviour to that described for CB8O.4, except that the stability of the SmC_A phase has increased such that a SmC_A-N transition is observed, see Figure S4 and the N_TB phase has been extinguished. CB10O.s4, similarly as CB8O.s4, shows a I-SmC_A transition, and on cooling three additional smectic phases, see Figure S5. On cooling the SmC_A phase, a texture similar to that described for the SmX₃ phase shown by CB8O.4 is obtained, see Figure S4. On cooling this, a texture consisting of partially transparent plates is obtained which persists until the formation of scratch-like defects resembling that seen for the SmX₄ phase described earlier for CB8O.4. We suggest that CB10O.s4 exhibits an additional phase, SmX₅, intermediate between the SmX₃ and SmX₄ phases. CB10O.t4 exhibits a monotropic SmC_A phase and further characterisation of this phase is precluded by the onset of rapid crystallisation.

Discussion

In all nine groups of materials listed in Tables 1 and 3, the tert-butyl homologues show the highest melting point, and in all but two, the butyl homologue has the lowest melting point. For the CBO11O.X and CB10O.X series, the melting points of the sec-butyl members are slightly lower than those of the corresponding butyl homologues. This implies that, in general, branching the terminal chain in these nonsymmetric dimers improves their packing efficiency within the crystalline structure and this is a theme we will return to later.

All the dimers containing a butyl terminal chain show a nematic-isotropic transition with the single exception of CBO12O.4, for which a smectic A-isotropic transition is observed. By comparison, only four of the sec-butyl substituted homologues show a nematic phase, namely, CB5O.s4, CBO4O.s4, CBO5O.s4 and CBO6O.s4, and of the tert-butyl substituted dimers, just CB5O.t4 exhibits a nematic phase. It is important to note that nematic behaviour is only observed for the branched chain homologues containing shorter spacers, and on increasing spacer chain length the branched terminal chain homologues become exclusively smectic in nature. The average reduction in the value of T_NI on exchanging a butyl chain for a sec-butyl one is around 40 K. In the only set of materials for which a comparison is possible, CB5O.X, the reduction in T_NI on replacing a sec-butyl by and tert-butyl group is 7 K. This reduction is considerably smaller than the decrease in T_NI of 37 K between CB5O.4 and CB5O.s4. Twist-bend nematic behaviour is not observed for any of these dimers containing a branched terminal chain, and this is in accord with the view that chain branching tends to destabilise the N_TB phase [46]. A very different pattern of behaviour emerges if we now compare the Sm-I transitions. The only group for which all three dimers exhibit a SmA-I transition are the CBO12O.X dimers and the reduction in T_SmAI on moving from the butyl to sec-butyl chain is 17 K whereas replacing the sec-butyl chain by the tert-butyl chain actually increases T_SmAI by 6 K. Thus, the difference in T_SmAI between CBO12O.4 and CBO12O.t4 is only 9 K compared to the 44 K decrease in T_NI moving from CB5O.4 to CB5O.t4. Indeed, in each group of materials for which a smectic-isotropic transition is observed for both the sec- and tert-butyl substituted dimers, the latter show the higher clearing temperatures, on average by 9 K.

Conventional wisdom accounts for the reduction in the clearing temperature on branching a terminal chain in terms of the associated reduction in shape anisotropy, and, for a smectic-isotropic transition, the resulting reduction in the lateral intermolecular interactions. For these particular nonsymmetric dimers, we must take into account the potential disruption of the favourable unlike mesogenic unit interaction. Within this overarching framework, it is surprising that we do not see a larger reduction in T_NI on moving from CB5O.s4 to CB5Ot4. Even more surprising are the observations of a much smaller decrease in T_SmAI passing from CBO12O.4 to CBO12O.s4, and an increase on moving to CBO12O.t4.

To account for these observations, we must now consider the intercalated nature of the local molecular packing in both the nematic and smectic phases. We have seen that the formation of these intercalated smectic phases is thought to be driven by a specific favourable interaction between the unlike mesogenic units [1,2], and in such an arrangement, the terminal chains are accommodated in a volume governed by the length of the flexible spacer, see Figure 1. In all nine groups of dimers reported here, the terminal chains are shorter than the spacer and thus intercalated smectic behaviour is possible and indeed observed. For the shortest even-membered spacers, CB5O.X and CBO4O.X, smectic and nematic behaviour is seen for the butyl-substituted dimers and for CB5O.s4. In the nematic phase, the primary effect of chain branching is to reduce the shape anisotropy and disrupt the specific interaction between the unlike mesogenic groups. Both these effects serve to reduce T_NI. In the smectic phases, however, the shorter, branched terminal chain may fill the volume in the intercalated structure more efficiently and enhance the stability of the phase, and this accounts for the increase in the smectic-isotropic transition temperatures moving from a sec-butyl to a tert-butyl terminal chain for dimers having the longer spacers. This greater packing efficiency also accounts for the higher melting temperatures shown by the tert-butyl homologues.

The greater tendency for the tert-butyl group to promote smectic behaviour than the sec-butyl chain is evident in the behaviour of binary mixtures of each with the corresponding butyl terminated dimer. Figure 9 shows, for example, the transition temperatures of the binary mixtures of CBO6O.4 with the corresponding branched dimers. As the mole fraction of CBO6O.4 is reduced in both sets of mixtures, T_NI falls as would be expected. Surprisingly, however, the N-I trendline for both sets of mixtures is essentially identical indicating that the combined effect of the sec or tert-butyl chains in reducing the structural anisotropy and disrupting the intermolecular interactions compared to CBO6O.4 are more or less the same in the nematic phase. By comparison, however, the SmA-N/I trendline increases as the mole fraction of CBO6O.4 is reduced for both sets of mixtures, but now the two lines have very different gradients with that associated with the tert-butyl mixtures being much steeper. Indeed, the SmA-N trendline intersects the N-I trendline such that the 0.8 mole fraction CBO6O.t4 mixture shows a SmA-I transition, as does pure CBO6O.t4, whereas the corresponding sec-butyl mixture and CBO6O.s4 both show nematic phases over a temperature range of about 20 K. This enhanced stabilisation of the smectic A phase by the tert-butyl chain is further evident in the behaviour of binary mixtures of CBO12O.4 with the corresponding branched dimers, see Figure 10. All three CBO12O.X dimers exhibit a smectic A-isotopic transition. Addition of either CBO12O.s4 or CBO12O.t4 to CBO12O.4 decreases T_SmAI, but the reduction is larger for the sec-butyl homologue. This further supports the view that the tert-butyl group can fill space more effectively in the intercalated structure.

Figure 9. (Colour online) The dependence of the transition temperatures on the mole fraction of CBO6O.4 for binary mixtures of CBO6O.4 with CBO6O.s4 (red circles) and with CBO6O.t4 (blue squares). the triangles indicate the transition temperatures for CBO6O.4. Filled symbols denote N-I transitions and open symbols either SmA-N or SmA-I transitions. Only Sm-N/I and N-I transitions are shown for the sake of clarity.

Figure 10. (Colour online) The dependence of the smectic A-isotropic transition temperatures, T_SmAI, on the mole fraction of CBO12O.4 for binary mixtures of CBO12O.4 with CBO12O.s4 (red circles) and with CBO12O.t4 (blue squares). The triangle indicates T_SmAI for CBO6O.4. Only SmA-I transitions are shown for the sake of clarity.

Finally, it is interesting to note that the longer odd-membered dimers have a strong tendency to exhibit tilted intercalated phases, whereas their even-membered counterparts show orthogonal phases. This has been attributed to the difference in shape between odd- and even-membered dimers [31]. Specifically, the bent odd-membered dimers experience greater difficulty packing into an intercalated arrangement and this provides the driving force for the formation of the intercalated smectic C phase in which the tilt direction alternates between the layers such that globally the tilt angle is zero, but non-zero within a layer [31].

Conclusions

The very general observation is that in a low molar mass liquid crystal branching a terminal chain reduces the liquid crystal transition temperatures. For the nonsymmetric dimers reported here, this is indeed the case on comparing their values of T_NI. Thus, the value of T_NI falls by around 40 K on passing from the n-butyl to sec-butyl-substituted dimer. Somewhat surprisingly, for the only pair of dimers for which a comparison is possible, the decrease between the sec-butyl and tert-butyl homologues is just 7 K. By contrast, however, the tert-butyl homologues show higher values of T_SmI than their sec-butyl counterparts, and in the only direct comparison possible, the value of T_SmAI for tert-butyl dimers is only marginally lower than that of the corresponding n-butyl homologue. This highly surprising observation may be accounted for in terms of the intercalated structure of the smectic phases formed by these dimers, and specifically the better packing efficiency afforded by the tert-butyl group within this structure in which the ability to accommodate the branched terminal chain increases with spacer length.

Acknowledgements

EG and DP thank the National Science Centre (Poland) for financial support under the grant no. 2016/22/A/ST5/00319. RW gratefully acknowledges The Carnegie Trust for the Universities of Scotland for funding the award of a PhD scholarship 2015–2018

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the National Science Centre (Poland) [2016/22/A/ST5/00319].