arXiv:cond-mat/0310745v1 [cond-mat.soft] 30 Oct 2003

Textural Analysis of a Mesophase with Banana Shaped Molecules.

Yu.A. Nastishin1,2, M.F. Achard1, H.T. Nguyen1, M. Kleman3

1Centre de Recherche Paul Pascal, CNRS, Université Bordeaux 1, Pessac, France

2Institute of Physical Optics, Lviv, Ukraine

3Laboratoire de Minéralogie-Cristallographie de Paris, Université Pierre-et-Marie-Curie,Paris, France

 

1. Introduction

The existence of a liquid crystal (LC) state in materials with bent- (banana-) shaped molecules was first documented in 1929 by Vorlander [1,2].  For a long time no special interest was taken in such mesogens and they were simply called ‘bad rod’ phases.  The quantity of papers studying such a type of LC’s increased rapidly when the authors of reference [3] reported ferroelectric properties in such materials.  Several hundreds banana compounds have been synthesized since [4-15].  Because of their spectacular textures (in some cases electrically switchable) that differ with those of conventional rod-like smectics, these mesophases were identified as novel LC phases and labeled ‘Bn’ [4], with n =1,2,…,8 refereeing to the time sequence of their experimental discoveries.  Some of them were later identified as 3D crystals.  However there are X-ray and textural evidences that some Bn phases exhibit either 1D or 2D discrete translation ordering.  Because in many cases the Bn classification is not based on structural information, the identification of mesophases in newly synthesized banana-shaped compounds might be ambiguous and confusing.

These difficulties are especially pronounced for the B7 phase, the most intricate and most poorly understood of the B phases. The first mesophase with the B7 label was found in nitro derivatives (n-OPIMB-NO2 [10] of the parent series of banana shaped compounds.  The X-ray patterns for unaligned samples show a diffuse scattering in the wide-angle region indicating a liquid-like local order. Several peaks are observed at small angles. The most intense, with several harmonics, corresponds to a length comparable to the molecular length and can be assigned to a smectic periodicity. However the presence of other incommensurate peaks excludes a simple layered structure and the observation of reflections at very small angles suggest a structure of modulated layers at a larger scale [8].  Comparable X-ray patterns have been found in other compounds (DnF3 [8]) labeled B7bis; it was remarked that in partially oriented samples some of the additional X-ray peaks point in directions that are different from the main peak direction, suggesting additional translation ordering in the plane of the layers.  Furthermore, recent electron microscopy observations of freeze-fractured specimens presented by Clark et al. [16] show that the structure of the B7 phase of MHOBOW shows periodically modulated layers with a modulation length of about 50 nm; the direction of the modulation is preserved through the layers stacking.  Therefore the liquid like character of the layers is 1D.  Some recent results indicate that the layer structure is made of interdigitated (frustrated) columns [17].  A lamello-columnar model has also been advanced on a theoretical basis [18].  According to [18a], the B7 phase, although made of achiral molecules, has no mirror planes: it is a chiral phase, divided into domains of opposite chirality (having the same free energy density) separated by walls, but showing no helicity, i.e. no twist of the columnar directions in the ground state [i].  The same authors [18b] claim that the local order might be tetrahedratic.

The samples of this phase appear to be highly defective when the traditional alignment techniques are used.  In particular one observes impressive assemblies of helical nuclei (ribbon like) growing from the isotropic phase.  Because the interpretation of the X-ray data for powder samples is somewhat ambiguous, the present identification of the B7 phase actually contents oneself with the evidence of these helical ribbons (HR).  The reason to choose such a feature as a basis for the mesophase identification is related to the fact that the banana molecules are not chiral; and indeed it was observed at the time of the first identification of the B7 phase [10], that the spontaneous growth of HR’s from achiral molecules has never been documented for other LC phases.  However later it was found that the HR’s can also be formed in banana smectics with simple fluid layers (SmSG), on cooling of the isotropic phase [11].  Hence, the observation of helical domains is not a sufficient criterion to assign a B7 label to a mesophase.  Notice also that helical domains are frequently met in condensed matter physics.  Some solid crystals, e.g. sodium ammonium and quartz single crystals grow either in the left or right form; on the other hand, the presence of HR’s or of other helical features is documented in a wealth of organic systems with molecular chirality (phospholipidic membranes [19, 20] and biopolymers with helical nuclei [21]) or without, (helical features in chrysotiles [22,23]).  It is also well known that polymeric (e.g., polyethylene) spherulites may take shapes which imply the presence of twisted ribbons [24]. 

We shall rely in the sequel on a model with lamello-columnar ordering and a strong anisotropy of the bond strength between columns; the strength of bond is different depending whether the two neighboring columns belong to the same layer or to two adjacent layers. The B7 phase investigated in [5] does not form free-standing films as smectics usually do.  Contrariwise, free-standing strands are stable.  This is an argument supporting some kind of prevalence of the columnar structure of the B7 phase, over the layered structure.  In other B7 phases, free-standing films might be preferred, indicating different physical properties in different materials.  An intralayer modulation period much larger than the smectic periodicity is also present in the lamellar phase of nucleosome core particles, stacked into columns [25], but in this case the particles are chiral.

The textural analysis, as a technique capable to derive the internal structure of a phase via observed textures, actually did not contribute yet to the collection of essential information about the structure of the B7 phase.  Our aim is to relate the partially successful structural information we have presented above to the polarization microscopy observations.  To achieve this goal we present our own observations in the next section, after a presentation of the materials that we have used, focusing the attention to the observations that shed light on the structure of the B7bis phase in the compound D14F3.  We find some new details that were not reported yet.  The description of the textures is accompanied by the analysis of the observed signatures. Simple geometrical models inspired by previous models devised in other contexts, one relating to screw dislocations with a giant Burgers vector, the other one to developable domains, give an adequate description of the observed textures; they are presented in Section 3.  Conclusions are summarized in Section 4.

2. Materials and Experimental Conditions

As mentioned above, the first B7 phase was found in 2-nitro substituted resorcinol derivatives n-OPIMB-NO2 [10] (Fig. 1, compound 1a).  Similar textures and comparable X-ray patterns have been found in DnF3 series (Fig. 1, compound 2, with n = 9-14) in which the position of the azomethine and ester connecting groups have been exchanged with reference to n-OPIMB structure and a fluorine atom substituted on the external phenyl rings [8]. In the present case, the B7 mesophase of DnF3 materials present a ‘ferroelectric’ switching while the n-OPIMB-NO2 mesophase has no electric response.  Again, one notices that there are differences in the physical properties of B7 phases observed in different materials.  In addition, there is no experimental evidence of a complete miscibility of the B7 phases of n-OPIMB-NO2 and of DnF3 compounds [8] and thus, to avoid misunderstanding, the authors proposed to use the ‘B7bis’ label for D9F3-D14F3 materials.  However we find that the B7 and the B7bis phases have the same structure, hence the same defects and textures [8].

A comparison of the textures and defects above with those pertaining to two purely lamellar banana phases will also be presented: the B2 mesophase found in the non substituted n-OPIMB series [3] (compound 1b), which corresponds to a switchable SmC phase, and the lamellar phase of the 10-OPIMB-F3 (compound 1c).  The mesophase of this last material was first designated as a B7 since helical defects have been observed [26, 27], but reinvestigation by X-ray analysis unambiguously shows a simple layer structure [8,11,16]. 

The optical textures of the mesophases were observed through a polarizing microscope (Leitz Diavert) equipped with a hot stage (Mettler FP-82HT) and an automatic controller (Mettler FP-90).  The temperature was controlled and stabilized with 0.1°C accuracy.  The samples were observed on regular glass slides, treated for homeotropic or planar alignment, or on glycerol substrates. The thickness of the samples prepared between two substrates was quite large (about 1mm) allowing to examine B7 nuclei freely floating in the isotropic phase.

3. Textural Analysis.

3.1. fast cooling. helical ribbons

Here and everywhere in this paper fast cooling means that the sample was left to cool down without special setting of the cooling rate.  The cooling rate was estimated to be a few tenths of a degree per minute (» 0.3-0.5 °C/min).  The D14F3 compound exhibits a B7bis phase immediately below the isotropic phase [8].  When quickly cooled from isotropic phase, the B7bis phase nucleates in the form of numerous flexible HR’s. There are HR’s of both helicities. To be more precise, there are two types of ribbons: slim (S) and fat (F)[ii].  They can be easily distinguished under the polarization microscope by their textures (compare Fig. 2a,b and c-h).  They both look like double helices.  S-ribbons (Fig. 2a,b) are very flexible; between crossed polarizers they look gray, because of their low birefringence.  Contrariwise, F-ribbons can be of very different shapes (Fig. 2c-f); they are rigid, rod-like and highly birefringent; they show intense colors between crossed polarizers.  They also differ by the ratio of their pitch over the diameter p/D (see below).  We think that the F-ribbons are modifications of the S-ribbons, which occur on their further growth. 

The final textures (Fig. 3) are not similar to any texture of conventional smectics.  The distortions do not relax and defects do not seem to be mobile.  In particular, there is no evidence of focal conic domains (FCD’s).  This latter remark is of some importance; it is well known that the existence of liquid layers is at the origin of FCD textures, eg., in SmA’s and even in SmC’s, although in this last case the normal to the layers is not an axis of continuous rotational symmetry. One can infer that, reciprocally, the absence of FCD’s is the signature of the absence of liquid smectic layers.  In fact, this claim has to be qualified: in our case, the layers do exist, but they are 1D fluid, not 2D as it is in conventional SmA’s or SmC’s.  This difference suffices to prevent the formation of well-developed FCD’s.  

We failed to prepare a uniformly oriented cell using the substrates treated for homeotropic as well as for planar alignment.  It seems that the anchoring conditions do not have any influence on the texture of the cell. The final textures (Fig.3) are totally predetermined by the nucleation process, namely by the size and shape of the growing nuclei. This feature indicates that the relaxation of distortions is not of a smectic type.  Indeed, observing the nucleation of smectic phases from their isotropic phases one can notice that the boundaries between neighboring nuclei disappear when they touch each other during growth.  The boundaries of separate smectic nuclei are totally washed out from the texture quite quickly by the relaxation, which is of 2D fluid type for smectics.  In the case of the B7bis phase the nuclei do not merge together.  Newly born HR’s often wind about already existing HR’s without merging (Fig. 2d,g,h).  When a growing helical nucleus meets another nucleus it may stop to grow, or might break and start to grow in another direction.  When two nuclei come into contact, their shapes do not change at all or change very little.  The final texture is composed of set of ‘frozen’ nuclei (Fig.3).  For example the banana leaf nucleus (Fig.3c), that we attempted to grow carefully, did not change shape after contact with other nuclei.  We are then led to conclude that for the B7bis phase the relaxation of distortions is quite a slow process and might be of a fluid type but not more than in one dimension, as in columnar phases.

3.2. slow cooling. numerous nucleus shapes, mostly developable domains.

When quenching the sample from the high temperature isotropic phase region, as discussed above, the only nucleus shapes one observes are HR’s.  But when the sample is cooled slowly from the isotropic phase prepared at a lower temperature, the set of observed shapes is much richer.  The minimal cooling rate that could be imposed by our hot stage was 0.1 °C/min.  To obtain a slower cooling rate, the sample was cooled down by 0.1°C steps with annealing pauses of several minutes between steps.  Such a cooling procedure is possible because the temperature range of the biphasic region of banana materials is usually quite large (more than 1°C).

(i)- we have observed that the F-HR’s (Fat Helical Ribbons), Fig. 2 c-h, relax slowly in the course of time to a variety of shapes.  The relaxation of these geometries, whatever their shape, displays a unique feature, namely that the F-HR’s transform into figures with no helicity, which in the nicest case (probably in the absence of contact with the substrates) have an axisymmetric, spindle, shape, Fig. 4a.  But if, during the transformation, these objects meet the substrate, they then form flat domains that in their final form have the shape of a banana leaf, see Fig. 3c. 

Our observations show that the transformation of the F-helical structure (nonzero pitch) into spindles or flat domains (in both last cases there is no helicity left) is not continuous, but is based on a process that is akin to a Volterra process; the comparison will be given its full scope in section 3.4.  Fig. 4b demonstrates that at some moments during annealing (as well as on cooling) the HR spontaneously splits along a meridian half-plane (called below the cut surface, see Fig. 4b and the schematical drawings below the photographs). In the sequel, we call such an intermediate object (second and third photos from the left in fig. 4b) an opened ribbon.   It is evident that when the F-HR is cut as described above the ribbon is not helical anymore.  We believe that the transformation F-HR to an opened ribbon occurs when the HR achieves a critical size above which it does not grow any longer.  In contrast to the HR, the growth of an opened ribbon is not limited in size.  On growing the opened ribbon might touch the substrate.  It then adopts a flat geometry dictated by the finite thickness of the cell, forming a banana leaf as mentioned above.  A banana leaf is often observed in thin samples (with a thickness of a few microns), where growing nuclei touch both substrates from the very beginning of their growth.  In some cases when the initial HR is not very thick and located far from the substrates the sides of the opened ribbon can again come together, transforming the opened ribbon into a spindle with a k = +1 central disclination, Fig.4a, 4c, Fig.5.  Fig. 4c demonstrates the final moment of the transformation from an opened ribbon to a spindle.  When moving the focus of the microscope from the center to the surface, Fig. 4c, it is found that the line shown by the white arrow is seen only when the microscope is focused in the central part.  Contrariwise, the two other lines, shown by black arrows in Fig. 4c, are well seen at any focusing between the center and the surface.  These two lines indicate the position of the cut surface of the opened ribbon.  This cut surface shortens in the course of time during annealing, (as well as on cooling).  Simultaneously the central line propagates. Hence the opened ribbon spontaneously closes, transforming into a spindle, as schematically shown in Fig.4c.  We observed that, when growing (or melting), the impurities attached to the surface of a spindle follow a trajectory spiraling in a plane perpendicular to the axis of the object.  Thus, we conclude that the spindles are nothing else but developable domains (DD), well known in columnar phases [29], differing then slightly from objects of revolution. It is important to stress that the reverse heating of the DD’s never results in the appearance of HR’s even up to complete melting. Therefore the transformation HR®DD is not reversible, indicating that the HR’s correspond to a metastable state.

(ii)- Finally, one can observe, as a result of a very slow cooling from the isotropic phase, spiraling stripes forming domains that are very similar to DD’s (Fig.6).  The domain shown by white arrows in Fig. 6 has a spiraling form like a DD and is definitely not a FCD.  The structure of a DD is schematically shown in the insert in Fig.4c.

In some cases the annealing transformations described above result in the appearance of spherical objects with double conical defects with the apex located in the center of the nuclei (Fig. 5).  These objects have all the features of a special type of developable domain (schematically represented in Fig. 5b), first devised in [29], Fig. 8 (see also [30], Fig. 10.14), and called there a singular point.

Fig.7 shows that, in contrast to the smectic myelinic forms, the sections of the B7bis nuclei are not circular but rather well faceted (shown by arrows in Fig. 7).  The well-developed polygons of Fig. 7 are akin to the textures of a thermotropic columnar phase with columns starting at the substrate and propagating into the bulk.  Fig.7 indicates a 2D positional ordering in the B7bis phase.  

3.3. growth processes.

In this section we discuss the growth processes of the HR’s and of the other shapes that have been observed, and compare them with the usual growth processes of LC’s.

            (i)- HR’s (both types), slow cooling. The LC phase under investigation exhibits unidirectional growth: the growth rate along the axis of a HR is much higher than it is for all other directions.  In fact up to the moment when the transformation HR®DD takes place, the lateral growth is not visible, and the HR’s apparently have a constant width. This indicates that the LC structure has one preferential space direction to which molecules can be attached much more easily than to other directions.

The HR growth from the isotropic phase is at variance with the usual picture of the growth of Sm phases, which looks multidirectional. Let us recall that the usual nuclei which grow from the isotropic phase in a Sm are Friedel’s bâtonnets.  The layers are perpendicular everywhere to the interface between the bâtonnet and the isotropic phase and are arranged inside into numerous fragmented FCD’s of various sizes and orientations; the bâtonnets grow by attaching molecules to the layer cross-sections at the interface.  Thus, the growth looks eventually multidirectional.  Other nucleus shapes are documented for SmA’s (myelinic forms that grow along their cylinder axis in heterogeneous compounds, see [31] and references within; the same was observed for a one-component compound [32]).  But in all those cases the local growth occurs by attaching new molecules to a layer cross-section.

The HR nucleation is definitely at variance with this picture.  We observe no FCD’s, no bâtonnets, no coalescence where the nuclei enter into contact, and the lateral size of the nuclei is finite and well defined. A B7 phase is definitely different from a Sm phase.  

(ii)- Nucleus Shapes in Other Growth Conditions.  Let us point out that if the liquid crystal domains are not totally melted, in an incomplete heating process, the remaining scraps serve as seeds for the growth under further cooling.  In such a case the shape of the seed from which it originates predetermines the shape of the growing nucleus[iii]. We remind that for smectics the shape of a seed does not influence much the shape of the growing nuclei: smectics do not exhibit growth memory and the relaxation processes are akin to those of a 2D fluid.  As a result the nucleus shape transforms so much during the growth process that usually one cannot recognize the original shape of the seed in the shape of the growing smectic nucleus.

3.4. other Bn phases, presence of FCD’s.

 Our observations of the B7bis phase do not relate to the existence of a classic smectic order.  We especially point out that we failed to find FCD’s, which could be the main signature of the smectic structure. For conventional smectics the FCD’s can be easily prepared by placing the material on a glass substrate or on glycerol.

The texture of the B7bis film floating on glycerol is displayed Fig. 9.  Instead of FCD’s we find numerous DD’s, whereas for the B2 phase of the banana material 12-OPIMB (Fig.1b) we really find a texture consisting of FCD’s, with a typical Apollonius packing (Fig. 10) [33,34]. The FCD’s are broken in a way much similar to those of the smectic C phase. Another banana material, namely 10-OPIMB-F3, also forms broken FCD’s (Fig.11a) and Grandjean terraces decorated  by FCD’s (Fig. 11b). We also find a Schlieren texture in this material (Fig.11c), indicating that the main director is tilted with respect to the smectic layers.

The B7 phases often built easily either free films, or free strands. One may assume that the different cases come from different anchoring energy anisotropies with the isotropic phase, favoring either layer contact with the isotropic phase (i.e. free films) or ‘columnar’ contact with the isotropic phase (i.e. free strands).  Of course, the more favored is the ‘column’ contact with the isotropic phase, the more (meta)stable are the HR’s.

Summarizing, we conclude that the textures for the B7bis presented in this paper and the B7 reported in [10] are typical of a helical columnar texture with layers.  The textures of helical columnar phases without layers have been theoretically studied by Kamien and Nelson [35]; they have not been observed yet, except maybe in [25].  In this later case the textures show a set of twist boundaries widely separated one from the next.  Liquid crystals with a lamello-columnar structure have been reported in [36,37].

3.5.  tentative models for observed defects.

(i)- Helical ribbons. We claim that the geometry of a S-HR (Slim Helical Ribbon) is that one of a screw dislocation with a giant Burgers vector, limited to a cylindrical region enclosed by equal helical disclination lines of strength k = 1/2, see Fig.3a,b.  These two disclinations constitute a mode of splitting of the dislocation, whose Burgers vector is equal to the pitch of the helices.  Such a model is not new; it has already been developed in [38a] (see also [38b], chapter 5), in order to explain the double helices that had been observed by C. Williams [39] in a thermotropic SmA cooled from a nematic phase. 

In this model, the smectic layers take the shape of nested parallel helicoids. The pitch of the helicoids is the Burgers vector b of the screw dislocation.  The focal surfaces of this set of surfaces are singular along two helical lines, also of pitch b, rejected at a finite distance b/2p of the common axis of the helicoids.  The helicoids are folded about these lines, which are precisely the k = 1/2 disclinations, see the two solid lines in Fig.8, which is plotted using equations given in [38].

The thin helical lines of the mesh scribed on each helicoid in Fig.8 can be considered as a schematic representation of the columns, which are in the direction of the HR axis; this is in accord with the unidirectional growth.  Observe also that in such a model the modulation suffers a helical twist from a layer to the next.

According to this model the distance D between two disclination lines in a HR has to be equal to b/p.  Thus, the ratio p/D is expected to be equal to p.  We have performed measurements for 26 HR’s.  For 16 of them, which are of S type, we find p/D» p.  Note that the middle photograph in Figure 9b from reference [8] as well as a ribbon in Fiq.6 from reference [28], called by the authors a ‘single spiral’, also represent a S-ribbon, in excellent agreement with the model of a giant screw dislocation.  For the 10 other ribbons, which are F, this ratio is smaller than 1.  This confirms the observation according to which F-ribbons are modifications of S-ribbons by lateral growth, whereas the screw character is preserved. 

The good agreement of the p/D values obtained for S-ribbons confirms the model according to which they are formed by parallel layers, stacked on the central ruled helicoid as it is shown in Fig.8.

(ii)- DD singular point. Such an object is an intriguing example of structural chirality.  According to the model in [29], the columns are lines in 3D space, exhibiting a torsion of a given sign, either left or right, so that the configuration looks twisted, although it is not: n.curln = 0, where n is the columnar director.  In a sense, this object shows structural chirality, no structural helicity (if we assign the concept of helicity to the pseudo-scalar n.curln, i.e., torsion.  The fact that this defect has been found only now can be related to the chiral and columnar nature of the B7 phase.  Another reason is that according to [29] the interior of the conical region of the defect occupies quite a large volume.  It is reasonable to expect that this region (dashed volume in Fig. 5b) can be filled by the isotropic phase. 

The following remarks stress why the tentative models fit our observations:

 (a)- the layers are composed of columns that rotate helically from one layer to the next .  Were the layers in the HR’s cylindrical (and not helicoidal), the columns would occupy different geometries in neighboring layers, and the layer modulation would be generically incommensurate with the perimeter of the cylinder; this geometry would be energetically very unsatisfactory. 

(b)- consider two HR’s in contact; there is no way that the directions of possibly merging layers do match, which is an obstacle to coalescence.

(c)- size limitation: in the model of [38], the analytical continuation of the helicoids outside the cylindrical region bound by the disclinations, is of positive Gaussian curvature, G > 0, while it is of negative Gaussian curvature, G < 0, outside.  We believe that this is an essential ingredient of the size limitation. Recall that in thermotropic smectics the FCD’s physical parts have G < 0, although the Dupin cyclides possess analytical continuations with G > 0. 

(d)- observe that unidirectional growth is natural for columnar phases: the probability to attach additional molecules to the column end, i.e. along the fluid direction is much higher than when doing it from the columns side.

(e) – the interfaces with the isotropic phase of the nuclei (mostly DDs) obtained on slow cooling are well faceted, indicating 2D positional order, hence visualizing the columnar structure. 

4. Conclusion.

Due to their complex structure, which involves at the same time a smectic layering and a 1D modulation of the layers (we call columns the elements of this modulation), the B7 and B7bis phases have remarkable distinctive features, which are apparent in a careful study of the (similar) textures that they display under cooling from the isotropic phase. 

Fast quenching from the higher temperature range of the isotropic liquid phase: most fascinating are the helical ribbons that then appear.  These helical ribbons have a finite width that scales as their pitch; they do not coalesce in larger ribbons when in contact.  These helical ribbons are screw dislocations with a giant Burgers vector, split into two helical disclination lines of strength k = 1/2.  The pitch of the helical ribbons is the Burgers vector b of the screw dislocation. Measured and predicted values of the p/D ratio are in agreement.  We refer to this state as a metastable state where the columns show helicity from one layer to the next, i.e. a natural (metastable) pitch to which b is related, most probably linearly.  Such structural features are not compatible with a compact space packing; they are frustrated.

Slow cooling: the samples do not grow helical ribbons, but the defects of the structure are akin to defects in a columnar liquid crystal, namely developable domains or nuclei with well faceted crossections, where the columns show no twist at all.  We refer to the state without helicity as the ground state.

Annealing: If the HR’s are carefully annealed at a temperature close to the temperature of the transition, they transform into DD’s looking at first sight as figures of revolution (which we baptized spindles) or into flat domains in the form of banana leaf. The discontinous disappearance of helicity under annealing is a rather unexpected phenomenon.  In the resulting geometry (DD’s), the columns belonging to neighboring layers are parallel: helicity has disappeared, there is no twist, i.e. n. curln = 0.  Furthermore the reverse heating never results into the reverse transformation from spindles to HR’s.  Therefore, these spindles are textures of the ground state.

A more detailed account of the geometrical model of HR’s, with a discussion of its energy, will be given in a forthcoming publication.

Acknowledgements.

We are pleased to thank Prof. Oleg Lavrentovich and Dr. Vladimir Dmitrienko for a critical reading of the manuscript.  One of us (MK) acknowledges stimulating discussions with Dr Françoise Livolant on the subject of helical and twisted ribbons. 

Figure Captions:

Fig.1. Molecular structures and phases of the studied materials : the upper part concerns the parent series of  banana materials [3] with its nitro- [10 ] and fluoro- [28, 29] derivatives; the lower part shows the formula of DnF3 compounds [8] in which the position of the linking groups has been exchanged. The B2 label corresponds to a tilted fluid lamellar phase, thus to the smectic C family.  The compound 10-OPIMB-F3 also exhibits a tilted fluid lamellar phase [8, 11, 23], but there is no proof that it is actually the same ‘B2’ phase as 12-OPIMB.  B7 and B7bis correspond to a more complex structure (see text).

 

Fig.2. Helical ribbons growing after fast cooling from the isotropic phase.

 

Fig.3. Final textures of the B7bis s phase after quenching; the thickness of the cell is of a few microns: a,c-f - crossed polarizers; b- without analyzer.

           

            Fig.4. Transformation of HR’s into developable domains (DD), (thick cell, » 1mm): (a) DD’s are similar to figures of revolution when they do not touch the substrates; (b) the intermediate stage (transformation of a HR into an opened ribbon); (c) final stage of the transformation of an opened ribbon into a DD.  The schematic representation of these transformations are shown below the corresponding photographs. Insertion in figure c): the crossection of a DD has the configuration of a developable domain.

 

Fig.5. Some other examples of nuclei in the form of figures of revolution obtained after annealing of helical ribbons (no analyzer).

 

Fig.6. Developable domains obtained under slow cooling.

 

Fig.7. Non-helical nuclei with well faceted cross-sections observed under slow cooling; cell thickness, » 1mm; no analyzer except for the two bottom figures, where the polarizers are crossed.

 

Fig.8.  Smectic layers with a screw dislocation split into two disclinations of strength 1/2.  Note that the parallel layers can be made of parallel equidistant helical columns (shown by helical lines).

 

Fig.9. The texture of the B7bis phase on a substrate of glycerol. Bar = 25mm.

 

Fig.10. Focal Conic Domain texture in the B2 phase of the material B12q on a substrate of glycerol. Bar = 50mm.

 

Fig.11. Liquid crystal textures of the material 3FB10q on a substrate of glycerol at temperatures below the clearing point: (a)- FCD’s, (b)- Grandjean terraces decorated by FCD’s, (c)- Schlieren texture.


References

[1]. D. Vorlander, A. Apel, Ber. Dtsch. Chem. Ges., 62, 2831 (1929).

[2]. G. Pelzl, I. Wirth, W. Weissflog, Liq. Cryst., 28, 969 (2001).

[3]. T. Niori, T. Sekine, J. Watanabe, T. Furukawa, H. Takezoe, J. Mater. Chem., 6,1231 (1996).

[4]. G. Pelzl, S. Diele, W. Weissflog, Adv. Mater, 11, 707 (1999).

[5]. D. M. Walba, E. Korblova, R. Shao, J. M. Maclennan, D. R. Link, M. A. Glaser, N. A. Clark, Science, 288, 2181 (2000).

[6]. T. Sekine, T. Niori, M. Sone, J. Watanabe, S. W. Choi, Y. Takanishi, H. Takezoe, Jpn. J. Appl. Phys, 36, 6455 (1997)

[7]. J.-P. Bedel, J.-C. Rouillon, J.-P. Marcerou, M. Laguerre, M.-F. Achard and H. T. Nguyen, Liq. Cryst., 27, 103 (2000).

[8]. J.-P. Bedel, J.-C. Rouillon, J.-P. Marcerou, M. Laguerre, M.-F. Achard and H. T. Nguyen, Liq. Cryst., 27, 1411 (2000).

[9]. A. Jakli, C. Lischka, W. Weissflog, G. Pelzl, A. Saupe, Liq. Cryst. 27, 1405 (2000).

[10]. G. Pelzl, S. Diele, A. Jakli, Ch. Lischka, I. Wirth, W. Weissflog, Liq. Cryst. 26, 135 (1999).

[11]. a) W. Weissflog, H. Nadasi, U. Dunemann, G. Pelzl, S. Diele, A. Eremin and H. Kresse, J. Mater. Chem., 11, 2748 (2001) ; b) A. Eremin, S. Diele, G. Pelzl, H. Nadasi and W. Weissflog, Phys. Rev. E67, 021702 (2003).

[12]. J.-C. Rouillon, J.-P. Marcerou, M. Laguerre, H. T. Nguyen and M.-F. Achard, J. Mater. Chem., 11, 2946 (2001).

[13]. B. K. Sadashiva, H. N. Shreenivas Murthy and S. Dhara, Liq. Cryst., 28, 483 (2001).

[14]. J.-P. Bedel, J.-C. Rouillon, J.-P.Marcerou, M. Laguerre, H. T. Nguyen and M.-F. Achard, Liq. Cryst., 28, 1285 (2001).

[15]. J.-P. Bedel, J.-C. Rouillon, J.-P. Marcerou, M. Laguerre, H. T. Nguyen and M.-F. Achard, J. Mater. Chem., 12, 2214 (2002).

[16]. N. A. Clark et al, Boulder Workshop, August 2002, Boulder Banana Workshop Archive, http://anini.colorado.edu/bananas/

[17]. M. Nakata, Y. Tanikishi, K. Ishikawa, J. Watanabe, H. Takezoe and N. A. Clark, 9th Intern. Ferroelectric Liquid Crystal Conf., Dublin (2003) P-80.

[18]. a)- H. R. Brand, P. E. Cladis and H. Pleiner, Europhys. Lett.,  57, 368 (2002); b)- H. R. Brand, H. Pleiner and P. E. Cladis, Euro. Phys. J., E7, 163 (2002).

[19]. R. Oda, I. Huc, M. Scmutz, S. J. Candau and F. C. MacKintosh, Nature, 399, 566 (1999).

[20]. M. S. Spector, A. Singh, P. B. Messersmith and J. M. Schnur, Nanoletters, 1, 375 (2001).

[21]. a)- C. Y. Li, S. Z. D. Cheng, J. J. Ge, F. Bai, J. Z. Zhang, I. K. Mann and F. W. Harris, Phys. Rev. Lett. 83, 4558 (1999); b)- C. Y. Li, D. Yan, S. Z. D. Cheng, F Bai, J. J. Ge, B. H. Calhoun, T. He, L.-C. Chien, F. W. Harris and B. Lotz, Phys. Rev., B60, 12675 (1999).

[22]. S. Amelinckx, B. Devouard and A. Baronnet Acta Cryst A52, 850 (1996).

[23]. B. Devouard, A. Baronnet, G. van Tendeloo and S. Amelinckx, Eur. J. Miner., 9, 539 (1997).

[24]. H. D. Keith and F. J. Padden, Macromolecules, 29, 7776 (1996).

[25]. A. Leforestier, J. Dubochet, and F. Livolant, Biophys. J., 81, 2414 (2001).

[26]. C. K. Lee and L.C. Chien, Liq. Cryst. 26, 609 (1999).

[27]. G. Heppke, D. D. Parghi and H. Sawade, Ferroelectrics, 243, 269 (2000)

[28]. I. Dierking, H. Sawade and G. Heppke, Liq. Cryst., 28, 1767 (2001).

[29]. M. Kleman, J. de Phys., 41, 737 (1980).

[30]. M. Kleman, O. D. Lavrentovich, Soft Matter Physics: An Introduction, Springer 2003.

[31]. R. Pratibha and N. V. Madhusudana, J.Phys.II, France, 2, 383 (1992).

[32]. S. L. Arora, P. Palffy-Muhoray, and R. A. Vora, Liq. Cryst., 5, 133 (1989)

[33]. R. Bideaux, N. Boccara, G. Sarma, L. de Sèze, P.-G. de Gennes and O. Parodi  J. de Phys., 34,  661 (1973).

[34]. C. Blanc and M. Kleman, Phys. Rev., E62, 6739 (2000).

[35]. R. D. Kamien and D. R. Nelson, Phys. Rev., E53, 650 (1996).

[36]. P. Davidson, A.-M. Levelut, H. Strzelecka, V. Gionis, J. Phys. Lett., 44, L-823 (1983).

[37]. R. Richardson, S. A. Ponomarenko, N. I. Boiko and V. P. Shibayev, Liq. Cryst., 26, 101 (1999).

[38]. a)- M. Kleman, J.Phys., 46, 1193 (1985); b)- M. Kleman, Points, Lines and Walls, Wiley 1983.

[39]. C. E. Williams, Phil. Mag., 32 313 (1975).

 

[i]. We shall specialize the term chirality to conformations (molecular or structural) with no mirror plane; the term helicity will connote the presence of molecular or structural twist.  It is most useful in the description of phases made of bent-shape molecules to have these two concepts at hand.

[ii]. The evidence for two types of ribbons has been first asserted in [28]; but this publication is essentially devoted to an investigation of the fractal structure of the assembly of helical nuclei, not to a detailed study of the transition from one type of ribbon to another.

[iii]. This fact suggests an analogy with the principles of solid crystal growth techniques that are based on the growth from a seed (top-seeding techniques).

 

 


Fig 1

 

Fig 2

Fig 3

Fig 4 a) b)

Fig 4 c)

Fig5 a)

Fig5 b)

 

 

Fig 6

 

 

Fig 7

 

Fig 8

Fig 9

Fig 10

Fig 11