1. Introduction

Liquid crystal (LC) materials find a wide range of applications in optoelectronics and photonics. The applications typically make use of the material’s large, broadband optical birefringence caused by the anisotropic nature of the molecules, combined with the high sensitivity of the molecular orientation to externally applied electric, magnetic, and optical fields. Recently, heterogeneous LC systems have attracted much scientific attention due to their potential for additional useful properties over the conventional LCs. By combining different constituents, it is possible to design novel smart materials with unique properties that the initial substances do not have. To this end, liquid crystals have been doped with various types of additives: azo-dyes [1], nanoparticles (ferroelectric, ferromagnetic, metal, etc.) [2–6], and polymers, as in polymer-dispersed and polymer-stabilized liquid crystals [7,8].

Nanoparticles have shown to be effective dopants in diluted LC suspensions. They do not disturb the orientation of the LCs but increase their sensitivity to electric, magnetic or light fields in compounds containing, respectively, ferroelectric, ferromagnetic or plasmonic metal nanoparticles. Nanoparticles can even provide spontaneous vertical LC alignment [6]. Doping polymers into LCs helps stabilize the system and provides materials that at the same time have unique anisotropic properties and a limited mobility for the LC molecules. Adding of dye molecules in an LC matrix usually does not change the basic matrix properties, such as elasticity, clearing temperature, electro-conductivity, refractive index or the other dielectric properties, provided that the dye concentration is kept low (1-2 wt%). The most significant change in this case is the appearance of an absorption band at a desired wavelength range of the spectrum. Azo-dye-doped liquid crystals demonstrate a collective molecular orientational response to polarized light matching in wavelength the absorption band of the dye. This effect, caused, e.g., by light-induced anchoring [9–11], makes possible several applications, for instance, in optical storage and processing of information (binary images, polarization grating recording [12–14], etc.).

Most light-induced effects in dye-doped LCs originate from the effective trans-cis isomerisation processes of the azo-dye molecules upon irradiation. This process is widely used not only in LCs, but also in materials based on other organic matrices, e.g., polymers. Materials in which azo-dye molecules are bound into a polymer have made it possible to realize fascinating applications relying on the photo-induced motions of the azo-dye molecules [15]. By simple irradiation with linearly polarized light, the azo-dyes are known to re-orient themselves perpendicular to the light polarization direction, which one can observe as light-induced birefringence [15]. Using an interference pattern of light, one can obtain periodical surface-relief gratings (SRGs) [16] applicable, e.g., in diffractive optical components and waveguide couplers. The bond between the photoactive molecules and the matrix can be covalent or non-covalent (ionic, coordination or hydrogen bond). Recently, interesting results have been obtained in hydrogen-bonded polymer-dye complexes by using light to very effectively induce strong birefringence or SRG formation in thin film samples [17,18]. H-bonding has clear advantages over the stronger bonding types. As opposed to the cooperative binding in ionic complexes leading to a 1:1 stoichiometry between the polymer and the dye, hydrogen-bonding leads to a random complexation between the constituents [19], which provides a means to continuously tune the dye-to-polymer ratio. Compared to traditional covalently-bonded side-chain dye-polymers, H-bonding is much more cost-effective and enables modular tunability of the complexes.

Non-covalent interactions between molecules of different substances have also stimulated the preparation of new LC composites. However, most publications so far related to non-covalent LC systems are limited to the so-called functionalized LC polymers, i.e., LC polymer matrices modified with low-molar-mass dopants, such as photochromic molecules [19,20].

In this paper, we propose and investigate into a new heterogeneous liquid crystalline material where hydrogen-bonded polymer-azo-dye complexes are doped at low concentration (c < 1 wt. %) in an LC bulk. The resulting dilute nano-suspension combines the unique anisotropy properties of liquid crystals with the stability of polymers and the light sensitivity and light-induced mobility of azo-dyes. In particular, we discuss the appearance of a surprising self-orientation property in the heterogeneous system and a spontaneous anchoring transition from planar to homeotropic alignment that the original components of the material do not possess.

2. Materials and experiments

2.1 Preparation of LC with polymer-azo-dye complex

The experiments were carried out with a nematic liquid crystal 4-Pentyl-4'-cyanobiphenyl (5CB; PI Chemicals Co., Ltd, Fig. 1(a) ) doped with a hydrogen-bonded polymer-azo-dye complex at a low concentration. The polymer-dye complex P4VP(CHAB)_x was prepared from the polymer poly-4-vynil-pyridine (P4VP, Mn: 1000 g/mol, Mw: 1200 g/mol; Polymer Source Inc., Fig. 1(b)) and the azo-dye 4-cyano-4'-hydroxyazobenzene (CHAB, BEAM Co., Fig. 1(c)). Both materials were separately dissolved in dimethylformamide (DMF, 98%, Fluka). After being stirred for 24 hours, the solutions were filtered through 0.2 μm syringe filters, and then mixed at two different concentrations in order to obtain complexation degrees of x = 0.5 and 1.0.

 

Fig. 1 Chemical structures of the liquid crystal 5CB (a), polymer P4VP (b), and azo-dye CHAB (c).

Download Full Size | PDF

The presence of a proton accepting pyridine ring in the structure of the polymer ensures the formation of donor-acceptor hydrogen bonds with the hydroxyl group of the CHAB dye [21]. CHAB was selected from a large variety of azo-dyes due to its rod-like shape and chemical structure that is very similar to 5CB [22].

The polymer-azo-dye complex P4VP(CHAB)_x was stirred for 24 hours, added to the LC at a concentration of less than 1 wt %, and then the DMF solvent was evaporated away at room temperature in 50 h. This finalized the preparation of the liquid crystal suspension doped with the hydrogen-bonded polymer-azo-dye complex, 5CB + c % P4VP(CHAB)_x (c < 1 wt. %). For comparison, samples of 5CB doped either with a pure polymer or an azo-dye were also prepared.

The suspension was poured in a sandwich-type symmetrical glass cell, the inner surfaces of which were covered with rubbed polyimide films (polyimide poly(4,4'-oxydiphenylene-pyromellitimide) [23] to provide strong anchoring and planar alignment of the LC. The cell thickness, set by Teflon stripes, was measured to be 27 μm. Cells filled either with the complex-doped LC, dye-doped LC, polymer-doped LC or pure LC from isotropic state (T ≈50 °C) were cooled slowly down to room temperature. The orientation of the doped LCs in the cells was checked with polarizing microscopy. The alignment was found to be planar with high fidelity in each case.

2.2 Anchoring transition

We observed that the cell filled with the complex-doped LC was not stable in orientation with time. Usually, the rubbed polyimide surface is considered to provide a strong anchoring for 5CB (W ~3 × 10⁻⁴ [24] ÷ 5 × 10⁻⁵ J/m² [25]), and, thus, should secure a stable orientation for the suspension. Indeed, immediately after preparation of the cell, the suspension had a perfect planar alignment along the rubbing direction. However, with time the complex-doped LC started to undergo a clear anchoring transition. After 200 hours the sample appeared to be totally homeotropic that was studied in polarizing microscope (Fig. 2 ). The typical conoscopic pattern of homeotropic alignment was also observed. Thus, the planar orientation of the suspension turns to a homeotropic one without any external action as the cell ages. In a cell with higher concentration of the complex in the suspension, e.g. 5CB + 1% P4VP(CHAB)_0.5 the anchoring transition occurs faster. This surprising phenomenon was observed only in the complex-doped LCs, but not in LCs doped either with the pure polymer or dye at the same concentration. The samples of 5CB + 0.25% CHAB and 5CB + 0.25% P4VP remained planar aligned for the whole period of observation.

 

Fig. 2 Photographs of the LC cell obtained between crossed polarizers at 45° to the rubbing direction. The LC cell was filled with 5CB + 0.5% P4VP(CHAB)_0.5. Time after the cell preparation: (a) 70 h, (b) 100 h, (c) 160 h, (d) 190 h, and (e) 220 h.

Download Full Size | PDF

We tested the thermostability of the alignment of the LC suspension. After the anchoring transition had taken place, we heated the cells to 100 °C for 30 minutes, thus destroying the orientation, and then cooled the sample back to the room temperature again. As confirmed by polarization optical microscopy, all the samples recovered their homeotropic orientation. Cycles of heating and cooling were repeated for several times and the homeotropic orientation was recovered after every cycle. This confirms that doping 5CB with the complex 0.5% P4VP(CHAB)_0.5 leads to a stable homeotropically orienting liquid crystal suspension.

Even in a cell made of bare glass (or quartz) plates without any aligning layers, the complex-doped LC self-oriented to a perfect homeotropic alignment, contrary to samples of pure LC, or LC doped with either dye or polymer, which all oriented rather randomly in the same cell. After filling the cell from isotropic state, the homeotropic alignment in the complex-doped LC cell appeared immediately after the sample had cooled down to the nematic state during an hour.

We also checked the critical concentration of the complex added to the LC that would still result in a self-oriented material. We decreased the concentration of P4VP(CHAB)_0.5 in the LC down to 0.15 wt % and still observed the self-orienting property of the suspensions. Below that concentration the orientation of the suspension on pure glass became random.

To understand the self-orienting property we carried out additional experiments.

2.3 Nematic to isotropic phase transition temperature

The phase transition temperature from nematic to isotropic phase of the complex-doped LC was studied and compared with the samples containing pure LC or LC with the other dopants. To justify the comparison, the concentration of the dye or polymer in each of the doped samples was the same. Flat optical capillaries of 500 × 50 μm in dimension were filled with the pure, dye-doped, polymer-doped, and complex-doped LC samples, placed side by side on a Linkam TMS 94 heating stage, and the phase transition was then observed using a Leica DM 4500P polarizing optical microscope. The heating and cooling rates were set to 0.2 °C/min. The results are gathered in Table 1 .

Table 1. The Phase Transition Temperature from Nematic to Isotropic State of the Pure, Dye-Doped, Polymer-Doped, and Complex-Doped LC Samples

View Table

We observed that doping with pure azo-dye CHAB and pure P4VP polymer slightly destabilized the nematic phase, which could be noticed as a decrease of the nematic to isotropic phase transition temperature. The respective reduced phase transition temperatures, 0.98 and 0.97, still remain very close to unity, so that the perturbation of the liquid-crystalline order is not remarkable. These dopants behaved like other molecular additives decreasing the interaction between the LC molecules. The measurements showed that adding polymer alone also caused broadening of the phase transition, indicating a co-existence of two phases at the transition. However, when doped with the hydrogen-bonded polymer-azo-dye complex, the phase transition temperature range was significantly broadened and the upper limiting temperature of the nematic to isotropic phase transition was increased compared to pure 5CB. Although adding pure dye or polymer to the cell seems to disturb the intrinsic LC order, adding a complex consisting of both dye and polymer is seen to stabilize the mesophase. It should be noted that the degree of the complexation also plays a significant role in the re-structuring of the LC. We observed that the suspension with a complexation degree of x = 1.0 (all the polymer side fragments were occupied with azo-dye molecules) led to a worse structuring of the LC compared to that at the complexation degree of x = 0.5. This means that in the new structuring of the heterogeneous system all three components (LC, polymer, dye) play a significant role.

2.4 LC order parameter in complex-doped LC

The director of the LC can tell only about the preferred orientation of the molecules, but nothing can be said how perfect the mesophase ordering is. Usually, an order parameter is determined, which quantifies the degree of the distant orientational order, $S=\frac{1}{2}〈3{\cos }^2\Theta -1〉$, where Θ is the angle between the axis of the separate molecule and the LC director, and averaging is made over the whole ensemble [26]. The degree of ordering can usually be approximated by measuring the dichroism in the LC absorbance in IR or UV.

The absorption spectra of the complex-doped LC for polarized light were measured with a HR4000CG-UV-NIR spectrometer (Ocean Optics Inc.). The spectra for a freshly prepared cell filled with the complex-doped LC showed that the LC order parameter could not be calculated in the UV region due to overlapping absorptions of 5CB (λ_max ≈275 nm [27]) and the polymer P4VP (λ_max ≈256 nm [28]). Nevertheless, the LC order parameter can also be estimated through the dichroism of the dye, if the dye molecule dopant is isomorphic to the LC [29]. Due to the remarkable resemblance of the chemical structures of 5CB and CHAB, the order parameter of 5CB can with good accuracy be approximated by measuring that of CHAB molecules in the LC bulk. The dye order parameter, S_dye, was calculated from the measured absorption spectra (Fig. 3 ) as $S_{LC}\approx S_{dye}=(A_{\max ,\left|\right|}-A_{\max ,\perp })/(A_{\max ,\left|\right|}+2A_{\max ,\perp })$, where $A_{max,\left|\right|}$ and $A_{max,\perp }$ are the absorption maxima of the dye for light polarized parallel and perpendicular to the liquid crystal director, respectively.

 

Fig. 3 Polarized absorption spectra of 5CB doped with 0.5% P4VP(CHAB)_0.5 just after filling the planar cell. The absorbance, A, was measured for light polarized parallel (||) and perpendicular ($\perp$) to the LC director. λ_max denotes the wavelength of maximum absorption of the azo-dye CHAB.

Download Full Size | PDF

Even though this method does not give an exact value for the LC order parameter S_LC, it provides a convenient and, actually, more reliable means to monitor changes in the order parameter [30] than more tedious methods measuring the absolute value of S_LC.

The dye dichroism was determined at the maximum of the azo-dye absorbance at λ_max = 373 nm. The LC order parameter in the 5CB + 0.5% P4VP(CHAB)_0.5 suspension was estimated to be 0.51. This value is to be compared with the order parameter measured in the dye-doped LC (same dye concentration) in the same symmetric planar cell, S_dye = 0.42, which is much smaller. For the suspension with the higher complexation degree, x = 1.0, the order parameter in the polymer-dye complex 5CB + 0.5% P4VP(CHAB)_1.0 appeared to be smaller than for the x = 0.5 sample. This result correlates with that obtained in the phase transition temperature measurements.

During the anchoring transition the difference in the maximum absorbance of the suspension for light polarized parallel and perpendicular to the rubbing direction decreased to zero for cell ages t > 150 h (see Fig. 4 ). Since for the homeotropic orientation this difference in the dye molecule absorption is intrinsically zero, we can conclude that in an aged cell all dye molecules are oriented perpendicular to the cell surfaces. Therefore, the 5CB LC molecules being isomorphic to the CHAB dye molecules are also oriented perpendicular to the surfaces, and at the end of the anchoring transition the suspension is homeotropically aligned.

 

Fig. 4 Dependence of maximum absorbance A_max (at λ_max) of 5CB + 0.5% P4VP(CHAB)_0.5 on the cell’s age. The measurements were done with light polarized parallel (||) and perpendicular ($\perp$) to the rubbing direction.

Download Full Size | PDF

3. Discussion

Spontaneous vertical alignment has also been achieved by Shie-Chang Jeng et al. [6] by adding POSS-nanoparticles into an LC and by Wen-Zheng Chen et al. [31] by doping an LC with POSS and methyl red dye. The role of the POSS additive in those works remained, however, somewhat unclear, i.e., whether the spontaneous vertical alignment was induced by adsorption of dissolved materials on the substrate or by the bulk effect of the dissolved materials in the LC.

Several facts should be taken into account in order to explain the extraordinary behavior of the complex-doped LC suspension developed here. (1) The increase of the clearing temperature correlates with the enhancement of the order parameter of the LC. The hydrogen-bonded complex changes the ordering and the whole structure of the LC matrix, leading to an improved orientation of the liquid crystal despite the low concentration. (2) All three components (LC, polymer, azo-dye) play an important role in the new structuring of the cell. The comparison of the suspensions of different complexation degrees showed that for an effective anchoring transition and a stable self-orientation not all the pyridine rings of the polymer should be occupied with the azo-dye (x = 0.5). Evidently, empty space between the polymer side fragments is needed for the LC molecules to interact with the dye molecules. (3) The rather long time required for the anchoring transition to take place and the ability of the system to be aligned on an untreated surface mean that there is a modification of the surfaces with time.

Summarizing the experimental data, we suggest a schematic model to describe the observed results. After filling the cell, aligned adsorbed layer of nematic molecules that formed just at filling [32] orients the bulk LC molecules along the rubbing direction generating a planar texture. The dye molecules are oriented parallel to the LC molecules due to the strong coupling between the two. The polymer attached to the dye with H-bonds is located in the bulk of the LC. After the filling, the polymer molecules start to get spontaneously adsorbed on the substrate and aligned along the surface. The dye molecules attached to the pyridine rings appear to be oriented perpendicular to the main chain of the polymer and thus to the surface. Due to the strong guest-host interaction between the LC and the dye molecules, the liquid crystal director prefers orientation along the new easy axis ignoring the original axis provided by the aligning rubbed polyimide surface (Fig. 5 ). The same model can explain the self-orientation occurring on pure glass. The adsorption of the polymer-dye complexes on the surface defines a new easy orientation axis for the LC.

 

Fig. 5 A schematic view of the anchoring transition in the complex-doped liquid crystal in time. d_LC – LC director.

Download Full Size | PDF

The rather long time for the anchoring transition to happen on the rubbed polyimide surface can be explained by the time it takes for the intensive molecular exchange between the aligning surface and the LC bulk [33]. This process critically depends on the relative sizes of the LC molecules and the polymer-dye complexes, as well as on their desorption and adsorption coefficients.

In order to verify the generality of the concept, we also prepared a suspension with a liquid crystal mixture, E7 doped with 0.5% of the complex P4VP(CHAB)_0.5. With this material the self-orientation also occurred but over a longer period of time. The anchoring transition from planar to homeotropic, for example, took 400 h. We also tried a polymer with a longer chain to enter the complex (P4VP, Mn: 7000 g/mol, Mw: 8100 g/mol). An eight times longer polymer chain slowed down the anchoring transition to happen within a period of appr. 16 days (380 h) for the case of 5CB doped with 0.5% P4VP(CHAB)_0.5. Except for the longer time, the other features of the self-orientation process did not change.

The results with the different LCs and polymers with different chain lengths give evidence of the generality of the alignment effect. They also show that the adsorption model relying on the molecular exchange process on the aligning surface is reasonable since a suspension with a longer polymer, as well as with LC having larger molecules requires a longer time for the transition. The reason why the homeotropic alignment during the anchoring transition starts to grow from the two edges of the cell is still under investigation.

As a further notice, we observed that the system turned out to be highly sensitive to linearly polarized light within the absorption band of the azo-dye, causing re-orientation of the LC perpendicular to the polarization vector of the light. Moreover, the complex-doped LC appeared to be a strongly nonlinear medium. These initial observations open up exciting perspectives for possible practical uses of this interesting photosensitive self-orienting material that does not require any additional processing or treatment.

4. Conclusion

We have developed a heterogeneous LC material that is composed of three functional components: liquid crystal, polymer and azo-dye. The polymer is for adsorbing onto the surface, the azo-dye for aligning the LC and for photosensitivity, and the LC for its unique anisotropy properties. The new suspension can be used in photonic and optoelectronic applications as a simple and very efficient photosensitive smart material that does not require any additional aligning processing or treatment. This can be of great advantage, e.g., in environments where aligning by orientants or rubbing is restricted by the small size or special shape of a surface (e.g. fibres, micro-lens arrays etc.). Our experiments also showed that the concentration of the complex can be decreased down to 0.15 wt. % and still have the self-orienting effect take place.