1. Introduction

Liquid Crystals (LC) are widely used in various types of displays and modulators and the operation of those devices is based on the electrically variable orientation of the director (average orientation of molecular axes) of the LC [1]. The initial alignment of the director in the LC cell remains a delicate issue up to day. Along with the traditional mechanical rubbing method, many alternative techniques have been developed to allow LC orientation with higher yield, for example, by oblique deposition of inorganic materials [2], by ion beam etching [3, 4] and light exposition [5, 6]. Photo-alignment possesses obvious advantages in comparison with the usual “rubbing” treatment of the substrates of LC cells, such as (i) elimination of electrostatic charges and impurities as well as mechanical damage of the surface; (ii) possibility to produce a good LC alignment in curved surfaces, multi-domain pixels, photonic bandgaps, thin tubes, LC droplets and other sophisticated cases, when usual “rubbing” method does not work. The application of photo-aligning technology results in a potential increase of manufacturing yield where fine tiny pixels are driven independently by thin film transistors. Promising results were obtained by using light induced photo-isomerization of azo-dye derivatives, which appears to be both efficient and very stable [7]. While the direct doping of the azo-dye in the volume of the LC [8–10] allows a rather efficient optical control of the LC alignment, its drawbacks are the low percentages of dye doping and the dynamic (non remnant) character of reorientation. Surface mediated control of LC alignment appeared to be more stable in time [11–15]. In this case, an azo-dye may be adsorbed on the substrate, deposited as Langmuir-Blodgett film [11,12], simply spin coated as thin layer [5–7], or added into polymer host [14]. Recently the technique of LC photo-alignment has been used to enhance the diffraction efficiency of a polarization grating recorded in azo-dye guest-host polymer film and then assembled with LC layer [14]. Similar polarization gratings were demonstrated by use of linear photopolymerizable polymer alignment layer [15].

In this paper we use a light polarization gratings to induce photo-isomerization in pure azo-dye layers deposited onto substrates of a LC cell and serving as a command layer to create corresponding LC alignment. As a result the permanent polarization grating is formed in the LC cell (the diffraction from azo-dye command layer is negligibly small). Polarizing microscopy and electro-optical studies of such grating have been carried out. Particularly the dependences of diffraction efficiency on the polarization state of incident light and on the applied voltage were studied.

2. Material system

The chemical formula of the azo-dye SD1 used for the fabrication of the command layer is presented in Fig. 1. One of the possible photo-aligning mechanisms in azo-dye films is a pure reorientation of azo-dye molecules. When the azo-dye molecules are optically pumped by a polarized light beam, the probability of absorption is proportional to cos² θ, where θ is the angle between the absorption oscillator of the azo-dye molecules and the polarization direction of the light (Fig. 1). Therefore, the azo dye molecules, which have their absorption oscillators (chromophores) parallel to the light polarization will most probably get excited and relaxed multiple times, which results in their reorientation from the initial alignment. This results in an excess of chromophores in a direction at which the absorption oscillator is perpendicular to the polarization of the light. For the excitation wavelength used the chromophore is parallel to the long molecular axis of the azo-dye (Fig. 1), i.e. the azo-dye molecules are tending to align their long axes perpendicular to the UV-light polarization. The function f(θ) of the statistical distribution of the azo-dye axes along various orientations θ, which is f =l/4π in the initial state, will tend to f= δ(θ-π/2) for a sufficiently long exposure. Hence, a thermodynamic equilibrium in the new oriented state will be established. Consequently, dichroism or birefringence is photo-induced and the associated order parameter, as a measure of this effect, goes to the saturation value, which can be very large in these materials.

The LC cell consists of two glasses covered with indium tin oxide (ITO) electrode layers. A thin film (thicknesses between 9–20 nm) of azo-dye SD1 having maximum absorption at 365 nm [6] was spin-coated directly onto ITO electrode of the substrate. Such substrates were used to assemble LC “sandwich” cells using a spacer providing a gap of 25 um. Those cells were filled (after photo-exposition) with commercially available nematic LC E7.

 

Fig. 1. The structure of sulfuric azo-dye SD1. The absorption oscillator is parallel to the molecular axis. Under the action of polarized light the SD1 molecules tend to reorient perpendicularly to the polarization of the activated light.

Download Full Size | PDF

3. Experimental set-up

The empty cell was exposed to periodically modulated (in space) polarization light pattern. Those polarization gratings were formed using a holographic setup with orthogonal states of polarizations of “interfering” beams (see Fig. 2). The CW Ar⁺-ion laser (operating at 488 nm) beam was expanded by the beam expander (BE) to ensure a homogeneous exposure over the whole examined region (beam diameter was about 6mm). A beam splitter (BS) was used to obtain the two symmetrically incident “interfering” beams. The angle between those beams was adjusted to provide spatial modulation period Λ=15μm. A Glan prisms (PG) and a wave plates (λ/2 and λ/4) were used to obtain, for example, right circularly (RCP) and left circularly (LCP) polarized beams of equal intensities. Experiments were carried out for various polarization modulation configurations, however only the case of RCP and LCP polarized “interfering” beams (hereafter called RCP+LCP), which generates periodically rotated plane polarized field [16], will be presented in this work. The total average power density of recording beams on the sample surface was 350 mW/cm² and the time of exposition was 10 min. A plane-polarized CW He-Ne laser beam (operating at 632.8nm) was used as a probe for real-time and post-exposure monitoring of diffractive gratings formed in the LC cell. This beam was used at normal incidence for all experiments. The polarization state of the probe laser beam was controlled with quarter wave plate.

The induced periodic LC director modulations were analyzed also by means of a polarization microscope.

 

Fig. 2. Experimental setup; λ/2 - half wave plate for power balance between interfering beams, BE - beam expander, BS - beam splitter, PG - Glan prism, λ/4 - quarter wave plates.

Download Full Size | PDF

4. Results and discussion

The relief profile of azo-dye command layer deposited onto substrate was first studied using an Atomic Force Microscope (AFM). In these experiments the pure glass plate without ITO was spin-coated with azo-dye and then was exposed to the interference pattern corresponding to RCP+LCP. A single laser beam exposure (with no spatial intensity modulation) was also carried out. Figure 3 presents AFM scans of the pure glass substrate (a), the substrate coated with azo-dye (b) and then exposed to light (c, d). One can see that the character of azo-dye distribution on the substrate before and after exposition is almost the same (compare Fig. 3(b) with Fig. 3(c,d)). There was no observable periodic relief modulation in azo-dye layer exposed to interference-patterned light (Fig. 3(c)), and the structure is rather similar to that of uniformly irradiated layer (Fig. 3(d)). So, no surface relief diffraction grating is expected in such azo-dye layers. To check this issue the monitoring of diffraction from the sample was carried out during exposition and after it. No diffraction was observed at these stages.

For the next step, the quality of LC photo-alignment by azo-dye was studied. In these experiments the azo-dye coated substrate was exposed during 12 min to linear polarized single beam from Ar⁺-ion laser (488 nm) and having a power density 175 mW/cm². Then the substrate was assembled with a second substrate coated with a surfactant CTAB which is known to provide a good homeotropic LC alignment. Such cells were filled with nematic LC E7 at room temperature. Figure 4 presents microscope photos of the cell placed between crossed polarizers. The difference between exposed and non-exposed areas of the cell can be clearly observed. In exposed area, the LC cell has a uniform hybrid alignment since one substrate is coated with CTAB, while the second one is coated with azo-dye which provides a planar alignment after exposition. There is no good alignment in non-exposed area. The high contrast between the two photos in exposed area indicates excellent quality of homogeneous (planar) LC alignment generated on the azo-dye coated substrate.

Finally, fresh (non-exposed and empty) LC cells were fabricated in the same way, as described above, but ITO coated glass substrates were used. Then the cells were exposed to optical polarization grating created by two “interfering” beam with opposite circular polarization (RCP+LCP). There was no observable diffraction from the cells without LC. However, strong diffraction was observed after filling such exposed cells with LC (see Fig. 5). A strong diffraction was also observed when the LC was present in the cell during exposition. These results show the modulation of LC alignment in the cell which is responsible for the diffraction. Thus, as it can be seen in 'Fig. 5, multiple diffracted orders are generated. This diffraction was studied for various voltages applied to the LC cell. The AC voltage of frequency 1 kHz was varied between 0 and 90 volts rms. It appeared that the diffraction may be strongly controlled by means of applied voltage.

 

Fig. 3. Atomic Force Microscope comparative scans of the command substrate: a) pure glass substrate, b) non-exposed SD1 layer on the glass substrate, c) SD1 layer exposed with “polarization interference” patterned light, d) SD1 layer exposed with uniform plane polarized light (at normal incidence).

Download Full Size | PDF

 

Fig. 4. Microscope images of the LC cell with one substrate coated with CTAB layer and the second substrate coated with azo-dye layer and uniformly (non-patterned) irradiated with single laser beam. The polarizers of the microscope are crossed. The polarization of exposing beam was (a) parallel to one of polarizers of microscope; (b) 45° degree with respect to polarizers of microscope. The central circular structure is an artifact due to the lamp illumination.

Download Full Size | PDF

 

Fig. 5. Diffraction patterns from the LC polarization grating for different polarizations of the probe beam and voltages. a) U=0V, linear polarized (p polarization) probe; b) U=15V, linear polarized (p polarization) probe; c) U=15V, right circular polarized (RCP) probe; d) U=15V, left circular polarized (LCP) probe.

Download Full Size | PDF

It is well known that polarization gratings can be used to selectively diffract and detect various polarization states of light [17, 18]. Light intensity distribution between zero and ±1 diffracted orders was studied as a function of the state of polarization (SOP) of the incident probe beam and applied voltage. The SOP of the probe beam was changed from Right Circular to Left Circular using a quarter-wave plate. The polarization selectivity of our grating is significantly increased with applied voltage. Figure 6 shows the results of quantitative studies of the diffraction efficiencies of 0 and ±1 orders as a function of the SOP of the incident probe beam for voltages U=0V and U=15V. In both cases, the first-order diffraction efficiencies strongly depend on the polarization state of the incident wave. The good contrast ratio between efficiencies for circularly polarized light diffracted into -1 and +1 orders is observed; this ratio can be improved by applying an electric field (see Fig.6(b)). The zero-order diffraction efficiency is independent on the polarization state of the incident beam, and is also controlled by electric field. Further increase of the applied voltage leads to almost complete erasing of polarization grating. We measured the ±1 orders diffraction efficiencies at U=90V to be below 1%. That can be explained by the fact that all molecules of LC are reoriented to uniform homeotropic state at high voltage and there is no more modulation of LC alignment in the cell. The detailed analysis of the diffraction efficiency on applied voltage is under study.

To better understand the character of modulation of the LC director in the cell, we have studied the microscope textures of obtained gratings (see Fig. 7). Both cell substrates were coated with azo-dye layers. The cell was exposed to RCP+LCP interference pattern and then filled with nematic LC. The thickness of the cells was 25 μm. One can see two different periodic structures in the cell - wide dark and bright lines with fuzzy edges and thin darks lines with high contrast edges. The first periodic line structure is observed only when the cell is placed between crossed polarizers of microscope (Fg. 7(a)) and disappears with removing the analyzer (Fig. 7(b)). This structure represents pure polarization grating in LC cell. The dark wide lines in Fig. 7(a) indicate the regions where the LC is oriented at 0° or 90° with respect to the polarizer (P) or analyzer (A); in bright regions the LC is aligned at an angle between 0° and 90° with maximum brightness at 45°. Such structure of LC alignment reproduces an azimuthal distribution of azo-dye molecules in command layers. The second periodic structure of thin dark lines is observed both in Fig. 7(a) and Fig. 7(b). That indicates the disclination lines in LC orientation. These disclinations may be attributed to the existence of a shift between patterned command structures on the top and bottom substrates of the cell. Such shift induces the competing right-handed and left-handed twisted domains in LC layer. These disclinations are under investigation.

 

Fig. 6. Diffraction efficiencies of the zero-order (dashed lines) and ±1 orders (solid and doted lines) as a function of the polarization state of incident light. a) no voltage is applied to the sample, b) applied voltage is 15V. The state of polarization is changed from RCP to LCP via linear polarization (p polarization).

Download Full Size | PDF

 

Fig. 7. Microscope photos of LC polarization grating: a) between crossed polarizers, b) without analyzer.

Download Full Size | PDF

5. Conclusions

Remnant high efficiency polarization gratings are created in nematic liquid crystal cell by using a command layer of azo-dye molecules deposited on the cell substrates and exposed to two “interfering” beams with opposite circular polarizations. The diffraction efficiency is controlled by electric field applied across the LC cell. Obtained polarization gratings can be used for electrically controlled discrimination and detection of polarized components of light.