Leaky-wave exploration of two-stage switch-on in a nematic pi-cell

L. Z. Ruan, J. R. Sambles
2005 Applied Physics Letters  
The two-stage switch-on dynamics of a nematic pi-cell are explored in detail using a convergent beam fully-leaky guided mode technique. The cell shows an initial switch-on with a time scale in the range several ms to tens of ms ͑depending on drive voltage͒ from the symmetrical H state to a new and semistable H state. It then slowly changes ͑over several hundred ms͒ to the final stable asymmetrical H state. Nematic liquid crystals form the basis of the majority of flat screen displays. Much
more » ... displays. Much interest has focused on pi-cells, 1-3 in which the director has parallel surface tilts, with liquid crystals of positive dielectric anisotropy. There are three different director states in such cells, a horizontal state, or H state, a twisted vertical state, or T state and a planar vertical state or V state, depending upon the surface tilts and the elastic constants of the material. Under application of an electric field, there are possible transitions between the different states which have been considered in device applications. [3] [4] [5] There exists a very well-accepted theory concerning flow and dynamic effects in nematic liquid crystals which was developed by Ericksen 6,7 and Leslie. 8 in the 1960's, and later simplified by Berreman and van Doorn. 9,10 However, there are only a few detailed experimental confirmations of many of its predictions. Some simple experiments 11,12 have confirmed the theoretically predicted "backflow" in a twisted nematic cell during switch-off by recording an optical "bounce" effect in transmission. However almost all experiments have been based on transmission or reflection observations during switching and as such the optical signal from the liquid crystal cell is an integrated response. Only recently, by using a convergent beam optical guided wave technique 13,14 has the switching process 15,16 in a twisted nematic cell been studied in detail providing convincing detailed evidence of the validity of the Ericksen-Leslie theory. In this present study, a convergent beam system is used to investigate the switch-on dynamics of a pi-cell, finding a two-stage switching process. The cell is comprised of two low index glass plates ͑n = 1.5170 at 632.8 nm͒, each coated inside by a thin ͑ϳ50 nm͒ layer of indium tin oxide ͑ITO͒, on top of which are surface aligning layers of rubbed polyimide. The rubbing directions are parallel and the plates are spaced by 5.0 m beads. A monodomain is obtained by filling the cell with liquid-crystal E7 ͑Merck͒ in the isotropic phase and slowly cooling to room temperature. This cell is inserted between the two glass hemispheres, optical contact being achieved with matching fluid all having the same index as the glass substrates. This complete assembly is placed so that the center of the sample is at the focus of a monochromatic light beam ͑see Fig. 1͒ . The light beam is collimated into a less-coherent parallel beam of diameter about 5 cm by using one rotating and one fixed diffuser and a beam expander. This beam passes through a polarizer, a horizontal slit aperture, and a pair of converging lenses, being focused to a spot at the center of the cell to be studied. Both reflectivity and transmissivity signals from the cell are recorded, through a second polarizer, by a charge coupled device camera ͑DALSA͒. To obtain data most sensitive to the director profile in the cell, it is set such that the rubbing direction at the surface has an angle of about 45°with the incident plane. In addition, because the angular region of the fully-leaky guided modes which is most sensitive to the director alignment is in the high in-plane wave vector area, the sample is arranged such that the angle between the central axis of the convergent beam and the cell normal is 70°. Choice of input polarizer ͑p, transverse magnetic, or s, transverse electric͒ and output polarizer allows a variety of angle dependent signals to be recorded. Data are fully analyzed for the static guided mode data at 0 V, 2.8 V, and 4.7 V ͓root-mean-square ͑10 kHz͒ voltages͔, and also for the dynamic switch-on data for 2.8 V and 4.7 V. Figure 2 shows how one of the signals, T ss , varies with time after switch-on of 4.7 V. It is apparent that in this dynamic process there are two quite different steps, first there is a relatively rapid change to a nearly unchanging semistable state. This state remains for a long time slowly relaxing to a different final state. The fast switch-on time of the first step is about 22 ms for 2.8 V and 7 ms for 4.7 V ͑Fig. 2͒, while the time over which the cell remains in the semistable state is about 200 ms for 2.8 V and 300 ms for 4.7 V applied fields. The cell reaches its final state after about 500 ms for 2.8 V and 750 ms for 4.7 V. Using multilayer optical theory 17 to fit the reflection and transmission data, the full director profile in the cell is obtained at many time steps ͑ϳ1000 sets of angle dependent data recorded͒ during the switch-on. From the data taken a͒ Electronic mail: j.r.sambles@exeter.ac.uk FIG. 1 . Schematic of the convergent beam system. APPLIED PHYSICS LETTERS 86, 052502 ͑2005͒
doi:10.1063/1.1806544 fatcat:7mhkpqtapjeexaptdrclrciffe