High Frequency Electrohydrodynamical Instabilities in Nematic Liquid Crystals

We have measured the thresholds of the electrohydrodynamic (EHD) instabilities appearing in the planar texture of the nematic liquid crystal MBBA (p-methoxybenzilidene-p-nbutyl-aniline) above the charge relaxation frequency. We observe two distincts instabilities, a slow convective hydrodynamic motion (made visible by dust motion) over distances comparable to the sample thickness, and the much shorter wavelength dielectric stripes (which give rise at higher voltage to the chevrons). At fixed temperature, the convective instability has a lower threshold than the dielectric stripes. These two instabilities are not coupled probably because of their very different spatial and time dependence. For fixed frequency, the temperature dependence of the convective threshold shows a continuity at T, (nematic to isotropic transition), with an isotropic Felici-like instability. On the contrary, the dielectric threshold diverges below T, as expected in the Carr-Helfrich mechanism. Our results confirm the Carr-Helfrich mechanism, fo explain the dielectric EHD instability. The claim by Barnik et al. that the high frequency EHD dielectric instability is explained by the Felici isotropic mechanism might result from a confusion between the two kinds of instabilities. Electrohydrodynamical (EHD) instabilities in nematic liquid crystals have arisen in the past a great interest, because of their applications to displays [l]. Following an idea by Carr [2], Helfrich [3] suggested a possible mechanism for DC excitation, extended by the Orsay Group [4] to the AC regime. This mechanism was based on a parametric coupling between space charges and bend distortion through the application of an electric field. It explained, in particular, the socalled dielectric regime, for an excitation frequency larger than the charge relaxation frequency (of the order of 10-100 Hz in a typical nematic liquid crystal). This mode was understood as a bend oscillation synchronous to the applied electric field oscillation, (*) Associated with C.N.R.S. around a quasi-static space charge distribution. The wave-vector of this mode was predicted, and found, much larger than the inverse thickness of the sample (on the contrary of the low frequency regime resulting in the so-called Williams domains which appear with a spatial period comparable to the sample thickness). Just above threshold, the thin periodic lines (period -a few microns) which are identified with the dielectric oscillations undergo a distortion and produce a typical chevron pattern. A lot of work has been devoted to these instabilities confirming the validity of the Carr-Helfrich-Orsay (CHO) model. Let us mention, for instance, the magnetic field dependence of the spatial period of the dielectric oscillations [5], and the observation of pretransitional bend oscillations by Rayleigh scattering [6]. Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1979365 HIGH FREQUENCY INSTABILrrIES IN NEMATIC LIQUID CRYSTALS C3-335 More recently, a Russian Group [7] underwent a systematic study of threshold values associated with high frequency EHD instabilities. A temperature dependence of the measured threshold has been published. which showed no discontinuity at the transition temperature Tc (from the nematic to the isotropic phase) with EHD instabilities observed in the isotropic phase. These last instabilities can be explained by one of the models described by Felici [gal and Atten [8b] for isotropic liquids. The conclusion of the Russian Group is that the high frequency dielectric regime was not due to the C H 0 mechanism, but was the particular appearance, in the nematic phase, of a Felici-like instability characteristic of isotropic liquids. The purpose of the present paper is to resume threshold experiments for the EHD instabilities in the dielectric regime, to ascertain which model, C H 0 or isotropic Felici, explains really the onset of high frequency instabilities. 1. The experimental set-up. We have used the standard geometry which is now classical for the observation of this kind of instabilities. We use a planar nematic sample of MBBA (methoxybenzilidene butyl aniline), with a T, of the order of 43 OC. This compound is not very pure but has the same Tc as the MBBA used in the Russian work [7]. The planar alignement is obtained by rubbing tin oxide coated glass plates. The thickness of the sample can be adjusted between 20 and 200 pm. The sample is placed inside a temperature controled oven (f 0.2 OC) and observed under a polarizing microscope. An AC generator applies a variable amplitude (V) and frequency (m) sine-wave voltage across the nematic sample. Some dust particules in the field of observation can show by their individual or correlated motion the existence of a field induced flow in the sample. 2. Experimental results. We first chose a rather thin sample (d = 26 pm) to reproduce the observations of reference [7]. At low frequency, we observe regular Williarns domains. The cut-off frequency, close to the charge relaxation frequency f,, is found at about 40 Hz. Due to sample aging, f, can increase up to 130 Hz after one month. Increasing the frequency, we do observe the threshold dependence of the dielectric oscillations proportional to wl'* (see Fig. 1, dots). This threshold is very sharp and can be determined with a one per cent accuracy. The interesting feature is that, below this threshold, we can see already some flow in the nematic, through the displacement of dust particules. Increasing the applied voltage from zero, for a given frequency, we first observe a slow erratic and uncorrelated motion of individual dust particules, always along the nematic director. For a higher voltage, we observe a correlated slow motion of almost all the dust particules with motion M BBA