Defect trajectories and domain-wall loop dynamics during two-frequency switching in a bistable azimuthal nematic device

Bistable azimuthal nematic alignment textures have been created in micrometer-scale channels for which one sidewall is smooth and straight and the other possesses a symmetric sawtooth morphology. The optical textures have been observed during dynamic switching between the two stable states in response to dual frequency ac waveform driving of a highly dispersive nematic liquid crystal. The switching processes involves collapsing of filamentlike director reorientation (tilt-wall) loops and the associated motion and annihilation of surface defects along and close to the edge at the sawtooth sidewall. The predictions from both the $n$-director-based Ericksen-Leslie theory and the $Q$-tensor theory are in good agreement with the experimental observations.

Figure 1

(a) A 3D representation of the sawtooth cell. (b) A plan section of the sawtooth channel.

Figure 2

Two stable states (a) planar and (b) HAN as observed through a microscope with crossed polarisers with corresponding director profiles.

Figure 3

Switching the cell (a) from the planar state to a HAN state with a low frequency ac field and (b) from a HAN state to the planar state using a high frequency ac field.

Figure 4

(Color online) Comparison of experimental results of switching the cell from the planar state to the HAN state with the computed $\theta$-model director profile, the associated transmission image and the transmission calculated from the $Q$-tensor model.

Figure 5

(a) The direction of rotation for the liquid crystal molecules near the surface depends on the surface orientation. (b) 0.1 s after the voltage is applied when the upper surface has no pretilt. (c) 0.1 s after the voltage is applied when the upper surface has a small pretilt of 0.01 radians, preferring the $\theta =-\pi /2$ state.

Figure 6

The loop shrinks until it disappears near the middle of a grating edge.

Figure 7

(a) The $x$-coordinate of the loop edges versus time for experimental data (dashed line) and numerical results (solid line). (b) The volume of the loop decreases linearly after the voltage is applied for experimental data (dashed line) and numerical results (solid line). (c) The elastic energy of the system decreases while the voltage is on until it reaches a constant value, then decreases as the cell relaxes into the HAN state.

Figure 8

(Color online) Enlarged plot of the director and corresponding transmission plots when switching from the planar state to the HAN state. (a) Before the pulse is applied. (b) After the pulse is applied and the loop has decreased in size. (c) After the loop has disappeared. (d) The final state after the pulse is removed.

Figure 9

When a voltage is applied across a long channel a thin line of high distortion can appear connecting the upper and lower surfaces.

Figure 10

(Color online) Comparison of experimental results for switching the cell from the HAN state to the planar state with the $\theta$-model director profile, the transmission profile resulting from this director profile and with the $Q$-tensor model.

Figure 11

(a) The $x$-coordinate of the loop edges versus time as the voltage is applied for experimental data (dashed line) and numerical results (solid line). (b) The volume of the loop decreases in a linear fashion as the voltage is applied for experimental data (dashed line) and numerical results (solid line). (c) The elastic energy of the system also decreases while the voltage is on until it reaches a constant value, then decreases as the cell relaxes into the planar state.

Figure 12

(Color online) Enlarged plot of the loop annihilation from the HAN state to the planar state. (a) As the voltage is applied the ends of the loops begin to move together along the surface edge. (b) The transmission representation of the director profile in (a).