Low-power control of haze using a liquid- crystal phase-grating device with two- dimensional polymer walls

We propose a two-dimensional (2D) polymer-walled liquid-crystal (LC) phasegrating device, which can be used to control the haze with a very low power. 2D polymer walls can be formed in an LC cell through ultraviolet light irradiation while applying an inplane electric field through phase separation induced by the spatial elastic energy difference. The transparent and translucent states can be realized by applying vertical and in-plane electric fields to the 2D polymer-walled LC cell, respectively. The cell can be operated with a very low power as the transparent [translucent] state is maintained even after the applied vertical [in-plane] electric field is removed. It consumes power only during state switching. The fabricated device exhibits outstanding performances, such as a very low operating voltage (< 10 V), low haze (< 2%) in the transparent state, high haze (> 90%) in the translucent state, and short switching time (< 2 ms), compared to those of other bistable LC devices, which can be used to control the haze. © 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement


Introduction
Low-power-consumption techniques for electronic devices are very important as they can extend the battery life of portable appliances or reduce the energy loss by electronic components in vehicles, showrooms, laboratories, buildings, etc. Liquid crystals (LCs) whose optical properties can be easily controlled by applying a weak external electric field are widely used in various electronic devices, particularly in display applications [1,2]. Recently, extensive studies based on the intrinsic features of the LC materials have been carried out for smart window or window display applications. Unlike traditional LC display devices, in which power is consumed by both backlight and driving circuits, smart window and window display devices use natural light without the requirement of a separate backlight so that the power is consumed only for driving of the device. The devices can be switched between transparent and translucent states by controlling the haze value through either light scattering [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20] or diffraction [21][22][23][24]. Usually, LCs can be switched by applying an electric field; they relax to the initial state when the applied electric field is removed. Such monostable LC devices require a continuous application of an electric field to maintain either the transparent or translucent state.

Formation of 2D polymer walls
In order to induce a large spatial elastic energy difference for phase separation of the LC/RM mixture, we can employ the effect of 2D confinement on the switching of nematic LCs. In the 2D-confined cells, when an electric field is applied to an LC cell with interdigitated electrodes, LC molecules are reoriented in the opposite direction at the edges of the interdigitated electrodes and domain boundaries where there is no change in the azimuth or polar angle of the LC director emerge [40][41][42][43][44][45]. At these domain boundaries, the LC molecules are not rotated by the elastic torque caused by neighboring rotated molecules, which leads to a large spatial elastic energy difference. Therefore, with the help of the phase separation induced by the spatial elastic energy difference, fine polymer walls could be formed at the domain boundaries of the cell through the UV-curing process without photomasks [46].
The polymer walls formed in the LC cell can contribute to the orientation of the LCs in the lateral direction so that a bistable LC phase-grating device could be realized using threeterminal electrodes [46]. However, the haze in the translucent state of a one-dimensional polymer-walled LC phase-grating cell may not be sufficiently high for practical applications [21][22][23].
In order to obtain a high haze in the translucent state of the device, we employed a 2D grating LC cell as the reaction template, as shown in Fig. 1, by which a large spatial phase difference can be induced regardless of the azimuth angle [22]. The common electrode on  Fig. 1 in Fig. 1 Fig. 3 In order to verify the bistable switching ability of the 2D polymer-walled cell, we measured the diffraction efficiency for the zeroth order and analyzed POM images and diffraction patterns, as shown in Fig. 5. We used a linearly polarized He-Ne laser beam (wavelength λ = 543.5 nm) as a light source whose polarization direction was perpendicular to the interdigitated electrodes. We measured the far-field intensity of the zeroth order using a photodiode placed 22.5 cm away from the LC cell.
When a vertical electric field was applied, the efficiency of the zeroth order increased with the applied voltage, as shown in Fig. 5(a). Owing to the vertical electric field applied to the cell, the LC molecules aligned vertically so that the spatial phase difference was reduced, which increased the light intensity of the zeroth order. This was confirmed through the POM images and diffraction patterns, as shown in Fig. 5(b). In addition to the change in the light intensity of the zeroth order, diffraction patterns with different angles were observed when a vertical voltage was applied to the cell. This was attributed to the formation of thicker polymer structures at the centers of the interdigitated electrodes than those at the middles of the gaps between them. The LC molecules near the middles of the gaps between the interdigitated electrodes can be switched with a lower voltage, whereas the LC molecules near the centers of the interdigitated electrodes require a higher voltage for switching owing to the strong anchoring attributed to the thicker polymer structures. Therefore, when a lower voltage is applied to the cell, only the LC molecules near the middles of the gaps between the interdigitated electrodes start to switch so that the grating period is doubled, which yielded diffraction patterns with reduced angles, as shown in Fig. 5(b). When the applied voltage was increased to 10 V, the LC molecules in the entire region were vertically aligned so that the incident light was not transferred to the higher orders and thus the diffraction disappeared.
When the applied vertical voltage was removed, the vertically aligned LC molecules tended to remain vertically oriented by the vertical anchoring owing to the alignment layer on each substrate. However, a slight change in the LC orientation was observed when the applied vertical voltage was removed. When a relatively low voltage was applied to the cell and subsequently removed, some LC molecules were relaxed by the in-plane anchoring provided by the polymer structure, as shown in Fig. 5(c), which led to a decrease in the zeroth-order light intensity. With the increase in the applied voltage to 10 V, a small change in the efficiency of the zeroth order was observed after the applied voltage was removed. If a voltage wave whose duration is not sufficiently long is applied to the cell, the intensity of the zeroth order may decrease as a fraction of the LCs could not remain vertically oriented and thus relax. In our experiment, when the duration was longer than 0.1 s, a small change in the efficiency of the zeroth order was observed after the applied vertical voltage was removed. In order to eliminate the effect of the pulse duration of the applied electric field on the light diffraction efficiency, all of the measurements were performed at 5 s after the vertical voltage was applied or removed. ric field is app rdless of the a rease in the zer hat the diffracti wn in Fig. 6 Fig. 7(c In order to the specular compared the When the ap induced so th increased, res translucent sta 0.9%, respect In the tran was slightly l index mismat specular trans 75.1% and 1.9 walled cell, th an even lower 8(a) and 8(b) lower than tha in-plane anch at a lower app small change was removed walled cell un and 92.2%, w respectively. T state even afte of the pure L the backgroun regardless of transparent st 7. Measured (a) s mer-walled cell, sw o evaluate the transmittance em with those pplied in-plane hat the specul spectively. At 9 ate were 4.5% ively. nsparent state, ower [higher] tch between th smittance and 9%, respective he voltage-depe r voltage than , respectively. at of the pure L horing by the p plied voltage a in optical perf d, as shown in nder an in-pla whereas those The 2D polym er the applied i C cell. Therefo nd view by the the azimuth an tate, as shown pecular transmitta witched from the tr optical perform and haze as of a pure LC e voltage was lar transmittan 9 V, the specul and 94.1%, wh the specular t than that of th he LC and po haze of the 2D ly. When an in endent specula that of the pur The operating LC cell. The re polymer structu as the polymer formance was o Figs. 8(a) and ane electric fie e after the app mer-walled cell in-plane voltag fore, in the tran e strong diffrac ngle, whereas in Fig. 8 In order to confirm the switching time of the fabricated LC cells, we investigated their dynamic switching behaviors, as shown in Fig. 10. We defined turn-on [turn-off] time as the transient time from 90% [10%] to 10% [90%] of the specular transmittance. In order to measure the response time of the pure LC cell, a voltage of 9 V was applied to the cell for turn-on and then removed after several seconds. The measured turn-on and turn-off times were 2.57 and 4.79 ms, respectively. In the pure LC cell, LC molecules are confined by not only the two substrates but also virtual walls. The distance between these virtual walls is approximately one quarter of the pitch of the interdigitated electrodes [23]. Therefore, the pure LC cell had a short response time despite the large cell gap of 20 μm, which is attributed to the 2D confinement effect.
In contrast to the pure LC cell, the turn-on and turn-off switchings of the 2D polymerwalled cell require in-plane and vertical electric fields, respectively. We applied a vertical voltage of 10 V to the cell for turn-off switching and in-plane voltage of 9 V for turn-on switching. The measured turn-on and turn-off times were 1.64 and 0.56 ms, respectively. The total response time of the 2D polymer-walled cell is approximately 70% shorter than that of the pure LC cell. The 2D polymer-walled cell exhibited a faster turn-on switching even at a lower applied voltage owing to the in-plane anchoring by the polymer walls. Moreover, the turn-off switching of the 2D polymer-walled cell did not rely on the slow relaxation of the LCs but was controlled by applying an electric field, which yielded the very short turn-off time. Finally, we compared the optical performance of the 2D polymer-walled cell with those of other bistable LC devices, which can be used to control the haze value, as summarized in Table 1. Ion-doped and polymer-stabilized ChLC devices based on light scattering exhibited relatively low haze values in their translucent states, and required very high voltages for switching between the transparent and translucent states. In addition, the turn-on and turn-off times were not sufficiently small for display applications. In these light-scattering devices, for a higher haze in the translucent state and faster switching, we can increase the polymer concentration, but the haze in the transparent state would also increase and the voltage for switching between the transparent and translucent states would further increase.
In contrast, a bistable SmA-LC device based on light diffraction exhibited a very high haze in the translucent state while maintaining a haze-free transparent state. Nevertheless, it required a high voltage for switching between the transparent and translucent states and had a long response time owing to the very high rotational viscosity of the SmA LC. The very narrow temperature range of the SmA LC may also be a common concern for practical applications.
On the other hand, the 2D polymer-walled cell is robust against such problems owing to the use of the nematic LC, which has already been commercialized in display applications. The 2D polymer-walled cell based on light diffraction exhibited excellent optical performances in both transparent and translucent states. Moreover, for the 2D polymer-walled cell, the voltage required for switching between the transparent and translucent states was significantly lower and the response time was significantly shorter than those of other bistable LC technologies.

Conclusion
We demonstrated a 2D polymer-walled LC device, which could be operated under a very low power as it could maintain the transparent or translucent state after the applied vertical or inplane voltage was removed. The 2D polymer walls were formed in the LC cell through UV irradiation while applying an in-plane electric field owing to the phase separation induced by the spatial elastic energy difference. The transparent [translucent] state could be maintained after the applied vertical [in-plane] field was removed. The proposed device exhibited outstanding characteristics, such as a very low operating voltage (< 10 V), low haze (< 2%) in the transparent state, high haze (> 90%) in the translucent state, and short switching time (< 2 ms), compared to those of other bistable LC devices, which can be used to control the haze value. We believe that this device is promising for power-saving smart window or window display applications.

Materials
All of the chemicals were purchased from commercial suppliers and used without further purification. In the preparation of the LC/RM mixture, we used Merck E7 as the host LC, with a dielectric anisotropy of ∆ε = 13.8 (ε ∥ = 19, ε ⊥ = 5.2), optical anisotropy of ∆n = 0.2253 (n e = 1.746, n o = 1.522 at 589 nm, 20 °C), elastic constants of K 11 = 11.1 pN, K 22 = 10.3 pN, K 33 = 17.1 pN, and rotational viscosity of γ 1 = 250 mPa·s. A UV-curable monomer RM257 containing a small amount of the photoinitiator Irgacure 651 was used in our experiment. RM257 has a rod-like structure and can be easily aligned with LCs. The LC/RM mixture was composed of 98.0% host LC and 2.0% UV-curable monomer. The LC/RM mixture was stirred for 24 h, followed by an ultrasonic-wave treatment for 2 h.

Fabrication of the 2D polymer-walled LC cell
On each glass substrate, transparent interdigitated pixel electrodes and flat common electrodes were formed with an insulating layer (oxide) sandwiched between them. The widths and gaps of the interdigitated electrodes were 2.8 and 6 μm, respectively. The thickness of the insulting layer was approximately 150 nm. A thin polyimide layer was then spin-coated onto the inner surface of each substrate and baked at 230 °C for 1 h. Subsequently, silica spacers with a diameter of 20 μm were employed to maintain the cell gap uniformity. The interdigitated electrodes were formed on both substrates, positioned at right angles to each other. Finally, the prepared LC/RM mixture was injected into the cell through capillary action. Under the applied electric field, the LC cell was exposed to the spatially uniform UV light radiation (wavelength λ = 315-400 nm) with a small intensity of 2 mW/cm 2 to polymerize the RM. The polymer walls were formed through curing of the RM under an applied electric field based on the phase separation due to the spatial elastic energy difference.

Methods and measurements
In order to confirm the spatial elastic energy distributions and phase profiles of the output light of the LC cell, numerical calculations were performed using the commercial software package TechWiz LCD 2D (Sanayi System Co., Ltd, Korea). For the numerical estimation of the LC director distributions, the Ericksen-Leslie equation coupled with the Laplace equation was solved using the finite-element method. The optical analysis was performed using the extended 2 × 2 Jones matrix method.
In order to observe the polymer structure, nonreactive LCs were removed with an organic solvent (cyclohexanone), while the polymer structure remained in the cell. We then carefully separated the two substrates of the cell after the LC removal. Optical texture observations on each separated substrate were carried out using a POM (Nikon L-UEPI, Japan) with a digital camera (Toshiba IK-637K, Japan), SEM (Model Supra 40 VP, Zeiss, Germany), and AFM (Model Nanosurf Easyscan 2, Nanosurf Inc., USA).
In order to confirm the diffraction characteristics of the fabricated cell, we measured the diffraction efficiency for the zeroth order while increasing the applied voltage. A linearly polarized He-Ne laser beam (wavelength λ = 543.5 nm) was used as the probe beam. The intensity of the zeroth order in the far-field was detected by a photodiode (Model 818-BB-21, Newport, USA) placed at a distance of 22.5 cm. During the optical measurement, the LC cell was driven by a square-wave voltage at a frequency of 1 kHz. The applied voltage was controlled by a LabVIEW (National Instruments) system.
In order to measure the optical performances of the fabricated cells, we used a haze meter (HW-65W, Murakami Color Research Laboratory, Tokyo, Japan). Specular [diffuse] transmittance T s [T d ] is the ratio of the intensity of the light that emerges from the LC cell, which is parallel (within a small angle range of 2.5°) [not parallel] with the light entering the cell, to that of the light entering the LC cell. The total transmittance T t is the sum of the specular transmittance T s and diffuse transmittance T d . The haze H can be calculated as the ratio H = T d /T t .