1. Introduction

Holographic Polymer-Dispersed Liquid Crystals (HPDLCs) are diffraction gratings made of slices of pure polymer alternated to films of polymer containing droplets of Nematic Liquid Crystals (NLC), representing good materials for many applications ranging from switchable windows to projection displays [1]. In particular, these structures are the focus of extensive research in the display industry. Devices, in spite of their good diffraction properties and low costs, yet exhibit some intrinsic drawbacks [2]. If the droplet size of the NLC component inside the polymer matrix is comparable with the wavelength of the impinging light, a strong scattering from the sample is observed. The effect can be reduced by reducing the droplet size, but this yields to an increasing in the voltage values needed to control the grating [3]. Furthermore, it has been shown that, when an external field (temperature, electric field, etc.) is used to drive a generic structure containing NLC droplets, (PDLC), its optical response depends, in fact, on the history of the sample. There are different evidences of this phenomenon [4]: Hysteresis, observed when the transmission at a given applied voltage depends on the previous voltage value. Persistence, which is found when the sample takes several seconds to relax to its original condition after the field is turned off. Memory effects, which affects a sample that exhibits a semi-permanent persistence. All these effects have been reported by several authors [5, 6], who studied the correlation between droplets size, morphology, and memory effects in the electro optical response of devices. Some years ago, a new kind of holographic grating has been realized in composite materials, which consists of films of regularly aligned liquid crystals separated by uniform polymer slices (POLICRYPS) [7]. Fabrication of devices is carried out by utilizing a very well stabilized optical setup [8], which ensures a uniform and sharp morphology of the realized samples. Moreover, a detailed investigation showing the good optical and electro-optical characteristics of these structures has been presented [9]. Utilization of POLICRYPS structures for applications has been also considered, and it turned out that investigation of these structures is of interest not only in the field of switchable diffraction gratings, but also for switchable micro-lasers, switchable optical phase modulators, switchable Bragg filters [10], and display applications [11]. In all cases, the overall quality of the device depends on a good knowledge of the electro-optical response of the POLICRYPS structure, which represents the “core” of the device itself. In particular, where display applications are concerned, a gray scale reproducibility is one of the key parameters. In this paper we show that the presence of memory effects in POLICRYPS gratings can result, in fact, negligible due to the absence of NLC droplets. By ramping up and down the sample with an applied voltage, it is possible to define hysteresis as the maximum difference between the upward and downward transmittance response, expressed as a percentage of the whole transmission range. We show that hysteresis measured in this way depends on the rate fields are applied and removed. Since minimization of hysteresis is an important goal for almost all applications, we have performed an electro-optical characterization of POLICRYPS gratings, by detecting the switching curves for the applying and removing cases of the external electric field in different experimental conditions. The relationship between the hysteresis area and the delay time (time between two consecutive applied voltages) has been analyzed, by comparing the area obtained by using a “manually driven method” and a “software driven method”, where the delay time can be finely controlled. Dependence of this area on the frequency of the applied electric field has been also analyzed.

2. Hysteresis observation

POLICRYPS is a structure obtained by following a well defined sequence of steps [7]. An empty cell, made of two indium-tin-oxide (ITO) glass plates, is filled with a homogeneous mixture of BL001 Nematic Liquid Crystal (by Merck) and NOA 61 pre-polymer (by Norland). The grating structure is then obtained by means of the typical setup for grating fabrication described in details elsewhere [7], implemented with a new technique for high polymerization process stability [8]. For the realized grating, calculation of the Bragg parameter gives ρ = Λ²/λL = 0.28, where L = 8.5 μm is the cell gap, Λ = 1.22 μm is the fringe spacing and λ = 0.633 μm represents the probe wavelength; the value of ρ indicates that we are operating in Bragg grating conditions [12]. For our experiments, we have arranged the set-up shown in Fig. 1. The sample temperature is kept constant at room conditions (25 °C) during all the experiments. To control the impinging intensity, the focalized light (spot diameter 0.5 mm) from a He-Ne laser (λ=0.633 μm) passes through a variable attenuator, made of a half waveplate (HWP) and a linear polarizer. Then, the incident polarization can be changed by using a second HWP. For POLICRYPS, a strong correlation exists between the diffraction efficiency and the polarization of the impinging beam. Therefore, in order to maximize the index grating contrast (and then the diffraction efficiency [13]), a TE polarization direction is fixed; in this case the light experiences the difference between the NLC extraordinary index (up to n_e ~ 1.7) and the polymer one (n_p ~ 1.54). Diffracted and transmitted beams are detected by two photodiodes. The electro-optical response is investigated by exploiting a low frequency (v = 1KHz) square-wave voltage applied across the ITO electrodes of the cell. In order to avoid application of a bi-polar field, an offset (DC voltage) component equal to half of the amplitude value (Vpp) is added in all experiments. Preliminary measurements have been performed by varying manually the duration (τ) of the single train of square waves. Results have indicated that, the hysteresis value depends on the value of τ, and tends to increase when τ is increased. A typical result is shown in Fig. 2, where the applied voltage in increased along the lower curve and decreased along the upper one. In this case we have a quite high hysteresis (about 34%) obtained for a τ value of about 20s. By switching off the voltage, a small persistence is still present, resulting in different transmittance values. This difference can be reduced by waiting for few minutes.

 

Fig. 1. Experimental setup for the electro-optical characterization of POLICRYPS grating. P polarizer, HWP half wave plate, 0T and 1T zero and first diffraction order beams. The POLICRYPS image refers to a typical sample observed at the optical microscope in crossed polarizer condition.

Download Full Size | PDF

 

Fig. 2. Hystersis curve of the electro-optical response of a POLICRYPS grating (delay time τ manually set at about 20s)

Download Full Size | PDF

3. Hysteresis characterization

We have performed a detailed characterization of the observed hysteresis effect by varying the delay time τ in a controlled way. We have realized a Labview based virtual instrument (VI) which enables to remote control the generation of external voltages by means of a GPIB connection; in this way, we can set the waveform shape, along with output impedance, frequency, and delay time τ. The software is realized in such a way that each step of the applied voltage ramp is made up of a series of square waves of time duration τ, which can be set in advance; also the amplitude of the ramp and the starting point can be easily selected. A typical waveform output is shown in Fig. 3, where inset (a) represents a particular of the applied square wave step while inset (b) shows the diffracted and transmitted intensity. Transmittance values reported in Fig. 4(a) were obtained by increasing (decreasing) the applied voltage from 0 to 60 V (60 to 0V) in steps of 3V with a delay time τ = 2 sec between two consecutive steps; in this case we find a hysteresis percentage area of 12%. Figure 4(b) and 4(c) show that, by lowering the delay time to 500 and 100 ms, the hysteresis decreases to 9% and 5% respectively, with an 85% reduction of the first measured value reported in Fig. 2. The observed effect can be attributed to the following reason: It is known that the degree of orientational order induced in the NLC molecular director by boundary conditions is quite high only in the proximity of the boundaries (polymer walls). In the bulk, only the action of an external applied field can induce a high order degree; this increases monotonically with the amplitude of the applied field and the time the field acts on the sample. As a consequence, the average degree of orientational order during the increasing voltage ramp increases by starting from the low one induced only by boundaries and, for each field value, it reaches levels that are higher the longer is the time interval τ the field is applied; on the contrary, during the decreasing voltage ramp, the order decreases by starting from the high level reached at the end of the increasing ramp. Therefore, it is evident that the average order degree of the NLC molecules necessarily depends on the history of the sample as well as on the value of τ. This hysteresis of the order degrees in the used experimental geometry yields an hysteresis in the electro-optical response of the sample.

 

Fig. 3. Typical applied waveform for hysteresis characterization.

Download Full Size | PDF

The zero field values in the transmission curves are not exactly the same for all measurement; the reason is in the high numbers of electro-optical cycles which have been realized on the sample, since in each cycle the sample retains some memory of the previous one. In order to avoid this effect, before each cycle, the sample should be heated above the isotropic temperature (to completely remove the memory) and cooled down to room temperature. We have not followed this procedure, since our aim was only to understand if, and in which way, the hysteresis is affected by the delay time τ. Anyway, no relevant thermal hysteresis was observed by detecting the transmittance variations versus temperature. At the same time, we have assumed that hysteresis effects due to the deformation of the polymer slices are negligible, due to the low shrinkage of the used NOA61 pre-polymer (less than 3–4%, as given by Norland). A remarkable characteristics of POLICRYPS gratings is the presence of free charges inside the mixture. Conducting effects due to charge transport are well known in conventional liquid crystal devices [14] and become more complex in composite materials. Migration of free ions in POLICRYPS structures induces the rise of an internal field which is opposite to the external one, thus causing an effective field decreasing across the NLC sample. However, the free ion migration inside the sample is different in the increasing and decreasing ramps and can be hypothesized as a further cause which contributes to the hysteresis effect. Since use of driving signals at high frequency prevents ion migration within the cell, we have measured the transmittance by varying that frequency. Figure 5 shows the behavior of hysteresis area versus the delay time for three different frequency values. Hysteresis increases with the delay time, as well as with the frequency (due to the ions’ effect). Anyway, a low frequency voltage increases the effect of the internal field created by migrated ions; this induces an increasing of the switching voltages thus affecting the possibility of applications.

 

Fig. 4. Hystersis curve of the electro-optical response of a POLICRYPS grating. (a) delay time τ= 2s - (b) delay time τ= 500ms - (c) delay time τ= 100ms

Download Full Size | PDF

 

Fig. 5. Hysteresis area versus the delay time for different frequency values of the external applied voltage.

Download Full Size | PDF

4. Conclusion

In conclusion, we have investigated the hysteresis effect in the electro-optical response of a POLICRYPS grating. We have characterized the hysteresis area as a function of the delay time between two consecutive steps of different applied voltage; by means of a “home made” Lab-View code, the hysteresis has been lowered to values of same few percentages. Dependence of hysteresis area on the frequency of the external applied voltage has shown that effects of migration of free ions in the sample are not negligible.