Polymer network liquid crystal (PNLC) was one of the most potential liquid crystal for submillisecond response phase modulation, which was possible to be applied in submillisecond response phase only spatial light modulator. But until now the light scattering when liquid crystal director was reoriented by external electric field limited its phase modulation application. Dynamic response of phase change when high voltage was applied was also not elucidated. The mechanism that determines the light scattering was studied by analyzing the polymer network morphology by SEM method. Samples were prepared by varying the polymerization temperature, UV curing intensity and polymerization time. The morphology effect on the dynamic response of phase change was studied, in which high voltage was usually applied and electro-striction effect was often induced. The experimental results indicate that the polymer network morphology was mainly characterized by cross linked single fibrils, cross linked fibril bundles or even both. Although the formation of fibril bundle usually induced large light scattering, such a polymer network could endure higher voltage. In contrast, although the formation of cross linked single fibrils induced small light scattering, such a polymer network cannot endure higher voltage. There is a tradeoff between the light scattering and high voltage endurance. The electro-optical properties such as threshold voltage and response time were taken to verify our conclusion. For future application, the monomer molecular structure, the liquid crystal solvent and the polymerization conditions should be optimized to generate optimal polymer network morphology.
© 2014 Optical Society of America
Liquid crystal has attracted growing research interests in non-display application field since the past several decades [1,2]. Due to the advantage of its low operating voltage, large birefringence, high aperture of liquid crystal devices, small power consumption, small volume and weight, phase only liquid crystal spatial light modulator was not only applied to adaptive optics but also to the field of beam steering [3,4]. However, many such applications was located in the near-infrared and mid-infrared wavelength region and usually large liquid crystal cell thickness was required to generate 2π phase change (one full wavelength). The response time will increase quadratically with increasing of the cell thickness and large cell thickness was not anticipated in real phase modulation application. To increase the response speed of liquid crystal spatial light modulators, many new submillisecond response liquid crystals were proposed, such as blue phase liquid crystal, ferroelectric liquid crystal and polymer network liquid crystal [5–10]. Among these emergence liquid crystals, polymer network liquid crystal possesses the advantage of not only submillisecond response speed but also relatively large birefringence over many other new materials . Hence polymer network liquid crystal possessed great application potential in future spatial light modulator.
However, until now the light scattering and high voltage endurance has limited its further application for phase modulation application, especially in the case when high voltage was required to activate 2π on state. The scattering was usually induced by the mismatch of refractive index in liquid crystal cell, in which the liquid crystal director was mainly aligned by the polymer network. To reduce the scattering intensity of polymer network liquid crystal, the common sense was to reduce the unit domain size of the formed polymer network according to light scattering theory. Y.H. Fan has proposed to increase the UV curing intensity to reduce the unit domain size of the polymer network . Based on this assumption, J. Sun increased the UV curing intensity to more than 200mW/cm2 to reduce the polymer network unit domain size, which was expected to be smaller than the light wavelength and induced 3% light scattering intensity at 1.06μm wavelength . High voltage endurance was also concerned. Due to formation of polymer network, the threshold was usually on the order of tens of voltage. The voltage required to generate 2π phase change was so high that electro-striction of polymer network may be induced and prolonged the response time to the order of several seconds. Although optimization of liquid crystal solvent such as viscosity and dielectric anisotropy was proposed to improve the optical performance of PNLC phase modulator, it was believed that the light scattering and dynamic response of 2π phase change was tightly associated with the polymer network morphology in PNLC phase modulator. It was well known that the polymer network morphology has been extensively studied by considering the monomer structure, monomer concentration, monomer solubility and polymerization temperature [13–19]. It was also well known that the planar fibril like rather than the rice-grain like polymer network morphology induced smaller light scattering and larger optical birefringence. However, detailed mechanism about the relationship between planar polymer network morphology and optical performance such as light scattering and high voltage endurance has not been elucidated. Additionally, the dynamic response of phase change in PNLC phase modulator was not reported to our knowledge. It was also believed to be associated with polymer network morphology. In this paper, we will focus on the optical property of PNLC phase modulator including light scattering and dynamic response of phase change. By SEM analysis, the polymer network morphology effect on the light scattering and dynamic response of phase change in PNLC phase modulator was mainly discussed.
2. Materials and experimental setup
The experimental procedure to fabricate polymer network liquid crystal phase modulator was shown as following. 7%wt RM257 and 0.5%wt photo-initiator was dissolved in the liquid crystal solvent. The selected liquid crystal solvent was HPC851900 from Jiangsu Hecheng Corporation in China. The prepared liquid crystal mixture was then filled into empty cell by capillary method. The cell gap was 14μm which was ensured by glass spacers. Fabry-Perot interference was also employed to calculate and confirm the cell thickness. A series of 10 PNLC phase modulator samples were prepared by varying the polymerization conditions, including the UV curing intensity, polymerization time and polymerization temperature. The polymerization conditions of these samples were listed in Table 1. The UV light source for UV curing was halide lamp from UVITRON International corporation.
The measurement of light scattering was essential in this study. Fiber optical spectrometer was employed to measure the light scattering intensity of the incident light. Non-polarized white light covering 700nm-1600nm wavelength region was collimated and incident directly on the PNLC samples. Collecting optical fiber was placed 40cm apart from the PNLC samples and connected with the spectrometer. Transmitted white light out of PNLC samples was collected by the collecting fiber. The collecting aperture of the fiber was 4mm. The calculated collecting angle was 10mrad. Such a configuration confirmed that almost all the scattered light was not collected and only non-scattered light was collected. A standard liquid crystal cell before UV curing and polymerization was used as normalization, in which liquid crystal solvent and monomer mixture was filled. It was worth to point out that the light scattering was dependant on the applied voltage, because the reorientation of liquid crystal director in the cell was not uniform due to the formation of polymer network. In this work, only the maximum light scattering was measured when the applied voltage go through all the voltage range that was enough to generate full 2π phase modulation. To investigate the dynamic response of phase change in PNLC phase modulator, the samples were placed between crossed polarizer. According to the V(voltage)-T(transmission) curve, the voltage required to generate 2π phase change was calculated. Photo-electric InGaSb tube detector was employed to record the dynamic response process.
The experimental method for obtaining the polymer network morphology of the PNLC phase modulator was as following. The fabricated PNLC cell was immersed in the liquid nitrogen for 3 minutes to be frozen. The frozen PNLC cell was immediately broken up by external force. In such a method, the polymer network morphology of the PNLC phase modulator was not damaged and retained its original morphology for accurate analysis. After broken up, the PNLC was immersed in acetone for 20 minutes to clear off the liquid crystal solvent and only allow the polymer network to leave behind. The dried polymer network was then coated with Au film for SEM analysis. The used SEM apparatus was Apollo 300.
3. Morphology effect on the light scattering property of PNLC phase modulators
Figure 1 gave the light scattering dispersion of PNLC phase modulator under different polymerization temperature. The UV curing intensity was 350mW/cm2 at 365nm. It was found that the maximum scattering intensity will be decreased by increasing the wavelength of incident light. Additionally the maximum scattering intensity will be increased by increasing the polymerization temperature. The maximum scattering intensity at 1.06μm was increased from 6% to 18% when the polymerization temperature was increased from 293K to 323K. Oscillations in the scattering dispersion of sample cured at 313K were induced by Fabry-Perot effect. The maximum scattering intensity was believed to be tightly associated with the morphology of each polymer network liquid crystal phase modulator.
Figure 2 gave the polymer network morphology of the samples in Fig. 1 under different polymerization temperature. As published before in some other references [13–19], the basic morphology of polymer network by RM 257 monomer was consisted of cross-linked fibril bundles and indicated planar morphology when the RM257 concentration was around 7%wt. Although the morphology of our sample was similar as shown in Fig. 2, the unit domain size and bundle diameter differed very much from each other. When the polymerization temperature was 323K, apparent fibril bundles were found which was consisted of tightly cross linked fibrils and some LC solvent. The cross linked fibrils in fibril bundle was linked so tightly that hardly void area was found. It was shown that the bundle diameter was around 250nm, and the void size between fibril bundles was around 150nm. When the polymerization temperature was decreased to 313K, the diameter of the formed fibril bundles was found to be decreased to be around 200nm and larger void size around 200nm of fibril bundles was also found. When the polymerization temperature was further decreased to be 293K, fibril bundles as large as that of sample DTemp-313 or sample DTemp-323 were hardly found. Only a few or even single fibrils were found to form the fibril bundles. As a result, the unit domain size of polymer network morphology and its void size of fibril was largely reduced. The light scattering intensity was tightly associated with the unit domain size of polymer network. The largely reduced unit domain size of sample DTemp-293 account for the smaller light scattering compared to that of sample DTemp-313 or sample DTemp-323 which was polymerized under temperature of 313K or 323K. According to Sun’s theory to express the light scattering of polymer network liquid crystal as shown in formula 1 :
Figure 3(a) gave the light scattering dispersion of PNLC phase modulator under different UV curing intensity. Figure 3(b) gave the light scattering dependence on the UV curing intensity at 1.06μm wavelength. It was indicated that the relationship between the UV curing intensity and the maximum light scattering intensity was complicated. Namely, the maximum light scattering intensity of PNLC was not proportional to the applied UV curing intensity. The diameter of the polymer network single fibril was semi-quantitatively expressed by the following Eq. (2) :Eq. (1), the maximum light scattering of PNLC phase modulator should be reduced just by increasing the UV curing intensity, which had been proved to be effective in reducing the light scattering of PNLC phase modulator [6,11]. But the hypothesis was in contrast with the experimental results shown in Fig. 3. It was indicated that a scattering peak was found when the curing intensity was around 170mW/cm2 and the light scattering intensity does not necessarily decrease with increasing the UV curing intensity. To elucidate the detailed mechanism, morphology of each sample was taken by SEM method.
Figure 4 gave the polymer network morphology of all the samples shown in Fig. 3, which were denoted as sample DInten-350 for 350mW/cm2, sample DInten-168 for 168mW/cm2, sample DInten-128 for 128mW/cm2, sample DInten-78 for 78mW/cm2, sample DInten-58 for 58mW/cm2 and sample DInten-25 for 25mW/cm2 as shown in Table 1. The relationship between polymer network morphology and UV curing intensity was also complicated as shown in Fig. 4 and Table 2, which would probably account for the complicated relationship between the light scattering and UV curing intensity. Although all the morphology was tightly associated with the single fibrils, the fibrils in these samples behaved in different manner. When the curing intensity was 25mW/cm2, coarse polymer network which was consisted of bundled fibrils (fibril bundle) were found in the SEM morphology. The diameter of the fibril bundles and the void size between them were both around 400nm. By increasing the UV curing intensity to 58mW/cm2, the fibril bundle characteristic of polymer network morphology was found to gradually disappear. Only cross-linked fibrils were left and no apparent fibril bundles were found. Similar phenomenon was also found when the UV curing intensity was 78mW/cm2, namely the fibril bundle characteristic of polymer network morphology was found to disappear. But further increase of UV curing intensity would induce formation of fibril bundles again, for example the fibril bundle characteristic was apparent when the curing intensity was 168mW/cm2 as shown in Fig. 4(b). The fibril bundle diameter was comparable to that of sample DInten-25, but the formed fibril bundles were not as dense as that of sample DInten-25 and were consisted of more finer single fibrils. Although fibril bundle characteristics were also found when the UV curing intensity was 128mW/cm2, the fibril bundles were so loose that the morphology exhibit a cross linked fibrils and the increase of light scattering was hence ignorable. From the above result, it was concluded that the formation of apparent fibril bundles usually induced larger unit domain size of polymer network and hence larger maximum light scattering intensity. In contrast, the formation of only cross linked single fibrils would induce smaller unit domain size of polymer network morphology and hence smaller light scattering intensity was induced. These conclusions were verified by the above experimental result.
The polymer network morphology dependence on the UV curing intensity was also interesting. It was reported that the competition between the diffusion limitation and the reaction limitation determine the final polymer network morphology, which was the physical origin of Eq. (2). It seemed that the reaction limitation may include the interaction between the fibrils or between the fibril bundles, which compete with each other and was responsible for the complicated morphology dependence on the UV curing intensity.
Figure 5 gave the light scattering dispersion of PNLC phase modulator under different UV curing time. With the curing time increasing from 20min to 60min, the maximum light scattering intensity at 1.064μm was decreased from 15% to 9%. The polymer network morphology was also shown Fig. 6.It was indicated that the formed fibrils grew denser by increasing the UV curing time which was also shown in Table 2. The polymer network morphology exhibited both cross linked single fibril and fibril bundle characteristics when the UV curing time was 20min. The looser fibrils and fibril bundle characteristic in the polymer network morphology took the main charge in the larger light scattering in sample DTime-20 compared to that of sample DTime-60. It was expected that the fibril bundle diameter of the polymer network would be reduced by increasing the UV curing time because of the further polymerization between the formed fibril bundles. Correspondingly the polymer network morphology exhibited more cross linked single fibril characteristics. The evolution of the polymer network when increasing UV curing time would induce smaller unit domain size of polymer network and corresponding smaller light scattering intensity. The experimental results and the expectation were in accordance with our conclusion in the above discussion.
The above results indicated that the light scattering intensity of polymer network liquid crystal phase modulator was mainly determined by the polymer network morphology, namely the unit domain size and void size between them. Not only cross linked single fibril but also bundled fibrils were found to construct the polymer network morphology. The formation of fibril bundles will increase the light scattering. In another aspect, not only the molecular structure and LC solvent determine the morphology, but also the polymerization conditions take main charge on the polymer network morphology. Increasing the reaction rate by increasing the curing intensity will increase the interaction rate both between fibrils and fibril bundles. The competition among these mechanisms will determine the final formed polymer network morphology in PNLC phase modulators.
The optical phase delay dependence on the applied voltage was also measured. Figure 7 gave the phase delay dependence on the applied voltage for samples with different curing intensity. It was indicated that the threshold voltage of PNLC phase modulator was also not necessarily dependant on the UV curing intensity. We attributed it to the morphology dependence on the UV curing intensity. The relationship between the PNLC morphology and threshold voltage were concluded in Table 2. It was believed that the smaller the polymer network unit domain size was, the larger the threshold voltage was . The threshold voltage behavior also verified our discussion on the polymer network morphology characteristics.
4. Dynamic response of phase change in PNLC phase modulators and its morphology effect
Dynamic response of phase change in PNLC was an important indicator for real phase modulator. The PNLC phase modulator should take advantage of both fast response and low light scattering. Figure 8 gave the on state response of π phase change and 2π phase change in PNLC phase modulator whose light scattering was only 9%. It was indicated that the response time of π phase change as shown in Fig. 8(a) was 120μs for 10%-90% phase change of the whole phase modulation. But only 1.24π phase change was observed during the first 20ms as shown in Fig. 8(b) when the applied signal voltage was so chosen that the generated phase change would be 2π. A longer time window of 2 seconds scale was chosen to observe the dynamic response of 2π phase change. The dynamic response of 2π phase change was physically divided into two processes, namely fast response process on the order of submillisecond and slow response process on the order of several seconds. It was indicated that the residual phase delay was gradually increased to 2π on the time scale of several seconds after fast response process. The slow response process deteriorated the fast response property of PNLC phase modulator which does not make sense in achieving fast response by using PNLC phase modulator.
The slow response process of PNLC was attributed to the applied high voltage which was responsible for electro-striction effect. Phase ripple was also found as shown in the inset of Fig. 8(c). The response of the liquid crystal director reorientation in the polymer network was so fast that the phase change will follow the fast change of the applied rectangular voltage signal. To prove this conclusion on the phase ripple phenomenon, the response of another sample with 3ms response speed was also shown in Fig. 8(d). It was indicated that the phase ripple was suppressed greatly because the reorientation of liquid crystal directors in polymer network could not follow the fast change of the applied rectangular voltage. To verify the electro-striction effect of polymer network, a higher voltage up to more than 200V was applied on the above PNLC cell. At the initial stage, the transparent PNLC cell in the visible range become opaque at once as the high voltage was applied. But after tens of seconds, the opaque PNLC cell gradually turned out to be transparent. A similar phenomenon would also be found when another high voltage was applied after removal of the former high voltage. We attribute this phenomenon to the above mentioned electro-striction of the polymer network. It was reasonable to predict that the phase change achieved in the fast response process determined the extent of electro-striction effect.
As mentioned above, it was probable that the electro striction effect may have slowed the response speed of PNLC phase modulator and was the physical origin of the slow process shown in Fig. 8(c). It was well known that higher dielectric anisotropy was optimal to reduce the threshold voltage and avoid the occurrence of electro-striction effect. But in some case, the liquid crystal with large dielectric anisotropy was not stable under UV curing. As discussed in last section, the morphology determined the light scattering of polymer network liquid crystal. It was also probably that the morphology was related tightly with the electro-striction effect. According to this hypothesis, it was possible to control the electro-striction effect by controlling the polymerization condition in polymer network liquid crystal.
Figure 9 gave the dynamic response of on state 2π phase change in PNLC phase modulators under different polymerization conditions. It was indicated in Fig. 9(a) that PNLC phase modulator cured at 293K achieved only 1.6π phase change in the fast response process of 2π on state process compared to that of 1.8π under higher curing temperature of 313K and 323K. The morphology of these samples in Fig. 2 may account for this phenomenon. The fibril bundles in sample DTemp-313 and DTemp-323 would help to reduce electro-striction effect. Electro-striction effect of polymer network in sample DTemp-293 was easily induced because of the formation of cross linked fibrils, which was responsible for the smaller phase change achieved in the fast response process. Bundles of fibrils were more stable compared to the single cross linked fibrils. Similar dynamic response in the samples with different UV curing intensity was also found as shown in Fig. 9(b). Sample DInten-168 and sample DInten-25 has achieved their maximum phase change after fast response process, namely 1.8π and 1.7π phase change separately. But the other samples have only obtained 1.25π to 1.46π phase change after fast response. Usually several seconds time was required to obtain the residual phase to achieve full 2π phase transition of on state 2π phase change. As shown in Fig. 4, the polymer network morphology in sample DInten-168 and sample DInten-25 exhibit more fibril bundle characteristics in which fibril bundles were cross linked, while the others of sample DInten-350, sample DInten-128, DInten-78 and DInten-58 exhibit more single fibril characteristics in which only fibrils but not fibril bundles were cross linked and the cross linked fibrils made up of polymer network. The cross linked fibril bundles were usually able to endure higher electric field to avoid electro-striction effect compared to that of cross linked single fibrils.
The dynamic response of on state 2π phase change in sample DTime-20 and sample DTime-60 was also taken as shown in Fig. 9(c). It was indicated that sample DTime-20 has achieved full 2π phase change in the fast response process. But sample DTime-60 only achieved 1.75π in the fast response process. It was indicated in Fig. 6. that the polymer network morphology in sample DTime-20 exhibit more fibril bundle characteristics compared to that in sample DTime-60, which accounted for the higher achieved phase change in the fast response process. These experimental results indicated that the PNLC phase modulators with more fibril bundle characteristics would be able to endure higher voltage and smaller electro-striction effect would be induced.
The response time of fast response process would be additive verification of the morphology characteristic as shown in Table 2. Especially the response time gave more information on the void size of the polymer network morphology. It was concluded from the response time of the fast response process that the experimental SEM results were convincible. However, due to coupling of the fast response process and the slow response process, the response time of the fast response process could be accurately measured.
The polymer network morphology of PNLC phase modulator was mainly characterized by cross linked single fibrils, their bundles or both. By varying the polymerization conditions such as polymerization temperature, UV curing intensity or polymerization time, the polymer network morphology could be controlled, which exhibited cross linked fibrils characteristics or cross linked fibril bundles characteristics. The interaction between fibrils or between fibril bundles would influence the final polymer network morphology. The formation of cross linked fibrils usually induced smaller unit domain size and smaller light scattering. But the high voltage endurance was poor. In contrast, the formation of cross linked fibril bundles or larger polymer network diameter possessed the advantage of higher voltage endurance, but usually induced larger unit domain size and higher light scattering. Additionally, the light scattering was not only influenced by the unit domain size but also by the void size between them. There was a tradeoff between the light scattering and high voltage endurance considering the morphology effect.
This work was supported by Natural Science Foundation of China(61378095), the Foundation of Science and Technology Development, the China Academy of Engineering Physics (2012B0401055), and the Development Foundation of Institute of Fluid Physics (SFZ20120304).
References and links
1. H. D. Tholl, “Novel laser beam steering techniques,” Proc. SPIE 6397, 639708 (2006). [CrossRef]
2. J. Gibson, B. Duncan, P. J. Bos, and V. Sergen, “Wide angle beam steering for infrared countermeasures application,” Proc. SPIE 4723, 100–111 (2002). [CrossRef]
3. B. Duncan, P. J. Bos, and V. Sergan, “Wide angle achromatic prism beam steering for infrared countermeasure application,” Opt. Eng. 42(4), 1038–1047 (2003). [CrossRef]
4. H. Hu, L. Hu, Z. H. Peng, Q. Q. Mu, X. Y. Zhang, C. Liu, and L. Xuan, “Advanced single-frame overdriving for liquid-crystal spatial light modulators,” Opt. Lett. 37(16), 3324–3326 (2012). [CrossRef] [PubMed]
6. J. Sun, H. Q. Xianyu, Y. Chen, and S.-T. Wu, “Submillisecond response polymer network liquid crystal phase modulators at 1064nm wavelength,” Appl. Phys. Lett. 99(2), 021106 (2011). [CrossRef]
7. D. Engström, M. J. O’Callaghan, C. Walker, and M. A. Handschy, “Fast beam steering with a ferroelectric-liquid-crystal optical phased array,” Appl. Opt. 48(9), 1721–1726 (2009). [CrossRef] [PubMed]
10. K. M. Chen, S. Gauza, H. Q. Xianyu, and S. T. Wu, “Submillisecond gray level response time of a polymer stabilized blue phase liquid crystal,” Displ. Tech. Lett. 6(2), 49–51 (2010). [CrossRef]
12. Y. H. Fan, Y. H. Lin, H. W. Ren, S. Gauza, and S. T. Wu, “Fast response and scattering free polymer network liquid crystals for infrared light modulators,” Appl. Phys. Lett. 84(8), 1233–1235 (2004). [CrossRef]
13. I. Dierking, “Recent developments in polymer stabilized liquid crystals,” Polym. Chem. 1(8), 1153–1159 (2010). [CrossRef]
14. C. V. Rajaram, S. D. Hudson, and L. C. Chien, “Morphology of polymer stabilized liquid crystals,” Chem. Mater. 7(12), 2300–2308 (1995). [CrossRef]
15. Y. H. Yin, W. Li, H. Cao, J. B. Guo, B. F. Li, S. M. He, C. B. Ouyang, M. Cao, H. Huang, and H. Yang, “Effects of monomer structure on the morphology of polymer network and the electro-optical property of reverse mode polymer stabilized cholesteric texture,” J. Appl. Polym. Sci. 111(3), 1353–1357 (2009). [CrossRef]
16. I. Dierking, L. L. Kosbar, A. Afzali, and A. C. Lowe, “Network morphology of polymer stabilized liquid crystals,” Appl. Phys. Lett. 71(17), 2454–2456 (1997). [CrossRef]
17. A. S. Sonin and N. A. Churochkina, “Liquid crystals stabilized by polymer networks,” Polym. Sci. Ser. A 52(5), 463–482 (2010). [CrossRef]
18. I. Dierking, “Polymer network stabilized liquid crystals,” Adv. Mater. 12(3), 167–181 (2000). [CrossRef]
19. C. V. Rajaram, S. D. Hudson, and L. C. Chien, “Effects of polymerization temperature on the morphology and electrooptic properties of polymer stabilized liquid crystals,” Chem. Mater. 8(10), 2451–2460 (1996). [CrossRef]
20. R. Q. Ma and D. K. Yang, “Freedericksz transition in polymer-stabilized nematic liquid crystals,” Phys. Rev. E Stat. Phys. Plasmas Fluids Relat. Interdiscip. Topics 61(2), 1567–1573 (2000). [CrossRef] [PubMed]