Open Access
Add to BibSonomy
Abstract
We report the design, fabrication, and characterization of an optically switchable polarizing beam splitter with a prism/azobenzene liquid crystal/prism hybrid structure. The beam splitter can operate in the polarization-splitting mode and the non-splitting mode. The switching between the modes is realized by the photoisomerization-induced phase transitions in the azobenzene liquid crystal, featuring all-optical control, bistability, and fast response. Such an active polarization-handling element is highly desirable as it not only simplifies and compacts sophisticated optical systems but also increases the degree of freedom in optical circuit design.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
For years, polarization-handling devices are of great and increasing importance in a variety of fields spanning from optics [1] to biology [2], chemistry [3], and atmospheric science [4]. The polarizing beam splitter (PBS) is one of such devices that separates two orthogonal polarization states (S and P polarizations in most cases) into different propagation directions [5–8]. It has been widely used in optical signal processing, such as demultiplexing [9], logic operations [10,11], polarizing interferometry [12], and further applied to quantum computing [10,13,14].
Conventional PBSs separate optical polarizations by creating anisotropy in refraction or reflection. They are generally grouped into three categories [5]—Cat.-1: refractive PBS based on birefringent prisms (e.g., Wollaston prism [15]), Cat.-2: reflective PBS based on Brewster windows (e.g., MacNeille cube polarizers [16,17]), and Cat.-3: reflective PBS based on birefringent prisms (e.g., Glan–Taylor prism [18]). Refractive PBSs (Cat.-1) usually require long interaction lengths to acquire noticeable separation of the two polarized beams (e.g., centimeters-long crystal for millimeter separation), owing to the inherently small birefringence of the crystal employed. Brewster type PBSs (Cat.-2) need multiple layers to achieve full polarization splitting, and so the reflected beam is easily spread out. The Cat.-3 PBSs exploit the anisotropy in the critical angle for total internal reflection (TIR) at the interface between the first birefringent crystal and the isotropic dielectric. By properly selecting the angle of incidence, one polarization is totally reflected while the other is mostly transmitted. The split angle can be very large, so that the required interaction length is much shorter than the Cat.-1 PBSs. Besides the traditional PBSs, polarization splitting elements based on directional coupler [19], multimode-interference coupler [20], photonic-crystal heterointerface [21], and dielectric metamaterial [7] have also been developed, mainly for silicon photonics applications.
To date, most of the PBSs are made passive; that is, the polarization-splitting properties are fixed once they are fabricated. PBS can be more flexible and functional by enabling all-optical, dynamic, and multistable switching. Here we propose an active PBS based on a prism-bound azobenzene liquid crystal (azo-LC), of which the polarization-splitting function can be controlled by light-induced phase transition of the liquid crystal (LC). As will be demonstrated, the switching is optically-driven, bistable, and high-speed (milliseconds). The interaction length is of the same scale as the beam size, making it potentially feasible for on-chip circuitry. Moreover, the PBS is highly reliable, of which the quality of the performance remains almost unchanged throughout 100 cycles of consecutive switching.
2. Working principle and design
The all-optical PBS is made by sandwiching an azo-LC that consists of 80 wt% 5CB (HCCH) and 20 wt% 1205 (BEAM Co.) between two surface-treated N-SF11 equilateral prisms (Thorlabs). 5CB is a rodlike photo-insensitive nematic LC, and 1205 is a mesogenic azobenzene [22] having two isomers: a rodlike trans form and a boomerang-shaped cis form. Like most azobenzenes, the trans-1205 is thermodynamically more stable than the cis form. 1205 can be photoisomerized from the trans to cis state under ultraviolet or violet exposure, and the reverse isomerization can be activated by using longer-wavelength visible light. In this work, we use a green laser at λ = 532 nm to induce the trans–cis isomerization and a violet laser at λ = 404 nm to induce the cis–trans isomerization. When driving the azo-LC to the cis-rich state, the nematic–isotropic transition temperature drops from 43.9°C to 13.0°C. This is due to the fact that the structural incongruity between the cis-1205 and 5CB molecules causes a drop in the bulk orientational order of the azo-LC. Therefore, within the 30°C range (43.9–13.0°C), the azo-LC can be isothermally switched between the nematic and isotropic phases by light. Note that the upper and lower limits of the operating temperature are determined by the clearing points of the azo-LC in the two photostationary states and can be tuned by adjusting the mixing ratio of 1205 to 5CB. Furthermore, once the cis-rich state is reached, the state can be kept for hours upon the removal of the optical pump, owing to the slow spontaneous cis–trans relaxation of 1205.
The nematic and isotropic phases of the azo-LC correspond to the polarization-splitting and non-splitting modes of the PBS, respectively. In the polarization-splitting mode (PSM), S- and P-polarized light are split into two propagation directions—one being reflected and the other being transmitted. In the non-splitting mode (NSM), the S and P polarizations are both reflected. The polarization-splitting function is designed to perform in the nematic phase of the azo-LC, whereas the NSM is switched on when the LC is transitioned to the isotropic phase [Fig. 1].
To realize such operation, the incident angle of the probe beam at the air-to-PBS interface, θA/P in Fig. 2(a), is critical. The proposed PBS is based on TIR at an optically anisotropic interface. Instead of using birefringent prisms like a Cat.-3 PBS, we choose to make the dielectric layer birefringent with a LC. TIR takes place when an optical beam travels from a high-index medium (with a refractive index ni) to a low-index medium (nt) at an incident angle larger than the critical angle θc = sin–1(nt/ni). In the nematic phase, the azo-LC (ne = 1.713 and no = 1.529 at λ = 633 nm) is aligned to the surface normal of the two high-index prisms (nP ≈1.78). To achieve the polarization splitting, an incident angle at the prism-to-LC interface (θP/LC) is chosen so that the S component of the incident beam is totally reflected, while the P component is transmitted through the LC:
The θc(S) and θc(P) are determined by the ordinary index (no) and effective index (neff) of the LC, respectively. The effective index is dependent on the angle between the beam and the director axis of the LC (θLC):In the isotropic liquid phase, the refractive index becomes nISO = 1.588 regardless of the polarization state. The beam could therefore be completely reflected or mostly transmitted depending on the θP/LC relative to θc,ISO.
Fig. 2 (a) Device configuration and definitions of parameters. Simulated transmittance and reflectance of the PBS as functions of the incident angle for azo-LC in the (b) nematic and (c) isotropic phases, respectively. The superscript shows the polarization state.
To find the operating range of θA/P, the angular dependences of transmittance (Tout) and reflectance (Rout) for S-wave and P-wave are simulated, as shown in Figs. 2(b) and 2(c). Tout (Rout) is defined as the ratio of the output to input intensity with the output beam at the other (same) side of the input with respect to the LC layer. Propagating in the PBS, light experiences reflection and/or refraction at every interface. The beam that eventually leaves the PBS by transmission encounters four interfaces, while the reflected one encounters three. Hence,
where the superscript (j) stands for the polarization state (S or P), the T and R on the right-hand side denote the transmittance and reflectance at each interface, respectively, and the subscripts are the interface indices. The T and R are calculated using the following formulae [23]: Figure 2(b) reveals that, when the LC layer is optically anisotropic (i.e., nematic), the S-wave and P-wave are fully split into the reflected and transmitted beams, respectively, for 0° < θA/P < 26°. Upon switching to the isotropic state [Fig. 2(c)], regardless of the probe polarization, the light is completely reflected at an incident angle between 6° and 62° because both the S- and P-waves experience TIR at the prism-to-LC interface. Hence, the θA/P that supports reversible PSM↔NSM switching ranges from 6° to 26°. However, Tout(P) and Rout(S) drop as the angle of incidence θA/P is increased. We thus set θA/P ≈8° in the following experiments, so that the PBS can support high output intensities and tolerate an angular error of 2°. Moreover, the split angles are 51° in the PBS (ψP) and 44° in the air (outside the PBS; ψA) at θA/P = 8°. To achieve full spatial separation of 1 mm-thick S- and P-polarized beams, the required length L is ~2.32 mm (see Fig. 2(a) for the definitions of ψP, ψA, and L).3. Experimental demonstration
For a proof-of-concept demonstration, the proposed PBS was fabricated by filling the azo-LC into an air gap of ~5 μm between two N-SF11 prisms via capillary action. The inner surfaces of the prism cell were pretreated with DMOAP (Sigma-Aldrich) coating to enable the homeotropic alignment of the azo-LC. It is noteworthy that the cell gap has to be larger than the penetration depth of a few 100’s nm, d > λ [4π(nP2sin2θP/LC − nLC2)]−1, but the gap uniformity is not critical because the operating mechanism is based on TIR. Figure 3(a) depicts the experimental setup used in this work. To analyze the polarizations of the transmitted and reflected light, a 633 nm red laser was used as a probe, of which the incident polarization was determined by two linear polarizers and a quarter-wave plate in between. The azo-LC PBS was placed on a rotation stage for the ease of controlling the angle of incidence. A 404 nm violet laser and a 532 nm green laser were employed for PSM→NSM and NSM→PSM switching, respectively. Rotatable polarizers, photodetectors, and power meters were applied to characterize the output polarizations. Figures 3(b) and 3(c) reveal that a probe polarized at 45° can be split into P- and S-polarized outputs when operating the azo-LC in the nematic phase, whereas the incident polarization was well preserved and coupled out through total reflection upon the all-optical switching to the isotropic liquid state. In the PSM [Fig. 3(b)], the Tout for P- and S-waves were 79.0% and 0.1%, respectively, while the Rout of P-wave was 3.9% and that of S-wave was 80.9%. Therefore, the polarization extinction ratios—PER = 10 log(P1/P0), where P1/P0 is the power ratio of the orthogonal polarizations—were 29.0 dB for the P-polarization (P1/P0 = Tout(P)/Tout(S)) and 13.2 dB for the S-polarization (P1/P0 = Rout(S)/Rout(P)), which are adequate for many applications. By contrast, in the NSM [Fig. 3(c)], the Rout of P- and S-waves were both as high as 80.8%, and so the PER was ~0. It is noteworthy that similar results were also obtained with incident light polarized at –45° as well as circularly polarized inputs.
Fig. 3 (a) Schematic of the experimental setup. Polar plots of measured transmittance and reflectance for azo-LC in the (b) nematic and (c) isotropic phases, respectively.
The switching dynamics of the azo-LC PBS were also monitored. Figures 4(a) and 4(b) depict that the P-wave can be switched from transmission to reflection in ~3 ms under violet excitation at Iviolet ≈1.5 W/cm2, while the S-wave remains transmitted. The reverse switching from the NSM to the PSM was achieved in ~120 ms with the green pump at Igreen ≈2.5 W/cm2 [Figs. 4(c) and 4(d)]. In a typical guest-host mixture of azobenzene and photostable LCs, the switching speed is governed by two processes: photoisomerization of the guest azobenzene and the successive orientational coupling to the photostable LC host. For the isothermal nematic–isotropic switching (PSM→NSM), the response time for the former process (τ) is inversely proportional to the pump intensity (I), absorption constant (α), and quantum efficiency (q): τ ~(α⋅q⋅I)–1 [24]. The latter process stems from the structural incongruity between the boomerang-shaped cis-1205 and rodlike 5CB molecules [25]. Upon photoisomerization, the intermolecular torque of a cis-azobenzene acting on the nematic host randomizes the LC director, of which the time constant is typically on the order of milliseconds. The response time for the isotropic–nematic switching (NSM→PSM) is on a similar time scale but may be retarded by the cis–trans photoisomerization process, owning to the weaker absorption in green (τ ~1/α), cf. Fig. 4(e). For applications that require fast switching, nanosecond response could be achieved by employing a nematic azobenzene solely (instead of an azobenzene/photostable LC mixture) with pulsed excitation to eliminate the slow orientational coupling process [26].
Fig. 4 Time-resolved transmittance (Tout) and reflectance (Rout) of the PBS experiencing switching from polarization-splitting mode (PSM) to non-splitting mode (NSM) and the reverse, respectively: (a) P-wave / PSM→NSM, (b) S-wave / PSM→NSM, (c) P-wave / NSM→PSM, and (d) S-wave / NSM→PSM. (e) Absorption spectra of 1205 in the trans and cis states, respectively. (f) Dynamics of Tout and Rout in the 100th switching cycle: PSM→NSM (top), NSM→PSM (bottom).
The reliability of the fabricated PBS has been examined. Comparing Fig. 4(f) to Figs. 4(a) and 4(c), it reveals that the quality of optical performance is maintained throughout 100 cycles of consecutive switching. This is attributed to the azo-LC’s remarkable tolerance to high pump intensity. In Ref [26], 532 nm, 8 ns-long laser pulses with energy densities up to ~0.9 J/cm2 were chosen to drive azo-LCs similar to 1205. The maximum peak intensity therein is on the order of 100’s MW/cm2, which is significantly higher than that used in the present study, ~2.5 W/cm2.
The switching between the PSM and NSM is not only fast but also ‘nonvolatile’, a feature which is of crucial importance in reducing energy consumption and circumventing problems caused by instabilities of the power supply. Both the polarization-splitting and non-splitting states are fairly stable. To be precise, the trans form of 1205 is more stable than the cis counterpart. The corresponding PSM is hence permanent, whereas the cis-rich NSM is less stable. Nonetheless, as depicted in Fig. 5, the NSM can be retained for several hours at room temperature (25°C) without the aid of the external excitation, due to the long-lived cis-isomer of 1205. To get insight about the cis–trans spontaneous relaxation process, we have also examined the evolution of the azo-LC’s clearing temperature (Tc) during the relaxation (thermal isomerization) process. The Tc dropped from ~44°C to ~13°C under violet irradiation. Upon three hours of dark relaxation, the clearing point rose to ~18°C, and continued to increase to ~24°C by another four hours. The PBS eventually self-transitioned back from the NSM to the PSM. A metastable state lifetime of more than 7 hours long is adequate for many bistable operation purposes. The stability can be further improved through material design or a clever choice of the operating temperature.
Fig. 5 Field-off stability of the non-splitting mode in ~3.4 hours upon 1 s-long exposure to 1 W/cm2 violet light, demonstrated using a P-polarized probe. The arrow points out the time at which the PSM→NSM switching was driven.
Compared to the state-of-the-art technology, the PERs in the present work, 29.0 dB for P-wave and 13.2 dB for S-wave, are comparable with the electrically switchable interferometric PBSs (~15 dB) [27] and the passive photonic-crystal PBSs (~23 dB for TE wave and ~18 dB for TM wave) [28]. The inevitable escape of P-wave via reflection at the prism–LC interface, ~3.9%, sets the upper limit of the PER of the reflected beam; and, this holds true for most TIR-based PBSs, such as the Glan–Taylor prism. The insertion loss of our PBS is ~1.0 dB, which is smaller than the aforementioned interferometric and photonic-crystal PBSs (1.5–3.5 dB). Additionally, the photonic-crystal and interferometric PBSs are based on optical interference. Interference-based mechanisms are, in general, highly dispersive, hence leading to a narrow operating bandwidth (usually ~100 nm or less). The proposed PBS, by contrast, supports a wide operating spectrum from ~0.6 μm to ~2.2 μm. To be more precise, the TIR mechanism enables broadband operation from the visible regime to ~1.75 μm at a fixed angle of incidence θA/P = 8°. For λ > 1.75 μm, the angle of incidence needs to be adjusted (at λ = 1.75 μm, the available θA/P ranges from 7.9° to 22.7°, calculated with the dispersions of the LC and prism taken into account). The absorption by the azobenzene 1205 [Fig. 4(e)] and that by the N-SF11 prisms lead to increases in the insertion loss for λ < 0.6 μm and λ > 2.2 μm, respectively, thus setting the lower and upper limits of the operating wavelength. Moreover, the optical switchability and bistability of the azo-LC PBS are unique among the current polarization-splitting techniques.
4. Conclusion
In summary, an all-optical PBS has been designed and experimentally demonstrated by sandwiching an azo-LC between two N-SF11 prisms. The PBS can be optically switched between the PSM and NSM. In the PSM, the azo-LC is set in the nematic phase, giving rise to a polarization-dependent critical angle of TIR and thus splitting the S- and P-polarized light into different directions. Specifically, the P-wave is allowed to be transmitted through the PBS while the S-wave is reflected at the prism-to-LC interface. The large split angle of over 40° in air makes it compact and thus feasible for use in on-chip systems. The PERs of the PBS are 29.0 dB for P-wave and 13.2 dB for S-wave. Upon switching to the NSM, the azo-LC is transitioned to the isotropic phase, the probe beam experiences TIR at the prism-to-LC interface regardless of the polarization state. The switching is accomplished by the light-driven isomerization of the azobenzene constituent in the LC and the resulting isothermal transition between the nematic and isotropic phases. With a yet-to-be-optimized setup, fast all-optical switching (in milliseconds) with persistent (hours-long) field-off bistability has been demonstrated. In conjunction with the ease of fabrication, compact size, tether-free control, and spatial addressability, the azo-LC PBS presents itself as a promising route to advanced optical processing and opens new avenues in applied optics and photonics.
Acknowledgment
The authors would like to thank the Ministry of Science and Technology of the Republic of China, Taiwan, for financially supporting this research under contract MOST106-2112-M-110-003-MY3.
References and links
1. M. Papaioannou, E. Plum, and N. I. Zheludev, “All-optical pattern recognition and image processing on a metamaterial beam splitter,” ACS Photonics 4(2), 217–222 (2017). [CrossRef]
2. S. Zhang, L. J. Gibson, A. B. Stilgoe, I. A. Favre-Bulle, T. A. Nieminen, and H. Rubinsztein-Dunlop, “Ultrasensitive rotating photonic probes for complex biological systems,” Optica 4(9), 1103–1108 (2017). [CrossRef]
3. C. Macias-Romero, I. Nahalka, H. I. Okur, and S. Roke, “Optical imaging of surface chemistry and dynamics in confinement,” Science 357(6353), 784–788 (2017). [CrossRef] [PubMed]
4. X. Song, X. Zhai, L. Liu, and S. Wu, “Lidar and Ceilometer Observations and Comparisons of Atmospheric Cloud Structure at Nagqu of Tibetan Plateau in 2014 Summer,” Atmosphere 8(1), 9 (2017). [CrossRef]
5. D. F. Vanderwerf, “Polarization Properties of Prisms and Reflectors,” in Applied Prismatic and Reflective Optics (Society of Photo-Optical Instrumentation Engineers, 2010), pp. 61–70.
6. D. Dai, L. Liu, S. Gao, D.-X. Xu, and S. He, “Polarization management for silicon photonic integrated circuits,” Laser Photonics Rev. 7(3), 303–328 (2013). [CrossRef]
7. B. Shen, P. Wang, R. Polson, and R. Menon, “An integrated-nanophotonics polarization beamsplitter with 2.4 × 2.4 μm2 footprint,” Nat. Photonics 9(6), 378–382 (2015). [CrossRef]
8. S. Zhu, Y. Liu, L. Shi, X. Xu, S. Yuan, N. Liu, and X. Zhang, “Tunable polarization beam splitter based on optofluidic ring resonator,” Opt. Express 24(15), 17511–17521 (2016). [CrossRef] [PubMed]
9. T. Pfau, R. Peveling, J. Hauden, N. Grossard, H. Porte, Y. Achiam, S. Hoffmann, S. K. Ibrahim, O. Adamczyk, S. Bhandare, D. Sandel, M. Porrmann, and R. Noe, “Coherent Digital Polarization Diversity Receiver for Real-Time Polarization-Multiplexed QPSK Transmission at 2.8 Gb/s,” IEEE Photonics Technol. Lett. 19(24), 1988–1990 (2007). [CrossRef]
10. T. B. Pittman, B. C. Jacobs, and J. D. Franson, “Probabilistic quantum logic operations using polarizing beam splitters,” Phys. Rev. A 64(6), 062311 (2001). [CrossRef]
11. Y. Zhang, Y. Zhang, and B. Li, “Optical switches and logic gates based on self-collimated beams in two-dimensional photonic crystals,” Opt. Express 15(15), 9287–9292 (2007). [CrossRef] [PubMed]
12. J. C. Bird, F. Liang, B. H. Solheim, and G. G. Shepherd, “A polarizing Michelson interferometer for measuring thermospheric winds,” Meas. Sci. Technol. 6(9), 1368–1378 (1995). [CrossRef]
13. D. Gottesman and I. L. Chuang, “Demonstrating the viability of universal quantum computation using teleportation and single-qubit operations,” Nature 402(6760), 390–393 (1999). [CrossRef]
14. B. P. Lanyon, J. D. Whitfield, G. G. Gillett, M. E. Goggin, M. P. Almeida, I. Kassal, J. D. Biamonte, M. Mohseni, B. J. Powell, M. Barbieri, A. Aspuru-Guzik, and A. G. White, “Towards quantum chemistry on a quantum computer,” Nat. Chem. 2(2), 106–111 (2010). [CrossRef] [PubMed]
15. R. A. Soref and D. H. McMahon, “Optical Design of Wollaston-Prism Digital Light Deflectors,” Appl. Opt. 5(3), 425–434 (1966). [CrossRef] [PubMed]
16. S. M. Macneille, “Beam splitter,” US Patent 2403731 (1946).
17. M. Banning, “Practical methods of making and using multilayer filters,” J. Opt. Soc. Am. 37(10), 792–797 (1947). [CrossRef] [PubMed]
18. R. Appel, C. D. Dyer, and J. N. Lockwood, “Design of a broadband UV-visible α-barium borate polarizer,” Appl. Opt. 41(13), 2470–2480 (2002). [CrossRef] [PubMed]
19. H. Fukuda, K. Yamada, T. Tsuchizawa, T. Watanabe, H. Shinojima, and S. Itabashi, “Ultrasmall polarization splitter based on silicon wire waveguides,” Opt. Express 14(25), 12401–12408 (2006). [CrossRef] [PubMed]
20. H. Jung Moo, R. Hyun Ho, P. Soon Ryong, J. Jae Wan, L. Seung Gol, L. El-Hang, P. Se-Geun, W. Deokha, K. Sunho, and O. Beom-Hoan, “Design and fabrication of a significantly shortened multimode interference coupler for polarization splitter application,” IEEE Photonics Technol. Lett. 15(1), 72–74 (2003). [CrossRef]
21. P. Rani, Y. Kalra, and R. K. Sinha, “Complete photonic bandgap-based polarization splitter on silicon-on-insulator platform,” J. Nanophotonics 10(2), 026023 (2016). [CrossRef]
22. U. A. Hrozhyk, S. V. Serak, N. V. Tabiryan, and T. J. Bunning, “Optical Tuning of the Reflection of Cholesterics Doped with Azobenzene Liquid Crystals,” Adv. Funct. Mater. 17(11), 1735–1742 (2007). [CrossRef]
23. E. Hecht, Optics, 5th ed. (Pearson, 2016).
24. N. V. Tabiryan and I.-C. Khoo, “Recent Advances in Nematic Liquid Crystal Nonlinear Optics,” in Handbook of Liquid Crystals (Wiley-VCH Verlag GmbH & Co., 2014), pp. 453–473.
25. C.-W. Chen, T.-H. Lin, and I. C. Khoo, “Dynamical studies of the mechanisms for optical nonlinearities of methyl-red dye doped blue phase liquid crystals,” Opt. Express 23(17), 21650–21656 (2015). [CrossRef] [PubMed]
26. U. A. Hrozhyk, S. V. Serak, N. V. Tabiryan, L. Hoke, D. M. Steeves, and B. R. Kimball, “Azobenzene liquid crystalline materials for efficient optical switching with pulsed and/or continuous wave laser beams,” Opt. Express 18(8), 8697–8704 (2010). [CrossRef] [PubMed]
27. N. Abadía, D. Xiangyang, Q. Lu, W. H. Guo, E. El-Fiky, D. V. Plant, and J. F. Donegan, “Novel polarization beam splitter based on p-i-n structure for an indium phosphide platform,” in 2017 19th International Conference on Transparent Optical Networks (ICTON, 2017), paper Tu.A5.1. [CrossRef]
28. U. G. Yasa, M. Turduev, I. H. Giden, and H. Kurt, “High Extinction Ratio Polarization Beam Splitter Design by Low-Symmetric Photonic Crystals,” J. Lightwave Technol. 35(9), 1677–1683 (2017). [CrossRef]
