An integrated optic polarization splitter with large fabrication tolerance and high reliability is required for optical signal processing in quantum-encrypted communication systems. A polarization splitter based on total internal reflection from a highly birefringent polymer—reactive mesogen—is proposed and demonstrated in this work. The device consists of a mode expander for reducing the wave vector distribution of the guided mode, and an interface with a large birefringence. Several polymers with suitable refractive indexes were used for fabricating the device. We obtained a polarization splitter with a low crosstalk (less than −30 dB), and a large fabrication tolerance.
© 2016 Optical Society of America
As the data traffic over the internet is growing daily, and the trends of internet services are diversifying, the need for quantum-encrypted communication arises [1–4]. In quantum communication, for detecting a specific polarization state (such as the orthogonal Bell-state polarization) used for encryption, various optical components including polarization splitters are necessary. In this work, as the first step toward developing integrated optical devices for quantum communication, we propose a polarization splitter with high splitting ratio, low insertion loss, and wide fabrication tolerance providing good production yields.
Several integrated optic polarization splitters based on optical materials such as silicon, silica, and polymers have been demonstrated. In the silicon-waveguide polarization splitter based on directional couplers, the effective index contrast between the TE and TM modes was increased by modifying the waveguide geometry [5, 6]. However, a minor change in the geometry of the waveguide resulted in poor polarization-splitting characteristics, indicating that it has small fabrication tolerance. In silica-waveguide polarization splitters, by using a stress optic effect, amorphous silicon film was deposited on one arm of the Mach–Zehnder interferometer for applying a stress on the waveguide to impose polarization-dependent phase retardation. However, precise adjustment of the relative phase difference required additional trimming processes . In indium phosphide waveguides, the polarization splitter was composed of a polarization converter and a directional coupler; however, the splitting ratio was degraded because of the stringent fabrication tolerances [8, 9].
Polymeric optical waveguide devices have some advantages such as convenient refractive-index control and large freedom in the device structure due to the various fabrication methods [10, 11]. Birefringence of a polymer is associated with the structure of an organic molecule, and a large birefringence can be obtained by orienting the molecules in a certain direction. Then, this large birefringence can be utilized to form a polarization splitter based on an asymmetric Y-branch waveguide [12, 13]. A recent work, incorporating a long-chain oligomer molecule, exhibited low insertion losses with a high splitting ratio. Even though it has been tested for an optical communication system, the device had long-term stability issues because the imposed birefringence was degraded when the polymer was formed as a narrow strip in the device .
For the polarization splitter to be useful for quantum-encryption purposes, it should have a polarization crosstalk less than −30 dB, low insertion loss, and good reliability. The lower crosstalk will provide the better signal extinction ratio. We propose a polarization splitter incorporating a UV curable liquid crystal, reactive mesogen (RM). The RM is a widely used material in liquid-crystal displays. The large birefringence of RM, obtained through molecular alignment, enables the creation of a reflection-type polarization splitter, which is not feasible using any other material. At the interface between the RM and another polymer, the TE and TM polarizations experience significantly different Fresnel reflections due to the large birefringence of RM. Consequently, TE polarization experiences total internal reflection (TIR) at the interface, while TM polarization passes through the interface. An optimum device structure is designed, and the characteristics of the fabricated polymer device are compared with the design results.
2. Operating principle and device design
RM is a rod-like liquid crystalline material with crosslinking end groups . It can be used to form a thin film with a large birefringence by coating and UV-curing the liquid oligomer . A waveguide structure can be fabricated through a photolithography process. The refractive indexes of an RM film of 5-μm thickness were measured using a prism coupler, and were found to be 1.6475 and 1.5205 for the TE and TM polarizations, respectively.
The proposed polymer-waveguide polarization splitter consists of a tapered waveguide, a waveguide mode expander, and a TIR interface as shown in Fig. 1. A low birefringence waveguide with polarization-independent propagation is constructed using a CO-polymer produced by ChemOptics Co. The CO-polymer waveguide is connected to the RM waveguide
through the taper. When TE-polarized light is launched at the input port, the light guided through the CO-polymer waveguide encounters the tapered RM waveguide. It is then coupled into the RM waveguide, which is expanded to increase the mode width, and is internally reflected at the TIR interface. On the other hand, the TM-polarized light passes through the TIR interface with no reflection because the CO-polymer is designed to have the same refractive index as the RM for TM polarization.
TIR phenomenon ideally occurs for a plane-wave incident on an interface at an angle larger than the critical angle. However, the guided light confined in a narrow optical waveguide contains a broad range of wave vectors corresponding to the various plane waves traveling with different angles. The width of the wave-vector distribution is inversely proportional to the guided mode width. Therefore, by increasing the width of the guided mode by using a mode expander, one can reduce the width of the wave-vector distribution, in order to produce a reflection that is close to an ideal TIR.
The mode expander was designed by the effective-index method and the two-dimensional (2-D) beam propagation method (BPM). For TE-polarized light, the output modal powers were calculated as shown in Fig. 2, to find the amount of the higher order mode coupling. The input and output waveguide widths were 6 μm and 30 μm, respectively. The core layer thickness was 4 μm, with a refractive index of 1.520. A longer taper results in a smaller second-higher-order mode power as the adiabatic transition becomes more efficient. The RM-thickness change affects the effective index and the vertical mode width; however, it has no significant effect on the adiabatic transition characteristics. It was confirmed that the mode expansion loss could be less than 0.1 dB for an expander longer than 600 μm.
When the expanded guided mode reaches the TIR interface, the TE modes should be reflected while TM modes should be transmitted. In the 2-D BPM simulation, the RM had refractive indices of 1.6457 and 1.5205 for the TE and TM polarizations, respectively. As shown in Fig. 3, for a large angle of incidence, TIR occurred with a negligible loss. As the thickness of the RM increases, the birefringence was enhanced; this resulted in strong reflections even for the smaller angle. For the thickness of the RM layer larger than 0.8 μm, the effective-index difference at the TIR interface became sufficiently large to produce a reflection loss as low as 0.05 dB for an incidence angle of 80°.
The mode-conversion loss of the taper was calculated by 3-D BPM simulations, and the results are shown in Fig. 4. The CO-polymer waveguide has a core dimension of 6 × 4 μm2, and the RM waveguide has a core width of 6 μm. For an RM thickness of 1 μm and a taper length of more than 400 μm, the coupling loss was reduced to less than 0.2 dB. For a sufficiently long taper length of 1000 μm, the coupling loss was saturated at a certain value. This was caused by the mode mismatch between the input and output modes, as depicted in the figure. The TE-polarized light subjected to TIR is concentrated into the RM layer at the taper. As the RM thickness reduced, the evanescent field extended deep into the cladding so that the mode size increased, to reduce the coupling loss. However, if the RM layer became too thin, the TIR efficiency decreased as shown in Fig. 3. Hence, the RM thickness should be optimized to be around 1.0 μm.
3. Fabrication of the integrated optic device
The polymer-waveguide polarization splitter was fabricated using an RM material and low birefringence CO-polymers available from ChemOptics Co. For the TM polarization, the refractive index of the RM should be matched to that of the CO-polymer waveguide core. A CO-polymer with a refractive index of 1.520 was used for the core, and one with a refractive index of 1.500 was used for the cladding.
The fabrication procedure is schematically presented in Fig. 5. A lower cladding layer of CO-150 with a thickness of 8 μm was formed on a silicon wafer by spin-coating and UV curing. Then, a first core layer of CO-152 was formed with a thickness of 2.1 μm. For alignment-assistance purposes, polyimide was coated with a thickness of 0.2 μm, cured at 250°C for 30 min, and rubbed under a velvet roller. Then, RM solution was spin-coated on the rubbed polyimide surface to have 1.3 μm thickness. Polyimide rubbed with a velvet cloth provides an anchoring force when the liquid crystal solution is coated on it. As the liquid crystal molecules of an anisotropic shape are aligned along the rubbing direction, it will produce a thin film with a strong optical birefringence . After UV curing, the RM film exhibited a higher refractive index for the TE polarization. The birefringence of the RM film depends on the alignment uniformity. If the RM alignment is not uniform throughout the depth of the RM film, the birefringence of RM might be low. We observed that the alignment is easier for a thin RM film where the anchoring force on the liquid crystal is easily transferred.
After coating a photoresist layer on the aligned RM layer, an optical waveguide pattern was formed through a photolithography process and oxygen plasma etching. The scanning electron microscope (SEM) photograph of the tapered RM waveguide is shown in Fig. 6(a). CO-152 was coated once again over the RM pattern as the second core layer. The CO-polymer waveguide pattern was formed using a Cr metal mask and oxygen plasma etching. Figure 6(b) shows an optical-microscope image of the TIR interface covered by the CO-polymer waveguide. Significant difference of the color was observed at the interface. A cladding of CO-150 was coated to finish the waveguide.
4. Characterization of the fabricated device
A 1550-nm distributed feedback (DFB) laser was used for characterizing the device. The TE/TM polarizations were adjusted using a fiber-optic polarization controller. The output light from the device was captured using a charge-coupled device (CCD) as shown in Fig. 7. Depending on the polarization, the output light appeared at different positions.
Polarization splitters with different angles of incidence were fabricated for the comparison between the design and the experimental results. The measured characteristics of the devices fabricated in two batches are summarized in Fig. 8(a), and are similar in their overall performances. For TE input, the polarization splitter exhibited the lowest crosstalk of −33.9 dB, and an insertion loss of 7.6 dB in the device with an angle of 76°. For TM input, a crosstalk of −30.1 dB and an insertion loss of 7.5 dB were obtained. As the angle increased, the reflected power for the TM polarization increased because of the partial Fresnel reflection; this resulted in a poor crosstalk.
Through cutback measurements, the propagation losses of the RM and the CO-polymer waveguides were measured as 3.2 dB/cm and 1.1 dB/cm at 1550 nm, respectively. The fiber coupling loss by the mode mismatch was calculated to be 3.2 dB, and could be reduced by incorporating a small-core fiber. The absorption loss of the RM waveguide was rather large because of the C–H vibration overtone absorption near 1550 nm. A fluorinated polymer with a large birefringence is preferred to reduce the absorption loss. Furthermore, as long as the birefringence is not changed, the device would have better performance at 1300 nm and 850 nm because of the lower absorption loss.
For a large angle of incidence, degradation of the crosstalk may be caused by the refractive-index difference of the RM pattern, which is dependent on the uniformity of the molecular alignment. If the alignment is not uniform enough, the refractive index of RM for the TM polarization will increase while that for the TE polarization will decrease. For various TM refractive indices, the transmission and reflection powers of the polarization splitter were calculated using the BPM simulation as shown in Fig. 8(b). From these results, one can find that the result for an index of 1.536 is the closest to the experimental result. Therefore, we can conclude that the RM material had a higher TM refractive index than that measured in the prism coupling method, in which a thick RM layer (5 μm) was used because of the limit of measurable film thickness. Since more reliable refractive-index data were obtained from this experiment, in our next experiment, we can adjust the angle of incidence to obtain a lower crosstalk and provide the larger tolerance.
A polarization splitter with a TIR interface was demonstrated by incorporating a UV curable liquid crystal. At the TIR interface formed by the highly birefringent RM material, the TE polarization was internally reflected, while the TM polarization passed through. An optimum device structure was fabricated based on various simulation results. A crosstalk of less than −30 dB was achieved in the device with an incidence angle of 76°. The experimental results were compared with the design results, and it was observed that the actual refractive index of the RM material for TM polarization was higher than the expected value. The corrected refractive index will be used in our next experiment to obtain a lower crosstalk. The device performance was not very sensitive to the changes in the RM material thickness or the incident-angle variations, which is important for producing a polarization splitter with a large fabrication tolerance with a good production yield. The fabricated device will be integrated with other waveguide devices for demonstrating an integrated optic device for quantum signal processing.
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (2014R1A2A1A10051994).
References and links
2. J. C. F. Matthews, A. Politi, A. Stefanov, and J. L. O’Brien, “Manipulation of multiphoton entanglement in waveguide quantum circuits,” Nat. Photonics 3(6), 346–350 (2009). [CrossRef]
3. A. Martin, A. Issautier, H. Herrmann, W. Sohler, D. B. Ostrowsky, O. Alibart, and S. Tanzilli, “A polarization entangled photon-pair source based on a type-II PPLN waveguide emitting at a telecom wavelength,” New J. Phys. 12(10), 103005 (2010). [CrossRef]
5. I. Kiyat, A. Aydinli, and N. Dagli, “A compact silicon-on-insulator polarization splitter,” IEEE Photonics Technol. Lett. 17(1), 100–102 (2005). [CrossRef]
6. 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]
7. M. Okuno, A. Sugita, K. Jinguji, and M. Kawachi, “Birefringence control of silica waveguides on Si and its application to a polarization-beam splitted/switch,” J. Lightwave Technol. 12(4), 625–633 (1994). [CrossRef]
8. L. M. Augustin, R. Hanfoug, J. J. G. M. van der Tol, W. J. M. de Laat, and M. K. Smit, “A compact integrated polarization splitter/converter in InGaAsP-InP,” IEEE Photonics Technol. Lett. 19(17), 1286–1288 (2007). [CrossRef]
9. L. M. Augustin, J. J. G. M. van der Tol, R. Hanfoug, W. J. M. de Laat, M. J. E. van de Moosdijk, P. W. L. van Dijk, Y. S. Oei, and M. K. Smit, “A single etch-step fabrication-tolerant polarization splitter,” J. Lightwave Technol. 25(3), 740–746 (2007). [CrossRef]
10. M.-C. Oh, W.-S. Chu, J.-S. Shin, J.-W. Kim, K.-J. Kim, J.-K. Seo, H.-K. Lee, Y.-O. Noh, and H.-J. Lee, “Polymeric optical waveguide devices exploiting special properties of polymer materials,” Opt. Commun. 362, 3–12 (2016). [CrossRef]
11. Z. Zhang and N. Keil, “Thermo-optic devices on polymer platform,” Opt. Commun. 362, 101–114 (2016). [CrossRef]
12. M.-C. Oh, S.-S. Lee, S.-Y. Shin, W.-Y. Hwang, and J.-J. Kim, “Polymeric waveguide polarization splitter based on poling-induced birefringence,” Electron. Lett. 32(4), 324–326 (1996). [CrossRef]
13. M.-C. Oh, M.-H. Lee, and H.-J. Lee, “Polymeric waveguide polarization splitter with a buried birefringent polymer,” IEEE Photonics Technol. Lett. 11(9), 1144–1146 (1999). [CrossRef]
14. J.-W. Kim, K.-J. Kim, M.-C. Oh, J.-K. Seo, Y.-O. Noh, and H.-J. Lee, “Polarization-splitting waveguide devices incorporating perfluorinated birefringent polymers,” J. Lightwave Technol. 29(12), 1842–1846 (2011). [CrossRef]
15. H. Thiem, P. Strohriegl, M. Shkunov, and I. McCulloch, “Photopolymerization of reactive mesogens,” Macromol. Chem. Phys. 206(21), 2153–2159 (2005). [CrossRef]
16. W.-S. Chu, S.-M. Kim, J.-W. Kim, K.-J. Kim, and M.-C. Oh, “Polarization converting waveguide devices incorporating UV-curable reactive mesogen,” J. Opt. Soc. Korea 15(3), 289–292 (2011). [CrossRef]
17. J. Hoogboom, J. A. A. W. Elemans, A. E. Rowan, T. H. M. Rasing, and R. J. M. Nolte, “The development of self-assembled liquid crystal display alignment layers,” Philos. Trans. Royal Soc. A 365(1855), 1553–1576 (2007). [CrossRef] [PubMed]