Open Access
Add to BibSonomy
Abstract
A Polarization Rotator-Splitter (PRS) based on thin-film lithium niobate (TFLN) is demonstrated in this work. The PRS consists of a partially etched polarization rotating taper and an adiabatic coupler, which enables the input TE0 and TM0 to be output as TE0 from two ports, respectively. The fabricated PRS using standard i-line photolithography achieved large polarization extinction ratios (PERs) of > 20 dB across the whole C-band. Excellent polarization characteristics are maintained when the width is changed by ±150 nm. The on-chip insertion losses of TE0 and TM0 are less than 1.5 dB and 1 dB, respectively.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Lithium niobate (LiNbO3, LN) is widely studied for its low loss, large transparency window, and excellent electro-optical properties [1]. Compared with Titanium-diffusion waveguides or proton-exchange waveguides, etched TFLN waveguides has a smaller footprint, stronger optical confinement, and nonlinear effect [2,3]. With the successful realization of various TFLN photonic devices, such as high-performance electro-optic modulators [4,5] and coherent optical modulators [6], TFLN has become a promising platform for future electro-optic integrated devices in high-speed communication systems [7,8]. LN is anisotropic with the largest electro-optical coefficient along the crystal z-axis [9]. Polarization sensitivity is an important factor restricting the development of photonic integration. The PRS is an ideal solution to this problem, ensuring the light remains in the wanted polarization state [10]. In addition, with the development of high-order modulation schemes, polarization multiplexing can further improve the optical communication capacity [11]. As the core polarization control device in the transmitting and receiving end of a coherent optical transceiver, PRS is worthy of further study.
The PRS realizes beam rotation and splitting of the orthogonal polarization mode, mainly based on mode coupling and evolution. Mode coupling requires couplers that satisfy both mode hybridization and phase conditions [12–15], and significant asymmetry results in stringent size tolerance. Based on the mode evolution, an adiabatic taper is used to realize the conversion of TM0 and TE higher-order mode. Then a coupler is used to separate the TE fundamental mode from the higher-order mode, such as a directional coupler [16], adiabatic coupler [17], asymmetric Y-junction [18], and MMI coupler [19]. Considering the device’s process difficulty and tolerance characteristics of devices, the PRS based on mode evolution and adiabatic coupler is more competitive.
Recently PRSs [20–24] on the TFLN platform have been demonstrated as summarized in Table 1. It is seen that the PRSs have achieved good performances with high polarization extinction ratio, large bandwidth, and low insertion loss. By adiabatically tapering waveguide structures, the efficient PRS operation has been demonstrated based on the LNOI platform over an experimentally measured bandwidth of 130 nm [24]. However, its polarization extinction performance needs further improvement. In addition, the dimensional tolerance of the reported devices does not meet the requirements of the standard photolithography process, and all the reports are based on electron beam lithography (EBL) technology. Therefore, it remains a question if a high performance TFLN PRSs can be fabricated by standard photolithography. Considering photolithography plays a very important role in integrated photonics, in order to meet the requirements of the photolithography process, it is necessary to design PRS devices with large fabrication tolerance.
Table 1. Performances of PRSs on TFLN
In this work, a high-performance and large tolerance PRS based on TFLN is designed and characterized. The adiabatic taper was used to realize the mode conversion of TM0 to TE1, and the adiabatic coupler was used to separate TE1 and TE0. Compared with multimode interference (MMI) and asymmetric directional coupler (ADC), the adiabatic coupler is less sensitive to size and wavelength. It can meet the requirements of large bandwidth and low production accuracy. The PRS is prepared using standard i-line photolithography, which is more competitive with lower cost and faster production than EBL. The fabricated PRS on a 300-nm-thick x-cut TFLN wafer achieved PERs of > 20 dB and on-chip insertion losses of < 1.5 dB across the whole C-band. It maintained good polarization extinction characteristics when the design width was changed ±150 nm.
2. Design and simulation
Considering the waveguide size errors caused by photolithography and etching, obtaining PRS with large fabrication tolerance is essential. PRS based on adiabatic mode evolution has large fabrication tolerance and wavelength insensitivity. The PRS structure is shown in Fig. 1, inset is the cross-sectional structure. The red area represents the LN waveguide with a height H = 0.3 μm, the pink area represents the LN slab with a height Hslab = 0.1 μm and the sidewall angle θ=75°. The symmetrical SiO2 cladding is used for easy integration with other devices.
Fig. 1. The schematic diagram of the PRS. It consists of an adiabatic taper, an adiabatic coupler, and the residual mode filter region. Inset is the cross-sectional of the waveguide.
The partially etched ridge waveguide structure was used to enhance the hybridization. Figure 2(a) shows the effective index (neff) of the ridge waveguide with width variations, and it shows that hybridization exists near the width w0 = 1.2 μm. Figure 2(b) shows the variation TE polarization fraction (Ex) with width. When the width w < 1 μm, the second largest effective indices (mode 2) is TM0, and the third largest (mode 3) is TE1. When the width w > 1.4 μm, mode 2 is TE1, and mode 3 is TM0. When the width w is between 1 μm and 1.4 μm, TM0 and TE1 polarization states could not be distinguished. In region ${\mathbf{I} }$, the ridge waveguide linear widens. The neff difference between modes 2 and 3 allows a TM0 input to remain in mode 2 and adiabatically evolve into TE1. At the same time, the TE0 input retains its polarization state, as shown in Fig. 2(c). According to the hybrid mode distribution and the possible width error of ±200 nm in i-line photolithography, wstart = w0-0.3 μm and wend = w0 + 0.3 μm are chosen as starting and cutoff widths of the adiabatic taper. The change of mode conversion efficiency with the length of region ${\mathbf{I} }$ is shown in Fig. 2(d).
Fig. 2. (a) The neff of ridge waveguide and (b) the ${E_x}$ fraction as the taper width varies. (c) Transmission simulation diagram with TE0 and TM0 as inputs. (d) Conversion efficiency as the length of taper varies.
The adiabatic coupler is also based on mode evolution, in which the input mode transmits adiabatically along the continuous changes of the coupler. The TE0 input will keep the first largest neff mode and transmit to the through waveguide as TE0 output. The TE1 input will keep the second largest neff mode and transmit to the cross waveguide as TE0 output, as shown in Fig. 3(a). ${n_{eff,{w_1},T{E_1}}} = {n_{eff,{w_2},T{E_0}}},\; {n_{eff,{w_3},T{E_1}}} = {n_{eff,{w_6},T{E_0}}}$ and ${n_{eff,{w_4},T{E_1}}} = {n_{eff,{w_5},T{E_0}}}$ can improve coupling efficiency and shorten the overall length of the adiabatic coupler, as shown in Fig. 3(b). In general, The TE0 input is output from the through port, while the TE1 input is output from the cross port in TE0 mode, as shown in Fig. 3(c). The change of mode coupling efficiency with the length of region ${\mathbf{II} }$ is shown in Fig. 3(d).
Fig. 3. (a) The neff of mode varies with the transmission of the adiabatic coupler. (b) The neff of the full-etched waveguide as width varies. (c) Transmission simulation diagram with TE0 and TM0 as inputs. (d) Coupling efficiency as the length of the coupler varies.
The polarization extinction characteristics of the through port are mainly affected by residual TM0 and TE1. After the adiabatic coupler, the first 1 1MMI was added to filter out the TE1, as shown in Fig. 4(a)(b). As shown in Fig. 4(c), the filtering effect of MMI is ∼18 dB, and the reflection of TE0 is <20 dB. Then the mode conversion area and the second 1 1MMI were added to filter the residual TM0. The polarization extinction characteristics of the cross port are mainly affected by TE0 input crosstalk, and the PER of the cross port is >35 dB at the current structure. Figure 4(d) shows the simulation results of the polarization extinction ratio of C and L bands, which are both greater than 30 dB in the wavelength range of 120 nm.
Fig. 4. Simulation diagram of (a) TE0 and (b) TE1 input the MMI. (c) Transmission and reflection as the length of MMI vary. (d) Simulation results of PER for two ports in the C and L bands.
3. Measurement set-up
Figure 5 shows a microscope image of the fabricated device, consisting of three regions: polarization rotation, adiabatic coupler, and filter.
The PER is defined as:
The test system of PRS is shown in Fig. 6(a), polarization controller HP8169A is mainly used in the test system.
Fig. 6. (a) Schematic diagram of the measurement set-up. (b) Diagram of polarization state changes of light through a polarization controller and an equivalent optical fiber.
Switching between the TE0 mode input and TM0 mode input is realized by adjusting the angle of the quarter-wave plate (θλ/4) and that of the half-wave plate (θλ/2) of the polarization controller. The angles are determined by finding the minimum output power of each port: The angles corresponding to the minimum power output of the cross port represent the TE input, and the angles corresponding to the minimum power output of the through port represent the TM input. This method has been verified by reverse testing of a distributed feedback laser whose output is TE polarized. These two orthogonal input modes correspond to angles that satisfy the following:
This relationship can be derived using the Jones matrix. The Jones matrix of a quarter-wave plate and half-wave plate can be expressed as:
Matrix derivations will not be performed in this article. Equation (2) can also be intuitively understood from Fig. 6(b), where the optical fiber part is equivalent to a half-wave plate and a quarter-wave plate. In Fig. 6(b), the angles of TM input are assumed as θλ/4 = 10° and θλ/2 = 30°.
By testing the straight waveguide, the coupling loss from the fiber to the device is obtained. According to our previous work, the transmission loss of the straight waveguide is about 0.23 dB/cm [25], which is negligible. So half of the insertion loss of the straight waveguide is the coupling loss. As shown in Fig. 7(a), the coupling losses of TE0 and TM0 are about 4.1 dB and 2.2 dB respectively. Due to the larger mode field of TM0 mode in waveguide, the coupling loss is smaller.
Fig. 7. (a) Measured coupling loss of TE0 and TM0. (b) Measured transmission of the designed PRS. (c) Comparison of simulated and tested polarization extinction ratios. (d) On-chip insertion loss of the designed PRS.
Figure 7(b) shows the transmission of the designed PRS, which has compensated for the difference in coupling loss between inputs of different polarization states. The PER of both ports is greater than 20 dB across the whole C-band. The device is insensitive to wavelength. Due to limited test light source conditions, the PER is estimated to remain high over a wider wavelength range, as shown in Fig. 4(d). Compared with the simulation results, as shown in Fig. 7(c), the polarization extinction performance of the device is reduced by about 15 dB, mainly due to the change in width and thickness during the manufacturing process.
By comparing the transmission power of the device and the straight waveguide, the effect of coupling loss is eliminated, and the on-chip insertion loss of PRS is obtained. Figure 7(d) shows the on-chip insertion loss of the PRS. The on-chip insertion losses of TE0 and TM0 are less than 1.5 dB and 1 dB respectively. The on-chip insertion loss of TE0 is mainly caused by MMI. We will optimize the size of MMI to reduce the on-chip insertion loss of TE0.
Width tolerance is the critical index of device application. The structures with different widths were added to the layout design. The waveguide sizes of different devices were tested by scanning electron microscope (SEM), and the relative width difference was consistent with the designed width difference, as shown in Fig. 8. The measured width is larger than the actual width of the LN waveguide due to SiO2 cladding. The performance with ±150 nm width variation was tested, as shown in Fig. 9. The PER was greater than 18 dB and 15 dB, respectively. The on-chip insertion loss was less than 0.6 dB and 4 dB, respectively. The on-chip insertion loss of TE0 can reflect the insertion loss of MMI. It is found that the loss of MMI is minimum in the structure with a width change of +150 nm and maximum in the structure with a width change of -150 nm. The result provides a reference for the design of the MMI dimension in the future optimization.
Fig. 9. Measured transmission and on-chip insertion loss for the width being (a)(b) δ=+150 nm (c)(d) δ= -150 nm.
4. Conclusion
In summary, we demonstrated a fully adiabatic TFLN-based PRS with large fabrication tolerance. The proposed device is fabricated by standard i-line photolithography. The PER of both ports is greater than 20 dB in the wavelength range of 40 nm, reflecting the excellent wavelength characteristics of the adiabatic device. In addition, when the width is changed by ±150 nm, the extinction ratio of the through port and the cross port can be greater than 18 dB and 15 dB, respectively. The on-chip insertion loss of TE0 and TM0 is less than 1.5 dB and 1 dB, respectively. The on-chip insertion loss of TE0 can be reduced by optimizing the MMI. The PRS proposed above is an attractive candidate for polarization multiplexing and coherent optical transceiver based on the TFLN photonic integration platform.
Funding
National Key Research and Development Program of China (2022YFB2802901); National Natural Science Foundation of China (62274073); Key Research and Development Program of Hubei Province (2021BAA001).
Disclosures
The authors declare no conflicts of interest.
Data availability
The data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. T. Volk and M. Wöhlecke, Lithium Niobate: Defects, Photo refraction and Ferroelectric Switching, vol. 115, Series: Springer Series in Materials Science (Springer, 2009).
2. G. Poberaj, H. Hu, W. Sohler, and P. Günter, “Lithium niobate on insulator (LNOI) for micro-photonic devices,” Laser Photonics Rev. 6(4), 488–503 (2012). [CrossRef]
3. Cheng Wang, Mian Zhang, Xi Chen, Maxime Bertrand, Amirhassan Shams-Ansari, Sethumadhavan Chandrasekhar, Peter Winzer, and Marko Lončar, “Integrated lithium niobate electro-optic modulators operating at CMOS-compatible voltages,” Nature 562(7725), 101–104 (2018). [CrossRef]
4. M. He, M. Xu, Y. Ren, J. Jian, Z. Ruan, Y. Xu, S. Gao, S. Sun, X. Wen, L. Zhou, L. Liu, C. Guo, H. Chen, S. Yu, L. Liu, and X. Cai, “High-performance hybrid silicon and lithium niobate Mach–Zehnder modulators for 100 Gbit s −1 and beyond,” Nat. Photonics 13(5), 359–364 (2019). [CrossRef]
5. Ye Liu, Heng Li, Jia Liu, Su Tan, Qiaoyin Lu, and Weihua Guo, “Low Vπ thin-film lithium niobate modulator fabricated with photolithography,” Opt. Express 29(5), 6320–6329 (2021). [CrossRef]
6. M. Xu, M. He, H. Zhang, J. Jian, Y. Pan, X. Liu, L. Chen, X. Meng, H. Chen, Z. Li, X. Xiao, S. Yu, S. Yu, and X. Cai, “High-performance coherent optical modulators based on thin-film lithium niobate platform,” Nat. Commun. 11(1), 3911 (2020). [CrossRef]
7. S. Saravi, T. Pertsch, and F. Setzpfandt, “Lithium Niobate on Insulator: An Emerging Platform for Integrated Quantum Photonics,” Adv. Opt. Mater. 9(22), 2100789 (2021). [CrossRef]
8. Mian Zhang, Cheng Wang, Di Zhu Prashanta Kharel, and Marko Lončar, “Integrated lithium niobate electro-optic modulators: when performance meets scalability,” Optica 8(5), 652–667 (2021). [CrossRef]
9. 刘晔, “高速薄膜铌酸锂 Mach-Zehnder 调制器的研究,” Ph.D. dissertation (Huazhong University of Science and Technology, 2021).
10. Hiroshi Fukuda, Koji Yamada, Tai Tsuchizawa, Toshifumi Watanabe, Hiroyuki Shinojima, and Sei-ichi Itabashi, “Silicon photonic circuit with polarization diversity,” Opt. Express 16(7), 4872–4880 (2008). [CrossRef]
11. S. Zhalehpour, J. Lin, M. Guo, H. Sepehrian, Z. Zhang, L. A. Rusch, and W. Shi, “All-Silicon IQ Modulator for 100 GBaud 32QAM Transmissions,” in Optical Fiber Communication Conference Postdeadline Papers 2019, (Optica Publishing Group, 2019), paper Th4A.5.
12. Liu Liu, Yunhong Ding, Kresten Yvind, and Jørn M. Hvam, “Silicon-on-insulator polarization splitting and rotating device for polarization diversity circuits,” Opt. Express 19(13), 12646–12651 (2011). [CrossRef]
13. Jing Wang, Ben Niu, Zhen Sheng, Aimin Wu, Xi Wang, Shichang Zou, Minghao Qi, and Fuwan Gan, “Design of a SiO2 top-cladding and compact polarization splitter-rotator based on a rib directional coupler,” Opt. Express 22(4), 4137–4143 (2014). [CrossRef]
14. Hang Guan, Ari Novack, Matthew Streshinsky, Ruizhi Shi, Qing Fang, Andy Eu-Jin Lim, Guo-Qiang Lo, Tom Baehr-Jones, and Michael Hochberg, “CMOS-compatible highly efficient polarization splitter and rotator based on a double-etched directional coupler,” Opt. Express 22(3), 2489–2496 (2014). [CrossRef]
15. Yule Xiong, Dan-Xia Xu, Jens H. Schmid, Pavel Cheben, Siegfried Janz, and Winnie N. Ye, “Fabrication tolerant and broadband polarization splitter and rotator based on a taper-etched directional coupler,” Opt. Express 22(14), 17458–17465 (2014). [CrossRef]
16. Daoxin Dai and John E. Bowers, “Novel concept for ultracompact polarization splitter-rotator based on silicon nanowires,” Opt. Express 19(11), 10940–10949 (2011). [CrossRef]
17. Wesley D. Sacher, Tymon Barwicz, Benjamin J. F. Taylor, and Joyce K. S. Poon, “Polarization rotator-splitters in standard active silicon photonics platforms,” Opt. Express 22(4), 3777–3786 (2014). [CrossRef]
18. Jing Wang, Ben Niu, Zhen Sheng, Aimin Wu, Wei Li, Xi Wang, Shichang Zou, Minghao Qi, and Fuwan Gan, “Novel ultra-broadband polarization splitter-rotator based on mode-evolution tapers and a mode-sorting asymmetric Y-junction,” Opt. Express 22(11), 13565–13571 (2014). [CrossRef]
19. Jing Wang, Minghao Qi, Yi Xuan, Haiyang Huang, You Li, Ming Li, Xin Chen, Qi Jia, Zhen Sheng, Aimin Wu, Wei Li, Xi Wang, Shichang Zou, and Fuwan Gan, “Proposal for fabrication-tolerant SOI polarization splitter-rotator based on cascaded MMI couplers and an assisted bi-level taper,” Opt. Express 22(23), 27869–27879 (2014). [CrossRef]
20. H. Luo, Zhiyan Chen, and Hao Li, “High-Performance Polarization Splitter-Rotator Based on Lithium Niobate-on-Insulator Platform,” IEEE Photonics Technol. Lett. 33(24), 1423–1426 (2021). [CrossRef]
21. Xuanhao Wang, An Pan, Tingan Li, Cheng Zeng, and Jinsong Xia, “Efficient polarization splitter-rotator on thin-film lithium niobate,” Opt. Express 29(23), 38044–38052 (2021). [CrossRef]
22. C. Deng, Wanghua Zhu, and Yu Sun, “Broadband Polarization Splitter-Rotator on Lithium Niobate-on-Insulator Platform,” IEEE Photonics Technol. Lett. 35(1), 7–10 (2023). [CrossRef]
23. Ranfeng Gan, Lu Qi, Ziliang Ruan, Jie Liu, Changjian Guo, Kaixuan Chen, and Liu Liu, “Fabrication tolerant and broadband polarization splitter-rotator based on adiabatic mode evolution on thin-film lithium niobate,” Opt. Lett. 47(19), 5200–5203 (2022). [CrossRef]
24. Zhaoxi Chen, Jingwei Yang, Wing-Han Wong, Edwin Yue-Bun Pun, and Cheng Wang, “Broadband adiabatic polarization rotator-splitter based on a lithium niobate on insulator platform,” Photonics Res. 9(12), 2319–2324 (2021). [CrossRef]
25. H. Li, Q. Chen, Y. Liu, Y. Tang, Q. Lu, M. Lu, and W. Guo, “High performance thin-film lithium niobate MZ modulator ready for massive production,” in Optical Fiber Communication Conference (OFC) 2022, S. Matsuo, D. Plant, J. Shan Wey, C. Fludger, R. Ryf, and D. Simeonidou, eds., Technical Digest Series (Optica Publishing Group, 2022), paper M2D.5.
