Open Access
Add to BibSonomy
Abstract
A silicon-based polarization-insensitive optical filter is proposed and demonstrated. For the present on-chip polarization-insensitive optical filter, there is a dual-polarization mode (de)multiplexer, a TE-type multimode waveguide grating (MWG) with triangular corrugations and a TM-type MWG with rectangular corrugations. Here the triangular corrugations are introduced to reduce the undesired reflection and suppress the Fabry-Parot resonance. Furthermore, lateral-shift apodization is introduced for both two types of MWGs to suppress the sidelobes. For the fabricated device, the measured 3 dB-bandwidth is as large as ∼11 nm and the excess loss is ∼1.5 dB for both polarizations, while the sidelobe suppression ratios are 23 dB and 17 dB for TE and TM polarizations, respectively.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Silicon photonic integrated devices have become very promising because of the CMOS compatibility and high integration density for many applications [1], including the short-distance optical interconnects in data centers and fiber-to-the-home (FTTH) systems [2,3]. As it is well known, a high-performance on-chip optical filter is one of the most important devices in wavelength-division multiplexing (WDM) optical communication systems. Currently, there are many silicon-based on-chip optical filters realized with different structures, including micro-ring resonators [4], arrayed-waveguide gratings [5], multimode interference couplers [6] and Mach-Zehnder interferometers [7]. One might notice that an optical filter based on Bragg grating is promising owing to their flexible wavelength selectivity, flat-top bandpass spectra and ultra-wide free spectral range for WDM filters in data centers and FTTH systems [8–10].
Since the polarization state of light in optical fiber communication systems is often random, it is desired to develop various polarization-insensitive components [11], including optical filters. As it is well known, silicon-on-insulator (SOI) nanophotonic waveguides usually have a huge birefringence due to the ultra-high index-contrast and submicron cross-section, and thus silicon photonic devices are often seriously polarization-sensitive. Currently, several polarization-insensitive optical filters have been demonstrated in recent years. A general solution to achieve polarization independence is to utilize the polarization diversity technology by using the combination of a polarization beam splitter and a polarization rotator or a polarization splitter-rotator [12,13]. In this way, the device size and the circuit complexity increase greatly. Another approach is using square SOI waveguides with relaxed polarization sensitivity [14,15]. However, the fabrication tolerance is very low and the fabrication is usually not easy. Recently, polarization-rotating Bragg gratings [16,17] provide another solution for polarization-independent optical filtering. However, the fabrication is complicated because a two-step etching process or a high fabrication accuracy is needed.
In this paper, we propose and demonstrate a polarization-insensitive grating-based optical filter on silicon, which composed of a dual-polarization mode (de)multiplexer, a TE-type multimode waveguide grating (MWG) and a TM-type MWG. The dual-polarization mode (de)multiplexer is based on a dual-core adiabatic coupler to drop the TE1 and TM1 modes simultaneously. For the TE- and TM-type MWGs, lateral-shift apodization is used to suppress the sidelobes. In particular, triangular corrugations are introduced to reduce the undesired reflection and suppress the Fabry-Parot resonance. The experimental result shows that the fabricated polarization-insensitive filter has a large bandwidth of ∼11 nm and a low excess loss of ∼1.5 dB for both two polarizations. The sidelobe suppression ratios (SLSRs) are 23 dB and 17 dB for TE- and TM-polarizations, respectively.
2. Structure and design
Figure 1(a) shows the proposed polarization-independent optical filter on silicon, which consists of a TE-type MWG with triangular corrugations, a TM-type MWG with rectangular corrugations and a dual-polarization mode (de)multiplexer based on an adiabatic dual-core taper. Its working principle is described as follows. As shown in Fig. 1(a), the TE-type MWG is used to convert the launched TE0 mode (forward) to the TE1 mode reflected backward when the wavelength is around the Bragg wavelength. Then the backward TE1 mode is converted to the TE0 mode at the drop port through the dual-polarization mode (de) multiplexer. For the TM0 mode launched at the input port, it goes through the TE-type MWG with low loss and low crosstalk, which is possible by manipulating the polarization dependence of the TE-type MWG. Then the forward TM0 mode is converted to the backward TM1 mode through the TM-type MWG when the wavelength is around the Bragg wavelength, which is designed optimally to be the same as that of the TE-type MWG. After that, the backward TM1 mode goes through the TE-type MWG once more with low loss, and is finally converted to the TM0 mode at the drop port through the dual-polarization mode (de)multiplexer.
Fig. 1. Schematic configurations. (a) The proposed polarization-insensitive filter; (b) The MWG with triangular corrugations; (c) The MWG with rectangular corrugations; (d) The mode (de)multiplexer based on an adiabatic dual-core taper; (e) The longitudinal apodization for the grating with triangular corrugations; (f) The longitudinal apodization of the grating with rectangular corrugations.
In the present design, we choose SOI waveguides with a 220 nm-thick top-silicon layer and a 2 µm-thick oxide buffer. The upper-cladding is chosen to be a 1.2 µm-thick SU-8 layer for the fabrication convenience. One could also choose some other cladding materials (e.g., SiO2), while some minor modifications are needed for achieving optimal design. For the dual-polarization mode (de)multiplexers shown in Fig. 1(d), the adiabatic dual-core taper is designed according to the method proposed in our previous work [18]. Figure 2(a) shows the calculated dispersion curves for TE and TM polarizations of SOI waveguides with different core widths at the central wavelength of 1550 nm. In order to drop the TE1 and TM1 modes simultaneously, the core widths at the input/output ends of waveguides A and B for adiabatic couplers are chosen as (wa1, wb1) = (480, 300) nm, and (wa2, wb2) = (900, 120) nm, respectively. With this design, the core width wai of waveguide A increases from 480 nm to 900 nm, while the core width wbi of waveguide B decreases from 300 nm to 120 nm. We choose the taper lengths (L01, L12, L23) = (20, 35, 15) µm and the gap widths (wg1, wg2, wg3) = (1.2, 0.2, 0.5) µm. The calculated transmissions of the backward TE1 and TM1 modes launched from the right side are shown in Figs. 2(b) and 2(c) (see the curves for the case of Δw = 0). It can be seen that the designed dual-polarization mode (de)multiplexer has a low excess loss (0.05–0.5 dB) and a high extinction ratio (ER) of >23 dB in the wavelength band from 1500 nm to 1600 nm. We also simulate the light propagation in the designed mode (de)multiplexers when operating with the TE1 and TM1 modes, respectively, as shown in insets of Figs. 2(b) and 2(c), which show that the designed mode (de)multiplexer works very well with both polarizations. An analysis for the fabrication tolerance of the designed mode (de)multiplexer is also given by assuming that there are variations of the core widths, i.e., Δw = ±20 nm, as shown in Figs. 2(b) and 2(c). It can be seen that the designed mode (de)multiplexer can work well even when the waveguide-width variation is as large as Δw = ±20 nm.
Fig. 2. (a) Calculated dispersion curves of an SOI nanowire with hco = 220 nm. Here, the thick solid and the hollow curves are for TE and TM polarizations, respectively. Simulated results for the transmissions of the designed mode (de)multiplexer when the variation of the core width is assumed as Δw = ±20 nm. (b) TE polarization and (c) TM polarization. Inset: simulated light propagation in the designed adiabatic dual-core taper when operating at TE- (b) and TM- (c) polarizations at central wavelength 1550 nm.
The grating structures are designed according to the phase matching condition between the forward fundamental mode and the backward higher-order mode for one of the polarizations, so that the TE0 (TM0) mode launched at the input port can be converted to the TE1 (TM1) mode reflected backward. Here the phase-matching condition is given as [9] neff1+neff2=λB/Λ, where neff0 and neff1 are respectively the effective indices of the TE0 (TM0) and TE1 (TM1) modes in the MWG, λB is the Bragg wavelength and Λ is the grating period. Here grating tapers are introduced to connect the input/output section and the grating section [see Figs. 1(b) and 1(c)] to suppress the undesired reflection loss at the front/back ends of the grating section [19]. For example, here the period number for the grating taper is chosen as Ntp1 = Ntp2 = 20. In order to improve the SLSRs, the grating is apodized longitudinally [9]. For example, the superposition of the gratings are modulated with a Gaussian function of the position z in the propagation direction [see Figs. 1(e) and 1(f)], i.e., Δs = exp (−b(z−L/2)/L2)L/2, where Δs is the longitudinal shift, b is the apodization strength and L is the length of the Bragg grating. For the TE-type MWG, one should minimize the reflection and suppress the Fabry-Parot resonance between these two MWGs in cascade when the TM0 and TM1 modes propagate. Here we propose a TE-type MWG with triangular corrugations instead of rectangular corrugations, as shown in Figs. 1(b) and 1(c). In this design, abrupt discontinuities are avoided and the minimum feature sizes are relaxed, which is desired for the fabrication [20].
Figures 3(a) and 3(b) respectively show the calculated spectral responses for the MWGs with triangular corrugations when launching the TM0 and TM1 modes. In order to give a comparison, we also calculate the spectral responses for the MWGs with rectangular corrugations when launching the TM0 and TM1 modes, as shown in Figs. 3(c) and 3(d). Here, the parameters for the triangular corrugations are chosen as follows: the corrugation depth δ = 200 nm, the grating period Λ = 308 nm, the grating width W = 1100 nm, the apodization strength b = 15, the grating number N = 200 and the taper grating number Ntp = 20. For the rectangular corrugations, the parameters are chosen as follows: the corrugation depth δ = 180 nm the grating period Λ = 312 nm, the grating width W = 1100 nm, the duty cycle η = 0.5, the apodization strength b = 15, the grating number N = 200 and the taper grating number Ntp = 20. It can be seen that the MWG with triangular corrugations has lower reflection power. For example, the reflection power is reduced from ∼−24 dB to ∼−30 dB for the TM0 mode and from ∼−24 dB to ∼−35 dB for TM1 mode. As a result, the F-P resonance between the TE-type MWG and the TM-type MWG can be neglected.
Fig. 3. Calculated spectral responses at the reflect/through ports for the MWG with triangular corrugations when launching the (a) TM0 and (b) TM1 modes, respectively. Calculated spectral responses at the reflect/through ports for the MWG with rectangular corrugations when launching the (c) TM0 and (d) TM1 modes, respectively.
Figure 4(a) shows the calculated spectral responses for the TE-type MWG when choosing the corrugation depths δ1 = 180, 190 and 200 nm, respectively. The corresponding periods are Λ1 = 306, 308 and 310 nm, respectively. Here, the other parameters for the MWG are W1 = 1100 nm, b1 = 15, N1 = 200 and Ntp1 = 20. From Fig. 4(a), it can be seen that the 3 dB-bandwidth of the design optical filter is ∼10 nm and the SLSR is >22 dB. Figure 4(a) also shows that the spectral response is not sensitive to the corrugation depth δ1, which indicates that the fabrication tolerance is large.
Fig. 4. (a) Calculated spectral responses of the transmissions at the drop port with different corrugation depths for TE-type Bragg, i.e., δ1 = 180, 190 and 200 nm. (b) Calculated spectral responses of the transmissions at the drop port with different corrugation depths for TM-type Bragg, i.e., δ2 = 430, 440 and 450 nm.
Since TM-polarization modes have relatively weak interaction with the sidewall gratings, the TM-type MWG is designed by choosing rectangular corrugations with large corrugation depths in order to achieve the spectral responses similar to that of the TE-type MWG. Figure 4(b) shows the calculated spectral responses for the TM-type MWG when choosing the corrugation depths δ2 = 430, 440 and 450 nm, respectively. The corresponding periods are Λ2 = 338, 440 and 442 nm, respectively. Here, the other parameters for the TM-type MWG are W2 = 1600 nm, b2 = 15, N2 = 200 and Ntp2 = 20. With this design, the TM-type MWG has a 3 dB-bandwidth of ∼10 nm, an excess loss of ∼0.2 dB, and an SLSR of >24 dB, which is similar to the performances of the designed TE-type MWG.
3. Fabrication and measurement
The designed polarization-insensitive filter was fabricated on an SOI wafer with a 220 nm-thick top-silicon layer and a 2 µm thick buried-dioxide layer. The processes of electron-beam lithography (Raith 150 II) and ICP (inductively coupled plasma) dry-etching were applied to etch through the silicon core. A 1.2-µm SU-8 polymer was spin-coated on the top to be the upper-cladding. The microscope and scanning electron microscopic (SEM) images of the fabricated polarization-insensitive filter are shown in Figs. 5(a)–5(e). In order to characterize the device for both TE and TM polarizations, we fabricated the optical filters with identical designs on the same chip, while their grating couplers are designed for TE and TM polarizations. Figure 6(a) shows the measurement setup used for the characterization of the fabricated devices. Here a broad-band amplified spontaneous emission (ASE) light source was used and the polarization state of light was modified to be TE- and TM-polarized by using a fiber-based polarizer and a polarization controller (PC). An optical spectrum analyzer (OSA) was used to receive the power output from the devices.
Fig. 5. (a) Microscope image of the fabricated polarization-insensitive optical filter on silicon. SEM images of the TM-type grating taper (b), the TE-type grating taper (c), the TM-type MWG (d) and the TE-type MWG (e).
Fig. 6. (a) Measurement setup. (b) Measured spectral responses of the polarization-independent optical filters when TE- or TM-polarized light is launched.
Figure 6(b) shows the normalized spectral responses at the drop port of the fabricated polarization-insensitive optical filter for both TE and TM polarizations. Here the measured transmissions of the straight waveguides with TE- or TM-type grating couplers are used for the normalization. From the measured results shown in Fig. 6(b), it can be seen that the 3 dB-bandwidths for TE- and TM- polarizations are 11 nm and 12 nm, respectively, which are both similar to the theoretical simulation shown in Figs. 4(a) and 4(b). The minimal excess losses for TE- and TM- polarization modes are ∼0.6 dB and ∼1.5 dB in the 1 dB-bandwidth, respectively. The SLSRs for TE and TM polarizations are ∼23 dB and ∼17 dB, respectively. There is a slight shift of 0.7 nm between the center wavelengths of TE and TM polarizations. The performance of the present optical filter can be improved further by optimizing the fabrication processes.
4. Conclusion
In conclusion, in this paper we have proposed and realized a silicon-based polarization-insensitive optical filter with a low loss and a large 3 dB-bandwidth. The present optical filter has been designed with a dual-polarization mode (de)multiplexer, a TE-type MWG and a TM-type MWG. Lateral-shift apodization is introduced for the MWGs to suppress the sidelobes. In particular, the TE-type MWG is designed with triangular corrugations to reduce any undesired reflection. The fabricated optical filter shows polarization-insensitive performances with large 3 dB-bandwidths of 11–12 nm, low losses of 0.6–1.5 dB, decent SLSRs of 17–23 dB for both polarizations. Furthermore, the 3 dB-bandwidth for the present optical filter can be adjusted by modifying the corrugation depths and thus will be useful for many applications.
Funding
Natural Science Foundation of Zhejiang Province (LD19F050001, LZ18F050001); National Natural Science Foundation of China (NSFC) (1171101320, 61725503); National Major Research and Development Program (2018YFB2200200).
References
1. W. Bogaerts and L. Chrostowski, “Silicon photonics circuit design: methods, tools and challenges,” Laser Photonics Rev. 12(4), 1700237 (2018). [CrossRef]
2. Y. A. Vlasov, “Silicon CMOS-integrated nano-photonics for computer and data communications beyond 100G,” IEEE Commun. Mag. 50(2), S67–S72 (2012). [CrossRef]
3. H. H. Chang, Y. H. Kuo, R. Jones, A. Barkai, and J. E. Bowers, “Integrated hybrid silicon triplexer,” Opt. Express 18(23), 23891–23899 (2010). [CrossRef]
4. W. Bogaerts, P. De Heyn, T. Van Vaerenbergh, K. De Vos, S. K. Selvaraja, T. Claes, P. Dumon, P. Bienstman, D. Van Thourhout, and R. Baets, “Silicon microring resonators,” Laser Photonics Rev. 6(1), 47–73 (2012). [CrossRef]
5. W. Bogaerts, P. Dumon, D. Van Thourhout, D. Taillaert, P. Jaenen, J. Wouters, S. Beckx, V. Wiaux, and R. G. Baets, “Compact wavelength-selective functions in silicon-on-insulator photonic wires,” IEEE J. Sel. Top. Quantum Electron. 12(6), 1394–1401 (2006). [CrossRef]
6. Y. Hu, R. M. Jenkins, F. Y. Gardes, E. D. Finlayson, G. Z. Mashanovich, and G. T. Reed, “Wavelength division (de)multiplexing based on dispersive self-imaging,” Opt. Lett. 36(23), 4488–4490 (2011). [CrossRef]
7. S. Dwivedi, H. D’heer, and W. Bogaerts, “A compact all-silicon temperature insensitive filter for WDM and bio-sensing applications,” IEEE Photonics Technol. Lett. 25(22), 2167–2170 (2013). [CrossRef]
8. W. Shi, H. Yun, C. Lin, M. Greenberg, X. Wang, Y. Wang, S. T. Fard, J. Flueckiger, N. A. F. Jaeger, and L. Chrostowski, “Ultra-compact, flat-top demultiplexer using anti-reflection contra-directional couplers for CWDM networks on silicon,” Opt. Express 21(6), 6733–6738 (2013). [CrossRef]
9. J. F. Jiang, H. Y. Qiu, G. C. Wang, Y. Li, T. G. Dai, D. B. Mu, H. Yu, J. Y. Yang, and X. Q. Jiang, “Silicon lateral-apodized add-drop filter for on-chip optical interconnection,” Appl. Opt. 56(30), 8425–8429 (2017). [CrossRef]
10. D. J. Liu, M. Zhang, and D. X. Dai, “Low-loss and low-crosstalk silicon triplexer based on cascaded multimode waveguide gratings,” Opt. Lett. 44(6), 1304–1307 (2019). [CrossRef]
11. D. X. Dai, L. Liu, S. M. Gao, D. X. Xu, and S. L. He, “Polarization management for silicon photonic integrated circuits,” Laser Photonics Rev. 7(3), 303–328 (2013). [CrossRef]
12. H. Fukuda, K. Yamada, T. Tsuchizawa, T. Watanabe, H. Shinojima, and S. I. Itabashi, “Silicon photonic circuit with polarization diversity,” Opt. Express 16(7), 4872–4880 (2008). [CrossRef]
13. H. Y. Wang, J. C. Dai, H. Jia, S. Z. Shao, X. Fu, L. Zhang, and L. Yang, “Polarization-independent tunable optical filter with variable bandwidth based on silicon-on-insulator waveguides,” Nanophotonics 7(8), 1469–1477 (2018). [CrossRef]
14. Y. Onawa, H. Okayama, D. Shimura, S. Miyamura, H. Yaegashi, and H. Sasaki, “Polarisation-insensitive Si wire waveguide add/drop wavelength filter using reflective mode conversion grating and mode split coupler,” Electron. Lett. 48(20), 1297 (2012). [CrossRef]
15. X. L. Deng, L. S. Yan, H. Y. Jiang, X. Feng, W. Pan, and B. Luo, “Polarization-insensitive and tunable silicon Mach-Zehnder wavelength filters with flat transmission passband,” IEEE Photonics J. 10(3), 1–7 (2018). [CrossRef]
16. H. Okayama, Y. Onawa, D. Shimura, H. Yaegashi, and H. Sasaki, “Polarization rotation Bragg grating using Si waveguide with top surface groove grating and terrace,” Jpn. J. Appl. Phys. 57(4), 042201 (2018). [CrossRef]
17. H. Yun, L. Chrostowski, and N. A. F. Jaeger, “Narrow-band, polarization-independent, transmission filter in a silicon-on-insulator strip waveguide,” Opt. Lett. 44(4), 847–850 (2019). [CrossRef]
18. D. X. Dai, C. L. Li, S. P. Wang, H. Wu, Y. C. Shi, Z. H. Wu, S. M. Gao, T. G. Dai, H. Yu, and H. K. Tsang, “10-channel mode (de)multiplexer with dual polarizations,” Laser Photonics Rev. 12, 1700109 (2018). [CrossRef]
19. D. J. Liu, H. Wu, and D. X. Dai, “Silicon multimode waveguide grating filter at 2 µm,” J. Lightwave Technol. 37(10), 2217–2222 (2019). [CrossRef]
20. E. Durán-Valdeiglesias, C. Alonso-Ramos, S. Guerber, D. Oser, X. L. Roux, D. Benedikovic, N. Vulliet, E. Cassan, D. Marris-Morini, C. Baudot, F. Boeuf, and L. Vivien, “Dual-polarization O-band silicon nitride Bragg filters with high extinction ration,” in Europe Conference on Integrated Optics (ECIO) (2018).
