Open Access
Add to BibSonomyAbstract
A compact, broadband, and low-loss TE-pass polarizer using transparent conducting oxides (TCOs) embedded in the center of the strip waveguide and deposited on its top is proposed and analyzed in detail. With the tunable permittivity of TCO, epsilon-near-zero (ENZ) of its real part and significant increase of its imaginary part can be achieved around the wavelength of 1.55 μm under a certain electron concentration. By introducing this ENZ material into the strip waveguide, huge polarization dependence can be realized, that is, the TE mode is almost not affected due to its quite weak interaction with TCOs, while the TM mode is extremely confined in the accumulation layers of TCO with high absorption loss, leading to a great reduction in length for the present polarizer. Moreover, the top TCO layer is applied to further enhance the polarizer performance. Results show that a polarizer of only 4.5 μm in length with an extinction ratio (ER) of 25.26 dB and an insertion loss of 0.21 dB is achieved at 1.55 μm, and its bandwidth can be extended to ~140 nm for an ER>20 dB. In addition, the ER can also be increased only by enlarging the length of the TCO-based polarizer.
© 2016 Optical Society of America
1. Introduction
Silicon photonics has drawn tremendous interest since it has a huge potential to realize on-chip photonic integrated circuits (PICs) with compact size, high performance, energy-efficient and high-yield, where these advantages mainly come from the commonly used silicon-on-insulator (SOI) platform with high-index-contrast and complementary metal-oxide-semiconductor (CMOS)-compatible processing [1]. On the other hand, however, this intrinsic high-index-contrast feature of the SOI platform inevitably induces strong polarization-dependence, which has greatly limited SOI-based PICs from large-scale application in optical communications [2]. To overcome this drawback, polarization diversity scheme composed of polarization splitters [3,4] and rotators [5,6] is normally leveraged to achieve a polarization-transparent circuit [7], but this scheme is relatively complex including a significantly increased system footprint. Noteworthy that, if the polarization-division-multiplexing transmission is not necessary, a more simple and cost-effective approach is to use polarizer that can efficiently output the desired polarization state and strip off the unwanted one [8]. Over the years, many kinds of waveguide-type polarizers have been proposed utilizing various structures such as tapered waveguides [9], shallowly-etched ridge waveguides [10], subwavelength gratings [11,12], hybrid plasmonic waveguides (HPWs) [13–15], and employing special materials such as graphene [16] and vanadium dioxide (VO2) [17]. Their principles can be roughly categorized as: mode attenuation [10,12,15–17], mode cutoff [9,11,13], and resonant coupling [14]. The mode attenuation usually requires a relatively long device length to obtain well performance, e.g., ~1 mm [10], 60 μm [12], and 30 μm [15]. As to the mode cutoff, significantly reduced device length can be achieved, e.g., ~10 μm [9], and this length can be further shrunk using HPW structure whose active region length, for instance, is only 1 μm at the expense of relatively high insertion loss (IL) (2.2 dB) and low extinction ratio (ER) (16 dB) [13]. For the resonant coupling, the polarizer shows a strong performance-dependence on the structural dimensions, particularly for the device length, thus leading to stringent fabrication requirements [14]. In addition, using multilayer graphene structure embedded in the slot region of the horizontal slot waveguide, a polarizer with ER of ~30 dB and IL of 0.2 dB is achieved in a length of 7 μm, but this device will pose huge challenges for the fabrication process [16]. Vanadium dioxide (VO2) has also been employed to design polarizer, while its performance relies on the insulator-to-metal phase transition with a critical temperature of ~68°C, requiring relatively high energy consumption and the obtained performance is needed to be improved substantially (IL ~3 dB, ER ~12 dB) [17]. Therefore, it is still greatly desired and required to explore new schemes to shorten the device length with low IL and high ER, as well as acceptable fabrication requirements.
Recently, transparent conducting oxides (TCOs), e.g., indium tin oxide (ITO), aluminum zinc oxide (AZO) and gallium zinc oxide (GZO), have gathered significant attention as promising active materials due to their unique properties such as epsilon-near-zero (ENZ) effect in the near infrared and electrically-tunable permittivity [18,19]. Accordingly, TCOs have been successfully applied to a variety of electro-absorption modulators in silicon photonics [20–25], which have large modulation depth, ultracompact size (even in the nanometer scale [25]) and low energy consumption due to their extremely strong light-matter interaction and absorption in the ultra-thin electron accumulation layer at the ENZ point, revealing great potentials to realize on-chip ultra-dense PICs. It is remarkable that, however, these features of TCOs are dependent on the polarization state, which can also be employed to design novel polarization-handling devices with enhanced performance.
In this paper, we propose and numerically analyze a compact, broadband and low-loss TE-pass polarizer based on mode attenuation by utilizing an ITO layer embedded in the center of the silicon-based strip waveguide and a second ITO layer deposited on its top, where high-k dielectric Hafnium dioxide (HfO2) is inserted between ITO and strip waveguide as the insulator layer to induce an electron accumulation layer at the ITO/HfO2 interface under bias. With the tunable permittivity of ITO, ENZ of its real part can be achieved around the wavelength of 1.55 μm with a certain electron concentration and then the TM mode will be strongly confined in the accumulation layers of ITO with high absorption loss. After a short propagation length, the input TM mode will be strongly attenuated. As to the TE mode, its interaction with ITO layers is quite weak and thus its propagation behavior almost cannot be affected. Within the design, the second ITO layer is employed to further enhance the device performance, especially for improving ER through extremely attenuating TM mode in a short length. From results, the polarizer length is only 4.5 μm with an ER of 25.26 dB and an IL of 0.21 dB at 1.55 μm, where the bandwidth can be enlarged to ~140 nm for an ER>20 dB. Compared with polarizers based on mode attenuation reported earlier, the suggested device has more compact size [10,12,15], which is even slightly smaller than the one using multilayer graphene structure with comparable performance [16], and the device performance (including IL, ER, and bandwidth) is superior to the one using VO2 as the plasmonic material [17]. Moreover, considering the fabrication, the proposed polarizer needs film depositions and etching processes on the SOI platform, and the layer thicknesses and numbers are clearly thicker and lower than those in [16], respectively, which can be beneficial to relax the fabrication requirements and achieve a high performance on-chip polarizer.
2. Device structure and principle
Figure 1 shows the schematic of our proposed TE-pass polarizer. An ITO layer with thickness of h1 sandwiched by two layers of HfO2 with identical thickness of h2 is embedded in the center of the strip waveguide. The device’s operating principle is based on different mode confinements and loss characteristics for both polarizations. When the TM mode is launched into this polarizer, it can be extremely confined in the ultra-thin accumulation layers of ITO if ENZ effect is well satisfied under a certain electron concentration, where the corresponding absorption loss is also very high due to nearly two orders of magnitude increased for the imaginary part of the refractive index of ITO in the accumulation layer (more details will be presented in the following analysis). Therefore, the input TM mode will rapidly attenuate, leading to a great reduction in length. For the TE mode, its propagation behavior may not be affected by the embedded structure because of its polarization direction parallel to the ITO surface and ultra-thin thicknesses of the accumulation layers of ITO, only ~1 nm on one side of ITO, as determined by Thomas-Fermi screening theory [26], corresponding to a low-loss transmission. To further improve the device performance, we introduce a second ITO layer with thickness of h4 deposited atop of the strip waveguide and a thin HfO2 layer with thickness of h3 is also inserted between them. Based upon this technique, the absorption loss of TM mode can be greatly increased due to the increased accumulation layer, whereas the TE mode suffers little influence since it is mostly distributed in the core region and the added ITO layer especially for the generated accumulation layer has little interaction with it. Finally, at the output port, we can only get the TE mode with high ER and low IL, where the TM mode is effectively absorbed.
Fig. 1 Schematic diagram of the proposed TE-pass polarizer, where its cross-section view of the polarizer region is also illustrated. h1 and h2 represent the thicknesses of ITO and HfO2 layers embedded in the center of the strip waveguide, h3 and h4 represent the thicknesses of HfO2 and ITO layers deposited on the top of the strip waveguide, respectively.
Within our design, the tunable permittivity of ITO and the operating condition of its ENZ effect are pivotal. Accordingly, the permittivity of ITO is analyzed, which can be described by Drude model [20–25],
where ε∞ is the high-frequency dielectric constant, ω is the angular frequency, γ is the electron scattering frequency, n is the electron concentration, m* is the effective mass of the electron, and e and ε0 are the electron charge and the vacuum permittivity, respectively. For ITO, ε∞ = 3.9, γ = 1.8 × 1014 rad/s, and m* = 0.35 me (where me is the electron mass) [26]. Figures 2(a) and 2(b) plot the calculated real and imaginary parts of ITO’s permittivity at different electron concentrations n within the wavelength range from 1.2 to 1.9 μm. For bulk material (n~1019 cm−3), its permittivity shows a quite low dispersion. When the electron concentration increases, however, dispersion can be clearly observed and the real and imaginary parts of the permittivity present opposite behaviors. Particularly, as n reaches to ~6.47 × 1020 cm−3, the real part of permittivity approaches zero (ENZ) at the wavelength of 1.55 μm (~0.0018) and the corresponding imaginary part is increased to ~0.5774 which is almost two orders of magnitude increased compared with that of n~1019 cm−3 (Im(ε) ~0.0063), thus leading to high absorption loss. Under this condition, huge enhancement of the electric field of the TM mode can be formed at the accumulation layers of ITO due to the continuity of the normal component of the electric displacement field and the absorption loss is also highest in these accumulation layers [20, 21]. As a consequence, light-matter interaction and absorption will reach the maximum in the nanometer scale, resulting in a huge attenuation for the TM mode. To achieve this required electron concentration in the accumulation layers of ITO, we employ the current injection through electrode and the electrode configurations are shown in Fig. 2(c), where bias voltages U between p-type silicon and ITO layers are applied. In addition, other computational parameters under the following analysis are: refractive indices of Si, HfO2 and SiO2 are taken as 3.478, 2.071 and 1.444 at the wavelength of 1.55 μm, and the waveguide width and height are set to be w = 400 nm and h = 250 nm, respectively.Fig. 2 (a) Real and (b) imaginary parts of ITO’s permittivity versus wavelength for different electron concentrations n, where the dotted lines indicate the required electron concentration to generate ENZ in the accumulation layer at the wavelength of 1.55 μm. The corresponding electrode configurations are illustrated in (c).
3. Results and discussion
To accurately analyze the modal characteristics of the polarizer, a full-vectorial mode solver based on the finite-difference frequency-domain method [27] is utilized. Figure 3 depicts the real and imaginary parts of effective indices of guided modes as a function of the electron concentration n for both polarization modes, where an ITO layer sandwiched by two HfO2 layers is embedded in the center of the strip waveguide (h1 = 10 nm, h2 = 5 nm). One can see that the variation range of the real part of effective index of TE mode is very small including a quite small value of its imaginary part (nearly order of 10−4), corresponding to a low-loss mode, while both real and imaginary parts of effective index of TM mode are sensitive to the electron concentration [21]. It is worthwhile to note that, if n reaches to ~6.47 × 1020 cm−3 (marked by gray line), the value of real part of effective index of TM mode is nearly located at its fastest-changing region and that of imaginary part reaches to the maximum, corresponding to the highest absorption loss. Moreover, the figure insets show the corresponding electric field distributions under several typical electron concentrations, and we can clearly observe that the TM mode is strongly confined in the accumulation layers of ITO at n~6.47 × 1020 cm−3 while the TE mode is mainly distributed in the whole core region almost without influence of n. To better show different mode confinement features, we also calculate the mode confinement factor Γ for both polarization modes (Γ = Pacc/Ptotal, where Pacc represents the power in the accumulation layers of ITO, Ptotal represents the total power [28]). At n~6.47 × 1020 cm−3, the mode confinement factors are calculated to be ~0.4123 for TM mode and ~0.0034 for TE mode, respectively, revealing that huge mode confinement for TM mode in the nanometer scale is obtained from the ENZ effect. Therefore, this strong polarization dependence provides us a route to design TE-pass polarizer and we analyze the transmission features of this waveguide structure by using a 3D finite-difference time-domain (FDTD) method [29], as illustrated in Fig. 4, where the polarizer length is chosen to be 4.5 μm and the inset shows the device structure [Fig. 4(a)]. The ER and IL of the device are calculated to evaluate the device performance, and their definitions are expressed as follows:
where Pyx stands for the power at the x port (1: input port, 2: output port) with y polarization state (TE or TM polarization mode). Note that although IL is relatively low (~0.13 dB @1.55 μm), the highest ER is lower than 20 dB and 100 nm bandwidth will make ER of the polarizer reduce to 14.6 dB, thus its performance is highly required to be improved. In addition, Figs. 4(b) and 4(c) show the field evolution of TE and TM modes through the device at the wavelength of 1.55 μm and residual TM mode can be observed at the output port. To effectively enhance the device performance, a simple approach is to increase the polarizer length to further absorb and attenuate the TM mode, but it will make the polarizer relatively long and not be beneficial to the on-chip ultra-dense integration [2,3]. Thus, a more suitable choice is to increase the absorption efficiency for the TM mode and the corresponding ER will also be increased naturally, while IL had better to be kept within the low value.Fig. 3 Real and imaginary parts of the effective indices of guided modes as a function of the electron concentration n for (a) the TE mode and (b) the TM mode in the polarizer region. The insets show the electric field distributions at n = 0.1 × 1020, 6.47 × 1020, 1.0 × 1021cm−3 and gray lines represent the electron concentration of the ENZ point.
Fig. 4 (a) Wavelength dependence of the polarizer with an ITO layer embedded in the waveguide center which is illustrated in the inset, and field evolution of (b) the TE (Ex) and (c) TM (Ey) modes along the propagation direction through the polarizer, where the polarizer length is 4.5 μm.
With the requirements of compact size and high performance, we propose an improved TE-pass polarizer with a second ITO layer deposited on the top of the strip waveguide where an HfO2 layer is also employed as an insulator layer and the electrode configurations are similar with those of Fig. 2(c), only adding another electrode to contact the second ITO layer. The improved design can introduce an extra high-loss accumulation layer to better absorb the TM mode while it almost does not affect the propagation behavior of the TE mode and the increased fabrication processes are relatively simple compared with those of embedded ITO layer in the strip waveguide. Figure 5 depicts the ER and IL of the improved design as its thicknesses of ITO and HfO2 vary (h1 and h2 in the central layer, h3 and h4 on the top layer). From Fig. 5(a), the device performance gradually deteriorates with the increase of h1 and h2 since the absorption loss of TM mode is decreased (imaginary part of its effective index decreases through further modal analysis) and these increased thicknesses will introduce higher IL for the TE mode. Accordingly, we make a compromise from both considering the performance and the fabrication requirements [22, 23], and the chosen values are h1 = 10 nm and h2 = 5 nm under the following analysis. For the top structure, h3 and h4 have optimum values about 5 and 50 nm, respectively, as shown in Fig. 5(b). Moreover, the calculated ER is 25.26 dB, which is increased by ~6.7 dB, and the IL is slightly increased to 0.21 dB, including low reflection loss (<-34 dB for TE mode, <-25 dB for TM mode). The wavelength dependence of the improved polarizer is also analyzed, where different electron concentrations of the accumulation layers of ITO are considered, as shown in Fig. 6. Within the calculated wavelength range from 1.2 to 1.9 μm, the curves of ER show a blue shift and their peak values are also increased with the increase of the electron concentration, which are consistent with the feature of the tunable permittivity of ITO, that is to say, when n increases, its real part especially the ENZ point is shifted to the short wavelength and its imaginary part is increased [Figs. 2(a) and 2(b)]. Meanwhile, IL is also increased due to larger imaginary part of the permittivity of ITO [Fig. 3(a)]. If ER>20 dB is required, the bandwidth is around 140 nm (from 1.474 to 1.614 μm) for n~6.47 × 1020 cm−3, where the corresponding IL is lower than 0.22 dB within the same bandwidth. Besides, the bandwidths for n~7.0 × 1020 and 7.5 × 1020 cm−3 are around 174 nm (from 1.4 to 1.574 μm) and 195 nm (from 1.34 to 1.535 μm) locating at the central wavelengths of 1.49 and 1.43 μm, respectively, which illustrate good tunability and device performance of the design.
Fig. 5 ER and IL of the improved polarizer as functions of (a) the ITO layer thickness h1 for different HfO2 thicknesses h2 in the waveguide center and (b) the layer thicknesses of HfO2 h3 and ITO h4 on the waveguide top.
Fig. 6 Wavelength dependence of the improved polarizer for different electron concentrations of the accumulation layers of ITO, where the horizontal dash-dotted line represents ER of 20 dB which is employed to estimate the bandwidth.
For the device fabrication, a SOI wafer with a 250 nm thick top silicon layer is employed. Firstly, the strip waveguide is patterned with deep ultraviolet lithography or e-beam lithography, and an inductively coupled plasma etching process is followed to make silicon-based strip waveguide with a width of 400 nm. Then, second lithography and etching processes are employed to define the polarizer region with a longitudinal length of 4.5 μm, where the polarizer region is plasma etched with 135 nm depth. Next, a thin HfO2 layer (h2 = 5 nm), an ITO layer (h1 = 10 nm), and a second thin HfO2 layer (h2 = 5 nm) are successively deposited on the etched strip waveguide in the polarizer region, where the thin HfO2 layer is deposited using atomic-layer-deposition (ALD) due to its high precision [21,23] and the ITO layer is deposited using magnetron sputtering [23]. Subsequently, a silicon film with a height of 115 nm is grown by epitaxy technology, e.g., molecular beam epitaxy, in the polarizer region, and at this point, the waveguide height of the polarizer region is same with that of the input/output strip waveguide. Finally, two top layers of HfO2 (h3 = 5 nm) and ITO (h4 = 50 nm) are also successively deposited on the silicon film in the polarizer region, and the upper cladding of SiO2 film is deposited using plasma enhanced chemical vapor deposition (PECVD) process. It is noted that, however, unwanted film layers outside of the waveguide core generated by the film deposition processes should be cleared away before depositing the final upper cladding. Within the fabrication process, the key films’ thicknesses can be precisely controlled using ALD, magnetron sputtering, and PECVD [23]. Accordingly, we investigate the variations of the vertical position of the central ITO layer and the horizontal position (in z-direction) of the top ITO layer generated by fabrication errors. Figure 7(a) shows the device performance as these positions offset from their designed values. One can see that the TE mode is hardly affected by these offsets and the largest deviation of ER which is dominant by the offset of the central ITO layer (−30 nm) is only ~2 dB within the calculated range, where ± 150 nm offset of the top ITO layer almost cannot cause obvious performance variation, thus indicating that the suggested polarizer has a relatively large tolerance for ITO’s position. Moreover, fabrication errors can also be compensated by tuning the permittivity of ITO suitably [21, 22]. Figure 7(b) shows the dependence of ER and IL on the polarizer length L. It is noted that both ER and IL increase monotonously with the polarizer length, but the growth rate of ER is much higher and IL is within the acceptable scope. This means that the proposed polarizer has the capability to significantly increase the ER only by increasing the length of the polarizer, for instance, as L increases to 8.0 μm, ER = 34.74 dB, IL = 0.43 dB also with an increased bandwidth ~254 nm (from 1.414 to 1.668 μm), and meanwhile the design offers good flexibility in use, that is, we can choose the suitable length of the polarizer depending on the requirements with respect to ER and IL.
Fig. 7 (a) ER and IL of the device as functions of the central layer offset and the top layer offset, and (b) length dependence of the designed polarizer.
To demonstrate the field evolution of the designed TE-pass polarizer, 3D FDTD simulation is also performed, as plotted in Fig. 8, where h1 = 10 nm, h2 = 5 nm, h3 = 5 nm, h4 = 50 nm, n = 6.47 × 1020 cm−3, L = 4.5 μm, and λ = 1.55 μm. The input TE mode can directly pass through the polarizer nearly without influence of the added ITO layers. As to the input TM mode, it attenuates rapidly once it enters into the polarizer which is greatly absorbed by the accumulation layers of ITO, where the inset shows the field evolution in y-z cross-section and we can clearly observe the strong confinement of TM mode in the ultra-thin regions and mode attenuation along the propagation direction. Therefore, a compact, broadband and low-loss TE-pass polarizer is realized. In addition, the polarization conversion behavior cannot be obviously observed in such a relatively short length since the device structure is symmetric with respect to the central vertical line without introducing asymmetric structures and the mode characteristics between TE and TM modes have huge difference [Figs. 3(a) and 3(b)], where the polarization conversion even cannot be effectively realized using mode hybridization [30].
Fig. 8 Field evolution of (a) the TE (Ex) and (b) TM (Ey) modes through the designed TE-pass polarizer in x-z cross section (y = 150 nm), where insets show the field evolution through the device in y-z cross section (x = 0).
4. Conclusion
In summary, we have proposed and numerically analyzed a compact, broadband and low-loss TE-pass polarizer using TCOs embedded in the center of the strip waveguide and deposited on its top. The electron concentration of the accumulation layers of TCO has been investigated to realize ENZ effect around the wavelength of 1.55 μm. Accordingly, the TM mode can be strongly confined in the accumulation layers of TCO with high absorption loss due to induced ENZ effect and significantly increased imaginary part of the modal effective index, leading to a great reduction in polarizer length. The TE mode however is hardly affected by the introduced TCO and directly transmits along the waveguide and outputs due to structural polarization-dependence. Moreover, the top TCO layer is applied to enhance the polarizer performance through adding an extra high-loss accumulation layer for the TM mode. The results show that the present TE-pass polarizer, with a length of only 4.5 μm, has a broad bandwidth of ~140 nm for an ER>20 dB, where ER and IL are 25.26 dB and 0.21 dB at wavelength of 1.55 μm, respectively, and the fabrication tolerance is also relatively large in terms of TCO’s position offset. Furthermore, the proposed polarizer can achieve an ultra-high ER (~35 dB) only by enlarging its length to 8 μm. With these advantages, the suggested TE-pass polarizer has promising applications for realizing on-chip high performance PICs where TE polarization is needed.
Acknowledgments
This work was supported by the National Natural Science Foundation of China (NSFC) (No. 11574046, 60978005), the Jiangsu Provincial Natural Science Foundation (No. BK20141120), and the Scientific Research Foundation of Graduate School of Southeast University (No. YBJJ1556).
References and links
1. H. Subbaraman, X. Xu, A. Hosseini, X. Zhang, Y. Zhang, D. Kwong, and R. T. Chen, “Recent advances in silicon-based passive and active optical interconnects,” Opt. Express 23(3), 2487–2510 (2015). [CrossRef] [PubMed]
2. D. Dai, J. Bauters, and J. E. Bowers, “Passive technologies for future large-scale photonic integrated circuits on silicon: polarization handling, light non-reciprocity and loss reduction,” Light Sci. Appl. 1(3), e1 (2012). [CrossRef]
3. X. Guan, H. Wu, Y. Shi, and D. Dai, “Extremely small polarization beam splitter based on a multimode interference coupler with a silicon hybrid plasmonic waveguide,” Opt. Lett. 39(2), 259–262 (2014). [CrossRef] [PubMed]
4. Y. Xu and J. Xiao, “Compact and high extinction ratio polarization beam splitter using subwavelength grating couplers,” Opt. Lett. 41(4), 773–776 (2016). [CrossRef] [PubMed]
5. A. Xie, L. Zhou, J. Chen, and X. Li, “Efficient silicon polarization rotator based on mode-hybridization in a double-stair waveguide,” Opt. Express 23(4), 3960–3970 (2015). [CrossRef] [PubMed]
6. Y. Xu, J. Xiao, and X. Sun, “A compact hybrid plasmonic polarization rotator for silicon-based slot waveguides,” IEEE Photonics Technol. Lett. 26(16), 1609–1612 (2014). [CrossRef]
7. T. Barwicz, M. R. Watts, M. A. Popovic, P. T. Rakich, L. Socci, F. X. Kartner, E. P. Ippen, and H. I. Smith, “Polarization-transparent microphotonic devices in the strong confinement limit,” Nat. Photonics 1(1), 57–60 (2007). [CrossRef]
8. M. Z. Alam, J. S. Aitchison, and M. Mojahedi, “Compact and silicon-on-insulator-compatible hybrid plasmonic TE-pass polarizer,” Opt. Lett. 37(1), 55–57 (2012). [CrossRef] [PubMed]
9. Q. Wang and S.-T. Ho, “Ultracompact TM-pass silicon nanophotonic waveguide polarizer and design,” IEEE Photonics J. 2(1), 49–56 (2010). [CrossRef]
10. D. Dai, Z. Wang, N. Julian, and J. E. Bowers, “Compact broadband polarizer based on shallowly-etched silicon-on-insulator ridge optical waveguides,” Opt. Express 18(26), 27404–27415 (2010). [CrossRef] [PubMed]
11. X. Guan, P. Chen, S. Chen, P. Xu, Y. Shi, and D. Dai, “Low-loss ultracompact transverse-magnetic-pass polarizer with a silicon subwavelength grating waveguide,” Opt. Lett. 39(15), 4514–4517 (2014). [CrossRef] [PubMed]
12. Y. Xiong, D.-X. Xu, J. H. Schmid, P. Cheben, and W. N. Ye, “High extinction ratio and broadband silicon TE-pass polarizer using subwavelength grating index engineering,” IEEE Photonics J. 7(5), 7802107 (2015). [CrossRef]
13. Y. Huang, S. Zhu, H. Zhang, T.-Y. Liow, and G.-Q. Lo, “CMOS compatible horizontal nanoplasmonic slot waveguides TE-pass polarizer on silicon-on-insulator platform,” Opt. Express 21(10), 12790–12796 (2013). [CrossRef] [PubMed]
14. Y. Xu and J. Xiao, “A compact TE-pass polarizer for silicon-based slot waveguides,” IEEE Photonics Technol. Lett. 27(19), 2071–2074 (2015). [CrossRef]
15. X. Sun, M. Z. Alam, S. J. Wagner, J. S. Aitchison, and M. Mojahedi, “Experimental demonstration of a hybrid plasmonic transverse electric pass polarizer for a silicon-on-insulator platform,” Opt. Lett. 37(23), 4814–4816 (2012). [CrossRef] [PubMed]
16. X. Yin, T. Zhang, L. Chen, and X. Li, “Ultra-compact TE-pass polarizer with graphene multilayer embedded in a silicon slot waveguide,” Opt. Lett. 40(8), 1733–1736 (2015). [CrossRef] [PubMed]
17. L. Sánchez, S. Lechago, and P. Sanchis, “Ultra-compact TE and TM pass polarizers based on vanadium dioxide on silicon,” Opt. Lett. 40(7), 1452–1455 (2015). [CrossRef] [PubMed]
18. V. J. Sorger, N. D. Lanzillotti-kimura, R.-M. Ma, and X. Zhang, “Ultra-compact silicon nanophotonic modulator with broadband response,” Nanophotonics 1(1), 17–22 (2012). [CrossRef]
19. V. E. Babicheva, A. Boltasseva, and A. V. Lavrinenko, “Transparent conducting oxides for electro-optical plasmonic modulators,” Nanophotonics 4(1), 165–185 (2015).
20. A. P. Vasudev, J.-H. Kang, J. Park, X. Liu, and M. L. Brongersma, “Electro-optical modulation of a silicon waveguide with an “epsilon-near-zero” material,” Opt. Express 21(22), 26387–26397 (2013). [CrossRef] [PubMed]
21. S. Zhu, G. Q. Lo, and D. L. Kwong, “Design of an ultra-compact electro-absorption modulator comprised of a deposited TiN/HfO₂/ITO/Cu stack for CMOS backend integration,” Opt. Express 22(15), 17930–17947 (2014). [CrossRef] [PubMed]
22. J. Baek, J.-B. You, and K. Yu, “Free-carrier electro-refraction modulation based on a silicon slot waveguide with ITO,” Opt. Express 23(12), 15863–15876 (2015). [CrossRef] [PubMed]
23. H. W. Lee, G. Papadakis, S. P. Burgos, K. Chander, A. Kriesch, R. Pala, U. Peschel, and H. A. Atwater, “Nanoscale conducting oxide PlasMOStor,” Nano Lett. 14(11), 6463–6468 (2014). [CrossRef] [PubMed]
24. H. Zhao, Y. Wang, A. Capretti, L. D. Negro, and J. Klamkin, “Broadband electroabsorption modulators design based on epsilon-near-zero indium tin oxide,” IEEE J. Sel. Top. Quantum Electron. 21(4), 3300207 (2015).
25. Z. Lu, W. Zhao, and K. Shi, “Ultracompact electroabsorption modulators based on tunable epsilon-near-zero-slot waveguides,” IEEE Photonics J. 4(3), 735–740 (2012). [CrossRef]
26. A. V. Krasavin and A. V. Zayats, “Photonic signal processing on electronic scales: electro-optical field-effect nanoplasmonic modulator,” Phys. Rev. Lett. 109(5), 053901 (2012). [CrossRef] [PubMed]
27. J. Xiao, H. Ni, and X. Sun, “Full-vector mode solver for bending waveguides based on the finite-difference frequency-domain method in cylindrical coordinate systems,” Opt. Lett. 33(16), 1848–1850 (2008). [CrossRef] [PubMed]
28. K. Okamoto, Fundamentals of Optical Waveguides (Academic, 2006).
29. D. M. Sullivan, Electromagnetic Simulation Using the FDTD Method (IEEE, 2000).
30. D. Dai and J. E. Bowers, “Novel concept for ultracompact polarization splitter-rotator based on silicon nanowires,” Opt. Express 19(11), 10940–10949 (2011). [CrossRef] [PubMed]
