Open Access
Add to BibSonomy
Abstract
Carrier-injection-based efficient polarization controller with a strained InGaAsP multiple-quantum-well (MQW) layer is demonstrated on a regrowth-free InP platform. We employ a straight-line device configuration by cascading an asymmetric polarization rotator (PR) to provide a fixed polarization conversion and a polarization-dependent phase shifter (PD-PS) to enable tunable polarization rotation. Based on a novel design concept, both the PR and PD-PS sections are integrated monolithically without active-passive integration. Using the fabricated device, we experimentally demonstrate efficient polarization conversion over the entire Poincaré sphere with a total current of less than 40 mA. With the capability of monolithically integrating other InP-based active components, the demonstrated scheme should be attractive for various applications, such-as low-cost coherent communication, microwave photonics, and quantum key distribution.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Polarization handling is becoming important in diverse applications, such as optical communication [1–5], microwave photonics [6,7], and quantum key distribution [8–11]. For example, in a so-called coherent-lite optical communication system, which is expected as a viable method to transmit coherent signals without using power-consuming digital signal processing (DSP) [1–3], high-speed polarization controllers are indispensable. For these applications, monolithically integrated polarization controllers are highly desired to achieve a significant improvement in operating speed as well as a reduction in size, cost, and power consumption.
Among various types of polarization controllers (PC) integrated on silicon and InP platforms [12–20], those based on a straight-line waveguide configuration [14–20] are attractive due to the simplicity and compactness. By cascading multiple stages of polarization rotator (PR) and polarization-dependent phase shifter (PD-PS) sections, conversion to an arbitrary state of polarization (SOP) can be achieved [14,17,18]. The PR section is implemented by an asymmetric waveguide that supports two eigenmodes with tilted electric/magnetic field vectors. In contrast, PD-PS is realized by a symmetric waveguide with an electrode to enable active tuning of the phase offset between the transverse-electric (TE) and transverse-magnetic (TM) components. For efficient and high-speed operation, the use of a strained multiple-quantum-well (MQW) active layer is effective due to the large inherent birefringence, which can be tuned through a current injection or a reverse bias. On the other hand, such MQW structure is generally unfavorable at the PR section since the inherent birefringence of MQW disturbs the optimal rotation of SOP on the Poincarè sphere. In the previous demonstration, therefore, butt-joint active-passive integration technique had to be used to insert an MQW stack only at the PD-PS sections [19].
In this paper, to realize a highly efficient electrically tunable PC by a simple fabrication process, we demonstrate an all-active PC with a compressively strained MQW layer on a regrowth-free InGaAsP/InP platform. Using the fabricated device, efficient conversion of the SOP over the entire Poincaré sphere is demonstrated with a total current of less than 40 mA.
2. Principle
Figure 1 illustrates the schematic of the monolithic InP-based polarization controller considered in this work, which comprises the PR and PD-PS sections with an MQW layer in a straight-line configuration. As shown in Fig. 1(b), the PR section has an asymmetric cross-section and is used to provide a fixed rotation to the Stokes vector (SV) on the Poincaré sphere. While a variety of asymmetric waveguides may be used as PR, the half-ridge waveguide shown in Fig. 1(b) provides a unique advantage that low-loss integration of active components is feasible owing to the thick InP cladding [17–19,21,22]. The PD-PS section, on the other hand, has a symmetric cross-section as shown in Fig. 1(c) and enables tunable rotation of the SV around the S1 axis under an external electrical control. By cascading two stages of these PR and PD-PS sections, we can convert an input TE or TM mode into an arbitrary SOP at the output [17,18].
Fig. 1. Schematic of the monolithic polarization controller on InP with strained MQW: (a) the entire structure, (b) the PR section, and (c) the PD-PS section. As the light propagates inside the respective waveguides, its Stokes vector rotates about Ω on the Poincaré sphere.
For efficient operation of the polarization controller, PD-PS needs to have a large phase difference between the TE and TM modes when an electrical bias is applied. To this end, the use of a compressively strained MQW active layer is effective. Due to the splitting of the heavy-hole (HH) and light-hole (LH) energy levels in a strained MQW, a large polarization-dependent band-filling effect can be obtained under current injection [19]. On the other hand, such MQW layer is generally unfavorable at the PR section, since it hinders the necessary fixed polarization rotation. Here, however, we demonstrate that by properly adjusting the waveguide structure at the PR section, necessary polarization rotation can be obtained even with the MQW layer, thus avoiding the need for active-passive integration.
As the light propagates through the PR section, its SV ${\textbf S}$ rotates about a birefringence vector ${\boldsymbol {\mathrm {\Omega}}}$ as [18,23]
Here, $\psi $ represents the effective tilt angle of the two eigenmodes of the asymmetric waveguide as illustrated in Fig. 1(b). On the other hand, ${\beta _1}$ and ${\beta _2}$ are the propagation constants of these eigenmodes and are related to the half-beat length by ${L_\pi } = \pi /|{{\beta_1} - {\beta_2}} |$. Assuming a PR section with a length L, ${\textbf S}$ rotates by an angle $|{\boldsymbol {\mathrm {\Omega}}}|L$. In order to realize a perfect PC that can convert input TE or TM light into an arbitrary SOP, the SV after the first PR [PR1 in Fig. 1(a)] needs to lie on the S2-S3 plane. We can understand from Fig. 1(b) that such a condition can only be satisfied if the vector ${\boldsymbol {\mathrm {\Omega}}}$ is located inside the gray shaded region in Fig. 2, or in other words, if $\psi $ is within the range between 22.5° and 67.5°.
Fig. 2. Required regimes for Ω and ΩAW on the S1-S2 plane to achieve polarization conversion over the entire Poincaré sphere.
When an MQW layer is inserted inside the asymmetric waveguide, we need to consider the additional birefringence induced by the MQW itself. To this end, we model the PR section by successively cascading infinitely thin layers of (i) an asymmetric waveguide without considering the birefringence of MQW and (ii) a waveplate having an effective birefringence of the MQW layer. Such approximation should be verified as long as the thickness of each layer Δd is small enough [23]. From Eq. (1), the change in Stokes vector after propagating one period of such thin layers can be expressed as
3. Device design and fabrication
First, the MQW stack needs to be designed properly to obtain a large polarization-dependent phase shift under current injection, while keeping the absorption below an acceptable level. In this work, the offset-QW design is selected, where a strained InGaAsP/InGaAsP MQW active layer is located on top of the bulk InGaAsP core layer to provide a modest confinement.
Figure 3 shows the designed epitaxial layer profile, together with the simulated spectra of absorption and effective refractive index change of the waveguide mode under carrier injection. A commercial software (Harold from Photon Design [24]) is employed. Assuming a PD-PS with 1000 µm length and 2.5 µm width, and the material parameters from literatures [25–28], the carrier density of 2.3 × 1018 cm−3 should be attained at a forward-biased current of around 18 mA. Due to the 0.8% compressive strain in the well, we can see in Fig. 3 that the absorption edges for the TE and TM modes are largely shifted and located at around 1400 nm and 1300 nm, respectively. As a result, under a current injection, a large polarization-dependent refractive index modulation of 8.0 × 10−4 is obtained with a negligible absorption below 3.3 cm−1 for both polarization modes at 1550 nm wavelength.
Fig. 3. (a) Epitaxial layer profile of the designed device. (b, c) Simulated spectra of (b) absorption and (c) effective refractive index change of the waveguide mode for TE and TM polarization modes under carrier injection.
Using the refractive index of the designed epitaxial layer for the TE and TM modes, ${{\boldsymbol {\mathrm {\Omega}}}_{\textrm{MQW}}}$ is derived by Eq. (2). Then, the asymmetric waveguide structure at the PR section is designed, so that ${{\boldsymbol {\mathrm {\Omega}}}_{\textrm{AW}}}$ lies inside the green shaded regime in Fig. 2. To this end, the eigenmode analysis based on the finite-element method is employed. Figure 4 shows the angle $\psi $ and the half-beat length ${L_\pi }$ of the eigenmodes at 1550 nm as a function of W and d [see Fig. 1(b) for definitions].
Fig. 4. Tilted angle of eigenmodes $\; \psi $ (°) (a) and the half-beat length ${L_\pi }$ (µm) at the PR section (b) as a function of W and d at 1550 nm wavelength.
Considering the fabrication tolerance, W and d are selected to be 1060 nm and 375 nm, respectively. In this case, $\psi $ = 42.1° and ${L_\pi }$ = 297 µm, so that ${{\boldsymbol {\mathrm {\Omega}}}_{\textrm{AW}}}$ lies inside the required regime as shown in Fig. 2. Finally, following the procedure described in Section 2, L is determined to be 150 µm to satisfy the condition of converting the SV to a state on the S2-S3 plane at the PR1 output. As for the PD-PS sections, the width and the length at PD-PS1 are set to be 2.5 µm and 2 mm, respectively. The length of PD-PS2 is reduced to 1 mm since only π phase shift is required at PD-PS2. From the calculated sensitivities of the eigenmodes shown in Fig. 4, we estimate that the W and d need to be controlled within ±40 nm and ±50 nm, respectively, to achieve nearly arbitrary polarization conversion, covering more than 96% of the entire surface of the Poincaré sphere.
The device was fabricated by a simple regrowth-free process, similar to the previous work [17]. After the formation of ridge waveguides by CH4/H2-based inductively-coupled plasma reactive ion etching (ICP-RIE), the asymmetric structures at the PR sections were formed by a self-aligned dry-etching process based on the angled electron-beam evaporation of SiO2 [21]. After the passivation with an SiO2 layer and the planarization using a polyimide layer, the top and bottom Ti/Au contacts were formed. Figure 5 shows the top photograph of the entire device as well as the cross-sectional scanning-electron microscope (SEM) images at respective sections. The total device length was 4.5 mm. Due to some fabrication errors, W and d of the actual device were 1160 nm and 315 nm, respectively.
Fig. 5. (a) Top photograph of the entire device. Cross-sectional SEM images at (b) PR and (c) PD-PS sections.
4. Measurement
We used a similar experimental setup as in [21] to characterize the fabricated device. In order to avoid the effects of reflections at the device facets, we employed an incoherent light source with the center wavelength of 1550 nm and the 3-dB bandwidth of 2 nm, which was generated by spectrally slicing the amplified spontaneous emission from an erbium-doped fiber amplifier (EDFA) with an arrayed-waveguide grating (AWG). The light was polarized and aligned to the TE mode at the input of the device by using a polarizer followed by a half-wave plate and a quarter-wave plate. The SOP of the output light from the device was measured by a polarization analyzer (Thorlabs, TXP5004).
First, the PR and PD-PS sections are individually characterized. By observing the output SOP from a test structure with only the PR section, $\psi $ and ${L_\pi }$ of the actual fabricated device are derived to be 45.0° and 380 µm, respectively. The small deviations from the designed values (42.1° and 297 µm) are attributed to the slightly over-etched core layer and non-vertical sidewalls, as we can confirm from Fig. 5(a). Indeed, for the actual structure considering these effects, $\psi $ and ${L_\pi }$ of the eigenmodes are calculated to be 45.0° and 349 µm, which agree well with the measured values. Figure 6(a) shows the polarization-dependent refractive index change as a function of current density, measured for the PD-PS structure shown in Fig. 5(b). We can confirm a reasonable agreement between the measurement and the calculation, which was derived from the numerical results presented in Fig. 3(c). The response time of the PD-PS is evaluated by using a test Mach-Zehnder interferometer with a PD-PS attached on one arm. Figure 6(b) shows the observed output optical waveform when a square pulse is applied. The rise and fall times are measured to be 2.64 ns and 10.2 ns, respectively. In addition to the carrier lifetime, this bandwidth may be limited by the parasitic capacitance of the electrode pads and the cabling. The operational speed should be enhanced by utilizing the polarization-dependent quantum-confined Stark effect (QCSE) [29,30] or the carrier-depletion effect [31] under a reverse bias.
Fig. 6. (a) Measured polarization-dependent refractive index changes at PD-PS as a function of current density together with the simulated result. (b) Measured temporal response of a test Mach-Zehnder interferometer with PD-PS attached on one arm.
We then evaluate the performance of the entire chip shown in Fig. 1(a). Figure 7(a) shows the measured output SOP as we increase the current injection to PD-PS1 (I1) from 0 to 25 mA, while the current to PD-PS2 (I2) is fixed to 0 mA. As expected, the output SV rotates around a circular trajectory. Then, Fig. 7(b) shows the case when we change both I1 and I2 as shown in the inset. For the sake of clarity, we plot each Stokes component of the same measured data as a function of I2 in Fig. 7(c)–(e). In all cases, the total current is less than 40 mA. We can see that the SV rotates in a nearly orthogonal direction by PD-PS2 with an efficiency of π rotation at 14 mA, which is comparable to the previous work [19] fabricated by the butt-joint active-passive integration. As a result, nearly entire surface of the Poincaré sphere is covered by tuning both PD-PS1 and PD-PS2. The existence of small circular regions around the S1 axis that cannot be converted to is attributed to the residual S1 component at the output of PR1. To attain complete coverage, the polarization rotation at the PR1 needs to be improved by minimizing the fabrication errors. This should be achieved, for example, by using an etching-stop layer [32]. Alternatively, we could insert one more PR and PD-PS sections to add a redundancy and compensate for the fabrication errors [16].
Fig. 7. Measured SOP of the output light with increasing I1 and I2. (a) I1 is increased from 0 to 25 mA, while I2 is fixed to 0 mA. (b) I1 and I2 are varied within the ranges shown in the inset. (c, d, e) Same experimental data as (b), but S1, S2, and S3 are plotted as a function of I2 for various I1.
The total on-chip loss of the current device was measured to be around 2 dB. From a reference sample, the propagation loss at the PD-PS section under 0 mA (which is equivalent to a passive symmetric waveguide) was derived to be 3.1 dB/cm and 3.5 dB/cm for the TE and TM modes, respectively. As a result, the total loss at the PD-PS accounted for around 1 dB with a negligible polarization dependent loss (PDL). However, this loss increased by around 0.9 dB under a current injection of 14 mA, corresponding to π polarization rotation in PD-PS2. The excess loss should be reduced by a further optimization of the MQW design.
5. Conclusion
We have designed, fabricated, and demonstrated a carrier-injection-based polarization controller with a strained MQW layer on a regrowth-free InGaAsP/InP platform. We employed a straight-line configuration comprised of an asymmetric PR section to provide a fixed polarization rotation and a symmetric PD-PS section to enable tunable polarization rotation. By using a novel design concept to optimize the waveguide structure with a birefringent MQW layer at the PR section, both the PR and PD-PS sections were integrated monolithically on a compact InP chip, thus avoiding the need for active-passive integration. Using the fabricated device, efficient conversion of the SOP to nearly entire surface of the Poincaré sphere was demonstrated with a total current of less than 40 mA. With the high-speed response time in the nanoseconds order and the capability of monolithic integration with other InP active devices, the demonstrated device should be useful for various applications.
Acknowledgement
Portions of this work were presented at the Optical Fiber Communication Conference (OFC) in 2020, “Monolithic polarization controller on regrowth-free InGaAsP/InP platform with strained MQW layer” (paper T4H.4).
Disclosures
The authors declare no conflict of interest.
References
1. M. Morsy-Osman, M. Sowailem, E. El-Fiky, T. Goodwill, T. Hoang, S. Lessard, and D. V. Plant, “DSP-free ‘coherent-lite’ transceiver for next generation single wavelength optical intra-datacenter,” Opt. Express 26(7), 8890–8903 (2018). [CrossRef]
2. J. K. Perin, A. Shastri, and J. M. Kahn, “Design of low-power DSP-free coherent receivers for data center links,” J. Lightwave Technol. 35(21), 4650–4662 (2017). [CrossRef]
3. T. Gui, X. Wang, M. Tang, Y. Yu, Y. Lu, and L. Li, “Real-time demonstration of 600 Gb/s DP-64QAM self homodyne coherent bi-direction transmission with un-cooled DFB laser,” in proceeding of Optical Fiber Communication Conference (OFC2020) Postdeadline, Th4C.3 (2020).
4. K. Kikuchi and S. Kawakami, “Multi-level signaling in the Stokes space and its application to large capacity optical communications,” Opt. Express 22(7), 7374–7387 (2014). [CrossRef]
5. D. Che, A. Li, X. Chen, Q. Hu, Y. Wang, and W. Shieh, “Stokes vector direct detection for short-reach optical communication,” Opt. Lett. 39(11), 3110–3113 (2014). [CrossRef]
6. Y. Zhang and S. Pan, “Broadband microwave signal processing enabled by polarization-based photonic microwave phase shifters,” IEEE J. Quantum Electron. 54(4), 1–12 (2018). [CrossRef]
7. J. Capmany, J. Mora, I. Gasulla, J. Sancho, J. Lloret, and S. Sales, “Microwave photonic signal processing,” J. Lightwave Technol. 31(4), 571–586 (2013). [CrossRef]
8. C. Ma, W. D. Sacher, Z. Tang, J. C. Mikkelsen, Y. Yang, F. Xu, T. Thiessen, H. Lo, and J. K. S. Poon, “Silicon photonic transmitter for polarization-encoded quantum key distribution,” Optica 3(11), 1274–1278 (2016). [CrossRef]
9. H. Cai, C. M. Long, C. T. DeRose, N. Boynton, J. Urayama, R. Camacho, A. Pomerene, A. L. Starbuck, D. C. Trotter, P. S. Davids, and A. L. Lentine, “Silicon photonic transceiver circuit for high-speed polarization-based discrete variable quantum key distribution,” Opt. Express 25(11), 12282–12294 (2017). [CrossRef]
10. P. Sibson, J. E. Kennard, S. Stanisic, C. Erven, J. L. O’Brien, and M. G. Thompson, “Integrated silicon photonics for high-speed quantum key distribution,” Optica 4(2), 172–177 (2017). [CrossRef]
11. D. Bunandar, A. Lentine, C. Lee, H. Cai, C. M. Long, N. Boynton, N. Martinez, C. DeRose, C. Chen, M. Grein, D. Trotter, A. Starbuck, A. Pomerene, S. Hamilton, F. N. C. Wong, R. Camacho, P. Davids, J. Urayama, and D. Englund, “Metropolitan quantum key distribution with silicon photonics,” Phys. Rev. X 8(2), 021009 (2018). [CrossRef]
12. W. D. Sacher, Y. Huang, L. Ding, T. Barwicz, J. C. Mikkelsen, B. J. Taylor, G.-Q. Lo, and J. K. Poon, “Polarization rotator-splitters and controllers in a Si3N4 on-SOI integrated photonics platform,” Opt. Express 22(9), 11167–11174 (2014). [CrossRef]
13. P. Dong, X. Chen, K. Kim, S. Chandrasekhar, Y. Chen, and J. H. Sinsky, “128-Gb/s 100-km transmission with direct detection using silicon photonic Stokes vector receiver and I/Q modulator,” Opt. Express 24(13), 14208–14214 (2016). [CrossRef]
14. D. C. Hutchings and B. M. Holmes, “A waveguide polarization toolset design based on mode beating,” IEEE Photonics J. 3(3), 450–461 (2011). [CrossRef]
15. B. M. Holmes, M. A. Naeem, D. C. Hutchings, J. H. Marsh, and A. E. Kelly, “A semiconductor laser with monolithically integrated dynamic polarization control,” Opt. Express 20(18), 20545–20550 (2012). [CrossRef]
16. J. D. Sarmiento-Merenguel, R. Halir, X. Le Roux, C. Alonso-Ramos, L. Vivien, P. Cheben, E. Durán-Valdeiglesias, I. Molina-Fernández, D. Marris-Morini, D.-X. Xu, J. H. Schmid, S. Janz, and A. Ortega-Moñux, “Demonstration of integrated polarization control with a 40 dB range in extinction ratio,” Optica 2(12), 1019–1023 (2015). [CrossRef]
17. Y. Kawabata, M. Zaitsu, T. Tanemura, and Y. Nakano, “Proposal and experimental demonstration of monolithic InP/InGaAsP polarization modulator,” in proceeding of Eur. Conf. Opt. Comm. (ECOC2014), Mo.3.C.2 (2014).
18. T. Tanemura and Y. Nakano, “Compact InP Stokes-vector modulator and receiver circuits for short-reach direct-detection optical links,” IEICE Trans. Electron. E101.C(7), 594–601 (2018).
19. M. Kazi, S. Ghosh, H. Sodabanlu, K. Suzuki, M. Sugiyama, T. Tanemura, and Y. Nakano, “High-speed carrier-injection-based polarization controller with InGaAlAs/InAlAs multiple-quantum wells,” IEEE Photonics Technol. Lett. 29(22), 1951–1954 (2017). [CrossRef]
20. S. An and O.-K. Kwon, “Active control of polarization state of the light in InP waveguide,” Opt. Express 27(26), 37806–37815 (2019). [CrossRef]
21. M. Zaitsu, T. Tanemura, A. Higo, and Y. Nakano, “Experimental demonstration of self-aligned InP/InGaAsP polarization converter for polarization multiplexed photonic integrated circuits,” Opt. Express 21(6), 6910–6918 (2013). [CrossRef]
22. T. Suganuma, S. Ghosh, M. Kazi, R. Kobayashi, Y. Nakano, and T. Tanemura, “Monolithic InP Stokes vector receiver with multiple-quantum-well photodetectors,” J. Lightwave Technol. 36(5), 1268–1274 (2018). [CrossRef]
23. R. Ulrich, “Representation of codirectional coupled waves,” Opt. Lett. 1(3), 109–111 (1977). [CrossRef]
24. https://www.photond.com/products/harold.htm
25. L. A. Coldren and S.W. Corzine, Diode Lasers and Photonic Integrated Circuits, Wiley (1995).
26. J. Piprek, Semiconductor Optoelectronic Devices, Academic Press (2003).
27. Ioffe Institute Website, “Semiconductor on NSM” (Electronic Archive, 2001) http://www.ioffe.ru/SVA/NSM/Semicond/
28. G. P. Agrawal and N. K. Dutta, Semiconductor Lasers, Springer (1993).
29. R. A. Griffin, S. K. Jones, N. Whitbread, S. C. Heck, and L. N. Langley, “InP Mach–Zehnder modulator platform for 10/40/100/200-Gb/s operation,” IEEE J. Select. Topics Quantum Electron. 19(6), 158–166 (2013). [CrossRef]
30. P. Zhou, Y. Nakano, and T. Tanemura, “Design of InGaAsP phase modulator with asymmetrically coupled quantum wells for efficient polarization modulation,” Jpn. J. Appl. Phys. 59(8), 082004 (2020). [CrossRef]
31. H. Ohe, H. Shimizu, and Y. Nakano, “InGaAlAs multiple quantum-well optical phase modulators based on carrier depletion,” IEEE Photonics Technol. Lett. 19(22), 1816–1818 (2007). [CrossRef]
32. A. E. Elfiqi, R. Kobayashi, R. Tanomura, T. Tanemura, and Y. Nakano, “Fabrication-tolerant half-ridge InP/InGaAsP polarization rotator with etching-stop layer,” IEEE Photon. Technol. Lett. 32(11), 663–666 (2020). [CrossRef]
