Optimizing graded-index few-mode fiber for space division multiplexing.

We investigate the graded-index few-mode fiber (GI-FMF) to realize a 4-LP-mode (i.e. LP01, LP11, LP21, and LP02) fiber for mode-division-multiplexed transmission. This study optimizes the GI-FMF for both, first, for large effective indices differences (Δneff), and second, for low differential mode delay (DMD) between any two LP modes, for different optimized parameters. Thus, it shows that GI-FMF is suitable for both weakly-coupled few-mode fiber (WC-FMF) as well as strongly-coupled few-mode fiber (SC-FMF) via adjusting the profile parameter (α), refractive index difference between core and cladding (nco - nclad), and core radius (a). We report the optimized parameters for WC-GI-FMF with large effective indices difference (Δneff) of 0.6 × 10-3 and low |DMD| of 5.4 ns/km while the minimum effective mode area (Min.|Aeff|) is 80 µm2 and bending loss (BL) of the highest order mode is 0.005 dB/turn (much lower than 10 dB/turn) at a 10 mm bend radius. Here, we could break down the degeneracy between LP21 and LP02 mode, which remains a challenging task in GI-FMF. To the best of our knowledge, this is the lowest DMD (5.4 ns/km) ever reported for such a weakly-coupled (Δneff = 0.6 × 10-3) 4-LP-mode FMF. Similarly, we optimized the parameters for SC-GI-FMF with Δneff of 0.1 × 10-3 and the lowest DMD of 0.9 ns/km while Min.|Aeff| is ≫ 100 µm2 and BL of higher order mode is 6 dB/turn (< 10 dB/turn) at 10 mm bend radius. Further, we investigate narrow air trench-assisted SC-GI-FMF to reduce the DMD and achieve the lowest DMD of 16 ps/km for a 4-LP-mode GI-FMF with a minimum Δneff of 0.7 × 10-5.

Analysis of a SiGe Mach-Zehnder modulator for 400G data communication.

In this paper, the performance of a silicon (Si) Mach-Zehnder modulator (MZM) is enhanced by implanting germanium (Ge) on Si, forming a graded Si-Ge (SiGe) core. A process simulation study is done, and the effect of substrate temperature, implantation energy, and pre-amorphization on the Ge composition and developed in-plane stress is observed. The dependence of active dopant concentration and defect-cluster formation on the annealing conditions is discussed. A comparison of the process simulated SiGe phase shifter with a Si phase shifter shows 2.27× higher phase shift at -5V for 1550 nm wavelength of operation. A dual-arm drive with quadrature operation is investigated for both SiGe and Si MZMs. A traveling-wave electrode is used to enhance the modulation bandwidth. The SiGe MZM achieves better performance in terms of modulation bandwidth, modulation speed, fiber transmission length, energy per bit, and dispersion tolerance compared to the Si MZM. The dual-arm driven SiGe MZM can achieve 62 Gbps error-free 2 km fiber transmission with energy per bit of 1.92 pJ/bit and 3 dB bandwidth of 62.8 GHz at -2.5V bias using 3Vpp on-off keying modulation.

2D modeling of silicon optical PN phase shifter.

In this paper, a 2D model of a silicon lateral PN optical phase shifter is presented, which can be used for multiple-parameter study and optimization of device performance without the need for any commercial numerical tools. The model shows good agreement with technology computer-aided design (TCAD) simulation and can be used to calculate the phase shift, absorption loss, modulation efficiency, and insertion loss of the phase shifter. Multiple-parameter study includes the waveguide dimensions, operating wavelength, cladding material, doping concentrations, junction offset, and applied voltage. The model employs the effective index method to determine the mode properties and construct the 2D mode field. The PN diode is modeled by taking into account the fringing electric field at the core-cladding interface, which results in a wider depletion region near the interface. Multiple-loss components are discussed, and the scattering loss and free-carrier absorption are modeled using Payne-Lacey and Soref models, respectively. The model uses 2D modal overlap with 2D carrier distribution across the waveguide to calculate the phase shifter performance metrics. The algorithm used to model the 2D nature of the PN diode depletion region is presented in detail and uses mathematical and analytical formulas instead of numerical methods, making the model faster and easy to implement, with accuracy on par with commercial tools.