A full 3D model of the modulation efficiency of a submicron complementary metal–oxide–semiconductor (CMOS)-compatible interleaved-junction optical phase shifter

• Published in: Journal of computational electronics Vol. 18; no. 4; pp. 1379 - 1387
• Main Authors: Shaikh, Abdurrahman Javid, Packeer, Fauzi, Baig, Mirza Muhammad Ali, Sidek, Othman
• Format: Journal Article
• Language: English
• Published: New York Springer US 01-12-2019; Springer Nature B.V
• Subjects: CMOS; Design optimization; Efficiency; Electrical Engineering; Engineering; Figure of merit; Finite difference time domain method; Finite element method; Geometry; Mathematical and Computational Engineering; Mathematical and Computational Physics; Mathematical models; Maxwell's equations; Mechanical Engineering; Modelling; Modulation; Modulators; Optical and Electronic Materials; P-n junctions; Phase shifters; Photonics; Prediction models; Propagation; Semiconductors; Silicon; Simulation; Theoretical; Three dimensional models; Topology; 3D optoelectronic modeling; Integrated optics; Optical modulators; Interleaved; Silicon-on-insulator (SOI); Silicon photonics; Optical phase shifters; junctions
• ISSN: 1569-8025; 1572-8137
• DOI: 10.1007/s10825-019-01366-8

The optimization of the performance of optical modulators requires reasonably accurate predictive models for key figures of merit. The interleaved
PN
junction topology offers the maximum mode/junction overlap and enables the most efficient modulators for depletion-mode operation. Due to the structure of such devices, accurate modeling must be fully three dimensional (3D), representing a nontrivial computational problem. A rigorous 3D model for the modulation efficiency of a silicon-on-insulator interleaved-junction optical phase modulator with submicron dimensions is presented herein. The drift–diffusion and Poisson’s equations are solved on a 3D finite-element mesh, while Maxwell’s equations are solved using the finite-difference time-domain method on 3D Yee cells. The entire modeling process is presented in detail, and all the coefficients required by the model are presented. The model validation suggests < 10% root-mean-square (RMS) error.