ECE PhD Prospectus Defense: Ayman Abdelhakeem

ECE PhD Prospectus Defense: Ayman Abdelhakeem

Title: Advanced Diffraction Grating Design in Silicon Photonics for Optical I/O and Atmospheric Remote Sensing

Presenter: Ayman Abdelhakeem

Advisor: Professor Miloš Popović

Chair: Professor Roberto Paiella

Committee: Professor Miloš Popović, Professor Tianyu Wang, Professor Roberto Paiella, Dr. Nathan Dostart

Google Scholar: https://scholar.google.com/citations?user=1mqGF8wAAAAJ&hl=en&authuser=2

Abstract: Silicon photonics has the ability to realize systems on a chip containing hundreds and thousands of components, bringing the concept of the integrated circuit revolution of the 1960s and 70s to light based circuits — photonic integrated circuits. This technology is currently being fast-tracked into high-volume CMOS manufacturing by the pressures of needing to interconnect AI compute hardware optically to scale AI supercomputer performance. At the same time, light is an ideal medium for sensing in many domains. The merger of integration and sensing into single-chip sensing instruments may revolutionize many sensing technologies. This thesis research has focused on photonic integrated circuits based on waveguide/grating structures to bring about such advances.

One thrust of this thesis harnesses silicon photonics for chip-scale spectropolarimetry for satellite-based atmospheric remote sensing — a collaboration supported by NASA through the AITHENA program, under a research initiative titled HPC-Enabled Photonic Integrated Circuit Modeling for a Near-Infrared On-Chip Spectrometer. Modern remote sensing instruments traditionally treat optical design and signal retrieval as separate stages, where the sensing front-end captures measurements that are subsequently processed by independent reconstruction algorithms. This separation limits overall performance, since the optical system is not optimised jointly with downstream retrieval tasks. This research proposes an integrated approach where optical encoding and signal reconstruction are co-designed for task-driven performance, realized through a silicon nitride photonic integrated circuit that simultaneously encodes polarisation state and high-resolution spectral information on a single chip, targeting compact spaceborne platforms. As part of this collaboration, simulation workflows for photonic integrated circuits were developed and validated, with integration into NASA Langley's high-performance computing infrastructure, including the K cluster. By embedding the full optical measurement chain into a unified design framework, this work addresses key challenges in miniaturised remote sensing, including fabrication tolerance, coupling efficiency, and resolution performance.

A significant challenge in silicon photonics is how to couple light on and off chip with minimum insertion loss — a problem that becomes especially acute as AI hardware infrastructure fast-tracks silicon photonics into high-volume CMOS manufacturing. The same grating diffraction physics that underlies the on-chip spectropolarimeter is brought to bear on this problem through a novel grating coupler-based fiber-to-chip coupling approach, developed through my internship at Ayar Labs. This unified grating framework connects both thrusts of the thesis — from satellite remote sensing to AI computing infrastructure — through the same fundamental principle of engineering waveguide-grating structures to control and manipulate light. This work has the potential for adoption in foundry-scale manufacturing at TSMC and potentially other foundries, broadening the impact of this research toward high-volume photonic integration.

When 2:00 pm - 4:00 pm on 22 June 2026
Building PHO 339