From DC to Daylight: Harnessing Electromagnetic Fields for Bioelectronics, Wireless Communications, and Silicon Photonics

April 18, 2018

Dr. Constantine Sideris

Caltech

Maxwell’s equations are responsible for explaining the fundamental operating principles behind most of today’s technology. In this talk, we will explore how understanding and controlling electromagnetic fields can lead to significant impact across a multitude of applications over a wide frequency range on the electromagnetic spectrum. Starting from the low-frequency end of the spectrum, I will present the design and implementation of a new integrated magnetic biosensor. The magnetic biosensor is fabricated in a standard CMOS foundry process without any post-fabrication processing and can perform in-vitro detection of DNA, proteins, and cells by utilizing magnetic nanoparticles as labels. We will discuss three different, improved sensor designs, which address sensor gain uniformity, enable multiplex target detection, and compensate sensor electrical and thermal drift based on spatial and temporal manipulations of the magnetic fields. I will present initial in-vitro biodetection experiments, and discuss future research directions moving towards in-vivo sensing with wearable and implantable devices, as well as actuation via targeted therapeutics. Next, we will look into the RF domain and develop maximal performance bounds for antennas. I will present a rapid simulation technique which, when coupled with heuristic optimization algorithms, can quickly and effectively produce new antenna structures de-novo with little or no manual intervention. The efficacy of these techniques will be shown in the context of a 3D printed coupling antenna for a dielectric waveguide communication link. Moving higher in frequency, we will explore the near-infrared (NIR) part of the spectrum in the context of silicon photonic device optimization. I will present on-going work in designing grating coupler and power splitting devices with arbitrary splitting ratios by using adjoint optimization and highly efficient integral equation techniques. We will also explore exciting future directions in these research areas, leveraging modern computation and efficient numerical algorithms as well as holistic co-design of circuits and electromagnetics.

Published on April 18th, 2018Last updated on April 26th, 2018