UC San Diego

ABSTRACT OF THE DISSERTATION

Approaching the Quantum Limit of Photodetection in Solid-State Detectors with Internal Signal Amplification

by

David Hall

Doctor of Philosophy in Electrical Engineering (Nanoscale Devices and Systems)

University of California, San Diego, 2017

Professor Yu-Hwa Lo, Chair

This thesis offers three physical device concepts to overcome limitations in conventional solid-state photodetectors. Tradeoff between sensitivity, bandwidth and dynamic range is a pressing issue for detectors that must meet the demands of imaging and communications applications. Single photon sensitivity has long been possible in a solid-state detector, but with no dynamic range at the device level, and bandwidth on the order of 10 MHz, far below the requirements for modern communications. Detectors reaching 40 GHz bandwidth and exceeding 20 dB dynamic range are now in wide use, but the fundamental physics of these devices limits the sensitivity to only 1000 photons in practical operating conditions. This thesis presents theoretical and experimental progress in three main thrusts that are promising in developing a detector that can maintain large dynamic range and high bandwidth while reaching the ultimate sensitivity limit of a single photon.

The first device covered is a device which incorporates multiple gain mechanisms in a single device. Rather than relying solely on avalanche gain, this device incorporates bipolar gain which relaxes the excess noise burden experienced by conventional avalanche detectors. Negative feedback is also shown to play a role in regulating the gain and reducing noise.

The second and third device concepts utilize localized states in disordered material to exploit a host of physical phenomena. Localized states are known to increase carrier excitation probability by relaxing the k-selection rule of momentum conservation. Auger excitation and electron phonon coupling play a role in this new internal signal amplification mechanism, called Cycling Excitation Processes (CEP).

CEP is first presented in crystalline silicon, where localization is achieved through heavy doping and partial compensation. Impressive performance in terms of gain, responsivity and noise are promising for realizing the goal of a next generation solid-state detector approaching the quantum limit.

Despite the merits, CEP in crystalline silicon suffers from high dark current due to heavy doping. Amorphous silicon is presented as an alternative means of achieving localization to realize CEP. High speed and high sensitivity capability is demonstrated. While the work is ongoing as of the writing of this thesis, a clear research plan is in place for realizing an amorphous silicon CEP detector capable of high bandwidth and single photon detection.