Ph.D. Final Defense - Dnyanesh Deepak Sarawate
Title:
Electric-Field-Driven Ion Trapping and Tunable Relaxation Kinetics in Two-dimensional Transistors for Edge Computing
Abstract:
Two challenges that modern electronics face include processing sensor data directly on local devices (i.e., edge computing) and protecting hardware from reverse engineering. These two applications demand different memory behaviors: processing time-dependent signals requires a short-term, fading memory, whereas protecting hardware requires programmable doping that persists without constant power. One way to achieve time-dependent memory is through the use of ions in an electric-double-layer (EDL). EDL gating provides ultra-high interfacial capacitance density (µF/cm²) by drifting mobile ions in an electrolyte to a semiconductor surface under an applied voltage. This dissertation first characterizes the time-dependent pulse dynamics of a well-studied PEO:LiClO₄ electrolyte on graphene and WSe₂ transistors. In a dual-gate configuration, pulsed inputs generate a fading memory for processing ECG signals, where 30/30 ms pulse timing at 305 K maintains baseline stability (+1.57% net retention) and current modulation up to 21%. Next, this work demonstrates the use of electric fields generated by the EDL (~V/nm) to trigger localized chemistry in a custom solid polymer electrolyte, enabling both programmable relaxation kinetics and non-volatile doping in 2D transistors. This custom copolymer, poly(DEAEMA-co-PEGMA), is synthesized to undergo an electric-field-triggered Menshutkin crosslinking reaction. In graphene FETs, +5 V programming induces persistent n-type shifts. Low-temperature readouts (T = 200 K) freeze mobile bulk ions to isolate trapped interfacial charge, confirming a sheet carrier density of ~10¹² cm⁻² with ≈80% retention over seven weeks; control devices with no crosslinker show no such retention. Finally, applying this electrolyte to WSe₂ transistors yields a programmable temporal element. Increasing the programming bias from +1 V to +4 V systematically tunes the mean relaxation time (〈τ〉) from 2.4 s to 85 s (>35x). These results establish field-sensitive electrolytes as a new class of electrolyte and a candidate material platform for adaptive edge computing and reconfigurable electronics.
Chair:
Dr. Susan Fullerton
Department of Chemical and Petroleum Engineering, University of Pittsburgh
Department of Electrical and Computer Engineering, University of Pittsburgh
Committee Members:
Dr. Brian Tackett
Department of Chemical and Petroleum Engineering, University of Pittsburgh
Dr. Lei Li
Department of Chemical and Petroleum Engineering, University of Pittsburgh
Dr. Nathan Youngblood
Department of Electrical and Computer Engineering, University of Pittsburgh