Interface engineering in organic field-effect transistors for neuromorphic and memory applications

Abstract

The rapid growth of artificial intelligence, edge computing, and distributed sensing has created a strong need for electronic hardware that can process and store information with lower energy consumption and reduced data movement. Conventional von Neumann computing systems physically separate memory and processing units, which introduces latency, bandwidth limitations, and high energy costs during data-intensive tasks. This dissertation investigates organic field-effect transistors (FETs) as tunable platforms for memory and neuromorphic applications, with emphasis on how contact interfaces, ferroelectric dielectrics, semiconductor–dielectric interfaces, and nanoscale charge-storage elements can be engineered to produce stable, low-voltage, and programmable device behavior. The first part of this work addresses charge injection at the metal–semiconductor contact. Ultrathin aluminum oxide (Al2O3) deposited by atomic layer deposition (ALD), selfassembled monolayers, and ultraviolet–ozone treatment are examined as contact-engineering strategies for reducing Schottky barrier height, promoting partial Fermi-level de-pinning, and lowering contact resistance. These studies establish a reliable transistor platform in which the effects of dielectric and interface engineering can be evaluated more clearly. The dissertation then examines polymer ferroelectric and polar dielectrics as active gate-stack components in organic FETs. Devices based on PVDF-TrFE and PVDF-HFP are studied through controlled poling, dielectric-thickness scaling, and semiconductor–dielectric interface modification. Vertical and lateral poling are used to investigate how dipole alignment influences charge transport, mobility, hysteresis, and device stability. Thin dielectric layers are explored to reduce operating voltage, while interface trap density and the density of states are evaluated to clarify the role of interfacial disorder. Ultrathin Al2O3 interlayers are introduced at the semiconductor–dielectric interface to passivate traps, improve charge transport, and enhance operational stability. Electrical measurements are supported by structural and morphological characterization, including atomic force microscopy, scanning electron microscopy, transmission electron microscopy, and differential phase-contrast scanning transmission electron microscopy, linking device performance to changes in the gate stack and interface. Building on these optimized transistor structures, this work investigates organic FETs as synaptic devices for neuromorphic hardware. Ferroelectric polymer gate stacks and donor–acceptor organic semiconductors are used to realize pulse-dependent conductance modulation, memory windows, and synaptic plasticity. Device-level metrics are connected to neural image-recognition simulations, demonstrating how conductance tuning, dynamic range, and update nonlinearity influence learning performance. Structure–function relationships in pyridyl-triazole copolymers are also examined to show how molecular design and microstructure affect synaptic behavior. The final part of the dissertation explores sub-nm platinum nanoparticles embedded at the semiconductor–dielectric interface as localized charge-storage elements. These nanoparticle-enabled organic transistors exhibit enlarged memory windows, multilevel state tuning, and both electrical and optically assisted programming. Current signatures consistent with room-temperature Coulomb-blockade-like behavior suggest that nanoscale charge confinement can provide an additional mechanism for programmable memory and neuromorphic functionality. Overall, this dissertation establishes an interface-centered design framework for organic transistor platforms in which charge injection, ferroelectric polarization, interfacial trapping, and nanoscale charge storage are systematically controlled. The results show that reliable neuromorphic and memory behavior in organic FETs requires integrated optimization of contacts, dielectrics, semiconductors, and interfaces. This framework provides a foundation for scalable, flexible, low-power organic electronics for future in-memory, adaptive, and neuromorphic computing systems.