This work presents a defect charging mechanism in 5‐nm‐thick amorphous Al2O3 thin‐films fabricated on plastic, which leads to multistate memory effects, and thus the realization of synaptic thin‐film transistors (TFTs) for neuromorphic applications. First, the Al2O3 thin‐films are characterized in metal–insulator–metal stacks. These devices exhibit ferroelectric‐like behavior, which is visible in the small‐signal capacitance and the surface charge density. Furthermore, the quantum‐mechanical simulation of the current–voltage characteristic leads to a physical model with trap charges close to the anode interface where deep‐level traps are identified by fitting the experimentally obtained resonant tunneling peaks. The trap charge lifetime and frequency behavior is evaluated in InGaZnO4 TFTs, where the 5‐nm‐thick Al2O3 layer is employed as gate dielectric. At an operating voltage as low as ±2 V, a charge trapping retention up to ≈3 h and a discernable ON/OFF read‐out with a factor >3 at 2 kHz are achieved. When subjected to a train of gate–source voltage pulses, the TFTs show charge integration properties which emulate facilitating and depressing behaviors of biological synapses. These results indicate that thin low‐temperature defect‐rich metal‐oxide dielectrics may be candidates for low‐voltage memory applications and neuromorphic circuits on unconventional substrates.

No abstract

Abstract:We present a survey of the effect of vacancies on quantum transport in graphene, exploring conduction regimes ranging from tunnelling to intrinsic transport phenomena. Vacancies, with density up to 2%, are distributed at random either in a balanced manner between the two sublattices or in a totally unbalanced configuration where only atoms sitting on a given sublattice are randomly removed. Quantum transmission shows a variety of different behaviours, which depend on the specific system geometry and disorder distribution. The investigation of the scaling laws of the most significant quantities allows a deep physical insight and the accurate prediction of their trend over a large energy region around the Dirac point.