Synaptic Plasticity in Electro-Polymerized Pedot Based Memristors for Neuromorphic Application

Uniform thin film of PEDOT is prepared using electropolymerization over a large area, which reduces cell-to-cell switching variation in memristive devices.Synaptic properties like short term plasticity (STP), long term plasticity (LTP), and spike-rate-dependent plasticity (SRDP) is demonstrated.Hebbian learning rule Spike time dependent plasticity (STDP) showed up to 75% change in synaptic weight under wide range of input signals, which makes it a suitable material for neuromorphic computation applications.Abstract:A polymer based memristor as an artificial synapse has been one proficient approach to mimic biological synapse. It remembers the last state and modulate the output accordingly with subsequent voltage signals at low voltages and at high processing speeds. A functional layer of poly(3,4-ethylenedioxythiophene) (PEDOT) polymer is deposited by electro-polymerization using cyclic voltammetry. Here, we studied Pt/PEDOT/Al based memristor, which exhibited excellent artificial synapse like plasticity. We discuss the basic function of remembering and forgetting of the devices through synapticity and plasticity. STDP, which is the most important Hebbian learning rule for learning and memory showed up to 75% change in synaptic weight under wide range of input signals. The devices show excellent plasticity behavior mimicking biosynaptic behavior, which makes them a suitable system for neuromorphic applications