Cation Coordination as a Molecular Handle to Program Synaptic behavior in OrganicMemristors
Year: 2026
Type: preprint
Abstract: Organic mixed ionic-electronic conductors offer a route to soft, low-voltage neuromorphic electronics because ionic motion can directly modulate electronic conductance, providing a materials-level analogue of synaptic plasticity.However, the molecular origin of this ion-electron coupling remains poorly understood, limiting the rational design of artificial synapses for time-dependent tasks such as visual recognition.Here, we show that synaptic dynamics in Organic mixed ionic-electronic memristors can be programmed by chemically controlling salt dissociation within the active matrix.The devices combine Super Yellow, alkali-metal triflate salts (XOTf; X = Li, Na or K), and trimethylolpropane ethoxylate, an oligoether that coordinates the cations with different strengths.This coordination regulates the voltage-dependent population of mobile triflate anions available to compensate holes injected into Sy, thereby controlling electrochemical p-doping and the memristive response.As a result, cation identity systematically tunes the activation threshold, ON/OFF ratio, switching kinetics, relaxation dynamics and volatility.Under pulsed stimulation, these programmed ion dynamics enable short-term and long-term plasticity, as well as spike-timing-dependent plasticity, with conductanceupdate amplitudes following the trend SyLi > SyNa > SyK.Learning rules extracted from the devices were implemented in spiking neural-network simulations for digit recognition, yielding accuracies of 87%, 86% and 76% for SyLi, SyNa and SyK, respectively, after one training epoch using 1000 MNIST images.This corresponds to a 60-fold reduction in training data relative to the full MNIST benchmark commonly used in neuromorphic simulations.These results link molecular control of salt dissociation to programmable ion-electron coupling, enabling efficient organic synapses for time-resolved neuromorphic vision.
Source: ChemRxiv
Institutions: Universitat de València, Trinity College Dublin
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7212154013