High-Performance Organic Memristors With Molecule-Mediated Ion Movements

The human brain is a highly integrated,parallel,and low-consumption event-driven information storage and analysis system with an efficiency that cannot be achieved by the current separated storage and processor systems based on the von Neumann architecture.It is evident that memristors have gradually been widely used in storage and neuromorphic computing.In recent years,organic memristors have become one of the most promising components for next-generation information processing and computing,and organic materials are widely used in the research of organic memristors by virtue of their convenient preparation process,low cost,flexibility,biocompatibility,and diversity of molecular structure design,the variety of novel organic materials helps to develop and prepare organic memristors with diverse structures and adjustable performance.However,the random and uncontrollable ion movements inside the active layer of the device leads to the low productivity,high operation voltages,high power consumption,and unstable high/low resistance state.Therefore,it is necessary to rationally design novel molecule-mediated ion transport strategy to enable the devices to achieve high productivity,low operation voltages,stable high/low resistance states,and high robustness.Until to now,there are few effective strategies,and specific platforms with a view to molecule with specific coordination groups-regulating ions have been rarely reported,let alone systematic characterizations.This work is based on the mechanism of ion transport assisted by deliacate redox coordination reaction,a series of organic memristors based on poly（9-vinylcarbazole）（PVK）and 7,7,8,8-tetracyanoquinodimethane（TCNQ）materials are designed and prepared.The performance,mechanism,advanced applications in neuromorphic computing and neuronal simulation of memristors are investigated in depth to provide a technical reserve for the theoretical development and practical multifunctional applications of organic memristors.The specific research contents are as follows:（1）A novel coordination-assisted electronic/ionic memristor is investigated to delicately regulate the ionic migration.As a general coordination strategy,this work introduced the well-known redox coordination reaction between silver and TCNQ into PVK matrix,and prepared high-performance ITO/PVK:TCNQ/Ag devices with crossbar"sandwich"structure.Raman spectra and Raman mapping have demonstrated that the introduced Ag can react with the cyanide of the TCNQ molecule in the active layer,and the resulting charge transfer complex Ag-TCNQ is uniformly distributed in the active layer and generating a stable Ag conductive filaments’rupture/formation process.By this novel mechanism,the device has a low operation voltage（～0.55 V/-0.21 V）and low power(2.07×10^-5 W),stable switching cycles（10³ cycles）and retention stability of high and low resistance states（10⁴s）.The charge transfer complex under the electric field can precisely regulate the migration of metallic silver ions through the coordination bonds between metal ions and TCNQ cyano,which solves the uncontrollable and random transmission drawback of ion migration in conventional ECM memristors.（2）The distribution of TCNQ molecules and molecular domain separation within the polymeric matrix are regulated by solvents with different polarities,thus regulating the formation shape and distribution of Ag conducting filaments.With the help of high polarity solvent acetonitrile,this work can expand the embedded TCNQ molecular domains,increasing the phase separation between TCNQ and PVK,and obtaining stable multi-level storage and continuous renewal of the conductive states.Further with the help of the medium-polarity solvent tetrahydrofuran,this work can modulate the discrete Ag conductive filaments mediated by small TCNQ molecular domains to achieve the threshold switching behavior.Raman spectroscopy and mapping,XRD,depth profiling XPS and in-situ nanoscale C-AFM were used to characterize the distribution and mechanism of Ag-TCNQ.（3）The memristor with programmable multi-level conductance states can be applied in artificial neural network（ANN）based neuromorphic computing.The devices obtained continuous varied conductance states under continuous weak pulses,revealing their potential function as artificial synapses,and obtained high accuracy in image recognition in SLP networks based on a supervised learning framework.Furthermore,the threshold-switching memristor can simulate the dynamic potentials（APs）function of neurons.Depolarization（potential increase）occurs gradually under successive voltage pulses until a peak threshold is reached,after which the ion channels spontaneously close and the potentials repolarize to the initial resting state.This work have also achieved integrating-and-firing action potential,which can be achieved under both positive and negative voltage pulses.This part of the work expands the application capabilities of the device and provides a solid foundation for the preparation of multifunctional neuromorphic devices.