Molecular Structure Engineering of Polyelectrolyte Bilayer-Based Memristors: Implications for Linear Potentiation and Depression Characteristics

Critical to artificial intelligence's future is emulating biological synapses with memristors. Moreover, according to a wide variety, low cost, simple fabrication, and good flexibility, organic materials provide a competitive approach in memristor and synapse emulation, especially in devices where ions carry current. Polyelectrolytes with different molecular structures were used as functional layers in this study to enhance the memory and synapse performance of polyelectrolyte bilayer-based memristors, per-formed by ions, and polyelectrolyte chain migration caused a potential drop change at the interface of the similar to 15 nm thick polyelectrolyte bilayer and its electrodes. Consequently, the memory device with strong polyelectrolyte sodium poly(styrene sulfonic acid) (PSS) and poly(diallyl dimethylammonium chloride) (PDAC), which were prepared by spin-coating, shows outstanding resistive switching performance and synapse functionalities than those with weak polyelectrolytes and/or polyelectrolytes without a ring structure. Particularly, the indium tin oxide (ITO)/PSS/PDAC/ITO device shows almost linear potentiation and depression characteristics by applying continuous pulse voltage, which results in high performance on the artificial neural network simulation as 90% on the Mixed National Institute of Standards and Technology data set. The steric hindrance between the two polyelectrolytes with the five-ring structure can be attributed to the causation of linear conductance update. Furthermore, it shows short-and long-term plasticity during potentiation and depression, which is essential for the development of neuromorphic systems with complex cognitive capabilities.

Automated summary

By using organic materials as functional layers, this study enhances the memory and synapse performance of polyelectrolyte bilayer-based memristors. The device with strong polyelectrolytes showed outstanding resistive switching performance and synapse functionalities, achieving 90% performance on the artificial neural network simulation.