Interface Barrier-Modulated Neuromorphic Behavior in Mo/CdIn2S4/ZnO–Al Structures Based on Metastable Defects

Abstract

Neuromorphic computing is an increasingly accelerating research field that still faces several hardware challenges, thus driving materials science research. This contribution presents prototype neuromorphic devices based on cadmium–indium sulfide (CdIn₂S₄). We have found that the compound, previously studied in photocatalysis and photovoltaics, has yet-to-be-explored memristive properties. We manufactured a series of devices with varying stoichiometry and unveiled their switching and neuromorphic behavior. Our measurements indicate that barriers on the CdIn₂S₄ interfaces and native CdIn₂S₄ metastable defects control the behavior of the devices. Due to its compelling optoelectronic properties, CdIn₂S₄ creates vast opportunities to use it as an active layer for optically controlled neuromorphic devices. We propose a disparate switching mechanism based on metastable defects that could be utilized in neuromorphic device technology.