Abstract
This study explores the area-dependent resistive switching (RS) characteristics of Gd(0.2)Ca(0.8)MnO(3) (GCMO)-based memristors with aluminum (Al) and gold (Au) electrodes, emphasizing their potential for neuromorphic computing applications. Using a combination of electrical measurements and X-ray photoelectron spectroscopy (XPS), we demonstrate that the high-resistance (HRS) and low-resistance (LRS) states exhibit predictable scaling with device area, with HRS resistances ranging from 10(7) to 10(8) Ω and LRS from 10(5) to 10(7) Ω, supporting the hypothesis of interface-type RS. XPS depth profiling revealed notable differences in AlO (x) interfacial layer composition between HRS and LRS, with a higher oxide content and a widened interfacial region in HRS. Additionally, the multistate RS capability of up to ten distinct levels was achieved by modulating applied voltages, highlighting GCMO's suitability as a material for synaptic weight storage in artificial neural networks. Our findings underscore GCMO's promise for energy-efficient, scalable memristor-based systems.