Improved switching performance and reliability of sol–gel-derived Y2O3-based RRAM devices with HNO3 incorporation
摘要
The rapid advancement of artificial intelligence has increased the demand for high-performance memory, thereby emphasizing the need for next-generation non-volatile memory solutions such as resistive random-access memory (RRAM). In this study, Y2O3-based RRAM devices were fabricated on ITO-coated glass substrates using the sol–gel method. The concentration of oxygen vacancies in the Y2O3 films was controlled by incorporating HNO3, a well-known strong oxidizing agent. The structural, chemical, and electrical properties of the films were analyzed as a function of HNO3 concentration. As the HNO3 amount increased, the resulting films became thinner and more uniform, exhibiting minimal cracks or pinholes, thereby achieving high-quality Y2O3 layers. Additionally, the oxidizing nature of HNO3 reduced the concentration of oxygen vacancies in the Y2O3 films. RRAM devices based on HNO3-treated Y2O3 demonstrated lower SET voltages, attributed to the reduced film thickness and enhanced high-resistance state (HRS) resulting from fewer oxygen vacancies, which otherwise serve as leakage current pathways. The lower vacancy concentration also suppressed random and excessive conductive filament growth between the electrodes. This led to increased HRS values and improved endurance and retention characteristics of the non-volatile RRAM devices. The Y2O3 RRAM devices fabricated with HNO3-modified precursors exhibited a high HRS/LRS ratio, exceeding ~ 106, endurance up to 103 cycles, and data retention beyond 104 s without notable degradation. This study provides valuable insights into optimizing the sol–gel deposition of Y2O3 via HNO3 modulation for RRAM architectures, representing a significant step toward developing reliable and stable non-volatile memory technologies.