<p>Resistive switching (RS) in oxide-based memristive devices is strongly affected by the stochastic formation and rupture of oxygen vacancies (<i>V</i><sub>o</sub>) conductive filaments, which limits device-to-device uniformity, cycle-to-cycle stability, and overall reliability. To mitigate such randomness, HfO<sub>2</sub> thin films were prepared via a sol–gel route at different post-annealing temperatures, and Pt point defects or interlayer defects were further introduced. The synergistic effects of annealing temperature and Pt-related defects on RS characteristics were systematically evaluated. Electrical measurements reveal that HfO<sub>2</sub> films annealed at 500&#xa0;°C exhibit the most desirable RS behavior and overall performance. At lower annealing temperatures, excessive organic residues and insufficient film densification suppress the formation of stable conductive filaments, resulting in weak switching behavior. In contrast, excessively high temperatures induce grain coarsening and grain boundary reduction, and may simultaneously generate excess <i>V</i><sub>o</sub>, leading to degraded insulation (i.e., reduced high-resistance state (HRS)) and thus weakened RS performance. Furthermore, a moderate density of Pt point defects induces local electric-field distortion, which is believed to promote directional <i>V</i><sub>o</sub> migration, and enhance filament formation reproducibility, thereby contributing to improved switching stability. However, when the defect density becomes too high (strong Pt aggregation or layer-like Pt intercalation), the film insulation is compromised, HRS resistance decreases, and RS performance deteriorates. In this study, the sample with 20&#xa0;s Pt sputtering exhibited the optimal RS performance. Overall, rational design of artificial defect structures combined with appropriate thermal treatment can help suppress the stochastic nature of filament evolution and consequently improve switching stability and device reliability. These findings provide experimental guidance and mechanistic insight into defect-engineering strategies for highly stable RS memory devices.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Performance optimization of HfO2-based resistive switching devices via Pt interlayer defect engineering

  • Zhiquan Wang,
  • Chengang Dong,
  • Xin Wang

摘要

Resistive switching (RS) in oxide-based memristive devices is strongly affected by the stochastic formation and rupture of oxygen vacancies (Vo) conductive filaments, which limits device-to-device uniformity, cycle-to-cycle stability, and overall reliability. To mitigate such randomness, HfO2 thin films were prepared via a sol–gel route at different post-annealing temperatures, and Pt point defects or interlayer defects were further introduced. The synergistic effects of annealing temperature and Pt-related defects on RS characteristics were systematically evaluated. Electrical measurements reveal that HfO2 films annealed at 500 °C exhibit the most desirable RS behavior and overall performance. At lower annealing temperatures, excessive organic residues and insufficient film densification suppress the formation of stable conductive filaments, resulting in weak switching behavior. In contrast, excessively high temperatures induce grain coarsening and grain boundary reduction, and may simultaneously generate excess Vo, leading to degraded insulation (i.e., reduced high-resistance state (HRS)) and thus weakened RS performance. Furthermore, a moderate density of Pt point defects induces local electric-field distortion, which is believed to promote directional Vo migration, and enhance filament formation reproducibility, thereby contributing to improved switching stability. However, when the defect density becomes too high (strong Pt aggregation or layer-like Pt intercalation), the film insulation is compromised, HRS resistance decreases, and RS performance deteriorates. In this study, the sample with 20 s Pt sputtering exhibited the optimal RS performance. Overall, rational design of artificial defect structures combined with appropriate thermal treatment can help suppress the stochastic nature of filament evolution and consequently improve switching stability and device reliability. These findings provide experimental guidance and mechanistic insight into defect-engineering strategies for highly stable RS memory devices.