Woo Kyung Seok, Williams R Stanley, Kumar Suhas
Sandia National Laboratories, Livermore, California 94550, United States.
Department of Electrical and Computer Engineering, Texas A&M University, College Station, Texas 77843, United States.
Chem Rev. 2025 Jan 8;125(1):294-325. doi: 10.1021/acs.chemrev.4c00454. Epub 2024 Dec 19.
Since the early 2000s, the impending end of Moore's scaling, as the physical limits to shrinking transistors have been approached, has fueled interest in improving the functionality and efficiency of integrated circuits by employing memristors or two-terminal resistive switches. Formation (or avoidance) of localized conducting channels in many memristors, often called "filaments", has been established as the basis for their operation. While we understand some qualitative aspects of the physical and thermodynamic origins of conduction localization, there are not yet quantitative models that allow us to predict when they will form or how large they will be. Here we compile observations and explanations of channel formation that have appeared in the literature since the 1930s, show how many of these seemingly unrelated pieces fit together, and outline what is needed to complete the puzzle. This understanding will be a necessary predictive component for the design and fabrication of post-Moore's-era electronics.
自21世纪初以来,随着晶体管尺寸缩小的物理极限逐渐逼近,摩尔定律即将终结,这激发了人们通过采用忆阻器或两端电阻开关来提高集成电路功能和效率的兴趣。许多忆阻器中局部导电通道(通常称为“细丝”)的形成(或避免)已被确立为其工作基础。虽然我们了解传导局域化的物理和热力学起源的一些定性方面,但尚未有定量模型能让我们预测它们何时会形成或会有多大。在这里,我们汇编了自20世纪30年代以来文献中出现的关于通道形成的观察结果和解释,展示了这些看似不相关的部分是如何组合在一起的,并概述了完成这一难题所需的条件。这种理解将是后摩尔时代电子产品设计和制造中必要的预测要素。