Using an Insulator-Metal Transition to Overcome the Fundamental Limits of Non-Volatile Memory Based on Ferroelectric Field Effect Transistors

利用绝缘体-金属转变克服基于铁电场效应晶体管的非易失性存储器的基本限制

• 批准号: 1914730
• 负责人: Nikhil Shukla
• 金额: $ 33.61万
• 依托单位: University of Virginia Main Campus
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-09-15 至 2024-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1914730&HistoricalAwards=false
• 关键词: Using; Insulator; Metal; Transition; Overcome

Nontechnical:The era of Big Data is creating enormous challenges to computing systems and networks. As the computing infrastructure evolves to meet these challenges, memory technology plays an increasingly important role. Over one hundred Zeta Bytes of data is expected to be generated annually by the year 2025. Future generations of computers will therefore need high performance memory that is fast, energy efficient, reliable, and compact. While existing technologies have some of the desired properties, they crucially lack others. For example, ferroelectric field effect transistors (FETs) based on hafnium oxide, a recently discovered material, exhibit very promising properties. They are persistent, retaining information even when turned off, have nanosecond switching speeds, and a small footprint. However, practical realization of this technology is impeded by fundamental issues which compromise reliability and constrain the operation of devices at low voltages. The proposed research aims at addressing this challenge by integrating a new functional material into ferroelectric FETs. Vanadium dioxide exhibits an insulator-to-metal phase transition, allowing it to be switched back and forth between an insulating and conductive states. This property will used to overcome the fundamental barriers associated with existing materials and enable a memory technology that significantly exceeds the performance of the current state-of-the-art. The project provides a natural platform for participating students to be involved in cross-disciplinary research at the intersection of electrical engineering and materials science. Outreach efforts will expose undergraduates and high school students to modern electronics, and broaden their experience and understanding of the opportunities in the area.Technical:This research project aims to overcome the fundamental design trade off involved in writing to, and reading from doped hafnium oxide ferroelectric field effect transistor-based non-volatile memory by replacing the conventional silicon channel of the transistor with an insulator-metal phase transition oxide, VO2. The ferroelectric state dependent abrupt resistance switching across the phase transition in VO2 that will be engineered in this device, will help de-convolute the read and write constraints, and enable the memory cell to be written at low voltages without adversely affecting the read margin, as well as also improve the reliability. A fundamental aspect of this research will focus on stabilizing the ferroelectric phase of doped hafnium oxide on the VO2 under a constrained thermal budget. Furthermore, through a materials-device co-design approach, the project will seek to experimentally demonstrate a ferroelectric transistor-based memory that can operate at a low write voltage and energy, provide large read distinguishability, and exhibit large endurance and reliability. Physics based models and simulations that capture the operation of the device as well as account for its compatibility with array level operation will be developed to support and guide the experimental effort. The results of the proposed research stand to have immense implications towards realizing a universal memory to support and accelerate the data revolution.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

非技术性：大数据的时代正在为计算系统和网络带来巨大的挑战。随着计算基础架构的发展以应对这些挑战，记忆技术起着越来越重要的作用。预计将每年在2025年每年生成一百多个Zeta字节。尽管现有技术具有一些所需的特性，但它们至关重要的是缺乏其他理想的特性。例如，基于氧化物Hafnium（一种最近发现的材料）具有非常有前途的特性的铁电场效应晶体管（FET）。它们是持久的，即使关闭，也可以保留信息，具有纳秒切换速度，并且占地面积很小。但是，这项技术的实际实现会受到损害可靠性并限制低压设备操作的基本问题。拟议的研究旨在通过将新功能材料整合到铁电FET中来应对这一挑战。二氧化钒表现出绝缘体到金属相变，从而使其可以在绝缘状态和导电状态之间来回切换。该属性将用来克服与现有材料相关的基本障碍，并实现大大超过当前最新表现的记忆技术。该项目为参与的学生提供了一个自然平台，以便在电气工程与材料科学的交集中参与跨学科研究。外展工作将使本科生和高中生接触现代电子产品，并扩大他们对该地区机会的经验和理解。技术：该研究项目旨在克服涉及写入和阅读的基本设计，并从基于晶体的非电向传统的跨跨跨型硅酸盐的传统置于跨度的sillisor sillator sillator sillator sillator sillator sillator sillator sillator sillor sillator sillator sillator sillator sillator sillator sillator，氧化物，vo2。将在该设备中设计的vo2中的铁电状态依赖性突然电阻切换，将有助于否定读写约束，并使记忆单元能够以低电压写入而不会不利地影响读取余量，并提高可靠性。这项研究的一个基本方面将着重于在受限的热预算下稳定VO2上掺杂的Hafnium氧化物的铁电相。此外，通过材料设备的共同设计方法，该项目将试图在实验上证明基于铁电晶体管的内存，该内存可以在低写入电压和能量下运行，提供较大的可读性，并具有较大的耐力和可靠性。基于物理学的模型和仿真，将开发捕获设备的操作以及对其与阵列水平操作的兼容性的说明，以支持和指导实验性工作。拟议的研究结果具有对实现普遍记忆来支持和加速数据革命的巨大影响。该奖项反映了NSF的法定使命，并被认为是值得通过基金会的智力优点和更广泛的影响评估的评估来支持的。

项目成果

Ferroelectric-based Accelerators for Computationally Hard Problems

用于解决计算难题的铁电加速器

• DOI: 10.1145/3453688.3461745
• 发表时间: 2021
• 期刊: GLSVLSI '21: Proceedings of the 2021 on Great Lakes Symposium on VLSI
• 作者: Bashar, Mohammad Khairul;Vaidya, Jaykumar;Surya Kanthi, R. S.;Lee, Chonghan;Shi, Feng;Narayanan, Vijaykrishnan;Shukla, Nikhil
• 通讯作者: Shukla, Nikhil

Ultra-Compact, Scalable, Energy-Efficient $VO_{2}$ Insulator-Metal-Transition Oxide Based Spiking Neurons for Liquid State Machines

用于液体状态机的超紧凑、可扩展、节能的 $VO_{2}$ 基于绝缘体金属过渡氧化物的尖峰神经元

• DOI: 10.1109/vlsi-soc46417.2020.9344078
• 发表时间: 2020
• 期刊: 2020 IFIP/IEEE 28th International Conference on Very Large Scale Integration (VLSI-SOC
• 作者: Ganguly, Samiran;Shukla, Nikhil;Ghosh, Avik W.
• 通讯作者: Ghosh, Avik W.