MRS Meetings and Events

 

EL20.01.02 2023 MRS Fall Meeting

Thermodynamic Origin of Nonvolatility in Resistive Memory

When and Where

Nov 28, 2023
11:00am - 11:15am

Hynes, Level 3, Room 301

Presenter

Co-Author(s)

Jingxian Li1,Yiyang Li1

University of Michigan, Ann Arbor1

Abstract

Jingxian Li1,Yiyang Li1

University of Michigan, Ann Arbor1
Resistive memory or memristor is a highly potential memory and computing unit for next-generation information storage, in-memory computing, and neuromorphic computing. Valence change memory (VCM) is a promising memristor that stores information through the distribution of oxygen vacancy point defects in transition metal oxides.<br/><br/>The ability to store and retain information is a critical function of nonvolatile memory. It is widely believed non-volatility in resistive memory is a result of the slow diffusion kinetics of oxygen vacancies. In this work, we combine materials characterization, device measurements, ab initio, and continuum modeling and show instead that information retention results from phase separation that results from materials thermodynamics. We definitively show for the first time that the amorphous refractory metal oxides undergo spinodal decomposition into two phases with different metal to oxygen ratios. Moreover, the stability of this spinodal decomposition controls whether a resistive memory cell will be able to retain information over time.<br/><br/>By applying phase separation, a foundational principle of materials science, to the field of resistive memory, we explain for the first time why these materials are able to function as nonvolatile memory. We further highlight the critical importance of considering nonideal thermodynamic interactions such as phase separation for future electronic materials dominated by point defects.

Keywords

Auger electron spectroscopy (AES)

Symposium Organizers

Gina Adam, George Washington University
Sayani Majumdar, Tampere University
Radu Sporea, University of Surrey
Yiyang Li, University of Michigan

Symposium Support

Bronze
APL Machine Learning | AIP Publishing

Publishing Alliance

MRS publishes with Springer Nature