Dec 4, 2024
11:45am - 12:00pm
Sheraton, Third Floor, Fairfax B
Sanghyo Lee1,Jinseok Ryu2,Hein Philipp3,Manfred Martin3,Miyoung Kim1
Seoul National University1,Diamond Light Source2,RWTH Aachen University3
Sanghyo Lee1,Jinseok Ryu2,Hein Philipp3,Manfred Martin3,Miyoung Kim1
Seoul National University1,Diamond Light Source2,RWTH Aachen University3
There has been growing interest in next-generation memory semiconductors, particularly those exhibiting resistive switching phenomena. Recent advancements in AI have accelerated this trend, drawing more attention to the potential of these materials. Among the various materials being researched, recent studies have shown that amorphous GaOx exhibits non-filamentary memristive switching behavior when sandwiched between two ion-blocking electrodes. This behavior is believed to be related to the movement of oxygen ions within the material when an electric field is applied. Previous studies have included electrical property measurements, structural analysis, and numerical simulations, but TEM studies on the material have not been actively conducted beyond structural analysis.<br/>In-depth TEM analysis of atomic and electronic structures can provide insights into the relationship between resistive switching and changes in the bulk oxygen concentration profile, directly imaging how these phenomena are interconnected. In this study, we induced resistive switching in an a-GaOx layer using in-situ biasing TEM. We employed a TEM holder from Nanofactory to induce changes in the oxygen concentration in the film and quantitatively confirmed changes in electrical conductivity by measuring resistance before and after switching.<br/>4D-STEM and STEM-EELS analyses provide the spatial distribution of short-range ordering, stoichiometry, and electronic structures. Particularly, the drift of oxygen vacancies under applied bias redistributes the local composition of the a-GaOx film. Given that the disproportion occurs below a critical value of x, local mapping of the radial distribution function and the radial variance profile by 4D-STEM, in correlation with different values of x, is crucial for understanding the switching behavior. This is further examined in conjunction with the electron-loss near edge structures.<br/>Our study provides direct insights into how changes in atomic and electronic structures influence resistive switching, advancing the understanding of the switching mechanism and highlighting its potential for next-generation memory device applications.