This project is jointly funded by the Electronic and Photonic Materials Program and the Ceramics Program, both in the Division of Materials Research.
Non-technical Description: This project addresses fundamental materials science mechanisms responsible for the reversible and nonvolatile changes of resistivity in resistive switching devices used in novel high-density computer memories. The research activities rely to large extent on advanced electron microscopy techniques. Better understanding of such mechanisms allows for improved quality and reliability of resistive switches and thus leads to more powerful computers. The project involves graduate and undergraduate students at Carnegie Mellon University. Each year, two freshmen work on semester-long research projects involving memristors, through a university-funded program. One of them is expected to continue the research to senior year with a fellowship from the Semiconductor Research Corporation. Principal Investigators actively collaborate with semiconductor companies such as GlobalFoundries, Applied Materials, and Intel. The collaboration brings summer internship opportunities for graduate and undergraduate students in the project.
This project explores the fundamental mechanisms of bipolar resistive switching phenomena in metal/oxide/metal systems. It is known that the small diameter conducting filament forms in the high resistance functional oxide film during a conditioning process referred to as electroforming. The model of electroforming and resistive switching developed during the PI's prior NSF project asserts that either one of these processes has two stages. In the first stage, the electron transport instability leads to appearance of negative differential resistance in the device characteristics, which in turn causes formation of the high current density filament. This stage of the filament formation is fully reversible, i.e., the filament dissolves if the voltage is reduced. If the current density exceeds a critical value, in the second stage the localized Joule heating leads to nonvolatile changes in the oxide structure. This project focuses on the fundamental mechanisms in both stages of filament formation using advanced electron microscopy techniques, i.e., in-situ imaging of the nanoscale metal-insulator-metal devices under bias. The imaging techniques include electron energy loss spectroscopy, differential diffraction, high-angle annular dark field, and electron holography, in collaboration with Arizona State University. The experimental findings are used to delineate several possible mechanisms of electronic instability and to assess the nature of the nonvolatile changes in the oxide layer.
