Faculty and Staff Directory
Yuriy V. Pershin
|Department:||Physics and Astronomy
College of Arts and Sciences
|Office:||Jones, Room 503|
Emerging memory devices (theory and experiment), unconventional computing, and 2D materials.
Emerging memory devices:
Мemristive, memcapacitive, and meminductive systems are two-terminal circuit elements whose basic characteristics (resistance, capacitance, and inductance) retain the memory of the past states through which the elements have evolved. The memory features of these elements are related to corresponding internal states (e.g., atomic structure, spin polarization, etc.), which can be influenced by an external control parameter like the voltage, charge, current, or flux. We are designing memory circuit elements, developing their models, and fabricating them in the lab. Our work relies on the use of various theoretical and computational approaches and often focuses on various aspects of nonlinear circuit dynamics. Electrochemical metallization cells, ferroelectric capacitors, and emulators are in the focus of our experimental efforts.
Memory circuit elements have received a great deal of attention in the context of in-memory computing and neuromorphic architectures as they combine information processing and storage functionalities in simple device structures of nanoscale dimensions. We are investigating new approaches to computing based on the collective dynamics of memory circuit elements. Examples of our work include an experimental demonstration of Pavlovian learning with a memristive neural network, memristive Pascaline, etc.
Currently, there is a strong interest in mechanical and electromechanical properties of graphene from both fundamental and application points of view. Recently, we have introduced the notion of graphene kinks and antikinks, which are topological excitations of buckled graphene membranes. Their remarkable properties – nanoscale size, stability, high propagation speed, and low energy dissipation – make them well-positioned to be used in applications involving nanoscale motion. We are developing a fundamental understanding of graphene kinks and antikinks, and designing methods for kink motion control using such tools as molecular dynamics simulations.