Physics Colloquium, May 20, 2003

Spin Transport and Spin Transfer Effects in Magnetic Nanostructures

Robert A. Buhrman

Director, Center for Nanoscale Systems
John Edson Sweet Professor of Engineering
School of Applied and Engineering Physics
Cornell University

The spin-dependent electron transport properties of thin film metallic multilayer structures that incorporate ferromagnetic layers, particularly when patterned into magnetic nanostructure configurations, are currently the focus of substantial scientific interest and the foundation for important present-day and future technological applications. To better understand and exploit these spin transport properties has required both the development of new measurement techniques and the ability to fabricate new types of device structures with controllable nanoscale dimensions. In this talk I will first discuss experiments that my colleagues and I have been pursuing at Cornell for the study of ballistic and near-ballistic spin-dependent transport in magnetic nanostructures, in both the thermal electron and hot electron regimes. I will then discuss the recently demonstrated phenomena of spin transfer. This effect occurs when a spin polarized current flows from one ferromagnetic layer through a normal spacer layer and impinges onto a second ferromagnetic layer. Due to spin-dependent scattering at the normal metal - ferromagnet interface, spin momentum is transferred from the current to the second ferromagnet and a net torque is exerted on its magnetic moment. This spin transfer effect can be utilized to reversibly switch a nanomagnet and/or excite it into microwave oscillation. Thus spin transfer is providing new ways to study both spin transport at the nanoscale and the dynamic behavior of thin film nanomagnets. Since spin transfer becomes increasingly more effective than the magnetic field as a means of exciting and switching a nanomagnet as the dimensions of the nanomagnet decrease, this new phenomena has considerable promise for being the foundation of a ultra-high density non-volatile memory technology. I will close by discussing some of the scientific challenges and questions that must be addressed before such a possibility can be realized.

3.30 p.m., Smith Laboratory, Room 1005

Refreshments served in Smith 1094 at 3:00 p.m.