Electric Field Control of Magnetization Switching in Magnetic Tunnel Junctions

2019-11-07T14:48:56Z (GMT) by Mukund Bapna
Magnetic tunnel junctions (MTJs) are the fundamental building blocks for technology such as magnetic random access memory (MRAM) and hard disk drives (HDD). MTJs store information in form of magnetization direction of its recording layer with respect to its reference layer. The process through which magnetization reversal can be achieved is thus fundamental and of great interest for MRAM technology. Magnetization switching can be achieved for example using a magnetic field, spin-transfer torque or by spin-orbit torque. The first commercially available MRAM were based on switching the MTJs with magnetic fields. Due to architectural and energy concerns with field driven operations it was difficult to meet the demand of ever increasing computational needs. The discovery of spin transfer torque (STT) in 1996 led to a surge in research and development of STT based memory that later became a mainstream technology. MRAM based on STT technology were commercially available around 2006. However, the need of large current densities ~106 A/cm2 to perform read and write operations resulted in significant heat dissipation for high bit densities and with bit sizes less than 50 nanometers. This still continues to impede the development of STT based MRAM technology for future computational needs. The MRAM technology is still evolving, and new phenomenology and materials are been extensively explored by industries, universities and research institutes across the globe. In this thesis I present the experimental study of device level physics on patterned magnetic tunnel junctions that are less than 100 nm in size. In Chapter 1, I describe the technique of Conductive atomic force microscopy (C-AFM) which is central to the research presented in this thesis. C-AFM provides with a sharp conductive probe (~20 nm tip radius of curvature) that can allow for electrical measurements on nano-size devices that otherwise need complicated interconnects. The C-AFM accelerates research at prototype state by eliminating the need of hardwiring the nanoscale devices.