NSF-NRI Supplement: Spintronic Logic Devices
Funding: NSF/NRI
Personnel
This project is still in very early stages, so what follows below is basically straight from the proposal!William Buter Arunava Gupta Subhadra Gupta Patrick LeClair Tim Mewes Dipanjan Mazumdar Vijay Sankar
Intellectual Merit
Electrons behave like tiny bar magnets due to their inherent spin degree of freedom. The orientation of these tiny bar magnets - electron spin up and spin down states - can used to store digital information, even without electrical power. This rapidly growing area of research is commonly known as “spintronics.” Already, the magnetic information storage industry has benefited greatly from spintronic read sensors for hard disks. In principle, the use of electron spin degree of freedom leads to increased density of information per device at ultra-low power of operation. From a technical standpoint, many challenges exist that need to be solved to make ‘spin’ electronics competitive, and in particular, make it compatible with existing semiconductor technology.
Many exciting ideas for truly novel spintronics devices — such as reprogrammable logic or quantum information processing — have yet to be realized. One fundamental reason is that more efficient methods of highly polarized electron spin injection and detection must be realized. If currents consisting of purely spin up or spin down electrons can be generated and detected, a wide variety of transformative spin electronics devices could be realized. This research aims at perfecting the injection and detection of polarized electrons using spin filters to create spintronic programmable logic devices.
We propose to explore spin-based devices centered on the robust, but so far unexploited effect of spin filter tunneling. Spin filter tunneling with magnetic insulators is a demonstrated mechanism for producing currents composed of only one electron spin, allowing near-perfect spin injection and detection. Spin filtering with nearly 100% efficiency has already been demonstrated with ferromagnetic Eu-based compounds, such as EuO and EuS. More recent candidate materials, offering operation above room temperature, have been identified and demonstrated with some success.
We propose methods to integrate efficient spin filter materials with semiconductor structures for spin-based devices. The focus will be on demonstrating components that can be readily integrated within existing paradigms, which we believe are within current technologies. The primary initial hurdles are materials development and proof-of-concept devices, and this will be our main focus. Proof-of-concept devices will be aimed squarely at magnetic field-programmable logic, which could have a transformative effect on computing - particularly in the mobile sector where size and power consumption are paramount.
Programmable logic is based on the idea of generic hardware that can be programmed for a specific functionality. In essence, it is the vision of entire chips whose end functionality is completely configurable in real time. Existing technologies, such as field-programmable gate arrays (FPGAs), suffer from prohibitively large power consumption and programming speeds - both of which are areas of primary advantage for spin-based technologies. We propose a specific plan of materials and device research to achieve our goals of efficient spin-based logic devices.