Modeling of spin transport in MTJ devices

Deepanjan Datta, Purdue University

Abstract

Spin-Transfer Torque Random Access Memory (STT-RAM) is a promising candidate for the next generation universal memory technology, with a combination of non-volatility, high endurance, high speed, and better scalability compared to existing memory technologies. The key building block of the operation of STT-RAM is the Magnetic Tunnel Junction (MTJ) device, which uses spin polarized currents instead of magnetic fields to read/ write magnetic bits. While the read process is largely determined by the charge current, the spin current is critical to the write process. This thesis presents a Non-equilibrium Green's Function (NEGF) based transport model for MTJ devices that allows us to model both charge and spin currents accurately. Although there are several experiments describing the bias and switching behavior of MTJ, none of the existing theoretical models correctly provide quantitative agreement with experiments. Using our model, we show: (1) good agreement with diverse experimental measurements like resistance and spin torques; (2) provide simple explanation for bias dependences of the torques; and (3) propose an asymmetric STT device, which has the potential of switching in non-reciprocal way. Furthermore, we present an analytical expression for the spin current starting from the current operator expression; we discuss about the formalism of Non-Equilibrium spin current in the coherent transport regime and discuss initial results for a possible approach to developing a compact model for STT-RAM cell integrable with CMOS circuits.

Degree

Ph.D.

Advisors

Datta, Purdue University.

Subject Area

Electrical engineering|Nanoscience

COinS