Cross-Sectional Area Dependence of Tunnel Magnetoresistance, Thermal Stability, and Critical Current Density in MTJ

Abstract

Tunnel magneto-resistance (TMR), thermal stability, and critical switching current are important metrics of a magnetic tunnel junction (MTJ). In this work, a detailed study of these metrics is conducted for the down-scaling of the transverse dimensions of the MTJ. The quantum transport and the magnetization dynamics simulations are performed using nonequilibrium Green's function in the mode-space approach and Object Oriented Micromagnetic Framework (OOMMF), respectively. The study of areal size quantization effects on the TMR shows that most of the contribution to the TMR comes from lower energy sub-bands and that the TMR saturates for dimensions above 50 nm. An anomalous behavior is observed in the bias dependence of TMR for the lower energy sub-bands and is explained in terms of the modified Slonczewski's analytical model for conductance around zero bias. The study of TMR scaling is extended to consider non-idealities by introducing elastic dephasing into our simulations. It is shown that with down-scaling of diameters, dephasing affects the zero bias TMR predominantly below 20 nm. Further, TMR is also studied in terms of sensitivity to the variations in the interface layer and the asymmetric reduction of TMR with bias and its reversal at higher bias is observed. OOMMF simulations of the larger stack, including the free layer, are carried out to understand the qualitative link between magnet switching behavior, thermal stability, and critical current density with area scaling. It is shown that the area dependence of thermal stability and critical current follow each other qualitatively and the scaling of both these metrics is correlated to different regimes of magnetization switching such as macrospin behavior or formation of metastable complex textures. The implications of scaling, on the various MTJ metrics, are discussed in terms of the application domain.

Voltage-Controlled Domain Wall Motion-Based Neuron and Stochastic Magnetic Tunnel Junction Synapse for Neuromorphic Computing Applications

“…We use effective mass tight binding and mode space approach method to formulate the MTJ device as shown in Fig. 2(b) [36]. The complete device Hamiltonian is expressed as below- [37] 𝐻 𝐷 = 𝐻 𝐿𝐹𝑀 + 𝐻 𝐼1 + 𝐻 𝑇𝐵 + 𝐻 𝐼2 + 𝐻 𝑅𝐹𝑀 (6) where, HD is the complete device Hamiltonian consisting of the HLFM, HI, HTB, HRFM corresponding to the Hamiltonians of left FM, interface, TB, and right FM respectively.…”

Magnetic tunnel junction based implementation of spike time dependent plasticity learning for pattern recognition

“…The current, for each transverse mode, is modeled by a 1D independent transport channel. We describe the effective mass Hamiltonian for both FM layers and TB in terms of onsite potential U onsite and hopping parameter t is given by: where , m, and a are reduced Planck's constant, effective mass of electron in respective material and lattice spacing [20]. The onsite potential U onsite depends upon the band structure of the material.…”

Interface imperfection effects on spin transfer torque switching: an atomistic approach

“…We performed micromagnetic simulations using the objectoriented micromagnetic framework (OOMMF) to determine the switching current density [28,29]. The roughness variation was incorporated to OOMMF based on the corresponding variation in damping constant.…”