Chaotic dynamics in spin torque nano oscillator driven by voltage feedback

TL;DR

This paper investigates how voltage feedback can induce chaotic dynamics in a spin torque nano oscillator with a multi-dimensional system, using a 3-terminal MTJ device, and analyzes the chaos through various nonlinear dynamics tools.

Contribution

It demonstrates the emergence of chaos in a spintronic device driven by voltage feedback, expanding understanding of nonlinear dynamics in multi-dimensional spin torque oscillators.

Findings

Chaotic regimes observed via Lyapunov exponents and bifurcation diagrams.

Voltage feedback delay and gain influence the transition to chaos.

Potential applications in random number generation and reservoir computing.

Abstract

Non-linear dynamics, including auto-oscillations, chaotic dynamics, and synchronization, are integral to physical and biological applications and can be excited in spintronic devices. In this study, we are interested in exploring the excitation of chaos using voltage feedback in a spin torque nano oscillator using a Magnetic Tunnel Junction (MTJ). According to the Poincar\'e-Bendixson theorem, chaos cannot arise in a two-dimensional system of MTJ featuring two dynamic variables describing the zenith and azimuth angles of magnetization. Hence, we prefer the feedback system as it creates a multi-dimensional system, making it interesting to explore the emergence of chaos in such systems. Such feedback is achieved by utilizing a 3-terminal device consisting of an MTJ with an in-plane pinned layer (PL) and an out-of-plane free layer (FL) geometry. When a DC current above the critical threshold is applied, the FL's oscillating magnetization generates an AC output voltage through the Tunnel Magneto Resistance (TMR) effect. A fraction of this voltage, fed back after a delay, modulates the FL's anisotropy via voltage controlled magnetic anisotropy (VCMA) effect, potentially driving precessional motion or chaotic dynamics or oscillator death based on the feedback delay and gain of the feedback circuit. The observed chaotic regime has been studied by evaluating the Lyapunov exponent, bifurcation diagrams, Fourier spectral analysis and reconstruction of the trajectory in embedding phase space. Such observed chaotic dynamics can find practical applications in random number generators and physical reservoir computing.