Emergent Non-Abelian Gauge Theory in Coupled Spin-Electron Dynamics

TL;DR

This paper derives a non-abelian gauge theory describing coupled spin-electron dynamics, revealing new geometrical spin torques and nutational motion effects beyond traditional adiabatic models.

Contribution

It introduces a novel non-abelian gauge theoretical framework for spin-electron systems, capturing complex geometrical effects and spin dynamics not accounted for in prior adiabatic theories.

Findings

Non-abelian gauge structure emerges in low-energy spin-electron dynamics.

Gauge-invariant spin-Berry curvature induces additional spin torques.

Numerical results show agreement with full theory beyond adiabatic approximation.

Abstract

A clear separation of the time scales governing the dynamics of "slow" and "fast" degrees of freedom often serves as a prerequisite for the emergence of an independent low-energy theory. Here, we consider (slow) classical spins exchange coupled to a tight-binding system of (fast) conduction electrons. The effective equations of motion are derived under the constraint that the quantum state of the electron system at any instant of time $t$ lies in the $n$-dimensional low-energy subspace for the corresponding spin configuration at $t$. The effective low-energy theory unfolds itself straightforwardly and takes the form of a non-abelian gauge theory with the gauge freedom given by the arbitrariness of the basis spanning the instantaneous low-energy sector. The holonomic constraint generates a gauge covariant spin-Berry curvature tensor in the equations of motion for the classical spins. In the non-abelian theory for $n>1$, opposed to the $n=1$ adiabatic spin dynamics theory, the spin-Berry curvature is generically nonzero, even for time-reversal symmetric systems. Its expectation value with the representation of the electron state is gauge invariant and gives rise to an additional {\em geometrical} spin torque. Besides anomalous precession, the $n\ge 2$ theory also captures the spin nutational motion, which is usually considered as a retardation effect. This is demonstrated by proof-of-principle numerical calculations for a minimal model with a single classical spin. Already for $n=2$ and in parameter regimes where the $n=1$ adiabatic theory breaks down, we find good agreement with results obtained from the full (unconstrained) theory.