Electron-phonon-coupled Langevin dynamics for strongly-correlated insulators

This paper derives generalized stochastic Landau-Lifshitz-Gilbert equations from first principles for spin-orbital coupled Mott insulators by incorporating electron-phonon interactions via a Keldysh path-integral formalism, thereby establishing a microscopic framework that accurately captures dissipative spin dynamics, thermal fluctuations, and non-equilibrium relaxation processes.

The Gist

The Big Picture: Why Do Materials "Cool Down"?

Imagine you are watching a crowded dance floor. The dancers are electrons, and the floorboards are the atoms of a material. In a special type of material called a Mott insulator, the dancers are so crowded and stubborn that they can't move freely to conduct electricity. Instead, they just spin and wiggle in place.

Scientists have long used a set of rules called the Landau-Lifshitz-Gilbert (LLG) equations to predict how these dancers spin. However, there's a problem with the old rules: they treat the "cooling down" process (dissipation) like a magic trick. They just say, "Okay, they lose energy," without explaining how or where that energy goes. It's like saying a car slows down because "friction exists," without mentioning the brakes or the road.

This paper introduces a new, more honest way to simulate these materials. The authors built a microscopic model that shows exactly how the dancers (electrons) interact with the floorboards (lattice vibrations/phonons) to lose energy and eventually settle down.

The New Tool: A "Microscopic" Dance Simulator

The authors created a new simulation method called electron-phonon coupled Langevin dynamics (epLD). Here is how it works, broken down into three parts:

1. The Dancers and the Floor (Electrons and Phonons)
In their simulation, the electrons aren't just spinning in a vacuum. They are constantly bumping into the floorboards. When an electron spins, it makes the floor vibrate. These vibrations are called phonons.

• The Analogy: Imagine a dancer (electron) spinning on a wooden stage. As they spin, they make the stage shake. The shaking isn't just a side effect; it's how the dancer loses their energy.

2. The Heat Bath (The Thermal Reservoir)
The floorboards themselves are connected to a giant, invisible "heat bath" (like a massive cooling system or the surrounding air).

• The Analogy: The shaking floorboards are connected to a giant sponge (the thermal bath) that soaks up the vibrations. This is how the energy leaves the system. The authors mathematically proved that this connection creates two things:
  • Damping: The floorboards resist the dancer's motion, slowing them down.
  • Noise: The sponge also jiggles randomly, giving the floorboards tiny, random kicks (thermal noise).

3. The Result: A Realistic Story
By connecting the dancers to the floor, and the floor to the sponge, the authors derived a new set of equations. These equations naturally produce the "friction" and "random jiggling" that the old rules had to guess at.

• The Result: When they ran the simulation, the system didn't just magically stop. It went through realistic stages:
  • Uncorrelated: At first, the dancers and floorboards are out of sync.
  • Dissipative: The dancers start transferring their energy to the floor, which passes it to the sponge.
  • Adiabatic: The dancers and floorboards start moving together in a synchronized rhythm.
  • Equilibrium: Finally, everything settles into a calm, steady state, just like a real material cooling down.

The "Hybrid" Surprise

One of the coolest findings in the paper is what happens when the dancers and the floorboards talk to each other very strongly.

• The Analogy: Imagine a dancer and a trampoline. If the dancer is light and the trampoline is stiff, they act separately. But if they are perfectly tuned to each other, they stop being two separate things and become a single, hybrid entity.
• The Finding: The authors showed that when the electron-phonon coupling is strong, the "dancers" (electronic excitations) and the "floorboards" (phonons) mix together. They create hybrid modes. The floorboards, which usually just vibrate in place, start to look like they are moving across the material (gaining "dispersion") because they are so tightly linked to the electrons. It's like the floorboards start dancing the same steps as the dancers.

Connecting Back to the Old Rules

The authors also checked if their fancy new simulation could do what the old, simpler rules (LLG) do.

• The Finding: They proved that if you take their complex, microscopic simulation and simplify it (by assuming the floor vibrations are very fast and the temperature is high), the equations turn into the exact same LLG equations that scientists have been using for decades.
• Why this matters: This confirms that the old rules are actually a "special case" of the new, more complete theory. It validates the old rules while showing us the deeper truth underneath them.

Summary

In short, this paper builds a microscopic bridge between the tiny world of electrons and the vibrating world of atoms.

• Old way: "Electrons lose energy because we say so."
• New way: "Electrons lose energy because they shake the floor, and the floor passes that energy to a heat bath, creating friction and random noise naturally."

This new framework allows scientists to simulate how these materials behave not just when they are calm, but when they are being heated up, cooled down, or hit with a laser pulse, providing a much more realistic picture of how real-world materials function.

Technical Summary

Technical Summary: Electron-Phonon-Coupled Langevin Dynamics for Strongly-Correlated Insulators

Problem Statement
The dynamical behavior of real materials, particularly Mott insulators where strong correlations suppress metallic behavior, remains a central challenge in condensed matter physics. While stochastic Landau-Lifshitz-Gilbert (LLG) equations are widely used to simulate spin dynamics in these systems, they typically introduce energy damping phenomenologically. Consequently, their validity for describing nonequilibrium processes and their connection to the microscopic origins of dissipation in real materials are unclear. Existing first-principles approaches that incorporate electron-phonon interactions are computationally demanding and struggle to access long-time nonequilibrium dynamics. There is a lack of a fully microscopic framework that consistently captures electron-phonon coupling, dissipation, and relaxation in strongly correlated systems.

Methodology
The authors develop a microscopic theory termed electron-phonon-coupled Langevin dynamics (epLD) to describe the nonequilibrium dynamics of spin-orbital coupled Mott insulators. The methodology proceeds as follows:

1. Model Hamiltonian: The system is modeled using a generalized Kugel-Khomskii Hamiltonian for localized electrons interacting via spin-orbital exchange, coupled to Einstein phonons (lattice vibrations) and an external thermal bath. The electronic degrees of freedom are treated within a multiorbital framework using $SU(N)$ coherent states, where $N = 2S + 1$.
2. Path Integral Formalism: The derivation utilizes the Keldysh path-integral formalism on the Keldysh contour. This approach treats the system, phonons, and thermal bath explicitly. The phonons are coupled to the bath to introduce damping and stochastic noise, satisfying the fluctuation-dissipation theorem.
3. Semiclassical Limit: By taking the semiclassical limit of the quantum transition amplitude on the Keldysh contour, the authors derive deterministic equations of motion augmented by damping and thermal noise terms. These terms are derived microscopically from the underlying electron-phonon coupling.
4. Markovian Approximation: The initial equations of motion are non-Markovian (history-dependent) due to memory effects in the retarded self-energy and colored noise. To facilitate numerical simulation, the authors introduce auxiliary variables to map the non-Markovian dynamics onto an equivalent Markovian process. Furthermore, in the high-temperature limit ($\hbar\omega \ll k_B T$), the noise correlations are approximated as white noise, resulting in a set of coupled first-order differential equations suitable for numerical integration.

Key Contributions and Results
The paper presents a robust theoretical framework and validates it through numerical benchmarks on a two-orbital spin-1 chain coupled to Einstein phonons.

• Derivation of Generalized Equations: The authors derive generalized stochastic equations of motion (Eq. 69) that explicitly couple electronic variables (parametrized by $SU(N)$ coherent states) to local phonon modes. These equations include microscopically derived damping and noise terms, replacing the phenomenological damping of standard LLG equations.
• Nonequilibrium Relaxation: Numerical integration demonstrates realistic cooling dynamics. The system exhibits distinct regimes: an initial uncorrelated phase, a dissipative phase where energy transfers from electrons to phonons and then to the bath, an adiabatic phase where electronic and phononic motions become correlated, and a final equilibrium state. The simulations accurately reproduce thermodynamic properties upon equilibration.
• Thermodynamic Benchmarking: The epLD results for thermodynamic observables (energy, magnetization, specific heat, and susceptibility) are compared against independent stochastic $U(3)$ Monte Carlo (u3MC) simulations. In the limit of vanishing electron-phonon coupling ($g \to 0$), the epLD results converge to the u3MC results, validating the formalism's consistency with known equilibrium properties.
• Hybridization of Modes: The study demonstrates that electron-phonon coupling induces hybridization between electronic and phononic modes in the excitation spectrum. As the coupling strength increases, flat phonon bands mix with quadrupole excitations, leading to band repulsion (anticrossing) and the formation of hybrid modes. This hybridization also affects the spin-dipole sector, splitting magnon bands.
• Recovery of LLG Equations: The paper demonstrates that the conventional stochastic LLG equations are recovered as a limiting case of the microscopic epLD theory. Specifically, for $SU(2)$ coherent states coupled to acoustic phonons in the weak damping and high-temperature limit, the derived equations map one-to-one onto the standard LLG equation with Gilbert damping and thermal noise.

Significance
The authors establish a reliable framework for capturing dissipative spin dynamics in strongly correlated systems, both in and out of equilibrium. By deriving damping and noise terms from first principles rather than introducing them phenomenologically, the approach bridges the gap between microscopic electron-phonon interactions and macroscopic dynamical behavior. The framework enables the simulation of realistic materials by allowing parameters (such as exchange interactions, phonon frequencies, and coupling constants) to be obtained from first-principles calculations. This provides a pathway to connect electronic band-structure calculations with microscopic dynamical behavior, offering a tool to study phenomena such as thermalization, decoherence, and nonequilibrium phase transitions in Mott insulators. The paper notes that while the current formulation focuses on localized electrons, the structure suggests potential extensions to itinerant fermionic systems and more realistic bath geometries.