🔬 mesoscale physics

Nonlinear quantum multi-spin dynamics without entanglement

This paper introduces a variational product-state framework for simulating nonlinear quantum multi-spin dynamics in materials with arbitrary spin $s$ by excluding intersite entanglement while retaining local quantum effects, thereby enabling efficient $\mathcal{O}(N)$ simulations and clarifying the relationship between quantum Landau-Lifshitz and Landau-Lifshitz-Gilbert dynamics.

Original authors: Ivan P. Miranda, Erik Sjöqvist

Published 2026-10-06

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ivan P. Miranda, Erik Sjöqvist

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Magnetism is a force we encounter every day, from the simple act of a compass needle finding north to the complex data storage inside a computer. At the heart of this phenomenon are tiny atomic magnets, known as spins, which behave like miniature bar magnets. For decades, scientists have described how these spins move and interact using classical physics, treating them like spinning tops that wobble and settle down. This approach works remarkably well for large collections of atoms, but it breaks down when we look at the quantum world, where particles can exist in multiple states at once and become deeply linked to one another in ways that defy everyday logic. The challenge has been to create a bridge between these two worlds: a way to simulate the complex dance of quantum spins in real materials without getting lost in an explosion of mathematical complexity that makes the calculations impossible to run on even the most powerful computers.

A team of researchers has now charted a new path through this difficulty by proposing a method to simulate quantum spin dynamics that deliberately ignores the most famous quantum trick of all: entanglement. Entanglement is the phenomenon where particles become so deeply connected that the state of one instantly influences the other, no matter how far apart they are. While this is a fundamental feature of quantum mechanics, the researchers argue that in many real-world magnetic materials, the environment is so noisy and chaotic that this deep connection is destroyed almost instantly. Instead of trying to track these fleeting quantum links, the team built a model where each spin is treated as an independent quantum object that still feels the influence of its neighbors, but does not share a single, unified quantum state with them. This approach allows them to keep the rich, non-classical behavior of individual spins while stripping away the computational burden of tracking the entire system as one giant, entangled whole.

The researchers developed two distinct ways to describe how these independent quantum spins lose energy and settle down, a process known as dissipation. In the classical world, there is a well-known rule for how a spinning top slows down, and a slightly different but equivalent rule that describes the same motion. In the quantum realm, however, the team found that these two rules are not always the same. They derived two new equations of motion: one based on a principle of least energy loss, and another based on a mechanical balance of forces. For simple quantum systems, these two equations predict the same behavior, just with a slight difference in the speed of time. But for more complex systems, where the spins are in a mixed state of different possibilities, the two equations diverge. They lead to different outcomes because the geometry of the quantum state space is more intricate than a simple circle or sphere; the "clock" that ticks for the system depends on the specific internal structure of the spin itself.

To test their ideas, the team ran detailed computer simulations on systems ranging from a single atom to vast lattices containing thousands of spins. They first showed that their method could capture a unique quantum effect where a spin, under the right conditions, could lose its magnetic direction entirely and become "anti-coherent," a state that a classical spinning top could never achieve. They then explored what happens when the spin size increases. They found that as the spins get larger, the quantum behavior begins to look more and more like the classical behavior, eventually matching the predictions of the old classical equations. However, this match is not perfect for every single spin; it is a statistical agreement that emerges when you look at the average behavior of a huge number of them. In simulations involving thousands of spins, the tiny differences between the quantum and classical predictions canceled each other out when viewed from a distance, making the macroscopic material appear to follow classical laws, even though the individual atoms were behaving in distinctly quantum ways.

The study also clarified why the two different quantum equations they derived behave differently. The difference is not due to the spins being entangled with each other, since the model explicitly forbids that. Instead, the difference arises from the internal structure of the individual spins themselves. When a spin is in a complex mixture of states, the way it loses energy depends on the specific gaps between its energy levels. One equation treats this loss as a smooth, geometric flow, while the other treats it as a mechanical friction. These two perspectives are only equivalent if the spin has a very simple internal structure; otherwise, they describe two different physical realities. This finding suggests that the choice of how to model energy loss in quantum materials is not just a mathematical detail, but a fundamental decision that changes the predicted behavior of the system.

Ultimately, this work provides a practical toolkit for simulating magnetic materials that are too complex for full quantum calculations but too quantum for simple classical models. By focusing on local quantum effects while ignoring the rapid decay of entanglement, the researchers have created a framework that can handle systems with thousands of spins, a scale that was previously out of reach. Their results show that while the microscopic world of individual atoms may be governed by complex quantum rules, the macroscopic world we observe often emerges as a classical average of these rules. This explains why classical models have been so successful in describing real magnets for so long, even though they miss the subtle quantum nuances happening at the atomic level. The researchers have not just found a new way to calculate; they have offered a clearer picture of how the quantum world gives rise to the classical one, revealing that the bridge between them is built on the collective averaging of countless individual, slightly different quantum stories.

Technical Summary: Nonlinear Quantum Multi-Spin Dynamics without Entanglement

Problem Statement
The simulation of quantum spin dynamics in magnetic materials faces a critical challenge: a fully quantum treatment using a single density operator $\varrho(t)$ for $N$ spins leads to an exponential increase in degrees of freedom, rendering simulations feasible only for very small systems (tens of spins). While classical Landau-Lifshitz (LL) and Landau-Lifshitz-Gilbert (LLG) equations successfully describe macroscopic magnetic phenomena, their quantum analogs are complicated by the emergence of entanglement and the non-trivial structure of higher-spin ( $s \ge 1$ ) density operators. Existing approaches often struggle to balance computational scalability with the retention of essential local quantum effects, particularly for mixed states where the equivalence between LL and LLG forms breaks down.

Methodology
The authors introduce a variational framework based on a product-state ansatz, $\varrho(t) = \bigotimes_{j=1}^N \varrho^{(j)}(t)$ , which explicitly excludes intersite entanglement while retaining full local quantum dynamics for arbitrary spin $s$ . This approach is motivated by the physical expectation that decoherence in materials rapidly disentangles spins, leaving local coherences intact.

The derivation proceeds in two main stages:

Conservative Dynamics: The authors apply the Dirac-Frenkel time-dependent variational principle (DF-TDVP) to the product-state manifold. By restricting local density operators to isospectral orbits (preserving the spectrum of $\varrho^{(j)}(0)$ ), they derive a local quantum Landau-Lifshitz-Gilbert (q-LLG) equation. The conservative part is driven by a local effective mean-field Hamiltonian, $h_j$ , which depends on the instantaneous states of all other spins.
Dissipative Dynamics: To incorporate dissipation, two distinct variational constructions are employed on the same manifold:
- Rayleigh-Lagrange-d'Alembert (LA) Extension: This yields the q-LLG equation, where dissipation is modeled via a Rayleigh function proportional to the square of the tangent velocity in state space.
- Onsager Gradient Flow: This yields the q-LL equation, where dissipation is determined by minimizing a functional combining the energy rate and a quadratic dissipation cost.

Both equations are reformulated using generalized coherence vectors (Bloch vectors for $s > 1/2$ ) associated with the generators of $SU(2s+1)$ .

Key Contributions

Derivation of q-LL and q-LLG: The paper establishes a rigorous variational derivation for both quantum LL and LLG equations applicable to arbitrary spin $s$ and mixed states, without entanglement.
Geometric Origin of Inequivalence: A central theoretical contribution is the clarification of why q-LL and q-LLG are generally inequivalent for mixed states. The authors demonstrate that this inequivalence arises from the geometric structure of the isospectral orbit. Specifically, the superoperator $L_j^2(\cdot)$ acts as a scalar only for states with exactly two distinct eigenvalues (e.g., qubits or specific mixed states). For generic mixed states with $d > 2$ , distinct spectral gaps lead to different relaxation rates for different tangent directions, preventing a simple time rescaling between the two dynamics.
Criterion for Equivalence: The paper provides a precise theorem stating the conditions under which a common time rescaling exists between q-LL and q-LLG, linking this to the uniformity of spectral gaps across the system.
Scalable Simulation Framework: The formulation allows for $O(N)$ simulations of quantum spin dynamics, making it feasible to study macroscopic systems with quantum local degrees of freedom.

Results
The authors validate the framework through three numerical benchmarks:

Nonlinearity and Anti-coherence: Simulations of single and two-spin systems ( $s=1$ ) demonstrate that the product-state dynamics can generate first-order anti-coherent states (where the dipole moment vanishes) asymptotically, a feature impossible for classical spins of fixed length. The results confirm that nonlinearity arises from both mean-field interdependence and the structure of the dissipative terms, even in the absence of entanglement.
Large-Spin Limit ( $s \to \infty$ ): For coherent initial states, the quantum dynamics converge to the classical LL/LLG results, with mismatches in initial velocity and stationary states scaling as $1/s$ . However, for mixed states with three distinct eigenvalues, the q-LL and q-LLG trajectories remain distinct even in the large- $s$ limit, as no single time rescaling can align them due to persistent spectral gap differences.
Large-System Limit ( $N \to \infty$ ): In systems with many spins, the authors observe a "macroscopic masking" effect. While local discrepancies between quantum and classical trajectories remain finite ( $O(1)$ ), spatial averaging leads to a vanishing macroscopic discrepancy scaling as $N^{-1/2}$ , provided correlations are short-ranged. This suggests that classical macroscopic behavior can emerge from quantum microscopic dynamics via self-averaging, even when local quantum corrections are significant.

Significance
The paper claims to provide a distinct route for extending classical atomistic spin dynamics into the quantum regime, specifically for materials where intersite entanglement is negligible. By separating local quantum effects (non-classical density operator structure) from many-body entanglement, the framework offers a computationally efficient tool for realistic simulations.

The authors emphasize that the inequivalence between q-LL and q-LLG is not a fundamental consequence of entanglement but a geometric property of the local state space. Furthermore, the results offer a potential explanation for the empirical success of classical LLG models in describing macroscopic magnetic observables: the quantum-classical discrepancy may be effectively masked by spatial averaging in large systems, even if the underlying microscopic dynamics remain distinctly quantum. This framework bridges the gap between fully quantum treatments and classical atomistic models, enabling the study of quantum effects in macroscopic magnetic materials without the prohibitive cost of full many-body simulations.

More like this