Improved upper and lower energy bounds for antiferromagnetic Heisenberg spin systems

TL;DR

This paper develops new upper and lower energy bounds for large antiferromagnetic Heisenberg spin systems, improving the accuracy of energy estimates where exact solutions are infeasible.

Contribution

It introduces novel bounds based on graph topology assumptions, namely n-cyclicity and n-homogeneity, enhancing previous methods for estimating minimal energies.

Findings

Upper bounds are closer to true energies than lower bounds.

Bounds are applicable to systems with specific topological properties.

Numerical analysis confirms the bounds' effectiveness.

Abstract

Large spin systems as given by magnetic macromolecules or two-dimensional spin arrays rule out an exact diagonalization of the Hamiltonian. Nevertheless, it is possible to derive upper and lower bounds of the minimal energies, i.e. the smallest energies for a given total spin S. The energy bounds are derived under additional assumptions on the topology of the coupling between the spins. The upper bound follows from "n-cyclicity", which roughly means that the graph of interactions can be wrapped round a ring with n vertices. The lower bound improves earlier results and follows from "n-homogeneity", i.e. from the assumption that the set of spins can be decomposed into n subsets where the interactions inside and between spins of different subsets fulfill certain homogeneity conditions. Many Heisenberg spin systems comply with both concepts such that both bounds are available. By investigating small systems which can be numerically diagonalized we find that the upper bounds are considerably closer to the true minimal energies than the lower ones.