Revisiting the Phase Transition of Spin-1/2 Heisenberg Model with a Spatially Staggered Anisotropy on the Square Lattice

TL;DR

This study re-examines the phase transition in a spin-1/2 Heisenberg model with staggered anisotropy using Monte Carlo simulations, proposing a new finite-size scaling approach that yields results consistent with the expected universality class.

Contribution

The paper introduces a novel finite-size scaling method based on fixed ratio of spatial winding numbers, improving analysis accuracy with limited large-volume data.

Findings

$ ho_{s2} L$ exhibits less correction than $ ho_{s1} L$

The new scaling approach confirms the critical exponent $ u$ matches $O(3)$ universality

Validation on ladder anisotropic Heisenberg model supports the method

Abstract

Puzzled by the indication of a new critical theory for the spin-1/2 Heisenberg model with a spatially staggered anisotropy on the square lattice as suggested in \cite{Wenzel08}, we re-investigate the phase transition of this model induced by dimerization using first principle Monte Carlo simulations. We focus on studying the finite-size scaling of $\rho_{s1} L$ and $\rho_{s2} L$, where $L$ stands for the spatial box size used in the simulations and $\rho_{si}$ with $i \in \{1,2\}$ is the spin-stiffness in $i$-direction. From our Monte Carlo data, we find that $\rho_{s2} L$ suffers a much less severe correction compared to that of $\rho_{s1} L$. Therefore $\rho_{s2} L$ is a better quantity than $\rho_{s1} L$ for finite-size scaling analysis concerning the limitation of the availability of large volumes data in our study. Further, motivated by the so-called cubical regime in magnon chiral perturbation theory, we additionally perform a finite-size scaling analysis on our Monte Carlo data with the assumption that the ratio of spatial winding numbers squared is fixed through all simulations. As a result, the physical shape of the system remains fixed in our calculations. The validity of this new idea is confirmed by studying the phase transition driven by spatial anisotropy for the ladder anisotropic Heisenberg model. With this new strategy, even from $\rho_{s1} L$ which receives the most serious correction among the observables considered in this study, we arrive at a value for the critical exponent $\nu$ which is consistent with the expected $O(3)$ value by using only up to $L = 64$ data points.