High Resolution Study of the 2D ANNNI Model Using a Two-replica Cluster Algorithm and Population Annealing

Shane Keiser, Jonathan Machta · 2026

Read High Resolution Study of the 2D ANNNI Model Using a Two-replica Cluster Algorithm and Population Annealing by Shane Keiser, Jonathan Machta (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-00882.txt.

Introduction

Keiser and Machta take a fine-grained look at a classic magnet model in this 2026 physics paper. The 2D ANNNI model adds axial next-nearest-neighbor couplings to the Ising model. It hosts an incommensurate floating phase, marked by a string of sharp peaks in specific heat. In finite systems those peaks blur together and resist clean study.

Their tools pair population annealing with a two-replica cluster algorithm. With them, they fully resolve the peak sequence of the floating phase. The two-replica method equilibrates the system far better than single-replica clusters or Metropolis moves. They trace its edge to how it shifts whole lines of defects between replicas. Resampling in the annealing step then drops replicas heavy with defects. The mix of moves and selection keeps the sample honest at low heat. The work gives simulators a sharper lens for floating phases in layered magnets.

Catalog record

CREATORShane Keiser, Jonathan Machta
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE51 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.00882

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

High Resolution Study of the 2D ANNNI Model Using a Two-replica Cluster Algorithm and Population Annealing Shane Keiser, Jonathan Machta Abstract The axial next-nearest-neighbor Ising (ANNNI) model in two dimensions is studied using population annealing combined with a two-replica cluster algorithm. We are able to fully resolve the sequence of sharp specific heat peaks that characterize the finite-size incommensurate floating phase. We also show that the two-replica cluster algorithm is much more effective in equilibrating the system than either single-replica cluster methods or the Metropolis algorithm when these are combined with population annealing. We argue that effectiveness of the new algorithm is due to its ability to move groups of defect lines between replicas combined with resampling in population annealing, which removes replicas from the population that have larger numbers of defect lines. I Introduction Materials with short range attractive and long range repulsive potentials often display equilibrium order with discrete translational symmetries. If the interactions are anisotropic, layered structures may form. Perhaps the simplest example of layering is found in the anisotropic next nearest neighbor Ising (ANNNI) model Fisher and Selke 1980 ; Selke 1988 . The ANNNI model has both ferromagnetic nearest neighbor interactions and competing antiferromagnetic next-nearest-neighbor (NNN) interactions in the axial direction. The ANNNI model has a rich phase diagram that features layered structures at low temperatures. We studied the 2D ANNNI model in the region in parameter space where the NNN interactions are sufficiently strong that the low temperature phase displays layered ordering with two spin-up layers followed by two spin-down layers in the axial direction. At higher temperatures there is believed to be an incommensurate floating (IC) phase with quasi-long-range order that is layered but with defect lines. Finally, at high temperature, there is a paramagnetic phase with short-range order. Although there are theoretical reasons Villain, J. and Bak, P. 1981 for expecting an IC phase, there is no proof of its existence, and the numerical results have been inconclusive, with disagreement on the existence Derian et al. 2006a ; Chandra and Dasgupta 2007 of the IC phase and its width in temperature Hu and Charbonneau 2021 . In this work, we find strong evidence for the existence of an IC phase but are unable to locate the transition temperature due to strong finite size corrections. The ANNNI model is difficult to simulate because of the frustration induced by the competing ferromagnetic and antiferromagnetic interactions. In addition to the standard Metropolis algorithm, several other computational methods have been applied to study the ANNNI model. In 3D, near the uniaxial Lifshitz point, Henkel and Pleimling applied a generalization of the Wolff algorithm Wolff 1989 for interactions of both signs to the ANNNI model Pleimling and Henkel 2001 . However, the Henkel–Pleimling cluster algorithm is not useful in the regime of strong NNN interactions for reasons discussed in Zheng et al. 2022 and in the present paper. The most successful computational methods to date for the ANNNI model use transfer matrices, either alone Derian et al. 2006a ; Hu and Charbonneau 2021 or in combination with Monte Carlo moves Matsubara et al. 1997 ; Sato and Matsubara 1999 ; Matsubara et al. 2017 . In the present work, we introduce a new computational method that uses population annealing Hukushima 2003 ; Machta 2010 ; Wang et al. 2015 together with two-replica cluster moves Swendsen and Wang 1986 ; Chayes et al. 1998 ; Redner et al. 1998 ; Li and Machta 2001 and single-spin flips. We find that this method is far more effective at equilibrating the 2D ANNNI model than either the Metropolis algorithm or the Wolff–Henkel–Pleimling algorithm in the strongly frustrated regime. In Sec. II we introduce and discuss the relevant features of the ANNNI model. Section III reviews population annealing and the two-replica cluster method, and describes how they are combined. This section also reviews the competing algorithms and introduces the observables to be measured. Results for both ANNNI model physics and algorithmic performance are presented in Sec. V . The paper closes with a discussion, Sec. VI . II Model The ANNNI model energy, E is described by \displaystyle E=-\left(\sum_{\langle i,j\rangle}\sigma_{i}\sigma_{j}-\kappa\sum_{[i,j]_{\text{z}}}\sigma_{i}\sigma_{j}\right), (1) where the sums are over nearest neighbors, \langle i,j\rangle and next-nearest neighbors in the axial z -direction, [i,j]_{\text{z}} . Our simulations are carried out on L\times L square lattices with periodic boundary conditions. The system size, L is chosen to be a multiple of 4 so that it admits layered structures without defects. At zero temperature and in any dimension d , as \kappa increases from zero, there is a transition from the twofold degenerate ferromagnetic ground state with energy per spin of (-d+\kappa) to a fourfold degenerate, \langle 2,2\rangle ground state. The \langle 2,2\rangle ground state displays a repeating pattern \uparrow\uparrow\downarrow\downarrow in the axial direction. The four degenerate \langle 2,2\rangle ground states differ by translations in the axial direction and have the energy per spin (-d+1-\kappa) . Comparing these energies, one sees that there is a transition from the ferromagnetic ground state to the \langle 2,2\rangle ground state at \kappa=1/2 . Although the ground state is the same in any number of dimensions, the 2D and 3D phase diagrams are quite different. In 3D there are many ordered phases characterized by different repeating patterns in the axial direction. In 2D, the subject of this study, it is believed that there are four distinct phases. For sufficiently low temperature and \kappa<1/2 there is a ferromagnetic phase, while for \kappa>1/2 there is an ordered \langle 2,2\rangle phase. At high temperature there is a paramagnetic phase. Intervening between the \langle 2,2\rangle phase and the paramagnetic phase there is believed to be a narrow critical phase, known as the incommensurate floating (IC) phase, which displays quasi-long range order. The transition between paramagnetic and IC phases is believed to be in the Brezinski–Kosterlitz–Thouless universality class, though the location of the transition line is poorly determined. The transition between the IC phase and the ordered phase is believed to be of the Pokrovsky–Talapov (PT) Pokrovsky and Talapov 1979 ; Selke 1988 incommensurate to commensurate type and is described by free fermion theory. Figure 1: Snapshots of two L=48 replicas at \beta=1.09 near the last specific heat peak with coexistence between a configuration in the IC phase (left panel) with four defect lines ( k=11/48 ) and a configuration in the \langle 2,2\rangle phase ( k=12/48 ). In the IC phase, typical configurations have regions of \langle 2,2\rangle order separated by defect lines consisting of layers with three or more like spins. A three-spin defect line translates the \langle 2,2\rangle phase type by one. After four such defect lines, the \langle 2,2\rangle phase returns to the same translation as before. Thus, in periodic boundary conditions with L a multiple of four, as is the case in our study, the number of three-spin defect lines must be a multiple of four. Two typical snapshots taken from our simulations are shown in Fig. 1 . These snapshots are taken near the transition between the IC and \langle 2,2\rangle phases. The left panel of Fig. 1 shows a configuration in the IC phase with four defect lines. The right panel shows a configuration that is representative of the \langle

Read the complete text