[Paper Review] Two dimensional Ising model with long-range competing interactions

Sergio A. Cannas, Pablo M. Gleiser+1 more|arXiv (Cornell University)|Feb 16, 2005

Theoretical and Computational Physics3 references3 citations

TL;DR

This paper reviews equilibrium and non-equilibrium properties of the 2D Ising model with competing short-range ferromagnetic and long-range antiferromagnetic dipolar interactions. It establishes that for δ > δₐ ≈ 0.425, the ground state is a striped phase, and identifies metastable striped states at low temperatures that lead to complex, slow coarsening dynamics due to free energy barriers, explaining apparent contradictions in prior dynamical observations.

ABSTRACT

The two-dimensional Ising model with competing short range ferromagnetic interactions and long range antiferromagnetic interactions is perhaps the most simple one containing the minimal microscopic ingredients necessary for an appropriate description of the macroscopic properties of ultrathin films and quasi--two--dimensional magnetic materials. Despite such relative simplicity, the frustration introduced by the competition between interactions generates complex behaviors that have eluded, up to now, a complete understanding of its general properties. In this work we review recent advances in the understanding of both equilibrium and non-equilibrium properties of the model. This includes a detailed description of several known properties of the thermodynamical phase diagram, as well as the existence of several types of metastable states and their influence in the low temperature dynamics.

Motivation & Objective

• To understand the thermodynamic phase diagram of the 2D Ising model with competing ferromagnetic and dipolar antiferromagnetic interactions.
• To resolve contradictions in low-temperature dynamical behavior reported in earlier studies.
• To investigate the role of metastable striped states in slowing down coarsening dynamics.
• To explore the implications of these dynamics for experimental systems like ultrathin magnetic films.
• To examine the possibility of a change in the order of the phase transition at large δ values.

Proposed method

• Analyzes the dimensionless Hamiltonian H = -δ∑<i,j>σiσj + ∑(i,j)σiσj/r³ij on a square lattice, with δ = J₀/Jd > 0.
• Uses mean-field and continuum approximations to study the phase diagram and ground state structure.
• Applies Monte Carlo simulations to probe equilibrium and non-equilibrium dynamics, including quenching protocols.
• Employs multiple dynamical analysis techniques to distinguish between logarithmic and algebraic domain growth behaviors.
• Introduces a metastability framework to explain diverging relaxation times in low-temperature regimes.
• Considers advanced sampling methods like Creutz cluster and multicanonical algorithms to mitigate finite-size and dynamical slowing-down effects.

Experimental results

Research questions

• RQ1 What is the nature of the ground state for the 2D Ising model with competing short-range ferromagnetic and long-range antiferromagnetic interactions?
• RQ2 How do metastable striped states influence the low-temperature coarsening dynamics of the system?
• RQ3 Why do different studies report conflicting dynamical behaviors—logarithmic vs. algebraic domain growth—under similar conditions?
• RQ4 Does the phase transition remain first-order for all δ > δₐ, or does it become continuous at large δ?
• RQ5 What is the role of defects in the high-temperature order-disorder transition in this model?

Key findings

• For δ > δₐ ≈ 0.425, the ground state is a striped phase with width h depending on δ, which is energetically favorable over ferromagnetic and checkerboard states.
• The system exhibits a metastability region at low temperatures where multiple metastable striped states of different widths coexist.
• Metastable states generate free energy barriers independent of domain size L, causing a transient slowing of coarsening dynamics that can mimic logarithmic growth for observation times shorter than the diverging relaxation time.
• Numerical cooling protocols from above the transition temperature support the metastability interpretation, showing delayed relaxation consistent with trapped dynamics.
• The phase transition is predicted to be first-order for all δ by continuum approximations, but Monte Carlo results suggest a possible change to continuous order at large δ, though the difference may be experimentally indistinguishable.
• The existence of multiple metastable phases at high δ and low T implies a complex dynamics with multiple characteristic time scales due to varying barrier heights.