Dynamical critical behavior of the two-dimensional three-state Potts model


Creative Commons License

VATANSEVER E., Barkema G. T., Fytas N. G.

Physical Review E, vol.110, no.1, 2024 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 110 Issue: 1
  • Publication Date: 2024
  • Doi Number: 10.1103/physreve.110.014135
  • Journal Name: Physical Review E
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Chemical Abstracts Core, Compendex, INSPEC, MEDLINE, zbMATH
  • Dokuz Eylül University Affiliated: Yes

Abstract

We investigate the dynamical critical behavior of the two-dimensional three-state Potts model with single spin-flip dynamics in equilibrium. We focus on the mean-squared deviation of the magnetization M (MSDM) as a function of time, as well as on the autocorrelation function of M. Our simulations reveal the existence of two crossover behaviors at times τ1∼Lz1 and τ2∼Lz2, separating three dynamical regimes. MSDM appears to shift from ordinary diffusion in the first regime, to anomalous diffusion in the second, and finally to be constant in the third regime. The magnetization autocorrelation function, on the other hand, is found to fluctuate between exponential decay, stretched-exponential decay, and then again exponential decay along these three regimes. This behavior is in agreement with the one reported recently for the two-dimensional Ising ferromagnet [Phys. Rev. E 108, 034118 (2023)2470-004510.1103/PhysRevE.108.034118], indicating that the existence of two dynamic critical exponents is not a peculiarity of the Ising model itself. A comparison of both MSDM and the magnetization's autocorrelation function suggests that within our numerical accuracy the exponents z1 and z2 are shared between the Ising and three-state Potts models at least for the particular case of single spin-flip dynamics studied here, even though their equilibrium universality classes are clearly distinct. Continuity of MSDM requires that α(z2-z1)=γ/ν-z1, in which α is the anomalous exponent in the intermediate regime. Since the ratio γ/ν is not shared between the two models, it follows that α is not shared either, an aspect well verified in our simulations. Finally, we also discuss the relevance of our main findings using another useful observable, namely the line magnetization Ml.