Critical Behavior of Coupled q-state Potts Models under Weak Disorder

P. Simon
DOI: https://doi.org/10.1209/epl/i1998-00200-6
1997-07-23
Abstract:We investigate the effect of weak disorder on different coupled $q$-state Potts models with $q\le 4$ using two loops renormalisation group. This study presents new examples of first order transitions driven by randomness. We found that weak disorder makes the models decouple. Therefore, it appears that no relations emerge, at a perturbation level, between the disordered $q_1\times q_2$-state Potts model and the two disordered $q_1$, $q_2$-state Potts models ($q_1\ne q_2$), despite their central charges are similar according to recent numerical investigations. Nevertheless, when two $q$-state Potts models are considered ($q>2$), the system remains always driven in a strong coupling regime, violating apparently the Imry-Wortis argument.
Disordered Systems and Neural Networks
What problem does this paper attempt to address?
This paper aims to explore the influence of weak disorder on the critical behavior of \(q\)-state Potts models (\(q\leq4\)) with different couplings. Specifically, the author uses the two - loop renormalization group method to study this problem. ### Main problems: 1. **Influence of weak disorder on coupled \(q\)-state Potts models**: The author attempts to understand how weak disorder changes the critical behavior of these models, especially whether they will lead to the transition from first - order phase transition to second - order phase transition. 2. **Relationship between different \(q\)-state Potts models**: The author hopes to explore the relationship between the disordered \(q_1\times q_2\)-state Potts model and two independent disordered \(q_1\) and \(q_2\)-state Potts models. ### Research background: - **Harris criterion**: Provides a method for judging whether disorder will change the universality class of phase transitions. - **Cardy's work**: Shows that after introducing weak disorder, \(N\) coupled Ising models will decouple into \(N\) independent Ising models. - **Pujol's research**: Extends to \(N\) 3 - state Potts models and discovers non - Ising - like second - order phase transitions (for \(N > 2\)). - **Numerical results**: Some numerical studies show that the 8 - state Potts model exhibits continuous phase transitions after introducing disorder, but the magnetic exponent is consistent with that of the pure Ising model; while other studies show that the magnetic exponent changes continuously with \(q\). ### Main findings: 1. **Model decoupling**: When considering two different Potts models, weak disorder makes the models decouple, that is, there is no obvious correlation (at least at the perturbative level). 2. **Strong - coupling regime**: When \(q_1 = q_2=3\) or \(4\), the system still enters the strong - coupling regime, which seems to violate the Imry - Wortis argument. 3. **Fixed - point analysis**: Multiple fixed points are found through the renormalization group equations, and their stabilities are analyzed. It is found that some fixed points are unstable and the system is finally driven into the strong - coupling regime. ### Conclusions: - Weak disorder can significantly change the critical behavior of coupled \(q\)-state Potts models, leading to model decoupling. - For the case of \(q_1 = q_2 = 3\) or \(4\), although there is disorder, the system still enters the strong - coupling regime, which contradicts the Imry - Wortis argument. - Numerical studies and further non - perturbative analyses are very important for verifying these results. Through the above research, the author shows the influence of weak disorder on coupled Potts models and proposes directions for further research, especially in numerical simulation and non - perturbative theory.