Table of Contents
Fetching ...

A dynamical approach to studying the Lee-Yang zeros for the Potts Model on the Cayley Tree

Diyath Pannipitiya, Roland Roeder

TL;DR

This work analyzes Lee-Yang zeros for the $q$-state Potts model on binary Cayley trees through a dynamical-systems lens. By constructing renormalization maps $R_{z,t,q}$ and $\hat R_{z,t,q}$ and employing the active/passive dichotomy, the authors precisely characterize the accumulation loci of zeros on the physical $z$-ray as the depth tends to infinity, yielding explicit thresholds $t_1(q)$, $t_2(q)$, and $t_3(q)$ and accumulation points $z_c(t,q)$ and $z_c^{\pm}(t,q)$ for ferromagnetic and antiferromagnetic regimes (with special cases for $q=2$). The analysis extends Ising results to Potts models (including unrooted trees) and provides a rigorous dynamical framework (Theorem D) linking zeros to iterates of renormalization maps, complemented by numerical plots and detailed lemmas. The methodology advances the mathematical understanding of phase-transition indicators in hierarchical lattices and offers a robust approach for studying Lee-Yang zeros via complex dynamics. Overall, the paper delivers explicit accumulation descriptions, rigorous proofs, and a unifying dynamical-system perspective for Potts-model zeros on Cayley trees.

Abstract

Let $Z_n(z,t)$ denote the partition function of the $q$-state Potts Model on the rooted binary Cayley tree of depth~$n$. Here, $z = {\rm e}^{-h/T}$ and $t = {\rm e}^{-J/T}$ with $h$ denoting an externally applied magnetic field, $T$ the temperature, and $J$ a coupling constant. One can interpret $z$ as a ``magnetic field-like'' variable and $t$ as a ``temperature-like'' variable. Physical values $h \in \mathbb{R}$, $T > 0$, and $J \in \mathbb{R}$ correspond to $t \in (0,\infty)$ and $z \in (0,\infty)$. For any fixed $t_0 \in (0,\infty)$ and fixed $n \in \mathbb{N}$ we consider the complex zeros of $Z_n(z,t_0)$ and how they accumulate on the ray $(0,\infty)$ of physical values for $z$ as $n \rightarrow \infty$. In the ferromagnetic case ($J > 0$ or equivalently $t \in (0,1)$) these Lee-Yang zeros accumulate to at most one point on $(0,\infty)$ which we describe using explicit formulae. In the antiferromagnetic case $(J < 0$ or equivalently $t \in (1,\infty)$) these Lee-Yang zeros accumulate to finitely many points of $(0,\infty)$, which we again describe with explicit formulae. The same results hold for the unrooted Cayley tree of branching number two. These results are proved by adapting a renormalization procedure that was previously used in the case of the Ising model on the Cayley Tree by Müller-Hartmann and Zittartz (1974 and 1977), Barata and Marchetti (1997), and Barata and Goldbaum (2001). We then use methods from complex dynamics and, more specifically, the active/passive dichotomy for iteration of a marked point, along with detailed analysis of the renormalization mappings, to prove the main results.

A dynamical approach to studying the Lee-Yang zeros for the Potts Model on the Cayley Tree

TL;DR

This work analyzes Lee-Yang zeros for the -state Potts model on binary Cayley trees through a dynamical-systems lens. By constructing renormalization maps and and employing the active/passive dichotomy, the authors precisely characterize the accumulation loci of zeros on the physical -ray as the depth tends to infinity, yielding explicit thresholds , , and and accumulation points and for ferromagnetic and antiferromagnetic regimes (with special cases for ). The analysis extends Ising results to Potts models (including unrooted trees) and provides a rigorous dynamical framework (Theorem D) linking zeros to iterates of renormalization maps, complemented by numerical plots and detailed lemmas. The methodology advances the mathematical understanding of phase-transition indicators in hierarchical lattices and offers a robust approach for studying Lee-Yang zeros via complex dynamics. Overall, the paper delivers explicit accumulation descriptions, rigorous proofs, and a unifying dynamical-system perspective for Potts-model zeros on Cayley trees.

Abstract

Let denote the partition function of the -state Potts Model on the rooted binary Cayley tree of depth~. Here, and with denoting an externally applied magnetic field, the temperature, and a coupling constant. One can interpret as a ``magnetic field-like'' variable and as a ``temperature-like'' variable. Physical values , , and correspond to and . For any fixed and fixed we consider the complex zeros of and how they accumulate on the ray of physical values for as . In the ferromagnetic case ( or equivalently ) these Lee-Yang zeros accumulate to at most one point on which we describe using explicit formulae. In the antiferromagnetic case or equivalently ) these Lee-Yang zeros accumulate to finitely many points of , which we again describe with explicit formulae. The same results hold for the unrooted Cayley tree of branching number two. These results are proved by adapting a renormalization procedure that was previously used in the case of the Ising model on the Cayley Tree by Müller-Hartmann and Zittartz (1974 and 1977), Barata and Marchetti (1997), and Barata and Goldbaum (2001). We then use methods from complex dynamics and, more specifically, the active/passive dichotomy for iteration of a marked point, along with detailed analysis of the renormalization mappings, to prove the main results.

Paper Structure

This paper contains 19 sections, 22 theorems, 95 equations, 15 figures.

Key Result

Theorem 1

For $t\in[0, 1]$, the complex zeros in $z$ of the partition function $Z(z, t)$ of the Ising model on any graph lie on the unit circle $\mathbb{T} = \{z\in \mathbb{C}:\ \mid z\mid = 1\}$.

Figures (15)

  • Figure 1: Left: rooted binary Cayley Tree $\Gamma_4$ of depth 4. Right: unrooted binary Cayley tree $\hat{\Gamma}_4$ of depth $4$.
  • Figure 2: The Lee-Yang zeros for the $5^{th}$ rooted binary Cayley Tree for the Ising model. Left: $t=0.0625<t_{\rm crit}=1/3$. Right: $t=0.5>t_{\rm crit}=1/3$.
  • Figure 3: The Lee-Yang zeros for the $5^{th}$ rooted binary Cayley Tree for the $3$-state Potts model. Left: $t=0.0625<t_{\rm crit}= \frac{1+\sqrt{73}}{36}\approx 0.265$. Right: $t=0.5>t_{\rm crit}= \frac{1+\sqrt{73}}{36}\approx 0.265$.
  • Figure 4: Graph of $z=z_c(t,3)$ (Theorem A). For $0 \leq t \leq 1/4$ we have $z_c(t,3) = 1$ (blue) and for $1/4 < t \leq t_2(q) \approx 0.2651$ we have $z_c(t,3) = \mathcal{Z}_q(t) < 1$ (red).
  • Figure 5: Graph of $z_c^-(t,3)$ and $z_c^+(t,3)$ ($Theorem\ B$).
  • ...and 10 more figures

Theorems & Definitions (52)

  • Theorem 1: Lee-Yang Circle Theorem
  • Remark 1
  • Remark 2
  • Remark 3
  • Remark 4
  • Remark 5
  • Definition 1
  • Remark 6
  • Lemma 1
  • proof
  • ...and 42 more