Table of Contents
Fetching ...

On the Lee-Yang property of some ferromagnets

Yuri Kozitsky

TL;DR

The paper addresses whether the Lee-Yang property for the full partition function $Z_N(h)$ can be induced by ferromagnetic couplings even when the single-spin law $Z_1(h)$ lacks it. It leverages the Lieb-Sokal Laguerre-entire function framework and analyzes the two-spin transform $F_{2N}$ built from a nonnegative interaction matrix $K$ and a Laguerre-type single-spin factor $\phi$, deriving explicit bounds on $\theta$ in terms of the minimum coupling $\varkappa$ to guarantee imaginary-axis zeros. The main result provides a theorem: under Assumption 1 with nonnegative $K$ and bounds $\theta \le \sqrt{\frac{e^{\varkappa}+e^{-\varkappa}}{2}}$ (and a stronger version with an additional condition) the function $F_{2N}$ is Laguerre and has imaginary zeros only; the particular regime $\theta \le 1$ yields Lee-Yang for the Blume-Capel and annealed diluted Ising models in specified parameter ranges. Corollaries show concrete parameter regimes for these models (e.g., $\Delta<\varkappa/2$ for Blume-Capel and thinning bounds for diluted Ising), and connect the results to standard lattices and Dyson hierarchical structures, thereby extending Lee-Yang-type behavior beyond single-spin measures.

Abstract

According to the Lieb-Sokal theorem, the partition function, $Z$, of a ferromagnetic spin model has the Lee-Yang property if the single-spin partition function has it. In this note, it is shown that for some spin models a ferromagnetic interaction can induce the Lee-Yang property of $Z$ even if the single-spin partition function fails to have it. In particular, this holds for the Blume-Capel model and for the annealed states of the $s=\pm 1$ site dilute Ising model with a neares-neighbor interaction on $\mathds{Z}^d$, as well as with interactions defined by a hierarchical structure similar to that of Dyson's hierarchical model.

On the Lee-Yang property of some ferromagnets

TL;DR

The paper addresses whether the Lee-Yang property for the full partition function can be induced by ferromagnetic couplings even when the single-spin law lacks it. It leverages the Lieb-Sokal Laguerre-entire function framework and analyzes the two-spin transform built from a nonnegative interaction matrix and a Laguerre-type single-spin factor , deriving explicit bounds on in terms of the minimum coupling to guarantee imaginary-axis zeros. The main result provides a theorem: under Assumption 1 with nonnegative and bounds (and a stronger version with an additional condition) the function is Laguerre and has imaginary zeros only; the particular regime yields Lee-Yang for the Blume-Capel and annealed diluted Ising models in specified parameter ranges. Corollaries show concrete parameter regimes for these models (e.g., for Blume-Capel and thinning bounds for diluted Ising), and connect the results to standard lattices and Dyson hierarchical structures, thereby extending Lee-Yang-type behavior beyond single-spin measures.

Abstract

According to the Lieb-Sokal theorem, the partition function, , of a ferromagnetic spin model has the Lee-Yang property if the single-spin partition function has it. In this note, it is shown that for some spin models a ferromagnetic interaction can induce the Lee-Yang property of even if the single-spin partition function fails to have it. In particular, this holds for the Blume-Capel model and for the annealed states of the site dilute Ising model with a neares-neighbor interaction on , as well as with interactions defined by a hierarchical structure similar to that of Dyson's hierarchical model.

Paper Structure

This paper contains 5 sections, 5 theorems, 43 equations.

Key Result

Proposition 1.1

LiebS Let $G$ be an exponential type entire function of $(x_1, \dots , x_N)\in \mathds{C}^N$ and $F$ be defined by the formula If $G(x_1 , \dots, x_N) \neq 0$ whenever $\Re x_i >0$ and $K_{ij}\geq 0$ for all $i,j=1, \dots N$, then $F(x_1 , \dots, x_N) \neq 0$ whenever all $\Re x_i >0$.

Theorems & Definitions (6)

  • Proposition 1.1
  • Proposition 1.2
  • Theorem 2.2
  • proof
  • Corollary 2.3
  • Corollary 2.4