The zeros of the partition function of the pinning model

TL;DR

This paper investigates the complex zeros of the pinning model's partition function, revealing their asymptotic distribution on a curve and analyzing their behavior near the critical point, with implications for Griffiths singularities.

Contribution

It provides a detailed analysis of the zeros of the partition function for specific inter-arrival laws, combining probabilistic and analytical techniques to understand their distribution and behavior.

Findings

Zeros asymptotically fill a closed curve touching the real axis at one point

Detailed behavior of zeros near the critical point is characterized

Results have implications for Griffiths singularities in disordered systems

Abstract

We aim at understanding for which (complex) values of the potential the pinning partition function vanishes. The pinning model is a Gibbs measure based on discrete renewal processes with power law inter-arrival distributions. We obtain some results for rather general inter-arrival laws, but we achieve a substantially more complete understanding for a specific one parameter family of inter-arrivals. We show, for such a specific family, that the zeros asymptotically lie on (and densely fill) a closed curve that, unsurprisingly, touches the real axis only in one point (the critical point of the model). We also perform a sharper analysis of the zeros close to the critical point and we exploit this analysis to approach the challenging problem of Griffiths singularities for the disordered pinning model. The techniques we exploit are both probabilistic and analytical. Regarding the first, a central role is played by limit theorems for heavy tail random variables. As for the second, potential theory and singularity analysis of generating functions, along with their interplay, will be at the heart of several of our arguments.