An ideal toy model for confining, walking and conformal gauge theories: the O(3) sigma model with theta-term

TL;DR

This paper introduces a two-dimensional O(3) sigma model with a theta-term as a toy model for four-dimensional gauge theories, capturing confining, walking, and conformal behaviors, and investigates the relevance of the theta parameter non-perturbatively.

Contribution

It demonstrates that the theta parameter is a relevant coupling in the toy model by analyzing topological charge correlations and establishing finiteness after renormalization.

Findings

The model is asymptotically free for all theta.

At small theta, the model exhibits confinement; at theta=pi, it has a non-trivial IR fixed point.

Divergences in topological susceptibility can be renormalized, making the theory finite.

Abstract

A toy model is proposed for four dimensional non-abelian gauge theories coupled to a large number of fermionic degrees of freedom. As the number of flavors is varied the gauge theory may be confining, walking or conformal. The toy model mimicking this feature is the two dimensional O(3) sigma model with a theta-term. For all theta the model is asymptotically free. For small theta the model is confining in the infra red, for theta = pi the model has a non-trivial infra red fixed point and consequently for theta slightly below pi the coupling walks. The first step in investigating the notoriously difficult systematic effects of the gauge theory in the toy model is to establish non-perturbatively that the theta parameter is actually a relevant coupling. This is done by showing that there exist quantities that are entirely given by the total topological charge and are well defined in the continuum limit and are non-zero, despite the fact that the topological susceptibility is divergent. More precisely it is established that the differences of connected correlation functions of the topological charge (the cumulants) are finite and non-zero and consequently there is only a single divergent parameter in Z(theta) but otherwise it is finite. This divergent constant can be removed by an appropriate counter term rendering the theory completely finite even at theta > 0.