Global symmetries, volume independence and continuity

TL;DR

This paper demonstrates that certain quantum field theories with global symmetries exhibit volume independence at large N due to twisted boundary conditions, and establishes adiabatic continuity between small and large compactification radii.

Contribution

It proves volume independence for $CP(N-1)$ and $O(N)$ models with twisted boundary conditions and clarifies the absence of phase transitions, supporting the concept of adiabatic continuity.

Findings

Twisted boundary conditions project out excited states at large N.

The $CP(N-1)$ model shows no phase transition at finite $L$.

Quantum kink-instantons govern the $ heta$ dependence at small $L$.

Abstract

We discuss quantum field theories with global $SU(N)$ and $O(N)$ symmetries for which the temporal direction is compactified on a circle of size $L$ with periodicity of fields up to a global symmetry transformation, i.e. twisted boundary conditions. Such boundary conditions correspond to an insertion of the global symmetry operator in the partition function. We argue that for a special choice of twists most of the excited states get projected out, leaving only either mesonic states or states whose energy scales with $N$. When $N\rightarrow \infty$ all excitations become suppressed at any compact radius and the twisted partition function gets a contribution from the ground-state only, rendering observables independent of the radius of compactification, i.e. volume independent. We explicitly prove that this is indeed the case for the $CP(N-1)$ and $O(N)$ non-linear sigma models in any number of dimensions. We further focus on the two-dimensional $CP(N-1)$ case which is asymptotically free, and demonstrate, unlike its thermal counterpart, the twisted theory has commuting $N\rightarrow\infty,L\rightarrow\infty$ limits and does not undergo a second-order phase transition at "zero-temperature" discussed by Affleck long ago. At finite $L$ the theory is described by an effective, zero-temperature quantum mechanics with smoothly varying parameters depending on $L$, eliminating the possibility of a phase transition at any $L$, which was conjectured by \"Unsal and Dunne. As $L$ is decreased at fixed and finite $N$ the relevant objects dictating the $\theta$ dependence are quantum kink-instantons, avatars of the small $L$ regime fractional instantons. These considerations, for the first time establishes the idea of adiabatic continuity advocated by \"Unsal et. al.