Post-inflationary behavior of adiabatic perturbations and tensor-to-scalar ratio

TL;DR

This paper investigates whether post-inflationary dynamics can boost the tensor-to-scalar ratio $r$ by damping adiabatic perturbations, using the $\delta N$ formalism and a comparison of isentropic fluids and multiple scalar fields. It shows that adiabatic perturbations are robust on superhorizon scales and cannot be generically damped by late-time entropy production or cancelled by curvaton-like isocurvature modes. Consequently, the final curvature perturbation tracks the initial inflationary perturbations, and significant suppression of scalar modes through post-inflationary processes is unlikely. Nonetheless, sizable tensor amplitudes remain viable in simple chaotic inflation frameworks, including supergravity realizations with shift symmetry and multi-field axion models, while many string-inspired models predict very small $r$ unless special mechanisms apply; thus detection of tensor modes would strongly favor simple chaotic inflation, guiding expectations for future CMB polarization experiments.

Abstract

We explain why it is so difficult and perhaps even impossible to increase the cosmological tensor-to-scalar perturbation ratio during the post-inflationary evolution of the universe. Nevertheless, contrary to some recent claims, tensor perturbations can be relatively large in the simplest inflationary models which do not violate any rules of modern quantum field theory.