Setting Initial Conditions for Inflation with Reaction-Diffusion Equation

TL;DR

This paper proposes using reaction-diffusion equation solutions to set initial conditions for inflation, showing inflation can start from small localized domains without fine-tuning over large regions.

Contribution

It introduces a novel approach using reaction-diffusion fronts to initiate inflation from small localized field domains, reducing the need for fine-tuned initial conditions.

Findings

Inflation can begin from a single small localized domain.

Reaction-diffusion fronts propagate slowly, allowing the field to change gradually.

Inflation occurs even when initial conditions are smaller than the Hubble size.

Abstract

We discuss the issue of setting appropriate initial conditions for inflation. Specifically, we consider natural inflation model and discuss the fine tuning required for setting almost homogeneous initial conditions over a region of order several times the Hubble size which is orders of magnitude larger than any relevant correlation length for field fluctuations. We then propose to use the special propagating front solutions of reaction-diffusion equations for localized field domains of smaller sizes. Due to very small velocities of these propagating fronts we find that the inflaton field in such a ${\it field domain}$ changes very slowly, contrary to naive expectation of rapid roll down to the true vacuum. Continued expansion leads to the energy density in the Hubble region being dominated by the vacuum energy, thereby beginning the inflationary phase. Our results show that inflation can occur even with a single localized field domain of size smaller than the Hubble size. We discuss possible extensions of our results for different inflationary models, as well as various limitations of our analysis (e.g. neglecting self gravity of the localized field domain).