Synthetic Gravitational Waves from a Rolling Axion Monodromy

TL;DR

This paper proposes a string theory-inspired model where a spectator axion's fast roll during inflation amplifies gauge fields, producing observable, parity-breaking gravitational waves at CMB and interferometer scales, with detectable non-Gaussianity.

Contribution

It introduces a novel mechanism linking axion dynamics to localized gravitational wave production, compatible with current observational constraints.

Findings

Generated observable GWs at CMB scales with parity violation.

Produced large GW signals at interferometer scales without violating scalar perturbation bounds.

Ensured model consistency through back-reaction and perturbativity constraints.

Abstract

In string theory inspired models of axion-like fields, sub-leading non-perturbative effects, if sufficiently large, can introduce steep cliffs and gentle plateaus onto the underlying scalar potential. During inflation, the motion of a spectator axion $\sigma$ on this potential becomes temporarily fast, leading to localized amplification of one helicity state of gauge fields. In this model, the tensor and scalar correlators sourced by the vector fields exhibit localized peak(s) in momentum space corresponding to the modes that exit the horizon while the roll of $\sigma$ is fast. Thanks to the gravitational coupling of gauge fields with the visible sector and the localized nature of particle production, this model can generate observable gravitational waves (GWs) at CMB scales while satisfying the current limits on scalar perturbations. The resulting GW signal breaks parity and exhibit sizeable non-Gaussianity that can be probed by future CMB B-mode missions. Depending on the initial conditions and model parameters, the roll of the spectator axion can also generate an observably large GW signature at interferometer scales while respecting the bounds on the scalar fluctuations from primordial black hole limits. In our analysis, we carefully investigate bounds on the model parameters that arise through back-reaction and perturbativity considerations to show that these limits are satisfied by the implementations of the model that generate GW signals at CMB and sub-CMB scales.