Models of inflation liberated by the curvaton hypothesis

TL;DR

This paper investigates how the curvaton hypothesis—where the primordial curvature perturbation is sourced by a field other than the inflaton—liberates inflation model-building. By decoupling the generation of perturbations from the inflaton dynamics, a wide range of slow-roll and thermal inflation scenarios become viable or easier to realize, including modular inflation and certain running-mass or brane-based models, and even thermal inflation with a modulus can generate cosmological perturbations. The authors show that many traditional constraints (e.g., on the inflaton potential steepness, loop corrections, and UV scales) are largely alleviated under the curvaton framework, while some models (notably D-term, simple chaotic, extended inflation, and trace-anomaly-inspired scenarios) remain constrained and do not benefit from liberation. They also analyze fast-roll inflation and thermal inflation, arguing that modular thermal inflation can produce the observed perturbations with a consistent cosmological history, whereas ordinary thermal inflation likely cannot. Overall, the curvaton liberation expands the inflationary model space, reduces fine-tuning, and offers novel pathways for constructing viable early-Universe scenarios with distinctive observational implications such as suppressed tensor modes and near-unity spectral index.

Abstract

It is usually supposed that inflation is of the slow-roll variety, and that the inflaton generates the primordial curvature perturbation. According to the curvaton hypothesis, inflation need not be slow-roll, and if it is the inflaton generates a negligible curvature perturbation. We find that the construction of slow-roll inflation models becomes much easier under this hypothesis. Also, thermal inflation followed by fast-roll becomes viable, with no slow-roll inflation at all.