Scalar perturbations during multiple-field slow-roll inflation
S. Groot Nibbelink, B. J. W. van Tent
TL;DR
The paper addresses scalar perturbations during slow-roll inflation with multiple real fields on a curved field manifold. It develops a covariant, field-trajectory–driven framework that uses a background-dynamics–induced basis and generalized slow-roll functions to analyze perturbations in three regimes around horizon crossing, deriving first-order analytic solutions for the coupled perturbations and a nontrivial particular solution that sources the adiabatic mode. The authors provide explicit expressions for adiabatic, isocurvature, and cross-correlators of the gravitational potential at recombination, and validate the theory with a quadratic-potential two-field example showing that multi-field effects can be substantial when ${\tilde{\eta}}^{\perp}$ is large. This framework is broadly applicable to high-energy theories with many scalar fields and non-minimal kinetic terms, enabling robust predictions for CMB and large-scale structure and clarifying when multi-field effects matter.
Abstract
We calculate the scalar gravitational and matter perturbations in the context of slow-roll inflation with multiple scalar fields, that take values on a (curved) manifold, to first order in slow roll. For that purpose a basis for these perturbations determined by the background dynamics is introduced and multiple-field slow-roll functions are defined. To obtain analytic solutions to first order, the scalar perturbation modes have to be treated in three different regimes. Matching is performed by analytically identifying leading order asymptotic expansions in different regions. Possible sources for multiple-field effects in the gravitational potential are the particular solution caused by the coupling to the field perturbation perpendicular to the field velocity, and the rotation of the basis. The former can contribute even to leading order if the corresponding multiple-field slow-roll function is sizable during the last 60 e-folds. Making some simplifying assumptions, the evolution of adiabatic and isocurvature perturbations after inflation is discussed. The analytical results are illustrated and checked numerically with the example of a quadratic potential.
