Systematics of multi-field effects at the end of warped brane inflation

TL;DR

This work investigates whether additional light scalar fields at the end of warped brane inflation can imprint significant curvature perturbations on CMB scales via the Lyth mechanism. It develops criteria for when end-of-inflation multi-field contributions are viable, and analyzes uplift schemes that control the end-of-inflation dynamics, applying them to a detailed D3-brane inflation model in a warped KS throat. The explicit case with a Kuperstein D7 embedding shows that, although inflation can persist toward the throat tip, the residual isometry directions become degenerate along the angular-stable trajectory, suppressing the Lyth end perturbation in this setup; however, other embeddings or throats with non-degenerate residual isometries could yield observable end-of-inflation effects. The study highlights the sensitivity of CMB predictions to infrared throat geometry and uplift choices, guiding future model-building to discriminate among multi-field end-of-inflation scenarios.

Abstract

We investigate in the context of brane inflation the possibility of additional light scalar fields generating significant power spectrum and non-Gaussianities at the end of inflation affecting the CMB scale observations. We consider the specific mechanism outlined by Lyth and describe the necessary criteria for it to be potentially important in a warped throat. We also discuss different mechanisms for uplifting the vacuum energy which can lead to different dominant contributions of the inflaton potential near the end of inflation. We then apply such criteria to one of the most detailed brane inflation models to date, and show that inflation can persist towards the tip of the throat, however for the specific stable inflationary trajectory, the light residual isometry direction becomes degenerate. We also estimate the effects for other inflationary trajectories with non-degenerate residual isometries.

Figures (2)

• Figure 1: (Left) the inflaton potential (\ref{['MathbbVphi']}) and (right) the slow-roll parameter $\varepsilon$ (\ref{['epsilonSR2']}). We normalize $M_{\rm Pl} = 1$ and for simplicity we set $A_0 = 1$. The point $r/\epsilon^{2/3} = 1$ denotes the tip from which there is no further radial displacement. As shown in the right panel, the potential is very flat near the tip.
• Figure 2: (Upper left) the inflaton potential (\ref{['delicatepotential']}) and the resulting slow-roll parameters, (upper right) $\varepsilon$ and (lower left) $\eta$. We show both cases where the Coulombic piece proportional to $r^{-4}$ is present (solid line) and absent (dotted line). As can be seen, without the Coulombic term inflation proceeds deep inside the throat, i.e. very small $r$ region, but (\ref{['delicatepotential']}) is no more valid there. In the lower right panel, we show $-\dot{H}/H^2$ which is exactly equivalent to the acceleration of the scale factor: $-\dot{H}/H^2 < 1$ means acceleration. Clearly, the criteria $|\eta| = 1$does not guarantee that inflation ends at the corresponding point, especially when the Coulombic term is negligible which is the scenario we discussed in the main text.