de Sitter-eating O-planes in supercritical string theory

TL;DR

This paper critically examines supercritical string theory models with de Sitter vacua, revealing that backreaction of O-planes causes significant geometric breakdowns, challenging the viability of such models.

Contribution

It provides a detailed analysis of backreaction effects in supercritical string models, highlighting issues with classical geometry and stability of dS vacua.

Findings

Supercritical dS vacua have large curvature and field strengths.

Backreaction of O-planes is large, causing geometric breakdown.

Classical descriptions are invalidated by backreaction effects.

Abstract

It has been proposed that flux compactifications of supercritical string theories (i.e., with spacetime dimension $D>10$) have dS vacua, with large $D$ acting as a control parameter for corrections to the classical spacetime effective action. In this paper, we provide a detailed analysis of the self-consistency of such models, focussing on $\alpha^\prime$ and backreaction corrections. We first show that all supercritical AdS, Minkowski and dS vacua in this setting have $\gtrsim \mathcal{O}(1)$ curvature and/or field strengths in the string frame. This may be in tension with suppressing $\alpha^\prime$ corrections unless the coefficients of the higher-derivative terms have a sufficiently strong large-$D$ suppression. We then argue that an additional and more severe problem arises in the dS case due to the backreaction of O-planes. In particular, we argue using a combination of geometric bounds and string-theory constraints that the O-plane backreaction is large in supercritical dS models. This implies that a large part of the naive classical geometry is eaten up by singular holes and thus indicates a breakdown of the classical description. Our finding resonates with several other recent results suggesting that string theory does not admit dS vacua in regimes where string and backreaction corrections are under control. As byproducts of our analysis, we derive a number of technical results that are useful beyond the specific applications in this paper. In particular, we compute the leading backreaction corrections to the smeared solution in a general flux compactification from $D$ to $d$ dimensions for an arbitrary distribution of O-planes and D-branes. We further argue for a general estimate for Green's functions on compact manifolds (and therefore for the backreaction corrections) in terms of their diameter, volume and dimension.