Brane-Antibrane Backreaction in Axion Monodromy Inflation

TL;DR

The paper addresses the stability of axion monodromy inflation models that involve NS5/anti-NS5 branes wrapped on homologous 2-cycles in distant warped throats. It uses worldsheet calculations to derive brane–antibrane interaction potentials, showing a dominant logarithmic interaction for 5-branes on homologous cycles (V ∝ log(R/Y)) alongside a 3-brane term (V ∝ 1/Y^4) that is subdominant in practice. The key result is that the 5-brane backreaction does not decouple at large separations, implying significant, nonlocal backreaction on the geometry and challenging throat-localized stabilization. Consequently, the bulk/string-scale physics sets the interaction energy, undermining the viability of large-field inflation models that rely on well-separated warped throats and tadpole cancellation within the throat.

Abstract

We calculate the interaction potential between D5 and anti-D5 branes wrapping distant but homologous 2-cycles. The interaction potential is logarithmic in the separation radius and does not decouple at infinity. We show that logarithmic backreaction is generic for 5-branes wrapping distant but homologous 2-cycles, and we argue that this destabilises models of axion monodromy inflation involving NS5 brane-antibrane pairs in separate warped throats towards an uncontrolled region.

Figures (2)

• Figure 1: The schematic brane configuration required for models of axion monodromy inflation. There is one NS5 brane and one $\overline{\hbox{NS}}5$ brane down separate warped throats, wrapping the homologous 2-cycle $\Sigma$. Inflation is driven by the vev $c \equiv \int_{\Sigma} C_2$ of a RR 2-form on this cycle.
• Figure 2: The $T^6/\mathbb{Z}_4$ orbifold.