G-Structures, Fluxes and Calibrations in M-Theory

TL;DR

The paper develops a general, geometry-first framework for supersymmetric warped M-theory backgrounds with flux on eight-manifolds, by allowing internal spinors of indefinite chirality and organizing the Killing spinor equations into a $G_2$-structure with controlled intrinsic torsion. The authors derive necessary and sufficient differential constraints on the $G_2$-structure data, reinterpret these conditions as generalized calibrations for dyonic M5-branes, and unravel how M5-branes wrapped on associative and SLAG three-cycles emerge in this setting (with explicit nonlinear PDEs from the Bianchi identities). Specializations to purely magnetic or purely electric flux produce AdS$_3$ and Spin(7) or weak $G_2$ holonomy structures, respectively, and the formalism recovers known solutions (e.g., dyonic M-branes, dielectric flows) while generating new examples. The work provides a unifying, covariant method to explore flux backgrounds, wrapped branes, and AdS$_3$ vacua in M-theory, with potential implications for holography and compactifications beyond special holonomy. It also clarifies the role of generalized calibrations in brane dynamics and offers a route to constructing novel geometries by combining $G$-structure data with flux backreaction and Bianchi identities.

Abstract

We study the most general supersymmetric warped M-theory backgrounds with non-trivial G-flux of the type R^{1,2} x M_8 and AdS_3 x M_8. We give a set of necessary and sufficient conditions for preservation of supersymmetry which are phrased in terms of G-structures and their intrinsic torsion. These equations may be interpreted as calibration conditions for a static ``dyonic'' M-brane, that is, an M5-brane with self-dual three-form turned on. When the electric flux is turned off we obtain the supersymmetry conditions and non-linear PDEs describing M5-branes wrapped on associative and special Lagrangian three-cycles in manifolds with G_2 and SU(3) structures, respectively. As an illustration of our formalism, we recover the 1/2-BPS dyonic M-brane, and also construct some new examples.