D-Branes in Landau-Ginzburg Models and Algebraic Geometry

TL;DR

This paper confirms Kontsevich's proposal for the category of B-branes in massive Landau-Ginzburg models, providing a complete algebraic description and implications for mirror symmetry and Fukaya categories.

Contribution

It verifies Kontsevich's candidate category for B-branes in Landau-Ginzburg models and offers a complete algebraic framework using Clifford algebra modules.

Findings

Confirmed Kontsevich's proposal for B-branes category

Described B-branes using Clifford algebra modules

Enabled computation of Fukaya categories for Fano varieties

Abstract

We study topological D-branes of type B in N=2 Landau-Ginzburg models, focusing on the case where all vacua have a mass gap. In general, tree-level topological string theory in the presence of topological D-branes is described mathematically in terms of a triangulated category. For example, it has been argued that B-branes for an N=2 sigma-model with a Calabi-Yau target space are described by the derived category of coherent sheaves on this space. M. Kontsevich previously proposed a candidate category for B-branes in N=2 Landau-Ginzburg models, and our computations confirm this proposal. We also give a heuristic physical derivation of the proposal. Assuming its validity, we can completely describe the category of B-branes in an arbitrary massive Landau-Ginzburg model in terms of modules over a Clifford algebra. Assuming in addition Homological Mirror Symmetry, our results enable one to compute the Fukaya category for a large class of Fano varieties. We also provide a (somewhat trivial) counter-example to the hypothesis that given a closed string background there is a unique set of D-branes consistent with it.