Table of Contents
Fetching ...

Extended Holomorphic Anomaly and Loop Amplitudes in Open Topological String

Johannes Walcher

TL;DR

The paper extends BCOV's holomorphic anomaly to open topological strings by incorporating D-brane boundary data through normal functions and Griffiths' infinitesimal invariant, establishing extended anomaly equations that include boundary degenerations. It develops a B-model framework combining tt* geometry, normal functions, and disk/annulus contributions, and demonstrates the formalism on the real quintic by solving low-genus open+closed amplitudes and fixing holomorphic ambiguities via large-volume, orbifold, and conifold expansions. The work also connects disk two-point data to the infinitesimal invariant and discusses open-string integrality conjectures, offering a systematic method to compute open+closed amplitudes on compact Calabi-Yau manifolds with D-branes. Together, these results illuminate the role of D-branes as boundary data in the vacuum bundle and pave the way for a full open-sector extension of topological string theory.

Abstract

Open topological string amplitudes on compact Calabi-Yau threefolds are shown to satisfy an extension of the holomorphic anomaly equation of Bershadsky, Cecotti, Ooguri and Vafa. The total topological charge of the D-brane configuration must vanish in order to satisfy tadpole cancellation. The boundary state of such D-branes is holomorphically captured by a Hodge theoretic normal function. Its Griffiths' infinitesimal invariant is the analogue of the closed string Yukawa coupling and plays the role of the terminator in a Feynman diagram expansion for the topological string with D-branes. The holomorphic anomaly equation is solved and the holomorphic ambiguity is fixed for some representative worldsheets of low genus and with few boundaries on the real quintic.

Extended Holomorphic Anomaly and Loop Amplitudes in Open Topological String

TL;DR

The paper extends BCOV's holomorphic anomaly to open topological strings by incorporating D-brane boundary data through normal functions and Griffiths' infinitesimal invariant, establishing extended anomaly equations that include boundary degenerations. It develops a B-model framework combining tt* geometry, normal functions, and disk/annulus contributions, and demonstrates the formalism on the real quintic by solving low-genus open+closed amplitudes and fixing holomorphic ambiguities via large-volume, orbifold, and conifold expansions. The work also connects disk two-point data to the infinitesimal invariant and discusses open-string integrality conjectures, offering a systematic method to compute open+closed amplitudes on compact Calabi-Yau manifolds with D-branes. Together, these results illuminate the role of D-branes as boundary data in the vacuum bundle and pave the way for a full open-sector extension of topological string theory.

Abstract

Open topological string amplitudes on compact Calabi-Yau threefolds are shown to satisfy an extension of the holomorphic anomaly equation of Bershadsky, Cecotti, Ooguri and Vafa. The total topological charge of the D-brane configuration must vanish in order to satisfy tadpole cancellation. The boundary state of such D-branes is holomorphically captured by a Hodge theoretic normal function. Its Griffiths' infinitesimal invariant is the analogue of the closed string Yukawa coupling and plays the role of the terminator in a Feynman diagram expansion for the topological string with D-branes. The holomorphic anomaly equation is solved and the holomorphic ambiguity is fixed for some representative worldsheets of low genus and with few boundaries on the real quintic.

Paper Structure

This paper contains 20 sections, 153 equations, 6 figures, 1 table.

Figures (6)

  • Figure 1: Left: The disk amplitude with $n+1$ boundary insertions captures the higher $A_\infty$ products \ref{['mn']}. Right: The disk amplitude with one bulk and one boundary insertion encodes the bulk to boundary obstruction map \ref{['obmap']}, and will also define the infinitesimal Abel-Jacobi map \ref{['infaj']}.
  • Figure 2: The two degenerations that contribute to the RHS of the holomorphic anomaly for closed strings.
  • Figure 3: The path-integral on the disk with a long strip attached defines a canonical open string vacuum.
  • Figure 4: The disk amplitude with two bulk insertions.
  • Figure 5: When adding boundaries, the two degenerations from Fig. \ref{['degenbcov']} remain unaffected.
  • ...and 1 more figures