The Nahm transform of multi-fractional instantons

TL;DR

This paper studies the Nahm transform of multi-fractional instantons in $SU(N)$ gauge theories on tori, revealing how these configurations map to dual instantons with related topological charges and exploring their properties under various torus period conditions.

Contribution

It explicitly constructs the Nahm transform for fractional instantons on tori, extending duality to detuned periods, and connects these findings to string theory and D-brane models.

Findings

Fractional instantons map to dual fractional instantons with related charges.

Explicit construction of Nahm transform for constant field strength fractional instantons.

Duality extends to detuned torus periods, linking solutions with different detuning parameters.

Abstract

We embed the multi-fractional instantons of $SU(N)$ gauge theories on $\mathbb T^4$ with 't Hooft twisted boundary conditions into $U(N)$ bundles and use the Nahm transform to study the corresponding configurations on the dual $\widehat{\mathbb T}^4$. We first show that $SU(N)$ fractional instantons of topological charge $Q={r \over N}$, $r \in \{1, 2,...,N-1\}$, are mapped to fractional instantons of $SU(\widehat N)$ of charge $\widehat Q = {r \over \widehat N}$, where $\widehat N = N q_1 q_3 - r q_3 + q_1$ and $q_{1,3}$ are integer-quantized $U(1)$ fluxes. We then explicitly construct the Nahm transform of constant field strength fractional instantons of $SU(N)$ and find the $SU(\widehat N)$ configurations they map to. Both the $\mathbb T^4$ instantons and their $\widehat {\mathbb T}^4$ images are self-dual for appropriately tuned torus periods. The Nahm duality can be extended to tori with detuned periods, with detuning parameter $\Delta$, mapping solutions with $\Delta >0$ on $\mathbb T^4$ to ones with $\widehat\Delta <0$ on $\widehat{\mathbb T}^4$. We also recall that fractional instantons appear in string theory precisely via the $U(N)$ embedding, suggesting that studying the end point of tachyon condensation for $\Delta \ne 0$ is needed -- and is perhaps feasible in a small-$\Delta$ expansion, as in field theory studies -- in order to understand the appearance and role of fractional instantons in $D$-brane constructions.