Table of Contents
Fetching ...

Branes And Supergroups

Victor Mikhaylov, Edward Witten

TL;DR

This work develops a two-sided D3-NS5 brane system where theta-terms induce a three-dimensional Chern-Simons theory for supergroups such as U(m|n) and OSp(m|2n). By applying S-duality, and in some cases a subsequent T-duality, it derives magnetic dual descriptions and strong/weak coupling relations, and, after lifting, connects to Khovanov-type homologies for supergroups. The paper also extends these constructions to orthosymplectic cases with orientifolds, analyzes line and surface operators under dualities, and discusses symmetry breaking and lifts to D4-D6 and M-theory, outlining two analogs of Khovanov homology for supergroups. Collectively, these results provide a rich, duality-driven framework linking higher-dimensional gauge theories, supergroup Chern-Simons theory, and categorified knot invariants. The framework has potential implications for knot theory via supergroup invariants and for understanding gauge/brane dualities in contexts involving boundary defects and orientifolds.

Abstract

Extending previous work that involved D3-branes ending on a fivebrane with $θ_{\mathrm{YM}}\not=0$, we consider a similar two-sided problem. This construction, in case the fivebrane is of NS type, is associated to the three-dimensional Chern-Simons theory of a supergroup U$(m|n)$ or OSp$(m|2n)$ rather than an ordinary Lie group as in the one-sided case. By $S$-duality, we deduce a dual magnetic description of the supergroup Chern-Simons theory; a slightly different duality, in the orthosymplectic case, leads to a strong-weak coupling duality between certain supergroup Chern-Simons theories on $\mathbb{R}^3$; and a further $T$-duality leads to a version of Khovanov homology for supergroups. Some cases of these statements are known in the literature. We analyze how these dualities act on line and surface operators.

Branes And Supergroups

TL;DR

This work develops a two-sided D3-NS5 brane system where theta-terms induce a three-dimensional Chern-Simons theory for supergroups such as U(m|n) and OSp(m|2n). By applying S-duality, and in some cases a subsequent T-duality, it derives magnetic dual descriptions and strong/weak coupling relations, and, after lifting, connects to Khovanov-type homologies for supergroups. The paper also extends these constructions to orthosymplectic cases with orientifolds, analyzes line and surface operators under dualities, and discusses symmetry breaking and lifts to D4-D6 and M-theory, outlining two analogs of Khovanov homology for supergroups. Collectively, these results provide a rich, duality-driven framework linking higher-dimensional gauge theories, supergroup Chern-Simons theory, and categorified knot invariants. The framework has potential implications for knot theory via supergroup invariants and for understanding gauge/brane dualities in contexts involving boundary defects and orientifolds.

Abstract

Extending previous work that involved D3-branes ending on a fivebrane with , we consider a similar two-sided problem. This construction, in case the fivebrane is of NS type, is associated to the three-dimensional Chern-Simons theory of a supergroup U or OSp rather than an ordinary Lie group as in the one-sided case. By -duality, we deduce a dual magnetic description of the supergroup Chern-Simons theory; a slightly different duality, in the orthosymplectic case, leads to a strong-weak coupling duality between certain supergroup Chern-Simons theories on ; and a further -duality leads to a version of Khovanov homology for supergroups. Some cases of these statements are known in the literature. We analyze how these dualities act on line and surface operators.

Paper Structure

This paper contains 82 sections, 345 equations, 14 figures, 2 tables.

Figures (14)

  • Figure 1: An NS5-brane (sketched as a vertical dotted line) with $m$ D3-branes ending on it from the left and $n$ from the right -- sketched here for $m=3$, $n=4$. The D3-branes but not the NS5-brane extend in the $x_3$ direction, which is plotted horizontally, and the NS5-brane but not the D3-branes extend in the $x_{4,5,6}$ directions, which are represented symbolically by the vertical direction in this figure.
  • Figure 2: A "quiver" associated to a chain of $w$ NS5-branes, with $n_0,\dots,n_w$ D3-branes in the regions bounded by the NS5-branes; $w=4$ in the example sketched here.
  • Figure 3: Dynkin diagram for the $\mathfrak{su}(m|n)$ superalgebra. The subscripts are expressions for the roots in terms of the orthogonal basis $\delta_\bullet$, $\epsilon_\bullet$. The superscripts represent the Dynkin labels of a weight. The middle root denoted by a cross is fermionic.
  • Figure 4: Dynkin diagram for the $\mathfrak{osp}(2m+1|2n)$ superalgebra, $m\ge1$. The subscripts are expressions for the roots in terms of the orthogonal basis $\delta_\bullet$, $\epsilon_\bullet$. The superscripts represent the Dynkin labels of a weight. The arrows point in the direction of a shorter root. The middle root denoted by a cross is fermionic. Roots of the $\mathfrak{sp}(2n)$ and $\mathfrak{so}(2m+1)$ subalgebras are on the left and on the right of the fermionic root. The site shown in grey and labeled $a_n$ is the long simple root of the $\mathfrak{sp}(2n)$ subalgebra, which does not belong to the set of simple roots of the superalgebra.
  • Figure 5: Examples of dominant weights for $\mathfrak{u}(3|4)$. a. A typical weight. b. A weight of atypicality two, which is part of a block of atypical weights. The block is labeled by $\widetilde{x}_1$, $\widetilde{x}_2$, and $\widetilde{y}_1$, which correspond to a dominant weight of $\mathfrak{u}(1|2)$. The weights that make up this block are parametrized by $z_1$ and $z_2$, which can be thought of as labels of a maximally atypical weight of $\mathfrak{u}(2|2)$.
  • ...and 9 more figures