Table of Contents
Fetching ...

Discrete Morse Theory for Khovanov Homology

Tuomas Kelomäki

Abstract

The standard methods for calculating Khovanov homology rely either on long exact/spectral sequences or on the algorithmic "divide and conquer" approach developed by Bar-Natan. In this paper, we employ an alternative and arguably simpler tool, discrete Morse theory, which is new in the context of knot homologies. The method is applied for 2- and 3-torus braids in Bar-Natan's dotted cobordism category, where Khovanov complexes of tangles live. This grants a recursive description of the complexes of 2- and 3-torus braids yielding an inductive result on integral Khovanov homology of links containing those braids. The result, accompanied with some computer data, advances the recent progress on a conjecture by Przytycki and Sazdanović which claims that closures of 3-braids only have 2-torsion in their Khovanov homology.

Discrete Morse Theory for Khovanov Homology

Abstract

The standard methods for calculating Khovanov homology rely either on long exact/spectral sequences or on the algorithmic "divide and conquer" approach developed by Bar-Natan. In this paper, we employ an alternative and arguably simpler tool, discrete Morse theory, which is new in the context of knot homologies. The method is applied for 2- and 3-torus braids in Bar-Natan's dotted cobordism category, where Khovanov complexes of tangles live. This grants a recursive description of the complexes of 2- and 3-torus braids yielding an inductive result on integral Khovanov homology of links containing those braids. The result, accompanied with some computer data, advances the recent progress on a conjecture by Przytycki and Sazdanović which claims that closures of 3-braids only have 2-torsion in their Khovanov homology.
Paper Structure (13 sections, 23 theorems, 91 equations, 17 figures, 2 tables)

This paper contains 13 sections, 23 theorems, 91 equations, 17 figures, 2 tables.

Key Result

Theorem 1.1

Let $L:{\mathbb {Z}}^2 \to \{\text{Links}\}$ be any function which creates a link $L(k,m)$ by uniformly gluing together a fixed tangle, a 3-torus braid with $k$ twists $(\sigma_1\sigma_2)^k$ and a 2-torus braid with $m$ twists $\sigma_1^m$; see Figure Link diagram for simplified Theorem. Then, Khova

Figures (17)

  • Figure 1: A link diagram $L(k,m)$ representing an example function $L$ for Theorem \ref{['simplified algorithm theorem for introduction']}. The boxes $(\sigma_1\sigma_2)^k$ and $\sigma_1^m$ exhibit the twisting of 3 and 2 strings for $k$ and $m$ times respectively, see Figures \ref{['T2 braid figure']} and \ref{['Braid diagram of T3']} for more specific illustrations.
  • Figure 2: A braid diagram for word $\sigma_1 \sigma_2 \sigma_2^{-1} \sigma_1 \sigma_2 \sigma_1^{-1}$ (left) and its braid closure (right).
  • Figure 3: Local relations of dotted cobordisms.
  • Figure 4: Khovanov complex of $\sigma_1^{-2}$ and one of its morphisms (gradings omitted). Contrary to our usual conventions, the braid $\sigma_1^{-2}$ is rotated 90 degrees clockwise so that the cobordism $f_4$ is easier to perceive.
  • Figure 5: Sign and smoothing conventions for crossings.
  • ...and 12 more figures

Theorems & Definitions (44)

  • Theorem 1.1
  • Conjecture 1.2: Przytycki, Sazdanović
  • Lemma 2.1
  • proof
  • Theorem 2.2: Discrete Morse theory
  • proof
  • Example 2.3
  • Proposition 2.4
  • proof
  • Lemma 3.1
  • ...and 34 more