Rouquier complexes are functorial over braid cobordisms (Q607281): Difference between revisions

This paper gives an explicit functor between the categories of braid cobordisms and Soergel bimodules complexes. In [J. Inst. Math. Jussieu 6, No. 3, 501--525 (2007; Zbl 1192.20004)], \textit{W. Soergel} considered a category of bimodules whose Grothendieck ring is isomorphic to the Hecke algebra. \textit{R. Rouquier} used these bimodules [in: de la Peña, José A. (ed.) et al., Trends in representation theory of algebras and related topics. Workshop on representations of algebras and related topics, Querétaro, México, August 11--14, 2004. Providence, RI: American Mathematical Society (AMS). Contemporary Mathematics 406, 137--167 (2006; Zbl 1162.20301)] to categorify the braid groups. To a braid \(\beta\), he associates, up to homotopy, a complex \(C_\beta\) of Soergel bimodules. The complex \(C_\beta\) proves to be closely related to the Khovanov-Rozansky homology of the closed braid \(\hat{\beta}\). The question whether Rouquier's construction lifts to the braid cobordism category is discussed in [\textit{M. Khovanov} and \textit{R. Thomas}, Homology Homotopy Appl. (2) 9, 19--94 (2007; Zbl 1119.18008)] but in a theoretical way, using strongly the fact that \(\Hom(C_\beta,C_\beta)\) is one dimensional. In [Diagrammatics for Soergel categories, \url{arXiv:0902.4700}], the first author and \textit{M. Khovanov} gave a diagrammatic presentation of the Soergel category which makes computations much more practicable. In this paper, the authors use this approach to define explicitly a functor from the braid cobordism category to the homotopy category of Soergel bimodules complexes. It deals with coefficients in a field but some hints are given for \(\mathbb Z\)-ones. The paper adresses the reader interested in explicit cobordism maps. The paper is organized as follows. The second part recalls the required elements of the Soergel category, both original and diagrammatic, Rouquier complexes and braid cobordisms. The third part defines the functor by giving formulae for every elementary braid cobordism. The fourth part is devoted to the proof of invariance under movie moves. The whole proof can be achieved by hand, but in order to save time and space, the authors use sparingly some techniques borrowed from [\textit{D. Clark}, \textit{S. Morrison} and \textit{K. Walker}, Geom. Topol. (3) 13, 1499--1582 (2009; Zbl 1169.57012)]. The last section motivates explicit maps as opposed to Khovanov and Thomas' theoretical arguments. It also discusses a possible refinement for \(\mathbb Z\)-coefficients.