Tilting sheaves on toric varieties (Q706125)

The paper delivers a contribution which supports a conjecture by A.~King: Every smooth, projective toric variety \(X\) is supposed to possess a full, strongly exceptional sequence, i.e.\ a tilting sheaf built from line bundles. The existence of such an object would lead to an equivalence between \(\mathcal D^b(\text{Coh }X)\) and the derived category of modules over a finite-dimensional algebra. The result of the paper consists of two theorems which show that the existence of a tilting sheaf is inherited via certain operations with \(X\). The first one takes the total spaces of projective bundles over toric varieties, the second one takes products of (non-neccessarily toric) varieties. The bundle claim is based on an earlier result obtained by \textit{D. O. Orlov} [Russ. Acad. Sci., Izv., Math. 41, No. 1, 133--141 (1993); translation from Izv. Ross. Akad. Nauk, Ser. Mat. 56, No. 4, 852--862 (1992; Zbl 0798.14007)] dealing with exceptional sequences instead. The additional feature of strongly exceptional sequences leads to the restriction to a toric base. To use the previous bundle construction as a step in an induction argument, the toric property has to be bequeathed, too. This leads to considering bundles that split into a direct sum of line bundles. Starting with \(\mathbb P^2\), the authors obtain (in this manner) an affirmative answer to King's conjecture in the case that \(X\) is a toric variety defined by a so-called splitting fan. This notion was introduced by Batyrev and can be defined combinatorially -- but it covers exactly the induction process mentioned above. Examples of toric varieties with splitting fan are those with Picard number two.