Tilting theoretical approach to moduli spaces over preprojective algebras (Q2637777)

The paper under review studies variations of moduli spaces of representations of preprojective \(K\)-algebras by applying tilting theory, to deal with minimal resolutions of Kleinian singularities. To a finite subgroup \(G\subset SL(2,K)\) a McKay quiver \(Q\) is associated, and a \(K\)-algebra \(\Lambda\) associated to \(Q\), called preprojective algebra. Different resolutions of quotient singularities \(\mathbb{A}^{n}/G\) are encoded by different moduli spaces \(\mathcal{M} _{\theta,d}({\Lambda})\) of \(\theta\)-semistable \(\Lambda\)-modules of dimension vector \(d\), in the sense of \textit{A. D. King} [Q. J. Math., Oxf. II. Ser. 45, No. 180, 515--530 (1994; Zbl 0837.16005)]. Hence, by studying variations of moduli spaces for different stability parameters \(\theta\) we can study quotient singularities. In section 2, tilting modules \(I_{w}\) (generalization in homological algebra of being torsion or torsion free for a module) are constructed over preprojective algebras. If we denote by \(\mathcal{S}_{\theta}(\Lambda)\subset Mod \Lambda\) the full subcategory of \(\theta\)-semistable \(\Lambda\)-modules, in section 3 relations between the moduli spaces are given, by showing that the functors \(Hom_{\Lambda}(I_{w},-)\) and \(-\otimes_{\Lambda}I_{w}\) induce equivalences between the categories \(\mathcal{S}_{\theta}(\Lambda)\) and \(\mathcal{S}_{w\theta}(\Lambda)\) (c.f. Theorem 3.13) which preserve \(S\)-equivalence classes, where \(w\) are elements of the Coxeter group. This induces a bijection between the sets of closed points in the moduli spaces, and in section 4 the equivalence can be extended to the respective derived categories to show that the bijection can be extended to an isomorphism of \(K\)-schemes (c.f. Theorem 4.20), by using the functors of points. Section 5 is devoted to use the previous results in the framework of Kleinian singularities to generalize some results of \textit{W. Crawley-Boevey} (see, e.g., [Am. J. Math. 122, No. 5, 1027--1037 (2000; Zbl 1001.14001)]). In section 6 a full example is provided. Note that the proofs work even in the case where \(Q\) is a non-Dynkin quiver with no loops. Combined with the homological nature of the proofs, this is why the authors expect to use the results in higher dimensions.