Pure subrings of regular local rings, endomorphism rings and Frobenius morphisms (Q1950619): Difference between revisions
From MaRDI portal
Created a new Item |
Added link to MaRDI item. |
||
links / mardi / name | links / mardi / name | ||
Revision as of 16:24, 1 February 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Pure subrings of regular local rings, endomorphism rings and Frobenius morphisms |
scientific article |
Statements
Pure subrings of regular local rings, endomorphism rings and Frobenius morphisms (English)
0 references
13 May 2013
0 references
For our purpose, let \(R\) denote some commutative Noetherian local complete normal domain of Krull dimension \(d\). All modules are finitely generated left module. Let \(X = \text{Spec}(R)\). A crepant resolution is a resolution of singularity \(f: Y \rightarrow X\) which preserves the canonical class i.e. \(f^* \omega_X \cong \omega_Y\). This notion was defined by \textit{M. Reid} [Algebraic Varieties and Analytic Varieties, Adv. Studies in Pure Mathematics 1, 131--180 (1983; Zbl 0558.14028)] to study resolution of threefolds. A crepant resolution exists for threefolds with Gorenstein quotient singularities [\textit{S. S. Roan}, Topology, 35, No. 2, 489--508 (1996; Zbl 0872.14034)] and two such resolutions are related by flops. It rarely exists for higher dimensions. A conjecture of Bondal-Orlov [``Semiorthogonal decomposition for algebraic varieties'', \url{arxiv:alg-geom/9506012}] stated that if two smooth varieties are related by flops then the derived category of coherent sheaves on them are equivalent. This conjecture was proved for the case of smooth three-folds by \textit{T. Bridgeland} [Invent. Math. 147, No. 3, 613--632 (2002; Zbl 1085.14017)]. \textit{M. Van den Bergh} gave a different proof of this result using categorical geometry in [Duke Math. J. 122, No. 3, 423--455 (2004; Zbl 1074.14013)]. In that paper he introduced the concept of non-commutative crepant resolution (henceforth NCCR for brevity) and showed that a three-fold with terminal singularities has a NCCR if and only if it has a crepant commutative resolution. If \(M\) be a torsion-free \(R\)-module, then \(\wedge :=\) End\(_R(M)\) is called a \(NCCR\) if it is a maximal Cohen-Macaulay (MCM) \(R\)-module and has finite global dimension. In what follows, \(S\) will denote a finitely generated \(R\)-module which is a complete regular local normal pure extension of \(R\). In the paper under review, the author proves (cf. Corollary 2.11) that if End\(_R(S)\) is MCM, then it is a NCCR. This important corollary is derived from a more general result in this direction (cf. Theorem 2.6). In section 4, the author proves that if End\(_R(S)\) is MCM and Frobenius map \(R \hookrightarrow R^{1/p}\) is finite (assuming characteristic \(p > 0\))then the ring End\(_R(R^{1/p^e})\) is a NCCR and Morita equivalent to End\(_R(S)\) for \(e \gg 0\) (cf. Corollary 4.2). In section 5, the author derives a non-commutative version of a result of \textit{J. Koh} and \textit{K. Lee} [J. Algebra 202, No. 2, 671--689 (1998; Zbl 0909.13008)] providing an upper bound on the projective dimension of a finitely generated left module (cf. Theorem 5.2).
0 references
noncommutative crepant resolutions
0 references
Frobenius morphisms
0 references
endomorphism rings
0 references
skew group rings
0 references
Cohen-Macaulay modules
0 references
quotient singularities
0 references
global dimension
0 references