Non-commutative projective Calabi-Yau schemes (Q2259186)

Noncommutative Calabi-Yau algebras have had considerable attention, most likely due to their connection with superstring theory. Almost all known noncommutative Calabi-Yau algebras are quiver algebras, which means that they are locally Calabi-Yau manifolds. This article constructs the first examples of non-trivial noncommutative projective Calabi-Yau schemes in the sense of \textit{M. Artin} and \textit{J. J. Zhang} [Adv. Math. 109, No. 2, 228--287 (1994; Zbl 0833.14002)]. Let \(k\) be a field, \(A=\bigoplus_{i=0}^\infty A_i\) a connected Noetherian graded \(k\)-algebra. Let \(\mathrm {gr}(A)\) be the subset of \(\mathrm {Gr}(A)\) consisting of finitely generated graded right \(A\)-modules. Let \(\mathrm {Tor}(A)\) be the subcategory of \(\mathrm {Gr}(A)\) consisting of torsion modules, and let \(\mathrm {tor}(A)=\mathrm {Tor}(A)\cap\mathrm {Gr}(A)\). Denote by \(\mathrm {Tails}(A)\) the abelian quotient category \(\mathrm {Tails}(A)=\mathrm {Gr}(A)/\mathrm {Tor}(A)\), the (canonical) quotient functor is denoted by \(\pi\). Also \(\mathrm {tails}(A)=\mathrm {gr}(A)/\mathrm {tor}(A)\), and if \(M\in\mathrm { Gr}(A)\), the notation is \(\mathcal M=\pi(M)\). Due to Artin and Zhang, the noncommutative projective scheme of a graded right Noetherian \(k\)-algebra \(A\) is defined as a triple \(\mathrm {proj}(A)=(\mathrm {tails}(A),\mathcal A,s)\) where \(s\) is the shift functor. Because \(\mathrm {Tails}(A)\) is an abelian category with enough injectives, the \(i\)-th derived functors \(\mathrm {Ext}^i_{\mathrm {Tails}(A)}(\mathcal M,\ast)\) of \(\mathrm {Hom}_{\mathrm {Tails}(A)}(\mathcal M,\ast)\) can be defined. The global section functor \(H^0(\mathrm {proj}(A),\ast)=\mathrm {Hom}_{\mathrm {Tails}(A)}(\mathcal A,\ast):\mathrm {Tails}(A)\rightarrow\mathrm {Vect}_k\) induces higher cohomologies \(H^i(\mathrm {proj}(A),\mathcal M)=\mathrm {Ext}^i_{\mathrm {Tails}(A)}(\mathcal A,\mathcal M)\), and the bifunctor \(\mathrm {Ext}^i_{\mathrm {tails}(A)}(\ast,\ast\ast)\) is defined as the restriction of \(\mathrm {Ext}^i_{\mathrm {Tails}(A)}(\ast,\ast\ast)\) on \(\mathrm {tails}(A)\). A notherian graded \(k\)-algebra \(A\) satisfies condition \(\;chi\) if \(\dim_k\mathrm {Ext}^i_{\mathrm {Tails}(A)}(k,M)<\infty\) for all \(i\geq 0\). The main result of the article is, \textit{verbatim}: Theorem: Let \(k\) be an algebraically closed field of characteristic zero and consider the following graded \(k\)-algebra \(A_n=k\langle x_1,\dots,x_n\rangle/(\sum_{k=1}^n x_k^n,\;x_ix_j=q_{ij}x_jx_i)_{ij},\) where the quantum parameters \(q_{ij}\in k^\times\) satisfy \(q_{ii}=q_{ij}^n=q_{ij}q_{ji}=1\). Then the quotient category \(\mathrm {Coh}(A_n)=\mathrm {gr}(A_n)/\mathrm {tor}(A_n)\) is a Calabi-Yau \((n-2)\) category if and only if \(\prod_{i=1}^n q_{ij}\) is independent of \(1\leq j\leq n.\) The graded \(k\)-algebra \(A_n\) is of the form \(A_n=B_n/(f_n)\) with \(B_n=k\langle x_1,\dots,x_n\rangle/x_ix_j=q_{ij}x_jx_i)_{i,j}\), \(f_n=\sum_{k=1}^n x_k^n.\) \(B_n\) is a Koszul Artin-Schelter (AS) regular algebras, and \(f_n\) is a normalizing element of degree equal to the global dimension of \(B_n\). Thus informally, \(\mathrm {proj}(A_n)\) is the noncommutative Fermat hypersurface in quantum \(\mathbb P^{n-1}\) which was previously studied in physics. In addition, the authors prove that there exists \(q_{ij}\)'s such that the graded \(k\)-algebra \(A_n\) is not a twisted coordinate ring of a Calabi-Yau \((n-2)\)-fold. The author claims that one of the motivations for the study comes from a virtual counting theory of stable sheaves on a polarized complex Calabi-Yau threefold. \textit{B. Szendrői} [Geom. Topol. 12, No. 2, 1171--1202 (2008; Zbl 1143.14034)] introduced a noncommutative version of quiver Calabi-Yau 3 algebras. This relies on the existence of the global Chern-Simons function on the moduli space of stable modules, and can not be generalized to the projective case. The author developed a virtual counting theory of stable modules over a noncommutative projective Calabi-Yau scheme, based on the work of Berend, Ciocan-Fontanine, Hwang, and Rose, and the algebra \(A_n\) above is an example of the general theory in the authors' PhD thesis. The article is explicit, and gives good examples and explanations of the general theory. Also, the main results are both nice and important.