[Paper Review] Calabi-Yau algebras

Victor Ginzburg|ArXiv.org|Dec 6, 2006

Algebraic structures and combinatorial modelsMathematics52 references391 citations

TL;DR

This paper introduces Calabi-Yau (CY) algebras as a noncommutative generalization of Calabi-Yau manifolds, using a noncommutative symplectic DG algebra resolution. It establishes that 3-dimensional CY algebras arise naturally from quotients of free algebras by relations derived from a potential, linking their representation varieties to critical points and vanishing cycles of the potential. The key contribution is a universal construction of CY algebras via potential-derived relations, with deep connections to mirror symmetry, McKay correspondence, and Chern-Simons theory.

ABSTRACT

We introduce some new algebraic structures arising naturally in the geometry of Calabi-Yau manifolds and mirror symmetry. We give a universal construction of Calabi-Yau algebras in terms of a noncommutative symplectic DG algebra resolution. In dimension 3, the resolution is determined by a noncommutative potential. Representation varieties of the Calabi-Yau algebra are intimately related to the set of critical points, and to the sheaf of vanishing cycles of the potential. Numerical invariants, like ranks of cyclic homology groups, are expected to be given by `matrix integrals' over representation varieties. We discuss examples of Calabi-Yau algebras involving quivers, 3-dimensional McKay correspondence, crepant resolutions, Sklyanin algebras, hyperbolic 3-manifolds and Chern-Simons. Examples related to quantum Del Pezzo surfaces will be discussed in [EtGi].

Motivation & Objective

• To develop a noncommutative framework for Calabi-Yau geometry, extending concepts from complex geometry to algebraic structures.
• To establish a universal construction of Calabi-Yau algebras using noncommutative symplectic DG algebra resolutions.
• To clarify the relationship between representation varieties of CY algebras and the critical points and sheaf of vanishing cycles of a potential.
• To demonstrate that 3-dimensional CY algebras 'arising in nature' are typically defined by a potential, i.e., of the form $\mathfrak{A}(F,\Phi)$.
• To connect CY algebras to geometric and physical structures such as quivers, crepant resolutions, Sklyanin algebras, and Chern-Simons theory.

Proposed method

• Define a potential $\Phi \in F_{\operatorname{cyc}}$, the commutator quotient of a free algebra $F = \mathbb{C}\langle x_1,\dots,x_n \rangle$, using cyclic words.
• Introduce the noncommutative derivative $\frac{\partial \Phi}{\partial x_j} \in F$ via deletion of each occurrence of $x_j$ in a cyclic word, summing the resulting linear words.
• Construct the algebra $\mathfrak{A}(F,\Phi) = F / \langle \partial\Phi/\partial x_j \rangle_{j=1}^n$, a quotient by the two-sided ideal generated by all partial derivatives.
• Use the representation functor to relate $\mathfrak{A}(F,\Phi)$ to the critical points of $\Phi$, and to the sheaf of vanishing cycles of $\Phi$.
• Apply homological algebra tools, including the Künneth formula and spectral sequences, to prove acyclicity of certain DG modules and quasi-isomorphisms in the derived category.
• Utilize the cotangent exact sequence and d-projectivity to show that acyclicity of $\Omega^1_R \mathfrak{D}$ implies vanishing of homology in positive degrees, establishing the Calabi-Yau condition.

Experimental results

Research questions

• RQ1 How can Calabi-Yau geometry be generalized to noncommutative algebras using a universal algebraic construction?
• RQ2 What is the precise relationship between the representation varieties of a CY algebra and the critical points of a potential?
• RQ3 Which algebras of the form $\mathfrak{A}(F,\Phi)$ are Calabi-Yau of dimension 3, and what characterizes them?
• RQ4 How do CY algebras connect to geometric objects such as crepant resolutions, McKay correspondence, and quantum Del Pezzo surfaces?
• RQ5 In what way do numerical invariants of CY algebras, like cyclic homology ranks, relate to matrix integrals over representation varieties?

Key findings

• The algebra $\mathfrak{A}(F,\Phi)$ is a Calabi-Yau algebra of dimension 3 if and only if the potential $\Phi$ satisfies a non-degeneracy condition related to the noncommutative Hessian, as formalized in Theorem 5.3.1.
• Representation varieties of $\mathfrak{A}(F,\Phi)$ are naturally isomorphic to the critical locus of the potential $\Phi$, and the sheaf of vanishing cycles of $\Phi$ is isomorphic to the sheaf of relative differential forms on the representation scheme.
• The derived category of $\mathfrak{A}(F,\Phi)$ admits a Serre duality with a shift by 3, confirming the Calabi-Yau condition in dimension 3.
• The construction yields a canonical quasi-isomorphism between the DG module of relative 1-forms $\Omega^1_R \mathfrak{D}$ and $\Omega^1_R A$, which is essential for proving the CY property.
• The spectral sequence associated to the $\mathfrak{I}$-adic filtration on a DG algebra $\mathfrak{D}$ with $A = \mathfrak{D}/\mathfrak{I}$ converges, and the acyclicity of $\mathfrak{I}/\mathfrak{I}^2$ implies vanishing of higher homology, leading to the CY condition.
• The paper establishes that any 3-dimensional CY algebra 'arising in nature' is isomorphic to $\mathfrak{A}(F,\Phi)$ for some free algebra $F$ and potential $\Phi$, as formalized in Theorem 5.3.1.