Cohomology ring of unitary $N=(2,2)$ full vertex algebra and mirror symmetry
Yuto Moriwaki
TL;DR
The paper provides a rigorous VOAs-based framework for 2d $N=(2,2)$ SCFTs, defining holomorphic/topological twists and introducing cohomology rings $H(F,d_B)$ and $H(F,d_A)$ that encode 2d TFTs and Frobenius-structure. It establishes unitarity-driven refinements to the cohomology (real $\mathbb{R}^2$-grading), proves Hodge duality and T-duality relations, and develops a spectral-flow construction that links the twisted sectors to CY-type invariants such as the Witten genus. A central contribution is the proposal and partial realization of Hodge-theoretic mirror symmetry at the SCFT/VOA level via mirror algebras $\hat{F}$, supported by tests on abelian varieties and a special K3. The framework aims to connect Calabi–Yau geometry with rigorous algebraic structures in full VOAs, offering a route to proving Hodge-theoretic mirror symmetry through sigma-model-inspired CY-type VOAs and their induced cohomological dualities. Overall, the work lays a mathematically precise foundation for extracting geometric invariants from unitary $N=(2,2)$ full VOAs and for exploring mirror phenomena through cohomology, dualities, and spectral flow in a fully algebraic setting.
Abstract
The mirror symmetry among Calabi-Yau manifolds is mysterious, however, the mirror operation in 2d N=(2,2) supersymmetric conformal field theory (SCFT) is an elementary operation. In this paper, we mathematically formulate SCFTs using unitary full vertex operator superalgebras (full VOAs) and develop a cohomology theory of unitary SCFTs (aka holomorphic / topological twists). In particular, we introduce cohomology rings, Hodge numbers, and the Witten index of a unitary $N=(2,2)$ full VOA, and prove that the cohomology rings determine 2d topological field theories and give relations between them (Hodge duality and T-duality). Based on this, we propose a possible approach to prove the existence of mirror Calabi-Yau manifolds for the Hodge numbers using SCFTs. For the proof, one need a construction of sigma models connecting Calabi-Yau manifolds and SCFTs which is still not rigorous, but expected properties are tested for the case of Abelian varieties and a special K3 surface based on some unitary $N=(2,2)$ full VOAs.
