Some constraints on Frobenius manifolds with a tt*-structure (Q627464)

The author studies some properties of the so-called Frobenius manifolds and tt*-structures. Frobenius manifolds were introduced by Dubrovin (1993) to generalize and formalize some previously considered structures in TFT (topological field theory) (E. Witten (1991), S. Ceccotti and C. Vafa (1991, 1994)) and unfolding theory of hypersurface singularities (K. Saito and M. Saito). Furthermore, a tt*-structure on a complex manifold \((M,J)\) is a real vector bundle \(E\to M\), endowed with a connection \(D\) and a section \(S\in\Gamma(T^*M\bigotimes \text{End}(E))\) which satisfy the so-called tt*-equation \(R^\theta=0\) \(\forall\theta\in {\mathbb R}\), where \(R^\theta\) is the curvature tensor of the connection \(D^{\theta}\) defined by \(D^{\theta}_X=D_X+(\cos\theta)S_X+(\sin\theta)S_{J(X)}\) \(\forall X\in\Gamma(TM)\). In particular, for \(E=TM\), a metric tt*-bundle is a possibly indefinite \(D\)-parallel-fibre metric \(g\) such that \(g(S_XY,Z)=g(Y,S_XZ)\) \(\forall X,Y,Z\in T_pM\), \(p\in M\). By Sato-Sato's work, it is known that the base space of a semi-universal unfolding of a hypersurface singularity can be equipped with the structure of a Frobenius manifold, and, by Ceccotti-Vafa's work, such manifold can be equipped with a tt*-geometry if the singularity is quasi-homogeneous. These structures are combined into a new structure called CDV-structure by C. Hertling (2003). The purpose of this paper is to give a necessary and sufficient condition for a Frobenius manifold to be a CDV-structure (Theorem 1). Furthermore, the author proves that, on a CDV-structure, the real structure \(\kappa\) cannot be flat and the underlying real \((1,1)\)-form of a natural semi-simple CDV-structure cannot be a symplectic form. In particular, the underlying Hermitian manifold of a non-trivial semi-simple positive CDV-structure cannot be a Kähler manifold. The paper, after a concise introduction, splits into four more sections. 2. Frobenius manifold and tt*-geometry. 3. Main results. 4. Proof of the theorems. 5. Applications. (5.1 Some consequences of Theorem 1. 5.2 Applications to harmonic Frobenius manifolds.) Reviewer's remark. It is interesting to emphasize that the geometric structures considered in this paper, and also some related ones as, e.g., the so-called massive Frobenius manifolds, and special geometry structures (also called topological-antitopological fusion structures) are of great importance in TFT. (See also, e.g., works by Alekseevsky-Cortés-Devchand (2002) and by Cortés-Schäfer (2004) and relations with Calabi-Yau manifolds and with Gromov-Witten invariants about the theory of integrable systems in the work by Dubrovin-Zhang (2008).) However, their physical applications require to handle quantum frameworks (see, e.g., Y. I. Manin (1999) and C. Sabbah (2002, 2008)), so they cannot be directly applied to encode quantum physical worlds. Despite the aspect that, in some sense, there is nowadays a strong contradiction to the physical motivations from which they originated, the geometrical theory of quantum (super) PDEs, as formulated by the reviewer of this paper, allows one to extend Frobenius manifolds and tt*-structures also in the category of quantum (super) manifolds. In fact, the geometric theory of noncommutative PDEs allows one to translate the very crucial aspects of integrability of such structures in noncommutative frameworks useful to encode quantum physical phenomena.