Abstract:In this paper, we construct examples of exact Lagrangians (of "locally conformally symplectic" type) in cotangent bundles of closed manifolds with locally conformally symplectic structures and give conditions under which the projection induces a simple homotopy equivalence between an exact Lagrangian and the $0$-section of the cotangent bundle. This line of questioning follows in the footsteps of Abouzaid and Kragh, and more generally of the Arnol'd conjecture. Notably, we will see that while exact Lagrangians cannot be spheres in this setting, a naive adaptation of the Abouzaid-Kragh theorem does not hold in this generalization.
What problem does this paper attempt to address?
The problems that this paper attempts to solve mainly focus on the rigidity constraints of "exact" Lagrangian submanifolds in locally conformally symplectic (lcs) geometry, especially in the case of having an "exact" lcs structure in the cotangent bundle of a closed manifold. Specifically, the paper explores the properties and behaviors of exact Lagrangian submanifolds under the framework of lcs geometry, especially their topological properties and the relationship with the base manifold.
### Main contributions of the paper:
1. **Construct new examples of exact Lagrangian submanifolds**:
- The paper constructs specific examples of exact Lagrangian submanifolds with a locally conformally symplectic structure in the cotangent bundle of a closed manifold.
- These examples show that in lcs geometry, the behavior of exact Lagrangian submanifolds is significantly different from that in standard symplectic geometry.
2. **Homotopy equivalence conditions induced by projection**:
- The paper gives the conditions under which an exact Lagrangian submanifold can induce a simple homotopy equivalence when projected onto the base manifold under certain conditions.
- Specifically, the paper proves that if certain specific conditions are met, the exact Lagrangian submanifold will maintain the equivalence of topological structure when projected onto the base manifold.
3. **Topological constraints**:
- The paper explores the constraints of the homology group of the base manifold on the homology group of the exact Lagrangian submanifold.
- For example, the paper proves that if the first cohomology group of the base manifold \( M \) is not zero, then the first cohomology group of the exact Lagrangian submanifold \( L \) is also not zero.
### Main results:
1. **Proposition 1.1**:
- For any closed manifold \( M \) with \( \dim(M)\neq 2 \), there exists a connected closed manifold \( L \) and an embedding \( i: L \to (T^*(M\times S^1_\theta), \lambda, d\theta) \) such that \( i(L) \) is a \( d\theta \)-exact Lagrangian submanifold, and \( \chi(L)\neq \chi(M) \) (where \( \chi \) represents the Euler characteristic).
- This proposition shows that the conclusion of the Abouzaid - Kragh theorem does not hold in lcs geometry, especially for 0 - exact Lagrangian submanifolds, the degree of projection can be zero.
2. **Theorem 1.1**:
- Let \( M \) be a closed manifold, \( \lambda \) be the canonical Liouville form of \( T^*M \), and \( \beta \in \Omega^1(M) \) be a closed form. If \( L \) is a connected closed \( \beta \)-exact Lagrangian submanifold in \( T^*M \), then:
\[
[\beta] \neq 0 \in H^1(M, \mathbb{R}) \implies [\beta] \neq 0 \in H^1(L, \mathbb{R})
\]
- This theorem means that as long as \( M \) is not a sphere, the exact Lagrangian submanifold \( L \) cannot be a sphere either.
3. **Theorem 1.2**:
- Let \( M \) be a connected closed manifold with dimension \( n \), \( \lambda \) be the canonical Liouville form of \( T^*M \), and \( \beta \in \Omega^1(M) \) be a closed form. If \( L \) is a connected closed manifold with dimension \( n \) and there is an embedding \( i: L \to (T^*M, \lambda, \beta) \) such that \( i^*\lambda = d(i^*\beta f) \) (where \( f \in C^