A note on the simplex-cosimplex problem (Q2403687)

This paper is devoted to the existence of triangular 2-faces in \(d\)-polytopes and their duals. (For instance, the hypercube does not have a triangular 2-face while the dual cross-polytope does.) Historically, and in a more general setting, \textit{G. Meisinger} et al. [Discrete Comput. Geom. 24, No. 2--3, 413--420 (2000; Zbl 0961.52003)] show that each \(d\)-polytope with \(d \geqslant 9\) has a 3-simplex as a quotient and raise a question of whether, for every \(k\), there is a polytope \(P^d\) such that either \(P^d\) or its polar dual \((P^d)^*\) has a \(k\)-simplex as a \(k\)-face. The paper under review gives a negative answer to the above question restricted to the class of polytopal spheres (in the topological sense), even in the case \(k = 2\). More precisely, the author constructs, for every \(d \geqslant 3\), polyhedral spheres \(S^d\) that are ``thick'' in the sense that neither \(S^d\) nor its dual \((S^d)^*\) have a triangular 2-face. The method of construction is closely related to the one used by Gao and Yau for proving the existence of nonpositive Ricci curvature structures in 3-sphere. It remains an open challenging problem, whether there exists a polytope \(P\) such that neither \(P\) nor its polar dual \(P^*\) have a triangular 2-face.