A remark about critical sets in \(\mathbb R^3\) (Q2418894)

By definition, a subset \(X \subset \mathbb R^n\) is called critical if there is a smooth function \(f \in C^\infty(\mathbb R^n)\) such that \(X\) is the set of critical points of \(f\), i.e., \(X\) is the set where the differential \(df\) vanishes. In the case \(n=1,2\) this problem was investigated in detail by \textit{A. Norton} and \textit{C. Pugh} [Mich. Math. J. 38, No. 3, 441--459 (1991; Zbl 0785.58016)]. The author analyzes the case \(n=3\). More precisely, let \(X\) be a compact and connected subset \(X \subset \mathbb R^3\) whose complement \(\mathbb R^3 \setminus X\) is connected and \(\dim_{\mathbb R} H^1(\mathbb R^3 \setminus X; \mathbb R) < \infty\). The author proves the following theorem: if \(X\) is critical, then \(X \subset \mathbb R^3\) is tame, i.e., \(X\) is homeomorphic to the interior of a compact manifold with boundary. As a consequence, this implies that the Whitehead continuum and solenoids in \(\mathbb R^3\) are not critical. The same is also true for the \(p\)-adic solenoids (Smale-Williams attractors), as it was conjectured earlier by \textit{M. Grayson} and \textit{C. Pugh} [Publ. Math., Inst. Hautes Étud. Sci. 77, 5--61 (1993; Zbl 0796.58012)]. In conclusion, a slightly weaker version of the main result for subsets in \(\mathbb R^n\) for any \(n\) is given.