A construction of complete complex hypersurfaces in the ball with control on the topology (Q2420207)

The main theorem of this paper states that for any complex hypersurface \(Z \subset \mathbb{C}^{n+1}\) and a compact set \(K \subset Z\) there exists a pseudoconvex Runge domain \(U \subset Z\) containing \(K\) which has a complete proper holomorphic embedding into the unit ball \(B_{n+1} \subset \mathbb{C}^{n+1}\). As an application of this theorem the authors prove that for \(n = 1\) a finite topology of \(D\) can be prescribed. The first such results are due to \textit{P. W. Jones} [Proc. Am. Math. Soc. 76, 305--306 (1979; Zbl 0418.32006)] who construcded complete bounded holomorphic embedding of the unit disc \(B_1 \subset \mathbb{C}\) to \(\mathbb{C}^3\) and a proper complete embedding of \(B_1\) to \(B_4\). Later the first author, Forstnerič and the third author [\textit{A. Alarcón} and \textit{F. Forstnerič}, Math. Ann. 357, No. 3, 1049--1070 (2013; Zbl 1288.32014); Invent. Math. 196, No. 3, 733--771 (2014; Zbl 1297.32009); Math. Ann. 363, No. 3--4, 913--951 (2015; Zbl 1343.53053); \textit{A. Alarcón} and the third author, Math. Ann. 355, No. 2, 429--455 (2013; Zbl 1269.53061)] constructed proper complete embeddings of open and bordered Riemann surfaces to \(B_3\). \textit{B. Drinovec Drnovšek} [J. Math. Anal. Appl. 431, No. 2, 705--713 (2015; Zbl 1326.32048)] proved a similar result for bounded strictly pseudoconvex domains in high-dimensional balls. The present paper is based on a refinement of the approach of the second author [Ann. Math. (2) 182, No. 3, 1067--1091 (2015; Zbl 1333.32018)], used to find an embedded complete closed complex hypersurface in \(B_{n+1}\). The proof of the main theorem is divided into several steps. First the obstacles \(\{T_j\}_{j \in \mathbb{N}_0}\) in \(B_{n+1} \setminus \frac{1}{2}\overline{B}_{n+1}\) are created. The (disjoint) sets \(T_j\), \(j \in \mathbb{N}\), are (real) (\(2n+1\))-dimensional balls tangent to spheres \(\mathbb{S}^{2n+1}(0,r_j)\), \(r_j < 1\), and they are positioned in such a way that any closed hypersurface in \(B_{n+1}\) avoiding them winds so wildly that it is complete. Next an exhaustion \(\{E_j\}_{j \in \mathbb{N}} \) of \(B_{n+1}\) by smoothly bounded strictly pseudoconvex domains is chosen such that \(T_i \subset E_j\), \(i \leq j-1\) and \(\overline{E_j} \cap E_i = \emptyset\), \(i \geq j\). Inductively a sequence of automorphisms \(\{\Phi_j\}_{j \in \mathbb{N}}\) is constructed such that \(\Phi_j\) approximates the identity on \(E_j\), \(\Phi_j \circ \dots \circ \Phi_1(Z)\) avoids \(T_0,\ldots, T_j\) and converges on a Runge domain \(\Omega \subset \mathbb{C}^{n+1} \) to a map which maps \(\Omega\) biholomorphically to the unit ball \(B_{n+1}\). The domain \(D\) is then a connected component of \(\Omega \cap Z\) containing \(K\).