Almost excellent unique factorization domains (Q2299425): Difference between revisions

An important area of research in Commutative Algebra is completions of local rings. Let us mention two important results: \begin{itemize} \item[(1)] In [Trans. Am. Math. Soc. 337, No. 1, 379--387 (1993; Zbl 0792.13011)], \textit{R. C. Heitmann} found necessary and sufficient conditions for a ring to be the completion of a local UFD. \item[(2)] In [J. Algebra 265, No. 1, 221--228 (2003; Zbl 1083.13508)], \textit{S. Loepp} characterized the completions of local excellent integral domains of characteristic zero. \end{itemize} Because of these results, the natural next question to consider is the question of when a complete local ring is the completion of an excellent UFD. This question is still open, however, there are some partial results. Let us mention two of them: \begin{itemize} \item[(1)] In [Commun. Algebra 33, No. 5, 1321--1336 (2005; Zbl 1072.13006)], \textit{J. Bryk} et al. modified Heitmann's method and showed that ``most'' complete local rings \(T\) are completions of a UFD \(A\) whose formal fibers are geometrically regular except for those over some height-one prime ideals. \item[(2)] In [Can. J. Math. 62, No. 4, 721--736 (2010; Zbl 1198.13021)], \textit{A. Boocher} et al. continued to study the relationship between a local ring \(A\) and its completion \(\widehat{A}\) by examining the formal fibers of the ring \(A\), but they considered a different kind of property on \(A\). They too adapted techniques used in the above mentioned paper by Heitmann. \end{itemize} In the paper under review, the authors also obtain some results related to the above mentioned natural question. Namely, let \((T,\mathfrak{m})\) be a complete local (Noetherian) domain of depth \(>1\), which contains rationals, satisfies \(|T|=|T/\mathfrak{m}|\), and whose set of all principal height-one prime ideals has the same cardinality as \(T\). The authors construct a universally catenary local UFD \(A\) such that the completion of \(A\) is \(T\) and such that there exist uncountably many height-one prime ideals \(\mathfrak{q}\) of \(T\) having the property that \((T/(\mathfrak{q}\cap A)T)_\mathfrak{q}\) is a field. Furthermore, if \(T\) is a normal domain, they were able to make \(A\) ``close'' to excellent in the sense that the formal fiber at every prime ideal of \(A\) of height not equal to \(1\) is geometrically regular, and uncountably many height-one prime ideals of \(A\) have geometrically regular formal fibers.