Unique representation domains. II. (Q2457282)

Given a star operation \(*\) of finite type, a domain \(R\) is said to be a \(*\)-unique representation domain (\(*-URD\)) if each \(*\)-invertible \(*\)-ideal can be uniquely expressed as a \(*\)-product of pairwise \(*\)-comaximal ideals with prime radicals. When \(*=t\), a \(*-URD\) is simply called a \(URD\) (Unique Representation Domain). The paper under review deals with \(*-URD\)'s and it contains three sections. In the first section, the authors characterize the \(PVMD\)'s that are \(URD\)'s. Mainly, a \(PVMD\) is a \(URD\) if and only if every nonzero principal ideal of \(R\) has at most finitely many minimal primes. In particular, a ring of Krull type is a \(URD\). The second section deals with a general approach. The main theorem states that if \(*\) is a star operation of finite type, then \(R\) is a \(*-URD\) if and only if \(*-\mathrm{Spect}(R)\) is treed and each proper \(*\)-ideal of finite type has only finitely many minimal primes, if and only if \(*-\mathrm{Spect}(R)\) is treed and each proper \(*\)-invertible \(*\)-ideal has only finitely many minimal primes if and only if \(*-\mathrm{Spect}(R)\) is treed and each nonzero proper principal ideal has only finitely many minimal primes. The last section discusses extensions to polynomial rings and ways of constructing \(URD\)'s. We note that a polynomial ring \(D[X]\) is a \(URD\) if and only if \(D\) is a \(UMT\) domain which is a \(URD\). Several applications and examples are constructed.