Ranks of soft operators in nowhere scattered \(\mathrm{C}^*\)-algebras (Q7028447)

The authors study the problem of realizing tracial states (or, rather, quasitraces) as ranks of operators. In particular, one of the main objects of study is \(\text{QT}(A)\), the cone of lower-semicontinuous, extended-valued 2-quasitraces on \(A\). As the authors mention in the introduction, each such quasi-trace \(\tau\) defines a map\N\[\Nd_\tau \colon (A \otimes \mathbb{K})_+ \to [0, \infty], \text{ where } d_\tau(a) := \lim_{n \to \infty} \tau(a^{1/n}).\N\]\NThe \textit{rank problem} is to determine which functions on \(\text{QT}(A)\) arise as above. The paper, nevertheless, approaches this from a more general setting: that of studying C*-algebras with the \textit{global Glimm property}, see Subsection~2.3 in the text.\N\N{Definition (see 2.3).} \(A\) has the \textit{global Glimm property} if for every \(a \in A_+\) and \(\varepsilon > 0\), there is some \(b \in \overline{aAa}\) such that \(b^2 = 0\) and \((a-\varepsilon)_+ \in \overline{\text{span}}\, AbA\).\N\NThe above was given in \N[\textit{E.~Kirchberg} and \textit{M.~Rørdam}, Adv. Math. 167, No.~2, 195--264 (2002; Zbl 1030.46075), \NDefinition~4.12], and always implies that \(A\) is nowhere scattered (i.e., \(J/I\) is never elementary for any pair of ideals \(I \subseteq J \subseteq A\)). Another notion of interest for the paper is that of \textit{soft} element, where \(a \in A\) is \textit{soft} if it generates a hereditary C*-algebra with no non-zero trivial quotients. The main result of the paper then reads as follows:\N\N{Theorem A.} Let \(A \cong A \otimes \mathbb{K}\) have the \textit{global Glimm property}. Then any \(d_\cdot(a)\) can be realized by a soft element, that is, for every \(a \in A\) there is some soft \(b \in A\) such that \(b\) is Cuntz-subequivalent to \(a\) and, moreover, \(d_\tau(a) = d_\tau(b).\)\N\NThe theorem above has a number of consequences (the most relevant ones listed as Theorem~B, C, D and E in the manuscript). For instance, Theorem~B proves that \(\text{QT}(A)\) is naturally isomorphic to the set of functionals on the soft part of the Cuntz semigroup of \(A\), denoted by \(\text{Cu}(A)_{\mathrm{soft}}\). Likewise, Theorem~D shows some consequences of Theorem~A in the \textit{Toms-Winter Conjecture}. In particular, it shows that the dimension of \(\text{Cu}(A)\) is either that of \(\text{Cu}(A)_{\mathrm{soft}}\) or one more.