Tensor products and regularity properties of Cuntz semigroups

Abstract

The Cuntz semigroup of a C*-algebra is an important invariant in the structure and classification theory of C*-algebras. It captures more information than K-theory but is often more delicate to handle. We systematically study the lattice and category theoretic aspects of Cuntz semigroups. Given a C*-algebra A, its (concrete) Cuntz semigroup Cu(A) is an object in the category Cu of (abstract) Cuntz semigroups, as introduced by Coward, Elliott and Ivanescu. To clarify the distinction between concrete and abstract Cuntz semigroups, we will call the latter Cu-semigroups. We establish the existence of tensor products in the category Cu and study the basic properties of this construction. We show that Cu is a symmetric, monoidal category and relate Cu(A B) with Cu(A)CuCu(B) for certain classes of C*-algebras. As a main tool for our approach we introduce the category W of pre-completed Cuntz semigroups. We show that Cu is a full, reflective subcategory of W. One can then easily deduce properties of Cu from respective properties of W, e.g. the existence of tensor products and inductive limits. The advantage is that constructions in W are much easier since the objects are purely algebraic. We also develop a theory of Cu-semirings and their semimodules. The Cuntz semigroup of a strongly self-absorbing C*-algebra has a natural product giving it the structure of a Cu-semiring. We give explicit characterizations of Cu-semimodules over such Cu-semirings. For instance, we show that a Cu-semigroup S tensorially absorbs the Cu-semiring of the Jiang-Su algebra if and only if S is almost unperforated and almost divisible, thus establishing a semigroup version of the Toms-Winter conjecture.

0

Turn this paper into a lesson

ArcXiv compiles a structured reading guide from this paper's metadata: plain-English importance, contributions, prerequisite concepts, which sections to read first, flashcards, and a quiz. Grounded in the abstract, never invented.

Discussion (0)

Sign in to join the discussion.

Loading comments…