Uniqueness for embeddings of nuclear $C^*$-algebras into type II$_{1}$ factors
Shanshan Hua, Stuart White
TL;DR
This work establishes uniqueness up to unitary conjugacy for unital, full, and nuclear maps from a separable, exact $C^*$-algebra $A$ with the UCT into ultraproducts of finite von Neumann factors, and specializes to II$_1$-factors. The authors develop a KK- and KL- uniqueness framework tailored to trace–kernel extensions, using a new $K_1$-injectivity result for Paschke duals under relatively purely large extensions, and then apply Schafhauser–style lifting to obtain unitary equivalence from trace and total $K$-theory data. The approach allows passing to ultraproducts and separable subextensions, yielding a novel uniqueness theorem for quasidiagonal approximations and a norm-approximate unitary equivalence result for embeddings into II$_1$ factors. Central to the method are the trace–kernel extension, a relatively purely large ideal in the multiplier algebra, and a $KL_{ ext{nuc}}$-based obstruction vanishing grounded in the universal multicoefficient theorem. The results significantly advance embeddings into II$_1$- and ultraproduct factors and provide robust tools for quasidiagonality and operator-algebra classification in this setting.
Abstract
Let $A$ be a separable, unital and exact $C^*$-algebra satisfying the universal coefficient theorem. We prove uniqueness theorems up to unitary conjugacy for unital, full and nuclear maps from $A$ into ultraproducts of finite von Neumann factors: any two such maps agreeing on traces and total $K$-theory are unitarily equivalent. There are two consequences. Firstly if one takes the factors to be a sequence $(M_{k_n})_{n}$ of matrix algebras, we obtain a uniqueness result for quasidiagonal approximations of $A$. Secondly, when $(\mathcal M,τ_{\calM})$ is a II$_1$ factor, a pair $φ,ψ:A\to\mathcal M$ of unital, injective and nuclear maps are norm approximately unitarily equivalent if and only if $τ_{\calM}\circφ=τ_{\calM}\circψ$. The main strategy is to use Schafhauser's classification of lifts along the trace--kernel extension. Since our codomains may lack the tensorial absorption properties needed in this work, the main new ingredient is a suitable $KK$-uniqueness theorem tailored to our situation. This is inspired by $KK$-uniqueness theorems of Loreaux, Ng and Sutradhar.
