Boundaries of reduced \(C^\ast\)-algebras of discrete groups (Q2397741)

\textit{M. Hamana} [Tohoku Math. J. (2) 30, 439--453 (1978; Zbl 0391.46054); ibid. 37, 463--487 (1985; Zbl 0585.46053)] introduced the notion of injectivity in the category of Banach modules over a Banach algebra, which was introduced in the category of Banach spaces. He also constructed the injective envelope of a Banach module equipped with the action of a discrete group \(G\). In the paper under review, after an introduction and preliminaries, in Section 3, the authors construct the Furstenberg boundary using Hamana's theory of equivariant injective envelopes. They consider the minimal \(C^*\)-subalgebra of \(\ell^\infty(G)\) that arises as the image of a unital positive \(G\)-equivariant projection on \(\ell^\infty(G)\); its existence and uniqueness was proved by Hamana. Moreover, they construct the algebra \(C(\partial_F G)\) of continuous functions on Furstenberg's universal \(G\)-boundary \(\partial_F G\) as a \(G\)-injective envelope in the sense of Hamana; and prove that the Hamana boundary \(\partial_F G\) can be identified with Furstenberg's universal \(G\)-boundary \(\partial_F G\). This construction of the Furstenberg boundary implies some powerful rigidity results for \(G\)-equivariant maps on \(\partial_F G\). \textit{N. Ozawa} [Bull. Lond. Math. Soc. 39, No. 1, 35--38 (2007; Zbl 1119.46042)] considered the problem of constructing a more tight nuclear embedding of the reduced \(C^*\)-algebra \(C^*_r({\mathbb F}_n)\) of the free group \({\mathbb F}_n\) on \(n\) generators. This \(C^*\)-algebra is known to be exact. Ozawa proved that there is a canonical nuclear \(C^*\)-algebra \(N(C^*_r({\mathbb F}_n))\) such that \[ C^*_r({\mathbb F}_n)\subseteq N(C^*_r({\mathbb F}_n))\subseteq I(C^*_r({\mathbb F}_n)), \] where \(I(C^*_r({\mathbb F}_n))\) denotes the injective envelope of \(C^*_r({\mathbb F}_n)\). More generally, Ozawa conjectured that it should be possible to construct such an embedding for any exact \(C^*\)-algebra. In Section 4, the authors prove that \(G\) is exact precisely when the \(G\)-action on \(\partial_F G\) is amenable, and use this fact to prove Ozawa's conjecture for the reduced \(C^*\)-algebra \(C^*_r(G)\) of every discrete exact group \(G\). More precisely, the authors show that, for a discrete exact group \(G\), there is a canonical nuclear \(C^*\)-algebra \(N(C^*_r(G))\) such that \[ C^*_r(G)\subseteq N(C^*_r(G))\subseteq I(C^*_r(G)), \] where \(I(C^*_r(G))\) denotes the injective envelope of \(C^*_r(G)\). Furthermore, they show that the algebra \( N(C^*_r(G))\) is simple if \(C^*_r(G)\) is simple, and prime if and only if \(C^*_r(G)\) is prime. In Section 5, the authors investigate rigidity and the injective envelope; this investigation is motivated by a key ingredient in Ozawa's paper, a rigidity result [loc. cit., Proposition 3] for unital positive equivariant maps between spaces of functions on the hyperbolic boundary of a hyperbolic group. They prove a generalization of this result which imposes slightly weaker requirements. The proof of their result is significantly different than Ozawa's. In fact, their techniques also apply to groups that are not necessarily hyperbolic or even relatively hyperbolic, including certain mapping class groups. The last section of this paper is dedicated to \(C^*\)-simplicity. In fact, the authors consider the connection between the \(C^*\)-simplicity of a group and the ergodic properties of its action on the Furstenberg boundary. They prove that Tarski monster groups are \(C^*\)-simple, which provides another solution to a problem of de la Harpe (recently answered by \textit{A. Yu. Olshanskii} and \textit{D. V. Osin} [Groups Geom. Dyn. 8, No. 3, 933--983 (2014; Zbl 1354.22009)]] about the existence of \(C^*\)-simple groups with no free subgroups. The authors' construction of the Furstenberg boundary provides an appropriate quantum-group-theoretic analogue of the Furstenberg boundary; and many of their results also hold for certain locally compact quantum groups. They intend to pursue these issues in a later paper in that setting.