Injective envelopes of C ∗ C^{*} -algebras as maximal rigid multiplier covers
Tomasz Kania · 2025
Read Injective envelopes of C ∗ C^{*} -algebras as maximal rigid multiplier covers by Tomasz Kania (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-24441.txt.
Introduction
Tomasz Kania writes alone on the structure of C-star algebras. These algebras blend algebra with limits and star symmetry. They sit at the heart of quantum theory and modern analysis. Hamana built for each such algebra a kind of best outer shell. That shell, the injective envelope, adds room without adding distortion.
Kania looks at covers, larger algebras that hold the first one faithfully. Some covers are rigid, meaning maps on them are forced to be trivial. He proves Hamana's shell is the largest rigid cover of all. He also shows any largest rigid cover must match that shell. For algebras of functions on a space, this gives a classic cover. The result recasts a deep construction in clean order terms.
Catalog record
| CREATOR | Tomasz Kania |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 19 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-24441
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Injective envelopes of C ∗ C^{*} -algebras as maximal rigid multiplier covers Thanks: Institute of Mathematics, Czech Academy of Sciences (RVO: 67985840). Funding received from NCN Sonata-Bis 13 (2023/50/E/ST1/00067) is acknowledged with thanks. Tomasz Kania Abstract . We present a noncommutative analogue of Błaszczyk’s elegant two-step construction of the Gleason cover: first maximise irreducibility, then compactify . The key insight is that extremal disconnectedness (respectively, injectivity) emerges naturally from a maximality principle rather than being imposed by hand. We work with A A -multiplier covers : non-unital C ∗ C^{*} -algebras E E equipped with a faithful, nondegenerate ∗ * -homomorphism ι : A → M ( E ) \iota:A\to M(E) . Covers are ordered by A A -preserving ucp \mathrm{ucp} maps between multiplier algebras in the forward direction. A cover is rigid if every ucp \mathrm{ucp} endomorphism of M ( E ) M(E) fixing A A pointwise is the identity. We prove that if ( E max , ι max ) (E_{\max},\iota_{\max}) is maximal among rigid covers, then the inclusion A ↪ M ( E max ) A\hookrightarrow M(E_{\max}) is rigid and essential; consequently, I ( A ) ≅ M ( E max ) I(A)\ \cong\ M(E_{\max}) canonically over A A by Hamana’s characterisation. In the commutative case A = C ( X ) A=C(X) , we recover Błaszczyk’s picture: for a maximal regular refinement ( X , T ∗ ) (X,T^{\ast}) with irreducible identity, G ( X ) = β ( X , T ∗ ) G(X)=\beta(X,T^{\ast}) and I ( C ( X ) ) ≅ C ( G ( X ) ) ≅ M ( C 0 ( U ) ) I(C(X))\cong C(G(X))\cong M(C_{0}(U)) for any dense open U ⊆ G ( X ) U\subseteq G(X) . Thus the paradigm maximise first, then compactify provides a unified conceptual framework on both sides of Gelfand duality. 2020 Mathematics Subject Classification Primary 46L05; Secondary 54G05, 46L30 1. Introduction The injective envelope of a C^{*} -algebra, introduced by Hamana in his foundational papers [ 5 , 6 ] , has become a central tool in noncommutative analysis. For a compact Hausdorff X , injectivity on the commutative side amounts to extremal disconnectedness: I(C(X))\cong C(K) with K extremally disconnected. Gleason’s classical construction [ 4 ] of K is intricate; Błaszczyk [ 1 ] later gave a strikingly concise route: first maximise the regular topology under an irreducibility constraint, then compactify to obtain G(X) . Indeed, in Błaszczyk argument, maximality is the driver and extremal disconnectedness is the consequence. Our aim is to transpose this to the noncommutative setting. The multiplier algebra M(E) is the natural analogue of the Čech–Stone compactification: M(C_{0}(Y))\cong C(\beta Y) . We introduce A -multiplier covers (E,\iota) and a rigidity notion paralleling Hamana’s. The punchline is that a maximal rigid cover forces M(E) to be a rigid essential extension of A , hence identifies canonically with I(A) . Dictionary. Topology C^{\ast} -algebra irreducible surjection Y\twoheadrightarrow X rigid/essential embedding A\hookrightarrow B extremally disconnected compact K injective C^{\ast} -algebra B Gleason cover G(X) injective envelope I(C(X)) Čech–Stone compactification Y\mapsto\beta Y multiplier algebra E\mapsto M(E) maximal regular refinement T^{\ast} maximal rigid A -multiplier cover (E_{\max},\iota_{\max}) {\lx@inpgf@ignorespaces(X,T^{\ast})} {\lx@inpgf@ignorespaces G(X)=\beta(X,T^{\ast})} {\lx@inpgf@ignorespaces(X,T)} \scriptstyle{\lx@inpgf@ignorespaces\beta} \scriptstyle{\lx@inpgf@ignorespaces\mathrm{id}\ \text{irreducible}} \scriptstyle{\lx@inpgf@ignorespaces q} {\lx@inpgf@ignorespaces E_{\max}} {\lx@inpgf@ignorespaces M(E_{\max})} {\lx@inpgf@ignorespaces I(A)} \scriptstyle{\lx@inpgf@ignorespaces\cong} Organisation. Section 2 recalls Hamana’s characterisation of I(A) and the multiplier-ideal technology we need. Section 3 introduces multiplier covers, defines the forward order by A - preserving \mathrm{ucp} maps, and proves that maximality forces essentiality via a lifting argument through I(A) . It also contains an intrinsic transfinite existence proof of maximal rigid covers using minimal A^{1} - projections and the Choi-Effros product. The main identification M(E_{\max})\cong I(A) is established in Section 4 , and Section 5 revisits the commutative picture, recovering Błaszczyk. We end with remarks on functoriality and local multipliers. 2. Preliminaries 2.1. Hamana’s injective envelope An operator system \mathsf{E} is injective if any \mathrm{ucp} map \phi:\mathcal{S}\to\mathsf{E} from an operator system \mathcal{S}\subseteq\mathcal{T} extends to a \mathrm{ucp} map \tilde{\phi}:\mathcal{T}\to\mathsf{E} . Theorem 2.1 . For every operator system A there exists an injective operator system I(A) and a unital complete order embedding j:A\hookrightarrow I(A) such that: (a) if A\subseteq\mathsf{E}\subseteq I(A) with \mathsf{E} injective, then \mathsf{E}=I(A) ; (b) if \Phi:I(A)\to I(A) is \mathrm{ucp} and \Phi|_{A}=\mathrm{id} , then \Phi=\mathrm{id} . Moreover, I(A) is unique up to a unique complete order isomorphism over A . If A is a C^{\ast} -algebra, then I(A) is a C^{\ast} -algebra and j is a * -monomorphism. Equivalently, I(A) is the unique rigid essential extension of A : whenever q:I(A)\to\mathsf{F} is a complete quotient map and q|_{A} is a complete isometry, then q is injective. 2.2. Multiplier algebras and quotienting ideals If E is a (not necessarily unital) C^{\ast} -algebra, its multiplier algebra M(E) is the C^{\ast} -algebra of double centralisers; see Busby [ 2 ] or Pedersen [ 8 , Ch. 3] . If Y is locally compact Hausdorff, then M(C_{0}(Y))\cong C_{b}(Y)\cong C(\beta Y) . We shall use the following standard fact. Lemma 2.2 . Let E be a C^{\ast} -algebra and I\lhd M(E) a closed two-sided ideal. Put K\ :=\ \overline{IE+EI}\ \lhd\ E. Then the quotient map q_{K}:E\to E/K induces a surjective * -homomorphism \pi_{K}:\ M(E)\longrightarrow M(E/K) whose kernel contains I . Conversely, for a closed ideal K\lhd E , the strict extension M(E)\twoheadrightarrow M(E/K) has kernel \{m\in M(E):mE\subseteq K,\ Em\subseteq K\} . Proof. See [ 8 , §3.12] and [ 2 ] . For x\in I one has xE,Ex\subseteq K , hence x\in\ker\pi_{K} . The converse kernel description is the usual characterisation of multipliers vanishing modulo K . ∎ Remark 2.3 . If B is unital and J\lhd B , the canonical map B\to M(J) given by left/right multiplication is a surjective * -homomorphism with kernel \mathrm{Ann}(J):=\{b\in B:bJ=Jb=0\} ; hence M(J)\ \cong\ B/\mathrm{Ann}(J). Thus M(J)=B iff J is essential in B . In particular, even if B is a W^{*} -algebra (e.g. a product of factors), M(J) need not be W^{*} , because \mathrm{Ann}(J) need not be weak ∗ -closed. This is consistent with Hamana’s examples where I(A) is monotone complete but not von Neumann [ 7 ] . 3. Multiplier covers and maximality Definition 3.1 . Let A be a C^{*} -algebra. An A -multiplier cover is a pair (E,\iota) where E is a (not necessarily unital) C^{*} -algebra and \iota:A\to M(E) is a faithful, nondegenerate * -homomorphism; we identify A with \iota(A)\subseteq M(E) . The cover is rigid if every \mathrm{ucp} map \Phi:M(E)\to M(E) with \Phi|_{A}=\mathrm{id}_{A} equals the identity. Given two A -multiplier covers (E_{1},\iota_{1}) and (E_{2},\iota_{2}) , we write (E_{1},\iota_{1})\ \preccurlyeq\ (E_{2},\iota_{2}) if there exists an A -preserving \mathrm{ucp} map \Theta:M(E_{1})\to M(E_{2}) such that \Theta\circ\iota_{1}=\iota_{2} . A rigid cover maximal with respect to \preccurlyeq is called a maximal rigid A -multiplier cover . Intrinsic existence via minimal A^{1} -projections Throughout this subsection we embed A faithfully and nondegenerately into \mathcal{B}(H) and write A^{1} for its unitisation inside \mathcal{B}(H) . By an A^{1} -map we mean a unital completely positive map fixing A^{1} pointwise. An A^{1} -projection is an idempotent A^{1} -map. We endow the space of maps \mathcal{B}(H)\to\mathcal{B}(H) with the point-ultraweak topology. In what follows, the