A Spectral Theorem for Imprimitivity C*-bimodules

After recalling in detail some basic definitions on Hilbert C*-bimodules, Morita equivalence and imprimitivity, we discuss a spectral reconstruction theorem for imprimitivity Hilbert C*-bimodules over commutative unital C*-algebras and consider some of its applications in the theory of commutative full C*-categories.

💡 Research Summary

The paper provides a detailed spectral reconstruction theorem for imprimitivity Hilbert C*-bimodules over commutative unital C*-algebras and explores its consequences for the theory of commutative full C*-categories. After a thorough review of the basic notions—Hilbert C*-bimodules, internal inner products, completeness, left and right actions, Morita equivalence, and the definition of imprimitivity—the authors focus on the case where the underlying algebras are commutative and unital. By the Gelfand–Naimark correspondence, such algebras can be written as C(X) and C(Y) for compact Hausdorff spaces X and Y.

The central result, called the “Spectral Reconstruction Theorem,” asserts that any imprimitivity C*-bimodule E over C(X)–C(Y) is completely determined by a homeomorphism φ : X → Y and a complex line bundle L over the graph Γφ ⊂ X × Y. More precisely, E is isomorphic (as a bimodule with its left and right C*-valued inner products) to the space C(Γφ, L) of continuous sections of L. The left C(X)-action corresponds to pointwise multiplication by functions on the first coordinate, while the right C(Y)-action corresponds to multiplication on the second coordinate. The proof proceeds by first extracting a *‑isomorphism between the algebras from the bimodule inner products, translating this algebraic isomorphism into the topological homeomorphism φ, and then constructing L by local trivializations of E along φ. An explicit isometry between E and the section space is then exhibited, verifying that the bimodule structure is preserved.

Having identified the geometric data (φ, L) underlying an imprimitivity bimodule, the authors turn to the associated linking (or transfer) C*-algebra T(E) = K(E) ⊕ C(X) ⊕ C(Y). The spectrum of T(E) is shown to be the closed relation given by Γφ, and the K‑theory of T(E) decomposes into contributions from the K‑theories of C(X) and C(Y) together with the class of L in the Picard group of Γφ. This description clarifies how Morita equivalence between C(X) and C(Y) is encoded in topological data and how invariants such as dimension functions or index pairings are transferred across the bimodule.

The paper then situates these results within the framework of commutative full C*-categories. In such a category, objects are algebras of the form C(X) and 1‑morphisms are precisely imprimitivity bimodules. By the spectral theorem, every 1‑morphism can be identified with a line bundle over the graph of a homeomorphism between the underlying spaces. Consequently, the categorical composition corresponds to the fiberwise tensor product of line bundles, and categorical equivalences reduce to homeomorphisms together with isomorphism classes of line bundles. This yields a concrete topological model for the entire category, allowing one to read off categorical invariants (e.g., automorphism groups, natural transformations) directly from the underlying spaces and their Picard groups.

Finally, the authors discuss prospects for extending the theorem beyond the commutative setting. They suggest that replacing the Gelfand spectrum with the primitive ideal space of a non‑commutative C*-algebra and replacing line bundles with Hilbert bundles of modules over the primitive spectrum could lead to a non‑commutative analogue of the spectral reconstruction. Such a generalization would potentially provide a new geometric perspective on Morita equivalence and imprimitivity in the broader C*-algebraic landscape.

In summary, the work bridges operator‑algebraic notions of imprimitivity with classical topological concepts of homeomorphisms and line bundles, delivering a clear geometric picture of Morita equivalence for commutative C*-algebras, elucidating the structure of linking algebras, and furnishing a concrete model for commutative full C*-categories. This synthesis not only deepens our understanding of existing theory but also opens avenues for future research in non‑commutative geometry and categorical operator algebras.