| Rigged Hilbert space | |
|---|---|
| Name | Rigged Hilbert space |
| Field | Mathematics; Quantum mechanics |
| Introduced | 1950s–1960s |
| Inventor | Israel Gelfand and collaborators |
| Related | Hilbert space; Schwartz space; distributions |
Rigged Hilbert space
A Rigged Hilbert space is a mathematical construction that extends a Hilbert space by adjoining spaces of test functions and distributions to accommodate non-normalizable states and generalized eigenvectors. It provides a framework to treat continuous spectra and resonances rigorously in quantum mechanics, making precise manipulations that appear in physicists' bra–ket notation and spectral decompositions.
A rigged Hilbert space formalizes the intuition behind Dirac's bra–ket formalism and the need to handle objects like plane waves and delta-normalized states which are not elements of the standard Hilbert space of square-integrable functions. The construction remedies pathologies encountered when applying the spectral theorem to operators with continuous spectrum, such as the momentum operator or the Hamiltonian of scattering systems. It is motivated by works of Paul Dirac, John von Neumann, and later mathematical formalizations by Israel Gelfand and collaborators.
Formally a rigged Hilbert space is a triple Φ ⊂ H ⊂ Φ′, often called a Gelfand triple or Gelfand triple. Here Φ is a dense subspace of the separable Hilbert space H endowed with a finer, nuclear topology; Φ′ denotes the continuous dual of Φ, the space of continuous linear functionals on Φ, equipped with the strong dual topology. Typical choices make Φ into a nuclear space so that the inclusion Φ → H is continuous and the embedding H → Φ′ identifies H with its anti-dual. The Gelfand triple enables duality pairings ⟨φ, F⟩ between φ ∈ Φ and F ∈ Φ′ that extend the inner product of H.
Key mathematical ingredients were developed in the context of functional analysis and distribution theory; the theory draws on concepts from topological vector space theory and the study of unbounded operators. The nuclearity condition ensures desirable mapping and tensor product properties used in spectral expansions.
In the rigged Hilbert space, operators that are symmetric or self-adjoint on H can admit generalized eigenvectors in Φ′ corresponding to points of the continuous spectrum. These generalized eigenvectors are distributional objects analogous to Dirac deltas or plane waves; they satisfy eigenvalue equations in the weak sense: for a densely defined operator A, ⟨Aφ, F⟩ = λ⟨φ, F⟩ for all φ ∈ Φ. This framework makes rigorous the use of bras and kets |λ⟩ and ⟨λ| as elements of Φ′ and the space of continuous linear functionals on Φ, respectively.
Prominent mathematical results include expansions of vectors in terms of generalized eigenvectors and the justification of resolution of the identity over continuous spectra using distributional measures. These results connect to the spectral measures from the spectral theorem but extend them to accommodate non-L^2 objects.
Rigged Hilbert spaces refine the spectral analysis of self-adjoint operators such as the Schrödinger operator and the momentum operator. In scattering theory and quantum dynamics, the Gelfand triple allows one to represent the continuous part of the spectrum by a continuous basis of functionals and to derive time evolution formulas for scattering states. The approach complements and clarifies the Stone's theorem on one-parameter unitary groups and functional calculus for unbounded operators by specifying domains and dual actions explicitly.
Mathematical physicists applied rigged Hilbert space techniques to justify the manipulations in textbooks by L. D. Landau, R. Shankar, and others, and to relate rigorous spectral decompositions to formal expansions used in quantum field theory and non-relativistic quantum mechanics.
Rigged Hilbert spaces have proven useful in the mathematical treatment of scattering theory, providing a setting for incoming and outgoing states as elements of distinct distribution spaces and enabling a rigorous formulation of the S-matrix and wave operators. They are a natural context for defining and analyzing resonances and Gamow vectors—generalized eigenvectors with complex eigenvalues modeling unstable states—popular in studies by the Ludwig Boltzmann-inspired approaches and by researchers such as A. Bohm and collaborators.
In concrete models (e.g., potential scattering), one constructs riggings adapted to the asymptotic behavior to obtain analytic continuation of the resolvent and meromorphic structure associated with resonances. Applications also appear in the mathematical study of quantum scattering theory at institutes like CERN and research groups in mathematical physics.
Common examples of Φ include the Schwartz space S(R^n) of rapidly decreasing smooth functions, the space of compactly supported smooth functions C_c^∞, or Sobolev-type spaces when dealing with differential operators. For the free particle on R^n, the Gelfand triple S(R^n) ⊂ L^2(R^n) ⊂ S′(R^n) (with S′ the space of tempered distributions) is standard and yields plane-wave generalized eigenvectors in S′. For bound-state problems, polynomially weighted test spaces or Hardy-space riggings are used to isolate analytic continuation properties essential for resonance theory.
Rigged Hilbert spaces bridge abstract functional analysis and practical distribution theory by embedding Schwartz distributions into operator-theoretic frameworks suitable for quantum applications. They rely on concepts introduced by Laurent Schwartz (tempered distributions) and extend the Riesz representation theorem context to dual pairings involving non-Hilbert elements. The approach has spawned further developments in spectral theory, nuclear space theory (as in work by Alexander Grothendieck), and modern treatments of operator extensions, domain questions, and perturbation theory in mathematical physics.
Category:Mathematical physics Category:Functional analysis