| spectral theory | |
|---|---|
| Name | Spectral theory |
| Field | Mathematics, Quantum mechanics |
| Notable people | David Hilbert, John von Neumann, Marshall Stone, Israel Gelfand |
spectral theory
Spectral theory studies the spectrum of linear operators, i.e., the set of values that generalize eigenvalues, and provides tools to decompose operators into simpler components. In the context of Quantum mechanics, it underpins the mathematical formulation of observables, state evolution, and measurement by linking self-adjoint operators to measurable quantities. Spectral methods enable analysis of both discrete and continuous energy levels, essential for understanding atoms, molecules, and scattering phenomena.
Spectral theory originated in functional analysis and operator theory and became central to quantum physics through the work of David Hilbert and John von Neumann. In quantum theory, physical quantities are modeled by linear operators on a Hilbert space, and the spectrum of an operator corresponds to possible measurement outcomes. This correspondence is formalized by the spectral theorem and by von Neumann's mathematical foundations in his book "Mathematical Foundations of Quantum Mechanics". The distinction between point spectrum (eigenvalues), continuous spectrum, and residual spectrum reflects physical categories such as bound states, scattering states, and pathological cases.
An operator in spectral theory typically means a bounded or unbounded linear map on a Hilbert space or Banach space. The spectrum of an operator T, denoted σ(T), consists of complex λ for which T − λI fails to be invertible. The spectrum is partitioned into the point spectrum (eigenvalues), continuous spectrum, and residual spectrum; for self-adjoint operators the residual spectrum is empty. Important operator classes include self-adjoint, unitary, normal, compact, and bounded operators. Key theorems such as the Fredholm alternative and results on compact operators (e.g., eigenvalue accumulation only at zero) guide classification and physical interpretation.
The spectral theorem gives canonical forms for normal and self-adjoint operators, allowing them to be represented as integrals over projection-valued measures. For a self-adjoint operator A, one writes A = ∫ λ dE(λ) where E(·) is a projection-valued measure on the spectrum. This representation enables the Borel functional calculus: for measurable f, define f(A) = ∫ f(λ) dE(λ). The theorem has classical formulations by Marshall Stone (Stone's theorem on one-parameter unitary groups) and von Neumann, and is foundational for expressing time evolution via the unitary group U(t)=e^{-iHt/ħ} generated by the Hamiltonian H. The spectral measure is closely related to the resolution of the identity used in expansions of quantum states.
In canonical quantum mechanics, observables correspond to self-adjoint operators on a Hilbert space (e.g., position operator, momentum operator, spin operators). The spectrum of an observable gives possible measurement outcomes; for example, the discrete spectrum of the hydrogen Schrödinger equation Hamiltonian yields atomic energy levels historically derived by Erwin Schrödinger and computed via separation of variables. Projection operators from the spectral decomposition implement quantum projective measurements in the von Neumann measurement postulate. The spectral approach also clarifies degenerate spectra, commuting families of observables, and simultaneous diagonalization via the functional calculus.
Continuous spectrum components describe extended states and scattering phenomena studied in scattering theory. The Lippmann–Schwinger equation, the concept of asymptotic completeness, and the S-matrix formalism relate spectral properties of Hamiltonians to experimentally observed cross sections. Resonances appear as complex poles of a continued resolvent (or of the S-matrix) and are mathematically treated through analytic continuation, complex scaling (dilation analyticity), or the theory of Gamow vectors. Important contexts include the study of few-body scattering at institutions such as CERN and research by figures like Ludwig Faddeev and Enrico Fermi.
Numerical spectral analysis uses methods from computational linear algebra and numerical analysis to approximate eigenvalues and spectral measures. Techniques include the Rayleigh–Ritz method, finite element method, spectral methods, and matrix diagonalization algorithms like the Lanczos algorithm and Arnoldi iteration. For Schrödinger operators, discretizations (e.g., finite difference, finite element) approximate bound states, while complex absorbing potentials and exterior complex scaling approximate resonances. Software and libraries often used include packages built on BLAS/LAPACK and domain-specific codes in computational chemistry and condensed-matter physics, such as those developed for DFT calculations.
Many quantum observables are unbounded operators requiring careful domain considerations; core results are found in the theory of closed and self-adjoint extensions (e.g., von Neumann's deficiency indices). The abstract framework of C*-algebras and von Neumann algebras generalizes spectral theory, providing tools for statistical mechanics and quantum field theory; the Gelfand–Naimark theorem and Gelfand representation connect commutative C*-algebras to spectral measures. Spectral decomposition in noncommutative settings underlies the classification of states, superselection sectors, and modular theory (Tomita–Takesaki), with applications in algebraic quantum field theory and the study of equilibrium via the KMS states.
Category:Mathematical physics Category:Operator theory