| Dirac delta | |
|---|---|
| Name | Dirac delta function |
| Caption | Schematic of delta as limit of peaked functions |
| Domain | Real numbers |
| Codomain | Distributions |
| Introduced | 1926 |
| Author | Paul Dirac |
Dirac delta
The Dirac delta, often called the delta "function", is a distribution that vanishes everywhere except at a single point and integrates to one. In Quantum mechanics and broader Quantum Physics, it is essential for expressing pointlike states, normalization of continuum eigenstates, and representing interactions idealized as localized in space or time. Its use underpins many theoretical constructions from scattering theory to quantum field operators.
The Dirac delta is defined as a linear functional δ on a space of test functions (typically Schwartz space or compactly supported smooth functions) satisfying δ(φ)=φ(0). Formally, it is not a function in the classical sense but a distribution introduced in physics by Paul Dirac and formalized by Laurent Schwartz. Key properties include sifting: ∫_{-∞}^{∞} δ(x-a) φ(x)\,dx = φ(a), scaling δ(ax)=δ(x)/|a|, and translation δ(x-a). Fourier transform relations connect δ with plane waves: the transform of δ is a constant, and inversely the transform of 1 is δ. The delta's derivative δ' acts on test functions via integration by parts, enabling representation of point sources in differential equations such as the Schrödinger equation with delta potentials. Rigorous treatment uses tempered distribution theory and the machinery of functional analysis on spaces like Hilbert space and L^2 space.
In nonrelativistic Quantum mechanics, δ appears in continuum normalization: momentum eigenstates satisfy ⟨p|p'⟩ = δ(p-p') and position eigenstates satisfy ⟨x|x'⟩ = δ(x-x'). These relations, used in conjunction with the Dirac bra–ket notation formulated by Paul Dirac, formalize completeness relations such as ∫ |x⟩⟨x| dx = I. The delta function models idealized point interactions, e.g., the delta potential V(x)=gδ(x), which yields analytically tractable bound and scattering states and serves as a pedagogical model in courses and texts by authors like L. D. Landau and Rudolf Peierls. In the operator formalism, δ distributions mediate the kernel representation of operators: the position representation of an operator Ĥ has kernel H(x,x') often expressed using δ to encode local potentials or contact terms. Use of δ must respect self-adjointness and domain issues in spectral theory and when defining unbounded operators in Hilbert space.
Position eigenkets |x⟩ are not elements of Hilbert space L^2(ℝ^n) but live in a rigged Hilbert space (Gelfand triple) linking the Hilbert space to its dual of distributions. This rigged construction, developed in part through work at institutions like Princeton University and in literature by I. M. Gel'fand, legitimizes δ as a linear functional on test vectors. The spectral decomposition of self-adjoint operators yields projection-valued measures; for continuous spectrum components the density involves δ-distributions. In quantum measurement theory, δ-like projectors represent ideal position measurements; however, realistic measurement apparatuses correspond to approximate δ's given by square-integrable wavepackets, consistent with positive operator-valued measure (POVM) descriptions in quantum information.
Green's functions solve inhomogeneous linear differential equations with source terms represented by δ, e.g., (E - Ĥ)G(x,x') = δ(x-x'). In time-dependent problems, propagators K(x,t;x',t') satisfy (iħ∂_t - Ĥ)K = iħ δ(t-t')δ(x-x'), encoding causality and initial conditions. In scattering theory and computational physics, Green's functions expressed with δ-sources are central to the Lippmann–Schwinger equation and construction of the S-matrix used in scattering experiments at facilities like CERN. Regularization of singular Green's functions often employs distributional identities and careful boundary conditions; textbooks by John R. Taylor and resources from Los Alamos National Laboratory illustrate practical implementations.
Physically, no true pointlike δ exists; it is realized as a limit of sequences such as normalized Gaussians, Lorentzians, or sinc functions. Regularization schemes—cutoffs, mollifiers, or point-splitting—replace δ by smooth approximations to control divergences in computations, a common necessity in quantum field theory renormalization at institutions like CERN or in condensed matter applications at Bell Labs. Experimental realizations approximate δ-like localization via tightly focused laser beams, atomic traps in ultracold atom experiments at laboratories such as Max Planck Institute for Quantum Optics, and scanning tunneling microscopy tips modeling near-point probes. Emphasizing social equity, access to advanced experimental platforms remains concentrated; democratizing tools and data is important for broad participation in foundational experiments.
In measurement theory, δ underlies idealized projective measurements of position and energy eigenvalues in pedagogical treatments. In scattering, δ enforces energy and momentum conservation in transition amplitudes, appearing in Fermi's golden rule and cross section formulas via δ(E_f-E_i). In quantum field theory, δ functions express local interaction vertices, commutation relations [φ(x), π(y)] = iħ δ(x-y), and momentum-space conservation δ^{(4)}(∑ p_i) in perturbative amplitudes computed with Feynman diagrams and regularized by renormalization techniques developed by theorists at Institute for Advanced Study and elsewhere. Careful handling of δ within distributions preserves causality and locality while highlighting the need to address divergences with principled, equitable access to computational resources and transparent methods.
Category:Mathematical physics Category:Quantum mechanics Category:Distributions (mathematics)