Coulomb Operator¶

The Hartree–Fock one-electron operator that averages the pairwise Coulomb kernel over an occupied-orbital density and multiplies a test orbital by the resulting local repulsive potential.

Version: v2 · 2026-08-30 · History
Domain-specific #: 1571
Origin domain: quantum chemistry
Subdomain: Hartree–Fock electronic structure theory
Aliases: Direct Coulomb Operator

Core Idea¶

In Hartree–Fock electronic-structure theory, the Coulomb operator is the one-electron direct-interaction operator obtained by averaging the electron–electron kernel over an occupied orbital's probability density. In atomic units, for a fixed normalized spin orbital $\chi_j(x_2)$, define

\[ J_j(x_1) = \int \chi_j^*(x_2)\frac{1}{r_{12}}\chi_j(x_2)\,dx_2 = \int \frac{|\chi_j(x_2)|^2}{r_{12}}\,dx_2 . \]

Its action on a test spin orbital $f$ is multiplication by that potential:

\[ [\hat J_j f](x_1)=J_j(x_1)f(x_1). \]

Outside atomic units, the kernel carries the electrostatic constant $e^2/(4\pi\epsilon_0 r_{12})$. The integral removes the source electron's coordinate $x_2$ while retaining the acted-on coordinate $x_1$. Thus a two-particle repulsion is converted into a one-particle mean field: the electron described by $f$ experiences the average repulsive potential generated by the probability distribution $|\chi_j|^2$.^[1]^[2]

The Coulomb operator is a constitutive term of the Hartree–Fock Fock operator. In spin-orbital notation,

\[ \hat f(1)=\hat h(1)+\sum_{j\in\mathrm{occ}}\bigl(\hat J_j(1)-\hat K_j(1)\bigr), \]

where $\hat h$ contains one-electron kinetic and nuclear-attraction terms and $\hat K_j$ is the exchange operator. Closed-shell spatial-orbital notation commonly writes $\hat f=\hat h+\sum_a(2\hat J_a-\hat K_a)$ because each occupied spatial orbital carries two electrons; density-matrix and spin conventions change visible coefficients without changing the retained identity.^[3]^[4]

The node survives as an autonomous domain-specific abstraction because its roles are stable across analytical derivations, basis-set implementations, restricted and unrestricted Hartree–Fock formulations, and neighboring mean-field methods. It is not merely “an operator involving Coulomb's law.” It is the precise reduction source orbital density + pairwise kernel → local mean potential → one-electron multiplication operator, used as the direct part of a self-consistent electronic-structure equation.

Structural Signature¶

The Coulomb operator has the following mandatory roles:

The source density. A normalized occupied spin orbital contributes $\rho_j(x_2)=|\chi_j(x_2)|^2$; a summed form uses an electron density or density matrix built from the occupied orbital set.
The Coulomb kernel. The interaction between coordinates is $r_{12}^{-1}$ in atomic units, or $e^2/(4\pi\epsilon_0r_{12})$ with SI constants.
The contraction over the source coordinate. Integrating $\rho_j(x_2)/r_{12}$ over $x_2$ yields the potential $J_j(x_1)$ at the retained coordinate.
The acted-on one-electron function. The operator accepts an orbital or other suitable one-particle test function $f(x_1)$.
Local multiplicative action. Once the source density is fixed, $\hat J_j$ acts by $f(x_1)\mapsto J_j(x_1)f(x_1)$.
Direct mean-field role. The operator represents the classical-looking direct electron–electron repulsion component of the Fock operator, paired with but distinct from exchange.
Self-consistent dependence. The source orbitals that define $J$ are themselves solutions of equations containing $J$; an SCF procedure updates the density, rebuilds the operator, and repeats until the input and output orbitals agree.^[1]^[5]

For fixed $\chi_j$, $\hat J_j$ is linear in the acted-on function:

\[ \hat J_j(af+bg)=a\hat J_jf+b\hat J_jg. \]

It is nevertheless a functional of the source orbital or density. Changing $\chi_j$ changes the multiplier $J_j(x_1)$, so the full map from the current orbital set to the next Fock operator is nonlinear and motivates self-consistency.

The corresponding Coulomb integral is

\[ J_{ij}=\langle \chi_i|\hat J_j|\chi_i\rangle = \iint \frac{|\chi_i(x_1)|^2|\chi_j(x_2)|^2}{r_{12}}\,dx_1dx_2 =(ii|jj). \]

It is symmetric in $i,j$ and nonnegative for the repulsive kernel. The recognition invariant is the operator construction, not merely the scalar integral: the density is first contracted into a potential that then acts on a one-electron function.

What It Is Not¶

Not the bare two-electron repulsion operator. $r_{12}^{-1}$ acts on a function of two electron coordinates. $\hat J_j$ is a one-electron mean-field operator obtained after integrating that kernel against a source density.
Not the exchange operator. $\hat K_j$ acts as $[\hat K_j f](x_1)=\chi_j(x_1)\int \chi_j^*(x_2)\frac{f(x_2)}{r_{12}}\,dx_2.$ The test function appears inside the integral and the source orbital appears at the output coordinate, so exchange is nonlocal. Coulomb is multiplication by a local potential for fixed density.^[1]
Not the whole Fock operator. The Fock operator combines the core Hamiltonian with sums of Coulomb and exchange contributions. $J$ is one component.
Not the core Hamiltonian. Kinetic energy and electron–nuclear attraction belong to $\hat h$; $J$ represents direct electron–electron repulsion.
Not a generic electrostatic potential. Nuclear Coulomb attraction, an externally imposed field, and the potential of a classical charge distribution use the same kernel but do not by themselves instantiate the occupied-orbital Hartree–Fock operator.
Not a compact operator by definition. “Compact operator” names a functional-analytic property. Multiplication by a Coulomb potential is not classified as this node because of compactness, and singularity/domain questions depend on the function space and Hamiltonian setting.
Not electron correlation beyond mean field. Hartree–Fock includes direct Coulomb and exchange effects within a single determinant but omits the remaining correlation energy. The Coulomb operator does not encode instantaneous correlated avoidance between electrons.
Not an unrestricted self-repulsion claim. A Coulomb term for an orbital can appear algebraically, but the corresponding exchange term cancels same-spin self-interaction in Hartree–Fock. Interpreting $J_i$ alone as a physical electron repelling itself ignores the paired structure.

Scope of Application¶

The home scope is nonrelativistic molecular and atomic Hartree–Fock theory, including restricted, unrestricted, and generalized spin-orbital formulations. The operator also appears as the direct or Hartree contribution in closely related self-consistent-field and hybrid electronic-structure methods. Its matrix representation is central to finite-basis implementations, while real-space and numerical-orbital programs construct the same mean potential in different computational forms.^[4]^[5]

In a finite atomic-orbital basis $\{\phi_\mu\}$, the Coulomb contribution is commonly assembled from the density matrix $P$ and electron-repulsion integrals:

\[ J_{\mu\nu} = \sum_{\lambda\sigma} P_{\lambda\sigma}(\mu\nu|\lambda\sigma), \]

with the exact factors depending on whether $P$ is a spin density, total closed-shell density, or another declared convention. The basis representation does not create a new abstraction: it is the matrix of the same density-contracted Coulomb operator.^[3]

The scope does not include every object called a “Coulomb operator” across physics. A truncated Coulomb interaction in periodic simulation, a relativistic Dirac–Coulomb Hamiltonian, a many-particle Coulomb interaction operator, or a Coulomb Green operator may use overlapping words and kernels while having different domains and roles. This node retains the quantum-chemistry Fock-operator meaning fixed by the source article and authoritative textbooks.

Clarity¶

The decisive diagnostic is which coordinate remains after integration. The bare pair interaction $r_{12}^{-1}$ retains both electron coordinates. The orbital-specific Coulomb operator integrates coordinate 2 against $|\chi_j(x_2)|^2$, leaving a scalar potential of coordinate 1. That potential then multiplies whatever function the operator acts on. If both coordinates remain, one still has a two-electron operator; if the test orbital remains inside the integral and a different orbital appears outside, one has exchange.

“Local” also needs precision. $\hat J_j$ is local as an operator in coordinate representation because its output at $x_1$ is $J_j(x_1)f(x_1)$; it does not require values of $f$ at other points. But computing $J_j(x_1)$ requires integrating the source density over all $x_2$. Local operator action is therefore not the same as a local-density approximation or short-range interaction.

Finally, distinguish the operator from its expectation value. $J_{ij}$ is a scalar repulsion integral between two orbital densities. $\hat J_j$ is the function-mapping object whose expectation in $\chi_i$ produces that scalar. This distinction matters in SCF equations, where the operator must act on candidate orbitals before energies are extracted.

Manages Complexity¶

The exact electronic Hamiltonian contains a sum of pairwise interactions $\sum_{i<j}r_{ij}^{-1}$ on a many-electron wavefunction. Hartree–Fock restricts the wavefunction to a Slater determinant and variationally turns the many-body problem into coupled one-electron equations. The Coulomb operator is the direct half of that compression: for each electron, the detailed coordinates of the others are replaced by their average orbital densities. The resulting field can be combined with the core and exchange operators in a Fock eigenproblem.^[6]^[1]

This reduction does not make the computation trivial. The source density depends on the unknown orbitals, so one first guesses orbitals or a density matrix, builds $J$ and $K$, solves the Fock equations, constructs a new density, and iterates. Basis-set implementations contract a four-index electron-repulsion tensor or an approximation to it with the density matrix. Integral screening, density fitting, multipole methods, and related algorithms reduce cost, but they preserve the defining contraction.^[3]^[5]

The abstraction therefore manages two forms of complexity at once. Conceptually, it replaces explicit pair configurations by a mean repulsive field. Computationally, it isolates a reusable density-to-potential-to-matrix operation that electronic-structure codes can optimize independently from diagonalization and exchange construction.

Abstract Reasoning¶

Several properties follow directly from the formula. For normalized $\chi_j$ and nonnegative Coulomb kernel, $J_j(x_1)\ge0$ and

\[ \langle f|\hat J_j|f\rangle = \iint \frac{|f(x_1)|^2|\chi_j(x_2)|^2}{r_{12}}\,dx_1dx_2 \ge 0. \]

Thus the direct term is repulsive in the usual electronic Hamiltonian sign convention. Because $J_j$ is real for an ordinary density, its multiplication operator is symmetric on an appropriate domain. These statements concern the Coulomb term, not the complete Fock energy, which also contains attraction and exchange.

The density dependence licenses additivity. If a total source density is a sum $\rho=\sum_j n_j|\chi_j|^2$, then the corresponding potential is

\[ J[\rho](x_1) = \int\frac{\rho(x_2)}{r_{12}}dx_2 = \sum_j n_jJ_j(x_1). \]

This explains both the orbital sum in the Fock operator and the density-matrix contraction in a basis. It also exposes convention hazards: occupation numbers, spin sums, and whether $P$ already includes a factor of two must be declared before coefficients are compared.

Self-consistency can be predicted from the same structure. If the orbitals change, their density changes; then $J$ changes; then the Fock eigenvectors generally change again. A calculation that builds $J$ once from an arbitrary initial density and never updates it does not solve the Hartree–Fock stationary equations unless it happens already to be self-consistent.

Knowledge Transfer¶

Within electronic-structure theory, the construction transfers literally across basis sets and numerical representations. Gaussian orbitals, Slater orbitals, finite elements, plane waves, or grids change how the integral is evaluated, not the role chain. Restricted and unrestricted theories redistribute spin and occupation factors, while the direct density contraction remains.

The same density-to-Hartree-potential construction occurs in Kohn–Sham density-functional theory and other mean-field equations, often under the names Hartree potential, direct potential, or Coulomb contribution. That is a legitimate within-domain transfer when the source density, Coulomb kernel, local multiplication action, and self-consistent role are explicit. It should not erase method boundaries: the exchange-correlation operator in Kohn–Sham DFT is not the Hartree–Fock exchange operator, and the full Fock and Kohn–Sham operators are not aliases.

Outside quantum many-body modeling, the portable skeleton is already captured by Aggregation and Transformation: integrate a distributed source through a kernel, then use the resulting field as an operator. Calling a demographic average, network influence score, or smoothed image a “Coulomb operator” merely because it aggregates distant contributions would be metaphorical. The domain-specific name requires electron-density and Coulomb-repulsion semantics.

Examples¶

Orbital-specific direct field. Given a normalized occupied spin orbital $\chi_j$, construct $J_j(x_1)$ by integrating $|\chi_j(x_2)|^2/r_{12}$. Acting on two different trial functions $f$ and $g$ multiplies each by the same $J_j(x_1)$ as long as the source orbital remains fixed. This makes the fixed-source operator linear even though rebuilding it from new orbitals is part of a nonlinear SCF loop.

Two occupied spin orbitals. For a determinant containing $\chi_1$ and $\chi_2$, the equation for $\chi_1$ contains the direct field of $\chi_2$. Algebraically one may sum over all occupied $j$ because $(J_1-K_1)\chi_1=0$: the same-spin self-Coulomb and self-exchange actions cancel. This is why inspecting $J$ alone and declaring physical self-repulsion is misleading.^[1]

Closed-shell spatial-orbital convention. In restricted Hartree–Fock, each occupied spatial orbital represents an alpha and a beta spin orbital. The direct contribution appears with an occupation factor, commonly $2J_a$, while exchange acts only between like-spin components and appears as $K_a$. For a doubly occupied helium-like spatial orbital, the formal combination leaves the mean repulsion associated with the other, opposite-spin electron. The example shows why formula coefficients must be read together with the spin and density convention.

Finite-basis SCF build. From current orbital coefficients construct $P_{\lambda\sigma}$, contract it with $(\mu\nu|\lambda\sigma)$ to obtain $J_{\mu\nu}$, combine core, Coulomb, and exchange matrices into $F$, solve the Roothaan–Hall generalized eigenproblem, and rebuild $P$. The Coulomb operator is the same abstraction before and after discretization; $J_{\mu\nu}$ is its basis representation.^[3]^[6]

Structural Tensions¶

Mean-field tractability versus correlated motion. Averaging the source electron over its orbital density yields coupled one-particle equations but discards instantaneous dynamical correlation. Post-Hartree–Fock methods repair what the direct-plus-exchange single-determinant field omits.
Local action versus global construction. $J$ multiplies the test function locally at $x_1$, yet its potential requires the source density over all space through the long-range kernel. “Local” describes operator action, not cheap or spatially local construction.
Linearity versus self-consistency. For fixed density, $J$ is a linear operator on test functions. As part of Hartree–Fock, it depends on the orbitals being solved for, making the full stationary problem nonlinear.
Physical interpretation versus algebraic partition. The direct term has a classical electrostatic interpretation, but only the combined Hartree–Fock expression handles antisymmetry and same-spin self-interaction cancellation. Individual operator terms should not be reified independently of their pairing.
Exact kernel versus approximate build. The formal operator uses $1/r_{12}$ exactly within the nonrelativistic model. Practical codes may approximate its matrix through screening, density fitting, grids, or multipoles. An approximation remains a Coulomb build only if it targets the same density contraction and controls the introduced error.
Notation economy versus convention drift. $J_j$, $J$, $J_{\mu\nu}$, Hartree potential, and direct matrix can refer to orbital, total-density, operator, potential, or basis representations. Occupation and spin factors prevent unqualified formulas from being safely interchanged.

Structural–Framed Character¶

Assessment: strongly structural (0.93 structural / 0.07 framed). The defining integral, source density, retained coordinate, multiplicative action, and Fock-operator role give an unusually exact recognition test. Given the same orbitals and physical constants, independent derivations produce the same operator up to notation and representation.

The small framed component concerns model and convention choices: restricted versus unrestricted orbitals, spin-orbital versus spatial-orbital notation, basis representation, density normalization, boundary conditions, relativistic modifications, and numerical approximation. These alter factors or implementation but do not turn the node into a socially framed category. The classification remains domain-specific because its literal roles belong to quantum electronic-structure theory, not because its internal definition is vague.

Structural Core vs. Domain Accent¶

The portable structural core is a transformation built by kernel aggregation: a distributed source is integrated against a distance-dependent kernel to produce a field, and that field acts on an input function. Transformation captures the rule-governed map; Aggregation captures the contraction of a density; Superposition supports the sum of orbital contributions.

The domain accent is the identity. The source is an occupied electron orbital or electron density, the kernel is Coulomb repulsion, the output is a local one-electron potential, the acted-on object is an orbital, and the operator enters a self-consistent Fock equation alongside exchange. Removing those roles yields a general potential operator or integral construction, not the Coulomb operator retained here.

This is why the candidate is not a prime. The same mathematical skeleton appears in gravitation, electrostatics, and kernel methods, but those are neighboring instantiations of broader abstractions. Literal Coulomb Operator recurrence stays within quantum many-electron mean-field theory.

Transformation — instantiated literally. For fixed source density, $\hat J$ is a rule-governed linear mapping from a one-electron test function to another function. It is the recommended minimal parent.
Aggregation — the source density is integrated over one coordinate to obtain a lower-dimensional potential field.
Superposition — total Coulomb fields and matrices are sums of orbital or density-matrix contributions because the Coulomb contraction is linear in density.
Iteration and Feedback — in SCF solution, orbitals generate $J$, the resulting Fock operator generates new orbitals, and convergence seeks a fixed point.
Approximation — the mean-field replacement and practical integral approximations must be distinguished: the former defines Hartree–Fock's model class, while the latter approximates computation of a fixed model.

Only the single strict placement under Transformation is proposed in this bundle. The other relations are explanatory and do not justify extra edges.

Relationships to Other Abstractions¶

Current abstraction Coulomb Operator Domain-specific

Parents (1) — more general patterns this builds on

Coulomb Operator is a kind of Transformation Prime

instantiated literally.

Hierarchy path (1) — routes to 1 parentless root

Coulomb Operator → Transformation → Function (Mapping)

Neighborhood in Abstraction Space¶

Coulomb Operator sits in a sparse region of the domain-specific corpus (94^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Collective Dynamics & Molecular Operators (6 abstractions)

Nearest neighbors

Roothaan–Hall Equations — 0.81
Eigenstate Thermalization Hypothesis — 0.77
Hydrogen Bond — 0.76
AKLT Model — 0.76
Fredholm Kernel — 0.76

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Exchange Operator. Exchange has the source orbital outside and the acted-on function inside the integral. It is nonlocal, depends on same-spin antisymmetry, and lacks the Coulomb operator's simple classical charge-density interpretation. The two appear together in Hartree–Fock and cancel same-spin self-action on an occupied orbital.

Fock Operator. The Fock operator is the full effective one-electron operator $h+J-K$ under the declared spin convention. Coulomb is its direct electron–electron component.

Hartree potential. The total-density potential $v_H(\mathbf r)=\int\rho(\mathbf r')/|\mathbf r-\mathbf r'|d\mathbf r'$ is structurally the summed Coulomb multiplier. Usage overlaps substantially, especially in DFT. Hartree potential is best reviewed as a scoped related surface rather than an unrestricted alias for the orbital-specific $J_j$ operator.

Coulomb integral. $(ii|jj)$ or $J_{ij}$ is a scalar matrix element of $\hat J_j$ in orbital $i$, not the operator itself. Electron-repulsion integrals $(\mu\nu|\lambda\sigma)$ are basis primitives contracted to build a Coulomb matrix.

Two-Electron Coulomb Operator. The pairwise $r_{12}^{-1}$ term acts on a two-particle wavefunction and is part of the exact many-electron Hamiltonian. The Hartree–Fock Coulomb operator is a one-body mean field derived from it.

Coulomb Hamiltonian or Dirac–Coulomb Hamiltonian. These are full many-particle Hamiltonians with kinetic, external, and interaction terms, possibly relativistic. They are not aliases for $\hat J$.

Compact Operator. Compactness is a topology-dependent property of a linear map between normed spaces. The quantum-chemical name Coulomb operator identifies a kernel-derived physical role, not that functional-analytic class.

References¶

[1] C. David Sherrill, An Introduction to Hartree–Fock Molecular Orbital Theory, Georgia Institute of Technology (2000), especially equations 20–28. https://vergil.chemistry.gatech.edu/static/content/hf-intro.pdf registry ↩a ↩b ↩c ↩d ↩e

[2] Attila Szabo and Neil S. Ostlund, Modern Quantum Chemistry: Introduction to Advanced Electronic Structure Theory, Dover Publications (1996 reprint), Chapter 3. ISBN 978-0-486-69186-2. https://store.doverpublications.com/products/9780486691862 registry ↩

[3] Q-Chem, Q-Chem User's Manual, section “Hartree–Fock Calculations,” equations 4.14–4.24. https://manual.q-chem.com/4.3/sect0037.html registry ↩a ↩b ↩c ↩d

[4] Trygve Helgaker, Poul Jørgensen, and Jeppe Olsen, Molecular Electronic-Structure Theory, Wiley (2000), Chapter 10, “Hartree–Fock Theory.” https://doi.org/10.1002/9781119019572.ch10 registry ↩a ↩b

[5] Susi Lehtola, Frank Blockhuys, and Christian Van Alsenoy, “An Overview of Self-Consistent Field Calculations Within Finite Basis Sets,” Molecules 25(5), 1218 (2020). https://doi.org/10.3390/molecules25051218 registry ↩a ↩b ↩c

[6] C. C. J. Roothaan, “New Developments in Molecular Orbital Theory,” Reviews of Modern Physics 23, 69–89 (1951). https://doi.org/10.1103/RevModPhys.23.69 registry ↩a ↩b