Density-Functional Theory Foundations

Introduction

An interacting many-electron wavefunction depends on the coordinates of all electrons, making it difficult to represent directly. Ground-state density-functional theory asks whether the three-dimensional electron density n(r) can instead serve as the basic variable for the ground-state energy. Hohenberg and Kohn established the formal foundation, and Kohn and Sham supplied a practical auxiliary-orbital construction. The theory is exact in principle for its specified ground-state problem, but practical calculations rely on approximate exchange–correlation functionals. That distinction is central when interpreting a DFT prediction.

Core explanation

The electron density n(r) describes how much electronic number density is present around each position r. For an N-electron system, integrating n(r) over all space gives N. It does not specify individual electron paths, and an orbital picture is one way of constructing it rather than the only physical description. The 1964 Hohenberg–Kohn paper established, under its ground-state conditions, that the density determines the external potential up to an additive constant, and hence the ground-state properties of the interacting electron system. A variational principle then identifies the correct ground-state density by minimizing an energy functional.

The formal statement does not hand us the exact functional for a complicated molecule. In particular, electron kinetic energy and electron–electron interaction effects are difficult to express directly in terms of density. The Kohn–Sham construction introduces an auxiliary system of noninteracting electrons represented by one-electron orbitals, chosen to reproduce the interacting system's ground-state density under the formal theory. This makes a large portion of kinetic energy manageable through orbitals. The remaining difference, together with exchange and correlation effects beyond the explicitly treated classical Coulomb term, is placed in an exchange–correlation energy functional E xc[n].

The energy is often presented schematically as E[n] = T s[n] + V ext[n] + J[n] + E xc[n]. T s is kinetic energy of the auxiliary noninteracting system, V ext is attraction to the nuclei or other external potential, and J is the classical electrostatic self-interaction of the density. E xc contains the missing quantum exchange and correlation pieces, including the difference between true and auxiliary kinetic energies. A specific approximation to E xc is what usually distinguishes practical Kohn–Sham DFT methods. Saying “DFT includes correlation” is therefore incomplete: an approximate functional models it, sometimes well and sometimes poorly for the chosen problem.

Like Hartree–Fock, Kohn–Sham calculations are self-consistent. A trial density generates an effective potential, the one-electron equations yield orbitals, and their occupations generate a new density. Iteration continues until the input and output densities agree to numerical tolerances. Convergence proves only that the chosen approximate equations were solved consistently. It does not prove the functional captures dispersion, charge transfer, a transition state or a strongly correlated electronic state accurately. Basis and grid quality introduce additional independent numerical questions.

Kohn–Sham orbitals are useful computational objects and can support chemical interpretation, but they should not automatically be equated with observed ionization energies or optical excitation energies. Their eigenvalues arise from the auxiliary effective potential. Special exact-theory relations exist for particular frontier quantities under stated assumptions, yet ordinary approximate orbital gaps often differ from measured excitation gaps. A DFT total-energy difference between properly defined states is usually a more direct route to a ground-state reaction energy than subtracting arbitrary orbital eigenvalues.

Ground-state DFT's formal foundation also does not make every application a ground-state problem. Excited states, temperature-dependent ensembles, open shells and periodic solids have additional formulations or approximations. Geometry optimization with a chosen ground-state functional is a prediction within that functional's domain. To compare with a laboratory enthalpy or rate, add relevant nuclear-motion, thermal, solvent and kinetic considerations. The next pages examine functional families and their characteristic limitations.

Step-by-step reasoning

1. Define nuclei, electron count, charge, spin and the desired ground-state observable. 2. Choose a practical exchange–correlation approximation and an appropriate spatial basis or grid. 3. Solve the Kohn–Sham density–potential–orbital loop self-consistently. 4. Check numerical convergence, occupation and stability of the intended electronic state. 5. Form the relevant total-energy or property difference, including needed geometry and thermal treatment. 6. Assess functional, basis and physical-model errors separately using benchmarks relevant to the question.

Visual explanation

Draw a many-electron wavefunction box with coordinates for many electrons and an arrow toward a three-dimensional density n(r). Then draw a loop: trial density → effective Kohn–Sham potential → auxiliary orbitals → reconstructed density → convergence check. Put E xc[n] as a labeled component of the potential and energy, with an asterisk noting “approximated in practical calculations.” This diagram makes the density foundational while showing why orbitals still appear in most computational implementations.

Real-world analogy

Imagine predicting a crowd's total distribution in a city without following every person's detailed history. A density map can answer many planning questions, and an auxiliary model can generate a similar map more easily. The analogy is limited: electrons obey quantum mechanics, exchange effects and correlations cannot be inferred from an ordinary crowd map, and the exact functional is a mathematical result rather than a simple population statistic.

Real-world example

A materials chemist computes the relative energies of two crystal structures using the same functional and numerical settings. The Kohn–Sham loop converges for both, and the calculated total-energy difference favors one structure. Before claiming it is the observed room-temperature phase, the chemist checks k-point and basis convergence, vibrational free-energy effects and whether the functional reproduces related structural energy differences. Self-consistency supplies a stable model result, not a complete phase diagram.

Why?

Why introduce orbitals if DFT is based on density? The exact ground-state density principle does not provide a convenient explicit expression for the kinetic energy of interacting electrons. The Kohn–Sham auxiliary orbitals make a large kinetic-energy contribution calculable while still using density to define the target system and the remaining functional terms. Thus orbital use is part of a practical construction, not a reversal of the density-based foundation.

Common misconception

“DFT is automatically exact because its theorem is exact.” The theorem supports an exact functional in principle; real calculations use approximations to exchange–correlation energy. Another error is thinking the Kohn–Sham orbitals are the actual interacting many-electron wavefunction. A third is reading a converged DFT calculation as a benchmark for every reaction class. Finally, a small Kohn–Sham orbital gap is not automatically an experimental excitation energy.

Worked example

Suppose a hypothetical DFT calculation yields E(A) = −100.200 and E(B) = −100.195 hartree for two states at the same defined geometry convention. The model predicts B above A by 0.005 hartree, about 13.1 kJ mol⁻¹. If a larger basis changes this difference to 12.5 kJ mol⁻¹, basis sensitivity appears small for a broad 10 kJ mol⁻¹ distinction. But if a second plausible functional predicts B below A by 2 kJ mol⁻¹, functional choice dominates the conclusion. The correct report would flag the disputed ordering rather than claiming DFT's formal exactness resolves it. These energies are invented for reasoning practice.

Quick check

1. What quantity is fundamental to ground-state DFT in the Hohenberg–Kohn formulation? Answer: The ground-state electron density, a function of position that integrates to the total electron number. 2. Why is E xc an approximation in most practical calculations? Answer: The exact exchange–correlation functional is not available in a convenient general form, so a model functional is selected.

Exam focus

State what density n(r) represents and distinguish a functional from an ordinary function. Summarize the Hohenberg–Kohn ground-state idea under its conditions and the purpose of auxiliary Kohn–Sham orbitals. Name the main schematic energy terms, identifying exchange–correlation as the practical approximation. Explain why numerical SCF convergence and formal exactness of the theory do not prove an approximate calculation is chemically accurate.

Advanced insight

The exact Kohn–Sham system reproduces density, not necessarily every excited-state or orbital property of the interacting system. Approximate functionals may also violate conditions satisfied by the exact functional, leading to systematic errors such as delocalization or missing long-range dispersion. Error cancellation can make a functional successful for one reaction class while failing another. Good validation therefore measures the target observable over representative systems and reports the functional name, basis, integration settings and state definitions.

Summary

Ground-state DFT uses electron density as its central variable. Kohn–Sham theory provides auxiliary orbitals to construct that density and evaluate much of the kinetic energy. Practical accuracy depends heavily on an approximate exchange–correlation functional and on numerical and physical-model choices. A converged DFT result is a prediction under those choices, not an automatic consequence of an exact theorem.

Practice questions

1. What does the integral of n(r) over all space equal for an N-electron system? Answer: It equals N, the total number of electrons. 2. Are Kohn–Sham orbitals themselves the exact many-electron wavefunction? Answer: No. They belong to an auxiliary one-electron construction that produces the target density in the formal theory. 3. Which part of the schematic Kohn–Sham energy expression carries the key practical approximation? Answer: The exchange–correlation functional E xc[n]. 4. If two functionals reverse a small energy ordering while basis convergence is good, what uncertainty deserves attention? Answer: Functional or method error, because the predicted ordering depends on the exchange–correlation approximation.