The Born–Oppenheimer approximation is an approximation in quantum mechanics that separates the motion of electrons from that of atomic nuclei in a molecule or solid. Because nuclei are much heavier than electrons, their motion often occurs on longer timescales. Electronic states can therefore be calculated with the nuclei held at fixed positions, after which nuclear motion is treated on the resulting electronic energy surface. The approximation underlies much of molecular structure theory, spectroscopy, and computational chemistry, but becomes unreliable when coupling between electronic states substantially affects nuclear motion.
Physical basis
A molecular quantum state depends on the positions of all its electrons and nuclei. Solving their motion simultaneously is difficult because the particles interact through electrostatic forces. The Born–Oppenheimer approximation organizes this problem around the large difference in their masses: even a proton is about 1,836 times as massive as an electron.
For many bound molecular states, this mass difference supports a hierarchy of characteristic energies and timescales. Electronic excitations generally involve higher energies than molecular vibrations, while rotational level spacings are often smaller still. This hierarchy is not universal, but it helps explain why electronic, vibrational, and rotational structure can often be analyzed separately.
The familiar statement that “electrons adjust instantaneously to the nuclei” is a useful shorthand, not a literal mechanism. More precisely, the approximation assumes that nuclear motion does not produce appreciable transitions out of the electronic state being followed. Its accuracy depends not only on nuclear masses but also on electronic energy gaps and on how strongly electronic states change with nuclear geometry.
Nuclei are fixed only during the electronic calculation. They may subsequently be treated quantum mechanically or, with an additional approximation, using classical mechanics.
Mathematical formulation
Fixed-nuclei electronic problem
In a nonrelativistic description without external fields, the molecular Hamiltonian can be written schematically as
where denotes all electronic coordinates, all nuclear coordinates, and is the nuclear kinetic energy operator:
Here is the mass of nucleus , and is the reduced Planck constant. The electronic Hamiltonian contains electronic kinetic energy and the electron–electron, electron–nucleus, and nucleus–nucleus Coulomb interactions. Including nuclear repulsion in is convenient because its eigenvalues then give the full fixed-nuclei potential energy.
At each nuclear geometry, the electronic Schrödinger equation is solved:
The nuclear positions enter as parameters rather than electronic dynamical variables. Each eigenvalue defines a potential energy surface.
Separation of the molecular wave function
For a single electronic state, the molecular wave function is approximated by
where describes nuclear motion.
Neglecting terms arising from nuclear derivatives of the electronic wave function gives
The nuclei therefore move in an effective potential generated by the electronic calculation. This equation can describe nuclear vibration, rotation, tunneling, and reaction dynamics without explicitly evolving the electrons.
This is an approximate dynamical separation, not an assertion that electrons and nuclei are independent. The electronic factor still depends parametrically on nuclear positions.
Couplings omitted by the approximation
A more general expansion is
With a complete electronic basis, including continuum contributions where needed, this is a representation of the full molecular wave function rather than an approximation. Substitution into the molecular Schrödinger equation produces coupled nuclear equations:
with
where
The brackets indicate integration over electronic coordinates. These terms arise because nuclear kinetic energy differentiates the electronic factor as well as the nuclear factor.
Off-diagonal terms couple nuclear motion on different electronic surfaces. The simplest Born–Oppenheimer treatment neglects them and also omits diagonal derivative corrections. More refined single-surface treatments retain some diagonal terms while still neglecting transitions between surfaces.
Potential energy surfaces and molecular structure
A potential energy surface connects electronic calculations to familiar chemical concepts:
- Local minima represent stable or metastable molecular geometries on a specified electronic surface.
- Curvature near a minimum determines harmonic vibrational frequencies when combined with nuclear masses.
- Saddle points can characterize transition-state geometries along a chemical reaction pathway.
- Separated-fragment limits describe dissociation channels and their relative energies.
For a diatomic molecule, the internal potential depends only on internuclear distance and can be represented by a curve. For a nonlinear molecule with nuclei, removing overall translation and rotation generally leaves internal coordinates; a linear molecule has .
A surface may be calculated on a grid, fitted to an analytical expression, or evaluated as nuclear motion proceeds. The resulting nuclear problem is often simplified further using a harmonic oscillator model or other approximations. Such choices are separate from the Born–Oppenheimer approximation itself.
Historical development and terminology
Max Born and J. Robert Oppenheimer introduced the approximation in 1927, shortly after the development of modern quantum mechanics. Their treatment used the electron-to-nuclear mass disparity to organize molecular energy contributions and explain the approximate separation of electronic, vibrational, and rotational motion.
The later Born–Huang formulation made explicit the expansion over multiple electronic states and the derivative couplings between them. It provides a standard framework for identifying terms omitted in a single-surface description.
Terminology is not completely uniform. Born–Oppenheimer approximation often means the simplest product ansatz with derivative terms neglected. Adiabatic approximation may mean neglecting intersurface transitions while retaining a diagonal correction. In electronic-structure discussions, however, these terms are sometimes used more loosely.
Here, “adiabatic” refers to following an electronic state as nuclear coordinates change. It is distinct from an adiabatic process in thermodynamics, which involves no heat transfer.
Applications
Electronic structure and chemical bonding
Holding nuclei fixed makes it possible to calculate electronic energies, charge distributions, and chemical bonding for specified geometries. Molecular orbital theory and many practical electronic-structure methods are commonly used within this framework.
Two different sources of error must be distinguished:
- error from the Born–Oppenheimer separation;
- error in the approximate solution of the electronic problem.
An inaccurate treatment of electron correlation or an inadequate electronic basis can give a poor potential energy surface even when the separation of electronic and nuclear motion is otherwise appropriate.
Spectroscopy and isotope effects
Nuclear energy levels on an electronic surface provide the basis for interpreting vibrational and rotational spectra, including infrared and Raman spectra.
At the fixed-nuclei, nonrelativistic Born–Oppenheimer level, different isotopes with the same nuclear charge share the same electronic surface. Their nuclear masses nevertheless produce different vibrational frequencies, rotational constants, and zero-point energies. Precision calculations can require corrections beyond this shared-surface description.
Molecular dynamics and reaction pathways
In Born–Oppenheimer molecular dynamics, the electronic problem is solved at successive nuclear configurations. In the simplest classical treatment, forces on nuclei are obtained from
The nuclei then follow classical trajectories. Classical nuclear motion is an additional approximation: Born–Oppenheimer separation itself does not exclude quantum nuclear effects such as tunneling.
Potential energy surfaces also support calculations of reaction pathways, barriers, and rate-related properties in chemical kinetics.
Solids
In condensed-matter physics, an analogous separation supports calculations of equilibrium crystal structures, lattice vibrations, and phonons. Electrons are treated as responding to ionic configurations. Effects involving substantial electron–phonon energy exchange or rapidly changing electronic occupations may require treatments beyond a static single-surface picture.
Breakdown and corrections
Small electronic energy gaps
The mass disparity alone does not guarantee accuracy. For distinct, nondegenerate electronic states, derivative coupling satisfies
A small energy gap can therefore produce strong coupling, although symmetry and the numerator also matter. Whether transitions occur additionally depends on nuclear motion through the coupled region.
The Born–Oppenheimer approximation is consequently vulnerable near avoided crossings and conical intersections, where electronic surfaces approach closely or become degenerate. Such regions are important in photochemistry and in the redistribution of electronic excitation into nuclear motion.
Diagonal corrections and geometric phase
For a normalized, nondegenerate electronic state that can locally be chosen real, a common diagonal Born–Oppenheimer correction is
It corrects the effective nuclear potential for part of the effect of the electronic wave function’s nuclear-coordinate dependence. It does not restore transitions between electronic states.
Near a conical intersection, an electronic state can also acquire a Berry phase when nuclear coordinates traverse a closed loop. This can affect nuclear interference and boundary conditions. A globally valid description may therefore need geometric information beyond an ordinary scalar potential surface.
Beyond single-surface dynamics
When transitions between electronic states are important, nonadiabatic dynamics methods evolve electronic and nuclear degrees of freedom together. Approaches include coupled quantum wave-packet propagation, surface-hopping methods, and mean-field electronic–nuclear dynamics, each with its own approximations.
No single mass ratio or energy-gap threshold determines validity in every system. Accuracy depends on the electronic states involved, nuclear masses and velocities, the region of configuration space explored, and the observable being calculated. An approximation adequate for an equilibrium geometry may be inadequate for an excited-state lifetime or a precision spectroscopic transition.