A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
The Born-Oppenheimer (BO) approximation separates nuclear and electronic motion by exploiting the large mass difference between electrons and nuclei: nuclei move so slowly relative to electrons that electrons instantaneously adjust to any nuclear configuration. This allows the total molecular wavefunction to be factored into an electronic part (solved for fixed nuclear positions) and a nuclear part (moving on the electronic potential energy surface). The BO approximation is the conceptual foundation for potential energy surfaces, molecular geometry, and most of computational chemistry. It breaks down in cases of closely spaced electronic states (conical intersections) or very fast nuclear dynamics.
Understand the physical reasoning first — electrons move ~1000× faster than nuclei — before tackling the mathematical separation of the Hamiltonian. Then see how the electronic energy as a function of geometry becomes the potential for nuclear motion.
The Schrödinger equation for a molecule is, in principle, a single equation involving all particles — every electron and every nucleus. For even a small molecule like water with 10 electrons and 3 nuclei, this means solving a coupled 39-dimensional problem (three spatial coordinates per particle). This is intractable as written. The Born-Oppenheimer approximation makes chemistry computationally possible by exploiting one physical fact: a proton is roughly 1,836 times heavier than an electron, and heavier nuclei are even more massive. Because of this enormous mass difference, electrons move thousands of times faster than nuclei. From the electrons' perspective, the nuclei are essentially frozen in place; from the nuclei's perspective, the electrons adjust instantaneously to any nuclear rearrangement.
This timescale separation justifies a mathematical factorization. You first freeze the nuclei at some fixed geometry — say, the two oxygen-hydrogen distances and the H-O-H angle in water — and solve the electronic Schrödinger equation for just the electrons in the field of those stationary nuclei. This gives you the electronic energy at that geometry. Then you move the nuclei slightly to a new geometry and solve the electronic problem again. Repeating this for many geometries maps out the potential energy surface (PES): a landscape where the x-axes are nuclear coordinates and the height is the electronic energy. The nuclei then move on this surface according to their own (nuclear) Schrödinger equation or, in many applications, classical Newton's equations.
The concept of a potential energy surface — the central object in all discussions of molecular geometry, reaction paths, and transition states — exists only because of the Born-Oppenheimer approximation. Without it, you cannot separate "the shape of the molecule" from "the behavior of the electrons," because both would be entangled in one inseparable wavefunction. The BO approximation is what allows you to say a molecule "has a geometry" at all, and to draw reaction coordinate diagrams with energy barriers between reactants and products.
The approximation does break down in important cases. When two electronic states come very close in energy at a particular nuclear geometry — a situation called a conical intersection — the electrons can no longer adjust instantaneously because there is no clear "ground state" to relax into. At these points, nuclear and electronic motion become coupled again, and the molecule can jump between electronic states. This breakdown is not a curiosity: it governs photochemistry, vision (the cis-trans isomerization of retinal), and many ultrafast processes. Recognizing where the Born-Oppenheimer approximation holds and where it fails is essential for knowing when standard computational methods will give reliable results and when more sophisticated non-adiabatic treatments are needed.