Hubbard model
/ HUB-erd /
Picture a board game played on a grid of dots, where tokens can hop from one dot to a neighbor, but with one house rule: if two tokens ever land on the same dot, you pay a penalty. That's almost the entire rulebook. Despite being this bare-bones, the game produces astonishingly rich and unpredictable patterns — and that is exactly the appeal.
The Hubbard model strips the physics of interacting electrons down to just two ingredients on a lattice of atoms. First, a 'hopping' term that lets an electron jump to a neighboring atom, which tends to spread electrons out and make a metal. Second, an 'on-site repulsion' that charges an energy cost whenever two electrons sit on the same atom, which tends to lock electrons in place. The whole behavior of the model is set by the tug-of-war between these two numbers.
It matters because this toy captures the essence of strongly correlated materials: tune the ratio of the two terms and the model can describe a metal, a Mott insulator, magnetism, and is suspected to hold the secret of high-temperature superconductivity. The honest caveat: despite its simplicity, the two-dimensional Hubbard model has never been solved exactly — it remains one of the great open problems of physics.
Physicists now build real Hubbard models out of ultracold atoms: lasers trap atoms in a grid of light, the atoms hop between sites and feel a repulsion when they meet, and researchers watch the magnetism and Mott-insulating states emerge directly — a tabletop simulation of the equation no computer can fully solve.
Ultracold atoms in a laser grid act out the Hubbard model that no computer can solve exactly.
The model is named for John Hubbard, who wrote it down in the 1960s, though others arrived at similar ideas independently. Its fame comes not from realism — it ignores most of real chemistry — but from being the simplest model that still refuses to give up its secrets.