Mott Insulators

Core Idea

A Mott insulator is a material that band theory predicts should be metallic (partially filled band) but is insulating due to strong electron-electron Coulomb repulsion. When the on-site repulsion U exceeds the bandwidth W, it becomes energetically prohibitive for electrons to hop between sites (each hop creates a doubly-occupied site costing energy U), and the system develops a charge gap despite the absence of a band gap. The Mott metal-insulator transition (MIT) can be driven by changing U/W through pressure, temperature, or chemical doping. Mott insulators are the parent compounds of many exotic materials, including cuprate high-T_c superconductors, colossal magnetoresistance manganites, and frustrated quantum magnets.

Explainer

The concept of the Mott insulator represents one of the most important failures — and subsequent triumphs — of theoretical condensed matter physics. Standard band theory, which treats electrons as independent particles moving in a periodic potential, predicts that any material with a partially filled band should be metallic. Yet many transition metal oxides, rare earth compounds, and organic conductors with partially filled bands are insulating. Nevill Mott explained this in the 1940s-60s: when the electron-electron Coulomb repulsion U is large enough compared to the bandwidth W, electrons become localized to avoid the energetic cost of sharing a site, and a correlation-driven gap opens.

The simplest picture uses the Hubbard model at half-filling. Each site has one electron. To conduct, an electron must hop to a neighboring site, creating a doubly-occupied site at cost U. If U >> W (the bandwidth from hopping), this cost is prohibitive and the electrons are stuck — each one pinned to its site. The single band splits into two Hubbard bands: the lower Hubbard band (removing an electron from a singly-occupied site, creating a hole) and the upper Hubbard band (adding an electron to create double occupancy). The gap between them is approximately U - W, and it is a many-body correlation gap, not a single-particle band gap.

The Mott metal-insulator transition occurs when U/W passes through a critical value of order 1. This can be tuned by pressure (increasing t and W by squeezing atoms closer), by temperature (thermal fluctuations can delocalize electrons), by doping (removing or adding electrons from the half-filled configuration), or by chemical substitution (changing U or t). The transition in V_2O_3, the canonical Mott system, is first-order at low temperatures (with hysteresis and a volume collapse) and ends at a critical point around 400 K, above which a continuous crossover replaces the sharp transition. Dynamical mean-field theory (DMFT) provides the modern theoretical framework for the Mott transition, capturing the competition between coherent quasiparticle formation and local moment physics.

Mott insulators are far more than an intellectual curiosity — they are the parent compounds of some of the most technologically important and scientifically puzzling materials. The cuprate high-T_c superconductors (La_{2-x}Sr_xCuO_4, YBa_2Cu_3O_7) are doped Mott insulators: the parent compound is an antiferromagnetic Mott insulator, and doping with holes produces d-wave superconductivity at temperatures up to 130 K. Colossal magnetoresistance manganites are Mott systems where magnetic field-driven delocalization produces enormous resistance changes. Frustrated Mott insulators on triangular and kagome lattices, where antiferromagnetic order is geometrically incompatible, are candidates for quantum spin liquid states — exotic phases with fractionalized excitations and topological order. Understanding Mott physics is thus central to the search for new quantum materials.