Loading field map
The central challenge of electronic and magnetic materials research is to understand how atomic and electronic structure gives rise to electrical conductivity, magnetism, and related phenomena, and to use that understanding to design materials with tailored properties. Over the past century, the explanatory frameworks used to meet this challenge have undergone profound shifts—from classical field theories to quantum mechanics, from localized to itinerant electron pictures, and from a focus on charge alone to the inclusion of spin and topological order. Each new framework emerged by addressing a limitation of its predecessors, and today several frameworks coexist, each best suited to a different class of materials or problem.
The first systematic attempts to explain electronic and magnetic behavior relied on classical physics. Classical Electromagnetism and Lorentz Electron Theory (1890–1920) treated electrons as tiny charged particles moving in response to applied fields, providing a qualitative account of optical dispersion and the Hall effect. However, it could not explain the temperature dependence of conductivity or the existence of ferromagnetism. The Drude Model (1900–1930) extended this picture by introducing collisions between electrons and a stationary lattice, yielding the first microscopic theory of electrical and thermal conduction in metals. Drude’s model successfully predicted the Wiedemann–Franz law but failed at low temperatures and could not account for the sign of the Hall coefficient in many metals. Weiss Molecular Field Theory (1907–1930) addressed ferromagnetism by postulating an internal “molecular field” that aligns atomic magnetic moments. This phenomenological approach explained the Curie temperature and hysteresis, but left the origin of the molecular field mysterious. All three frameworks were macroscopic and classical; they provided useful engineering approximations but lacked the quantum underpinnings needed for predictive power.
The advent of quantum mechanics transformed the subfield, but it also introduced a lasting division between two competing pictures of electrons in solids: delocalized band states and localized atomic orbitals.
Band Theory of Solids (1928–present) replaced the classical free-electron model by showing that electrons in a periodic lattice occupy energy bands separated by gaps. This framework explained why some materials are metals, others insulators, and still others semiconductors, and it provided the foundation for understanding doping and carrier transport. Band theory treats electrons as extended plane waves weakly perturbed by the lattice—a delocalized picture that works well for simple metals and conventional semiconductors.
At the same time, Heisenberg Exchange Interaction Model (1928–present) offered a quantum explanation for ferromagnetism: the exchange interaction between neighboring localized spins. This model gave a microscopic basis for Weiss’s molecular field and correctly predicted the existence of spin waves. It treated magnetic moments as firmly attached to atomic sites, contrasting sharply with band theory’s delocalized electrons.
Heitler–London Localized-Moment Model (1930–1970) narrowed the Heisenberg approach by applying valence-bond theory to solids, emphasizing that electrons in magnetic insulators are tightly bound to individual atoms. This model excelled for materials like transition-metal oxides but could not describe metallic ferromagnets where electrons roam freely.
Stoner Model for Itinerant Ferromagnetism (1938–present) took the opposite stance: it treated ferromagnetism as a band-splitting phenomenon in a metal, where the exchange interaction shifts the spin-up and spin-down bands relative to each other. The Stoner model successfully described weak ferromagnets like nickel and palladium alloys, but it overestimated Curie temperatures and failed for strongly correlated systems.
Hubbard Model (1963–present) attempted to bridge the localized and itinerant pictures. By adding a strong on-site Coulomb repulsion term to a tight-binding band model, Hubbard captured the transition from metallic to insulating behavior as correlations increase. The Hubbard model became the canonical framework for strongly correlated electron systems, and its extensions—such as dynamical mean-field theory—remain central to modern computational materials science. It did not replace band theory or the Heisenberg model but instead revealed a continuum between them, with the ratio of interaction strength to bandwidth as the key parameter.
The invention of the transistor in 1947 launched the Semiconductor Band Engineering Paradigm, which applies band theory to design electronic devices. This framework treats semiconductors as materials whose band gaps, doping levels, and heterojunctions can be precisely controlled to manipulate charge carriers. It is an engineering-oriented extension of band theory, focusing on practical outcomes such as diodes, lasers, and integrated circuits. The semiconductor paradigm coexists with more fundamental models; it rarely concerns itself with strong correlations or magnetism, instead assuming that band theory is sufficient for the materials it targets (silicon, III–V compounds, etc.). Its success has made it the dominant framework in industrial research, but it leaves open questions about correlated oxides, magnetic semiconductors, and quantum materials.
Beginning in the late 1980s, researchers began to treat spin and topology as design variables on par with charge, opening new frontiers.
Spintronics (1988–present) emerged from the discovery of giant magnetoresistance, which exploits the spin of electrons to control current. This framework adds the spin degree of freedom to the semiconductor paradigm, enabling devices such as spin valves and magnetic tunnel junctions. Spintronics does not reject band theory but extends it by incorporating spin-dependent transport and magnetization dynamics.
Multiferroics Paradigm (2003–present) goes further by coupling magnetic order with ferroelectricity, allowing one ferroic property to control the other. This framework challenges the traditional separation of magnetic and dielectric materials and seeks materials where multiple order parameters coexist and interact. Its conceptual innovation is the deliberate design of cross-coupling between ferroic orders, a goal that requires going beyond single-order-parameter models.
Topological Insulators and Weyl Semimetals Paradigm (2005–present) introduced a new classification of electronic phases based on topological invariants rather than band gaps alone. Topological insulators are insulating in the bulk but conduct on their surfaces via protected states, while Weyl semimetals host exotic quasiparticles that mimic relativistic fermions. This framework transforms the band-theory picture by showing that some materials cannot be adiabatically connected to ordinary insulators without closing the gap. It has revived interest in band topology and provided a fresh lens for understanding quantum Hall effects, spin–orbit coupling, and surface states.
Today, no single framework dominates the entire subfield. Band theory remains the workhorse for conventional semiconductors and simple metals, and it is the starting point for most electronic-structure calculations. Hubbard-model extensions (especially dynamical mean-field theory) are essential for strongly correlated materials such as high-temperature superconductors and heavy-fermion compounds. Spintronics and topological materials are active frontiers where new phenomena are discovered regularly, and multiferroics is a growing niche for multifunctional materials.
What the leading frameworks agree on is that quantum mechanics is indispensable, that electronic structure determines properties, and that computational methods (density functional theory, DMFT, etc.) are powerful tools for prediction. They disagree on the relative importance of correlations versus band structure, on whether localized or itinerant descriptions are more fundamental for a given material, and on whether topological classification will eventually subsume traditional band-insulator categories or remain a special case. The Hubbard model and band theory are in living disagreement over how to treat intermediate-correlation systems, while topological frameworks coexist with both by adding a new dimension to the classification.
This pluralism is a sign of maturity: the subfield now has a toolbox of frameworks, each calibrated for a specific range of materials and questions. The challenge for students is not to choose one correct theory but to understand which framework applies when, and how they can be combined to design the next generation of electronic and magnetic materials.