Loading field map
The central challenge of electronic and magnetic materials engineering has always been the same: how to design a solid so that its response to an electric field, a magnetic field, or both is precisely the one needed for a device. Over the past 120 years, the frameworks that engineers use to meet that challenge have changed dramatically—from the empirical mixing of metals to the quantum-mechanical engineering of electron bands, spin textures, and topological states. Each new framework did not simply discard its predecessors; it preserved, narrowed, or coexisted with them, creating a layered toolkit that today spans metallurgical recipes, solid-state physics, computational simulation, and data-driven discovery.
The first systematic framework for electronic and magnetic materials was Ferromagnetic Alloy Design, which dominated the first half of the twentieth century. Engineers working within this framework treated magnetic performance as a problem of bulk composition and thermomechanical processing. By varying the proportions of iron, nickel, cobalt, and other elements, they produced alloys with useful combinations of permeability, coercivity, and saturation magnetization. The method was largely empirical: a new alloy was cast, rolled, heat-treated, and tested, and the results were codified into composition-property charts. This framework produced the first transformer steels and permanent-magnet alloys, but it offered little understanding of why a given composition worked.
Running alongside Ferromagnetic Alloy Design, but from a completely different intellectual tradition, was Semiconductor Rectifier Physics. Where the alloy designers thought in terms of bulk composition, the rectifier physicists thought in terms of interfaces and barriers. Working with copper-oxide and selenium rectifiers, and later with germanium and silicon point-contact diodes, they developed a framework that explained rectification as a phenomenon of charge-carrier depletion at a metal-semiconductor junction. This framework was narrower in scope—it addressed only the asymmetry of current flow—but it was far more mechanistic. It introduced the concepts of majority and minority carriers, depletion layers, and barrier heights that would later become the foundation of all semiconductor device physics. For the first half of the century, these two frameworks coexisted with almost no overlap, one serving the power and telecommunications industries, the other serving the emerging electronics industry.
The invention of the transistor in 1947 triggered a transformation that reshaped both sides of the field. Semiconductor Device Physics emerged as a comprehensive framework that extended the earlier rectifier physics into a full theory of charge transport in doped crystals. Engineers could now design not just rectifiers but amplifiers, switches, and logic gates by controlling doping profiles, junction geometries, and carrier lifetimes. This framework absorbed the earlier rectifier physics entirely, preserving its core concepts while vastly expanding their range. By the 1960s, semiconductor device physics had become the dominant framework for all electronic materials, and it remains a living tradition today, especially in the design of silicon-based integrated circuits.
On the magnetic side, the same period saw a split that has persisted ever since. Magnetic Domain Theory and Hysteresis Engineering provided a framework that explained the macroscopic magnetic behavior of a material in terms of the nucleation, motion, and pinning of magnetic domain walls. This framework made it possible to engineer hysteresis loops—the relationship between applied field and magnetization—by controlling grain size, inclusions, and crystallographic texture. Out of this framework grew two complementary but rival design philosophies. Hard Magnetic Materials focused on maximizing coercivity and energy product, producing permanent magnets for motors, generators, and speakers. Soft Magnetic Materials focused on minimizing coercivity and hysteresis loss, producing materials for transformer cores, inductors, and magnetic shielding. Both frameworks remain active today, and their rivalry is a productive one: the same domain theory that explains how to pin domain walls in a hard magnet also explains how to let them move freely in a soft magnet.
While domain theory was explaining magnetic behavior at the micrometer scale, a different revolution was taking place at the atomic scale. Band Structure Engineering, which emerged in the 1960s, extended semiconductor device physics by treating the electronic band structure itself as a designable quantity. Instead of simply doping a fixed material, engineers could now grow heterostructures—layers of different semiconductors—to create quantum wells, barriers, and superlattices with tailored band alignments. This framework made possible the high-electron-mobility transistor, the quantum-well laser, and the entire field of optoelectronics. Band structure engineering did not replace semiconductor device physics; it added a new layer of design freedom, allowing engineers to control not just where carriers moved but how they behaved quantum-mechanically.
In the 1980s, a further extension emerged that challenged the primacy of charge as the only relevant electronic degree of freedom. Spintronics introduced the idea that the spin of the electron, not just its charge, could be used to store and process information. The landmark discovery of giant magnetoresistance in 1988 showed that a multilayer structure could change its electrical resistance dramatically depending on the relative alignment of magnetic layers. Spintronics absorbed the domain theory of magnetic materials and the band theory of electronic transport into a single framework, creating devices—spin valves, magnetic tunnel junctions, spin-transfer-torque oscillators—that are now standard in hard-disk read heads and magnetic random-access memory.
Two other frameworks from this period expanded the field in directions that were initially seen as exotic but later became central. High-Temperature Superconducting Materials, discovered in 1986, presented a profound challenge: the cuprate materials conducted electricity without resistance at temperatures far above those predicted by conventional BCS theory. This framework coexists with band structure engineering and spintronics, but it has not been absorbed by them. Instead, it remains a distinct research tradition focused on understanding and exploiting unconventional pairing mechanisms, with applications in high-field magnets, power cables, and quantum sensing. Multiferroic Materials, which emerged in the 1990s, addressed a different kind of coexistence: materials that are simultaneously ferromagnetic and ferroelectric. This framework extends domain theory by asking how magnetic and electric domain structures interact, and it has opened a path toward devices in which an electric field controls magnetism or a magnetic field controls polarization.
By the early 2000s, the scientific frameworks for electronic and magnetic materials were mature, but the process of designing a new material remained slow and expensive. Two new methodological frameworks emerged to address this bottleneck. Integrated Computational Materials Engineering (ICME) treats the design of a material as a multiscale simulation problem, linking quantum-mechanical calculations at the atomic scale to phase-field models at the microstructural scale and finite-element simulations at the device scale. ICME does not replace any of the earlier scientific frameworks; it provides an infrastructure that integrates them. A band-structure calculation, a domain-wall mobility model, and a device-level heat-transfer simulation can all be combined within an ICME workflow to predict how a new material will perform before it is ever synthesized.
Materials Informatics, which emerged at roughly the same time, takes a different approach. Instead of building physics-based simulations, it uses statistical learning and machine learning to discover patterns in large datasets of material compositions, structures, and properties. Materials informatics is not a tool within ICME; it is a parallel methodology that can sometimes bypass physics-based modeling entirely. Where ICME excels at predicting the behavior of well-understood systems, materials informatics excels at exploring uncharted composition spaces and suggesting candidates that no physical model would have predicted. The two frameworks are complementary, and many research groups now use them in combination: informatics to generate hypotheses, ICME to test them with simulation, and then experiment to validate the results.
The most recent frameworks have shifted the focus from designing materials with useful properties to designing materials with exotic quantum states. Two-Dimensional Materials, launched by the isolation of graphene in 2004, created a framework in which materials are reduced to single atomic layers, revealing properties that are absent in the bulk. This framework extends band structure engineering to the ultimate limit of confinement, and it has produced a family of materials—transition-metal dichalcogenides, hexagonal boron nitride, phosphorene—with tunable band gaps, strong excitonic effects, and valley-selective optical properties.
Topological Insulators, which emerged around 2005, challenged a core assumption of band structure engineering: that a material's electronic properties are determined solely by its band gap. Topological insulators have a band gap in their interior but conducting states on their surface that are protected by time-reversal symmetry. This framework introduced a new classification of materials based on topological invariants rather than band gaps, and it has since expanded to include topological semimetals, topological superconductors, and magnetic topological insulators. Topological insulators do not replace band structure engineering; they add a new dimension to it, showing that the topology of the band structure matters as much as its energy gaps.
Quantum Materials, the most recent framework, is the broadest. It encompasses materials in which strong correlations, topology, or quantum fluctuations produce emergent phenomena that cannot be understood within any single earlier framework. Quantum materials include high-temperature superconductors, topological insulators, heavy-fermion systems, and quantum spin liquids. The framework is not a unified theory but a recognition that the old categories—metal, insulator, magnet, superconductor—are no longer sufficient. Quantum materials engineering draws on band structure engineering, spintronics, and multiferroic frameworks simultaneously, and it often requires the computational infrastructure of ICME and the pattern-finding power of materials informatics to make progress.
Today, the leading frameworks coexist in a complex division of labor. Semiconductor device physics remains the workhorse for silicon-based electronics. Band structure engineering dominates optoelectronics and high-speed transistors. Spintronics provides the read heads and memory elements that make modern computing possible. Hard and soft magnetic materials continue to serve power and actuation applications. High-temperature superconductors and multiferroics are active research frontiers with growing commercial niches. ICME and materials informatics are now standard tools in both academic and industrial materials development. Two-dimensional materials, topological insulators, and quantum materials are the most active areas of fundamental research.
The major disagreements are not about which framework is correct but about which approach is most efficient for a given problem. Should a new magnetic material be discovered by high-throughput synthesis and informatics, or by domain-theory-guided simulation within ICME? Should a topological insulator be designed by band-structure engineering of known compounds, or by searching for new topological phases with quantum materials methods? These are not settled questions, and the field is pluralistic: different groups, different applications, and different material classes call for different combinations of frameworks. What unites them is the recognition that the old empirical approach, while still useful, has been supplemented by a layered set of tools that span from the quantum-mechanical to the statistical, from the atomic to the device scale.