forty years and still no stronger magnet

Problem statement

The strongest commercially available permanent magnets — neodymium-iron-boron (Nd₂Fe₁₄B), discovered in the early 1980s (1982 per Coey's review; announced publicly in 1983–84) — have not been surpassed in maximum energy product (BHmax) since: ARPA-E's own program overview is titled "No New Magnets Developed for Over 40 Years." This performance ceiling limits the power density and efficiency of electric motors, generators, and actuators across transportation, power generation, and industrial applications. Worse, Nd₂Fe₁₄B magnets depend on rare earth elements (neodymium, dysprosium) whose supply chain is concentrated in China — around 60% of global rare earth mining output in 2024, about 91% of separation and refining, and roughly 94% of sintered permanent magnet production (IEA). The compositional space of multi-element magnetic materials is vast but largely unexplored — ARPA-E's program materials describe the Nd₂Fe₁₄B discovery as accidental, and estimate that of roughly 100,000 relevant four-element systems for magnets, perhaps 1% have been examined.

Why this matters

Permanent magnets are critical components in EV motors, wind turbine generators, industrial robots, and defense systems. The global permanent magnet market is roughly $30 billion annually, with NdFeB accounting for about 62% of it by value (per ARPA-E's MAGNITO program overview), and is growing with electrification trends. Every increment of magnet performance translates directly to smaller, lighter, more efficient motors — a stronger magnet enables the same torque from a smaller motor, reducing vehicle weight and energy consumption. The rare earth supply chain vulnerability has been treated as a national and economic security concern by the U.S. government, including in 2025 executive orders on critical minerals. ARPA-E's MAGNITO program — launched August 2025 alongside the ROCKS critical-minerals program under an announcement of $60 million for the two programs — explicitly aims to reinvigorate the materials science of magnets using modern computational and experimental tools, targeting materials with saturation magnetization above 2.5 T and a doubling of BHmax beyond 800 kJ/m³.

What’s been tried and why it hasn’t worked

The ARPA-E REACT (Rare Earth Alternatives in Critical Technologies) program, launched in 2011 with $31.6 million across 14 projects, funded alternatives to rare earth magnets, focusing on substitution strategies (iron nitride, manganese-based compounds, cerium-based compounds) that achieved respectable performance but never matched Nd₂Fe₁₄B's energy product. Iron nitride (α″-Fe₁₆N₂) has a giant saturation magnetization (~2.9 T) and a theoretical energy product far exceeding Nd₂Fe₁₄B's, but scalable bulk magnets have not been achieved — demonstrations remain limited to thin films, foils, and compacted powders performing well below the theoretical values (Wang, 2020). Manganese-based alloys (Mn–Bi, Mn–Al) have been developed as rare-earth-free "gap magnets," but their energy products remain far below Nd₂Fe₁₄B's. The fundamental challenge is that the physics of permanent magnetism requires a rare combination of properties: high saturation magnetization, high coercivity, and high anisotropy, usually achievable only with specific crystal structures that are difficult to predict and harder to synthesize in bulk. Traditional materials discovery approaches screened candidate compositions one at a time, barely scratching the surface of the multi-element composition space.

What would unlock progress

Modern computational tools (density functional theory at scale, machine learning interatomic potentials) can now screen millions of candidate compositions for magnetic properties, dramatically expanding the search space. High-throughput synthesis and characterization methods (diffusion couples spanning composition gradients, combinatorial thin-film approaches, Kerr microscopy for rapid magnetic screening) can survey many compositions in parallel. MAGNITO's program materials highlight the ternary-to-quinary (three-to-five-element) composition space — hundreds of thousands of largely unexamined systems — where emergent magnetic properties might arise from complex crystal structures that were previously too difficult to predict or discover by intuition. The convergence of computational prediction, high-throughput experimentation, and rapid magnetic-property characterization creates a window for systematic magnet discovery.

Entry points for student teams

A team could use computational screening (DFT or ML-based surrogate models) to identify candidate ternary or quaternary compositions with high predicted magnetization, then compare predictions against known experimental data to validate the approach. Alternatively, a team could survey the patent and literature landscape for iron nitride synthesis methods and identify the specific thermodynamic barriers to bulk Fe₁₆N₂ production. Materials science, computational physics, and condensed matter physics skills are central.