Data-driven concavity engineering for magnetic activation in rigid microspheres

Concave architectures harness geometric curvature to amplify interfacial activity and local fields, presenting encouraging prospects for catalytic, optical, and electromagnetic applications. However, their controllable synthesis in rigid metals remains challenging due to pronounced anisotropic surface energy and crystallographic constraints. Departing from conventional thermodynamically driven mechanisms, we propose a broadly applicable stress manipulation strategy during droplet evaporation to induce and regulate surface concavity in rigid microspheres. Such deliberately engineered concave configuration, guided by machine learning optimization, enables enhanced magnetic activity in electromagnetic systems. Reprogrammed magnetic domains orchestrate localized energy redistribution and magnetic activation, meanwhile effectively reducing spatial occupancy. The optimal FeCo alloys outperformed commercial magnetic powders, demonstrating a 28% increase in high-frequency magnetic response and a 79.5% reduction in density, while maintaining high saturation magnetization (236.6 emu/g). This mechanism successfully decouples magnetism from density, allowing for better design of lightweight, multifunctional materials.