Abstract
This study systematically benchmarks the performance of four single soft magnetic powders-water-atomized Fe-Si-Cr (FeSiCr), silica-coated reduced iron powder (RIP), silica-coated carbonyl iron powder (CIP), and phosphate-coated CIP (CIP-P)-to establish quantitative relationships between powder attributes, deformation substructure, and high-frequency loss for molded power inductors (100 kHz-1 MHz). We prepared toroidal compacts at 200 MPa and characterized them by initial permeability (μ(i)), core-loss (P(cv)(f)), partitioning (P(cv)(f) = K(h)f + K(e)f(2), K(h), Ke: hysteresis and eddy-current loss coefficients), and EBSD (electron backscatter diffraction)-derived microstrain metrics (Kernel Average Misorientation, KAM; low-/high-angle grain-boundary fractions). Corrosion robustness was assessed using a 5 wt% NaCl, 35 °C, 24 h salt-spray protocol. Our findings reveal that FeSiCr achieves the highest μ(i) across the frequency band, despite its lowest compaction density. This is attributed to its coarse particle size (D(50) ≈ 18 µm) and the resulting lower intragranular pinning. The loss spectra are dominated by hysteresis over this frequency range, with FeSiCr exhibiting the largest K(h), while the fine, silica-insulated Fe powders (RIP/CIP) most effectively suppress K(e). EBSD analysis shows that the high coercivity and hysteresis loss in CIP (and, to a lesser extent, RIP) are correlated with dense, deformation-induced subgrain networks, as evidenced by higher mean KAM and a lower low-angle grain boundary fraction. In contrast, FeSiCr exhibits the lowest KAM, with strain confined primarily to particle contact regions. Corrosion testing ranked durability as FeSiCr ≳ CIP ≈ RIP ≫ CIP-P, which is consistent with the Cr-rich passivation of FeSiCr and the superior barrier properties of the SiO(2) shells compared to low-dose phosphate. At 15 A, inductance retention ranks CIP (67.9%) > RIP (55.7%) > CIP-P (48.8%) > FeSiCr (33.2%), tracking a rise in effective anisotropy and-for FeSiCr-lower M(s) that precipitate earlier roll-off. Collectively, these results provide a microstructure-informed selection map for single-powder formulations. We demonstrate that particle size and shell chemistry are the primary factors governing eddy currents (K(e)), while the KAM-indexed substructure dictates hysteresis loss (K(h)) and DC-bias superposition characteristics. This framework enables rational trade-offs between magnetic permeability, core loss, and environmental durability.