Search

The metastable α‘‐Fe16N2 phase may have a magnetic moment up to 50% higher than that of pure bulk α‐Fe. This article addresses the following issues. (i) Can epitaxial films of α‘‐Fe16N2 be prepared phase pure? Yes, but there are some doubts. (ii) Can powders of α‘‐Fe16N2 be prepared phase pure? Not yet. (iii) Is the Mossbauer spectrum due to α‘‐Fe16N2, to martensite, or to something else? Most assign it to α‘‐Fe16N2. (iv) What is the specific saturation magnetic moment of α‘‐Fe16N2? Some claim it is close to that of α‐Fe, most claim that it is much larger. (v) Is the high moment due to α‘‐Fe16N2, or to some other phase?

The α‘‐Fe16N2 phase was synthesized by a nitriding, quenching, and tempering process starting from Fe2O3. Acicular γ‐Fe2O3 was first reduced to α‐Fe under H2, then it was converted to γ‐austenite at 650–700 °C by nitriding using a NH3‐H2 mixed gas. A martensitic transformation to α’‐martensite occurred when the sample was quenched in liquid nitrogen. Finally, the α‘‐Fe16N2 phase was formed from the α’‐martensite by tempering in the temperature range 120–200 °C. In order to increase the extent of the martensitic transformation, acicular Fe2O3 particles with 10–15 at % MnO2 were prepared by oxidizing Fe++ and Mn++ in alkaline aqueous solution. When a quenched sample, which contains martensite, α‐Fe, and austenite, was tempered at 200 °C for different times, the magnetic moment first increased (transformation of α’‐martensite to α‘‐Fe16N2); the highest magnetic moment was obtained at about 60 min of tempering. Using even longer tempering times, the magnetic moment decreased (decomposition of α‘‐Fe16N2 to α‐F...

Fe2N particles were prepared by nitriding acicular γ‐Fe2O3 particles (precoated with sodium silicate to prevent the particles from sintering) at about 450 °C for 3–6 h under a mixed stream of H2‐NH3 (1:4 molar ratio). X‐ray diffraction showed mostly Fe4N and small amounts of α‐Fe and Fe2O3, but no Fe8N(α‘‐Fe16N2). The saturation magnetic moment density σS increased to 140–160 emu/g and the coercivity Hc increased to about 1000 Oe. Although the ultimate goal of preparing Fe8N particles was not attained, an impressive combination of magnetization and coercivity was achieved. To obtain Fe8N by chemical doping, the nitrogen content of Fe4N was decreased by coating with metal hydroxides of Ti, Nd, and Sm, and reducing under H2 at 300 °C: the magnetization dropped precipitously, and no Fe8N was formed.

To increase the coercivity of α‐Fe particles, needlelike goethite crystals were first coated with different Ln3+ ions (Ln=La,Nd,Sm,Gd,Dy,Eu) plus B3+, then reduced to α‐Fe under H2. The magnetic moment of the particles depends upon the Ln and B contents and temperature. Both powder x‐ray diffraction and Mossbauer spectra indicate the existence of a crystalline FeO phase; the FeO content decreased as the reduction temperature increased. A phase with no hyperfine splitting was also detected in the Mossbauer spectra, which may be superparamagnetic α‐Fe or amorphous iron oxide; the amount of this phase decreased as the reduction temperature increased. The temperature dependence of the coercivity of these B‐ and Ln‐coated particles was steeper than for particles coated with silica: this suggests a coercivity mechanism other than shape anisotropy.

Elongated metal particles have been prepared by depositing transition metals into the pore channels of anodic aluminum oxide (alumite), with well‐defined pores that build up into uniform and parallel channels. The highest coercivity for freshly prepared α‐Fe particles in the alumite substrate reached 2700 Oe in the easy direction. The remanance curves of alumite films containing electrodeposited metals (Fe and Co) were measured to study the magnetostatic interaction of the particles. The possible reversal mechanisms of the metal particles in the alumite substrates were examined by the rotational hysteresis integral and the angular variation of the coercivity.