Metastable compositionally and magnetically modulated state of Fe-Ni Invar and the associated super-moment dynamics from Mössbauer spectroscopy

• Published in: Journal of magnetism and magnetic materials 1987-03, Vol.66 (1), p.121-152
• Main Authors: Rancourt, D.G., Smit, H.H.A., Thiel, R.C.
• Format: Article
• Language: English
• Subjects: Condensed matter: electronic structure, electrical, magnetic, and optical properties; Cross-disciplinary physics: materials science; rheology; Electron states; Exact sciences and technology; Magnetic resonances and relaxations in condensed matter, mössbauer effect; Materials science; Metals, semimetals and alloys; Mössbauer effect; other γ-ray spectroscopy; Physics; Specific materials
• ISSN: 0304-8853
• DOI: 10.1016/0304-8853(87)90137-5

The presence of magnetic and chemical clustering, in Fe 65Ni 35 Invar, had previously been inferred from various measurements which we review. We independently confirm both types of clustering and measure the associated supermoment dynamics for the first time. The temperature dependence of the fluctuation frequency is understood in terms of inter-cluster surface exchange energy and leads to a spherical cluster radius of 7.2 Å. The observed super-moment excitations are sufficient to explain the anomalous temperature dependence of the magnetization and are slow enough, compared to lattice response times, to cause the magnetovolume anomalies. The magnetic clusters are seen to persist far above T c - at least up to 600 K. Using a two-cluster-type approximation, the mean compositional morphology is obtained as: 39 at% Fe 85Ni 15/61 at% Fe 53Ni 47. Also, the alloy which is most often referred to as “Invar”, is found to contain ⪅ 2 at% of α-Fe precipitate. A change of + 1 at% in the amount of this precipitate is found to cause a change of ≃ + 50 K in the T c measured on first heating. The three most common models for Invar - itinerate model, two γ-state model, and mixed exchange model - are discussed in terms of both data available in the literature and our own results. The mixed exchange approach is shown to be the most reasonable. Its crudeness is argued to arise from two factors: (i) since detailed structural data were not available, it has assumed the wrong fine scale schemical morphology, and (ii) it uses only three values of J ij whereas we expect a broad distribution of exchange energies to arise from a Slater-Bethe-Bozorth-type interaction whose strength is a function of local composition.