Pressure dependence on the remanent magnetization of Fe-Ni alloys and Ni metal

We measured the acquisition of magnetic remanence of iron-nickel alloys (${\mathrm{Fe}}_{64}{\mathrm{Ni}}_{36}$, ${\mathrm{Fe}}_{58}{\mathrm{Ni}}_{42}$, and ${\mathrm{Fe}}_{50}{\mathrm{Ni}}_{50}$) and pure Ni under pressures up to 23 GPa at room temperature. Magnetization decreases markedly for ${\mathrm{Fe}}_{64}{\mathrm{Ni}}_{36}$ between 5 and 7 GPa yet remains ferromagnetic until at least 16 GPa. Magnetization rises by a factor of 2–3 for the other compositions during compression to the highest applied pressures. Immediately upon decompression, magnetic remanence increases for all Fe-Ni alloys while magnetic coercivity remains fairly constant at relatively low values (5–20 mT). The amount of magnetization gained upon complete decompression correlates with the maximum pressure experienced by the sample. Martensitic effects best explain the increase in remanence rather than grain-size reduction, as the creation of single domain sized grains would raise the coercivity. The magnetic remanence of low Ni Invar alloys increases faster with pressure than for other body-centered-cubic compositions due to the higher magnetostriction of the low Ni Invar metals. Thermal demagnetization spectra of ${\mathrm{Fe}}_{64}{\mathrm{Ni}}_{36}$ measured after pressure release broaden as a function of peak pressure, with a systematic decrease in Curie temperature. Irreversible strain accumulation from the martensitic transition likely explains the broadening of the Curie temperature spectra, consistent with our x-ray diffraction analyses.

Figure 1

Magnetic and physical properties of the iron-nickel alloys used in this study. (a) Magnetic susceptibility. (b) Curie temperature trends for bcc, α-${\mathrm{Fe}}_{1-x}{\mathrm{Ni}}_x$ (red curve; Chuang et al. [22]) and fcc, γ-${\mathrm{Fe}}_{1-x}{\mathrm{Ni}}_x$ (blue curve; Crangle and Hallam [7]). Black squares represent the data from our samples derived from the second derivative of the curves in Fig. 1. (c) Magnetization normalized to the initial (room temperature) value versus temperature for the four Fe-Ni metals used in this study. (d) X-ray powder diffraction patterns demonstrating a pure fcc phase for all compositions. The top three patterns were measured with a STOE STADI P diffractometer using Mo-$K{\alpha }_1$ radiation ($\lambda =0.709$Å); bottom two were measured with an Agilent Gemini diffractometer using Mo-$K\alpha$ radiation ($\lambda =0.7107\AA$). Bottommost spectrum was obtained upon pressure release (ambient conditions) from 16.3 GPa.

Figure 2

(a)–(c) Backfield magnetization acquisition curves for three independent experiments on ${\mathrm{Fe}}_{64}{\mathrm{Ni}}_{36}$ powder. (d), (e) Backfield magnetization curves for ${\mathrm{Fe}}_{58}{\mathrm{Ni}}_{42}$ and ${\mathrm{Fe}}_{50}{\mathrm{Ni}}_{50}$. (f), (g) Backfield magnetization acquisition curves for independent experiments on ${\mathrm{Ni}}_{100}$ powder. Some pressure steps are omitted to better distinguish the curves; all pressure steps are given in Table 1.

Figure 3

(a)–(c) Normalized SIRM and (d)–(f) coercivity of remanence (Bcr) as a function of pressure for ${\mathrm{Fe}}_{64}{\mathrm{Ni}}_{36}$, ${\mathrm{Fe}}_{58}{\mathrm{Ni}}_{42}$, ${\mathrm{Fe}}_{50}{\mathrm{Ni}}_{50}$, and ${\mathrm{Ni}}_{100}$. SIRM was normalized after shape correction. Arrows indicate the pressure path; data from Table 1.

Figure 4

(a) Remanent magnetization normalized to the initial (25 °C) value versus temperature for ${\mathrm{Fe}}_{64}{\mathrm{Ni}}_{36}$ at 0 GPa (starting material) and measured after decompression from the indicated peak pressure (heating runs only). The inset shows the temperature after 50% and 95% loss in magnetization as a function of pressure as well as the Curie temperature defined by the second derivative. (b) Comparison of the first and second x-ray peaks between the noncompressed (0 GPa) sample with two that were pressure cycled to peak pressures of 6.8 and 16.3 GPa—the one to 16.3 GPa was measured after heating to 407 °C; that to 6.8 GPa was never heated.

Figure 5

Normalized SIRM versus Ni concentration during compression to 5 GPa; pure iron data from Wei and Gilder [25].