High-pressure magnetic properties of antiferromagnetic samarium up to 30 GPa using a SQUID-based vibrating coil magnetometer

Samarium (Sm) has antiferromagnetic (AFM) ordering at the hexagonal site at $T_{\mathrm{N}}\left(\text{hex}\right)=106$ K, and at the cubic site at $T_{\mathrm{N}}\left(\text{cub}\right)=14$ K at ambient pressure. The structural transition from the so-called Sm-type structure to the double hexagonal close-packed (dhcp) structure occurs at approximately 6 GPa. According to electrical resistance measurements, the shift of $T_{\mathrm{N}}$(hex) toward the low-temperature side occurs simultaneously with the shifting of $T_{\mathrm{N}}$(cub) toward the high-temperature side with increasing pressure. We conducted dc magnetic measurements on Sm at high pressure up to 30 GPa using a SQUID-based vibrating-coil-magnetometer to pursue $T_{\mathrm{N}}$(cub) at high pressure: The magnetic measurements revealed stepwise anomalies below 2 GPa. A ferromagnetic (FM) anomaly was observed in the dhcp structure, suggesting that spins at both structural sites formed the FM magnetization. In the face-centered cubic (fcc) structure at above 12 GPa, a reduction in the magnetic signal occurred. In the distorted fcc structure at above 20 GPa, sufficient suppression of the FM moment was observed and afterward the diamagnetic signal suggesting the possibility of superconductivity was observed.

Figure 1

(a) $T$-dependence of magnetization $M$ of 186.6-mg polycrystalline Sm composed of many fragments with a rectangular shape of approximately 0.5 mm $\times$ 0.5 mm at $H_{\mathrm{dc}}$ = 5 kOe and ambient pressure, measured using a commercial SQUID magnetometer. The measurement was performed in the warming process to 150 K after ZFC and the following FC process. (b) $P\text{−}T$ phase diagram for Sm based on previous electrical resistance measurements [15]. The red circles present $T\mathrm{s}$ with magnetic anomaly at $H_{\mathrm{dc}}$ = 5 kOe, shown in (a). The $P\text{−}T$ region covered in the present experiment is shaded in light blue. The $P$ range of four runs is shown by the blue bars.

Figure 2

(a) Overview of the setup of SQUID-VCM: The use of diamond anvils with different culet size (i.e., 500 and 550 $\mu \mathrm{m}$) allows the CuBe-NiCrAl composite gasket to be deformed away from the detection coil. (b) Picture of the detection coil made of NbTi wire (30-$\mu \mathrm{m}$ diameter) coated with Cu. The diameter of the wire, including the Cu and insulating layers, is 50 $\mu \mathrm{m}$. The 20-turn detection coil was prepared around the bore with a diameter of 2 mm; the outside has a 10-turn compensation coil.

Figure 3

$T$-dependence of $V_{\mathrm{VCM}}$ for the first run. $V_{\mathrm{VCM}}$ proportional to the magnetization is normalized with the dc magnetic field $H_{\mathrm{dc}}$, resulting in a relative dc magnetic susceptibility. The first run covers a pressure range of up to 6 GPa (near the border between the Sm-type and dhcp structures [(a) 0, (b) 2.0, and (c) 5.9 GPa]. A series of magnetic anomalies for Sm is presented with closed triangles.

Figure 4

$T$-dependence of $V_{\mathrm{VCM}}$ for Sm in the second run. $V_{\mathrm{VCM}}$ proportional to the magnetization is normalized with the applied dc magnetic field $H_{\mathrm{dc}}$. In the second run, the pressure range increased to 23.1 GPa, below which the structure changed from Sm-type $\to$ dhcp $\to$ fcc $\to$ distorted fcc structures: [(a) 6.3–20.5 GPa and (b) 23.1GPa]. The $H_{\mathrm{dc}}$ in (a) is 17 Oe at 6.3 GPa, 13 Oe at 8.3 and 12.8 GPa, and 10 Oe at 20.5 GPa. In (a), the $T$ position of the diamagnetic signal of Pb at $P$ = 6.3 GPa is marked with a red broken line and a black arrow. A series of magnetic anomalies for Sm is marked with closed triangles.

Figure 5

$T$-dependence of $V_{\mathrm{VCM}}$/$H_{\mathrm{dc}}$ for the third run, covering the $P$ range to 21.1 GPa (at maximum 30.8 GPa), below which the structure changes as Sm-type $\to$ dhcp $\to$ fcc $\to$ distorted fcc. The $H_{\mathrm{dc}}$ value is 30 Oe except for 24 Oe at 0 GPa.

Figure 6

$T$-dependence of $V_{\mathrm{VCM}}$/$H_{\mathrm{dc}}$ for the third run up to 30.8 GPa. The $H_{\mathrm{dc}}$ value is 30 Oe except for 24 Oe at 0 GPa. The low-$T$ data in Fig. 5 are enlarged in (a) with addition of the data for $P>$ 21.1 GPa. (b) For 24.8 GPa, there are two data points in the warming processes after ZFC and FC at $H_{\mathrm{dc}}$ = 30 Oe.

Figure 7

$T$-dependence of $V_{\mathrm{VCM}}$ for Sm in the fourth run. $V_{\mathrm{VCM}}$ proportional to the magnetization is normalized with the applied dc magnetic field $H_{\mathrm{dc}}$. $H_{\mathrm{dc}}$ was increased up to 110 Oe. In the fourth run, the pressure was focused to $P=22.8$ and 27.3 GPa, corresponding to the pressure range of the distorted fcc structures.

Figure 8

$H_{\mathrm{dc}}$ dependence of offset temperature $T_{\mathrm{sc}}$ at $P=22.8$ GPa and 27.3 GPa in the fourth run. Solid curves represent $H_{\mathrm{dc}}=H_{\mathrm{c}}\left(0\right)\{1-{[T/T_{\mathrm{sc}}\left(0\right)]}^2\}$ with $T_{\mathrm{sc}}$(0) = 6.14 (blue) and 6.25 K (red).

Figure 9

$P\text{−}T$ phase diagram determined by the present experiments. Previous results published in the literature on the $R$ measurements [15] are plotted with solid black circles. The characteristic temperatures $\#2$ and $\#4$, determined via three runs, are plotted with open and closed symbols, respectively. The trends in the shifting of $\#2$ and $\#4$ are presented with red broken curve and dotted line. The offset temperatures, likely suggesting superconducting transition temperature, determined via three runs are plotted with solid diamonds. The intensity of $V_{\mathrm{VCM}}$ is represented by colors ranging from red (positive) to blue (negative). The red region denotes the FM state, while the blue region denotes the diamagnetic state.