Compensated Semimetal LaSb with Unsaturated Magnetoresistance

By combining angle-resolved photoemission spectroscopy and quantum oscillation measurements, we performed a comprehensive investigation on the electronic structure of LaSb, which exhibits near-quadratic extremely large magnetoresistance (XMR) without any sign of saturation at magnetic fields as high as 40 T. We clearly resolve one spherical and one intersecting-ellipsoidal hole Fermi surfaces (FSs) at the Brillouin zone (BZ) center $\mathrm{\Gamma }$ and one ellipsoidal electron FS at the BZ boundary $X$. The hole and electron carriers calculated from the enclosed FS volumes are perfectly compensated, and the carrier compensation is unaffected by temperature. We further reveal that LaSb is topologically trivial but shares many similarities with the Weyl semimetal TaAs family in the bulk electronic structure. Based on these results, we have examined the mechanisms that have been proposed so far to explain the near-quadratic XMR in semimetals.

Figure 1

Crystal structure and magnetotransport of LaSb. (a) Schematic crystal structure of LaSb. (b) XRD pattern on the (001) surface of single crystal. The inset shows a typical cuboid single crystal. (c) Resistivity as a function of temperature at magnetic field $H=0$ (black curve) and 9 T (red curve). The inset illustrates the directions of $H$ and $I$ in the magnetotransport measurements. (d) $MR(\%)=[R(H)\text{−}R(0)]/R(0)\times 100\%$ plotted as a function of field up to 40 T at $T=2\mathrm{K}$ (red curve), where $R(H)$ and $R(0)$ represent the resistivity at magnetic field $H$ and at zero field, respectively. Blue curve is the fitting to the two-band model considering slightly imperfect carrier compensation ($n_e/n_h=0.998$). Black curve is the fitting to the three-band model with perfect carrier compensation. (e) Simulated MR as a function of the ratio $n_e/n_h$ based on the two-band model. Solid circles represent the experimental MR at $T=2$ and 200 K under 9 T assuming that the suppression of MR at high temperatures is attributed to carrier imbalance. (f) FFT spectrum of the Shubnikov–de Haas oscillations shown in the Supplemental Material [39]. The inset shows that the peak positions of $F_{\gamma }$ and $F_{2\beta }$ are extracted by the fitting to two Gaussian functions.

Figure 2

Experimental and calculated FSs of LaSb. (a) Schematic of the first and second 3D BZs. The cyan and green areas illustrate the locations of the mapping data in (b) and (c), respectively. (b),(c) FS intensity plots obtained by integrating the photoemission intensity within $E_F\pm 10\mathrm{meV}$ recorded with $hv=53$ and 83 eV, close to the $k_z=0$ and $\pi$ planes, respectively, at $T=30\mathrm{K}$. $a^{\prime }$ is the half of lattice constant $a(=6.5\AA )$ of the face-center-cubic unit cell. Cuts 1–4 in (b) indicate the momentum locations of the measured bands in Fig. 3. (d) Same as (b), but measured at $T=200\mathrm{K}$. Dashed lines represent the (001) surface BZ. (e) Fermi wave vectors extracted from the data recorded with $hv=53$ (black dots) and 83 eV (green dots) at $T=30\mathrm{K}$ and $hv=53\mathrm{eV}$ at $T=200\mathrm{K}$ (red dots). Solid lines are FSs derived from QOs. (f) Calculated 3D FSs with the MBJ potential.

Figure 3

Near-$E_F$ band dispersions along high-symmetry lines measured at $k_z\sim 0$ plane ($hv=53\mathrm{eV}$). (a),(b) Photoemission intensity plot along $\mathrm{\Gamma }-X$ [cut 1 in Fig. 2] and corresponding 2D curvature intensity plot [48], respectively. (c) Energy distribution curves of (a) around $X$. (d),(e) Same as (a),(b) but rotated by 45° [cut 2 in Fig. 2]. (f),(g) Same as (a),(b) but along $X-W-X$ [cut 3 in Fig. 2]. (h) Momentum distribution curves of (f) around the right $X$. (i),(j) Same direction as in (d),(e) but around $X(-\pi ,-\pi )$ [cut 4 in Fig. 2]. The black dots in (c) and (h) indicate the peak positions, which track the dispersions of the electron band along and perpendicular to the long axis of the ellipsoidal $\gamma$ FS, respectively. The linear electron band at $X$ in (a)–(c) and the hole bands around $X$ in (i) and (j) are a result of the band folding with a wave vector $Q=(\pi ,\pi ,0)$.