Growth-Controlled Twinning and Magnetic Anisotropy in CeSb$_2$

TL;DR

CeSb$_2$ presents a rich magnetism masked by in-plane twinning in layered diantimonides. By combining Sb-rich flux growth with slow cooling and comprehensive magnetometry, the authors extract the intrinsic in-plane anisotropy, finding an easy-axis saturation near $M_{ ext{easy}} \approx 1.8~\mu_{\text{B}}/\text{Ce}$ at high field while the hard axis remains strongly suppressed and nearly linear. They establish a consistent low-temperature phase diagram along the true easy axis and show that high-quality, untwinned crystals are achieved by avoiding the proposed high-temperature transition, with no clear evidence for a distinct $\beta$ phase. These results provide essential constraints for crystal-electric field models and improve understanding of the interplay between anisotropic magnetism and unconventional superconductivity in CeSb$_2$.

Abstract

Cerium diantimonide (CeSb$_2$) is a layered heavy-fermion Kondo lattice material that hosts complex magnetism and pressure-induced superconductivity. The interpretation of its in-plane anisotropy has remained unsettled due to structural twinning, which superimposes orthogonal magnetic responses. Here we combine controlled crystal growth with magnetization and rotational magnetometry to disentangle the effects of twinning. Nearly untwinned high-quality single crystals reveal the intrinsic in-plane anisotropy: the in-plane easy axis saturates at $M_{\text{easy}}(4~\text{T}) \approx 1.8~μ_{\text{B}}$/Ce, while the in-plane hard axis magnetization is strongly suppressed, nearly linear, and comparable to the out-of-plane response. These results resolve long-standing discrepancies in reported magnetic measurements, in which in-plane metamagnetic transition fields and saturation magnetization varied significantly across previous studies. Growth experiments demonstrate that avoiding the proposed $α$-$β$ structural transition $-$ through Sb-rich flux and slower cooling $-$ systematically reduces twinning. However, powder X-ray diffraction and differential thermal analysis measurements show no clear evidence of a distinct $β$ phase. Our results establish a consistent magnetic phase diagram and provide essential constraints for crystal-electric field models, enabling a clearer understanding of the interplay between anisotropic magnetism and unconventional superconductivity in CeSb$_2$.

Figures (15)

• Figure 1: Layered crystal structure of CeSb$_2$ showing Ce -- Sb bilayers (green: Ce, brown: Sb) and Sb sheets along the $c$ axis. Twinning is schematically explained (blue). Structure model generated using the VESTA software package Momma_VESTA_2011.
• Figure 2: Top: In-plane $M(H)$ at 2.5 K for various crystals, showing m$_1$/m$_2$ behavior due to twinning. Symbols denote the different growth conditions: squares - crystals that crossed the proposed high-temperature transition; circles - crystals that may have crossed it; triangles - crystals that avoided it. Bottom: Same data normalized to the saturation magnetization $M_S=M(2.2~\text{T})$; m$_1$ curves collapse, while m$_2$ curves show an additional linear term. Inset:$M(H)$ around small fields; a tiny hysteresis is visible at $\mu_0H=0~\text{T}$ for all crystals.
• Figure 3: Extraction of the intrinsic hard-axis magnetization for Crystal 5 by scaling and subtracting the m$_1$ curve from the m$_2$ curve. A faint hysteresis is visible in the hard-axis response.
• Figure 4: Schematic of in-plane magnetization at $T=2.5$ K for varying twinning ratios, including the untwinned anisotropy ($x=1$). The $a$ axis is assigned as the easy axis and the $b$ axis as the hard axis (similar to the $c$-axis response reported by Bud'ko et al.Budko_CeSb2_magnetic_1998) Miyake_CeSb2_magnetic_switch_2025, though the opposite assignment cannot be excluded Shan_FM_ladder_q1D_2025.
• Figure 5: Top: Angular dependence of the magnetization $M(\theta)$ in the field-polarized (FP) regime compared to model fits. Bottom:$M(\theta)$ in the paramagnetic (PM) regime with corresponding model fits. Insets: Schematic illustrations of the expected angular dependence for each regime.