Emergent Random Spin Singlets in Disordered Spin-1/2 perovskite BaCu$_{1/3}$Ta$_{2/3}$O$_3$

TL;DR

This study investigates BaCu$_{1/3}$Ta$_{2/3}$O$_3$ as a disordered spin-1/2 system, combining X-ray diffraction and EXAFS to reveal local chemical order that shapes magnetic exchange pathways. The material shows no magnetic ordering down to 0.1 K; susceptibility and magnetization exhibit scaling consistent with broad exchange distributions, yet specific heat measurements reveal deviations from ideal random-singlet expectations. A data-driven distributed-exchange dimer model reconstructs a broad but non-singular P(J) with a peak near $J\approx 4$ K and additional features around 70 K and 0.1 K, consistent with the observed thermodynamics and EXAFS-derived Cu–Ta disorder. The results point to a class of three-dimensional spin-disordered materials where structural disorder yields broad but finite exchange distributions, avoiding the infinite-randomness fixed point characteristic of RS phases.

Abstract

We investigate the disordered perovskite BaCu$_{1/3}$Ta$_{2/3}$O$_3$, where Cu (spin-1/2) and Ta randomly occupy a pseudo-cubic lattice. Synchrotron X-ray diffraction and X-ray absorption spectroscopy establish the local nature of the disorder, revealing the presence of structurally constrained magnetic exchange paths. No magnetic ordering or spin freezing is observed down to 0.1 K. The low-temperature magnetic and thermodynamic behavior is captured by a broad but non-singular distribution $P(J)$ of exchange couplings $J$. These results open the possibility of realizing a disordered quantum ground state where the exchange randomness is broad yet intrinsically bounded, departing from the conventional infinite-randomness fixed point driven random-singlet phase.

Figures (4)

• Figure 1: (a) The Rietveld refinement of X-ray powder diffraction patterns of BCTO, (b) a schematic of the unit cell of the BCTO crystal structure. Here, A$_b$ and N are the absorbing atom and the nearest-neighbour atom, respectively, as discussed in the XAS (see text).
• Figure 2: (a) Normalized Ta-LIII and Cu-K edge XANES spectra, with the energy scale given relative to the edge energy (E–Eo) to facilitate comparison; (b) $k^2$-weighted experimental EXAFS data (dots) and corresponding best-fit curves (solid lines) at the Cu and Ta edges for the BCTO sample; (c) moduli of the $k^2$-weighted $\vert \rm{FT}\vert$ of the experimental data (dots) and best fit curves (solid lines) for the various spectra; the curves are vertically offset for clarity. (d) homo-atomic (red) and hetero-atomic (blue) neighbour probabilities around the absorber plotted as a function of absorber concentration x, for both random distribution (dashed lines) and chemical order (solid lines) models (see text), compared with the experimental data, where x = 0.33 (0.66) corresponds to the Cu (Ta) edge results.
• Figure 3: (a) Magnetic susceptibility $\chi$ and $\chi^{-1}$ plotted as a function of temperature; (b) Isothermal magnetization $M(H)$ at various temperatures. The solid lines in (a) and (b) are fits to the data (see text); (c) and (d) show data scaling behavior for $\chi$(T) and $M(H)$, respectively. Inset in (c) shows $\chi(T)$ on a double-log plot. The units are same as in (a). The arrow indicates departure from T$^{-\gamma}$ at low temperatures.
• Figure 4: (a) The temperature variation of specific heat plotted as $\rm c_p/T$ (phonon contribution is also shown); (b) $(\mu_0 H)^{\gamma} c_p/T$ ($\gamma = 0.67$) plots against T/$\mu_0H$ for various applied fields are shown not to obey the scaling behavior. Inset in (b) shows $c\rm_{mag}$ varying linearly with T (red line). T$^{0.33}$ (blue line) is also shown to emphasize non-compliance with the randoms-singlet ground state; (c) exchange-coupling distribution $P(J)$ obtained here (inset: double-logarithmic plot of the same); (d) $c_{\rm mag}(T)/T$ versus $T$ plots for various H. The solid lines in (d) are fit to the data (inset shows the temperature variation of $\rm S_{mag}$ for $\mu_0H = 0, 9$ T).