Iron displacements and magnetoelastic coupling in the antiferromagnetic spin-ladder compound BaFe${}_2$Se${}_3$

We report long-range-ordered antiferromagnetism concomitant with local iron displacements in the spin-ladder compound BaFe${}_2$Se${}_3$. Short-range magnetic correlations, present at room temperature, develop into long-range antiferromagnetic order below $T_N$ $=$ 256 K, with no superconductivity down to 1.8 K. Built of ferromagnetic Fe${}_4$ plaquettes, the magnetic ground state correlates with local displacements of the Fe atoms. These iron displacements imply significant magnetoelastic coupling in Fe$X_4$-based materials, ingredients hypothesized to be important in the emergence of superconductivity. This result also suggests that knowledge of these local displacements is essential for properly understanding the electronic structure of these systems. As with the copper oxide superconductors two decades ago, our results highlight the importance of reduced dimensionality spin-ladder compounds in the study of the coupling of spin, charge, and atom positions in superconducting materials.

Figure 1

(a) The structure of BaFe${}_2$Se${}_3$ consists of double chains of edge-sharing FeSe${}_4$ tetrahedra, which extend in and out of the page. The chains are well separated from each other by Ba ions. (b) The double chains in BaFe${}_2$Se${}_3$ are cutouts of the two-dimensional layers of edge-sharing Fe$X_4$ tetrahedra found in the iron-based superconductors. Each double chain consists of pairs of Fe atoms (“rungs”) tiled in a chain along one direction, forming a ladder structure.

Figure 2

(a) In addition to sharp nuclear Bragg reflections, NPD data at 300 K on BaFe${}_2$Se${}_3$ exhibit a broad diffuse scattering peak attributable to magnetic correlations. In their stead, sharp magnetic Bragg peaks corresponding to long-range magnetic order develop by 200 K. (b),(c) The magnetic structure that best fits the NPD data at 5 K is obtained with the all-block configuration with spins aligned along the crystallographic $a$ axis. (d) An all staggered configuration provides a significantly worse fit to the data, indicated by the systematic discrepancies in the difference between the data and the predicted intensities.

Figure 3

(a) The temperature dependence of the magnetic susceptibility for BaFe${}_2$Se${}_3$ shows a discontinuity at 256 K, near the same temperature that long-range magnetic order develops (indicated by the moment per Fe, extracted from Rietveld refinements of NPD data). Error bars for the NPD moments are less than the size of the symbols, and the scatter in the Δ$M$/Δ$H$ data is from the subtraction process used to remove the ferromagnetic contribution from <1$\%$ Fe metal. (b) A low-field (μ${}_{0}H$ $=$ 10 Oe), zero-field-cooled magnetization curve for BaFe${}_2$Se${}_3$ conclusively demonstrates that the material is not a bulk superconductor above 1.8 K (perfect diamagnetism corresponds to −1 on the $y$ axis). The black and orange (gray) lines are guides to the eye.

Figure 4

Local structure analysis with NPDF unambiguously illustrates that the Fe atoms within the spin ladder in BaFe${}_2$Se${}_3$ are each displaced by ∼0.05 Å from their ideal positions, in either a block or staggered configuration. These local iron displacements, not visible in the average crystal structure, increase in magnitude as magnetic order develops, indicating strong magnetoelastic coupling in the iron chalcogenide family of materials.