Beyond Stoichiometry: Layer-Selective Cation Disorder Enables Ultralow Lattice Thermal Conductivity in GeBi4 X 7 and GeBi6 X 10 ( X = Se, Te)

Abstract The nominal compounds GeBi4X7 and GeBi6X10 (X = Se, Te) exhibit pronounced cation disorder, posing challenges in achieving single-phase materials at their nominal stoichiometries and thereby complicating functional optimization. In this work, through controlled cation composition regulation, we successfully synthesize single-phase compounds derived from nominal GeBi4Te7 and GeBi6Te10, eliminating the secondary phases previously reported in these compounds. The actual stoichiometries of these single phases are determined to be Ge0.75Bi4.25Te7 and Ge0.65Bi6.35Te10, respectively. Additionally, two isostructural single-phase compounds are discovered in the Se-based counterparts, with actual compositions of Ge0.72Bi4.28Se7 and Ge0.62Bi6.38Se10. Rietveld refinement analyses reveal that the cation layers adjacent to the van der Waals gaps are preferentially occupied by Bi, whereas cation disorder predominantly arises in the central mixed-cation layer within the GeBi2X4 stack. This layer-selective preference, combined with cation disorder enabled by the nearly identical electronegativities of Ge and Bi, drives the overall stoichiometry toward the Bi-rich regime. With respect to thermoelectric properties, the complex-stacking single-phase materials derived from nominal GeBi4X7 and GeBi6X10 show distinct advantages over the simple-stacking GeBi2X4 counterparts and conventional Bi2X3 matrices. Experimental results indicate that the incorporation of additional Bi2X3 stacks markedly enhances phonon scattering and reduces the lattice thermal conductivity, while exerting minimal impact on carrier mobility. Of particular significance, the single-phase material Ge0.65Bi6.35Te10 derived from nominal GeBi6Te10 achieves a remarkably low lattice thermal conductivity of ∼0.35 W m–1 K–1 and a dimensionless thermoelectric figure of merit of ∼0.62 at 525 K. Predictions based on the single parabolic band model indicate a substantial potential for improving the thermoelectric performance of Ge0.65Bi6.35Te10 through the optimization of carrier concentration, suggesting that this compound represents a promising alternative to the traditional Bi2Te3 matrix.