Expanding the Composition Range of Transition Metal Dichalcogenide Alloys, Fe_x(Fe_yV_1–ySe₂)

Abstract

A series of Fe x (Fe y V 1−y Se 2 ) compounds were prepared with 0 < x < 0.08 and 0 < y < 0.8 from precursors composed of designed sequences of atomically thin elemental layers deposited with targeted compositions. The largest value of y achieved for a single-phase sample was 0.8, which is over 2.5 times larger than previously reported for Fe y V 1−y Se 2 compounds. Nucleation of the 1T-structured dichalcogenide alloys occurred during the deposition, and slow solid-state diffusion rates during growth at low annealing temperatures prevented the segregation of the elements into alternative compounds such as pyrite-or marcasite-structured FeSe 2 . The lattice parameters and the amounts of x and y in each sample depended on both the initial compositions and the annealing conditions (temperature, time, and Se partial pressure). Since similar lattice parameters can be obtained for different values of x and y, multiple experimental techniques (composition data from XRF, thickness determined from XRR, and Laue oscillations and lattice parameters from XRD) were needed to unambiguously determine the site and extent of Fe incorporation for each sample. The x and y values determined from the combined techniques agreed with a direct measurement of the ratio of intercalated to substituted Fe via cross-sectional STEM-EDS. The results presented here suggest that low-temperature annealing of intimately mixed precursors provides a general route to extending the solid solution range of alloys, as low diffusion rates limit the partitioning of the precursor into a mixture of compounds. Unambiguously determining the amounts of intercalated and substituted dopants is a significant challenge and cannot be accomplished with lattice parameters alone, explaining the often contradictory results found in the literature.