The Crystal Structures of Pb5Sb4S11 (Boulangerite) – A Phase Transition Explains Seemingly Contradictory Structure Models

Abstract

In the literature, many seemingly contradictory structure models of boulangerite, ideally Pb5Sb4S11, and the related mineral falkmanite have been reported. These can be explained by a phase transition between a low‐temperature (LT) modification (P21/c, a ≈ 8.07 Å, b ≈ 23.5 Å, c ≈ 21.6 Å, β ≈ 100.7°) and a disordered high‐temperature (HT) modification [Pnma, a = 23.530(2) Å; b = 4.0385(8) Å and c = 21.273(2) Å, R1(obs) = 0.039] that occurs above ca. 350–400 °C. The almost completely ordered distribution of lead and antimony cations in the LT modification involves a superstructure; the corresponding reflections are obvious in electron and X‐ray diffraction patterns. Partial ordering may render them more or less diffuse. They vanish upon heating above the transition temperature, also for originally ordered natural boulangerite (from Trepča, Kosovo), and appear again after annealing below the phase transition temperature (e.g. at 330 °C). The HT phase is characterized by pronounced cation disorder. Cation ordering can be described by group‐subgroup relationships which may also serve for a unified description of various structure models from literature. Starting from the orthorhombic HT phase, a translationengleiche followed by an isomorphous symmetry reduction yields the degrees of freedom required for cation ordering. The non‐standard setting A21/d11 is optimal for a unified comparison of all known structure models. Varying degrees of ordering reported in literature and observed for various new samples suggest that the transition is kinetically controlled. These relationships also contribute to the long‐lasting discussion about the nomenclature of minerals with the sum formula Pb5Sb4S11 or very similar ones. A precise structure determination of a single crystal (grown by chemical vapor transport) using synchrotron radiation along with detailed electron diffraction investigations and high‐resolution electron microscopy confirms the statistically disordered structure model of the quenched high‐temperature phase. Bond‐valence sums (BVS) and interatomic distances confirm the interpretation of cation split positions as the superposition of the corresponding independent cation positions in the LT phase. Measurements of transport properties indicate very low n‐type electrical and thermal conductivities for both modifications. The order‐disorder transition significantly affects the Seebeck coefficient.