Temperature-induced barium de-trapping from a double-well potential in Ba6Ge25

TL;DR

This study investigates how temperature affects barium atom trapping in Ba6Ge25, revealing a transition linked to Ba de-trapping from a double-well potential that influences the material's structural, transport, and magnetic properties.

Contribution

It introduces a model explaining temperature-induced Ba de-trapping from a double-well potential and its impact on Ba6Ge25's properties, supported by neutron diffraction data.

Findings

Ba atoms occupy split sites that grow apart with decreasing temperature.

A transition occurs around 200-250K involving Ba de-trapping.

Electrical conductivity drops due to increased structural disorder.

Abstract

The crystal structure of barium-germanium clathrate Ba6Ge25 was studied using neutron powder diffraction in the temperature range 20-300K. The compound was found to be cubic (S.G. P4_1 23) in the entire temperature range. However, the fully-ordered model of the crystal structure (no split sites) is marginal at room temperature, and clearly fails at low temperature. A much better description of the crystal structure below 250K is given in terms of two split Ba sites, with random occupancies, for two out of three types of cages present in the Ba6Ge25 structure. The Ba atoms were found to interact strongly with the Ge host. The separation of the split Ba sites grows with decreasing temperature, with a sudden increase on cooling through the 200-250K temperature range, accompanied by an expansion of the entire crystal structure. We propose a simple model for this transition, based on temperature-induced de- trapping of Ba from a deep double-well potential. This transition is associated with sizeable anomalies in the transport and magnetic properties. The most significant of these effects, that is, the drop in electrical conductivity on cooling, can be easily explained within our model through the enhanced structural disorder, which would affect the relaxation time for all portions of the Fermi surface. We suggest that the other anomalies (increase in the absolute value of the negative Seebeck coefficient, decrease in the magnetic susceptibility) can be explained within the framework of the one-electron semi- classical model, without any need to invoke exotic electron-electron interaction mechanisms.