Dislocations and other extended defects, such as grain boundaries and stacking faults, can impart strong effects on the thermal and electronic transport properties of thermoelectric materials. Yet, our understanding of the structure of such defects, even for such widely used materials as bismuth telluride, is still quite limited.

In this presentation, I will discuss our work investigating the structure of extended crystal defects in Bi₂Te₃ and related tetradymite-structured thermoelectric materials. As will be discussed, the anisotropic, layered crystal structure of Bi₂Te₃ has important implications for the resulting atomic-scale core structures of defects in this material. I will begin with a brief overview of the tetradymite system, placing the structure of Bi₂Te₃ in relationship to other layered chalcogenide materials and discussing the dependence of the layering sequence on the metal to chalcogen ratio.

Next, I will discuss HAADF-STEM observations of the atomic structure of dislocations in Bi₂Te₃. I will focus on two very different types of dislocation: one with Burgers vector of type (1/3)<2 -1 -1 0>, which lies parallel with the basal plane, and the other with Burgers vector of type (1/3)<0 1 -1 -1>, which has a large dislocation component normal to the basal plane of c/3, or one full quintuple unit. The first type of dislocation possesses a stoichiometric core that is glissile on the basal plane. The second type possesses a much more complex dislocation core consisting of bismuth-rich faulted region consistent with Bi₃Te₄.

I will also discuss our current understanding of grain boundaries in this class of materials. Arrays of discrete dislocations can accommodate small misorientations between adjacent crystallites of a few degrees. Examples will be shown for low angle boundaries composed of both (1/3)<2 -1 -1 0> and (1/3)<0 1 -1 -1> type dislocations. For large misorientations, however, simple dislocation descriptions break down and it is more physically meaningful to consider a given interface with reference to a low-energy, reference grain boundary structure. Perhaps the simplest such interface in Bi₂Te₃ is the (0001) basal twin. Our electron microscopic observations and density functional theory calculations have shown that the structure of the Bi₂Te₃ basal twin consists simply of a reversal of the basal plane stacking sequence and that this reversal occurs at the Te⁽¹⁾ type planes. As I will discuss it is also possible to go beyond the simple structure of an ideal (0001) twin by considering the introduction of disconnections (defects possessing both dislocation and step content) to this interface. Finally, I will discuss a more complex mode of twinning, on {1 0 -1 5} and {0 1 -1 10}, which has previously been reported in electroplated films of Bi₂Te₃. Our HAADF-STEM observations provide insight concerning the atomic-scale structure of such twinning, pointing to the formation of structural units consistent with local packets of rock-salt structured material.

Sandia National Laboratories is a multi-mission laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000.