Elementary considerations on the local symmetry in optoelectronic materials and their phase change behavior

Crystalline structures of elemental solids can be rationalized in terms of the competition between ions and electrons: ions try to increase local symmetry and thus packing fraction, while electrons want to reduce it. If the latter win, layered structures, network, or molecular solids form and the opening of an electronic gap is favoured. In this work, it will be discussed how this competition can affect the thermodynamic behavior of phase change materials (PCMs), in particular that of Ge-Sb-Te alloys: their technologically relevant metastable crystalline structures can be derived from (hypothetical, metallic) simple cubic crystals near half-filling via a symmetry breaking, such as a Peierls distortion in Sb-rich PCMs or ordering of chemical species onto sublattices on the GeTe-Sb2Te3 pseudo-binary line, leading to the formation of sigma-bonded networks. Local symmetry and density become even smaller and the gap opens up even more in the glass, for example, when the group IV element germanium undergoes a coordination change from (distorted) octahedral in the crystal to tetrahedral. This coordination change leaves the sigma-bonded network intact, as will be demonstrated by analysis of first-principle simulations. Based on local symmetry arguments, simple rules for the number of electron holes and/or vacancies in metastable crystalline structure of PCMs can be derived and the response of Ge-Sb-Te alloys to pressure be predicted: crystalline alloys will amorphise under pressure when there are more Te than Ge atoms and increase their conductivity. Conversely, disordered alloys will crystallize if the number of Ge atoms exceeds that of Te. The possibility to switch the latter PCMs reversibly with pressure will be discussed. Lastly, unusual relaxation dynamics of PCMs are identified from first-principle calculations: when a solid is streched to its amorphisation point, the ionic energy (which is minimized in the crystal) increases with time as opposed to the dominating electronic energy. At the same time, coordination statistics become increasingly distinct with age from those in the crystal, i.e., the glass initially relaxes away from the crystalline phase.