A fully tetrahedral and highly corner-sharing network model of ZnCl2 glass and its comparison to SiO2 glass.

Zinc chloride, ZnCl2, is intermediate between a strong and a fragile glass former. During computational simulations, it is therefore important to account for ion polarizability. This, together with the lack of suitable interatomic potential parameters, is the likely cause for the lack of modeling studies on ZnCl2 glass which contain a high degree of ZnCl4 tetrahedral units. Through using accurate interatomic potential parameters and applying the adiabatic core-shell model, the first fully tetrahedral model of ZnCl2 glass was obtained. The Cl-Zn-Cl bond angle of 109° reproduced the ideal tetrahedral bond angle, and the calculated total neutron and x-ray structure factors closely replicated experimental findings. While 86% of the ZnCl4 tetrahedral units were corner-sharing, 14% were found to be edge-sharing. This led to two distinct contributions in both the Zn-Cl-Zn bond angle distribution and in the Zn⋯Zn nearest neighbour peaks being seen. These are not apparent in studies based on neutron diffraction. By comparing the intermediate glass former ZnCl2 to the strong glass former SiO2, marked differences in ring statistics became apparent. The Zn-Cl-Zn bond angle of around 110° enabled 3-membered rings to form in significant proportions. In contrast, 3-membered rings were only present in SiO2 glass as defects. By calculating the ZnCl2 and SiO2 partial structure factors, strong similarities became visible after scaling according to nearest neighbour distances. Although it was apparent that the main contributions to the first sharp diffraction peak (FSDP) came from cation-anion correlations, the relative scaling of the FSDP positions in ZnCl2 and SiO2 glass was not understood.