I'm working with an amorphous system. With oxygen, my system has both covalent and ionic bond forming cations. I utilise the Wannier centre approach to define a covalent bond, and my theory is that if the Wannier centre is near to the line connecting the cation and anion, it is a covalent bond.
However, this method fails to identify an ionic bond as there is no region of high electronic charge density which can be localized to obtain Wannier center. How can an ionic bond be defined using a computational technique?
Let's imagine I have an AO2 system in which A is known to form a covalent bond. When we add the modifier B2O, we get a binary glass. Now, if no minima can be found using the Pair distribution function, how should I analyse the local environment, such as the coordination number of B cation?