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?

  • 1
    $\begingroup$ Take a look at this questions and the answers there: mattermodeling.stackexchange.com/q/901/24 In between, in DFT, you don't define the bonds, you have to calculate the electronic density and them, from analyze it, get conclusions. $\endgroup$
    – Camps
    Commented Jun 20, 2022 at 14:16
  • 1
    $\begingroup$ Maybe I am wrong but ionic bonds are not strictly speaking defined as the coulomb interaction is always present between two ions. $\endgroup$
    – Greg
    Commented Jun 20, 2022 at 15:53
  • $\begingroup$ I gave my +1 long ago, but can you update us? Were the comments by Greg and Camps useful? Did you figure out an answer? Is this still a problem for you? $\endgroup$ Commented Dec 31, 2022 at 3:22