Question regarding orbital hybridization in DFT

[sidiq]

Hello All!

I was wondering if anyone can tell me if or if not the wavefunctions used to construct the electron density in density functional theory contains hybridizied orbitals.

I have been working with a planewave pseudopotential code and in the input file of a pseudopotential, say Nickel, it only contains values for l=0 (s-orbital), l=1 (p-orbital) and l=2 (d-orbital). If I would want to calculate NiO, I would include another pseudopotential input file, Oxygen, which contains simply the s and p orbital terms.

Thus, as I understand it, the wavefunctions for Ni and O are built only from their respective atomic orbitals. My question is, do they get hybridizied? Is there an algorithm for these atomic orbitals to somehow be made into a t2g or eg orbital, thus initializing a new wavefunction to be fed into the DFT calculation?

Thank you for any help and I welcome any corrections to my understanding.

 

[DrDu]

Hybridization isn't a concept of DFT or MO based theories but of valence bond theory, although certainly s, p, and d- orbitals can all make a contribution to a given MO in general.
There is code to use symmetry addapted linear combinations (SALC's) of AO's in computations. Details depend on the program you are using. The manual is your friend, here.

 

[sidiq]

Thank you for the reply DrDu. It was really insightful for me.