Spin Calculations give different results in Plane-wave DFT

[bhaskarchilukuri]

Hello Everyone,

I am trying to calculate the lowest energy spin-state (singlet and triplet) of a transition metal complex using VASP.
I know that, in VASP which uses Plane Wave DFT, NUPDOWN keyword gives the difference in number of electrons between spin up and spin down components.
The metal-ligand complex optimizes well both for neutral/singlet (NUPDOWN = 0) and excited/triplet (NUPDOWN = 2).
The metal-ligand complex is placed at the center of a big box (lattice) of vacuum essentially making the complex isolated form any periodic interactions.
I have used the hybrid B3LYP functional with GGA=B3 pseudo-potentials for my calculations. What I see is that metal binds to the ligand in both singlet and triplet states.


Question:
Using the same functional and pople basisset, I performed a calculation using molecular DFT codes like GAMESS, GAUSSIAN or ADF (without periodic boundary conditions).
In the the same complex, the metal would not bind to the ligand when in Singlet/Neutral state (it keeps pushing the ligand away from the metal even with loose or strict optimization conditions).
The transition metal and ligand would however only bind in Triplet state but not in Singlet state. This also true in experiment.
I have the same functional in either case (VASP and Molecular DFT) although the potentials may be different because molecular DFT codes use basis-sets while VASP uses plane-wave potentials.
But I would expect the same results in either case.

Do anyone know why this might be happening ?
Is it because of differences between molecular and plane-wave DFT ? If so, can you please explain the specific things that may be causing this anomaly?

Thank you

[alex]

Hi bhaskarchilukuri,

I'd suspect your setting in the 3D case. Your box may not yet be big enough, or maybe you are trapped by symmetry.

Hth

alex