Is it possible to calculate electronic energy of a cation X+ without relaxation of its orbitals compared to those of netural X in DFT, HF, ...?! In other words, can one model a hypothetical atom with arbitrary orbital energies and then calculate its electronic enery?(keeping in mind the quantum mechanics doesn't allow the levels to have arbitrary energies!) I want to calculate vertical ionization potential (for atoms and molecules) using DFT or HF, and I need to reduce the effects of orbital relaxation, self interaction errors, derivative discontinuity and ... to obtain IP equal to negative value of HOMO (Koopman's theorem). However, Koopman's theorem refers to cations that are neither electronically nor geometrically relaxed.
Do 1-SCF calculation with fixed (previously optimised for neutral molecule) geometry.
Thank you for your answers.
Dear Prof. Kislyakov, 1-SCF (Single point) calculations only fixed the geometry but not electronic structure. Orbitals of X+ will be relaxed upon the single point calculations compared to those of neutral X.
It is inevitable. All electronic transitions have the same time scale.
I understand what you want to do: (1) Store the orbitals for X on file (2) Read them back in for a calculation on X+ (3) Evaluate the energy of X+ with the orbitals for X....no SCF iterations whatsoever.
The problem is how to tell the electronic structure program not to do ANY iterations and just evaluate the energy for the cation with the orbitals for the neutral. I tried to do that a number of years ago with Gaussian03 (or was it G98?) by playing with the IOPs etc. Even specifying just 1 SCF evaluation and a ridiculously low SCF convergence criterion (so that the calc might 'think' it was converged right at the beginning) I was never quite sure whether the calculation had in fact iterated once --- or the energy of X+ had just been evaluated (using guess=read) as desired.
If you are currently using Gaussian, play with the IOP's and see wif you can set it up right. Otherwise, I would look at the manuals for GAMESS or ORCA to see if I could do the calcs there.
Dear Prof. Krogh-Jespersen
Thank you for your valuable comment. I will try this procedure with G09.
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling