Timeline for How closely should the energy components of ab initio SCF results at a stationary nuclear configuration preserve the virial theorem?
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 30, 2017 at 11:47 | comment | added | wang1908 | Thanks for looking into this, it's very nice to have someone to compare notes:-) Here is the input of the calculation, for the purpose of the project I worked on UHF rather than RHF chosen, not sure where it caused the differences. molecule h2o { O H 1 0.96 H 1 0.96 2 104.5 } set basis 3-21g set scf_type pk set g_convergence GAU_VERYTIGHT set reference uhf E, wfn = optimize('scf',return_wfn=True) | |
| Aug 30, 2017 at 5:22 | comment | added | jheindel | @Wang-X-Y Ehhh... I ran some calculations on water with a number of different basis sets and there is very little change between them. In fact sometimes the smaller basis set had a ratio closer to one, but really the difference is quite small. I think I'll post these sometime tomorrow because pretty much my answer above isn't right. I found it very compelling, but the fact HF is a mean-field theory invalidates my reasoning. | |
| Aug 30, 2017 at 0:08 | comment | added | jheindel | Yes I think you're correct. It makes sense. I'm completely guessing here but it could be that if smaller basis sets show a larger deviation, and they should follow the virial theorem exactly, that this is the result of some kind of intramolecular basis set superposition error... ? | |
| Aug 30, 2017 at 0:05 | history | edited | jheindel | CC BY-SA 3.0 |
deleted 5 characters in body
|
| Aug 29, 2017 at 22:19 | comment | added | wang1908 | Thanks for the insightful answer. Although SCF is single-body mean field theory, the wave function thus derived is an approximation to the exact wave function of the all-electron Hamiltonian, in which case n = 1 in your formula, even after electron-electron repulsion is taken into account. The above deviation is only significant when a small basis set, say 3-21G, is used. For 6-31G deviation is rather negligible. You are probably right to the point that for smaller basis set, it does not evaluate to a parameter ζ=1, but I do not fully understand how. | |
| Aug 28, 2017 at 6:51 | history | answered | jheindel | CC BY-SA 3.0 |