Hi, I am examining a ~70 atom Nickel(III) complex with doublet low spin and quintet high spin using ub3lyp method. I have been examining the %HF exchange at different values for both low/high spin to see how consistently my method ascertains the ground state after an optimization.
I have been encountering instabilities in the wavefunction after my optimizations. To correct this, I run stable=opt and then reoptimize using guess=check from the stable=opt checkpoint file, as well as switching to scf=(tightlineq,qc,maxcycle=512) in the optimization. Unfortunately, for my highest HF exchange (35%), even after this optimization converges, I still have instabilities in the wavefunction.
Right now I am just repeating stable=opt and hoping the geometry converges with a stable wavefunction, but are there other measures that can be taken to help ensure a stable wavefunction?
Thank you!
Did you try it?
first optimize structure by opt=(calcall) freq keywords then check stability by stable (not stable=opt) keyword
Hey there,
I tried opt=(tight,calcfc), scf=(tightlineq,qc), and guess=check using the checkpoint from the stable=opt calculation. It still had instabilities.
In my large system, calcall exceeds a week, the walltime on the server I use. Do you know any other adjustments I can make?
Thank you
First, how many total electrons does your "molecule" have? If the DOUBLET calculation with charge=0 works, then this means the total number of electrons is odd. The combination of odd number electrons and QUINTET is not allowed, so Gaussian should stop its calculation. Is it a typo of QUARTET?
Second, STABLE's judgement is not always perfect. See the discussion of "Wavefunction Stability", p.71, in Gaussian's text book , "Exploring Chemistry with Electronic Structure Methods, 3rd Edition".
For modeling of metal complex, I hope, you can find some idea from EXAMPLE 9.7, p447.
have a good study!
do you check stability after optimization( opt=(tight,calcfc), scf=(tightlineq,qc)) only by stability keyword(not stable=opt)
if you did
Try to generate the stable wave function by stable=op and (guess=alter or guess=mix)
the instability can relate to symmetry of wavefunction
- Badges
- Science topic