Problem with opt convergence using XTB

rpestana94

Greetings, I'm trying to do an optimization using XTB, in my system, I have an NMA molecule with 17 molecules of solvent, for the solvent, I'm using ChCl3, H2O, DMSO, acetonitrile, hexane, toluene, methanol, and tetrahydrofuran.

The final goal is to do a frequency calculation, with all solvents I can do it but when I'm using tetrahydrofuran, the opt does not converge with the 17 molecules, for some reason converges just with 11 molecules. The energy from one step to another goes higher and higher and the structure looks fine. I don't know what is happening, if any of you have any idea of what is going on I really appreciate it.

I'm using the default parameter and executing xtb like this:
xtb file.pdb -I xtb.inp --gbsa Thf --ohess --uhf 1

Thanks in advance.

The picture show how the energy goes higher and higher

sorry for the bad quality of the picture.

sobereva

NMA is a singlet molecule, why did you set --uhf 1?

rpestana94

sobereva 发表于 2021-7-13 03:14
NMA is a singlet molecule, why did you set --uhf 1?

I'm using --uhf 1 because I have NMA with 17 molecules of solvent.

But the problem where solve by adjusting the Electronic temperature of convergence of MD, my advisor give the solution and to be honest I don't know why this work, professor sob do you have a paper or an explanation of what is this parameter?

Thanks in advance

sobereva

rpestana94 发表于 2021-7-13 22:54
I'm using --uhf 1 because I have NMA with 17 molecules of solvent.

But the problem where solve ...

-uhf 1 corresponds to doublet, this impossibly corresponds to your actual case. You should fix this problem first.