TURBOMOLE Users Forum

TURBOMOLE Modules => Jobex: Structure Optimization and Molecular Dynamics => Topic started by: jbaltrus on May 19, 2010, 07:40:02 PM

Title: -ex statpt, large gradient, small energy, no convergence
Post by: jbaltrus on May 19, 2010, 07:40:02 PM

I will give this problem a chance on the forum.

I have this floppy system (or is it?..) that I optimize in excited state. It would not converge since gradient is large while energy is negligible. In fact, I am getting this problem in most of the molecules I work with, but only in excited state optimization and not in ground state optimization.

Is there anything I am missing or can tweak to get it optimized? I have been offered to get rid of hessaprox already since I start with the converged ground state geometry, I do not know if that will affect convergence, though

thanks

Jonas

$convinfo
 energy change  : actual value =  -0.2001E-09 threshold =   0.1000E-05
 geom. gradient : actual value =   0.2354E-01 threshold =   0.1000E-03
$convergence not reached

Title: Re: -ex statpt, large gradient, small energy, no convergence
Post by: jbaltrus on May 19, 2010, 07:40:40 PM

this is exactly the problem that Beverly in one post below has