<html><head></head><body><div style="font-family: Verdana;font-size: 12.0px;"><div><tt class="letterText">Dear All,<br/>
<br/>
I try to do some structure optimizations and frequency<br/>
calculations at the CASPT2 level for several lanthanide<br/>
fluorides. For a few molecules I have the problem that the<br/>
GRAD(0) does not converge although the energy, GRAD(ORB)and<br/>
GRAD(CI) are already converged.<br/>
Sometimes this happens during optimization sometimes just<br/>
in the frequency analysis. I already tried analytical as<br/>
well as numerical gradients, with no success.<br/>
Does anyone knows what's the problem and how to handle this<br/>
or how I can change the threshhold for Grad(0)?<br/>
<br/>
Thanks in advance,<br/>
Thomas</tt></div></div></body></html>