| 12:18:33 PM PDT - Sat, Oct 11th 2014 |
|
Dear all,
I need advice on troubleshooting transition state search. First, I run frequency calculation run to get the hessian and immediately encounter problems since frequencies are nowhere to be found in the output whereas the *.vib.cm.vdos has three columns and neither of them are negative, even though I constructed a well known guess where at least one mode should be negative. Then I have problems that I don't know how to select the combination of MODDIR VARDIR and FIRSTNEG, NOFIRSTNEG. Either option makes my geometry explode after the first step into something unreasonable. Saddest thing is that I can't see internal coordinates anywhere even though I specify autoz. My input is periodic unit for NWPW calculations, can those proceed in internal coords so I can pick a mode to follow? As it is, there is no clear way to design TS search
Jonas
|