Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly

Alex Granovsky
gran@classic.chem.msu.su

the post with the output file you mentioned is:

http://classic.chem.msu.su/cgi-bin/ceilidh.exe/gran/gamess/forum/?C370ca56b1cGU-7358-884-00.htm

The email of the author of the post is shown in the post, so you can contact him directly if you like.

Regards,
Alex Granovsky

On Mon Aug 16 '10 4:25am, Jonas Baltrusaitis wrote
--------------------------------------------------
>Vladimir,

>what is nneg=2 for? from the manual it is the amount of negatives in Hessian. but shouldn't TS have only one negative, e.g. specifying nneg=2 sort of tells software to look for something that is not TS?

>Another question is more conceptual and related with the first one: how can  2+2 TS involving movement of 4 atoms have only one negative frequency? It should be two (what I am getting during my highest energy linear scan guess), but that's not true TS

>Finally, I found a very similar example to mine on this website,

>http://classic.chem.msu.su/gran/gamess/forum/attach/mp2saddlebf.out

>it's saddle point of 2+2 transition but the history of this file is unclear, e.g. initial scan to determine guess for TS is not to be found. I wonder if somebody could share details on that file very first steps that lead to a successful TS search

>thanks

>JOnas

>On Tue Jun 29 '10 12:28pm, Vladimir A. Mironov wrote
>----------------------------------------------------
>>The point of maximum energy along the reaction coordinate is a good guess for TS search in Firefly. As far as I understand, you have already obtained it and want to find "saddle" with two negative frequencies. You can specify
>>method=schlegel nneg=2
>>in the $statpt group. Alternatively, you may use symmetry, if it is possible for your molecule.

>>The search of "true" saddle points in 2+2 reaction is tricky. Surface scan along the one of negative modes may help you. In this case you should specify the displacement vector from the hessian output and set all atom masses to 1.

>>On Tue Jun 29 '10 4:55am, Jonas Baltrusaitis wrote
>>--------------------------------------------------
>>>I have these two wonderful molecule before and after reaction and the reaction is 2+2 internal reorganization. I would like to calculate transition state but to all fairness I fail due to several facts:
>>>-it's 2+2 organization so there is 4 total atoms involved. whenever I calculate initial guess Hessian (with some other software) it always gives 2 negative frequencies, one for symmetric one for antisymmetric stretch of those 4 atoms. That might mean that there is 2 TS where 2 atoms react...

>>>But I digress... Briefly, in FF I can't get decent initial guess. If I take constrained optimization between initial and final structures, calculate Hessian and start TS search, structure clearly goes the wrong way.

>>>How do I obtain good initial guess for this TS? In Gaussian there is several TS methods where one input initial and final coords. Is there such a thing in FF?

>>>Any other comments?

>>>thanks

>>>Jonas