Ceilidh ... Re^2: $IRC keyword definition

Firefly and PC GAMESS-related discussion club

Re^2: $IRC keyword definition

Dear Dr. Granovsky,
thanks for you suggestions. In my case the problem turned out to be the precision in locating the transition state because I needed to switch from the default convergence (1e-5) to 1e-6.
I also verified that the keyword FORWRD leads to different products depending on the given value (.T. or .F.).

Many thanks for the help
Regards
Luca Maidich

On Mon Jan 31 '11 12:09pm, Alex Granovsky wrote
-----------------------------------------------
>Dear Luca,

>>I've encountered a problem during an IRC job. The problem is that starting from the saddle point (located by a Saddle plus Hessian job) and using the keyword FORWRD=.T. and FORWRD=.F. brings me to the same products.

>Usually, the reasons causing the problem you have encountered are as follows:
>

The imaginary frequency is rather small, and
>
The saddle point was not located with high precision required in this situation
>

>The solution to this problem is to reoptimize saddle point with tighter OPTTOL. This should be fast as you already have Hessian and geometry, and should take only few additional geometry optimization steps.

>Hope this helps,
>Alex Granovsky

Sat Feb 5 '11 4:31pm

This message read 821 times