| 8:54:47 AM PDT - Thu, Jun 7th 2012 |
|
Point 2: With the geometry:actlist set ( http://nwchemgit.github.io/index.php/Release61:Geometry#Applying_constraints_in_geometry_opt... ), the geometry optimization will only take into account the gradients and hessian of the moving atoms. By default it will follow the first negative mode. See http://nwchemgit.github.io/index.php/Release61:Geometry_Optimization#Initial_Hessian and the section after that on how the initial mode is defined and found, and how you can modify if needed.
In general, if an analytic form of the gradient or hessian is not available, the code will do this automatically. Hence, no need to specify anything. But, for numerical: just add numerical to the end of the task line, i.e.
task dft frequencies numerical
Thanks,
Bert
Quote:Noraplanas Jun 6th 8:25 pmThanks for your answer.
Regarding point 2: My concern is if we are not able to calculate the frequencies in the numerical framework ( point 1) how is it possible to find the transition state without being sure we have the correct imaginary frequency.
Additionally, I haven't found infromation on how to run numrical frequencies in the manual... could somebody point me where to look?
Thanks!
Nora |