## Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly

Re: Performing a transition state calculation
Hi!

wxMacMolPlt nicely reads the firefly output vibrational modes of a hessian calculation (which you perform by adding the keywords  hssend=.t. in \$STATPT group of the optimization run and reading a bit more in the manual for understanding pits and caveats).

For wxMacMolPlt compatibility, you just have to use the "-legacy" option. Just run firefly as:
firefly801 -i inp.inp -p -f -run -legacy > out.out

cheers

On Tue Nov 18 '14 6:57pm, Kevin wrote
-------------------------------------
>Hello, fellow fireflies,

>I'm still somewhat new to the quantum chemical scene, so forgive my naïveté.

>I'm attempting a Firefly emulation of Jans Jensen's excellent tutorial on conducting a GAMESS transition state calculation of flouride anion reacting with chloromethane (found here: http://molecularmodelingbasics.blogspot.com/2009/08/finding-transition-state-sn2-reaction.htmlhttp://molecularmodelingbasics.blogspot.com/2009/08/finding-transition-state-sn2-reaction.html). The procedure is essentially as follows:

>1) While holding constant the fluorine to carbon intermolecular distance, optimize the geometry.
>2) Determine the imaginary frequency "frame".
>3) Use this imaginary frequency "frame" data as the initial geometry guess of a saddle point calculation.

>The input and execution of the initial optimization, for the purpose of finding an imaginary frequency, are no problem. Where I get hung up is in taking the output to serve as the initial geometry for a saddle point calculation. I understand that I need both the \$hess as well as the frequency (specifically, the imaginary frequency) data. Dr. Jensen, in his tutorial, simply opens up the GAMESS .log output in wxMacMolPlt, chooses the imaginary frequency "frame" of interest, and uses this to directly produce a new input file for the SADPOINT calculation. By contrast, wxMacMolPlt does not, as far as I can tell, read the "frames" of an optimization of a Firefly .out. Perhaps Avogadro can substitute, but Avogadro crashes whenever I attempt to open even a simple Firefly calculation (does anyone else have this problem?).

>While I can visualize the frames nicely in ViewMol3D, it lacks MacMolPlt's function to produce an input file. This, then, is the crux of my problem: What data from the raw output of optimization can/should I take to implement into a saddle point input, as Dr. Jensen does to perform his second, saddle point calculation, following optimization?

>For instance, while there are over 20 "frames" in the optimization sequence, there are only two \$hess groupings. So which would I choose? Additionally, how do I implement the imaginary frequency information into the saddle point calculation input file?

>I hope I've framed my question clearly enough. I've attached the output file I produced. Please let me know if I can clarify anything.