Normal Mode Analysis+IR/Raman

[Rajini]

Dear friends,
I need some help regarding running Gaussian 03. I have a complex molecule (64 atoms with 6 types of atom)...and finished geometry optimization successfully using method b3lyp with basis set pVDZ.. it took 4 days to complete this optimization procedure...Now i want to do frequency calculation with the same method and basis set...For this freq. calculation will it take the same time as it took for geometry optimization..more? or less? or the same?..When i did my geom. opt. after 1 days it got aborted due to lack memory...so i took the last value (i.e., coordinates values) from the aborted file and then repeated..so like this i did 4 times and finally everything converged successfully...Now if i do this freq. calculation using the converged values...if it get aborted...what should i use from the aborted file and repeat it?
Also for another molecule i successfully made a freq. calculation...but now i don't know how to interpret it..i just want to know how to plot Raman and IR spectrum...which values in X and X should i take for plotting from the result output file?is there any thing to tell about Normal mode analysis from this output file?or how to assign certain modes to certain vibrations peaks??
Thanks for ur reply...
Rajini

 

[TeethWhitener]

For this freq. calculation will it take the same time as it took for geometry optimization..more? or less? or the same?..

It will take much longer. Gaussian's procedure for frequency calculations is to calculate the potential gradient at each atom (by calculating the single point energy at 3 points in each dimension =9 points for N atoms = 9N). There are a few tricks it uses to speed this process up, but plan on frequency calculations taking at least 10 times longer than geometry optimizations.

Now if i do this freq. calculation using the converged values...if it get aborted...what should i use from the aborted file and repeat it?

Gaussian provides some ability to restart failed gradient calculations. I haven't done it in years, so I don't remember precisely what to do, but it should be in the documentation.

Also for another molecule i successfully made a freq. calculation...but now i don't know how to interpret it..i just want to know how to plot Raman and IR spectrum...which values in X and X should i take for plotting from the result output file?is there any thing to tell about Normal mode analysis from this output file?

Simply doing a frequency calculation will return all normal modes for the molecule. To calculate IR/Raman spectra, you also need to tell the program to calculate the matrix elements corresponding to these processes (transition dipole moment for IR and polarizability tensor for Raman). Gaussian can do all this; you just need to go through the documentation to find how to implement it. The output will be a list of oscillator strengths corresponding to normal modes. One tip: if your molecule has symmetry, now is definitely the time to use it. It vastly speeds up frequency calculations.

or how to assign certain modes to certain vibrations peaks??

Gaussview (or another good visualization program) can do this. The Gaussian output file actually contains this info in the form of a matrix of displacements for each of the 3N normal modes in a N-atom molecule (including translations/rotations). A visualization program will translate this matrix into a lovely little movie of the different vibrational modes.