## Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly

Re^4: Is it possible to get UV-VISIBLE AND RAMAN SPECTRUM USING FIREFLY???

Thomas Pijper
thomaspijper@hotmail.com

Dear Siddheshwar,

UV/vis absorption spectroscopy measures the transmission of light through a sample, which is inversely related to the absorption. Spectra are typically plotted with wavelengths on the x-axis and absorption units (which can be converted to molar absorptivity) on the y-axis. There is no such thing as UV/vis emission spectroscopy AFAIK, possibly you are thinking of fluorescence spectroscopy (which is a different thing entirely).

Using TDDFT, you can calculate the energies and corresponding excitation wavelengths of vertically excited electronic states. These are typically a good approximation to experimentally observed absorption bands. However, whereas experimental bands are typically very wide, the results obtained with linear response TDDFT produce sharp lines only. These can be convoluted with Gaussian shapes in order to obtain a spectrum that is more like an experimental spectrum. One program that can do this is GaussSum (free). See the following site for a working example:

http://gausssum.sourceforge.net/GaussSum2.2/Example/index.html

The real-time TDDFT functionality in the next version of Firefly should be able to produce much more accurate spectra that do not need this postprocessing with Gaussian shapes.

As for your input, I would recommend to start with NSTATE=10. Possibly, a higher value is needed depending on how far into the UV you would like to calculate the spectrum. Only singlet states need to be considered. ISTATE does not have a function as you will not need a gradient or properties (and these can currently not be calculated analytically with Firefly). The Tamm-Dancoff approximation can help to speed up the calculation. Also, I recommend the use of a functional with a large portion of exact exchange, such as BHHLYP. Finally, make sure your ground state geometry is acceptable -- it should have been optimized to a minimum, though not necessarily using the same functional/basis set.

Kind regards,
Thom

On Wed Feb 12 '14 3:33pm, Siddheshwar Chopra wrote
--------------------------------------------------
>Dear Sir,
>Could you tell me what would be the sufficient commands to get a decent UV spectrum.
>Is this sufficient?? \$TDDFT NSTATE=3 ISTSYM=0 ISTATE=1 \$END

>If NSTATE=3, then what does ISTATE mean?
>Could you give me some clue about it? As far I understood we can get ABSORPTION spectra.. Is it possible to get EMISSION spectra for a particular excitation wavelength?

>Regards,

>On Wed Feb 12 '14 8:07am, Siddheshwar Chopra wrote
>--------------------------------------------------
>>Dear Sir,
>>I am for the first time going to try UV-Visible spectra. Could you tell me what kind of spectra does it calculate?? Is it Absorption or Emission UV-Vis spectra??
>>In both the cases, how do I specify the range of wavelengths on x-axis in i/p file? And also what are the y-axis data? Are they the absorption Intensities/ emission Int. or something else?
>>Could you provide me a sample i/p file for say Benzene with all the above data?

>>Regards,
>>
>>
>>On Mon Apr 22 '13 2:55pm, Thomas wrote
>>--------------------------------------
>>>Dear Siddheshwar,

>>>Yes, there is. RAMAN spectra can be obtained with RUNTYP=RAMAN, though you will need to do a few calculations prior to this:

>>>- First, you need to do a geometry optimization.
>>>- Secondly, you need to do a Hessian calculation on the optimized geometry (RUNTYP=HESSIAN). Note that it is possible to do this in the same run as the geometry optimization by specifying \$STATPT HSSEND=.T. \$END
>>>- Finally, you can run a RUNTYP=RAMAN job. You should specify \$FORCE RDHESS=.T. \$END and provide the complete \$HESS group, that you obtained in the previous step, in your input.

>>>As for UV/vis spectra, these can be calculated with TDHF, TDDFT, CIS, and (X)MCQDPT2. The first three methods print out oscillator strengths. (X)MCQDPT2 on the other hand prints out transition dipole moments which you will have to convert to oscillator strengths yourself.
>>>
>>>
>>>Kind regards,
>>>Thom
>>>
>>>
>>>
>>>On Sun Apr 21 '13 9:56pm, Siddheshwar Chopra wrote
>>>--------------------------------------------------
>>>>Dear sir,
>>>>Do we have an option in firefly8 tp get uv visible and raman spectrum??
>>>>Regards,

Thu Feb 13 '14 0:02am