Hello Brajendra,

The easy way to calculate the excitation states is with TD-SCF method. First you can optimize the molecule with its spin and charge in ground state (depending on molecule an excitation don't show a change in geometry.

TD-SCF sometimes don't gives the most accurate results but I recommend to read the book: Cramer J. C., Essentials of Computational Chemistry. There there are explained other methods for more accurate excitation calculation as truncated-CI, MCSCF, etc. depending on you study. Also recommend see the link of gaussian.

In output of STATESEXAMPLE.log you could see an example of excitations with its information (energy, symmetry...)

I hope it helps you,

Joaquim Rius

http://www.gaussian.com/g_tech/g_ur/l_keywords09.htm

Mr. Kusmariya,

You can use the following combinations of methods+keywords:

1. Optimization of Sx states in gas-phase (GP) and solution

Keywords/method: td=(nstates=6,root=1) SCRF=PCM (or scrf=cpcm,solvent=water)

This is an example for number of states 6 and a state of interest 1. They chan be changed changing the corresponding numbers.

The computations in solution have involved PCM approach in this example in water, using "scrf=cpcm,solvent=water". If you use only "SCRF=PCM", than you should open the *.gjf file with Gaussian, and to add at the end in the "molecular specification section" "2.238 3". The "2.238" is dielectric constant, for this example is CCl4, but you can change this value (attachment, page 3).

2. Optimization of Sx/Tx states in GP and solution:

- Method/keyword: rcis=(nstates=6,root=1)/sdd or RCIS(50-50)/SDD TD test

As basis set here is ECPs (SDD) which is very suitable for coordination compounds

The change of the methods you can carry out using the shown in the attachment approach (page 3) as well.

- Method/keyword: RPA(Triplets,Nstates=6,Root=3)/SDD

This is an example like point 1, computing only in this case Ts, but it can be changed. The computation is in GP.

- Method/keyword: RPA(Triplets,Nstates=6,Root=3)/SDD SCRF=PCM

This approach involves also PCM, and thus you should add by the mentioned above manner a solvent type, for example "2.238 3", as shown in point 1, too.

NB! Optimization of the geometry is in GS (Keyword: "Opt" for geometry optimization). There has not significant difference between the ground and excited state geometries. Those are processes joined with valence electrons.

In this context, if you would like to predict excitations of different tautomers, resonance forms, you need optimization in GS to each of them.

Please pay attention to an attachment where are shown experimental and theoretical EAs, Fs at different conditions (attachment pages 1,2).

To compare the accuracy of those methods you can use data in [Ref. 1] attachment, for example, containing EAs and Fs of structurally similar derivatives in different solvents and concentrations.

However, to compare theoretical accuracy to data of coordination compounds, it is better to use experimental spectra of 4f-elements, for example, because of they have well distinguished and not so broad profiles, rather than to use Fs specra of transition metal complexes.

thanks Joaquim MRius Bartra and Bojidarka B. Ivanova for your kind suggestions

In your situation, it could simply be specifying an alternate spin multiplicity in Gaussian.  If the ground state is a singlet, optimize the triplet and you will arrive at a geometry for an excited state.  For spectra, you could optimize two states and calculate the energy difference between them.  Alternatively, calculate a vertical excitation energy by optimizing the ground state and re-calculating the energy of the ground state without re-optimizing the geometry.