Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly

Solntsev Pasha
solntsev@univ.kiev.ua

Ok, now geometry situation is clear for me. So what about emission energy? Is it possible to compute this energy?

>Note that Kasha's rule http://en.wikipedia.org/wiki/Kasha%27s_rulehttp://en.wikipedia.org/wiki/Kasha%27s_rule says that the emission is only possible from the lowest-energy state. Is state #2 from your TDDFT calculation the first or the second excited state? I'm asking about the numbering of the states: do you use the numeration as in CIS and TDDFT (state #1 is the first excited state) or as in MCSCF (state #1 is the ground state)?

So, if we are talking about luminescence properties we must always study first excited state,am i right?

>What is the nature of the transition of interest (pi->pi* or n->pi*)? TDDFT and CIS often give different order of occurrence of these states (e.g., in pure-GGA TDDFT n->pi* transition may be lower than pi->pi* transition, while in CIS it is quite the reverse). Make sure that you take the correct state for optimization. In addition, if you need excited state geometry optimization in TDDFT, you may use GAMESS-US.

ok. I will use GAMESS US to compute tddft-gradient.

>However, higher-level methods, such as MCSCF/XMCQDPT and EOM-CCSD, are more reliable. While EOM-CC is not implemented in Firefly and can be applied only to relatively small molecules (no more than 2 benzene rings), MCSCF/XMCQDPT can be applied to much larger molecules even on PC and allows for geometry optimization (although not as fast as by CIS). See the tutorials on GAMESS-US website: http://www.msg.chem.iastate.edu/tutorials/excited.pdfhttp://www.msg.chem.iastate.edu/tutorials/excited.pdf, http://www.msg.chem.iastate.edu/tutorials/mcscf1.pdfhttp://www.msg.chem.iastate.edu/tutorials/mcscf1.pdf, http://www.msg.chem.iastate.edu/tutorials/mcscf2.pdfhttp://www.msg.chem.iastate.edu/tutorials/mcscf2.pdf

I want understand tddft and cis methods towards UV-vis, then i will try MCSCF/XMCQDPT.

Thanks for answer.

Best, Pavel.


>>UV-Vis:
>>1) I made geometry optimization with symmetry (C2v), DFT(PBE96/6-31G*) => get geometry of main state.
>>2) I made TDDFT single point energy run (using optimized coordinates) at the same basis set to get excitation energies. Ok, state #2 is state of my interest. So, now i get information about Absorption properties of my molecule(oscillation strength, energy).

>>Luminescence:
>>I need to know energy of following states

>> S1*
>> /\ \
>> ||   S1
>> ||   ||
>> ||   \/
>> ||   S0*
>> || /
>> S0

>>Where:

>>S0 - main state
>>S1* - excited "unrelaxed" state
>>S1 - relaxed excited state
>>S0* - unrelaxed main state.
>>"==>" - emission and excitation, respectively.

>>So, from step 1 and 2 i got S0 and S1* states. They have same geometry but different properties(energy, symmetry ...)

>>3) I know, i can't use TDDFT to get geometry of S1 state. So i chose cis approximation.
>>3a) $CONTRL RUNTYP=optimize CITYP=cis $END
>>    $cis ncore=0 nstate=2 $end
>>ok, i got equilibrium geometry of S1 state.

>>4) my main problem: How can i get energy and geometry of S0* state?
>>Any suggestions?  

>>Also, how can i understand istate=-2 notation in tddft/cis groups?
>>I read from  DFT link:
>>"The state-tracking feature of the PC GAMESS/Firefly' TDDFT code can be activated by selecting negative value of istate in the $TDDFT group. It is intended for geometry optimization of the excited states in the case of root flipping.".
>>I don't understand how can i use this option for geometry optimization (tddft does not support gradient yet).
>>
>>
>>Best, Pavel.
>>
>>
>>