Thomas Pijper
thomaspijper@hotmail.com
>I checked the excitation state transition and it belongs to the HOMO----->LUMO Transition (I guess it is optical band gap). Then how can we call it fluorescence emission wavelength???
Fluorescence is the relaxation of an electronically excited state to the ground state by release of a photon. So, the S0-S1 energy gap will correspond to the emission wavelength.
>Also there are several other Highly probable transitions from TDDFT o/p file which could result in emission...
That is not possible. Kasha's rule states that emission predominantly takes place from the first excited state. Note that the oscillator strengths printed by Firefly should be ignored - they provide information on excitation, not emission.
>Additionally how do we conclude that these are RADIATIVE transitions???
There are always multiple pathways by which an excited state can relax. Some of them are indeed nonradiative. However, the amount of energy that needs to be released is always the same. That is why the S0-S1 gap is pretty much all you need to know. (Unless relaxation involves intersystem crossing, but that is a different story altogether.)
>As you advised I will read up about CASSCF/MCSCF.. As far as I have read, we need to supply the active space from TDDFT run to run CASSCF calcs.. But what is so special about these calcs is a matter of study...
TDDFT is unrelated to CASSCF, it is an entirely different method. An active space indeed needs to be constructed, but this is typically from HF or DFT ground state orbitals. I advise you have a look at Jensen's 'Introduction to Computational Chemistry'. Herein, a lot of methods in CC are explained, amongst which CASSCF and TDDFT.
Kind regards,
Thom
On Wed Mar 26 '14 9:02pm, Siddheshwar Chopra wrote
--------------------------------------------------
>Dear Thomas,
>Yes I could locate the TDDFT EXCITATION ENERGIES and found the S0-S1 gap. But Thomas my doubt is still there.. I checked the excitation state transition and it belongs to the HOMO----->LUMO Transition (I guess it is optical band gap). Then how can we call it fluorescence emission wavelength???
>Also there are several other Highly probable transitions from TDDFT o/p file which could result in emission...
>Additionally how do we conclude that these are RADIATIVE transitions???
>Again my curiosity is how to find fluorescence emission ???
>As you advised I will read up about CASSCF/MCSCF.. As far as I have read, we need to supply the active space from TDDFT run to run CASSCF calcs.. But what is so special about these calcs is a matter of study...
>Regards,
>On Wed Mar 26 '14 4:06pm, Thomas Pijper wrote
>---------------------------------------------
>>Dear Siddheshwar,
>>> Please tell me how should I find S0-S1 gap?
>>In the output, locate the "EQUILIBRIUM GEOMETRY LOCATED" string. Then, scroll upwards to the output of the last TDDFT calculation. You'll come across a table with the header "TDDFT EXCITATION ENERGIES", the first entry in this table provides the energy of the first excited state (the S1->S0 energy).
>>> I am a bit confused with CASSCF/MCSCF calculations?
>>Compared to TDDFT, those calculations are quite complicated to execute. Before attempting them, I suggest you have a look at an introductory textbook on computational chemistry as well as study some input examples. Last but not least, carefully read the Firefly manual.
>>
>>
>>Kind regards,
>>Thom
>>
>>
>>On Wed Mar 26 '14 3:26pm, Siddheshwar Chopra wrote
>>--------------------------------------------------
>>>Dear Sir,
>>>I am glad to inform you that I have performed TDDFT calculation for the conditions suggested by you. And I have got the message "EXECUTION OF FIREFLY TERMINATED NORMALLY" at the end too (Is it called converging?. Please tell me how should I find S0-S1 gap? Is it the same HOMO/LUMO gap? Could you briefly describe the steps to get it?
>>>I am a bit confused with CASSCF/MCSCF calculations? It seems I have to add active space to these calcs. to get the most probable transitions. Please make me understand this.
>>>Regards,
>>>On Wed Mar 26 '14 1:54pm, Thomas Pijper wrote
>>>---------------------------------------------
>>>>Dear Siddheshwar,
>>>>For calculating the fluorescence emission wavelength, you should just follow the advice I gave you in an earlier message. Just perform an TDDFT excited state geometry optimization with ISTATE=1 and NSTATE=3, starting from your ground state optimized geometry. When the calculation has converged, the S0-S1 energy gap for the optimized geometry corresponds to the fluorescence emission wavelength (in absence of solvation).
>>>>
>>>>
>>>>Kind regards,
>>>>Thom
>>>>
>>>>
>>>>On Wed Mar 26 '14 1:32pm, Siddheshwar Chopra wrote
>>>>--------------------------------------------------
>>>>>Dear Thomas,
>>>>>Let me discuss in points::
>>>>>1) I ran TDDFT calcs for ground state configuration...So that means I got correct UV abs. spectra..Is it right?
>>>>>2) NSTATE was kept high because it gave me more number of points for which osc. strength was calculated.. More the number of points, better the graph.. Is it right? In my case I kept it high.
>>>>>3) Now please guide me how to obtain fluorescence spectra once I have the UV abs. spectra.. I have the active space ready with me.. I just know these transitions.. How do I run for these calculations now? Any commands would be welcome..
>>>>>Also how do I now ensure that the lowest excited state is optimized before the emission spectra run?
>>>>>Regards,
>>>>>On Wed Mar 26 '14 1:03pm, Thomas Pijper wrote
>>>>>---------------------------------------------
>>>>>>Dear Siddheshwar,
>>>>>>TDDFT geometry optimizations are quite simple. In addition to the usual parameters for an optimizations (RUNTYP=OPTIMIZE, the $STATPT group, etc), one only needs to specify the state of interest with the ISTATE keyword of the $TDDFT group. ISTATE=1 specifies the first excited state. NSTATE should ideally be set a bit higher, in case the desired state is not yet the lowest in energy at the start of the diagonalization procedure. NSTATE=10 is too high IMO, NSTATE=3 should be sufficient for most cases.
>>>>>>> But doesn't this run give me absorption spectra??
>>>>>>UV/vis absorption data derived from energies and oscillator strengths is valid only for TDDFT calculations on the optimized ground state geometry. Geometries that do not correspond to a minimum on the S0 PES should not be used when calculating UV/vis spectra.
>>>>>>
>>>>>>
>>>>>>Kind regards,
>>>>>>Thom
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>On Wed Mar 26 '14 12:17pm, Siddheshwar Chopra wrote
>>>>>>---------------------------------------------------
>>>>>>>Dear Thomas,
>>>>>>>Yes I am referring to the fluorescence emission. I have as you suggested optimized the system for ground state.. Then I ran for TDDFT using the following commands(some omitted):
>>>>>>> $CONTRL CITYP=TDDFT SCFTYP=RHF $END
>>>>>>> $CONTRL DFTTYP=B3LYP COORD=UNIQUE NZVAR=1 $END
>>>>>>> $BASIS GBASIS=N31 NGAUSS=6 $END
>>>>>>> $STATPT opttol=10E-7 NSTEP=100 Method=GDIIS $END
>>>>>>> $SCF fdiff= .f. diis=.t. $END �
>>>>>>> $DFT HFX(1)=0.15 LMAX=41 NRAD=99 $end
>>>>>>> $ZMAT DLC=.t. AUTO=.t. $END
>>>>>>> $SCF NCONV=6 $END
>>>>>>> $TDDFT NSTATE=10 ISTSYM=0 ISTATE=1 TDA=.t. $END
>>>>>>> $DATA
>>>>>>>And I copied the ground state coordinates to a TDDFT i/p file. Are the above commands correct?? But how do I ensure that the first excited state is optimized? How to check for that? But doesn't this run give me absorption spectra?? I could find only excitation energies vs osc. strengths.
>>>>>>>Please help.
>>>>>>>Regards,
>>>>>>>
>>>>>>>
>>>>>>>On Wed Mar 26 '14 11:45am, Thomas Pijper wrote
>>>>>>>----------------------------------------------
>>>>>>>>Dear Siddheshwar,
>>>>>>>>Are you referring to emission as in fluorescence emission? Fluorescence emission bands can be calculated with relative easy with Firefly 8.1.0. The steps are as follows:
>>>>>>>>1. Optimize the ground state geometry of your system.
>>>>>>>>2. Use TDDFT to optimize the geometry of the first excited state, which effectively allows the ground state geometry to relax after excitation. The energy difference between the S0 and S1 states of the relaxed geometry corresponds to your emission band. Note that, as with UV/vis spectra calculated by TDDFT, this will give a single energy value whereas an experimentally observed spectral band will always have a certain width.
>>>>>>>>As far as I know, it is currently not possible to include the effects of solvation in the calculation.
>>>>>>>>
>>>>>>>>
>>>>>>>>Kind regards,
>>>>>>>>Thom
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>On Wed Mar 26 '14 10:42am, Siddheshwar Chopra wrote
>>>>>>>>---------------------------------------------------
>>>>>>>>>Dear Sir,
>>>>>>>>>I read a paper which used Firefly and discussed about TDDFT calculations. There was a mention of LUMO to HOMO transitions in the same work. It said "by �combining �both �the �experimental �data �and �the �calculation
>>>>>>>>>results �with �the �Franck-Condon �principle, �the �simplified �mechanism �for downward vertical transition can be predicted as" ...... And it was termed as Emission band.
>>>>>>>>>Sir, is it possible to study emission spectra using Firefly? If yes how can we go about it?
>>>>>>>>>Regards,
[ This message was edited on Wed Mar 26 '14 at 11:09pm by the author ]