Can Dalton calculate oscillator strength

Find answers or ask questions regarding Dalton calculations.
Please upload an output file showing the problem, if applicable.
(It is not necessary to upload input files, they can be found in the output file.)

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Can Dalton calculate oscillator strength

Post by xiongyan21 » 10 Jul 2020, 16:03

Some reference in JACS suggested that the first singlet excitation energy of gaseous uracil is about 5.0 eV, and with B3lyp and aug-cc-pvdz, Dalton 2018's 4.650 eV agrees with GAMESS's 4.650 eV(VMN5) and NWCHEM7.0.0's 4.690 eV. GAMESS PBE0 gives 0.22eV deviating from 5.0 eV.
The diagnostic from GAMESS for the nature of the excitation is almost the same as that from Dalton2018,. The oscilator strength by GAMESS TDDFT of the excitation is zero, where that from Dalton is vanishly small and the total one from NWCHEM7.0.0 TDDFT is 0.0000025340, but that from GAMESS eomccsd in cr-eomccsd(t) is not zero, e.g., the y-component is around 0.18, and of course those, more than one kind of expressions, can give excitation energies around 5.0 eV. The diagnostic from both GAMESS and Dalton2018 cannot ascertain this is a Rydberg state. In the regard of eomccsd, the computation of excitation energy completely fails.

Can Dalton calculate oscillator strength?

Very Best Regards!
Attachments
uracil.dal
(149 Bytes) Downloaded 47 times
Last edited by xiongyan21 on 27 Jul 2020, 08:17, edited 8 times in total.

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 11 Jul 2020, 12:01

With B3lyp, the first gasous excitation energy of the enol form hemexazole by B3lyp TDDFT from Dalton2018 agrees with that from EOMCCSD with GAMESS, both agreeing perfectly with that from NWCHEM7.0.0, and from GAMESS using the same method and basis set, which can be compared with that from cr-eomccsd(t) by GAMESS, not only one kind,; the oscillating strength from EOMCCSD is very small, which supports perhaps Rydberg nature, and the diagnostic from Dalton2018, almost identical with that from GAMESS, also supports this. In the regard of the calculation of excitation energy, EOMCCSD perhaps fails.
Using water as the solvent, the first excitation energy with NWCHEM7.0.0 employing B3lyp, where the oscillator strength is as the following,
Dipole Oscillator Strength 0.0793540906
Electric Quadrupole 0.0000004605
Magnetic Dipole 0.0000000259
Total Oscillator Strength 0.0793545770, ABOUT 60 TIMES OF ITS GASEOUS ONE,

is around 0.20eV deviating from that using GAMESS。

The excitation energy of the second state in water using NWCHEM7.0.0 and B3lyp is around 0.32 eV deviating from the experimental value, where the oscillator strength changes from a very small value to around 0.17.

Very Best Regards!
Last edited by xiongyan21 on 15 Jul 2020, 06:44, edited 21 times in total.

hjaaj
Posts: 388
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Can Dalton calculate oscillator strength

Post by hjaaj » 11 Jul 2020, 13:31

Please post your output file.

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 11 Jul 2020, 14:26

Dear Prof. Jensen
I will use ethene optimized by Dalton only using 3-21G to compare, and will put the log file from Dalton2018 when it is finished. Of course, this is just for the comparison of different softwares, and cannot agree with experimental data.

Very Best Regards!

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 12 Jul 2020, 07:47

I should declare this is just for the comparison of different softwares, not considering the requirement of the standard agreeing with experimental data.

First, ethene is optimized using 3-21G and mp2 with Dalton2018, then B3lyp TDDFT and 6-31+G(d) are employed to calculated the first three excited states.

The excitation energies obtained from Dalton2018 agree very well with those form NWCHEM7.0.0 and GAMESS, and the diagnostics here also are approximately the same as those from GAMESS.
Using GAMESS, the oscillator strengths of the first two excitations from TDDFT and eomccsd(degenerate) are all zero, meaning "dark", where the transition dipoles are either zero or small, but the excitation energies from cr-eomccsd(t) are much higher, indicating the quality of excitation energy calculations are not up to the standard.


Very Best Regards!
Attachments
etheneg.log
(442.59 KiB) Downloaded 41 times
Last edited by xiongyan21 on 26 Aug 2020, 15:15, edited 5 times in total.

hjaaj
Posts: 388
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Can Dalton calculate oscillator strength

Post by hjaaj » 12 Jul 2020, 21:10

When you want dipole length oscillator strengths, you must add the keyword

Code: Select all

.DIPLEN
under

Code: Select all

*LINEAR
.

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 13 Jul 2020, 03:02

Dear Prof. Jensen
Thanks a lot.
The oscillator strengths obtained form Dalton2018 using TDDFT for the first two compared here for ethene also agree with those from TDDFT and EOMCCSD from GAMESS in the sense of they are all very small, not in terms of the values, and those from NWCHEM7.0.0 TDDFT are Total Oscillator Strength 0.0000033990 and Total Oscillator Strength 0.0000033993.

It is very interesting that solvation, here, with water, makes the second excited state of the enol form of hemexazole perhaps change its nature and its absorption, according to the calculation from TDDFT by GAMESS. I have not trIed this with solvation by Dalton2018.

Expeoriments are required to verify it.


Very Best regards!
Attachments
etheneg1.log
(444.64 KiB) Downloaded 41 times
Last edited by xiongyan21 on 24 Sep 2020, 00:50, edited 5 times in total.

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 18 Jul 2020, 12:32

By the way, here the geometry is from GAMESS optimization; perhaps the best way to compare certain results in solvation of different softwares is to use geometries optimized by each sofware, respectively.

Very Best Regards!

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 22 Jul 2020, 09:56

I wish to calculate the excited states energies of hemexazole in water
The input is
**DALTON INPUT
.RUN RESPONSE
*PCM
.SOLVNT
H2O
.NPCMMT
0
.NESFP
7
.ICESPH
2
.PCMCAV
.INA
1
2
3
4
5
6
7
.RIN
1.9
1.4
1.4
1.7
1.9
1.8
1.9
.AREATS
0.6
**WAVE FUNCTIONS
.DFT
B3LYP
**RESPONSE
*LINEAR
.SINGLE RESIDUE
.DIPLEN
.ROOTS
3
**END OF DALTON INPUT

The error is

Overall default print level: 0
Print level for DALTON.STAT: 1

HERMIT 1- and 2-electron integral sections will be executed
"Old" integral transformation used (limited to max 255 basis functions)
Wave function sections will be executed (SIRIUS module)
Dynamic molecular response properties section will be executed (RESPONSE module)
--- SEVERE ERROR, PROGRAM WILL BE ABORTED ---
Date and time (Linux) : Wed Jul 22 16:45:54 2020
Host name : hp-HP-Z2-Mini-G3-Workstation

Reason: ILLEGAL KEYWORD IN PCMINP

Total CPU time used in DALTON: 0.00 seconds
Total wall time used in DALTON: 0.00 seconds


QTRACE dump of internal trace stack

========================
level module
========================
2 DALTON_GNRLI
1 DALTON main
========================


What integral print level should be used? What is the illeagal word here?

Very Best Regards!

hjaaj
Posts: 388
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Can Dalton calculate oscillator strength

Post by hjaaj » 22 Jul 2020, 12:37

Look in the output before asking, please, the output tells which keyword is illegal!

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 22 Jul 2020, 12:50

Dear Prof. Jensen
The hole output is listed above, where there is no illeagal words listed.

Very Best Regard!
Last edited by xiongyan21 on 25 Jul 2020, 03:55, edited 5 times in total.

hjaaj
Posts: 388
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Can Dalton calculate oscillator strength

Post by hjaaj » 22 Jul 2020, 13:07

OK, then please upload the total output file.

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 22 Jul 2020, 13:21

Dear Prof. Jensen
Here, the initial geometry is prepared by Avogadro.

Very Best Regards!
Attachments
hemexazole3.log
(14.83 KiB) Downloaded 39 times
b3lype.dal
(251 Bytes) Downloaded 37 times
Last edited by xiongyan21 on 22 Jul 2020, 14:18, edited 1 time in total.

hjaaj
Posts: 388
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Can Dalton calculate oscillator strength

Post by hjaaj » 22 Jul 2020, 13:41

I see. It is an error that the information is not written to the output file, I think you would find it in stdout if it was saved.

Anyway, looking into your problem, your error is that you used ".PCMCAV" instead of "*PCMCAV", as far as I can see (I am not PCM expert).

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 22 Jul 2020, 13:55

Dear Prof. Jensen
Thanks a lot. That is a problem...
I really do not notice it because I do not wish you to do proof-reading.
I will try it tomorrow using a geometry optimized by b3lyp
and aug-cc-pVDZ.
Very Best Regards!

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 23 Jul 2020, 06:38

Dear Prof. Jensen
I have finished the calculation of the excited states of hymexazole optimized, all by aug-cc-pVDZ and b3lyp, and found in water, the first one is around 0.14 eV deviating from that obtained by GAMESS stated previously and 0.12eV from that in the reference, and the diagnostic also supports the perhaps nature change.

Experiments is required to verify it.

By the way, its enol form geometry in water has been searched in water with b3lyp and aug-cc-pvdz with GAMESS, and the following frequency analyses in water then in gas exhibit two and one small imaginary frequencies, respectively, but the excitation energies, oscillator strengths, and diagnostic are close to those of the mp2 and 6-311++G(2d,2p) optimized one. The small frequencies,although do not make significant sense here, should also be eliminated when doing in this way, but it may difficult to converge.

The previously optimized hymexazole in vacuum gives a discrepancy in CH3CN of about 0.14 eVs when compared with experiments, and supports the explanation of the excitation states calculated before.

Very Best Regards!
Last edited by xiongyan21 on 24 Sep 2020, 00:45, edited 6 times in total.

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 25 Jul 2020, 03:57

By the way, the geometry of a pesticide, CAS 1185987-44-3, has been optimized using GAMESS with b3lyp and 6-311++G(2d,2p).

I will try to use Dalton2018 to do this, if the hard disk capacity is enough, because it seems it will store much more on hard disk, and of course it will help to clean the scratch.

Very Best Regards!
Last edited by xiongyan21 on 21 Aug 2020, 11:24, edited 12 times in total.

taylor
Posts: 597
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Tianjin University
Country: China

Re: Can Dalton calculate oscillator strength

Post by taylor » 25 Jul 2020, 07:21

I admit I am now completely lost as to exactly what it is you want to calculate, but your earlier posting was a DFT response calculation. If that is what you want to do and you are concerned about disk space, surely you can specify .DIRECT in the **DALTON INPUT section?

Best regards
Pete

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 25 Jul 2020, 07:52

That will be slower, and may not work if the molecule is complex, basis set is large, and calculation level is high.

Very Best Regard!

taylor
Posts: 597
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Tianjin University
Country: China

Re: Can Dalton calculate oscillator strength

Post by taylor » 25 Jul 2020, 10:02

It will certainly need more operations to run direct, but that does not necessarily make it slower, because there is no input/output, and I/O can be very time-consuming in walltime. I do not understand the assertion that it "may not work if the molecule is complex", because there is nothing about running direct that is related to the complexity of the molecule. Nor the assertion "or if the basis set is large": how can this cause it not to work? If one is not exceeding some dimensions fixed at build time, I cannot see how the basis set size can make a direct calculation fail. The calculations level being "high", well, the .DIRECT functionality is implemented for SCF, DFT, and CC methods, so the only level for which it is not possible is MCSCF.

I'm sure our experience is in no way unusual, and we have run response calculations with up to 2000+ basis functions (direct) with Dalton. If you need even larger (DFT response) calculations you should check out the possibilities of LSDalton. We have used this for up to 20,000 basis functions.

Best regards
Pete

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 25 Jul 2020, 10:45

Upload you log file, or render your reference.

Very Best Regards!

taylor
Posts: 597
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Tianjin University
Country: China

Re: Can Dalton calculate oscillator strength

Post by taylor » 25 Jul 2020, 14:10

I would have thought there were one or two references in the Dalton and LSDalton manuals that would cover these matters...

Best regards
Pete

hjaaj
Posts: 388
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Can Dalton calculate oscillator strength

Post by hjaaj » 25 Jul 2020, 19:08

xiongyan21 wrote:
23 Jul 2020, 06:38
I have finished the calculation of the excited states of hemexazole optimized, all by aug-cc-pVDZ and b3lyp, and found in water, the first one is around 0.14 eV deviating from that obtained by GAMESS stated previously and 0.12eV from that in the reference, and the diagnostic also supports the nature change.
The excitation energies are by default converged to ca. 0.01 eV, because it is waste of CPU time to converge much more than the intrinsic error in the model. THUS 0.14 eV versus 0.12 eV sounds fine. If you want better agreement just for the fun of it, then you must specify tighter convergence thresholds both in wave function and response sections in Dalton, and perhaps also in GAMESS (I do not know the standard defaults in GAMESS).

I guess it could also be because of some minor differences in the PCM part.

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 26 Jul 2020, 10:53

Sorry, that is hymexazole.

With Dalton2018, I am also doing the optimization, and will try to calculate it.

Very Best Regards!
Last edited by xiongyan21 on 21 Aug 2020, 11:26, edited 4 times in total.

xiongyan21
Posts: 286
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: Can Dalton calculate oscillator strength

Post by xiongyan21 » 26 Jul 2020, 14:27

Dear Dr. Jensen
I think the initial geometry may also make a difference, because in the case of GAMESS, it was optimized with mp2 and 6-311++G(2d,2p).
Very Best Regards!

Post Reply

Who is online

Users browsing this forum: No registered users and 9 guests