Calculating oscillator strengths


Just Got Here
Hi,

When I try to calculate oscillator strengths using CC2, it prints out the oscillator strength for the first excited state, but not for the others. Does anyone know, how to get the oscillator strengths for the other excited states?

There are three very similar previous questions on this forum, but none of them have attracted a response.
http://nwchemgit.github.io/Special_AWCforum/st/id3440/TCE_transition_dipole_moment...
http://nwchemgit.github.io/Special_AWCforum/st/id2510/oscillator_strength_for_EOM-...
http://nwchemgit.github.io/Special_AWCforum/st/id1601/Bug%3A_Oscillator_strengths_...

Thanks for the help,

All the best,

Matthias




Here's my input:




  TITLE "Egg without spinach."
MEMORY 2500 mb

  GEOMETRY
O -0.21588246 0.16922727 0.78153341
H 0.74571464 0.21510024 0.82795358
H -0.47503217 0.82191249 1.44197301
END

  BASIS SPHERICAL
H library cc-pVDZ
O library cc-pVDZ
END

  SCF
SINGLET
RHF
END

  TCE
HF
CC2
FREEZE atomic
NROOTS 5
THRESH 1.0e-4
DIPOLE
END
set tce:thresheom 1.0e-4
set tce:threshl 1.0e-3

  TASK TCE ENERGY




And here's the relevant section of the output:




EOM-CCSD transition moments / hartree
--------------------------------------------
X 0.0000000 Y 0.3691437 Z 0.0000000
Oscillator Strength 0.0268294
--------------------------------------------

EOM-CCSD dipole moments / hartree & Debye
------------------------------------
X 0.0000000 0.0000000
Y 0.0000000 0.0000000
Z 0.2477967 0.6298413
Total 0.2477967 0.6298413
------------------------------------

Singles contributions
6a1 (alpha) --- 5b2 (alpha) -0.9513669003

Doubles contributions

Excited state root  2
Excitation energy / hartree = 0.872579877010030
/ eV = 23.744116662209223

Excited state root  3
Excitation energy / hartree = 1.153901938036031
/ eV = 31.399282696456101

Excited state root  4
Excitation energy / hartree = 1.268696577055734
/ eV = 34.523005088977740

Excited state root  5
Excitation energy / hartree = 1.416550171084003
/ eV = 38.546307800889565

Task  times  cpu:      737.0s     wall:      844.7s




P.S. I'm running NWChem 6.8.

Forum Vet
Actually, the calculation gets stuck at
CCSD iterations
-----------------------------------------------------------------
Iter Residuum Correlation Cpu Wall V2*C2
-----------------------------------------------------------------
When HF is deleted and io ga is added, it still gets stuck.
QA test tce_c2_cc2 also gets stuck.

I arbitrarily choose QA tests tce_n2, tce_cr_eom_t_ch_rohf, tce_eomccsd_gradient, tce_eomsd_somsol1,tce_eomsd_eomsol2,tce_lr_ccsd_tq, tce_c20_triplet, tce_hyperpolar_ccsd_small, tce_helium_imaginary and tce_eaccsd_ozone to eaxamine whether TCE can still work on my computer without the recompilation of NWChem6.8, fortunately all being able to finish. Thus, perhaps there is a bug in cc2 of TCE.

QA test tce_c20_triplet has many warnings during ROHF iterations, diffferent from that in the test.
QA test tce_hyperpolar_ccsd_smallgets stuck when running paralelly.

The output of QA test tce_helium_imaginary is not provided, and I do not know the physical meaning of the calculation, also whether it is correct.

QA test tce_rohf_uccsd can pass, and when I modify it into ccsd(t), it also can pass.

I have reported these and put all the log files on Github.

By the way, in order to obtain osccilator strengths, one should do left-hand eom-ccsd, but it seems NWChem only can do right-hand iterations, whereas GAMESS has this capability.

Forum Vet
Github issue with fix at
https://github.com/nwchemgit/nwchem/issues/143


Forum >> NWChem's corner >> Running NWChem