SOC between lowest lying singlet and triplet excited state

Questions
#1

Dear Q-chem users and support,
I am studying excited states based on TD-DFT, while i successfully managed to calculated the SOC, i am not totally confident on my results, i already used the ground state geometry of the lowest lying singlet for starts and the ground state of the complex. Both cases resulted in higher vertical energies related to the selected excited states than the ones calculated before. Then i do not know which ground state geometry should is use for this study. Hope this description helps out.

Thanks in advance

Best regards
SGJ

#2

The objective of your study is not entirely clear from this description. It would be helpful if you posted your inputs and provided more details on the physics you are trying to model.

#3

I’m trying to accurately compute the SOC between the first singlet and first triplet from an Re(I) complex using its ground state geometry to get an insight into its ISC rate. The problem is that i can not replicate the first singlet and first triplet vertical energies or transitions. Since this values are not “correct” the SOC values can not be used for my study.
pd: I have tried with and without solvent effects.

An example input line is:
$molecule

$end

$rem
EXCHANGE b3lyp
BASIS general
ECP general
ecp_fit true
CIS_N_ROOTS 6
MAX_SCF_CYCLES 600
MAX_CIS_CYCLES 50
SCF_ALGORITHM diis
MEM_STATIC 300
MEM_TOTAL 2000
SYMMETRY false
SYM_IGNORE true
CIS_SINGLETS true
CIS_TRIPLETS true
CALC_SOC true
SET_ITER 300
$end

$basis
H
6-31G(d,p)

C
6-31G(d,p)

N
6-31G(d,p)

O
6-31G(d,p)

Cl
6-31G(d,p)

Re
lanl2dz

$end

$ecp
Re
lanl2dz

$end

Thanks in advance and sorry for the poor description.

#4

Thanks, this description makes a lot more sense. When you say you cannot reproduce “correct” excitation energies, what are you comparing to? Experiment, another excited state method, or another software?

#5

I’m comparing them with another software results (gaussian16), sorry for the delayed response.

#6

Thanks, if you observe that different programs produce different TDDFT excitation energies, here some things to check:

  • Ensure that the geometry is the same and verify that nuclear repulsion energy is exactly the same
  • Compare total SCF energy between the outputs, although keep in mind that using LANL2DZ will cause Q-Chem to invoke ECP reconstruction (https://manual.q-chem.com/5.3/Ch8.S11.html). ECP reconstruction is expected to have insignificant effect on energy differences. It is also possible to evaluate ECP integrals in Q-Chem by quadrature via ECP_QUAD=TRUE in $rem, which should produce an absolute SCF energy comparable to what other codes produce.
  • If the SCF energy is a close match, the next step is to compare TDDFT excitation energies.

If you’d like more assistance navigating the differences between Q-Chem and G16, please feel free to send full outputs to support@q-chem.com. We will review them in confidence.

1 Like