Inconsistency in SS-PCM/TDDFT results: Polar and Non-polar solvents

Questions
#1

Hi developers and users,
I am seeking to perform SS-PCM/TDDFT calculations to study the vertical excitation of the 4-Amino-4’-nitrobiphenyl molecule in various polar and non-polar solvents.
The input goes as follows:
$molecule
read opt_BP.txt
$end

$rem
METHOD wB97X-D
BASIS 6-31+G(d,p)
JOB_TYPE sp
max_scf_cycles 200
SCF_CONVERGENCE 8
THRESH 14
INCDFT FALSE
PRINT_GENERAL_BASIS TRUE
PRINT_ORBITALS 10
pcm_print 1
RPA 2
cis_singlets TRUE
cis_triplets FALSE
cis_n_roots 10
NTO_PAIRS 2
SYMMETRY_IGNORE TRUE
SYMMETRY FALSE
cis_state_deriv 10
CIS_DYNAMIC_MEM TRUE
CIS_RELAXED_DENSITY TRUE
CIS_MOMENTS TRUE
SOLVENT_METHOD PCM
TDDFT_PCM FALSE
$end

$pcm
Theory IEFPCM
NONEQUILIBRIUM TRUE
ChargeSeparation Pekar
StateSpecific Perturb
TdNonEq 1
$end

$solvent
Dielectric 78.4
OpticalDielectric 1.78
$end

@@@

$molecule
read
$end

$rem
METHOD wB97X-D
BASIS 6-31+G(d,p)
JOB_TYPE sp
max_scf_cycles 200
SCF_CONVERGENCE 8
THRESH 14
INCDFT FALSE
PRINT_GENERAL_BASIS TRUE
PRINT_ORBITALS 10
pcm_print 1
RPA 2
cis_singlets TRUE
cis_triplets FALSE
cis_n_roots 10
NTO_PAIRS 2
SYMMETRY_IGNORE TRUE
SYMMETRY FALSE
cis_state_deriv 10
CIS_DYNAMIC_MEM TRUE
CIS_RELAXED_DENSITY TRUE
CIS_MOMENTS TRUE
SOLVENT_METHOD PCM
TDDFT_PCM FALSE
$end

$pcm
Theory IEFPCM
NONEQUILIBRIUM TRUE
ChargeSeparation Pekar
StateSpecific Perturb
TdNonEq 2
$end

$solvent
Dielectric 78.4
OpticalDielectric 1.78
$end

While performing the job,the molecule exhibits a CT state at the S10 state in vacuum. However, in water, this CT state shifts to the S9 state with a decrease in excitation energy. When performing the same calculation in n-Hexane, the CT state further shifts to the S8 state, showing an even lower excitation energy than in water. I have also conducted similar calculations in various other solvents such as benzene, ethanol, and formamide, and have observed a consistent trend: the red shift of the CT state is more pronounced in non-polar solvents compared to polar solvents, which is contrary to what is normally expected.

I am not sure if this discrepancy is due to errors or missing keywords in the input file causing this inconsistency in the results.

I would greatly appreciate any assistance or insights you might offer to help resolve this issue. Thank you for your time and consideration.

#2

can you try removing the TdNonEq keyword in all of the jobs? That is associated with the constrained variational principle of X. Y. Li, which is rather different from the solvation models developed by Herbert, You, Mewes, and Dreuw (which are what most of the rest of those keywords are for). We should perhaps do some renaming…

#3

Dear Prof. Herbert,

I followed your suggestion and removed the ‘TdNonEq’ keyword. Now, the program is yielding better results with both polar and non-polar solvents. Thank you for your guidance.

#4

Okay good. This is not the first time I’ve thought that was confusing, so I’m going to submit a ticket to do some renaming (and add comments to the manual) for a future release.