Facing issue in optimization

Questions
#1

Hi,
I want to optimize geometry at first singlet excited state (S1) using SF-TDDFT, but I am facing issues. Firstly I optimized geometry without a solvent method at S1, and It gets optimized. Then I took the optimized coordinates and put the job with the solvent method (PCM). Please suggest to me how I will optimize it.

$molecule
0 3
50 Atoms (C,N,H)
$end

$rem
BASIS = cc-pVDZ
EXCHANGE = LRC-wPBEh
JOB_TYPE = OPTIMIZATION
OMEGA = 141
MAX_CIS_CYCLES = 700
SCF_CONVERGENCE = 8
MAX_SCF_CYCLES = 700
THRESH = 14
UNRESTRICTED = true
GEN_SCFMAN = true
MOM_START = 6
SYMMETRY_IGNORE = true
SYMMETRY = false
SPIN_FLIP = 1
CIS_N_ROOTS = 4
CIS_STATE_DERIVATIVE = 3
GEOM_OPT_MAX_CYCLES = 100
MEM_TOTAL = 10000
SOLVENT_METHOD = PCM
$end

$solvent
DIELECTRIC 7.425700
OPTICALDIELECTRIC 1.974025
$end

At the end of output
684 -1145.5311244334 9.85e-06
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
685 -1145.5310104245 1.07e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
686 -1145.5308398509 1.28e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
687 -1145.5310222208 1.06e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
688 -1145.5310838641 1.03e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
689 -1145.5311146973 1.04e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 95.99 / 96
690 -1145.5308440893 1.02e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
691 -1145.5301723763 1.90e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
692 -1145.5297224383 2.85e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
693 -1145.5299011479 2.75e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
694 -1145.5300529642 2.61e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
695 -1145.5300801302 2.57e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
696 -1145.5300301038 2.67e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
697 -1145.5299398709 2.83e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
698 -1145.5298593251 2.99e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
699 -1145.5297268325 3.10e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
700 -1145.5298243116 2.94e-05
Maximum Overlap Method Active
MOM overlap: 98.00 / 98
Maximum Overlap Method Active
MOM overlap: 96.00 / 96
gen_scfman_exception: SCF failed to converge

Q-Chem fatal error occurred in module libgscf/gen_scfman/gen_scfman_main.C, line 218:

Error in gen_scfman

Please submit a crash report at Q-Chem Crash Reporter

#2

The proximate reason for this failure is that the SCF fails to converge, probably because the MOM algorithm cannot find a consistent set of MOs. The deeper problem is that I have no idea what you are trying to accomplish with this input. Why turn on MOM starting in the 6th SCF cycle? It does not make sense. In fact, why use MOM at all without a $occupied section or for this kind of job? Do you have some kind of degeneracy?

#3

Dear @jherbert
I turned on MOM because earlier, the SCF started oscillating between 10-5 and 10-6. then I gave mom_start = 5 it did not work the same problem happened, then I gave mom_start = 6.
Please correct me where I am going wrong.

#4

You are probably better off just letting it run without MOM. THRESH is set tightly, which is a good idea for large molecules, you might consider use of BASIS2 (= STO-3G or something small) if you have convergence problems.

#5

Thanks for the suggestion @jherbert
I gave this input:

$rem
BASIS = cc-pVDZ
BASIS2 = STO-3G
EXCHANGE = omegaB97XD
GUI = 2
JOB_TYPE = OPTIMIZATION
OMEGA = 140
MAX_CIS_CYCLES = 700
SCF_CONVERGENCE = 8
MAX_SCF_CYCLES = 700
THRESH = 14
UNRESTRICTED = true
GEN_SCFMAN = true
SYMMETRY_IGNORE = true
SYMMETRY = false
GEOM_OPT_MAX_CYCLES = 100
SPIN_FLIP = 1
CIS_N_ROOTS = 4
CIS_STATE_DERIVATIVE = 3
MEM_TOTAL = 2000
SOLVENT_METHOD = PCM
$end

$solvent
DIELECTRIC 7.425700
OPTICALDIELECTRIC 1.974025
$end

I got this Error:

Exchange: 0.2220 Hartree-Fock + 1.0000 wB97X-D + LR-HF
Correlation: 1.0000 wB97X-D
Using SG-2 standard quadrature grid
Dispersion: Grimme D
using 20 threads for integral computing

OpenMP Integral computing Module
Release: version 1.0, May 2013, Q-Chem Inc. Pittsburgh

/root/qchem/bin/qchem: line 126: 6851 Aborted (core dumped) ${QCPROG_S} ${inp} ${scr}
remove work dirs /root/qchem/scratch/qchem6833.0 – /root/qchem/scratch/qchem6833.3

#6

Use GEN_SCFMAN = FALSE in $rem. (In the latest version of Q-Chem this is printed as the error but perhaps that message is not present in your version.)

#7

Hi @jherbert
I tried with this input:

$rem
BASIS = cc-pVDZ
BASIS2 = STO-3G
EXCHANGE = omegaB97XD
JOB_TYPE = OPTIMIZATION
OMEGA = 140
MAX_CIS_CYCLES = 700
SCF_CONVERGENCE = 8
MAX_SCF_CYCLES = 700
THRESH = 14
UNRESTRICTED = true
symmetry_ignore = true
symmetry = false
geom_opt_max_cycles = 100
GEN_SCFMAN = false
SPIN_FLIP = 1
CIS_N_ROOTS = 4
CIS_STATE_DERIVATIVE = 3
MEM_STATIC = 6000
MEM_TOTAL = 10000
SOLVENT_METHOD = PCM
$end

$solvent
DIELECTRIC 7.425700
OPTICALDIELECTRIC 1.974025
$end

In 14 optimization cycle, it starts fluctuating like this

168 -1146.4545159922 8.33E-05
169 -1146.4543566366 8.92E-05
170 -1146.4542118425 9.65E-05
171 -1146.4540884653 1.00E-04
172 -1146.4542001680 9.52E-05
173 -1146.4541178516 9.88E-05
174 -1146.4540915067 9.89E-05
175 -1146.4540660669 1.00E-04
176 -1146.4539802728 1.02E-04
177 -1146.4534338920 1.18E-04
178 -1146.4529243754 1.32E-04
179 -1146.4536505976 1.09E-04
180 -1146.4546939368 4.01E-05
181 -1146.4554474269 1.63E-05
182 -1146.4556774171 1.53E-05
183 -1146.4561482021 2.03E-05
184 -1146.4566687497 3.11E-05
185 -1146.4561692659 2.96E-05
186 -1146.4558832012 3.53E-05
187 -1146.4557550575 3.84E-05
188 -1146.4556599681 4.16E-05
189 -1146.4557884856 3.82E-05
190 -1146.4558268386 3.78E-05
191 -1146.4558837907 3.72E-05
192 -1146.4559894942 3.79E-05

#8

You have a problematic SCF. You need to try different SCF guesses and differerent SCF algorithms until you get it to converge, maybe a different functional. That’s about all that can be said in general.