Molden input not read by Molden

Questions
#1

Hi,

I tried the “molden_format” = true in Qchem 5.4 to visualize the optimization steps but also the orbitals of the resulting structure but I have not succeded. The reason for it is that molden does not read the resulting file that I created when copying from the output file from Qchem the Molden input section.
This file starts with:

[Molden Format]
[Atoms] (Angs)
C 1 6 4.06583245 1.20884252 1.17652596
C 2 6 3.25870420 0.12055580 0.73117011
C 3 6 3.60320922 -0.53067111 -0.53350781
C 4 6 4.66611825 -0.07117893 -1.25724327
C 5 6 5.43753222 1.02116426 -0.80262743
C 6 6 5.12800430 1.66352699 0.44767549
C 7 6 2.15264969 -0.31255870 1.49137530
C 8 6 1.30032384 -1.32918005 1.11745631
C 9 6 0.17135649 -1.79271709 1.95389783
C 10 6 0.35407587 -2.81746692 2.88184261
C 11 6 -0.69615021 -3.24484664 3.67896723
C 12 6 -1.96783632 -2.69476223 3.55728941
C 13 6 -2.14899265 -1.68325615 2.62429399
C 14 6 -1.08834896 -1.22785530 1.84111979
N 15 7 -3.00991819 -3.19469881 4.39998961
C 16 6 -4.11675032 -2.28668370 4.62697240
C 17 6 6.52153788 1.48186928 -1.57725899
N 18 7 7.41492005 1.86857066 -2.20239185
C 19 6 -3.49005037 -4.50670882 3.98721557
H 20 1 1.33128217 -3.27270395 2.99666293
H 21 1 -0.53051195 -4.01504935 4.42341412
H 22 1 -1.25888056 -0.42334989 1.13455701
H 23 1 -3.11950225 -1.21972083 2.50153844
H 24 1 1.45168672 -1.82069665 0.16058846
H 25 1 1.95811660 0.18482404 2.43695504
H 26 1 4.93702043 -0.53962542 -2.19492558
H 27 1 5.73608059 2.48990554 0.78623784
H 28 1 3.03372656 -1.38067961 -0.87858587
H 29 1 3.81074241 1.68054599 2.11884231
H 30 1 -2.65370675 -5.19327226 3.86094429
H 31 1 -4.05605146 -4.47657488 3.04021642
H 32 1 -4.14440722 -4.91394273 4.76193606
H 33 1 -3.74240369 -1.31457270 4.95122791
H 34 1 -4.74534096 -2.68894090 5.42416391
H 35 1 -4.76057505 -2.13421445 3.74310446
[GTO]
1 0
S 6 1.000000
4.56324000E+03 1.96665000E-03
6.82024000E+02 1.52306000E-02
1.54973000E+02 7.61269000E-02
4.44553000E+01 2.60801000E-01
1.30290000E+01 6.16462000E-01
1.82773000E+00 2.21006000E-01

And finishes:

35
Step 720
C 4.06583245 1.20884252 1.17652596
C 3.25870420 0.12055580 0.73117011
C 3.60320922 -0.53067111 -0.53350781
C 4.66611825 -0.07117893 -1.25724327
C 5.43753222 1.02116426 -0.80262743
C 5.12800430 1.66352699 0.44767549
C 2.15264969 -0.31255870 1.49137530
C 1.30032384 -1.32918005 1.11745631
C 0.17135649 -1.79271709 1.95389783
C 0.35407587 -2.81746692 2.88184261
C -0.69615021 -3.24484664 3.67896723
C -1.96783632 -2.69476223 3.55728941
C -2.14899265 -1.68325615 2.62429399
C -1.08834896 -1.22785530 1.84111979
N -3.00991819 -3.19469881 4.39998961
C -4.11675032 -2.28668370 4.62697240
C 6.52153788 1.48186928 -1.57725899
N 7.41492005 1.86857066 -2.20239185
C -3.49005037 -4.50670882 3.98721557
H 1.33128217 -3.27270395 2.99666293
H -0.53051195 -4.01504935 4.42341412
H -1.25888056 -0.42334989 1.13455701
H -3.11950225 -1.21972083 2.50153844
H 1.45168672 -1.82069665 0.16058846
H 1.95811660 0.18482404 2.43695504
H 4.93702043 -0.53962542 -2.19492558
H 5.73608059 2.48990554 0.78623784
H 3.03372656 -1.38067961 -0.87858587
H 3.81074241 1.68054599 2.11884231
H -2.65370675 -5.19327226 3.86094429
H -4.05605146 -4.47657488 3.04021642
H -4.14440722 -4.91394273 4.76193606
H -3.74240369 -1.31457270 4.95122791
H -4.74534096 -2.68894090 5.42416391
H -4.76057505 -2.13421445 3.74310446

Is it possible that I am doing something wrong? My input was:

JOBTYPE = opt
MECP_OPT = true
SPIN_FLIP = true
cis_triplets= false
cis_singlets= true
print_orbitals= true
MECP_METHODS = mecp_direct
GEOM_OPT_MAX_CYCLES = 5200
GEOM_OPT_TOL_DISPLACEMENT=100
CIS_N_ROOTS = 5
MECP_STATE1 = [0,1] ! Looking for state of charge 0, 1st singlet ie/ the GS !
MECP_STATE2 = [0,2] ! Second singlet
METHOD = CAM-B3LYP
BASIS = 6-311G(d,p)
SCF_CONVERGENCE = 10
SYM_IGNORE = true
MOLDEN_FORMAT = true

Thank you in advance

#2

Can you post full input including $molecule? Note that MolDen does not read Q-Chem output, you need to extricate the MolDen file from the Q-Chem output and save it in a separate file (but it sounds like you may be doing that already). In the meantime, simple example is

$molecule
0 1
        O         0.0000000000    0.0000000000    0.1278971150
        H        -0.7611715735    0.0000000000   -0.4710730487
        H         0.7611715735    0.0000000000   -0.4710730487
$end

$rem
method hf
basis  sto-3g
molden_format true
jobtype opt
$end

MolDen file should look like

[Molden Format]
[Atoms] (Angs)
  O      1    8      0.00000000      0.00000000      0.14439600
  H      2    1     -0.75806327      0.00000000     -0.49147712
  H      3    1      0.75806327      0.00000000     -0.49147712
[GEOCONV]
energy
      -74.9645178635
      -74.9657665202
      -74.9658786340
...some lines removed for brevity...
Step 5
  O      0.00000000      0.00000000      0.14439600
  H     -0.75806327      0.00000000     -0.49147712
  H      0.75806327      0.00000000     -0.49147712

and this opens for me (and lets me step through the optimization) using what I believe is Molden 5.7.

1 Like
#3

Hi, thank you very much for your swift reply. It was something related to the instalation of Molden in my machine. However, and something I want to ask you is: my job was a S0-S1 conical intersection search. I think that the orbitals that are saved in molden format are for the ground state. Am I right? Is it physically correct to use them as for the conical intersection?

Thank you again.

#4

In a certain sense, the ground-state orbitals are the only ones that you have so it’s certainly not incorrect to use them in this (or any) situation, but you do need to be prepared for a lot of configuration mixing. The largest TDDFT amplitudes x_ia (meaning i → a excitation, i = occ and a = virt) are printed out.

If you want “better” orbitals than the ground state, you can use natural transition orbitals (NTOs). These are state-specific, however - in fact, they are the natural orbitals of the excited-state density matrix. (For a discussion, see [2204.10135] Density Functional Theory for Electronic Excited States). There are various ways to get these out of Q-Chem, please see the Q-Chem manual. They can be visualized either with MolDen or with IQmol.