input files for geomtery optimization + normal mode calculations, DFT/B3LYP/6-31G(d,p))

Darek -

unread,

Nov 17, 2025, 3:22:17 AMNov 17

to cp2k

Dear CP2K community,

I would like to use CP2K software to reproduce the results obtained with GAUSSIAN.

I would like to (i) optimize the geometry and (ii) calculate the normal modes (perform a vibrational analysis) for the molecule C20H22N8O6S2, characterised with the following atomic coordinates:

S -0.9348288137 1.7376071546 -0.2570119375
S -8.5585608658 -1.6464676835 0.3682299649
O -1.7225075170 -2.4392359635 0.7342302791
O 5.6622419631 1.3122457843 1.1113174377
O 0.9652905974 -3.1942561175 -0.3363487203
O 2.1658590079 -2.3323000791 1.3665937632
O -4.9053238253 0.1673494117 -2.0463926347
O -4.4544493435 2.7626511207 0.2379754834
N 6.3545677018 -0.4587005407 -0.2173058006
N -0.4169350061 -0.9115933817 -0.5569211969
N 9.4893184194 1.4033386030 -0.1592059976
N -3.5658274844 -0.0051180853 -0.2025982838
N -5.6885165154 2.1721371832 0.5259183181
N -8.1321722706 0.8440732729 0.7945817183
N -7.0363906121 -1.0486351883 -0.0393578921
N -10.2999029178 0.2054481393 1.4678591822
C 7.7930165612 -0.2143135270 -0.1716658051
C -1.1573409766 0.1792764349 -1.1905621796
C -2.4777133786 -0.6180481502 -0.8958502950
C 8.3918274207 -0.1136621170 1.2686455991
C 8.2402637597 1.1016089745 -0.8605544139
C 5.7629413554 -1.4298637284 -1.1360479512
C 9.1719475958 1.2144171484 1.2655995449
C 4.2744882080 -1.4762775994 -0.7213993948
C -1.5663836612 -1.5577557543 -0.0680720881
C 4.0716025858 -0.1943625484 0.0315586064
C 5.4238465941 0.3368927791 0.4035861552
C 0.8933804250 -0.8577504747 -0.0585808185
C 1.5674965874 0.3257798306 0.0133754186
C 0.9014451990 1.6520119865 -0.3441922521
C 2.9759425706 0.5127570608 0.3634760559
C 1.4231751897 -2.1709072151 0.4227074326
C -4.7170438553 0.3479662498 -0.8559959974
C -5.7802339700 0.9914077791 0.0248863957
C -7.0237244808 0.2344333237 0.2587938635
C -9.0780969978 -0.0503761848 0.9219919910
H 8.2557508097 -1.0585499776 -0.6901466240
H -0.9003302047 0.3233656666 -2.2436882539
H -2.8751271072 -1.0778735441 -1.8043237475
H 9.0589202460 -0.9618997498 1.4505744426
H 7.6084764024 -0.1201450754 2.0277323384
H 8.4138713866 0.9747068173 -1.9356106950
H 7.4570829920 1.8630843159 -0.7155959013
H 6.2534029772 -2.4044071516 -1.0347403146
H 5.8815186564 -1.1032004513 -2.1801586531
H 8.5304379131 2.0138676841 1.6714600705
H 10.0909812875 1.1704579948 1.8605596531
H 4.1026327029 -2.3209826434 -0.0425246521
H 3.5956582393 -1.6006612694 -1.5729579051
H 9.8046010548 2.3522733933 -0.3518809070
H 1.2572463082 2.4200491362 0.3506538043
H 1.2241001402 1.9694008035 -1.3440223345
H 3.1977496713 1.4412646320 0.8921322369
H -3.4461485491 0.2673126885 0.7641005027
H 1.2586806877 -4.0116269657 0.1086545502
H -4.5381784035 3.6359508941 0.6535919460
H -10.5286245204 1.1873855850 1.5623283595
H -11.0747019402 -0.3991506813 1.2341985330

Could you help me in preparing the synatctically correct input files to accomplish this task?
CP2K software offers mamy default options and thus I am unsure which keywords must necessarily be included in the input files.
For instance, I am not sure whether I should include the &MOTION block in the input file to optimize the geomtery.
My intention is to perform the calculations using an all-electron potential (not pseudopotential) and obtain the normal modes saved in the MOLDEN format.

Hence, I would like to kindly ask for the support of an experienced CP2K user for the support in the preparation of working input files.

Thank you in advance,

Darek

Marcella Iannuzzi

unread,

Nov 19, 2025, 5:03:16 AMNov 19

to cp2k

Dear Derek,

Have a look at the tutorials and materials available on line

For instance: https://manual.cp2k.org/trunk/methods/optimization/geometry.html

https://manual.cp2k.org/trunk/methods/dft/index.html

Regards

Marcella

Message has been deleted

Darek

unread,

Nov 22, 2025, 6:51:02 AMNov 22

to cp2k

I have known this websistes; nevertheless, I have still some unvertainties:

- the RUN_TYPE should be obviously set to GEO_OPT. Hence, I wonder whether adding the &MOTION block with &GEO_OPT sub-block is necessary if I prefer to rely on default parameters of CP2K?

- what would be a better approach to geometry optimization - to rely on traditional diagonalization or utilize orbital transformation (OT) methods, e.g. DIIS?

- the input file needs to contain information about the unit cell parameters. Meanwhile, I am interested in the analysis of a single molecule in a vacuum. Despite the fact that I sepcified PERIODOC as NONE, the program still urges me to speicfy the dimensions of a unit cell. Should I worry about it?

- the next issue relates to the PIOSSON section. Should I add "PERIODIC none" in this section? Then, what is the most recommended PIOSSON_SOLVER for a non-periodic system (a single molecule)?

- is there a possibility to reuse the SCF results from the first step (geometry optimization) in the second step (normal mode analysis)? These calculations are related to each other by copying the optimized coordinates from the xyz file generated in the first step to the input file of vibrational analysis. Maybe there exists a better way to connect them?

Marcella Iannuzzi

unread,

Nov 23, 2025, 4:18:25 AMNov 23

to cp2k

Hi Derek

- The default options in GEO_OPT are in general good enough

- Any wfn optimisation method is ok for GEO_OPT, as long as the forces are correct, i.e. sufficiently tight SCF converge

- QS uses grids defined by the simulation cell

- WAVELET or MT

- No, copy the last coordinates into a new xyz and pass it to the TOPOLOGY section. The wfn RESTART can be used