Conflict of results

[Monu Joy]

Hi there

I have cell-optimized a MOF structure (CYCU-3) using cp2k and it converged nicely but the resultant structure looks a little weird, which means the ligands are kind of twisted and I thought that could be the minimum structure. Just to double-check, I used another software CASTEP but that resulted in a reasonable geometry (with respect to the experimental structure). This is not for a comparison between cp2k and CASTEP but I would like to learn where I was wrong in the cp2k input description/model. The input geometry and the cut-off value (925) for both cp2k and CASTEP are the same but CASTEP used PBESol where cp2k used PBE. In both cases, the Hexagonal symmetry was fixed.

I know both packages are completely different in various means, especially pseudopotentials and plane waves, etc. 

Any help would be highly appreciated. 

P.S. All the input/output structures and files attached.     

-Monu Joy

[Marcella Iannuzzi]

Dear Monu Joy

Your input seems OK. 

It might be an issue of the functional or of the pseudo potentials. Have you tried PBESol with cp2k?

Things that I would test are : increasing the PW cutoff, increasing the system size in z, the basis sets, and maybe also another vdw. 

Kind regards

Marcella

[Monu Joy]

Thanks, Marcella, for the suggestions. I just increased the Z direction by MULTIPLE_UNIT_CELL  1 1 2. Albeit it's not yet converged, the intermediate geometries are reasonable (no twisting so far). 

By the way, I wanted to try PBESol as well but it ended up getting an error "invalid value for enumeration:PBESol" and when I checked the manual, I haven't seen PBESol so I am wondering where I am wrong in the input...

Thanks in advance!

-Monu

--
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/3cde1027-258b-4196-8665-a08e8b60aae6n%40googlegroups.com.

[Monu Joy]

Forgot to attach the input file, sorry about that.

Please see attached.

-Monu

[Monu Joy]

Thank you very much for your response. Should I use it like the following...? If yes, I am getting an error: "found an unknown keyword PARAMETRIZATION in section XC_FUNCTIONAL" 

      &XC

      &XC_FUNCTIONAL PBE

        LIBXC

        PARAMETRIZATION PBESOL

      &END XC_FUNCTIONAL

      &XC

      &XC_FUNCTIONAL PBE

        PARAMETRIZATION PBESOL

      &END XC_FUNCTIONAL

-Monu

On Tuesday, 12 January, 2021, 04:52:44 pm GMT-5, Travis <polla...@gmail.com> wrote:

Hi,

Set PBEsol here,

https://manual.cp2k.org/cp2k-6_1-branch/CP2K_INPUT/FORCE_EVAL/DFT/XC/XC_FUNCTIONAL/PBE.html#list_PARAMETRIZATION

The vdW section recognizes 'PBEsol' as an option. The same process is done for revPBE too. 

-T

https://groups.google.com/d/msgid/cp2k/a5a865d8-8513-407f-a395-137318794fdan%40googlegroups.com

.

[Lucas Lodeiro]

Hi Monu,

You have to use & to open the new subsections needed. 

      &XC
        &XC_FUNCTIONAL 

          &PBE
            PARAMETRIZATION PBESOL

          &END PBE
        &END XC_FUNCTIONAL

      &END XC

Regards!

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/2044864971.844088.1610488854928%40mail.yahoo.com.

[Monu Joy]

Thank you so much Lucas, appreciated...

-Monu

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAOFT4PLoiRAk6YvCMsqGphWvxBGhKwfwTobK5eZkwQKnKBCYeg%40mail.gmail.com.

[Leopold Talirz]

Since Andres Ortega from our group noticed this with a similar Al MOF, I just wanted to confirm:

Thanks, Marcella, for the suggestions. I just increased the Z direction by MULTIPLE_UNIT_CELL  1 1 2. Albeit it's not yet converged, the intermediate geometries are reasonable (no twisting so far). 

This is indeed the source of the issue here - doubling the cell along z results in a reversal of the twisting of the ligands, even when starting from the twisted geometry (I checked).

In our case, the twisted structure was the correct minimum for the unit-cell only calculation, with a substantially lower energy (~250meV per twisted phenyl), while for the doubled cell the twisted structure was slightly higher in energy (order of 10meV per twisted phenyl).

This then re-raises the original question of the thread, which is why CASTEP seems to yield a different result. I believe the answer is in the CASTEP output:

                           -------------------------------

                              k-Points For BZ Sampling

                           -------------------------------

                       MP grid size for SCF calculation is  1  1  2

                            with an offset of   0.000  0.000  0.000

                       Number of kpoints used =             1

CASTEP chooses a 2-kpoint grid along z (which reduced to just a single k-point calculation through symmetry).

From the CASTEP manual [1]:

> For cubic cells and for the C direction of hexagonal cells the even and odd grids for the Monkhorst-Pack scheme give the same number of k-points. However, the even grid provides better sampling and will always be used automatically under these conditions. This ensures that a good grid with k-point separation at (or less than) the specified target can be achieved more economically. As a consequence, such lattices may have much finer separations than requested as odd grids have been excluded - even though they would have been closer to the specified separation - and the better even grids have been taken in preference.

I suspect that forcing CASTEP to do a Gamma-point calculation would result in the same twisting as seen in CP2K (in our case, we checked with CRYSTAL).

For our MOF, just to exclude any possibility of issues with OT, we also tested this with DIAGONALIZATION and KPOINTS in cp2k and got the same result as with the 1x1x2 supercell.

The lesson is simply that ~6.5A can be too short along the direction of these inorganic chains/rods in MOFs.

Best,

Leopold

[1] http://www.tcm.phy.cam.ac.uk/castep/documentation/WebHelp/CASTEP.html#modules/castep/dlgcastepelecoptkpoints.htm?Highlight=monkhorst

 

[Monu Joy]

Even though the doubling of the z-axis (MULTIPLE_UNIT_CELL 1 1 2) seemed fine in the initial cycles, it ended up being twisted on convergence. Then I have tried the K-point scheme (SCHEME MONKHORST-PACK 2 2 2), where the result was the same, twisted. However, the structure seems reasonable (no twisting) in CASTEP even with the gamma point.

But of course, the culprit could be the small z-axis, but I am not sure how to tackle this in CP2K.

-Monu

--

You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+uns...@googlegroups.com.
To view this discussion on the web visit

https://groups.google.com/d/msgid/cp2k/a018da49-5c42-4ade-b9f5-1c1b68579a91n%40googlegroups.com

.

[Leopold Talirz]

Hi Monu,

thanks for following up - I did run a CP2K cell optimization starting from your relaxed "twisted" CP2K geometry. You find the trajectory and movie attached

Does this agree with your results?

It certainly seems to "untwist" the ligands (see movie) - I had to stop the optimization at some point though because convergence became very slow.

Some stats from the total energy:

First total energy (single-point): -2160.041714658635101

First optimization step: -2160.136731180092738

50th optimization step: -2160.2817551900

100th optimization step: -2160.2996248085

I.e. after the first "untwisting" that lowers total energy by  ~260mHa (quite significant), the final 50 steps lower the total energy by less than 3 mHa.

This is with cp2k 8.1 [2] and with minimal modifications to the input file (see tar file). It's not clear to me why you would get a different result - both MULTIPLE_UNIT_CELL 1 1 2 and 2 2 2 K-points should give this result.

As to why the CASTEP gamma point calculation does not yield the twisted structure: it is possible that this involves breaking a symmetry that is for some reason preserved in the CASTEP calculation (don't know).

Can you try starting your CASTEP calculation from the twisted CP2K geometry and see where it ends up?

If it remains twisted, you can then compare total energies to the untwisted result, which should tell you which one is the actual minimum.

Best,

Leo

[1] Try  "grep i 925-pos-1.xyz" on the trajectory

[2] Our previous calculations were with cp2k v5.1 though.

[Monu Joy]

Hi Leo

Thanks for the help. I started a CASTEP calculation on the CP2K twisted (relaxed) geometry and it converged as twisted, and this CASTEP relaxed twisted geometry is more stable than the untwisted version by a factor of 0.00030737 eV. Gamma point was used for both twisted and untwisted calculations.

Moreover, I have tried the CP2K calculation as you suggested to untwist the geometry but it converged as twisted. I am using CP2K 7.1 and the only change that I have made in the input was the removal of the following lines:  

        #COORD_FILE CP2K_Optimzed structure_CIF.cif
        #COORD_FILE_FORMAT CIF

Are there any significant changes between 7.1 and 8.1 in terms of dealing with symmetry...?

-Monu

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/1240ce91-eb54-48ff-9c13-0e4e1f6ce735n%40googlegroups.com.

[Sarah-K]

Hi there,

I'm a fresh user of CP2K. Planning to use it for MOFs cell optimizations. I have a basic/stupid question regarding the input file. If I wanna optimize the unit cell parameters, does the input file have to include all atoms in a unit cell, or I can minimize  the number of atoms in a way?

Thanks, S