Ar lattice constant calculation

[sumit agrawal]

Hii everyone,

I want to calculate a lattice constant for the Ar crystal of (2*2*2 unit cell). I am attaching my input and out files. Where I will find the lattice constant ?

Please help me!!

[Fabian Ducry]

Hi

You have only performed a single point energy calculation (RUN_TYPE ENERGY). You cannot optimize the lattice constant with this kind of simulation. Please refer to the tutorials on how to perform such a calculation, e.g. here: https://www.cp2k.org/events:2019_cp2k_workshop_ghent:index under "Geometry and cell optimization of NaCl clusters"

Cheers,

Fabian

--
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/CALSDoYbVb-H5Bp3K5twhNotUESTnixKMaWBnBKNMZr7xzC2%3DSw%40mail.gmail.com.

[Lucas Lodeiro]

Hi Sumit,

First of all, you cannot find the lattice constant with this type of calculation. You ran an ENERGY calculation. If you want to do it, you need to run a CELL_OPT calculation.

In your current calculation, you use a 20 A lattice constant cube. Also, I think your initial cell dimension is not correct, since the lattice constant of Argon is 5.26 A, then a 2x2x2 supercell is not 20 A sized. Also, your system has 63 atoms... ¿?.

Other related topics are: a 2x2x2 supercell is not sufficient to compute a periodical Argon system... You need a bigger supercell, or use kpoints. Also, as this system just interacts by dispersion, probably the PBE functional is not sufficient at all. You need to account the dispersion terms by means of Grimes D3 or other.

Regards

[sumit agrawal]

Dear Lucas Lodeiro and Fabian,

Thanks for your nice suggestions. I am new to cp2k. I have just started learning this. I have some query regarding my procedure for preparation for crystal structure. I am following below procedure to prepare the crystal structure:

First, I have prepared the crystal structure (2x2x2) with the lattice dimension a=5.2229 A, b=5.2229 A and c=5.2229 A. It contains 63 atoms. After that I removed some atoms from x, y and z directions so that it become periodic in x, y and z direction. After this, it contains 32 atoms. 

Now I save this lattice in XYZ format. This will be my initial coordinate system. Am I correct?

After this I took this coordinate file and ran for the cp2k calculation.

Now in cp2k calculations for this lattice structure what will be my cube size? (Here you mention it will not be 20 A. )

One more query, here I am starting with a lattice dimension with 5.2229 A . So after Cell_OPTIMIZATION, will the geometry of my crystal change or not?

Yes, dispersion correction must be required for this type of calculation, I will do that.

Here I am attaching my crystal structure images.

Thanks 

Sumit

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CAOFT4PKMq7MQYztPnmT3GL5zaPS%3DQErOF4RWR1ch3zmT2gnDEg%40mail.gmail.com.

[Fabian Ducry]

Dear Sumit

A supercell of (2x2x2) has to contain 2*2*2=8 times the number of atoms of the unit cell. So It can never have an odd number as in your case 63. You should follow a tutorial on how to prepare such a structure, e.g. with ASE. Here, for example, you find instruction on how to build a crystal: http://exciting-code.org/lithium-atomic-simulation-environment. Once you have the ASE Atoms object building the supercell is as easy as atoms*(2,2,2). When you have a correct crystal, make sure that the &CELL section in the cp2k input matches the lattice dimensions of the crystal. You cannot just use an arbitrary number.

CELL_OPT will optimize both the atom positions as well as the lattice dimensions.

Fabian

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CALSDoYaG6E_PftB%2Bj2xsfsbbfmaFvoW0VvZUxT1CS2kOoz1Rzg%40mail.gmail.com.

[sumit agrawal]

Dear Fabian,

Thanks for the quick response. Actually I prepared a lattice structure (2x2x2) in VESTA software. Is it the correct tool to prepare lattice structure, any idea? 

I tried again to prepare a lattice structure. So 2x2x2 will contain 32 atoms. Then I performed cell_opt in cp2k. I am attaching my coordinate file, input file and out file. 

Where will I find the lattice constant in the output file?

One more query: will the lattice constant for 1x1x1 unit cell and nxnxn unit cell be same or different?

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/8845fe89-4df1-745e-b8a9-5f536e6434f0%40gmail.com.

[Fabian Ducry]

cp2k prints "CELL| Vector" after each iteration of the optimizer. From this you can calculate the lattice constant.

The lattice constant usually refers to the dimension of a unit cell, therefore it does not change when you create a supercell. The cell dimensions, however, differ between the unit and supercell.

Fabian

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/CALSDoYbEU_9OwK%3DMBiOjwBmbb6Lm%2BtiyRCKgtHxTNhK%3DV_gpQw%40mail.gmail.com.

[sumit agrawal]

Thanks Fabian!!!

From the output file I can see a=b=c=11.0732 A. Is the lattice constant 11.0732 A or half of this which is 5.5366 A, since it is a 2x2x2 unit cell ?

To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/be4fa357-daa9-6398-b0f2-deb1441e30ae%40gmail.com.