Cluster Expansion using ATAT in Two Dimensions (2D)?

Hello everyone,

I am a newbie to ATAT. I looked into manual but could not anything on using ATAT for two dimensional (2D) materials. While ago, when I started using ATAT I was successful implementing it for 3D materials, alloys etc.! However, I could not find any option to force ATAT to search for clusters in just 2D (in-plane) as in 2D materials, rather than trying to expand the unitcell in all three dimensions (change the volume of the unitcell) to predict new structures as it normally does.

My issue appears when I try to use maps command (or automated "pollmach runstruct_vasp" command) in a 2D material (like Li in graphene), ATAT predicts a few structures and then stuck and unable to predict more.

Any help/comment or clue is highly appreciated,
Regards

Note that maps has a -2d option!
Just make sure that your artificial 3rd dimension has a long lattice parameter so that no clusters will be generated along that direction.
It’s not a perfect solution but may help you get there.

Dear Prof. van de Walle,
I tried to use ATAT to generate different carbon sheets using ‘maps -2d’. It generates a structure for me, but when I run ‘runstruct_vasp &’ in that directory, the vasp calculation is not initiated. It just gets stuck at the very same structure. Here are my lat.in and vasp.wrap files. i am unable to figure out where I have gone wrong. I would be grateful if you could help me out.

lat.in:
2.0 2.0 7.5 90 90 120
1.0 0.0 0.0
-0.5 0.866025404 0.0
0.0 0.0 2.0
2.459975478 1.420267504 7.500000000 C,C

vasp.wrap:
[INCAR]
PREC = high
ISTART = 0
ICHARG = 2
ISMEAR = 0
SIGMA = 0.1
NSW = 50
IBRION = 2
ISIF = 3
EDIFF = 1.0D-5
EDIFFG = -0.005
ENCUT = 500
LCHARG = .FALSE.
LWAVE = .FALSE.
KPPRA = 100
DOSTATIC

[KPOINTS]
K-points
0
Gamma
15 15 1
0 0 0

Thanks a lot!

Cheers,
Aaditya

Dear Aaditya,

I suppose that you are using the ATAT to do cluster expansion for the graphene in 2D dimension? I am trying to figure our why do you set the possible atoms of atom position 1 in the lat.in file for two identical atoms C, C? In my opinion, this should be set like two different types of atoms, like A, B.

Best,
Pengfei

  1. the C,C in your lat.in is strange. I don’t see what you are trying to accomplish.
  2. The fact that vasp is not running could be due to many things. Look at the files in the directory: which ones are there, any errors, etc.