Optimization issue with mos2

[hsuya]

I am trying to optimize mos2 monolayer using LCAO however the resulting structure has S-S bond formation. Can you suggest why this is happening and what is the alternative?

[hsuya]

the log file for the above is not getting uploaded so please check this link https://drive.google.com/open?id=1m7OKuPCVioAQDdNRxNLZAE4Aepml2s7J

[Petr Khomyakov]

Why do you think this reconstruction at the edge should not have happened? Your structure does not look like a monolayer, but rather as a narrow nanoribbon. That means there exist 2 edges, and reconstruction is an edge effect I suppose. I would guess that reconstruction may also depend on the initial structure assumed for your ribbon edges.   As you may see the 2 edges of the ribbon are actually different, and reconstruct in a different manner.

[hsuya]

Okay I might have created the initial structure in the wrong way. Also do I have to maintain the b/a and c/a ratio in the lattice parameter for this or will it come to proper values after optimization.

[Petr Khomyakov]

Your structure looks like a 2D quantum dot, i.e., it is no longer a periodic crystal as such. I do not think c/a ratio matters for that.  What matters is vacuum padding around the dot and the boundary conditions for the 6 facets of the simulation cell. I would recommend using non-periodic boundary conditions to avoid interaction between the image cells. 

[hsuya]

okay I will try. Actually I am trying to do geometry optimization of a mos2 4x4x1 supercell and I am stuck for past 10 days. Don't know what to do next nothing is working. I have tried few methods as reported in literature

[Petr Khomyakov]

Strictly-speaking, your structure does not look like 4x4 MoS2 supercell. The question is whether you want to study infinite 2D sheet with 4x4 supercell or 2D quantum dot built from the 4x4 supercell?

[hsuya]

I am sorry...after listening to you I realized my structure was not correct. This is what I am working on.