SiNR geometry optimization

[Heinz]

Hello
I followed the silicene geometry optimization step in the tutorial: http://quantumwise.com/publications/tutorials/mini-tutorials/209-opening-a-band-gap-in-silicene-and-bilayer-graphene-with-an-electric-field and I did as said in the tutorial for silicene crystal structure.

Now I would like to try optimizing silicene nanoribbon geometry. For this aim I generated a graphene nanoribbon in the builder and then changed all C atoms to Si. But I am confused in the geometry optimization parameters. When it is a GNR or SiNR, the builder sees it as a "unit cell" (lattice type). In this case should I unconstrain in both x y and z directions? Moreover when it is "unit cell" type lattice in the builder, lattice parameters dialog box does not permit to change lattice parameters, I can only change primitive unit vectors. We know that lattice constant is around 3.86A for silicene so how can I adjust this parameter so that geometry optimization takes less time. I guess it will take a very long time if I cannot change the lattice constant to 3.86A?

In addition: For graphene nanoribbons and similar molecules in which direction shall I constrain in the optimization window? I unconstrained in x y and z for SiNR and atoms fell apart as in the attached figure 

Any help is apreciated a lot.

[Anders Blom]

Indeed you must change the lattice constant before the optimization, since if you use the C-C bond length the atoms will be too close and the forces very large, which gives problems.

There is a simple relation between the lattice vectors in graphene (or the orthorhombic supercell version of it) and the lattice constants, you can work it out on paper, there are a few sqrt(3)/2 here and there

There is no need to optimize the C lattice constant since you have vacuum all around. Probably the crystal fell apart because of the wrong lattice constants.

[Heinz]

Thank you very much.

Similarly I also tried to optimize the geometry of graphene unit cell where two hydroxyls are connected as given in the attached script. Here, atoms also fell apart but I didn't change the lattice constants. Just built the graphene nanosheet m=2 and n=2 and then added O and H then tried to optimize.

Could you please help me to optimize this geometry? How can I adjust the settings of the optimization so that the atoms won't fall apart?

[Anders Blom]

One word: k-points

[Heinz]

Only in x- direction or all directions?

[Anders Blom]

In the directions where the structure is periodic, of course.

[Heinz]

Thanks very much.