Hydrogen termination of Si(100)

[supreme_general]

Good day to everyone!

I am a new user of VASP, and my advisers told me to do a calculation involving the surface energy of Si(100) surface with 5 slabs and with the bottom layer terminated with H, in order to reduced the dangling bonds. I had no problem setting up the supercell. My only dilemma is on how to find the coordinates of the H atoms and where will I put it in the POSCAR.

I hope I can find an answer to this! Thank you!

[Oscar_rp]

Dear Supreme_general,

You may be able to open your POSCAR file with for example MOLDEN
http://www.cmbi.ru.nl/molden/molden.html
(there are many orther programs..)
that is a free software and read POSCAR files and modify/add the H atoms at the dangling bonds at the surface.

You can save it as a new POSCAR

Cheers,
O.

[forsdan]

Sorry, but I don't understand your question. Do you mean how to optimize the positions of the hydrogen atoms on the surface? If so, then please have a look at the ionic relaxation tags NSW, IBRION, (ISIF) in the manual and add appropriate values to the INCAR file. Of course if you don't have any information about how many hydrogen atoms and where to put them, you will have to investigate several configurations.

If you are referring to how to specify coordinates to the POSCAR file then please have a look at the cooresponding POSCAR section in the manual for instructions.

Cheers,
/Dan
<span class='smallblacktext'>[ Edited Wed May 26 2010, 10:28AM ]</span>

[supreme_general]

Dear Oscar_rp,

Thank you very much for that information! I'll try it to see if it will work!

Dear forsdan,

The problem is that, I would do some surface calculations and the first step in doing that is to relax the surface, and calculate the surface energy. My advisers told me that there are two methods to do it, one is the symmetric method wherein the top and bottom slabs of the supercell will be treated as surfaces, while the other one is the asymmetric method wherein only the top slab will be allowed to relax while fixing the bottom layers to represent the bulk. Since the first method is computationally expensive, they said that the last method is better. And to save more computational time, I should place H atoms on the bottom slab to neutralize dangling bonds and thus creating the bulk region. The problem is that, I do not know how should I place the H atoms next to Si atoms. They told me that the bond length for Si-H is different, and the positions will also differ compared if I just add Si. That is basically the problem.

I hope you get it. Thank you for your interest in the problem.

[forsdan]

Ok. I got it. However, I have not personally used such an approach for surface calculations. Nevertheless, the way I probably would do it is to first find out geometric positions of the H atoms from earlier studies (which I am quite confident should exist in this case) or by investigating the site preference (or simply using a real-space analysis to show where the dangeling bonds are). The site preference finding would be performed by freezing the Si slab and only optimize the H atom distances to the surface. Then I would use the obtained structure as a starting point for the surface energy calculations.

Well, at least this is my first thought without looking deeper into the literature of how the positioning of the H atoms usually is done.

Cheers,
/Dan

<span class='smallblacktext'>[ Edited Thu May 27 2010, 10:14AM ]</span>

[Danny]

Also check the influence of your H on the back of your slab to be sure it has no big modifications as a result (which sometimes happens, eg for a Ge slab I needed to make it as thick as a symmetric slab to get comparable accuracy. Note that a symmetric slab has the advantage of extra symmetry, reducing the computational cost by reducing the number of K-points in the IBZ)

Cheers
Danny