Dear all, Thank you for your time. As I mentioned before I have taken my initial parameters from a successful 2D MoS2 calculation. I am going to apply your suggestions and If I was successful I would post the results here.
Best regards, On 1/11/15, Salvador Barraza-Lopez <[email protected]> wrote: > Adding to Andrei's response, you may want to ensure that your electronic > bands and lattice parameters for bulk MoS2 do reproduce published results > (you are probably trying to go too fast). > Best regards, > -Salvador > ________________________________________ > From: [email protected] [[email protected]] on behalf of > [email protected] [[email protected]] > Sent: Sunday, January 11, 2015 11:53 AM > To: [email protected] > Cc: siesta-l > Subject: Re: [SIESTA-L] Help regarding structural relaxations for AMoS2NR > >> Dear all, >> >> I am trying to relax an armchair MoS2 nanoribbon. My initial >> coordinates are taken from the MoS2 sheet. My .fdf file together with >> my intial positions are attached. The problem is that after about 70 >> steps of CG relaxation, all the atoms are coming together and the >> distances between atoms are much shorter than the initial ones. The >> relaxed structure after about 70 steps is attached in elec.XV file. I >> have successfully done relaxations for the MoS2 sheet but it seems >> that I am not able to relax the nanoribbon. Any helps would be >> apprecited. >> >> Bests, >> Mohammad, >> > > Dear Mohammad, > some suggestions: > > 1. If you start a complicated system and a mess comes out, > it might be a good idea to try to sort the things out. You have a bit > too much of everything: many atoms, 70 CG steps, empty supercell whose > dimensions you relax (that does not make sense anyway) etc. > How about trying just bulk MoS2 first? > Then, if it not not quite to expectations, you can pull different > levers: basis, cutoffs, XC etc. If everything works fine, you can > proceed to your nanoribbons or whatever. > > 2. If you expect help on a quite vague issue, > it might be a good idea to provide at least the output file. > There are not many cases when the error is seen right away from > the input file, and to expect the colleagues to repeat your calculations > just out of curiosity might be too optimistic. > > Have a nice day, > > Andrei Postnikov >
