Hi everyone, I am experiencing a behavior from siesta which I consider rather weird - however, one of you could have an explanation for this.
I have done a very simple calculation - a water molecule in a box of 8x8x8 angstroms - and saved the charge density. I had an old code that converted the RHO file to xsf, so I used it to convert the RHO file. My surprise is that only part of the cell has zero charge density around the water molecule. So, I decided to use Andrei Postnikov's code to convert the file and check if this was something stupid I did in my own code. I compiled it using EXACTLY the same ifort flags I used to compile siesta and the result is the same: only a portion of the unit cell has zero charge density around the molecule, and the rest of the cell, far away from the molecule, has a charge density of about 1.2 e/Bohr^3. The pseudopotentials used were those provided in the water molecule example (LDA), and I used a rather high cutoff, 300 Ry. I have also seen this behavior on other calculations involving vacuum slabs and never bothered much. However, today I started doing stuff related to Bader analysis (inspired by all the posts of Alexander :) ) and I suspect that this is one factor that may be leading to absurd results in my analysis using the code of Henkelman et al. If anyone wishes to help me, I am enclosing the pseudos, the fdf file and output of my calculation showing the compilation flags. So, if anyone can tell me potential problems and/or confirm that the problem indeed happens in machines other than mine, they could help me have a Xmas break even happier than it will certainly be... :o) Cheers, Marcos -- Dr. Marcos Verissimo Alves Post-Doctoral Fellow Unité de Physico-Chimie et de Physique des Matériaux (PCPM) Université Catholique de Louvain 1 Place Croix du Sud, B-1348 Louvain-la-Neuve Belgique ------ Gort, Klaatu barada nikto. Klaatu barada nikto. Klaatu barada nikto.
