Dear sir, Thank you for responding > please note that convergence parameters can be _very_ arbitrary > between different codes and they can be in very strange units.
sure, I think I have convert the units properly. I want to use 1.0e-5 eV/atom as the CASTEP do for my Energy convergence. Do you think I should multiple 12 for my cell has 12 atoms? Maybe I did it wrong... > before going any further with running big calculations. try > smaller ones. I think it's small enough. It can be finished within an hour. > and have you checked the requirements for each > pseudopotential? is your stress tensor converged wrt. wavefunction > _and_ density cutoff? ...and k-points? you are using vanderbilt > ultrasoft pseudopotentials and they can require unexpectedly high > density cutoffs. Many piece of works published on journals said 25 and 200 Ryd is enough. I'll test a 400 eV ( 30 Ryd ) later. > if i remember correctly. you are also not mentioning > whether the reference you are comparing to is using the _same_ > functional that you are using. different functionals > tend to over/underestimate lattice parameters differently. Here is part of the reference: J. Phys. Chem. B 2006, 110, 7463-7472 Calculations were carried out in the framework of density functional theory by using the Vienna ab-initio simulation package (VASP). Ultrasoft pseudopotentials were used to describe the ionic cores, and a plane-wave basis set with a cutoff energy of 400 eV was used to expand the valence electrons. The nonlocal exchange and correlation energies were calculated with the Perdew-Wang (PW91) functional of the generalized gradient approximation (GGA). Calculations including spinpolarization have been performed for all structures presented here. The results were compared with those from non-spinpolarized calculations in the case of a zero magnetic moment. The consistence of the results from both calculations ensures that the system is in a true nonmagnetic state. The geometries were relaxed using a conjugate-gradient or quasi-Newton scheme as implemented in VASP. The atomic structures were relaxed until the forces on all unconstrained atoms were less than 0.03 eV/?. Similar setups have been employed in the study of a wide range of systems including metal and metal-oxide surfaces by the authors and many others. The calculated bulk anatase TiO2 lattice parameters a = b = 3.803 A, c = 9.603 A, c/a = 2.525, which agree well with the experiments, were used to construct the periodic slabs for the (101) surface calculations. ............. Also other literatures, such as Journal of Molecular Structure (Theochem) 709 (2004) 73?78. As the above paragraph in literature said, what kind of pseudopotentials should be used?Ti.pw91-nsp-van.UPF ? Or Ti.pw91-sp-van_ak.UPF? Or Ti.pbe-sp-van_ak.UPF ? I note that in castep the corresponding name of pseudopotentials are O_00PBE.usp, Ti_00PBE.usp. But there is another option for us to choose x-c functional. While QE 4.0 doesn't has, QE read from pseudopotential files. So Ti.pbe-xxxxx.UPF and O.pbe-xxx.UPF maybe not suitable. What do you think witch is more suitable for calculation described by the reference? By the way, I note that CASTEP has find 32 symmetry operations of my cell. My cell is a standard anatase crystall. But Q-E only can find 4 at most since I adjust parameters of ~/espresso-4.0/PW/eqvect.f90. Do this matter effect my results significantly? Vega Lew PH.D Candidate in Chemical Engineering State Key Laboratory of Materials-oriented Chemical Engineering College of Chemistry and Chemical Engineering Nanjing University of Technology, 210009, Nanjing, Jiangsu, China _________________________________________________________________ News, entertainment and everything you care about at Live.com. Get it now! http://www.live.com/getstarted.aspx -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.democritos.it/pipermail/pw_forum/attachments/20080704/86d3827f/attachment.htm
