These are pseudopotentials that I've either created or collected. I have tested them and believe that for my research they produce accurate results. I do not guarantee that any of these will work for research projects other than mine, i.e., you should test these yourself before you use them for any calculation that you're wagering your career on. Any comments, corrections, or suggestions would be appreciated.
-- Scott
The below potentials seem to work well for perovskite PbTiO3. They are norm-conserving and use the LDA (Ceperley and Alder).1 I'm happy with the softness and transferability of the O and Ti, although I'm sure that there is room for improvement. The pseudopotentials are given as input files for the 2.0.6 version of the Opium pseudopotential generator.
Pb
I am using the Pb pseudopotential available from the
Rappe group's library.
If you can't find the library from this link, write me and
I'll send you my copy of the param file.
Ti
The input file is ti.param and the opium report file is here
ti.rpt.
O
The input file is o.param and the opium report file is here
o.rpt. I should note that I started with the Rappe group's
GGA oxygen potential and then tweaked it to work using the LDA exchange-correlation.
Tests were conducted using Abinit. The energy was converged to 10 meV and the forces to better than 5 meV/Angstrom when a plane wave cut-off of 45 Hartree was used. The Brillouin zone is sampled using a 6x6x6 Monkhorst-Pack grid.
The cubic lattice parameter is calculated to be 3.915 Angstrom, which is within 1.5% of the experimental value 3.696 Angstrom. (insert reference) The tetragonal lattice parameters, are calculated to be a=3.8876 Angstrom and c=4.0792 Angstrom. The ratio of which is c/a=1.0493. This compares to experiment.....
The polarization of the tetragonal structure is calculated to be......