Robert, I guess so. However, you could try including only the 5p orbitals in the valence and applying NLCC to try to account for the 5s if using an ionic pseudo really makes you uncomfortable. I checked the pesudos on the PWSCF website and while they include the 5s orbitals as semicore, they are ultrasoft ones - if I'm not mistaken they don't have the same restrictions as TM pesudos in what regards the angular momentum channelsf that can be included in the valence. So it might be worth a shot.
On the other hand, aren't there any papers out there for calculations with La using norm-conserving pseudos that you could check to see if the recipe (that is, channels and rc's included) is given? Nothing on the siesta list of publications? Marcos On Mon, Jun 21, 2010 at 9:01 PM, Koch, Robert J <[email protected]> wrote: > The only way I'm able to do this is by dropping the 6s2 states (to include > semicore 5s2). Does this mean when I use this PP in a calculation it will be > ionic? > > -Rob > ________________________________________ > From: Marcos Veríssimo Alves [[email protected]] > Sent: Monday, June 21, 2010 1:03 PM > To: [email protected] > Subject: Re: [SIESTA-L] Lanthanum Pseudo > > Robert, I'm afraid that you'll have to includie semicore states > then... I'm not sure that NLCC can always account for the cases it > should, everything has a breakdown case. > > Marcos > > On Mon, Jun 21, 2010 at 6:45 PM, Koch, Robert J <[email protected]> wrote: >> Hello all! >> >> I am generating a PP for Lanthanum in the ground state. The cross >> excitation energies are way off between the AE and PT series. Can anyone >> provide advice to fix this? I've tried incorporating core corection, but I >> don't seem to be able to choose "good" core radii, since I am not getting >> anything which is bringing me close to where my excitation energies should >> be (the AE values). I've read Louie, Froyen, Cohen, Phys. Rev. B 26, 1738 >> (1982), but it hasn't helped. >> >> I've tried including the 5p states as semicore, but this makes things much >> worse. Note that in this configuration I cannot incorporate 5s states as >> redundant L channels with different n values are not allowed. >> >> Any advice is greatly appreciated! >> >> ATM 3.2.2 21-JUN-10 Lanthanum GS-5d1 6s2 &v&d >> ATM 3.2.2 21-JUN-10 Lanthanum 5d0 6s2 4f1 &v&d >> ATM 3.2.2 21-JUN-10 Lanthanum 5d0 6s1 4f2 &v&d >> ATM 3.2.2 21-JUN-10 Lanthanum 5d0 6s0 4f3 &v&d >> &d total energy differences in series >> &d 1 2 3 4 >> &d 1 0.0000 >> &d 2 0.6703 0.0000 >> &d 3 0.3236 -0.3467 0.0000 >> &d 4 0.6380 -0.0323 0.3144 0.0000 >> *----- End of series ----* spdfg &d&v >> ATM 3.2.2 21-JUN-10 Lanthanum GS-5d1 6s2 &v&d >> ATM 3.2.2 21-JUN-10 Lanthanum 5d0 6s2 4f1 &v&d >> ATM 3.2.2 21-JUN-10 Lanthanum 5d0 6s1 4f2 &v&d >> ATM 3.2.2 21-JUN-10 Lanthanum 5d0 6s0 4f3 &v&d >> &d total energy differences in series >> &d 1 2 3 4 >> &d 1 0.0000 >> &d 2 1.4060 0.0000 >> &d 3 0.4306 -0.9754 0.0000 >> &d 4 0.7982 -0.6078 0.3676 0.0000 >> *----- End of series ----* spdfg &d&v >> >> >> >> pe Lanthnum >> tm2 3.00 >> n=La c=rvr >> 0.0 0.0 0.0 0.0 0.0 0.0 >> 11 4 >> 6 0 2.00 0.00 >> 6 1 0.00 0.00 >> 5 2 1.00 0.00 >> 4 3 0.00 0.00 >> 3.50 4.10 3.50 1.00 1.00 1.15 >> >> >> Best, >> Rob Koch >> >> >
