developers:pseudos:psp45_format
Differences
This shows you the differences between two versions of the page.
Next revision | Previous revision | ||
developers:pseudos:psp45_format [2015/02/25 13:27] – created Matteo Giantomassi | developers:pseudos:psp45_format [2020/08/10 18:07] (current) – Xavier Gonze | ||
---|---|---|---|
Line 1: | Line 1: | ||
+ | ==== This file contains information about formats of pspcod=4 and pspcod=5 pseudopotentials. ==== | ||
+ | |||
+ | Pspcod=4 corresponds to the case of Teter pseudopotentials | ||
+ | generated in Louvain-La-Neuve using the code ATOM. Pspcod=5 corresponds | ||
+ | to " | ||
+ | it is possible to treat these pseudopotentials (in the version 1.5 and later). | ||
+ | However, one has to be careful about the formats of these pseudopotentials. | ||
+ | Indeed, to be read by the ABINIT code, they have to be | ||
+ | slightly modified. Let's take examples to explain these modifications: | ||
+ | |||
+ | A. pspcod = 4 case (Teter pseudopotentials). The example corresponds to Lead. | ||
+ | |||
+ | 1. OLD HEADER | ||
+ | |||
+ | (Xe+4f14)+6s1.8 5d10 6p0.2 5f0.05; | ||
+ | 82.00000 | ||
+ | | ||
+ | .000 | ||
+ | | ||
+ | 2 2 2 0 | ||
+ | |||
+ | 2. NEW HEADER (READABLE BY ABINIT) | ||
+ | |||
+ | (Xe+4f14)+6s1.8 5d10 6p0.2 5f0.05; | ||
+ | 82.00000 | ||
+ | 4 | ||
+ | 0 0 0 | ||
+ | .000 .000 .000 .000 | ||
+ | 1 0 0 | ||
+ | .000 .000 .000 .000 | ||
+ | 2 0 0 | ||
+ | .000 .000 .000 .000 | ||
+ | 3 0 0 | ||
+ | .000 .000 .000 .000 | ||
+ | .000 | ||
+ | |||
+ | This header corresponds exactly to these of the TM pseudoptentials (pspcod=1). | ||
+ | In addition, in the new form of the Teter pseudopotentials, | ||
+ | | ||
+ | | ||
+ | |||
+ | -------------------------------------------------------------------------------- | ||
+ | |||
+ | B. pspcod = 5 case (" | ||
+ | The example corresponds to Oxygen. | ||
+ | |||
+ | 1. OLD HEADER | ||
+ | |||
+ | Compromise psp for oxygen with rc=1.5 ec=25 double reference | ||
+ | | ||
+ | | ||
+ | 0.999999999999999955E-06 | ||
+ | | ||
+ | | ||
+ | |||
+ | 2. NEW HEADER | ||
+ | |||
+ | Compromise psp for oxygen with rc=1.5 ec=25 double reference | ||
+ | | ||
+ | | ||
+ | | ||
+ | | ||
+ | | ||
+ | | ||
+ | | ||
+ | | ||
+ | |||
+ | |||
+ | |||
+ | Note that this header is almost the same as for the pspcod=1 and pspcod=4 case. | ||
+ | The only difference lies in the presence of " | ||
+ | the Hamman grid). Here also, we indicate in addition the number of each angular | ||
+ | momentum before the set of data corresponding to the considered angular momentum. | ||
+ | |||
+ | ------------------------------------------------------------------------------- | ||
+ | |||
+ | Inside ABINIT, a pseudopotential with format 4 will be treated by | ||
+ | the routine psp1in.f, that calls psp1lo.f (local part), | ||
+ | psp1nl.f (non-local part), and psp1cc (core correction). | ||
+ | |||
+ | As a matter of numerical accuracy, note that the integral | ||
+ | of (V(r)+Zion/ | ||
+ | allowed radius (usually about 100 a.u.), without cut-off. | ||
+ | V(r)+Zion/r should tend rapidly to zero | ||
+ | for large radii (beyond 5 a.u.), but this correct behaviour will | ||
+ | not be enforced by psp1lo . If the tail of V(r) is inaccurate | ||
+ | (i.e. if the pseudopotential is in single precision), there | ||
+ | will be large inaccuracies in the integral, because of the r^2 factor. | ||
+ | |||
+ | By contrast, a pseudopotential with format 5 will be treated by | ||
+ | the routine psp5in.f, that calls psp5lo.f (local part), | ||
+ | psp5nl.f (non-local part), and psp5cc (core correction). | ||
+ | |||
+ | The integral of (V(r)+Zion/ | ||
+ | from 0 to the highest allowed radius (usually about 100 a.u.), | ||
+ | except that beyond 20 a.u. no value of abs(V(r)) larger | ||
+ | than 2.0d-8 is tolerated. This will allow to cut off spurious | ||
+ | behaviour of pseudopotential whose data file is written in single precision. | ||