Erik van Lenthe. vanlenthe在scm.com
Mon 12 10:19:54 CET 2019

Dear Cristian Celis,

To change a basis set to a larger frozen core is not completely trivial.
One needs to make an input for the Dirac program (that makes the frozen 
core density) and make a new basis set file.

The input for the dirac program is described here
To change the input for dirac is easy.
cp $ADFHOME/atomicdata/Dirac/Am.5d $ADFHOME/atomicdata/Dirac/Am.6p
change the 4th line in Am.6p from
1  300  0  0  0  0  21
1  300  0  0  0  0  24
such that the 6s, 6p1/2 and 6p3/2 are in the core.
Do the same for ADFHOME/atomicdata/Dirac.gaussian if you want a finite 
cp $ADFHOME/atomicdata/Dirac.gaussian/Am.5d 
change the 4th line in Am.6p from
1  300  0  0  0  0  21
1  300  0  0  0  0  24

The basis set file format is described here:
Read this part carefully.
Example to make a TZ2P Am.6p, in which the basis set is almost identical 
to the Am.5p
cp $ADFHOME/atomicdata/ZORA/TZ2P/Am.5d $ADFHOME/atomicdata/ZORA/TZ2P/Am.6p
Move  the '6S    5.350' and '6P    4.750' from valence basis functions 
to core-orthogonalization function.
These basis functions should be at the correct place !

CORE    5  4  3  1
CORE    6  5  3  1

Add the 6s coefficients to the core DESCRIPTION
   -0.68683337538258664324E-03 -0.21144719036565971793E-02
   -0.21741125502845102396E-01  0.55411161057029156146E-02
    0.33381730888860879081E-01  0.61937643433275152427E-01
   -0.23542795141327677033E+00  0.28641794170225015614E+00
    0.49386386724569508733E+00 -0.93817771255972903788E+00
   -0.48774362568029742959E-01  0.79535760007537836813E+00
    0.33031149593468805969E+00 -0.91808020218060130014E-03
    0.29942286843769412227E-02 -0.10295882073977163061E-02
Add the 6p coefficients to the core DESCRIPTION
    0.56843843729799939032E-02  0.54257416086278389966E-01
   -0.50721844016195692662E-01 -0.29957488185829805882E+00
    0.33877546318883555720E+00  0.22665840017401472917E+00
   -0.55950898318960473610E+00  0.56748631065508814419E-01
    0.75278628399493907608E+00  0.33554681350054915701E+00
These coefficients should be at the correct place !

One should be careful with such a large frozen core, that the overlap 
with other frozen core on other atoms is not loo large.
The 6s and 6p of Am are relatively diffuse.
ADF does not correct for frozen core overlap errors.

Best regards,
Erik van Lenthe

You wrote:
> Dear ADF community,
> I wonder if it is possible to manually modify the current available 
> frozen cores (fc) for actinides such as Am, where the largest fc 
> includes up to the 5d shell. I want to include the 6s and 6p shells, 
> so going from Am.5d to Am.6p.
> Is there any way to do that?
> Thanks in advance,
> *Cristian Celis-Barros*
> _______________________________________________
> ADFlist mailing list
> ADFlist at scm.com
> //lists.tofoba.com/mailman/listinfo/adflist

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <//lists.tofoba.com/pipermail/adflist/attachments/20191202/a74f4215/attachment-0001.html>