Sample directory: adf/e_Frags_PtCl4H2
The (scalar) relativistic option (Pauli formalism) is used because of the presence of the relativistic Pt atom. The complex is built from fragments H2 and PtCl42–.
The program dirac is applied to generate the corepotentials file for all involved atom types, including Hydrogen. The latter has no frozen core, let alone a relativistic one, but the corepotentials file also contains the total (relativistic) atomic potential. The (relativistic) atomic total potential is used in some types of relativistic options only, but it is a good idea to simply always run DIRAC for all the atoms whenever you do a relativistic calculation.
$ADFBIN/dirac <$adfresources/Dirac/Cl.2p $ADFBIN/dirac <$adfresources/Dirac/Pt.4d $ADFBIN/dirac <$adfresources/Dirac/H mv TAPE12 t12.rel
The script above generates one TAPE12 file. The second and third dirac runs recognize the presence of the TAPE12 file (with the standard name 'TAPE12') that resulted from the earlier ones and they write their resulting data to the tail of it.
$ADFBIN/adf <<eor Create H file=$ADFRESOURCES/DZP/H XC LDA vwn GGA becke perdew End Relativistic Scalar CorePotentials t12.rel & H 3 End End Input eor mv TAPE21 t21H
The final calculations of the molecule and larger fragments are performed with gga ('NonLocal') xc corrections. Although it is not necessary to use the same settings in the Create runs, applying them looks 'nicer' and gives a better approximation of the bond energy of the molecule with respect to the basic atoms. Here it serves to show that also in a Create run various options can be used.
$ADFBIN/adf <<eor create Cl file=$ADFRESOURCES/DZP/Cl.2p xc lda vwn GGA becke perdew end relativistic scalar corepotentials t12.rel & Cl 1 end end input eor mv TAPE21 t21Cl $ADFBIN/adf <<eor Create Pt file=$ADFRESOURCES/DZ/Pt.4d XC lda vwn GGA becke perdew End Relativistic scalar CorePotentials t12.rel & Pt 2 End End Input eor mv TAPE21 t21Pt
It is important to use the relativistic option in the creation of the fragments if the final molecule will use it as well. The corepotentials file is attached and the input indicates that the section on that file for Cl is #1, while the Pt data are in section #2.
Now, all basic atoms have been generated. We go on to generate the two larger fragments H2 and PtCl42- from which we are going to build the final complex.
$ADFBIN/adf <<eor Title H2 R=1.68a.u. NoPrint sfo,frag,functions Units length bohr End Atoms H 0.0 0.0 0.84 H 0.0 0.0 -0.84 End Fragments H t21H End XC lda vwn GGA becke perdew End Relativistic Scalar CorePotentials t12.rel & H 3 End End Input eor mv TAPE21 t21H2
The result file TAPE21 is renamed and saved to serve as fragment file.
$adf <<eor title PtCl4 (2-) noprint sfo,frag,functions units length bohr end ATOMS Pt 0 0 0 Cl 4.361580 0.000000 0 Cl 0.000000 4.361580 0 Cl -4.361580 0.000000 0 Cl 0.000000 -4.361580 0 end fragments Pt t21Pt Cl t21Cl end xc lda vwn GGA becke perdew end relativistic scalar corepotentials t12.rel & Cl 1 Pt 2 end charge -2 end input eor mv TAPE21 t21PtCl4
The key charge is used to specify the net total charge. The default for the net total charge is the sum-of-fragment-charges. The fragments (Pt and Cl atoms) have been computed neutrally, but we want to calculate the PtCl4 complex as a 2- ion.
Finally we compute PtCl4H22- from the fragments PtCl42- and H2.
$ADFBIN/adf <<eor title PtCl4 H2 units length bohr end integration 4.0 xc lda vwn GGA becke perdew end relativistic scalar corepotentials t12.rel & H 3 Cl 1 Pt 2 end ATOMS Pt 0 0 0 f=PtCl4 Cl 4.361580 0.000000 0.00000000 f=PtCl4 Cl 0.000000 4.361580 0.00000000 f=PtCl4 Cl -4.361580 0.000000 0.00000000 f=PtCl4 Cl 0.000000 -4.361580 0.00000000 f=PtCl4 H 0.0 0.0 5.58 f=H2 H 0.0 0.0 7.26 f=H2 end fragments PtCl4 t21PtCl4 H2 t21H2 end end input eor
Note that, although the key charge is not supplied, the molecule is not neutral: the default charge (that is, omitting the keys charge, occupations) is the sum-of-fragments: the fragments here are H2 and PtCl42-, yielding a net charge for the molecule of minus two.
Note the f= fragment specification in the Atoms block. No fragment-numbering suffix (/n) is required because there is only one fragment of each fragment type.




