The ADF package contains a series of sample runs. Provided are UNIX scripts to run the calculations and the resulting output files.
The examples serve:
Where references are made to the operating system (OS) and to the file system on your computer,
the terminology of a UNIX type OS is used and a hierarchical structure of directories is assumed.
All sample files are stored in subdirectories under $ADFHOME/examples/, where $ADFHOME is the main directory of the ADF package. There are two main subdirectories in examples/: adf/ for calculations with the molecular code ADF (and related utility programs) and band/ for calculations with the periodic structures code BAND. Each sample run has its own directory (under adf/ or band/ respectively). For instance, $ADFHOME/examples/band/NaCl/ contains an BAND calculation on the NaCl bulk crystal. Each sample subdirectory contains:
Notes:
Many of the provided samples have been devised to be short and simple, at the expense of physical or chemical relevance and precision or general quality of results. They serve primarily to illustrate the use of input, necessary files, and type of results. The descriptions have been kept brief. Extensive information about using keywords in input and their implications is given in the User's Guides (ADF and BAND) and the Utilities and Property Programs documents (NMR, DIRAC, and other utility programs).
When you compare your own results with the sample outputs, you should check in particular (as far as applicable):
General remarks about comparisons:
Default settings of print options result in a considerable amount of output. This is also the case in some of the sample runs, although in many of them quite a bit of 'standard' output is suppressed by inserting applicable print control keys in the input file. Consult the User's Guide about how to regulate input with keys in the input file.
Sample directory: band/NiCu_XC/
This is an important example featuring many important keywords.
The line:
Skip DOS
reduces the length of the output considerably.
A spin-unrestricted
calculation can be done by using the keyword
Unrestricted
The Becke Perdew generalized gradient approximation (GGA) for the exchange-correlation functional is invoked through the XC block:
XC GGA Always Becke Perdew End
The subkey ALWAYS indicates that the Becke-Perdew xc potential should be used during the SCF. If this is not specified, the GGA energy will be calculated post-SCF using the LDA density.
$ADFBIN/band << eor Title Surface alloy: Cu slab with one surface Cu replaced by Ni (1:1) Comment Technical Low quadratic K space integration Reasonable real space integration accuracy Features Lattice : 2D Unit cell : 4 atoms, sqrt(2) x sqrt(2), 2 layers Basis : NO+STO w/ core Options : XC functional in SCF End Skip DOS Convergence Degenerate default End DIIS NCycleDamp 15 DiMix 0.25 End SCF Mixing 0.10 End Accuracy 3.5 Kspace 3 Unrestricted XC GGA Always Becke Perdew End Units Length Angstrom End Define latt=3.61 ! FCC lattice parameter halflatt=latt/2 End Lattice latt 0 0 0 latt 0 End Atoms Ni 0 0 0 End Atoms Cu :: layer 1 halflatt halflatt 0 :: layer 2 0 halflatt -halflatt halflatt 0 -halflatt End ... EndInput eor
Sample directory: band/LiMetaGGA/
Example for a post SCF metaGGA calculation.
$ADFBIN/band << eor Title Li bulk KSpace 3 accuracy 5 DIIS dimix 0.3 ncycledamp 0 end scf mixing 0.6 end xc metaGGA postscf end define ha=6.60/2 end Lattice -ha ha ha ha -ha ha ha ha -ha end atoms Li 0.0 end BasisDefaults BasisType DZ Core Large end end input eor
Sample directory: band/Pt_slab/
This example can of course be compared directly to the Cu slab. This example is important, as SCF convergence is frequently difficult in slab calculations. The specifications in the CONVERGENCE, SCF, and DIIS blocks are typical. Such settings are recommended in slab calculations with convergence problems.
The DEGENERATE subkey specifies that bands with the same energy should have the same occupation numbers. This helps SCF convergence. The same is true for the values for the MIXING subkey in the SCF block and the DIMIX subkey in the DIIS block. Please note that the recommended value for Mixing is approximately half of the value for Dimix.
Another important feature in BAND is that it enables relativistic treatments for heavy nuclei. Both the ZORA scalar relativistic option and spin-orbit effects have been implemented. The line
Relativistic ZORA SPIN
specifies that in this case both the scalar relativistic effects (ZORA) and spin-orbit effects (SPIN) will be taken into account. Whereas the ZORA keyword does not make the calculation much more time-consuming, the same cannot be said for the spin-orbit option. Usually the ZORA keyword will give the most pronounced relativistic effects and the spin-orbit effects will be a fairly minor correction to that. We therefore recommend scalar ZORA as a good default method for treating heavy nuclei.
The DEPENDENCY keyword means that the calculation should continue even if the basis is nearly linearly dependent (as measured by the eigenvalues of the overlap matrix).
$ADFBIN/band << eor Title Platinum slab Comment Technical Low quadratic K space integration Low real space integration accuracy Features Lattice : 2D Unit cell : 3 atoms, 1x1 Basis : NO+STO w/ core Options : Spinorbit ZORA End Convergence Degenerate 1.0E-03 End SCF Iterations 60 Mixing 0.06 End DIIS NCycleDamp 15 DiMix 0.15 End KSpace 3 Accuracy 3 Relativistic ZORA SPIN Dependency Basis=1E-8 Define latt=7.41 lvec=latt/SQRT(2.0) ysh=lvec/SQRT(3.0) dlay=latt/SQRT(3.0) End Lattice SQRT(3.0)*lvec/2.0 0.5*lvec SQRT(3.0)*lvec/2.0 -0.5*lvec End Atoms Pt 0 0 0 :: layer 1 -ysh 0.0 -dlay :: layer 2 ysh 0.0 -2.0*dlay :: layer 3 End ... END INPUT eor
Sample directory: band/BetaIron/
With the UNRESTRICTED keyword we do a spin polarized calculation. Normally this would converge to the ferromagnetic solution.
With the SpinFlip keyword we make sure that we start with an antiferrmagnetic density.
For antiferromagnetic iron we need a larger unit cell of two atoms. Because these atoms appear to the program as symmetry equivalent we have to specify them as sepearate types.
$ADFBIN/band << eor
TITLE Beta iron
UNITS
length Angstrom
angle Degree
END
ATOMS Fe
0.0 0.0 0.0
end
! second iron as new type to break the symmetry
atoms Fe
-1.435 -1.435 1.435
END
Lattice
-1.435 1.435 1.435
1.435 -1.435 1.435
2.87 2.87 -2.87
End
CONVERGENCE
CRITERION 1.0e-4
Degenerate default
SpinFlip 2 ! Flip (startup) spin density at second atom
END
BasisDefaults
BasisType DZ
Core Large
End
UNRESTRICTED
end input
eor
Sample directory: band/Graphene_Dispersion/
A normal GGA would give no interaction between two graphene sheets.
Use the dispersion option in the XC keyblock.
In the first run we use BP86-D, and in the second BLYP-D3.
$ADFBIN/band << eor
TITLE Dispersion energy with two parallel graphene sheets
kspace 3
accuracy 4.5
XC
gga scf bp86
dispersion default
end
dependency basis=1e-6
UNITS
length Angstrom
angle Degree
END
ATOMS C
0.0 0.0 0.0
0.0 0.0 -3.355
1.23 0.7101408312 0.0
-1.23 -0.7101408311 -3.355
END
Lattice
2.46 0.000000 0
1.23 2.130422493 0
End
CONVERGENCE
CRITERION 1.0e-4
Degenerate default
END
DIIS
CLARGE 10
CHUGE 30
DIMIX 0.3
NVCTRX 10
NCYCLEDAMP 5
END
SCF
iterations 50
mixing 0.2
END
BasisDefaults
BasisType TZP
Core Large
End
end input
eor
rm RUNKF
$ADFBIN/band << eor
TITLE Grimme3 dispersion energy with two parallel graphene sheets
kspace 3
accuracy 4.5
XC
gga scf blyp
dispersion Grimme3
end
dependency basis=1e-6
Gradients
End
UNITS
length Angstrom
angle Degree
END
ATOMS C
0.0 0.0 0.0
0.0 0.0 -3.355
1.23 0.7101408312 0.0
-1.23 -0.7101408311 -3.355
END
Lattice
2.46 0.000000 0
1.23 2.130422493 0
End
CONVERGENCE
CRITERION 1.0e-4
Degenerate default
END
DIIS
CLARGE 10
CHUGE 30
DIMIX 0.3
NVCTRX 10
NCYCLEDAMP 5
END
SCF
iterations 50
mixing 0.2
END
BasisDefaults
BasisType TZP
Core Large
End
end input
eor
Sample directory: band/HonPerovskite_Solvation/
We want to model H adsorption on a Perovskite surface in a solution, modelled by a COSMO surface.
We create only the COSMO surface above the slab with the RemovePointsWithNegativeZ option.
$ADFBIN/band << eor
TITLE Hydrogen on Perovksite wit solvation
PeriodicSolvation
nstar 3
SymmetrizeSurfacePoints true
RemovePointsWithNegativeZ true
end
screening
rmadel 30 ! to speed up the calculation
end
Solvation
Surf Esurf
charge method=inver
end
UNITS
length Angstrom
angle Degree
END
atoms H
0 0 0.9
end
ATOMS Ca
0.0 0.0 0.0
0.0 3.535533906 -3.535533906
END
ATOMS Ti
-2.5 -3.535533906 0.0
-2.5 4.440892099e-16 -3.535533906
END
ATOMS O
0.0 -3.535533906 0.0
2.5 1.767766953 -1.767766953
2.5 -1.767766953 -1.767766953
END
diis
dimix 0.3
nvctrx 20
end
Lattice
5.0 0.000000 0
0.000000 7.071067812 0
End
CONVERGENCE
CRITERION 1.0e-4
Degenerate default
END
BasisDefaults
BasisType SZ
Core Large
End
scf
iterations 30
end
XC
LDA SCF VWN
END
Dependency basis=1e-6
end input
eor
Sample directory: band/EField/
With the EFIELD keyword you can specify a static electric field in the z-direction.
$ADFBIN/band << eor title field in au kspace 1 accuracy 5 integeration accint 6 end radialDefaults nr 10000 end Gradients End EField Unit a.u. eZ=-0.03 end lattice 15 0 0 0 15 0 end atoms H 0 0 0 end atoms Li 0 1 3.0 end BasisDefaults BasisType TZP Core Large end end input eor
Sample directory: band/NiO_Hubbard/
With the UNRESTRICTED keyword we do a spin polarized calculation.
With the HubbardU key block we set up the GGA+U calculation. You need to specify per atom type (only two here, Ni, and O) the U and the l-value to which it should be applied.
$ADFBIN/band << eor
TITLE NiO
UNITS
length Angstrom
angle Degree
END
HubbardU
printOccupations true
Enabled true
uvalue 0.3 0.0 ! Type 1 (Ni) will get U=0.3
lvalue 2 -1 ! Will be used for the d-orbitals of Ni
End
ATOMS Ni
0.0 0.0 0.0
END
ATOMS O
2.085 2.085 2.085
END
Lattice
0.000000 2.085 2.085
2.085 0.000000 2.085
2.085 2.085 0.000000
End
CONVERGENCE
CRITERION 1.0e-4
Degenerate default
END
DIIS
CLARGE 10
CHUGE 10
DIMIX 0.2
NVCTRX 20
NCYCLEDAMP 0
END
SCF
iterations 50
mixing 0.1
END
BasisDefaults
BasisType TZP
Core Large
End
XC
GGA Becke Perdew
END
UNRESTRICTED
Dependency basis=1e-6
end input
eor
Sample directory: band/ZnS_ModelPotential/
With the XC sub key model we invoke the so-called TB-mBJ model potential, which increases band gaps for solids.
$ADFBIN/band << eor
TITLE ZnS
accuracy 4.5
kspace 3
UNITS
length Angstrom
angle Degree
END
ATOMS Zn
0.0 0.0 0.0
END
ATOMS S
1.3525 1.3525 1.3525
END
Lattice
0.000000 2.705 2.705
2.705 0.000000 2.705
2.705 2.705 0.000000
End
CONVERGENCE
Degenerate default
END
DIIS
CLARGE 10
CHUGE 10
DIMIX 0.2
NVCTRX 20
NCYCLEDAMP 0
END
BasisDefaults
BasisType DZ
Core Large
End
XC
model TB-mBJ
END
Dependency basis=1e-8
end input
eor
Sample directory: band/BasisDefaults/
This example shows some of the flexibility of the BasisDefaults key. The defaults are set to a DZ basis and a Large core. As the example shows, it is possible to override the defaults per atom type by specifying subkeys in Atoms blocks.
$ADFBIN/band << eor Title CO + H2: fine tuning the basis defaults KSpace 1 Accuracy 4 Define far=20 End Lattice far End ! So we have cheap defaults BasisDefaults BasisType DZ Core Large End Atoms C ! This C has no frozen core 0 Core None End Atoms O ! This O with a larger basis 0 2.13 BasisType TZ2P End Atoms H ! This one also with a larger basis 4 0 BasisType V End Atoms H ! Let us use the default settings for this atom 4 1.43 End END INPUT eor
Sample directory: band/Cu2_Confine/
This example illustrates two keywords to speed up BAND calculations and reduce the required disk space to some extent, without significant loss of accuracy.
The TAILS keyword implies that cutoffs will be applied to the (basis) functions. Basis functions will be assumed to be zero outside the radius where the remaining relative norm of the function is smaller than 0.00001, in this example. The shape of the function is slightly modified, in a 'soft' way, beyond the radius where the relative norm of the function outside the radius is 0.01. The Coulomb option is always recommended in this line as it gives more accurate results for core functions.
The CONFINEMENT option can be specified for each atom type separately. It is based on a soft cut-off related to a distance criterion. In this sense it is quite different from the tails option. The CONFINEMENT option can be useful in itself, for avoiding near linear dependency in the basis. However, its use in making BAND calculations more efficient is primarily in combination with the tails option.
$ADFBIN/band << eor Title Copper dimer - confinement and tails options KSpace 1 Accuracy 4.5 Dependency Basis=1E-10 Fit=1E-7 Units Angle Radian End SCF Mixing 0.05 Iterations 20 End Convergence Degenerate default End DIIS NCycleDamp 0 DiMix 0.15 End Define bbb=20 ccc=4.2 End Lattice bbb bbb 0 -bbb bbb 0 End Atoms Cu 0 0 0 0 0 ccc End .... END INPUT eor
The output file prints some information on how much was gained because of the tails and confinement options in various routines, for example:
Compression due to negligible tails in baspnt
max. percentage stored per node 98.23%
min. percentage stored per node 95.45%
Sample directory: band/RestartSCF/
This very simple example shows how you can continue with an unfinished calculation. It consists of two runs. After the first run the RUNKF file is saved, and the renamed file is used in the second run. The second run is almost a copy for the first, except for the Restart key. You can, however, change more, as long as the basis set remains the same.
# ----------------------------- first run -------------------------- $ADFBIN/band << eor Title B chain KSpace 3 Accuracy 4 skip dos XC GGA Becke Perdew END UNRESTRICTED SCF Mixing 0.4 Iterations 40 End Convergence Degenerate default End DIIS NCycleDamp 0 DiMix 0.5 End Define ddd=4.0 ccc=1.5 End Lattice ddd End Atoms B 0 End BasisDefaults BasisType TZ2P Core Large End END INPUT eor mv RUNKF BChain.runkf rm Points # ----------------------------- second run -------------------------- $ADFBIN/band << eor Title B chain restart KSpace 3 Accuracy 4 XC GGA Becke Perdew END UNRESTRICTED Restart BChain.runkf& scf end SCF Mixing 0.4 Iterations 40 End Convergence Degenerate default End DIIS NCycleDamp 0 DiMix 0.5 End Define ddd=4.0 zzz=3 End Lattice ddd End Atoms B 0 End BasisDefaults BasisType TZ2P Core Large End END INPUT eor
Sample directory: band/BeO_tape41/
If we save the RUNKF file of a calculation we can restart it to calculate things on a grid.
In the second run we restart from the file BeO.runkf. We specify to use a regular grid and ask the program to calculate a bunch of properties on that grid.
$ADFBIN/band << eor Title BeO Comment in the "ideal" most symmetric configuration: u=3/8, and c=sqrt(8/3)*a End KSpace 3 accuracy 4 Dependency basis=1e-9 fit=1e-8 DIIS dimix 0.2 ncycledamp 0 end scf mixing 0.4 end xc gga scf bp86 end Define uuu=3/8 aaa=5.10 ccc=sqrt(8/3)*aaa End Coordinates Natural ATOMS Be 0 0 0 1/3 1/3 1/2 END ATOMS O 0 0 uuu 1/3 1/3 uuu+1/2 END Lattice aaa 0 0 0.5*aaa 0.5*sqrt(3)*aaa 0 0 0 ccc End BasisDefaults BasisType DZ Core large end end input eor mv RUNKF BeO.runkf $ADFBIN/band << eor Title BeO Comment in the "ideal" most symmetric configuration: u=3/8, and c=sqrt(8/3)*a End Restart BeO.runkf & DensityPlot End Grid Type Coarse End DensityPlot FITDENSITY_deformation_scf FITDENSITY_total_scf ATOMIC_density ATOMIC_coulombPot COULOMBPOTENTIAL_multipole_deformation_scf COULOMBPOTENTIAL_scf COULOMBPOTENTIAL_shortrange_deformation_scf XCPOTENTIAL_scf End KSpace 3 accuracy 4 Dependency basis=1e-9 fit=1e-8 DIIS dimix 0.2 ncycledamp 0 end scf mixing 0.4 end xc gga scf bp86 end Define uuu=3/8 aaa=5.10 ccc=sqrt(8/3)*aaa End Coordinates Natural ATOMS Be 0 0 0 1/3 1/3 1/2 END ATOMS O 0 0 uuu 1/3 1/3 uuu+1/2 END Lattice aaa 0 0 0.5*aaa 0.5*sqrt(3)*aaa 0 0 0 ccc End BasisDefaults BasisType DZ Core large end end input eor
Sample directory: band/NaCl/
A bulk crystal computation for Sodium Chloride (common salt), with a subsequent DOS analysis, using a Restart facility to use the results from a preceding calculation.
Calculations on periodic systems are carried out with the BAND program. Its input format has recently been changed substantially. It is now more similar in style to ADF. Old BAND input files are no longer compatible with the new version however.
The BAND input still follows slightly different conventions from the ADF input, for historical reasons.
The COMMENT keyword allows users to provide some information about the run which may be of use later. Usually a brief summary of the run is given here.
Numerical integration precision is controlled with the key Accuracy (in ADF: Integration)
The accuracy for integrals over the Brillouin Zone is set by the Kspace key. The latter should, generally, take as value an odd number (3, 5...) to invoke the accurate quadratic tetrahedron integration procedure. For even values it will revert to the linear tetrahedron method, which is almost always inferior in accuracy.
The key Angstroms specifies that geometric data, such as lattice constants are in angstrom units.
Since there are 3 data records in the Lattice block, the calculation will assume 3-dimensional periodicity, with lattice vectors as indicated. Note that lattice vectors are undefined up to linear combinations among themselves. Internally, the program will recombine the input vectors so as to minimize the size of the actually used vectors.
The input line Coordinates NATURAL means that atomic positions are input as coefficients in terms of the lattice vectors, rather than as absolute (Cartesian) coordinate values.
For each of the atoms in the calculations, Na and Cl here, there must be data blocks to specify various items. First, their positions in the crystal unit cell (key Atoms). Second, the single isolated atom computation that will serve as start-up (Dirac). Third, any Slater-type orbital basis functions (BasisFunctions) for that atom. Fourth, the fit functions (FitFunctions) for the calculation of the Coulomb potential and. The third item (BasisFunctions) is optional and not present in this example.
It is recommended to include the numerical atomic orbitals that are computed by the Herman-Skillman type subprogram DIRAC as basis functions for the periodic structure calculation. This is effectuated by putting the word VALENCE in the Dirac data blocks. If that is done, additional STO basis functions (key BasisFunctions) are optional and are used to increase the basis set flexibility. In absence of the numerical (DIRAC/VALENCE) orbitals, a minimal STO set is necessary of course, lest we wouldn't have any basis set at all.
In an equivalent ADF calculation, basis and fit functions would be provided through the Create runs, which pick up the basis and fit functions from a database file. The Create runs would also serve to provide the start-up density, as the DIRAC runs do in BAND.
The basis and fit sets that one has to insert into the BAND input files can be taken from the corresponding ADF database. Note, however, that ADF does not use any numerical orbitals. Since it is recommended to include such numerical orbitals in a BAND calculation, one has to adjust the STO-type basis set for BAND, in comparison with ADF, so as to avoid linear dependency with the numerical orbitals. As a general guideline: for each of the included numerical orbitals (the occupied valence orbitals of the DIRAC calculation), one should remove one STO of the appropriate (n,l)-value. This keeps the overall size and flexibility of the basis at the same level and is usually sufficient to avoid dependency troubles.
The RUNKF key, early in the input, specifies that this standard result file from BAND must be saved under the name 't21.NaCl'. This file will be used in the follow-up calculation of Density-of-States properties.
Note, finally, that the data blocks of block type keys in the input for BAND end with a record 'END', as in ADF, whereas previously '**' was used in BAND to end a record.
$ADFBIN/band << eor Title Title NaCl (from neutral atoms) Comment Technical Hybrid K space integration (3D) Low real space integration accuracy Natural coordinates Lengths in Angstrom Parameters Dirac procedure Features Lattice : 3D Unit cell : 2 atoms Basis : NO w/ core Options : Save restart file End Accuracy 3.5 Kspace 3 Units Length Angstrom End Lattice 0 2.75 2.75 2.75 0 2.75 2.75 2.75 0 End ATOMS NA 0 End Coordinates Natural Atoms Cl .5 .5 .5 End AtomType Na Dirac Na 4 1 Radial 2000 RMin 1E-4 RMax 60 VALENCE 1 0 2 0 2 1 3 0 1.0 SubEnd FitFunctions 1 0 18.9 2 0 30.3 2 0 15.5 3 0 14.9 3 0 8.9 4 0 7.8 4 0 5.1 4 0 3.3 5 0 2.8 5 0 1.9 5 0 1.3 2 1 14.3 3 1 9.9 4 1 6.7 4 1 3.4 5 1 2.4 5 1 1.3 3 2 10.5 4 2 5.4 5 2 3.0 5 2 1.3 4 3 5.8 5 3 1.7 5 4 2.0 SubEnd End AtomType Cl Dirac Cl 5 3 VALENCE 1 0 2 0 2 1 3 0 3 1 5.0 SubEnd FitFunctions 1 0 29.1 2 0 49.5 2 0 26.1 3 0 25.8 3 0 15.8 4 0 14.2 4 0 9.4 4 0 6.2 5 0 5.4 5 0 3.8 5 0 2.6 2 1 21.2 3 1 16.5 4 1 12.4 4 1 6.8 5 1 5.1 5 1 3.1 3 2 16.6 4 2 9.4 5 2 5.5 5 2 2.6 4 3 8.7 5 3 3.3 5 4 4.0 SubEnd End End Input eor mv RUNKF t21.NaCl rm Points
The next run has largely the same input and provides a restart of the previous run.
The key RESTARTDOS tells the program to pick up the indicated file as restart file and to use it for DOS analysis purposes.
The DOS key block details the energy grid (and range) and the file to write the data to. The optional keys GROSSPOPULATIONS and OverlapPopulations invoke the computation of, respectively, gross populations and overlap populations (i.e. for each of these the density-of-states values in the user-defined energy grid).
$ADFBIN/band << eor
Title Title NaCl (from neutral atoms) DOS analysis (restart)
Comment
Technical
Hybrid K space integration (3D)
Low real space integration accuracy
Natural coordinates
Lengths in Angstrom
Parameters Dirac procedure
Features
Lattice : 3D
Unit cell : 2 atoms
Basis : NO w/ core
Options : Use restart file for DOS
Analysis: DOS, PDOS, COOP
End
Restart t21.NaCl &
DOS
End
Accuracy 3.5
Kspace 3
SCF
Iterations 15
End
Units
Length Angstrom
End
Lattice
0 2.75 2.75
2.75 0 2.75
2.75 2.75 0
End
DOS
File NaCl.dos
Energies 1000
Min -0.5
Max 0.5
End
GROSSPOPULATIONS
FRAG 1
FRAG 2
SUM
1 0
2 0
ENDSUM
End
OVERLAPPOPULATIONS
LEFT
FRAG 1
RIGHT
FRAG 2
LEFT
1 0
1 1
RIGHT
2 0
2 1
End
Atoms NA
0
End
Coordinates Natural
Atoms Cl
.5 .5 .5
End
AtomType Na
Dirac Na
4 1
Radial 2000
RMin 1E-4
RMax 60
VALENCE
1 0
2 0
2 1
3 0 1.0
SubEnd
FitFunctions
1 0 18.9
2 0 30.3
2 0 15.5
3 0 14.9
3 0 8.9
4 0 7.8
4 0 5.1
4 0 3.3
5 0 2.8
5 0 1.9
5 0 1.3
2 1 14.3
3 1 9.9
4 1 6.7
4 1 3.4
5 1 2.4
5 1 1.3
3 2 10.5
4 2 5.4
5 2 3.0
5 2 1.3
4 3 5.8
5 3 1.7
5 4 2.0
SubEnd
End
AtomType Cl
Dirac Cl
5 3
VALENCE
1 0
2 0
2 1
3 0
3 1 5.0
SubEnd
FitFunctions
1 0 29.1
2 0 49.5
2 0 26.1
3 0 25.8
3 0 15.8
4 0 14.2
4 0 9.4
4 0 6.2
5 0 5.4
5 0 3.8
5 0 2.6
2 1 21.2
3 1 16.5
4 1 12.4
4 1 6.8
5 1 5.1
5 1 3.1
3 2 16.6
4 2 9.4
5 2 5.5
5 2 2.6
4 3 8.7
5 3 3.3
5 4 4.0
SubEnd
End
End Input
eor
Finally, we copy the contents of the DOS result file to standard output
echo Contents of DOS file cat NaCl.dos
Sample directory: band/CnHn/
This example illustrates how a one-dimensional periodic system can be treated by specifying only one lattice vector. It further shows how variables can be defined with the DEFINE keyword. The rest more or less speaks for itself. The Kspace integration is taken very accurate, whereas real space integration (ACCURACY keyword) is not so accurate.
Here and in the following BAND examples, we will leave out some space consuming parts of the input file which have been discussed already. Please check the actual input files if you wish to repeat one of the calculations.
$ADFBIN/band << eor Title Polymer Comment Technical Quadratic k space integration (1D) Low real space integration accuracy Definitions of variables Features Lattice : 1D, polymer Unit cell : 4 atoms Basis : NO+STO w/ core End Kspace 5 Accuracy 3 Units Length Angstrom Angle Radian End Define dCCd=1.3386 dCCs=1.4510 dCH=1.0770 aCCC=124.5/180*pi arC=aCCC-pi/2 aCCH=119.2/180*pi ! double bonded CC arH=aCCH-pi/2 End Lattice dCCd+sin(arC)*dCCs cos(arC)*dCCs 0.0 End Atoms C dCCd/2 0.0 0.0 -dCCd/2 0.0 0.0 End Atoms H dCCd/2+sin(arH)*dCH -cos(arH)*dCH 0.0 -dCCd/2-sin(arH)*dCH cos(arH)*dCH 0.0 End
A larger unit cell can of course be specified as well. In the second part of the example a supercell of 5 units is used. Another new feature introduced in this example is the TAILS keyword, which similar to ADF implies that distance cut-offs are applied to make the calculation cheaper. At the moment no big gains are yet to be expected from this, but this situation is expected to change in future versions of the code.
in the BASIS key. This subkey is actually mandatory at the moment if the TAILS keyword is used.
$ADFBIN/band << eor Title Polymer with big unit cell (5 units) Comment Technical Low quadratic k space integration (1D) Low real space integration accuracy Definitions of variables Features Lattice : 1D, polymer Unit cell : 4 atoms Basis : NO+STO w/ core End Kspace 3 Accuracy 3 Units Length Angstrom Angle Radian End Tails Bas=1E-2 Define dCCd=1.3386 dCCs=1.4510 dCH=1.0770 aCCC=124.5/180*pi arC=aCCC-pi/2 aCCH=119.2/180*pi ! double bonded CC arH=aCCH-pi/2 Latx(nlatt)=nlatt*(dCCd+sin(arC)*dCCs) Laty(nlatt)=nlatt*(cos(arC)*dCCs) Laty(nlatt)=nlatt*(cos(arC)*dCCs) End Lattice Latx(5) Laty(5) 0.0 End Atoms C dCCd/2 0.0 0.0 -dCCd/2 0.0 0.0 dCCd/2+Latx(1) Laty(1) 0.0 -dCCd/2+Latx(1) Laty(1) 0.0 dCCd/2+Latx(2) Laty(2) 0.0 -dCCd/2+Latx(2) Laty(2) 0.0 dCCd/2+Latx(3) Laty(3) 0.0 -dCCd/2+Latx(3) Laty(3) 0.0 dCCd/2+Latx(4) Laty(4) 0.0 -dCCd/2+Latx(4) Laty(4) 0.0 End Atoms H dCCd/2+sin(arH)*dCH -cos(arH)*dCH 0.0 -dCCd/2-sin(arH)*dCH cos(arH)*dCH 0.0 dCCd/2+sin(arH)*dCH+Latx(1.0) -cos(arH)*dCH+Laty(1.0) 0.0 -dCCd/2-sin(arH)*dCH+Latx(1.0) cos(arH)*dCH+Laty(1.0) 0.0 dCCd/2+sin(arH)*dCH+Latx(2.0) -cos(arH)*dCH+Laty(2.0) 0.0 -dCCd/2-sin(arH)*dCH+Latx(2.0) cos(arH)*dCH+Laty(2.0) 0.0 dCCd/2+sin(arH)*dCH+Latx(3.0) -cos(arH)*dCH+Laty(3.0) 0.0 -dCCd/2-sin(arH)*dCH+Latx(3.0) cos(arH)*dCH+Laty(3.0) 0.0 dCCd/2+sin(arH)*dCH+Latx(4.0) -cos(arH)*dCH+Laty(4.0) 0.0 -dCCd/2-sin(arH)*dCH+Latx(4.0) cos(arH)*dCH+Laty(4.0) 0.0 End
Sample directory: band/Pd/
The main things to note are the specification of a scalar relativistic ZORA calculation, and the fact that symmetry information will be printed on output.
This example also shows incidentally how to override optimization memory parameters KGRPX and CPVector.
$ADFBIN/band << eor
Title Pd bulk
Comment
Technical
Hybrid k space integration (3D)
Reasonable real space integration accuracy
Definitions of variables
Features
Lattice : 3D
Unit cell : 1 atom
Basis : NO+STO w/ core
Options : spin restricted, scalar relativistic,
numerical fit functions
Full symmetry
End
KGRPX 2
CPVector 256
Kspace 5
Accuracy 4.0
Print SYMMETRY
Relativistic ZORA
Define
halflatt=7.35/2
End
Lattice :: FCC
0 halflatt halflatt
halflatt 0 halflatt
halflatt halflatt 0
End
Atoms Pd
0.0
End
...
END INPUT
eor
Sample directory: band/Ne/
$ADFBIN/band << eor Title Ne matrix 4.2 K Comment Technical Hybrid k space integration (3D) (high) Reasonable real space integration accuracy Definitions of variables Features Lattice : 3D Unit cell : 1 atom Basis : NO+STO w/o core End Kspace 9 Accuracy 4.0 Units Length Angstrom End Define halflatt=4.43/2 End Lattice :: FCC 0 halflatt halflatt halflatt 0 halflatt halflatt halflatt 0 End Atoms Ne 0.0 0.0 0.0 End AtomType Ne Dirac Ne 3 0 Valence 1S 2S 2P SubEnd BasisFunctions 1S 12.45 2S 3.65 2S 1.55 2P 1.45 2P 5.50 3D 2.0 4F 3.0 SubEnd FitFunctions 3S 2.9000 3S 4.0900 3S 5.7700 3S 8.1400 2S 8.3300 2S 12.5200 2S 18.8100 1S 17.2000 3P 3.1000 3P 4.8000 3P 7.4200 2P 8.4100 2P 14.1000 3D 2.9000 3D 5.6700 3D 11.1000 4F 3.9500 4F 8.0000 5G 5.0000 SubEnd End END INPUT eor
Sample directory: band/H_ref/
This example consists of several atomic energy calculations:
Only the first run is given here:
$ADFBIN/band << eor Title Formation energy of the H-atom w.r.t. spherical atom Comment Technical Quadratic K space integration High real space integration accuracy Features Lattice : 2D, quasi atom Unit cell : 1 atom, 1x1 Basis : NO+STO End Kspace 5 Accuracy 5.0 Convergence Criterion 1E-6 End Define far=20.0 End Lattice far 0.0 0.0 0.0 far 0.0 End Atoms H 0.0 0.0 0.0 End AtomType H Dirac H 1 0 VALENCE 1S 1 SubEnd BasisFunctions 1S 1.58 2P 1.0 SubEnd FitFunctions 1S 3.16 1S 2.09 1S 1.38 2S 1.50 2P 4.00 2P 2.65 2P 1.75 3D 4.00 3D 2.50 4F 3.00 5G 4.00 SubEnd End END INPUT eor
Sample directory: band/H_on_Pt/
This example is in many ways similar to the Pt slab example. We refer to the explanations in that example for details. Suffice it to say here that the convergence criteria have again been chosen such that also difficult slab calculations have a good chance to converge. Further, this calculation is once again at the spin-orbit relativistic level. The geometry is such that two layers of Pt are covered by a hydrogen layer.
$ADFBIN/band << eor Title Hydrogen on platinum Comment Technical Low quadratic K space integration Low real space integration accuracy Features Lattice : 2D Unit cell : 4 atoms, 1x1 Basis : NO+STO Options : Spinorbit ZORA End SCF Mixing 0.1 Iterations 100 End Convergence Degenerate default Criterion 1e-6 End DIIS NCycleDamp 0 DiMix 0.15 End KSpace 3 Accuracy 3 Relativistic ZORA SPIN Dependency Basis=1E-8 Define latt=7.41 lvec=latt/SQRT(2.0) ysh=lvec/SQRT(3.0) dlay=latt/SQRT(3.0) height=3.0 End Lattice SQRT(3.0)*lvec/2.0 0.5*lvec SQRT(3.0)*lvec/2.0 -0.5*lvec End Atoms Pt 0 0 0 :: layer 1 -ysh 0.0 -dlay :: layer 2 End Atoms H 0.0 0.0 height End ... END INPUT eor
Sample directory: band/BNForce/
This example shows how to calculate the gradient of the energy with respect to atomic displacements, using the GRADIENTS key block.
$ADFBIN/band << eor
Title BN zincblende structure (force calculation)
comment
Calculate Atomic Forces (energy gradient)
Frozen core approximation
TZ2P basis
A/K -> 5/3
2 atoms per unit cell
end
dependency basis=1e-8
Kspace 3
Accuracy 5.0
Units
Length Angstrom
End
Define
a = 3.60
End
GRADIENTS
END
Lattice
0.0 0.5*a 0.5*a
0.5*a 0.0 0.5*a
0.5*a 0.5*a 0.0
End
Atoms B
0.0 0.0 0.0
end
Atoms N
0.2404*a 0.2404*a 0.2404*a
end
BasisDefaults
BasisType TZ2P
Core Large
End
END INPUT
eor
In the output you will find the result
FINAL GRADIENTS
1 B 0.019676 0.019676 0.019676
2 N -0.019677 -0.019677 -0.019677
Sample directory: band/H2BulkGeo/
This example shows how to optimize the geometry, using the GEOOPT key block, including a restart of a previous optimization. It consists of two runs (The example is designed for speed and not for accuracy.)
$ADFBIN/band << eor
Title H2 bulk geometry optimization
Comment
Optimized for speed / Accuracy low
Basis very small
Demonstrates the GeoOpt block
SCF iterations limited to make it faster
End
KSpace 2
Accuracy 3.1
screening
RMadel 9
end
Dependency Basis=1e-8
GeoOpt
Iterations 5
Converge Grad=1e-6
RestartSCF true
End
SCF
Mixing 0.2
Iterations 2
End
Convergence
Degenerate default
End
DIIS
NCycleDamp 0
DiMix 0.5
End
Units
Angle Radian
End
Define
aaa=1
End
ATOMS H
0
aaa
End
Lattice
5.0*aaa 0 0
0 5.0*aaa 0
0 0 5.0*aaa
End
AtomType H
DIRAC H
1 0
VALENCE
1S 1
SubEnd
BasisFunctions
1S 0.69
2P 1.25
SubEnd
FitFunctions
1S 3.16
1S 2.09
1S 1.38
2S 1.50
2P 4.00
2P 2.65
2P 1.75
3D 4.00
3D 2.50
:: 4F 3.00
:: 5G 4.00
SubEnd
End
END INPUT
eor
mv RUNKF H2BulkGeo.runkf
rm Points
Note that the RUNKF file is saved. In the next run we use the result file to continue the geometry optimization
# ----------------------------- second run --------------------------
$ADFBIN/band << eor
Title H2 bulk geometry optimization
Comment
Restarting the geometry optimization
Optimized for speed / Accuracy low
Basis very small
Demonstrates the GeoOpt block
SCF iterations limited to make it faster
End
KSpace 2
Accuracy 3.1
screening
RMadel 9
end
Dependency Basis=1e-8
Restart H2BulkGeo.runkf&
GeometryOptimization
end
GeoOpt
Iterations 5
Converge Grad=1e-6
RestartSCF true
End
SCF
Mixing 0.2
:: Iterations 2
End
Convergence
Degenerate default
End
DIIS
NCycleDamp 0
DiMix 0.5
End
Units
Angle Radian
End
Define
aaa=1
End
ATOMS H
0
aaa
End
Lattice
5.0*aaa 0 0
0 5.0*aaa 0
0 0 5.0*aaa
End
AtomType H
DIRAC H
1 0
VALENCE
1S 1
SubEnd
BasisFunctions
1S 0.69
2P 1.25
SubEnd
FitFunctions
1S 3.16
1S 2.09
1S 1.38
2S 1.50
2P 4.00
2P 2.65
2P 1.75
3D 4.00
3D 2.50
:: 4F 3.00
:: 5G 4.00
SubEnd
End
END INPUT
eor
Sample directory: band/CO_MetaGGA_Force/
This example shows how to calculate the gradient of the energy with respect to atomic displacements for a metaGGA functional, using the GRADIENTS and XC key block.
$ADFBIN/band << eor Title CO chain KSpace 3 accuracy 5 DIIS dimix 0.2 ncycledamp 0 end scf mixing 0.3 end xc metaGGA scf force tpss end gradients end define ha=6.60/2 end Lattice 6.0 0 0 end atoms C 0.0 end atoms O 2.20 end BasisDefaults BasisType DZ Core Large end end input eor
Sample directory: band/H2SlabFreqTS/
First run: H2 slab frequency calculation near minimum. Second run: transition state search, using as restart the results of the first run.
$ADFBIN/band << eor
Title H2 slab frequency calculation near minimum
Comment
Optimized for speed / Accuracy low
Basis very small
Demonstrates the frequencies runtype
End
KSpace 2
Accuracy 4
screening
RMadel 9
end
Dependency Basis=1e-8
RunType
Frequencies
End
GeoOpt
RestartSCF true
End
SCF
Mixing 0.2
End
Convergence
Degenerate default
End
DIIS
NCycleDamp 0
DiMix 0.5
End
Units
Length Angstrom
Angle Radian
End
ATOMS H
:: TS coords
:: 0.408818 -0.059284 0.374229
:: 0.305357 0.059284 -0.374229
:: Local minimum coords:
-0.4 0 0.1
0.4 0 -0.1
End
Lattice
2.645886 0 0
0 2.645886 0
End
AtomType H
DIRAC H
1 0
VALENCE
1S 1
SubEnd
BasisFunctions
1S 0.69
2P 1.25
SubEnd
FitFunctions
1S 3.16
1S 2.09
1S 1.38
2S 1.50
2P 4.00
2P 2.65
2P 1.75
3D 4.00
3D 2.50
SubEnd
End
END INPUT
eor
mv RUNKF H2SlabFreq.runkf
rm Points
Here comes the second run where the Hessian (from the frequency run) is used to find the transtition state
$ADFBIN/band << eor
Title H2 slab TS search
Comment
Optimized for speed / Accuracy low
Basis very small
Demonstrates the GeoOpt block
End
KSpace 2
Accuracy 5.0
screening
RMadel 9
end
Dependency Basis=1e-8
RunType
TS
End
GeoOpt
Iterations 50
Converge Grad=1e-3
RestartSCF true
InitialHessian H2SlabFreq.runkf
End
SCF
Mixing 0.2
End
Convergence
Degenerate default
End
DIIS
NCycleDamp 0
DiMix 0.5
End
Units
Length Angstrom
Angle Radian
End
ATOMS H
-0.4 0 0.1
0.4 0 -0.1
End
Lattice
2.645886 0 0
0 2.645886 0
End
AtomType H
DIRAC H
1 0
VALENCE
1S 1
SubEnd
BasisFunctions
1S 0.69
2P 1.25
SubEnd
FitFunctions
1S 3.16
1S 2.09
1S 1.38
2S 1.50
2P 4.00
2P 2.65
2P 1.75
3D 4.00
3D 2.50
SubEnd
End
END INPUT
eor
mv RUNKF H2SlabTS.runkf
rm Points
Sample directory: band/Silicon/
The time-dependent DFT functionality is an important functionality. It enables you to calculate frequency-dependent dielectric functions for 1-dimensional and 3-dimensional periodic systems. In the present example, a standard geometry for bulk Silicon is given. The accuracy and kspace variables can keep their normal values. The important part in this example is of course the RESPONSE block. It specifies that 7 frequencies should be treated, with an even spacing between 0.0 hartree and 0.25 hartree. In this example scalar ZORA relativistic effects are switched on with the isz line in the RESPONSE block.
$ADFBIN/band << eor TITLE Silicon ACCURACY 5 KSPACE 2 DEPENDENCY BASIS 1e-10 UNITS LENGTH ANGSTROM END RESPONSE nfreq 7 strtfr 0d0 endfr 25d-2 isz 1 END DEFINE AAA=5.43 HA=AAA/2 END LATTICE 0 HA HA HA 0 HA HA HA 0 END ATOMS Si 0.0 0.0 0.0 HA/2 HA/2 HA/2 END ... END INPUT eor
For Silicon the real and imaginary parts of the dielectric function: epsilon(omega) = 1 + 4 Pi Chi(omega) are calculated.
In the output file, the results will look something like the fragment below. The output specifies for which frequency the dielectric function is determined, and then proceeds to print the values for the 3x3 tensors.
The real and imaginary parts are printed separately. At this frequency, the imaginary part is still zero. Because of the high symmetry of the system, the real part is a constant times the unit matrix except for numerical noise.
** Frequency ** 0.833333E-01 au 2.26756 eV ** ** Start the SCF procedure ** ***** Real ****** ** Chi_jj X ** -12.8363 0.142802E-18 0.547977E-17 ** Chi_jj Y ** 0.202883E-17 -12.8363 0.121052E-17 ** Chi_jj Z ** 0.124042E-16 0.215311E-17 -12.8363 ***** Imag ****** ** Chi_jj X ** 0.000000E+00 0.000000E+00 0.000000E+00 ** Chi_jj Y ** 0.000000E+00 0.000000E+00 0.000000E+00 ** Chi_jj Z ** 0.000000E+00 0.000000E+00 0.000000E+00 *****************
After each frequency has been treated, the results are summarized in a table, for each component separately. For the xx-component, this looks as in the table below. The frequency/energy is again printed in two different units, the Dielectric Function is printed in a.u. The values for Chi, which are trivially related to those printed here, are summarized in a separate table.
=================================================================
== Frequency === Dielectric Function ==
== a.u. == e.V. === Re == Im ==
============XX-dir===============================================
0.416667E-01 1.13378 16.1119 0.000000E+00
0.833333E-01 2.26756 23.7904 0.000000E+00
0.125000 3.40134 15.8529 35.8574
0.166667 4.53512 -3.49949 20.2221
0.208333 5.66890 -6.60897 12.3661
0.250000 6.80268 -6.42943 6.87957
============YY-dir===============================================
0.416667E-01 1.13378 16.1119 0.000000E+00
0.833333E-01 2.26756 23.7904 0.000000E+00
0.125000 3.40134 15.8529 35.8574
0.166667 4.53512 -3.49949 20.2221
0.208333 5.66890 -6.60897 12.3661
0.250000 6.80268 -6.42943 6.87957
============ZZ-dir===============================================
0.416667E-01 1.13378 16.1119 0.000000E+00
0.833333E-01 2.26756 23.7904 0.000000E+00
0.125000 3.40134 15.8529 35.8574
0.166667 4.53512 -3.49949 20.2221
0.208333 5.66890 -6.60897 12.3661
0.250000 6.80268 -6.42943 6.87957
Results of the test calculation (red/blue) are plotted in next Figure together with experimental data (yellow/green). The results for the seven specified frequencies are given. It should be obvious that more frequencies are needed (resulting in longer run times) to obtain a smooth curve in which peaks cannot be missed because of too coarse interpolation.
Sample directory: band/He_crystal/
$ADFBIN/band << eor TITLE DIAMOND ACCURACY 5 KSPACE 2 Dependency Basis=1.e-6 UNITS LENGTH ANGSTROM END RESPONSE nfreq 7 strtfr 0d0 endfr 7d-1 END DEFINE AAA=3.57 HA=AAA/2 END LATTICE 0 HA HA HA 0 HA HA HA 0 END ATOMS C 0.0 0.0 0.0 HA/2 HA/2 HA/2 END BasisDefaults BasisType DZ End END INPUT eor
Sample directory: band/Diamond/
$ADFBIN/band << eor TITLE DIAMOND ACCURACY 5 KSPACE 2 Dependency Basis=1.e-6 UNITS LENGTH ANGSTROM END RESPONSE nfreq 7 strtfr 0d0 endfr 7d-1 END DEFINE AAA=3.57 HA=AAA/2 END LATTICE 0 HA HA HA 0 HA HA HA 0 END ATOMS C 0.0 0.0 0.0 HA/2 HA/2 HA/2 END BasisDefaults BasisType DZ End END INPUT eor
Sample directory: band/H_chain
The input for a one-dimensional system is no different from that for a three-dimensional system. No tests have been performed on two-dimensional systems yet in combination with the TDDFT code. It is therefore not expected to work. Besides the number of frequencies, and the beginning and end frequency of the frequency range, the RESPONSE block also contains stricter than default convergence criteria (cnvi and cnvj) for the fit coefficients describing the density change.
$ADFBIN/band << eor Title H2-chain ACCURACY 5 KSPACE 3 DEPENDENCY BASIS 1e-10 RESPONSE nfreq 10 strtfr 0d0 endfr 15d-3 cnvi 1d-5 cnvj 1d-5 END DEFINE HX = 4.5 END LATTICE HX END ATOMS H 1.0 0.001 0.0 -1.0 -0.001 0.0 END ... END INPUT eor
The output for this calculation will give rise to a table like the following one:
=================================================================
== Frequency === Polarizability(xx) ==
== a.u. == e.V. === Re == Im ==
=================================================================
0.166667E-02 0.453512E-01 127.119 0.00000
0.333333E-02 0.907023E-01 127.201 0.00000
0.500000E-02 0.136054 127.338 0.00000
0.666667E-02 0.181405 127.530 0.00000
0.833333E-02 0.226756 127.778 0.00000
0.100000E-01 0.272107 128.083 0.00000
0.116667E-01 0.317458 128.446 0.00000
0.133333E-01 0.362809 128.868 0.00000
0.150000E-01 0.408161 129.351 0.00000
=================================================================
== Frequency === Chi_JJ (xx) ==
== a.u. == e.V. === Re == Im ==
=================================================================
0.00000 0.00000 -2.74118 0.00000
0.166667E-02 0.453512E-01 -2.74153 0.00000
0.333333E-02 0.907023E-01 -2.74259 0.00000
0.500000E-02 0.136054 -2.74436 0.00000
0.666667E-02 0.181405 -2.74685 0.00000
0.833333E-02 0.226756 -2.75005 0.00000
0.100000E-01 0.272107 -2.75399 0.00000
0.116667E-01 0.317458 -2.75866 0.00000
0.133333E-01 0.362809 -2.76409 0.00000
0.150000E-01 0.408161 -2.77028 0.00000
=================================================================
Sample directory: band/Cu_resp_NR_ALDA/
This example shows how to calculate the dielectric function for a metal. The RESPONSE keyblock is as usual. In the manual it is explained that the KSPACE parameter should be chosen with care.
$ADFBIN/band << eor Title: response of Cu within ALDA Comment Technical Hybrid k space integration (3D) Reasonable real space integration accuracy Definitions of variables Features Lattice : 3D Unit cell : 1 atom Basis : NO+STO w/ core (DZ/Cu.2p) Options : ALDA End kspace 7 Accuracy 4.0 RESPONSE nfreq 2 strtfr 0.10d0 endfr 0.25d0 END Units Length Angstrom End Define halflatt=3.61/2 End Lattice :: FCC 0 halflatt halflatt halflatt 0 halflatt halflatt halflatt 0 End Atoms Cu 0.0 End BasisDefaults BasisType DZ Core Large End END INPUT eor
Sample directory: band/TiF3a/
Sample calculation for an ESR A-tensor calculation. The calculation should be spin unrestricted and have the ATENSOR keyword.
$ADFBIN/band << eor
Title tif3
comment
end
Kspace 1
Accuracy 5.0
Units
Length Angstrom
End
Define
a = 25
RTIF = 1.78
RY = RTIF*SQRT(3)/2
End
Unrestricted
ATENSOR
END
Lattice
a 0.0 0.0
0.0 a 0.0
0.0 0.0 a
End
Atoms Ti
0.0 0.0 0.0
end
Atoms F
RTIF 0 0
-RTIF/2 RY 0
-RTIF/2 -RY 0
end
AtomType Ti
DIRAC Ti
7 0
VALENCE
1S
2S
2P
3S
3P
4S
3D 2
SubEnd
BASISFUNCTIONS
1S 26.300000
2S 5.600000
2P 12.000000
3S 4.650000
3P 5.400000
3P 2.300000
3D 4.950000
3D 1.040000
4S 1.900000
4S 0.800000
4P 1.200000
4F 2.300000
SubEnd
FITFUNCTIONS
1S 52.600000
2S 58.320000
2S 39.240000
3S 38.750000
3S 27.730000
4S 26.240000
4S 19.530000
4S 14.540000
5S 13.490000
5S 10.330000
5S 7.910000
6S 7.260000
6S 5.680000
6S 4.450000
6S 3.480000
7S 3.170000
7S 2.530000
7S 2.010000
7S 1.600000
2P 38.300000
3P 25.820000
4P 17.610000
5P 12.200000
5P 7.070000
6P 4.970000
6P 3.010000
7P 2.150000
3D 28.600000
4D 19.530000
5D 13.550000
5D 7.860000
6D 5.540000
6D 3.360000
7D 2.400000
4F 13.900000
5F 7.920000
5F 3.800000
6F 2.240000
5G 8.400000
5G 4.390000
5G 2.290000
SubEnd
End
AtomType F
DIRAC F
3 0
VALENCE
1S
2S
2P 5
SubEnd
BASISFUNCTIONS
1S 10.880000
2S 3.240000
2S 0.740000
2P 4.540000
2P 1.240000
3D 2.000000
4F 3.000000
SubEnd
FITFUNCTIONS
1S 21.760000
2S 24.390000
2S 16.560000
2S 11.240000
2S 7.630000
3S 7.600000
3S 5.480000
3S 3.950000
3S 2.850000
3S 2.050000
3S 1.480000
2P 14.400000
2P 7.530000
3P 5.900000
3P 3.420000
3P 1.980000
3D 9.700000
3D 6.160000
3D 3.910000
3D 2.480000
4F 6.500000
4F 3.750000
5G 4.500000
SubEnd
End
END INPUT
eor
Sample directory: band/TiF3g/
Sample calculation for an ESR g-tensor calculation. This makes only sense for a relativistic spin orbit calculation.
$ADFBIN/band << eor
Title tif3
comment
1d structure
end
Kspace 1
Accuracy 5.0
Relativistic ZORA SPIN
esr
end
Units
Length Angstrom
End
Define
a = 25
RTIF = 1.78
RY = RTIF*SQRT(3)/2
End
Lattice
a
End
Atoms Ti
0.0 0.0 0.0
end
Atoms F
0 RTIF 0
-RY -RTIF/2 0
RY -RTIF/2 0
end
AtomType Ti
DIRAC Ti
7 0
VALENCE
1S
2S
2P
3S
3P
4S
3D 2
SubEnd
BASISFUNCTIONS
1S 26.300000
2S 5.600000
2P 12.000000
3S 4.650000
3P 2.450000
3D 4.950000
3D 1.040000
4S 1.650000
4P 1.200000
SubEnd
FITFUNCTIONS
1S 52.600000
2S 58.320000
2S 39.240000
3S 38.750000
3S 27.730000
4S 26.240000
4S 19.530000
4S 14.540000
5S 13.490000
5S 10.330000
5S 7.910000
6S 7.260000
6S 5.680000
6S 4.450000
6S 3.480000
7S 3.170000
7S 2.530000
7S 2.010000
7S 1.600000
2P 38.300000
3P 25.820000
4P 17.610000
5P 12.200000
5P 7.070000
6P 4.970000
6P 3.010000
7P 2.150000
3D 28.600000
4D 19.530000
5D 13.550000
5D 7.860000
6D 5.540000
6D 3.360000
7D 2.400000
4F 13.900000
5F 7.920000
5F 3.800000
6F 2.240000
5G 8.400000
5G 4.390000
5G 2.290000
SubEnd
End
AtomType F
DIRAC F
3 0
VALENCE
1S
2S
2P 5
SubEnd
BASISFUNCTIONS
1S 10.880000
2S 3.220000
2P 3.520000
SubEnd
FITFUNCTIONS
1S 21.760000
2S 24.390000
2S 16.560000
2S 11.240000
2S 7.630000
3S 7.600000
3S 5.480000
3S 3.950000
3S 2.850000
3S 2.050000
3S 1.480000
2P 14.400000
2P 7.530000
3P 5.900000
3P 3.420000
3P 1.980000
3D 9.700000
3D 6.160000
3D 3.910000
3D 2.480000
4F 6.500000
4F 3.750000
5G 4.500000
SubEnd
End
END INPUT
eor
Sample directory: band/PE-NMR/
With the NMR keyblock you can specify for which atom you want the shielding tensor.
The NMR option is not implemented with the frozen core approximation, hence we set core to NONE.
$ADFBIN/band << eor TITLE PE NMR nmratom=1 ms0=1. end Units Length Angstrom Angle Degree end xc GGA Always Becke Perdew end DEPENDENCY BASIS 1e-10 KSPACE 3 Accuracy 5 define aCCC = 112.777 aHCH = 109.47 alf2 = aCCC/2 bet = aHCH/2. rcc=1.533 rch=1.09 z3 = rch*cos(bet) y3 = rch*sin(bet) T = 2*rcc*sin(alf2) C2x = rcc*sin(alf2) C2z = -rcc*cos(alf2) end Lattice T 0. 0. end Atoms C 0. 0. 0. C2x 0. C2z end Atoms H 0. y3 z3 0. -y3 z3 C2x y3 C2z-z3 C2x -y3 C2z-z3 end BasisDefaults BasisType TZ2P Core NONE End end input eor
Sample directory: band/SnO_EFG/
The calculation of the electric field gradient is invoked by the EFG keyblock.
Because Sn is quite a heavy atom we use the scalar relativistic option.
$ADFBIN/band << eor Title SnO EFG kspace 3 accuracy 5 relativistic ZORA programmer FlioSinglePrecision false end EFG End Units Length Angstrom End coordinates natural dependency basis=1e-8 diis ncycledamp 0 dimix 0.2 end define aaa=3.8029 ccc=4.8382 vvv=0.2369 end lattice aaa 0 0 0 aaa 0 0 0 ccc end atoms O 0 0 0 0.5 0.5 0 end atoms Sn 0 0.5 vvv 0.5 0 -vvv end BasisDefaults BasisType DZ Core none End end input eor
Sample directory: band/Frags_COCu/
This example illustrates the usage of fragments in a BAND
calculation for analysis purposes. It takes two runs to do the DOS analysis in a fragment basis, and an extra two runs to get the deformation density with respect
to the fragment densities.
The setup involves first the computation of the free CO overlayer, which is to be absorbed on a Cu surface. To suppress (most of the) interactions between the CO molecules, i.e. to effectively get the molecular CO, the KSpace parameter is set to 1 (= no dispersion), and the lattice parameters are set so large that the CO molecules are far apart. The standard result file RUNKF is saved under the name 'CO.runkf'.
# ----------------------------- CO molecule -------------------------- $ADFBIN/band << eor Title The CO fragment Comment Technical Zero order k space integration Good real space integration accuracy Definitions of variables Features Lattice : 2D, large lattice vectors Unit cell : 2 atoms, 1x1, quasi molecular Basis : NO+STO w/ core End PRINT EIGENS Kspace 1 ! neglect dispersion Accuracy 4 Define bond=2.18 far=25 End Lattice ! CO molecules far apart far 0.0 0.0 far End Atoms C 0 0 0 End Atoms O 0 0 bond End BasisDefaults BasisType DZ Core Large End End Input eor mv RUNKF CO.runkf
Now we can use the result file to do a DOS analysis for CO on a copper surface treating the molecule as a fragment. With Fragment%Labels we assign names to the different symmetry orbitals. The Density-of-States analysis details are given with the keys DOS (energy grid, result file with DOS data) and, optionally, GrossPopulations and OverlapPopulations.
# ----------------------------- CO + Cu slab --------------------------
$ADFBIN/band << eor
Title Cu slab with CO adsorbed
Comment
Technical
Quadratic K space integration (low)
Good real space integration accuracy
Definitions of variables
Features
Lattice : 2D
Unit cell : 3 atoms, 1x1
Basis : NO+STO w/ core
Options : Molecular fragment
Analysis: DOS, PDOS, COOP
End
KSpace 3
Accuracy 4
! fragment specification
Fragment CO.runkf
1 1
2 2
Labels ! let us give them some labels
2Sigma
2Sigma*
1Pi_x
1Pi_y
3Sigma
1Pi_x*
1Pi_y*
3Sigma*
SubEnd
End
! use fragment basis in dos
DosBas
Fragment 1
End
DOS ! Analysis
File pdos.CO_Cu
Energies 500
Min -0.750
Max 0.300
End
GrossPopulations
3 2 ! All metal d states
Sum ! ALl metal sp states
3 0
3 1
EndSum
Frag 1 ! All CO states
Sum ! CO 1pi
FragFun 1 5
FragFun 1 6
EndSum
FragFun 1 7 ! CO 5-sigma
End
OverlapPopulations
Left ! Metal d with CO
3 2
Right
Frag 1
End
Define
dist=3.44
bond=2.18
End
Lattice
4.822 0.0
0.0 4.822
End
Atoms C
0 0 dist
End
Atoms O
0 0 dist+bond
End
Atoms CU
0.0 0.0 0.0
End
BasisDefaults
BasisType DZ
Core Large
End
End Input
eor
After this run we copy the computed DOS data from the DOS result file to standard output. We also save the restart file for later use.
echo Contents of DOS file cat pdos.CO_Cu mv RUNKF COCu.runkf
Next we want to know the deformation density with respect to the two fragments: 1) The CO molecule and 2) the bare Cu surface. We haven't done the bare Cu surface yet, so that is what happens next.
# ----------------------------- Cu slab -------------------------- $ADFBIN/band << eor Title Cu slab Comment Technical Quadratic K space integration (low) Good real space integration accuracy Definitions of variables Features Lattice : 2D Unit cell : 3 atoms, 1x1 Basis : NO+STO w/ core Options : End Kspace 3 Accuracy 4 Define dist=3.44 bond=2.18 End Lattice 4.822 0.0 0.0 4.822 End Atoms CU 0.0 0.0 0.0 End BasisDefaults BasisType DZ Core Large End End Input eor mv RUNKF Cu.runkf
Now we are all set to do our final calculation. We have the two fragment files CO.runkf and Cu.runkf, and the restart file COCu.runkf. Next we want to know the deformation density with respect to the two fragments: 1) The CO molecule and 2) the bare Cu surface. The visualization options like OrbitalPlot and Densityplot require a regular set of points (a grid). Here is how it works
# ----------------------------- CO + Cu slab restart -------------------------- $ADFBIN/band -n 1 << eor Title Cu slab with CO adsorbed (restart density plot) Kspace 3 Accuracy 4 Restart COCu.runkf & DensityPlot End Grid Type Coarse End DensityPlot scf End ! fragment specification Fragment CO.runkf 1 1 2 2 End Fragment Cu.runkf 1 3 End Define dist=3.44 bond=2.18 End Lattice 4.822 0.0 0.0 4.822 End Atoms C 0 0 dist End Atoms O 0 0 dist+bond End Atoms CU 0.0 0.0 0.0 End BasisDefaults BasisType DZ Core Large End End Input eor
This particular restart options does not work in parallel, hence the "-n 1" on the first line.The result of the last run is a file named TAPE41. Normally you would save that to COCu.t41
mv TAPE41 COCu.t41
and view it with adfview. On the TAPE41 file are now three fields shown in adfview as
being the deformation density of CO+Cu with
respect to the atoms, and the same for the two fragments CO and the Cu slab. In adfview you can add the fields of the two fragments, and then create another
field that holds the difference.
Sample directory: band/Cu_slab/
A two-dimensional infinite (periodic boundary conditions) slab calculation is performed for Cu. The dimensionality is simply defined by the number of records in the Lattice data block. In a 2-dimensional calculation the lattice vectors are put in the xy-plane. (In a one-dimensional calculation (polymer), the lattice vector is taken along the x-axis in the program.)
Slab calculations for metals frequently suffer from SCF convergence problems, as a result of the open valence band(s). To help the program converge it is often useful or even necessary to use some special features, such as the ElectronicTemperature key. This particular key requires a numerical value (0.025 in the example) and implies that a finite-temperature electronic distribution is used, rather than a sharp cut-off at the Fermi level. The numerical value is the applied energy width, in hartree units.
The so-modified electronic distribution also affects the energy. The 'true' zero-T energy is computed, approximately, by an interpolation formula. The interpolation is not very accurate and one should try to use as small as possible values for the ElectronicTemperature key so as to avoid increasing uncertainty in the results. The program prints, in the energy section of the output file, the finite-T correction term that has been applied through the interpolation formula. This gives at least an indication of any remaining uncorrected deviation of the outcome from a true zero-T calculation.
In the second run the runkf file of the first run is used to do an orbital plot restart. Normally you would rename the resulting TAPE41 to "myslab.t41" and watch the orbitals with adfview.
# ----------------------------- first run -------------------------- $ADFBIN/band << eor Title Cu slab Comment Technical Quadratic K space integration Good real space integration accuracy Features Lattice : 2D Unit cell : 1 atom, 1x1 Basis : NO+STO w/ core Options : ElectronicTemperature (temperature effect) End Kspace 5 Accuracy 4 Convergence ElectronicTemperature 0.025 End Lattice 4.822 0.0 0.0 4.822 End Atoms Cu 0.0 0.0 0.0 End BasisDefaults BasisType DZ End EndInput eor mv RUNKF CuSlab.runkf rm Points # ----------------------------- orbital plot -------------------------- export NSCM=1 $ADFBIN/band -n 1 << eor Title Cu slab orbital plot Comment Technical Quadratic K space integration Good real space integration accuracy Features Lattice : 2D Unit cell : 1 atom, 1x1 Basis : NO+STO w/ core Options : ElectronicTemperature (temperature effect) End Kspace 5 Accuracy 4 Restart CuSlab.runkf& OrbitalPlot End Grid Type Coarse End OrbitalPlot 1 Band 2 4 ! k-point 1, bands 2 to 4 3 -0.1 0.1 ! k-point 3 orbitals within 0.1 Hartree from Fermi Level End Convergence ElectronicTemperature 0.025 End Lattice 4.822 0.0 0.0 4.822 End Atoms Cu 0.0 0.0 0.0 End BasisDefaults BasisType DZ End EndInput eor echo "Begin TOC of tape41" $ADFBIN/dmpkf -n 1 TAPE41 --toc echo "End TOC of tape41"
Sample directory: band/NaCl_ionic/
A bulk crystal computation for Sodium Chloride (common salt), starting with charged atoms.
$ADFBIN/band << eor Title NaCl (from charged atoms) Comment Technical Hybrid K space integration (3D) Low real space integration accuracy Natural coordinates Lengths in Angstrom Parameters Dirac procedure Features Lattice : 3D Unit cell : 2 atoms Basis : NO w/ core Options : Ionic atoms in startup Note: Seems very unstable, highly sensitive to DIIS parameters (and simple damping) and well parameters. End Accuracy 3.5 Kspace 3 Diris NCycleDamp 15 End Units Length Angstrom End Lattice 0 2.75 2.75 2.75 0 2.75 2.75 2.75 0 End ATOMS Na 0 End Coordinates Natural Atoms Cl .5 .5 .5 End AtomType Na Dirac Na 4 1 RADIAL 2000 RMin 1E-8 RMax 200.0 VALENCE 1 0 2 0 2 1 3 0 0 SubEnd BasisFunctions 3S 1.75 3P 1.75 SubEnd FitFunctions 1 0 18.9 2 0 30.3 2 0 15.5 3 0 14.9 3 0 8.9 4 0 7.8 4 0 5.1 4 0 3.3 5 0 2.8 5 0 1.9 5 0 1.3 2 1 14.3 3 1 9.9 4 1 6.7 4 1 3.4 5 1 2.4 5 1 1.3 3 2 10.5 4 2 5.4 5 2 3.0 5 2 1.3 4 3 5.8 5 3 1.7 5 4 2.0 SubEnd End AtomType Cl Dirac Cl 5 3 RADIAL 2000 RMin 1E-8 RMax 450.0 RWell 150.0 VWell -2.0 VALENCE 1 0 2 0 2 1 3 0 3 1 SubEnd FitFunctions 1 0 29.1 2 0 49.5 2 0 26.1 3 0 25.8 3 0 15.8 4 0 14.2 4 0 9.4 4 0 6.2 5 0 5.4 5 0 3.8 5 0 2.6 2 1 21.2 3 1 16.5 4 1 12.4 4 1 6.8 5 1 5.1 5 1 3.1 3 2 16.6 4 2 9.4 5 2 5.5 5 2 2.6 4 3 8.7 5 3 3.3 5 4 4.0 SubEnd End End Input eor
Sample directory: band/Li2O_Bader/
To get the Atoms In Molecules (or Bader) analysis use the GridBasedAIM block key.
The grid-based AIM method is very fast, but a bit inaccurate. Therefore we force extra integration points.
Title Li2O bulk (fluorite structure) KSpace 3 accuracy 4 Integration accsph 6 accpyr 6 end GridBasedAIM End Dependency basis=1e-9 fit=1e-8 tails bas=0 core=0 fit=0 DIIS dimix 0.2 ncycledamp 0 end scf mixing 0.4 end xc gga scf bp86 end define ha=8.73/2 qa=8.73/4 end Lattice 0 ha ha ha 0 ha ha ha 0 end atoms O 0.0 end atoms Li qa qa qa 3*qa qa qa end BasisDefaults BasisType TZ2P Core small end end input eor
Sample directory: band/Li_BZPlot/
First run: automatic generation of band structure data on the result file (to be viewed with the GUI). We use a little bit of interpolation for smoother curve.
$ADFBIN/band << eor Title Li Bulk kspace 5 units length angstrom end ATOMS Li Li 0.0 0.0 0.0 END BZStruct Interpol 1 Enabled true Automatic true end Lattice -1.745 1.745 1.745 1.745 -1.745 1.745 1.745 1.745 -1.745 End SCF Mixing 0.3 end diis dimix 0.3 ncycledamp 0 end BasisDefaults BasisType DZ Core Large end end input eor
Second run: specifying the path through the BZ zone by hand. We set automatic to false and then specify the path with the BZPath keyblock, using one or more path sub keys. This run will actually produce exactly the same path as the automatic one.
$ADFBIN/band << eor
Title Li Bulk
kspace 5
units
length angstrom
end
ATOMS Li
Li 0.0 0.0 0.0
END
BZStruct
Enabled true
Automatic false
end
bzpath
kmesh 2
path
0.25 0.25 0.25
0 0 0.5
0 0 0
0.5 -0.5 0.5
0.25 0.25 0.25
0 0 0
subend
path
0 0 0.5
0.5 -0.5 0.5
subend
end
Lattice
-1.745 1.745 1.745
1.745 -1.745 1.745
1.745 1.745 -1.745
End
SCF
Mixing 0.3
end
diis
dimix 0.3
ncycledamp 0
end
BasisDefaults
BasisType DZ
Core Large
end
end input
eor
| BasisDefaults [1] | H2BulkGeo [1] | Ne [1] |
| BeO_tape41 [1] | H2SlabFreqTS [1] | NiCu_XC [1] |
| BetaIron [1] | H_chain [1] | NiO_Hubbard [1] |
| BNForce [1] | H_on_Pt [1] | Pd [1] |
| CnHn [1] | H_ref [1] | PE-NMR [1] |
| CO_MetaGGA_Force [1] | He_crystal [1] | Pt_slab [1] |
| Cu2_Confine [1] | HonPerovskite_Solvation [1] | RestartSCF [1] |
| Cu_resp_NR_ALDA [1] | Li2O_Bader [1] | Silicon [1] |
| Cu_slab [1] | Li_BZPlot [1] | SnO_EFG [1] |
| Diamond [1] | LiMetaGGA [1] | TiF3a [1] |
| Frags_COCu [1] | NaCl [1] | TiF3g [1] |
| Graphene_Dispersion [1] | NaCl_ionic [1] | ZnS_ModelPotential [1] |
