ADF development version subreleases
r18135: Added Python Interpreter and Foray Build Tool
An embedded Python interpreter has been added for Mac and Linux (32-bit). The Foray build tool has also been included. Details on how to use these tools have been added to the file documentation/text/Foray.text.
r18016: Meta-Hybrid implementation in ADF
SCF implemented for the metaHybrids M06, M06-2X, and TPSSh. All electron basis sets should be used.
XC METAHYBRID M06 End
XC METAHYBRID TPSSH End
r17989: Analytical frequencies for LDA PW92 implemented
LDA PW92 is supported for analytical second derivatives.
r17937: use OpenBabel to guess bonds
When OpenBabel is available within the ADF distribution, it will be used to guess the bonds. This should give more reliable results then the previous GuessBonds code, and should return proper bond orders as well.
r17890: bug-fix QUILD: set up was not properly saved
r17870: bug-fix QUILD: convergence parameters for Quild were not handled properly
r17838: Open shell spin-orbit (non-)collinear MetaGGA in ADF
SCF and optimizations are now also possible for the metaGGA's M06L and TPSS in case of open shell spin-orbit coupling in the (non-)collinear approximation. All electron basis sets should be used.
r17784: DOS module for the ADF-GUI added
The DOS module is the old BANDdos module that does no longer exist. The DOS module can handle both ADF and BAND result files (.t21 and .runkf).
The name of the executable is adfdos.
When opening a BAND runkf file, it works exactly as BANDDos used to do.
With a .t21 file from ADF the total dos (TDOS), or the GPDOS for selected atoms and optionally spdf-shells per atom (just as BANDdos did). Thus, the behaviour with ADF and BAND result files is identical.
Technically the DOS module uses the dos program when handling .t21 files. For a detailed explanation of the total dos and GPDOS see the manual for the dos program.
r17780: Add Mopac option for use with Quild in ADFinput
To use, set up a quild calculation and choose Mopac as description in the Quild panel. Use the Mopac panel to set the details if anything non-standard. Note that you need to use Mopac2009 or later.
r17768: Bugfix ADFmovie: bond distance info failed when bonds were created or destroyed during movie
r17732: Add MM option for use with Quild in ADFinput
Currently, the MM program can only be run as part of Quild. The MM program used is the NewMM program distributed with ADF. The input generation is currently not fool-proof, so the user must check the MM input, make sure the force field used is correct and so on.
It has been tested on small organic systems containing C, H and O atoms only.
r17715: Add interfae for Quild and Solvent molecules
Use the Regions panel to define regions. Not all options have been implemented yet.
Use the Quild panel to set up a quild calculation. The input descriptions for the subsystems are available from the button bar on the bottom.
To add explicit solvent molecules, give the command in the Solvent panel. Alternatively one can use the pull-down command in the Regions panel to add the solvent molecules to a particular region.
The newly created molecules will just be part of your usual input, or you can use Quild to give the solven a different treatment.
r17710: Mopac and DFTB interface improved
You can use the pre-optimizer with either of these methods, as before. Now you can also make detailed changes to the Mopac or DFTB input, run that set up interactively or as part of your job.
r17674: Density of States (DOS) utility enhancements
The DOS utility has been improved in many ways including much more simple input and speed improvements. There has also been a bug fixed that affected pdos values calculated on the SFO basis.
Changes to the input file format
Comments in the input. The DOS utility now recognizes ADF-style comments in the input file. Any line beginning with an exclamation mark "!" or a double colon "::" is considered a comment and is ignored.
TAPE21 keyword. An optional TAPE21 keyword has been added. It can be used to specify a result file from an ADF calculation instead of copying it to the current working directory:
TAPE21 /path/to/t21/file.t21
If this keyword is ommitted, DOS will read the molecular data from a TAPE21 file in the current directory, the same as before.
Number ranges. It is now possible to use ranges of numbers in the SFO and BAS specifications inside the PDOS, GPDOS and OPDOS keywords. For example, instead of specifying a long list of consequentive BAS functions as
GPDOS BAS 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 BAS 101 102 103 104 105 106 END
one can write it using ranges, which is completely equivalent:
GPDOS BAS 1:16 101:106 END
ATOM and ATYPE keywords. New keywords have been added for specifying sets of functions inside the PDOS, GPDOS and OPDOS input blocks. It is now possible to select functions based on the atom number or type and the type of atomic orbitals. Some examples:
! Select all 3d functions of the first atom GPDOS ATOM 1 3d END ! Select all p-type functions of the first atom GPDOS ATOM 1 p END ! Select all functions of the first atom GPDOS ATOM 1 END
Please note that atomic numbers are specified in the input order (as opposite to the ADF's internal atom order). Below are some examples of selecting atom types, which is equivalent to specifying all atoms of the same type one by one.
! Select all functions of all platinum atoms GPDOS ATYPE Pt END ! Select all f-type functions of all platinum atoms GPDOS ATYPE Pt f END ! Select all 6s functions of all platinum atoms GPDOS ATYPE Pt 6s END
Only one ATOM or ATYPE specification per line is allowed but it is possible to put any number of them into one *DOS section, for example:
GPDOS ATOM 1 3d ATOM 2 p ATOM 3 END
Note 1: Mixing ATYPE and ATOM keywords inside the same block is allowed but
it is not allowed to use ATYPE or ATOM keyword together with BAS and/or SFOs.
Note 2: In contrast to the BAS keyword, which produces wrong results
when used with d- and f-type functions, using the ATOM and ATYPE keywords in such
cases yields correct results. The reason is that the s and p
components of d- and f-sets of primitive functions, respectively,
are projected out before calculating partial and overlap DOS.
Bug fix in the PDOS of SFOs
In the PDOS values calculated on SFOs, there was a bug caused by the fact that the SFO overlap matrix, as saved by ADF on TAPE21, is not normalized. This does not pose any problem for the GPDOS and OPDOS because the eigenvectors are scaled accordingly so the total result is the same. However, the PDOS formulas assume that the overlap matrix is normalized and thus produce wrong results if it's not. After the fix, the SFO overlap matrix is normalized and the eigenvectors are re-scaled before use.
Speed improvements
The speed of DOS calculations has been greatly improved by using BLAS calls where possible. This yields about an order of magnitude speed-up for a 2000 by 3000 matrix. For example, calculating GPDOS of all 64 carbon atoms of a 161-atomic organometallic complex takes about 8 seconds on a 2.5GHz Intel Core 2 Duo T9300 processor.
r17613: Basis key has option to suppress output of create runs
To change where the output from ADF create runs and the Dirac program goes, use the CreateOutput option. If it is not present, it will go to standard output. The speciial value 'None' makes it disappear, and any other value will be used as a file name in which to save the output:
BASIS CreateOutput None END BASIS CreateOutput YourFileName END
r17553: MetaGGA during SCF and geometry optimization in ADF
Optimizations are now also possible for the metaGGA's M06L and TPSS. Note: the implementation is still being tested. All electron basis sets should be used. M06L needs high integration accuracy for reasonable gradients. For TPSS moderate integration accuracy for reasonable gradients is sufficient.
r17447: MetaGGA during SCF in ADF
Self consistency of the meta GGA is implemented as was previously done in BAND. Only single points calculations. The available MetaGGA's at the moment are:
XC METAGGA M06L End
XC METAGGA TPSS End
r17396: bug fix: PBE functional now default uses LDA PW92 and PBEc from Burke, examples
Bug fixed in bug fix r17379. Examples are updated. Warning: LDA VWN is supported for analytical second derivatives, but not the required LDA PW92, which is not solved yet.
r17379: bug fix: PBE functional now default uses LDA PW92 and PBEc from Burke
Before this bug-fix the 'correct' PBE functional could be obtained with explicitely specifying the LDA PW functional and the correct PBEc correlation functional:
XC LDA PW92 GGA PBE USEBURKEROUTINES End
Now this is default if one uses
XC GGA PBE End
The old ('incorrect') defaults can be calculated with
XC LDA VWN GGA PBE USESPROUTINES End
r17371: Compiled-in limit on the max number of MPI processes
As of now the max number of processes specified at the configure time has no effect. The library responsible for parallel communication in programs of the ADF package use dynamic memory allocation and thus can handle any number of tasks.
r17333: Bugfix converting KF files
Bugfix converting KF files: now variables can have length greater than the used length after conversion. Before, the variables would be truncated upon conversion to the used length.
r17312: Increase the max number of atoms to 30000
r17293: bug fix COSMO-RS: temperature results for different temperatures
Approximate ADF revision number. COSMO-RS revision number r172. Bug fixed in case more than 1 temperature is calculated in case of 'solvent vapor pressure', 'solubility', and 'boiling point', calculations. Only the results for the first temperature were correct.
r17252: bug fix graphs (xyplot): zooming and translating was broken
r17243: bug fix Bader analysis unrestricted calculation
The alpha and beta density are added and next the Bader analysis is performed.
r17209: Draw help lines in ADFspectra VCD plot
Horizontal lines are now drawn at 60 and 120 degrees. VCD angles within these lines are non-robust modes, those outside these lines are robust modes.
r17118: Plot UI (in ADFmovie, ADFspectra and Cosmo-RS) slightly changed
The main purpose of the change is better and more flexible handling of multiple curves within one plot. This is currently used in Cosmo-RS and ADFspectra (for VCD plots).
To adjust axes details, click next to the axes (outside the plot field). Thus, you have one dialog for the Y axes, and another for the X axes. If a second Y axes is present it also has its own dialog.
To adjust curve details (show points, colors, and so on) or global plot details, click below the X axes as this is combined with the X axes details.
Zooming with the mouse outside the plot window will zoom exclusively the X or one of the Y axes, depending on where the mouse is located.
r17109: Bugfix ADFview: handle many orbitals in field popup menus properly
r17089: ADFspectra now displays the angle in case of a VCD spectrum
r17075: ADFinput now supports setting masses per atom (for different isotopes)
Right click on an atom, and use the pop-up menu to set the mass of that atom (or the whole selection). The masses are used by ADF (using the AtomProps key), and are relevant for frequencies and releated data.
r17064: Bug fix ADFspectra: VCD information was not read correctly from .t21
r17057: Add AtomProps key to make it easier to use different isotopes
First make special atomic fragments in the ATOMS block (by modifying the element name). Next, use the AtomProps key to define extra parameters to the CREATE line for particular fragments:
Atoms
N 0.000000 0.000000 0.010272
H -0.471582 -0.816803 0.407861
H.D 0.943163 0.000000 0.407861
H.T -0.471582 0.816803 0.407861
End
AtomProps
H.D m=2.014101778
H.T m=3.01604927
End
Basis
Type TZP
End
In this particular example the AtomProps is used to set the masses of the H.D and H.T atoms. You can also use it to set the nuclear charge (q), see the create run options in the ADF manual.
NOTE:The AtomProps key can only be used in combination with the BASIS key.
r17030: ADFspectra vertical axes fixed, and option to keep height or area of the peaks constant
r17015: Double click in ADFjobs on .out, .err and .logfile now open in ADF-GUI
To open those files in a text editor, keep the control key pressed while double-clicking these files.
r17010: ADFmovie: option to scale displacements
Now you can scale the normal modes within ADFmovie (the displacements with respect to the molecular structure), either while viewing the movie or while viewing displacement vectors. To do this, use the Scale Displacements command in the View menu.
r17006: ADFmovie: fix displacement vectors
Displacement vectors are mainly used to visualize normal modes via ADFspectra.After showing and hiding displacement vectors, you could not show and hide them again without restarting ADFmovie.
r16981: Create fix2008 branch for ADF2008.01 release
r16741: More changed the way scratch files are handled by the master
The master will create its logfile and TAPE13 files in the cwd but place all other files in its scratch directory. For all tasks: if SCM_TMPDIR is not defined at all then the scratch sub-directories will be created in cwd.
r16658: Bugfixes for Prepare tool in ADFjobs (did not work on Linux)
r16652: Changed the way scratch files are handled
Changed the way scratch files are handled. As of now, instead of appending __<number> each temporary file name is prepended with the temp directory path. This makes copying stuff by the master from the CWD unnecessary and makes SCM_USETMPDIR obsolete. The files are still copied back from the temp dir to $SCM_WD at the end of the run.
r16616: Add ESR A-tensor and g-tensor to BANDinput
r16613: Fix the ADF-GUI (error about restraintsfixed)
Fixes a bug making the ADF-GUI impossible to use, introduced in r16595.
r16558: Fix ADFoutput to handle new geometry optimizer output
r16525: Bug fixes in ADFspectra related to unit changes
r16493: Add visualization of Bader properties to GUI
Atomic scalar properties are now available, just like other atomic charges. Also in the output browser a quick link to the Bader output has been added.
r16488: Bugfix in ADFinput which prevented generation of .run scripts
The bug was introduced at version r16363 ...
r16486: Updated HP-MPI version to the latest available
The HP-MPI library has been updated to the latest currently available: 2.2.7 on Linux and 1.1p1 on Windows
r16385: First implementation of Mobile Block Hessian method added
This is not a user-friendly release, and is largely untested. It calculates the MBH curvatures using finite-differencing on the ADF gradients.
To use this option, choose run type MBH, and give the names of the frozen blocks.
ATOMS C 0.0000 0.0000 0.0000 b=b1 H 0.6316 0.6316 0.6316 b=b1 H 0.6316 -0.6316 -0.6316 b=b1 H -0.6316 0.6316 -0.6316 b=b1 H -0.6316 -0.6316 0.6316 END Symmetry nosym Geometry mbh b1 End
r16363: Bader AIM analysis added (the bader panel) to ADFinput
r16236: Normal mode analysis after partial Hessian calculation
Full normal mode analysis is performed after analytical frequencies calculation with a NUC keyword present. This is done to assist users who perform partial Hessian calculations for a subsequent transitions search, to find the mode corresponding to their reaction coordinate.
r16228: Transit option now available in new optimization branch
A transit calculation option has been added in the new optimization branch. This is capable of performing both linear transits, and non-linear transits, and is the default when the 'Linear Transit' or 'Transit' sub-block is included in the 'Geometry' block.
The new transit code works differently to the old: the transit is represented as a sequence of constrained optimizations. A 'Constraints' block is used to delineate the constraints applied at each stage of the transit.
To perform a linear transit, start and end values are supplied.
Constraints angle 2 1 3 start=100.0 end=120.0 End Geometry Transit 4 Optim Deloc End
In the example above, 4 stages are required; ADF will interpolate the start and end values supplied for the angle between atoms 2, 1, and 3. Note that 'Transit' can now be used in place of 'Linear Transit', due to the more general nature of the new transit calculations.
Non-linear transits are possible, and can even be combined with linear transits in other coordinates. To perform a non-linear transit in a particular coordinate, explicit values must be given.
Constraints dist 1 2 0.8 0.9 1.1 1.15 angle 2 1 3 start=100.0 end=120.0 End Geometry Transit 4 Optim Deloc End
In the example above, 4 values are given for the distance between atoms 1 and 2. This distance constraint will be applied simultaneously with the linear transit constraints for the angle, with other degrees of freedom optimized at each stage of the transit.
It is worth noting that fixed constraints can also be used in a transit.
Constraints dist 1 2 0.8 0.9 1.1 1.15 angle 2 1 3 100.0 End Geometry Transit 4 Optim Deloc End
In this example, the angle between atoms 2, 1, and 3 will be fixed at 100.0 degrees at all stages of the transit.
Finally, it should be pointed out that 'partial constraints' are used by default in the transit calculations. These constraints are not required to be fully met at each intermediate geometry, but are fully met at the converged geometries. You can use fully converged constraints by supplying the FULLCONVERGE option to the 'Constraints' subblock of the 'Geometry' block (not to be confused with the 'Constraints' block at root level).
r16192: ADF-GUI use lines to show bonds
Check the Bond Lines Only box in the View menu (Molecule Ball&Sticks) to visualize bonds using simple lines instead of tubes. This will speed-up display of large systems.
You can set the default that you like in the Preferences.
r16188: ADFjobs may dynamically import queues from remote systems
You can set up queues on some remote system in the normal way. Next, when you lauch ADFjobs on a different machine you can configure it to dynamically load the queues defined on the remote system. To do this, open the Preferences, go to ADFjobs, and specify the host and username for the remote system.
By default ADFjobs will search in the remote .scm_gui directory. This is the location in which ADFjobs normally saves the queue definitions. However, if on the remote machine the environment variable SCM_QUEUES is defined ADFjobs will try to get the queues from $SCM_QUEUES. This makes it easy for a system administrator to preconfigure queues to be used by all ADFjobs users.
r16175: Changes in 'Constraints' input block
The 'NewConstraints' block has now been changed to 'Constraints', and the old 'Constraint' block is now 'LinearConstraints'.
In the 'Constraints' block, constraint values are now optional (where appropriate). For example, if a distance constraint is imposed, and no value supplied, the initial value is used for the constraint. The same applies to other constraints (eg angle, coordinate, fixed atom).
The 'Constraints' block is now also included in BAND, with the same functionality as in ADF.
r16119: Core density and its derivatives from Densf
In order to get frozen core density one should specify CORE
after the Density keyword:
Density Core
Similarly specifying CORE after DenGrad or DenHess
will tell Densf to calculate the core density gradient or the Hessian, respectively.
Please note that in order to get total density gradient and/or Hessian in the same calculation,
the DenGrad and/or DenHess keyword must be specified again on a
separate line. For example, specifying
DenGrad CORE DenGrad
will result in both core and the total exact SCF density gradient.
The calculated frozen core density and its derivatives are saved on TAPE41 in the section
Core. The variables inside the section are named exactly the same way as the
corresponding variables in other sections: Density, DensityGradX,
DensityHessXX, etc.
r16038: kf.py now has methods that return details of the table of contents of a KF file
r16026: ADFinput now picks up atomic charges after a DFTB preoptimization
r16025: Magnetizability, ORD and polarizability options improved
The different options that used to be present in the Polarizability panel have been reorganized. Also new functionality offered by the AORESPONSE method in ADF has been made available. This includes magnetizability, and lifetime options.
r16025: Raman in ADF-GUI improved
A new Raman panel has been added to make it easier to set up raman calculations (both full and range scan). Also ADFspectra has been fixed to handle raman spectra when symmetry is used.
r16000: DENSF can now read grid as list of points and calculate density gradient and Hessian
When specifying inline grid the Grid keyword should look as follows:
Grid Inline x1 y1 z1 x2 y2 z2 ... xN yN zN End
Here, x#, y#, and z# are coordinates of points at which requested properties will be calculated. This feature may be used, for example, by external programs to calculate various properties at a number of points exactly and avoid interpolation with its inaccuracy. This feature should be used only when the output file has a TAPE41 format.
The following keywords may be used to calculate partial derivatives of the exact electron density:
DenGrad DenHess
In order to calculate partial derivatives of the fitted density, one should add fit after the
corresponding keyword:
DenGrad fit DenHess fit
The calculated properties are saved to TAPE41 in the section SCF variables
DensityGrad# and DensityGrad## where # is one of
X, Y, or Z. For example, an XY component of the density
hessian is saved in SCF%DensityHessXY. Derivatives of the fitted density are saved
to variables with names prepended with Fit, for example, SCF%FitDensityHessXY.
r15995: Calculate Frequencies at the end of a geometry optimization or transition state search automatically
At the end of your optimization calculation (eihter a TS or geometry), you can calculate the frequencies using the anaylitical frequencies method. To do this, go to the detail tab in ADFinput (use the ... shortcut at the right of the task name in the Main tab), and check the frequencies check box at the bottom of all options.
r15975: Using Hessian from Restart Tape with New Optimizer Branch
You can now use a hessian file stored on a TAPE21 as the initial hessian in a geometry run (eg transition state search) with the new optimizer. The input setup is exactly the same as with the old optimizer, ie, you simply need to use the 'Restart' block.
r15931: NewConstraints are possible in BAND
It is now possible to specify geometry constrains that will be enforced during geometry optimization in BAND the same way as it has been already possible with ADF. The constraints are currently implemented only as 'strict', i.e. they are enforced from the first geometry update. The atoms are numbered from 1 as they appear in input. Note: Should the NewConstraints block be renamed in ADF, it will also change its name in BAND.
The constraints are specified as follows:
NewConstraints Atom N1 X Y Z ! Atom N1 is frozen at the specified coordinates Dist N1 N2 Value ! Distance N1-N2 is kept fixed at the specified value Angle N1 N2 N3 Value ! Angle N1-N2-N3 is kept fixed at the specified value Dihed N1 N2 N3 N4 Value ! Dihedral angle N1-N2-N3-N4 is kept fixed at the specified value End
Here, N1,...,N4 are atom numbers as they appear in the input;
Value is a value of the constraint in Angstroms or degrees.
r15917: Vibration Frequency Calculations in BAND
It is now possible to calculate vibrational frequencies of atoms in unit cell by means of finite difference method. This is done by specifying:
RunType Frequencies End
Options may be specified using a Frequencies input block:
Frequencies Step 0.001 Nuc 1 2 3 4 5 End
Here, Step sets the finite difference step size in Angstrom (0.001 by default).
The Nuc keyword lists atoms that will be included in the frequency calculation,
which makes it possible to calculate only part of the second derivatives matrix. For example,
this option can be used save time when one is interested in vibrations of an adsorbed
molecule by allowing only atoms of the molecule (and possibly the nearest atoms of the surface)
to move.
r15894: Partial Hessian option added to ADFinput
Go to the Frequencies panel, select some atoms, and press the + to add them to the list of atoms for which to calculate the Hessian.
r15894: Add constraints for new geometry optimizer to ADFinput
The Geometry Constraints panel can be used to create Atom, Bond, Angle or Diheral constraints for the new optimizer.
If the old optimizer is used, the Atom constraints will be ignored (for now), and the others will be added as restraints.
r15852: Added examples for the new AOResponse functionality
It is now possible to calculate optical rotation (ORD) spectra using the latest AOResponse functionality by Jochen Autschbach and coworkers. There have been added a few test examples, description thereof follows below.
For details see the following papers:
Krykunov, M.; Autschbach, J., Calculation of static and dynamic linear
magnetic response in approximate time-dependent density functional
theory, J. Chem. Phys. 2007, 126, 024101-12.
Krykunov, M.; Kundrat, M. D.; Autschbach, J., Calculation of CD spectra
from optical rotatory dispersion, and vice versa, as complementary tools
for theoretical studies of optical activity using time-dependent density
functional theory, J. Chem. Phys. 2006, 125, 194110-13
Krykunov, M.; Autschbach, J., Calculation of origin independent optical
rotation tensor components for chiral oriented systems in approximate
time-dependent density functional theory, J. Chem. Phys. 2006, 125,
034102-10.
Autschbach, J.; Jensen, L.; Schatz, G. C.; Tse, Y. C. E.; Krykunov, M.,
Time-dependent density functional calculations of optical rotatory
dispersion including resonance wavelengths as a potentially useful tool
for determining absolute configurations of chiral molecules, J. Phys.
Chem. A 2006, 110, 2461-2473.
Krykunov, M.; Autschbach, J., Calculation of optical rotation with
time-periodic magnetic field-dependent basis functions in approximate
time-dependent density functional theory, J. Chem. Phys. 2005, 123,
114103-10.
Baev, A.; Samoc, M.; Prasad, P. N.; Krykunov, M.; Autschbach, J.,
A Quantum Chemical Approach to the Design of Chiral Negative Index
Materials, Optics Express 2007, 15, 5730-5741.
Krykunov, M.; Banerjee, A.; Ziegler, T.; Autschbach, J., Calculation of
Verdet constants with time-dependent density functional theory.
Implementation and results for small molecules, J. Chem. Phys. 2005, 122,
074105-7.
Input options: Property to calculate
Polarizability
By default, the polarizability is calculated. This can be modified using one of the keys below.
Optical Rotation
OPTICALROT
Specify OPTICALROT to calculate optical rotatory dispersion spectrum instead of polarizabilities.
Optical Rotation with Velocity Gauge
VELOCITYORD
This option should be used instead of OPTICALROT with GIAO if the finite lifetime effects need to be taken into account (LIFETIME option).
Magnetizabilities
MAGNETICPERT
Calculate static or time-dependent magnetizability.
Input options: Calculation parameters
Frequency of the Perturbation Field
To calculate time-dependent properties, one needs to specify frequency of perturbation
field. This can be done using one of the options below.
FREQUENCY Nfreq freq1 freq2 ... freqN units
Here Nfreq specifies the number of frequencies that follow. The last item on the line specifies the units and is one of EV, HARTREE, ANGSTROM.
FREQRANGE startfreq endfreq Nfreq units
Here Nfreq specifies the number of frequencies in the closed range startfreq thru endfreq. The last item on the line specifies the units and is one of EV, HARTREE, ANGSTROM.
Damping (Resonance Peak Width)
LIFETIME width
Specify the peak width in Hartree
XC Kernel
ALDA
Use VWN kernel. This option is the default.
XALPHA
Use Xα kernel instead of the default VWN one.
Technical paramaters and expert options
SCF {NOCYC} {NOACCEL} {CONV=conv} {ITER=niter}
Specify CPKS parameters such as the degree of convergence and the maximum number of interations:
- NOCYC - disable self-consistence altogether;
- NOACCEL - disable convergence acceleration;
- CONV - convergence criterion for CPKS. The default value is 10-6.
The value is relative to the uncoupled result (i.e. to the value without self-consistence). - ITER - maximum number of CPKS iterations, 50 by default.
Specifying ITER=0 has the same effect as specifying NOCYC.
GIAO - include the Gauge-Independent Atomic Orbitals (GIAO). This option should not be used with damping
(LIFETIME keyword) and the VELOCITYORD option should be used instead.
FITAODERIV - use fitted AO Derivatives.
COMPONENTS {XX} {XY} {XZ} {YX} {YY} {YZ} {ZX} {ZY} {ZZ}
Limit the tensor components to the specified ones. Using this option may save the computation time.
Examples
DMO_ORD_aoresponse
This test example consists of two ORD calculations: a with and without velocity gauge.
H2O_magnet
This test demonstrates how to calculate static magnetizability of a water molecule.
H2O_TD_magnet
This test demonstrates how to calculate dynamic magnetizability of a water molecule.
r15824: Added Swiss Cheese gradient smoothing option
A new variant of the standard Voronoi-based integration grid has been added to improve gradient smoothness during geometry optimizations (and other runs where the geometry is varied). The algorithm works as follows: A standard Voronoi-based integration grid is generated, but without the spherical integration regions that usually surround each atom. After this 'sea' of points has been formed, spherical holes are made at the atomic centers by mapping points radially outward (hence the name 'swiss cheese'). These holes are then filled up with more points.
This explains how the grid is generated, but not how it smoothes the gradients. When the geometry is varied, the sea of points is conserved. That is, the original positions of the points are stored, and used at each geometry; the points are mapped radially to create the atom-centered holes, but the variation in the positions and weights of the points varies smoothly. The 'balls' of points surrounding each center move with the atoms. So the overall picture is similar to ball bearings moving through a sea of oil: the oil moves aside to make way for the ball bearings, and moves back in to fill any space left behind.
To use this option, you use the 'SMOOTH' sub-block in the 'GEOMETRY' block, and supply as argument the keyword 'SWISSCHEESE'.
Geometry smooth swisscheese End
Optionally, you can add an extra decimal argument that gives the maximum distance (in Angstrom) that an atom may move before the integration is regenerated from scratch (default is 0.2 Angstrom).
Geometry smooth swisscheese 0.5 End
The larger the number given, the more effective the smoothing; however, using a too large number could lead the integration grid to become 'inappropriate' for the system geometry, leading to poor integration accuracy.
WARNING: The radial mapping of points reduces the accuracy of the integration, because it effectively warps the integration space. For this reason, more points are needed when using swiss cheese smoothing. So for the same accuracy parameter, a swiss cheese calculation is more expensive than a standard calculation.
r15818: Show scalar values by atomic radii in GUI
Atomic scalar values (charges, radii, shieldings and so on) can be represented by atom sizes. The absolute value of the scalar values is used, and the full range of scalar values is mapped into the minimum to maximum range from the preferences. No mapping is performed if the scalar values already fit in the range.
r15813: ADFmovie: export to .amv file of current movie added
r15810: Color legend added when atoms are colored by scalars
r15807: Capped-Bonds-Only visualization for molecule added
To use it, check the 'Capped Bonds Only' box in the View menu. Use the bigger and smaller menu comamands from the view menu to adjust the diamater of the bond to your taste. Uncheck the 'Show Atoms' in the View menu to remove the caps and get a pipe-representation.
Note that this representation of your molecule makes it very hard or impossible to select atoms.
r15798: Franck-Condon Factors program added
fcf is an auxiliary program which can be used to calculate Franck-Condon factors from two vibrational mode calculations.
fcf requires an ascii input file where the user specifies the TAPE21 files from two adf vibrational mode calculations, carried out for two different electronic, spin or charge states of the same molecule. These calculations can be either numerical or analytical. The number of vibrational quanta that have to be taken into account for both states in the evaluation of the Franck-Condon factors have to be specified.
fcf produces a (binary) KF file TAPE61, which can be inspected using the kf utilities. Furthermore, fcf writes the frequencies, vibrational displacements and electron-phonon couplings for both states too the standard output, including any error messages.
The input for fcf is keyword oriented and is read from the standard input. fcf recognizes only two different keywords and both must be specified to perform the calculation. All input therefore has the following form:
STATES state1 state2
QUANTA l1 l2
STATES
The filenames of two TAPE21 files resulting from a numerical or analytical frequency calculation. The calculations must have been performed on the same molecule, i.e. the type, mass and order of occurence of all the atoms (or fragments) has to be the same in both files.
QUANTA
The maximum number of vibrational quanta to be taken into account for both states. Franck-Condon factors will be calculated for every permutation of up to and including l1/l2 quanta over the vibrational modes.
r15779: GUIPrefs module added
The preferences for all GUI modules are now combined and handled by the GUIPrefs module. Selecting 'Preferences' from the menu (currently the SCM menu, that might change) will start this module.
r15745: Bugfix in SCF Coulomb potential display in ADFview (was a bug in densf)
r15736: Virtual orbitals in ADFview may use a different color scheme by default
Normally when you save the color scheme (using the Color details pulldown menu in the control line for a field) it applies to all colors as before. However, if you are currently displaying a virtual orbital the saved colors will apply to virtual orbitals only. If no special color scheme for virtual orbitals has been saved the same color scheme as for occupied orbitals (and anything else) will be used.
r15730: Atom numbering in the ADF-GUI and BAND-GUI now starts at 1
Now the atom numbers match the atom numbers in the input order of ADF.
r15724: Cube file export in ADFview added
An option to export the fields calculated by ADFview as Gaussian Cube files (.cub files) has been added.
Use the 'Export Fields As Cube Files' command from the File menu to use it.
r15685: BAND changes in DOS
1) Calculate DOS on the fly, preventing an old problem that sometimes more nodes were required to perform this task.
2) Store data for the GUI module BANDdos on the .runkf file.
3) Normally the DOS is evaluated at a grid of energie values with separation Delta using an exact delta function. Plots thus produced look quite rough and narrow peaks can be missing entirely. It is more convenient to integrate the dos over each energy interval Delta E. This leads to a more comprehensible DOS. Very narrow peaks show up as a peak of 1 electron (or multiples of that, depending on degeneracy). The input option is Dos::IntegrateDeltaE. The default is now true. To get the old-style DOS it should be set to false.
r15661: ADFlevels Show Interaction modified
On startup, ADFlevels will show interactions for the visible levels of your final molecule.
If no levels are selected, the 'Show Interactions' menu command will show all interactions to the currently visible levels of your final molecule. This is fast and normally what you want to see. To see also the other interactions from visible fragment levels, you need to select them explicitly and use the 'Show Interactions' menu command again.
If levels are selected, the 'Show Interactions' menu comamnd will show all interactions to the selected levels, no matter if they are visible or not.
r15637: BANDdos module added
Visualize different kinds of DOS when present on a .runkf file. You will need a recent BAND version, and the name of the BANDdos executable is banddos.
The total dos is displayed by default. Select atoms to get the corresponding partial DOS. Use the popup menus on the atoms (right click, or left clickandhold) and select some partial DOS by funtion or L value.
r15622: Color bonds in GUI by atom types
You can change the bond coloring: the colors of the atom types will be used. To activate, check the 'Color Bonds By Atom' check box in the View menu.
In the preferences panel (in ADFinput) you can set and save the default value.
r15585: DFTB Writes Results to KF File
The DFTB code now writes final results to the DFTB.kf file.
r15527: ADFjobs search fields combined
All search terms must occur, and the search is done in the job name and queue name fields.
r1505: Extent MOPAC panel: user can specify any MOPAC key (for experts only)
r15489: Resizing ADFjobs: job name will use additional space
This is useful if you have jobs with long names
r15477: Zooming and Translation added in BANDstructure
Drag with left mouse to translate, zoom with right mouse, control left, or scroll wheel. This is identical to the controls in ADFlevels, ADFspectra and the ADFmovie plots.
r15446: Visualization of NBOs in ADFview added
If they are present on the .t21 file, you can visualize them. Also, the NBOs and NLMOs in the ADFview menus have more informative labels. In ADFinput, use the Orbitals pane to perform the NBO analysis.
r15443: Modifications and additions to the Grimme dispersion-corrected functionals
For description of the first implementation see below under "r15257: Grimme Dispersion Correction". In this revision, a possibility to modify the van der Waals radii scaling parameter has been added. The DISPERSION keyword is now as follows:
DISPERSION {s6scaling} {RSCALE=r0scaling}
here, s6scaling is the global scaling parameter s6 and r0scaling is the scaling parameter for VdW radii, which is equal to 1.1 by default.
This revision also adds all GGA-D energies from the J Comp Chem (2006) paper to the list of Post-SCF energies printed with the METAGGA keyword. These are: BLYP-D, PBE-D, BP86-D, TPSS-D, B3LYP-D, and B97-D.
r15430: Mopac panel now also works with Mopac2006 (as well as Mopac2007)
r15395: Add Mopac panel in ADFinput to perform interactive Mopac calculations
Select the Mopac panel to perform an interactive mopac calculation, with more options available then using the pre-optimzer button. If the geometry of the molecule is changed it will be reflected in the molecule in ADFinput. The MOPAC charges property of the atoms in the ADF-GUI will also be updated, and optionally the bonds may be updated according to the calculated MOPAC bond orders. Currently only MOPAC2007 is supported from this panel.
r15390: Import Coordinates in ADFinput now can read MOPAC AUX file
r15384: Fix in algorithm to get job ID when submitting with ADFjobs: now works with LoadL
r15379: Added option to calculate energies/gradients for geometry runs via external program
A subblock has been added to the GEOMETRY block to allow energies and gradients to be calculated by an external program for use in a geometry optimization using the new branch. A simple input looks like this:
Title WATER Geometry Optimization with External program
Basis
Core large
Type SZ
end
Atoms
O 0.000000 0.000000 0.000000
H 0.000000 -0.689440 -0.578509
H 0.000000 0.689440 -0.578509
End
Geometry
externprogram externprog.exe coord=coords.inp energy=energy.out grad=grads.out
End
Note that you need to supply information about atomic fragments, such as the basis set, even though these are not actually used in the calculations.
The EXTERNPROGRAM subblock must have as first argument the name of the executable program or script that gets run to generate the energy and gradients. It also takes keyed parameters for the coordinates file name, the gradients file name, and the energy file name. These files are used to communicate values between ADF and the external program.
When ADF is ready to perform an energy and gradient calculation, it writes the current cartesian coordinates to the file name given in the input. The format is similar to the ATOMS block in the ADF input file: it has one atom per line, with the the element symbol given, followed by the x, y, and z coordinates.
ADF will then run the executable program, and then read in the energy and gradients from the file names given in the input file. The external program is thus responsible for reading the coordinates (in atomic units) written by ADF from file, generating the corresponding energy and gradients (in atomic units), and writing these to the appropriate files. ADF will then take another geometry step, and the process will repeat.
r15363: Try to improve importing of coordinates from PDB files in the ADF-GUI
r15356: Turn of printing of SFO EIG and SFO OVL by default
To turn back on (and get old output), add the following to your input:
EPRINT
SFO eig ovl
END
r15284: Use empty fields in ADFinput for default values
Many input fields in ADFinput are now empty. The corresponding values will not be written in the ADF input file, and as a result the default value as determined by ADF will be used. This will ensure that defaults of ADFinput and ADF match, and also generates cleaner run scripts.
r15257: Grimme Dispersion Correction
The Grimme's dispersion correction has been added as part of the XC functional. The correction is added to SCF energy, gradients and the Hessian matrix elements when those are calculated. It is switched on by specifying a DISPERSION keyword in the XC input block or by selecting the corresponding GGA-D in ADFinput. The DISPERSION keyword has an optional parameter that specifies the global scaling factor with which the correction is added. If the parameter is omitted then 1.0 is used except for the following GGA's: BLYP, PBE, or BP86. For these functionals, the factors proposed by Grimme in J. Comp. Chem. (2006) p. 1787 are used. They are: 0.75 (PBE), 1.2 (BLYP), 1.05 (BP86).
r15253: Option to user MOPAC hessian as startup hessian in geometry optimizations
To activate, check the Mopac Hessian option in the Task:GeometryOptimization panel in ADFinput.
r15242: High-order quasi-newton based optimize option, and implicit hessian updates
Two new experimental algorithms have been added: the first is an implicit hessian update algorithm. The standard explicit scheme involves beginning an optimization with an approximated hessian at the initial geometry, which is estimated using a lower level of theory such as a force field, and then updating this hessian each step using the gradients and steps calculated during the optimization.
The implicit scheme is similar, but an approximate hessian is generated at each step using the lower level of theory, and then updated with the preceding steps and gradients which have been stored during the optimization. In theory, the implicit scheme should be better, because the hessian 'guess' used is evaluated for the current geometry, rather than the initial geometry. In practice, not enough testing has been done to know if this is really the case.
To use the implicit hessian update, you append the work 'implicit' to the end of the hessupd subkey in the geometry block:
geometry hessupd bfgs implicit end
The high-order quasi-newton scheme is designed to work in collaboration with the implicit hessian update. In the standard quasi-newton scheme, the approximate hessian is used together with the calculated forces to estimate the location of the minimum of the true PES by approximating it with a single quadratic surface. In the high-order scheme, a steepest descent propagation is used, with the true PES approximated locally at each step by a quadratic surface. This surface taken from the approximate hessian given by the implicit hessian update method at each point. The hessian is geometry dependent, and varies along the path followed. The gradient is integrated along the path until it is close to zero. This approach effectively introduces higher-order terms via the implicit hessian update method, which come from the lower level of theory used to build the approximate hessian. It is expected to work particularly well with cartesian coordinates, for which the surface may be highly anharmonic, but it is not well tested.
To use the high-order quasi-newton scheme, you should set the hessian updater to use the implicit scheme, and add a quasinewton subblock:
geometry quasinewton highorder hessupd bfgs implicit end
r15237: Changed the default geometry optimizer, integration settings, and gradient convergence
The new optimization branch is now the default for geometry optimizations and transition state searches. Other types of runs (eg LT, IRC) make use of the old branch. The new optimization branch uses delocalized coordinates by default.
In order to use the old optimizer, use the 'branch' key in the geometry block:
geometry branch old optim cartesian end
The default gradient convergence for geometry optimizations has been changed to 1.e-3. The integration parameters for geometry optimizations are now 4 4, rather than 4 3.
r15171: Changed DFTB parameter naming scheme
The DFTB naming scheme for parameter files, which are kept in atomicdata/DFTB, has been changed to the scheme used in the DFTB+ code. That is, the (mixed-case) name of each atom, separated by a hyphen, and including the extension 'skf' (Eg. C-Co.skf)
r15128: ADFspectra mouse interaction reworked
Now ADFspectra uses the same code as used for other plots. Thus the mouse interactions has changed. Use the scroll bar, or the right mouse button, or control-left mouse button to zoom. Drag with left mouse button to scroll.
r15120: Improvements to DFTB performance
The calculation of gradients in the DFTB code is now orders of magnitude faster, due to new optimizations.
r15073: Add Boys localized orbitals and NBO analysis to ADF-GUI
In ADFinput you can enable calculating them in the Orbitals panel. When they are present on a TAPE21 result file, you can also visualize the Boys localized orbitals and the NLMO's from the NBO analysis in ADFview.
r15060: In ADFjobs, a single click on the job button
ADFjobs already had a short-cut: double click on a job and it will open either in ADFinput or in ADFtail. This is now even more convenient: you get the same effect by clicking once on the job button (on the left side).
r15046: Bugfix ADFlevels: sometimes electrons were not drawn (especially for partially filled levels)
r15043: ADFlevels mouse interaction improved (and very different from what it was ...)
Drag with left mouse to translate, use scroll wheel to zoom. Alternatively, drag with right mouse outside a stack to zoom. Drag with left mouse and shift key to select levels with marquee. Finally, control-drag with left mouse to zoom for those who have a single button mouse without scroll wheel.
r15036: Fixes for report facility in ADFjobs on windows
r15019: Spin and Occupation panel in ADFinput
In ADFinput you can use the Spin and Occupation panel to set the occupation details. ADFinput will perform a fast test run adf to get required information about symmetry and occupations if no matching .t21 is available.
r15016: Control menu choice in SCM menu
If you choose an option from the SCM menu while pressing the control key, the choosen module will launch but with no file specfied. Thus it is a simple way to launch modules.
r14978: Enable ScaLapack for Itanium2/Linux and PC/Windows platforms
Diagonalization of the Fock matrix has up to now been performed on the master node only. With ScaLapack, it is now possible to distribute the work among all nodes in a parallel run. ScaLapack as implemented in Intel(R) MKL 10.1 is currently used in Itanium2/Linux/HP-MPI version of ADF. The Windows/HP-MPI ADF version uses standard ScaLapack.
r14953: Added option to read initial hessian from text file
You can now use the 'inithess' subkey in the geometry block to read a hessian from a text file, to use with the new optimizer branch. The only argument is the name of the file containing the initial hessian. The hessian must be given in full, in non-mass-weighted cartesian coordinates, and in atomic units (hartree/bohr**2).
r14929: Add scripting tools: kf.py
kf.py is a Python interface to read and write KF files (the binary files from the ADF package). For reading, adfreport is easier and more reliable (it will be updated when data on the binary files is changed). To use, add $ADFHOME/scripting to your PYTHONPATH, and a working installation of the ADF package is required (kf.py uses the KF utilities).
r14897: Update ADFLaunch to work with Mac OS X 10.5 (Leopard)
r14868: Fix bug in atom ordering of atom charges (except Hirshfeld) in ADF-GUI
r14861: Generating sets of jobs, reporting results, and scripting
In ADFjobs the Tools/Prepare menu can be used to generate sets of jobs with different options. This is a GUI to the adfprep program.
The Tools/Report commands may be used to generate reports from a set of selected jobs. Reports can include any information present on the .t21 result files, including pictures of orbitals. This is a GUI for the adfreport program.
adfprep is intended to facilitate scripting: it makes it very easy to construct proper adf jobs from within a script. Execute 'adfprep' from the command line without arguments to get more information on how to use it.
adfreport is intended to facilitate scripting: it makes it very easy to get results calculated by adf in your own script. The results are taken from a .t21 result file. Execute 'adfreport' from the command line without arguments to get more information on how to use it.
Check the Bakerset and ConvergenceTestCH4 examples from the examples/adf directory to see some (very simple) scripts using adfprep and adfreport.
r14849: Import coordinates now tries to figure out the proper units
The import coordinate command in adfinput now tries to determine if bohr or angstroms are used, based on the number of bonds found using a distance criterium.
r14845: ADF-GUI: color atom by scalar property
ADFinput, ADFview and ADFmovie now can color the atoms by a selected scalar property, for example the calculated atomic charge. Use the popup menus to do this. Currently the color range is fixed.
r14781: Laplacian of the exact and fitted SCF density
It is now possible to calculate Laplacian on the total SCF density (exact or fitted) using the densf utility. The corresponding densf keyword is
Laplacian {fit}
The fit argument is optional and if it is present densf will calculate Laplacian of the fitted density.
The Laplacian feature is also supported by ADFview.
r14768: Show multiple atom info
In ADFinput, ADFmovie, ADFview and BANDinput, show multiple atom properties (name, charge, radius etc) at the same time. Also the updating of the displayed information has been fixed.
r14677: MOPAC as pre-optimizer in ADFinput
ADFinput now can also use MOPAC as a pre-optimizer. Currently MOPAC2007 (from www.openmopac.net) and MOPAC2006 (from Fujitsu) are supported. Atom charges as calculated can be displayed. With MOPAC2007 the bonds in ADFinput may also be updated based on the bond orders calculated.
MOPAC2007 needs to be installed in /opt/mopac, MOPAC2006 needs to be installed in /opt/mopac2006. If you use different install locations you need to update the $ADFBIN/mopac.scm file.
You can also use the SCM_MOPAC environment variable to select the MOPAC you wish to use, even on a remote machine. For example:
SCM_MOPAC=ssh bigmachine adfhome/bin/mopac.scm export SCM_MOPAC
will start mopac on bigmachine using the normal mopac.scm script.
Use the Preferences panel from ADFinput to select which pre-optimizer to use.
r14341: Added block constraints
Block constraints allow the internal degrees of freedom of a block of atoms to be frozen, so that the block moves as a whole. Block constraints are now available when using the new branch optimizer with the NewConstraints input block. To apply block constraints, you add block labels to atoms in the Atoms block, and then add the block constraint in the NewConstraints input block.
ATOMS
1.C -0.004115 -0.000021 0.000023 b=b1
2.C 1.535711 0.000022 0.000008 b=b2
3.H -0.399693 1.027812 -0.000082 b=b1
4.H -0.399745 -0.513934 0.890139 b=b1
5.H -0.399612 -0.513952 -0.890156 b=b1
6.H 1.931188 0.514066 0.890140 b=b2
7.H 1.931432 0.513819 -0.890121 b=b2
8.H 1.931281 -1.027824 0.000244 b=b2
END
NEWCONSTRAINTS
BLOCK b1
BLOCK b2
END
GEOMETRY
OPTIM DELOCAL
END
This comes from the example GO_newconstraints.
r14166: New constraints now fully enforced
The constraints in the new optimizer branch were initially not fully enforced. Although they had to be converged at the end of an optimization, they might not be met during the optimization. This has been changed, such that they are now enforced by default. They do not have to be fully met in the input, but if the input geometry is far from meeting the constraints, a large, erratic first geometry step may result.
You can avoid fully enforcing constraints, effectively returning to the old behavior, by adding a 'constraints' subblock to the geometry block:
GEOMETRY CONSTRAINTS PARTIALCONVERGE END NEWCONSTRAINTS ... END
Although the default, you can explicitly indicate that you want to fully enforce constraints at each step by using
FULLCONVERGE in place of PARTIALCONVERGE.
r14142: DFTB as pre-optimizer in ADFinput
ADFinput now can also use DFTB as pre-optimzer. The DFTB program part of the ADF package will be used, and you will need to download the parameter files yourself.
Use the Preferences panel from ADFinput to select which pre-optimizer to use.
r13856: Introduced new Density Functional Tight Binding (DFTB) program
The DFTB program is orders of magnitude faster than DFT, but requires parameter files to be installed for all pair-wise combinations of atoms in a molecule. Unfortunately, parameters are not available for some elements, but calculations for many common molecules should be possible.
To use DFTB, you need to request parameter files at the DFTB.org web site, and put these files in the directory atomicdata/DFTB.
The input for DFTB is similar to ADF, but not exactly the same. At this point, it is possible to perform single point, geometry optimization, transition state search, and frequency calculations. All output is written to standard output; there are no keyed-files (eg TAPE21) at this time.
Examples of DFTB calculations can be found in the examples/dftb directory. Here is a run script for one such example:
"$ADFBIN/dftb" << eor
Geometry
RunType GO
Optim Delocal
Converge Grad=0.0001
End
Atoms
C 0.000000 0.000000 0.000000
C 1.402231 0.000000 0.000000
C 2.091015 1.220378 0.000000
C 1.373539 2.425321 0.004387
C -0.034554 2.451759 0.016301
C -0.711248 1.213529 0.005497
O -0.709522 3.637718 0.019949
C -2.141910 1.166077 -0.004384
O -2.727881 2.161939 -0.690916
C -0.730162 4.530447 1.037168
C -0.066705 4.031914 2.307663
H -0.531323 -0.967191 -0.007490
H 1.959047 -0.952181 -0.004252
H 3.194073 1.231720 -0.005862
H 1.933090 3.376356 -0.002746
O -2.795018 0.309504 0.548870
H -2.174822 2.832497 -1.125018
O -1.263773 5.613383 0.944221
H -0.337334 4.693941 3.161150
H 1.041646 4.053111 2.214199
H -0.405932 3.005321 2.572927
End
eor
r13415: Reference point (August 21, 2007)
Version ADF2007.01 split from the development version. Bug fixes in ADF2007.01 are also included. The svn version numbers in the development version that correspond to these bug fixes will be approximately the same as those in the ADF2007.01 version. See the 2007.01 release notes for details.
