NBO need the FULLFOCK key, and there seems to be a problem with the fock key.
The nuclear gradient was no longer correct for spin-unrestricted functionals.
This did not work properly for spin-restricted calculations.
It did not work properly if the orbital belongs to an irrep which has a name containing a quote.
The Bader properties were read in input-order from t21, which must be the internal ADF order. Thus if the internal order is different from the input order the Bader atomic properties were mixed up when displayed with the ADF-GUI.
Clearly wrong negative temperatures (in Kelvin) could appear in the iterative scheme, typically can happen if the boiling point is low.
ADFinput wrote an invalid hessupd key (with no hessian update method specified) when opening .adf files created with old versions. A simple work-around is to delete the hessupd key from the run script.
There were some issues with copy-pasting in the ADF-GUI: when pasting using a Windows machine the text was pasted twice, and the text field lost focus when moving to the menu. Both have been fixed.
When showing results from a logfile, the final bond energy shown in ADFmovie (in graphs and in the main window) was the final LDA bond energy. The correct bond energies (including GGA) were missing.
This bug was introduced in ADF2007, thus also in ADF2007 the ELF was calculated incorrectly for the SCF solution.
The request frequencies were not written to the input file.
There was an undocumented constraint to freeze a single coordinate of an atom.
Constraints Coord 9 3 :: freeze of atom 9 the third coordinate End
but this effectively froze all three the coordinates.
Bug can show up in case of spin-orbit coupling and small basis sets.
The quality of saved movies in ADFmovie has been improved, using MP4 instead of MPG. Also, if you have a looping movie 10 iterations will be shown for smoother playback.
Semi-empirical (Swart) initial hessian. The default was and is the unit matrix.
GeoOpt InitialHessian Swart End
Also some bugfixes for geometry optimization and frequency run.
Basis set search in the database could lead to a frozen core smaller than intended by the user. For instance: "Core Small" would lead to "Core None". This problem happened when the default installation path contains a dot (which is the default).
On some intel platforms systems with heavy atoms could fail with the error message Dirac: TOO BAD.
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.
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.
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.
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.
This fixes the error Failed to create intermediate z-matrix for delocal coords.
When using constraints during a geometry optimization, the estimation of the Hessian was based on the assumtion that normal rigid motions should be applied. These are now disabled in such a case.
For very large unit cells the Brillouin zone becomes very small. In such cases the relative error could become quite larger, leading to a program termination of BAND. Now it is also tested if the absolute error is significant.
Do not always store the history of density matrices during a geometry optimization. This is now only needed when using a meta GGA functional.
This bug was causing ADF to hang during geometry optimization. It was showing up only when the restart file was supplied with a relative path, which means that the file is broadcast to the slave nodes by the master. Since not everything is broadcast by the master but only necessary information the data needed by the optimizer was available on the master node only.
Running a guess calculation corrupted the integration parameters
In the 2008 release a bug was introduced in the gradient code. Not all systems are affected by it. Only systems with symmetry that have atoms of the same type not being equivalent. A consequence was that (for the systems) affected the energy went up during a geometry optimization.
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.
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.
When a user performes a non-ZORA calculation, but no basis set is present for the requested element, ADF searches also the all-electron ZORA basis sets as alternatives. However, in that situation (only) ADF always used the largest all-electron ZORA basis set, even if a smaller all-electron ZORA basis set is present of the requested quality. This has been fixed.
Fixed an issue on 64-bit Linux that was causing all GUI modules to crash with a "Floating point exception" message.
The alpha and beta density are added and next the Bader analysis is performed.
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.
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.
During a geometry optimization the symmetry of the system could slowly get lost (causing errors in the symmetry detection routines). Now always symmetrize the nuclear gradient.
Calculations combinating ZORA and MetaGGA stopped prematurely due to an error in the run configuration.
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.
Now you can choose to keep the peak height or the peak area fixed when changing the peak width. Also, change of units of the x-axes has no longer any effect on the values shown on the y-axes.
To open those files in a text editor, keep the control key pressed while double-clicking these files.
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.
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.
Search:
Snapshots:
