XML-documentation elements

The whole documentation is appended to the respective SK file. It starts (and ends) with the Documentation element and is divided into four main sections: General, SK_table, E_rep and Misc. In the following the different fields in these sections are explained in detail:

  • General

    • Identifier

      String used to identify the SK file. Should be (but not necessarily is) equal to the name of the file. The naming convention {Element}-{Element}.skf is highly recommended. (A file for the interaction of Gold with Oxygen would be named Au-O.skf).

    • Author

      Author or list of authors of this SK file. You may add your email address.

    • Creation

      Rough date of creation.

    • Element

      Give element symbols not names here. For a diatomic file there are two elements (Element1, Element2). For monoatomic files only one element is required (Element1).

    • Md5sum

      MD5 checksum of SK file without documentation. This value can be obtained by issuing the command awk '$1~"<Docu" {x=1} x!=1 {print}' <SK-file> | md5sum for the complete file. Leave blank if you do not know what you should do.

    • Svn_Id

      Identifier of the file within the SVN repository. Just enter $Id:$ here, the expansion will be performed by SVN.

    • Svn_URL

      Location of the file within the SVN repository. Just enter $URL:$ here, the expansion will be performed by SVN.

    • Compatibility

      List of SK files the present one is compatible with. This includes not only the files which were used to fit the present interaction but generally all files which are known to work with the present one. Mono- and diatomic files a treated differently.

      • Monoatomic files: Let us assume the interaction file A-A for atom A is to be documented. In the files A-C and A-D the parameters (compression radii, functional, scc/non-scc) for atom A are exactly like in A-A. Then the files A-C and A-D (but not the file C-D) need to appear in the list.

      • Diatomic files: The interaction file A-B for atoms A and B is to be documented. Here only files A-A and B-B are listed. (See also this). The identifiers of the files in the list are attributes of the Partner element which carries the following information:

        • Element: Exactly as above.

        • Md5sum: Exactly as above.

        • SvnLoc: Location of this file in the SVN repository. If unknown, this field is filled out by the Slater-Koster administrator.

  • SK_table

    • Code

      Code to generate the SK tables. Could be either twocnt or lcao or rlcao (relativistic lcao) by Prof. Seifert.

    • Functional

      Exchange-correlation functional used.

    • Superposition

      Superposition mode. One of dense (density superposition) and pot (potential superposition).

    • Basis

      Appears once for a monoatomic file, twice for a diatomic one. Do not forget to set the attribute atom correctly.

      • Shells

        List of shells included as LCAO basis. List items are separated by white space. Do not use comma, semicolon or alike.

      • Exponents

        Exponents used in the expansion of atomic orbitals in Slater functions.

      • Power

        Maximum power n in the term \(r(l+n)\) found in the expansion of atomic orbitals in Slater functions. (Here it is implicitly assumed that the same number of exponents and also the same n is used for every angular momentum. If this is not the case, please add a comment in the Notes.)

      • Potential

        Power n of the compression potential \((r/r0)^n\).

      • Density

        Compression radius of the density.

      • Wavefunction

        Compression radii of the LCAO wavefunctions in the order given by Shells. List items are separated by white space. Do not use comma, semicolon or alike.

    • Hubbard

      Only for monoatomic files. One of atom,ang, td and none. If atom, only the s-Type Hubbard parameter is meaningful and was used in the parameterization. If ang, all three Hubbard values are meaningful and were used in the parameterization. If td, p/d-Hubbards are modified to allow for calculation of excited states. If none, no Hubbards are present.

  • E_rep

    • Storage

      Either spline if the repulsive potential is stored as a spline, or poly if it is stored as a polynomial.

    • SCC

      Either yes if charge self-consistency was used in the fitting procedure or no if not.

    • Ab_initio

      The first entry in this element refers to the code (e.g. Gaussian, NRLMOL…) used to generate the reference data for the repulsive potential. The second entry is given in pretty free format as XC-functional/basis representation, where the basis representation can be the name of standard LCAO basis, the name of a predefined basis (i.e. keyword) in the code you use, a plane wave cutoff or a grid parameter. If you don’t use predefined basis sets, provide a rough description (e.g. if it contains a split valence, polarization and/or diffuse functions) in the Notes.

    • Fit_systems

      Molecules or bulk systems used in the fitting of the present interaction. Can be a list, if e.g. single and double bonds were fitted separately.

  • Misc

    • Accuracy

      Results of tests performed with this SK file or the set to which this file belongs (e.g. Boron files

    • Failures

      Known failures of this SK file or the set to which this file belongs (e.g. Boron files).

    • Publications

      • Fit

        A publication which describes the fitting of this SK file or the set to which this file belongs. A diploma/PhD thesis or a link to a document on the WWW are also fine.

      • Other

        Other publications in which the present file was used. Multiple Other elements are marked with the id attribute.

    • Notes

      Any information which does not fit in the above fields but still is important. E.g: In conjunction with which problem was this file created? Does it provide improvements over existing files? Are there any particularities in the fitting procedure, any non standard approaches? Does this file contain additional information like dipole matrix elements, wave function coefficients or spin Hubbards?