Notes on using a .cif file

From FEFF

The ‘*.cif’ file import feature is currently only implemented for k-space calculations of crys- tals. We intend to enable it for molecules also; please contact us for collaboration if you have an interest in this. feff uses the CIFTBX library to read ‘.cif’ files. This should allow you to use any valid ‘.cif’ file. ”Warning - library not found” warnings at runtime can usually be ignored. The Open Crystallography Database is an excellent resource for ‘.cif’ files that can be accessed from anywhere. The Inorganic Crystal Structure Database is another good resource for those with an academic subscription. We recommend saving the ‘.cif’ file to the working directory containing also ‘feff.inp’.

Use the CIF CARD as in the Cr2GeC example below. You cannot use an ATOMS card in the same ‘feff.inp’ file. It is no longer necessary to use a POTENTIALS card. feff automatically assigns a potential type to each crystallographically inequivalent atom in the unit cell. These potentials are given the default angular cutoffs (lmax) based on atomic number.  Inspecting ‘pot.inp’ or ‘log1.dat’ shows the list of assigned potentials. Note that the current scheme could be problematic for large unit cells, where there might be many crystallographically inequivalent atoms of the same atomic number. Calculating a large number of potentials will make the calculation slow and potentially unstable (the SCF algorithm may have a hard time reaching convergence). Furthermore, in our experience respecting strict crystallographic equivalence does not always improve the accuracy of a calculation. For such large systems, it may be preferable to make all atoms of the same atomic number equivalent, or to determine equivalence based on first-shell coordination only. Please contact us; we will gladly implement such functionality. If both a CIF card and a POTENTIALS card are present in ‘feff.inp’, the program does the following. First, it determines if the list of potentials in the POTENTIALS card matches exactly that generated from the CIF file. If it does not match, the POTENTIALS card is ignored and a warning is printed to the screen. If the list of potentials does match, the program takes the options for the potentials (i.e., lmax1 lmax2 xnatph spinph ) and uses these values instead of the defaults generated from the CIF file. This way, the user can still control the angular momentum cutoff and set the spin-related variables.

Finally, a word of warning on using the TARGET card to indicate the corehole atom. Combined with the CIF card, this counts an atom in a list of the crystallographically inequivalent atoms in the unit cell (e.g., 2 C atoms for graphite) in the order given in the CIF file. However, combined with LATTICE/ATOMS cards, it counts an atom in a list of all atoms in the unit cells (e.g., 4 C atoms for graphite, 2 of which are equivalent by symmetry to the other 2) in the order given in the ATOMS card. Hence, the value of the TARGET card must be reevaluated if you switch from one representation to the other. Note that the JFEFF GUI currently does not display the content of the CIF file; you must look at the file yourself to find out in what order the atoms are listed and identify which one you want to place the core hole on. If it is the third atom listed in the CIF file, use ”TARGET 3” in ‘feff.inp’ or in the JFEFF GUI.

developer's resources