Published: 31 March 2021| Version 1 | DOI: 10.17632/w2mxnsgwgt.1
Kirill Migdal,


Here we saved the input files used for ABINIT, VASP, and WIEN2k calculations of orthorhombic, bcc, bct and hcp uranium. Also, we provided data for the charge-density distributions in the form proposed by the VASP developers. We have tried to publish input files for all points with different lattice parameters. In rare cases, we have few examples of input files used for a calculation of a fixed lattice by one of three calculation packages. The list of the VASP input files contains INCAR, KPOINTS and POSCAR. We do not provide POTCAR which is also need to reproduce our data due to the license restrictions. As for the input files used for our ABINIT calculations, we include here *.in and *.files files. Again, we have no opportunity to publish the GTH pseudopotential which is designated in *.files as "U-q14". We carried out calculations with the use of this pseudopotential thanks to kind permission of Vladimir Stegailov and Grigory Smirnov. You may ask them if you interested in investigations based on the use of this pseudopotential. In the case of the data for WIEN2k calculations, we have no such limitations. Here, we present structural file (*.struct), k-points grid file (*.klist) and both files (*.in1 and *.in2) used to implement an initial guess for electronic levels and some important parameters as Rmt*Kmax. Some general parameters used for all lattice structures in our WIEN2k calculations you can find in the file "d" (see the folder 'WIEN2k scripts'). Another part of this database is the data for charge-density distributions. All data are given here in the form accepted by the VASP developers (CHARGCAR). Initial data of WIEN2k (*.rho3d) and ABINIT (*.xf obtained by the use of cut3d utility) calculations were translated using the Python scripts written by the authors. We have provided these scripts too. The supercells we used for heated uranium were obtained previously as discussed in the article referenced below (DOI:10.1134/S0018151X17050121).


Steps to reproduce

In all cases of calculations for bcc uranium only one step is necessary. Generally speaking, calculations for the other 3 types of uranium lattices should be carried out in two steps because we have two unknown lattice parameters for each fixed density. In some cases we have provided two sets of input files for a first and a second stage of a calculation with a certain density and lattice type (see the folders "step_1" and "step_2") . In all other cases the data provided as input files correspond to the previously found minima of the total energy. Thus, the first stage of a search of stable lattice parameters may be omitted here. If your aim is to reproduce the charge-density data presented here, we may ask you to pay attention to the dimensions of a FFT grid in addition to the main list of parameters (k-grid`s dimensions , a cutoff energy, a number of empty levels). There are some differencies between plotting of charge-density distribution in VASP, ABINIT and WIEN2k because dimensions of charge-density array written to *.rho3d and *xsf files can be more than dimensions provided in the input files. For example, if the ABINIT parameter "nfft 3*100" , the real dimension of the array in the *.xsf file (don`t forget about cut3d!) is 101x101x101. Result of a VASP calculation written to a CHGCAR have the same dimensions as given in the corresponding INCAR. You need to use Python3 to run the scripts used for translation of ABINIT *.xsf or WIEN2k *rho3d files. Only "math" library is required by these scripts. The name of translated *.xsf or *rho3d file should be noted in the main procedure of these scripts. The files "d" or "d.txt" can be useful to run the command "init_lapw" with the option "-b". You may need to change your "init_lapw" script which is the part of WIEN2k package to specify a new k-grid by each dimension. The last but not least is to remember to use the Fermi energies from the file "efermi-for_U-in_w2k.txt" instead of values written to the *.in1 files by default. The usage of the default Efermi value (0.5 Ry) for the WIEN2k calculations often leads to the problem with determination of electronic levels at high compressions of uranium. The behavior of Fermi energy proposed in the file "efermi-for_U-in_w2k.txt" is defined by the simple relation for ideal Fermi-gas (Ef ~ (electron density)^(2/3)).


Moskovskij fiziko-tehniceskij institut


Materials Science, Actinide, Density Functional Theory Approaches