4.2. Standards and Reference Energies

download:pdf

4.2.1. The MedeA Standard 500

MedeA uses a standard for Precision VASP calculations, automatically applying the directive PREC = Accurate and a plane wave cutoff of 500 eV (900 eV for hard potentials where the PAW potential definition in the POTCAR file has ENMAX > 450 eV). The MedeA VASP GUI option Standard 500 is recommended when an overall precise plane cutoff and setting is required. Using a standard cutoff of 500 eV within a project facilitates the comparison of data from compounds containing different elements like, e.g., in the calculation of heat of formations or defect energies.

VASP parameters covered by Precision Standard 500 :

  • VASP precision is set to accurate
  • Planewave cutoff energy is set to 500 eV (900 eV for hard PAW potentials which have ENMAX > 450 eV)

VASP parameters recommended to be set in addition, to comply with Standard 500 :

  • Reciprocal space projection
  • Convergence criterion for SCF cycle: 10-7 eV
  • Convergence criterion for geometry optimization: 0.001 eV/\({\mathring{\mathrm{A}}}\)
  • Conjugate gradient algorithm for geometry optimization

For crystalline structures:

  • K-spacing of 0.2 \({\mathring{\mathrm{A}}}\)-1, shift origin to \({\Gamma}\)
  • For non-local exchange based functionals (e.g. hybrid functionals): k-spacing of 0.5 \({\mathring{\mathrm{A}}}\)-1, shift origin to \({\Gamma}\), and set X, Y, and Z axis components of the k-mesh factors to 2
  • K-space integration using the tetrahedron method including Bl\({\ddot{\mathrm o}}\)chl corrections

For molecular structures:

  • Box ensuring a minimum of about 8 \({\mathring{\mathrm{A}}}\) empty space between the molecule and its nearest-images over the periodic boundary condition
  • \({\Gamma}\)-point only
  • K-space integration using Fermi smearing with a 0 eV smearing width

4.2.2. Reference Energies for the Calculation of the Heat of Formation

The computation of the electronic contribution to heats of formation for compounds requires reference energies for all constituent elements in their standard state. Below is a list of model structures for all elements in their standard state, which is applied for calculating the energy of formation as a property from the MedeA VASP 6 GUI. For a few cases, where, e.g., van der Waals interactions have to be included, the standard state structures are replaced by other structures that are correctly covered by standard DFT functionals. Correction energies are then used to obtain the energy of the standard state structures.

In structure optimization calculations for the reference systems, in general a non-magnetic Hamiltonian is used, however, spin-polarization is required for O2, Cr (antiferromagnetic), \({\gamma}\)-Mn (antiferromagnetic), Fe, Co, Ni, and lanthanides heavier than Ce for potentials including f electrons as valence states.

For solids, all cell parameters and internal degrees of freedom need to be relaxed. For atoms and molecules a fixed box of 10 \({\mathring{\mathrm{A}}}\) in each dimension is used (for S8 molecules a cubic box of 15 \({\mathring{\mathrm{A}}}\)), for molecules the atomic positions need to be relaxed.

For computing energies of formation, the application of the Standard 500 setting as outlined above is recommended to obtain a suitable accuracy.

Table 50 Summary of models and technical details for the calculations of reference energies of the elements
Elements Structure Correction, techn. details
H H2 molecule  
He He atom  
Li bcc  
Be \({\alpha}\)-Be, hcp  
B \({\alpha}\)-B, R-3m  
C diamond, Fd-3m -1.897 kJ/mol, because graphite is the low temperature phase
N N2 molecule  
O O2 molecule spin-polarized, because molecule exists in triplet state
F F2 molecule  
N Ne atom  
Na bcc  
Mg hcp  
Al fcc  
Si Fd-3m  
P Cmca, black phosphorus  
S S8 molecule -13.04875 kJ/mol, condensation
Cl Cl2 molecule  
Ar Ar atom  
K bcc  
Ca \({\alpha}\)-Ca, fcc  
Sc \({\alpha}\)-Sc, hcp  
Ti \({\alpha}\)-Ti, hcp  
V bcc  
Cr \({\alpha}\)-Cr, bcc, antiferromagnetic spin-polarized
Mn \({\gamma}\)-Mn P4/mmm, antiferromagnetic -4.348 kJ/mol = \({\alpha}\)-Mn, spin-polarized
Fe \({\alpha}\)-Fe, bcc spin-polarized
Co \({\beta}\)-Co, hcp spin-polarized
Ni fcc spin-polarized
Cu fcc  
Zn hcp  
Ga \({\alpha}\)-Ga, Cmca  
Ge Fd-3m  
As R-3m grey metallic
Se P3121 grey metallic
Br Br2 molecule -22.85 kJ/mol, condensation
Kr Kr atom  
Rb bcc  
Sr \({\alpha}\)-Sr, fcc  
Y hcp  
Zr \({\alpha}\)-Zr, hcp  
Nb bcc  
Mo bcc  
Tc hcp  
Ru hcp  
Rh fcc  
Pd fcc  
Ag fcc  
Cd hcp  
In I4/mmm  
Sn \({\alpha}\)-Sn, Fd-3m transition temperature of 286 K to \({\beta}\)-Sn, I41/amd
Sb R-3m  
Te P3121  
I Cmca  
Xe Xe atom  
Cs bcc  
Ba bcc  
La \({\alpha}\)-La, dhcp  
Ce \({\gamma}\)-Ce, fcc  
Pr dhcp  
Nd dhcp  
Pm dhcp  
Sm R-3m  
Eu bcc  
Gd hcp  
Tb hcp  
Dy hcp  
Ho hcp  
Er hcp  
Tm hcp  
Yb fcc  
Lu hcp  
Hf hcp  
Ta \({\alpha}\)-Ta, bcc  
W \({\alpha}\)-W, bcc  
Re hcp  
Os hcp  
Ir fcc  
Pt fcc  
Au fcc  
Hg R-3m 2.3 kJ/mol melting (JANAF)
Tl \({\alpha}\)-Tl, hcp  
Pb fcc  
Bi R-3m  
Po Pm-3m  
At no structure  
Rn Rn atom  
Fr no structure  
Ra bcc  
Ac fcc  
Th \({\alpha}\)-Th, fcc  
Pa I4/mmm  
U \({\alpha}\)-U, Cmcm  
Np Pnma  
Pu \({\alpha}\)-Pu, P121/m1  
Am dhcp  
Cm dhcp  
download:pdf