Besides prof. Postnikov advices, if you are trying to compare your results with previous calculations, you need to check:
- the slab structure (check the plane/direction for the slab) - the structure optimization and/or constraints (there are different ways to use and define the slab behavior) - the type of pseudopotentials - the size of the basis set - the convergence thresholds - the cut-off values - the initial position of the B atom - parameters that can affect the calculated energy values (electronic temperature, mixer method/weight/etc.) []'s, Camps On Thu, Jan 15, 2026 at 6:00 PM Andrei Postnikov < [email protected]> wrote: > Dear Roland, > some suggestions: > 1. Check the structure. It is difficult to judge from your input file; > make a visualisation from working XV in order to see that everything is > correct. > From my experience, surprises due to structure input errors are not > uncommon. > 2. The 3x3 lateral cell size seems rather small to simulate adsorption > of an isolated atom. In principle this might be a factor > responsible for a difference from the expected value. > Ideally, a convergence with respect to supercell size has to be tested. > 3. As a reference energy for desorbed case, move the boron atom away from > the surface > within the same cell, retaining the Cu atoms at their positions. This will > minimize > systematic errors. Check the BSSE later on. > 4. The relaxation at the surface > with and without the boron atom adsorbed might be different. Again, > the lateral size might be too small for correctly incorporating the > relaxation > around the adsorbed atom. (This is just a guess; I don't know the system). > > Good luck > > Andrei > > > to get the adsorption energy, the boron energy from boron crystal is not > a good reference. I'd suggest > > > ----- Le 14 Jan 26, à 9:59, Roland Coratger [email protected] a > écrit : > > > Dear all, > > > > I am trying, as a training exercise, to recover the adsorption energy of > > a boron atom on a Cu(111) slab, which according to the literature should > > be around -2 eV. The energy is given by: E(ads) = E(slab+B) - E(slab) - > > E(B). For E(B), if I use a B atom in the slab’s box, the energy is very > > negative and unrealistic (around -4 eV). If I use the energy of a B atom > > from the 3D boron crystal, the energy becomes positive (around +2 eV), > > so there is no adsorption. Below you will find my input file for the > > slab+B system. I use the same parameters for the other two energies. The > > BSSE correction (a few tenths of an eV) does not change the observed > > trend. Am I making a mistake somewhere and/or do you have any > > suggestions to help me recover the correct value? > > > > Thank you in advance for you help. > > > > Regards, > > > > Roland. > > > > _______________________________________ > > SystemName CuB test > > SystemLabel cu_b > > NumberOfAtoms 46 > > NumberOfSpecies 2 > > > > XC.functional GGA > > XC.authors PBE > > > > MaxSCFIterations 200 > > > > %block ChemicalSpeciesLabel > > > > 1 29 Cu # Species index, atomic number, species label > > 2 5 B # Species index, atomic number, species label > > > > %endblock ChemicalSpeciesLabel > > > > PAO.FixSplitTable T > > PAO.EnergyShift 20 meV > > PAO.SplitNorm 0.15 > > MeshCutoff 300.000000 Ry > > ElectronicTemperature 50.000000 K > > > > # > > MD.TypeOfRun CG # Broyden also possible > > MD.NumCGsteps 200 > > > > # > > SolutionMethod diagon > > SCF.DM.Converge true # Converge SCF step wrt density > > matrix (default: 1e-4) > > SCF.H.Converge true > > DM.NumberPulay 3 > > DM.History.Depth 3 > > > > #SCF Mixer -> Density pour les systèmes difficiles > > > > SCF.Mix Hamiltonian > > > > # Mixer 0.5 reduit le nombre de pas pour des systèmes faciles > > # Mixer 0.001 augmente le nombre de pas pour des systèmes difficiles > > > > SCF.Mixer.Weight 0.05 > > SCF.Mixer.History 6 > > SCF.Mixer.Method Pulay > > MaxSCFIterations 100 > > > > SCF.DM.Tolerance 5.0E-5 eV > > SCF.H.Tolerance 0.0005 eV > > > > > > MD.MaxStressTol 0.0025 eV/Ang**3 > > > > # Nouvelle ligne pour la force entre atomes > > > > MD.MaxForceTol 0.01 eV/Ang > > > > > > # Use old data to save time > > MD.UseSaveXV > > MD.UseSaveDM > > > > # Save atomic coordinates at each step > > WriteCoorStep .true. > > WriteMDHistory .true. > > > > > > PAO.BasisType split > > PAO.BasisSize DZP > > > > LatticeConstant 1.0000 Ang > > > > %block LatticeVectors > > 7.65797 0.00000 0.00000 > > 3.82898 6.63199 0.00000 > > 0.00000 0.00000 24.00000 > > %endblock LatticeVectors > > > > AtomicCoordinatesFormat Ang > > > > %block AtomicCoordinatesAndAtomicSpecies > > > > 3.829 0.7369 1.80 2 # Atome de B en site > cfc > > > > 0.0 0.0 0.0 1 > > 1.2763 2.2107 0.0 1 > > 2.5527 4.4213 0.0 1 > > 2.5527 0.0 0.0 1 > > 3.829 2.2107 0.0 1 > > 5.1053 4.4213 0.0 1 > > 5.1053 0.0 0.0 1 > > 6.3816 2.2107 0.0 1 > > 7.658 4.4213 0.0 1 > > > > 0.0 1.4738 -2.0842 1 > > 1.2763 3.6844 -2.0842 1 > > 2.5527 5.8951 -2.0842 1 > > 2.5527 1.4738 -2.0842 1 > > 3.829 3.6844 -2.0842 1 > > 5.1053 5.8951 -2.0842 1 > > 5.1053 1.4738 -2.0842 1 > > 6.3816 3.6844 -2.0842 1 > > 7.658 5.8951 -2.0842 1 > > > > 1.2763 0.7369 -4.1685 1 > > 2.5527 2.9476 -4.1685 1 > > 3.829 5.1582 -4.1685 1 > > 3.829 0.7369 -4.1685 1 > > 5.1053 2.9476 -4.1685 1 > > 6.3816 5.1582 -4.1685 1 > > 6.3816 0.7369 -4.1685 1 > > 7.658 2.9476 -4.1685 1 > > 8.9343 5.1582 -4.1685 1 > > > > 0.0 0.0 -6.2527 1 > > 1.2763 2.2107 -6.2527 1 > > 2.5527 4.4213 -6.2527 1 > > 2.5527 0.0 -6.2527 1 > > 3.829 2.2107 -6.2527 1 > > 5.1053 4.4213 -6.2527 1 > > 5.1053 0.0 -6.2527 1 > > 6.3816 2.2107 -6.2527 1 > > 7.658 4.4213 -6.2527 1 > > > > 0.0 1.4738 -8.3369 1 > > 1.2763 3.6844 -8.3369 1 > > 2.5527 5.8951 -8.3369 1 > > 2.5527 1.4738 -8.3369 1 > > 3.829 3.6844 -8.3369 1 > > 5.1053 5.8951 -8.3369 1 > > 5.1053 1.4738 -8.3369 1 > > 6.3816 3.6844 -8.3369 1 > > 7.658 5.8951 -8.3369 1 > > > > %endblock AtomicCoordinatesAndAtomicSpecies > > > > %block kgrid_Monkhorst_Pack > > 12 0 0 0. > > 0 12 0 0. > > 0 0 1 0. > > %endblock kgrid_Monkhorst_Pack > > > > SaveTotalPotential T > > SaveTotalCharge T > > SaveElectrostaticPotential T > > -- > SIESTA is supported by the Spanish Research Agency (AEI) and by the > European H2020 MaX Centre of Excellence > (https://urldefense.com/v3/__http://www.max-centre.eu/__;!!D9dNQwwGXtA!StckWOSOOjpmsvjRweSqVEWqzGdIdPNy3eQF6OMu8DaMBJ1iNwSlSTbn0KsoLz3n_tNZZs7qzsBH$ > ) >
