We know that the transition metal oxides, like CuO are difficult to model using DFT and hence we use DFT+U with an appropriate U value to correctly describe the electronic structure of the transition metal oxides. You have suggested an U value of 4.0 for Copper in copper oxides in your previous papers like “A. Jain, G. Hautier, S.P. Ong, C. Moore, C.C. Fischer, K.A. Persson, G. Ceder, Formation Enthalpies by Mixing GGA and GGA+U calculations, Physical Review B, vol. 84 (2011), 045115”.
But, why have the oxides of copper been calculated with GGA and not with GGA+U here in the materials project?
I appreciate your help. Thanks!
Thank you for bringing this to our attention. The MP reaction calculator shows a bad Cu2O + O -> CuO reaction enthalpy wrt experiment, which would justify the use of +U.
Cu2O is Cu1+, which is a d10 system. Technically(…) one should not need to use +U for full shell d orbitals. Perhaps this is why MP did not add +U for Cu to begin with, i.e., whomever was doing the calibration required two oxides both with partially filled d shells.
We will look into re-running copper oxide calculations with GGA+U.
Other oxides of copper like CuAl2O4 with Cu2+ has partially filled d orbitals though. Just to let you know that there is no correction for copper in your current +U and oxide corrections.
Thank you for your reply! I will be waiting for your results. Great initiative and website by the way!