My Community

Dear users,

I am calculating the cohesive energy of a series of transition metal (TM) oxides and I have the following questions: because the energy of the TMO is calculated within the LDA+U framework should I use the same LDA+U for the free atoms? Does it make sense to subtract LDA+U/GGA+U energies of oxides from LDA/GGA energies of free atoms?
Thanks,
Zsolt

Dear Zsolt,


I'm not sure about calculating the cohesive energy of TM oxides starting from energies of free atoms. But closely related is calculating the formation enthalpy from O2 energy and the energy of the transition metal. Prof. Ceder group addressed the issues related to this in several papers. Key ones are:

1) O2 molecule overbinding:

http://journals.aps.org/prb/abstract/10 ... .73.195107

2) DFT vs. DFT+U (DFT is preferred for the metal but DFT+U is preferred for the oxide):

http://journals.aps.org/prb/abstract/10 ... .84.045115

Other groups might have other published work which I'm unaware of.