Page 1 of 1

The energy calculated using the MSRPBEl functional is a large positive value.

Posted: Sun Nov 02, 2025 3:23 am
by tianhui_liu

Dear All,

I used the MSRPBEl function to calculate the energy of Au, and obtained a value of 1719 eV. This value appears unusually large, so I would like to ask for possible explanations. For comparison, the energy calculated using the PBE function is –110 eV. The relevant files have been uploaded for reference.

Many thanks in advance.
Zi Wang


Re: The energy calculated using the MSRPBEl functional is a large positive value.

Posted: Mon Nov 03, 2025 11:10 am
by marie-therese.huebsch

The total energy (TOTEN) alone is completely meaningless. It can be positive or negative, small or large since it is given in comparison to a fictisious system: the sum of the isolated atoms evaluated with PBE (for the PBE potentials or CA for the LDA potententials). That is the reference, because this is how the pseudopotentials are created. But these technical details are maybe not the most relevant for you, so I want to try to provide more practical advise.

You can compare the total energy of two structures. In your case, you may want to relax the ionic positions of the Au surface with a defect. Here is a useful page about surface calculations: Surfaces, thin films, and 2D materials

You can also first do a simpler calculation and compute the volume vs total energy for bulk Au using PBE and MSRPBEl and do an equation of state fit to obtain the equilibrium lattice constant and bulk modulus. For that you should

  • Prepare >=4 structures with different volume using the scaling factor in the second line of the POSCAR file

  • Given the total energies and volumes you can fit the equatiuon of state of your choice, e.g., using pymatgen.analysis.eos

This should help to gain trust in MSRPBEl if the lattice constant and bulk modulus are resonable compared to experiment.

Does this answer your question?


Re: The energy calculated using the MSRPBEl functional is a large positive value.

Posted: Wed Nov 05, 2025 3:50 pm
by tianhui_liu

Dear Dr.Marie-Therese Huebsch,

Thank you very much for your kind and patient reply. The Python library you recommended works perfectly and has helped me resolve my issue. I truly appreciate your assistance.

Best regards,
Zi Wang


Re: The energy calculated using the MSRPBEl functional is a large positive value.

Posted: Thu Nov 06, 2025 7:47 am
by marie-therese.huebsch

Great, I am glad to hear it. I will lock this topic now. Feel free to open a new thread if you have other questions.

All the best!