Should I do DFT+U approach in AIMD tor transition metal??

[Seunghwan_Kwon]

Should I do DFT+U approach in AIMD tor transition metal??

For example, in the case of NiO AIMD (ab initio molecular dynamics) simulation, should we include DFT+U?

Also, if there are multiple transition metal compounds in AIMD simulation, how do we set U parameter?

For instance, if NiO and NiC are formed in AIMD, there will be multiple U parameter for each compound.

Then, how do we set U parameter?

[jonathan_lahnsteiner2]

Dear Seunghwan_Kwon,

If you want to use DFT+U in AIMD, you first have to determine the U parameter.
There is a tutorial on the vasp wiki discussing how to obtain the U parameter for NiO
wiki/index.php/Calculate_U_for_LSDA%2BU
If you have several transition metal compounds, you have to do this step for all transition metal compounds in your system.
For the example you are stating, NiO and NiC you would have to the the determination for both Ni's, the NiO and NiC and treat them
as different species in your POSCAR and POTCAR.
If DFT+U is needed for your AIMD simulation strongly depends on the system and what properties of the
system you want to study. This can not be answered in general.

I hope this helps to answer your question

All the best Jonathan