computation does not yet treat fractional occupied orbitals

[yulei wang]

Hi, 

I want to ask whether MUTIPLICITY can be used with SMEAR? When I use them together, the WARNING appears as ‘WARNING: S**2 computation does not yet treat fractional occupied orbitals’. Can anyone give me some suggestions to deal with it. And here are my input file and out file. 

Thanks very much.

Best Regards,

Wang

[Marcella Iannuzzi]

Dear Wang

you can fix the number of electrons per spin channel when using smearing by activating the keyword

FIXED_MAGNETIC_MOMENT

in the section SMEAR. 

Whether for metallic systems this physically makes sense is to be demonstrated. 

The states around the Fermi energy of metallic systems would anyway show fractional occupation, which prevents the S**2 computation

If your system is not metallic, do not use SMAER. 

Kind regards,

Marcella

[yulei wang]

Dear Marcella,

 

Thanks for your reply. I am studying the adsorption of nitrogen molecule on Ru surface.

 So, under these circumstances should I add SMEAR or not? And if I should, how can I setting FIXED_MAGNETIC_MOMENT in the case of Involving two different kinds of atoms?

 

Best Regards,

Wang

[Marcella Iannuzzi]

Dear Wang,

yes, you should use smearing. 

The multiplicity is a global property and it is not assigned per atom or per kind. 

Anyway, I would first converge the electronic structure without imposing constraints. 

In principle the resulting electronic distribution should correspond to the DFT minimum energy. 

Kind regards

Marcella