[QE-users] pw.x - choosing starting_magnetization value for specific expected total magnetization

[Guido Fratesi]

Dear Abdul,

the starting magnetization is here defined as (N_UP-N_DW)/(N_UP+N_DW) so 
a value of 1 means fully polarized, N_UP=N and N_DW=0.

Eg if you want 3muB for an atom having 8 electrons you need to put a 
value of 3/8.

To obtain the various minima you may have to sample various initial 
configurations. It's not at all guaranteed that the calculation will 
keep within the basin that you want, and you may need to reduce the 
mixing beta, or possibly to constraint the magnetization during the SCF 
cycle, and possibly run a following calculation restarting with that 
density but without the constraint.

Are you using DFT+U? If not, adding a very small U (1e-5) and setting 
verbosity=high will not change the solution but allow you to track the 
occupation of the atomic orbitals in detail during the SCF loop.

HTH,
Guido


-- 
Guido Fratesi

Associate Professor
Physics Department "Aldo Pontremoli"
University of Milan
Via Celoria 16, 20133 Milan, Italy

Phone: +39 02 503 17348
email: guido.fratesi at fisica.unimi.it
web:   https://sites.google.com/site/guidofratesi/