[QE-users] sorry for the n! problem with the SCAN functional

[zhouchao]

Dear QE Developer,


I apologize for bothering you with a question that has likely been asked many times on this forum. I am running into the following error message with the input_dft='scan' interface using the libxc library: "S matrix not positive definite or / the charge is wrong". I have searched the forum for solutions to this issue and have found that the problem does not arise when using the PBE functional.


Unfortunately, my goal is to accurately describe the structure and energy of my system using the SCAN functional. I have attempted to modify the encut energy and diagonalization methods with no success. I am using ONCV norm conserving pseudopotentials, and have included my input file below for reference.


&CONTROL


    calculation='vc-relax', disk_io='low', prefix='CrB',


    pseudo_dir='./', outdir='./tmp', verbosity='high'


    tprnfor=.true., tstress=.true., forc_conv_thr=1.0d-5


/


&SYSTEM


    ibrav= 0,


    celldm(1) = 1.8897261328856432, ! a.u. to Angst


    nat= 3, ntyp= 2,


    input_dft='scan0',


    occupations = 'smearing', smearing = 'gauss', degauss = 0.03


    ecutwfc= 150, ecutrho = 600,


/


&ELECTRONS


    conv_thr = 1.0d-9


    mixing_beta = 0.8


    diagonalization = 'ppcg'


/


 &IONS


    ion_dynamics = 'bfgs'


    upscale = 20


/


 &CELL


    cell_dynamics = 'bfgs'


    press = 0


/


ATOMIC_SPECIES


  Cr 51.966 Cr.UPF


  B  10.81 B.UPF


CELL_PARAMETERS (alat=  1.88972613)


     2.904617312   0.000000000   0.000000000


  -1.452308656   2.515472381   0.000000000


  -0.000000000  -0.000000000   2.825090285


ATOMIC_POSITIONS (crystal)


Cr 0  0 0


B 0.3333333429999996  0.6666666870000029  0.5000000000000000


B 0.6666666269999979  0.3333333129999971  0.5000000000000000


K_POINTS {automatic}


  20 20 12 0 0 0






I would greatly appreciate any advice or guidance you can offer on how to address this issue and successfully use the SCAN functional to accurately describe my system's structure and energy.


Thank you for your time and expertise.


Best regards,


zhouchao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20230617/e60fc579/attachment.html>