The SCF doesnot converge, using diagonalization

Zhendong Guo zhendon... at
Sat Nov 29 21:09:19 UTC 2014

Dear Everyone
    I am  trying to calculate the eigenvalue of LUMO and HOMO for 
kappa-Al2O3. The system owns the periodic boundary condition and includes 
80 atoms.
I tried using the following in the section of DFT to output the eigenvalues 
of MOs. 
        &END MO
         NHOMO 5
         NLUMO 5
        &END MO_CUBES" 

>From the tutorials of Cp2K, I notice:
Section MO 

*Controls the printing of the molecular orbitals.Note that this is only 
functional with diagonalization based methods, in particular not with OT*So, 
I turn "OT" off in my input file and use the following setting in the 
section of SCF   
      SCF_GUESS                ATOMIC 
      EPS_SCF                     1.0E-7
      MAX_SCF                     300  
      MAX_ITER_LUMOS       10000      
      EPS_LUMOS                1.0E-7
      &MIXING  T
        ALPHA 0.4
        NBROYDEN 8

*      &END MIXING*But, I found the SCF calculation didnot converge after 
300 steps when I checked the output files.

*So, Could you kindly give me some advise to make the SCF calculation 
converge correctly?*

*The input file and the xyz file for the system are also attached, please 
help me check them when you are free.zhendong*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: k-Al2O3.inp
Type: chemical/x-gamess-input
Size: 2897 bytes
Desc: not available
URL: <>
-------------- next part --------------
A non-text attachment was scrubbed...
Type: chemical/x-xyz
Size: 4427 bytes
Desc: not available
URL: <>

More information about the CP2K-user mailing list