[CP2K-user] [CP2K:22449] DFT+U geometry optimization of CuAl₂O₄ slab + H₂O never converges — what should I do next?

[双磊]

Hello,

  I am using CP2K 2026.2 to compute the adsorption energy of H₂O on a 
CuAl₂O₄ slab. The system contains 227 atoms (224-atom slab + 3-atom H₂O) 
with PERIODIC XY, DZVP-MOLOPT-SR-GTH, PBE+D3(BJ), UKS, MULTIPLICITY 33, and 
U_eff = 3.0 eV on Cu 3d (L 2). Only the three H₂O atoms are free, giving 9 
degrees of freedom.

Two problems:

   1. 
   
   The warning DFT+U energy contribution is negative appears throughout. 
   The &PLUS_U output shows single-spin d occupations of 9.3–10.4, whereas 
   the physical limit is 5. The total spin is stable at 32.0 and S² = 272.1 
   (ideal 272). I have tried MAGNETIZATION 1.0, MULLIKEN_CHARGES, 
   ENFORCE_OCCUPATION, and U = 2.0, but none removes the warning cleanly.
   2. 
   
   GEO_OPT never reaches GEOMETRY OPTIMIZATION COMPLETED. BFGS collapses, 
   LBFGS reports false convergence, and CG oscillates.
   
My question is: given this situation, what should I do next? Specifically:

   - 
   
   Is there a known working SCF/optimizer setup for DFT+U geometry 
   optimization of transition-metal-oxide slabs with adsorbed molecules?
   - 
   
   Should I accept a looser MAX_FORCE (e.g. 1e-2) and treat the result only 
   as a relative adsorption energy?
   
  I would appreciate a concrete recommendation rather than a full diagnosis 
— I have spent a lot of time on parameter tuning and need to know which 
direction is worth pursuing.

     Below is my calculation input file.Thanks for your advice.

Thank you,
Mike

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cp2k/cb57f0fd-9396-498c-9251-bc035345514en%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20261030/ac1933e6/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: POSCAR
Type: application/octet-stream
Size: 13609 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20261030/ac1933e6/attachment-0002.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ads_final_u3.inp
Type: chemical/x-gamess-input
Size: 2192 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20261030/ac1933e6/attachment-0001.inp>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: coords_227_fixed.inc
Type: application/octet-stream
Size: 14528 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20261030/ac1933e6/attachment-0003.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: coords_227_fixed.xyz
Type: chemical/x-xyz
Size: 13656 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20261030/ac1933e6/attachment-0001.xyz>