[CP2K-user] QM water molecules tend to accumulate at the borders of the QM cell in periodic QM/MM simulation

[Sal]

Dear CP2K developers,


I am running periodic QM/MM simulation (BLYP/TZV2P with GTH-BLYP) that 
includes water, protein and lipid. 

Problem is that few  water  molecules are pushed against the edge even with 
a hard quadratic wall (K=40 kcalmol/angstrom^2).

QM cell dimensions are chosen to be ~2 A larger than QM size.

I am  wondering if you know any specific parameter setting (e. g. # of LIB 
in GEEP, NGRIDS, GPW cutoffs, ...) would cause this.





Thanks for your time,

Sal





More detail about my QM/MM setup:




    &MGRID 

      COMMENSURATE

      &RS_GRID

        DISTRIBUTION_TYPE DISTRIBUTED

      &END RS_GRID

      CUTOFF 360

      NGRIDS 4

      REL_CUTOFF 60

    &END MGRID                                                      


    &QS                                                             

      METHOD GPW                                                      

      EPS_DEFAULT 1.0E-10                                             

      MAP_CONSISTENT TRUE                                           

      EXTRAPOLATION ASPC                                              

      EXTRAPOLATION_ORDER 3                                           

    &END QS                                                         


    &SCF                                                            

      MAX_SCF 30 

       EPS_SCF 1.0E-6

      SCF_GUESS RESTART

#      SCF_GUESS ATOMIC

      &OT T

          MINIMIZER DIIS

         N_DIIS 7

      &END OT

      &OUTER_SCF

         MAX_SCF 3

         EPS_SCF 1.0E-6

      &END OUTER_SCF 


    &XC                                                            

      &XC_FUNCTIONAL BLYP

      &END XC_FUNCTIONAL

    &VDW_POTENTIAL                                                

         POTENTIAL_TYPE  PAIR_POTENTIAL

         &PAIR_POTENTIAL                                                 

           TYPE  DFTD3

           PARAMETER_FILE_NAME  dftd3.dat

           REFERENCE_FUNCTIONAL BLYP

         &END PAIR_POTENTIAL

       &END VDW_POTENTIAL

    &END XC                                                         


      

    

.....

.....

.....


   E_COUPL GAUSS

    USE_GEEP_LIB  9

    &PERIODIC 

       GMAX 1.0

       &MULTIPOLE 

          EWALD_PRECISION 1.0E-6

#          NGRIDS  16 16 24

          ANALYTICAL_GTERM 

          RCUT  14.0

       &END MULTIPOLE

    &END PERIODIC


    &CELL

      ABC [angstrom] 15 16 22


      PERIODIC XYZ 

    &END CELL

    CENTER NEVER

    &WALLS 

      type quadratic

      k [kcalmol/angstrom^2] 40.0

      WALL_SKIN  1.0 1.0 1.0

    &END WALLS 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20181206/3043c89c/attachment.htm>