COSMO in NWCHEM 6.5


Just Got Here
I have the following input file:

start aspartame

memory total 128000 mb

permanent_dir /work2/quantum/sarom-ssok1/restart
scratch_dir /local/scratch/sarom-ssok1

title aspartame-CCD

basis spherical print segment
 * library cc-pVDZ
end

scf
 rhf
 singlet
 diis
 diisbas 10
 direct
 maxiter 200
end

geometry units angstroms print xyz
 symmetry c1
  O       -1.29008        2.50910       -1.16264
  O        2.02006        0.01437       -2.49609
  O        0.38582        1.80735        0.09905
  O        2.65897        0.40477        2.61855
  O        1.30169       -1.12556        1.80314
  N        0.42248       -0.65567       -1.05616
  N        3.91674       -1.69196       -1.45480
  C       -0.54562        0.30039       -1.52981
  C       -1.97681       -0.26864       -1.55017
  C        2.75433       -1.32890       -0.67164
  C       -2.53703       -0.67976       -0.20275
  C        1.70427       -0.58683       -1.50619
  C       -0.41752        1.60398       -0.75065
  C        3.20259       -0.44586        0.50238
  C       -2.34784       -1.97047        0.28682
  C       -3.26134        0.22242        0.57506
  C       -2.85568       -2.34651        1.52240
  C       -3.77183       -0.15086        1.81011
  C        2.26666       -0.43825        1.67887
  C       -3.56856       -1.43698        2.28955
  C       -1.22453        3.78977       -0.55463
  H       -0.27836        0.54910       -2.55434
  H       -2.62712        0.47667       -2.00486
  H       -1.95972       -1.13146       -2.21625
  H        0.31100       -0.97230       -0.11156
  H        2.31114       -2.23665       -0.26328
  H        4.16053       -0.82375        0.86295
  H        3.37008        0.58635        0.19166
  H       -1.79995       -2.69186       -0.30566
  H       -3.43502        1.22497        0.20646
  H        3.65340       -2.37685       -2.14328
  H        4.20417       -0.88829       -1.99052
  H       -2.69527       -3.35351        1.88410
  H       -4.33430        0.56381        2.39661
  H       -3.96731       -1.72967        3.25171
  H       -1.40547        3.71359        0.51543
  H       -2.00042        4.38290       -1.02879
  H       -0.24873        4.24042       -0.72201
  H        2.02518        0.37002        3.32684
end

task scf energy


which seems to produce a run with COSMO solvation.

 ----------------------------------------------
                DIIS ROHF

 Convergence threshold     :          1.000E-04
 Maximum no. of iterations :          200
 Integral*density screening:          1.000E-07
 DIIS subspace             :           10

 ----------------------------------------------

     COSMO solvation phase
         1  0       -1023.594128336433       0.25E+01  0.90E+00
         2  0       -1023.878365421912       0.16E+01  0.24E+00
         3  1       -1023.939783864741       0.12E+01  0.10E+00
         4  2       -1023.964399749986       0.91E+00  0.88E-01
         5  3       -1024.001853116476       0.47E-01  0.95E-02
         6  4       -1024.002023141146       0.13E-01  0.18E-02
         7  5       -1024.002046521194       0.66E-02  0.42E-03
         8  6       -1024.002053294979       0.23E-02  0.38E-03
         9  7       -1024.002054180684       0.11E-02  0.76E-04
        10  8       -1024.002054294537       0.25E-03  0.20E-04


       Final RHF  results 
       ------------------ 

         Total SCF energy =  -1024.002054309373
      One-electron energy =  -4930.060794326499
      Two-electron energy =   2130.888794026391
 Nuclear repulsion energy =   1775.169945990735


This happens for NWCHEM 6.5 with all the new patches applied and NWCHEM 6.3

Even if I add:

cosmo
 off
end


It still produces a run with COSMO solvation.

How do I produce a pure gas phase run without any form of solvation?

Forum Vet
Thank you very much for this bug report.
The bug surfaces when you use DIIS in the SCF modules.
Therefore, if you remove the two options diis and diisbas 10, you will get a gas phase calculation.
By the way, using the DIIS scheme in the SCF module is not an approach I would recommend.

Just Got Here
Thank you Edo.

Do you mind telling me when the DIIS scheme is appropriate?


Forum >> NWChem's corner >> Running NWChem