calculations not reaching convergence criteria
Just Got Here
9:53:48 AM PST - Tue, Feb 19th 2013
I have problems with my calculations. a few days ago geometrynya wrong, then after I can confirm I got a problem regarding the geometry convergen. following the error message:
For further details see manual section:
17:17:scf_gradient: scf energy failed:: 0
scf_gradient: scf energy failed        0

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

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

 current input line : 

    0: 

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

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

This type of error is most commonly associated with calculations not reaching convergence criteria

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

For more information see the NWChem manual at http://nwchemgit.github.io/index.php/NWChem_Documentation


And this is my input file :
start cuglisin
echo
title "cuglisin"

charge 2

geometry units angstroms
zmatrix
Cu1
N2 Cu1 2.17111
O3 Cu1 2.12250 N2 96.84716
C4 O3 1.77122 Cu1 97.13324 N2 344.56316
C5 N2 1.48375 Cu1 108.43496 O3 23.99584
O6 Cu1 2.03184 N2 109.32959 O3 226.20577
C7 O6 1.50955 Cu1 91.29682 N2 227.08930
N8 Cu1 2.08117 N2 123.06724 O3 114.96379
C9 N8 1.52207 Cu1 110.95727 N2 116.04611
H10 C5 1.12074 N2 110.69817 Cu1 246.43302
H11 C5 1.16053 N2 107.23551 Cu1 87.31678
H12 C9 1.43839 N8 105.06279 Cu1 257.52503
H13 C9 1.16881 N8 95.49391 Cu1 126.93195
O14 C4 1.27935 O3 123.74053 Cu1 194.97418
O15 C7 1.17281 O6 108.98918 Cu1 180.79760
H16 N2 1.48603 Cu1 161.79963 O3 120.34326
H17 N8 1.59571 Cu1 138.85009 N2 304.89306
end
end

basis
  • library 6-31G
end

driver
maxiter 200
end

scf
doublet
uhf
end

task scf optimize
  1. task scf freq

Forum Vet
11:48:12 AM PST - Tue, Feb 19th 2013
Fragment Guess
Velyana
A good solution to your convergence problem is to use the Fragment Guess described in the NWChem manual at
http://nwchemgit.github.io/index.php/Hartree_Fock_Theory#Superposition_of_fragment_molecular...

Please have a look at the modified input file at the end of this posting.
In order to write this, I simply had to look at your molecular structure in a molecule viewer, identify the fragments,
then use a bit of intuition and use a trial and error approach to identify charge and multiplicity for each fragment.
Cheers, Edo
start cuglisin

echo
title "cuglisin"

charge 2

geometry big
 C                    0.28945382     2.81607562    -0.35707429
 O                   0.44365020     3.97345300    -0.87998681
 N                   -1.21118910     1.27281318     1.24447189
 C                   -1.15423910     2.61914048     0.62344566
 H                  -0.65306174     3.35065051     1.30884516
 H                  -2.40382216     1.99727708     1.75541574
 H                  -1.97085841     2.65259489    -0.20047399
Cu                   0.05199018    -0.06753972     0.09488464
 O                    1.31792913     1.42309499    -0.72998685
 O                   -1.11389771    -1.39465845    -0.90902526
 C                   -0.24533519    -2.54642195    -0.46429929
 N                    1.29617602    -1.48664811     0.97204001
 C                    0.86532236    -2.89659824     0.59375065
 H                   0.17100643    -3.39801344     1.74937774
 H                   1.91455851    -3.23352559     0.20426345
 O                  -0.66785734    -3.54112174    -0.91985462
 H                   2.46974811    -1.55280481     2.05122593
end
geometry co
 C                    0.28945382     2.81607562    -0.35707429
 O                   0.44365020     3.97345300    -0.87998681
end 
geometry cug1
 N                   -1.21118910     1.27281318     1.24447189
 C                   -1.15423910     2.61914048     0.62344566
 H                  -0.65306174     3.35065051     1.30884516
 H                  -2.40382216     1.99727708     1.75541574
 H                  -1.97085841     2.65259489    -0.20047399
end
geometry cucu
Cu                   0.05199018    -0.06753972     0.09488464
end
geometry cug2
 O                    1.31792913     1.42309499    -0.72998685
 O                   -1.11389771    -1.39465845    -0.90902526
 C                   -0.24533519    -2.54642195    -0.46429929
 N                    1.29617602    -1.48664811     0.97204001
 C                    0.86532236    -2.89659824     0.59375065
 H                   0.17100643    -3.39801344     1.74937774
 H                   1.91455851    -3.23352559     0.20426345
 O                  -0.66785734    -3.54112174    -0.91985462
 H                   2.46974811    -1.55280481     2.05122593
end
basis
 *   library 6-31G
end

driver
maxiter 200
end

charge 0
set geometry co
scf
 uhf
 singlet
 vectors input atomic output co.mos
end
task scf ignore

charge 0
set geometry cug1
scf
 uhf
 singlet
 vectors input atomic output cug1.mos
end
task scf ignore

charge -2
set geometry cug2
scf
 print mulliken
 uhf
 singlet
 vectors input atomic output cug2.mos
end
task scf ignore

set geometry cucu
charge 0
 scf
 doublet 
 vectors input atomic output cucu.mos
end

task scf

charge -2
set geometry big
scf
 uhf
 doublet 
 vectors input fragment co.mos cug1.mos cucu.mos cug2.mos output cug.mos
 maxiter 65
end
task scf


Forum >> NWChem's corner >> Running NWChem