From NWChem
Viewed 1147 times, With a total of 2 Posts
|
|
|
Clicked A Few Times
Threads 2
Posts 8
|
|
| 1:35:24 AM PST - Sun, Nov 29th 2015 |
|
Hello,
I am having difficulty getting the SCF iterations to look like reasonable convergence is happening. I am used to handling typically difficult cases (e.g. anions and diffuse functions) in other programs (G09, ORCA, GAMESS). There seems to be very little information in the documentation or forums for dealing with presumably common convergence issues. The problem I am having is getting a medium sized system to converge. The guess looks reasonable but the SCF cycles always make some absurdly large step as shown below on the 5th SCF cycle. I have tried removing DIIS, changing levelshifting, changing integration grids, changing thresholds, turning off symmetry and reasonable combinations of those. The integrated density seems strange as there should be 342 electrons, a seemingly large discrepancy from the 342.2 or so. I had no issue with the geometry optimization using the def2-SVP basis set and I expect some potential issues relating using the larger def2-TZVP basis set and linear dependencies, but typically this problem results in the SCF nearly converging and oscillating.
Any ideas or tips would be greatly appreciated.
Thanks
Brad Rose
SCF cycles:
convergence iter energy DeltaE RMS-Dens Diis-err time
---------------- ----- ----------------- --------- --------- --------- ------
d=90,ls=0.0 1 -2069.2799221782 -8.49D+03 1.62D-03 143.5
Grid integrated density: 342.355072410021
Requested integration accuracy: 0.10E-06
d=90,ls=0.0 2 -2069.3288457699 -4.89D-02 1.60D-03 195.5
Grid integrated density: 342.319599138210
Requested integration accuracy: 0.10E-06
d=90,ls=0.0 3 -2069.3722852981 -4.34D-02 2.15D-03 247.5
Grid integrated density: 342.287673419738
Requested integration accuracy: 0.10E-06
d=90,ls=0.0 4 -2069.4109961586 -3.87D-02 4.15D+01 299.4
Grid integrated density: 342.222503260402
Requested integration accuracy: 0.10E-06
d=90,ls=0.0 5****************** -1.09D+06 2.43D+02 361.9
Grid integrated density: 342.200624706760
Requested integration accuracy: 0.10E-06
d=90,ls=0.0 6****************** -3.90D+07 2.19D+02 428.0
Grid integrated density: 342.180837722986
Requested integration accuracy: 0.10E-06
Sample input:
title "b3lyp-d3"
dft
xc b3lyp
disp vdw 3
direct
grid fine
tolerances tight
convergence ncydp 3 damp 90 nolevelshifting
end
basis
* library def2-tzvp
end
geometry
C 1.005318590 1.005318590 0.000000000
C -0.367972140 1.373290730 0.000000000
C -1.373290730 0.367972140 0.000000000
C -1.005318590 -1.005318590 0.000000000
C 0.367972140 -1.373290730 0.000000000
C 1.373290730 -0.367972140 0.000000000
C -2.755662630 0.738377580 0.000000000
C -0.738377580 2.755662630 0.000000000
C 2.017285050 2.017285050 0.000000000
C 2.755662630 -0.738377580 0.000000000
C 0.738377580 -2.755662630 0.000000000
C -2.017285050 -2.017285050 0.000000000
C -1.647866770 -3.390989410 0.000000000
C -0.268400220 -3.760616360 0.000000000
C 2.112749590 -3.122589190 0.000000000
C 3.122589190 -2.112749590 0.000000000
C 3.760616360 0.268400220 0.000000000
C 3.390989410 1.647866770 0.000000000
C 1.647866770 3.390989410 0.000000000
C 0.268400220 3.760616360 0.000000000
C -2.112749590 3.122589190 0.000000000
C -3.122589190 2.112749590 0.000000000
C -3.760616360 -0.268400220 0.000000000
C -3.390989410 -1.647866770 0.000000000
C 4.400325600 2.664691500 0.000000000
C 5.143139500 -0.107527730 0.000000000
C 4.507853330 -2.478448000 0.000000000
C 5.484062450 -1.469450100 0.000000000
C 4.844177260 -3.878109980 0.000000000
C 3.878109980 -4.844177260 0.000000000
C 2.478448000 -4.507853330 0.000000000
C 1.469450100 -5.484062450 0.000000000
C 0.107527730 -5.143139500 0.000000000
C -0.936453130 -6.134235560 0.000000000
C -2.256125580 -5.780630400 0.000000000
C -2.664691500 -4.400325600 0.000000000
C -4.014612350 -4.014612350 0.000000000
C -4.400325600 -2.664691500 0.000000000
C -5.780630400 -2.256125580 0.000000000
C -6.134235560 -0.936453130 0.000000000
C -5.143139500 0.107527730 0.000000000
C -5.484062450 1.469450100 0.000000000
C -4.507853330 2.478448000 0.000000000
C -4.844177260 3.878109980 0.000000000
C -3.878109980 4.844177260 0.000000000
C -2.478448000 4.507853330 0.000000000
C -1.469450100 5.484062450 0.000000000
C -0.107527730 5.143139500 0.000000000
C 0.936453130 6.134235560 0.000000000
C 2.256125580 5.780630400 0.000000000
C 2.664691500 4.400325600 0.000000000
C 5.780630400 2.256125580 0.000000000
C 6.134235560 0.936453130 0.000000000
H 5.901683530 -4.157440700 0.000000000
H 4.157440700 -5.901683530 0.000000000
H 1.752655390 -6.540998960 0.000000000
H -0.649607490 -7.189728210 0.000000000
H -3.032287510 -6.551291020 0.000000000
H -4.788343570 -4.788343570 0.000000000
H -6.551291020 -3.032287510 0.000000000
H -7.189728210 -0.649607490 0.000000000
H -6.540998960 1.752655390 0.000000000
H -5.901683530 4.157440700 0.000000000
H -4.157440700 5.901683530 0.000000000
H -1.752655390 6.540998960 0.000000000
H 6.540998960 -1.752655390 0.000000000
H 7.189728210 0.649607490 0.000000000
H 6.551291020 3.032287510 0.000000000
C 4.014612350 4.014612350 0.000000000
H 4.788343570 4.788343570 0.000000000
H 3.032287510 6.551291020 0.000000000
H 0.649607490 7.189728210 0.000000000
end
task dft optimize
|
|
|
-
Edoapra Forum:Admin, Forum:Mod, bureaucrat, sysop
|
|
|
Forum Vet
Threads 9
Posts 1472
|
|
| 5:35:13 PM PST - Tue, Dec 1st 2015 |
|
Spherical basis
|
Here is a slightly modified input file. The main change is the use of the spherical functions (the default is cartesian)
This input will converge using NWChem 6.6
start b3lyp-d3
title "b3lyp-d3"
dft
xc b3lyp
disp vdw 3
direct
grid fine
tolerances tight
end
basis spherical
* library def2-tzvp
end
geometry
C 1.005318590 1.005318590 0.000000000
C -0.367972140 1.373290730 0.000000000
C -1.373290730 0.367972140 0.000000000
C -1.005318590 -1.005318590 0.000000000
C 0.367972140 -1.373290730 0.000000000
C 1.373290730 -0.367972140 0.000000000
C -2.755662630 0.738377580 0.000000000
C -0.738377580 2.755662630 0.000000000
C 2.017285050 2.017285050 0.000000000
C 2.755662630 -0.738377580 0.000000000
C 0.738377580 -2.755662630 0.000000000
C -2.017285050 -2.017285050 0.000000000
C -1.647866770 -3.390989410 0.000000000
C -0.268400220 -3.760616360 0.000000000
C 2.112749590 -3.122589190 0.000000000
C 3.122589190 -2.112749590 0.000000000
C 3.760616360 0.268400220 0.000000000
C 3.390989410 1.647866770 0.000000000
C 1.647866770 3.390989410 0.000000000
C 0.268400220 3.760616360 0.000000000
C -2.112749590 3.122589190 0.000000000
C -3.122589190 2.112749590 0.000000000
C -3.760616360 -0.268400220 0.000000000
C -3.390989410 -1.647866770 0.000000000
C 4.400325600 2.664691500 0.000000000
C 5.143139500 -0.107527730 0.000000000
C 4.507853330 -2.478448000 0.000000000
C 5.484062450 -1.469450100 0.000000000
C 4.844177260 -3.878109980 0.000000000
C 3.878109980 -4.844177260 0.000000000
C 2.478448000 -4.507853330 0.000000000
C 1.469450100 -5.484062450 0.000000000
C 0.107527730 -5.143139500 0.000000000
C -0.936453130 -6.134235560 0.000000000
C -2.256125580 -5.780630400 0.000000000
C -2.664691500 -4.400325600 0.000000000
C -4.014612350 -4.014612350 0.000000000
C -4.400325600 -2.664691500 0.000000000
C -5.780630400 -2.256125580 0.000000000
C -6.134235560 -0.936453130 0.000000000
C -5.143139500 0.107527730 0.000000000
C -5.484062450 1.469450100 0.000000000
C -4.507853330 2.478448000 0.000000000
C -4.844177260 3.878109980 0.000000000
C -3.878109980 4.844177260 0.000000000
C -2.478448000 4.507853330 0.000000000
C -1.469450100 5.484062450 0.000000000
C -0.107527730 5.143139500 0.000000000
C 0.936453130 6.134235560 0.000000000
C 2.256125580 5.780630400 0.000000000
C 2.664691500 4.400325600 0.000000000
C 5.780630400 2.256125580 0.000000000
C 6.134235560 0.936453130 0.000000000
H 5.901683530 -4.157440700 0.000000000
H 4.157440700 -5.901683530 0.000000000
H 1.752655390 -6.540998960 0.000000000
H -0.649607490 -7.189728210 0.000000000
H -3.032287510 -6.551291020 0.000000000
H -4.788343570 -4.788343570 0.000000000
H -6.551291020 -3.032287510 0.000000000
H -7.189728210 -0.649607490 0.000000000
H -6.540998960 1.752655390 0.000000000
H -5.901683530 4.157440700 0.000000000
H -4.157440700 5.901683530 0.000000000
H -1.752655390 6.540998960 0.000000000
H 6.540998960 -1.752655390 0.000000000
H 7.189728210 0.649607490 0.000000000
H 6.551291020 3.032287510 0.000000000
C 4.014612350 4.014612350 0.000000000
H 4.788343570 4.788343570 0.000000000
H 3.032287510 6.551291020 0.000000000
H 0.649607490 7.189728210 0.000000000
end
task dft optimize
|
|
|
|
|
|
Clicked A Few Times
Threads 2
Posts 8
|
|
| 7:02:29 AM PST - Thu, Dec 3rd 2015 |
|
Thanks for the tip. It does indeed solve the problem.
It seems that some parameters are undocumented, for example after quite a lot of searching on the forums and google I found that
set lindep:n_dep 0
will allow me to set the linear dependance cutoff. I was unable to find this in the documentation. Is there anywhere these 'advanced' features are listed/discussed?
Thanks.
Brad Rose
|
|
|
AWC's:
2.5.10 MediaWiki - Stand Alone Forum Extension
Forum theme style by: AWC
