DFT optimized failed and Bad termination exit code 255
Just Got Here
10:43:37 PM PDT - Sat, Aug 18th 2018
Hello all,

I'm new to NWChem 6.8 and facing issues to run Raman calculation (solvation). The main problem is it does not reach the convergence criteria (dft optimize failed). I tried maxiter up to 1000 but similar issue is still found.

I also get the below error. Can anyone assist me to fix these issues?

=======================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= PID 21764 RUNNING AT compute-3-10.localdomain
= EXIT CODE: 255
= CLEANING UP REMAINING PROCESSES
= YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
=======================================================================

Below is the input file :
echo

start molecule

title "solvation"
charge 0

geometry units angstroms noautosym noautoz
  C       -2.16792895    -0.76286432    -0.03368798

  C       -0.84576223    -0.82921793    -0.05399091

  H       -2.67099034     0.18265039     0.04473144

  H       -2.77136442    -1.65109566    -0.09507467

  C       -0.09318252     0.46372095     0.03650672

  C       -0.06766891    -2.11581233    -0.16170888

  H       -0.74672312    -2.95798497    -0.21903223

  H        0.58193686    -2.25345048     0.69603582

  H        0.56243634    -2.11910020    -1.04469127

  O       -0.59478922     1.54163458     0.12791735

  O        1.21721017     0.28045421     0.00407983

  C        2.03870872     1.43317391     0.08584505

  H        3.05456309     1.07076390     0.04463603

  H        1.86053152     1.95677305     1.01453598

  H        1.83930801     2.09756937    -0.74320229
end

basis
 * library 6-311G**
end

cosmo
 dielec 6.533
end

dft
 xc b3lyp

 odft

 mult 1

 direct
end

driver
 maxiter 200
end

hessian
 profile
end

set prop:newaoresp 1
set cphf:thresh 1.0d-6
set cphf:acc 1.0d-16
set cphf:maxiter 60

property
 response 1 0.0719800279 # hartrees = 633 nm

 damping 0.004

 convergence 1e-6
end

raman
 normal

 lorentzian
end

driver ; tight ; end
task dft optimize
task dft raman numerical

Appreciate any help.

Sophia

Forum Vet
4:32:15 AM PDT - Sun, Aug 19th 2018
This input causes the delta E to oscillate between D-9 and D-10, and the calculation not to converge, where your mistake does not appear, yet.

By the way, I am not very clear about the usage of Hessian in a solvent, because the vibrational analysis may be invalid for an nonstationary point on a PES, which may be automatically printed in GAMESS.

Thanks.

Forum Vet
3:04:33 PM PDT - Sun, Aug 19th 2018
Please try this input file to get a converged DFT solution
start molecule

title "solvation"
charge 0

geometry units angstroms autosym 0.2
  C       -2.16792895    -0.76286432    -0.03368798
  C       -0.84576223    -0.82921793    -0.05399091
  H       -2.67099034     0.18265039     0.04473144
  H       -2.77136442    -1.65109566    -0.09507467
  C       -0.09318252     0.46372095     0.03650672
  C       -0.06766891    -2.11581233    -0.16170888
  H       -0.74672312    -2.95798497    -0.21903223
  H        0.58193686    -2.25345048     0.69603582
  H        0.56243634    -2.11910020    -1.04469127
  O       -0.59478922     1.54163458     0.12791735
  O        1.21721017     0.28045421     0.00407983
  C        2.03870872     1.43317391     0.08584505
  H        3.05456309     1.07076390     0.04463603
  H        1.86053152     1.95677305     1.01453598
  H        1.83930801     2.09756937    -0.74320229
end

basis spherical
 * library 6-311G**
end

cosmo
 dielec 6.533
end

dft
 xc b3lyp
 mult 1
 direct
 smear
 convergence ncydp 0 damp 45 dampon 1d99 dampoff 1d-4
end
set quickguess t

task dft optimize
...

Just Got Here
8:08:32 AM PDT - Mon, Aug 20th 2018
It worked.
Thank you, edoapra.

Gets Around
10:52:26 AM PDT - Mon, Aug 20th 2018
What is "quickguess"? I cannot find it in the documentation pages.
John Keller


Forum >> NWChem's corner >> Running NWChem