I am trying to import the wavefunction from another quantum code into nwchem and am running into problem. My goal is to generate a wavefunction in an external electric field to use in a rt-td-dft calculation.
To start with, I tested with the ground state. I have succeeded in translating the wavefunction into the ascii format needed for asc2mov but when I read it in and run a dft with noscf (or without noscf) the total energy (starting total energy) is too high by approx 6 hartrees but the orbital energies are the same as the other code to within 4 decimal places. I have confirmed that the converged wavefunction in both codes give the same energy to withing 5 decimal places (same level of theory, basis set, grid). Why would the total energy differ when the orbital energies are at the converged values indicating that it read in the converged wavefunction from the other code properly? If I let it go through the scf, it will converge to the correct energy. Any help is greatly appreciated.
my nwchem input file:
start ag
MEMORY 1700 mb
title "ag"
geometry "ag" units angstroms noautosym nocenter noautoz
Ag 0.0 0.0 0.00
Ag 0.0 0.0 2.68
Ag 0.0 0.0 5.36
Ag 0.0 0.0 8.04
Ag 0.0 0.0 10.72
Ag 0.0 0.0 13.40
end
BASIS "ao basis" PRINT
Ag library "LANL2DZ ECP"
END
ECP
Ag library "LANL2DZ ECP"
END
set geometry "ag"
dft
direct
iterations 225
xc becke88 perdew86
convergence energy 1e-8
grid lebedev 90 14 ssf euler
tolerances tight
vectors input "ag.movecs"
noscf
end
task dft energy
|