hi, Ryan,
By default NWChem only prints out large MO coefficients, so if you use GabEdit
to read the NWChem output file ,what you get is incomplete about the MOs. To plot
the MOs correctly you can use either the dplot directive to generate cube files or
use the keyword moldenfile to generate molden file. There is documentation about this.
best
Yu
Quote:Ryan81 Nov 30th 7:29 pmHi all,
I am new to NWChem and performing a simple DFT calculation and comparing the results to those obtained from GAMESS and Gaussian with which I have more experience.
I am running a single-point energy calculation at the B3LYP/6-31G(d) level, and I get the same orbital energies as I do when I perform the same calculation with GAMESS and Gaussian. However, when I view the orbitals from the NWChem output file in Gabedit and Jmol, the orbitals look very different from those in GAMESS and Gaussian (the results from GAMESS and Gaussian agree with each other).
I have performed an energy calculation in NWChem using a geometry optimized with NWChem and a separate energy calculation with a geometry optimized with GAMESS, and I get the same results (that is, NWChem gives the same eigenvalues as GAMESS and Gaussian but different orbital plot). This tells me that it's not an issue with the geometry optimization.
Why are the orbitals incorrect when plotted from the NWChem output? I find it strange that the orbital energies are pretty much the same (agree within 6 or so significant figures) while the plotted orbitals differ so severely. Any suggestions as to how I can calculate/plot the orbitals in NWChem to agree with the orbitals produced from GAMESS and Gaussian?
Thank you! |