Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

New open-shell molecular electronic structure program: What to do with it?

  1. Dec 1, 2010 #1


    User Avatar
    Science Advisor

    Hello forum members,
    I got a nifty new multireference CI program at my hands and now I'm wondering what to do with it. The program is able to correlate general MCSCF wave functions, and as such basically any kind of molecular electronic structure imaginable; in particular also electronic structures in which DFTs blow up (as long as the molecule is small enough).

    The new program could do things like potential energy surface scans during bond breaking/formation, (general) excited state calculations, 1-electron properties of molecules (electric moments and stuff), and strongly correlated ground state wave functions (say, in transition metal complexes). My main problem now is: I'm not really that good with real molecular physics or chemistry (I'm in method development), and I can't come up with anything truely interesting to apply it to. Does anyone have a suggestion? Otherwise I'd probably go for ligand effects in small transition metal clusters to demonstrate its efficacy.

    Things the new program can likely do (with enought patience):
    - treat molecules with up to 15-30 atoms (depends on the atoms and the accuracy goals)
    - treat active spaces up to ~15-25 orbitals
    - treat reference spaces up to around 1-10 million CSFs
    - additionally treat up to around 30-100 closed-shell orbitals (i.e., a large inactive space)

    Especially the closed-shell part (it's really good with these) should significantly extend the capabilities of all previous MRCI programs. Yet it is written with with routine applications in mind (say, to get reference values for model clusters for checking which DFT to use), and I wonder if there is anything more interesting to investigate. Any suggestions?
  2. jcsd
  3. Dec 1, 2010 #2


    User Avatar
    Science Advisor

    Maybe something like http://pubs.acs.org/doi/pdf/10.1021/ct9006234" [Broken] (although a smaller system or model would be required of course, perhaps a biomimetic model); So I'm thinking maybe metallo-organic systems/organic catalysts, and problems in multistate reactivity.

    I was recently told there's a lot of low-lying fruit for excited-state methods in organic/metallo-organic catalysis problems, although I don't know any examples off-the-top-of-my-head.
    Last edited by a moderator: May 5, 2017
  4. Dec 2, 2010 #3


    User Avatar
    Science Advisor

    Thanks, alxm. That looks interesting. I'll definitely investigate this direction!
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: New open-shell molecular electronic structure program: What to do with it?
  1. Inner shell electrons (Replies: 2)