I am trying to perform CASSCF calculations for a series of diradicals. As input orbitals I have used two types of natural orbitals. One is unrestricted natural orbitals and the other MP2-based natural orbitals. From both sets I am getting different trend for the properties I am trying to look into. Can anyone suggest me which is the better way to know which results are reliable ? I have used ORCA 4.0 version.
2
-
3$\begingroup$ Since you have to reconverge the CAS orbitals anyway, it doesn't really matter which starting orbitals you choose, as long as they are sensible choices with respect to the spin state. They should give you the identical solution in the end, if you have chosen the same active space. The context of your question is not enough to actually help you with your problem. Highly related: Is it reasonable to use natural orbitals of an unrestricted HF calculation as a basis to start a CASSCF calculation for a radical $\endgroup$– Martin - マーチン ♦Commented Jul 31, 2018 at 12:36
-
$\begingroup$ If you need more help with this topic, I suggest you edit in more information, like how you determine your active space, what kind of molecule, etc. (I know that is sometimes difficult, when involved in active research, but you can try with a more simple model compound that shows the same behaviour.) $\endgroup$– Martin - マーチン ♦Commented Aug 1, 2018 at 9:07
Add a comment
|