sanya
sanya@photonics.ru
No, SCF accelerator and guess have no effect on the final SCF energy if the SCF solution is stable. If the solution is unstable, you don't have to use single-reference SCF at all. I hope it is not your case.
>(2)In relation to (1), if YES, would that be methodologically valid in computational chemistry?
The answer is NO, so you don't have to worry about it.
>(3)(assuming that spin contamination is not severe) THEORETICALLY, would there be significant difference between VIE obtained from open shell and close shell calculation with the same input file (of course except for SCFTYP, MULT and ICHARG) compared to VIE obtained from open shell with (B) input and from close shell with (A) input?
It's no difference at all. You may check it with some small molecule, for which you may perform calculations with both input types.
However, I think it is no good in using close shell energy for VIE calculation. I'd better use UMP2 energy both for close shell and open shell. Try and see.
>MERRY CHRISTMAS AND HAPPY NEW YEAR TO EVERYONE!
Oh, no! it's only November! :)