The answers of What should be done with "Is my calculation correct ?"-type questions? clear out that questions need not be deleted by the OP/Community.
Note: This is the Part-2 of the same issue but instead of editing it, I am starting this new thread so that the comments,answers and votes on ...
i know gaussian has eigentest which, if it finds a negative eigenvalue it terminates the program. anyway to have gaussian optimize a molecule and to KEEP GOING if it finds ni=1 as opposed to ni=0?
You see. The geometry optimization tries to find not a minimum, but rather a stationary point. TS is also a stationary point. Thus, finding TS is 100% valid result of geometry optimization.
Try to manually shift atoms in the direction of imaginary frequency. GaussView can do that easily.
Yes, but do you know that this grid is used only in the SCF procedure and that Coupled Perturbed Hartree-Fock (CPHF) equations used to calculate frequencies use a different grid?
You can try to add something like Perturbed CPHF=(Grid=Fine)
or CPHF=(Grid=UltraFine)
I do not know which grid is used in CPHF when SuperFine grid is used in SCF.
But when one uses UltraFine grid in SCF, only SG1 grid is used in CPHF.