Hi i've got a problem with my code (purely made for educational purpose to understand quantum/comp chemistry). For many hours i've been trying to see where i made mistake. My overlap matrix gives wrong (mostly bcs 1,2 and 2,1 are good, and diagonal elements integrate to unity (thank god)) off dia...

Can I estimate the bond energy by running a single Gaussian calculation of the fragments at very long separation (say, 40 angstroms) in the singlet spin state? Or I have to calculate each fragment without the other? What multiplicity I use for each job? doublet? Which approach is better? Option 1...