Hi,

I'm trying to model the potential energy surface of the N-H bond length in NH4+ spanning a bond length from, say, 0.6 Å to 10 Å to study the behavior of the potential energy as well as partial charges as the proton departs. I do this using flexible scanning in Gaussian 16. I have noticed a lot of weird behavior here:

  • When I try simple RHF/6-311+G(d,p), I get a seemingly optimum PES curve resembling the one from Morse potential. However, when I try including electronic correlation via MP2, B3LYP, or even CCSD(T), sometimes with aug-cc-pVTZ, I always get an energy catastrophe where, after bond breakage, the potential energy of the system goes down toward negative infinity as the "bond" length increases.
  • In both RHF and higher levels, the Mulliken partial charge on the departing proton never reaches +1. The maximum it gets is +0.8 even at distances of 30 A.
  • Except for RHF, I always noticed that the remaining NH3 fragment would lose its trigonal pyramidal geometry as H+ goes away, and becomes trigonal planar. which is counterintuitive.
  • On the contrary, none of this weird behavior was noticed when I did the same type of calculation on H3O+. Any clues?

    More Ahmed Taha Ayoub's questions See All
    Similar questions and discussions