Complete vs Restricted Active Space Perturbation Theory Calculation of the Cr2 Potential Energy Surface

In this paper, we calculate the potential energy surface (PES) and the spectroscopic constants of the chromium dimer using the recently developed restricted active space second-order perturbation (RASPT2) method. This approach is benchmarked against available experimental measurements and the complete active space second-order perturbation theory (CASPT2), which is nowadays established as one of the most accurate theoretical models available. Dissociation energies, vibrational frequencies, and bond distances are computed at the RASPT2 level using several reference spaces. The major advantage of the RASPT2 method is that with a limited number of configuration state functions, it can reproduce well the equilibrium bond length and the vibrational frequency of the Cr dimer. On the other hand, the PES is well described only at short distances while at large distances, it compares very poorly with the CASPT2. The dissociation energy is also ill-behaved, but its value can be largely improved using a simple workaround that we explain in the text. In the paper, we also address the effect of the Ionization Potential Electron Affinity (IPEA) shift (a parameter introduced in the zeroth-order Hamiltonian in the CASPT2 method to include the effect of two-electron terms) and show how its default value of 0.25 is not suitable for a proper description of the PES and of the spectroscopic parameters and must be changed to a more sound value of 0.45.