On the Spectroscopically Determined Potential Energy Surfaces for the Electronic Ground States of NO2 and H2O

Previous spectroscopically determined potentials for both H216O and NO2 are discussed. It is shown that a recent H216O potential energy surface due to D. Xie and G. Yan (1996. Chem. Phys. Lett. 248, 409), which was determined by fits to vibrational term values alone and was claimed to be more accurate than other published spectroscopically determined potentials for this system, actually gives unacceptably poor results for rotationally excited water. Reasons for this failure are discussed and the dangers of relying on vibrational term values alone are emphasized. Previous spectroscopic potentials for ground state NO2 are all found to have problems with unphysical minima ("holes"). Starting from the potential energy surface for the ground (&Xtilde;2A1) electronic state of NO2 constructed by S. A. Tashkun and P. Jensen (1994. J. Mol. Spectrosc. 165, 173) using the approximate MORBID approach a suitable starting point for fits using an exact kinetic energy operator approach was constructed. Least-squares fits to 17 potential parameters gives a potential which reproduces 173 vibrational term values with a standard deviation of only 2.8 cm-1 in the low-energy region (<10 000 cm-1). For many even levels below, and all levels above, approximately 10 000 cm-1 the first excited electronic state (&Atilde;2B2) perturbs the vibrational energy levels of the ground state. We were unable to fit these levels. Tests show that the resulting effective potential surface has no problems with unphysical holes and gives a reasonable representation of the rotational structure of the low-lying vibrational states of NO2. Copyright 1997 Academic Press. Copyright 1997Academic Press