Triatomic molecule, vibration-rotation Hamiltonians

We find it convenient to reverse the historical ordering and to stait with (neatly) exact nonrelativistic vibration-rotation Hamiltonians for triatomic molecules. From the point of view of molecular spectroscopy, the optimal Hamiltonian is that which maximally decouples from each other vibrational and rotational motions (as well different vibrational modes from one another). It is obtained by employing a molecule-bound frame that takes over the rotations of the complete molecule as much as possible. Ideally, the only remaining motion observable in this system would be displacements of the nuclei with respect to one another, that is, molecular vibrations. It is well known, however, that such a program can be realized only approximately by introducing the Eckart conditions [38]. 

An alternative form of exact nonrelativistic vibration-rotation Hamiltonian for triatomic molecules (ABC) is that used by Handy, Carter (HC), and... 

We have discussed up to now vibrational spectra of linear and bent triatomic molecules. We address here the problem of rotational spectra and rotation-vibration interactions.3 At the level of Hamiltonians discussed up to this point we only have two contributions to rotational energies, coming from the operators C(0(3]2)) and IC(0(412))I2. The eigenvalues of these operators are... 

Jensen, P. (1983), The Nonrigid Bender Hamiltonian for Calculating the Rotation-Vibration Energy Levels of a Triatomic Molecule, Comp. Phys. Rep. 1,1. 

In a paper in 1979, Carl Ballhausen [1] expressed the belief that today we realize that the whole of chemistry is one huge manifestation of quantum phenomena, but he was perfectly well aware of the care that had to be taken to express the relevant quantum theory appropriately. So in an earlier review [2] that he had undertaken with Aage Hansen, he scorned the usual habit of chemists in naming an experimental observation as if it was caused by the theory that was used to account for it. Thus in the review they remark that a particular phenomenon observed in molecular vibration spectra is presently refered to as the Duchinsky effect. The effect is, of course, just as fictitious as the Jahn-Teller effect. Their aim in the review was to make a start towards rationalization of the nomenclature and to specify the form of the molecular Hamiltonian implicit in any nomenclature. In an article that Jonathan Tennyson and I published in the festschrift to celebrate his sixtieth birthday in 1987 [3], we tried to present a clear account of a molecular Hamiltonian suitable for treating the vibration rotation spectrum of a triatomic molecule. In an article that I wrote that appeared in 1990 [4], I discussed the difficulty of deciding just how far the basic chemical idea of molecular structure could really be fitted into quantum mechanics. 

The anharmonic potential energy is usually easier to represent in internal coordinates than in normal mode coordinates. However, what restricts the use of internal coordinates is the complicated expression for the vibrational/rotational kinetic energy in these coordinates (Pickett, 1972). It is difficult to write a general expression for the vibrational/rotational kinetic energy in internal coordinates and, instead, one usually considers Hamiltonians for specific molecules. For a bent triatomic molecule confined to rotate in a plane, the internal coordinate Hamiltonian is (Blais and Bunker, 1962) ... 

Bunker, P.R., Moss, R.E. Effect of the breakdown of the Born-Oppenheimer approximation on the rotation-vibration Hamiltonian of a triatomic molecule, J. Mol. Spectry. 1980, 80,217-28. 

Each vibration a molecule exhibits corresponds to two quadratic terms in the Hamiltonian as in Equation 7.35. There is a quadratic momentum term and a quadratic term in the associated displacement coordinate. Thus, each vibrational mode contributes 2 X NkT/2 to U. We can collect this information concisely by saying that for a molecular gas, U is 3NkT/2 (translation), plus either NkT/2 for linear molecules or 3NkT/2 for nonlinear molecules (rotation), plus NkT times the number of modes of vibration. The gas of an ideal diatomic molecule has U = 7NkT/2 as in Equation 11.59. A linear triatomic molecule has U = 13NkT/2 because there are three vibrational modes, one of which is twofold degenerate. 

---