Bond breaking perturbation theory

All this applies to weak and medium strong H-bonds like those encountered for alcohols and many other systems up to carboxylic acid dimers or about 32-42 kJ/mol. (8 or 10 kcal/mol.) Unfortunately vibrational spectra of systems with very strong H-bonds could, with a few exceptions, only be measured in condensed phases. Factors that come in when such systems are examined are potential surfaces with two minima with, in certain cases, the possibility of tunnelling, or flat single minima Most of these systems are likely to be so anharmonic that second order perturbation theory breaks down and the concept of normal vibrations becomes itself question-nable. Many such systems are highly polarizable and are strongly influenced by the environment yielding extremely broad bands 92). Bratos and Ratajczak 93) has shown that even such systems can be handled by relaxation theories. 

Multireference Formalisms. - Whilst the generalization of MPn theory1111 -and, in particular, because of its efficiency, MP2 theory -is obviously an important requirement if many-body perturbation theory is to be applied to bond breaking processes, radicals, excited states and the like where a multireference formalism is mandated, a robust theory that be applied routinely to a wide range of problems has been elusive for over 25 years (see, for example, the discussion of the problems associated with multireference perturbation theory in my monograph53 Electron correlation in molecules published in 1984). 

Due to the inherent multireference nature of bond-breaking problems, it seems rather unlikely that single-determinant CC methods of any sort will ever provide a satisfactory solution to these problems. It seems much more sensible to use MR approaches to study problems of this sort when they are relevant for a particular chemical application. For small molecules, MRCI offers a suitable option, while CASSCF calculations corrected by perturbation theory are more generally useful for larger systems. 

At this point we can gain additional information from perturbation theory, not easily available from orbital correlation diagrams. If we wish to join two olefin molecules together to form cyclobutane, it is clearly necessary to transfer electron density from a filled orbital of hx symmetry, to an empty orbital of bz symmetry (see Fig. 10). This is the only way in which we can break the carbon-carbon n bonds and convert them into suitable a bonds. Mango and Schachtschneider originally accomplished the necessary change by moving electrons from a filled b orbital (concentrated on the olefins) into the empty 61 orbital (concentrated on the metal), and from the filled bz orbital (concentrated on the metal) into the empty 62 orbital (concentrated on the olefins). 

---