Methods CIPSI

As far as the molecular calculation is concerned, the use of an ab initio method is necessary for an adequate representation of the open-shell metastable N (ls2s) + He system with four outer electrons. The CIPSI configuration interaction method used in this calculations leads to the same rate of accuracy as the spin-coupled valence bond method (cf. the work on by Cooper et al. [19] or on NH" + by Zygelman et al. [37]). 

Electron correlation was treated by the CIPSI multi-reference perturbation algorithm ([24,25] and refs, therein). The Quasi Degenerate Perturbation Theory (QDPT) version of the method was employed, with symmetrisation of the effective hamiltonian [26], and the Maller-Plesset baricentric (MPB) partition of the C.I. hamiltonian. 

The results of the above cited applications [18-28,45] have clearly shown that CS INDO method is fairly successful in combining equally satisfactory predictions of electronic spectra and potential surfaces (especially along internal rotation pathways) of conjugated molecules, a goal never reached by other NDO-type procedures. CS INDO shares, at least partly, the interpretative advantages of the CIPSI-PCILO-CNDO procedure [32,33,36,37], coming from using the same hybrid AO basis sets, but improves its predictive capabilities as far as spectroscopic and photochemical properties are concerned. 

Both methods may use a reduced optimally selected Cl instead of a CAS-SDCI (for example, a selected Cl using a CIPSI procedure [10,18]). In such a case, the computational cost of ec-CCSD falls down dramatically, while the final energies hardly change [35]. 

Comparisons with other theoretical methods are important. Our Cu calculations are based on the pseudopotential of Pellisier [35], who applied it to the Cu2 system. A controversy arose when CASSCF-CCI calculations on the same system seemed to imply [4lQ] that the CIPSI calculations matched the experimental energy too well for a non-relativistic method. Pellisier [43] replied by showing his pseudopotential included relativistic effects. 

The data of Table 3b have been computed by replacing the active space of the CASPT2 treatment by the CIPSI multireference space of the MRPT2 method documented in section 2.1. The selection of the components of the reference space by this technique enables us to reduce the gap between the Moller-Plesset and Epstein-Nesbet total energies to less than 0.01 a.u., and so to improve our evaluation of energy balances, disregarding a possible overestimation of the Moller-Plesset values due to intruder states, as it is the case in the ScNC system. 

Numerous theoretical studies on DMABN have been carried out, and many of them confirm the greater validity of the TICT model. The main body of such calculations, however, has been limited to the isolated system, while few examples including solvent effects can be quoted. " On the contrary, the phenomenon is strongly related to solvation and thus explicit considerations of solvent interactions are very important to get a more accurate understamding of the experimental evidence on the specific effects due to the presence of polar solvents. Here we summarize the results of the correlated study of DMABN both in vacuo and in solution we have published on the Journal of American Chemical Society. In this study we have used the multireference perturbation configuration interaction (Cl) method, known with the CIPSI acronym, which has been coupled to the PCM-IEF solvation continuum model. ... 

If it is below an a priori set threshold, the respective CSF is eliminated. There are other procedures available in which the perturbation theory is used for screening and selection of CSFs in a more sophisticated way. Among these the most popular has been the CIPSI method and the Cl energy extrapolation method. ... 

---