Dispersion energy, estimation

The last important contribution to intermolecular energies that will be mentioned here, the dispersion energy (dEnis). is not accessible in H. F. calculations. In our simplified picture of second-order effects in the perturbation theory (Fig. 2), d mS was obtained by correlated double excitations in both subsystems A and B, for which a variational wave function consisting of a single Slater determinant cannot account. An empirical estimate of the dispersion energy in Li+...OH2 based upon the well-known London formula (see e.g. 107)) gave a... 

As a result, some approaches to computing dispersion energy have involved using either experimental or theoretical data for gas-phase clusters to estimate the strength of dispersion interactions between different possible solute and solvent functional groups. However,... 

In contrast with the EHF energy, which can generally be estimated by ab initio methods for many systems of practical interest (see ref. 140 for a semiempirical treatment), the interatomic correlation energy is most conveniently approximated from the dispersion energy by accounting semiempiri-cally for orbital overlap and electron exchange effects. Thus the following... 

The multipole part can efficiently be estimated from the distributed multipole analysis191. In this way the electrostatic penetration contribution is obtained. One may note that the accuracy of the electrostatic term can be increased by keeping the penetration part from Eq. (1-184), and replacing the Hartree-Fock distributed multipole moments by some correlated, e.g. MP2 moments. Finally, the intramonomer correlation term and the dispersion energy can be evaluated from the expression,... 

The electronic properties of organic conductors are discussed by physicists in terms of band structure and Fermi surface. The shape of the band structure is defined by the dispersion energy and characterizes the electronic properties of the material (semiconductor, semimetals, metals, etc.) the Fermi surface is the limit between empty and occupied electronic states, and its shape (open, closed, nested, etc.) characterizes the dimensionality of the electron gas. From band dispersion and filling one can easily deduce whether the studied material is a metal, a semiconductor, or an insulator (occurrence of a gap at the Fermi energy). The intra- and interchain band-widths can be estimated, for example, from normal-incidence polarized reflectance, and the densities of state at the Fermi level can be used in the modeling of physical observations. The Fermi surface topology is of importance to predict or explain the existence of instabilities of the electronic gas (nesting vector concept see Chapter 2 of this book). Fermi surfaces calculated from structural data can be compared to those observed by means of the Shubnikov-de Hass method in the case of two- or three-dimensional metals [152]. 

Finally we should note a valuable approach developed by Pyper and coworkers " for ionic sohds in which HF methods are used first to obtain a set of crystal orbitals, the interactions between which are calculated as a function of distance including a full explicit evaluation of the exchange term (unlike the local density approximation used in the electron gas method). Estimates of the dispersive energy are then added to the resulting interaction energy. This approach is particularly successful for strongly ionic hahdes and oxides. 

As a final test of the influence of correlation, the CC geometry was reoptimized at the MP2 level, using a 6-31G basis set. The resulting changes, with respect to the Hartree-Fock structure, accounted for a further stabilization of only 1.1 kcal/mol, or a 6% increase, despite the contraction of one of the H-bonds from 3.05 to 2.92 A, a full 0.13 A. It was noted that an estimate of the dispersion energy by the empirical London formulation does not reproduce the correlation contribution to the interaction as computed by MP2. 

An analysis of a Pto.sRhos with dispersion 0.40 by energy-dispersive X-ray spectroscopy showed that the individual particles had approximately the overall composition. The CO/ Pt double-resonance spectrum of CO close to Pt in the Pto.sRho.s surface is shown in Fig. 40b. The mere existence of this double-resonance signal shows that there is platinum in the surfaces of these particles. Its position, however, is different from that of CO on a pure platinum surface, showing that these particles are alloys. From the analysis of the Pt/ CO double-resonance spectrum of platinum in Fig. 62b, it is found that a fraction 0.49 + 0.07 of the Pt atoms are attached to CO, whereas the dispersion is estimated to be between 0.40 and 0.67 (Sec-... 

In order to improve on the Hartree-Fock model, the use of perturbation theory is common. The first energy correction is obtained at second order and the corresponding method is calles second-order Moller-Plesset perturbation theory (MP2). MP2 calculations provide a first estimate for the correlation energy SE, which turned out to be also useful for estimates of the interaction energy in cases dominated by dispersive interactions (see the next section for an overview on how interaction energies can be calculated from total electronic energy estimates). 

In addition, scanning electron microscopy coupled with energy-dispersive energy analysis (SEM-EDX) has been implemented as a complementary analytical tool for various purposes (i) to estimate the mean particle size of the metallic particles and look at the eventual influence of the used precursors on these characteristics (ii) to investigate more deeply the composition and dispersion anomalies detected by XPS on certain catalysts (iii) to find experimental evidence for bismuth redeposition on the catalyst surface after use. 

Computation of correlation contributions is more problematic. Szalewicz et al. first calculated the dispersion energy Edisp ° > which corresponds to its polarization component, neglecting that part which is due to exchange of electrons. A calculation of this quantity in a finite basis set represents an upper bound to the exact value and is free of superposition error due to its formalism. The results in Figure 6 illustrate the convergence of this quantity. A basis-set limit within 0.1 kcal/mol of the lowest value of -1.93 kcal/mol was estimated. The reader should take note that this term is half as large as the SCF contribution, far from negligible. 

---