Dispersion-accounting DFT

To give an overview of the subject, quantum chemical methods that properly account for dispersion are discussed in more details in this chapter. These, however, are limited to several wave function theory and two dispersion-accounting density functional theory (DFT) methods. In fact, many dispersion-accounting DFT approaches have been developed recently but only few of them are widely used in modeling of transition metal-graphene systems. 

Therefore, some special dispersion-accounting [31] DFT methods are needed for reliable description of these interactions. These methods may initially account for dispersion in a nonempirical way (i.e., include the nonlocal term in that accounts for dispersion) or may incorporate empirical dispersion corrections (in a force-field manner). Our emphasis is on these two methods, which are commonly used in modeling of transition metal-containing graphene systems. Other methods designed to account for dispersion interactions are also available but are beyond the scope of our discussion. The reviews on dispersion-accounting DFT methods are available elsewhere [24, 28]. 

S22, S66, and S66 x 8 were developed and represent a number of Ej, values of organic molecule dimers calculated with state-of-the-art electron correlation methods [2]. Approximate (importantly, dispersion-accounting DFT methods) computational methods are often tested against these databases and quite often initially parameterized to reproduce these data. Therefore, these datasets represent benchmarks that help to validate a recently introduced method or to parameterize a new one. 

Reactive force field methodologies such as RFF, QM/MM approaches, and DFT methodologies continue to be used to study polymerization, though the level of activity has dropped since the 1990s. Perhaps this decline is due to a lack of agreement with experiment in efforts initiated, but not published. The lack of dispersion in DFT summarized in Section 7.2.1, the polymer chain conformational issues discussed in Section 7.2.2, and the difficulty in accounting for the counteranion in Section 7.2.2.4 are the most probable sources of disagreement with experiment. 

Crystal structure prediction The field of organic crystal structure prediction remains one of the best testing grounds for intermolecular potentials. Acciuades need not be as high as that needed for spectroscopic calculations, but the effects of molecular flexibility and many-body non-additivity need to be accounted for. See Price (2008, 2009) for recent reviews of this subject. For a description of dispersion-corrected DFT methods specially parametrized for organic crystals see Neumann and Perrin (2005). For a comprehensive examination of the role of detailed distributed multipole models in this field see Day et al. (2005). 

In view of this and in line with the DFT-D approach described by Grimme [118], we have added an atom-atom pair-wise additive potential of the form Csemi-empirical energy [19-21] in order to account for dispersion effects [43], Thus the dispersion corrected semi-empirical energy ( Pm3-d) is now given by ... 

While he recognizes that SCS-MP2 is perhaps his most important individual contribution, he asserts that the dispersion correction work is much more significant. There are many problems that cannot be solved without dispersion. We can live without double hybrids or SCS-MP2, but proper accounting for dispersion is critical within DFT. ... 

These imperfections have occasioned to review the spherical DFT approach with respect to a more correct description for fluids which consists of non-spherical particles. The paper applies a statistic thermodynamic approach [7, 8] which uses density functional formulation to describe the adsorption of nitrogen molecules in the spatial inhomogeneous field of an adsorbens. It considers all anisotropic interactions using asymmetric potentials in dependence both on particular distances and on the relative orientations of the interacting particles. The adsorbens consists of slit-like or cylinder pores whose widths can range from few particle diameters up to macropores. The molecular DFT approach includes anisotropic overlap, dispersion and multipolar interactions via asymmetric potentials which depend on distances and current orientations of the interacting sites. The molecules adjust in a spatially inhomogeneous external field their localization and additionally their orientations. The approach uses orientation distributions to take the latter into account. 

While it is possible to account for non-covalent interactions using specialized force fields, common density functionals do not correctly describe the long-range van der Waals (London dispersion) interactions. Efficient dispersion correction schemes for DFT have been developed [14-17], but so far their application in QM/MM refinements is scarce. The importance of London forces for biomolecu-lar structures has been established conclusively [18]. 

However, there is no doubt that periodic DFT calculations are the way forward and Fig. 6.25c shows the results of such a calculation for Mg2[FeH6]. This gives the modes in the correct positions and includes the external modes. However, the shape of the librational mode shows that it is strongly dispersed and this is not taken into account in this type of calculation. Clearly the next step is to carry out a periodic DFT calculation across the complete Brillouin zone to include the dispersion, as was done for the alkali metal hydrides ( 6.7.1). 

However, care has to be taken when applying DFT to hydrocarbon species in zeolites. The currently available functionals do not properly account for dispersion, which is a major stabilizing contribution for hydrocarbon-zeolite interactions. Due to the size of the systems it is difficult to apply wavefunction-based methods such as CCSDfT) or MP2. Thanks to an effective MP2/DFT hybrid approach and an extrapolation scheme energies, including the dispersion contribution, are now available for the different hydrocarbon species of Fig. 22.1 [50]. 

Despite all the advantages of the DFT method one should be aware of well known failure of this method. Exchange-correlation functionals currently available are not capable to account for dispersion interaction, e. g., interaction between zeolite channel wall and hydrocarbons cannot be properly described at the DFT level (sec section 4.3). 

Localization of stationary points along the reaction path for reactions taking place inside the zeolite pores is one of the greatest challenges in zeolite modeling. The reactions of hydrocarbons are particularly difficult to model since the hydrocarbon...zeolite interaction can be dominated by the dispersion interaction that is not properly accounted at the DFT level. Only one example is presented here. Clark et al. investigated the role of benzenium-lype carbenium ion in the bimolecular w-xylene disproportionation reaction in zeolite faujasite.163] The benzenium-type carbenium ion 1 was identified in zeolite catalyst for the... 

---