Exploring the applicability of Density Functional Theory approximations

© 2021 Nisha MehtaThe development of density functional approximations (DFAs) is a very active research field, with the number of approximations continuously growing. At the same time, there is a lack of published work that gives a fair evaluation and comparison of DFAs on an equal basis. Considering the vast and important Density Functional Theory (DFT) user community, the introduction of such a variety of techniques can lead to confusion and a low acceptance rate of newly proposed methods in general. In fact, from a user's point of view, it would be beneficial to have access to only a handful of methods that distinguish themselves through their accuracy, reliability, and general robustness in benchmark studies and subsequent applications. Therefore, the primary objective of this thesis is to identify ways to the dizzying array of DFAs. Throughout this thesis, a range of topics will be discussed. They originate from the idea of making computational techniques more reliable. The findings and applications of reliable and efficient computational techniques are at the heart of this thesis. The first project of the thesis is concerned with the thorough evaluation of the double-hybrid density functionals on a large and diverse GMTKN55 database for general main-group thermochemistry, kinetics, and noncovalent interactions. Our study represents one of the largest and most thorough double hybrid studies ever conducted and can serve as an important benchmark informing method developers and users alike. We then shed light on the communication gap between method developers and users, and show that despite tremendous scientific progress in the area, people's perceptions of their favourite approaches have not changed much. The third project provides the most comprehensive database for chalcogen bonding interactions (CHAL336), with high-level wave function based interaction energies. This is followed by the first thorough evaluation of the performance of DFAs for such interactions. Fourthly, we present a proof-of-concept study with the aim of reducing the cost of double hybrid DFT calculations. In this framework, we present the suitability of Grimme's geometric counterpoise correction scheme for basis set superposition errors in double hybrids. Finally, we utilise some of the most reliable and computationally efficient DFT methods to gain insights into cyclotricatecylene-based molecular capsules that can selectively extract ions from solution. Overall, this thesis encourages a shift in the computational procedures surrounding thermochemical problems, allowing users to make more informed decisions about using DFT methodologies in computations