Subsystem-Based Methods for Global and Local Optical Properties

Abstract

The accurate calculation of optical properties in complex systems is a challenging task due to the quantum-mechanical nature of optical processes and the computational demands involved. Subsystem-based methods offer a promising approach to tackle these challenges by partitioning the full system into smaller, computationally more efficient subsystems while maintaining the necessary accuracy. In this thesis, two distinct subsystem-based approaches for calculating optical properties are investigated: local embedding methods and global fragmentation schemes. In the local embedding methods, different polarizable embedding schemes were employed to calculate optical excitations on solvated para-nitroaniline and pentameric formyl thiophene acetic acid. By dissecting the individual interaction effects in a common theoretical framework and developing an extensive computational setup, a one-to-one comparison between polarizable embedding and frozen-density embedding was performed. This comparison provided valuable insights into the importance of separate interaction effects and highlighted the strengths and limitations of each approach. The results revealed the major significance of mutual ground-state polarization in the embedding schemes and the partial importance of dynamical environment effects. Specifically, a strong dependence on the underlying structural geometries could be observed. The global fragmentation schemes allow the calculation of optical properties for various zeolitic imidazolate frameworks (ZIFs). In an extensive study, a general computational protocol was established in order to obtain geometries and calculate refractive indices for ZIF-8, yielding accurate results with reduced computational demands. The implemented fragmentation schemes were then extended to predict optical properties for ZIFs with the same topology but substituted organic linkers. The results indicated the high potential of this approach for the rationalization and prediction of optical properties in empty ZIFs. Moreover, the framework was adapted to incorporate guest molecules into the porous structure of the ZIFs, providing rough estimates of the number of molecules incorporated per pore. Overall, this thesis introduces novel theoretical and computational frameworks for the accurate calculation of optical properties in complex systems. The presented approaches offer a promising direction for future research in the field of optical properties, enabling a deeper understanding of the interactions and their effects on the optical response of various molecules. This paves the way for an improved characterization and design of advanced materials with precise optical functionalities.