Lighting up nanoclusters: path to efficient quantum chemical approaches

Abstract

Noble metal nanomaterials represent one of the most actively pursued research areas in modern science, driven by their wide range of applications in medicine, catalysis, sensing, and optoelectronics. The intriguing optical features observed in these systems have sparked considerable interest, motivating investigations from both experimental and theoretical perspectives. This dissertation focuses on the theoretical front of this endeavor. In theoretical chemistry, the choice of computational method is a critical decision, one that must carefully balance the desired level of accuracy with the computational cost involved. In this dissertation, we center our approach around density functional theory (DFT), one of the most widely adopted and powerful methods for modeling noble metal nanoclusters, which we employ for ground-state calculations. From there, we take a step further into the excited-state regime through time-dependent density functional theory (TDDFT), a natural and elegant extension of DFT that opens a window into the absorption and photoluminescent properties of these systems. Together, these two complementary frameworks allow us to explore and understand light-matter interactions at their fundamental level. This dissertation is built around two central objectives, the first of which is addressed in Chapters 3 and 4, where we explore how surrounding environmental factors influence the optical properties of noble metal nanomaterials. In Chapter 3, we investigate the effect of static electric fields on three geometrically distinct Ag₁₀ dimer configurations: end-to-end, parallel, and 90° angled arrangements, using TDDFT. We systematically explore how the direction of the applied electric field affects the optical response of these systems, as well as the resulting changes in their molecular orbital characteristics. Building on this theme, Chapter 4 extends our investigation to the electronic structure and excitation characteristics of the Au₄₂(SCH₃)₃₂ nanorod, examined across different solvation media. This nanorod is of particular interest due to its experimentally and theoretically validated near-infrared (NIR) absorption peak and dual emission character in its photoluminescence spectra. In this chapter, we further probe its molecular orbital characteristics and investigate the sensitivity of this NIR absorption peak to changes in the surrounding environment, deepening our understanding of how solvation conditions modulate the optical behavior of this system. The second objective of this dissertation shifts focus towards reducing the computational cost of conventional theoretical frameworks without compromising their accuracy, addressed in Chapters 5 and 6. In Chapter 5, we introduce a machine learning force field (MLFF) for gold hydride nanoclusters, capable of efficiently predicting relative potential energies across diverse cluster configurations at DFT-level accuracy, offering a computationally tractable alternative where exhaustive configurational sampling would otherwise be expensive. Chapter 6 extends the TDDFT plus tight binding (TDDFT+TB) method, which is an approximation to TDDFT, to explore the excited-state potential energy surfaces of open-shell molecular systems, through the derivations and implementation of unrestricted analytical excited-state gradients. A key advantage of TDDFT+TB is its computational efficiency; by introducing approximations at the level of the two-electron integrals, which is the primary computational bottleneck, the method achieves a dramatic reduction in cost compared to conventional TDDFT, while retaining accuracy. Overall, this dissertation explores light-matter interactions from a theoretical perspective, investigating the influence of external environmental factors on the optical properties, while advancing efficient computational approaches that push the boundaries of theoretical chemistry without sacrificing accuracy.