A Multiscale Approach to Unravelling the Structure and Infrared Spectra of Astronomically Relevant Nanosilicates

[eng] Silicates are ubiquitously found both terrestrially and throughout the universe where they are often found as small particles. In the interstellar medium (ISM), these particles undergo high-energy processes that cause them to shatter into ultrasmall grains, resulting in a vast population of nanosilicates. These nanosized grains are likely to play a crucial role in various chemical processes in space, such as catalysing the formation of molecules. Infrared (IR) spectroscopy is the primary tool for identifying silicate grains and analysing their properties and chemical composition. Such analyses are typically performed by empirical comparisons of astronomical observations with spectra of laboratory-made samples. However, this approach only provides indirect information about atomistically detailed properties and typically employs bulk silicate samples. Consequently, most knowledge about nanosilicates relies on top- down extrapolations from bulk properties. As such, much of our current understanding of nanosilicate grains is likely to be unreliable due to the known strong size dependence of structure and properties of materials at the nanoscale. Thanks to the advances in technology and theoretical methods, computational modelling has revolutionized our ability to interpret experimental results and astronomical observations. In particular, such modelling can probe materials at conditions that are challenging to replicate in the laboratory. (e.g. at very small sizes and under extreme environments). In the case of nanosilicates, computational modelling offers a powerful alternative approach to improve our understanding of their properties and to potentially help confirm their presence and abundance in the universe together with astronomical observations and laboratory experiments. This thesis expands upon previous theoretical models to study nanosilicates aiming to provide a more comprehensive study of their properties. Here, the main particular focus is on improving our knowledge and understanding of these nanosystems through their IR spectra. The thesis is structured as follows: first, a general overview of silicates, their astronomical significance, and the important role of nanosilicate grains is presented, along with a review of the current state-of-the-art computational modelling approaches. Next the theoretical foundations of all the methodologies used to build our models are described in depth, as well as the different software used. Afterwards, the different research publications derived from this work are discussed providing, for each one, an introduction to the theme and motivation of the publication and a brief discussion of the results and the conclusions that can be extracted from them. First, we explore various methodologies for accurately computing vibrational spectra, emphasizing their ability to incorporate thermal effects. We then compute the heat capacities of small nanograins comparing them with previous estimates and analysing the implications for their IR emission. Next, we investigate the properties of slightly hydroxylated nanosilicates, particularly their ability to emit microwave radiation. This is followed by a collaborative study combining computational and observational modelling to assess the potential detectability of small nanosilicate grains using. We further discuss two additional collaborative works to study cationic and anionic nanosilicates synthetized in the lab, examining their relevance to both astronomy and atmospheric chemistry. Finally, we study the size dependency of different nanosilicate properties, including IR spectra through modelling nucleated nanosilicate growth. The thesis concludes with a summary of the main conclusions derived from these works.