In this thesis, a high vacuum modified magnetron sputtering system has been used to create very clean nanoparticles of silver, gold and nickel. These nanomaterials all have a diverse range of applications, and their usefulness in these applications is dependent upon their unique properties associated with their size and shape. The properties of nanoparticles can degrade and become essentially useless, by coalescence processes and via interaction with environmental contaminants, in particular oxygen. In this thesis, Transmission Electron Microscopy (TEM) techniques have been utilised to explore the properties of these nanoparticles, which are extremely clean compared to nanoparticles fabricated through many other techniques. Having cleaner samples means that TEM studies of the particles are more representative of the metals themselves, rather than any organic additives or other by products of synthesis. It also provides opportunities to study how these nanomaterials degrade, both naturally over time, or by forcing oxidation or coalescence using a TEM beam. In this thesis, conclusions have been made about how these forced processes compare with natural ageing of the materials. In particular, it was found that silver coalescence is dominated by morphological considerations, and proceeds by a 'collapsing' of morphologies from icosahedral to decahedral to tetragonal. Enhanced stability was observed for decahedral shapes, together with a resistance to coalescence and Ostwald Ripening processes. The coalescence of silver was compared to the coalescence of gold and both these processes were used to attempt to understand the wetting of silver on gold nanoparticles. The current understanding of this recently observed phenomenon has been furthered, giving consideration to the importance of facets and morphology in this process. In addition to coalescence of gold and silver nanoparticles, nickel nanocubes have been created in the mag- netron system, which is a novel route to their synthesis, and their oxidation has been studied during in-situ TEM experiments. The preferential oxidation of {111} facets over {100} facets has been observed and have therefore it has been concluded that different morphologies of nickel nanoparticles will oxidise differently. Since shape control is highly important in many applications of magnetic nanoparticles, and oxidation is the most important degradation route, this result is of great significance.