This thesis focuses on the development of two methods of nanoparticle synthesis, the applications of the nanoparticles and the characterisation, classification, and behaviour of the nanoparticle structures. Building on the previously developed Matrix Assembly Cluster Source (MACS) technology, gold nanoparticles were fabricated using a newly created MiniMACS system, altered by using butane as a matrix gas which deposits using readily available liquid nitrogen (at a temperature of 77 K). Direct sputtering of atoms from a target material was also developed in this system, displaying the versatility of the system by the ease of switching between the two methods. This Sputtered Atom Source (SAS) technique was calibrated and subsequently used to create nanoparticles by growth of gold clusters from atoms on a carbon surface then were classified according to the cluster structure. After analysis of over 600 gold clusters identified in aberration corrected Scanning Transmission Electron Microscopy (STEM) images, it was found that clusters smaller than 300 atoms tend to have a glassy/amorphous structure, while above this size decahedral and FCC (Face – Centred Cubic) motifs compete for the dominant structure, with FCC being slightly more favourable. Icosahedral was observed very rarely when carrying out this analysis. Dynamic analysis was also undertaken by studying a video of 500 frames of thesame two clusters showing a coalescence event. Before aggregation, the larger cluster analysed structurally consisted of less than the 309 atoms that represent an “ideal” magic number cluster and crosses this threshold after the event has taken place. The dominant structure went from a prevailing decahedral shape before coalescence to an amorphous/glassy configuration, with the second most commonly occurring structure being FCC. The decahedral shape appeared much less frequently post aggregation. The chamber housing the MiniMACS was adapted to include two probes to measure resistance in-situ while depositing atoms using the SAS method. Fabrication of gold contacts on a suitable silicon substrate was successfully achieved, allowing for the observation of a resistance profile during creation of a percolating network of gold clusters. From here, exposure to the atom beam could be correctly terminated at a point just below the percolation threshold, allowing for the creation of atomic filament growth and potential with potential to image these in the TEM. Calibration of a larger scale MACS system was undertaken to create silver nanoparticles deposited on both copper and silver foils for use in catalysis. Lead nanoparticles were also deposited into carbon fibre paper to fabricate an electrode for use in water purification. This feat demonstrated that the MACS method is approaching rates needed for research and development in industry of the production of fuel cells, creating clusters at a rate of around 0.05 mg/cm2/hour. This fabrication also demonstrated the depth the clusters can penetrate a porous structure, with strong signal found in the top 50 μm, and then a diminished signal at around 150 – 200 μm,numbers which correlate to the carbon paper’s pore size. This project has gone beyond that of the work carried out in the literature, whereby this reports on the first instance in which clusters created with the MACS method, with clusters been fabricated at a higher temperature than before, and as butane with a matrix gas. Surface grown clusters were also discovered using readily available instruments that the MACS uses, allowing for quick switching between the two techniques. The non-mass selected nature of the clusters allowed the observation of single atoms, allowing investigations into the number of atoms surface grown clusters contain, classified by their structure. Coating carbon fibre paper with lead clusters accomplished both a proof-of-concept reactive oxygen species for water purification, as well as demonstrating a high cluster yield. Finally, instrumentation development, beam calibration and exposure recipes have lead way to large cluster coverage on the scale of percolating thing films for use of imaging filaments in a memristor device. This work has laid the foundation for further work to be developed.