• bio-functionalised metal nanoparticles
• colloidal systems
• mesoscale model
• molecular dynamics

Gold nanoparticles have the capability to be functionalized with a variety of chemical groups in order to inhibit the aggregation of proteins into fibrillar amyloid deposits responsible of the amyloid diseases such as the degenerative arthritis. The efficiency of this process is reduced by the NPs self aggregation. Therefore, it is essential to determine the aggregation properties of NPs for different environmental conditions.
Molecular dynamics (MD) simulations permits to study how the NP behavior changes with their shape, size, charge and functional groups and by increasing the concentration within the different conditions found in the cell environment. However, in MD simulations performed with atomistic representations, the explored time and space scales are limited by the computational resources. This obstacle can be overcome by reducing the model resolution and by eliminating the degrees of freedom of the solvent water atoms and ions building a new effective potential for the NPs.
In this thesis work all-atoms MD simulations of isolated or couples of NPs in different environmental conditions are analyzed and used as the basis to costruct a Mesoscale model representing each NP as a single spherical object. The analysis allows to parametrize an interaction potential, decomposed in two terms, one long range term describing the electrostatics, and another short range term describing the NP size and hydrophobicity.
The obtained empirical potentials are then used in large-scale MD simulations, whose analysis gives information on the aggregation properties of the NP considered, for a given temperature and concentration. NPs which interact with a short range attractive and long range repulsive potential are found to exhibit clustered particle states, while the phase diagram of NPs in which repulsion dominates show the existence of a fluid, body-centered cubic and face-centered cubic phases, depending on the concentration and ionic strength values.
These results are compared and found to be in agreement with those present in literature for colloidal systems interacting with similar potentials. Additionally, the dependence of the parametrized interaction potential on parameters connected to ion concentration, NP size, charge and hydrophobicity, allow predicting the aggregation properties also for particles with different characteristics and to identify the optimal combinations of NP properties to have them acting as antiaggregats.