Elucidating the ligand shell structure and dynamics of Au683MBA32 gold nanocluster using molecular dynamics simulations

Synthesising novel gold nanoparticles and -clusters can be often easier than characterising them, and after experimental analysis many options for the possible molecular formula of the cluster may remain as equally valid options. These options may be investigated by quantum computational means, such as density functional theory, but since DFT includes the electronic interactions, this might take time and excessive effort if the amount of options is large. In this thesis a new way of easing the computational task is introduced by a novel method, that combines a guessing algorithm and molecular dynamics simulations. The algorithm produces possible structures by utilising existing structures of known clusters, and molecular dynamics are used to compute the dynamics and interactions in the structures to differentiate between a good guess and a bad one with lower computational cost. This method was utilised in this thesis on gold nanocluster Au683MBA32. Three models were generated using the algorithm with the help of the information gained in previous research (Pablo D. Jadzinsky et al. Science, 318 (5849):430–433, 2007 and Azubel et al. Science, 345(6199):909–912, 2014). The structures were simulated in a custom made GROMACS force-field (Pohjolainen et al. Journal of Chemical Theory and Computation, 12(3):1342-1350, 2016) for 50 ns in NPT conditions. The resulting trajectories were examined for indicators of internal stability and compared with experimental results.