Bridging experiments and simulations towards a mechanistic understanding of gold patchy nanoparticle formation

This is the dataset for the following manuscript "Bridging experiments and simulations towards a mechanistic understanding of gold patchy nanoparticle formation".

Abstract

The large-scale synthesis of tailored nanoparticles requires a detailed understanding of their formation processes. Yet, such mechanistic insights remain limited for anisotropic patchy nanoparticles (PNPs), which consist of a dielectric core particle partially coated with a metal patch. In this work , we unravel the multi-step formation mechanism of gold-on-polystyrene PNPs produced by a seed-mediated continuous flow process. Using a double T-mixer setup, we obtain kinetic data from inline UV-Vis-NIR extinction spectroscopy and correlate them with speciation calculations and ex situ Zeta potential and HAADF-STEM analyses. To resolve structural evolution, patch growth is quenched at selected residence times using glutathione, enabling microscopy of intermediates. The combined data support a two-channel picture of gold integration. One channel advances lateral spreading via incorporation at the patch rim, while a second contributes to thickening on the patch top surface. By tuning reaction kinetics via chloride and ascorbic acid concentrations, we shift the balance between the channels. Fast kinetics lead to transport-limited dendritic growth, whereas slower kinetics promote densification and, at the slowest rates, thickened patches with reduced spreading. Based on these insights, we develop a kinetic model for dense patch formation that reproduces trends across multiple syntheses and intermediate-time quenched samples. The model supports the hypothesis that the final patch shape emerges from the interplay between gold monomer fluxes toward the rim and top surface of the growing patch. Overall, this work advances the mechanistic understanding of anisotropic nanoparticle formation and provides a quantitative basis for targeted property design guided by mechanistic models.

The uploaded data are sorted by figure.