This project is concerned with behavior of 2D materials in liquid environments, specifically graphene oxide (GO) and reduced graphene oxide (rGO) in aqueous solutions. Reduced graphene oxide is of interest for large‐scale graphene deposition since it can be produced cost efficiently in large quantities. It contains, however, a very high density of defects that strongly impact the wetting properties, leading e.g., to changes in the rheological properties, not only quantitatively, but also qualitatively depending on concentration and pH levels. A variety of defects and defect clusters have been identified, including e.g., vacancies and vacancy clusters, adatoms, carboxyl side groups, line defects as well as grain boundaries. The functional groups are charged species which causes long-range electrostatic interactions, as well as intrinsic renormalization of elastic properties. The overall, long term, objective of this project is to systematically characterize the influence of these defects and build a scale-bridging model for the static and dynamic properties of rGO/GO dispersions. In this first phase, we seek an understanding of the wetting properties and their dependence on sp2/sp3 ratio and degree of disorder. The aims are (1) to quantitatively predict the wetting angles of water droplets on GO/rGO, with and without hydrophilic substrates and liquid environments, and (2) to obtain a sufficient understanding of the physical mechanisms to be able to proceed with further investigations of the properties of aqueous rGO/GO dispersions. The work planned for this application is thus a prerequisite for future studies to elucidate the mechanisms behind phenomena such as transitions to liquid crystalline phases and complex rheological behaviors, for instance thixotropy and shear thinning. In our approach, we will use molecular dynamics (MD) to simulate rGO/GO in contact with water in different settings, for instance changing substrates or other external environment parameters. For this purpose, we will use LAMMPS, which can benefit from CPU-parallelization as well as GPU-acceleration. Our preliminary benchmarks and calibrations on small systems (50 000 atoms) have shown that the interatomic force fields OPLSAA, TIP4P, together with long range Coulombic interactions (k-space) produce good results. For proceeding further, we will need to simulate systems with up to 10^6 atoms. Our current limited HPC resources (the physics department at Chalmers shares a joint allocation on the Vera-cluster on a first-come-first serve basis) are not sufficient for the required upscaling. Resource wise, the MD simulations are not memory intensive but require several cores to run efficiently. Post processing of data includes calculating standard quantities such as radial distribution functions, rms-fluctuations, etc. For this we will use Python scripts together with Ovito. These latter tasks are more memory intensive than the MD simulations, but the number of core-hours dedicated to post processing is significantly less than those for MD. The vast majority of postprocessing is done on ordinary workstations and not on the cluster.