Context <p>Waste plastics have a significant impact on natural resources and the environment. The focus is on the interfacial adhesion behavior of three types of waste plastics—polyethylene (PE), polypropylene (PP), and Poly acrylic (PA)—with neutral silica (SiO₂) aggregates, as well as the influence of graphene oxide (GO) on the asphalt-water-aggregate interface performance. The validity of the molecular dynamics simulations was assessed based on fundamental thermodynamic properties of asphalt, including density, glass transition temperature, cohesive energy density, and solubility parameters. The diffusion patterns of the four asphalt components on the SiO₂ aggregate surface were analyzed using the mean square displacement (MSD), diffusion coefficient, and relative concentration distribution. The bonding strength between various waste plastics and the asphalt-aggregate interface was quantified by calculating the adhesion work. The results indicate that van der Waals forces and electrostatic forces play a crucial role in the adhesion between asphalt and minerals. The adhesion work between aged asphalt and aggregates is highest when PE plastic is added, lowest when PP plastic is added, and intermediate when PA plastic added. Upon introducing water molecules at the asphalt-aggregate interface, the interfacial adhesion work decreased significantly, indicating that water molecules exert a substantial negative impact on interfacial adhesion performance. However, after coating the aggregate surface with graphene oxide, the adhesion work at the asphalt-water-aggregate interface increased significantly, suggesting that graphene oxide enhances water damage resistance by improving interfacial polarity. This study provides molecular-level theoretical support for the recycling of waste plastics in asphalt-based materials and for the optimization of interfaces.</p> Methods <p>Molecular dynamics simulations were performed using Materials Studio software with the COMPASS II force field. The Nose thermostat and Berendsen isobaric thermostat were used to control temperature and pressure, respectively, to maintain constant temperature and pressure conditions throughout the simulation. In non-bonded interaction calculations, electrostatic forces were calculated using the Ewald method, and van der Waals forces were calculated using the Atom Based method, with a cutoff radius of 15.5&#xa0;Å. The simulations were conducted using NVT and NPT ensembles, with all simulations set to fine computational precision. The constructed model was first subjected to geometric optimization (100,000 iterations), followed by annealing treatment (10 cycles with a temperature change gradient of 100&#xa0;K), and then the ideal molecular structure of the model was obtained through NPT ensemble simulations.</p>