Marine aerosols impact the climate by interacting with incoming solar radiation, serving as nuclei for cloud formation and providing interfaces for heterogeneous and multiphase atmospheric chemistry. Individual aerosols vary widely in their size, chemical profile and internal mixing states; however, current single-particle analysis techniques in the laboratory are not able to access the nanosecond to millisecond timescale dynamics within these aerosols, which ultimately guide their climate-relevant properties. Thus GPU-accelerated all-atom molecular dynamics (MD) simulations have become potent tools to complement laboratory investigations in this new field. In this work, the MD engines NAMD and GROMACS are used to model entire marine aerosols at the atomic level, with the goal to understand how aerosol surface properties are modulated by particle size and chemical composition. Model SSA are generated based on experimentally-determined aerosol lipidomic and proteomic information. They range in size from 5 to 200 nm in diameter and 1x10<sup>5</sup> to 1x10<sup>8</sup> atoms, system sizes which are only accessible with the aid of GPU-enabled supercomputing. The simulated particles vary in surface curvature as well as protein and lipid type and concentration. These simulations allow us to understand how particle size ultimately impacts the dynamics and interactions within aerosols at the micro scale, including protein and lipid partitioning, diffusion, and water transport across the interface, allowing us to make predictions about macroscale aerosol behavior. The variation in lipid and protein concentration—strongly modulated by the biochemistry at the ocean surface—allows us to connect ocean microbiology to aerosol properties. Integrating computational methods with experimental techniques advances our capacity to accurately model marine aerosols, inching us closer to unraveling the complexity of marine aerosol chemistry and how it shapes our climate.

Microphysical climate-relevant properties of model marine aerosols surfaces explored with GPU-accelerated all-atom molecular dynamics

2019 Fall National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.