| 04 Mar 2025
Growth of Atmospheric Freshly Nucleated Particles: A Semi-Empirical Molecular Dynamics Study
Abstract. When simulating new particle formation rates, collisions in the system are approximated as hard spheres without long-range interactions. This simplification may lead to an underestimation of the actual formation rate. In this study, we employ semi-empirical molecular dynamics (SEMD) at the GFN1-xTB level of theory to probe the sticking process of the monomers sulfuric acid (SA), methanesulfonic acid (MSA), nitric acid (NA), formic acid (FA), ammonia (AM), methylamine (MA), dimethylamine (DMA), and trimethylamine (TMA) onto freshly nucleated particles (FNPs). The FNPs considered are (SA)10 (AM)10, (SA)10 (MA)10, (SA)10 (DMA)10, and (SA)10 (TMA)10.
In general, we find that the hard-sphere kinetic approximation, which neglects long-range interactions, significantly underestimates the number of collisions leading to sticking. By calculating the sticking coefficient from SEMD simulations, we obtain enhancement factors of 2.3 and 1.5 for the SA+(SA)10 (AM)10 and AM+(SA)10 (AM)10 collisions, respectively. A comparison with OPLS all-atom force field simulations shows similar enhancement factors of 2.4 and 1.6 for the SA+(SA)10(AM)10 and AM+(SA)10(AM)10 collision, respectively
Compared to the force field simulations, SEMD exhibits a more isotropic sticking behavior, with the probability remaining near unity for small offsets before rapidly dropping to 0 % beyond a certain offset. In contrast, the force field simulations show a more gradual decline in sticking probability due to certain orientations still leading to sticking. The largest discrepancy between the two methods occur at lower collision velocities—below 200 m/s for SA and below 400 m/s for AM—where force field simulations, even for head-on collisions, predict low or zero sticking probability. This has previously been attributed to periodic repulsions between the rotating collision partners caused by fluctuations in their charge distributions. In contrast, SEMD simulations do not exhibit this behavior. Since these low velocities are not significantly populated in our simulations, both methods yield similar enhancement factors. However, for systems with larger effective masses, where such velocities are more prevalent, we would expect the two methods to diverge.
Competing interests: Jonas Elm is a member of the editorial board of Aerosol Research.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.
