Search for: All records

Abstract. Evidence has accumulated that secondary organic aerosols (SOAs) exhibit complex morphologies with multiple phases that can adopt amorphous semisolid or glassy phase states. However, experimental analysis and numerical modeling on the formation and evolution of SOA still often employ equilibrium partitioning with an ideal mixing assumption in the particle phase. Here we apply the kinetic multilayer model of gas–particle partitioning (KM-GAP) to simulate condensation of semi-volatile species into a core–shell phase-separated particle to evaluate equilibration timescales of SOA partitioning. By varying bulk diffusivity and the activity coefficient of the condensing species in the shell, we probe the complex interplay of mass transfer kinetics and the thermodynamics of partitioning. We found that the interplay of non-ideality and phase state can impact SOA partitioning kinetics significantly. The effect of non-ideality on SOA partitioning is slight for liquid particles but becomes prominent in semisolid or solid particles. If the condensing species is miscible with a low activity coefficient in the viscous shell phase, the particle can reach equilibrium with the gas phase long before the dissolution of concentration gradients in the particle bulk. For the condensation of immiscible species with a high activity coefficient in the semisolid shell, the mass concentration in the shell may become higher or overshoot its equilibrium concentration due to slow bulk diffusion through the viscous shell for excess mass to be transferred to the core phase. Equilibration timescales are shorter for the condensation of lower-volatility species into semisolid shell; as the volatility increases, re-evaporation becomes significant as desorption is faster for volatile species than bulk diffusion in a semisolid matrix, leading to an increase in equilibration timescale. We also show that the equilibration timescale is longer in an open system relative to a closed system especially for partitioning of miscible species; hence, caution should be exercised when interpreting and extrapolating closed-system chamber experimental results to atmosphere conditions. Our results provide a possible explanation for discrepancies between experimental observations of fast particle–particle mixing and predictions of long mixing timescales in viscous particles and provide useful insights into description and treatment of SOA in aerosol models. 

Wildfires, which have been occurring increasingly in the era of climate change, emit massive amounts of particulate matter (PM) into the atmosphere, strongly affecting air quality and public health. Biomass burning aerosols may contain environmentally persistent free radicals (EPFRs, such as semiquinone radicals) and redox-active compounds that can generate reactive oxygen species (ROS, including ·OH, superoxide and organic radicals) in the aqueous phase. However, there is a lack of data on EPFRs and ROS associated with size-segregated wildfire PM, which limits our understanding of their climate and health impacts. We collected size-segregated ambient PM in Southern California during two wildfire events to measure EPFRs and ROS using electron paramagnetic resonance spectroscopy. EPFRs are likely associated with soot particles as they are predominantly observed in submicron particles (PM 1 , aerodynamic diameter ≤ 1 μm). Upon extraction in water, wildfire PM mainly generates ·OH (28–49%) and carbon-centered radicals (∼50%) with minor contributions from superoxide and oxygen-centered organic radicals (2–15%). Oxidative potential measured with the dithiothreitol assay (OP-DTT) is found to be high in wildfire PM 1 , exhibiting little correlation with the radical forms of ROS ( r 2 ≤ 0.02). These results are in stark contrast with PM collected at highway and urban sites, which generates predominantly ·OH (84–88%) that correlates well with OP-DTT ( r 2 ∼ 0.6). We also found that PM generated by flaming combustion generates more radicals with higher OP-DTT compared to those by smoldering or pyrolysis. 

Gas-particle partitioning of secondary organic aerosols is impacted by particle phase state and viscosity, which can be inferred from the glass transition temperature ( T g ) of the constituting organic compounds. Several parametrizations were developed to predict T g of organic compounds based on molecular properties and elemental composition, but they are subject to relatively large uncertainties as they do not account for molecular structure and functionality. Here we develop a new T g prediction method powered by machine learning and “molecular embeddings”, which are unique numerical representations of chemical compounds that retain information on their structure, inter atomic connectivity and functionality. We have trained multiple state-of-the-art machine learning models on databases of experimental T g of organic compounds and their corresponding molecular embeddings. The best prediction model is the tgBoost model built with an Extreme Gradient Boosting (XGBoost) regressor trained via a nested cross-validation method, reproducing experimental data very well with a mean absolute error of 18.3 K. It can also quantify the influence of number and location of functional groups on the T g of organic molecules, while accounting for atom connectivity and predicting different T g for compositional isomers. The tgBoost model suggests the following trend for sensitivity of T g to functional group addition: –COOH (carboxylic acid) > –C(O)OR (ester) ≈ –OH (alcohol) > –C(O)R (ketone) ≈ –COR (ether) ≈ –C(O)H (aldehyde). We also developed a model to predict the melting point ( T m ) of organic compounds by training a deep neural network on a large dataset of experimental T m . The model performs reasonably well against the available dataset with a mean absolute error of 31.0 K. These new machine learning powered models can be applied to field and laboratory measurements as well as atmospheric aerosol models to predict the T g and T m of SOA compounds for evaluation of the phase state and viscosity of SOA. 

Secondary organic aerosol (SOA) plays a critical, yet uncertain, role in air quality and climate. Once formed, SOA is transported throughout the atmosphere and is exposed to solar UV light. Information on the viscosity of SOA, and how it may change with solar UV exposure, is needed to accurately predict air quality and climate. However, the effect of solar UV radiation on the viscosity of SOA and the associated implications for air quality and climate predictions is largely unknown. Here, we report the viscosity of SOA after exposure to UV radiation, equivalent to a UV exposure of 6 to 14 d at midlatitudes in summer. Surprisingly, UV-aging led to as much as five orders of magnitude increase in viscosity compared to unirradiated SOA. This increase in viscosity can be rationalized in part by an increase in molecular mass and oxidation of organic molecules constituting the SOA material, as determined by high-resolution mass spectrometry. We demonstrate that UV-aging can lead to an increased abundance of aerosols in the atmosphere in a glassy solid state. Therefore, UV-aging could represent an unrecognized source of nuclei for ice clouds in the atmosphere, with important implications for Earth’s energy budget. We also show that UV-aging increases the mixing times within SOA particles by up to five orders of magnitude throughout the troposphere with important implications for predicting the growth, evaporation, and size distribution of SOA, and hence, air pollution and climate.