Puy de Dôme Station (France): A Stoichiometric Approach to Compound Classification in Clouds.

Seven cloud water samples were collected from May to October 2018 at the Puy de Dôme station (PUY) in France and analyzed by positive‐ion atmospheric pressure photoionization [(+)APPI] Fourier transform ion cyclotron resonance mass spectrometry. The assigned formulas (ranging from 3,865 to 6,380) were attributed using the multidimensional stoichiometric constraint classification of Rivas‐Ubach et al. (2018, https://doi.org.10.1021/acs.analchem.8b00529) to six main categories (RUCs): LipidC, ProteinC, Amino‐sugarC, CarbohydrateC, NucleotideC, and OxyaromaticC. Back trajectories were calculated by the computing atmospheric trajectory tool (CAT) model to obtain information on the air mass history. Partial least square regressions were performed using chemical data, CAT back‐trajectory calculations and FT‐ICR MS data to analyze the environmental variability of the organic sample composition. ProteinC is correlated with the continental surface for air masses transported within the boundary layer, and Amino‐sugarC is strongly correlated with acetate, NO3− and NH4+, suggesting Anthropogenic sources for amino sugars and proteins. LipidC is correlated with the sea surface for air masses transported within the free troposphere, confirming the long‐range transport of marine biogenic sources. Concerning Oxy‐aromaticC, given the correlations with oxidants and pollutants, as well as anti‐correlations with local influence, we proposed a mechanism of oxidation from remote anthropogenic sources. Plain Language Summary: Clouds were sampled on top of the Puy de Dôme mountain in France to study their chemical composition. Cloud droplets were collected by impaction with samplers specifically designed for that. The seven samples, collected from May to October 2018, were characterized by a high‐resolution mass spectrometry method, revealing thousands of organic compounds carrying carbon, hydrogen, nitrogen, sulfur, and phosphorus atoms, belonging to both natural and anthropogenic sources. Using a recently developed classification method, organic compounds were shared into classes based on their respective numbers of elements to lipids, proteins, amino sugars, carbohydrates, and aromatic compounds. To better understand the variability of the molecular fingerprints of each collected sample, we applied statistical analysis which enabled us to link the history of the air‐masses calculated with a three‐dimensional kinematic trajectory code and the chemical composition of the clouds. For example, proteins were related to the time spent by the air mass above continental surfaces at low altitude (in the boundary layer). Amino sugars and aromatic compounds were strongly correlated with anthropogenic sources. Finally, lipids were correlated with the time spent over the ocean in the free troposphere, confirming potential long‐range transport from marine source. Key Points: Analyzing cloud water by high resolution‐mass spectrometry shows the presence of non‐polar, semi‐polar and phosphorus‐containing compoundsA multidimensional stoichiometric constraint classification was applied to assign molecular formulas according to six main categoriesA statistical study highlights correlations between these categories and air mass history, such as proteins with continental boundary layer [ABSTRACT FROM AUTHOR]