Development of a non-targeted LC-UHRMS approach for organic aerosol analysis: first application to urban and biogenically influenced air masses

The analysis of filter samples of atmospheric organic aerosols provides information about atmospheric processes and the origin of aerosol particles. However, limiting analysis to a few target compounds ignores a large proportion of the compounds present on the filters. Since the availability of high-resolution mass spectrometers, non-target analysis addresses parts of this problem. However, such an analysis can be very time-consuming. Since it is hardly possible to analyze all individual compounds manually, an automated or semi-automated evaluation of the data is required. This poster presents a method for the non-target analysis of atmospheric organic aerosol filter samples using UHPLC-Orbitrap-MS. The extracted filter samples are analyzed in a two-step process that provides maximum information. In the first step, a full-scan high resolution mass spectrum is measured, which is then analyzed with MZmine, capturing all compounds with their respective retention time and exact mass. Using this data, a second experiment is designed in which an isolated MS/MS spectrum (with stepped fragmentation energy) of all detected compounds is measured. With the MS/MS data of the measured compounds and a local database in combination with in-silico fragmentation, a reliable prediction of the chemical composition, functional groups and/or parts of the molecular structure is possible. The combination of these steps drastically improves the reliability of the prediction, as not only the exact mass of the molecule is considered, but also additional information about the fragmentation of the molecule is included. Python scripts automate the processes and create a comprehensible summary for each detected compound, minimizing the manual workload. For this contribution, filters of the ACROSS campaign 2022 (Rambouillet Forest, France), where urban and biogenically influenced air masses are present, were analyzed in the manner described above and a brief summary of the results is given.