Advanced Identification of Atmospheric Compounds

ongoing

Global warming and air pollution are urgent challenges facing humanity. Atmospheric aerosol particles contribute to both by affecting air quality, cloud formation, and the absorption and reflection of sunlight. Increasingly, the atmosphere is shaped by organic rather than inorganic chemistry. While initiatives like the European Green Deal have reduced inorganic emissions, volatile organic compounds (VOCs) from biogenic and anthropogenic sources are on the rise. These VOCs react to form oxidized compounds that drive aerosol particle formation. Understanding the molecular processes leading to particle formation is crucial for tackling climate change and improving air quality, with accurate compound identification being a key component of this effort. However, atmospheric science currently lacks the tools and infrastructure to identify most compounds in aerosol samples collected in field campaigns. The main challenge is the absence of reference standards for atmospheric compounds in mass spectral (MS) libraries, which are needed to match unknown spectra with known references. Experimentally creating such a library is difficult due to the lack of reference samples and the large number and reactive nature of atmospheric compounds. Through CLOUDMAP, I will address these challenges by creating an MS library using computational methods. I will develop a comprehensive dataset of mass spectra of pure atmospheric compounds via simulations, alongside a machine learning tool for enhanced compound identification. This tool will be used to identify unknown compounds from aerosol samples collected during field campaigns. By integrating quantum chemical MS simulations, machine learning, and atmospheric science, CLOUDMAP will advance the digitization of atmospheric chemistry, improve the understanding of aerosol formation, and contribute to both climate change mitigation and improved air quality.