Air pollution is a significant global health concern, causing millions of premature deaths annually. China, despite implementing mitigation strategies, still faces substantial air quality challenges, with a large fraction of fine particulate matter (PM2.5) composed of secondary inorganic aerosol (SIA) and SOA. While SIA reductions have been achieved, the expected health benefits haven't fully materialized, highlighting the critical need to understand SOA sources and formation pathways. The complexity of SOA formation, involving multiple gaseous organic compounds and atmospheric aging, obscures source identification. Traditional mass spectrometers fragment organic molecules, hindering precursor identification. However, recent advancements in soft-ionization mass spectrometry offer enhanced molecular speciation and time resolution, improving the potential to identify SOA sources. This study addresses the knowledge gap by using advanced techniques to identify and quantify SOA sources and their variability in Beijing, a megacity frequently affected by severe haze.

Previous research has established the significant contribution of secondary organic aerosols to haze in Beijing and its detrimental health impacts. Studies using aerosol mass spectrometry have provided insights into the bulk chemical composition of PM2.5 in Beijing, showing a substantial fraction of organic aerosol. However, the detailed sources and formation pathways of this organic aerosol have remained largely unknown, hindering the development of effective pollution control strategies. While some studies have attempted source apportionment using various techniques, the complexity of the chemical composition and the lack of near-molecular information have limited the accuracy and resolution of these efforts. The application of newly developed soft-ionization mass spectrometers has opened new possibilities for characterizing the molecular composition of organic aerosols with high time resolution, enabling a more precise identification of SOA sources. This study builds upon these advancements by combining advanced source apportionment techniques with highly time-resolved near-molecular characterization to unravel the intricacies of SOA formation in Beijing.

This study utilized a combination of techniques to characterize organic aerosol (OA) in Beijing. Quantitative OA measurements were performed using a Time-of-Flight Aerosol Chemical Speciation Monitor (TOF-ACSM), providing bulk chemical composition data including OA, nitrate, sulfate, chloride, and ammonium. A seven-wavelength aethalometer measured equivalent black carbon (eBC). Near-molecular information was obtained using a Filter Inlet for Gases and AEROsols coupled to an iodide Chemical Ionization Mass Spectrometer (FIGAERO-CIMS). This instrument uses soft chemical ionization to detect molecular ions and their chemical formulas. The data analysis involved positive matrix factorization (PMF) for source apportionment, separately applied to TOF-ACSM and FIGAERO-CIMS data. For FIGAERO-CIMS data, a multilinear regression (MLR) approach was used to quantify the mass loadings of different SOA types, using OA minus (HOA + COA) from the TOF-ACSM as a reference. Air mass back trajectories were also used to determine the geographical origin of the different SOA components. The study analyzed data from winter (November 2019 – January 2020), a COVID-19 lockdown period (January-April 2020), and summer (May-July 2020). Various data processing steps, including background correction and field blank subtraction, were applied to the FIGAERO-CIMS data, and uncertainty estimates were carefully considered. A sensitivity assessment was conducted to evaluate the impact of uncertainties in instrument response factors on the results. In addition to online measurements, offline filter samples were analyzed to quantify levoglucosan, methylbutanetricarboxylic acid (MBTCA), and pinic acid.

The study revealed substantial seasonal and compositional differences in OA sources. In winter, primary solid-fuel emissions (SFOA) dominated, comprising more than half of the OA during severe haze events. Secondary solid-fuel SOA (sfSOA) and aqueous-phase SOA (SOAaq) were also major contributors. During severe winter haze episodes, secondary solid-fuel SOA could comprise up to 80% of the OA. Air mass back trajectories identified the Beijing–Tianjin–Hebei region and rural mountainous areas west of Beijing as significant sources of wintertime solid-fuel SOA precursors. In summer, SOA was predominant, with aromatic emissions from the Xi’an–Shanghai–Beijing region being the main driver. Biogenic SOA played a minor role, even in summer. Aromatic-dominated SOA increased during the day, consistent with photochemical processes. The long-range transport of precursors was highlighted by the observation that air masses containing high amounts of aromatic SOA had ages of up to 2 to 3 days. During the COVID-19 lockdown, while primary emissions decreased, secondary PM2.5 formation continued, highlighting the importance of secondary formation processes. The study also found that the chemical composition of the different SOA factors varied substantially and provided characteristics for those various factors (e.g., sfSOA, aromSOAday, aromSOAnight, bioSOAday, bioSOAnight, SOAaq). The study identified specific compounds characterizing these categories (e.g., levoglucosan, small dicarboxylic acids, nitroaromatics, etc.), allowing for a more precise identification of the respective emission sources. The sensitivity analysis showed that the relative contribution of various SOA components is subject to uncertainty related to instrument response factors, but that this uncertainty does not affect the main conclusions of the study.

The findings highlight the crucial role of transported emissions in contributing to organic aerosol pollution in Beijing. The significant contribution of SOA from regions outside Beijing underscores the necessity of regional collaborative efforts for effective pollution mitigation. The dominance of solid-fuel SOA in winter and aromatic SOA in summer reflects the seasonal variations in emission sources and atmospheric processes. The persistence of secondary PM2.5 formation even during the COVID-19 lockdown emphasizes the importance of targeting secondary organic aerosol precursors for effective air quality improvement. The detailed molecular characterization of OA provided unprecedented insights into the chemical composition and sources of different SOA types, significantly improving the understanding of SOA formation pathways and informing more targeted pollution control strategies. The combined use of TOF-ACSM and FIGAERO-CIMS data provided a more comprehensive picture of OA sources than either instrument alone, demonstrating the benefits of integrating multiple measurement techniques.

This study provides comprehensive insights into the sources and formation of organic aerosol in Beijing, revealing the substantial contribution of transported emissions. The findings underscore the limitations of local-only mitigation efforts and highlight the urgent need for regional cooperation to address this large-scale pollution problem. Future research could focus on refining the quantification of individual sources, improving the understanding of chemical transformations and aging processes, and exploring the health implications of specific organic aerosol components identified in this study.

The study acknowledges limitations in the quantification of certain OA components due to uncertainties in instrument response factors. Although a sensitivity analysis was performed, there remains potential for inaccuracies in the quantification of specific OA types, and some organic aerosols are not captured by the instruments used. The air mass back trajectory analysis only considered transport up to 72 hours prior to measurement, which might not fully capture the long-range transport of some precursors. The study focused on a single urban location in Beijing, and the findings may not be fully generalizable to all areas within the city or other regions with different emission characteristics.