Atmospheric aerosols significantly impact Earth's radiative forcing and human health. While inorganic aerosols are well-understood due to abundant measurements, the composition of organic aerosols (OA), comprising 20–90% of fine aerosol mass, remains largely uncharacterized. Although considerable progress has been made in quantifying OA emissions and atmospheric processes, research has primarily focused on carbon, neglecting the significant contribution of nitrogen (N) to many OA molecules. The abundance, sources, and atmospheric behavior of nitrogenous organic aerosols (OAN) are poorly understood. High-resolution mass spectrometry has revealed hundreds of N-containing organic molecules in ambient OA, some of which absorb light and contribute to climate change, participate in new particle formation, and influence the global nitrogen cycle. Existing quantitative knowledge of aerosol ON is limited to a few compound groups or subgroups, often focusing on the water-soluble fraction. The bulk ON analysis, including both water-soluble and water-insoluble fractions, has been challenging due to limitations of existing methods. This study addresses this gap by applying a recently developed aerosol IN&ON analyzer that enables sensitive and simultaneous quantification of inorganic and organic nitrogen without pretreatment. This new method allows for a comprehensive and accurate assessment of ON in numerous samples, providing valuable insights into the magnitude and sources of bulk OAN. The objective is to quantify ON in aerosol filter samples from various locations in China, establish the relative quantities of aerosol ON and inorganic nitrogen (IN), and identify major ON sources through source apportionment analysis, ultimately improving our understanding of OA composition and environmental impacts.

Previous studies have focused on specific groups of nitrogen-containing organic compounds in aerosols (e.g., urea, amines, amino acids, nitroaromatic compounds, nitro-PAHs) or on quantifying subgroups like organic nitrates using techniques such as aerosol mass spectrometry (AMS) and thermal dissociation coupled with optical measurement (TD-LIF). However, these approaches are often limited to the water-soluble fraction (WSON) or suffer from inaccurate measurements of trace-level aerosol nitrogen using techniques like the Carlo Erba elemental analyzer (EA). The “difference method” used to quantify ON (Total N - IN) also introduces significant errors, especially when ON is a minor component. The limited existing data highlight the need for a more comprehensive and accurate method for quantifying total aerosol ON, encompassing both water-soluble and water-insoluble fractions, to understand the full extent of OAN's role in atmospheric chemistry and environmental impacts.

This study utilized a newly developed aerosol IN&ON analyzer employing programmed thermo-evolution and chemiluminescent detection coupled with multivariate curve resolution data treatment. This innovative method provides sensitive and simultaneous quantification of aerosol IN and ON without pretreatment, overcoming the limitations of previous methods. More than 600 PM2.5 aerosol filter samples were collected from 12 sites across China, representing diverse urban influences (two suburban sites in the North China Plain (NCP), three suburban/rural sites, and seven urban sites in the Pearl River Delta (PRD)). Sampling spanned a whole year at most sites. In addition to ON and IN, organic carbon (OC), other major aerosol chemical constituents, source-specific elements, and organic molecules were measured. The joint OC and ON data were used to estimate the contribution of OAN to OA using Monte Carlo (MC) simulations, incorporating the OAN-to-ON ratio (A), the ON-to-OC ratio (B), and the reciprocal of the OA-to-OC ratio (C). The distribution of B was derived from the measured data, while A and C were constrained using literature values and prior studies. Sensitivity analyses were performed to evaluate the robustness of the results using various distribution assumptions and confidence levels. Source apportionment of ON was conducted using positive matrix factorization (PMF) analysis of a subset of samples with comprehensive speciation data (from urban TW, rural NS, and suburban BJ sites). This involved resolving primary and secondary sources of ON, including biomass burning, primary biological particle emissions, vehicular emissions, cooking emissions, industrial emissions, soil dust, sea salt and secondary formation processes (sulfate-rich and nitrate-rich factors and biogenic/anthropogenic SOAs). The ON/OC ratios for individual ON sources were also calculated to provide valuable data for atmospheric modeling studies. The details of the chemical analyses and the PMF and MC simulation techniques are presented in the supplementary materials.

The average concentrations of PM2.5-bound ON ranged from 0.4 to 1.4 µg N m⁻³ across the 12 sampling sites, representing 17–31% of total aerosol nitrogen. Higher IN and ON levels were observed in northern China compared to southern China. Urban areas in the PRD region showed slightly higher ON levels than suburban/rural areas. Monte Carlo simulations, constrained by paired ON and OC measurements, estimated that N-containing organic molecules contributed 37–50% (with a 95% confidence interval of [12%, 94%]) to ambient organic aerosols. Source apportionment analysis revealed biomass burning (21–24%) and secondary formation (~30%) as the dominant ON sources. Primary biological aerosol particles (PBAP) were also a significant source (7–18%), particularly in non-urban atmospheres. Seasonal variations were observed, with biomass burning most significant in winter and PBAP emissions prominent in summer. Secondary formation linked to sulfate-rich factors was important throughout the year in southern China, while at the BJ site, the nitrate-rich factor's contribution to aerosol ON was notably higher, reflecting higher NOx levels. The PMF analysis provided ON/OC ratios for individual sources; biomass burning had ratios comparable to experimental measurements of biomass burning smoke, while PBAP emission and soil dust exhibited higher ratios than fossil fuel combustion sources. Secondary ON sources showed a wide range of ON/OC ratios. The sensitivity analyses showed that the estimated OAN/OA values were robust to different assumptions about the distributions of parameters and confidence levels.

This study provides the first comprehensive quantification of bulk aerosol ON across diverse environments in China, revealing a substantial contribution of OAN (37–50%) to OA mass. The large contribution of N-containing molecules to OA has likely been underappreciated in previous studies. The detailed source apportionment analysis identifies biomass burning, secondary formation, and primary biological aerosol as major ON sources. Seasonal variations in source contributions highlight the dynamic interplay between different emission sources and atmospheric processes. The findings address the significant gap in our understanding of OA composition by providing quantitative data on bulk ON and its sources. This information is crucial for refining atmospheric models and improving assessments of OA's environmental impacts, particularly concerning atmospheric nitrogen nutrient inputs to ecosystems. The identified sources and ON/OC ratios for individual sources offer valuable information for improving existing atmospheric models and facilitating more accurate estimations of atmospheric nitrogen deposition fluxes.

This study demonstrates the significant contribution of nitrogenous organic molecules to ambient organic aerosols, highlighting the need to incorporate nitrogen into OA modeling and analyses. The quantitative data on bulk ON and its major sources (biomass burning, secondary formation, primary biological aerosols) provide critical information for refining our understanding of OA composition and its impact on the environment. Further research should focus on integrating bulk ON measurements with the quantification of individual N-containing organic molecules, and the development of online ON measurement techniques is recommended for advancing the study of OAN sources and chemical processes.

Potential limitations include the loss of semi-volatile organic compounds during sampling, which might introduce biases in the measurements of OC and ON and subsequent estimations of OAN/OA. The lack of tracer data for cooking emissions and SOA at some sites prevented the accurate resolution of their contributions to ON in the PMF analysis. However, the sensitivity tests conducted to evaluate the robustness of the results and the strong correlation observed between IN measurements conducted soon after sample collection and those conducted two to four years later suggest that the findings are robust. Further research incorporating more comprehensive speciation data and addressing sampling artifacts is warranted.