Characteristics and gas-particle partitioning of amine-containing particles from firework emissions: New insights from machine learning and single-particle mass spectrometry

Firework displays during the Spring Festival (SF) trigger intense short-term emissions and severe air quality deterioration, yet their effects on particulate amines remain poorly understood. This study conducted field observations during the 2023 SF using single-particle aerosol mass spectrometry combined with source sample analysis to investigate the emission characteristics and transformation mechanisms of amine-containing particles. Results show that amine-containing particles increased by 96.2% during the firework period, accompanied by a peak size shift from 0.42 µm to 0.44 µm and enhanced internal mixing with sulfate. Firework emissions altered mixing states by elevating metals and mixed elemental-organic carbon particles but reducing biomass burning and secondary components. Source sample analysis confirmed fireworks as the primary contributor, with N-methylaniline emerging as a potential tracer distinguishing fireworks from firecrackers. Fireworks increased modified aerosol acidity (R_ra′) and altered the atmospheric acid-base balance. The multi-model comparative framework identified R_ra′ as the dominant driver governing gas-particle partitioning of amines during the FW period, where aerosol liquid water content (ALWC) and O₃ exerted secondary influences, and ALWC showed notable potential nonlinear threshold effects under certain humidity conditions. These findings establish fireworks as a crucial amine source and provide insights into modeling secondary aerosol formation under extreme pollution events.