Chemical Composition and Mixing State of Fine Particles During Three Dust Episodes in Hohhot, China: Implications for Sustainable Urban Air Quality Monitoring

Dust events degrade urban air quality, but fine-particle composition accompanying these events can vary substantially. We combined single-particle aerosol mass spectrometry (SPAMS) with adaptive resonance theory neural-network (ART-2a) clustering, hourly air-quality and meteorological observations, and 72 h HYSPLIT backward trajectories to analyze 21,311 single-particle mass spectra from 0.2 to 2.0 μm particles during three episodes in Hohhot, China, in April 2024. The episodes were associated with northwestern, predominantly northern, and westerly pathways, respectively. Elemental carbon (EC) represented 43.7%, 25.7%, and 23.8% of particles assigned to the eight major classes by number during Dust I, Dust II, and Dust III, respectively, whereas Dust-class fractions were 23.4%, 19.2%, and 17.7%. Dust I showed the highest EC fraction and stronger Na+ and EC-related signals but weaker nitrate association. Dust II had the highest levoglucosan-containing (LEV) fraction (22.5%), the largest OC-LEV and EC-Sec contributions within their respective parent classes, and the strongest nitrate signals, indicating greater biomass-burning influence and association with secondary inorganic species. Dust III showed the strongest sulfate signal and the highest OC-Sec, indicating enhanced internal mixing of organic carbon with secondary inorganic species. These case-specific contrasts reveal distinct mixtures of mineral, combustion-derived, biomass-burning-related, and secondary components. Integrating single-particle composition and mixing-state information with conventional monitoring can support more differentiated dust-event assessment, public-health risk communication, and sustainable urban air quality management. These contrasts should not be interpreted as invariant characteristics of the corresponding transport pathways.