Aerosol particles that catalyze ice nucleation alter the optical properties and precipitation cycles of clouds. Although mineral dust aerosol particles containing metal oxides are susceptible to the formation of oxygen vacancies (VO) on their surfaces, the impact of these defects on ice nucleation activity has not been addressed. To investigate the impact of VO sites, we conducted a droplet immersion freezing assay on zinc aluminate (ZnAl2O4) and magnesium aluminate (MgAl2O4) spinels annealed under air, nitrogen, and oxygen atmospheres. We observe that samples annealed under nitrogen promote ice nucleation at warmer temperatures compared to those treated in oxidizing atmospheres, with the effect being most pronounced for ZnAl2O4. To further understand these results, we investigated the immersion freezing of zinc oxide (ZnO) and magnesium oxide (MgO). Here, we observe that ZnO nucleates ice at substantially warmer temperatures than MgO after annealing under nitrogen. We hypothesize that the trends in ice nucleation activity are due to the varying concentrations of VO that form during the annealing process on the oxide surfaces, which tend to be higher in the absence of O2. Density functional theory (DFT) calculations support our hypothesis, indicating that VO is more stable on the surfaces of the Zn-containing oxides. The study suggests that oxygen vacancies, which are common defects on metal oxide surfaces that affect their adsorption and catalytic properties, can influence the efficiency with which mineral dust aerosol particles activate ice formation and affect cloud radiative forcing.
Reactive uptake of methylglyoxal (Mgly) on aerosol particles is an important source of secondary organic aerosol (SOA), yet its significance remains highly uncertain due to the poorly constrained uptake coefficients (γMgly). Here, we quantified γMgly on deliquesced pH-buffered ammonium nitrate (AN) and sulfate/nitrate/ammonium (SNA) aerosols via flow tube experiments by directly measuring SOA formation under variable relative humidity (RH, 75-92%) and pH (3.1-4.4). For AN aerosols, γMgly ranged from 1.92 × 10-4 to 7.29 × 10-4, increasing with enhanced RH due to salting-out effects. Moreover, γMgly decreased by a factor of 2 to 10 as pH rose from 3.15 to 4.4 with NH3 addition, suggesting that acid-catalyzed reactions dominate the Mgly uptake. The pH dependence was captured by a first-order reaction rate constant (kI = 102.44-0.85·pH, R2 = 0.93). This kinetic parameter, together with effective Henry's law constants, can be implemented to update the γMgly parametrization. Addition of sulfate aerosols was found to strongly suppress γMgly, reducing kI to 12-38% of the estimation on AN-alone aerosols at a similar pH. Our findings underscore the critical role of aerosol pH and composition in Mgly uptake and provide kinetic parameters to atmospheric models to improve predictions of Mgly SOA.

