How Cold Is Cold Enough? Refrigeration of the Next-Generation Impactor to Prevent Aerosol Undersizing

Heat transfer from impactor to aqueous aerosols causes underestimation of droplet size due to evaporation. Hence, pharmacopeia suggests cooling the impactor to 5°C, which is well below aerosol temperature. In this study, we assessed the droplet size at four different impactor temperatures under controlled ambient conditions to compare the compendial 5°C method with our in-house method, where the impactor is cooled to aerosol temperature. A single nebulizer/compressor unit was used throughout. It produced an aerosol at 17°C when operated at 50% RH and 23°C RT ambient conditions. Thirty-six experiments were conducted with saline, 9 each at impactor temperatures of 5°C, 10°C, 17°C, and 23°C. NaCl stage deposition was determined by conductometry, mass on stages by weighing. Moreover, a simulation was carried out to track aerosol temperature when entering the impactor. Measuring at 23°C yields a significantly smaller mass median aerodynamic diameter (MMAD) than at 5°C-17°C. Despite elevated water condensation in the impactor at 5°C and 10°C, there was no increase in MMAD compared with 17°C. Instead, droplet size determination at 5°C led to significantly smaller values than at 17°C, probably due to distorted volumetric impactor flow rates at different impactor temperatures. Reevaluation of data with flow rates adjusted for impactor temperature (14.1 L/min at 5°C vs. 15.0 L/min at 23°C) led to indistinguishable results at 5°C-17°C. A computational fluid dynamics (CFD) simulation confirmed rapid cooling of the incoming air within the inlet and stage 1 and, with it, the systematic droplet undersizing due to reduced volumetric airflow using a cooled impactor. As long as the impactor temperature is at or below aerosol temperature, no effects on droplet size can be observed. Measuring at aerosol temperature yields the same results as at 5°C, but prevents condensation. However, cooling the impactor well below ambient temperature can cause a systematic error in the volumetric flow rate through the impactor if not corrected accordingly.