Captive Aerosol Growth and Evolution (CAGE) chamber system to investigate particle growth due to secondary aerosol formation

Abstract

Environmental chambers are a commonly used tool for studying the production and processing of aerosols in the atmosphere. Most are located indoors and most are filled with air having prescribed concentrations of a small number of reactive gas species. Here we describe portable chambers that are used outdoors and filled with mostly ambient air. Each all-Teflon® 1 m3 Captive Aerosol Growth and Evolution (CAGE) chamber has a cylindrical shape that rotates along its horizontal axis. A gas-permeable membrane allows exchange of gas-phase species between the chamber and surrounding ambient air with an exchange time constant of approximately 0.5 h. The membrane is non-permeable to particles, and those that are injected into or nucleate in the chamber are exposed to the ambient-mirroring environment until being sampled or lost to the walls. The chamber and surrounding enclosure are made of materials that are highly transmitting across the solar ultraviolet and visible wavelength spectrum. Steps taken in the design and operation of the chambers to maximize particle lifetime resulted in averages of 6.0, 8.2, and 3.9 h for ∼ 0.06, ∼ 0.3, and ∼ 2.5 µm diameter particles, respectively. Two of the newly developed CAGE chamber systems were characterized using data acquired during a 2-month field study in 2016 in a forested area north of Houston, TX, USA. Estimations of measured and unmeasured gas-phase species and of secondary aerosol production in the chambers were made using a zero-dimensional model that treats chemical reactions in the chamber and the continuous exchange of gases with the surrounding air. Concentrations of NO, NO2, NOy, O3, and several organic compounds measured in the chamber were found to be in close agreement with those calculated from the model, with all having near 1.0 best fit slopes and high r2 values. The growth rates of particles in the chambers were quantified by tracking the narrow modes that resulted from injection of monodisperse particles and from occasional new particle formation bursts. Size distributions in the two chambers were measured intermittently 24 h d−1. A bimodal diel particle growth rate pattern was observed, with maxima of about 6 nm h−1 in the late morning and early evening and minima of less than 1 nm h−1 shortly before sunrise and sunset. A pattern change was observed for hourly averaged growth rates between late summer and early fall.