Radical chemistry in oxidation flow reactors for atmospheric chemistry research

Environmental chambers have been playing a vital role in atmospheric chemistry research for seven decades. In last decade, oxidation flow reactors (OFR) have emerged as a promising alternative to chambers to study complex multigenerational chemistry. OFR can generate higher-than-ambient concentrations of oxidants via H2O, O-2 and O-3 photolysis by low-pressure-Hg-lamp emissions and reach hours to days of equivalent photochemical aging in just minutes of real time. The use of OFR for volatile-organic-compound (VOC) oxidation and secondary-organic-aerosol formation has grown very rapidly recently. However, the lack of detailed understanding of OFR photochemistry left room for speculation that OFR chemistry may be generally irrelevant to the troposphere, since its initial oxidant generation is similar to stratosphere. Recently, a series of studies have been conducted to address important open questions on OFR chemistry and to guide experimental design and interpretation. In this Review, we present a comprehensive picture connecting the chemistries of hydroxyl (OH) and hydroperoxy radicals, oxidized nitrogen species and organic peroxy radicals (RO2) in OFR. Potential lack of tropospheric relevance associated with these chemistries, as well as the physical conditions resulting in it will also be reviewed. When atmospheric oxidation is dominated by OH, OFR conditions can often be similar to ambient conditions, as OH dominates against undesired non-OH effects. One key reason for tropospherically-irrelevant/undesired VOC fate is that under some conditions, OH is drastically reduced while non-tropospheric/undesired VOC reactants are not. The most frequent problems are running experiments with too high precursor concentrations, too high UV and/or too low humidity. On other hand, another cause of deviation from ambient chemistry in OFR is that some troposphericallyrelevant non-OH chemistry (e.g. VOC photolysis in UVA and UVB) is not sufficiently represented under some conditions. In addition, the fate of RO2 produced from VOC oxidation can be kept relevant to the troposphere. However, in some cases RO2 lifetime can be too short for atmospherically-relevant RO2 chemistry, including its isomerization. OFR applications using only photolysis of injected O-3 to generate OH are less preferable than those using both 185 and 254 nm photons (without O-3 injection) for several reasons. When a relatively low equivalent photochemical age (n similar to 1 d) and high NO are needed, OH and NO generation by organic-nitrite photolysis in the UVA range is preferable. We also discuss how to achieve the atmospheric relevance for different purposes in OFR experimental planning.