The billions of tons of mineral dust released into the atmosphere each year provide an important surface for reaction with gas-phase pollutants. These reactions, which are often enhanced in the presence of light, can change both the gas-phase composition of the atmosphere and the composition and properties of the dust itself. Because dust contains titanium-rich grains, studies of dust photochemistry have largely employed commercial titanium dioxide as a proxy for its photochemically active fraction; to date, however, the validity of this model system has not been empirically determined. Here, for the first time, we directly investigate the photochemistry of the complement of natural titanium-containing minerals most relevant to mineral dust, including anatase, rutile, ilmenite, titanite, and several titanium-bearing species. Using ozone as a model gas-phase pollutant, we show that titaniumcontaining minerals other than titanium dioxide can also photocatalyze trace gas uptake, that samples of the same mineral phase can display very different reactivity, and that prediction of dust photoreactivity based on elemental/mineralogical analysis and/or lightabsorbing properties is challenging. Together, these results show that the photochemistry of atmospheric dust is both richer and more complex than previously considered, and imply that a full understanding of the scope and impact of dust-mediated processes will require the community to engage with this complexity via the study of ambient mineral dust samples from diverse source regions.

Mixing ratios of the criteria air contaminant nitrogen dioxide (NO2) are commonly quantified by reduction to nitric oxide (NO) using a photolytic converter followed by NO-O3 chemiluminescence (CL). In this work, the performance of a photolytic NO2 converter prototype originally designed for continuous emission monitoring and emitting light at 395 nm was evaluated. Mixing ratios of NO2 and NOx (= NO + NO2) entering and exiting the converter were monitored by blue diode laser cavity ring-down spectroscopy (CRDS). The NO2 photolysis frequency was determined by measuring the rate of conversion to NO as a function of converter residence time and found to be 4.2 sec -1 . A maximum 96% conversion of NO2 to NO over a large dynamic range was achieved at a residence time of (1.50.3) sec, independent of relative humidity.Interferences from odd nitrogen (NOy) species such as peroxyacyl nitrates (PAN; RC(O)O2NO2), alkyl nitrates (AN; RONO2), nitrous acid (HONO), and nitric acid (HNO3) were evaluated by operating the prototype converter outside its optimum operating range (i.e., at higher pressure and longer residence time) for easier quantification of interferences. Four mechanisms that generate artifacts and interferences were identified: direct photolysis, foremost of HONO at a rate constant of 6% that of NO2, thermal decomposition, primarily of PAN, surface promoted photochemistry, and secondary chemistry in the connecting tubing. These interferences are likely present to a certain degree in all photolytic converters currently in use but are rarely evaluated or reported. Recommendations for improved performance of photolytic converters include operating at lower cell pressure and higher flow rates, thermal management that ideally results in a match of photolysis cell temperature with ambient conditions, and minimization of connection tubing length. When properly implemented, these interferences can be made negligibly small when measuring NO2 in ambient air.

Non-exhaust emissions have surpassed exhaust emissions in their contribution to traffic particulate matter (PM). One major type is brake wear and, although its physical and toxicological properties have been studied,...