Measuring molecular singlet oxygen ( ¹ O ₂ *) from atmospheric photosensitizers: Intercomparison of techniques, irradiation setups, data analysis and protocol recommendations

Abstract. Molecular singlet oxygen (1O2*) is the first excited state of molecular oxygen O2 and can be formed through indirect photochemistry during irradiation of chromophoric organic matter. Once formed in the particle and droplet phases in the atmosphere, 1O2* can be a competitive oxidant in the photochemical processing of organic matter. Yet, as more researchers study the atmospheric photochemistry of 1O2*, it is useful to establish protocols by evaluating and comparing experimental setups across laboratories. Here, we present 1O2* measurements from four photosensitizing molecules in four photoreactor setups at three research institutions, including two xenon lamps of different strengths and two multi-bulb UVA + UVB broadband systems. The production of 1O2* was investigated from perinaphthenone, lignin, and juglone, which are photosensitizers with atmospherically relevant light absorbing moieties, as well as from Rose Bengal, a standard photosensitizer. Two chemical actinometers, 2-nitrobenzaldehyde and p-nitroanisole/pyridine, were used to quantify photon fluxes and calculate rates of light absorbance for photosensitizers for each photoreactor. We compared two commonly used 1O2* quantification methods, chemical probe method using furfuryl alcohol, as well as direct 1O2* phosphorescence detection at 1270 nm. Rates of light absorbance across experimental setups for each photosensitizer ranged between 0.2 and 62 × 10-5 molphotons L-1 s-1, while 1O2* steady-state concentrations ranged between 0.01 and 129 × 10-11 M. Despite order of magnitude differences in rate of light absorbance and 1O2* steady state concentrations, normalizing to 1O2* quantum yields showed good inter-laboratory agreement but only for perinaphthenone (94 ± 9 % – 112 ± 17 %) and for Rose Bengal (67 ± 15 % – 87 % ± 5 %). 1O2* quantum yields for lignin and juglone increased with decreasing irradiation wavelength, consistent with a wavelength-dependence. Finally, we make five recommendations to improve the accuracy and reproducibility of 1O2* measurements. These include considering wavelength-dependent quantum yields, avoiding suppression of 1O2*, controlling and reporting photoreactor temperature, considering light scattering from nanoparticles, and conducting control experiments. These recommendations will help standardize 1O2* measurements in studying photochemical processing of atmospheric aerosols and droplets.