Our results reveal that iron carboxylate photochemistry induces aging by fragmentation, functionalization and oligomerization producing a wide spectrum of compounds. This chemical evolution of aerosol particles containing these compounds can potentially induce changes in the physical properties of the particles such as viscosity or absorption. We observed products of the photolysis of Fe(III) citrate but also products coming from radical chemistry and/or direct photolysis of the compounds produced. This photochemistry needs to be investigated in more detailed to better understand how it influences the physical properties of aerosol particles. We suggested that that the system can run into anoxic conditions in the bulk as a result of oxygen diffusion limitations changing the chemistry and the resulting compounds. When this applies to atmospheric aerosol particles, the chemistry driven in the bulk will extremely depend on the diffusion of oxygen within particles. The anoxic conditions lead to different chemical evolution of the particles such as the production of unsaturated compounds, as our HPLC and STXM results show, where HOx chemistry plays a minor role.
The release of OVOCs appears to be greater under humid conditions than under dry conditions due to the limitations of oxygen to get into the bulk and due to the changes in the diffusivity of the OVOCs themselves.
7 Conclusions and outlook
7.1 Conclusions
Brown carbon and iron carboxylate complexes are chromophores present in atmospheric aerosol parti-cles that photochemically trigger radical reactions, which induce the production of HOx radicals and the subsequent oxidation and secondary radical and non-radical processes, which contribute to aerosol aging.
The products resulting from these processes include oxygenated volatile organic compounds (OVOCs), func-tionalization and oligomerization products. The overarching goal of this thesis has been to understand the feedbacks between indirect photochemical radical production and aging triggered by chromophores men-tioned above within atmospheric aerosol particles and the physical properties of those particles such as diffusion or viscosity.
The HO2 radical production and release photochemically induced by BrC have been assessed by HO2
release measurements by using IC, BPh and BBA as proxies of photosensitizers forming excited triplet states, which act as oxidants towards organic aerosol components. Comparable HO2release was also observed from iron citrate initiated photochemistry. It has been demonstrated that the contribution of this photochemistry to the sources of HO2in the gas phase of the atmosphere is less relevant for polluted regions but it may be more important in unpolluted regions under biogenic influences, in twilight or at high latitudes when less UV for traditional radical production or direct photolysis is available and gas-phase radical production rates are lower.
It was demonstrated that the presence of triplet-scavenging, partially oxygenated organic H atom donors in the condensed phase can substantially contribute to the organic and HO2radical production, by scavenging triplet states of BrC faster than more oxidized organic compounds, such as citric acid. In a similar way, the excited triplet states act as a sink to drive reactive uptake of H atom donors, such as limonene, from the gas phase, and thereby promote the aerosol growth. The predicted OH production in the condensed phase by other, traditional, sources is around 15 M day−1 22. Uptake of HOx radicals from the gas phase amounts to about 1 M day−1. The upper limit of indirect photochemical HO2 production is around 20 M day−1. Therefore, the triplet induced radical production clearly outcompetes the traditional condensed phase radical sources and also radical uptake from the gas phase. Simultaneously, the HO2 produced from triplet scavenging is in turn scavenged by the same organic compounds, hence preventing the release of HO2
into the gas phase (as mentioned above), but keeping the oxidative capacity within aerosol particles.
In this thesis, also the the influence of halides on the photosensitized chemistry of BrC was studied. It was established previously that halides act as electron donor towards triplet states195. Therefore, halides can promote the organic and HO2 radical production in a similar way as organic donors. Additionally, halogen radicals and molecular halogens are produced, which can be recycled into halides by HO2 with an efficiency up to around 50 % therefore reducing the oxidative capacity in the aerosol phase. In turn, the other 50 % are released to the gas phase in the form of molecular halogen compounds. Using the example of iodine, it has been demonstrated that this photosensitized halogen activation is comparable to or larger than halogen activation by oxidation of iodide by ozone in aerosol particles, when the reactive uptake is low
due to diffusion limitations, as is the case in sea-spray derived organic-rich aerosol particles transported to the free troposphere.
A complex dependence of the photosensitized chemistry and radical production and release withRHwas observed. This complexity comes from the interplay between the changes in diffusion, intersystem crossing, radical reaction rates and a possible photosensitized surface reaction upon water content changes. In all systems investigated, radical production increased from dry to mediumRHconditions, reaching a maximum in the range 50−60 % RH, which was attributed to increasing diffusivity of all reactants and products.
Also the release of OVOC products from iron citrate induced citric acid degradation and from iron doped limonene SOA increased similarly. The decrease of photochemical turnover at higherRHwas generally larger than expected from the effect of dilution of reactions with increasing water activity and may be related to complex water activity-dependent reaction mechanisms not considered so far.
We observed the carbon chemical evolution and iron speciation under UV irradiation of samples containing Fe(III) citrate and citric acid including the degradation of carboxylate groups by direct photolysis of Fe(III) carboxylate complexes and the HOx radical chemistry triggered by the photolysis of Fe(III) carboxylate complexes. We also observed gradients in the iron oxidation state in single particles after irradiation, which indicates the existence of oxygen diffusion limitations. We conclude that Fe(III) carboxylate photochemistry can contribute to aerosol aging and potentially to changes in the physical properties of aerosol particles such as diffusion, viscosity or absorption. The products detected together with the iron speciation maps after irradiation point out to oxygen diffusion limitations within atmospheric particles aerosol which would induce an anoxic region in the bulk beyond the reacto-diffusive length of oxygen. Generalizing from the special case of the study cases with either high photosensitizer or high iron citrate concentrations, we could use the assessment of OH production for well-established OH sources for aqueous tropospheric aerosol particles, 15 M/day from Tilgner et al. (2013)22. We could assume that each organic radical resulting from reaction with OH is able to scavenge oxygen. An upper limit concentration of oxygen in atmospheric aerosol particles is around 10−3M. Then, the first order decay rate of oxygen would be 5.9 s−1. Assuming a diffusion coefficient of about 10−12 cm2 s−1 for oxygen11, the reacto-diffusive length (l = p
D/k216) would be about 40 nm.
By the support of our results, we claim the existence of anoxic region in the bulk of atmospheric aerosol particles in full sun light.
In order to address the impact of iron complex chemistry in a more realistic organic aerosol matrix, the OVOC release was measured, produced by photolysis and the subsequent HOx chemistry, while irradiating SOA produced by the oxidation of limonene by ozone as a function of iron content and RH. We noted that the presence of iron in low concentrations (∼ 4 %) can promote the OVOC production and release most likely due to the production of carboxylate complexes. On the other hand, high concentrations of iron produce a drop in the OVOC release. At high iron concentration, a larger fraction of iron may form less stable complexes in the SOA matrix. Therefore HOx radical chemistry can be shut down by cycling between Fe(II) and Fe(III). The OVOC release increases from dry conditions to 55 %RHby a factor of 4 most likely due to the faster diffusion of oxygen inside the samples and the subsequent increase of the HOx production due to photolysis. A diffusion model built by fitting the output of the model to the data demonstrates an increase of the diffusion coefficient of organic molecules along with the increase in the RH. We conclude
that the potential mass loss induced by photochemistry in SOA is around 4 % per day, which is a factor of 4 larger than the mass loss calculated by previous work at dry conditions299.