Aircraft-based in-situ aerosol mass spectrometry: Chemical characterization and source identification of submicron particulate matter in the free and upper troposphere and lower stratosphere

The composition of the atmosphere is frequently perturbed by the emission of gaseous and particulate matter from natural as well as anthropogenic sources. While the impact of trace gases on the radiative forcing of the climate is relatively well understood the role of aerosol is far more uncertain. Therefore, the study of the vertical distribution of particulate matter in the atmosphere and its chemical composition contribute valuable information to bridge this gap of knowledge. The chemical composition of aerosol reveals information on properties such as radiative behavior and hygroscopicity and therefore cloud condensation or ice nucleus potential. rnThis thesis focuses on aerosol pollution plumes observed in 2008 during the POLARCAT (Polar Study using Aircraft, Remote Sensing, Surface Measurements and Models, of Climate, Chemistry, Aerosols, and Transport) campaign over Greenland in June/July and CONCERT (Contrail and Cirrus Experiment) campaign over Central and Western Europe in October/November. Measurements were performed with an Aerodyne compact time-of-flight aerosol mass spectrometer (AMS) capable of online size-resolved chemical characterization of non-refractory submicron particles. In addition, the origins of pollution plumes were determined by means of modeling tools. The characterized pollution episodes originated from a large variety of sources and were encountered at distinct altitudes. They included pure natural emissions from two volcanic eruptions in 2008. By the time of detection over Western Europe between 10 and 12 km altitude the plume was about 3 months old and composed to 71 % of particulate sulfate and 21 % of carbonaceous compounds. Also, biomass burning (BB) plumes were observed over Greenland between 4 and 7 km altitude (free troposphere) originating from Canada and East Siberia. The long-range transport took roughly one and two weeks, respectively. The aerosol was composed of 78 % organic matter and 22 % particulate sulfate. Some Canadian and all Siberian BB plumes were mixed with anthropogenic emissions from fossil fuel combustion (FF) in North America and East Asia. It was found that the contribution of particulate sulfate increased with growing influences from anthropogenic activity and Asia reaching up to 37 % after more than two weeks of transport time. The most exclusively anthropogenic emission source probed in the upper troposphere was engine exhaust from commercial aircraft liners over Germany. However, in-situ characterization of this aerosol type during aircraft chasing was not possible. All long-range transport aerosol was found to have an O:C ratio close to or greater than 1 implying that low-volatility oxygenated organic aerosol was present in each case despite the variety of origins and the large range in age from 3 to 100 days. This leads to the conclusion that organic particulate matter reaches a final and uniform state of oxygenation after at least 3 days in the free troposphere. rnExcept for aircraft exhaust all emission sources mentioned above are surface-bound and thus rely on different types of vertical transport mechanisms, such as direct high altitude injection in the case of a volcanic eruption, or severe BB, or uplift by convection, to reach higher altitudes where particles can travel long distances before removal mainly caused by cloud scavenging. A lifetime for North American mixed BB and FF aerosol of 7 to 11 days was derived. This in consequence means that emission from surface point sources, e.g. volcanoes, or regions, e.g. East Asia, do not only have a relevant impact on the immediate surroundings but rather on a hemispheric scale including such climate sensitive zones as the tropopause or the Arctic.

Polycyclic aromatic hydrocarbons (PAHs) in river bank soils

Polycyclic aromatic hydrocarbons (PAHs) are ubiquitous and found in the atmosphere, aquatic environment, sediments and soils. For environmental risk assessments and the allocation of the polluter it is important to know the PAH sources. PAH contamination sites are usually the result of anthropogenic processes. Three major sources are known: i) petroleum, including crude oil and its refined products and coals (petrogenic PAHs), ii) burning of organic matter (pyrogenic PAHs) and iii) transformation products of natural organic precursors present in the environment (diagenetic processes). In one case elevated PAH concentrations were found in river bank soils when building a retention area along the Mosel River. The source of the PAHs in this area was unclear and required the investigation of possible sources. To evaluate the PAH distribution along the Mosel River, a section of ~ 160 km along the river and a short section along the Saar River were investigated within this study. Concentrations of the Σ16 EPA PAHs were as high as 81 mg kg-1 dry weight (dw). Additionally, coal particles were identified in some soils, which originated from mining activities in the Saarland region. PAH distribution patterns of the 16 EPA PAHs suggest a mainly pyrogenic origin and in some cases a mixture of pyrogenic and petrogenic origin. For a comprehensive investigation five sampling sites were selected. Two sites were located before the confluence of the Mosel and Saar River, one site at the confluence and two sites after the confluence. The examination included typical forensic methods such as PAH distribution patterns of 45 PAHs (including alkylated PAHs), calculation of PAH ratios, determination of PAH alkyl homologues, n-alkanes, principal component analysis (PCA) and coal petrography. The results revealed a mainly pyrogenic source at sampling sites before the confluence of the two rivers. At and after the confluence, a mixture of pyrogenic and petrogenic inputs were present. With the help of coal petrography, coal derived particles could be identified in these soils. Therefore, coal was suggested to be the petrogenic source. It could be shown that sites with diffuse sources of contaminants, like the bank soils of the Mosel River, are difficult to characterize. As previously mentioned for detailed source identifications, the use of various forensic methods is essential. Determination of PAH alkyl homologue series, biomarkers and isotopes are often recommended. Source identification was evaluated using three different methods (i.e. PAH distribution patterns of an extended PAH spectrum, PAH ratios and analyses of n-alkanes). It was assessed if these methods were sufficient for the initial steps in identifying sources of PAHs in selected samples, and if they could be used for decision-making purposes. Point- and non-point sources were identified by applying the three methods and it could be shown that these relatively simple methods are sufficient in determining the primary source. In a last step of this study two soils (one before the confluence of the Mosel and Saar rivers and one after the confluence), and one sediment of the Mosel River were evaluated by investigating the mutagenic potential of the soils and the sediment with a fluctuation version of the Ames-test. The study showed that coal bearing soils at the Mosel River do not exhibit a greater mutagenic potential than other soils or sediments without coal particles.

Copper and zinc stable isotope ratios as tracers of biogeochemical processes, sources and transport of Cu and Zn in soils

Copper and Zn are essential micronutrients for plants, animals, and humans; however, they may also be pollutants if they occur at high concentrations in soil. Therefore, knowledge of Cu and Zn cycling in soils is required both for guaranteeing proper nutrition and to control possible risks arising from pollution.rnThe overall objective of my study was to test if Cu and Zn stable isotope ratios can be used to investigate into the biogeochemistry, source and transport of these metals in soils. The use of stable isotope ratios might be especially suitable to trace long-term processes occurring during soil genesis and transport of pollutants through the soil. In detail, I aimed to answer the questions, whether (1) Cu stable isotopes are fractionated during complexation with humic acid, (2) 65Cu values can be a tracer for soil genetic processes in redoximorphic soils (3) 65Cu values can help to understand soil genetic processes under oxic weathering conditions, and (4) 65Cu and 66Zn values can act as tracers of sources and transport of Cu and Zn in polluted soils.rnTo answer these questions, I ran adsorption experiments at different pH values in the laboratory and modelled Cu adsorption to humic acid. Furthermore, eight soils were sampled representing different redox and weathering regimes of which two were influenced by stagnic water, two by groundwater, two by oxic weathering (Cambisols), and two by podzolation. In all horizons of these soils, I determined selected basic soil properties, partitioned Cu into seven operationally defined fractions and determined Cu concentrations and Cu isotope ratios (65Cu values). Finally, three additional soils were sampled along a deposition gradient at different distances to a Cu smelter in Slovakia and analyzed together with bedrock and waste material from the smelter for selected basic soil properties, Cu and Zn concentrations and 65Cu and 66Zn values.rnMy results demonstrated that (1) Copper was fractionated during adsorption on humic acid resulting in an isotope fractionation between the immobilized humic acid and the solution (65CuIHA-solution) of 0.26 ± 0.11‰ (2SD) and that the extent of fractionation was independent of pH and involved functional groups of the humic acid. (2) Soil genesis and plant cycling causes measurable Cu isotope fractionation in hydromorphic soils. The results suggested that an increasing number of redox cycles depleted 63Cu with increasing depth resulting in heavier 65Cu values. (3) Organic horizons usually had isotopically lighter Cu than mineral soils presumably because of the preferred uptake and recycling of 63Cu by plants. (4) In a strongly developed Podzol, eluviation zones had lighter and illuviation zones heavier 65Cu values because of the higher stability of organo-65Cu complexes compared to organo-63Cu complexes. In the Cambisols and a little developed Podzol, oxic weathering caused increasingly lighter 65Cu values with increasing depth, resulting in the opposite depth trend as in redoximorphic soils, because of the preferential vertical transport of 63Cu. (5) The 66Zn values were fractionated during the smelting process and isotopically light Zn was emitted allowing source identification of Zn pollution while 65Cu values were unaffected by the smelting and Cu emissions isotopically indistinguishable from soil. The 65Cu values in polluted soils became lighter down to a depth of 0.4 m indicating isotope fractionation during transport and a transport depth of 0.4 m in 60 years. 66Zn values had an opposite depth trend becoming heavier with depth because of fractionation by plant cycling, speciation changes, and mixing of native and smelter-derived Zn. rnCopper showed measurable isotope fractionation of approximately 1‰ in unpolluted soils, allowing to draw conclusions on plant cycling, transport, and redox processes occurring during soil genesis and 65Cu and 66Zn values in contaminated soils allow for conclusions on sources (in my study only possible for Zn), biogeochemical behavior, and depth of dislocation of Cu and Zn pollution in soil. I conclude that stable Cu and Zn isotope ratios are a suitable novel tool to trace long-term processes in soils which are difficult to assess otherwise.rn

Identification and dissolution behavior of the secondary uranium minerals in the corrosion products of Depleted Uranium (DU) ammunition formed in soils

In der vorliegenden Studie wurden verschiedene Techniken eingesetzt um drei Proben (4, 7, and 8) die aus denrnKorrosionsprodukten von aus dem Kosovo Krieg stammenden Munitionskugeln, bestehend aus abgereichertem Uranrn(Depleted Uranium - DU), zu untersuchen. Als erstes Verfahren wurde die Raman-Spektroskopie eingesetzt. Hierbeirnzeigte sichin den Proben, charakterisiert durch einen Doppelpeak, die Anwesenheit von Schoepitrn(UO2)8O2(OH)12(H2O)12. Der erste und zweite Peakzeigte sich im Spektralbereich von 840,3-842,5 cm-1rnbeziehungsweise 853,6-855,8 cm-1. Diese Werte stimmen mit den Literaturwerten für Raman-Peaks für Schoepitrnüberein. Des Weiteren wurde bei dieser Untersuchungsmethode Becquerelite Ca(UO2)6O4(OH)6(H2O)8 mit einemrnPeak im Bereich zwischen 829 to 836 cm-1 gefunden. Aufgrund des Fehlens des Becquerelitespektrums in derrnSpektralbibliothek wurde eine in der Natur vorkommende Variante analysiert und deren Peak bei 829 cm-1 bestimmt,rnwas mit den Ergebnissen in den Proben korrespondiert. Mittels Röntgenbeugung (X-Ray Diffraction, XRD) zeigtenrnsich in allen Proben ähnliche Spektren. Das lässt darauf schließen, dass das pulverisierte Material in allen Probenrndas gleiche ist. Hierbei zeigte sich eine sehr gute Übereinstimmung mit Schoepit und/oder meta-rnSchoepit(UO2)8O2(OH)12(H2O)10, sowie Becquerelite. Weiterhin war weder Autunit, Sabugalit noch Uranylphosphatrnanwesend, was die Ergebnisse einer anderen Studie, durchgeführt an denselben Proben, wiederlegt. DiernAnwesenheit von P, C oder Ca im Probenmaterial konnte ausgeschlossen werden. Im Falle von Calciumkann diesrnmit der Anwesenheit von Uran erklärt werden, welches aufgrund seines Atomradius bevorzugt in Becquerelite (1:6)rneingebaut wird. Die beiden Hauptpeaks für Uran lagen im Falle von U 4f 7/2 bei 382.0 eV und im Falle von U 4f 5/2rnbei 392 eV. Diese Werte mit den Literaturwerten für Schoepit und meta-Schoepitüberein. Die Ergebnissernelektronenmikroskopischen Untersuchung zeigen U, O, Ca, Ti als dominante Komponenten in allen Messungen.rnElemente wie Si, Al, Fe, S, Na, und C wurden ebenfalls detektiert; allerdings kann nicht ausgeschlossen werden,rndass diese Elemente aus dem Boden in der unmittelbaren Umgebung der Munitionsgeschosse stammen. Gold wurdernebenfalls gemessen, was aber auf die Goldarmierung in den Probenaufbereitungsbehältern zurückgeführt werdenrnkann. Die Elektronenmikroskopie zeigte außerdem einige Stellen in denen elementares Uran und Bodenmineralernsowie sekundäre Uranminerale auftraten. Die Elementübersicht zeigt einen direkten Zusammenhang zwischen U andrnCa und gleichzeitig keine Korrelation zwischen U und Si, oder Mg. Auf der anderen Seite zeigte sich aber einrnZusammenhang zwischen Si und Al da beide Konstituenten von Bodenmineralen darstellen. Eine mit Hilfe derrnElektronenstrahlmikroanalyse durchgeführte quantitative Analyse zeigte den Massenanteil von Uran bei ca. 78 - 80%,rnwas mit den 78,2% and 79,47% für Becquerelite beziehungsweise Schoepit aufgrund ihrer Summenformelrnkorrespondiert. Zusätzlich zeigt sich für Calcium ein Massenanteil von 2% was mit dem Wert in Becquerelite (2.19%)rnrecht gut übereinstimmt. Der Massenanteil von Ti lag in einigen Fällen bei 0,77%, was auf eine noch nicht korrodierternDU-Legierung zurückzuführen ist. Ein Lösungsexperiment wurde weiterhin durchgeführt, wobei eine 0,01 M NaClO4-rnLösung zum Einsatz kam in der die verbliebene Probensubstanz der Korrosionsprodukte gelöst wurde;rnNatriumperchlorate wurde hierbei genutzt um die Ionenstärke bei 0,01 zu halten. Um Verunreinigungen durchrnatmosphärisches CO2 zu vermeiden wurden die im Versuch für die drei Hauptproben genutzten 15 Probenbehälterrnmit Stickstoffgas gespült. Eine Modelkalkulation für den beschriebenen Versuchsaufbau wurde mit Visual MINTEQrnv.3.0 für die mittels vorgenannten Analysemethoden beschriebenen Mineralphasen im pH-Bereich von 6 – 10 imrnFalle von Becquerelite, und Schoepit berechnet. Die modellierten Lösungskurven wurden unter An- und Abwesenheitrnvon atmosphärischem CO2 kalkuliert. Nach dem Ende des Lösungsexperiments (Dauer ca. 6 Monate) zeigten diernKonzentrationen des gelösten Urans, gemessen mittels ICP-OES, gute Übereinstimmung mit den modelliertenrnSchoepit und Becquerelite Kurven. Auf Grund des ähnlichen Löslichkeitverhaltens war es nicht möglich zwichen denrnbeiden Mineralen zu unterscheiden. Schoepit kontrolliert im sauren Bereich die Löslichkeit des Urans, währendrnbecquerelit im basichen am wenigsten gelöst wird. Des Weiteren bleibt festzuhalten, dass ein Anteil an CO2 in diernverschlossenen Probenbehälter eingedrungen ist, was sich mit der Vorhersage der Modeldaten deckt. Die Löslichkeitrnvon Uran in der Lösung als Funktion des pH-Wertes zeigte die niedrigsten Konzentrationen im Falle einer Zunahmerndes pH-Wertes von 5 auf 7 (ungefähr 5,1 x 10-6 mol/l) und einer Zunahme des pH-Wertes auf 8 (ungefähr 1,5 x 10-6rnmol/l bei). Oberhalb dieses Bereichs resultiert jeder weitere Anstieg des pH-Wertes in einer Zunahme gelösten Uransrnin der Lösung. Der ph-Wert der Lösung wie auch deren pCO2-Wert kontrollieren hier die Menge des gelösten Urans.rnAuf der anderen Seite zeigten im Falle von Becquerelite die Ca-Konzentrationen höhere Werte als erwartet, wobeirnwahrscheinlich auf eine Vermischung der Proben mit Bodensubstanz zurückgeführt werden kann. Abschließendrnwurde, unter Berücksichtigung der oben genannten Ergebnisse, eine Fallstudie aus Basrah (Irak) diskutiert, wo inrnzwei militärischen Konflikten Uranmunition in zwei Regionen unter verschiedenen Umweltbedingungen eingesetztrnwurden.