Untersuchungen zur Ozonolyse einfacher Alkene in der Atmosphären-Simulationskammer SAPHIR

The ozonolysis, that is, the oxidation with O$_{3}$, is one of the most important removal pathways for alkenes in the troposphere. The ozonolysis competes with the removal of alkenes through the reaction with OH-radicals during the day and through the reaction with NO$_{3}$ during the night. The mechanism of the gas-phase ozonolysis is as yet not fully understood. It is suspected that during ozonolysis beside different stable products OH and HO$_{2}$ radicals are also formed, which play an important role in the tropospheric chemistry. The scope of this work was to find an answer to this and other questions related to the ozonolysis and to discuss their relevance within the tropospheric chemistry. Short chained alkenes (C$_{n}$ with n < 5) are the main components of the total alkenes in urban regions. Past investigations were carried out with reactant concentrations within the ppmv range. These values are clearly much higher than the concentrations found in the troposphere. The question arises wether the observations at high concentrations would apply to the lower tropospheric ppbv range. Within the scope of this work the ozonolysis of C$_{2}$ to C$_{4}$ alkenes with reactant concentrations between 20 ppbv and 200 ppbv were investigated. The atmospheric simulation chamber SAPHIR in the Research Center in Jülich, Germany, was used for this purpose. Experiments with CO as radical scavenger were carried out to determine the rate constants of the ozonolysis of the investigated alkenes. The results fell within the range presented in the literatur. The time profile of the ozonolysis reactions were in excellent agreement with the expected kinetics, which cancels the likelihood of any interfering chemical processes. The yields for OH and HO$_{2}$ radicals of the investigated alkenes was determined. The yield of OH obtained illustrated a water dependence not previously mentioned in literatur. This work produced also the first determinations for the HO$_{2}$ yields, which exceeded a value of 1.0 for the most alkenes, contradicting an assumed value of 0.2 used in models. In addition the yields of stable products such as HCHO, CH$_{3}$CHO and CO were also determined. With Z-2-Buten a water dependence of the CH$_{3}$CHO yield was also found. For some alkenes the obtained yields of HCHO and CH$_{3}$CHO were greater than found in previous experimental determinations. This clearly points to additional unidentified reactions pathways leading to HCHO and CH$_{3}$CHO. What was also proven was a CO dependence of the yields of HO$_{2}$ and CH$_{3}$CHO. In order to explain the former observations the standard mechanism of the ozonolysis was modified. This modification takes into consideration the competition between the unimolecular decomposition of two different Criegee Intermediates and their bimolecular reaction with either water or CO. The water dependence of the OH yield would lead to a lower OH concentration at night in dry regions as compared to humid regions and consequently to a lower overall removal of hydrocarbons. The importance of the higher yields for HO$_{2}$ radicals was investigated using a box model. The simulation showed for the first time that higher HO$_{2}$ yields lead to higher HO$_{2}$ concentrations at night. The recombination of the HO$_{2}$ radicals and/or reactions with other peroxiradicals thus result in higher H$_{2}$O$_{2}$ concentrations as well as higher concentrations of organic peroxides than expected