Flender/Flender Gear Units/Bevel-helical speed reducer B3
rements were then to be performed on flowing aerosols, with relatively large particle diameters in the [MCrIm range and comprising

pure PAHs, including investigation of the influence of the particle diameter. In afurtherstage,an artificial PAH test aerosol eneratorwas to be

used to produce microstruc- tured, monodisperse submicron aerosols wig defined coatings. The purpose of this was to measure the fluorescence

behaviour of various PAHS, individually adsorbed on trace-catching aerosols with article diameters between 8 nm and 1 nm, and to

determine the detection efficiency of tR sensor: Yet another task consisted in the investigation of the influence of various particle materials on the fluorescence process. Further experiments were intended to show whether individual components could be deter- mined in PAH mixtures adsorbed on particles. In the final stage, the aim was to analyse various real aerosol systems and thus to investigate the fluorescence on real matrices. 3. Results sensor for continuous recording of PAHs was designed and tested during this project. The various PAHs can be identified and quantified by varying the excitation wavelength and eval- uating the fluorescence spectra. Further differentiation is permitted by time-resolved fluor- escence signals. An initial overview of the fluorescence pro erties of PAHs was obtained by measurin 3 crys talline PAH samples. It was found that the guorescence spectra and the decay times %termin: ed should be sufficient for differentiation. Fluorescence maxima between 4 nm (.. phenanthrene) and 6 nm (.. rubrene) were measured for the various PAH crystallites. The decay times were between 1 ns (acenapthylene; excitation wavelength: 2 nm) and 1 ns (rubrene: excitation wavelen th 3 nm). Some crystallites, such as coronene and ovalene, displayed dependence of # fluorescence decay time on the emission wavelength. This implied the existence of various excited energy levels. lrlitial aerosol fluo