Atmospheric aerosols play a key role in the earth climate and biogeochemical cycles. Moreover, the global deaths caused by exposure to ambient particulate matter with a diameter less than 2.5 µm (PM2.5) reached 4.2 million in 2015 and is further increasing with the rapid development of mobile industrial societies. However, the observation of toxic trace compounds, their distribution, transport pathways, and degradation are so-far limited by technological challenges. Single-particle mass spectrometry (SPMS) can tackle many of these issues by obtaining the size and a chemical profile from individual particles in real time. However, the conventional laser ionization method in SPMS leads to strong fragmentation of important species, impeding their identification and provides only limited data on the particle composition.

Our research focuses on new technologies that acquire chemical information from individual, microscopic particles. Key is the understanding and optimization of complex laser-matter interactions for efficient and non-destructive ablation, ionization and detection of health-relevant trace components from particles. We exploit resonances between laser radiation and molecules or atoms to detect such species with masses in the order of 10^-16–10^-12 g per particle.

In single-particle mass spectrometry (SPMS) [6,7], airborne particles are accelerated in a gas expansion into vacuum. Their size is determined by their final velocity, measured by Mie scattering in a pair of laser beams (laser velocimetry). An electronic circuit calculates the arrival time of each individual particle in the ion source of the mass spectrometer and triggers the desorption- and ionization laser pulses that hit the particles ‘on-the-fly’ for the desired excitation and ionization scheme. The resulting ions are analyzed in a bipolar mass spectrometer, detecting positive and negative ions in opposite time-of-flight mass analyzers. Fast digitizes record the mass spectra of every individual particle and pattern recognition algorithms based on artificial neural networks classify the particles for source apportionment.