This dissertation presents the application of the molecular LIBS method, a novelapproach of Laser-Induced Breakdown Spectroscopy (LIBS), to optimize the detection of pitting chlorides in concrete structures, which are e.g. contaminated bydeicing salt in winter. Potentiometric titration as the standard method for chloride determination in building material analysis is costly and time-consuming. Ithas the decisive disadvantage that the determination of chloride concentrationis based on the total mass of the concrete and not on the cement content asrequired by the European standard EN 206. The imaging capabilities of LIBS forphase separation of the concrete meet this requirement. LIBS was already usedby BAM in 1998 in building material analysis, but the detection of chlorides withLIBS requires expensive helium purging and spectrometers outside the visiblespectral range to detect emissions of atomic chlorine. The approach of molecular LIBS is to quantify the emission of chloride-containing molecular radicalsformed during the cooling phase of the laser-induced plasma. The advantagescompared to conventional LIBS method are the emission in the visible spectralrange and the applicability without noble gas purging. In this thesis the influenceof the experimental components on the time behaviour of the relevant molecularemission bands is investigated, signal deviations due to plasma fluctuations aresignificantly reduced and for plasma analysis the molecular formation is simulated on atomistic scales and compared with standard methods. In simultaneousmeasurements, atomic and molecular Cl emission are directly compared and the quantification is optimized by data combination. Molecular LIBS will be extended to a quantifying and imaging method that can detect chlorides withoutnoble gas purging.