Chlorine determination in cement-bound materials with Laser-induced Breakdown Spectroscopy (LIBS) – A review and validation

Highlights: • Overcome misinterpretation of analyte concentrations in porous matrices • Comparison of different methods to determine limit of detection (LOD) and limit of quantification (LOQ) • Validation of chlorine determination with LIBS in cement-bound materials • The error of the LIBS analysis (0.595 ± 0.061 wt%) is comparable to standardised methods (0.40 ± 0.06 wt%). The determination of chloride is still one of the main tasks for the evaluation of reinforced concrete structures. The corrosion of the reinforcement induced by the penetrating chlorides is the dominant damage process affecting the lifetime of concrete structures. In the recent years different research groups demonstrated that LIBS can be a fast and reliable method to quantify chlorine in cement-bound materials. Because chlorine in concrete can only occur as solved ions in the pore solution or bound in salts or hydrated cement phases, the detected emission of chlorine can be correlated with the chloride concentration determined e.g. with potentiometric titration. This work inter alia describes the production of reference samples and possible side effects during the production process. Due to transport processes in the porous matrix of the cement a misinterpretation of the concentrations is possible. It is shown how to overcome these effects and higher precisions of the single measurements can be realised. Using the calibration method, blank sample method and noise method, three different ways of calculating the limit of detection (LOD) and limit of quantification (LOQ) are compared. Due to the preparation of the reference samples a precision of the whole calibration model of s_x0 = 0.023 wt% is determined. The validation of the model is based on different test sets, which are varying in their composition (different Cl-salts, water-to-cement ratios and additives). The determined mean error of the validation is 0.595 ± 0.063 wt%, which is comparable to standardised methods like potentiometric titration, direct potentiometry or photometry (0.40 ± 0.06 wt%) [1].