TY - CONF A1 - Abad Andrade, Carlos Enrique A1 - Florek, S. A1 - Becker-Ross, H. A1 - Okruss, M. A1 - Jakubowski, Norbert A1 - Tatzel, Michael A1 - Vogl, Jochen A1 - Recknagel, Sebastian A1 - Panne, Ulrich T1 - Shedding light on global warming: high-resolution optical spectroscopy for magnesium isotope analysis N2 - Magnesium is a major element in the hydrosphere and biosphere, and it is direct correlated with the carbon cycle. Therefore, the study of its isotopic fractionation in sediments and sea water helps to understand the earth’s climate and global warming. Mg has three stable isotopes, 24Mg, 25Mg and 26Mg, and traditionally isotope amount ratios are measured by mass spectrometric methods. Their drawbacks include the high costs for instruments and their operation, experienced operators and elaborate time-consuming chromatographic sample preparation. Recently, an optical spectrometric method has been proposed as faster and low-cost alternative for the analysis of isotope ratios: high-resolution continuum source graphite furnace molecular absorption spectrometry (HR-CS-GFMAS). For the determination of Mg isotope ratios in selected rock reference materials, the high-resolution molecular absorption spectrum of in-situ generated MgF molecule was studied applying multivariate analysis and the results compared with MC-ICP-MS. Samples were dissolved by acid digestion and Mg isotopes analyzed with and without matrix extraction. The absorption spectra were recorded for MgF for the electronic transition X 2Σ → B 2Σ+. The MgF spectrum is described as the linear combination of their isotopic components or isotopologues: 24MgF, 25MgF, and 26MgF. The isotope analysis was done by deconvolution of the MgF spectrum by partial least square regression (PLS). A PLS model was built and calibrated with enriched isotope spikes and certified reference materials. Spectra data was preprocessed by a derivate of second order and venetian blinds cross-validation was employed for finding the optimum latent variables. Finally, the model was refined by a genetic algorithm which identified the best subset of variables for a precise and accurate regression. Results are compatible with those obtained by MC-ICP-MS with an accuracy of ± 0.3‰ with uncertainties ranging between 0.02 to 0.6‰. This accuracy and precision discriminate the isotope fractionation in geological samples, and it is suitable for earth’s climate studies. T2 - Adlershofer Forschungsforum 2019 CY - Berlin, Germany DA - 11.11.2019 KW - Isotope analysis KW - Geological thermometer KW - Magnesium KW - Optical spectroscopy PY - 2019 AN - OPUS4-49890 LA - eng AD - Bundesanstalt fuer Materialforschung und -pruefung (BAM), Berlin, Germany ER -