Online NMR spectroscopy is a promising analysis method for reaction and process monitoring. This method facilitates non-invasive investigations of complex liquid multicomponent mixtures without prior calibration. In the present project, new analysis methods based on Bayesian statistics will be developed that enable the automatic and robust quantitative analysis of large series of NMR data. The asset of the Bayesian method is that it incorporates prior knowledge of the investigated NMR spectra in the analysis. Therefore, also difficult spectra (e.g. low signal-to-noise ratio, overlapping peaks, flowing samples) can be analysed quantitatively with this method where the existing approaches fail. The new analysis method will be optimised for applications in reaction and process monitoring by exploiting prior knowledge of the investigated sample, e.g. the stoichiometry of the occurring reactions or the flow velocity of the sample. Robust quantitative analysis of spectra of complex mixtures acquired with benchtop NMR spectrometers will be facilitated by taking prior knowledge into account that is gained from spectra of these samples acquired with high field NMR spectrometers. The new analysis methods developed in this project will contribute to extend the possible fields of applications of NMR spectroscopy for reaction and process monitoring.
DFG Programme
Research Grants
International Connection
New Zealand
