| Popis: |
Pyrolysis liquids are very complex and heterogeneous in composition. This makes them hard to comprehensively analyze, which is one of the hurdles that could hinder further advances in science and technology toward their valorization. Recently, renewed interest grew for quantitative recording of two-dimensional 1H-13C heteronuclear single-quantum correlation (HSQC) nuclear magnetic resonance (NMR). This makes 1H-13C HSQC NMR a valuable tool to fingerprint and quantitatively assess these complex liquids. However, data analysis of complex 1H-13C HSQC spectra lacks behind on these recent experimental developments. That is, 1H-13C HSQC spectra are often manually and ad hoc analyzed. This work, therefore, seeks to automate data analysis from 1H-13C HSQC spectra. We explored the use of image processing tools and identified their much underestimated potential. Indeed, many of the existing tools (often built-in software) were found to be applicable for noise detection/removal, generation/comparison of regions of interest, etc. Moreover, pseudo-Voigt peaks were fitted to the 1H-13C HSQC spectra, with an average R2 of 0.94. These fitted spectral peaks allowed for the generation of a peak list, as an input for multivariate analysis. This allowed for pinpointing differences in the chemical composition of the samples. Overall, a new echelon for easy analysis of 1H-13C HSQC spectra has been explored and demonstrated. |
