Speedy Component Resolution Using Spatially Encoded Diffusion NMR Data

• Published in: Magnetic resonance in chemistry Vol. 63; no. 1; pp. 49 - 61
• Main Authors: Lorandel, Benjamin, Rocha, Hugo, Cazimajou, Oksana, Mishra, Rituraj, Bernard, Aurélie, Bowyer, Paul, Nilsson, Mathias, Dumez, Jean‐Nicolas
• Format: Journal Article
• Language: English
• Published: England Wiley Subscription Services, Inc 01.01.2025; Wiley; John Wiley and Sons Inc
• Subjects: Algorithms; Analytical chemistry; Chemical Sciences; Coding; Data analysis; Data processing; Diffusion coefficient; diffusion NMR; Diffusion rate; DOSY; Mixtures; Multivariate analysis; NMR spectroscopy; OUTSCORE; SCORE; Spatial data; Spectra; Spectrum analysis; ultrafast NMR; ultrafast NMR; DOSY; diffusion NMR; OUTSCORE; SCORE; multivariate analysis; Diffusion NMR
• ISSN: 0749-1581; 1097-458X; 1097-458X
• DOI: 10.1002/mrc.5488

ABSTRACT Diffusion‐ordered NMR spectroscopy (DOSY) is a powerful tool to analyse mixtures. Spatially encoded (SPEN) DOSY enables recording a full DOSY dataset in just one scan by performing spatial parallelisation of the gradient dimension. The simplest and most widely used approach to processing DOSY data is to fit each peak in the spectrum with a single or multiple exponential decay. However, when there is peak overlap, and/or when the diffusion decays of the contributing components are too similar, this method has limitations. Multivariate analysis of DOSY data, which is an attractive alternative, consists of decomposing the experimental data, into compound‐specific diffusion decays and 1D NMR spectra. Multivariate analysis has been very successfully used for conventional DOSY data, but its use for SPEN DOSY data has only recently been reported. Here, we present a comparison, for SPEN DOSY data, of two widely used algorithms, SCORE and OUTSCORE, that aim at unmixing the spectra of overlapped species through a least square fit or a cross‐talk minimisation, respectively. Data processing was performed with the General NMR Analysis Toolbox (GNAT), with custom‐written code elements that now expands the capabilities, and makes it possible to import and process SPEN DOSY data. This comparison is demonstrated on three different two‐component mixtures, each with different characteristics in terms of signal overlap, diffusion coefficient similarity, and component concentration. The SCORE and OUTSCORE algorithms are used to analyse SPEN DOSY data, within the GNAT toolbox.