Application of Multi-Way Analysis to 2D NMR Data

Abstract

Two- and three-dimensional or even higher-dimensional NMR spectroscopy is changing from specialised techniques to more commonly used ones. As the complexity of the acquired NMR data increases, the task of analysing these data constantly becomes more and more demanding and new methods are required to facilitate the analysis. With one-dimensional NMR data multivariate data analysis has proven to be a strong tool, but how should one analyse higher-dimensional NMR data in order to extract as much relevant information as possible without having to break data down into smaller dimensions and thus lose the inherent structure? A class of multivariate data analytical techniques called multi-way analysis encompass techniques that have been designed to handle and analyse such data structures directly. In this paper, the theory of some of the most commonly used multi-way methods will be described and examples of their application to three-way arrays of NMR data reported in the literature will be given. The focus will be on the generic principles of multi-way analysis using three-way data in order to gently introduce the new concepts.