Time-Domain NMR Analysis of Grass and Hay Composition

Abstract

Time-Domain Nuclear Magnetic Resonance (TD-NMR) spectroscopy has proven to be a powerful, fast, robust, yet inexpensive and user-friendly tool for analyzing the relaxation behavior of bulk samples. This study explores the capabilities of TD-NMR and statistical software for the characterization and component analysis of grass and hay samples, with a focus on the identification of carbohydrates and crude protein content. These components are of primary interest to the equine and production animal nutrition industries. This pioneering project aims to assess the possibilities and limits for reliable and reproducible component analysis by TD-NMR in comparison to conventional wet chemical analyses. A protocol optimizing sample temperature, number of scans, and a pre-measurement equilibration time was developed for these samples. Three sample sets were analyzed multiple times using TD-NMR and used to create a 106-sample calibrated standard set. Compositional prediction tests were conducted using Bruker’s integrated OPUS software and SIMCA multi- and megavariate data analysis software, with data from conventional wet chemical analyses of each sample serving as the basis of comparison.1 A high degree of precision was achieved, and less than 10% variation in component predictions was observed across consecutive measurements of the same sample. For components such as Dry Matter (DM) and Acid Detergent Fiber (ADF), high accuracy was also achieved, and predicted values fell within 5% of wet chemistry reference values for the majority of unknowns. For other components such as Moisture, Crude Protein (CP), Water-Soluble Carbohydrates (WSC), and Ethanol-Soluble Carbohydrates (ESC), most predicted values were within 5% of reference values while others exceeded 10% relative error. Experimental results using different component permutations, with up to 6 components predicted simultaneously, suggest that TD-NMR could see future applications in the animal nutrition industry.