A rapid and multi-element method for the analysis of major nutrients in grass (Lolium perenne) using energy-dispersive X-ray fluorescence spectroscopy

Abstract

AbstractElemental analysis of grass (Lolium perenne) is essential in agriculture to ensure grass quality and animal health. Energy-dispersive X-ray fluorescence (EDXRF) spectroscopy is a rapid, multi-element alternative to current methods using acid digestion and inductively coupled plasma optical emission spectrometry (ICP-OES). Percentage phosphorus (P), potassium (K), magnesium (Mg) and calcium (Ca), determined from grass samples using EDXRF, were within 0.035, 0.319, 0.025 and 0.061, respectively, of ICP-OES values. Concordance correlation coefficients computed using agreement statistics ranged from 0.4379 to 0.9669 (values close to one indicate excellent agreement); however, the level of agreement for each element depended on the calibrations used in EDXRF. Empirical calibrations gave excellent agreement for percentage P, K and Ca, but moderate agreement for percentage Mg due to a weaker correlation between standards and intensities. Standardless calibration using the fundamental parameters (FP) approach exhibited bias, with consistently lower values reported for percentage P and Mg, when compared with ICP-OES methods. The relationship between the methods was plotted as scatter plots with the line of equality included, and although correlation coefficients indicated strong relationships, these statistics masked the effects of consistent bias in the data for percentage P and Mg. These results highlight the importance of distinguishing agreement from correlation when using statistical methods to compare methods of analysis. Agreement estimates improved when a matching library of grass samples was added to the FP method. EDXRF is a comparable alternative to conventional methods for grass analysis when samples of similar matrix type are used as empirical standards or as a matching library.