Multi-elemental analysis of fertilizer using laser-induced breakdown spectroscopy coupled with partial least squares regression

Laser-induced breakdown spectroscopy (LIBS) has been applied for the multi-elemental analysis of fertilizer. With a set of 11 fertilizer samples containing different levels of phosphorus and potassium, it was identified that the line intensity of the analyte does not follow a straightforward correlation with the element concentration with the presence of matrix effects. Instead, the line intensity of a given analyte element is not only related to that analyte, but also to other elements present in the samples. Further analysis reveals the correlations among the line intensities of the main components. Based on the correlation analysis, a set of calibration models were generated for phosphorus and potassium with the method of partial least squares (PLS) analysis, which is known as a multivariate calibration method. The prediction accuracy and reproducibility of these PLS models were then validated using independent LIBS measurements. The results show that the predicted concentrations with these models provided by LIBS measurements are in good agreement with the reference concentrations, which confirms that the LIBS technique has good potential for the in situ rapid determination of the main elements present in fertilizer.