Laser-induced breakdown spectroscopy (LIBS) and wavelength dispersive X-ray fluorescence (WDXRF) data fusion to predict the concentration of K, Mg and P in bean seed samples

[Display omitted] •Direct K, Mg and P determination in solid beans.•LIBS and WDXRF data fusion for calibration models proposition.•Multiple linear regression presented low error when compared with univariate calibration.•Trueness values around 80 and 120% for analytes determined. The present study aims to develop a fast and simple method for the determination of potassium (K), magnesium (Mg) and phosphor (P) in bean seed samples employing a data fusion strategy in the low-level with laser-induced breakdown spectroscopy (LIBS) and wavelength dispersive X-ray fluorescence (WDXRF) techniques combined with direct solid sample analysis. Univariate and multivariate (multiple linear regression, MLR) calibration and leave-one-out cross validation strategies were evaluated to build the calibration models correlated with reference values obtained by inductively coupled plasma optical emission spectrometry (ICP OES) after microwave-assisted acid digestion. The proposed calibration models for WDXRF and LIBS were tested using 14 samples, where the best results were obtained using the data fusion of both techniques. The standard error of cross validation (SECV) obtained were: 0.12% for K, 0.019% for Mg and 0.10% for P, and the trueness ranged between 89 and 124% for K, 82 to 125% for Mg and 73 to 128% for P. These values showed a good accuracy, precise and robustness of the method and a greater reliability of the results. In addition, the predicted concentrations ranged from 0.97 to 1.55% for K, 0.14 to 0.28% for Mg, and 0.27 to 0.82% for P.