Quantification of calcium in infant formula using laser-induced breakdown spectroscopy (LIBS), Fourier transform mid-infrared (FT-IR) and Raman spectroscopy combined with chemometrics including data fusion.

Laser-induced breakdown spectroscopy (LIBS), Fourier transform mid-infrared (FT-IR) and Raman spectroscopy combined with chemometrics were investigated to quantify calcium (Ca) content in infant formula powder (INF). INF samples (n = 51) with calcium content levels (ca. 6.5-30 mg Ca/100 kJ) were prepared in accordance with the guidelines of Commission Directive 2006/125/EC. Atomic absorption spectroscopy (AAS) was used as the reference method for Ca content determination. To predict Ca content in INF samples, partial least squares regression (PLSR) models that developed based on LIBS, Raman and FT-IR spectral data, respectively. The model developed using LIBS data achieved the best performance for the quantification of Ca content in INF (R2 (cross-validation (CV))-0.99, RMSECV-0.29 mg/g; R2 (prediction (P))-1, RMSEP-0.63 mg/g). PLSR models that developed based on data fusion of Raman and FT-IR spectral features obtained the second best performance (R2CV-0.97, RMSECV-0.38 mg/g; R2P-0.97, RMSEP-0.36 mg/g). This study demonstrated the potential of LIBS, FT-IR and Raman spectroscopy to accurately quantify Ca content in INF.