Feasibility Analysis of Lower Limit of Quantification of NIR for Solvent in Different Hydrogen Bonds Environment Using Multivariate Calibrations

Near infrared (NIR) transmission spectroscopy has been widely used for quantitative analysis in different solvents system. Lower limit of quantification (LLOQ) of NIR in solvent of different hydrogen bonding capability has not been reported. This paper was to investigate the LLOQ of NIR for citric acid in different solvents (ethanol, water, acetone, DMSO) combined with partial least squares (PLS) model and accuracy profile. The parameters of PLS model were optimized by cross-validation. PLS model in each solvent was validated using accuracy profile. The result showed that ethanol was the best quantitative solvent for citric acid, and the LLOQ was 2.791 mg/mL. In water, acetone and DMSO, the LLOQ was 4.542 mg/mL, 3.187 mg/mL and 6.563 mg/mL, respectively. This study provided approximate reference values on the LLOQ of NIR in different solvents system.