Second-order multivariate models for the processing of standard-addition synchronous fluorescence-pH data. Application to the analysis of salicylic acid and its major metabolite in human urine.

In the present work, we describe the determination of salicylic acid and its major metabolite, salicyluric acid, in spiked human urine samples, using synchronous fluorescence spectra measured in a flow-injection system with a double pH gradient. Because the fluorescent urine background constitutes a potentially interfering signal, it becomes necessary to achieve the second-order advantage. Moreover, due to significant changes in the signal of the analytes in the presence of the urine matrix, mainly for salicyluric acid, standard addition was required in order to obtain appropriate quantifications. Several second-order multivariate calibration models were evaluated for this purpose: PARAFAC and MCR-ALS in two different modes, and PLS/RBL.