Coupling Second-Order Excitation-Emission Spectrofluorimetric Data with Standard Addition method to Quantify Carvedilol in Real Samples

Prediction using pure standards is expected to be biased whenever the slope of the calibration is affected by the presence of sample matrix. Moreover, in the presence of unknown spectral interferents, first-order algorithms like partial least squares cannot be used. In this study, a method for determination of carvedilol (CAR) in tablet and urine samples is proposed by excitation-emission fluorescence spectroscopy (EEM). The multivariate curve resolution-alternating least-squares (MCR-ALS) coupled with trilinearity constraint exploiting the second order advantage is applied for the analysis of EEMs. Moreover, the combination of standard addition with MCR-ALS was used to correct the matrix effect. Indeed, by the proposed strategy, both matrix effect and the problem of the presence of unknown interferents in determination of CAR are overcome.       The MCR-ALS method was applied on EEMs under non-negativity and trilinearity constraints. For both samples, CAR was quantified at low mg l-1 level with an overall prediction error of -3.1% and -4.0% in urine and tablet samples, respectively.