Abstract
This paper demonstrates the high potential of excitation-emission phosphorescence matrices in combination with multivariate regression for determining drugs in a complex matrix. The simultaneous determination of naproxen and propranolol in urine by micelle-sensitized room temperature phosphorescence is made difficult by strong overlap between the three-dimensional phosphorescence spectra for the two analytes and by the inner filter effect arising from the presence of a biological fluid. To circumvent these shortcomings, we used three-dimensional phosphorescence spectroscopy in combination with multivariate calibration methodology. The ensuing method allowed us to determine both compounds at microgram-per-litre levels without the need for separation. Using a calibration matrix in combination with a factorial design comprising two levels per factor and a central star design ensured accurate calibration throughout the desired concentration intervals. The proposed method was validated by application to a test set of synthetic samples. Based on the results, the N-PLS model used affords the accurate simultaneous determination of both analytes in urine samples
Keywords: Multivariate calibration, Naproxen, Propranolol, Room temperature phosphorescence, Urine, Excitation, –, emission phosphorescence, Room temperature phosphorimetry (RTP), Chemometric Models, N-PLS/U-PLS Analysis, Validation
26