When spectral measurements are obtained from a large number of samples and wavelengths, the useful information contained in the data can be hidden, due to overlapping of spectral curves, complexity of the analysed samples and instrumental noise. In order to eliminate systematic variations unrelated to analytes concentration, a preprocessing of the analytical data seems necessary [1]. The aim of this work is the R&D study of a new preprocessing method able to improve the predictive ability of the multivariate calibration models, in particular in the assay of drug components present in lower content or in traces. The new method is based on the elaboration of the spectral curve, which is transformed in a new signal curve obtained by processing the area under the absorbance-wavelength curve. This preprocessing procedure has been successfully used to analyse pharmaceutical systems with components 60 or 120 times less concentrated than the other components. The CAP pretreatment has been compared with other preprocessing procedures. The chemometric models built by adopting the pretreatment CAP showed predictive parameters better than models using ordinary spectrophotometric data or derivative spectral data. The better performance was demonstrated in the validation of the calibration models on synthetic mixtures or commercial formulations.
UV data manipulation by cumulative area preprocessing (CAP) for resolution of multicomponent pharmaceutical systems
De Luca M;IOELE, Giuseppina;RAGNO, Gaetano;
2011-01-01
Abstract
When spectral measurements are obtained from a large number of samples and wavelengths, the useful information contained in the data can be hidden, due to overlapping of spectral curves, complexity of the analysed samples and instrumental noise. In order to eliminate systematic variations unrelated to analytes concentration, a preprocessing of the analytical data seems necessary [1]. The aim of this work is the R&D study of a new preprocessing method able to improve the predictive ability of the multivariate calibration models, in particular in the assay of drug components present in lower content or in traces. The new method is based on the elaboration of the spectral curve, which is transformed in a new signal curve obtained by processing the area under the absorbance-wavelength curve. This preprocessing procedure has been successfully used to analyse pharmaceutical systems with components 60 or 120 times less concentrated than the other components. The CAP pretreatment has been compared with other preprocessing procedures. The chemometric models built by adopting the pretreatment CAP showed predictive parameters better than models using ordinary spectrophotometric data or derivative spectral data. The better performance was demonstrated in the validation of the calibration models on synthetic mixtures or commercial formulations.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
