Aplicação de métodos quimiométricos de calibração e resolução multivariada de curvas em espectroscopia Raman para Análise qualitativa e quantitativa de polimorfismo em carbamazepina, clopidogrel e ezetimiba

Abstract

In this work were used chemometric tools like PLS, iPLS to perform multivariate calibration of ternary and binary mixtures of polymorphs of carbamazepine, ezetimibe and Clopidogrel. Initially were synthesized two polymorphs of carbamazepine and two polymorphs of clopidogrel from its respective commercial polymorphs, totaling three polymorphs for each of these drugs. For carbamazepine were synthesized the dihydrate polymorphs (CBZ DH) and polymorph I (CBZ I) from the Polymorph III (CBZ III), which were characterized by DSC and XRD, and Raman spectroscopy. Inittialy an iPLS model was built with the polymorphs. The samples were prepared according to a ternary mixture, totaling 10 samples for a set of calibration and a further 5 to a set of validation. At each sample was collected 10 Raman spectra at different points, subsequently an arithmetic average of the 10 spectra was calculated, resulting in an average spectrum with which is built the model calibration and validation presenting prediction error: (CBZ I) 6 2 mg; (CBZ III) 6.8 mg; (CBZ DH) 11.6 mg. In a second step the same procedure was conducted, but samples (300 mg) were in the presence of its excipients (microcrystalline cellulose, sodium croscarmellose, magnesium stearate and colloidal silicon dioxide). The samples were prepared by adding (250 mg), carbamazepine (50 mg), prediction errors were: (CBZ I) 4.8 mg; (CBZ III) 3.0 mg; (CBZ DH) 0.7 mg. For bisulfate Clopidogrel were synthesized polymorphs II (CBS II) and amorphous (amorphous CBS) from polymorph I (CBS I, commercial), the iPLS calibration model was constructed as done for carbamazepine, however were used mid-infrared and Raman spectroscopy and the samples were not prepared in the presence of its excipients. Prediction errors for the model with Raman spectroscopy was: (CBS I) 3.2 mg; (CBS II) 5.0 mg; (Amorphous CBS) 6.1 mg and mid-infrared spectroscopy was: (CBS I) 4.8 mg; (CBS II) 7.4 mg; (Amorphous CBS) 6.9 mg. For ezetimibe was synthesized the hydrate (EZT-H) from the commercial polymorph (EZT-An) and held a binary mixture of the two polymorphs, with the concentration of each polymorphs ranging 10% by weight in each sample, since in the presence of its excipients. Each sample weight 100 mg, with 10 mg of ezetimibe and 90 mg of excipients (croscarmellose sodium, lactose monohydrate, povidone and sodium lauryl sulphate), the same method was applied for the preparation of the samples validation set. The calibration model was built by iPLS with the following prediction errors: (EZT-An) 0.7 mg; (EZT-H) 0.6 mg. A new sample was prepared only with EZT-An and the excipients which was vacuum packed in a polyethylene container, carried out a monitoring for 24 hours where it was seen that in this period the hydration of Ezetmiba was already 72% due to the moisture of the excipients.