The ATR-FTIR quantitation of azithromycin in three products of commercial tablets was carried out on product specific quantitative regression models using powdered paracetamol as matrix modifier to overcome the variation of spectral response and influence of sample matrix. For each product, a PLS quantitative regression model was established using training infrared spectra obtained from reference mixtures (reference powders with known mass content (%, w/w) of azithromycin mixed homogenously with paracetamol to have mass percentage of azithromycin over total mass of azithromycin and paracetamol (PA) from 30% to 70%). The spectral data were collected in wavenumber range depending on commercial product within the wavenumber zone from 1300 cm−1 to 1750 cm−1 to build quantitative regression models. To quantify azithromycin in any commercial batch of the same product, the homogenized sample powder was mixed with paracetamol to have mixtures with PA value about 50% to record infrared spectrum. The actual amount of azithromycin would then be calculated from spectral response of unknown sample and the pre-established quantitative regression model. Each quantitative regression model was validated according to the current requirements of ICH guideline Q2R1 and those of AOAC International in term of specificity, accuracy, precision, long-term robustness and reliability. The validation results proved that the quantitative regression models were accurate, precise, reliable and robust, able to provide quantitative results of azithromycin in tablets equivalent to those provided by official HPLC method of USP44.
Read full abstract