A green and sustainable UV spectrophotometric approach for simultaneous determination of rosuvastatin, pravastatin, and atorvastatin in pharmaceuticals leveraging firefly algorithm-enhanced partial least squares regression.

This study aimed to develop a green and sustainable analytical method for the quantitative determination of three statins-rosuvastatin, pravastatin, and atorvastatin-using their UV spectral fingerprints. Partial Least Squares (PLS) regression combined with the Firefly Algorithm (FFA) for variable selection was employed to optimize the analysis. A partial factorial design was used to construct a 25-sample synthetic calibration set, while a central composite design served for external validation. The FFA-PLS approach demonstrated superior performance over traditional PLS models, achieving relative root mean square errors of prediction of 1.68%, 1.04%, and 1.63% for rosuvastatin, pravastatin, and atorvastatin, respectively, compared to 2.85%, 2.77%, and 3.20% for conventional PLS. FFA-PLS also enabled model simplification, reducing latent variables from 4, 3, and 4 to 2, 2, and 3 for the respective statins while requiring fewer wavelengths. Validation in accordance with ICH guidelines further confirmed the method's accuracy, precision, and selectivity. Besides, application to real pharmaceutical samples yielded mean recoveries ranging from 99.23% to 99.90%, with RSD% below 2%. Furthermore, comparative analysis with reported chromatographic methods revealed no significant differences in terms of mean and variance as calculated by a two-tailed -test and -test, respectively. Finally, environmental impact assessment metrics demonstrated the method's superior sustainability (AGREE score: 0.78 0.64 for HPLC; RGB12 whiteness index: 91.4% 75.8% for HPLC-UV). In conclusion, the proposed UV-PLS-FFA method offers an effective, accurate, and environmentally friendly alternative for the determination of statins in pharmaceutical samples, aligning with the principles of green chemistry and sustainability and has potential for broader applicability beyond the scope of this study.