Abstract
Two validated, accurate and selective chemometric models were developed to detect cyclobenzaprine center dot HCl (CZ) in the presence of its two major degradation products: dibenzocycloheptatrienone and anthraquinone. Additionally, a comparative study was conducted between two chemometric models, namely partial least squares regression (PLSR) and spectral residual augmented classical least squares (SRACLS) via handling UV spectra in the range of 210-340 nm. A 3 factor 4 level experimental design was applied to provide a training set of 16 mixtures. An independent test set consisting of 9 mixtures was established to confirm the prediction capacity of the developed models. The results ascertain the capability of developed multivariate calibration models to selectively determine CZ with high selectivity and accuracy (100.0 +/- 0.7% and 100.0 +/- 0.6% for PLSR and SRACLS, respectively). A statistical comparison between the obtained results and results from the official HPLC method was run, where no significant difference was found.