Akademik Veri Yönetim Sistemi
Journal of Chemometrics, cilt.40, sa.5, 2026 (SCI-Expanded, Scopus)
In this study, a new double-algorithmic chemometric architecture employing continuous wavelet transform (CWT) and parallel factor analysis (PARAFAC) was developed for wavelet-domain separation of overlapping UV spectra and co-quantification of analytes in laboratory-prepared binary mixtures and a commercial tablet formulation. Within this framework, one-dimensional UV absorbance spectra were transformed into the wavelet domain across multiple scale factors to generate structured pseudo-three-way data arrays. These multiway arrays were subsequently decomposed by PARAFAC, enabling the simultaneous extraction of pure spectral profiles, concentration profiles, and scale-dependent features for each component. Unlike prediction-based calibration approaches such as PLS and PCR, the proposed CWT-PARAFAC architecture enables true analytical separation of overlapping spectral components. The wavelet-domain representation was optimized by evaluating different wavelet families over multiple scale ranges, and those providing the best spectral consistency and recovery performance were selected for subsequent multiway analysis. The applicability of the proposed architecture was demonstrated using a representative pharmaceutical binary mixture containing moexipril hydrochloride (MOEX) and hydrochlorothiazide (HCTZ), whose UV absorption spectra strongly overlap. The CWT-PARAFAC framework provided effective analytical separation of overlapping spectral signals and enabled accurate and robust quantification of both analytes. Model performance was confirmed through validation studies, including linearity, precision, and recovery, and the method was successfully applied to the analysis of a commercial tablet formulation. Overall, the results demonstrate that the proposed CWT-PARAFAC double-algorithmic architecture enables effective analysis of complex multicomponent mixture systems exhibiting signal overlap and provides a practical and cost-effective alternative analytical approach to conventional high-cost analytical techniques.