Greenness and whiteness evaluated triple divisor-continuous wavelet transform spectrophotometric and advanced chemometric methods for the simultaneous determination of four photoinitiators in saliva and intravenous injection solution

• Published in: Sustainable chemistry and pharmacy 2023-12, Vol.36, p.101284, Article 101284
• Main Authors: Gamil, Mirna, El Zahar, Noha M., Magdy, Nancy, El-Kosasy, Amira M.
• Format: Article
• Language: English
• Subjects: Chemometrics; CWT; Intravenous; Photoinitiators; Saliva; Spectrophotometry; Triple divisor
• ISSN: 2352-5541; 2352-5541
• DOI: 10.1016/j.scp.2023.101284

Migration of photoinitiators from packaging materials and dental products in saliva presents an extensive real damage, particularly that emerges from packed food and dental products in saliva or from polyethylene packaging of the intravenous solutions which is extremely dangerous to human health and to the environment. In line with this, environmentally evaluated, advanced and reliable triple divisor combined with continuous wavelet transform (CWT) spectrophotometric method and nine chemometric methods were established and validated for the concurrent determination of four photoinitiators: Benzophenone, 4-Hydroxyacetophenone, Benzoin, and Diphenyliodonium hexafluorophosphate in saliva and in intravenous injection solutions. To our knowledge, this is the first spectrophotometric method for the analysis of the cited photoinitiators as well as the primary white method for the examination of those photoinitiators. Also, in this paper, the triple divisor method is combined with continuous wavelet transform for the first time. Triple divisor–CWT peak amplitudes were defined at 260.0, 275.8, 262.2 and 231.2 nm over a linear dynamic range of 0.50–13.00, 0.80–10.00, 1.5–30.00 and 1.00–28.00 μg/mL for Benzophenone, 4-Hydroxyacetophenone, Benzoin and Diphenyliodonium hexafluorophosphate, respectively. The nine chemometric methods incorporate the application of principle component regression (PCR), orthogonal partial least square (OPLS), ridge regression (RR) and artificial neural networks (ANN) methods along with different variable selections including genetic algorithm (GA-PCR, GA-OPLS and GA-RR) and interval selections (Interval-PCR and Interval-OPLS). Regarding the adopted chemometrics models, a multivariate multifactor experimental design based on the calibration curve range of the studied photoinitiators was considered to construct a calibration set consisting of eighteen mixtures while using seven mixtures for the validation set. The proposed methods were demonstrated to be sensitive, accurate, precise and reproducible throughout the validation process. The acquired results revealed that the proposed spectrophotometric and chemometric approaches could be effectively applied for the regular analysis of the cited photoinitiators in biological fluids as saliva and in intravenous solutions. The greenness of the proposed methods was assessed using two green metric tools enclosing Analytical Eco-scale & Analytical Greenness Assessment (AGREE), and a White Analytical