Quantitative structure-activity relationship(QSAR) models for color and COD removal for some dyes subjected to electrochemical oxidation

• Published in: Environmental technology Vol. 44; no. 16; pp. 2374 - 2385
• Main Authors: Hirpara, Katha S., Patel, Upendra D.
• Format: Journal Article
• Language: English
• Published: England Taylor & Francis 16-07-2023; Taylor & Francis Ltd
• Subjects: Atomic properties; Chemical oxygen demand; Color; Color removal; Coloring Agents - chemistry; Degradation; Dyes; Electrochemical oxidation; electrochemical oxidation (EO); Electrochemistry; Electrolysis; Goodness of fit; Hydrogen bonds; molecular descriptors; Molecular orbitals; Molecular Structure; normalization; Oxidation; Prediction models; QSAR; Quantitative Structure-Activity Relationship; Quantum chemistry; Statistical analysis; Structure-activity relationships; Wastewater; molecular descriptors; electrochemical oxidation (EO); dyes; QSAR; normalization
• ISSN: 0959-3330; 1479-487X
• DOI: 10.1080/09593330.2022.2028014

Electrochemical oxidation is an efficient method for the destruction of dyes in wastewater streams. The experimental conditions during electrochemical oxidation (EO) and molecular structure of a dye greatly influence the extent of degradation. The extent of degradation for a variety of dyes by EO can be predicted conveniently by the use of Quantitative structure-activity Relationship (QSAR) models. An abundant amount of published data on dye degradation by EO using highly variable experimental conditions lies unutilized to prepare QSAR models. In this study, an effort is made to use published experimental data on EO of aqueous dyes after applying an easy method of normalization, to prepare QSAR models for percent color and COD removal. Normalized color and COD removal were obtained by multiplying the reported removal by volume of reactor and concentration of dye; and divided by total current passed and the time of electrolysis. More than 15 molecular descriptors were computed using Schrodinger-suit 2018-3. The multiple linear regression (MLR) approach was used to develop normalized color and COD removal models. The quantum chemical descriptors: highest occupied molecular orbital energy (HOMO) and lowest unoccupied molecular orbital energy (LUMO), polar surface area (PSA), hydrogen bond donor count (HBD), and number of atoms were found significant. The statistical indices: goodness-of-fit, R
2
 > 0.75, and internal and external validations, Q
2
LOOCV
and Q
2
ext
, > 0.5, satisfied the criteria for predictive models and indicated that the method of normalization used in this study is adequate. Developed QSAR models are quite simple, interpretable, and transparent.