Green synthesis of ZrO2/PdO-NPs for photodegradation of anionic dyes: Photocatalytic activity and machine learning modelling

• Published in: Journal of molecular liquids Vol. 410; p. 125581
• Main Authors: da Silva, Mauricio Dalla Costa Rodrigues, Druzian, Daniel Moro, Brum, Luis Fernando Wentz, dos Santos, Cristiane, Pavoski, Giovani, Espinosa, Denise Crocce Romano, Ruiz, Yolice Patricia Moreno, Galembeck, André, da Silva, William Leonardo
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 15-09-2024
• Subjects: Anionic dyes; Gradient Boosting Regression; Green synthesis; Machine Learning; Wastewater treatment; Anionic dyes; Green synthesis; Wastewater treatment; Machine Learning; Gradient Boosting Regression
• ISSN: 0167-7322
• DOI: 10.1016/j.molliq.2024.125581

•C. Illinoiensis shell extract was used for green synthesis;•ZrO2/2.5_PdO-NPs achieved 97.67% removal of Congo Red (CR) dye;•ZrO2/2.5_PdO-NPs presented good photocatalytic stability after 6 cycles;•Machine learning study was effective to proposed a degradation mechanism for CR;•ZrO2/2.5_PdO-NPs also showed potential for degraded other anionic dyes;
The present work aims to synthesize and characterize zirconia dioxide impregnated with palladium nanoparticles (ZrO2/PdO-NPs) for the photodegradation of Congo red (CR), methyl orange (MO), erythochrome T black (EBT) and tartrazine yellow (TY) dyes. The extract of Carya illinoiensis shells was used for the production of ZrO2/(1.0, 2.5, and 5.0)_PdO-NPs. Heterogeneous photocatalysis was carried out in a batch system for 180 min followed by an experimental design, where kinetic studies, variations of the main process parameters, electron scavengers, reuse and machine learning were demonstrated to predict possible degraded molecules. The characterizations confirmed the synthesis of ZrO2/2.5_PdO-NPs with phases and groups of both constituents and properties suitable for heterogeneous photocatalysis, such as band gap energy of 2.16 eV, zeta potential of −12.93 ± 0.73 mV, zero-point charge of 6.99 ± 0.04 and demonstrating agglomerated particles with spherical and elliptical morphology and crystallite size in the range of 3 – 15 nm. The experimental design indicated the ideal condition for photocatalysis of [CR] = 20 mg/L, [ZrO2/2.5_PdO-NPs] = 1.75 g/L, and pH=10 with a removal of 99.7 % and a pseudo first-order model (k = 0.02459 min−1). The ZrO2/2.5_PdO-NPs showed a good photocatalytic stability after 6 cycles (about 15 % reduction in activity) and the production of O2●- significantly influenced the CR photodegradation. Regarding the degradation mechanism, the machine learning study indicated that the azo bonds in the CR dye were cleaved during the photocatalysis, forming smaller molecules that can be further mineralized. Tests with other dyes (MO, EBT, TY) demonstrated photodegradation of around 92 % to 60 %. Therefore, it was possible to synthesize supported metallic nanoparticles with great potential for application as nano catalysts for different dyes.