Answers are grounded in this paper and cite their sources.
Enhanced aqueous arsenite (As(III) removal by adsorption on marble waste powder (MWP) in batch and continuous mode was investigated. A predictive (ML) algorithm was developed to predict arsenic removal by MWP. This study pioneers the use of ML applications on MWP. The batch-scale data revealed that the adsorption of arsenite on MWP can be best described by the Liu Isotherm model, and non-linear pseudo-first-order kinetics were observed. The intraparticle diffusion model revealed adsorption occurred in more than one step, and further analysis indicated that mass transfer was the dominant step. Under favorable conditions, the regeneration potential of MWP was also observed. From laboratory experiments, three comprehensive datasets (Batch adsorption, continuous adsorption, and regeneration) were generated and used for predictive modelling. Different linear and non-linear ML models were first optimized with hyperparameter tuning using GridSearchCV and then trained and evaluated for their performance. Evaluation metrics and learning curves showed that non-linear ML models outperformed linear models. The extra trees model was the most accurate predictive model, with prediction accuracy of 91.1 %, 99.2 % and 66.8 % in datasets, respectively. Theoretical up-scaling suggests fixed-bed pilot-scale system of MWP can treat around 6000 liters of arsenic-contaminated water in 1.37 days before the breakthrough. • Marble waste powder is potential adsorbent for reducing arsenite contamination, ranging from 5 to 1000 μg/L, in water. • Pilot-scale fixed-bed column system of marble waste powder can treat approximately 6000 liters of water in 1.37 days. • The extra trees model can efficiently predict arsenite removal from water using marble waste powder. • Arsenite saturated marble waste powder can be regenerated in alkaline conditions.
No takes yet. Share an insight, caveat, or question.
Devi et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: