Batch and continuous reactor studies for the adsorption of As(III) from wastewater using a hybrid biochar loaded with transition metal oxides: Kinetics and mass transfer analysis

Abstract

As(III) presence in low concentration (1-5 mg/L) in water presents a challenging problem in its removal. In the present study, biochar prepared by pyrolysis of mustard cake and loaded with Fe-Mn binary oxides through hydrothermal technique was used for adsorptive removal of As(III) from water in batch and continuous mode. The synthesised biochar exhibited mesoporous structures in the range of 2-50 nm (based on BET analysis). The maximum adsorption capacity (95.7 mg/g) obtained using biochar loaded with both Fe-Mn oxides was found to be 1.4 times higher than that of pristine biochar. The adsorption equilibria was best described by Freundlich isotherm (based on R 2 and χ 2) suggesting that the As(III) adsorption was multilayered. The external mass transfer coefficients (β L = 10-5 cm 2 /s) were observed to be higher than film (D f = 10-7-10-9 cm 2 /s) and intraparticle (D i = 10-9 cm 2 /s) diffusivities in batch mode. In column studies, Thomas model gave the best correlation coefficient (R 2 > 0.95) and the adsorption was limited by external mass transfer. Kinetic rate constant decreased with increase in initial As(III) concentration and flow rate. The oxide loaded biochar exhibited reusability up to three times for As(III) removal.

Machine learning for the management of biochar yield and properties of biomass sources for sustainable energy

“…This enables the researcher to explore how well the model can adapt to new data, and prevents errors being introduced by a single training-testing split. 200 An investigator may employ cross-validation, holdout testing, or external validation with diverse datasets to confirm the accuracy and representativeness of a generalized model. 201,202 The technique of stratified sampling can be used for diverse data, testing for resilience to noise, and evaluating the importance of features.…”

“…When samples from different types of biomass and an extensive variety of experimental settings are used, the model can learn reliable patterns that can be applied to different situations. 200 The ability to explain how a model works is also important in terms of knowing how predictions are made. Techniques such as feature importance evaluation and SHapley Additive exPlanations (SHAP) values can be used to identify which factors have the most impact on the model's output.…”

“…Cross‐validation involves dividing the dataset into multiple subsets, training the model on certain portions of them, and testing its performance on the rest of the data. This enables the researcher to explore how well the model can adapt to new data, and prevents errors being introduced by a single training–testing split 200 . An investigator may employ cross‐validation, holdout testing, or external validation with diverse datasets to confirm the accuracy and representativeness of a generalized model 201,202 .…”

“…At the training stage, it is important to use a diverse and representative dataset to make it easier to generalize and transfer the model. When samples from different types of biomass and an extensive variety of experimental settings are used, the model can learn reliable patterns that can be applied to different situations 200 . The ability to explain how a model works is also important in terms of knowing how predictions are made.…”

Features of Arsenic Distribution in the Soils of Potash Mines

“…Arsenic is a chemical element of the first class of environmental hazard with high toxicity and carcinogenic properties that pose a serious risk to humans [2,3]. Arsenic can enter the human body through the consumption of arsenic-contaminated water or agricultural products grown on arsenic-contaminated soils.…”