Influence of torrefaction and pyrolysis on engineered biochar and its applicability in defluoridation: Insight into adsorption mechanism, batch adsorber design and artificial neural network modelling

Abstract

Natural and anthropogenic activities lead to massive contamination of groundwater with various organic and inorganic pollutants. Of these, fluoride is a major contaminant and elevated concentration of fluoride results in many health-related issues such as dental and skeletal fluorosis. This study aims to exploit an agricultural waste (rice husk) as a feedstock for synthesis of a value-added product (biochar), for effective defluoridation of drinking water. To that effect, thermal (torrefaction/pyrolysis) and chemical (iron/zinc) activation of rice husk-derived biochar were tested. Activation of biochar via pyrolysis followed by iron modification was capable of maximum fluoride removal (95.4 %). Optimized conditions for fluoride adsorption were time (120 min), biochar dose (4 g/L), solution pH (4) and initial F− concentration (5 mg/L). Additionally, a predictive modelling was carried out using an artificial neural network (ANN) model with four input variables and one output variable. Accuracy of the model was tested using error functions such as mean absolute error (0.002), root mean squared error (0.32) and coefficient of determination (0.99) for training, 3.78, 13.54 and 0.62 for validation and 5.10, 6.66 and 0.98 for testing respectively. Langmuir isotherm was best suited model among the isotherms tested with adsorption capacity of 4.45 mg/g. Rice husk derived biochar was found to be an economically viable alternative to commercially available adsorbents, which could also be employed for the remediation of other groundwater pollutants.