Landfill leachate treatment by activated carbon (AC) from banana pseudo-stem, iron oxide nanocomposite (IOAC), and iron oxide nanoparticles (IONPs)

This study investigated the efficiency of activated carbons (ACs) produced from banana pseudo-stem, iron oxide nanocomposite (IOAC), and iron oxide nanoparticles (IONPs) for chemical oxygen demand (COD, mg/L), dissolved organic carbon (DOC, mg/g), and color (mg/g) removal from landfill leachate. A continuous-flow bed column study was conducted to ascertain behavioral pattern of the adsorbents during adsorption experiments (variables: bed column height, flow rate, and influent concentration). Kinetics (pseudo-first (PFO) and pseudosecond (PSO) orders) and isotherms (Langmuir, Freundlich, and Temkin) models were used to describe adsorption trends of pollutants. An artificial neural network (ANN) was used to model the adsorption process. Risk assessment analysis on the treated effluent was determined based on risk quotients (RQ) from COD standards for sewage and effluents. The bed column study showed that percentage removal of COD, DOC, and color increased with increasing the bed height while empty bed contact time (EBCT) increased from 1.6 to 4.7 min as the column bed height increased from 7.5 to 12.5 cm respectively. The adsorption kinetics for all treatments were well-fitted with the PSO kinetic model suggesting that the removal rate of COD, DOC, and color is dependent on the availability of the sorption sites. The adsorption process was observed to have good linearity in the three isotherm models with an overall R2 > 0.90. The values from t for each dataset of AC, IOAC, and IONPs fit the qt Levenberg-Marquardt (LM) training algorithm with best validation performance, lowest mean square error (MSE) of 0.19789 at epoch 8 and highest coefficient of determination (R2) of 0.9922. The IONPs treatment showed the lowest RQ (0.5732-0.4723) from all dosages of adsorbents indicating medium risk to aquatic organisms.