Waste-derived activated carbons for effective adsorptive removal of strontium, barium, and binary pollutants: A response surface methodology study

Abstract

A double-pronged approach to pollution management is proposed by sustainably managing solid wastes and converting them to activated carbons that are then utilized for water treatment. In this study, gas-to-liquids (GTL) derived biosolids, carboard and their mixed samples are used to produce the activated carbons. In a laboratory batch study, a Box-Behnken experimental design was used with four factors and three levels to optimize the removal of single component strontium and barium, and the binary system of pollutants from an aqueous solution. The design incorporated response surface modeling (RSM) techniques with a total of 29 different experimental data points collected and analyzed in this study. The study was conducted considering four parameters: the initial pH of the solution (ranging from 4.0 to 8.0), temperature (ranging from 20 to 40 ℃), the percentage of cardboard (fixed at 0 %, 50 %, and 100 %), and the amount of adsorbent (between 0.05 and 0.5 g). These factors were assigned three levels, represented as −1, 0, and 1. A second-order polynomial regression equation was then developed to estimate the responses. The statistical significance of the independent variables and their interactions was assessed using analysis of variance (ANOVA) with a 95 % confidence level (α = 0.05). The results revealed that only temperature and dosage show significant effects on the responses and that optimum values of the selected parameters were obtained by solving the prediction equations, which were validated with less than 4 % error %. In an attempt to optimize the factors, a pH of 5.5, temperature of 40 ℃ and dose of 0.3 g is found for all three samples. Validation results for optimization also proved that varying the percentage cardboard showed little difference in the percentage removal of all the pollutants. The results from this study can be directly applied for any such systems trying to optimize these parameters and the prediction equations can be utilized effectively.