Modeling of the adsorptive removal of arsenic: A statistical approach

Arsenic in drinking water has been recognized as a serious community health problem because of its toxic nature and therefore, its removal is highly essential. A series of adsorption experiments (batch and column) were performed utilizing iron impregnated sugarcane carbon (Fe–SCC), a composite adsorbent, to remove arsenic from aqueous systems. Under optimized batch conditions, the Fe–SCC could remove up to 94.5% of arsenic from contaminated water. The artificial neural network (ANN) model was developed from batch experimental data sets which provided reasonable predictive performance (R2 = 0.964; 0.963) of arsenic adsorption. In batch operation, the adsorbent dose had the most significant impact on the adsorption process. For column operation, central composite design (CCD) in response surface methodology (RSM) was applied to investigate the influence on the breakthrough time for optimization and evaluation of interacting effects of different operating variables. The perturbation plot depicted that the breakthrough time is more sensitive to initial concentration and adsorbent dose than flow rate. The optimized result obtained from bar plot revealed that the Fe–SCC was an effective and economically feasible adsorbent; whereas more than 93% desorption efficiency showed the reusability of the adsorbent. The high arsenic adsorptive removal ability and regeneration efficiency of this adsorbent suggest its applicability in industrial/household systems and data generated would help in further upscaling of the adsorption process.