The removal of arsenic species from aqueous solution by indigenous microbes: batch bioadsorption and artificial neural network model

• Main Authors: Altowayti, Wahid Ali Hamood, Haris, Somayeh Asadi, Almoalemi, Hafedh, Shahir, Shafinaz, Zakaria, Zarita, Ibrahim, Sallehuddin
• Format: Article
• Published: Elsevier B.V. 2020
• Subjects: Q Science (General)
• Online Access: http://eprints.utm.my/id/eprint/93459/; http://dx.doi.org/10.1016/j.eti.2020.100830

Arsenic contamination of groundwater is a problem that affects millions of people across the world. Biological treatment of arsenic using microorganisms is an interesting alternative to conventional methods because it is most efficient, low cost and eco-friendly. In this study, three different bacterial species were isolated from arsenic laden tailing dam sludge at gold mining and the 16S rRNA sequencing data exposed their affiliation to three different genera, Bacillus thuringiensis strain WS3, Pseudomonas stutzeri strain WS9 and Micrococcus yunnanensis strain WS11. Considering the advantage of the different structures of these bacterial cell walls in adsorption, attempts were made to use individual and mixed dried biomass of these strains to achieve highest As (III) and As (V) removal under different conditions. Mixed dried biomass of WS3, WS9 and WS11 were found to be efficient in the removal of As (III) and As (V) up to 95% and 98%, respectively. Optimal conditions for arsenic removal were found; 8 and 6 h of contact time for As (III) and As (V), respectively, 7.5 (ppm) As (III) and 9 (ppm) As (V) concentration at 37 °C, pH 7, and 0.60 mg/ml of biomass dosage. In comparison of estimated coefficient of determination (R2), Langmuir isotherm model provided the best fit to the experimental data while the adsorption kinetic model followed the pseudo-second-order. FESEM–EDX analysis established diverse cell morphological changes with significant amounts of arsenic adsorbed onto the biomass compared to original biomass. FTIR analysis showed the involvement of hydroxyl, thiol, amide and amine functional groups in arsenic removal. Comparisons between actual and model adsorption results show that the artificial neural network model can predict adsorption efficiency with high accuracy (R2 > 0.98). Consequently, mixed dried biomass of WS3, WS9 and WS11 are strongly recommended for bio-treatment of toxic arsenic from the environment.