Open AccessAdsorptive removal of dairy industrial pollutants by chitosan zinc –oxide nanoadsorbent- A comparitive study of artificial neural network and quadratic Box-Behnken design
Author(s) -
B. L. Dinesha,
Sharanagouda Hiregoudar,
Udaykumar Nidoni,
K. T. Ramappa,
A.T. Dandekar,
M. V. Ravi,
K.B. Sankalpa
Publication year - 2021
Publication title -
journal of environmental biology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.247
H-Index - 48
eISSN - 2394-0379
pISSN - 0254-8704
DOI - 10.22438/jeb/42/6/mrn-1724
Subject(s) - chemical oxygen demand , biochemical oxygen demand , box–behnken design , adsorption , scanning electron microscope , zinc , response surface methodology , coefficient of determination , chemistry , diffractometer , materials science , nuclear chemistry , analytical chemistry (journal) , mathematics , chromatography , environmental engineering , environmental science , metallurgy , composite material , wastewater , organic chemistry , statistics
Aim: To investigate the effect of operational parameters on the adsorption of biological oxygen demand (BOD) and chemical oxygen demand (COD) on to Chitosan zinc oxide (CZnO) nanoadsorbent using cost-effective and eco-friendly nanoadsorbent based effluent treatment processes. Methodology: CZnO nanoadsorbent particle was synthesized using chemical precipitation method. The nano size <100 nm was achieved using high-speed cryo all mill, followed by the characterization using high-end instruments such as scanning electron microscope with elemental detection sensor (SEM-EDS), atomic force microscope (AFM), X-ray diffractometer (XRD) and Fourier transform inform infrared spectroscopy (FT-IR). Modeling and optimization of operational parameters were done with the artificial neural network (ANN) and Box-BehnkenDesign (BBD) statistical tools. Results: Optimized treatment combination for adsorption of BOD and COD were found at initial BOD and COD concentration of 100 and 200 mg l−1, pH of 7.0 and 2.0, adsorbent dosage of 1.25 mg l−1, contact time of 100 and 60 min. In these conditionsthe desirability values of 0.988 and 0.950 were found for BOD and COD adsorption. The maximum per cent reduction of BOD and COD by using CZnO nanoadsorbent was found to be 96.71 and 87.56. Two models such as Quadratic Box-Behnken and ANN were compared in term of sum of square errors (SSE), root mean square error (RMSE) and correlation coefficient (R2) values. Interpretation: The results obtained revel the well trained ANN model found to be more accurate in prediction of BOD and COD adsorption process parameters compared to BBD model.