Application of ANN, hypothesis testing and statistics to the adsorptive removal of toxic dye by nanocomposite

Abstract

Statistics can be used in a variety of ways to present, compute, and critically analyze experimental data. To determine the significance and validity of the experimental data, a variety of statistical tests are used. Using a synthesized CoO/NiO/MnO2 Nanocomposite, the present study used adsorption to remove the dye Bromophenol Blue (BPB) from a contaminated aqueous solution. In order to (a) determine the optimal pH of the solution, (b) confirm the experiment's success, and (c) investigate the effect of adsorbent dose on BPB dye removal from aqueous solutions. The experimental data were statistically analyzed through hypothesis testing using the t-test, paired t-test, and Chi-square test. The null hypothesis that the optimal pH value is 7 is accepted since tobserved (−1.979)<ttabulated (−2.262). Since χ2observed (1.052)< χ2tabulated (3.841), null hypothesis that the higher adsorbent dose helps in higher % removal of dye is accepted. Both the obtained Freundlich adsorption isotherm and the Langmuir isotherm's R2 values, which were both close to 1, indicate that the isotherms are favorable. Karl Pearson's relationship coefficient values for Langmuir and Freundlich adsorption isotherms found to be 0.9693 and 0.9994 respectively, which show a more significant level of connection between's the factors. The ANN model predicted adsorption percentage with regression value R is 0.996. ANN model result predict 99.60 % BPB dye adsorption using optimized parametric conditions. The ANN model produced values that were more precise, reliable, and reproducible, demonstrating its superiority.