Ultrasonic Assisted and Neural Network Model for Adsorption of of humic acid (HAs) by Synthesised CM-β-CD-Fe3O4NPs from Aqueous Solutions

Abstract

ABSTRACT In the present study, the applicability of CM-β-CD-Fe3O4NPs synthesis for eliminating humic acid rapidly from aqueous solutions. Identical techniques including BET, FT-IR, XRD and SEM have been utilised to characterise this novel material. The investigation showed the applicability of CM-β-CD-Fe3O4NPs as an available, suitable and low-cost adsorbent for the proper deletion of humic acid from aqueous media. Also, the impacts of variables including initial humic acid (HA) concentration (X1), pH (X2), adsorbent dosage (X3) and sonication time (X4) came under scrutiny using central composite design (CCD) under response surface methodology (RSM). The experiments have been designed utilising response surface methodology. In this current article, the values of 12 mg·L−1, 0.03 g, 7.0, 4.0 min were considered as the ideal values for humic acid (HA) concentration, adsorbent mass, pH value and contact time, respectively. The kinetics and isotherm studies proved the appropriateness of the second-order and Langmuir models for the kinetics and isotherm of the adsorption of humic acids (HAs) on the adsorbent. The adsorbent was proved to be recyclable more than once. Since almost 99.5% of humic acids (HAs) were deleted with ideal adsorption capacities of 105 mg·g−1 for humic acids (HAs). The use of artificial neural network model in predicting data with Levenberg–Marquardt algorithm, purelin or a linear transfer function at the output layer and tansig or a tangent sigmoid transfer function in the hidden layer with five neurons was helpful for deletion of humic acid (HAs), MSE (mean square error) of 0.33 for humic acid (HA) removal by CM-β-CD-Fe3O4NPs. The overall results confirmed that CM-β-CD-Fe3O4NPs could be a promising adsorbent material for humic acid (HA) removal from aqueous solutions.