Architecture of GO/CoFe2O4/ZnO nanocomposite for efficient fluoride removal: An approach using RSM, ANN and GRU modeling

The presence of fluoride in wastewater possesses a significant risk to the environment. Graphene oxide-based nanocomposites were prepared as potential adsorbents for fluoride removal from water. Various techniques including X-ray diffraction (XRD), Scanning Electron Microscopy (FE-SEM), Energy dispersive X-ray (EDX), Fourier Transform Infrared Spectroscopy (FTIR), RAMAN spectroscopy, Ultraviolet-Visible spectroscopy (UV-vis), Vibrating Sample Magnetometer (VSM), X-ray Photoelectron Spectroscopy (XPS), Brunauer Emmett Teller (BET), and Electrochemical Impedance Spectroscopy (EIS) were employed to investigate the morphological, structural, chemical composition, elemental analysis, surface area, optical band gap, etc. of the CoFe2O4, CoFe2O4/ZnO composite and GO/CoFe2O4/ZnO nanocomposite. XRD analysis confirmed a reduction in crystallite size to 15.7 nm of GO/CoFe2O4/ZnO. Batch adsorption studies were conducted for fluoride removal with input independent variables (pH, temperature, initial fluoride concentration, time, and an adsorbent GO/CoFe2O4/ZnO dosage). The Response Surface Methodology (RSM) and Gated Recurrent Unit (GRU) models predicted the fluoride removal process, with RSM achieving an optimal fluoride removal of 94.3%, GRU achieving 94.5% and Artificial neural network (ANN) attaining 93.72%. A kinetic investigation was carried out using Pseudo first order (PFO) and Pseudo second order (PSO). Results confirmed that the adsorption of fluoride followed the PSO kinetic model with an (R-2 = 0.999). Isotherm studies were conducted using Langmuir, Temkin, Freundlich, and Dubinin Radushkevich isotherms. Among them, the best model follows an adsorption mechanism, with (R-2 = 0.984) and an adsorption capacity of 10.68 mgg(-1) was Langmuir Isotherm. The real water sample of Ravi River, Pakistan also investigated for fluoride removal capability. This research contributes to the promotion of RSM and GRU modeling with high sustainability for water purification. The contemplated nano composite (GO/CoFe2O4/ZnO) is a potential adsorbent for the mechanism of water detoxification containing fluoride which may be useful for environmental remediation as well.

Automated summary

This study proposes graphene oxide-based nanocomposites as potential adsorbents for fluoride removal from water. Various characterization techniques were employed to investigate the properties of the nanocomposite. The Response Surface Methodology, Gated Recurrent Unit, and Artificial Neural Network models were used to predict the fluoride removal process. Kinetic and isotherm studies were conducted to understand the adsorption process. The research shows that the nanocomposite has high adsorption capacity and can effectively remove fluoride from real water samples.