Synthesis and characterization of mixed phase manganese ferrite and hausmannite magnetic nanoparticle as potential adsorbent for methyl orange from aqueous media: Artificial neural network modeling

Abstract In this study, a novel magnetic nanoscale adsorbent of mixed phase manganese ferrite and hausmannite (MNF–HM) was synthesized by a simple chemical precipitation method and its adsorptive property towards a toxic anionic dye methyl orange (MO) was investigated. This new adsorbent was characterized using X-ray diffraction (XRD), scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FTIR), point of zero charge (PZC), vibrating sample magnetometer (VSM) measurements and Brunauer–Emmett–Teller (BET) analysis. These crystalline mixed phase MNF–HM nanoparticles show high saturation magnetization (30.12 emu/g) as well as high specific surface area (100.62 m 2 /g) and pore volume, which makes it a magnetically separable efficient adsorbent with maximum adsorption capacity of 303.66 mg/g for MO at room temperature. The solution pH becomes the most influencing parameter and maximum adsorption efficiency of 97.4% is observed within 45 min at solution pH 2.0 with initial MO concentration of 50 mg/L and MNF–HM nanoparticle dose of 1.0 g/L. An artificial neural network (ANN) model was developed for accurate prediction of MO removal (%) and optimal ANN structure (4–7–1) shows minimum mean squared error (MSE) of 0.00071 and high determination coefficient (R 2 ) of 0.990. Isotherm modeling exhibited that Langmuir isotherm model is the most efficient one in explaining its adsorption behavior and the kinetic studies reveal that the adsorption process follows pseudo-second-order kinetic model with the involvement of intra-particle diffusion. Thermodynamic studies reveal that the adsorption process is feasible, spontaneous and exothermic in nature.