applying neural network model for adsorption methyl paraben (mp) dye using ricinus communis-capeed fe3o4 nps synthesized from aqueous solution

The applicability of the synthesized ricinus communis-capeed fe3o4nps as a novel adsorbent for eliminating methyl paraben (mp) from aqueous media was investigated. various techniques including brunauer emmett teller theory (bet), fourier transform infrared (ft-ir) spectroscopy, x-ray diffraction (xrd), scanning electron microscopy (sem), and energy dispersive x-ray (edx) were used to characterize this novel adsorbent. the maximum adsorption efficiency of (mp) dye onto ricinus communis-capeed fe3o4 nps was 98.6% at an optimum ph value of 7.0, the adsorbent dosage of 0.01 g, (mp) dye concentration of 15 mg/l, and contact time of 12 min were considered as the ideal values for (mp) dye. the adsorption data fitted well with the langmuir isotherm model with a correlation coefficient (r2 > 0.97), whereas the adsorption kinetics followed the pseudo-second-order kinetics. the use of an artificial neural network model in predicting data with the levenberg–marquardt algorithm, purlin, or a linear transfer function at the output layer, and training was helpful. ann model as a tool (mean square error) mseann = 0.0034, msefl = 0.023, and mseanfis = 0.0020 for removal of the (mp) dye onto ricinus communis-capeed fe3o4 nps synthesis. thermodynamic parameters of free energy (δg0), enthalpy (δh0), and entropy (δs0) of adsorption were determined using isotherms. ∆h0=59.58 kj/mol, ∆g0= -2.8324 kj/mol and ∆s0=221.15 kj/mol. k. the value of (δgo, δho, and δso) confirmed the sorption process was endothermic reflecting the affinity of ricinus communis-capeed fe3o4nps for removing (mp) dye onto ricinus communis-capeed fe3o4nps process requires heat. the maximum monolayer capacity (qmax) was observed to be 195.0 mg/g for (mp) dye at desired conditions.