Adiantum Capillus-Veneris\nExtract as a Sustainable\nInhibitor to Mitigate Corrosion in Acid Solutions: Experimental, Machine-Learning\nSimulation, and Multiobjective Optimization

Green corrosion inhibitors have been widely used as sustainable\nreplacements for synthetic organic inhibitors. The application of\nadiantum capillus-veneris (ACV) extract to mitigate mild steel corrosion\nin a hydrochloric acid solution was the main focus of this investigation.\nCorrosion inhibition was studied using electrochemical impedance spectroscopy\n(EIS) and polarization techniques. EIS curves were modeled using a\nshallow neural network. Subsequently, a multiobjective genetic algorithm\nwas employed to identify the optimal combination of concentration\nand time, represented by a Pareto front. EIS revealed an inhibitory\nefficacy of 88% at the optimal concentration of 800 ppm. Polarization\nresults showed that ACV acted as a mixed inhibitor, and at 800 ppm,\nthe corrosion current density decreased from 105 to 44 μA/cm<sup>2</sup>. Surface analytical techniques confirmed the corrosion-inhibitory\neffect of ACV. Results indicated that the sample selected from the\nlower lobe of the Pareto front, dominated by impedance magnitude,\noutperformed other tested samples. Furthermore, the machine learning-based\ncorrosion prediction model demonstrated a high accuracy. This work\nhighlighted the viability of machine learning in assessing corrosion\nresistance and improved the generalization capacity of optimizing\ncorrosion inhibitors.