This work reports the anti-corrosion behavior of liquid smoke from rice husk ash to unveil the contribution of its active compounds in 1 M HCl solution. In this study, the developed methodology to test, analyze, and model the novel type of green corrosion inhibitor for C1018 was characterized using Electrochemical impedance spectroscopy (EIS), Potentiodynamic polarization, and deep learning methods. The inhibitor structure was characterized by Fourier transform infrared analysis (FTIR) and Ultraviolet–visible spectroscopy (UV-Vis). The surface characterization of mild steel immersed in blank and 80 ppm solution inhibitor was performed using Atomic force microscopy (AFM) analysis. The corrosion test results show that the inhibitor is considered a mixed-type inhibitor to achieve the optimum inhibition of 80 ppm at 323 K, reaching up to 99% inhibition efficiency. The AFM results show a smoother surface given a lower skewness parameter at −0.5190 nm on the treated mild steel. The artificial neural network demonstrates the lower overfitting on the inhibited steel, a higher accuracy prediction of 81.08%, and a lower loss rate of 0.6001 to model the relationship between the EIS and Potentiodynamic polarization and the evolution of the passive layer on the treated mild steel. The experiment agrees well with the prediction result to model the adsorbed inhibitor. The work can be used as a guideline to pave the way for subsequent applicability in developing green corrosion inhibitors based on experimental and artificial intelligence approaches.
- deep learning corrosion inhibition
- green corrosion inhibitor
- liquid smoke inhibitor