(225t) Computer-Aided Molecular Design to Identify Corrosion Inhibitors for STEEL Reinforced Concrete

Current corrosion inhibitors used in concrete have their drawbacks as they can be toxic or cause side reactions side reactions leading to concrete structure cracking. Accordingly, there is a need to identify new candidates to be used in concrete structures that are effective, environmentally friendly, and cost-effective. To do so, a computer aided molecular design (CAMD) using atomic Signatures as molecular descriptors can be used to design new corrosion inhibitors for steel reinforce concrete. A Signature, in simple terms, is a codification system that describes the connectivity of an atom to its neighbor in a molecule to an extent of bonding called “height”. In this study, a dataset was constructed using 15 corrosion inhibitors consisting of amines and alkanol amines tested on carbon steel rebars in 0.6 M Cl^‒ simulated concrete pore solution. From this dataset, 18 unique height‒1 atomic Signatures were generated and correlated to the corrosion current density (i_corr) using a forward stepping multilinear regression with leave one out cross validation. The model was able to achieve a correlation coefficient (r²) and predictability (q²) of 0.91 and 0.67, respectively. Accordingly, the CAMD was initiated by seeking all possible combinations of the 18 unique height‒1 atomic Signatures using a brute force method, to generate new molecular structures. After 119,259 iterations, 3386 solutions were found of which 164 has an i_corr value between 0.6 and 0.05 µAcm^‒2. To validate the outcome of the CAMD, morpholine (one of the solutions that met the i_corr criteria) was purchased to identify its experimental i_corr utilizing electrochemical testing. The experimental i_corr value was 0.75 µAcm^‒2, which closely matched the predicted value of 0.599 µAcm^‒2 ‒ validating the model.