Rational design and concurrent optimization of antibacterial and anticorrosion bifunctionality for amino acid-surfactant pairs via a machine learning strategy

2D-EI enables concurrent optimization of antibacterial and anticorrosion performance without pre-set weights. • Bridging data-driven approaches with theoretical calculations reveals the key mechanistic drivers of synergistic effects. • Weak molecular interactions may favor the adsorption configurations of inhibitors on metal surfaces. • The optimized pairs with balanced performance exhibit stronger inhibition against microbial corrosion. Developing advanced antibacterial and anticorrosion bifunctional inhibitor formulations represents a promising approach to address microbial corrosion. Up to now, achieving this goal has been hindered by the following challenges: i) the gigantic search space of the optional components; ii) the inherent trade-off between antibacterial and anticorrosion bifunctionality; and iii) the ambiguous mechanism of achieving ideal formulations. Herein, we developed a data-driven framework that combines bi-objective global optimization with interpretable molecular interaction modules. This approach efficiently explored the vast formulation-performance landscape without pre-set optimization weights, enabling the discovery of the entire Pareto-optimal solutions of amino acid-surfactant pairs. Among these, an optimal formulation with balanced antibacterial and anticorrosion behavior showed the best performance in mitigating the microbial corrosion. Meanwhile, the crucial molecular interaction factors that determined the antibacterial and anticorrosion synergy of inhibitor pairs were found among the various possibilities. Notably, we demonstrated that relatively weak molecular interactions between amino acids and surfactants were paradoxically more favorable for the formation of ideal adsorption configurations on the metal surface. This work delivers a promising corrosion inhibitor discovery paradigm for addressing microbial corrosion, while offering a versatile closed-loop route for the rational design of multifunctional drug combinations.