Quantum-Accelerated Supercomputing Atomistic Simulations for Corrosion Inhibition

This work demonstrated a proof-of-concept use case for the emerging quantum-centric supercomputing approaches combining HPC resources with quantum computers. We presented a systematic implementation of hybrid quantum-classical computational method for accelerating atomistic simulations studying corrosion inhibition. We used aluminium surface with screened inhibitor molecule on top. We combined density functional theory (DFT) with quantum algorithm through an active space embedding scheme. We chosen inhibitor molecules of 1,2,4-Triazole and 1,2,4-Triazole-3-thiol. Our implementation leveraged the ADAPT-VQE algorithm with benchmarking against classical DFT calculations, achieving binding energies of -0.386 eV and -1.279 eV for 1,2,4-Triazole and 1,2,4-Triazole-3-thiol, respectively. The binding energy of the thiol derivative aligned with experimental observations regarding sulfur-functionalized inhibitors' which could improve corrosion protection. The methodology employed the orb-d3-v2 machine learning potential for rapid geometry optimizations, followed by accurate DFT calculations using CP2K with PBE functional and Grimme's D3 dispersion corrections. CP2K is a robust DFT package that can scale on HPC classical resources of CPUs and GPUs efficiently. We also benchmarked against smaller systems and revealed that StatefulAdaptVQE implementation achieves a 5-6x computational speedup while maintaining accuracy. This work contributes to the literature studying quantum-accelerated materials science applied to periodic systems, demonstrating the viability of hybrid quantum-classical approaches for studying surface-adsorbate interactions in corrosion inhibition applications. In which, can be transferable to other applications such as carbon capture on metal oxide frameworks and solid-state battery materials studies.