Quantum computing for corrosion simulation: workflow and resource analysis

npj | quantum information Article Published in partnership with The University of New South Wales https://doi.org/10.1038/s41534-025-01171-1 Quantum computing for corrosion simulation: workflow and resource analysis Check for updates 1 2 1 1 3 1234567890():,; 1234567890():,; Nam Nguyen , Thomas W. Watts , Benjamin Link , Kristen S. Williams , Yuval R. Sanders , Samuel J. Elman3 , Maria Kieferova3, Michael J. Bremner3,4, Kaitlyn J. Morrell5, Justin Elenewski5, Eric B. Isaacs2, Samuel D. Johnson2, Luke Mathieson3, Kevin M. Obenland5, Matthew Otten6 , Rashmi Sundareswara2 & Adam Holmes2 Corrosion is a pervasive issue that impacts the structural integrity and performance of materials across various industries, imposing a significant economic impact globally. In fields like aerospace and defense, developing corrosion-resistant materials is critical, but progress is often hindered by the complexities of material-environment interactions. While computational methods have advanced in designing corrosion inhibitors and corrosion-resistant materials, they fall short in understanding the fundamental corrosion mechanisms due to the highly correlated nature of the systems involved. This paper explores the potential of leveraging quantum computing to accelerate the design of corrosion inhibitors and corrosion-resistant materials, with a particular focus on magnesium and niobium alloys. We investigate the quantum computing resources required for high-fidelity electronic ground-state energy estimation (GSEE), which will be used in our hybrid classical-quantum workflow. Representative computational models for magnesium and niobium alloys show that 2292 to 38598 logical qubits and (1.04 to 1962) × 1013 T-gates are required for simulating the ground-state energy of these systems under the first quantization encoding using plane waves basis. Corrosion is a natural process that degrades materials through chemical or electrochemical reactions with their environment, impacting both the structural integrity and performance of engineering materials1. It imposes a significant economic burden, with global costs exceeding $2.5 trillion annually2. In aerospace and defense, corrosion-resistant materials reduce maintenance costs and enhance sustainability3. The development of corrosion-resistant materials relies on a combination of experimental and computational methods. Computational models can aid in the designing of corrosion inhibitors and the development of corrosion-resistant materials,4,5 with experiments validating these materials6. However, no single mathematical framework currently exists that can fully capture the complexity of the physical and chemical processes driving corrosion7,8. As a result, different computational models need to be investigated for various corrosion applications. This work introduces a complete workflow from first principles for simulating corrosion processes. The novel quantum-classical hybrid approach unifies previously unrelated classical and quantum algorithms. We present computational models for aqueous corrosion in magnesium and magnesium-aluminum alloy surfaces, which may be extendable to other metallic systems; and for high-temperature oxidation in niobium alloys, which are critical for extreme environments but highly susceptible to degradation. Our models for magnesium corrosion are based on nudged elastic band modeling, while our niobium algorithm involves a coupledcluster approximation to map the problem to a classical Ising model. In both cases, ground-state energy estimation is a key computational bottleneck. Traditional modeling techniques, like finite element analysis (FEA), are poorly suited to corrosion due to the highly localized, non-uniform environments in which it occurs9. Corrosion processes often span multiple length and time scales and can evolve over time, transitioning from localized pitting to more catastrophic modes such as stress corrosion cracking or corrosion fatigue8. Factors such as material properties, corrosive environments, and 1 Applied Mathematics, Boeing Research & Technology, Huntsville, USA. 2HRL Laboratories LLC, Malibu, CA, USA. 3Centre for Quantum Software and Information, School of Computer Science, Faculty of Engineering & Information Technology, University of Technology Sydney, Ultimo, Australia. 4Centre for Quantum Computation and Communication Technology, University of Technology Sydney, Ultimo, Australia. 5MIT Lincoln Laboratory, Lexington, MA, USA. 6Department of e-mail: ; ; ; Physics, University of Wisconsin - Madison, Madison, WI, USA. npj Quantum Information | (2026)12:27 1 Article https://doi.org/10.1038/s41534-025-01171-1 component assembly further complicate these interactions10,11. The lack of an integrated framework that combines chemistry, microstructure, and mechanical behavior into a single model necessitates significant simplifying assumptions12. At the atomic scale, the chemical reactions driving corrosion are governed by the electronic structure of materials, which can be described by the time-independent Schrödinger equation. Accurate solutions to the Schrödinger equation are crucial for predicting properties such as corrosion rates but are computationally prohibitive due to the superpolynomial scaling of the wavefunction size with the number of orbitals (N) and electrons (Ne)13. Many of these processes involve highly correlated electronic states,8 making exact diagonalization infeasible when N and Ne exceed 2514. For example, Fig. 1 shows a magnesium alloy model that incorporates 1000 water molecules and eight atomic layers, highlighting the scale and complexity of these systems. Simulating such a model with high accuracy is far beyond the capabilities of classical methods. Quantum computers offer a promising path forward, as they are uniquely suited for simulating strongly correlated chemical systems15,16. Recent advances in quantum algorithms for simulating chemical models,17,18 along with new quantum software libraries,19 make quantum computing increasingly practical. While quantum computers are not expected to completely negate high-performance computer costs, it will potentially enable high-fidelity computations on larger atomistic models and more accurate simulation of corrosion processes that are highly correlated. We conduct a comprehensive resource assessment for quantum modeling of aqueous and high-temperature corrosion processes, marking the first application of the pyLIQTR computational tool to a real-world materials science problem. Our work develops and implements a hybrid quantumclassical workflow that incorporates Ground-State Energy Estimation (GSEE) and details the required logical-level quantum hardware resources for high-fidelity simulations of magnesium and niobium alloy corrosion. In particular, we cost utility-scale instances and establish the feasibility of quantum-enhanced approaches in materials degradation studies by estimating the total number of logical qubits a (...truncated)