i Blazquez, Pere BancellsPedron, Federico NicolásTorres, Anthoni Alcarazde Moraes, Elizane EfigeniaCole, IvanMartins, Ernane de Freitas2026-06-112026-06-11ORCID:/0000-0001-6582-1457/work/217149640https://hdl.handle.net/1885/733810549The interaction between water and metallic interfaces is crucial in many fields, and accurate modeling requires good parametrizations using reference data. In classical molecular dynamics (MD), an important part of this interaction is described using the Lennard–Jones (LJ) potential. However, previously reported LJ parameters are not always optimal for capturing the metal/water interactions observed in ab initio descriptions such as density functional theory (DFT). Therefore, well-tailored LJ parameters are necessary to improve the description of water structuring metals in classical MD. The usual route for obtaining LJ parameters involves energetic analysis, where the energies of various structures are obtained via DFT calculations and then matched with the energies obtained using the LJ potentials by varying the sigma/epsilon parameters. Here, we show a different approach to fit LJ parameters for metal/water interactions, based on structural analysis. We report several classical MD simulations for gold/water, varying the sigma/epsilon parameters, comparing the resulting water structuring with that obtained using DFT. Additionally, we test these parameters in quantum mechanics/molecular mechanics (QM/MM) MD simulations, where electrostatic interactions are enabled. Our results demonstrate that the proposed approach can improve the LJ parameters reported in the literature and potentially develop parameters for more complex systems where the water structure above metallic surfaces plays a significant role. Finally, within this proposed approach, the water density profile obtained in hybrid QM/MM calculations, where water is treated as MM at a substantially reduced cost, closely matches the description it would have if treated as QM.Work supported by the Spanish MCIN/AEI/10.13039/501100011033 and by ERDF A way of making Europe through Grant PID2022-139776NB-C62. This project has received funding from EuroHPC JU-EU under the “MAX” project (Grant No. 101093374), and from Grant PCI2022-134972-2 funded by the Spanish MCIN/AEI/10.13039/501100011033 and the European Union NextGenerationEU/PRTR. In addition, this work was supported by the EU’s H2020 framework programme for research and innovation under grant agreements NFFA-Europe Pilot Transnational Access Facility (Grant No. 101007417), having benefited from the access provided by ICN2 in Barcelona within the framework of the NFFA proposals [ID801, ID548, and ID900]. The authors thankfully acknowledge the computer resources at MareNostrum and the technical support provided by Barcelona Supercomputing Center (QHS-2021-3-0022, QHS-2022-1-0030, QHS-2022-2-0035, QHS-2022-3-0030, QHS-2024-2-0035, and QHS-2024-3-0022). The ICN2 is supported by the Severo Ochoa Centres of Excellence programme, Grant No. CEX2021-001214-S, funded by MCIN/AEI/10.13039.501100011033.enPublisher Copyright: © 2026 by the authors.classical MDDFTfittinginterfacesLennard–Jonesmolecular modelingQM/MM202610.3390/nano16030160105029873818