Atomistic simulations of matter, especially those that leverage first-principles (ab initio) electronic structure theory, provide a microscopic view of the world, underpinning much of our understanding of chemistry and materials science. Over the last decade or so, machine-learned force fields have transformed atomistic modeling by enabling simulations of ab initio quality over unprecedented time and length scales. However, early machine-learning (ML) force fields have largely been limited by (i) the substantial computational and human effort required to develop and validate potentials for each particular system of interest and (ii) a general lack of transferability from one chemical system to the next. Here, we show that it is possible to create a general-purpose atomistic ML model, trained on a public dataset of moderate size, that is capable of running stable molecular dynamics for a wide range of molecules and materials. We demonstrate the power of the MACE-MP-0 model—and its qualitative and at times quantitative accuracy—on a diverse set of problems in the physical sciences, including properties of solids, liquids, gases, chemical reactions, interfaces, and even the dynamics of a small protein. The model can be applied out of the box as a starting or “foundation” model for any atomistic system of interest and, when desired, can be fine-tuned on just a handful of application-specific data points to reach ab initio accuracy. Establishing that a stable force-field model can cover almost all materials changes atomistic modeling in a fundamental way: experienced users obtain reliable results much faster, and beginners face a lower barrier to entry. Foundation models thus represent a step toward democratizing the revolution in atomic-scale modeling that has been brought about by ML force fields.

Calculating sublimation enthalpies of molecular crystal polymorphs is relevant to a wide range of technological applications. However, predicting these quantities at first-principles accuracy – even with the aid of machine learning potentials – is a challenge that requires sub-kJ mol⁻¹ accuracy in the potential energy surface and finite-temperature sampling. We present an accurate and data-efficient protocol for training machine learning interatomic potentials by fine-tuning the foundational MACE-MP-0 model and showcase its capabilities on sublimation enthalpies and physical properties of ice polymorphs. Our approach requires only a few tens of training structures to achieve sub-kJ mol⁻¹ accuracy in the sublimation enthalpies and sub-1% error in densities at finite temperature and pressure. Exploiting this data efficiency, we perform preliminary NPT simulations of hexagonal ice at the random phase approximation level and demonstrate a good agreement with experiments. Our results show promise for finite-temperature modelling of molecular crystals with the accuracy of correlated electronic structure theory methods.

Water's ability to autoionize into hydroxide and hydronium ions profoundly influences surface properties, rendering interfaces either basic or acidic. While it is well-established that protons show an affinity to the air-water interface, a critical knowledge gap exists in technologically relevant surfaces like the graphene-water interface. Here we use machine learning-based simulations with first-principles accuracy to unravel the behavior of hydroxide and hydronium ions at the graphene-water interface. Our findings reveal that protons accumulate at the graphene-water interface, with the hydronium ion predominantly residing in the first contact layer of water. In contrast, the hydroxide ion exhibits a bimodal distribution, found both near the surface and further away from it. Analysis of the underlying electronic structure reveals local polarization effects, resulting in counterintuitive charge rearrangement. Proton propensity to the graphene-water interface challenges the interpretation of surface experiments and is expected to have far-reaching consequences for ion conductivity, interfacial reactivity, and proton-mediated processes.

The Bernal-Fowler ice rules stipulate that each water molecule in an ice crystal should form four hydrogen bonds. However, in extreme or constrained conditions, the arrangement of water molecules deviates from conventional ice rules, resulting in properties significantly different from bulk water. In this study, we employ machine learning-driven first-principles simulations to identify a new stabilization mechanism in nanoconfined ice phases. Instead of forming four hydrogen bonds, nanoconfined crystalline ice can form a quasi-one-dimensional hydrogen-bonded structure that exhibits only two hydrogen bonds per water molecule. These structures consist of strongly hydrogen-bonded linear chains of water molecules that zig-zag along one dimension, stabilized by van der Waals interactions that stack these chains along the other dimension. The unusual interplay of hydrogen bonding and van der Waals interactions in nanoconfined ice results in atypical proton behavior such as potential ferroelectric behavior, low dielectric response, and long-range proton dynamics.

Boron, nitrogen, and carbon are neighbors in the periodic table and can form strikingly similar twin structures─hexagonal boron nitride (hBN) and graphene─yet nanofluidic experiments demonstrate drastically different water friction on them. We investigate this discrepancy by probing the interfacial water and atomic-scale properties of hBN using surface-specific vibrational spectroscopy, atomic-resolution atomic force microscopy (AFM), and machine learning-based molecular dynamics. Spectroscopy reveals that pristine hBN acquires significant negative charges upon contacting water at neutral pH, unlike hydrophobic graphene, leading to interfacial water alignment and stronger hydrogen bonding. AFM supports that this charging is not defect-induced. pH-dependent measurements suggest OH^- chemisorption and physisorption, which simulations validate as two nearly equally stable states undergoing dynamic exchange. These findings challenge the notion of hBN as chemically inert and hydrophobic, revealing its spontaneous surface charging and Janus nature, and providing molecular insights into its higher water friction compared to carbon surfaces.

We use machine-learning molecular dynamics with first-principles accuracy to probe water at graphene surfaces across slit widths ranging from the open-interface limit to angstrom-scale confinement.