Materials with negative thermal expansion (NTE) are essential for applications requiring precise control of thermal expansion. Owing to their exceptional chemical tunability, flexible architectures, and low-energy lattice vibrations, metal-organic frameworks (MOFs) represent a rich platform for exploring NTE. However, uncovering the structural motifs that govern NTE across the enormous MOF design space remains experimentally challenging, and large-scale first-principles phonon calculations are computationally prohibitive. Here, we comprehensively evaluate the factors influencing NTE in MOFs by utilizing a high-throughput workflow based on MACE-MP-MOF0, a machine learning interatomic potential fine-tuned for MOFs with near-ab initio accuracy, to construct PhononMOFdb, a database of phonons, inelastic neutron scattering spectra, bulk moduli, and heat capacities for over 12,000 MOFs. High-throughput screening of this database reveals that highly porous cubic topology frameworks with heavier, lower-valent metal nodes favor strong NTE, while linker functionalization provides a practical handle for tuning NTE magnitude and sign without compromising mechanical stability. Experimental validation via high-resolution temperature-dependent synchrotron powder X-ray diffraction on the Ce-UiO-66 MOF and its brominated variants confirms the design recipe and yields volumetric NTE coefficients surpassing current records. This work establishes a data-driven strategy for engineering NTE in MOFs, showing how machine learning-accelerated discovery and targeted experimental validation together unlock predictive materials design.

Silicon-based anodes promise an increase in energy density for Li-ion batteries, yet they suffer from a poor calendar life. Researchers have posited that fluorinated lithium salt forms reactive side products that destroy the solid electrolyte interphase (SEI), even without cycling. HF is one such reactive side product formed from trace water contamination in the electrolyte. Some electrolytes, such as localized high concentration electrolytes (LHCEs) may improve cell stability in highly reactive systems, such as Li metal and Si. In the present study, LHCEs containing 200-300 ppm of water retained up to 8% greater capacity (1200-1300 mAh/g Si) in calendar life tests over 200 days compared to dried electrolytes (< 20 ppm water). Calendar aging took place at 100% state of charge. Cells with 200-300 ppm water performed comparably to cells with 20 ppm water in cycle life tests (900-1000 mAh/g Si) . Even adding 1000 ppm water did not lead to rapid capacity fade in cells undergoing cycle life tests. Nano-FTIR spectroscopy revealed chemical and structural differences in the SEI for cells with 1000 ppm water compared to 200-300 ppm water. The SEI differences, including increased Li2O concentration, may have contributed to improved calendar life. This research reveals the capabilities of LHCEs to improve the calendar and cycle life of Si-based Li-ion batteries, despite the presence of a highly reactive contaminant.

We perform the first high-throughput search and evaluation of materials that can serve as excellent low-mass dark matter detectors. Using properties of close to 1000 materials from the Materials Project database, we project the sensitivity in dark matter parameter space for experiments constructed from each material, including both absorption and scattering processes between dark matter and electrons. Using the anisotropic materials in the dataset, we further compute the level of daily modulation in interaction rates and the resulting directional sensitivities, highlighting materials with prospects to detect the dark matter wind. Our methods provide the basic tools for the data-driven design of dark matter detectors, and our findings lay the groundwork for the next generation of highly optimized direct searches for dark matter as light as the keV scale. This represents a major step in the application of results from condensed matter physics to dark matter search design.

The materials science community seeks to support the FAIR principles for computational simulation research. The MatCore Project was recently launched to address this need, with the goal of developing an overall metadata framework and accompanying guidelines. This paper reports on the MatCore goals and overall progress. Historical background context is provided, including a review of the principles underlying successful core metadata standards. The paper also presents selected MatCore examples and discusses future plans.

Generative models and machine learning promise accelerated material discovery in MOFs for CO2 capture and water harvesting but face significant challenges navigating vast chemical spaces while ensuring synthetizability. Here, we present MOFGen, a system of Agentic AI comprising interconnected agents: a large language model that proposes novel MOF compositions, a diffusion model that generates crystal structures, quantum mechanical agents that optimize and filter candidates, and synthetic-feasibility agents guided by expert rules and machine learning. Trained on all experimentally reported MOFs and computational databases, MOFGen generated hundreds of thousands of novel MOF structures and synthesizable organic linkers. Our methodology was validated through high-throughput experiments and the successful synthesis of five "AI-dreamt" MOFs, representing a major step toward automated synthesizable material discovery.

NMR crystallography has emerged as a promising technique for the determination and refinement of atomic coordinates in crystal structures. The crystal structure of compounds containing quadrupolar nuclei, such as 27Al, can be improved by directly comparing solid-state NMR measurements to DFT computations of the electric field gradient (EFG) tensor. The non-negligible computational cost of these first-principles calculations limits the applicability of this method to all but the most well-defined structures. We developed a fast, low-cost machine learning model to predict EFG parameters based on local structural motifs and elemental parameters. We computed 8081 EFG tensors from 1681 27Al crystalline solids using DFT and benchmarked them against 105 experimentally measured 27Al sites. Surprisingly, simple local geometric features dominate the predictive performance of the resulting random-forest model, yielding an R2 value of 0.98 and an RMSE of 0.61 MHz for CQ, the quadrupolar coupling constant. This model accuracy should enable pre-refining future structural assignments before finally validating with first-principles calculations. Such a catalogue of 27Al NMR tensors can serve as a tool for researchers assigning complex NMR spectra influenced by the nuclear electric quadrupole interaction.

Cycloolefin resins (CORs) comprising dicyclopentadiene (DCPD) cross-linkers are high performance thermosets for diverse single-use applications. If it were possible to carry out deconstruction of DCPD thermosets through exclusive reformation of the cyclopentene ring in DCPD, then linear polyDCPD chains could re-enter subsequent manufacturing cycles and enable resin-to-resin circularity. Here, we develop a chemical recycling process whereby linear polyDCPD recyclates are generated from end-of-life commercial and model CORs, including copolymers, using a second-generation Hoveyda-Grubbs ruthenium(II) alkylidene catalyst for deconstruction via ring-closing metathesis. The properties of first-generation DCPD thermosets were reproduced across subsequent generations of recycling and reuse, where up to 84% of linear DCPD was retrievable from end-of-life thermosets after deconstruction. We further quantify how comonomer incorporation, the curing process, and aging affect round-trip material efficiency and the quality of regenerated thermosets.

Correction for “Atomate2: modular workflows for materials science” by Alex M. Ganose et al., Digital Discovery, 2025, 4, 1944–1973, https://doi.org/10.1039/D5DD00019J.

Machine learning interatomic potentials (MLIPs) have introduced a new paradigm for atomic simulations. Recent advancements have led to universal MLIPs (uMLIPs) that are pre-trained on diverse datasets, providing opportunities for universal force fields and foundational machine learning models. However, their performance in extrapolating to out-of-distribution complex atomic environments remains unclear. In this study, we highlight a consistent potential energy surface (PES) softening effect in three uMLIPs: M3GNet, CHGNet, and MACE-MP-0, which is characterized by energy and force underprediction in atomic-modeling benchmarks including surfaces, defects, solid-solution energetics, ion migration barriers, phonon vibration modes, and general high-energy states. The PES softening behavior originates primarily from the systematically underpredicted PES curvature, which derives from the biased sampling of near-equilibrium atomic arrangements in uMLIP pre-training datasets. Our findings suggest that a considerable fraction of uMLIP errors are highly systematic, and can therefore be efficiently corrected. We argue for the importance of a comprehensive materials dataset with improved PES sampling for next-generation foundational MLIPs.

The rapid development of foundation potentials (FPs) in machine learning interatomic potentials demonstrates the possibility for generalizable learning of the universal potential energy surface. The accuracy of FPs can be further improved by bridging the model from lower-fidelity datasets to high-fidelity ones. In this work, we analyze the challenge of this transfer learning (TL) problem within the CHGNet framework. We show that significant energy scale shifts and poor correlations between GGA and r2SCAN hinder cross-functional transferability. By benchmarking different TL approaches on the MP-r2SCAN dataset, we demonstrate the importance of elemental energy referencing in the TL of FPs. By comparing the scaling law with and without the pre-training on a low-fidelity dataset, we show that significant data efficiency can still be achieved through TL, even with a target dataset of sub-million structures. We highlight the importance of proper TL and multi-fidelity learning in creating next-generation FPs on high-fidelity data.

LiNi0.5Mn1.5O4 (LNMO) is a high-voltage spinel cathode with low nickel content, making it an attractive candidate for next-generation lithium-ion batteries (LIBs). However, its application is limited by interfacial instability with conventional carbonate-based electrolytes at high voltages. In this work, a localized saturated electrolyte (LSE) capable of stably operating up to 4.85 V is investigated. Molecular dynamics simulations and Fourier transform infrared spectroscopy reveal that adding "non-solvating" 1,1,2,2-tetrafluoroethyl-2,2,3,3-tetrafluoropropyl ether diluent in the saturated electrolyte, more PF6- anions are present in the first solvation shell of Li+, at the expense of solvent molecules. This tailored solvation environment promotes the formation of a robust, LiF-rich cathode-electrolyte interphase that mitigates transition metal dissolution and parasitic side reactions. The optimized LSE enables excellent cycling performance, with 95% capacity retention in Li|LNMO half-cells after 100 cycles and 94% retention in Li4Ti5O12|LNMO full cells after 250 cycles, even at a practically relevant LNMO cathode loading of approximate to 15 mg cm-2. These results highlight the benefits of electrolyte engineering and solvation structure control in advancing high-voltage LIB technologies.

Hydrolysis is a fundamental family of chemical reactions where water facilitates the cleavage of bonds. The process is ubiquitous in biological and chemical systems, owing to water's remarkable versatility as a solvent. However, accurately predicting the feasibility of hydrolysis through computational techniques is a difficult task, as subtle changes in reactant structure like heteroatom substitutions or neighboring functional groups can influence the reaction outcome. Furthermore, hydrolysis is sensitive to the pH of the aqueous medium, and the same reaction can have different reaction properties at different pH conditions. In this work, we have combined reaction templates and high-throughput ab initio calculations to construct a diverse data set of hydrolysis free energies. The developed framework automatically identifies reaction centers, generates hydrolysis products, and utilizes a trained graph neural network (GNN) model to predict ΔG values for all potential hydrolysis reactions in a given molecule. The long-term goal of the work is to develop a data-driven, computational tool for high-throughput screening of pH-specific hydrolytic stability and the rapid prediction of reaction products, which can then be applied in a wide array of applications including chemical recycling of polymers and ion-conducting membranes for clean energy generation and storage.

Kinetically trapping the high-temperature states through rapid cooling solidification is widely used for the synthesis of high-entropy alloys (HEAs), especially those with intrinsically immiscible elemental combinations1-4. However, strategies need to be developed to overcome the fundamental limitations of rapid cooling solidification in controlling the crystallinity, structure and morphology of HEAs. Here we introduce an isothermal solidification strategy for the synthesis of HEAs by rapidly altering the metal alloy composition through liquid-liquid interface reactions at low temperatures, for example, from 25 °C to 80 °C. We use gallium (Ga)-based metal as the sacrificial reagent and mixing medium. By directing the reactions to the interfaces between the Ga-based liquid metal and an aqueous metal ion solution, the foreign metal ions can be reduced at the interfaces and incorporated into the liquid metal quickly. HEAs with various crystallinity (single crystal, mesocrystal, polycrystal and amorphous), morphology (zero, two and three dimensions) and compositions can be achieved through the isothermal solidification. Ga can be completely consumed, resulting in Ga-free HEAs. If desired, Ga can be one of the metal elements in the final products. In situ liquid phase transmission electron microscopy (TEM) studies and theoretical analysis show the isothermal solidification mechanisms. Our direct observations show the enhanced mixing of liquid metal elements and the solidification process with fluctuating nucleation dynamics. The isothermal solidification marks a powerful strategy for HEA synthesis through an unexplored pathway of kinetically trapping the high-entropy states.

Understanding and facilitating pure magnesium nucleation/growth electrodeposition behavior with ultrahigh Coulombic efficiency is complicated by the phenomenon of solid electrolyte interphase (SEI) formation in state-of-the-art, halogen-free magnesium electrolytes. Defining the electrolyte properties necessary to achieve ideal electrodeposition/stripping (E/S) thus remains elusive. Here, we reveal for the first time, rapid magnesium electrodeposition behavior that forms densely aligned, micron-sized thin platelets by establishing a dynamic bare magnesium/electrolyte interface during high-rate net plating. This effectively “SEI-free” interface allows facile magnesium diffusion and migration in stripping with near-unity E/S efficiency under demanding conditions over long-term cycling. The intrinsic electrolyte stability of the salt/solvent at the molecular level is the key to forming such an interface. The efficacy of the dynamic bare interface and an electrodeposited, free-standing magnesium anode is demonstrated in a high-areal-capacity full cell. These findings provide new design principles and fundamental understanding of interfacial chemistry in multivalent metal batteries.

Tetragonal ZnGa 2 Te 4 thin films with direct bandgap ∼1.86 eV show −2 mA cm −2 photocurrent, highlighting their potential as efficient photocathodes for photoelectrochemical CO 2 reduction.

Accurate potential energy surface (PES) descriptions are essential for atomistic simulations of materials. Universal machine learning interatomic potentials (UMLIPs)^1-3 offer a computationally efficient alternative to density functional theory (DFT)^4 for PES modeling across the periodic table. However, their accuracy today is fundamentally constrained due to a reliance on DFT relaxation data.^5,6 Here, we introduce MatPES, a foundational PES dataset comprising ∼ 400,000 structures carefully sampled from 281 million molecular dynamics snapshots that span 16 billion atomic environments. We demonstrate that UMLIPs trained on the modestly sized MatPES dataset can rival, or even outperform, prior models trained on much larger datasets across a broad range of equilibrium, near-equilibrium, and molecular dynamics property benchmarks. We also introduce the first high-fidelity PES dataset based on the revised regularized strongly constrained and appropriately normed (r^2SCAN) functional^7 with greatly improved descriptions of interatomic bonding. The open source MatPES initiative emphasizes the importance of data quality over quantity in materials science and enables broad community-driven advancements toward more reliable, generalizable, and efficient UMLIPs for large-scale materials discovery and design.

High-throughput density functional theory (DFT) calculations have become a vital element of computational materials science, enabling materials screening, property database generation, and training of "universal" machine learning models. While several software frameworks have emerged to support these computational efforts, new developments such as machine learned force fields have increased demands for more flexible and programmable workflow solutions. This manuscript introduces atomate2, a comprehensive evolution of our original atomate framework, designed to address existing limitations in computational materials research infrastructure. Key features include the support for multiple electronic structure packages and interoperability between them, along with generalizable workflows that can be written in an abstract form irrespective of the DFT package or machine learning force field used within them. Our hope is that atomate2's improved usability and extensibility can reduce technical barriers for high-throughput research workflows and facilitate the rapid adoption of emerging methods in computational material science.

We present MP-ALOE, a dataset of nearly 1 million DFT calculations using the accurate r2SCAN meta-generalized gradient approximation. Covering 89 elements, MP-ALOE was created using active learning and primarily consists of off-equilibrium structures. We benchmark a machine learning interatomic potential trained on MP-ALOE, and evaluate its performance on a series of benchmarks, including predicting the thermochemical properties of equilibrium structures; predicting forces of far-from-equilibrium structures; maintaining physical soundness under static extreme deformations; and molecular dynamic stability under extreme temperatures and pressures. MP-ALOE shows strong performance on all of these benchmarks, and is made public for the broader community to utilize.

Phase stability predictions are central to computational materials discovery efforts and have been made possible by large databases of computed properties from high-throughput density functional theory (DFT) calculations. Such databases now contain millions of calculations at the generalized gradient approximation (GGA) level of theory, representing an enormous investment of computational resources. Although it is now feasible to carry out large numbers of calculations using more accurate methods, such as meta-GGA functionals, recomputing the entirety of a database with a higher-fidelity method is impractical and would not effectively leverage the value embodied in existing calculations. Instead, we propose in this work a general procedure by which higher-fidelity, low-coverage calculations (e.g., meta-GGA calculations for selected chemical systems) can be combined with lower-fidelity, high-coverage calculations (e.g., an existing database of GGA calculations) in a robust and scalable manner to yield improved phase stability predictions. We demonstrate our scheme using legacy GGA(+\textit{U}) calculations and new r$^2$SCAN meta-GGA calculations from the Materials Project and illustrate its application to solid and aqueous phase stability. We discuss practical considerations for constructing mixed phase diagrams and present guidelines for prioritizing high-fidelity calculations for maximum benefit.