The atomic simulation environment-a Python library for working with atoms

The Amsterdam Modeling Suite

In this paper, we present the Amsterdam Modeling Suite (AMS), a comprehensive software platform designed to support advanced molecular and materials simulations across a wide range of chemical and physical systems. AMS integrates cutting-edge quantum chemical methods, including Density Functional Theory (DFT) and time-dependent DFT, with molecular mechanics, fluid thermodynamics, machine learning techniques, and more, to enable multi-scale modeling of complex chemical systems. Its design philosophy allows for seamless coupling between components, facilitating simulations that range from small molecules to complex biomolecular and solid-state systems, making it a versatile tool for tackling interdisciplinary challenges, both in industry and in academia. The suite also emphasizes user accessibility, with an intuitive graphical interface, extensive scripting capabilities, and compatibility with high-performance computing environments.

Shadow Molecular Dynamics for a Charge-Potential Equilibration Model

We introduce a shadow molecular dynamics (MD) approach based on the Atom-Condensed Kohn-Sham second-order (ACKS2) charge-potential equilibration model. In contrast to regular flexible charge models, the ACKS2 model includes both flexible atomic charges and potential fluctuation parameters that allow for physically correct charge fragmentation and improved scaling of the polarizability. Our shadow MD scheme is based on an approximation of the ACKS2's flexible charge-potential energy function, in combination with extended Lagrangian Born-Oppenheimer MD. Utilizing this shadow charge-potential equilibration approach mitigates the costly overhead and stability problems associated with finding well-converged iterative solutions to the charges and potential fluctuations of the ACKS2 model in an MD simulation. Our work provides a robust and versatile framework for efficient, high-fidelity MD simulations of diverse physical phenomena and applications.

Defective Carbon Catalysts with Graphitic N-Modified Adjacent Pentagons as Active Sites for Boosted Oxygen Reduction Reaction in Seawater

Seawater electrocatalysis is highly desired for various energy storage and conversion systems, such as water splitting using seawater as an electrolyte and metal fuel cells. However, the adsorption of chloride ions (Cl-) on the active sites of cathodes would worsen the oxygen reduction reaction (ORR) activity and stability, thus lowering the battery performance. Herein, the coupling active sites of graphitic N-regulated adjacent pentagon defects in carbon nanosheets (GAP/CN) were first synthesized by a low-boiling-point metal-mediated partial N-removal strategy. Experimental and theoretical results affirm the advantageous cooperative effect between adjacent pentagons and graphitic N toward the ORR in a harsh seawater environment, where adjacent pentagons act as the authentic highly effective ORR active sites and surrounding graphitic N site serves as the structural regulator to weaken the binding strength of harmful Cl- to prevent catalyst poisoning. As a result, GAP/CN delivers excellent ORR activities in diverse electrolytes, including 0.1 M KOH (half-wave potential of 0.87 V), alkaline artificial seawater (half-wave potential of 0.87 V), and natural seawater (half-wave potential of 0.71 V), and also good long-term stability, which can be comparable to commercial Pt/C. This study provides valuable guidance for the rational design of ORR electrocatalysts for seawater-related energy-conversion devices.

Catalytic Activity of an Ensemble of Sites for CO2 Hydrogenation to Methanol on a ZrO2-on-Cu Inverse Catalyst

The significant increase in CO2 emissions from heavy fossil fuel utilization has raised serious concerns, highlighting the need for effective methods to convert CO2 into value-added chemicals. Here, we report a computational investigation on the catalytic activity of ZrO2-on-Cu inverse catalysts for CO2 hydrogenation to methanol, considering highly dispersed ZrO2 trimers on Cu (111). Such clusters present a large ensemble of formate-containing configurations, Zr3On(OH)m(OCHO)l, making the evaluation of the catalytic activity very challenging. We found that the sites on the various catalyst configurations exhibit markedly different activities for formate hydrogenation, despite their similar free energy and composition. To understand these differences in reactivity, we examined the structural and electronic nature of the low free-energy catalyst configurations and identified that the energy of the lowest unoccupied orbital of the reacting formate, modified by its binding with the catalytic site, is a descriptor for the reaction energy of the formate hydrogenation step. From there, we screened an ensemble of catalyst structures using this descriptor to predict highly active metastable catalyst configurations and computed the reaction pathways and transition states for formate hydrogenation. From this investigation, we distinguished reactive from nonreactive sites and formate species on the ZrO2/Cu inverse catalyst based on structural and electronic features. We showed that rare metastable configurations control the activity. Additionally, an efficient method for examining the reactivity of a large number of coexisting catalyst structures was developed.

Empowering materials science with VASPKIT: a toolkit for enhanced simulation and analysis

Driven by rapid advances in high-performance supercomputing, computational materials science has emerged as a powerful approach for exploring, designing, and predicting material properties at the atomic and molecular scales. Among the various computational tools developed in this field, the Vienna Ab initio Simulation Package (VASP) stands out as a widely adopted and highly versatile platform for performing first-principles density functional theory (DFT) calculations. VASP is widely used to explore electronic structures, phonon behavior, magnetic properties, thermodynamics and catalytic mechanisms across a diverse range of materials systems. Despite its robust capabilities, utilizing VASP requires expertise in setting up simulations and analyzing results, which can be time consuming and technically challenging. To address these barriers, VASPKIT was developed as a comprehensive toolkit to simplify the workflow for VASP users. VASPKIT streamlines both preprocessing and postprocessing tasks, enabling users to generate essential input files based on customizable parameters and automate computational workflows. The postprocessing features of VASPKIT allow for efficient analysis of electronic, mechanical, optical and catalytic properties, thereby substantially reducing the need for advanced programming expertise. This protocol provides a detailed guide to using VASPKIT, including practical examples to demonstrate its versatility and utility in conducting and analyzing DFT calculations. For instance, the computation of elastic constants, electronic band structures and density of states for a graphene system can typically be completed within half an hour, depending on the computational resources available. By offering step-by-step guidance, this protocol aims to further expand the accessibility and impact of VASPKIT in the field of computational materials science.

Generator of Neural Network Potential for Molecular Dynamics: Constructing Robust and Accurate Potentials with Active Learning for Nanosecond-Scale Simulations

Neural network potentials (NNPs) enable large-scale molecular dynamics (MD) simulations of systems containing >10,000 atoms with the accuracy comparable to ab initio methods and play a crucial role in material studies. Although NNPs are valuable for short-duration MD simulations, maintaining the stability of long-duration MD simulations remains challenging due to the uncharted regions of the potential energy surface (PES). Currently, there is no effective methodology to address this issue. To overcome this challenge, we developed an automatic generator of robust and accurate NNPs based on an active learning (AL) framework. This generator provides a fully integrated solution encompassing initial data set creation, NNP training, evaluation, sampling of additional structures, screening, and labeling. Crucially, our approach uses a sampling strategy that focuses on generating unstable structures with short interatomic distances, combined with a screening strategy that efficiently samples these configurations based on interatomic distances and structural features. This approach greatly enhances the MD simulation stability, enabling nanosecond-scale simulations. We evaluated the performance of our NNP generator in terms of its MD simulation stability and physical properties by applying it to liquid propylene glycol (PG) and polyethylene glycol (PEG). The generated NNPs enable stable MD simulations of systems with >10,000 atoms for 20 ns. The predicted physical properties, such as the density and self-diffusion coefficient, show excellent agreement with the experimental values. This work represents a remarkable advance in the generation of robust and accurate NNPs for organic materials, paving the way for long-duration MD simulations of complex systems.

Wiggle150: Benchmarking Density Functionals and Neural Network Potentials on Highly Strained Conformers

Accurate benchmarks are key to assessing the accuracy and robustness of computational methods, yet most available benchmark sets focus on equilibrium geometries, limiting their utility for applications involving nonequilibrium structures such as ab initio molecular dynamics and automated reaction-path exploration. To address this gap, we introduce Wiggle150, a benchmark comprising 150 highly strained conformations of adenosine, benzylpenicillin, and efavirenz. These geometries─generated via metadynamics and scored using DLPNO-CCSD(T)/CBS reference energies─exhibit substantially larger deviations in bond lengths, angles, dihedrals, and relative energies than other conformer benchmarks. We evaluate a diverse array of computational methods, including density-functional theory, composite quantum chemical methods, semiempirical models, neural network potentials, and force fields, on predicting relative energies for this challenging benchmark set. The results highlight multiple methods along the speed-accuracy Pareto frontier and identify AIMNet2 as particularly robust among the NNPs surveyed. We anticipate that Wiggle150 will be used to validate computational protocols involving nonequilibrium systems and guide the development of new density functionals and neural network potentials.

Learning Pairwise Interaction for Extrapolative and Interpretable Machine Learning Interatomic Potentials with Physics-Informed Neural Network

Achieving both robust extrapolation and physical interpretability in machine learning interatomic potentials (ML-IPs) for atomistic simulation remains a significant challenge, particularly in data-scarce areas such as chemical reactions or complex, multicomponent materials at extreme conditions. Here, we present a pairwise-decomposed physics-informed neural network (P2Net) that parametrizes an analytical bond-order potential (BOP) layer to decouple the energy contributions of atomic pairs. By leveraging fundamental physical principles, P2Net demonstrates excellence at extrapolating beyond its training regime and accurately capturing molecular geometries far from equilibrium. The pairwise energy decomposition further empowers the bond analyses for deprotonation and SN2 reactions, which is not easy with most ML-IPs. The atomic pair energy offers how to elucidate the evolution of interatomic interactions as a reaction proceeds. Our methodology highlights enhanced data efficiency in building ML-IPs and facilitates more informative postsimulation analysis, thereby broadening the applicability of ML-IPs to complex and reactive systems.

Modeling electronic absorption spectra with nuclear quantum effects in constrained nuclear-electronic orbital framework

Electronic absorption spectra serve as versatile and powerful tools in experiments. Accurate theoretical simulation of electronic absorption spectra is challenging because multiple factors such as environmental effects and nuclear quantum effects contribute to spectrum lineshapes. This work proposes a protocol to model electronic absorption spectra in the constrained nuclear-electronic orbital framework. Solvent effects, temperature effects, and particularly nuclear quantum effects can be taken into consideration in this unified framework. This protocol is applied to investigate the electronic absorption spectrum of the pyridine molecule in water. Nuclear quantum effects are found to induce a broadening and red shift of the absorption spectrum of pyridine.

Atomic Layer Restructuring of Gold Surfaces by N-Heterocyclic Carbenes over Large Surface Areas

Even highly planar, polished metal surfaces display varying levels of roughness that can affect their optical and electronic properties, impacting performance in state-of-the-art microelectronics. Current methods for smoothing rough metallic surfaces require either the removal or addition of substantial amounts of material using complex processes that are incompatible with 3-dimensional nanoscale features needed for state-of-the-art applications. We present a vapor-phase process that results in up to a 60% smoothing of nanometer-scale rough gold surfaces through a single exposure to a class of ligands called N-heterocyclic carbenes. This process does not require removal or addition of metal from the surface and provides smoothing at the Ångström scale. Smoothing occurs in a single deposition, giving quantifiable differences in the adsorption behavior of the resulting surfaces. The process takes place through an adatom-extraction-driven destabilization and restructuring of the surface in a self-limiting manner. This process is achieved without the use of harsh chemical etchants or mechanical intervention, takes only minutes, and can easily be integrated with vapor-phase processing in situ in microfabrication workflows. Our observations demonstrate atomic layer restructuring, a technique that compliments atomic layer deposition and atomic layer etching in the fabrication and processing of high-precision materials.

Mild-Condition Conversion of Methane to Acetic Acid over MoS2-Confined Rh-Fe Sites

The oriented conversion of CH4 to CH3COOH at low temperature, even room temperature, is both scientifically significant and industrially applicable for CH4 utilization, yet it is extremely challenging due to the difficulties associated with efficient CH4 activation and controllable C-C coupling. In this study, we for the first time achieve the room-temperature conversion of CH4 to CH3COOH using molecular O2 and CO over MoS2-confined Rh-Fe sites, which delivers an unprecedented CH3COOH selectivity of 90.3% and a productivity of 26.2 μmol gcat.-1 h-1 at 25 °C. Furthermore, the productivity of CH3COOH can be enhanced to 105.6 μmol gcat.-1 h-1 at 80 °C, while maintaining a high selectivity of 95.6%. Comprehensive experimental and theoretical investigation reveal the critical role of Rh-Fe synergy in the selective formation of CH3COOH. The confined Fe sites in MoS2 enable the activation of O2 to generate highly reactive Fe═O center for CH4 dissociation to CH3 species at room temperature, which then readily couple with adsorbed CO on adjacent Rh sites to form the key CH3CO intermediate for CH3COOH production. The unique structure of Rh-Fe sites offers synergistic catalytic properties that effectively balance C-H activation and C-C coupling, successfully addressing the trade-off between activity and selectivity in the carbonylation of CH4 to CH3COOH under mild conditions.

Enhancing the prediction of TADF emitter properties using Δ-machine learning: A hybrid semi-empirical and deep tensor neural network approach

This study presents a machine learning (ML)-augmented framework for accurately predicting excited-state properties critical to thermally activated delayed fluorescence (TADF) emitters. By integrating the computational efficiency of semi-empirical PPP+CIS theory with a Δ-ML approach, the model overcomes the inherent limitations of PPP+CIS in predicting key properties, including singlet (S1) and triplet (T1) energies, singlet-triplet gaps (ΔEST), and oscillator strength (f). The model demonstrated exceptional accuracy across datasets of varying sizes and diverse molecular features, notably excelling in predicting oscillator strength and ΔEST values, including negative regions relevant to TADF molecules with inverted S1-T1 gaps. This work highlights the synergy between physics-inspired models and machine learning in accelerating the design of efficient TADF emitters, providing a foundation for future studies on complex systems and advanced functional materials.

DeePMD-GNN: A DeePMD-kit Plugin for External Graph Neural Network Potentials

Machine learning potentials (MLPs) have revolutionized molecular simulation by providing efficient and accurate models for predicting atomic interactions. MLPs continue to advance and have had profound impact in applications that include drug discovery, enzyme catalysis, and materials design. The current landscape of MLP software presents challenges due to the limited interoperability between packages, which can lead to inconsistent benchmarking practices and necessitates separate interfaces with molecular dynamics (MD) software. To address these issues, we present DeePMD-GNN, a plugin for the DeePMD-kit framework that extends its capabilities to support external graph neural network (GNN) potentials.DeePMD-GNN enables the seamless integration of popular GNN-based models, such as NequIP and MACE, within the DeePMD-kit ecosystem. Furthermore, the new software infrastructure allows GNN models to be used within combined quantum mechanical/molecular mechanical (QM/MM) applications using the range corrected ΔMLP formalism.We demonstrate the application of DeePMD-GNN by performing benchmark calculations of NequIP, MACE, and DPA-2 models developed under consistent training conditions to ensure fair comparison.

Influence of Coverage Dependence on the Thermophysical Properties of Adsorbates and Its Impact on Microkinetic Models

This work focuses on the impact of lateral interactions on the thermophysical properties of adsorbates. We present different parametrizations for coverage-dependent enthalpy, entropy, and heat capacity in a mean-field microkinetic model. These models are tested against two systems, CO/Pt(111) and CO/Co(0001), using two different functionals. A detailed investigation into how coverage influences the thermophysical properties of CO* is presented. We place particular emphasis on studying the impact of coverage on the vibrational partition function and how this affects the entropy of adsorbates. Higher coverages typically lead to increased repulsive interactions, which should further constrain the large amplitude modes that contribute the most to the vibrational entropy. In some cases, however, the opposite effect occurred; the vibrational entropy actually increased because surface crowding forced adsorbates to different binding locations that had lower frequencies. Our results highlighted cases where coverage-dependent entropy should be included, such as for adsorbates with lateral vibrational modes and systems at high temperatures. These methods for including coverage-dependent properties into mean-field microkinetics in a thermodynamically consistent way are now available in the open-source software Cantera.

Discriminating High from Low Energy Conformers of Druglike Molecules: An Assessment of Machine Learning Potentials and Quantum Chemical Methods

Accurate and efficient prediction of high energy ligand conformations is important in structure-based drug discovery for the exclusion of unrealistic structures in docking-based virtual screening and de novo design approaches. In this work, we constructed a database of 140 solution conformers from 20 druglike molecules of varying size and chemical complexity, with energetics evaluated at the DLPNO-CCSD(T)/complete basis set (CBS) level. We then assessed a selection of machine learning potentials and semiempirical quantum mechanical models for their ability to predict conformational energetics. The GFN2-xTB tight binding density functional method correlates with reference conformer energies, yielding a Kendall's τ of 0.63 and mean absolute error of 2.2 kcal/mol. As putative internal energy filters for screening, we find that the GFN2-xTB, ANI-2x and MACE-OFF23(L) models perform well in identifying low energy conformer geometries, with sensitivities of 95 %, 89 % and 95 % respectively, but display a reduced ability to exclude high energy conformers, with respective specificities of 80 %, 61 % and 63 %. The GFN2-xTB method therefore exhibited the best overall performance and appears currently the most suitable of the three methods to act as an internal energy filter for integration into drug discovery workflows. Enrichment of high energy conformers in the training of machine learning potentials could improve their performance as conformational filters.

Electronic insights into the role of nuclear quantum effects in proton transfer reactions of nucleobase pairs

Double proton transfer in nucleobase pairs leads to point mutations in nucleic acids. A series of constrained nuclear-electronic orbital calculations combined with natural bond orbital and non-covalent interaction analyses, and kinetic studies have quantitatively revealed the importance of nuclear quantum effects (NQEs) in the reaction. Compared with the classical treatment of the nuclei, the probability of forming the tautomeric isomers of Cytosine-Guanine, when explicitly accounting for NQEs, increased by a factor of 8.0. This outcome can be attributed to enhancing the interaction between the orbitals at the reactive site due to NQEs, which increased the number of electrons occupying the antibonding orbitals.

The NOMAD mini-apps: A suite of kernels from ab initio electronic structure codes enabling co-design in high-performance computing

This article introduces a suite of mini-applications (mini-apps) designed to optimise computational kernels in ab initio electronic structure codes. The suite is developed from flagship applications participating in the NOMAD Center of Excellence, such as the ELPA eigensolver library and the GW implementations of the exciting, Abinit, and FHI-aims codes. The mini-apps were identified by targeting functions that significantly contribute to the total execution time in the parent applications. This strategic selection allows for concentrated optimisation efforts. The suite is designed for easy deployment on various High-Performance Computing (HPC) systems, supported by an integrated CMake build system for straightforward compilation and execution. The aim is to harness the capabilities of emerging (post)exascale systems, which necessitate concurrent hardware and software development - a concept known as co-design. The mini-app suite serves as a tool for profiling and benchmarking, providing insights that can guide both software optimisation and hardware design. Ultimately, these developments will enable more accurate and efficient simulations of novel materials, leveraging the full potential of exascale computing in material science research.

Eliminating nanometer-scale asperities on metallic thin films through plasma modification processes studied by molecular dynamics and AFM

We report the effects of reducing surface asperity size at the nanometer scale on metallic surfaces by plasma-assisted surface modification processes using simulations and experiments. Molecular dynamics (MD) simulations were conducted by irradiating various inert gas ions (Ne, Ar, Kr, and Xe) onto a cobalt slab with nanoscale asperities on the surface. The MD simulations showed that as the atomic number of the inert gas increased the surface asperity size was reduced more efficiently, while the etching rate decreased. The dependencies of the scattering behaviors on the inert gas ions originated from the mass exchange between the working gas ions and the slab atoms. Atomic force microscopy and X-ray fluorescence measurements were performed on hard disk media subjected to the surface modification processes. These measurements experimentally demonstrated that the density of nanoscale asperities was reduced with a lower etching rate as the atomic number of the inert gas increased, consistent with the simulation results. Through this study, we clarified that heavier working gases were more effective in reducing surface asperity size without significantly reducing the thickness of the material, which can contribute to better control of surface morphologies at the nanometer scale.

Controlling Reaction Pathways of Ethylene Hydroformylation Using Isolated Bimetallic Rhodium-Cobalt Sites

Designing efficient ligand-free heterogeneous catalysts for ethylene hydroformylation to produce C3 oxygenates is of importance for both fundamental research and practical applications, but it is often hindered by insufficient catalytic activity and selectivity. This work designs isolated rhodium-cobalt (Rh-Co) sites confined within a ZSM-5 zeolite to enhance ethylene hydroformylation rates and selectivity while maintaining catalyst stability. By adjusting the Co/Al ratio in Co-ZSM-5, different sizes of Co are formed; subsequent Rh introduction produces isolated Rh1Cox clusters with different Rh-Co coordination numbers (CNs). In-situ characterizations and density functional theory calculations reveal that a Rh-Co CN of 3, corresponding to an isolated Rh1Co3 site, provides optimal bindings to reaction intermediates and thus achieves the highest hydroformylation rates among supported Rh-based catalysts. This study demonstrates the role of coordination-tuning via a secondary metal in effectively controlling the reaction pathway over single Rh atom catalysts.

Point defect formation at finite temperatures with machine learning force fields

Point defects dictate the properties of many functional materials. The standard approach to modelling the thermodynamics of defects relies on a static description, where the change in Gibbs free energy is approximated by the internal energy. This approach has a low computational cost, but ignores contributions from atomic vibrations and structural configurations that can be accessed at finite temperatures. We train a machine learning force field (MLFF) to explore dynamic defect behaviour using Te+1 i and V +2 Te in CdTe as exemplars. We consider the different entropic contributions (e.g., electronic, spin, vibrational, orientational, and configurational) and compare methods to compute the defect free energies, ranging from a harmonic treatment to a fully anharmonic approach based on thermodynamic integration. We find that metastable configurations are populated at room temperature and thermal effects increase the predicted concentration of Te+1 i by two orders of magnitude - and can thus significantly affect the predicted properties. Overall, our study underscores the importance of finite-temperature effects and the potential of MLFFs to model defect dynamics at both synthesis and device operating temperatures.

Correlated Flat-Bottom Elastic Network Model for Improved Bond Rearrangement in Reaction Paths

This study introduces correlated flat-bottom elastic network model (CFB-ENM), an extension of our recently developed flat-bottom elastic network model (FB-ENM) for generating plausible reaction paths, i.e., collision-free paths preserving nonreactive parts. While FB-ENM improved upon the widely used image-dependent pair potential (IDPP) by addressing unintended structural distortion and bond breaking, it still struggled with regulating the timing of series of bond breaking and formation. CFB-ENM overcomes this limitation by incorporating structure-based correlation terms. These terms impose constraints on pairs of atom pairs, ensuring immediate formation of new bonds after breaking of existing bonds. Using the direct MaxFlux method, we generated paths for 121 reactions involving main group elements and 35 reactions involving transition metals. We found that CFB-ENM significantly improves reaction paths compared to FB-ENM. CFB-ENM paths exhibited lower maximum DFT energies along the paths in most reactions, with nearly half showing significant energy reductions of several tens of kcal/mol. In the few cases where CFB-ENM yielded higher energy paths, most increases were below 10 kcal/mol. We also confirmed that CFB-ENM reduces computational costs in subsequent precise reaction path or transition state searches compared to FB-ENM. An implementation of CFB-ENM based on the Atomic Simulation Environment is available on GitHub for use in computational chemistry research.

Melting, Bubblelike Expansion, and Explosion of Superheated Plasmonic Nanoparticles

We report on time-resolved coherent diffraction imaging of gas-phase silver nanoparticles, strongly heated via their plasmon resonance. The x-ray diffraction images reveal a broad range of phenomena for different excitation strengths, from simple melting over strong cavitation to explosive disintegration. Molecular dynamics simulations fully reproduce this behavior and show that the heating induces rather similar trajectories through the phase diagram in all cases, with the very different outcomes resulting solely from whether and where the stability limit of the metastable superheated liquid is crossed.

Cathodic Hydroxide Ions Induce Tetrose Formation during Glycolaldehyde Electroreduction to Alcohols: A Potential CO2-to-Carbohydrate Pathway

The electrochemical synthesis of organic compounds from CO2 can potentially alleviate climate change by hampering the atmospheric accumulation of greenhouse gases. The production of carbohydrates from CO2 reduction will have promising applications for the manufacturing of valuable, multi-carbon compounds that are traditionally produced from the petrochemical or agricultural industries. In this work, we analyzed the copper-catalyzed electrochemical reduction of glycolaldehyde, a commonly observed trace CO2RR product that has been previously proposed as an intermediate for alcohol formation. We determine that glycolaldehyde is not the main intermediate on polycrystalline copper-based electrocatalysts that selectively produce ethanol. In an unbuffered electrolyte, the cathodic hydroxide ions produced induce the coupling of glycolaldehyde to tetroses in the solution phase, yielding a maximum glycolaldehyde-to-sugar conversion of 47.2% under ambient conditions. Using in situ infrared spectroscopy coupled with density functional theory (DFT) calculations, we show that glycolaldehyde reduction to alcohols proceeds via adsorption of its enol tautomer, η2(C,C)─CHOH═CHOH. Our findings not only shed light on the C2 alcohol formation pathways during CO2RR, but also imply that a CO2 electrolyzer can potentially produce C4 carbohydrates via CO2 reduction to glycolaldehyde followed by C─C coupling in the solution phase, with only a high local pH needed to drive the tetrose formation step.

Form factor of prismatic particles for small-angle scattering analysis

Since the morphology of nanoparticles directly influences many of their properties, accurately determining their shape is crucial for targeted applications. In this work, we focus on nanoprisms due to their widespread use and the limitations of direct imaging techniques in accurately describing their polygonal cross section. Specifically, we introduce a new tool for small-angle scattering (SAS) analysis of nanoprisms that requires minimal computation time compared with all-atom simulations and other form factor analyses. A key innovation in this work is the implementation of the Lebedev quadrature for isotropic averaging, which allows for accurate form factor calculations using few sampling points. This form factor model is developed for any n-sided prism and is compared with small-angle X-ray scattering and transmission electron microscopy experimental data for gold and/or silver nanoprisms (n = 3, 4, 5). For small sizes, the nanoprism form factor model is compared with the result obtained with the Debye equation from atomic coordinates, showing a very good agreement. We explore the effects of the aspect ratio and cross-sectional shape of the nanoprisms on the form factor curves and discuss the limitations of our approach. Overall, our method combines precise shape determination with rapid computation time, paving the way for detailed characterization of nanoprisms using SAS techniques, potentially even during their growth.

Toward a unified benchmark and framework for deep learning-based prediction of nuclear magnetic resonance chemical shifts

The study of structure-spectrum relationships is essential for spectral interpretation, impacting structural elucidation and material design. Predicting spectra from molecular structures is challenging due to their complex relationships. Here we introduce NMRNet, a deep learning framework using the SE(3) Transformer for atomic environment modeling, following a pretraining and fine-tuning paradigm. To support the evaluation of nuclear magnetic resonance chemical shift prediction models, we have established a comprehensive benchmark based on previous research and databases, covering diverse chemical systems. Applying NMRNet to these benchmark datasets, we achieve competitive performance in both liquid-state and solid-state nuclear magnetic resonance datasets, demonstrating its robustness and practical utility in real-world scenarios. Our work helps to advance deep learning applications in analytical and structural chemistry.

Quantum Calculations of Hydrogen Absorption and Diffusivity in Bulk CeO2

CeO2 (ceria) is an attractive material for heterogeneous catalysis applications involving hydrogen due to its favorable redox activity combined with its relative impermeability to hydrogen ions and molecules. However, to date, many bulk ceria/hydrogen properties remain unresolved in part due to a scarcity of experimental data combined with quantum calculation results that vary according to the approach used. In this regard, we have conducted a series of density functional theory (DFT) calculations utilizing generalized gradient (GGA), metaGGA, and hybrid functionals as well as several corrections for electronic correlations, applied to a number of properties regarding hydrogen in bulk stoichiometric CeO2. Our calculations place reasonable bounds on the lattice constants, band gaps, hydrogen absorption energies, and O-H bond vibrational frequencies that can be determined by DFT. In addition, our results indicate that the activation energy barriers for hydrogen bulk diffusion are uniformly low (<0.15 eV) for the calculation parameters probed here and that, in general, the effect of hydrogen tunneling is small at ambient temperatures. Our study provides a recipe to determine fundamental physical chemical properties of Ce-O-H interactions while also determining realistic ranges for diffusion kinetics. This can facilitate the determination of future coarse-grained models that will be able to guide and elucidate experimental efforts in this area.

Fine-Tuning Graph Neural Networks via Active Learning: Unlocking the Potential of Graph Neural Networks Trained on Nonaqueous Systems for Aqueous CO2 Reduction

Graph neural networks (GNNs) have revolutionized catalysis research with their efficiency and accuracy in modeling complex chemical interactions. However, adapting GNNs trained on nonaqueous data sets to aqueous systems poses notable challenges due to intricate water interactions. In this study, we proposed an active learning-based fine-tuning approach to extend the applicability of GNNs to aqueous environments. The geometry optimization and transition state search workflows are designed to reduce computational costs while maintaining DFT-level accuracy. Applied to the CO2 reduction reaction, the workflow delivers a 2-3-fold acceleration in geometry optimization through a relaxed force threshold combined with DFT refinement. The versatility of the transition state search algorithm was demonstrated on key C-C coupling pathways, pinpointing *CO-*COH as the most energetically favorable pathway in aqueous systems of Cu and Cu-based Ag, Au, and Zn alloys. The Brønsted-Evans-Polanyi relationship remains robust under water-induced fluctuations, with alloyed metals such as Al, Ga, and Pd, along with Ag, Au, and Zn, exhibiting coupling efficiency comparable to that of Cu. Additionally, perturbation-based training on forces and energies extends the application of GNNs to aqueous ab initio molecular dynamics simulations, enabling efficient modeling of dynamical trajectories. This work presents novel approaches to adapting nonaqueous models for application in aqueous systems, highlighting GNNs' potential in solvated environments and laying a foundation for accelerating predictions of catalytic mechanisms under realistic conditions.

A Three-Tiered Hierarchical Computational Framework Bridging Molecular Systems and Junction-Level Charge Transport

The nonequilibrium Green's function (NEGF) method combined with ab initio calculations has been widely used to study charge transport in molecular junctions. However, the significant computational demands of high-resolution calculations for all device components pose challenges in simulating junctions with complex molecular junction structures and understanding the functionality of molecular devices. In this study, we developed a series of computational methods capable of effectively handling the molecular Hamiltonian, electrode electronic structures, and their interfacial coupling at different theoretical levels. As three-tiered hierarchical levels, they enable efficient charge transport computations ranging from individual molecules to complete junction systems, achieving an optimal balance between computational cost and accuracy. Moreover, integrated into a Question-Driven Hierarchical Computation (QDHC) framework, we show this three-tiered framework is able to address specific research objectives by isolating and analyzing the dominant factors governing charge transport, thus significantly enhancing the efficiency of analyzing charge transport mechanisms, as validated through a series of benchmark studies on diverse molecular junction systems.

Adaptive-precision potentials for large-scale atomistic simulations

Large-scale atomistic simulations rely on interatomic potentials, providing an efficient representation of atomic energies and forces. Modern machine-learning (ML) potentials provide the most precise representation compared to electronic structure calculations, while traditional potentials provide a less precise but computationally much faster representation and, thus, allow simulations of larger systems. We present a method to combine a traditional and a ML potential into a multi-resolution description, leading to an adaptive-precision potential with an optimum of performance and precision in large, complex atomistic systems. The required precision is determined per atom by a local structure analysis and updated automatically during simulation. We use copper as demonstrator material with an embedded atom model as classical force field and an atomic cluster expansion (ACE) as ML potential, but, in principle, a broader class of potential combinations can be coupled by this method. The approach is developed for the molecular-dynamics simulator LAMMPS and includes a load-balancer to prevent problems due to the atom dependent force-calculation times, which makes it suitable for large-scale atomistic simulations. The developed adaptive-precision copper potential represents the ACE-forces with a precision of 10 me V/Å and the ACE-energy exactly for the precisely calculated atoms in a nanoindentation of 4 × 106 atoms calculated for 100 ps and shows a speedup of 11.3 compared with a full ACE simulation.

Efficient exploration of reaction pathways using reaction databases and active learning

The fast and accurate simulation of chemical reactions is a major goal of computational chemistry. Recently, the pursuit of this goal has been aided by machine learning interatomic potentials (MLIPs), which provide energies and forces at quantum mechanical accuracy but at a fraction of the cost of the reference quantum mechanical calculations. Assembling the training set of relevant configurations is key to building the MLIP. Here, we demonstrate two approaches to training reactive MLIPs based on reaction pathway information. One approach exploits reaction datasets containing reactant, product, and transition state structures. Using an SN2 reaction dataset, we accurately locate reaction pathways and transition state geometries of up to 170 unseen reactions. In another approach, which does not depend on data availability, we present an efficient active learning procedure that yields an accurate MLIP and converged minimum energy path given only the reaction end point structures, avoiding quantum mechanics driven reaction pathway search at any stage of training set construction. We demonstrate this procedure on an SN2 reaction in the gas phase and with a small number of solvating water molecules, predicting reaction barriers within 20 meV of the reference quantum chemistry method. We then apply the active learning procedure on a more complex reaction involving a nucleophilic aromatic substitution and proton transfer, comparing the results against the reactive ReaxFF force field. Our active learning procedure, in addition to rapidly finding reaction paths for individual reactions, provides an approach to building large reaction path databases for training transferable reactive machine learning potentials.

Comparative Study of UMCM-9 Polymorphs: Structural, Dynamic, and Hydrogen Storage Properties via Atomistic Simulations

The structural and dynamic properties of two polymorphs of the metal-organic framework UMCM-9 (UMCM-9-α and -β) have been studied via molecular dynamics (MD) simulations in conjunction with density functional tight binding (DFTB) as well as the newly developed MACE-MP neural network potential (NNP). Based on these calculations, a novel UMCM-9-β polymorph is proposed that exhibits reduced linker strain and increased flexibility compared to UMCM-9-α, which is shown to be energetically less stable. UMCM-9-β exhibits enhanced diffusion of molecular hydrogen due to weaker host-guest interactions, whereas UMCM-9-α exhibits stronger interactions, leading to improved hydrogen adsorption. The results suggest that synthesis conditions may control the formation of both polymorphs: UMCM-9-β is likely to be the thermodynamic product, forming under stable conditions, while UMCM-9-α may be the kinetic product, forming under accelerated synthesis conditions. This study highlights the potential for optimizing MOFs for specific gas storage applications to achieve the desired structural and associated gas storage properties.

Direct Imaging of Chirality Transfer Induced by Glycosidic Bond Stereochemistry in Carbohydrate Self-Assemblies

Carbohydrates, essential biological building blocks, exhibit functional mechanisms tied to their intricate stereochemistry. Subtle stereochemical differences, such as those between the anomers maltose and cellobiose, lead to distinct properties due to their differing glycosidic bonds; the former is digestible by humans, while the latter is not. This underscores the importance of precise structural determination of individual carbohydrate molecules for deeper functional insights. However, their structural complexity and conformational flexibility, combined with the high spatial resolution needed, have hindered direct imaging of carbohydrate stereochemistry. Here, we employ noncontact atomic force microscopy integrated with a data-efficient, multifidelity structure search approach accelerated by machine learning integration to determine the precise 3D atomic coordinates of two carbohydrate anomers on Au(111). We observe that the stereochemistry of the glycosidic bond regulates on-surface chiral selection in carbohydrate self-assemblies. The reconstructed models, validated against experimental data, provide reliable atomic-scale structural evidence, uncovering the origin of the on-surface chirality from carbohydrate anomerism. Our study confirms that nc-AFM is a reliable technique for real-space discrimination of carbohydrate stereochemistry at the single-molecule level, providing a pathway for bottom-up investigations into the structure-property relationships of carbohydrates in biological research and materials science.

On-demand heralded MIR single-photon source using a cascaded quantum system

We propose a mechanism for generating single photons in the mid-infrared (MIR) using a solid-state or molecular quantum emitter. The scheme uses cavity quantum electrodynamics (QED) effects to selectively enhance a Frank-Condon transition, deterministically preparing a single Fock state of a polar phonon mode. By coupling the phonon mode to an antenna, the resulting excitation is then radiated to the far field as a single photon with a frequency matching the phonon mode. By combining macroscopic QED calculations with methods from open quantum system theory, we show that optimal parameters to generate these MIR photons occur for modest light-matter coupling strengths, which are achievable with state-of-the-art technologies. Combined, the cascaded system we propose provides a quasi-deterministic source of heralded single photons in a regime of the electromagnetic spectrum where this previously was not possible.

Crystal Structure Prediction Meets Artificial Intelligence

Crystal structure prediction (CSP) represents a fundamental research frontier in computational materials science and chemistry, aiming to predict thermodynamically stable periodic structures from given chemical compositions. Traditional methods often face challenges such as high computational costs and local minima trapping. Recently, artificial intelligence methods, represented by generative adversarial networks (GANs), variational autoencoders (VAEs), diffusion models, and large language models (LLMs), have revolutionized the traditional prediction paradigm. These computational frameworks efficiently extract chemical rules and structural features from crystal databases, significantly reducing computational costs while maintaining prediction accuracy. This Perspective systematically evaluates the advantages and limitations of various generative models, explores their synergies with conventional approaches, and discusses their future prospects in accelerating materials discovery and development, providing new insights for future research directions.

Overview on Building Blocks and Applications of Efficient and Robust Extended Tight Binding

The extended tight binding (xTB) family of methods opened many new possibilities in the field of computational chemistry. Within just 5 years, the GFN2-xTB parametrization for all elements up to Z = 86 enabled more than a thousand applications, which were previously not feasible with other electronic structure methods. The xTB methods provide a robust and efficient way to apply quantum mechanics-based approaches for obtaining molecular geometries, computing free energy corrections or describing noncovalent interactions and found applicability for many more targets. A crucial contribution to the success of the xTB methods is the availability within many simulation packages and frameworks, supported by the open source development of its program library and packages. We present a comprehensive summary of the applications and capabilities of xTB methods in different fields of chemistry. Moreover, we consider the main software packages for xTB calculations, covering their current ecosystem, novel features, and usage by the scientific community.

Propane Activation on Pt Electrodes at Room Temperature: Quantification of Adsorbate Identity and Coverage

C-H bond activation is the first step in manufacturing chemical products from readily available light alkane feedstock and typically proceeds via carbon-intensive thermal processes. The ongoing emphasis on decarbonization via electrification motivates low-temperature electrochemical alternatives that could lead to sustainable chemicals production. Platinum (Pt) electrocatalysts have shown activity towards reacting alkanes; however, little is known about propane electrocatalytic activation and conditions suitable for enabling selective oxidation to valuable products. Herein, we utilize a combination of electrochemical mass spectrometry (ECMS) and density functional theory (DFT) calculations to elucidate the potential dependence of propane activation on Pt electrocatalysts. Results show a strong dependence of adsorption on the applied potential in room-temperature aqueous acidic electrolyte, with a maximum coverage of propane-derived adsorbates at 0.30 V vs RHE. Using charge deconvolution and deuterated experiments, the mechanism of adsorption was elucidated, and C3H2 * was determined as the average dehydrogenated propane-derived adsorbate species. DFT calculations further corroborate these results, showing that the formation of deeply dehydrogenated species is energetically accessible at room temperature. The combined theoretical and experimental findings yield insights for selective activation of paraffinic C-H bonds at room temperature, aqueous conditions-a critical step towards decarbonized chemical manufacturing.

Alkali Cation Inhibition of Imidazolium-Mediated Electrochemical CO2 Reduction on Silver

Imidazolium-based ionic liquids have led to enhanced CO2 electroreduction activity due to cation effects at the cathode surface, stabilizing the reaction intermediates and decreasing the activation energy. In aqueous media, alkali cations are also known to improve CO2 reduction activity on metals such as Ag, with the enhancement attributed to electrical double layer effects and trending with the size of the alkali cation. However, the effect of a mixed catholyte solution of alkali cations in the presence of an imidazolium-based ionic liquid has not been well-explored. Herein, 1-ethyl-3-methylimidazolium tetrafluoroborate, [EMIM][BF4], in water was investigated with alkali salts to unravel the interaction effects for CO2 electroreduction on Ag. Although both [EMIM]+ and alkali cations have individually improved CO2 to CO conversion on Ag in water, electrochemical results showed that alkali cations hindered imidazolium-mediated CO2 electroreduction in most conditions. Li+, in particular, was sharply inhibitory compared to other alkali cations and strongly redirected the selectivity to hydrogen evolution. The nature of the alkali cation inhibition was investigated with spectroscopic techniques, including in situ surface-enhanced Raman spectroscopy (SERS) and dynamic electrochemical impedance spectroscopy (DEIS). Along with computational insights from density functional theory (DFT), the electrochemical and spectroscopic data suggest that alkali cations inhibit [EMIM]-mediated CO2 reduction by competing for surface adsorption sites, preventing the potential-dependent structural reorientation of imidazolium, and promoting hydrogen evolution by bringing solvated water to the cathode surface.

Machine-Learning-Accelerated Surface Exploration of Reconstructed BiVO4(010) and Characterization of Their Aqueous Interfaces

Understanding the semiconductor-electrolyte interface in photoelectrochemical (PEC) systems is crucial for optimizing the stability and reactivity. Despite the challenges in establishing reliable surface structure models during PEC cycles, this study explores the complex surface reconstructions of BiVO4(010) by employing a computational workflow integrated with a state-of-the-art active learning protocol for a machine-learning interatomic potential and global optimization techniques. Within this workflow, we identified 494 unique reconstructed surface structures that surpass conventional chemical intuition-driven, bulk-truncated models. After constructing the surface Pourbaix diagram under Bi- and V-rich electrolyte conditions using density functional theory and hybrid functional calculations, we proposed structural models for the experimentally observed Bi-rich BiVO4 surfaces. By performing hybrid functional molecular dynamics simulations with the explicit treatment of water molecules on selected reconstructed BiVO4(010) surfaces, we observed water dissociation from molecular water. Our findings demonstrate significant water dissociation on reconstructed Bi-rich surfaces, highlighting the critical role of bare and undercoordinated Bi sites (only observable in reconstructed surfaces) in driving hydration processes. Our work establishes a foundation for understanding the role of complex, reconstructed Bi surfaces in surface hydration and reactivity. Additionally, our theoretical framework for exploring surface structures and predicting reactivity in multicomponent oxides offers a precise approach to describing complex surface and interface processes in PEC systems.

Defects induce phase transition from dynamic to static rippling in graphene

Two-dimensional (2D) materials display nanoscale dynamic ripples that significantly impact their properties. Defects within the crystal lattice are the elementary building blocks to tailor the material's morphology. While some studies have explored the link between defective structures and rippling dynamics in 2D materials, a comprehensive understanding of this relationship has yet to be achieved. Here, we address this using machine learning-driven molecular dynamics simulations. Specifically, we find that above a critical concentration of defects, free-standing graphene sheets undergo a dynamic transition from freely propagating to static ripples. Our computational approach captures the dynamics with atomic resolution, and reveals that the transition is driven by elastic interactions between defects. The strength of these interactions is found to vary across defect types and we identify a unifying set of principles driving the dynamic-to-static transition in 2D materials. Our work not only rationalizes puzzling experimental results for defective 2D materials, but also paves the way to design two-dimensional devices with tailored rippling dynamics. These insights could lay the foundations for a class of disorder-based catalytic and interfacial materials.

Electrochemical CO2 reduction to liquid fuels: Mechanistic pathways and surface/interface engineering of catalysts and electrolytes

The high energy density of green synthetic liquid chemicals and fuels makes them ideal for sustainable energy storage and transportation applications. Electroreduction of carbon dioxide (CO2) directly into such high value-added chemicals can help us achieve a renewable C cycle. Such electrochemical reduction typically suffers from low faradaic efficiencies (FEs) and generates a mixture of products due to the complexity of controlling the reaction selectivity. This perspective summarizes recent advances in the mechanistic understanding of CO2 reduction reaction pathways toward liquid products and the state-of-the-art catalytic materials for conversion of CO2 to liquid C1 (e.g., formic acid, methanol) and C2+ products (e.g., acetic acid, ethanol, n-propanol). Many liquid fuels are being produced with FEs between 80% and 100%. We discuss the use of structure-binding energy relationships, computational screening, and machine learning to identify promising candidates for experimental validation. Finally, we classify strategies for controlling catalyst selectivity and summarize breakthroughs, prospects, and challenges in electrocatalytic CO2 reduction to guide future developments.

Automated and Efficient Sampling of Chemical Reaction Space

Machine learning interatomic potentials (MLIPs) promise quantum-level accuracy at classical force field speeds, but their performance hinges on the quality and diversity of training data. An efficient and fully automated approach to sample chemical reaction space without relying on human intuition, addressing a critical gap in MLIP development is presented. The method combines the speed of tight-binding calculations with selective high-level refinement, generating diverse datasets that capture both equilibrium and reactive regions of potential energy surfaces. By employing single-ended growing string and nudged elastic band methods, reaction pathways previously underrepresented in MLIP training sets, particularly near transition states are systematically explored. This approach yields datasets with rich structural and chemical diversity, essential for robust MLIP development. Open-source code is provided for the entire workflow, facilitating the integration of the approach into existing MLIP development pipelines.

Octahedral Tilt-Driven Phase Transitions in BaZrS3 Chalcogenide Perovskite

Chalcogenide perovskites are lead-free materials for potential photovoltaic or thermoelectric applications. BaZrS3 is the most-studied member of this family due to its superior thermal and chemical stability, desirable optoelectronic properties, and low thermal conductivity. Phase transitions in BaZrS3 remain underexplored in the literature, as most experimental characterizations of this material have been performed at ambient conditions where the orthorhombic Pnma phase is reported to be stable. In this work, we study the dynamics of BaZrS3 across a range of temperatures and pressures using an accurate machine learning interatomic potential trained with data from hybrid density functional theory calculations. At 0 Pa, we find a first-order phase transition from the orthorhombic to tetragonal I4/mcm phase at 610 K, and a second-order transition from the tetragonal to the cubic Pm3̅m phase at 880 K. The tetragonal phase is stable over a larger temperature range at higher pressures. To confirm the validity of our model we compare our results with a range of published experimental data and report a prediction for the X-ray diffraction pattern as a function of temperature.

Investigation of Surface Passivation Mechanisms in CrCoNi Alloys via Interpretable Machine Learning

This study investigates the formation and stability of passivation films in medium-entropy CrCoNi alloy using an integrated approach combining first-principles calculations and Gaussian Approximation Potential (GAP). By performing extensive structural relaxations coupled with random occupancy of the alloy and multiple potential adsorption sites, the results based on surface reconstruction are analyzed. Our results demonstrate that disordered atomic structures in medium entropy CrCoNi alloy exhibit improved oxide film stability compared to ordered structures, with lower system energies correlating with more robust passivation layers. Furthermore, by constructing physically meaningful descriptors and employing symbolic regression in complex surface environments, simple yet highly correlated rules are obtained, providing insights into the passivation strengthening of CrCoNi alloy from both global and local perspectives. Our work paves the way for exploring surface regularities in complex medium-entropy alloys with the combination of quantum mechanical accuracy and machine learning efficiency.

Using Conformational Sampling to Model Spectral and Structural Changes of Molecules at Elevated Pressures

Conformational sampling is nowadays a standard routine in computational chemistry. Within this work, we present a method to perform conformational sampling for systems exposed to elevated pressures within the CREST program, allowing us to model pressure-induced changes of molecular ensembles and structural parameters. For this purpose, we extend the molecular Hamiltonian with the PV (pressure times volume) term, using the solvent-accessible volume. The volume computation is performed within the new standalone library libpvol. A first application shows good agreement with experimental data and provides a reasonable explanation for severe pressure-induced structural and spectroscopic changes of the molecules dichloroethane and tetra(4-methoxyphenyl)ethylene.

Recombination of Autodissociated Water Ions in a Nanoscale Pure Water Droplet

The recombination of water ions has diverse scientific and practical implications, ranging from acid-base chemistry and biological systems to planetary environments and applications in fuel cell and carbon conversion technologies. While spatial confinement affects the physicochemical properties of water dynamics, its impact on the recombination process has rarely been studied. In this work, we investigate the dynamics of water, the water ion distribution, and the ion recombination process in water droplets as a function of droplet size through molecular dynamics simulations and adaptive quantum mechanical/molecular mechanical calculations. We compare the dynamics of recombination in water droplet sizes ranging from 100 to 18 000 waters, both in their interiors and on their surfaces. We found that the self-diffusion of water dramatically decreases in droplets with a diameter below 2.2 nm. Using a classical RexPoN force-field, we found that the ions in 1000 H2O's spend almost 50% of the time on the surface and 0.5 nm beneath it with a slight preference for OH- ion to reside longer on the surface. We estimate that, on average, recombination in these drops occurs at 400 ps in 1000 H2O's and 1 ns in 3000 H2O's. We also found that recombination is not limited by the local structure of the surface or the size of the droplet but can be influenced by the geometry of the water wire connecting the ions as they approach each other, which can often prevent recombination. Our results provide insights to the reaction microenvironments presented by nanoscopic water droplets.

Broadening the Scope of Neural Network Potentials through Direct Inclusion of Additional Molecular Attributes

Most state-of-the-art neural network potentials do not account for molecular attributes other than atomic numbers and positions, which limits its range of applicability by design. In this work, we demonstrate the importance of including additional electronic attributes in neural network potential representations with a minimal architectural change to TensorNet, a state-of-the-art equivariant model based on Cartesian rank-2 tensor representations. By performing experiments on both custom-made and public benchmarking data sets, we show that this modification resolves input degeneracy issues stemming from the use of atomic numbers and positions alone, while enhancing the model's predictive accuracy across diverse chemical systems with different charge or spin states. This is accomplished without tailored strategies or the inclusion of physics-based energy terms, while maintaining efficiency and accuracy. These findings should furthermore encourage researchers to train and use models incorporating these additional representations.

Ordered Pt3Mn Intermetallic Setting the Maximum Threshold Activity of Disordered Variants for Glycerol Electrolysis

Glycerol electrolysis is a promising strategy for generating hydrogen at the cathode and value-added products at the anode. However, the effect of the atomic distribution within catalysts on their catalytic performance remains largely unexplored, primarily because of the inherent complexity of the glycerol oxidation reaction (GOR). Herein, an ordered Pt3Mn (O-Pt3Mn) intermetallic compound and a disordered Pt3Mn (D-Pt3Mn) alloy are used as model catalysts, and their performance in the GOR and hydrogen evolution reaction (HER) is studied. O-Pt3Mn consistently outperforms D-Pt3Mn and commercial Pt/C catalysts. It can generate high-value glycerate at a notable production rate of 17 mM h-1 while achieving an impressively low cell voltage of 0.76 V for glycerol electrolysis, which is ∼0.98 V lower than that required for water electrolysis. Statistical analysis using theoretical calculations reveals that Pt-Pt-Pt hollow sites are crucial for the catalytic GOR and HER. The averaged adsorption energies of key intermediates (simplified as C*, O*, and H*) on diverse catalysts closely correlate with their experimentally observed activity. Our proposed linear models accurately predict these adsorption energies, exhibiting high correlation coefficients ranging from 0.97 to 0.99 and highlighting the significance of the distribution of the topmost and subsurface-corner Mn atoms in determining these adsorption energies. By sampling all possible Mn configurations within the fitted linear models, we confirm that O-Pt3Mn establishes the maximum activity threshold for the GOR and HER compared with any disordered variant. This study presents an innovative framework for exploring the effect of the atomic distribution within catalysts on their catalytic performance and designing high-performance catalysts for complex reactions.

Band-Gap Regression with Architecture-Optimized Message-Passing Neural Networks

Graph-based neural networks and, specifically, message-passing neural networks (MPNNs) have shown great potential in predicting physical properties of solids. In this work, we train an MPNN to first classify materials through density functional theory data from the AFLOW database as being metallic or semiconducting/insulating. We then perform a neural-architecture search to explore the model architecture and hyperparameter space of MPNNs to predict the band gaps of the materials identified as nonmetals. The top-performing models from the search are pooled into an ensemble that significantly outperforms the best single model. Uncertainty quantification is evaluated with Monte Carlo dropout and ensembling, with the ensemble method proving superior. The domain of applicability of the ensemble model is analyzed with respect to the crystal systems, the inclusion of a Hubbard parameter in the density-functional-theory calculations, and the atomic species building up the materials.

MORE-Q, a dataset for molecular olfactorial receptor engineering by quantum mechanics

We introduce the MORE-Q dataset, a quantum-mechanical (QM) dataset encompassing the structural and electronic data of non-covalent molecular sensors formed by combining 18 mucin-derived olfactorial receptors with 102 body odor volatilome (BOV) molecules. To have a better understanding of their intra- and inter-molecular interactions, we have performed accurate QM calculations in different stages of the sensor design and, accordingly, MORE-Q splits into three subsets: i) MORE-Q-G1: QM data of 18 receptors and 102 BOV molecules, ii) MORE-Q-G2: QM data of 23,838 BOV-receptor configurations, and iii) MORE-Q-G3: QM data of 1,836 BOV-receptor-graphene systems. Each subset involves geometries optimized using GFN2-xTB with D4 dispersion correction and up to 39 physicochemical properties, including global and local properties as well as binding features, all computed at the tightly converged PBE+D3 level of theory. By addressing BOV-receptor-graphene systems from a QM perspective, MORE-Q can serve as a benchmark dataset for state-of-the-art machine learning methods developed to predict binding features. This, in turn, can provide valuable insights for developing the next-generation mucin-derived olfactory receptor sensing devices.

Rational design of precatalysts and controlled evolution of catalyst-electrolyte interface for efficient hydrogen production

The concept of precatalyst is widely accepted in electrochemical water splitting, but the role of precatalyst activation and the resulted changes of electrolyte composition is often overlooked. Here, we elucidate the impact of potential-dependent changes for both precatalyst and electrolyte using Co2Mo3O8 as a model system. Potential-dependent reconstruction of Co2Mo3O8 precatalyst results in an electrochemically stable Co(OH)2@Co2Mo3O8 catalyst and additional Mo dissolved as MoO42- into electrolyte. The Co(OH)2/Co2Mo3O8 interface accelerates the Volmer reaction and negative potentials induced Mo2O72- (from MoO42-) further enhances proton adsorption and H2 desorption. Leveraging these insights, the well-designed MoO42-/Mo2O72- modified Co(OH)2@Co2Mo3O8 catalyst achieves a Faradaic efficiency of 99.9% and a yield of 1.85 mol h-1 at -0.4 V versus reversible hydrogen electrode (RHE) for hydrogen generation. Moreover, it maintains stable over one month at approximately 100 mA cm-2, highlighting its industrial suitability. This work underscores the significance of understanding on precatalyst reconstruction and electrolyte evolution in catalyst design.