arXiv subjects

Nima Karimitari

Publications and source records attributed to Nima Karimitari.

8 recordsLinked to original sources

arXiv2026

Systematic Fine-Tuning of MACE Interatomic Potentials for Catalysis

Once trained, machine-learned interatomic potentials (MLIPs) provide a fast and accurate way to study catalytic reaction pathways, but their performance strongly depends on the training set. Here, we compare nine MLIPs trained with different data sets and strategies, including from-scratch (FS) training and fine-tuning (FT) of large foundation models. The models are evaluated on reaction energies, $E_{r}$, and reaction energy barriers, $E_{a}$, for 141 reactions, including CO$_2$ reduction to C$_2$ and C$_3$ products, propane dehydrogenation, hydrogen intercalation on Pd, and out-of-distribution oxygen evolution reaction (OER) on metal oxides. FS models trained with 5%--10% perturbed high-energy configurations from molecular dynamics or contour exploration reduce the error by more than twofold compared with models trained only on relaxation trajectories. In contrast, FT MLIPs are less sensitive to sampling and transfer well to out-of-distribution reactions. An MLIP fine-tuned on metallic catalysts achieves a 0.30 eV MAE for OER on iridium oxide polymorphs, outperforming out-of-the-box MACE-MH-1 by 0.08 eV and the best FS model by 0.14 eV. A model fine-tuned to O and OH adsorption on metal oxides gives a 0.19 eV reaction-barrier MAE for out-of-distribution CO$_2$RR on Cu, comparable to an FS model trained on in-distribution C--C bond-breaking reactions. Finally, a large MLIP fine-tuned on 49,860 configurations gives the best overall performance across metallic and metal-oxide catalysts and was used to screen a large left-out set of bimetallic alloys, achieving a 0.15 eV MAE for $E_{r}$, even for adsorbates on unseen Miller-index surfaces such as (532). This work identifies the training configurations needed for accurate FS and FT MLIPs for catalytic reaction modeling.

physics.chem-ph↗

arXiv2025

Role of Wadsley Defects and Cation Disorder to Enhance MoNb12O33 Diffusion

Wadsley-Roth (WR) niobates have emerged as high-rate anode materials that can combine rapid ionic diffusion with good electronic conductivity. WR compounds have been defect-enhanced by limited annealing, however, such materials often contain multiple types of defects. In particular, both Wadsley defects (variable block size) and transition metal disorder have the potential to modify transport rates, however the corresponding effects are not well understood mechanistically. Here, MoNb12O33 (MNO) was calcined at two different temperatures to compare a defect-rich condition (MNO-800) with a proximal order-rich condition (MNO-900) as assessed through XRD, XANES, EXAFS, and STEM characterizations. Galvanostatically cycled lithium half cells of MNO-800 exhibited additional capacity (307 mAh/g at 0.1C, 4.66% higher) and improved high-rate capacity of 200 mAhg-1 at 10C. ICI-based overpotential analysis identified solid state diffusion as the dominant rate limiting process where MNO-800 correspondingly exhibited ~3X faster capacity-weighted diffusivity. A machine-learning interatomic potential was trained to density functional theory and then applied with molecular dynamics (MLIP-MD) to examine the possible roles of Wadsley defects and transition metal disorder. For both defect-types, Li was found to populate and activate fast diffusion paths from window sites at lower extents of lithiation as compared to the order-rich model.

cond-mat.mtrl-sci↗

arXiv2025

Combined Experimental and Computational Analysis of Lithium Diffusion in Isostructural Pair VNb9O25 and VTa9O25

Wadsley-Roth crystal structures are an attractive class of materials for batteries because lithium diffusion is facilitated by the ReO3-like block structure with electron transport enabled by edge-sharing along shear planes. However, clear structure-property relationships remain limited, making it challenging to develop improved materials. Here, the first lithiation of VTa9O25 is reported, enabling a direct isostructural comparison with the better-known VNb9O25. These materials have similar unit cell volumes and atomic radii yet exhibit different voltage windows, C-rate dependent capacities, and transport metrics. Time-dependent overpotential analysis reveals ionic diffusion as the primary bottleneck to high rate-performance in both cases, however, the lithium diffusivity for VNb9O25 was an order of magnitude faster than that for VTa9O25. These experimental trends aligned well with density functional theory calculations combined with molecular dynamics that show a factor of six faster diffusion in VNb9O25. Nudged elastic band calculations of the probable hopping pathways indicate that VNb9O25 consistently exhibits a lower activation barrier for lithium diffusion. Bader charge analysis reveals a larger net charge on Li in VNb9O25 due to the higher electronegativity of Nb which stabilizes the transition state and lowers the barrier. This stabilization arises from the stronger Coulombic interaction between Li and its coordinated O-environment. These materials behave similarly upon lithiation wherein the lattice vectors (corresponding to the block plane) increase until about 50% lithiation and then decrease. However, the electronic structure differs, indicating that VNb9O25 undergoes a insulator to metal transition at a lower state of charge compared with VTa9O25. Overall, this work establishes the role of the cation (Nb or Ta) on the electronic and transport properties during lithiation.

cond-mat.mtrl-sci↗

arXiv2025

Development of machine-learned interatomic potentials to predict structure, transport, and reactivity in platinum-based fuel cells

Machine-learned interatomic potentials (MLIPs) have rapidly progressed in accuracy, speed, and data efficiency in recent years. However, training robust MLIPs in multicomponent systems still remains a challenge. In this work, we train a MLIP to describe hydrated Nafion ionomers and platinum catalysts, which are important components of fuel cells, by constructing a diverse training set to describe the bulk polymer and interfacial catalyst-polymer interactions well. We find that active learning improves the initial dataset little in terms of reducing uncertainty and error, pointing towards a need for more effective methods to efficiently explore the relevant interactions in complex, multicomponent systems. We use our trained MLIP to study the properties of the platinum-Nafion system, including polymer structure, proton mobility in a bulk Nafion polymer and near a platinum-Nafion interface, and reactions near and far from the interface, finding excellent results for structure and reactions contained within our training set. Transport seems well described, with both vehicular transport and Grotthuss hopping captured, although converged calculations of diffusivities were not computed because they require calculations of tens of nanoseconds that are challenging with current state-of-the-art MLIPs. The combined insights that this model provides can be leveraged to optimize fuel cell performance, and the approach can be applied to other chemical processes and devices where structure, transport, and reactivity all contribute to overall observed performance.

physics.chem-ph↗

arXiv2024

Bridging electronic and classical density-functional theory using universal machine-learned functional approximations

The accuracy of density-functional theory (DFT) is determined by the quality of the approximate functionals, such as exchange-correlation in electronic DFT and the excess functional in the classical DFT formalism of fluids. The exact functional is highly nonlocal for both electrons and fluids, yet most approximate functionals are semi-local or nonlocal in a limited weighted-density form. Machine-learned (ML) nonlocal density-functional approximations are promising in both electronic and classical DFT, but have so far employed disparate approaches with limited generality. Here, we formulate a universal approximation framework and training protocol for nonlocal ML functionals, combining features of equivariant convolutional neural networks and the weighted-density approximation. We prototype this approach for several 1D and quasi-1D problems and demonstrate that a functional with exactly the same hyperparameters achieves excellent accuracy for the hard-rod fluid, the inhomogeneous Ising model, the exact exchange functional for electrons, the electron kinetic energy functional for orbital-free DFT, as well as for liquid water with 1D inhomogeneities. These results lay the foundation for a universal ML approach to exact 3D functionals spanning electronic and classical DFT.

physics.chem-ph↗

arXiv2024

Accurate Crystal Structure Prediction of New 2D Hybrid Organic Inorganic Perovskites

Low dimensional hybrid organic-inorganic perovskites (HOIPs) represent a promising class of electronically active materials for both light absorption and emission. The design space of HOIPs is extremely large, since a diverse space of organic cations can be combined with different inorganic frameworks. This immense design space allows for tunable electronic and mechanical properties, but also necessitates the development of new tools for in silico high throughput analysis of candidate structures. In this work, we present an accurate, efficient, transferable and widely applicable machine learning interatomic potential (MLIP) for predicting the structure of new 2D HOIPs. Using the MACE architecture, an MLIP is trained on 86 diverse experimentally reported HOIP structures. The model is tested on 73 unseen perovskite compositions, and achieves chemical accuracy with respect to the reference electronic structure method. Our model is then combined with a simple random structure search algorithm to predict the structure of hypothetical HOIPs given only the proposed composition. Success is demonstrated by correctly and reliably recovering the crystal structure of a set of experimentally known 2D perovskites. Such a random structure search is impossible with ab initio methods due to the associated computational cost, but is relatively inexpensive with the MACE potential. Finally, the procedure is used to predict the structure formed by a new organic cation with no previously known corresponding perovskite. Laboratory synthesis of the new hybrid perovskite confirms the accuracy of our prediction. This capability, applied at scale, enables efficient screening of thousands of combinations of organic cations and inorganic layers.

cond-mat.mtrl-sci↗

arXiv2023

A foundation model for atomistic materials chemistry

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 ML force fields have largely been limited by: (i) the substantial computational and human effort of developing and validating 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 get reliable results much faster, and beginners face a lower barrier to entry. Foundation models thus represent a step towards democratising the revolution in atomic-scale modeling that has been brought about by ML force fields.

physics.chem-ph↗

arXiv2023

Improving the reliability of machine learned potentials for modeling inhomogenous liquids

The atomic-scale response of inhomogeneous fluids at interfaces and surrounding solute particles plays a critical role in governing chemical, electrochemical and biological processes at such interfaces. Classical molecular dynamics simulations have been applied extensively to simulate the response of inhomogeneous fluids directly, and as inputs to classical density functional theory, but are limited by the accuracy of the underlying empirical force fields. Here, we deploy neural network potentials (NNPs) trained to ab initio simulations to accurately predict the inhomogeneous response of two widely different fluids: liquid water and molten NaCl. Although NNPs can be readily trained to model complex bulk systems across a range of state points, in order to appropriately model a fluid's response at an interface, inhomogeneous configurations must be included in the training data. We establish protocols based on molecular dynamics simulations in external atomic potentials in order to sufficiently sample the correct configurations of inhomogeneous fluids. We show that NNPs trained to inhomogeneous fluid configurations can predict several properties such as the density response, surface tension and size-dependent cavitation free energies in water and molten NaCl corresponding to ab initio interactions more accurately than empirical force fields. This work therefore provides a first demonstration and framework for extracting the response of inhomogeneous fluids from first principles for classical density-functional treatment of fluids free from empirical potentials.

cond-mat.mtrl-sci↗