@article{21381, abstract = {The lack of long-range electrostatics is a key limitation of modern machine learning interatomic potentials (MLIPs), hindering reliable applications to interfaces, charge-transfer reactions, polar and ionic materials, and biomolecules. In this Perspective, we distill two design principles behind the Latent Ewald Summation framework, which can capture long-range interactions, charges, and electrical response just by learning from standard energy and force training data: (i) use a Coulomb functional form with environment-dependent charges to capture electrostatic interactions, and (ii) avoid explicit training on ambiguous density functional theory partial charges. When both principles are satisfied, substantial flexibility remains: essentially any short-range MLIP can be augmented; charge equilibration schemes can be added when desired; dipoles and Born effective charges can be inferred or fine-tuned; and charge/spin-state embeddings or tensorial targets can be further incorporated. We also discuss current limitations and open challenges. Together, these minimal, physics-guided design rules suggest that incorporating long-range electrostatics into MLIPs is simpler and perhaps more broadly applicable than is commonly assumed.}, author = {Kim, Dongjin and Cheng, Bingqing}, issn = {1089-7690}, journal = {The Journal of Chemical Physics}, number = {6}, publisher = {AIP Publishing}, title = {{Long-range electrostatics for machine learning interatomic potentials is easier than we thought}}, doi = {10.1063/5.0316886}, volume = {164}, year = {2026}, } @article{18820, abstract = {Feature selection is essential in the analysis of molecular systems and many other fields, but several uncertainties remain: What is the optimal number of features for a simplified, interpretable model that retains essential information? How should features with different units be aligned, and how should their relative importance be weighted? Here, we introduce the Differentiable Information Imbalance (DII), an automated method to rank information content between sets of features. Using distances in a ground truth feature space, DII identifies a low-dimensional subset of features that best preserves these relationships. Each feature is scaled by a weight, which is optimized by minimizing the DII through gradient descent. This allows simultaneously performing unit alignment and relative importance scaling, while preserving interpretability. DII can also produce sparse solutions and determine the optimal size of the reduced feature space. We demonstrate the usefulness of this approach on two benchmark molecular problems: (1) identifying collective variables that describe conformations of a biomolecule, and (2) selecting features for training a machine-learning force field. These results show the potential of DII in addressing feature selection challenges and optimizing dimensionality in various applications. The method is available in the Python library DADApy.}, author = {Wild, Romina and Wodaczek, Felix and Del Tatto, Vittorio and Cheng, Bingqing and Laio, Alessandro}, issn = {2041-1723}, journal = {Nature Communications}, publisher = {Springer Nature}, title = {{Automatic feature selection and weighting in molecular systems using Differentiable Information Imbalance}}, doi = {10.1038/s41467-024-55449-7}, volume = {16}, year = {2025}, } @article{20926, abstract = {Most current machine learning interatomic potentials (MLIPs) rely on short-range approximations, without explicit treatment of long-range electrostatics. To address this, we recently developed the Latent Ewald Summation (LES) method, which infers electrostatic interactions, polarization, and Born effective charges (BECs), just by learning from energy and force training data. Here, we present LES as a standalone library, compatible with any short-range MLIP, and demonstrate its integration with methods such as MACE, NequIP, Allegro, CACE, CHGNet, and UMA. We benchmark LES-enhanced models on distinct systems, including bulk water, polar dipeptides, and gold dimer adsorption on defective substrates, and show that LES not only captures correct electrostatics but also improves accuracy. Additionally, we scale LES to large and chemically diverse data by training MACELES-OFF on the SPICE set containing molecules and clusters, making a universal MLIP with electrostatics for organic systems, including biomolecules. MACELES-OFF is more accurate than its short-range counterpart (MACE-OFF) trained on the same data set, predicts dipoles and BECs reliably, and has better descriptions of bulk liquids. By enabling efficient long-range electrostatics without directly training on electrical properties, LES paves the way for electrostatic foundation MLIPs.}, author = {Kim, Dongjin and Wang, Xiaoyu and Vargas, Santiago and Zhong, Peichen and King, Daniel S. and Inizan, Theo Jaffrelot and Cheng, Bingqing}, issn = {1549-9626}, journal = {Journal of Chemical Theory and Computation}, number = {24}, pages = {12709--12724}, publisher = {American Chemical Society}, title = {{A universal augmentation framework for long-range electrostatics in machine learning interatomic potentials}}, doi = {10.1021/acs.jctc.5c01400}, volume = {21}, year = {2025}, } @article{20990, abstract = {Modeling the response of material and chemical systems to electric fields remains a longstanding challenge. Machine learning interatomic potentials (MLIPs) offer an efficient and scalable alternative to quantum mechanical methods, but do not by themselves incorporate electrical response. Here, we show that polarization and Born effective charge (BEC) tensors can be directly extracted from long-range MLIPs within the Latent Ewald Summation (LES) framework, solely by learning from energy and force data. Using this approach, we predict the infrared spectra of bulk water under zero or finite external electric fields, ionic conductivities of high-pressure superionic ice, and the phase transition and hysteresis in ferroelectric PbTiO3 perovskite. This work thus extends the capability of MLIPs to predict electrical response –without training on charges or polarization or BECs– and enables accurate modeling of electric-field-driven processes in diverse systems at scale.}, author = {Zhong, Peichen and Kim, Dongjin and King, Daniel S. and Cheng, Bingqing}, issn = {2057-3960}, journal = {npj Computational Materials}, publisher = {Springer Nature}, title = {{Machine learning interatomic potential can infer electrical response}}, doi = {10.1038/s41524-025-01911-z}, volume = {11}, year = {2025}, } @article{20492, abstract = {The glassy thermal conductivities observed in crystalline inorganic perovskites such as Cs3Bi2I6Cl3 are perplexing and lacking theoretical explanations. Here, we first experimentally measure its thermal transport behavior from 20 to 300 K, after synthesizing Cs3Bi2I6Cl3 single crystals. Using path-integral molecular dynamics simulations driven by machine learning potentials, we reveal that Cs3Bi2I6Cl3 has large lattice distortions at low temperatures, which may be related to the large atomic size mismatch. Employing the Wigner formulation of thermal transport, we reproduce theexperimental thermal conductivities based on lattice-distorted structures. This studythus provides a framework for predicting and understanding glassy thermal transportin materials with strong lattice disorder.}, author = {Zeng, Zezhu and Fan, Zheyong and Simoncelli, Michele and Chen, Chen and Liang, Ting and Chen, Yue and Thornton, Geoff and Cheng, Bingqing}, issn = {1091-6490}, journal = {Proceedings of the National Academy of Sciences}, number = {41}, pages = {e2415664122}, publisher = {National Academy of Sciences}, title = {{Lattice distortion leads to glassy thermal transport in crystalline Cs3Bi2I6Cl3}}, doi = {10.1073/pnas.2415664122}, volume = {122}, year = {2025}, } @article{20011, abstract = {Heat transport in glasses over a wide temperature range is critical for applications in gate dielectrics and thermal insulators but remains poorly understood due to the challenges in modeling vibrational anharmonicity and configurational dynamics across the glass transition. Recent predictions show an unusual decrease in thermal conductivity (κ) with temperature in amorphous hafnia (a-HfO2), contrasting with the typical trend in glasses. Using molecular dynamics with a machine-learning-based neuroevolution potential, we compute κ of a-HfO2 from 50 K to 2000 K. At low temperatures, the Wigner transport equation captures both anharmonicity and quantum statistics. Above 1200 K, atomic diffusion invalidates the quasiparticle picture, and we resort to the Green–Kubo method to capture convective transport. We further extend the Wigner transport equation to supercooled a-HfO2, revealing the crucial role of low-frequency modes in facilitating heat transport. The computed κ, based on both Green–Kubo and Wigner transport theories, increases continuously with temperature up to 2000 K.}, author = {Zeng, Zezhu and Liang, Xia and Fan, Zheyong and Chen, Yue and Simoncelli, Michele and Cheng, Bingqing}, issn = {2639-4979}, journal = {ACS Materials Letters}, pages = {2695--2701}, publisher = {American Chemical Society}, title = {{Thermal transport of amorphous hafnia across the glass transition}}, doi = {10.1021/acsmaterialslett.5c00263}, year = {2025}, } @article{20452, abstract = {Accurate modeling of long-range forces is critical in atomistic simulations, as they play a central role in determining the properties of material and chemical systems. However, standard machine learning interatomic potentials (MLIPs) often rely on short-range approximations, limiting their applicability to systems with significant electrostatics and dispersion forces. We recently introduced the Latent Ewald Summation (LES) method, which captures long-range electrostatics without explicitly learning atomic charges or charge equilibration. We benchmark LES on diverse and challenging systems, including charged molecules, ionic liquids, electrolyte solutions, polar dipeptides, surface adsorption, electrolyte/solid interfaces, and solid-solid interfaces. Here we show that LES can reproduce the exact atomic charges for classical systems with fixed charges and can infer dipole and quadrupole moments, as well as the dipole derivative with respect to atomic positions, for quantum mechanical systems. Moreover, LES can achieve better accuracy in energy and force predictions compared to methods that explicitly learn from charges.}, author = {King, Daniel S. and Kim, Dongjin and Zhong, Peichen and Cheng, Bingqing}, issn = {2041-1723}, journal = {Nature Communications}, publisher = {Springer Nature}, title = {{Machine learning of charges and long-range interactions from energies and forces}}, doi = {10.1038/s41467-025-63852-x}, volume = {16}, year = {2025}, } @article{19495, abstract = {Machine learning interatomic potentials (MLIPs) often neglect long-range interactions, such as electrostatic and dispersion forces. In this work, we introduce a straightforward and efficient method to account for long-range interactions by learning a hidden variable from local atomic descriptors and applying an Ewald summation to this variable. We demonstrate that in systems including charged and polar molecular dimers, bulk water, and water-vapor interface, standard short-ranged MLIPs can lead to unphysical predictions even when employing message passing. The long-range models effectively eliminate these artifacts, with only about twice the computational cost of short-range MLIPs.}, author = {Cheng, Bingqing}, issn = {2057-3960}, journal = {npj Computational Materials}, publisher = {Springer Nature}, title = {{Latent Ewald summation for machine learning of long-range interactions}}, doi = {10.1038/s41524-025-01577-7}, volume = {11}, year = {2025}, } @article{20702, abstract = {Qualitative and quantitative orbital properties such as bonding/antibonding character, localization, and orbital energies are critical to how chemists understand reactivity, catalysis, and excited-state behavior. Despite this, representations of orbitals in deep learning models have been very underdeveloped relative to representations of molecular geometries and Hamiltonians. Here, we apply state-of-the-art equivariant deep learning architectures to the task of assigning global labels to orbitals, namely energies characterizations, given the molecular coefficients from Hartree–Fock or density functional theory. The architecture we have developed, the Cartesian Equivariant Orbital Network (CEONET), shows how molecular orbital coefficients are readily featurized as equivariant node features common to all graph-based machine-learned potentials. We find that CEONET performs well at predicting difficult quantitative labels such as the orbital energy and orbital entropy. Furthermore, we find that the CEONET representation provides an intuitive latent space for differentiating orbital character for the qualitative assignment of e.g. bonding or antibonding character. In addition to providing a useful representation for further integrating deep learning with electronic structure theory, we expect CEONET to be useful for automatizing and interpreting the results of advanced electronic structure methods such as complete active space self-consistent field theory. In particular, the ability of CEONET to infer multireference character via the orbital entropy paves the way toward the machine-learned selection of active spaces.}, author = {King, Daniel S. and Grzenda, Daniel and Zhu, Ray and Hudson, Nathaniel and Foster, Ian and Cheng, Bingqing and Gagliardi, Laura}, issn = {1091-6490}, journal = {Proceedings of the National Academy of Sciences}, number = {48}, publisher = {National Academy of Sciences}, title = {{Cartesian equivariant representations for learning and understanding molecular orbitals}}, doi = {10.1073/pnas.2510235122}, volume = {122}, year = {2025}, } @article{20704, abstract = {Generative models have advanced significantly in sampling material systems with continuous variables, such as atomistic structures. However, their application to discrete variables, like atom types or spin states, remains underexplored. In this work, we introduce a discrete flow matching model, tailored for systems with discrete phase-space coordinates (e.g., the Ising model or a multicomponent system on a lattice). This approach enables a single model to sample free energy surfaces over a wide temperature range with minimal training overhead, and the model generation is scalable to larger lattice sizes than those in the training set. We demonstrate our approach on the 2D Ising model, showing efficient and reliable free energy sampling. These results highlight the potential of flow matching for low-cost, scalable free energy sampling in discrete systems and suggest promising extensions to alchemical degrees of freedom in crystalline materials. The codebase developed for this work is openly available at https://github.com/tuoping/alchemicalFES.}, author = {Tuo, Ping and Zeng, Zezhu and Chen, Jiale and Cheng, Bingqing}, issn = {1549-9626}, journal = {Journal of Chemical Theory and Computation}, number = {22}, pages = {11427--11435}, publisher = {American Chemical Society}, title = {{Scalable multitemperature free energy sampling of classical Ising spin states}}, doi = {10.1021/acs.jctc.5c01248}, volume = {21}, year = {2025}, } @article{15311, abstract = {Materials with low thermal conductivity usually have complex crystal structures. Herein we experimentally find that a simple crystal structure material AgTlI2 (I4/mcm) owns an extremely low thermal conductivity of 0.25 W/mK at room temperature. To understand this anomaly, we perform in-depth theoretical studies based on ab initio molecular dynamics simulations and anharmonic lattice dynamics. We find that the unique atomic arrangement and weak chemical bonding provide a permissive environment for strong oscillations of Ag atoms, leading to a considerable rattling behaviour and giant lattice anharmonicity. This feature is also verified by the experimental probability density function refinement of single-crystal diffraction. The particularly strong anharmonicity breaks down the conventional phonon gas model, giving rise to non-negligible wavelike phonon behaviours in AgTlI2 at 300 K. Intriguingly, unlike many strongly anharmonic materials where a small propagative thermal conductivity is often accompanied by a large diffusive thermal conductivity, we find an unusual coexistence of ultralow propagative and diffusive thermal conductivities in AgTlI2 based on the thermal transport unified theory. This study underscores the potential of simple crystal structures in achieving low thermal conductivity and encourages further experimental research to enrich the family of materials with ultralow thermal conductivity.}, author = {Zeng, Zezhu and Shen, Xingchen and Cheng, Ruihuan and Perez, Olivier and Ouyang, Niuchang and Fan, Zheyong and Lemoine, Pierric and Raveau, Bernard and Guilmeau, Emmanuel and Chen, Yue}, issn = {2041-1723}, journal = {Nature Communications}, publisher = {Springer Nature}, title = {{Pushing thermal conductivity to its lower limit in crystals with simple structures}}, doi = {10.1038/s41467-024-46799-3}, volume = {15}, year = {2024}, } @article{15052, abstract = {Substrate induces mechanical strain on perovskite devices, which can result in alterations to its lattice dynamics and thermal transport. Herein, we have performed a theoretical investigation on the anharmonic lattice dynamics and thermal property of perovskite Rb2SnBr6 and Cs2SnBr6 under strains using perturbation theory up to the fourth-order terms and the unified thermal transport theory. We demonstrate a pronounced hardening of low-frequency optical phonons as temperature increases, indicating strong lattice anharmonicity and the necessity of adopting temperature-dependent interatomic force constants in the lattice thermal conductivity ( κL) calculations. It is found that the low-lying optical phonon modes of Rb2SnBr6 are extremely soft and their phonon energies are almost strain independent, which ultimately lead to a lower κL and a weaker strain dependence than Cs2SnBr6. We further reveal that the strain dependence of these phonon modes in the A2XB6-type perovskites weakens as their ibrational frequency decreases. This study deepens the understanding of lattice thermal transport in perovskites A2XB6 and provides a perspective on the selection of materials that meet the expected thermal behaviors in practical applications.}, author = {Cheng, Ruihuan and Zeng, Zezhu and Wang, Chen and Ouyang, Niuchang and Chen, Yue}, issn = {2469-9969}, journal = {Physical Review B}, number = {5}, publisher = {American Physical Society}, title = {{Impact of strain-insensitive low-frequency phonon modes on lattice thermal transport in AxXB6-type perovskites}}, doi = {10.1103/physrevb.109.054305}, volume = {109}, year = {2024}, } @article{18952, abstract = {A seventh blind test of crystal structure prediction was organized by the Cambridge Crystallographic Data Centre featuring seven target systems of varying complexity: a silicon and iodine-containing molecule, a copper coordination complex, a near-rigid molecule, a cocrystal, a polymorphic small agrochemical, a highly flexible polymorphic drug candidate, and a polymorphic morpholine salt. In this first of two parts focusing on structure generation methods, many crystal structure prediction (CSP) methods performed well for the small but flexible agrochemical compound, successfully reproducing the experimentally observed crystal structures, while few groups were successful for the systems of higher complexity. A powder X-ray diffraction (PXRD) assisted exercise demonstrated the use of CSP in successfully determining a crystal structure from a low-quality PXRD pattern. The use of CSP in the prediction of likely cocrystal stoichiometry was also explored, demonstrating multiple possible approaches. Crystallographic disorder emerged as an important theme throughout the test as both a challenge for analysis and a major achievement where two groups blindly predicted the existence of disorder for the first time. Additionally, large-scale comparisons of the sets of predicted crystal structures also showed that some methods yield sets that largely contain the same crystal structures.}, author = {Hunnisett, Lily M. and Nyman, Jonas and Francia, Nicholas and Abraham, Nathan S. and Adjiman, Claire S. and Aitipamula, Srinivasulu and Alkhidir, Tamador and Almehairbi, Mubarak and Anelli, Andrea and Anstine, Dylan M. and Anthony, John E. and Arnold, Joseph E. and Bahrami, Faezeh and Bellucci, Michael A. and Bhardwaj, Rajni M. and Bier, Imanuel and Bis, Joanna A. and Boese, A. Daniel and Bowskill, David H. and Bramley, James and Brandenburg, Jan Gerit and Braun, Doris E. and Butler, Patrick W. V. and Cadden, Joseph and Carino, Stephen and Chan, Eric J. and Chang, Chao and Cheng, Bingqing and Clarke, Sarah M. and Coles, Simon J. and Cooper, Richard I. and Couch, Ricky and Cuadrado, Ramon and Darden, Tom and Day, Graeme M. and Dietrich, Hanno and Ding, Yiming and DiPasquale, Antonio and Dhokale, Bhausaheb and van Eijck, Bouke P. and Elsegood, Mark R. J. and Firaha, Dzmitry and Fu, Wenbo and Fukuzawa, Kaori and Glover, Joseph and Goto, Hitoshi and Greenwell, Chandler and Guo, Rui and Harter, Jürgen and Helfferich, Julian and Hofmann, Detlef W. M. and Hoja, Johannes and Hone, John and Hong, Richard and Hutchison, Geoffrey and Ikabata, Yasuhiro and Isayev, Olexandr and Ishaque, Ommair and Jain, Varsha and Jin, Yingdi and Jing, Aling and Johnson, Erin R. and Jones, Ian and Jose, K. V. Jovan and Kabova, Elena A. and Keates, Adam and Kelly, Paul F. and Khakimov, Dmitry and Konstantinopoulos, Stefanos and Kuleshova, Liudmila N. and Li, He and Lin, Xiaolu and List, Alexander and Liu, Congcong and Liu, Yifei Michelle and Liu, Zenghui and Liu, Zhi-Pan and Lubach, Joseph W. and Marom, Noa and Maryewski, Alexander A. and Matsui, Hiroyuki and Mattei, Alessandra and Mayo, R. Alex and Melkumov, John W. and Mohamed, Sharmarke and Momenzadeh Abardeh, Zahrasadat and Muddana, Hari S. and Nakayama, Naofumi and Nayal, Kamal Singh and Neumann, Marcus A. and Nikhar, Rahul and Obata, Shigeaki and O'Connor, Dana and Oganov, Artem R. and Okuwaki, Koji and Otero-de-la-Roza, Alberto and Pantelides, Constantinos C. and Parkin, Sean and Pickard, Chris J. and Pilia, Luca and Pivina, Tatyana and Podeszwa, Rafał and Price, Alastair J. A. and Price, Louise S. and Price, Sarah L. and Probert, Michael R. and Pulido, Angeles and Ramteke, Gunjan Rajendra and Rehman, Atta Ur and Reutzel-Edens, Susan M. and Rogal, Jutta and Ross, Marta J. and Rumson, Adrian F. and Sadiq, Ghazala and Saeed, Zeinab M. and Salimi, Alireza and Salvalaglio, Matteo and Sanders de Almada, Leticia and Sasikumar, Kiran and Sekharan, Sivakumar and Shang, Cheng and Shankland, Kenneth and Shinohara, Kotaro and Shi, Baimei and Shi, Xuekun and Skillman, A. Geoffrey and Song, Hongxing and Strasser, Nina and van de Streek, Jacco and Sugden, Isaac J. and Sun, Guangxu and Szalewicz, Krzysztof and Tan, Benjamin I. and Tan, Lu and Tarczynski, Frank and Taylor, Christopher R. and Tkatchenko, Alexandre and Tom, Rithwik and Tuckerman, Mark E. and Utsumi, Yohei and Vogt-Maranto, Leslie and Weatherston, Jake and Wilkinson, Luke J. and Willacy, Robert D. and Wojtas, Lukasz and Woollam, Grahame R. and Yang, Zhuocen and Yonemochi, Etsuo and Yue, Xin and Zeng, Qun and Zhang, Yizu and Zhou, Tian and Zhou, Yunfei and Zubatyuk, Roman and Cole, Jason C.}, issn = {2052-5206}, journal = {Acta Crystallographica Section B}, number = {6}, pages = {517--547}, publisher = {International Union of Crystallography}, title = {{The seventh blind test of crystal structure prediction: Structure generation methods}}, doi = {10.1107/s2052520624007492}, volume = {80}, year = {2024}, } @article{15359, abstract = {Molecular dynamics (MD) simulations play an important role in understanding and engineering heat transport properties of complex materials. An essential requirement for reliably predicting heat transport properties is the use of accurate and efficient interatomic potentials. Recently, machine-learned potentials (MLPs) have shown great promise in providing the required accuracy for a broad range of materials. In this mini-review and tutorial, we delve into the fundamentals of heat transport, explore pertinent MD simulation methods, and survey the applications of MLPs in MD simulations of heat transport. Furthermore, we provide a step-by-step tutorial on developing MLPs for highly efficient and predictive heat transport simulations, utilizing the neuroevolution potentials as implemented in the GPUMD package. Our aim with this mini-review and tutorial is to empower researchers with valuable insights into cutting-edge methodologies that can significantly enhance the accuracy and efficiency of MD simulations for heat transport studies.}, author = {Dong, Haikuan and Shi, Yongbo and Ying, Penghua and Xu, Ke and Liang, Ting and Wang, Yanzhou and Zeng, Zezhu and Wu, Xin and Zhou, Wenjiang and Xiong, Shiyun and Chen, Shunda and Fan, Zheyong}, issn = {1089-7550}, journal = {Journal of Applied Physics}, number = {16}, publisher = {AIP Publishing}, title = {{Molecular dynamics simulations of heat transport using machine-learned potentials: A mini-review and tutorial on GPUMD with neuroevolution potentials}}, doi = {10.1063/5.0200833}, volume = {135}, year = {2024}, } @article{18452, abstract = {Diffusion models have recently emerged as powerful tools for the generation of new molecular and material structures. The key insight is that the noise in these models is related to the response of the atoms to displacement, and the denoising step is thus analogous to the geometry relaxation of atomistic systems starting from a random structure. Building on this, we present a generative method called Response Matching (RM), which leverages the fact that each stable material or molecule exists at the minimum of its potential energy surface. Any perturbation induces a response in energy and stress, driving the structure back to equilibrium. Matching this response is closely related to score matching in diffusion models. Another important aspect of state-of-the-art diffusion models is the incorporation of physical symmetries such as translation, rotation, and periodicity. RM employs a machine learning interatomic potential and random structure search as the denoising model, inherently respecting these symmetries and exploiting the locality of atomic interactions. RM handles both molecules and bulk materials under the same framework. Its efficiency and generalization are demonstrated on three systems: a small organic molecular data set, stable crystals from the Materials Project, and one-shot learning on a single diamond configuration.}, author = {Cheng, Bingqing}, issn = {1549-9626}, journal = {Journal of Chemical Theory and Computation}, number = {20}, pages = {9259--9266}, publisher = {American Chemical Society}, title = {{Response matching for generating materials and molecules}}, doi = {10.1021/acs.jctc.4c00998}, volume = {20}, year = {2024}, } @article{17322, abstract = {Machine learning interatomic potentials are revolutionizing large-scale, accurate atomistic modeling in material science and chemistry. Many potentials use atomic cluster expansion or equivariant message-passing frameworks. Such frameworks typically use spherical harmonics as angular basis functions, followed by Clebsch-Gordan contraction to maintain rotational symmetry. We propose a mathematically equivalent and simple alternative that performs all operations in the Cartesian coordinates. This approach provides a complete set of polynormially independent features of atomic environments while maintaining interaction body orders. Additionally, we integrate low-dimensional embeddings of various chemical elements, trainable radial channel coupling, and inter-atomic message passing. The resulting potential, named Cartesian Atomic Cluster Expansion (CACE), exhibits good accuracy, stability, and generalizability. We validate its performance in diverse systems, including bulk water, small molecules, and 25-element high-entropy alloys.}, author = {Cheng, Bingqing}, issn = {2057-3960}, journal = {npj Computational Materials}, publisher = {Springer Nature}, title = {{Cartesian atomic cluster expansion for machine learning interatomic potentials}}, doi = {10.1038/s41524-024-01332-4}, volume = {10}, year = {2024}, } @article{17278, abstract = {An azeotrope is a constant boiling point mixture, and its behavior is important for fluid separation processes. Predicting azeotropes from atomistic simulations is difficult due to the complexities and convergence problems of Monte Carlo and free-energy perturbation techniques. Here, we present a methodology for predicting the azeotropes of binary mixtures, which computes the compositional dependence of chemical potentials from molecular dynamics simulations using the S0 method and employs experimental boiling point and vaporization enthalpy data. Using this methodology, we reproduce the azeotropes, or lack thereof, in five case studies, including ethanol/water, ethanol/isooctane, methanol/water, hydrazine/water, and acetone/chloroform mixtures. We find that it is crucial to use the experimental boiling point and vaporization enthalpy for reliable azeotrope predictions, as empirical force fields are not accurate enough for these quantities. Finally, we use regular solution models to rationalize the azeotropes and reveal that they tend to form when the mixture components have similar boiling points and strong interactions.}, author = {Wang, Xiaoyu and Cheng, Bingqing}, issn = {1089-7690}, journal = {Journal of Chemical Physics}, number = {3}, publisher = {AIP Publishing}, title = {{Integrating molecular dynamics simulations and experimental data for azeotrope predictions in binary mixtures}}, doi = {10.1063/5.0217232}, volume = {161}, year = {2024}, } @article{12702, abstract = {Hydrocarbon mixtures are extremely abundant in the Universe, and diamond formation from them can play a crucial role in shaping the interior structure and evolution of planets. With first-principles accuracy, we first estimate the melting line of diamond, and then reveal the nature of chemical bonding in hydrocarbons at extreme conditions. We finally establish the pressure-temperature phase boundary where it is thermodynamically possible for diamond to form from hydrocarbon mixtures with different atomic fractions of carbon. Notably, here we show a depletion zone at pressures above 200 GPa and temperatures below 3000 K-3500 K where diamond formation is thermodynamically favorable regardless of the carbon atomic fraction, due to a phase separation mechanism. The cooler condition of the interior of Neptune compared to Uranus means that the former is much more likely to contain the depletion zone. Our findings can help explain the dichotomy of the two ice giants manifested by the low luminosity of Uranus, and lead to a better understanding of (exo-)planetary formation and evolution.}, author = {Cheng, Bingqing and Hamel, Sebastien and Bethkenhagen, Mandy}, issn = {2041-1723}, journal = {Nature Communications}, publisher = {Springer Nature}, title = {{Thermodynamics of diamond formation from hydrocarbon mixtures in planets}}, doi = {10.1038/s41467-023-36841-1}, volume = {14}, year = {2023}, } @article{12912, abstract = {The chemical potential of adsorbed or confined fluids provides insight into their unique thermodynamic properties and determines adsorption isotherms. However, it is often difficult to compute this quantity from atomistic simulations using existing statistical mechanical methods. We introduce a computational framework that utilizes static structure factors, thermodynamic integration, and free energy perturbation for calculating the absolute chemical potential of fluids. For demonstration, we apply the method to compute the adsorption isotherms of carbon dioxide in a metal-organic framework and water in carbon nanotubes.}, author = {Schmid, Rochus and Cheng, Bingqing}, issn = {1089-7690}, journal = {The Journal of Chemical Physics}, number = {16}, publisher = {AIP Publishing}, title = {{Computing chemical potentials of adsorbed or confined fluids}}, doi = {10.1063/5.0146711}, volume = {158}, year = {2023}, } @article{12879, abstract = {Machine learning (ML) has been widely applied to chemical property prediction, most prominently for the energies and forces in molecules and materials. The strong interest in predicting energies in particular has led to a ‘local energy’-based paradigm for modern atomistic ML models, which ensures size-extensivity and a linear scaling of computational cost with system size. However, many electronic properties (such as excitation energies or ionization energies) do not necessarily scale linearly with system size and may even be spatially localized. Using size-extensive models in these cases can lead to large errors. In this work, we explore different strategies for learning intensive and localized properties, using HOMO energies in organic molecules as a representative test case. In particular, we analyze the pooling functions that atomistic neural networks use to predict molecular properties, and suggest an orbital weighted average (OWA) approach that enables the accurate prediction of orbital energies and locations.}, author = {Chen, Ke and Kunkel, Christian and Cheng, Bingqing and Reuter, Karsten and Margraf, Johannes T.}, issn = {2041-6539}, journal = {Chemical Science}, publisher = {Royal Society of Chemistry}, title = {{Physics-inspired machine learning of localized intensive properties}}, doi = {10.1039/d3sc00841j}, year = {2023}, }