Active sparse Bayesian committee machine potential for isothermal-isobaric molecular dynamics simulations.

Recent advancements in machine learning potentials (MLPs) have significantly impacted the fields of chemistry, physics, and biology by enabling large-scale first-principles simulations. Among different machine learning approaches, kernel-based MLPs distinguish themselves through their ability to handle small datasets, quantify uncertainties, and minimize over-fitting. Nevertheless, their extensive computational requirements present considerable challenges. To alleviate these, sparsification methods have been developed, aiming to reduce computational scaling without compromising accuracy. In the context of isothermal and isobaric ML molecular dynamics (MD) simulations, achieving precise pressure estimation is crucial for reproducing reliable system behavior under constant pressure. Despite progress, sparse kernel MLPs struggle with precise pressure prediction. Here, we introduce a virial kernel function that significantly enhances the pressure estimation accuracy of MLPs. Additionally, we propose the active sparse Bayesian committee machine (BCM) potential, an on-the-fly MLP architecture that aggregates local sparse Gaussian process regression (SGPR) MLPs. The sparse BCM potential overcomes the steep computational scaling with the kernel size, and a predefined restriction on the size of kernel allows for fast and efficient on-the-fly training. Our advancements facilitate accurate and computationally efficient machine learning-enhanced MD (MLMD) simulations across diverse systems, including ice-liquid coexisting phases, Li10Ge(PS6)2 lithium solid electrolyte, and high-pressure liquid boron nitride.

Machine Learning of First-Principles Force-Fields for Alkane and Polyene Hydrocarbons.

Machine learning (ML) interatomic potentials (ML-IAPs) are generated for alkane and polyene hydrocarbons using on-the-fly adaptive sampling and a sparse Gaussian process regression (SGPR) algorithm. The ML model is generated based on the PBE+D3 level of density functional theory (DFT) with molecular dynamics (MD) for small alkane and polyene molecules. Intermolecular interactions are also trained with clusters and condensed phases of small molecules. It shows excellent transferability to long alkanes and closely describes the ab inito potential energy surface for polyenes. Simulation of liquid ethane also shows reasonable agreement with experimental reports. This is a promising initiative toward a universal ab initio quality force-field for hydrocarbons and organic molecules.

Anisotropic and amphoteric characteristics of diverse carbenes.

Despite its key importance in carbene chemistry, the amphoteric (i.e., both nucleophilic and electrophilic) behavior of the divalent carbon atom (:C) in carbenes is not well understood. The electrostatic potential (EP) around :C is often incorrectly described by simple isotropic atomic charges (particularly, as in singlet CF2); therefore, it should be described by the multipole model, which can illustrate both negative and positive EPs, favoring the positively and negatively charged species that are often present around :C. This amphotericity is much stronger in the singlet state, which has more conspicuous anisotropic charge distribution than the triplet state; this is validated by the complexation structures of carbenes interacting with Na+, Cl-, H2O, and Ag+. From the study of diverse carbenes [including CH2, CLi2/CNa2, CBe2/CMg2, CF2/CCl2, C(BH2)2/C(AlH2)2, C(CH3)2/C(SiH3)2, C(NH2)2/C(PH2)2, cyclic systems of C(CH2)2/C(CH)2, C(BHCH)2, C(CH2CH)2/C(CHCH)2, and C(NHCH)2/C(NCH)2], we elucidate the relationships between the electron configurations, electron accepting/donating strengths of atoms attached to :C, π conjugation, singlet-triplet energy gaps, anisotropic hard wall radii, anisotropic electrostatic potentials, and amphotericities of carbenes, which are vital to carbene chemistry. The (σ2, π2 or σπ) electronic configuration associated with :C on the :CA2 plane (where A is an adjacent atom) in singlet and triplet carbenes largely governs the amphoteric behaviors along the :C tip and :C face-on directions. The :C tip and :C face-on sites of σ2 singlet carbenes tend to show negative and positive EPs, favoring nucleophiles and electrophiles, respectively; meanwhile, those of π2 singlet carbenes, such as very highly π-conjugated 5-membered cyclic C(NCH)2, tend to show the opposite behavior. Open-shell σπ singlet (such as highly π-conjugated 5-membered cyclic C(CHCH)2) and triplet carbenes show less anisotropic and amphoteric behaviors.

Remarkably Enhanced Lattice Oxygen Participation in Perovskites to Boost Oxygen Evolution Reaction.

Enhancing the participation of the lattice oxygen mechanism (LOM) in several perovskites to significantly boost the oxygen evolution reaction (OER) is daunting. With the rapid decline in fossil fuels, energy research is turning toward water splitting to produce usable hydrogen by significantly reducing overpotential for other half-cells' OER. Recent studies have shown that in addition to the conventional adsorbate evolution mechanism (AEM), participation of LOM can overcome their prevalent scaling relationship limitations. Here, we report the acid treatment strategy and bypass the cation/anion doping strategy to significantly enhance LOM participation. Our perovskite demonstrated a current density of 10 mA cm-2 at an overpotential of 380 mV and a low Tafel slope (65 mV dec-1) much lower than IrO2 (73 mV dec-1). We propose that the presence of nitric acid-induced defects regulates the electronic structure and thereby lowers oxygen binding energy, allowing enhanced LOM participation to boost OER significantly.

Sparse Gaussian Process Regression-Based Machine Learned First-Principles Force-Fields for Saturated, Olefinic, and Aromatic Hydrocarbons.

Universal machine learning (ML) interatomic potentials (IAPs) for saturated, olefinic, and aromatic hydrocarbons are generated by using the Sparse Gaussian process regression algorithm. The universal potentials are obtained by combining the potentials for the previously trained alkane/polyene systems and the potentials generated with the presently trained cyclic/aromatic hydrocarbon systems, along with the newly trained cross-terms between the two systems. The ML-IAPs have been trained using the PBE + D3 level of density functional theory for the on-the-fly adaptive sampling of various hydrocarbon molecules and these clusters composed of small molecules. We tested the ML-IAPs and found that they correctly predicted the structures and energies of the ß-carotene monomer and dimer. Also, the simulations of liquid ethylene reproduced the molecular volume and the simulations of toluene crystals reproduced higher stability of the α-phase over the ß-phase. These ab initio-level force-fields could eventually evolve toward universal organic/polymeric/biomolecular systems.

Fast atomic structure optimization with on-the-fly sparse Gaussian process potentials.

We apply on-the-fly machine learning potentials (MLPs) using the sparse Gaussian process regression (SGPR) algorithm for fast optimization of atomic structures. Great acceleration is achieved even in the context of a single local optimization. Although for finding the exact local minimum, due to limited accuracy of MLPs, switching to another algorithm may be needed. For random gold clusters, the forces are reduced to ∼0.1 eV Å-1within less than ten first-principles (FP) calculations. Because of highly transferable MLPs, this algorithm is specially suitable for global optimization methods such as random or evolutionary structure searching or basin hopping. This is demonstrated by sequential optimization of random gold clusters for which, after only a few optimizations, FP calculations were rarely needed.

Reduced Potential Barrier of Sodium-Substituted Disordered Rocksalt Cathode for Oxygen Evolution Electrocatalysts.

Cation-disordered rocksalt (DRX) cathodes have been viewed as next-generation high-energy density materials surpassing conventional layered cathodes for lithium-ion battery (LIB) technology. Utilizing the opportunity of a better cation mixing facility in DRX, we synthesize Na-doped DRX as an efficient electrocatalyst toward oxygen evolution reaction (OER). This novel OER electrocatalyst generates a current density of 10 mA cm−2 at an overpotential (η) of 270 mV, Tafel slope of 67.5 mV dec−1, and long-term stability >5.5 days' superior to benchmark IrO2 (η = 330 mV with Tafel slope = 74.8 mV dec−1). This superior electrochemical behavior is well supported by experiment and sparse Gaussian process potential (SGPP) machine learning-based search for minimum energy structure. Moreover, as oxygen binding energy (OBE) on the surface closely relates to OER activity, our density functional theory (DFT) calculations reveal that Na-doping assists in facile O2 evolution (OBE = 5.45 eV) compared with pristine-DRX (6.51 eV).

Universal Machine Learning Interatomic Potentials: Surveying Solid Electrolytes.

We apply ab initio molecular dynamics (AIMD) with on-the-fly machine learning (ML) of interatomic potentials using the sparse Gaussian process regression (SGPR) algorithm for a survey of Li diffusivity in hundreds of ternary crystals as potential electrolytes for all-solid-state batteries. We show that models generated for these crystals can be easily combined for creating more general and transferable models which can potentially be used for simulating new materials without further training. As examples, universal potentials are created for Li-P-S and Li-Sb-S systems by combining the expert models of the crystals which contained the same set of elements. We also show that combinatorial models of different ternary crystals can be directly applied for modeling composite quaternary ones (e.g., Li-Ge-P-S). This hierarchical approach paves the way for modeling large-scale complexity by a combinatorial approach.

Two Liquid-Liquid Phase Transitions in Confined Water Nanofilms.

The stories behind supercooled bulk and confined water can be different. Bulk water has a metastable liquid-liquid phase transition at deeply supercooled conditions, but the existence of such a phenomenon in confined water is in question. Herein we show simulation results of first-order phase transitions between high- and low-density liquid (HDL and LDL) in confined water in both positive and negative pressures. A mid-density state between these two local states appears, which lets the transition show the hysteresis loop with transiently stable intermediate states. On the basis of Landau theory that we have adapted for mixing of moieties with high- and low-density states, we explain the phase transitions with the order parameter-dependent free energy change which is governed by second- to higher-order interactions among those moieties.

Supra-Binary Polarization in a Ferroelectric Nanowire.

The prediction and observation of supra-binary polarization in a ferroelectric nanowire (FNW) covered with a semicylindrical gate that provides an anisotropic electric field in the FNW are reported. There are gate-voltage-driven transitions between four polarization states in the FNW's cross-section, dubbed vertical-up, vertical-down, radial-in, and radial-out. They are determined by the interplay between the spatial depolarization energy and the free energy induced by an anisotropic external electric field, in clear distinction from the conventional film-based binary ferroelectricity. When the FNW is mounted on a biased graphene nanoribbon (GNR), these transitions induce exotic current-voltage hysteresis in the FNW-GNR transistor. This discovery suggests new operating mechanisms of ferroelectric devices. In particular, it enables intrinsic quaternary-digit information manipulation in parallel to the bit manipulation employed in conventional data storage.

First-order and continuous melting transitions in two-dimensional Lennard-Jones systems and repulsive disks.

In two-dimensional Lennard-Jones (LJ) systems, a small interval of melting-mode switching occurs below which the melting occurs by first-order phase transitions in lieu of the melting scenario proposed by Kosterlitz, Thouless, Halperin, Nelson, and Young (KTHNY). The extrapolated upper bound for phase coexistence is at density ρ∼0.893 and temperature T∼1.1, both in reduced LJ units. The two-stage KTHNY scenario is restored at higher temperatures, and the isothermal melting scenario is universal. The solid-hexatic and hexatic-liquid transitions in KTHNY theory, even so continuous, are distinct from typical continuous phase transitions in that instead of scale-free fluctuations, they are characterized by unbinding of topological defects, resulting in a special form of divergence of the correlation length: ξ≈exp(b|T-T_{c}|^{-ν}). Here such a divergence is firmly established for a two-dimensional melting phenomenon, providing a conclusive proof of the KTHNY melting. We explicitly confirm that this high-temperature melting behavior of the LJ system is consistent with the melting behavior of the r^{-12} potential and that melting of the r^{-n} potential is KTHNY-like for n≤12 but melting of the r^{-64} potential is first order; similar to hard disks. Therefore we suggest that the melting scenario of these repulsive potentials becomes hard-disk-like for an exponent in the range 12

Graphene and Graphene Analogs toward Optical, Electronic, Spintronic, Green-Chemical, Energy-Material, Sensing, and Medical Applications.

This spotlight discusses intriguing properties and diverse applications of graphene (Gr) and Gr analogs. Gr has brought us two-dimensional (2D) chemistry with its exotic 2D features of density of states. Yet, some of the 2D or 2D-like features can be seen on surfaces and at interfaces of bulk materials. The substrate on Gr and functionalization of Gr (including metal decoration, intercalation, doping, and hybridization) modify the unique 2D features of Gr. Despite abundant literature on physical properties and well-known applications of Gr, spotlight works based on the conceptual understanding of the 2D physical and chemical nature of Gr toward vast-ranging applications are hardly found. Here we focus on applications of Gr, based on conceptual understanding of 2D phenomena toward 2D chemistry. Thus, 2D features, defects, edges, and substrate effects of Gr are discussed first. Then, to pattern Gr electronic circuits, insight into differentiating conducting and nonconducting regions is introduced. By utilizing the unique ballistic electron transport properties and edge spin states of Gr, Gr nanoribbons (GNRs) are exploited for the design of ultrasensitive molecular sensing electronic devices (including molecular fingerprinting) and spintronic devices. The highly stable nature of Gr can be utilized for protection of corrosive metals, moisture-sensitive perovskite solar cells, and highly environment-susceptible topological insulators (TIs). Gr analogs have become new types of 2D materials having novel features such as half-metals, TIs, and nonlinear optical properties. The key insights into the functionalized Gr hybrid materials lead to the applications for not only energy storage and electrochemical catalysis, green chemistry, and electronic/spintronic devices but also biosensing and medical applications. All these topics are discussed here with the focus on conceptual understanding. Further possible applications of Gr, GNRs, and Gr analogs are also addressed in a section on outlook and future challenges.