LASP: Fast global potential energy surface exploration

Machine Learning-Driven Insights for Phase-Stable FA x Cs1-x Pb(I y Br1-y )3 Perovskites in Tandem Solar Cells

The inherent chemical tunability of perovskite materials has spurred extensive research into composition engineering within the perovskite community. However, identifying the optimal composition across a broad range of variations still remains a significant challenge. Conventional trial-and-error methods are prohibitively expensive and environmentally taxing for comprehensive screening. Here, we employed machine learning-accelerated atomic simulation to guide the design of stable perovskite solar cells absorbers. Our approach entailed training of a neural network (NN) potential using data generated from first-principles calculations, yielding a perovskite NN potential exhibiting high accuracy. Utilizing this NN potential, we constructed a phase diagram for FA x Cs1-x Pb(I y Br1-y )3 (where 0 ≤ x ≤ 1 and 0 ≤ y ≤ 1, FA denotes formamidinium cation). Integrating this with a band gap diagram, we successfully identified global optimal perovskite compositions for tandem applications with 1.7 and 1.8 eV band gaps. We have identified that all FA x Cs1-x Pb(I y Br1-y )3 with >1.8 eV band gaps are thermodynamically vulnerable to phase segregation and developed a strategy to stabilize thermodynamically unstable phases by suppressing phase segregation kinetics. Finally, theoretical predictions were confirmed by the corresponding experiments. Our results suggest that creating perovskites/Si tandem solar cells with 1.7 eV FA x Cs1-x Pb(I y Br1-y )3 encounters less severe challenges in addressing phase segregation issues than perovskites/perovskites tandem solar cells with 1.8 eV FA x Cs1-x Pb(I y Br1-y )3.

De Novo Inverse Design Superhard C-N Compounds via Global Machine Learning Interatomic Potentials and Multiobjective Optimization Algorithm

A major challenge in the field of superhard materials is the identification of compounds with a hardness exceeding that of diamond. In this study, we developed a variable-composition inverse material design (VC-IMD) approach for designing C-N superhard materials. In this approach, an improved multiobjective optimization algorithm is introduced, utilizing structure similarity constraint to prevent convergence toward local minima. Combined with active learning, it trains global machine learning interatomic potentials (g-MLIPs) while exploring target materials. By comparing several g-MLIPs and selecting the best, the resulting g-MLIPs achieved reasonable precision within three iterations. Through multiple searches, 38 novel and stable C-N superhard materials not present in major computational materials databases were identified. Notably, the material C3(P6422) with a hardness of 97.4 GPa was discovered, potentially exceeding that of diamond (94.0 GPa). This approach provided a new pathway for materials design with target properties.

Data-driven discovery of Pt single atom embedded germanosilicate MFI zeolite catalysts for propane dehydrogenation

Zeolite-confined metal is an important class of heterogeneous catalysts, demonstrating exceptional catalytic performance in many reactions, but the identification of a stable metal-zeolite combination with a simple synthetic method remains a top challenge. Here artificial intelligence methods, particularly global neural network potential based large-scale atomic simulation, are utilized to design Pt-containing zeolite frameworks for propane-to-propene conversion. We show that out of the zeolite database (>220 structure framework) and more than 100,000 Pt/Ge differently distributed configurations, there are only three Ge-containing zeolites, germanosilicate (MFI, IWW and SAO) that are predicted to be capable of stabilizing Pt single atom embedded in zeolite skeleton and at the meantime allowing propane fast diffusion. Among, the Pt1@Ge-MFI catalyst is successfully synthesized via a simple one-pot synthesis without a lengthy post-treatment procedure, and characterized by high-resolution experimental techniques. We demonstrate that the catalyst features an in-situ formed [GePtO3H2] active site under the reductive reaction condition that can achieve long-term (>750 h) high activity and selectivity (98%) for propane dehydrogenation. Our simple catalyst synthesis holds promise for scale-up industrial applications that can now be rooted in first principles via data-driven catalyst design.

Unraveling the mechanisms of ketene generation and transformation in syngas-to-olefin conversion over ZnCrO x |SAPO-34 catalysts

Ketene was identified as an intermediate in syngas-to-olefin (STO) conversion catalyzed by metal oxide-zeolite composites, which sparked a hot debate regarding its formation mechanism and catalytic roles. Here, we employed large-scale atomic simulations using global neural network potentials to explore the STO reaction pathways and microkinetic simulations to couple the reaction kinetics in ZnCrO x |SAPO-34 composite sites. Our results demonstrate that the majority of ketene (86.1%) originates from the methanol carbonylation-to-ketene route via nearby zeolite acidic sites, where methanol is produced through conventional syngas-to-methanol conversion on the Zn3Cr3O8 (0001) surface, while the minority of ketene (13.9%) arises from a direct CHO*-CO* coupling pathway (CHO* + CO* + H* → CHOCO* + H* → CH2CO + O*) on Zn3Cr3O8. The presence of the ketene pathway significantly alters the catalytic performance in the zeolite, as methanol carbonylation to ketene is kinetically more efficient in competing with conventional methanol-to-olefins (MTO) conversion and thus predominantly drives the product to ethene. Based on our microkinetic simulation, it is the methanol carbonylation activity in the zeolite that dictates the performance of STO catalysts.

Fine-Tuned Global Neural Network Potentials for Global Potential Energy Surface Exploration at High Accuracy

Machine learning potential (MLP), by learning global potential energy surfaces (PES), has demonstrated its great value in finding unknown structures and reactions via global PES exploration. Due to the diversity and complexity of the global PES data set, an outstanding challenge emerges in achieving PES high accuracy (e.g., error <1 meV/atom), which is essential to determine the thermodynamics and kinetics properties. Here, we develop a lightweight fine-tuning MLP architecture, namely, AtomFT, that can explore PES globally and simultaneously describe the PES of a target system accurately. The AtomFT potential takes the pretrained many-body function corrected global neural network (MBNN) potential as the basis potential, exploits and iteratively updates the atomic features from the pretrained MBNN model, and finally generates the fine-tuning energy contribution. By implementing the AtomFT architecture on the commonly available CPU platform, we show the high efficiency of AtomFT potential in both training and inference and demonstrate the high performance in challenging PES problems, including the oxides with low defect content, molecular reactions, and molecular crystals─in all systems, the AtomFT potentials enhance significantly the PES prediction accuracy to 1 meV/atom.

Machine Learning Potential for Copper Hydride Clusters: A Neutron Diffraction-Independent Approach for Locating Hydrogen Positions

Determining hydrogen positions in metal hydride clusters remains a formidable challenge, which relies heavily on unaffordable neutron diffraction. While machine learning has shown promise, only one deep learning-based method has been proposed so far, which relies heavily on neutron diffraction data for training, limiting its general applicability. In this work, we present an innovative strategy─SSW-NN (stochastic surface walking with neural network)─a robust, non-neutron diffraction-dependent technique that accurately predicts hydrogen positions. Validated against neutron diffraction data for copper hydride clusters, SSW-NN proved effective for clusters where only X-ray diffraction data or DFT predictions are available. It offers superior accuracy, efficiency, and versatility across different metal hydrides, including silver and alloy hydride systems, currently without any neutron diffraction references. This approach not only establishes a new research paradigm for metal hydride clusters but also provides a universal solution for hydrogen localization in other research fields constrained by neutron sources.

Influence of Synthesis Conditions on the Capacitance Performance of Hydrothermally Prepared MnO2 for Carbon Xerogel-Based Solid-State Supercapacitors

In this study, the potential to modify the phase structure and morphology of manganese dioxide synthesized via the hydrothermal route was explored. A series of samples were prepared at different synthesis temperatures (100, 120, 140, and 160 °C) using KMnO4 and MnSO4·H2O as precursors. The phase composition and morphology of the materials were analyzed using various physicochemical methods. The results showed that, at the lowest synthesis temperature (100 °C), an intercalation compound with composition K1.39Mn3O6 and a very small amount of α-MnO2 was formed. At higher temperatures (120-160 °C), the amount of α-MnO2 increased, indicating the formation of two clearly distinguished crystal structures. The sample obtained at 160 °C exhibited the highest specific surface area (approximately 157 m2/g). These two-phase (α-MnO2/K1.39Mn3O6) materials, synthesized at the lowest and highest temperatures, respectively, and containing an appropriate amount of carbon xerogel, were tested as active mass for positive electrodes in a solid-state supercapacitor, using a Na+-form Aquivion® membrane as the polymer electrolyte. The electrochemical evaluation showed that the composite with the higher specific surface area, containing 75% manganese dioxide, demonstrated improved characteristics, including 96% capacitance retention after 5000 charge/discharge cycles and high energy efficiency (approximately 99%). These properties highlight its potential for application in solid-state supercapacitors.

Do Rh-Hydride Phases Contribute to the Catalytic Activity of Rh Catalysts under Reductive Conditions?

Rh-hydride phases were believed to be key causes of the exceptional catalytic ability of Rh catalysts under H2 reductive conditions. Here, we utilize the large-scale machine-learning-based global optimization to explore millions of Rh bulk, surface, and nanoparticle structures in contact with H2, which rules out the presence of subsurface/interstitial H in Rh and Rh-hydride phases as thermodynamically stable phases under ambient conditions. Instead, an exceptional Rh-H affinity is identified for surface Rh atoms in Rh nanoparticles that can accommodate a high concentration of adsorbed H, with the surface Rh to H ratio reaching ∼2.5, featuring stable six-H-coordinated Rh, [RhH6]. Such [RhH6] species forming at edged Rh sites are found to be the key intermediates in the electrochemical hydrogen evolution reaction (HER) on Rh. Guided by theory, our synthesized Rh concave nanocubes with a high density of edged Rh sites achieve a Tafel slope of 28.4 mV dec-1 and a low overpotential of 36.1 mV at jECSA = 1 mA cm-2, which outperforms commercial Pt/C and other morphologies of Rh catalysts. Our results clarify the active phase in Rh-H nanosystems and guide the catalyst design by precise morphology control of nanocatalysts.

Local-Softening Stochastic Surface Walking for Fast Exploration of Corrugated Potential Energy Surfaces

Global potential energy surface (PES) exploration provides a unique route to predict the thermodynamic and kinetic properties of unknown materials, but the task is highly challenging for systems with tight covalent bonds. Here, we develop the local-softening stochastic surface walking (LS-SSW) method for scanning corrugated PESs. LS-SSW transforms the vibrational mode space of a system by adding pairwise penalty potentials with a self-adaption mechanism, which helps to delocalize and soften the strong local modes. This allows the stochastic surface walking (SSW) method to capture more efficiently the correct local atomic movement toward nearby minima and simultaneously reduce the barrier height of reactions. As a result, the local trapping time in searching for the corrugated PES is greatly reduced. LS-SSW can be applied generally to the reaction pathway sampling and the global PES exploration of both clusters and crystals, the high efficiency of which is demonstrated in searching the reaction pathways between C4H6 isomers, finding the global minimum of carbon clusters up to 360 atoms, and constructing the global PES of Fe7C3 material.

LASP to the Future of Atomic Simulation: Intelligence and Automation

Atomic simulations aim to understand and predict complex physical phenomena, the success of which relies largely on the accuracy of the potential energy surface description and the efficiency to capture important rare events. LASP software (large-scale atomic simulation with a Neural Network Potential), released in 2018, incorporates the key ingredients to fulfill the ultimate goal of atomic simulations by combining advanced neural network potentials with efficient global optimization methods. This review introduces the recent development of the software along two main streams, namely, higher intelligence and more automation, to solve complex material and reaction problems. The latest version of LASP (LASP 3.7) features the global many-body function corrected neural network (G-MBNN) to improve the PES accuracy with low cost, which achieves a linear scaling efficiency for large-scale atomic simulations. The key functionalities of LASP are updated to incorporate (i) the ASOP and ML-interface methods for finding complex surface and interface structures under grand canonic conditions; (ii) the ML-TS and MMLPS methods to identify the lowest energy reaction pathway. With these powerful functionalities, LASP now serves as an intelligent data generator to create computational databases for end users. We exemplify the recent LASP database construction in zeolite and the metal-ligand properties for a new catalyst design.

Two-dimensional crystalline platinum oxide

Platinum (Pt) oxides are vital catalysts in numerous reactions, but research indicates that they decompose at high temperatures, limiting their use in high-temperature applications. In this study, we identify a two-dimensional (2D) crystalline Pt oxide with remarkable thermal stability (1,200 K under nitrogen dioxide) using a suite of in situ methods. This 2D Pt oxide, characterized by a honeycomb lattice of Pt atoms encased between dual oxygen layers forming a six-pointed star structure, exhibits minimized in-plane stress and enhanced vertical bonding due to its unique structure, as revealed by theoretical simulations. These features contribute to its high thermal stability. Multiscale in situ observations trace the formation of this 2D Pt oxide from α-PtO2, providing insights into its formation mechanism from the atomic to the millimetre scale. This 2D Pt oxide with outstanding thermal stability and distinct surface electronic structure subverts the previously held notion that Pt oxides do not exist at high temperatures and can also present unique catalytic capabilities. This work expands our understanding of Pt oxidation species and sheds light on the oxidative and catalytic behaviours of Pt oxide in high-temperature settings.

The seventh blind test of crystal structure prediction: structure generation methods

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.

Leveraging Machine Learning Potentials for In-Situ Searching of Active sites in Heterogeneous Catalysis

This Perspective explores the integration of machine learning potentials (MLPs) in the research of heterogeneous catalysis, focusing on their role in identifying in situ active sites and enhancing the understanding of catalytic processes. MLPs utilize extensive databases from high-throughput density functional theory (DFT) calculations to train models that predict atomic configurations, energies, and forces with near-DFT accuracy. These capabilities allow MLPs to handle significantly larger systems and extend simulation times beyond the limitations of traditional ab initio methods. Coupled with global optimization algorithms, MLPs enable systematic investigations across vast structural spaces, making substantial contributions to the modeling of catalyst surface structures under reactive conditions. The review aims to provide a broad introduction to recent advancements and practical guidance on employing MLPs and also showcases several exemplary cases of MLP-driven discoveries related to surface structure changes under reactive conditions and the nature of active sites in heterogeneous catalysis. The prevailing challenges faced by this approach are also discussed.

In Situ Surfaced Mn-Mn Dimeric Sites Dictate CO Hydrogenation Activity and C2 Selectivity over MnRh Binary Catalysts

Massive ethanol production has long been a dream of human society. Despite extensive research in past decades, only a few systems have the potential of industrialization: specifically, Mn-promoted Rh (MnRh) binary heterogeneous catalysts were shown to achieve up to 60% C2 oxygenates selectivity in converting syngas (CO/H2) to ethanol. However, the active site of the binary system has remained poorly characterized. Here, large-scale machine-learning global optimization is utilized to identify the most stable Mn phases on Rh metal surfaces under reaction conditions by exploring millions of likely structures. We demonstrate that Mn prefers the subsurface sites of Rh metal surfaces and is able to emerge onto the surface forming MnRh surface alloy once the oxidative O/OH adsorbates are present. Our machine-learning-based transition state exploration further helps to resolve automatedly the whole reaction network, including 74 elementary reactions on various MnRh surface sites, and reveals that the Mn-Mn dimeric site at the monatomic step edge is the true active site for C2 oxygenate formation. The turnover frequency of the C2 product on the Mn-Mn dimeric site at MnRh steps is at least 107 higher than that on pure Rh steps from our microkinetic simulations, with the selectivity to the C2 product being 52% at 523 K. Our results demonstrate the key catalytic role of Mn-Mn dimeric sites in allowing C-O bond cleavage and facilitating the hydrogenation of O-terminating C2 intermediates, and rule out Rh metal by itself as the active site for CO hydrogenation to C2 oxygenates.

The Transformation Mechanism of Graphite to Hexagonal Diamond under Shock Conditions

The formation of a hexagonal diamond represents one of the most intriguing questions in materials science. Under shock conditions, the graphite basal plane tends to slide and pucker to form diamond. However, how the shock strength determines the phase selectivity remains unclear. In this work, using a DFT-trained carbon global neural network model, we studied the shock-induced graphite transition. The poor sliding caused by scarce sliding time under high-strength shock leads to metastable hexagonal diamond with an orientation relationship of (001)G//(100)HD+[010]G//[010]HD, while under low-strength shock due to long sliding distance cubic diamond forms with the orientation (001)G//(111)CD+[100]G//[110]CD, unveiling the strength-dependent graphite transition mechanism. We for the first time provide computational evidence of the strength-dependent graphite transition from first-principles, clarifying the long-term unresolved shock-induced hexagonal diamond formation mechanism and the structural source of the strength-dependent trend, which facilitates the hexagonal diamond synthesis via controlled experiment.

Structural transitions in liquid semiconductor alloys: A molecular dynamics study with a neural network potential

Liquid-liquid phase transitions hold a unique and profound significance within condensed matter physics. These transitions, while conceptually intriguing, often pose formidable computational challenges. However, recent advances in neural network (NN) potentials offer a promising avenue to effectively address these challenges. In this paper, we delve into the structural transitions of liquid CdTe, CdS, and their alloy systems using molecular dynamics simulations, harnessing the power of an NN potential named LaspNN. Our investigations encompass both pressure and temperature effects. Through our simulations, we uncover three primary liquid structures around melting points that emerge as pressure increases: tetrahedral, rock salt, and close-packed structures, which greatly resemble those of solid states. In the high-temperature regime, we observe the formation of Te chains and S dimers, providing a deeper understanding of the liquid's atomic arrangements. When examining CdSxTe1-x alloys, our findings indicate that a small substitution of S by Te atoms for S-rich alloys (x > 0.5) exhibits a structural transition much different from CdS, while a large substitution of Te by S atoms for Te-rich alloys (x < 0.5) barely exhibits a structural transition similar to CdTe. We construct a schematic diagram for liquid alloys that considers both temperature and pressure, providing a comprehensive overview of the alloy system's behavior. The local aggregation of Te atoms demonstrates a linear relationship with alloy composition x, whereas that of S atoms exhibits a nonlinear one, shedding light on the composition-dependent structural changes.

Stability and Dynamics of Zeolite-Confined Gold Nanoclusters

Nanoengineered metal@zeolite materials have recently emerged as a promising class of catalysts for several industrially relevant reactions. These materials, which consist of small transition metal nanoclusters confined within three-dimensional zeolite pores, are interesting because they show higher stability and better sintering resistance under reaction conditions. While several such hybrid catalysts have been reported experimentally, key questions such as the impact of the zeolite frameworks on the properties of the metal clusters are not well understood. To address such knowledge gaps, in this study, we report a robust and transferable machine learning-based potential (MLP) that is capable of describing the structure, stability, and dynamics of zeolite-confined gold nanoclusters. Specifically, we show that the resulting MLP maintains ab initio accuracy across a range of temperatures (300-1000 K) and can be used to investigate time scales (>10 ns), length scales (ca. 10,000 atoms), and phenomena (e.g., ensemble-averaged stability and diffusivity) that are typically inaccessible using density functional theory (DFT). Taken together, this study represents an important step in enabling the rational theory-guided design of metal@zeolite catalysts.

A novel sp3 carbon allotrope with 4+5+6+7+8 odd-even ring

In this study, we report a novel monoclinic phase of carbon that contains 4+5+6+7+8 member rings in P21/m symmetry, identified by applying the stochastic surface walking method combined with high dimensional neural network potentials. We demonstrate that this phase possesses lower energy than graphite above 21.5 GPa. The phonon spectra show that this structure is stable under ambient pressure. This phase is a super hard material with a shear hardness as high as 81.9 GPa while it possesses an indirect band gap of 3.16 eV. The energy barrier of graphite to the Y phase is 0.27 eV, slightly higher than that of the hexagonal diamond (0.21 eV) in a similar phase transition mechanism. Two types of thermodynamically stable interfaces can be formed with the hexagonal diamond (HD), namely (001)Y//(100)HD, [100]Y//[010]HD and (001)Y//(001)HD, [010]Y//[001]HD. Although the discrete bulk Y phase is hard to synthesize, a faulted structure between HD is possible because of the well-matched interface between Y and HD. Our work shows that the Y phase may be formed in some special conditions and enhances our understanding of the formation of novel carbon allotropes.

Global Optimization of Cation Ordering in Perovskites by Recommendation-Based Basin-Hopping

Cation ordering in multication perovskites is related to many important material properties and performances, but computational determination of the cation ordering remains a major challenge. Here, we propose a new computational approach by introducing a machine learning recommender system into the basin-hopping framework (RBH) for optimizing cation ordering. Taking the electrocatalyst Ba0.5Sr0.5Co0.8Fe0.2O3 (BSCF5582) as a showcase example, we found that the efficiency of RBH in identifying low-energy configurations outperforms the methods of cluster expansion and conventional basin-hopping. The RBH results revealed that the BSCF5582 catalyst tended to have a layered ordering of A-site cations and disordered B-site cations both in bulk and on the surfaces. Further, on the A-site-terminated surface, we found the segregation of large Ba atoms. Similarly, on the A-site- terminated surface of the recently developed Cs0.2Sr0.8Co0.4Fe0.6O3 (CSCF2846) catalyst, layered ordering at the A-site and surface enrichment of large Cs atoms were also found. The layered ordering was robust against thermal effects, as found from molecular dynamics simulations at 800 K. This work provides a new approach for thermodynamic global optimization of chemical ordering in multicomponent materials.

Many-Body Function Corrected Neural Network with Atomic Attention (MBNN-att) for Molecular Property Prediction

Recent years have seen a surge of machine learning (ML) in chemistry for predicting chemical properties, but a low-cost, general-purpose, and high-performance model, desirable to be accessible on central processing unit (CPU) devices, remains not available. For this purpose, here we introduce an atomic attention mechanism into many-body function corrected neural network (MBNN), namely, MBNN-att ML model, to predict both the extensive and intensive properties of molecules and materials. The MBNN-att uses explicit function descriptors as the inputs for the atom-based feed-forward neural network (NN). The output of the NN is designed to be a vector to implement the multihead self-attention mechanism. This vector is split into two parts: the atomic attention weight part and the many-body-function part. The final property is obtained by summing the products of each atomic attention weight and the corresponding many-body function. We show that MBNN-att performs well on all QM9 properties, i.e., errors on all properties, below chemical accuracy, and, in particular, achieves the top performance for the energy-related extensive properties. By systematically comparing with other explicit-function-type descriptor ML models and the graph representation ML models, we demonstrate that the many-body-function framework and atomic attention mechanism are key ingredients for the high performance and the good transferability of MBNN-att in molecular property prediction.

Enhanced Hydrogen Evolution Catalysis of Pentlandite due to the Increases in Coordination Number and Sulfur Vacancy during Cubic-Hexagonal Phase Transition

The search for new phases is an important direction in materials science. The phase transition of sulfides results in significant changes in catalytic performance, such as MoS2 and WS2. Cubic pentlandite [cPn, (Fe, Ni)9S8] can be a functional material in batteries, solar cells, and catalytic fields. However, no report about the material properties of other phases of pentlandite exists. In this study, the unit-cell parameters of a new phase of pentlandite, sulfur-vacancy enriched hexagonal pentlandite (hPn), and the phase boundary between cPn and hPn are determined for the first time. Compared to cPn, the hPn shows a high coordination number, more sulfur vacancies, and high conductivity, which result in significantly higher hydrogen evolution performance of hPn than that of cPn and make the non-nano rock catalyst hPn superior to other most known nanosulfide catalysts. The increase of sulfur vacancies during phase transition provides a new approach to designing functional materials.

Stabilization of Pd0 by Cu Alloying: Theory-Guided Design of Pd3Cu Electrocatalyst for Anodic Methanol Carbonylation

Electrocatalytic carbonylation of CO and CH3OH to dimethyl carbonate (DMC) on metallic palladium (Pd) electrode offers a promising strategy for C1 valorization at the anode. However, its broader application is limited by the high working potential and the low DMC selectivity accompanied with severe methanol self-oxidation. Herein, our theoretical analysis of the intermediate adsorption interactions on both Pd0 and Pd4+ surfaces revealed that inevitable reconstruction of Pd surface under strongly oxidative potential diminishes its CO adsorption capacity, thus damaging the DMC formation. Further theoretical modeling indicates that doping Pd with Cu not only stabilizes low-valence Pd in oxidative environments but also lowers the overall energy barrier for DMC formation. Guided by this insight, we developed a facile two-step thermal shock method to prepare PdCu alloy electrocatalysts for DMC. Remarkably, the predicted Pd3Cu demonstrated the highest DMC selectivity among existing Pd-based electrocatalysts, reaching a peaked DMC selectivity of 93 % at 1.0 V versus Ag/AgCl electrode. (Quasi) in situ spectra investigations further confirmed the predicted dual role of Cu dopant in promoting Pd-catalyzed DMC formation.

Thermal conductivity of irregularly shaped nanoparticles from equilibrium molecular dynamics

The computation of thermal conductivity for finite nanoparticulate systems, particularly those of irregular shapes, poses significant challenges. The nonequilibrium molecular dynamics (NEMD) methods has been extensively utilized in numerous prior studies for the computation of thermal conductivity of nanoparticles. One of our recent works (Donget al2021Phys. Rev.B103035417) proposed that equilibrium molecular dynamics (EMD) methods can be used for the simulation of thermal conductivity of finite-scale systems and demonstrated their equivalence to NEMD methods. In this study, we investigated the application of the (EMD) approach for the computation of thermal conductivity in zero-dimensional nanoparticles. In our initial step, we merged both methodologies to substantiate the equivalence in thermal conductivity calculation for cube and cylinder nanoparticles. After filtering the data, we confirmed the usefulness of EMD for evaluating the thermal conductivity of zero-dimensional materials. The NEMD method faces challenges in accurately predicting thermal conductivity in nanoparticle systems with a varying cross-sectional area along the transport direction, whereas EMD methods can be utilized to estimate thermal conductivity when the volume is known. In a subsequent study, we used the state-of-the-art machine learning potential to calculate the thermal conductivity of spherical nanoparticles and compared the results with those obtained using the classical Tersoff potential. Ultimately, we predicted the thermal conductivity of nanoparticles with various geometries in all directions. Our findings collectively demonstrate the simplicity and effectiveness of employing EMD methods for calculating thermal conductivity in nanoparticle systems, thereby opening up new avenues for investigating thermal transport properties in particle systems as well as nanopders.

A human-machine interface for automatic exploration of chemical reaction networks

Autonomous reaction network exploration algorithms offer a systematic approach to explore mechanisms of complex chemical processes. However, the resulting reaction networks are so vast that an exploration of all potentially accessible intermediates is computationally too demanding. This renders brute-force explorations unfeasible, while explorations with completely pre-defined intermediates or hard-wired chemical constraints, such as element-specific coordination numbers, are not flexible enough for complex chemical systems. Here, we introduce a STEERING WHEEL to guide an otherwise unbiased automated exploration. The STEERING WHEEL algorithm is intuitive, generally applicable, and enables one to focus on specific regions of an emerging network. It also allows for guiding automated data generation in the context of mechanism exploration, catalyst design, and other chemical optimization challenges. The algorithm is demonstrated for reaction mechanism elucidation of transition metal catalysts. We highlight how to explore catalytic cycles in a systematic and reproducible way. The exploration objectives are fully adjustable, allowing one to harness the STEERING WHEEL for both structure-specific (accurate) calculations as well as for broad high-throughput screening of possible reaction intermediates.

Construction of High Accuracy Machine Learning Interatomic Potential for Surface/Interface of Nanomaterials-A Review

The inherent discontinuity and unique dimensional attributes of nanomaterial surfaces and interfaces bestow them with various exceptional properties. These properties, however, also introduce difficulties for both experimental and computational studies. The advent of machine learning interatomic potential (MLIP) addresses some of the limitations associated with empirical force fields, presenting a valuable avenue for accurate simulations of these surfaces/interfaces of nanomaterials. Central to this approach is the idea of capturing the relationship between system configuration and potential energy, leveraging the proficiency of machine learning (ML) to precisely approximate high-dimensional functions. This review offers an in-depth examination of MLIP principles and their execution and elaborates on their applications in the realm of nanomaterial surface and interface systems. The prevailing challenges faced by this potent methodology are also discussed.

Machine-Learned Kohn-Sham Hamiltonian Mapping for Nonadiabatic Molecular Dynamics

In this work, we report a simple, efficient, and scalable machine-learning (ML) approach for mapping non-self-consistent Kohn-Sham Hamiltonians constructed with one kind of density functional to the nearly self-consistent Hamiltonians constructed with another kind of density functional. This approach is designed as a fast surrogate Hamiltonian calculator for use in long nonadiabatic dynamics simulations of large atomistic systems. In this approach, the input and output features are Hamiltonian matrices computed from different levels of theory. We demonstrate that the developed ML-based Hamiltonian mapping method (1) speeds up the calculations by several orders of magnitude, (2) is conceptually simpler than alternative ML approaches, (3) is applicable to different systems and sizes and can be used for mapping Hamiltonians constructed with arbitrary density functionals, (4) requires a modest training data, learns fast, and generates molecular orbitals and their energies with the accuracy nearly matching that of conventional calculations, and (5) when applied to nonadiabatic dynamics simulation of excitation energy relaxation in large systems yields the corresponding time scales within the margin of error of the conventional calculations. Using this approach, we explore the excitation energy relaxation in C60 fullerene and Si75H64 quantum dot structures and derive qualitative and quantitative insights into dynamics in these systems.

Abundance of Low-Energy Oxygen Vacancy Pairs Dictates the Catalytic Performance of Cerium-Stabilized Zirconia

Cerium-stabilized zirconia (Ce1-xZrxOy, CZO) is renowned for its superior oxygen storage capacity (OSC), a key property long believed to be beneficial to catalytic oxidation reactions. However, 50% Ce-containing CZO recorded with the highest OSC has disappointingly poor performance in catalytic oxidation reactions compared to those with higher Ce contents but lower OSC ability. Here, we employ global neural network (G-NN)-based potential energy surface exploration methods to establish the first ternary phase diagram for bulk structures of CZO, which identifies three critical compositions of CZO, namely, 50, 60, and 80% Ce-containing CZO that are thermodynamically stable under typical synthetic conditions. 50% Ce-containing CZO, although having the highest OSC, exhibits the lowest O vacancy (Ov) diffusion rate. By contrast, 60% Ce-containing CZO, despite lower OSC (33.3% OSC compared to that of 50% Ce-containing CZO), reaches the highest Ov diffusion ability and thus offers the highest CO oxidation catalytic performance. The physical origin of the high performance of 60% Ce-containing CZO is the abundance of energetically favorable Ov pairs along the ⟨110⟩ direction, which reduces the energy barrier of Ov diffusion in the bulk and promotes O2 activation on the surface. Our results clarify the long-standing puzzles on CZO and point out that 60% Ce-containing CZO is the most desirable composition for typical CZO applications.

Nature-Inspired Photocatalytic Azo Bond Cleavage with Red Light

The emerging field of photoredox catalysis in mammalian cells enables spatiotemporal regulation of a wealth of biological processes. However, the selective cleavage of stable covalent bonds driven by low-energy visible light remains a great challenge. Herein, we report that red light excitation of a commercially available dye, abbreviated NMB+, leads to catalytic cleavage of stable azo bonds in both aqueous solutions and hypoxic cells and hence a means to photodeliver drugs or functional molecules. Detailed mechanistic studies reveal that azo bond cleavage is triggered by a previously unknown consecutive two-photon process. The first photon generates a triplet excited state, 3NMB+*, that is reductively quenched by an electron donor to generate a protonated NMBH•+. The NMBH•+ undergoes a disproportionation reaction that yields the initial NMB+ and two-electron-reduced NMBH (i.e., leuco-NMB, abbreviated as LNMB). Interestingly, LNMB forms a charge transfer complex with all four azo substrates that possess an intense absorption band in the red region. A second red photon induces electron transfer from LNMB to the azo substrate, resulting in azo bond cleavage. The charge transfer complex mediated two-photon catalytic mechanism reported herein is reminiscent of the flavin-dependent natural photoenzyme that catalyzes bond cleavage reactions with high-energy photons. The red-light-driven photocatalytic strategy offers a new approach to bioorthogonal azo bond cleavage for photodelivery of drugs or functional molecules.

Amorphous Chloride Solid Electrolytes with High Li-Ion Conductivity for Stable Cycling of All-Solid-State High-Nickel Cathodes

Solid electrolytes (SEs) are central components that enable high-performance, all-solid-state lithium batteries (ASSLBs). Amorphous SEs hold great potential for ASSLBs because their grain-boundary-free characteristics facilitate intact solid-solid contact and uniform Li-ion conduction for high-performance cathodes. However, amorphous oxide SEs with limited ionic conductivities and glassy sulfide SEs with narrow electrochemical windows cannot sustain high-nickel cathodes. Herein, we report a class of amorphous Li-Ta-Cl-based chloride SEs possessing high Li-ion conductivity (up to 7.16 mS cm-1) and low Young's modulus (approximately 3 GPa) to enable excellent Li-ion conduction and intact physical contact among rigid components in ASSLBs. We reveal that the amorphous Li-Ta-Cl matrix is composed of LiCl43-, LiCl54-, LiCl65- polyhedra, and TaCl6- octahedra via machine-learning simulation, solid-state 7Li nuclear magnetic resonance, and X-ray absorption analysis. Attractively, our amorphous chloride SEs exhibit excellent compatibility with high-nickel cathodes. We demonstrate that ASSLBs comprising amorphous chloride SEs and high-nickel single-crystal cathodes (LiNi0.88Co0.07Mn0.05O2) exhibit ∼99% capacity retention after 800 cycles at ∼3 C under 1 mA h cm-2 and ∼80% capacity retention after 75 cycles at 0.2 C under a high areal capacity of 5 mA h cm-2. Most importantly, a stable operation of up to 9800 cycles with a capacity retention of ∼77% at a high rate of 3.4 C can be achieved in a freezing environment of -10 °C. Our amorphous chloride SEs will pave the way to realize high-performance high-nickel cathodes for high-energy-density ASSLBs.

Automated Exploration of Reaction Networks and Mechanisms Based on Metadynamics Nanoreactor Simulations

We developed an automated approach to construct a complex reaction network and explore the reaction mechanisms for numerous reactant molecules by integrating several theoretical approaches. Nanoreactor-type molecular dynamics was used to generate possible chemical reactions, in which the metadynamics was used to overcome the reaction barriers, and the semiempirical GFN2-xTB method was used to reduce the computational cost. Reaction events were identified from trajectories using the hidden Markov model based on the evolution of the molecular connectivity. This provided the starting points for further transition-state searches at the electronic structure levels of density functional theory to obtain the reaction mechanism. Finally, the entire reaction network containing multiple pathways was built. The feasibility and efficiency of the automated construction of the reaction network were investigated using the HCHO and NH3 biomolecular reaction and the reaction network for a multispecies system comprising dozens of HCN and H2O molecules. The results indicated that the proposed approach provides a valuable and effective tool for the automated exploration of the reaction networks.

The Identity of Nickel Peroxide as a Nickel Superoxyhydroxide for Enhanced Electrocatalysis

Nickel peroxides are a class of stoichiometric oxidants that can selectively oxidize various organic compounds, but their molecular level structure remained elusive until now. Herein, we utilized structural prediction using the Stochastic Surface Walking method based on a neural network potential energy surface and advanced characterization using the as-synthesized nickel peroxide to unravel its chemical identity as the bridging superoxide containing nickel hydroxide, or nickel superoxyhydroxide. Superoxide incorporation tunes the local chemical environment of nickel and oxygen beyond the conventional Bode plot, offering a 6.4-fold increase in the electrocatalytic activity of urea oxidation. A volcanic dependence of the activity on the oxygen equivalents leads to the proposed active site of the Ni(OO)(OH)Ni five-membered ring. This work not only unveils the possible structures of nickel peroxides but also emphasizes the significance of tailoring the oxygen environment for advanced catalysis.

Global Neural Network Potential with Explicit Many-Body Functions for Improved Descriptions of Complex Potential Energy Surface

The high dimensional machine learning potential (MLP) that has developed rapidly in the past decade represents a giant step forward in large-scale atomic simulation for complex systems. The long-range interaction and the poor description of chemical reactions are typical problems of high dimensional MLP, which are mainly caused by the poor structure discrimination of the atom-centered ML model. Herein, we propose a low-cost neural-network-based MLP architecture for fitting global potential energy surface data, namely, G-MBNN, that can offer improved energy and force resolution on a complex potential energy surface. In G-MBNN, a set of many-body energy terms based on the local atomic environment are explicitly included in computing the total energy─the total energy of the system is written as the sum of atomic energy and many-body energy contributions. These extra many-body energy terms are computationally low-cost and, importantly, can provide easy access to delicate energy terms in complex systems such as very short repulsion, long-range attractions, and sensitive angular-dependent covalent interactions. We implement G-MBNN in the LASP code and demonstrate the improved accuracy of the new framework in representative systems, including ternary-element energy materials LiCoOx, TiO2 with defects, and a series of organic reactions.

Regulating reconstruction of oxide-derived Cu for electrochemical CO2 reduction toward n-propanol

Oxide-derived copper (OD-Cu) is the most efficient and likely practical electrocatalyst for CO2 reduction toward multicarbon products. However, the inevitable but poorly understood reconstruction from the pristine state to the working state of OD-Cu under strong reduction conditions largely hinders the rational construction of catalysts toward multicarbon products, especially C3 products like n-propanol. Here, we simulate the reconstruction of CuO and Cu2O into their derived Cu by molecular dynamics, revealing that CuO-derived Cu (CuOD-Cu) intrinsically has a richer population of undercoordinated Cu sites and higher surficial Cu atom density than the counterpart Cu2O-derived Cu (Cu2OD-Cu) because of the vigorous oxygen removal. In situ spectroscopes disclose that the coordination number of CuOD-Cu is considerably lower than that of Cu2OD-Cu, enabling the fast kinetics of CO2 reaction and strengthened binding of *C2 intermediate(s). Benefiting from the rich undercoordinated Cu sites, CuOD-Cu achieves remarkable n-propanol faradaic efficiency up to ~17.9%, whereas the Cu2OD-Cu dominantly generates formate.

ChecMatE: A workflow package to automatically generate machine learning potentials and phase diagrams for semiconductor alloys

Semiconductor alloy materials are highly versatile due to their adjustable properties; however, exploring their structural space is a challenging task that affects the control of their properties. Traditional methods rely on ad hoc design based on the understanding of known chemistry and crystallography, which have limitations in computational efficiency and search space. In this work, we present ChecMatE (Chemical Material Explorer), a software package that automatically generates machine learning potentials (MLPs) and uses global search algorithms to screen semiconductor alloy materials. Taking advantage of MLPs, ChecMatE enables a more efficient and cost-effective exploration of the structural space of materials and predicts their energy and relative stability with ab initio accuracy. We demonstrate the efficacy of ChecMatE through a case study of the InxGa1-xN system, where it accelerates structural exploration at reduced costs. Our automatic framework offers a promising solution to the challenging task of exploring the structural space of semiconductor alloy materials.

In Situ Structure of a Mo-Doped Pt-Ni Catalyst during Electrochemical Oxygen Reduction Resolved from Machine Learning-Based Grand Canonical Global Optimization

Pt-Ni alloy is by far the most active cathode material for oxygen reduction reaction (ORR) in the proton-exchange membrane fuel cell, and the addition of a tiny amount of a third-metal Mo can significantly improve the catalyst durability and activity. Here, by developing machine learning-based grand canonical global optimization, we are able to resolve the in situ structures of this important three-element alloy system under ORR conditions and identify their correlations with the enhanced ORR performance. We disclose the bulk phase diagram of Pt-Ni-Mo alloys and determine the surface structures under the ORR reaction conditions by exploring millions of likely structure candidates. The pristine Pt-Ni-Mo alloy surfaces are shown to undergo significant structure reconstruction under ORR reaction conditions, where a surface-adsorbed MoO₄ monomer or Mo₂O _x dimers cover the Pt-skin surface above 0.9 V vs RHE and protect the surface from Ni leaching. The physical origins are revealed by analyzing the electronic structure of O atoms in MoO₄ and on the Pt surface. In viewing the role of high-valence transition metal oxide clusters, we propose a set of quantitative measures for designing better catalysts and predict that six elements in the periodic table, namely, Mo, Tc, Os, Ta, Re, and W, can be good candidates for alloying with PtNi to improve the ORR catalytic performance. We demonstrate that machine learning-based grand canonical global optimization is a powerful and generic tool to reveal the catalyst dynamics behavior in contact with a complex reaction environment.

(2×1) Reconstruction Mechanism of Rutile TiO2(011) Surface

Understanding the reconstruction kinetics of solid surfaces involving an ensemble of atomic movements is practically important but challenging due to the complexity of high-dimensional potential energy surfaces. Herein, we develop a step-deciding technique incorporated with the nudged elastic band method, which enables multidirection pathway sampling and ensures the capture of a minimum energy path (MEP). Using this approach, the (2×1) reconstruction mechanism of a rutile-TiO₂(011) surface, a classic and long-standing open problem in the fields of surface science and heterogeneous catalysis, is quantified, and the MEP is explicitly identified and explained. Following the least-bond-breaking rule, it gives a stepwise Ti-O bond cleavage mechanism with a collection of decoupled local structural relaxation modes at an overall barrier of 1.25 eV critically affected by initial Ti-O bond opening, which is much lower than the common synergy mechanism. Moreover, the adsorption-induced reconstruction is rationalized considering practical reaction conditions, where H atom adsorbate is shown to effectively stabilize the labile one-fold O_1c intermediate and promote the reconstruction kinetics. This work reveals the reconstruction mechanism regarding multiatom movements and provides a general method for the structural exploration of other complicated systems.

Machine Learning in Unmanned Systems for Chemical Synthesis

Chemical synthesis is state-of-the-art, and, therefore, it is generally based on chemical intuition or experience of researchers. The upgraded paradigm that incorporates automation technology and machine learning (ML) algorithms has recently been merged into almost every subdiscipline of chemical science, from material discovery to catalyst/reaction design to synthetic route planning, which often takes the form of unmanned systems. The ML algorithms and their application scenarios in unmanned systems for chemical synthesis were presented. The prospects for strengthening the connection between reaction pathway exploration and the existing automatic reaction platform and solutions for improving autonomation through information extraction, robots, computer vision, and intelligent scheduling were proposed.

Benchmark of general-purpose machine learning-based quantum mechanical method AIQM1 on reaction barrier heights

Artificial intelligence-enhanced quantum mechanical method 1 (AIQM1) is a general-purpose method that was shown to achieve high accuracy for many applications with a speed close to its baseline semiempirical quantum mechanical (SQM) method ODM2*. Here, we evaluate the hitherto unknown performance of out-of-the-box AIQM1 without any refitting for reaction barrier heights on eight datasets, including a total of ∼24 thousand reactions. This evaluation shows that AIQM1's accuracy strongly depends on the type of transition state and ranges from excellent for rotation barriers to poor for, e.g., pericyclic reactions. AIQM1 clearly outperforms its baseline ODM2* method and, even more so, a popular universal potential, ANI-1ccx. Overall, however, AIQM1 accuracy largely remains similar to SQM methods (and B3LYP/6-31G* for most reaction types) suggesting that it is desirable to focus on improving AIQM1 performance for barrier heights in the future. We also show that the built-in uncertainty quantification helps in identifying confident predictions. The accuracy of confident AIQM1 predictions is approaching the level of popular density functional theory methods for most reaction types. Encouragingly, AIQM1 is rather robust for transition state optimizations, even for the type of reactions it struggles with the most. Single-point calculations with high-level methods on AIQM1-optimized geometries can be used to significantly improve barrier heights, which cannot be said for its baseline ODM2* method.

Data-Driven Machine Learning for Understanding Surface Structures of Heterogeneous Catalysts

The design of heterogeneous catalysts is necessarily surface-focused, generally achieved via optimization of adsorption energy and microkinetic modelling. A prerequisite is to ensure the adsorption energy is physically meaningful is the stable existence of the conceived active-site structure on the surface. The development of improved understanding of the catalyst surface, however, is challenging practically because of the complex nature of dynamic surface formation and evolution under in-situ reactions. We propose therefore data-driven machine-learning (ML) approaches as a solution. In this Minireview we summarize recent progress in using machine-learning to search and predict (meta)stable structures, assist operando simulation under reaction conditions and micro-environments, and critically analyze experimental characterization data. We conclude that ML will become the new norm to lower costs associated with discovery and design of optimal heterogeneous catalysts.

Coupled oxygen desorption and structural reconstruction accompanying reduction of copper oxide

Understanding structural transformation and phase transition accompanying reactions in a solid as a catalyst or oxygen carrier is important to the design and optimization of many catalytic or chemical looping reaction processes. Herein, we combined density functional theory calculation with the stochastic surface walking global optimization approach to track the structural transformation accompanying the reduction of CuO upon releasing oxygen. We then used machine learning (ML) methods to correlate the structural properties of CuO_x with varying x. By decomposing a reduction step into oxygen detachment and structural reconstruction, we identified two types of pathways: (1) uniform reduction with minimal structural changes; (2) segregated reduction with significant reconstruction. The results of ML analysis showed that the most important feature is the radial distribution functions of Cu-O at a percentage of oxygen vacancy [C(O_V)] < 50% and Cu-Cu at C(O_V) > 50% for CuO_x formation. These features reflect the underlying physicochemical origin, i.e., Cu-O breaking and Cu-Cu formation in the respective stage of reduction. Phase diagram analysis indicates that CuO will be reduced to Cu₂O under a typical oxygen uncoupling condition. This work demonstrates the complexity of solid structural transformation and the potential of ML methods in studying solid state materials involved in many chemical processes.

Long-range ordered porous carbons produced from C60

Carbon structures with covalent bonds connecting C₆₀ molecules have been reported^1-3, but their production methods typically result in very small amounts of sample, which restrict the detailed characterization and exploration necessary for potential applications. We report the gram-scale preparation of a new type of carbon, long-range ordered porous carbon (LOPC), from C₆₀ powder catalysed by α-Li₃N at ambient pressure. LOPC consists of connected broken C₆₀ cages that maintain long-range periodicity, and has been characterized by X-ray diffraction, Raman spectroscopy, magic-angle spinning solid-state nuclear magnetic resonance spectroscopy, aberration-corrected transmission electron microscopy and neutron scattering. Numerical simulations based on a neural network show that LOPC is a metastable structure produced during the transformation from fullerene-type to graphene-type carbons. At a lower temperature, shorter annealing time or by using less α-Li₃N, a well-known polymerized C₆₀ crystal forms owing to the electron transfer from α-Li₃N to C₆₀. The carbon K-edge near-edge X-ray absorption fine structure shows a higher degree of delocalization of electrons in LOPC than in C₆₀(s). The electrical conductivity is 1.17 × 10^-2 S cm^-1 at room temperature, and conduction at T < 30  K appears to result from a combination of metallic-like transport over short distances punctuated by carrier hopping. The preparation of LOPC enables the discovery of other crystalline carbons starting from C₆₀(s).

Neural Network Accelerated Investigation of the Dynamic Structure-Performance Relations of Electrochemical CO2 Reduction over SnO x Surfaces

Heterogeneous catalysts, especially metal oxides, play a curial role in improving energy conversion efficiency and production of valuable chemicals. However, the surface structure at the atomic level and the nature of active sites are still ambiguous due to the dynamism of surface structure and difficulty in structure characterization under electrochemical conditions. This paper describes a strategy of the multiscale simulation to investigate the SnO _x reduction process and to build a structure-performance relation of SnO _x for CO₂ electroreduction. Employing high-dimensional neural network potential accelerated molecular dynamics and stochastic surface walking global optimization, coupled with density functional theory calculations, we propose that SnO₂ reduction is accompanied by surface reconstruction and charge density redistribution of active sites. A regulatory factor, the net charge, is identified to predict the adsorption capability for key intermediates on active sites. Systematic electronic analyses reveal the origin of the interaction between the adsorbates and the active sites. These findings uncover the quantitative correlation between electronic structure properties and the catalytic performance of SnO _x so that Sn sites with moderate charge could achieve the optimally catalytic performance of the CO₂ electroreduction to formate.

Growth Pattern of Large Morse Clusters with Medium-Range Potentials

Due to the extremely complex potential energy surfaces of large Morse clusters with medium-range potentials (i.e., ρ = 6 and 10), global optimization studies in the literature are limited to a cluster size (N) of ≤240. Starting from completely random structures, we successfully systematically studied Morse clusters with up to 700 atoms using our unbiased fuzzy global optimization (FGO) method. While all of the putative global minima reported previously have been efficiently obtained, new global minima with lower energies are identified for N values of 176, 258, 485, 561, 817, and 923 with ρ = 6 and for N values of 151, 202, 226, and 229 with ρ = 10. A detailed growth pattern and magic clusters are obtained. For the first time, we find that a central vacancy is present in Morse clusters containing 542, 543, 548, and 922 atoms with ρ = 6. FGO has achieved high performance in large clusters with different interatomic interaction ranges, thus showing great application potential in the global structure optimization of general clusters.