Cluster Structures Research Articles

A novel federated multi-view clustering method for unaligned and incomplete data fusion

Recently, federated multi-view clustering (FedMVC) has emerged as a powerful tool to uncover complementary cluster structures across distributed clients, gaining significant attention in the realm of data fusion. While FedMVC methods have adeptly addressed the challenges of feature heterogeneity among various clients, achieving notable success in controlled environments. Their applicability often hinges on the assumptions of strict alignment and data completeness across multi-view clients. These assumptions, unfortunately, are not always consistent with real-world conditions. Specifically, practical applications often come with (1) unaligned multi-view data and (2) missing data. Current FedMVC methods struggle to effectively address these challenges. To bridge this gap, this paper presents FCUIF, a novel method that eliminates the need for data alignment and completeness assumptions. To tackle unaligned data, FCUIF leverages both sample commonality and view versatility to adaptively generate alignment matrices, ensuring effective cross-view alignment. For the challenge of missing data, FCUIF uses an unsupervised technique to evaluate and refine imputation quality, efficiently handling various scenarios of incomplete multi-view data. Our extensive experiments using four public datasets demonstrate FCUIF’s superior performance when dealing with unaligned and incomplete multi-view data. The code is available at https://github.com/5Martina5/FCUIF.

Argon Nanocluster Filled Between Twisted Bilayer Graphene: Molecular Dynamics Simulation

In 2004, graphene was discovered by Novoselov and Geim. With the development of technology，twisted bilayer graphene (tBLG) has become a hot research topic. On the basis of predecessors, we further study the confined spaces between twisted bilayer graphene. The interlayer spacing between them is expected to modify properties of atoms and molecules confined at the atomic interfaces. The effects of twist angle, interlayer spacing of the tBLG and temperature (20 K–70 K) on the equilibrium structure of argon cluster are systematically studied by means of molecular dynamics simulation with the tBLG being considered to be fixed. As the interlayer spacings of tBLGs with different twist angles increase at a fixed temperature 20 K, the equilibrium structures of the argon cluster transform from a monoatomic layer to multilayer structure. For different twist angles at the fixed interlayer spacing, the structures of the argon cluster gradually form periodic pattern (similar to moiré pattern). The effects of temperature and non-equiliblium suddenly heating process on the structures of argon clusters between tBLGs are also studied by molecular dynamics simulation. Our results may provide a theoretical support for the preparation of new carbon-based intercalated composites and the application of graphene.

Genetically Fused Resilin-like Polypeptide-Coiled Coil Bundlemer Conjugates Exhibit Tunable Multistimuli-Responsiveness and Undergo Nanofibrillar Assembly.

Peptide-based materials are diverse candidates for self-assembly into modularly designed and stimuli-responsive nanostructures with precisely tunable compositions. Here, we genetically fused computationally designed coiled coil-forming peptides to the N- and C-termini of compositionally distinct multistimuli-responsive resilin-like polypeptides (RLPs) of various lengths. The successful expression of these hybrid polypeptides in bacterial hosts was confirmed through techniques such as gel electrophoresis, mass spectrometry, and amino acid analysis. Circular dichroism spectroscopy and ultraviolet-visible turbidimetry demonstrated that despite the fusion of disparate structural and responsive units, the coiled coils remained stable in the hybrid polypeptides, and the sequence-encoded differences in thermoresponsive phase separation of the RLPs were preserved. Cryogenic transmission electron microscopy and coarse-grained modeling showed that after thermal annealing in solution, the hybrid polypeptides adopted a closed loop conformation and assembled into nanofibrils capable of further hierarchically organizing into cluster structures and ribbon-like structures mediated by the self-association tendency of the RLPs.

A multiscale study on the microstructure and hardening models of the irradiation defects on reactor pressure vessel steels: Modelling and experiment

Irradiation defects are known to impede the dislocation motion in irradiated reactor pressure vessel (RPV) steels. However, consistent experimental studies of the respective irradiation defects and relevant theoretical studies of the interaction mechanism and hardening model are still lacking. In this study, the microstructure and irradiation-induced hardening of RPV steels are investigated using transmission electron microscopy (TEM)/energy-dispersive X-ray spectrometry (EDS) and molecular dynamics (MD) simulations. In A508-Ⅲ steel, 9R, twinned fcc, and other typical Cu-rich cluster structures are observed using TEM after the irradiation experiment. Based on the experimental results, MD simulations are performed to study the interactions between the dislocations and Cu clusters or voids. A mechanism diagram of the interaction mechanism coupling the effects of temperature and defect size is shown for both the Cu clusters and voids. A relevant hardening model is proposed based on the MD data, which is further discussed in relation to the macroscopic mechanical properties. Furthermore, the irradiation-induced hardening of the RPV steel is successfully predicted, which further demonstrates the reliability of the proposed unified hardening model.

Anti-fogging/dry-dust transparent superhydrophobic surfaces based on liquid-like molecule brush modified nanofiber cluster structures

Transparent protective coatings capable of preventing fog and dust accumulation have broad application prospect in photovoltaic systems, optical devices and consumer electronics. Although a number of superhydrophobic coatings have been developed for self-cleaning purpose over the past three decades, there is still a lack of surfaces that can simultaneously possess high transparency, remarkable superhydrophobicity, and excellent fog and dust resistance. In this study, we have prepared surfaces featuring sub-wavelength nanofiber cluster structures through a facile plasma etching method, and further modified the surface with liquid-like perfluoropolyether (PFPE) brushes. The prepared PFPE modified nanofibrous surface (PFPE-NS) exhibits superior optical transparency (transmittance 90.4 % ± 0.7 %) and water repellency, with a water contact angle as high as 171.0° ± 0.6° and sliding angle down to 0.5° ± 0.1° (5 µL). More importantly, benefitted from the nanofiber cluster structures and the slippery liquid-like surface chemistry, the adhesion and accumulation of fog droplets and dust particles on PFPE-NS is greatly inhibited. As a consequence, PFPE-NS can keep excellent optical clearness after 2 h fogging test and maintain an average transmittance above 87 % after 24 h dusting test. Our study provides a promising strategy through constructing liquid-like nanofibrous coating for optical protection that could be applicable in practical rainy, foggy, and dusty environments.

Structure-guided feature and cluster contrastive learning for multi-view clustering

Multi-view clustering (MVC) technology performs unsupervised clustering on data collected from multiple sources, and has received intense attention in recent years. However, most existing MVC methods fail to consider retaining view-specific information when learning multi-view consistent representations. Besides, the feature and cluster structures of multi-view data cannot be fully leveraged in clustering. In this paper, we propose a structure-guided feature and cluster contrastive learning (SGFCC) for multi-view clustering. Specifically, SGFCC utilizes autoencoders to achieve view-specific information reconstruction in feature space, and extracts high-level features for multi-view consistent representation learning to eliminate the effects of view-specific information and noise on consistent representation. To fully capture the similar clustering structure of high-level features and semantic features of samples across different views, we adopt a structure-guided feature-level and cluster-level contrastive learning strategy in our SGFCC model. Furthermore, we design a clustering layer to explore the cluster structure of high-level features. Different from most existing MVC methods, our method applies a non-fusion scheme that aggregates the semantic information of all views to obtain the final semantic labels. Extensive experiments on public datasets demonstrate that our method outperforms other competitors in clustering tasks.

Efficient Algorithm for K-Multiple-Means

K-Multiple-Means is an extension of K-means for the clustering of multiple means used in many applications, such as image segmentation, load balancing, and blind-source separation. Since K-means uses only one mean to represent each cluster, it fails to capture non-spherical cluster structures of data points. However, since K-Multiple-Means represents the cluster by computing multiple means and grouping them into specified c clusters, it can effectively capture the non-spherical clusters of the data points. To obtain the clusters, K-Multiple-Means updates a similarity matrix of a bipartite graph between the data points and the multiple means by iteratively computing the leading c singular vectors of the matrix. K-Multiple-Means, however, incurs a high computation cost for large-scale data due to the iterative SVD computations. Our proposal, F-KMM, increases the efficiency of K-Multiple-Means by computing the singular vectors from a smaller similarity matrix between the multiple means obtained from the similarity matrix of the bipartite graph. To compute the similarity matrix of the bipartite graph efficiently, we skip unnecessary distance computations and estimate lower bounding distances between the data points and the multiple means. Theoretically, the proposed approach guarantees the same clustering results as K-Multiple-Means since it can exactly compute the singular vectors from the similarity matrix between the multiple means. Experiments show that our approach is several orders of magnitude faster than previous clustering approaches that use multiple means.

Spectroscopic and Theoretical Identifications of Two Structural Motifs of (H2O)10 Cluster.

Precise characterization of archetypal systems of aqueous hydrogen-bonding networks is essential for developing accurate potential functions and universal models of water. The structures of water clusters (H2O)n (n = 2-9) have been verified recently through size-specific infrared spectroscopy with a vacuum ultraviolet free electron laser (VUV-FEL) and quantum chemical studies. For (H2O)10, the pentagonal prism and butterfly motifs were proposed to be important building blocks and were observed in previous experiments. Here we report the size-specific infrared spectra of (H2O)10 via a joint experimental and theoretical study. Well-resolved spectra provide a unique signature for the coexistence of pentagonal prism and butterfly motifs. These (H2O)10 motifs develop from the dominant structures of (H2O)n (n = 8, 9) clusters. This work provides an intriguing prelude to the diverse structure of liquid water and opens avenues for size-dependent measurement of larger systems to understand the stepwise formation mechanism of hydrogen-bonding networks.

Structural differences of Cu-Pd clusters with three potential parameters

Parameters of potential function have a great influence on the structures of bimetallic clusters. Three potential parameters, i.e., parameter sets C, P/N, and M, are applied to study the structural difference of Cu-Pd clusters within 147 atoms. The structural optimization is performed by adaptive immune optimization algorithm with constructed inner cores and iterated local search method. Results showed that for parameters C and P/N the structural motifs of 55-atom and 147-atom Cu-Pd clusters are almost the same, but the atomic distribution is different. As for clusters with 75 and 98 atoms, there exist obvious differences in geometric configurations.

Influence of carboxymethyl cellulose on the stability, rheology, and curcumin bioaccessibility of high internal phase Pickering emulsions

Recently, there has been a focus on using biopolymer-based particles to stabilize high internal phase Pickering emulsions (HIPPEs) due to the notable advances in biocompatibility and biodegradability. In this work, the complex particles of peanut protein isolate and carboxymethyl cellulose (CMC) with various substitution degrees (DS; 0.7 and 0.9) and weight average molecular weights (Mw; 90, 250, and 700 kDa) were prepared and characterized as novel stabilizers. For the obtained four types of morphologically distinct particles, the complex particles formed by CMC (0.9 DS and 250 kDa) showed cluster structures with an average size of 1.271 μm, equally biphasic wettability with three-phase contact angles of 91.5°, and the highest diffusion rate at the oil-water interface. HIPPEs stabilized by these particles exhibited more elastic behavior due to the smaller tanδ and higher viscosity, as well as excellent thixotropic recovery properties and stability against heating, storage, and freeze-thawing. Furthermore, confocal laser scanning microscopy verified that these particles formed a dense interfacial layer around the oil droplets, which could resist flocculation and coalescence between oil droplets during in vitro digestion. The improved bioaccessibility of curcumin-loaded HIPPEs made these delivery systems potentially apply in functional foods.

Atomic structure of clusters and GP-zones in an Al-Mg-Si alloy

The structures of atomic clusters and GP-zones preceding precipitation in an Al-0.70Mg-0.85Si-0.15Fe-0.25Mn (wt. %) alloy have been investigated by annular dark field scanning transmission electron microscopy imaging and density functional theory calculations. One analysed condition consisted of one-month natural aging, and another was in a pre-baked state after 24 h aging at 90 °C. To quantify the effect of clusters and GP-zones on microstructure development during a subsequent isothermal artificial aging at 185 °C, hardness evolution was measured, and precipitate statistics was performed at the peak hardness corresponding to 5 h aging. Several types of clusters and GP-zones have been observed and based on structure and visual appearance named 'Disordered Frank-Kasper', 'Square', 'Binocular', and 1β", 2β", 3β", 4β". The first four are most common in the natural aged condition and are associated with a well-known influence on the initial hardness evolution. The last three are fractions of the hardening β" phase and refer to the number of structural units they contain. These GP-zones have an increased density in the pre-baked condition, leading to a rapid increase in hardness during artificial aging and to a finer microstructure of high-density small needle precipitates at peak hardness.

Bi(III) Binding Stoichiometry and Domain-Specificity Differences Between Apo and Zn(II)-bound Human Metallothionein 1a.

Bismuth is a xenobiotic metal with a high affinity to sulfur that is used in a variety of therapeutic applications. Bi(III) induces the cysteine-rich metallothionein (MT), a protein known to form two-domain cluster structures with certain metals such as Zn(II), Cd(II), or Cu(I). The binding of Bi(III) to MTs has been previously studied, but there are conflicting reports on the stoichiometry and binding pathway, which appear to be highly dependent on pH and initial metal-loading status of the MT. Additionally, domain specificity has not been thoroughly investigated. In this paper, ESI-MS was used to determine the binding constants of [Bi(EDTA)]- binding to apo-MT1a and its individual αMT fragment. The results were compared to previous experiments using βMT1a and βαMT3. Domain specificity was investigated using proteolysis methods and the initial cooperatively formed Bi2MT was found to bind to cysteines that spanned across the traditional metal binding domain regions. Titrations of [Bi(EDTA)]- into Zn7MT were performed and were found to result in a maximum stoichiometry of Bi7MT, contrasting the Bi6MT formed when [Bi(EDTA)]- was added to apo-MT. These results show that the initial structure of the apo-MT determines the stoichiometry of new incoming metals and explains the previously observed differences in stoichiometry.

Dielectric Behavior and Prolate Growth Patterns of Silicon Clusters SiN with N = 12-30 by Cryogenic Electric Beam Deflection.

We present a comprehensive investigation of the dielectric behavior and geometric structures of cold neutral SiN clusters of intermediate size with N = 12-30 atoms. For this, cryogenic electric beam deflection experiments were carried out for the first time for Si clusters at nozzle temperatures below 30 K. In combination with quantum chemical calculations based on density functional theory and classical trajectory simulations of the rotational dynamics in the electric field, the geometric structures of the clusters are discriminated. Clusters with N < 15 favor a single-capped square antiprism as a nucleus for cluster growth, forming compact geometries in the molecular beam. Starting with 15 atoms, a prolate-like growth is observed. The prolate structures are based on stable building blocks which reappear for numerous sizes throughout the cluster growth. Finally, the transition from prolate to quasi-spherical shapes is shown to take place around Si29/Si30 as predicted theoretically by the literature. The influence of the exchange-correlation functional on the predicted structure and dielectric properties is discussed in detail for some clusters. Relaxation of the electric-dipole moment and therefore quenching of the observed electric response due to vibrational excitation and collisions with the background gas are also considered, which explains deviations between experiment and theory.

Pseudotetrahedral Organotin-Capped Chalcogenidometalate Supermolecules with Optical Limiting Performance.

The rational design of crystalline clusters with adjustable compositions and dimensions is highly sought after but quite challenging as it is important to understand their structural evolution processes and to systematically establish structure-property relationships. Herein, a family of organotin-based sulfidometalate supertetrahedral clusters has been prepared via mixed metal and organotin strategies at low temperatures (60-120 °C). By engineering the metal composition, we can effectively control the size of the clusters, which ranges from 8 to 35, accompanied by variable configurations: P1-[(RSn)4M4S13], T3-[(RSn)4In4M2S16] (R = nbutyl-Bu and phenyl-Ph; M = Cd, Zn, and Mn), T4-[(BuSn)4In13Cu3S31], truncated P2, viz. TP2-[(BuSn)6In10Cu6S31], and even T5-[(BuSn)4In22Zn6Cu3S52], all of which are the largest organometallic supertetrahedral clusters known to date. Of note, the arylstannane approach plays a critical role in regulating the peripheral ligands and further enriching geometric structures of the supertetrahedral clusters. This is demonstrated by the formation of tin-oxysulfide clusters, such as T3-[(RSn)4Sn6O4S16] (R = Bu, Ph, and benzyl = Be) and its variants, truncated T3, viz., TT3-[(BuSn)6Sn3O4S13] and augmented T3, viz., T3-[(Bu3SnS)4Sn6O4S16]. Especially, two extraordinary truncated clusters break the tetrahedral symmetry observed in typical supertetrahedral clusters, further substantiating the advantages offered by the arylstannane approach in expanding cluster chemistry. These organometallic supertetrahedral clusters are highly soluble and stable in common solvents. Additionally, they have tunable third-order nonlinear optical behaviors by controlling the size, heterometallic combination, organic modification, and intercluster interaction.

Machine-learning accelerated structure search for ligand-protected clusters.

Finding low-energy structures of ligand-protected clusters is challenging due to the enormous conformational space and the high computational cost of accurate quantum chemical methods for determining the structures and energies of conformers. Here, we adopted and utilized a kernel rigid regression based machine learning method to accelerate the search for low-energy structures of ligand-protected clusters. We chose the Au25(Cys)18 (Cys: cysteine) cluster as a model system to test and demonstrate our method. We found that the low-energy structures of the cluster are characterized by a specific hydrogen bond type in the cysteine. The different configurations of the ligand layer influence the structural and electronic properties of clusters.

Single-atom tailored atomically-precise nanoclusters for enhanced electrochemical reduction of CO2-to-CO activity

The development of facile tailoring approach to adjust the intrinsic activity and stability of atomically-precise metal nanoclusters catalysts is of great interest but remians challenging. Herein, the well-defined Au8 nanoclusters modified by single-atom sites are rationally synthesized via a co-eletropolymerization strategy, in which uniformly dispersed metal nanocluster and single-atom co-entrenched on the poly-carbazole matrix. Systematic characterization and theoretical modeling reveal that functionalizing single-atoms enable altering the electronic structures of Au8 clusters, which amplifies their electrocatalytic reduction of CO2 to CO activity by ~18.07 fold compared to isolated Au8 metal clusters. The rearrangements of the electronic structure not only strengthen the adsorption of the key intermediates *COOH, but also establish a favorable reaction pathway for the CO2 reduction reaction. Moreover, this strategy fixing nanoclusters and single-atoms on cross-linked polymer networks efficiently deduce the performance deactivation caused by agglomeration during the catalytic process. This work contribute to explore the intrinsic activity and stability improvement of metal clusters.

Synthesis and Single Crystal X-ray Diffraction Structure of an Indium Arsenide Nanocluster.

The discovery of magic-sized clusters as intermediates in the synthesis of colloidal quantum dots has allowed for insight into formation pathways and provided atomically precise molecular platforms for studying the structure and surface chemistry of those materials. The synthesis of monodisperse InAs quantum dots has been developed through the use of indium carboxylate and As(SiMe3)3 as precursors and documented to proceed through the formation of magic-sized intermediates. Herein, we report the synthesis, isolation, and single-crystal X-ray diffraction structure of an InAs nanocluster that is ubiquitous across reports of InAs quantum dot synthesis. The structure, In26As18(O2CR)24(PR'3)3, differs substantially from previously reported semiconductor nanocluster structures even within the III-V family. However, it can be structurally linked to III-V and II-VI cluster structures through the anion sublattice. Further analysis using variable temperature absorbance spectroscopy and support from computation deepen our understanding of the reported structure and InAs nanomaterials as a whole.

Geometrical Stabilities and Electronic Structures of Ru3 Clusters on Rutile TiO2 for Green Hydrogen Production.

In response to the vital requirement for renewable energy alternatives, this research delves into the complex interactions between ruthenium (Ru3) clusters and rutile titanium dioxide (TiO2) (110) interfaces, with the aim of enhancing photocatalytic water splitting processes to produce environmentally friendly hydrogen. As the world shifts away from traditional fossil fuels, this study utilizes the density functional theory (DFT) and the HSE06 hybrid functional to thoroughly assess the geometric and electronic properties of Ru3 clusters on rutile TiO2 (110) surfaces. Given TiO2's renown role as a photocatalyst and its limitations in visible light absorption, this research investigates the potential of metals like Ru to serve as additional catalysts. The results indicate that the triangular Ru3 cluster exhibits exceptional stability and charge transfer effectiveness when loaded on rutile TiO2 (110). Under ideal adsorption scenarios, the cluster undergoes oxidation, leading to subsequent changes in the electronic configuration of TiO2. Further exploration into TiO2 surfaces with defects shows that Ru3 clusters influence the creation of oxygen vacancies, resulting in a greater stabilization of TiO2 and an increase in the energy required for creating oxygen vacancies. Moreover, the attachment of the Ru3 cluster and the creation of oxygen vacancies lead to the emergence of polaronic and hybrid states centered on specific titanium atoms. These states are vital for enhancing the photocatalytic performance of the material within the visible light spectrum. This DFT study provides essential insights into the role of Ru3 clusters as potential supplementary catalysts in TiO2-based photocatalytic systems, setting the stage for practical experiments and the development of highly efficient photocatalysts for sustainable hydrogen generation. The observed effects on electronic structures and oxygen vacancy generation underscore the intricate relationship between Ru3 clusters and TiO2 interfaces, offering a valuable direction for future research in the pursuit of clean and sustainable energy solutions.

Adsorption Mechanisms of TM3 (TM = Mo, Ru, Au)-Decorated Tin Sulfide Monolayers for the Decomposition of Gas Components under Fault Conditions in Oil-Immersed Transformers.

Oil-immersed transformers play a pivotal role owing to their environmentally friendly characteristics, compact footprint, and cost-effectiveness. Ensuring the online monitoring of oil-immersed transformers is a fundamental measure to ensure the secure and stable operation of modern power systems. In this paper, metal particle cluster-doped SnS is firstly used in the adsorption and sensing of decomposition components (CO, C2H2) under fault conditions in oil-immersed transformers. The study comprehensively analyzed band structure, differential charge density, density of states, and molecular orbital theory to unveil the adsorption and sensing mechanisms of target gases. The findings suggest that the modification of metal particle clusters can enhance the surface electronic properties of single-layer SnS. In the regions of metal particle clusters and the gas-surface reaction area, electronic activity is significantly heightened, primarily attributed to the contribution of d-orbital electrons of the metal cluster structures. The modified SnS exhibits adsorption capacity in the following order: Ru3-SnS > Mo3-SnS > Au3-SnS. Additionally, the modified material demonstrates increased competitiveness for C2H2, with adsorption types falling under physical chemistry adsorption. Different metal elements exert diverse effects on the electronic distribution of the entire system, providing a theoretical foundation for the preparation of corresponding sensors. The findings in this work offer numerical insights for the further preparation and development of SnS nanosensors, concurrently shedding light on the online monitoring of faults in oil-immersed transformers.

LiDAR Dynamic Target Detection Based on Multidimensional Features.

To address the limitations of LiDAR dynamic target detection methods, which often require heuristic thresholding, indirect computational assistance, supplementary sensor data, or postdetection, we propose an innovative method based on multidimensional features. Using the differences between the positions and geometric structures of point cloud clusters scanned by the same target in adjacent frame point clouds, the motion states of the point cloud clusters are comprehensively evaluated. To enable the automatic precision pairing of point cloud clusters from adjacent frames of the same target, a double registration algorithm is proposed for point cloud cluster centroids. The iterative closest point (ICP) algorithm is employed for approximate interframe pose estimation during coarse registration. The random sample consensus (RANSAC) and four-parameter transformation algorithms are employed to obtain precise interframe pose relations during fine registration. These processes standardize the coordinate systems of adjacent point clouds and facilitate the association of point cloud clusters from the same target. Based on the paired point cloud cluster, a classification feature system is used to construct the XGBoost decision tree. To enhance the XGBoost training efficiency, a Spearman's rank correlation coefficient-bidirectional search for a dimensionality reduction algorithm is proposed to expedite the optimal classification feature subset construction. After preliminary outcomes are generated by XGBoost, a double Boyer-Moore voting-sliding window algorithm is proposed to refine the final LiDAR dynamic target detection accuracy. To validate the efficacy and efficiency of our method in LiDAR dynamic target detection, an experimental platform is established. Real-world data are collected and pertinent experiments are designed. The experimental results illustrate the soundness of our method. The LiDAR dynamic target correct detection rate is 92.41%, the static target error detection rate is 1.43%, and the detection efficiency is 0.0299 s. Our method exhibits notable advantages over open-source comparative methods, achieving highly efficient and precise LiDAR dynamic target detection.