Link prediction using low-dimensional node embeddings: The measurement problem.

Graph representation learning is a fundamental technique for machine learning (ML) on complex networks. Given an input network, these methods represent the vertices by low-dimensional real-valued vectors. These vectors can be used for a multitude of downstream ML tasks. We study one of the most important such task, link prediction. Much of the recent literature on graph representation learning has shown remarkable success in link prediction. On closer investigation, we observe that the performance is measured by the AUC (area under the curve), which suffers biases. Since the ground truth in link prediction is sparse, we design a vertex-centric measure of performance, called the VCMPR@k plots. Under this measure, we show that link predictors using graph representations show poor scores. Despite having extremely high AUC scores, the predictors miss much of the ground truth. We identify a mathematical connection between this performance, the sparsity of the ground truth, and the low-dimensional geometry of the node embeddings. Under a formal theoretical framework, we prove that low-dimensional vectors cannot capture sparse ground truth using dot product similarities (the standard practice in the literature). Our results call into question existing results on link prediction and pose a significant scientific challenge for graph representation learning. The VCMPR plots identify specific scientific challenges for link prediction using low-dimensional node embeddings.

scMMT: a multi-use deep learning approach for cell annotation, protein prediction and embedding in single-cell RNA-seq data.

Accurate cell type annotation in single-cell RNA-sequencing data is essential for advancing biological and medical research, particularly in understanding disease progression and tumor microenvironments. However, existing methods are constrained by single feature extraction approaches, lack of adaptability to immune cell types with similar molecular profiles but distinct functions and a failure to account for the impact of cell label noise on model accuracy, all of which compromise the precision of annotation. To address these challenges, we developed a supervised approach called scMMT. We proposed a novel feature extraction technique to uncover more valuable information. Additionally, we constructed a multi-task learning framework based on the GradNorm method to enhance the recognition of challenging immune cells and reduce the impact of label noise by facilitating mutual reinforcement between cell type annotation and protein prediction tasks. Furthermore, we introduced logarithmic weighting and label smoothing mechanisms to enhance the recognition ability of rare cell types and prevent model overconfidence. Through comprehensive evaluations on multiple public datasets, scMMT has demonstrated state-of-the-art performance in various aspects including cell type annotation, rare cell identification, dropout and label noise resistance, protein expression prediction and low-dimensional embedding representation.

Porous lanthanide metal-organic frameworks with metallic conductivity.

Metallic charge transport and porosity appear almost mutually exclusive. Whereas metals demand large numbers of free carriers and must have minimal impurities and lattice vibrations to avoid charge scattering, the voids in porous materials limit the carrier concentration, provide ample space for impurities, and create more charge-scattering vibrations due to the size and flexibility of the lattice. No microporous material has been conclusively shown to behave as a metal. Here, we demonstrate that single crystals of the porous metal-organic framework Ln1.5(2,3,6,7,10,11-hexaoxytriphenylene) (Ln = La, Nd) are metallic. The materials display the highest room-temperature conductivities of all porous materials, reaching values above 1,000 S/cm. Single crystals of the compounds additionally show clear temperature-deactivated charge transport, a hallmark of a metallic material. Lastly, a structural transition consistent with charge density wave ordering, present only in metals and rare in any materials, provides additional conclusive proof of the metallic nature of the materials. Our results provide an example of a metal with porosity intrinsic to its structure. We anticipate that the combination of porosity and chemical tunability that these materials possess will provide a unique handle toward controlling the unconventional states that lie within them, such as charge density waves that we observed, or perhaps superconductivity.

Solvent-Driven Self-Assembly of One-Dimensional Lepidocrocite Titanium-Oxide-Based Nanofilaments.

Herein, the self-assembly of one-dimensional titanium oxide lepidocrocite nanofilaments in 10 different water miscible organic solvents was investigated. The nanofilament snippets, with minimal cross sections of ∼5 × 7 Å2 and lengths around 30 nm, begin as an aqueous colloidal suspension. Upon addition, and brief mixing, of the colloidal suspension into a given solvent, a multitude of morphologies─seemingly based on the hydrophilicity and polarity of the solvent─emerge. These morphologies vary between sheets, highly networked webs, and discrete fibers, all with no apparent change in the lepidocrocite structure. On the micro- and nanoscale, the morphologies are reminiscent of biological, rather than inorganic, materials. The results of this work give insight into the self-assembly of these materials and offer new pathways for novel macrostructures/morphologies assembled from these highly adsorbent and catalytically active low-dimensional materials.

Anisotropic Electron-Phonon Interactions in 2D Lead-Halide Perovskites.

Two-dimensional (2D) hybrid organic-inorganic metal halide perovskites offer enhanced stability for perovskite-based applications. Their crystal structure's soft and ionic nature gives rise to strong interaction between charge carriers and ionic rearrangements. Here, we investigate the interaction of photogenerated electrons and ionic polarizations in single-crystal 2D perovskite butylammonium lead iodide (BAPI), varying the inorganic lamellae thickness in the 2D single crystals. We determine the directionality of the transition dipole moments (TDMs) of the relevant phonon modes (in the 0.3-3 THz range) by the angle- and polarization-dependent THz transmission measurements. We find a clear anisotropy of the in-plane photoconductivity, with a ∼10% reduction along the axis parallel with the transition dipole moment of the most strongly coupled phonon. Detailed calculations, based on Feynman polaron theory, indicate that the anisotropy originates from directional electron-phonon interactions.

Dimensionality Engineering of Magnetic Anisotropy from the Anomalous Hall Effect in Synthetic SrRuO3 Crystals.

Magnetic anisotropy in atomically thin correlated heterostructures is essential for exploring quantum magnetic phases for next-generation spintronics. Whereas previous studies have mostly focused on van der Waals systems, here we investigate the impact of dimensionality of epitaxially grown correlated oxides down to the monolayer limit on structural, magnetic, and orbital anisotropies. By designing oxide superlattices with a correlated ferromagnetic SrRuO3 and nonmagnetic SrTiO3 layers, we observed modulated ferromagnetic behavior with the change of the SrRuO3 thickness. Especially, for three-unit-cell-thick layers, we observe a significant 1500% improvement of the coercive field in the anomalous Hall effect, which cannot be solely attributed to the dimensional crossover in ferromagnetism. The atomic-scale heterostructures further reveal the systematic modulation of anisotropy for the lattice structure and orbital hybridization, explaining the enhanced magnetic anisotropy. Our findings provide valuable insights into engineering the anisotropic hybridization of synthetic magnetic crystals, offering a tunable spin order for various applications.

Structural approach to charge density waves in low-dimensional systems: electronic instability and chemical bonding.

The charge density wave (CDW) instability, usually occurring in low-dimensional metals, has been a topic of interest for longtime. However, some very fundamental aspects of the mechanism remain unclear. Recently, a plethora of new CDW materials, a substantial fraction of which is two-dimensional or even three-dimensional, has been prepared and characterised as bulk and/or single-layers. As a result, the need for revisiting the primary mechanism of the instability, based on the electron-hole instability established more than 50 years ago for quasi-one-dimensional (quasi-1D) conductors, has clearly emerged. In this work, we consider a large number of CDW materials to revisit the main concepts used in understanding the CDW instability, and emphasise the key role of the momentum dependent electron-phonon coupling in linking electronic and structural degrees of freedom. We argue that for quasi-1D systems, earlier weak coupling theories work appropriately and the energy gain due to the CDW and the concomitant periodic lattice distortion (PLD) remains primarily due to a Fermi surface nesting mechanism. However, for materials with higher dimensionality, intermediate and strong coupling regimes are generally at work and the modification of the chemical bonding network by the PLD is at the heart of the instability. We emphasise the need for a microscopic approach blending condensed matter physics concepts and state-of-the-art first-principles calculations with quite fundamental chemical bonding ideas in understanding the CDW phenomenon in these materials.

The Formation and Transformation of Low-Dimensional Carbon Nanomaterials by Electron Irradiation.

Low-dimensional materials based on graphene or graphite show a large variety of phenomena when they are subjected to irradiation with energetic electrons. Since the 1990s, electron microscopy studies, where a certain irradiation dose is unavoidable, have witnessed unexpected structural transformations of graphitic nanoparticles. It is recognized that electron irradiation is not only detrimental but also bears considerable potential in the formation of new graphitic structures. With the availability of aberration-corrected electron microscopes and the discovery of techniques to produce monolayers of graphene, detailed insight into the atomic processes occurring during electron irradiation became possible. Threshold energies for atom displacements are determined and models of different types of lattice vacancies are confirmed experimentally. However, experimental evidence for the configuration of interstitial atoms in graphite or adatoms on graphene remained indirect, and the understanding of defect dynamics still depends on theoretical concepts. This article reviews irradiation phenomena in graphene- or graphite-based nanomaterials from the scale of single atoms to tens of nanometers. Observations from the 1990s can now be explained on the basis of new results. The evolution of the understanding during three decades of research is presented, and the remaining problems are pointed out.

What Makes On-Chip Microdevices Stand Out in Electrocatalysis?

Clean and sustainable energy conversion and storage through electrochemistry shows great promise as an alternative to traditional fuel or fossil-consumption energy systems. With regards to practical and high-efficient electrochemistry application, the rational design of active sites and the accurate description of mechanism remain a challenge. Toward this end, in this Perspective, a unique on-chip micro/nano device coupling nanofabrication and low-dimensional electrochemical materials is presented, in which material structure analysis, field-effect regulation, in situ monitoring, and simulation modeling are highlighted. The critical mechanisms that influence electrochemical response are discussed, and how on-chip micro/nano device distinguishes itself is emphasized. The key challenges and opportunities of on-chip electrochemical platforms are also provided through the Perspective.

Thermal Property of Fullerene Fibers: One-Dimensional Material with Exceptional Thermal Performance.

The recent groundbreaking achievement in the synthesis of large-sized single crystal C60 monolayer, which is covalently bonded in a plane using C60 as building blocks. The asymmetric lattice structure endows it with anisotropic phonon modes and conductivity. If these C60 are arranged in form of 1D fiber, the improved manipulation of phonon conduction along the fiber axis could be anticipated. Here, thermal properties of C60-fiber, including thermal transfer along the C60-fiber axis and across the interlayer interface are investigated using molecular dynamic simulations. Taking advantage of the distinctively hollow spherical structure of C60 building blocks, the spherical structure deformation and encapsulation induced thermal reduction can be up to 56% and 80%, respectively. By applying external electronic fields in H2O@C60 model, its thermal conductivity decreases up to 60%, which realizes the contactless thermal regulation. ln particular, the thermal rectification phenomenon is discovered by inserting atoms/molecules in C60 with a rational designed mass-gradient, and its maximum thermal rectification factor is predicted to ≈45%. These investigations aim to achieve effective regulation of the thermal conductivity of C60-fibers. This work showcases the potential of C60-fiber in the realms of thermal management and thermal sensing, paving the way to C60-based functional materials.

Strategies for the Preparation of Single-Atom Catalysts Using Low-Dimensional Metal-Organic Frameworks.

As single-atom catalysts are important energy materials, their preparation and synthesis methods have become particularly important. The unique structures of low-dimensional metal-organic frameworks and their derivatives provide various strategies for preparing single-atom catalysts. This paper summarizes various strategies for the preparation of single-atom catalysts based on low-dimensional metal-organic frameworks and their derivatives.

Dual Defect Passivation at the Buried Interface for Printable Mesoscopic Perovskite Solar Cells with Reduced Open-Circuit Voltage Loss.

Numerous defects exist at the buried interface between the perovskite and adjacent electron transport layers in perovskite solar cells, resulting in severe non-radiative recombination and excessive open-circuit voltage (VOC) loss. Herein, a dual defect passivation strategy utilizing guanidine sulfate (GUA2SO4) as an interface modifier is first reported. On the one hand, the SO4 2- preferentially interacts with Pb-related defects, generating water-insoluble lead oxysalts complexes. Additionally, GUA+ diffuses into the perovskite and induces the formation of low-dimensional perovskite. These reactions effectively suppress trap states at the buried interface and perovskite boundaries in printable mesoscopic perovskite solar cells (p-MPSCs), thus increasing the carrier lifetime. Meanwhile, GUA2SO4 optimizes the interface energy band alignment, thus accelerating the charge extraction and transfer at the buried interface. This synergistic effect of trap passivation and interface energy band alignment modulation is strongly demonstrated by an increase in average VOC of 70 mV and the power conversion efficiency improvement from 17.51% to 18.70%. This work provides a novel approach to efficiently improve the performance of p-MPSCs through dual-targeted defect passivation at the buried interface.

Recent Progress of Low-Dimensional Metal-Organic Frameworks for Aqueous Zinc-Based Batteries.

Aqueous zinc-based batteries (AZBs) are promising energy storage solutions with remarkable safety, abundant Zn reserve, cost-effectiveness, and relatively high energy density. However, AZBs still face challenges such as anode dendrite formation that reduces cycling stability and limited cathode capacity. Recently, low-dimensional metal-organic frameworks (LD MOFs) and their derivatives have emerged as promising candidates for improving the electrochemical performance of AZBs owing to their unique morphologies, high structure tunability, high surface areas, and high porosity. However, clear guidelines for developing LD MOF-based materials for high-performance AZBs are scarce. In this review, the recent progress of LD MOF-based materials for AZBs is critically examined. The typical synthesis methods and structural design strategies for improving the electrochemical performance of LD MOF-based materials for AZBs are first introduced. The recent noteworthy research achievements are systematically discussed and categorized based on their applications in different AZB components, including cathodes, anodes, separators, and electrolytes. Finally, the limitations are addressed and the future perspectives are outlined for LD MOFs and their derivatives in AZB applications. This review provides clear guidance for designing high-performance LD MOF-based materials for advanced AZBs.

Fully connected autoencoder and convolutional neural network with attention-based method for inferring disease-related lncRNAs.

Since abnormal expression of long noncoding RNAs (lncRNAs) is often closely related to various human diseases, identification of disease-associated lncRNAs is helpful for exploring the complex pathogenesis. Most of recent methods concentrate on exploiting multiple kinds of data related to lncRNAs and diseases for predicting candidate disease-related lncRNAs. These methods, however, failed to deeply integrate the topology information from the meta-paths that are composed of lncRNA, disease and microRNA (miRNA) nodes. We proposed a new method based on fully connected autoencoders and convolutional neural networks, called ACLDA, for inferring potential disease-related lncRNA candidates. A heterogeneous graph that consists of lncRNA, disease and miRNA nodes were firstly constructed to integrate similarities, associations and interactions among them. Fully connected autoencoder-based module was established to extract the low-dimensional features of lncRNA, disease and miRNA nodes in the heterogeneous graph. We designed the attention mechanisms at the node feature level and at the meta-path level to learn more informative features and meta-paths. A module based on convolutional neural networks was constructed to encode the local topologies of lncRNA and disease nodes from multiple meta-path perspectives. The comprehensive experimental results demonstrated ACLDA achieves superior performance than several state-of-the-art prediction methods. Case studies on breast, lung and colon cancers demonstrated that ACLDA is able to discover the potential disease-related lncRNAs.

Density-independent plasmons for terahertz-stable topological metamaterials.

To efficiently integrate cutting-edge terahertz technology into compact devices, the highly confined terahertz plasmons are attracting intensive attention. Compared to plasmons at visible frequencies in metals, terahertz plasmons, typically in lightly doped semiconductors or graphene, are sensitive to carrier density (n) and thus have an easy tunability, which leads to unstable or imprecise terahertz spectra. By deriving a simplified but universal form of plasmon frequencies, here, we reveal a unified mechanism for generating unusual n-independent plasmons (DIPs) in all topological states with different dimensions. Remarkably, we predict that terahertz DIPs can be excited in a two-dimensional nodal line and one-dimensional nodal point systems, confirmed by the first-principle calculations on almost all existing topological semimetals with diverse lattice symmetries. Besides n-independence, the feature of Fermi velocity and degeneracy factor dependencies in DIPs can be applied to design topological superlattice and multiwalled carbon nanotube metamaterials for broadband terahertz spectroscopy and quantized terahertz plasmons, respectively. Surprisingly, high spatial confinement and quality factor, also insensitive to n, can be simultaneously achieved in these terahertz DIPs. Our findings pave the way for developing topological plasmonic devices for stable terahertz applications.

Structures, fundamental properties, and potential applications of low-dimensional C60 polymers and other nanocarbons: a review.

Since carbon (C) atom has a variety of chemical bonds via hybridization between s and p atomic orbitals, it is well known that there are robust carbon materials. In particular, discovery of C60 has been an epoch making to cultivate nanocarbon fields. Since then, nanocarbon materials such as nanotube and graphene have been reported. It is interesting to note that C60 is soluble and volatile unlike nanotube and graphene. This indicates that C60 film is easy to be produced on any kinds of substrates, which is advantage for device fabrication. In particular, electron-/photo-induced C60 polymerization finally results in formation of one-dimensional (1D) metallic peanut-shaped and 2D dumbbell-shaped semiconducting C60 polymers, respectively. This enables us to control the physicochemical properties of C60 films using electron-/photo-lithography techniques. In this review, we focused on the structures, fundamental properties, and potential applications of the low-dimensional C60 polymers and other nanocarbons such as C60 peapods, wavy-structured graphene, and penta-nanotubes with topological defects. We hope this review will provide new insights for producing new novel nanocarbon materials and inspire broad readers to cultivate new further research in carbon materials.

We review the structures, fundamental properties, and applications of low-dimensional C₆₀ polymers and other related nanocarbons such as C₆₀ peapods, wavy-structured graphene, and penta-nanotubes from a standpoint of topological defects.

Metal-Free Hydrazinium Halide Perovskitoid Single Crystals for X-ray Detection.

Metal-free perovskitoids (MFPs) with N2H5+ as B-site component possess higher crystal density and hydrogen bonding networks and have been recently expanded into X-ray detection. However, research on this material is in its infancy and lacks an understanding of the function of halide components on physical properties and device performance. Here, N2H5-based MFP single crystals (SCs) with different halides are fabricated, and the influence of halides on the crystal structure, band nature, charge transport characteristics, and final device performance is actively explored. Based on theory and experiments, the tolerance factor and octahedral factor jointly determine the octahedral composition. Further, halides with different electronegativities and ionic radii also affect octahedral distortion and energy band bending, further influencing carrier transport and device performance. Finally, a sensitivity of 1284 µC Gyair-1 cm-2 and low detection limits (LoD) of 5.62 µGyair s-1 were obtained by the Br-based device due to its superior physical properties.

Role of Functional Groups in Tuning Luminescence Signature of Solution-Processed Graphene Quantum Dots: Experimental and Theoretical Insights.

Zero-dimensional graphene quantum dots (GQDs) present unique optoelectronic properties in the large-spectrum range from UV to visible. However, the origin of luminescence in GQDs is still a debatable question. Therefore, the present work investigates the features of trap-mediated and edge-state-functionalized group-associated luminescence enhancement of GQDs. The attached functional groups' involvement in the upsurge of photoluminescence has been discussed theoretically as well as experimentally. In addition, the role of the aromatic ring, the functional group attached, and their positions of attachment to the aromatic ring to tune the emission wavelength and Raman modes have been elucidated theoretically as well as experimentally. We found that in the case of the -OH group attached outside of the aromatic ring, the long-range π hybridization dominates, which suggests that the emission from this model can be dictated by long-range π hybridization. In particular, we found that oxygen-containing functional groups attached outside of the aromatic ring are the main source of the luminescence signature in GQDs. Furthermore, density functional theory (DFT) indicates that the -OH functional group attached outside of the aromatic ring perfectly matched with our experimental results, as the experimental bandgap (2.407 eV) is comparable with the theoretical simulated bandgap (2.399 eV) of the -OH group attached outside of the aromatic ring.

Boosting Blue Self-Trapped Exciton Emission in All-Inorganic Zero-Dimensional Metal Halide Cs2ZnCl4 via Zirconium (IV) Doping.

Low-dimensional metal halides with efficient luminescence properties have received widespread attention recently. However, nontoxic and stable low-dimensional metal halides with efficient blue emission are rarely reported. We used a solvothermal synthesis method to synthesize tetravalent zirconium ion-doped all-inorganic zero-dimensional Cs2ZnCl4 for the first time. Bright blue emission in the range of 370 nm-700 nm with a emission maximum at 456 nm was observed in Zr4+:Cs2ZnCl4 accompanied by a large Stokes shift, which was due to self-trapped excitons (STEs) caused by the lattice vibrations of the twisted structure. Simultaneously, the PLQY of Zr4+:Cs2ZnCl4 achieve an impressive 89.67%, positioning it as a compelling contender for future applications in blue-light technology.

On Casimir and Helmholtz Fluctuation-Induced Forces in Micro- and Nano-Systems: Survey of Some Basic Results.

Fluctuations are omnipresent; they exist in any matter, due either to its quantum nature or to its nonzero temperature. In the current review, we briefly cover the quantum electrodynamic Casimir (QED) force as well as the critical Casimir (CC) and Helmholtz (HF) forces. In the QED case, the medium is usually a vacuum and the massless excitations are photons, while in the CC and HF cases the medium is usually a critical or correlated fluid and the fluctuations of the order parameter are the cause of the force between the macroscopic or mesoscopic bodies immersed in it. We discuss the importance of the presented results for nanotechnology, especially for devising and assembling micro- or nano-scale systems. Several important problems for nanotechnology following from the currently available experimental findings are spelled out, and possible strategies for overcoming them are sketched. Regarding the example of HF, we explicitly demonstrate that when a given integral quantity characterizing the fluid is conserved, it has an essential influence on the behavior of the corresponding fluctuation-induced force.