A5056731969- Advanced battery technologies research
- Advanced Nanomaterials in Catalysis
- Advanced Graph Neural Networks
- Electrochemical sensors and biosensors
- Machine Learning in Materials Science
- Electrocatalysts for Energy Conversion
- Advanced biosensing and bioanalysis techniques
- Advanced Photocatalysis Techniques
- Natural Language Processing Techniques
- Electrochemical Analysis and Applications
- Electronic and Structural Properties of Oxides
- Fault Detection and Control Systems
- Privacy-Preserving Technologies in Data
- Advanced Battery Materials and Technologies
- Clay minerals and soil interactions
- X-ray Diffraction in Crystallography
- Complex Network Analysis Techniques
- Concrete and Cement Materials Research
- Mobile Crowdsensing and Crowdsourcing
- ZnO doping and properties
- Topic Modeling
- TiO2 Photocatalysis and Solar Cells
- Catalytic Processes in Materials Science
- Expert finding and Q&A systems
- Transition Metal Oxide Nanomaterials
Beijing Computational Science Research Center
2018-2025
University of Zurich
2023-2024
Zhengzhou University
2023-2024
University of Alberta
2024
Zhejiang University
2022-2023
Beihang University
2021
Exploring high-performance zeolite-supported metal catalysts is of great significance. Herein, we develop a strategy for fabricating isolated single atomic site in Y zeolite (M-ISAS@Y, M = Pt, Pd, Ru, Rh, Co, Ni, Cu) by situ separating and confining metal-ethanediamine complex into β-cages during the crystallization process followed thermal treatment. The M-ISAS are stabilized skeletal oxygens zeolite, crystallinity, porosity, large surface area well inherited M-ISAS@Y. As demonstration,...
Abstract Although oxygen vacancies (O v s) play a critical role for many applications of metal oxides, controllable synthetic strategy anisotropic O s remains great challenge. Here, novel is proposed to achieve the regional dual structure with at both surface and in interior TiO 2 by constructing amorphous domains. The as‐prepared black domains exhibits superior activity degrading rhodamine B (RhB) solutions, which can instantly decompose RhB just shake. First‐principle simulations reveal...
Anomaly detection in graphs has attracted considerable interests both academia and industry due to its wide applications numerous domains ranging from finance biology. Meanwhile, graph neural networks (GNNs) is emerging as a powerful tool for modeling data. A natural fundamental question that arises here is: can abnormality be detected by networks? In this paper, we aim answer question, which nontrivial. As many existing works have explored, seen filters signals, with the favor of low...
Anti-money laundering (AML) systems play a critical role in safeguarding global economy. As money is considered as one of the top group crimes, there crucial need to discover sub-network behind particular transaction for robust AML system. However, existing rule-based methods discovery heavily based on domain knowledge and may lag modus operandi launderers. Therefore, this work, we first address problem with neural network approach, propose an framework AMAP equipped adaptive proposer. In...
Polarons in metal oxides are localized charge carriers that significantly influence material properties. The critical role of polarons photocatalytic processes arises from their spatial distribution and dynamic In this study, we propose a physically meaningful descriptor based on the potential charged particles polarization field to quantify relative stability polaron configurations during migration. Taking widely studied Rutile-phase TiO2 as typical model, focused electron addition, where...
Abstract Determination of the atomic structure inorganic single-walled nanotubes with complex stoichiometry remains elusive due to too many coordinates be fitted respect X-ray diffractograms inherently exhibiting rather broad features. Here we introduce a methodology reduce number variables and enable resolution for stoichiometry. We apply it recently synthesized methylated aluminosilicate aluminogermanate imogolite nominal composition (OH) 3 Al 2 O Si(Ge)CH . Fitting scattering diagrams,...
Metallic zinc-based anodes often encounter challenges such as dendrite growth, side reactions, and by-product generation, leading to diminished reversibility. This study introduces a novel approach by introducing methyl acetate cosolvent into an aqueous electrolyte based on Zn(OTf)2 through salting-in effect. Regulating the solvation sheath structure of Zn2+ enables formation cathode interphase anode interphase, enhancing stability vanadium-base zinc metal anode, respectively. The...
In electrochemical experiments, the number of electrons electrode immersed in electrolyte is usually variable. Additionally, numbers adsorbed substances on surface electrode, solvent molecules, and counter charge ions near-surface region can also vary. Treating solid-liquid interfaces with typical fixed electron density functional theory (DFT) approach tends to be a challenge. This addressed by using grand canonical ensemble approaches. We present implementation two approaches open-source...
The discovery of an original structure the water at inner surface inorganic aluminogermanate nanotubes and its specific dynamics are reported, based on density functional theory molecular inelastic neutron scattering.
Time series modeling has attracted great research interests in the last decades. Among literature, shapelet-based models aim to extract representative subsequences, and could offer explanatory insights. In order capture shapelet dynamics evolutions, we propose a novel framework of bridging time representation learning graph modeling, with two different implementations. We first formulate process extracting time-aware shapelets, then briefly introduce key idea transforming data into evolution...
Transition metal hydr(oxy)oxides (TMHs) are considered efficient electrocatalysts for the oxygen evolution reaction (OER) under alkaline conditions. Toward identification of potential descriptors to circumvent scaling relation limit OER, first-principles calculations were used quantify effects on overpotential different s (Mg), p (Al), and d (Ti, V, Cr, Fe, Co, Sc, Zn) electron dopants in Ni-based TMHs. Both adsorbate mechanism (AEM) lattice oxygen-mediated (LOM) examined. The results...
We present the implementation of Hubbard (
A convenient and highly sensitive colorimetric sensing platform for Cr( vi ) assay based on a Cu-curcumin nanozyme.
The advancement of Large Language Models (LLMs) has remarkably pushed the boundaries towards artificial general intelligence (AGI), with their exceptional ability on understanding diverse types information, including but not limited to images and audio. Despite this progress, a critical gap remains in empowering LLMs proficiently understand reason graph data. Recent studies underscore LLMs' underwhelming performance fundamental reasoning tasks. In paper, we endeavor unearth obstacles that...
We investigate termination effects in aluminosilicate (AlSi) and aluminogermanate (AlGe) imogolite nanotubes (NTs) by means of semi-local range-corrected hybrid Density Functional Theory (DFT) simulations. Following screening identification the smallest finite model capable accommodating full relaxation NT terminations around an otherwise geometrically electrostatically unperturbed core region, we quantify discuss physical truncation on structure, relative energy, electrostatics electronic...
Nickel–iron (Ni–Fe) hydroxides have received much attention as abundant and efficient electrocatalysts for the oxygen evolution reaction (OER) under alkaline conditions.
Polarons play a vital role for many energy-conversion and storage processes; it is accordingly imperative to understand their system-specific formation reactivity. In first-principles simulations, the of polarons can be modeled by either breaking symmetry initial structure or imposing an implicit explicit constraining potential on local orbital occupation. this work, we propose subspace occupancy-constraining (SOCP) approach simulate through occupancy number relevant orbitals. We illustrate...
In this paper, we introduce "InfiAgent-DABench", the first benchmark specifically designed to evaluate LLM-based agents in data analysis tasks. This contains DAEval, a dataset consisting of 311 questions derived from 55 CSV files, and an agent framework LLMs as agents. We adopt format-prompting technique, ensuring be closed-form that can automatically evaluated. Our extensive benchmarking 23 state-of-the-art uncovers current challenges encountered addition, have developed DAAgent,...
This work proposes DyExpert, a dynamic graph model for cross-domain link prediction. It can explicitly historical evolving processes to learn the evolution pattern of specific downstream and subsequently make pattern-specific predictions. DyExpert adopts decode-only transformer is capable efficiently parallel training inference by \textit{conditioned generation} that integrates both modeling trained extensive graphs across diverse domains, comprising 6M edges. Extensive experiments on eight...
Recently, large language models (LLMs) have demonstrated superior capabilities in understanding and zero-shot learning on textual data, promising significant advances for many text-related domains. In the graph domain, various real-world scenarios also involve where tasks node features can be described by text. These text-attributed graphs (TAGs) broad applications social media, recommendation systems, etc. Thus, this paper explores how to utilize LLMs model TAGs. Previous methods TAG...
We present Expert-Token-Routing, a unified generalist framework that facilitates seamless integration of multiple expert LLMs. Our represents LLMs as special tokens within the vocabulary meta LLM. The LLM can route to an like generating new tokens. Expert-Token-Routing not only supports learning implicit expertise from existing instruction dataset but also allows for dynamic extension in plug-and-play manner. It conceals detailed collaboration process user's perspective, facilitating...