A5011972821- X-ray Diffraction in Crystallography
- Topic Modeling
- Natural Language Processing Techniques
- Speech and dialogue systems
- Theoretical and Computational Physics
- Fault Detection and Control Systems
- Advanced Condensed Matter Physics
- Computational and Text Analysis Methods
- Nuclear Physics and Applications
- Spectroscopy and Chemometric Analyses
- Crystallization and Solubility Studies
- Multiferroics and related materials
- Quantum Dots Synthesis And Properties
- Ferroelectric and Piezoelectric Materials
- Magnetic and transport properties of perovskites and related materials
- Neural Networks and Applications
- Machine Learning in Materials Science
- nanoparticles nucleation surface interactions
- Evaluation Methods in Various Fields
- High-pressure geophysics and materials
- Advanced X-ray and CT Imaging
- Metallic Glasses and Amorphous Alloys
- Advanced Photocatalysis Techniques
- Geochemistry and Geologic Mapping
- Perovskite Materials and Applications
Columbia University
2013-2015
Tsinghua University
1995-2014
Newcastle University
1995
A strategy is described for regularizing ill posed structure and nanostructure scattering inverse problems (i.e. solution) from complex material structures. This paper describes both the philosophy of approach, a software implementation, DiffPy Complex Modeling Infrastructure (DiffPy-CMI).
We report the structure of methylammonium lead(II) iodide perovskite in mesoporous TiO2, as used high-performance solar cells. Pair distribution function analysis X-ray scattering reveals a two component nanostructure: one with medium range crystalline order (30 atom %) and another only local structural coherence (70 %). The nanostructuring correlates blueshift absorption onset increases photoluminescence. Our findings underscore importance fully characterizing controlling for improved cell...
Luminescent semiconducting quantum dots (QDs) are central to emerging technologies that range from tissue imaging solid-state lighting. However, existing samples heterogeneous, which has prevented atomic-resolution determination of their structures and obscured the relationship between atomic electronic structures. Here we report synthesis, isolation, structural characterization three cadmium selenide QDs with uniform compositions (Cd35Se20(X)30(L)30, Cd56Se35(X)42(L)42, Cd84Se56(X)56(L)56;...
The xPDFsuite software program is described. It for processing and analyzing atomic pair distribution functions (PDF) from X-ray powder diffraction data. provides a convenient GUI SrXplanr PDFgetX3, allowing the users to easily obtain 1D pattern raw 2D images then transform them PDFs. also bundles PDFgui which allows create structure models fit experiment specially useful working with large numbers of datasets such as high throughout measurements. Some key features are: real time PDF...
The atomic pair distribution function (PDF) analysis of X-ray powder diffraction data has been used to study the structure small and ultra-small CdSe nanoparticles. A method is described that uses a wurtzite zinc-blende mixed phase model account for stacking faults in particles. mixed-phase successfully describes nanoparticles larger than 2 nm yielding fault density about 30%. However, ultrasmall smaller nm, models cannot fit experimental PDF showing significantly modified from particles...
The analytical form of the magnetic pair distribution function (mPDF) is derived for first time by computing Fourier transform neutron scattering cross section from an arbitrary collection moments. Similar to atomic applied study structure, mPDF reveals both short-range and long-range correlations directly in real space. This experimentally accessible yields even when they are only ordered. evaluated various example cases build intuitive understanding how different patterns will appear mPDF.
We report the use of X-ray diffraction in combination with computed tomography to provide quantitative information a coin cell Li-ion battery and commercial Ni/MH AAA for first time. This technique allows structural be garnered opens up possibility tracking nanostructural changes operandi. In case cylindrically wound, standard cell, we were able map all different phases complex geometry, including anode, cathode, current collector casing, as well amorphous such binder separator. battery,...
Discovery of new complex oxides that exhibit both magnetic and ferroelectric properties is great interest for the design functional magnetoelectrics, in which research driven by technologically exciting prospect controlling charges fields spins applied voltages, sensors, 4-state logic, spintronics. Motivated notion a tool-kit oxide design, we developed chemical synthesis strategy single-phase multifunctional lattices. Here, introduce class multiferroic hollandite Ba-Mn-Ti not apparent...
We demonstrate a remarkable equivalence in structure measured by total X-ray scattering methods between very small metallic nanoparticles and bulk glasses (BMGs), thus connecting two disparate fields, shedding new light on both. Our results show that for nanoparticle diameters <5 nm the of Ni changes from fcc to characteristic BMG-like structure, despite them being formed single element, an effect we call nano-metallic glass (NMG) formation. However, high-resolution TEM images NMG clusters...
This paper explores optimal methods for obtaining one-dimensional (1D) powder pattern intensities from two-dimensional (2D) planar detectors with good estimates of their standard deviations. We describe to estimate uncertainties when the same image is measured in multiple frames as well a single frame. show importance considering correlation diffraction points during integration and re-sampling process data analysis. find that correlations between adjacent pixels can lead seriously...
This paper describes a methodology for semi-supervised learning of dialogue acts using the similarity between sentences. We suppose that sentences with same act are more similar in terms semantic and syntactic information. However, previous work on sentence mainly modeled as bag-of-words then compared different groups words corpus-based or knowledge-based measurements word similarity. Novelly, we present vector-space representation, composed embeddings, is, related distributed...
Topic modeling has been a widely used tool for unsupervised text analysis. However, comprehensive evaluations of topic model remain challenging. Existing evaluation methods are either less comparable across different models (e.g., perplexity) or focus on only one specific aspect quality document representation quality) at time, which is insufficient to reflect the overall performance. In this paper, we propose WALM (Words Agreement with Language Model), new method that comprehensively...
Topic modeling is a fundamental task in natural language processing, allowing the discovery of latent thematic structures text corpora. While Large Language Models (LLMs) have demonstrated promising capabilities topic discovery, their direct application to suffers from issues such as incomplete coverage, misalignment topics, and inefficiency. To address these limitations, we propose LLM-ITL, novel LLM-in-the-loop framework that integrates LLMs with many existing Neural (NTMs). In global...
The primary challenge of multi-label active learning, differing it from multi-class lies in assessing the informativeness an indefinite number labels while also accounting for inherited label correlation. Existing studies either require substantial computational resources to leverage correlations or fail fully explore dependencies. Additionally, real-world scenarios often addressing intrinsic biases stemming imbalanced data distributions. In this paper, we propose a new learning strategy...
The key task in spoken language understanding research is the semantic tagging of sequences. Deep belief networks have recently shown great performance word-labeling tasks while conditional random field has been a successful approach to model probabilities sequences global fashion. In contrast CRFs, DBNs are optimized based on tag-by-tag likelihood locally normalized way and may suffer from label bias problem. this paper, we combine DBN CRF by employing top hidden layer DBN. This DBN-CRF...
One of the key components in spoken dialog systems is semantic slot-filling, a sequence tagging task. There are several state-of-the-art supervised approaches to model slot-filling problem such as conditional random fields (CRF), support vector machine (SVM) and stochastic finite state transducers (SFST). A general way improve their performance use unsupervised word embeddings extra input features. In this paper we evaluate two kinds on all three for slot-filling. We near baselines, find...
Ni-Pd nanoparticles synthesized for CO catalysis are characterized by transmission electron microscopy and total X-ray scattering. The sizes of these can be tuned to size with great control over the monodispersity samples. pair distribution functions reveal a local ordering within highly disordered atomic structure nanoparticles. PDFs show size-dependent deviation from typical bulk face centered cubic (fcc) materials. long-range isotropic disorder non-fcc fitted using an exponentially damped...