A5100750437- Machine Learning in Materials Science
- Rare-earth and actinide compounds
- Inorganic Chemistry and Materials
- Magnetic Properties and Synthesis of Ferrites
- X-ray Diffraction in Crystallography
- Graphene research and applications
- Diamond and Carbon-based Materials Research
- Computational Drug Discovery Methods
- Characterization and Applications of Magnetic Nanoparticles
- Carbon Nanotubes in Composites
- biodegradable polymer synthesis and properties
- Nuclear Materials and Properties
- Enzyme Catalysis and Immobilization
- Biochemical and Molecular Research
- Enzyme Structure and Function
- Iron oxide chemistry and applications
- Magnetic Properties of Alloys
- Biofuel production and bioconversion
- Advanced Thermoelectric Materials and Devices
- Water Quality Monitoring and Analysis
- Polysaccharides and Plant Cell Walls
- Thermal Expansion and Ionic Conductivity
- Powder Metallurgy Techniques and Materials
- Iron-based superconductors research
- Nerve injury and regeneration
Columbia University
2023-2025
Cooper Union
2022-2023
Kookmin University
2015-2017
Hanbat National University
2014
Pusan National University
1996-2013
Korea Advanced Institute of Science and Technology
2006
Sungkyunkwan University
2004
Materials informatics employs data-driven approaches for analysis and discovery of materials. Features also referred to as descriptors are essential in generating reliable accurate machine-learning models. While general data can be obtained through public commercial sources, features must tailored specific applications. Common featurizers suitable generic chemical problems may not effective features-property mapping solid-state materials with ML Here, we have assembled the Oliynyk property...
Crystal structure classification of binary intermetallic structures with 1:3 stoichiometry was done simple machine learning algorithms. The successful crystal segregation is attributed to the novel set descriptors comprising both compositional and structural features. dataset includes 97 features, a total 2366 reported compounds adopting six different types. unsupervised method based on principal component analysis (PCA) followed by clustering using K-means applied cluster belonging Using...
Abstract Machine learning models as part of artificial intelligence have enjoyed a recent surge in answering long-standing challenge thermoelectric materials research. That is to produce stable, and highly efficient, for their application devices commercial use. The enhancements these offer the potential identify best solutions challenges accelerate research through reduction experimental computational costs. This perspective underscores examines advancements approaches from community...
A new ternary rare-earth indide, ErCo2In, was synthesized by arc-melting and subsequent annealing at 1070 K for 720 h. The compound extends the RECo2In (RE = Y, Pr, Nd, Sm, Gd, Tb, Dy, Ho) series. Single-crystal X-ray diffraction revealed ErCo2In to crystallize in TbCo2In-type (a coloring variant of PrCo2Ga-type) structure type oP8, space group Pmma, Wyckoff sequence f2ea, а 4.999(4), b 4.029(3) c 7.078(5) Å. crystal characterization with methods further supplemented DFT materials...
Crystal structure classification of binary intermetallic structures with 1:3 stoichiometry was done machine learning algorithms. The data set included 97 features and a total 2366 reported compounds adopting six different types. An unsupervised method based on principal component analysis (PCA) followed by clustering using the K-means applied to cluster belonging With recommendation engine, we predicted expansion clusters then identified cluster/structure-type overlap. PuNi3-type among...
A new ternary rare-earth indide, ErCo2In, was synthesized by arc-melting and subsequent annealing at 1070 K for 720 h. The compound extends the RECo2In (RE = Y, Pr, Nd, Sm, Gd, Tb, Dy, Ho) series. Single-crystal X-ray diffraction revealed ErCo2In to crystallize in TbCo2In-type (a coloring variant of PrCo2Ga-type) structure type oP8, space group Pmma, Wyckoff sequence f2ea, а 4.999(4), b 4.029(3) c 7.078(5) Å. crystal characterization with methods further supplemented DFT materials...
The support vector machine model produced the best results with a root mean square error of 1.54 × 10 −6 K −1 . was applied to 3 593 726 possible AA′BB′O compositions, resulting in 150 451 predictions confidence region.
Materials informatics uses data-driven approaches for the study and discovery of materials. Features or descriptors are crucial components in generating reliable accurate machine-learning models. While general data can be acquired through public commercial sources, features must tailored a specific application. Common featurizers suitable generic chemical problems, but may not ideal solid state Here, we have assembled Oliynyk property list feature generation which works well on limited...
Traditional and non-classical machine learning models for solid-state structure prediction have predominantly relied on compositional features (derived from properties of constituent elements) to predict the existence its properties. However, lack structural information can be a source suboptimal property mapping increased predictive uncertainty. To address challenge, we introduce strategy that generates combines both with minimal programming expertise required. Our approach utilizes...
Traditional and non-classical machine learning models for solid-state structure prediction have predominantly relied on compositional features (derived from properties of constituent elements) to predict the existence its...
New ternary rare-earth indides RE23Co6.7In20.3 (RE = Gd–Tm, Lu) have been synthesized by arc-melting the elements under argon and subsequent annealing at 870 K for 1200 h. Single-crystal X-ray diffraction revealed Er23Co6.7In20.3 to crystallize in a new structure type oP100, space group Pbam Wyckoff sequence h11g13da with 23.203(5), b 28.399(5), c 3.5306(6) Å. The crystal structures of Tb, Ho, Er Tm) were determined from single powder data further investigated DFT methods. compounds belong...
In this study, we investigated the growth properties of carbon nanowalls (CNW) depending on substrate types. We deposited metal films Si substrates via RF magnetron sputtering with use four-inch W, Cu and Ni targets. A microwave plasma enhanced chemical vapor deposition (PECVD) system was used to grow CNWs metal-coated using H2 CH4 gases. The vertical superficial conditions grown types were characterized by field emission scanning electron microscopy (FE-SEM). Raman analysis investigate...
Transparent mesoporous silica plates doped with rare-earth metal oxide were prepared using solvent-evaporation method based on the self-organization between structure-directing agent and silicate in a non-aqueous solvent. A triblock copolymer, Pluronic (F127 or P123), was used as agent, while tetraethyl orthosilicate (TEOS) source. The pore diameter surface area of plate optimized conditions ca 40 600 m2 g(-1), respectively, for both agent. Rare-earth oxides (Eu, Tb, Tm oxide) mesochannel...
In this study, the coating of synthesized carbon nanowalls (CNWs) with various metal layers (Ni, Cu, and W) was investigated. CNWs were by microwave plasma enhanced chemical vapor deposition (PECVD) a methane (CH4) hydrogen (H2) gas mixture on p-type Si wafer, then coated films using an RF magnetron sputtering system four-inch targets. Different times (5, 10, 20, 30 min) established to obtain different thicknesses which coated. Field emission scanning electron microscopy (FE-SEM) used...