In recent years,material genome has been a hot topic in the field of material science.The emergence of the term“material genome”was largely inspired by the successful Human Genome Project.Trad让ionally,the discovery...In recent years,material genome has been a hot topic in the field of material science.The emergence of the term“material genome”was largely inspired by the successful Human Genome Project.Trad让ionally,the discovery and development of new materials and new processes depend on scientific intuition and a lengthy trial-and-error process.For years,material scientists have been longing to find some sort of basic building blocks whose structure and defects may determine the properties of materials,similar to the genome in the field of biology.展开更多
Discovering new superconductors via traditional trial-and-error experimental approaches is apparently a time-consuming process,and the correlations between the critical temperature(Tc) and material features are still ...Discovering new superconductors via traditional trial-and-error experimental approaches is apparently a time-consuming process,and the correlations between the critical temperature(Tc) and material features are still obscure.The rise of machine learning(ML) technology provides new opportunities to speed up inefficient exploration processes,and could potentially uncover new hints on the unclear correlations.In this work,we utilize open-source materials data,ML models,and data mining methods to explore the correlation between the chemical features and Tcvalues of superconducting materials.To further improve the prediction accuracy,a new model is created by integrating three basic algorithms,showing an enhanced accuracy with the coefficient of determination(R2) score of 95.9 % and root mean square error(RMSE) of 6.3 K.The average marginal contributions of material features towards Tcvalues are estimated to determine the importance of various features during prediction processes.The results suggest that the range thermal conductivity plays a critical role in Tcprediction among all element features.Furthermore,the integrated ML model is utilized to screen out potential twenty superconducting materials with Tcvalues beyond 50.0 K.This study provides insights towards Tcprediction to accelerate the exploration of potential high-Tcsuperconductors.展开更多
Studying the complexity of the electronic phase diagram is at the heart of understanding strongly correlated system in general,with high Tc superconductors as the most known examples.High temperature superconductivity...Studying the complexity of the electronic phase diagram is at the heart of understanding strongly correlated system in general,with high Tc superconductors as the most known examples.High temperature superconductivity has a wide range of application potentials in power transmission,nuclear magnetic resonance,magnetic levitation transportation,aerospace,information and communication technologies,etc.Understanding its mechanism remains a long-standing challenge,due to its complex material structures and interlays among different phases such as charge density wave,antiferromagnetic and superconducting phases.As a result,this has greatly hindered its further development.展开更多
High-throughput powder X-ray diffraction(XRD)with white X-ray beam and an energy-dispersive detector array is demonstrated in this work on a CeO;powder sample on a bending magnet synchrotron beamline at the Shanghai S...High-throughput powder X-ray diffraction(XRD)with white X-ray beam and an energy-dispersive detector array is demonstrated in this work on a CeO;powder sample on a bending magnet synchrotron beamline at the Shanghai Synchrotron Radiation Facility(SSRF),using a simulated energy-dispersive array detector consisting of a spatially scanning silicon-drift detector(SDD).Careful analysis and corrections are applied to account for various experimental hardware-related and diffraction angle-related factors.The resulting diffraction patterns show that the relative strength between different diffraction peaks from energy-dispersive XRD(EDXRD)spectra is consistent with that from angle-resolved XRD(ARXRD),which is necessary for analyzing crystal structures for unknown samples.The X-ray fluorescence(XRF)signal is collected simultaneously.XRF counts from all pixels are integrated directly by energy,while the diffraction spectra are integrated by d-spacing,resulting in a much improved peak strength and signal-to-noise(S/N)ratio for the array detector.In comparison with ARXRD,the diffraction signal generated by a white X-ray beam over monochromic light under the experimental conditions is about 104 times higher.The full width at half maximum(FWHM)of the peaks in q-space is found to be dependent on the energy resolution of the detector,the angle span of the detector,and the diffraction angle.It is possible for EDXRD to achieve the same or even smaller FWHM as ARXRD under the energy resolution of the current detector if the experimental parameters are properly chosen.展开更多
基金This work is supported in part by the National Key Research and Development Program of China(2018YFB0703600).
文摘In recent years,material genome has been a hot topic in the field of material science.The emergence of the term“material genome”was largely inspired by the successful Human Genome Project.Trad让ionally,the discovery and development of new materials and new processes depend on scientific intuition and a lengthy trial-and-error process.For years,material scientists have been longing to find some sort of basic building blocks whose structure and defects may determine the properties of materials,similar to the genome in the field of biology.
基金financial supports from the Fund of Science and Technology on Reactor Fuel and Materials Laboratory(JCKYS2019201074)the Affiliated Hospital of Putian University,the Shenzhen Fundamental Research Program(JCYJ20220531095404009)+1 种基金the Shenzhen Knowledge Innovation Plan-Fundamental Research(Discipline Distribution)(JCYJ20180507184623297)the Major Science and Technology Infrastructure Project of Material Genome Big-science Facilities Platform supported by Municipal Development and Reform Commission of Shenzhen。
文摘Discovering new superconductors via traditional trial-and-error experimental approaches is apparently a time-consuming process,and the correlations between the critical temperature(Tc) and material features are still obscure.The rise of machine learning(ML) technology provides new opportunities to speed up inefficient exploration processes,and could potentially uncover new hints on the unclear correlations.In this work,we utilize open-source materials data,ML models,and data mining methods to explore the correlation between the chemical features and Tcvalues of superconducting materials.To further improve the prediction accuracy,a new model is created by integrating three basic algorithms,showing an enhanced accuracy with the coefficient of determination(R2) score of 95.9 % and root mean square error(RMSE) of 6.3 K.The average marginal contributions of material features towards Tcvalues are estimated to determine the importance of various features during prediction processes.The results suggest that the range thermal conductivity plays a critical role in Tcprediction among all element features.Furthermore,the integrated ML model is utilized to screen out potential twenty superconducting materials with Tcvalues beyond 50.0 K.This study provides insights towards Tcprediction to accelerate the exploration of potential high-Tcsuperconductors.
文摘Studying the complexity of the electronic phase diagram is at the heart of understanding strongly correlated system in general,with high Tc superconductors as the most known examples.High temperature superconductivity has a wide range of application potentials in power transmission,nuclear magnetic resonance,magnetic levitation transportation,aerospace,information and communication technologies,etc.Understanding its mechanism remains a long-standing challenge,due to its complex material structures and interlays among different phases such as charge density wave,antiferromagnetic and superconducting phases.As a result,this has greatly hindered its further development.
基金supported by the National Key Research and Development Program of China,China(2017YFB0701900)High-Level Special Funds(G02256401 and G02256301)+1 种基金supported by the fund of the Guangdong Provincial Key Laboratory(2018B030322001)the Guangdong-Hong Kong-Macao Joint Laboratory(2019B121205001)。
文摘High-throughput powder X-ray diffraction(XRD)with white X-ray beam and an energy-dispersive detector array is demonstrated in this work on a CeO;powder sample on a bending magnet synchrotron beamline at the Shanghai Synchrotron Radiation Facility(SSRF),using a simulated energy-dispersive array detector consisting of a spatially scanning silicon-drift detector(SDD).Careful analysis and corrections are applied to account for various experimental hardware-related and diffraction angle-related factors.The resulting diffraction patterns show that the relative strength between different diffraction peaks from energy-dispersive XRD(EDXRD)spectra is consistent with that from angle-resolved XRD(ARXRD),which is necessary for analyzing crystal structures for unknown samples.The X-ray fluorescence(XRF)signal is collected simultaneously.XRF counts from all pixels are integrated directly by energy,while the diffraction spectra are integrated by d-spacing,resulting in a much improved peak strength and signal-to-noise(S/N)ratio for the array detector.In comparison with ARXRD,the diffraction signal generated by a white X-ray beam over monochromic light under the experimental conditions is about 104 times higher.The full width at half maximum(FWHM)of the peaks in q-space is found to be dependent on the energy resolution of the detector,the angle span of the detector,and the diffraction angle.It is possible for EDXRD to achieve the same or even smaller FWHM as ARXRD under the energy resolution of the current detector if the experimental parameters are properly chosen.
