Employing a comprehensive structure search and high-throughput first-principles calculation method on 1561 compounds,the present study reveals the phase diagram of Lu-H-N.In detail,the formation energy landscape of Lu...Employing a comprehensive structure search and high-throughput first-principles calculation method on 1561 compounds,the present study reveals the phase diagram of Lu-H-N.In detail,the formation energy landscape of Lu-H-N is derived and utilized to assess the thermodynamic stability of each compound that is created via element substitution.The result indicates that there is no stable ternary structure in the Lu-H-N chemical system,however,metastable ternary structures,such as Lu_(20)H_(2)N_(17)(C2/m)and Lu_(2)H_(2)N(P3m1),are observed to have small E_(hull)(<100 meV/atom).It is also found that the energy convex hull of the Lu-H-N system shifts its shape when applying hydrostatic pressure up to 10 GPa,and the external pressure stabilizes a couple of binary phases such as LuN_9 and Lu_(10)H_(21).Additionally,interstitial voids in LuH_(2)are observed,which may explain the formation of Lu_(10)H_(21)and LuH_(3-δ)N_ε.To provide a basis for comparison,x-ray diffraction patterns and electronic structures of some compounds are also presented.展开更多
The Cs V_(3)Sb_(5) kagome lattice holds the promise for manifesting electron correlation,topology and superconductivity.However,by far only three Cs V_(3)Sb_(5)-like kagome materials have been experimentally spotted.W...The Cs V_(3)Sb_(5) kagome lattice holds the promise for manifesting electron correlation,topology and superconductivity.However,by far only three Cs V_(3)Sb_(5)-like kagome materials have been experimentally spotted.We enlarge this family of materials to 1386 compounds via element species substitution,and the further screening process suggests that 28 promising candidates have superior thermodynamic stability,hence they are highly likely to be synthesizable.Moreover,these compounds possess several unique electronic structures,and can be categorized into five non-magnetic and three magnetic groups accordingly.It is our hope that this work can greatly expand the viable phase space of the Cs V_(3)Sb_(5)-like materials for investigating or tuning the novel quantum phenomena in kagome lattice.展开更多
The recent report of near-ambient superconductivity in the nitrogen-doped lutetium hydride has attracted considerable attention.Subsequent follow-up studies confirmed the pressure-induced color changes in both N-free ...The recent report of near-ambient superconductivity in the nitrogen-doped lutetium hydride has attracted considerable attention.Subsequent follow-up studies confirmed the pressure-induced color changes in both N-free and N-doped LuH_(2) but failed to reproduce superconductivity. It remains a puzzle why the samples in the original report exhibited pronounced resistance anomaly reminiscent of the superconducting transition. Here, we show that percolation of metallic grains with high conductivity through the insulating surfaces in cold-pressed LuH_(2) samples can occasionally produce sharp resistance drops, which even display magnetic field and/or current dependences but stay far from zero resistance. The insulating surface of LuH2grain should be attributed to the modification of hydrogen stoichiometry or the contamination by oxygen/nitrogen, resulting in an increase of resistance by over six orders of magnitude. Such an effect is more significant than that discovered recently in LaH_(3±x), which may indicate that LuH_(2) can be a potential superionic conductor. Our results call for caution in asserting the resistivity drops as superconductivity and invalidate the background subtraction in analyzing the corresponding resistance data.展开更多
Transition metal dichalcogenides(TMDs)are a class of materials with various useful properties,and it is worthwhile to have a thorough evaluation of the characteristics of the TMDs,most importantly,their structural sta...Transition metal dichalcogenides(TMDs)are a class of materials with various useful properties,and it is worthwhile to have a thorough evaluation of the characteristics of the TMDs,most importantly,their structural stability and exfoliability,in a systematic fashion.Here,by employing high-throughput first-principles calculations,we investigate the vast phase space of TMDs,including 16 bulk phases and 6 monolayer phases for all possible TMD combinations[comprising(3d,4d,5d)transition-metal cations and(S,Se,Te)anions],totaling 1386 compounds.Through the‘bird-view’of the as-large-as-possible configurational and chemical space of TMDs,our work presents comprehensive energy landscapes to elucidate the thermodynamic stability as well as the exfoliability of TMDs,which are of vital importance for future synthesis and exploration towards large-scale industrial applications.展开更多
随着数据科学和材料科学的进步,人们如今可构建出较为准确的人工智能模型,用于材料性质预测.本文中,我们以170,714个无机晶体化合物的高通量第一性原理计算数据集为基础,训练得到了可精确预测无机化合物形成能的机器学习模型.相比于同...随着数据科学和材料科学的进步,人们如今可构建出较为准确的人工智能模型,用于材料性质预测.本文中,我们以170,714个无机晶体化合物的高通量第一性原理计算数据集为基础,训练得到了可精确预测无机化合物形成能的机器学习模型.相比于同类工作,本项研究以超大数据集为出发点,构建出无机晶体形成能的高精度泛化模型,可外推至广阔相空间,其中的Dense Net神经网络模型精度可以达到R^(2)=0.982和平均绝对误差(MAE)=0.072 eV atom^(-1).上述模型精度的提升源自一系列新型特征描述符,这些描述符可有效提取出原子与领域原子间的电负性和局域结构等信息,从而精确捕捉到原子间的相互作用.本文为新材料搜索提供了一种高效、低成本的结合能预测手段.展开更多
This study introduces a novel artificial intelligence(AI)force field,namely a graph-based pre-trained transformer force field(GPTFF),which can simulate arbitrary inorganic systems with good precision and generalizabil...This study introduces a novel artificial intelligence(AI)force field,namely a graph-based pre-trained transformer force field(GPTFF),which can simulate arbitrary inorganic systems with good precision and generalizability.Harnessing a large trove of the data and the attention mechanism of transformer algorithms,the model can accurately predict energy,atomic force,and stress with mean absolute error(MAE)values of 32 me V/atom,71 me V/A,and 0.365 GPa,respectively.The dataset used to train the model includes 37.8 million single-point energies,11.7 billion force pairs,and 340.2 million stresses.We also demonstrated that the GPTFF can be universally used to simulate various physical systems,such as crystal structure optimization,phase transition simulations,and mass transport.The model is publicly released with this paper,enabling anyone to use it immediately without needing to train it.展开更多
基金Chinese Academy of Sciences(Grant Nos.CAS-WX2023SF0101 and XDB33020000)the National Key R&D Program of China(Grant Nos.2021YFA1400200 and 2021YFA0718700)。
文摘Employing a comprehensive structure search and high-throughput first-principles calculation method on 1561 compounds,the present study reveals the phase diagram of Lu-H-N.In detail,the formation energy landscape of Lu-H-N is derived and utilized to assess the thermodynamic stability of each compound that is created via element substitution.The result indicates that there is no stable ternary structure in the Lu-H-N chemical system,however,metastable ternary structures,such as Lu_(20)H_(2)N_(17)(C2/m)and Lu_(2)H_(2)N(P3m1),are observed to have small E_(hull)(<100 meV/atom).It is also found that the energy convex hull of the Lu-H-N system shifts its shape when applying hydrostatic pressure up to 10 GPa,and the external pressure stabilizes a couple of binary phases such as LuN_9 and Lu_(10)H_(21).Additionally,interstitial voids in LuH_(2)are observed,which may explain the formation of Lu_(10)H_(21)and LuH_(3-δ)N_ε.To provide a basis for comparison,x-ray diffraction patterns and electronic structures of some compounds are also presented.
基金the financial support from the Chinese Academy of Sciences(Grant Nos.ZDBS-LY-SLH007,XDB33020000,and CAS-WX2021PY-0102)the National Natural Science Foundation of China(Grant No.12174428)。
文摘The Cs V_(3)Sb_(5) kagome lattice holds the promise for manifesting electron correlation,topology and superconductivity.However,by far only three Cs V_(3)Sb_(5)-like kagome materials have been experimentally spotted.We enlarge this family of materials to 1386 compounds via element species substitution,and the further screening process suggests that 28 promising candidates have superior thermodynamic stability,hence they are highly likely to be synthesizable.Moreover,these compounds possess several unique electronic structures,and can be categorized into five non-magnetic and three magnetic groups accordingly.It is our hope that this work can greatly expand the viable phase space of the Cs V_(3)Sb_(5)-like materials for investigating or tuning the novel quantum phenomena in kagome lattice.
基金supported by the National Natural Science Foundation of China (Grant Nos. 12025408, 11921004, 11834016, and 11888101)the Beijing Natural Science Foundation (Grant No. Z190008)+1 种基金the National Key R&D Program of China (Grant Nos. 2021YFA1400200, and 2021YFA1400300)the Strategic Priority Research Program of CAS (Grant No. XDB33000000)。
文摘The recent report of near-ambient superconductivity in the nitrogen-doped lutetium hydride has attracted considerable attention.Subsequent follow-up studies confirmed the pressure-induced color changes in both N-free and N-doped LuH_(2) but failed to reproduce superconductivity. It remains a puzzle why the samples in the original report exhibited pronounced resistance anomaly reminiscent of the superconducting transition. Here, we show that percolation of metallic grains with high conductivity through the insulating surfaces in cold-pressed LuH_(2) samples can occasionally produce sharp resistance drops, which even display magnetic field and/or current dependences but stay far from zero resistance. The insulating surface of LuH2grain should be attributed to the modification of hydrogen stoichiometry or the contamination by oxygen/nitrogen, resulting in an increase of resistance by over six orders of magnitude. Such an effect is more significant than that discovered recently in LaH_(3±x), which may indicate that LuH_(2) can be a potential superionic conductor. Our results call for caution in asserting the resistivity drops as superconductivity and invalidate the background subtraction in analyzing the corresponding resistance data.
基金supported by National Key R&D Program of China(2021YFA0718700)The computational resource is provided by the Platform for Data-Driven Computational Materials Discovery of the Songshan Lake laboratorythe financial support from Chinese Academy of Sciences(Grant Nos.ZDBS-LYSLH007,XDB33020000,and CAS-WX2021PY-0102).
文摘Transition metal dichalcogenides(TMDs)are a class of materials with various useful properties,and it is worthwhile to have a thorough evaluation of the characteristics of the TMDs,most importantly,their structural stability and exfoliability,in a systematic fashion.Here,by employing high-throughput first-principles calculations,we investigate the vast phase space of TMDs,including 16 bulk phases and 6 monolayer phases for all possible TMD combinations[comprising(3d,4d,5d)transition-metal cations and(S,Se,Te)anions],totaling 1386 compounds.Through the‘bird-view’of the as-large-as-possible configurational and chemical space of TMDs,our work presents comprehensive energy landscapes to elucidate the thermodynamic stability as well as the exfoliability of TMDs,which are of vital importance for future synthesis and exploration towards large-scale industrial applications.
基金the financial support from the Chinese Academy of Sciences(CAS-WX2021PY-0102,ZDBS-LY-SLH007,and XDB33020000)。
文摘随着数据科学和材料科学的进步,人们如今可构建出较为准确的人工智能模型,用于材料性质预测.本文中,我们以170,714个无机晶体化合物的高通量第一性原理计算数据集为基础,训练得到了可精确预测无机化合物形成能的机器学习模型.相比于同类工作,本项研究以超大数据集为出发点,构建出无机晶体形成能的高精度泛化模型,可外推至广阔相空间,其中的Dense Net神经网络模型精度可以达到R^(2)=0.982和平均绝对误差(MAE)=0.072 eV atom^(-1).上述模型精度的提升源自一系列新型特征描述符,这些描述符可有效提取出原子与领域原子间的电负性和局域结构等信息,从而精确捕捉到原子间的相互作用.本文为新材料搜索提供了一种高效、低成本的结合能预测手段.
基金supported by the National Natural Science Foundation of China(12025407 and 11934003)Chinese Academy of Sciences(CAS-WX2023SF-0101,XDB33020000,XDB33030100)National Key R&D Program of China(2021YFA0718700,2021YFA1400200)。
文摘This study introduces a novel artificial intelligence(AI)force field,namely a graph-based pre-trained transformer force field(GPTFF),which can simulate arbitrary inorganic systems with good precision and generalizability.Harnessing a large trove of the data and the attention mechanism of transformer algorithms,the model can accurately predict energy,atomic force,and stress with mean absolute error(MAE)values of 32 me V/atom,71 me V/A,and 0.365 GPa,respectively.The dataset used to train the model includes 37.8 million single-point energies,11.7 billion force pairs,and 340.2 million stresses.We also demonstrated that the GPTFF can be universally used to simulate various physical systems,such as crystal structure optimization,phase transition simulations,and mass transport.The model is publicly released with this paper,enabling anyone to use it immediately without needing to train it.
