Crystallographic stability is an important factor that affects the stability of perovskites.The stability dictates the commercial applications of lead-based organometal halide perovskites.The tolerance factor(t)and oc...Crystallographic stability is an important factor that affects the stability of perovskites.The stability dictates the commercial applications of lead-based organometal halide perovskites.The tolerance factor(t)and octahedral factor(μ)form the state-of-the-art criteria used to evaluate the perovskite crystallographic stability.We studied the crystallographic stabilities of halide and chalcogenide perovskites by exploring an effective alternative descriptor,the global instability index(GII)that was used as an indicator of the stability of perovskite oxides.We particularly focused on determining crystallographic reliability by calculating GII.We analyzed the bond valence models of the 243 halide and chalcogenide perovskites that occupied the lowest-energy cubic-phase structures determined by conducting the first-principles-based total energy minimization calculations.The decomposition energy(ΔHD)reflects the thermodynamic stability of the system and is considered as the benchmark that helps assess the effectiveness of GII in evaluating the crystallographic stability of the systems under study.The results indicated that the accuracy of predicting thermodynamic stability was significantly higher when GII(73.6%)was analyzed compared to the cases when t(55%)andμ(39.1%)were analyzed to determine the stability.The results obtained from the machine learning-based data mining method further indicate that GII is an important descriptor of the stability of the perovskite family.展开更多
With the rapid development of artificial intelligence and machine learning(ML)methods,materials science is rapidly entering the era of data-driven materials informatics.ML models serve as the most crucial component,cl...With the rapid development of artificial intelligence and machine learning(ML)methods,materials science is rapidly entering the era of data-driven materials informatics.ML models serve as the most crucial component,closely bridging material structure and material properties.There is a considerable difference in the prediction performance of different ML methods for material systems.Herein,we evaluated three categories(linear,kernel,and nonlinear methods)of models,with twelve ML algorithms commonly used in the materials field.In addition,halide perovskite was chosen as an example to evaluate the fitting performance of different models.We constructed a total dataset of 540 halide perovskites and 72 features,with formation energy and bandgap as target properties.We found that different categories of ML models show similar trends for different target properties.Among them,the difference between the models is enormous for the formation energy,with the coefficient of determination(R2)range 0.69-0.953.The fitting performance between the models is closer for bandgap,with the R^(2)range 0.941-0.997.The nonlinear-ensemble model shows the best fitting performance for both the formation energy and the bandgap.It shows that the nonlinear-ensemble model,constructed by combining multiple weak learners,effectively describes the nonlinear relationship between material features and target property.In addition,the extreme gradient boosting decision tree model shows the most superior results among all the models and searches for two new descriptors that are crucial for formation energy and bandgap.Our work provides useful guidance for the selection of effective machine learning methods in the data-mining studies of specific material systems.展开更多
基金supported by the National Natural Science Foundation of China(62004080 and 92061113)the Postdoctoral Innovative Talents Supporting Program(BX20190143)the China Postdoctoral Science Foundation(2020M670834)。
文摘Crystallographic stability is an important factor that affects the stability of perovskites.The stability dictates the commercial applications of lead-based organometal halide perovskites.The tolerance factor(t)and octahedral factor(μ)form the state-of-the-art criteria used to evaluate the perovskite crystallographic stability.We studied the crystallographic stabilities of halide and chalcogenide perovskites by exploring an effective alternative descriptor,the global instability index(GII)that was used as an indicator of the stability of perovskite oxides.We particularly focused on determining crystallographic reliability by calculating GII.We analyzed the bond valence models of the 243 halide and chalcogenide perovskites that occupied the lowest-energy cubic-phase structures determined by conducting the first-principles-based total energy minimization calculations.The decomposition energy(ΔHD)reflects the thermodynamic stability of the system and is considered as the benchmark that helps assess the effectiveness of GII in evaluating the crystallographic stability of the systems under study.The results indicated that the accuracy of predicting thermodynamic stability was significantly higher when GII(73.6%)was analyzed compared to the cases when t(55%)andμ(39.1%)were analyzed to determine the stability.The results obtained from the machine learning-based data mining method further indicate that GII is an important descriptor of the stability of the perovskite family.
基金supported by the National Natural Science Foundation of China(Grants Nos.62125402 and 92061113)。
文摘With the rapid development of artificial intelligence and machine learning(ML)methods,materials science is rapidly entering the era of data-driven materials informatics.ML models serve as the most crucial component,closely bridging material structure and material properties.There is a considerable difference in the prediction performance of different ML methods for material systems.Herein,we evaluated three categories(linear,kernel,and nonlinear methods)of models,with twelve ML algorithms commonly used in the materials field.In addition,halide perovskite was chosen as an example to evaluate the fitting performance of different models.We constructed a total dataset of 540 halide perovskites and 72 features,with formation energy and bandgap as target properties.We found that different categories of ML models show similar trends for different target properties.Among them,the difference between the models is enormous for the formation energy,with the coefficient of determination(R2)range 0.69-0.953.The fitting performance between the models is closer for bandgap,with the R^(2)range 0.941-0.997.The nonlinear-ensemble model shows the best fitting performance for both the formation energy and the bandgap.It shows that the nonlinear-ensemble model,constructed by combining multiple weak learners,effectively describes the nonlinear relationship between material features and target property.In addition,the extreme gradient boosting decision tree model shows the most superior results among all the models and searches for two new descriptors that are crucial for formation energy and bandgap.Our work provides useful guidance for the selection of effective machine learning methods in the data-mining studies of specific material systems.
基金supported by the National Natural Science Foundation of China(61722403,92061113,and 12004131)the Interdisciplinary Research Grant for Ph Ds of Jilin University(101832020DJX043)。