Liner Programming And Adaptive Machine Learning Approach For Accelerated Halide Perovskite Photovoltaic Materials Screening

Stability is the primary concern in the design of new materials.The stability of materials can be divided into thermodynamic stability and dynamic stability,and thermodynamic stability is commonly used to measure stability.Decomposition energy can be used to quantitatively describe the thermodynamic stability of materials.The traditional method is to calculate the decomposition energy corresponding to one or some specific decomposition pathways.However,this method can not accurately judge the stability of materials and predict some "new" materials that do not exist in experiments.Thus,this work took halide perovskites as an example and used linear programming to calculate the optimal decomposition pathway and the lowest decomposition energy(?Hd)of the material to predict its thermodynamic stability.In the calculation process,determining the competitive secondary phases is the most critical step.To make the calculation results consistent with the experimental results,the competitive secondary phases is derived from the ICSD and PDF cards.Based on this study,a set of methods based on linear programming,material database and first-principles calculations has been developed,which can calculate the decomposition energy and predict the thermodynamic stability of materials,easily and quickly.The figure of merit is of crucial importance in materials design to search for candidates with optimal functionality.In the field of photovoltaics,the bandgap(Eg)is a well-recognized figure of merit for screening solar cell absorbers subject to the Shockley-Queisser limit.In this study,the bandgap as the figure of merit is challenged since an ideal solar cell absorber requires multiple criteria such as stability,optical absorption,and carrier lifetime.Multiple criteria make the quantitative description of material candidates difficult and computationally time-consuming.Based on the previous work,we combine thermodynamic stability(?Hd)and Eg into a unified figure of merit and use Bayesian optimization(BO)to accelerate materials screening.We have found that,in comparison to an exhaustive search via multiple parameters,BO based on the unified figure of merit can screen optimal candidates(Eg-PBE between 0.6-1.2 eV,?Hd>-29 meV/atom)more efficiently.Therefore,the proposed method opens a viable route for the search of optimal solar cell absorbers from a large amount of material candidates with less computational cost.