Effect of alloying on the carrier dynamics in high-performance perovskite solar cells

High-efficiency solar cells often require light absorbers prepared from alloys, such as Cd Te_(1-x)Se_(x),CuIn_(x)Ga_(1-x)Se_(2), Cu_(2)ZnSnS_(4-x)Se_(x), and(Cs_(x)FA_(1-x))Pb(I_(1-y)Br_(y))_(3). However, how alloying affects solar cell performance is poorly understood, and determining common features associated with alloying is of significant interest. Herein, we studied the correlation between the A/X site compositional ratio and the photogenerated carrier dynamics using mixed halide perovskites(Cs_(x)FA_(1-x))Pb(I_(1-y)Br_(y))_(3)as examples.Nonadiabatic molecular dynamics calculations demonstrated that charge carrier recombination is highly sensitive to the compositional ratio at the A/X-site. The enhanced lifetime is attributable to the suppression of atomic fluctuations, the weakening of electron-phonon coupling, and a reduction in the electrontransition probability between band edges. The optimal Br concentration was determined to be ~18%, in agreement with experimental observations. This study not only advances our understanding of why mixed perovskites usually exhibit superior experimental photoelectric properties, but also provides a route for optimizing the carrier lifetimes and efficiencies of perovskite solar cells.

Machine learning in materials design:Algorithm and application

Traditional materials discovery is in ‘trial-and-error’ mode, leading to the issues of low-efficiency, high-cost, and unsustainability in materials design. Meanwhile, numerous experimental and computational trials accumulate enormous quantities of data with multi-dimensionality and complexity, which might bury critical ‘structure–properties’ rules yet unfortunately not well explored. Machine learning(ML), as a burgeoning approach in materials science, may dig out the hidden structure–properties relationship from materials bigdata, therefore, has recently garnered much attention in materials science. In this review, we try to shortly summarize recent research progress in this field, following the ML paradigm:(i) data acquisition →(ii) feature engineering →(iii) algorithm →(iv) ML model →(v) model evaluation →(vi) application. In section of application, we summarize recent work by following the ‘material science tetrahedron’:(i) structure and composition →(ii) property →(iii) synthesis →(iv) characterization, in order to reveal the quantitative structure–property relationship and provide inverse design countermeasures. In addition, the concurrent challenges encompassing data quality and quantity, model interpretability and generalizability, have also been discussed. This review intends to provide a preliminary overview of ML from basic algorithms to applications.

Designing solar-cell absorber materials through computational high-throughput screening

Although the efficiency of CH3 NH3 PI3 has been refreshed to 25.2%,stability and toxicity remain the main challenges for its applications.The search for novel solar-cell absorbers that are highly stable,non-toxic,inexpensive,and highly efficient is now a viable research focus.In this review,we summarize our recent research into the high-throughput screening and materials design of solar-cell absorbers,including single perovskites,double perovskites,and materials beyond Perovskites.BazrS3(single perovskite),Ba2 BiNbS6(double perovskite),HgAl2 Se4(spinel),and IrSb3(skutterudite)were discovered to be potential candidates in terms of their high stabilities,appropriate bandgaps,small carrier effective masses,and strong optical absorption.

High-throughput computational screening of oxide double perovskites for optoelectronic and photocatalysis applications

Oxide double perovskites A2 B’B"O6 are a class of emerging materials in the fields of optoelectronics and catalysis.Due to the chemical flexibilities of perovskite structures,there are multiple elemental combinations of cations A,B’,and B",which leading to tremendous candidates.In this study,we comprehensively screened stable oxide double perovskite A2 B’B"O6 from a pool of 2,018 perovskite candidates using a high-throughput computational approach.By considering a tolerance factor(t)-octahedral factor(μ) phase diagram,138 candidates with Fm 3 m, P21/c,and R3 c phases were selected and systematically studied via first-principles calculations based on density functional theory.The screening procedure finally predicted the existence of 21 stable perovskites,and 14 among them have never been reported.Verification with existing experimental results demonstrates that the prediction accuracy for perovskite formability is approximately 90%.The predicted oxide double perovskites exhibit quasi-direct bandgaps ranging from 0 to 4.4 eV with a significantly small direct-indirect bandgap difference,balanced electron and hole effective masses,and strong optical absorptions.The newly predicted oxide double perovskites may enlarge the pool of material candidates for applications in optoelectronics and photocatalysis.This study provides a route for computational screening of novel perovskites for functional applications.

Defect calculations with quasiparticle correction: A revisited study of iodine defects in CH3NH3PbI3

Defect levels in semiconductor band gaps play a crucial role in functionalized semiconductors for practical applications in optoelectronics;however,first-principle defect calculations based on exchange-correlation functionals,such as local density approximation,grand gradient approximation(GGA),and hybrid functionals,either underestimate band gaps or misplace defect levels.In this study,we revisited iodine defects in CH_(3)NH_(3)PbI_(3) by combining the accuracy of total energy calculations of GGA and single-electron level calculation of the GW method.The combined approach predicted neutral Im_(i) to be unstable and the transition level of Im_(i)(+1/-1)to be close to the valence band maximum.Therefore,Im_(i) may not be as detrimental as previously reported.Moreover,Vm I may be unstable in the-1 charged state but could still be detrimental owing to the deep transition level of Vm I(+1/0).These results could facilitate the further understanding of the intrinsic point defect and defect passivation observed in CH_(3)NH_(3)PbI_(3).

Substrate screening for superclean graphene growth using firstprinciples calculations

Suppressing the formation of amorphous surface carbon and contaminants during the preparation of graphene by chemical vapor deposition remains an ongoing issue.Herein,we analyzed how substrate characteristics affect graphene quality by simulating margin extension,the nucleation process,and defect pegging configurations on mono-crystalline oriented metal substrates with the aim of enhancing graphene cleanliness.Defect formation energy and nucleation potential,which are indirect substrate–graphene interaction features,were found to appropriately evaluate graphene quality.The crystallographic orientation of the metal substrate was discovered to be critical for producing superclean graphene.A low graphene defect density and high nucleation rate on the Cu(100)facet guarantee growth of high-quality graphene,especially in terms of suppressing the formation of amorphous carbon.In addition,rapid kink growth and self-healing on the Cu(100)facet facilitate rapid graphene synthesis,which is also promoted by rapid kink splicing and margin self-repair on this facet.This study provides theoretical insight useful for the synthesis of superclean graphene.

Invisible vapor catalysis in graphene growth by chemical vapor deposition

Vapor catalysis was recently found to play a crucial role in superclean graphene growth via chemical vapor decomposition(CVD).However,knowledge of vapor-phase catalysis is scarce,and several fundamental issues,including vapor compositions and their impact on graphene growth,are ambiguous.Here,by combining density functional theory(DFT)calculations,an ideal gas model,and a designed experiment,we found that the vapor was mainly composed of Cui clusters with tens of atoms.The vapor pressure was estimated to be~10^(-12)-10^(-1)1 bar under normal low-pressure CVD system(LPCVD)conditions for graphene growth,and the exposed surface area of Cui clusters in the vapor was 22-269 times that of the Cu substrate surface,highlighting the importance of vapor catalysis.DFT calculations show Cu clusters,represented by Cu17,have strong capabilities for adsorption,dehydrogenation,and decomposition of hydrocarbons.They exhibit an adsorption lifetime and reaction flux six orders of magnitude higher than those on the Cu surface,thus providing a sufficient supply of active C atoms for rapid graphene growth and improving the surface cleanliness of the synthesized graphene.Further experimental validation showed that increasing the amount of Cu vapor improved the as-synthesized graphene growth rate and surface cleanliness.This study provides a comprehensive understanding of vapor catalysis and the fundamental basis of vapor control for superclean graphene rapid growth.

Harnessing data using symbolic regression methods for discovering novel paradigms in physics

In recent years,machine-learning methods have profoundly impacted research in the interdisciplinary fields of physics.However,most machine-learning models lack interpretability,and physicists doubt the credibility of their conclusions because they cannot be combined with prior physical knowledge.Therefore,this review focuses on symbolic regression,which is an interpretable machine-learning method.First,the relevant concepts of machine learning are introduced in conjunction with induction.Next,we provide an overview of symbolic regression methods.Subsequently,the recent directions for the application of symbolic regression methods in different subfields of physics are outlined,and an overview of the ways in which the applications of symbolic regression have evolved in the realm of physics is provided.The major aim of this review is to introduce the basic principles of symbolic regression and explain its applications in the field of physics.

Bayesian optimization based on a unified figure of merit for accelerated materials screening:A case study of halide perovskites

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(E_g)is a well-recognized figure of merit for screening solar cell absorbers subject to the Shockley-Queisser limit.In this paper,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.Taking halide perovskites as an example,we combine thermodynamic stability(ΔHd)and Eginto 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(E_g,PBEbetween 0.6–1.2 eV,ΔHd>-29 meV per 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.