Predicting Perovskite Bandgap and Solar Cell Performance with Machine Learning

Gok, Elif Ceren; Yildirim, Murat Onur; U., Haris, M. P.; Eren, Esin; Pegu, Meenakshi; Hirandu, Naveen; Huang, Peng; Kazim, Samrana; Öksüz, Ayşegül Uygun; Ahmad, Shahzada

doi:10.1002/solr.202100927

Cited by 21 publications

(20 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[ 25 ] Shahzada. Ahmad et al used the random forest (RF) algorithm to predict the bandgap and PCE of PSCs based on UV–vis absorption and J–V spectra data [ 26 ] The aforementioned articles mostly use ML tools to screen unknown materials or predict their performance. The explanations of ML models and the mining of potential physical laws between features are rarely involved.…”

Section: Introductionmentioning

confidence: 99%

How Machine Learning Predicts and Explains the Performance of Perovskite Solar Cells

et al. 2022

View full text Add to dashboard Cite

Characterizing the electrical parameters of perovskite solar cells (PSCs) usually requires a lot of time to fabricate complete devices. Here, machine learning (ML) is used to reduce the device fabrication process and predict the electrical performance of PSCs. Using ML algorithms and 814 valid data cleaned from 2735 peer‐reviewed publications, ML prediction models are built for bandgap, conduction band minimum, valence band maximum of perovskites, and electrical parameters of PSCs. These prediction models have excellent accuracy, and the root mean square error of the prediction models for bandgap and power conversion efficiency (PCE) reaches 0.064 eV and 1.58%, respectively. Among the many factors that affect the performance of PSCs, those factors play a major role in the lack of comprehensive explanation. Through the prediction model of electrical parameters and Shapley Additive explanations theory, the factors affecting the PCE of PSCs are explained and analyzed. It can not only verify the objective physical laws from the perspective of ML, but also conclude that among the 13 features, the content of formamidinium/NH2CHNH2+ plays the most important role in improving the PCE of PSCs. These results show that ML has great application possibilities in the PSC field.

show abstract

Section: Introductionmentioning

confidence: 99%

How Machine Learning Predicts and Explains the Performance of Perovskite Solar Cells

et al. 2022

View full text Add to dashboard Cite

show abstract

“…The remaining PSCs in the same material group were trained to contribute to a material discovery. [22] For example, to create a realistic scenario, we assumed that there is no information available…”

Section: Resultsmentioning

confidence: 99%

“…The remaining PSCs in the same material group are trained to contribute to a material discovery. [ 24 ] For example, to create a realistic scenario, we assumed that there is no information available for 4PyPTPDAn, yet we have an access to all information from the rest of the material group: wavelength and its respective absorbance values for both 2PyPTPDAn and 3PyPTPDAn; and also open‐circuit potential ( V OC ) and its corresponding short‐circuit current density ( J SC ) values for both 2PyPTPDAn and 3PyPTPDAn. We computed an ML approach for predicting the optical bandgap and PSCs performance of HTMs with various substitution groups.…”

Section: Resultsmentioning

confidence: 99%

Automated Machine Learning Approach in Material Discovery of Hole Selective Layers for Perovskite Solar Cells

et al. 2022

Self Cite

View full text Add to dashboard Cite

In the rapidly emerging field of perovskite solar cells, rational hole selective layer development is considered a double-engine of this progress. To tap into the full potential and accelerate the commercialization path, Machine Learning (ML) is being tasked for perovskite screening and on organic semiconductors core group based on phthalocyanines, pyridine, triazatruxene, thiophenes, carbazole, and phenothiazine to yield efficient solar cells. However, sincere efforts have not led to the design of hole selective layers. Here we demonstrate how ML can be applied to the advancement of hole transport materials (HTMs) to accelerate the development. We evaluated the influence of various HTMs with various groups on the optoelectronic features and photovoltaic performance and validated it using both the random forest model and AutoML framework General Automated Machine Learning Assistant (GAMA). To this end, we utilized the GAMA to predict the suitability of hole selective layers and it returned a 0.0542 ± 0.0470 RMSE score for 15 different materials on average. We (i) established correlations between experimental and predicted results; and (ii) implemented GAMA for HTM suitability prediction. This paves the way for judicious and effective ways to overcome the bottleneck in the development of HTMs. In particular, the prediction approach on optoelectronic features and photovoltaic performance from GAMA is an effective, reliable, and fast methodology and is pioneering in the field of hole transport material screening.

show abstract

“… 36 Only a few models have tried to predict solar cell performance. 37 − 39 All of these ML models focused on the perovskite, and the reduced number of HTMs limited the scope of the models to link the role of the HTM with the performance of the cell. However, in recent years, the number and types of HTMs have shown a significant increase, and we are currently able to gather hundreds of experimental PCE values from the literature.…”

Section: Introductionmentioning

confidence: 99%

“…To date, the ML models applied in the context of PSCs , have been mainly focused on the prediction of perovskite properties, like the band gap, , stability, , ionic conductivities, and other transport properties . Only a few models have tried to predict solar cell performance. − All of these ML models focused on the perovskite, and the reduced number of HTMs limited the scope of the models to link the role of the HTM with the performance of the cell. However, in recent years, the number and types of HTMs have shown a significant increase, and we are currently able to gather hundreds of experimental PCE values from the literature.…”

Section: Introductionmentioning

confidence: 99%

Data-Driven Analysis of Hole-Transporting Materials for Perovskite Solar Cells Performance

et al. 2022

View full text Add to dashboard Cite

We have created a dataset of 269 perovskite solar cells, containing information about their perovskite family, cell architecture, and multiple hole-transporting materials features, including fingerprints, additives, and structural and electronic features. We propose a predictive machine learning model that is trained on these data and can be used to screen possible candidate hole-transporting materials. Our approach allows us to predict the performance of perovskite solar cells with reasonable accuracy and is able to successfully identify most of the top-performing and lowest-performing hole-transporting materials in the dataset. We discuss the effect of data biases on the distribution of perovskite families/architectures on the model’s accuracy and offer an analysis with a subset of the data to accurately study the effect of the hole-transporting material on the solar cell performance. Finally, we discuss some chemical fragments, like arylamine and aryloxy groups, which present a relatively large positive correlation with the efficiency of the cell, whereas other groups, like thiophene groups, display a negative correlation with power conversion efficiency (PCE).

show abstract

Predicting Perovskite Bandgap and Solar Cell Performance with Machine Learning

Cited by 21 publications

References 22 publications

How Machine Learning Predicts and Explains the Performance of Perovskite Solar Cells

How Machine Learning Predicts and Explains the Performance of Perovskite Solar Cells

Automated Machine Learning Approach in Material Discovery of Hole Selective Layers for Perovskite Solar Cells

Data-Driven Analysis of Hole-Transporting Materials for Perovskite Solar Cells Performance

Contact Info

Product

Resources

About