Prediction of Solar Cell Materials via Unsupervised Literature Learning

Despite the significant advancement of the data-driven studies for physical science, the textual data that are numerous in the literature are not fully embraced by the physics and materials community. In this manuscript, we successfully employ the natural language processing (NLP) technique to unsupervisedly predict the existence of solar cell types including the dye-sensitized solar cells and the perovskite solar cells based on literatures published prior to their first discovery without human annotation. Enlightened by this, we further identify possible solar cell material candidates via NLP starting with a comprehensive training database of 3.2 million paper abstracts published before 2021. The NLP model effectively predicts the existing solar cell materials, while an uncommon solar cell material namely PtSe2 is suggested as an appropriate candidate for the future solar cells. Its optoelectronic properties are comprehensive investigated via first-principles calculations to reveal the decent stability and optoelectronic performance of the NLP-predicted candidate. This study demonstrates the viability of the textual data for the data-driven materials prediction and highlights the NLP method as a powerful tool to reliably predict the solar cell materials.

An Assessment Framework for Solar Cell Material Based on a Modified Fuzzy DEMATEL Approach

We propose an evaluation system to choose appropriate materials for solar cells. A fuzzy DEMATEL information procedure was used for decision-making to gather information and analyze the casual relationship. These data acquired were partitioned into causal and impact bunches, empowering users to gather an improved understanding of the intelligent relationship among them, as well as making recommendations for changes to upgrade their general execution. The proposed approach can deliver a compelling fabric choice assessment with satisfactory criteria that fit the respondent’s discernment designs; particularly, these evaluation dimensions are interlaced. Recommendations are given to assist government authorities to plan a Taiwan solar cell industry approach and for industries to develop commerce techniques for improvement in the solar cell field.

Piezophototronic Solar cell based on Third Generation Semiconductor Materials

By applying the outward uniform strain on the non-centrosymmetric piezoelectric semiconductor, the polarization charges on the material surface are induced. Polarization charges are often generated within the crystals provided that the applied strain is non-uniform. The strain applied has an effect on electronic transport and can be utilized to modulate the properties of the material. The effect of multiway coupling between piezoelectricity, semiconductor transport properties, and photoexcitation results in piezo-phototronic effects. Recent studies have shown the piezoelectric and semiconductor properties of third-generation semiconductors have been used in photodetectors, LEDs, and nanogenerators. The third-generation piezoelectric semiconductor can be used in high-performance photovoltaic cells. A third-generation piezo-phototronic solar cell material is theoretically explored in this manuscript on the basis of a GaN metal-semiconductor interaction. This study aims to determine the effects of piezoelectric polarization on the electrical performance characteristics of this solar cell material. Performance parameters such as Power Conversion Efficiency, Fill Factors, I-V Characteristics, Open Circuit Voltage, and Maximum Output Power have been evaluated. The piezophototronic effect can enhance the open-circuit current voltage by 5.5 percent with an externally applied strain by 0.9 percent. The study will open a new window for the next generation of high-performance piezo-phototronic effects.