Using a Neural Network to Improve the Optical Absorption in Halide Perovskite Layers Containing Core-Shells Silver Nanoparticles

Machine learning for nanoplasmonics

Plasmonic nanomaterials have outstanding optoelectronic properties potentially enabling the next generation of catalysts, sensors, lasers and photothermal devices. Owing to optical and electron techniques, modern nanoplasmonics research generates large datasets characterizing features across length scales. Furthermore, optimizing syntheses leading to specific nanostructures requires time-consuming multiparametric approaches. These complex datasets and trial-and-error practices make nanoplasmonics research ripe for the application of machine learning (ML) and advanced data processing methods. ML algorithms capture relationships between synthesis, structure and performance in a way that far exceeds conventional simulation and theory approaches, enabling effective performance optimization. For example, neural networks can tailor the nanostructure morphology to target desired properties, identify synthetic conditions and extract quantitative information from complex data. Here we discuss the nascent field of ML for nanoplasmonics, describe the opportunities and limitations of ML in nanoplasmonic research, and conclude that ML is potentially transformative, especially if the community curates and shares its big data.

Optical Characterization and Prediction with Neural Network Modeling of Various Stoichiometries of Perovskite Materials Using a Hyperregression Method

Quaternary perovskite solar cells are being extensively studied, with the goal of increasing solar cell efficiency and securing stability by changing the ratios of methylammonium, formamidinium, I₃, and Br₃. However, when the stoichiometric ratio is changed, the photoelectric properties reflect those of different materials, making it difficult to study the physical properties of the quaternary perovskite. In this study, the optical properties of perovskite materials with various stoichiometric ratios were measured using ellipsometry, and the results were analyzed using an optical simulation model. Because it is difficult to analyze the spectral pattern according to composition using the existing method of statistical regression analysis, an artificial neural network (ANN) structure was constructed to enable the hyperregression analysis of n-dimensional variables. Finally, by inputting the stoichiometric ratios used in the fabrication and the wavelength range to the trained artificial intelligence model, it was confirmed that the optical properties were similar to those measured with an ellipsometer. The refractive index and extinction coefficient extracted through the ellipsometry analysis show a tendency consistent with the color change of the specimen, and have a similar shape to that reported in the literature. When the optical properties of the unmodified perovskite are predicted using the verified artificial intelligence model, a very complex change in pattern is observed, which is impossible to analyze with a general regression method. It can be seen that this change in optical properties is well maintained, even during rapid variations in the pattern according to the change in composition. In conclusion, hyperregression analysis with n-dimensional variables can be performed for the spectral patterns of thin-film materials using a simple big data construction method.

Instantaneous Property Prediction and Inverse Design of Plasmonic Nanostructures Using Machine Learning: Current Applications and Future Directions

Advances in plasmonic materials and devices have given rise to a variety of applications in photocatalysis, microscopy, nanophotonics, and metastructures. With the advent of computing power and artificial neural networks, the characterization and design process of plasmonic nanostructures can be significantly accelerated using machine learning as opposed to conventional FDTD simulations. The machine learning (ML) based methods can not only perform with high accuracy and return optical spectra and optimal design parameters, but also maintain a stable high computing efficiency without being affected by the structural complexity. This work reviews the prominent ML methods involved in forward simulation and inverse design of plasmonic nanomaterials, such as Convolutional Neural Networks, Generative Adversarial Networks, Genetic Algorithms and Encoder-Decoder Networks. Moreover, we acknowledge the current limitations of ML methods in the context of plasmonics and provide perspectives on future research directions.

Cation Substitution Effects on the Structural, Electronic and Sun-Light Absorption Features of All-Inorganic Halide Perovskites

All-inorganic perovskites (such as CsPbI3) are emerging as new candidates for photovoltaic applications. Unfortunately, this class of materials present two important weaknesses in their way to commercialization: poor stability and...

Improved optical properties of perovskite solar cells by introducing Ag nanopartices and ITO AR layers

Embedded noble metal nanostructures and surface anti-reflection (AR) layers affect the optical properties of methylammonium lead iodide (CH3NH3PbI3) perovskite solar cells significantly. Herein, by employing a combined finite element method and genetic algorithm approach, we report five different types of CH3NH3PbI3 perovskite solar cells by introducing embedded Ag nanoparticles within the CH3NH3PbI3 layer and/or top ITO cylinder grating as an AR layer. The maximum photocurrent was optimized to reach 23.56 mA/cm2, which was 1.09/1.17 times higher than Tran's report/ flat cases. It is also comparable with values (23.6 mA/cm2) reported in the literature. The calculations of the electric field and charge carrier generation rate of the optimized solar cell further confirms this improvement than flat cases. It attributes to the synergistic effect of the embedded Ag nanoparticles and ITO AR layer. The results obtained herein hold great promise for future boosting the optical efficiency of perovskite solar cells.

Artificial Neural Network-Based Prediction of the Optical Properties of Spherical Core-Shell Plasmonic Metastructures

The substitution of time- and labor-intensive empirical research as well as slow finite difference time domain (FDTD) simulations with revolutionary techniques such as artificial neural network (ANN)-based predictive modeling is the next trend in the field of nanophotonics. In this work, we demonstrated that neural networks with proper architectures can rapidly predict the far-field optical response of core-shell plasmonic metastructures. The results obtained with artificial neural networks are comparable with FDTD simulations in accuracy but the speed of obtaining them is between 100-1000 times faster than FDTD simulations. Further, we have proven that ANNs does not have problems associated with FDTD simulations such as dependency of the speed of convergence on the size of the structure. The other trend in photonics is the inverse design problem, where the far-field optical response of a spherical core-shell metastructure can be linked to the design parameters such as type of the material(s), core radius, and shell thickness using a neural network. The findings of this paper provide evidence that machine learning (ML) techniques such as artificial neural networks can potentially replace time-consuming finite domain methods in the future.

The effects of the chemical composition on the structural, thermodynamic, and mechanical properties of all-inorganic halide perovskites

A systematic ab-initio study of all-inorganic perovskites with formula CsPb_1−bSn_b(I_1−xBr_x)₃ has been performed, elucidating the connection of that composition with their structural, thermodynamics and mechanical properties.