Artificial intelligence-guided and first-principles design and optimization of Rb₂YAgX₆ (X = F, Cl) double perovskite photodetectors

Engineering notes

Engineering notes will be added by the Power for Solar editorial team.

Chinese explanation / 中文解读

中文解读待补充：本站会优先为光伏效率、钙钛矿太阳能电池、储能技术、太阳能热利用、BIPV、并网技术等高价值论文补充中文说明。

Original abstract

Halides perovskites of lead have already shown very good optoelectronic behavior, but because of their toxicity and environmental issues, lead-free materials have been sought faster. The paper investigates the performance of Rb 2 YAgX 6 (X = F, Cl) lead-free double perovskite photodetectors by means of a hybrid, first-principles and artificial intelligence (AI)-directed computational model. The stability of the structure, electronic band structure, density of states, and optical properties of Rb 2 YAgX 6 (X = F, Cl) were studied using Density Functional Theory (DFT). The analysis shows that Rb 2 YAgF 6 has a stable cubic structure and has a bandgap of about 2.1 to 2.3 eV, which qualifies it to be used in the photodetection of the visible light. Simulations carried out by SETFOS also indicate that the performance of the device is highly dependent on the thickness of the absorbers, the concentration of doping, and the density of interface defects. Optical examination reveals the presence of intense absorption in the wavelength of 350 to 600 nm with an absorption coefficient of about 10 5 cm − 1 . In order to enhance the performance of the photodetector, AI-guided machine learning model was built on simulated data sets. Some of the device parameters that are considered during the optimization are the thickness of the absorber layer (200–600 nm), carrier mobility (10–50 cm 2 V −1 s − 1 ), carrier defect density (10 13 -10 16 cm − 1 ) and interface recombination velocity (10 3 -10 5 cm − 1 ). The trained model forecasts the effects of these parameters on such critical device parameters as responsivity, external quantum efficiency, and specific detectivity. The optimized design has a responsivity of about 0.38 AW − 1 , an external quantum efficiency of over 75% and a detectivity of the order of 10 12 Jones. These findings reveal that AI-assisted optimization allows much better performance of a device at a lower cost of computation. The combined first principles and AI-based design approach offers a good route with which to arrive at high-performance and environment-friendly double perovskite photodetectors.

Links and sources

• PDF from original source

Comments