Análise estatística de processos para a otimização de fotoeletrodos aplicados em células solares sensibilizadas por corante

Abstract

Photovoltaic solar energy is an option to supply the growing consumption of electricity, with low environmental impact. Dye-sensitized solar cells are among the alternatives studied by the researchers, due to their ease and low manufacturing cost. The development of this technology depends on improving the performance of these devices, making studies with the objective of investigating the operation and proposing improvements very important. Implementing a 2 3 Factor analysis, it was possible to analyze modifications of three parameters of the photoelectrode simultaneously. The statistical analysis showed that the change in the composition of the film, mixing two different semiconductor compounds, was relevant, resulting in an increase of 93.12% in the value of the short-circuit current density and an increase of 28.71% in the circuit voltage open. The results of electronic microscopy show that a more uniform layer was obtained in photoelectrodes made with two semiconductors, expanding their surface area. Also, the reproducibility of the thickness of the films on the photoelectrodes was analyzed, using the blade coating technique. The coefficient of variation of thicknesses can be reduced by controlling the speed and force at the time of deposition of the semiconductor layer. In order to control the force and speed, an automated equipment for film deposition was built, with adjustment of these variables, allowing this study to be carried out. Thickness measurements, made with a profilometer, reveal less variability in films manufactured by applying a lower speed and a higher force intensity. Photoelectrodes constructed with greater thickness also presented low standard deviation values, indicating good reproducibility. Photovoltaic tests show the importance of controlling the thickness of the photoelectrode layers, where the best results were obtained in films of 9.91 ± 0.154 μm, resulting in an efficiency of 4.06%.