| Summary: | 碩士 === 國立東華大學 === 材料科學與工程學系 === 103 === In this study, we use electrochemical impedance spectroscopy (EIS) analysis to confirm the corresponding reaction mechanisms to five resonance peaks in the Bode plots of dye-sensitized solar cells. In the future, if anyone want to improve the efficiency of dye sensitized solar cell, we can compare measurement results by electrochemical impedance spectroscopy analysis before modification and after modification, following the same approach. By referencing this study, one can find out after the modification which component associated with certain mechanism will be affected by the modification, and to tell whether modification was successful. This provides a great help to our group in continuing the study of dye-sensitized solar cells in the future.
From the analysis of Bode plots we found five resonance peaks A, B, C, D, and E, associated with five mechanisms, respectively. These five mechanisms were deduced from a rigorous process of elimination and differentiation from seven possible mechanisms that has passed a preliminary elimination process. The corresponding mechanisms are listed here. The first is the electron recombination at the interface between ITO and TiO2 (mechanism 1), responsible for A peak, where its response frequency is about 10-1 ~ 100 Hz. The second is the electron recombination on titanium dioxide surface with the electrolyte (mechanism 3), responsible of B peak, where its response frequency is about 100 ~ 102 Hz. The third is the charge redistribution above the interface between titanium dioxide and electrolyte solution when increase the applied voltage (mechanism 5), responsible for C peak, where its response frequency is about 101 ~ 103 Hz. The fourth is electron recombination at the interface betwwen electrolyte solution and the Pt counter-electrode (mechanism 7), responsible for D peak, where its response frequency is about 103 ~ 104 Hz. The fifth is electron recombination with holes in ground state level of dye in the dye molecule (mechanism 4), responsible for E peak, where its response frequency is about 104 ~ 105 Hz.
In addition, the resonance peak associated with electron transport in TiO2 layer (mechanism 2) cannot be resolved in our measured Bode plots, because of the very small phase difference angle in our measured Bode plots. Similary, the resonance peak associated with ion transport in the electrolyte (mechanism 6), with response frequency about 10-1 ~ 100 Hz,
vii
cannot be resolved easily either, since it has also a very small phase difference angle that can only be observed under magnification.
We identified some reaction mechanisms and their reaction rates that were yet reported in the literatures or clearly explained using Bode plots. In this study, we successfully, through different experiments using Bode plots and lots of discussions, found the corresponding reaction rate associated with three mechanisms. These mechanisms are the electron recombination in the interface between ITO and TiO2 (mechanism 1), the redistribution of charge above the interface between titanium dioxide and electrolyte solution (mechanism 5), and ion transport in the electrolyte solution (mechanism 6). The confirmation of the presence of the mechanism of the charge redistribution above the interface between titanium dioxide and electrolyte solution (mechanism 5) is our special contribution since the corresponding resonance peak was difficult to be seen in Bode plots measured only at VOC, as a common practice adopted in conventional studies.
With the help of found formula in the literature, we confirmed the decease of peak phase difference angle is caused mainly by the decrease of equivalent impedance of the interface.
We also learned that the voltage drop across the TiO2/electrolyte interface is most significant compared with those across other interfaces, in response to increase of the applied voltage, from the observation of the trend of phase difference angle change first increase then decrease with applied voltage of B peak, for the ever dye-immersed devices.
From the capacitance vs. voltage diagram for dye-immersed devices, we can see capacitance for B peak increases steadly, unlike that of D peak, because the dielectric layer associated with capacitance of peak B is substantively present while that for peak D is not, which makes the capacitance associated with peak D eventually declined at high voltages.
In addition, that the phase difference for C peak will be smaller as the of the applied voltage increases was observed. This trend let us know that the interface capacitance will also be smaller as the applied voltage increases, and charge stored at the interface between TiO2 and electrolyte solution will be varied with the increase of voltage. We further studied the voltage dependence of interface charge distribution at the TiO2/electrolyte solution interface by considering three factors that may affect the concentration of I - , I3 - at the interface and predicted that the concentration
viii
of I-, I3- at the interface is higher that that away from the interface at zero bias.
Finally, we compared the Bode plots of devices with electrode without dye soaking and that with dye soaking, and found that the device without dye soaking has an earlier declination of capacitance of D peak with voltage as the film thickness increases. This is because that the devices without dye soaking produce larger current on the Pt counter electrode surface which cause the destruction of the electrical double layer on the surface.
Keyword : Spin coating, Electrochemical impedance spectroscopy, Titanium dioxide, Bode plots, Dye-sensitized solar cell, resonance peak
|
|---|
