Interface architecture between TiO2/perovskite, perovskite/hole transport layer, and perovskite grain boundary (Conference Presentation)

Abstract

In order to examine the interface structure of TiO2/perovskite layer, quartz crystal microbalance sensor (QCM) was used. On the QCM sensor, TiO2 layer was fabricated and the PbI2 solution in Dimethylformamide (DMF) was passed on the QCM sensor to estimate the adsorption density of the PbI2 on the titania2. The amount of PbI2 adsorption on TiO2 surface increased as the adsorption time and leveled off at a certain time. PbI2 still remained even after the solvent only (DMF) was passed on the TiO2 layer on QCM (namely rinsing with DMF), suggesting that the PbI2 was tightly bonded on the TiO2 surface. The bonding structure was found to be Ti-O-Pb linkage by XPS analysis. We concluded that the Ti-OH on the surface of TiO2 reacts with I-Pb-I to form Ti-O-Pb-I and HI (Fig.1 B). The surface trap density was measured by thermally stimulated current (TSC) method. Before the PbI2 passivation, the trap density of TiO2 was 1019 cm3. The trap density decreased to 1016/cm3 after the PbI2 passivation, suggesting that the TiO2 surface trap was passivated with I-Pb-I. The passivation density was tuned by the concentration of PbI2 in DMF, by which TiO2 layer was passivated. Perovskite solar cells were fabricated on the passivated TiO2 layer with various PbI2 passivation densities by one step process (mixture of PbI2 + MAI in DMF). It was found that Jsc increased with an increase in the Ti-O-Pb density. We concluded that the interface between TiO2 and perovskite layer has passivation structure consisting of Ti-O-Pb-I which decreases the trap density of the interfaces and supresses charge recombination. The effect of Cl anion on high efficiency is still controversial when perovskite layer is prepared by one step method from the mixture of MAI and PbCl2. It was found that adsorption density of PbCl2 on TiO2 surface was much higher than that of PbI2 from the experiment using QCM sensor. After the surface was washed with DMF, Cl and Pb were detected. These results suggest that the TiO2 surface was much more passivated by PbCl2 than by PbI2. This may explain partially the high efficiency when the perovskite layer was fabricated by one step process consisting of MAI and PbCl2 solution. We also observed that the crystal size increased with an increase in the amount of Cl anion which of course one of the explanation of the high efficiency. The interface of hole transport layer/perovskite layer, and between perovskite layer /perovskite layer (grain boundary) was passivated with organic amines. The passivation was also effective for increasing Voc and Jsc. This was explained by the results of transient absorption spectroscopy that the charge recombination time between hole transport payer/perovskite layer increased from 0.3 μsec to 60 μsec.