Efficient two-photon water splitting photoelectrode

Abstract

This thesis presents work related to the development and fabrication of tandem photoelectrochemical (PEC) water splitting devices. These PEC cells have the potential to help mitigate the effects of solar power intermittency by converting sunlight into chemicals. The main focus of the work has been fabrication of high band gap cells intended to work in conjunction with Si devices. The materials employed as photoabsorbers have been methylammonium lead halide perovskites, which display a wide range of beneficial material properties. These include high absorption in the visible range, band gap tunability and high defect tolerance. The major shortcoming of the otherwise ideal class of material is their poor stability towards moisture and elevated temperatures. The projects included in the thesis are somewhat different in nature and range from device fabrication to the study of more fundamental material properties. The first project focused on tandem solar cell deign and fabrication. Tandem cells have the potential to be much more effcient than single band gap absorbers and their use is thus desired. Monolithic tandem devices, employing a Si bottom cell, require a top cell with a band gap of around 1.7 eV to achieve maximum effciency. The initial aim of the project was thus to fabricate single band gap perovskite absorbers with the appropriate band gap. This was achieved using the low pressure vapour assisted solution process (LP-VASP) perovskite deposition method and by varying the halide composition of the CH3NH3PbI(3-x)Brx structure. The layers were subsequently utilized to fabricate devices, which reached a power conversion effciency (PCE) of up to 15:2%, while using the normal cell architecture. Devices using the inverted conguration was also fabricated, although their performance was poorer. An atomic layer deposition (ALD) recipe for fabricating thin films of TiO2 was furthermore developed and used to deposit thin layers on top of inverted perovskite devices. Inclusion of the ALD TiO2 layers, as well as Ti contacts, was found to increase the open circuit voltages of the cells, yielding up to 1.15 V, while decreasing hysteresis. The ALD layers additionally increased the moisture resistance of the devices, although not to a sufficient degree. Further improvement of the lifetime of perovskite cells submerged in water was investigated by using a special encapsulation method. Although a functioning device was never produced, the method shows great promise. In addition to the protection schemes, sputtering was utilized to deposit ITO layers to serve as transparent top contacts. The layers obtained showed good transmission in the visible range and featured sheet resistances down to 22.3 Ω/sq. Finally, attempts at improving the performance of the inverted configuration cells were carried out, unfortunately, without clear results.The second project investigated evaporation of methylammonium iodide (MAI) using quartz crystal microbalance (QCM) measurements and mass spectrometry. Co-evaporation of MAI and PbCl2 has been shown to be a possible route of perovskite deposition, though the resultant cell performance has generally been inferior. The origin of the issues have often been attributed to difficulties controlling the MAI deposition rate. From the QCM experiments it was found that the sticking of the MAI was extremely low, especially at elevated substrate temperatures. The sticking remained low, even when PbCl2 was deposited on the QCM crystals. From the mass spectrometry it was found that MAI decomposes upon evaporation, primarily into CH3NH2 and HI. To form the perovskite from a PbCl2 film high partial pressures of MAI are consequently needed to supply a net influx of the compounds. The results have consequences for the design of deposition chambers, and additionally, provide fundamental insights into the kinetics of perovskite deposition, as well as the nature of MAI evaporation.The approach of using QCM sensors together with mass spectrometry was likewise used in the third and final project to study the thermal decomposition of various methylammonium lead halide perovskites. CH3NH3PbI(3-x)Brx was deposited on a QCM crystal and by heating the sensor, the rate of reevaporation could be estimated. Using temperatures of 85 °C and 102 °C resulted in reevaporation rates of 2.44x10-4 nm=s and 1.42x10-3 nm=s, respectively. For the mass spectrometry investigations the pure CH3NH3PbI3 was used, although CH3NH3PbI(3-x)Brx was expected to yield similar mass spectra. The results of the experiments indicated that the organic part of the perovskite starts evaporating already at 100 °C, leading to the decomposition of the perovskite. The low thermal stability limits the possible protection layers that can be applied on top of the perovskite devices and lowers the lifetime of the cells.