TiO2 and NiO:Cu Carrier-Selective Barrier Layers for Heterojunction Solar Cells

Abstract

Next-generation solar cell designs will figure significantly in the approaching green energy mix. This thesis makes contributions to both the modeling and the materials of these devices. External quantum efficiency-based measurements of PEDOT/n- Si/TiO2 solar cells are used, in conjunction with modeling, to extract the surface recombination velocity of the n-Si/CVD-TiO2 interface. It is also shown that even very thin layers of ALD-TiO2 can effectively passivate crystalline silicon, though even the thinnest layers contribute measurably to the hole barrier. The increase in surface recombination velocity after ALD-NiO overlayer deposition is linked to changes in the TiO2 stoichiometry as measured by XPS. We also present significant contributions to the transient electrical modeling of double-heterojunction silicon solar cells. It is shown how these measurements can be used to determine both the front- and back-interface recombination velocities in fully fabricated solar cells. Physics-based modeling is combined with full device simulations to account for geometric effects, and the full results are applied to measurements of PEDOT/n-Si and PEDOT/n-Si/TiO2 heterojunction solar cells, showing that emitter efficiency is an increasingly significant problem in these devices as the back interface is made more ideal. Next, a thorough account is given of a novel ALD-NiO deposition process, outlining how changes in deposition parameters change the content of the resulting films. A process is also developed to fabricate ALD-CuO, and the two are integrated to give, for the first time, Cu-doped NiO deposited by ALD. The copper is shown to increase film conductivity, an effect corroborated by spectroscopic band measurements. ALD-NiO is shown to depin the interfacial Fermi level in NiO/c-Si diodes; integrated into perovskite solar cells, it gives results comparable to the more standard solution-deposited NiO films, with a small VOC improvement. Finally, the last chapter of this thesis leverages patent data to examine how work or collaboration across economic sectors influences research choices and output. First, it is shown that university-industry collaborations tend to produce more valuable patents. Next, using a novel data set of those who have patented in each of the military, commercial, and academic sectors, it is shown that these peoples’ least influential contributions tend to come when they do military work.