Novel Pathways to High-Efficiency Chalcopyrite Photovoltaic Devices: A Spectroscopic Investigation of Alternative Buffer Layers and Alkali-Treated Absorbers

摘要

Within the past few years, breakthroughs in Cu(In,Ga)Se2 (CIGSe) thin-film photovoltaic device efficiencies (on a laboratory scale) were achieved utilizing alkali-treated (KF) absorbers. Na incorporation in the CIGSe absorber, either diffused from the substrate or deliberately deposited, affects the surface electronic properties of the CIGSe absorber. The role of Na, however, is still not fully understood with some studies suggesting that Na also passivates defects at the grain boundaries. Replacing Na with K offered an efficiency boost resulting in KF treatments becoming the new “hot topic” in the chalcopyrite field, both in terms of understanding how the treatment changes the absorber along with studying the differences between alternative KF deposition methods. To provide insight on these issues, x-ray (XPS) and ultraviolet (UPS) photoelectron spectroscopy, inverse photoemission spectroscopy (IPES), as well as x-ray emission spectroscopy (XES) are utilized to investigate two sample sets. The first set (Chapter 4) compares the effects of both KF and NaF treatments on absorbers taken from the production line of STION and the National Renewable Energy Laboratory. The purpose here is to compare how similar alkali-treatments affect chalcopyrite devices from different sources along with comparing the alkali-treatments themselves. The second sample set (Chapter 5) investigates effects of KF treatments when incorporated utilizing different deposition techniques.The most recent world record efficiency for CIGSe thin-film devices was not achieved with the KF-treatment, but with the replacement of the traditional CdS buffer layer (between the absorber and transparent front electrode) with Zn(O,S), a material offering the possibility of increasing the current collection in the shorter wavelength region of the solar spectrum. To further optimize these photovoltaic devices, an understanding of the interactions between the absorber and the buffer layer is crucial. For example, record CdS/CIGSe devices have a flat conduction band alignment at the buffer/absorber interface, while, in contrast, the less efficient CdS/Cu(In,Ga)S2 device exhibits a cliff-like conduction band offset, impeding electron transport. Thus, a determination of the conduction band offset is, among other aspects, of significant importance.When using Zn(O,S) as the buffer layer, it should be noted that the bandgap of a Zn(O,S) alloy exhibits a strong bowing effect as the O:S ratio varies. With the ability to change the O:S ratio and alter the bandgap, it is thus important to understand the chemical and electronic structure of the Zn(O,S)/CIGSe interface in high-efficiency devices through direct and independent analysis of the heterojunction formation, the valence band, and the conduction band. This is the first non-destructive analysis of the interface using XPS, UPS, IPES, and XES investigating samples with varying buffer layer thickness. A comprehensive and all-experimental depiction of the electronic level alignment (Chapter 6) and chemical interactions (Chapter 7) at the interface will be presented.