Routes Towards Improved Open Circuit Voltage in Sb2Se3 Solar Cells

Sb2Se3 is a defect tolerant, earth-abundant, stable semiconductor which has shown promise for low cost thin film photovoltaic power generation. It has a large absorption coefficient, suitable band gap, simple binary composition and benign grain boundaries compared to typical 3D materials. These attributes have helped solar cell efficiencies to surpass 10% with relatively little research effort in comparison to other absorber technologies. Despite this encouraging progress, the rate of improvement appears to have slowed recently as record devices approach the detailed balance limit of short circuit current density. This means that further improvements are likely to originate primarily from increasing the open circuit voltage (Voc) and/or fill factor (FF).<br/><br/>Interfaces within a solar cell play a key role in both the Voc and FF, and therefore we have carried out a detailed study of the partner layers in the Sb2Se3 device structure either side of the absorber layer. We have replaced the standard CdS front contact with TiO2, which has a wider band gap and is far more robust, making it more amenable to the subsequent high temperature processing steps required for large grain, crystalline Sb2Se3 thin films. The initial solution processed TiO2 layer has been replaced with a sputter deposited film which improved device uniformity and increased FF, however was prone to run-to-run variations in TiO2 film quality and the inconsistent appearance of ‘S’-shape JV curves. This issue has now been resolved through careful control of the sputtering process, which has led to exceptionally high short circuit current densities and significantly improved FF, however the impact on Voc has been limited. The back interface was then studied to determine the role of the hole transport material (HTM) in Sb2Se3 devices, and assess whether this could be an area for improvement. Several commonly used organic HTM layers were compared against one another, and to a simple Au contact. The direct Sb2Se3-Au contact is limited by a secondary barrier causing ‘rollover’ in JV curves, despite seemingly favourable band positions. Each of the HTM layers demonstrated the ability to eliminate this secondary barrier, thereby lowering the series resistance and improving the fill factor. Whilst this increases the champion efficiency, the improvement is relatively minor, especially considering the gradual oxidation of the Sb2Se3 back surface produces a similar self-passivation effect. Instead, the HTM layers have a far more drastic effect on the average efficiency, which is attributed to a pinhole blocking effect by the solution processed organic films. Nonetheless, Voc remains below 500mV for all devices tested.<br/><br/>After thoroughly investigating the partner layers, we are now turning our attention to the Sb2Se3 layer itself. We are currently exploring several routes to improve the quality of the absorber layer. This includes a comprehensive study of dopants and defects in Sb2Se3, whereby the growth of high purity single crystals with the controlled addition of selected extrinsic dopants will inform the most suitable compositions to target in thin films, both in terms of stoichiometry and p-type dopant incorporation. This is particularly important given the apparent chalcogen-poor stoichiometry reported by most groups, especially for PVD processed films, as well as the lack of rigorous control over doping type typically found in literature reports. The transfer of knowledge gained from this single crystal study towards full devices will be accelerated via the use of a novel dual source close space sublimation chamber. This will allow co-deposition of different source material with varying stoichiometry and dopant concentration, and therefore allow a wide range of material compositions to be rapidly deposited. These compositions will then be assessed in terms of their implied Voc in order to carry out a loss analysis to identify the different contributions to the remaining deficit.