Thin film CdTe based solar cells have the potential for high efficiency and have been investigated for 30 years due to their variations in flexibility in manufacturing technology, rapid deposition, and an excellent match to the solar spectrum. Despite these promising attributes, these devices have not reached their full potential. Their large scale implementation is limited by several factors including back contact and stability issues. The device performance is not well co-related with different design parameters and to date there is no generally accepted model for carrier transport.
Fabricating ohmic contacts to a p-type CdTe cells, due to its high work function, has been a long-standing problem in the photovoltaic industry. Many different contact materials and processes have been developed, but still, the contacts at their best are quasi-ohmic. In this work, a semi-transparent back contact using Cu doped ZnTe film was developed by galvanic deposition. The bifacial device configuration with transparent ZnTe:Cu back contacts permitted us to perform bifacial device characterization by illuminating the device through either the front or back contact or simultaneously through both. This approach to device characterization using a bifacial device is novel contribution in the field of thin film CdTe photovoltaic characterization techniques because it allows us to determine the fundamental minority carrier transport parameters such as the diffusion length (L) and depletion width (W) and separate effect of front junction from the back contact. The cells used for this project were produced by different methods but had similar efficiencies ∼10-12%, even though the processing conditions and back contact fabrication techniques and materials varied.
Bifacial current voltage (JV) analysis of illuminated devices indicated that the back contact was photosensitive. Experimental results prove that CdTe device operation is controlled solely by the primary CdS/CdTe heterojunction. Detailed bifacial spectral analysis on devices made with different CdTe thicknesses and different applied biases was used to quantify the transport parameters L and W. CdTe devices with absorber thickness in the range of 3-8μm were investigated. The diffusion lengths derived were in the range of 0.6-0.7μm. In CdTe devices the absorption co-efficient is in the range of 105 1/cm, W in the range of 2-4μm. As a result the absorption depth αW > 1, and all the carrier generation and absorption occur in the depletion region even for forward bias of 0.5V. Bifacial spectral response characterization results prove that for CdTe solar cells, the operation is determined by voltage dependent current collection and not diffusion length.
Laboratory simulated accelerated test life conditions were used to investigate the cell degradation phenomenon, with direction towards identifying device degradation mechanisms. Stressing results show that primary degradation in xvii efficiency occurs due to degradation in the open circuit voltage (Voc) and the fill factor (FF). Cyclic stress, where in the devices were subjected to alternating light and dark light bias cycles and switching applied bias during stress experiments performed on CdTe devices allowed us to identify and partially explain transient degradation and recovery mechanism in CdTe devices. Results from stressing experiments performed on CdTe devices made with different back contact materials showed that cells with ZnTe:Cu based contacts had minimum degradation in Voc. Bifacial JV analysis performed on these stressed devices revealed a photo-sensitive back contact.
Temperature dependence of the open circuit voltage, with different back contacts was studied under different illumination conditions. Measurements showed there were two distinct regions, one above 220K where open circuit voltage linearly increases with temperature, and one below 200K where Voc becomes nearly independent of both temperature and light intensity. A model explaining this "Voc saturation" behavior has been proposed in this document.
Photovoltaic device modeling results obtained using AMPS (Analysis of microelectronic and photonic structures) suggest that the dominant recombination mechanism is the SRH recombination through midgap states.