| Popis: |
This dissertation represents a collection of related studies that employ the various capabilities of spectroscopic ellipsometry (SE) to characterize and gain insights into the properties of the materials of relevance to advanced cadmium telluride (CdTe) photovoltaic technology. The structural, optical, and electrical properties of the component layers of the CdTe solar cell have been investigated using different SE data collection modes and analysis techniques. The component layers of the CdTe solar cell are deposited on soda lime and TEC-15 glass substrates in the superstrate configuration, i.e., with the solar irradiance entering through the glass substrate. The key layers include a transparent conducting oxide front contact, typically pyrolytic fluorine-doped tin oxide (SnO2:F); a high resistivity transparent layer (HRT) of either pyrolytic SnO2, sputtered MgxZn1 xO (MZO), or both; semiconductor layers of either n-type cadmium sulfide (CdS), cadmium selenide (CdSe), or both; p-type CdTe; a p+ back contact interlayer typically copper based; and a metallic conducting back contact layer, such as gold. In this research, SE-deduced component layer properties have been correlated with the CdTe device performance parameters. Applying various SE capabilities not only identifies the correlations between the material properties and solar cell performance but also enables optimization of the preparation conditions (e.g., substrate temperature) and resulting properties (e.g., thickness) of the CdTe device components. First, we have employed UV-VIS SE and ex-situ mapping SE results to correlate between the CdSe optical and structural properties with the CdTe solar cell performance. The effects of varying CdSe layer thickness on the CdTe solar cell performance have been studied, focusing on the TEC-15/HRT/CdSe/CdTe/Cu/Au cell structure. A set of four 6.5 cm x 6.5 cm TEC-15/HRT structures were coated with different nominal thicknesses of CdSe for incorporation within the CdTe solar cell. The CdSe replaces the conventional n-type CdS window layer. The CdSe layer has been deposited at room temperature on TEC-15/HRT glass and the optical properties and thicknesses of the resultant films were studied by mapping SE. By using the CdSe layer in place of CdS, the goal is to increase the short circuit current (JSC) of the CdTe solar cell due to both short and long-wavelength collection, but without a significant loss in the product of the open-circuit voltage (VOC) and fill-factor (FF). The room temperature CdSe bandgap is ranged from 1.748 eV to 1.772 eV for the CdSe effective thickness range of 91 – 145 nm, where the effect of in-plane film stress appears to cause this correlation, which is in good agreement with the unalloyed CdSe bandgap the found in literature. Although CdSe has a larger bandgap (~ 1.75 eV) than CdTe (~ 1.5 eV), alloying CdSe with CdTe is considered the best approach for reducing the bandgap due to the bowing effect observed when alloying up to 50 at.% CdSe with CdTe to obtain ~1.35 eV. For the sputtered CdSe/CdTe solar cell, the maximum temperature of 390 C is reached upon CdCl2 treatment of the TECTM-15/CdSe/CdTe as-deposited structure. At this temperature Se and Te interdiffusion into CdTe and CdSe, respectively, and this leads to an alloy with a narrower bandgap than CdTe at the top (sun-side) of the device. As a result, collection can occur at photon energies lower than the bandgap of the CdTe. Because carrier collection can also occur from the top-most Se-rich region of the device, the device responds well in the blue to the near-ultraviolet range as well, where conventional devices are negatively impacted by absorption losses within the CdS. In this research, four 81 81 array dot cells have been fabricated on 6.5 cm x 6.5 cm TECTM-15/HRT glass by depositing CdSe at the center of the substrate platform for CdTe deposition, in order to assure relative uniformity in the CdTe layer thickness. A reduction in the VOC-FF product for the CdSe/CdTe solar cells compared to the standard CdS/CdTe devices motivated a plan to prepare an alternative n-type semiconductor structure for sputtered CdTe solar cells using the bilayer combination of CdS/CdSe. With this bilayer combination for CdTe solar cells, both the CdS and CdSe layers must be optimized. Ex-situ M-SE has been used to face the challenge of thickness optimization and guide the design of the bilayer. A variety of CdS depositions were performed on 6.5 cm x 6.5 cm TECTM-15/HRT structures with intended center thickness ranging from 6 nm to 100 nm. Because of the smaller size of the substrates, the CdS thickness was more uniform over the area compared to the CdSe thickness. Two center CdSe thicknesses were explored, i.e., 110 nm and 175 nm, which gave larger ranges of thickness due to the greater non-uniformity associated with the CdSe deposition. The optical responses of all layers from this and previous studies were used to interpret the variation in the measured performance for different CdS/CdSe/CdTe devices. The results show clear correlations between the performance parameters of the CdS/CdSe/CdTe solar cells and the effective layer thicknesses of both the CdS and CdSe. Adding a 13 nm thick CdS layer and thinning the CdSe layer from 150 to 100 nm can provide an optimum device performance that, due to improvements in both VOC and JSC, is higher than that of the best CdSe/CdTe cell. Furthermore, by reducing the thickness of the inactive CdS window layer of the conventional cell and adding the 100 nm thick CdSe layer results in cell performance higher than the standard sputtered CdS/CdTe developed over many years. It should be emphasized that this higher performance is achieved with the low-cost TECTM-15/HRT substrate and an all-sputtered, low-temperature process with a maximum of 390C during the CdCl2 treatment.Finally, the properties of n-type magnesium zinc oxide (MgxZn1-xO or MZO) thin films have been explored. These films have been deposited to play roles as novel front contact materials in CdTe solar cells. Ex-situ SE measurements have been applied to determine the structural and optical properties of MgxZn1-xO layers of different compositions deposited by sputtering on glass substrates. The resulting dielectric functions from the SE analyses were parameterized as a function of the fundamental bandgap, which in turn can be expressed in terms of composition. As a result of the parametric form developed here, the dielectric function of the MgxZn1-xO can be predicted based on a specification of the bandgap or composition x. Using the parametric form of the MgxZn1-xO dielectric function, M-SE has been used to characterize the uniformity in the structural properties and bandgap of MgxZn1 xO film deposited on 16 cm x 16 cm samples of TECTM-15 by magnetron sputtering from MgO and ZnO targets. Due to the co-sputtering method, the non-uniformity of the bandgap and composition can be observed, and by varying the power to the two targets, the Mg content will be varied, and the bandgap tuned to an optimum center value on the substrate based on correlations with solar cell performance. The solar cell fabrication process studied here involves the deposition of CdS/CdTe on TECTM-15/MZO structures. For tracking the changes in the optical properties due to interactions between layers in this process, infrared spectroscopic ellipsometry (IR-SE) measurements have been applied to the as-deposited TECTM-15/MZO/CdS at different stages of the fabrication process. The IR-SE measurement is considered a powerful method to determine the free-carrier and the phonon mode parameters of complex thin-film structures. Although no free carriers are detected in the MZO layers deposited under conditions used for solar cells with x ~ 0.3 and a 3.6 eV bandgap, decreases in the resistivity and increases in scattering time are observed for the TECTM-15 SnO2:F upon MZO over-deposition and annealing at 250C, the standard temperature for CdS/CdTe deposition. Thus, improvements in the SnO2:F are induced by subsequent processing. In a final demonstration of SE for analysis of solar cell structures incorporating MZO, through-the-glass (TG-SE) measurements have been applied to as-deposited TECTM-15/MZO/CdS/CdTe stack to determine the structural and optical parameters of the layers in the device stack. Among the structural parameters determined include SnO2:F, MZO, and CdTe bulk layer thicknesses, as well as SnO2:F/MZO, MZO/CdS, CdS/CdTe interface roughness, and CdTe surface roughness thicknesses. Because of the relatively large roughness on the MZO surface of ~ 60 nm compared with the standard SnO2 HRT, all detected CdS resides within the MZO/CdS and CdS/CdTe interface layers for the sample studied in detail. The electronic properties determined include the SnO2:F resistivity and scattering time, the MZO bandgap which provides the Mg content, and the CdTe bandgap which provides the in-plane stress in the layer. |
