Home Research Classes Astronomy Sporting Clays Affiliations My Library Links

Photovoltaic devices can be used to generate electricity from a clean, renewable energy source, the sun. However, one large problem must be overcome before widespread use of photovoltaic devices can become commonplace. Current photovoltaic devices have low conversion efficiency. Extrinsic defects such as Cu and Na can play important roles in decreasing the conversion effiency of thin-film polycrystalline solar cells. In CdTe, Cu is often used as part of a back contact preparation but Cu easily migrates to the CdS/CdTe junction region and into the CdS during processing and perhaps during cell operation.  Furthermore, its chemical state and electrical activity can be changed by moderate light soak and recovered again by subsequent heat treatments. In our work, we have used ion implantation with photoluminescence, electroluminescence, and x-ray spectroscopies (Fluorescence and Absorption) to improve our understanding of Cu impurities in CdTe grown on either float glass, quartz or kapton substrates. These measurements have been a collaborative effort with the University of Toledo Physics Department (Al Compaan and Xiangxin Liu).

Synchrotron x-ray fluorescence was used to determine the amount of Cu impurities in the CdTe films and substrates (Quartz, BSG, SLG). Films of CdTe 2-3 mm thick were magnetron sputtered onto borosilicate glass slides.  The Cu Ka fluorescence at 8.05 keV from these films was excited by the x-ray beam at 9.5 keV, above the K-absorption edge of Cu at 8.98 keV.

X-ray Absorption (XAS) measurements on these samples were used to determine the local structure around the Cu impurities. The XAS measurements clearly show that the local structure around the Cu atoms was different between CdCl2 treated samples and untreated samples on a quartz substrate. This suggests that the local structure of the impurity in the substrate may be a large factor in the device properties. This is also implied by photoluminescence differences in the BSG and SLG substrates. We now have an excellent handle on Cu in CdTe and are now starting to explore the role of Cu in CdS.

Jeff at SSRL: Here I am standing in front of the XPD chamber that I used to collect my Thesis DATA.