Inorganic Photovolatics and Transparent Conducting Oxides

While most conducting materials (like copper) are not transparent, and most transparent materials (like glass) are not conducting, there is an important class of materials that are both transparent and conducting. These materials are important to the solar cell architecture as they provide a transparent window which allows sunlight to pass through while also allowing the electricity to conduct out of the cell. Presently there are few materials known to possess both properties, and these materials, such as indium-tin-oxide, are very expensive. We are studying novel transparent conducting oxide materials made from earth-abundant elements, such as Zn, Ni, and Co. Advances here will help both inorganic and organic PV materials.

This work is a DOE-EFRC funded project in collaboration with NREL, Northwestern, and Oregon State University. Our collaborators synthesized samples with various compositions under different conditions. Specifically, powder samples are fabricated under almost equilibrium conditions at Northwestern, and polycrystalline and textured thin films are made at NREL using sputtering and PLD. At SSRL, we use the Anomalous X-ray Diffraction (AXRD) to correlate the cation site occupancy to the electronic properties such as conductivities. This technique allows us to probe precisely where the cations are sitting in these complex oxide structures. So far, the focus is to investigate transition metal oxides with the spinel structure, AB2O4. The two different cations can reside on either the tetrahedral interstices or the octahedral interstices of oxygen lattice. By comparing AXRD experimental results to the structure factor calculations, we can determine the degree of inversion in the spinel samples, and thus the cation site occupancy.

Fig. 1 This shows the data analyzed for tetrahedral sites near the Ni absorption K edge for one stoichiometric NiCo2O4 thin film. The red circles are integrated intensities measured at SSRL whereas the lines are calculated intensities with different degrees of inversion, ν. The best fit is 82% for the Ni224 data, indicating that only 18% of the tetrahedral sites are occupied by the Ni where as the rest are occupied by Co.

Fig. 2 A recent photo of the team of EFRC Center for Inverse Design working on beamline 7-2. From left to right: Linda Lim, another graduate student from Mike Toney’s group, Paul Ndione, our collaborator from the National Renewable Energy Laboratory, and me.