摘要

Transparent conductive oxides (TCOs) are a known failure mode in a variety of thin film photovoltaic (PV) devices, through mechanisms such as resistivity increase and delamination. Degradation science studies of these materials, as well as most PV systems, have primarily utilized industry standard qualification protocols, which are not designed to be used as lifetime prediction tests. This work applies a data science approach to this engineering challenge, utilizing commercially available TCOs and subjecting them to an array of stressors, including environmental and material stressors. Optical, electrical and surface sensitive TCO property metrics were monitored and analyzed en mass. Different degradation mechanisms and modes were observed when different stressor combinations were applied; TCO surfaces are sensitive to the proportion of water and light in an exposure, yellowing of the TCO only occurs when humidity and UV light are combined, and PEDOT:PSS (poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate)) application results in hazing and roughening of aluminum-doped zinc oxide (AZO). Using multi-variate analytics and plotting critical material properties against one another in a mechanistic plot, trade-offs between properties and the activation of different degradation mechanisms become readily apparent. In addition to a survey of failure modes of TCOs, a possible solution to the degradation of AZO was examined: the application of an organofunctional silane layer. The application of a thin APTES (3-aminopropyltriethoxysilane) film nearly eliminated the observed edge effects and greatly reduced the resistivity increase caused by damp heat exposure of AZO.

  • 出版日期2015-12