Quantification of Valleys of Randomly Textured Substrates as a Function of Opening Angle: Correlation to the Defect Density in Intrinsic nc-Si:H

Optical and electrical properties of hydrogenated nanocrystalline silicon (nc-Si:H) solar cells are strongly influenced by the morphology of underlying substrates. By texturing the substrates, the photogenerated current of nc-Si:H solar cells can increase due to enhanced light scattering. These textured substrates are, however, often incompatible with defect-less nc-Si:H growth resulting in lower Voc and FF. In this study we investigate the correlation between the substrate morphology, the nc-Si:H solar-cell performance, and the defect density in the intrinsic layer of the solar cells (i-nc-Si:H). Statistical surface parameters representing the substrate morphology do not show a strong correlation with the solar-cell parameters. Thus, we first quantify the line density of potentially defective valleys of randomly textured ZnO substrates where the opening angle is smaller than 130° (ρ<130). This ρ<130 is subsequently compared with the solar-cell performance and the defect density of i-nc-Si:H (ρdefect), which is obtained by fitting external photovoltaic parameters from experimental results and simulations. We confirm that when ρ<130 increases the Voc and FF significantly drops. It is also observed that ρdefect increases following a power law dependence of ρ<130. This result is attributed to more frequently formed defective regions for substrates having higher ρ<130.