摘要： State-of-the-art solar cells with passivated surfaces fabricated on block-cast multicrystalline silicon (mc-Si) wafers show a pronounced degradation in efficiency under illumination at elevated temperature, as it typically occurs during operation in a solar module. This effect, frequently named 'Light and elevated Temperature-Induced Degradation' (LeTID), has been attributed to the activation of a specific, hitherto unrevealed bulk defect in mc-Si. Recent experimental results of several labs have indicated that hydrogen is somehow involved in the responsible defect physics, without however providing any direct evidence so far. In this article, we present experimental data unambiguously showing a direct positive correlation of the extent of LeTID with the hydrogen content introduced into the silicon bulk during firing of the silicon wafers coated with hydrogen-rich silicon nitride (SiNx:H) layers. Additional experiments including the pronounced impact of phosphorus gettering on the LeTID extent and the dependence of the degradation and regeneration on the wafer thickness support the involvement of a second species, with most indications pointing towards a metallic impurity. Several approaches of completely avoiding the instability in mc-Si solar cells are derived from the presented defect model, including 1) tuning of the SiNx:H layer properties to minimize the in-diffusion of hydrogen into the wafer and 2) the thinning of the mc-Si wafer, improving the getterability of the metal impurity component toward the surfaces.