Insights into mechanism for performance degradation in silicon solar cells under light and elevated temperature

Light- and elevated-temperature-induced degradation (LeTID) is one of the important reliability challenges for monocrystalline Si photovoltaics, yet its microscopic origin is unclear. Using electron paramagnetic resonance (EPR) on Ga-doped Czochralski Si, complemented by photoluminescence (PL) imaging to qualitatively track changes due to LeTID, a partially hydrogenated vacancy cluster is postulated as the LeTID-active defect. In the fully degraded state, EPR shows a silicon dangling-bond signal with a g -value of 2.006 and a spin density of ∼5 × 10 12 ± 10 12 cm −3 . Ab initio simulations were performed using density functional theory (DFT) to obtain g -values and energy levels for a Si dangling bond in different local environments. Consistent with experiments, DFT simulations predict average g -values over the range of ∼2.006–2.007 with energy levels of ∼0.2–0.4 eV from the valence band edge. The extent of LeTID was ∼2 times higher in the wafers obtained from the tail end of the ingot compared to the seed or middle regions, likely due to the differences in the density of intrinsic defects such as vacancies and vacancy clusters. We show that firing creates a shallow precursor state whose concentration rises with the peak firing temperature. The proposed defect model is consistent with dark annealing experiments, based on which we propose the possibility of two competing pathways for recovery in the lifetime: recovery via loss of H from partially hydrogenated vacancy clusters, or permanent regeneration through complete hydrogenation of the defects. The latter pathway offers a possible mitigation strategy. This work advances an atomic-level framework for LeTID, bridging fundamental defect physics with practical approaches for solar cell stabilization.