Answers are grounded in this paper and cite their sources.
Simulation-guided comparison of coatings shows improved efficiency in silicon solar cells, indicating better optical performance.
This study presents a simulation-guided and experimentally supported comparison of CNT–PVA and TiO₂–ZnO–CNT–PVA antireflection coatings for silicon solar cells, with the aim of improving optical coupling while avoiding exaggerated performance claims. Under standard test conditions using a QuickSun solar simulator, the uncoated reference cell exhibited an efficiency of 16.07% and a short-circuit current density of 35.88 mA/cm². The fabricated CNT–PVA and Hybrid 1 L coated cells were experimentally measured and achieved efficiencies of 20.45% and 22.22%, respectively, confirming the same performance ranking predicted by the model. The improvement is attributed mainly to reduced front-surface reflection, improved refractive-index matching, and current-driven photovoltaic enhancement, while the nearly unchanged fill factor indicates that the primary coating benefit is optical rather than electrical. Because wavelength-resolved EQE was not measured in the present work, this optical-current interpretation is supported by reflectance and I–V evidence but remains to be confirmed by future EQE characterization. A calibrated optical–electrical framework based on hierarchical effective-medium modeling and the transfer-matrix method was used to project thickness-dependent behavior over the 50–114 nm range. Within this framework, hybrid designs outperformed CNT–PVA, and the best projected Hybrid 2 L case reached 23.13% near 114 nm; however, this two-layer architecture remains a model-guided candidate pending direct coated-cell validation. A substrate-aware hardness model was used only for comparative mechanical screening and not as a substitute for nanoindentation. Overall, the contribution lies in the calibrated comparative framework and direct device-level evaluation of selected hybrid architectures rather than in claiming a wholly new ARC class.
No takes yet. Share an insight, caveat, or question.
Darwash et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: