Experimentally anchored model for sputtered alumina encapsulation of perovskite films

Moisture ingress remains a major barrier to metal-halide perovskite photovoltaics, while scalable encapsulation requires a quantitative relation between barrier design and protected-film lifetime. We present an experimentally anchored transport–degradation–optical framework for RF-magnetron-sputtered Al 2 O 3 on MAPbI 3 films. The model couples thickness-dependent, defect-mediated water-vapour transport to a cumulative moisture-dose response and transfer-matrix optical modelling. A reference transport coefficient was calibrated using the film-level T 80 of a 10.35 nm encapsulated film aged at 85 ∘ C and 60% relative humidity. The film retained 80% of its combined structural, absorption and photoluminescence proxy for 937 h, compared with 49 h for the unencapsulated control; both values are single-specimen film-level estimates. The calibration implies an effective diffusivity of 1.45 × 10 − 19 m 2 s − 1 and a model-inferred water-vapour transmission rate of 2.59 × 10 − 4 g m − 2 day − 1 at the test condition. At 85 ∘ C and 85% relative humidity, the model predicts T 80 values of 661 h for 10.35 nm and 6158 h for 15 nm; the latter gives a predicted 3.9 × 10 − 5 g m − 2 day − 1 , below the widely used 10 − 4 barrier benchmark. Deterministic sensitivity analysis gives a 15 nm T 80 of 4.7–8.8 kh for ± 30 % variation in the calibrated coefficient and preserves benchmark compliance for plausible changes in the defect-coalescence length. The layer introduces negligible modelled optical loss. The framework converts film-ageing data into a testable thickness window for film-level packaging, although the transmission rates are model predictions that await direct permeation measurement, and the lifetimes remain moisture-limited film estimates rather than device or module lifetimes.