Data-Driven Analysis of Transparency–Efficiency Trade–Offs and Indoor Spectral Matching in Perovskite Photovoltaics

Abstract

Semitransparent and indoor perovskite photovoltaics (PVs) require coordinated control of optical transmission, photocurrent, voltage retention, and spectral response. Here, we quantitatively analyze published Cs-, Br-, and Pb-based device to clarify the transparency–efficiency relationships and indoor spectral-matching behavior. Visible transmittance (T_VIS) correlates negatively with power conversion efficiency (PCE; r = −0.89) and short-circuit current density (J_SC; r = −0.53), whereas the total bandgap-referenced voltage deficit, ΔV, exhibits a broad, leverage-sensitive distribution without a compelling monotonic dependence across T_VIS. ΔV is therefore treated as a composite descriptor rather than a direct measure of nonradiative recombination. Devices achieving 12%–16% PCE typically cluster at 20%–40% T_VIS, while the only one record above 80% T_VIS reports ≈0.5% PCE at T_VIS ≈ 85% (n = 1), precluding definition of a universal high-transparency boundary. Equal-illuminance spectral power distribution (SPD) analysis combined with modeled external quantum efficiency (EQE) edges further indicate that absorbers near 1.9−2.1 eV provide greater modeled photocurrent than a 2.30 eV bromide-rich baselines under common white LEDs. Overall, transparency produced mainly by bandgap widening or uniform absorber thinning compromises indoor photon harvesting. Efficient semitransparent indoor devices should instead retain spectrally matched active absorbers and create transmission through area-selective or photonic designs while minimizing contact and interface losses.