Scripts for Rapid Parameter Estimation from Photoluminescence Decays of Halide Perovskite Thin Filmshttps://doi.org/10.26165/JUELICH-DATA/2IEHBFHeumann, RobinRudolph, TobyGaosheng, HuangKirchartz, ThomasDreessen, ChrisJülich DATA2026-09-092026-09-09T08:20:52ZThe dataset entails synthetic physics simulations and experimental photoluminescence measurements by HEUMANN, RUDOLPH, HUANG, KIRCHARTZ and DREESSEN (Forschungszentrum Jülich), supporting neural-network-accelerated Shockley–Read–Hall (SRH) trap characterisation of semiconductor thin films. The basic methodology is described in "Heumann et al. (2026): Rapid Parameter Estimation from Photoluminescence Decays of Halide Perovskite Thin Films , Advanced Energy Materials, [DOI]". The dataset contains scripts using Python toolsets for extracting trap parameters from time-resolved (TRPL) and steady-state (SSPL) photoluminescence measurements of semiconductor thin films using Shockley–Read–Hall recombination models with radiative recombination included. These scripts provide CMA-ES parameter fitting with an optional neural-network surrogate for the transient solver, and a Bayesian grid pipeline that maps the posterior over the full parameter space and reports credible intervals. The most current version of the scripts can be found in the GitHub repository https://github.com/roheumann/rapid-trpl-bayes. At its core, the dataset provides a large synthetic training corpus of differential carrier lifetime (τ_diff) vs. quasi-Fermi level splitting (QFLS) curves, computed by numerically solving the SRH transient ODE over a quasi-random Sobol-sampled parameter space and one to two trap levels for the material. Complementing these simulations, the dataset includes experimental time-resolved (TRPL) and steady-state (SSPL) photoluminescence measurements from perovskite thin-film samples, which serve as targets for parameter fitting and Bayesian uncertainty quantification. A further component comprises 1D spatial PDE diffusion simulations (Beer–Lambert excitation profile, film thickness 450 nm, Eg = 1.625 eV) run across five representative physical scenarios and five optical absorption coefficients (α = 10⁴–10⁶ cm⁻¹), from which a homoskedastic model-error sigma (σ_trpl = 0.4936 in log₁₀ space) is derived; this value parameterises the Gaussian log-likelihood in the Bayesian inference stage and accounts for the approximation error of the 0D point model relative to the full 1D diffusion model. Pre-trained neural-network surrogate model artefacts are provided alongside, enabling τ_diff curve prediction at orders-of-magnitude lower computational cost than the ODE solver. The data provided contains relevant variables of the semiconductor photophysical recombination cycle, including trap density (N_t), trap energy level (E_t), electron and hole SRH lifetimes (τ_n, τ_p), radiative recombination coefficient (k_rad), bandgap energy (E_g), carrier excitation density, quasi-Fermi level splitting, differential decay time and photoluminescence quantum yield. (2026)Computer and Information SciencePhysicsperovskite solar cells, Bayesian inference, transient photoluminescence, steady-state photoluminescence, uncertainty quantificaiton, machine learning, Shockley-Read-Hall recombination, charge-carrier lifetime, parameter estimationdoi2026-09-01Heumann, RobinHeumann, RobinRudloph, TobyHuang, GaoshengKirchartz, ThomasDreessen, Chris2026-09-01CC0CC0 Waiver