This repo contains the data and algorithms for manuscript "High-throughput micro-scale bandgap mapping for perovskite-inspired materials with complex composition space".
All hyperspectral imaging hypercubes, spatially-resolved spectra and bandgap analysis results can be found at https://osf.io/nx2ae/.
The bandgap extraction algorithm in this manuscript was adapted from Siemenn, A. E. et al. Using scalable computer vision to automate high-throughput semiconductor characterization. Nat. Commun. 15, 4654 (2024). The original code is available at https://github.com/PV-Lab/Autocharacterization-Bandgap. The main changes made for better fitting and generalizability are summirized in the table below:
| This work | Previous work | |
|---|---|---|
| Range | Full data range | Manual selection of target bandgap range |
| Resolution | Spatially-resolved bandgaps (N x N spectra) | Average bandgap of each droplet (1 spectrum) |
| Fitting | Linear regression on maximum difference | Linear regression on detected peaks |