Interpretable machine learning diagnosis of degradation mechanisms in lead-free solar cells from temperature- and illumination-dependent electrical measurements
Organizations: Department of Electrical and Electronic Engineering, Bangladesh University of Engineering and Technology, Dhaka 1205, Bangladesh
Abstract
Machine learning coupled with device simulation is widely used to optimize thin-film solar cells, but almost exclusively in the forward direction, from device parameters to predicted performance. The inverse problem encountered after fabrication, namely identifying the mechanism that limits a given device, is still addressed by manual inspection of characterization data. In this work, both directions are examined for lead-free solar cells using approximately 42,000 SCAPS-1D simulations. Replacing the Spiro-OMeTAD hole transport layer with CuI increases the simulated power conversion efficiency from 19.6% to 21.6%, almost entirely through the fill factor. Surrogate models trained and tested on a factorial parameter grid substantially overstate their predictive accuracy. An XGBoost regressor attains on held-out grid points but only 0.58 on independently sampled devices, whereas training on a Latin hypercube design of comparable size restores . For the inverse problem, a multi-label classifier identifies which of six degradation mechanisms are present from simulated temperature- and illumination-dependent open-circuit-voltage measurements. It achieves a macro score of 0.938, compared to 0.687 for a calibrated physics-rule baseline. The classifier does not, however, generalize to combinations of mechanisms absent from the training set.
Figures & tables
| Parameter | FTO | TiO 2 | RbSn 0.5 Ge 0.5 I 3 | CuI |
|---|---|---|---|---|
| Thickness ( ) | 0.4 | 0.02 | 0.35 | 0.1 |
| Band gap ( ) | 3.3 | 3.0 | 1.2 | 3.1 |
| Electron affinity ( ) | 4.1 | 4.0 | 4.0 | 2.1 |
| Relative permittivity | 9 | 9 | 6.5 | 6.5 |
| ( ) | ||||
| ( ) |
| Model | Held-out grid (1,024) | Off-grid (300) |
|---|---|---|
| XGBoost | 0.9997 | 0.579 |
| Random forest | 0.9987 | 0.811 |
| MLP | 0.9979 | 0.994 |
| Quadratic | 0.9689 | 0.955 |
| Linear | 0.8024 | 0.725 |
| Input | Model | macro |
|---|---|---|
| 15 physics features (Section 2.4 ) | XGBoost | 0.935 0.001 |
| 15 physics features | MLP | 0.915 0.007 |
| raw curves only (13 values) | XGBoost | 0.811 0.003 |
| raw curves only (13 values) | 1-D CNN | 0.857 0.007 |
| information-matched raw curves (33) | XGBoost | 0.855 0.005 |
| information-matched raw curves (33) | MLP | 0.863 0.012 |
| Pair | Pub. | Pred. | Dev. | Range |
|---|---|---|---|---|
| FASnI 3 / TG 45 | 5.34 | 94.6 | ||
| FASnI 3 / SP 46 | 2.38 | 10.8 |