Doping TiO2: AI finally untangles the skein of descriptors
AI-generated hypothesis · Pre-publication · To be tested experimentally
Table of contents — full brief
- Hypothesis and mechanismCausal chain, key assumptions, residual unknowns
- State of the artVerified references and counter-evidence (DOIs)
- Falsifiable predictionsQuantitative bounds, statistical tests, H0
- Experimental protocolThree phases — in silico → minimal → full
- Impact analysisNovelty, residual gaps, available data
- Panel reviewFive personas + meta-review
Verified references
5 of 11 references- DOI: 10.1021/acs.chemrev.2c00061 ↗
Machine Learning for Electrocatalyst and Photocatalyst Design and Discovery.
2022 - DOI: 10.1021/acsami.3c18490 ↗
Knowledge-Driven Experimental Discovery of Ce-Based Metal Oxide Composites for Selective Catalytic Reduction of NOx with NH3 through Interpretable Machine Learning.
2024 - DOI: 10.3390/MET11081159 ↗
Application of Machine Learning Algorithms and SHAP for Prediction and Feature Analysis of Tempered Martensite Hardness in Low-Alloy Steels
2021 - DOI: 10.1002/advs.75815 ↗
Interpretable Machine Learning Framework for Nb─Si Based Alloy Design with Enhanced Fracture Toughness
2026 - DOI: 10.1007/s10853-025-11154-4 ↗
Review: machine learning approaches for diverse alloy systems
2025
+ 6 more references
Detailed panel scores
The protocol explicitly defines falsifiable predictions with quantitative thresholds and null hypotheses, which constitutes good practice for the testing of hypotheses.
The hypothesis exploits SHAP for interpretability in photocatalyst design, an opportune and pertinent approach given the growing emphasis on explainable AI in materials science.
The hypothesis explicitly defines quantitative thresholds (R² ≥ 0.8, SHAP ≥ 0.05, rank stability ≥ 0.9), which is commendable for testability.
Addresses a critical R&D bottleneck in photocatalysts: the identification of key descriptors of activity under visible light, which may accelerate the design of efficient doped TiO2 for solar fuel production and environmental remediation.
A clear hypothesis with quantifiable success criteria (R² ≥ 0.8, SHAP threshold, rank stability) that align with the norms typically used to evaluate ML models, which facilitates the panel's assessment of feasibility.
Receive the next SPORE hypotheses
Once or twice a month, in your inbox. No spam, one-click unsubscribe.
Your data stays private. No third-party sharing. GDPR-compliant.