Efficiency and Accessibility Are Not Enough: Chemical-Validity Filtering in High-Throughput Screening of Organic Photovoltaic Materials
- 1Department of Physics, Faculty of Science, University of Yaounde, Po. Box 812, Yaounde, Cameroon
- 2Department of Physics, Faculty of Science, University of Douala, Po. Box 24157, Douala, Cameroon
Abstract
Descriptor-based screening for organic photovoltaics ranks molecules on computed properties such as frontier-orbital energies and synthetic accessibility, but such descriptors do not test whether a highly ranked molecule is a chemically sensible material. We screened 17,458 molecules from the PubChemQC database using PCESAScore—Scharber-model efficiency minus the RDKit synthetic-accessibility score—and examined what the metric actually returns. Seven molecules cross the viability threshold, but three are not organic semiconductors (molecular oxygen, a magnesium carbonate salt, a quinhydrone cocrystal) and three more carry reactive groups (two azides, one nitroso). A transparent chemicalvalidity filter and a reactive-group screen reduce the viable set from seven to four to a single stable candidate. The top-ranked molecule, an azidotriazine, reaches a 36.1% nominal efficiency—above the Scharber model’s single-junction ceiling—from a 0.91 eV gap, and carries a 9.26 D dipole and a negative computed reorganization energy: signatures of an unusable material, not a lead. Across the dataset, heteroatom composition correlates only weakly with orbital energies (|r| ≤ 0.34), so no simple design rule holds. Descriptor-based materials screens require an explicit chemical-validity and stability gate; the gap between the pre-and post-gate viable sets is a result in itself.
Related
- Charge-Transfer-State Descriptors from Donor–Acceptor Interfaces Enable Improved Prediction of Indoor Organic Photovoltaic Efficiency
- Microscale high-throughput screening-assisted discovery of a photo/copper dual catalytic system for benzylic C(sp2)–C(sp3) Suzuki–Miyaura coupling: data mining and mechanistic insights
- Prediction of MOF Adsorption Isotherms and High-Throughput Screening Using a Physics-Informed Membranes Graph Neural Network
- A Fluorescence‐Based High‐Throughput Screening Method for EGFR Inhibitors Using an EGFR‐Targeted Environment‐Sensitive Fluorescent Probe
- Research on the Intervention of Bile Stasis-Induced Pruritus Using Small Molecules Targeting MRGPRX4 Through High-Throughput Screening
BibTeX
@misc{kongo2026efficiency,
title = {{Efficiency and Accessibility Are Not Enough: Chemical-Validity Filtering in High-Throughput Screening of Organic Photovoltaic Materials}},
author = {Patrick Sorrel Mvoto Kongo and Steve Cabrel Teguia Kouam and Jean-Pierre TCHAPET NJAFA and Serge Guy Nana Engo},
howpublished = {ChemRxiv},
year = {2026},
month = sep,
doi = {10.26434/chemrxiv.15006402/v3},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15006402/v3},
note = {CC BY},
}
RIS
TY - UNPB
TI - Efficiency and Accessibility Are Not Enough: Chemical-Validity Filtering in High-Throughput Screening of Organic Photovoltaic Materials
AU - Patrick Sorrel Mvoto Kongo
AU - Steve Cabrel Teguia Kouam
AU - Jean-Pierre TCHAPET NJAFA
AU - Serge Guy Nana Engo
PY - 2026
DA - 2026/09/15
PB - ChemRxiv
DO - 10.26434/chemrxiv.15006402/v3
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15006402/v3
AB - Descriptor-based screening for organic photovoltaics ranks molecules on computed properties such as frontier-orbital energies and synthetic accessibility, but such descriptors do not test whether a highly ranked molecule is a chemically sensible material. We screened 17,458 molecules from the PubChemQC database using PCESAScore—Scharber-model efficiency minus the RDKit synthetic-accessibility score—and examined what the metric actually returns. Seven molecules cross the viability threshold, but three are not organic semiconductors (molecular oxygen, a magnesium carbonate salt, a quinhydrone cocrystal) and three more carry reactive groups (two azides, one nitroso). A transparent chemicalvalidity filter and a reactive-group screen reduce the viable set from seven to four to a single stable candidate. The top-ranked molecule, an azidotriazine, reaches a 36.1% nominal efficiency—above the Scharber model’s single-junction ceiling—from a 0.91 eV gap, and carries a 9.26 D dipole and a negative computed reorganization energy: signatures of an unusable material, not a lead. Across the dataset, heteroatom composition correlates only weakly with orbital energies (|r| ≤ 0.34), so no simple design rule holds. Descriptor-based materials screens require an explicit chemical-validity and stability gate; the gap between the pre-and post-gate viable sets is a result in itself.
ER -