Bandgap tuned indaceno based polymer design by machine learning for chemical space by descriptor designing and data guidelines
[ X ]
Tarih
2025
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Elsevier Ltd
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
This research introduces an innovative approach for developing bandgap-tuned polymers by machine learning assisted modeling of indaceno polymer chemical space. Using a dataset of indaceno donor moieties, the descriptors that encode fundamental molecular features are designed and trained to predict their bandgaps. After three modeling rounds of breaking retrosynthesis in Python, 1000 new polymers with their bandgaps are designed. The study shows that descriptors like valence electrons, Labute ASA, and Morgan Density has a significant impact on model performance to highlight their key role. Additionally, the synthetic accessibility scores of newly designed polymers reach a maximum of 17 for RDkit based descriptors to indicate their ease of synthesis and promising practical applicability. This work not only deepens the understanding of indaceno polymers but also lays the groundwork indaceno based polymer design of through data-driven approaches.
Açıklama
Article number : 109899
Anahtar Kelimeler
Bandgap, Indaceno, Machine learning, Polymer design, Reverse engineering
Kaynak
Materials Science in Semiconductor Processing
WoS Q Değeri
Q2
Scopus Q Değeri
Q1
Cilt
200
Sayı
Künye
Sumrra, S. H., Güleryüz, C., Hassan, A. U., Afzaal, M., Siddiqa, A., El Azab, I. H., ... & Mahmoud, M. H. (2025). Bandgap tuned indaceno based polymer design by machine learning for chemical space by descriptor designing and data guidelines. Materials Science in Semiconductor Processing, 200, 109899. 10.1016/j.mssp.2025.109899
