Inferring effective electrostatic interaction of charge‐stabilized colloids from scattering using deep learning

• Published in: Journal of applied crystallography 2024-08, Vol.57 (4), p.1047-1058
• Main Authors: Tung, Chi-Huan, Chen, Meng-Zhe, Chen, Hsin-Lung, Huang, Guan-Rong, Porcar, Lionel, Chang, Ming-Ching, Carrillo, Jan-Michael, Wang, Yangyang, Sumpter, Bobby G., Shinohara, Yuya, Do, Changwoo, Chen, Wei-Ren
• Format: Article
• Language: English
• Subjects: charge‐stabilized colloids; Colloids; Deep learning; Electrostatic properties; Molecular interactions; potential inversion; Scattering; small‐angle scattering; structure factors
• ISSN: 1600-5767; 0021-8898; 1600-5767
• DOI: 10.1107/S1600576724004515

An innovative strategy is presented that incorporates deep auto‐encoder networks into a least‐squares fitting framework to address the potential inversion problem in small‐angle scattering. To evaluate the performance of the proposed approach, a detailed case study focusing on charged colloidal suspensions was carried out. The results clearly indicate that a deep learning solution offers a reliable and quantitative method for studying molecular interactions. The approach surpasses existing deterministic approaches with respect to both numerical accuracy and computational efficiency. Overall, this work demonstrates the potential of deep learning techniques in tackling complex problems in soft‐matter structures and beyond. A deep learning approach has been developed for analyzing small‐angle scattering data, effectively addressing the potential inversion problem in colloids. The method is validated using both simulation results and experimental spectra of charged silica suspensions and is shown to outperform existing approaches in accuracy and efficiency. This study highlights the potential of deep learning in soft‐matter research.