Uncertainty-Aware AI-Enabled Synthesis of Transparent Conducting Materials

Interdisciplinary Areas: Data and Engineering Applications, Autonomous and Connected Systems, Innovation and Making, Future Manufacturing, Micro-, Nano-, and Quantum Engineering, Others

Project Description

Transparent conductors such as Indium Tin Oxide (ITO) are integral to a wide array of modern technologies, including solar cells, touchscreens, smart windows, and optoelectronic devices. While AI is undoubtedly a promising and emerging approach for the discovery of new transparent conductors, its integration into the design process presents several challenges, hindering its wide adaptation. One of the primary challenges is the limitation of reliable data, common in the scientific domain. The scarcity and unreliability of data can lead to models that may overfit the polymer training data, thereby producing materials that may lack uniqueness, stability, and satisfactory levels of conductivity and transparency. A second challenge, mainly due to the complexity of the polymer space, is that the AI's predictions about the properties of new polymers are inevitably uncertain. Thus, one needs to generate a large number of candidate polymers that require manual synthesis and testing by human scientists, which is prohibitively time-consuming and resource-intensive.

The objective of this project is to establish an algorithmic foundation toward developing a multi-modal model with certified uncertainty quantification, to address these challenges and accelerate the automated discovery and synthesis of transparent conductors.

The postdoctoral researcher will perform theoretical, numerical, and experimental work at the interface of chemical and material sciences and machine learning. The Postdoc will have access to a synthetic laboratory, a processing/fabrication room, a measurement room, and an innovation room, which is available to the Polymer Innovation for Advanced Organic Electronics (PI-AOE) group. Additionally, new lab spaces for automated discovery will be available.

Start Date

Flexible in Feb 2025 - Aug 2025

Postdoc Qualifications

PhD in Electrical Engineering, Chemical Sciences and Engineering, Material Sciences and Engineering, Computer Science and Engineering, or closely related fields.
Experience in computational and experimental chemistry, Artificial Intelligence, Machine Learning.

Co-advisors

Jianguo Mei
Professor of Chemistry
jgmei@purdue.edu
https://www.jianguomei.com/

Abolfazl Hashemi
Assistant Professor of Electrical and Computer Engineering (ECE)
abolfazl@purdue.edu
https://abolfazlh.github.io/

Eugenio Culurciello
Interim Director, Purdue Institute for Physical Artificial Intelligence
Professor of Biomedical Engineering
Professor of Psychological Sciences, Health and Human Sciences, Mechanical Engineering and Electrical & Computer Engineering (courtesy)
euge@purdue.edu
https://culurciello.github.io/

Bibliography

X. Lei, W. Ye, Z. Yang, D. Schweigert, H.-K. Kwon, and A. Khajeh, “A self-improvable polymer discovery framework based on conditional generative model,” arXiv preprint arXiv:2312.04013, 2023.

. Yang, W. Ye, X. Lei, D. Schweigert, H.-K. Kwon, and A. Khajeh, “De novo design of polymer electrolytes with high conductivity using gpt-based and diffusion-based generative models,” arXiv preprint arXiv:2312.06470, 2023.

C. Zeni, R. Pinsler, D. Z¨ugner, A. Fowler, M. Horton, X. Fu, S. Shysheya, J. Crabb´e, L. Sun, J. Smith, B. Nguyen, H. Schulz, S. Lewis, C.-W. Huang, Z. Lu, Y. Zhou, H. Yang, H. Hao, J. Li, R. Tomioka, and T. Xie, “Mattergen: a generative model for inorganic materials design,” 2024.

. Xia, K. Wu, P. Deng, R. Liu, Y. Zhang, H. Guo, Y. Cui, Q. Pei, L. Wu, S. Xie, S. Chen, X. Lu,
S. Hu, J. Wu, C.-K. Chan, S. Chen, L. Zhou, N. Yu, H. Liu, J. Guo, T. Qin, and T.-Y. Liu, “Target-aware molecule generation for drug design using a chemical language model,” bioRxiv, 2024.

R. Sarabia-Riquelme, L. E. Noble, P. Alarcon Espejo, Z. Ke, K. R. Graham, J. Mei, A. F. Paterson, and M. C. Weisenberger, “Highly conductive n-type polymer fibers from the wet-spinning of n-doped pbdf and their application in thermoelectric textiles,” Advanced Functional Materials, p. 2311379.