Symposium MT03: Machine Learning Methods, Data and Automation for Sustainable Electronics

Nathaniel Park, IBM Research

Using Domain Specific Language to Enable Artificial Intelligence for Polymer and Catalyst Design

Written by Kazi Zihan Hossain

Artificial intelligence (AI) is reshaping different domains, including materials science and polymer chemistry. However, new polymer development is being obstructed due to a lack of well-developed open-source data. Additional efforts are often required to represent the existing experimental data with available models properly. To overcome the issue, Nathaniel Park and colleagues from IBM research have developed and presented a domain-specific language, Chemical Markdown Language (CMDL), specifically for polymers. CMDL allows researchers to use simple syntax to document complex polymeric structures and experimental results through the IBM Materials Notebook. They demonstrated the capability of CMDL by creating new polymers and catalysts through generative AI and validating them through experimentation. The research team is also improving the CMDL to integrate more features to allow seamless integration of AI in materials development.