Collaborative Research: DMREF: Data-Driven Design of Hybrid Membrane Materials for Organic Liquid Separations

About This Grant

Chemicals that are free of impurities are critical to everyday products such as electronics, medicine, and food. However, separating a chemical mixture into its pure constituents is energy intensive and expensive. This project will develop new membrane materials that can separate chemical mixtures at lower cost and use less energy. The research team will combine advanced data science with lab experiments to speed up materials discovery. The project will focus on separating a liquid mixture of small molecules called paraffins and olefins. This specific separation is especially important to the chemical industry because these molecules are used to make fuels and plastics. The results of this project will be new membrane materials, and better computer programs for finding these materials. Additional benefits to society will come from training science and engineering students in data science, undergraduate research and training, and public outreach at science festivals. This project combines researchers with expertise in polymer synthesis, materials science, chemical engineering, and data science. The goal is to discover new organic-inorganic (hybrid) membrane materials that can separate organic liquid mixtures. The research team will combine high-throughput physical experimentation with machine learning (ML) models to create new data-driven frameworks for membrane material discovery and optimization. This project will focus on using combinatorial chemistry to create structurally-tunable microporous polymers. These polymers will be combined with newly-developed inorganic vapor infiltration techniques to create a wide range of organic-inorganic hybrid membranes. These hybrid membranes will be designed for chemical stability and selectivity to achieve difficult organic liquid separations, including the separation of olefin and paraffin mixtures. Data-informed ML models will be developed to establish feasibility of the polymer synthesis, chemical stability, and permeation selectivity. The corresponding data-driven workflow will identify promising materials that can separate a given liquid mixture, and are also easy to manufacture. The most promising membrane material candidates will be tested to validate predictions. The experimental results will be fed back to improve simulation predictions. The project will also support a multi-disciplinary undergraduate research program that will train students in lab automation. Public outreach includes a demonstration module that visibly separates colored dyes using membranes. This will create awareness of how these “hidden” manufacturing processes are important to human well-being and economic security. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.