Scientific Achievement
Computer simulations reveal that solute diffusion strongly correlates with simple geometric measures of porous triblock copolymer membranes.
Image: Triblock copolymer and computational scheme to study relationship between morphology and transport.
Significance and Impact
This machine-learned relationship will be used to design polymers for fabricating water-treatment membranes with targeted properties.
Image: Richness of diffusivities D in quenched morphologies of various polymer architectures.
Research Details
- The performance of porous membranes depends strongly on their (mesoscale) pore structure, but this relationship is insufficiently understood to do effective engineering.
- Many membrane structures were simulated, and solute diffusion was measured to generate a large data set. A relationship between the two was discovered using image processing and data science approaches.
- The data-driven model accurately predicts diffusivity from morphology and can help find optimal pore structures.
M.P. Howard et al. Macromolecules 53, 2336 (2020).
Work was performed at University of Texas at Austin and University of California, Santa Barbara.