December 6, 2023

New Polymer Membranes and AI Predictions Aim to Reduce Energy and Water Use in Oil Refining

Researchers at Georgia Tech have developed a new type of polymer membrane that could revolutionize the way crude oil is processed, significantly reducing the energy and water required while extracting more valuable materials. Known as DUCKY polymers, these membranes have the potential to transform multiple industries, from biofuels and biodegradable plastics to pulp and paper products. The team at Georgia Tech has also developed artificial intelligence (AI) tools that can predict the performance of these polymer membranes, expediting the development of new ones.

The oil refining process involves the separation of various components in crude oil to produce fuels, plastics, textiles, and other materials. This initial separation accounts for approximately 1% of global energy consumption. Traditionally, distillation has been used for this purpose, which is a water- and energy-intensive process. The researchers at Georgia Tech are aiming to replace this process with membrane separation technology, which uses electricity instead of heat or combustion, making it more environmentally friendly.

The key innovation behind the team’s polymer membranes is a new family of polymers created from spirocyclic monomers. These monomers assemble in chains with 90-degree turns, forming a material with unique properties that selectively bind and allow desirable molecules to pass through. The DUCKY polymers are created using click chemistry, a Nobel Prize-winning reaction called CuAAC, giving them their peculiar name.

While each characteristic of the polymer membranes is not novel on its own, their combination makes them highly effective. Researchers from ExxonMobil collaborated with the team and confirmed the membrane’s effectiveness by extracting more valuable materials from the leftover sludge after the distillation process.

The development of AI models is another breakthrough in this research. By using machine learning algorithms and mass transport simulations, the team can predict the performance of polymer membranes in complex separations. This allows for the design of new polymers with specific permeation performance, potentially accelerating the material design process, which is currently a trial-and-error approach.

Training the AI algorithms involved collecting and analyzing a vast amount of data on solvent diffusion through polymers, resulting in a comprehensive dataset. The AI models’ predictions were validated with experiments, including tests on the DUCKY polymer membranes, which showed a close correlation between predicted and actual results.

The combination of new polymer membranes and AI predictions has the potential to transform the oil refining industry, reducing energy, carbon, and water footprints while enabling the production of new and valuable materials. The research conducted at Georgia Tech is a significant step towards more sustainable and efficient industrial processes.

1. Source: Coherent Market Insights, Public sources, Desk research
2. We have leveraged AI tools to mine information and compile it