Even the most effective water filters allow some contaminants to pass, but creating better materials and then evaluating them takes time and effort.
According to a new study published in ACS Central Science, artificial intelligence (AI) may hasten the creation of promising materials.
In a proof-of-concept experiment, they simulated several arrangements of water-attracting and water-repelling groups along a porous membrane of a filter and discovered the best configurations that should permit water to pass through freely while slowing some impurities.
Water for drinking and other applications is cleaned using filter systems, which can be anything from little faucet attachments to large industrial systems.
Current membrane filters, however, struggle to remove small, neutral compounds like boric acid, a popular insecticide used on crop plants, or highly unclean water.
This is due to the fact that synthetic porous materials are typically only capable of sorting chemicals according to their size or charge.
But because of the various sorts of functional groups, or groupings of atoms, that line the channels in biological membranes, pores constructed of proteins, such as aquaporin, can distinguish water from other molecules by both size and charge.
Steps towards a new AI filter technology
M. Scott Shell and colleagues planned to utilise computers to construct the interior of a carbon nanotube pore to filter water containing boric acid after being inspired to accomplish the same with a synthetic porous material.
The scientists created a model of a carbon nanotube channel with hydroxyl groups (which attract water) and/or methyl groups (which repel it) attached to each atom on the inner wall.
Then, to determine how rapidly water and boric acid would pass through the pore, scientists created and tested tens of thousands of functional group patterns using optimization techniques and machine learning, a form of artificial intelligence.
Here’s what they found:
One or two rows of hydroxyl groups sandwiched between methyl groups, forming rings around the middle of the pore, were shown to be the best designs.
Water traversed the pore in these simulations almost two times as quickly as boric acid. Another set of simulations demonstrated that the improved carbon nanotube designs may also be used to separate other neutral solutes like phenol, benzene, and isopropanol from water.
AI can not be dispensed with
According to the researchers, this work shows how AI may be used to create water purification membranes with brand-new features, which could serve as the foundation for a brand-new kind of filter system.
They continue by saying that the strategy may be modified to create surfaces with special interactions with water or other molecules, like coatings that fend off fouling.
The Center for Materials for Water and Energy Systems (M-WET), an Energy Frontier Research Center, and additional U.S. Department of Energy funding are acknowledged by the authors.
National Science Foundation Graduate Research Fellowship, Materials Research Science and Engineering Center, California NanoSystems Institute, and National Science Foundation.
The U.S. Congress granted a charter to the nonprofit American Chemical Society (ACS). The goal of ACS is to promote chemistry as a whole and its practitioners for the benefit of the planet and all of its inhabitants.
Through its several research options, peer-reviewed publications, scientific conferences, eBooks, and weekly news publication Chemical & Engineering News, the Society is a world leader in fostering excellence in science education and facilitating access to chemistry-related information and research.
Although ACS itself does not perform chemical research, ACS journals are among the most quoted, trusted, and read in the scientific literature.
As a pioneer in scientific information solutions, the CAS branch collaborates with international innovators to hasten scientific advances by curating, linking, and analysing content from around the globe. The principal locations of ACS are in Columbus, Ohio, and Washington, D.C.
Source:ACS
