Nanoporous materials are emerging attractive to serve some biggest challenges of today’s society. This is because nanoporous materials are serving from absorption of carbon dioxide or methane from air to sensing toxic compounds in air and store hydrogen gas for fuel.
In fact, with the tiny, nanoscaled-sized pores of nanoporous materials, they are useful for many sustainable applications. On the contrary, nanoporous materials are fabricated by chemists in labs by adding molecule by molecule which makes them cumbersome and expensive to develop.
In a new development, a research team at the Oregon State University and Washington State have developed a unique computer algorithms that plays a game of 20 questions to quickly narrow down thousands of possible molecular designs to find the optimal one with minimal effort and minimal cost.
Meanwhile, the mixture of different chemical elements in nanoporous materials remains a challenge that needs to be composed and to figure out the best combination.
Interestingly, nanoporous materials have a huge potential of various types of building blocks and arrangements that can be mixed endlessly. If new configurations of elements and structures of nanoporous materials are to be tried in a laboratory it would be very expensive. Therefore, the computational challenge remains to figure out the right combination of elements that have the desired properties. This is where AI-based algorithm comes into picture.
For the POC of the study, the best candidate of nanoporous materials that can absorb methane narrowed down. Researchers evaluated 120 possible nanoporous materials and found the best known candidates from a library of 7,000 materials which are considerably better than traditional algorithms have performed.
Importantly, the algorithm developed by the researchers are able to find the best material with fewer number of evaluations.
