Predicting the mechanical properties of metal-organic frameworks using Machine Learning

A new study suggests that machine learning can predict the mechanical properties of metal-organic frameworks, materials that will to be of great importance in the near future.

Metallic-Organic Frameworks or MOFs are complex molecular structures created by the connection of metallic and organic atoms. They have crystalline structure, developing in every direction in space, which is highly porous making them ideal to function as a storage. An impressive fact is that a sugar-cube-sized MOF has a massive surface area about the size of 6 football fields.

MOFs can be utilized for many purposes, including storing toxic gases or extracting the water from the air. "That MOFs are so porous makes them highly adaptable for all kinds of different applications, but at the same time their porous nature makes them highly fragile," Dr. David Fairen-Jimenez from Cambridge's Department of Chemical Engineering and Biotechnology, leader the research and co-author of the study, stated.

They can be customized into numerous combinations and, thus, it's not possible to derive the mechanical properties of each one. However, scientists developed a machine learning algorithm in order to determine the properties of more than 3,000 existing MOFs. 

MOFs are produced in powder form. They are unstable materials that may collapse during the manufacturing process due to their porosity. This causes severe financial loss and time waste. Therefore, predicting their mechanical properties ensures that scientists will produce the most durable of them. "We are now able to explain the landscape for all the materials at the same time. This way, we can predict what the best material would be for a given task." Dr. Fairen-Jimenez said.

The team has developed a website where other researchers can evaluate the properties of their MOFs using the produced algorithm. Their aim is to reduce uncertainty in the MOF field. "It allows researchers to access the tools they need in order to work with these materials: it simplifies the questions they need to ask," Dr. David Fairen-Jimenez, added.