Scientists use computational modeling to design ‘ultrastable’ MOFs
(Nanowerk News) Materials known as metal-organic frameworks (MOFs) have rigid, cage-like structures suitable for a wide range of applications, from gas storage to drug delivery. By changing the building blocks that go into the material, or the way they are assembled, researchers can design MOFs suitable for different uses.
However, not all possible MOF structures are stable enough to be used for applications such as reaction catalysis or gas storage. To help researchers figure out which MOF structures are best suited for a particular application, MIT researchers have developed a computational approach that lets them predict which structures will be most stable.
Using their computational model, the researchers have identified about 10,000 possible MOF structures that they classify as “highly stable”, making them good candidates for applications such as converting methane gas to methanol.
“When people come up with a hypothetical MOF material, they don’t need to know beforehand how stable it is,” said Heather Kulik, an MIT professor of chemistry and chemical engineering, and the study’s senior author. “We used our data and machine learning models to generate building blocks that expected high stability, and when we reassembled them in much more diverse ways, our data set was enriched with materials with higher stability than the previous set. from hypothetical materials that people have made.
MIT graduate student Aditya Nandy is the lead author of the paper, which appears in the journal Affairs (“Ultrastable MOF database reconstituted from stable fragments with a machine learning model”).
Scientists are interested in MOFs because they have a porous structure that makes them suitable for applications involving gases, such as gas storage, separating similar gases from one another, or converting one gas into another. Recently, scientists have also begun to explore its use for delivering drugs or imaging agents into the body.
The two main components of MOF are secondary building units — organic molecules that incorporate metal atoms such as zinc or copper — and organic molecules called linkers, which link the secondary building units together. These parts can be joined together in a variety of ways, much like LEGO building blocks, says Kulik.
“Because there are so many different types of LEGO blocks and ways you can assemble them, it creates a combinatorial explosion of possibilities for different organic metal framework materials,” he says. “You can really control the entire structure of the metal organic framework by picking and choosing how you assemble the various components.”
Currently, the most common way to design MOFs is through trial-and-error. Recently, researchers have started trying computational approaches to design these materials. Most such studies are based on predictions of how well a material will perform for a given application, but do not always take into account the stability of the resulting material.
“A MOF material that is really good for catalysis or for gas storage will have a very open structure, but once you have this open structure, it can be very difficult to ensure that it is also stable over long term use,” said Kulik.
In the 2021 study (JACK, “Using Machine Learning and Data Mining to Leverage Community Knowledge to Engineering Stable Metal-Organic Frameworks”), Kulik reports on a new model he devised by mining several thousand papers on MOFs to find data about the temperature at which certain MOFs will decompose and whether certain MOFs can withstand the conditions necessary to remove the solvents used to synthesize them. He trained a computer model to predict these two features — known as thermal stability and activation stability — based on molecular structure.
In the new study, Kulik and his students used the model to identify about 500 MOFs with very high stability. Then, they broke those MOFs down into the most common building blocks — 120 secondary and 16 connecting building units.
By recombining these building blocks using about 750 different architectural types, including many not normally included in such models, the researchers generated about 50,000 new MOF structures.
“One of the things that is unique about our set is that we see a lot more diverse crystal symmetries than has ever been seen before, but (we do so) use these building blocks that come only from very stable MOFs and are experimentally synthesized. , ”says Kulik.
The researchers then used their computational model to predict how stable each of these 50,000 structures was, and identified about 10,000 that they considered highly stable, both for thermal stability and activation stability.
They also screen structures for their “delivery capacity” — a measure of a material’s ability to store and release gases. For this analysis, the researchers used methane gas, because capturing methane can be useful for removing it from the atmosphere or converting it into methanol. They found that the 10,000 ultrastable materials they identified had a good delivery capacity for methane and were also mechanically stable, as measured by their predicted elastic modulus.
“Designing MOFs requires considering different types of stability, but our model allows predictions of thermal stability and activation at almost no cost,” says Nandy. “By also understanding the mechanical stability of these materials, we are providing a new way to identify promising materials.”
The researchers also identified certain building blocks that tend to result in more stable materials. One of the secondary building units with the best stability is a molecule containing gadolinium, a rare earth metal. The other is a porphyrin which contains cobalt — a large organic molecule made of four interconnected rings.
Students in Kulik’s lab are now working to synthesize some of these MOF structures and test them in the lab for their stability and potential catalytic and gas-separating abilities. Researchers have also made them ultrastable materials database available to researchers interested in testing it for their own scientific applications.