Zeolites are among the most highly prized materials of the modern chemical economy and many manufacturers may have been making them without even knowing it. These microporous crystals are exceptionally versatile adsorbents that have found a wide range of uses, from isolating petroleum hydrocarbons and stabilizing radioactive waste to concentrating oxygen for medical applications, holding nitrogen for slow release in soil and even straining ammonia from the humble home aquarium.

Although some zeolites occur naturally, they can be synthesized inexpensively from very simple ingredients, including aluminum, sand and oxygen. These sophisticated materials take the form of elaborate metal-organic frameworks that are normally created from soluble metal salts of one sort or another. “These nitrates or chlorides can be fairly toxic, aggressive ingredients,” says McGill University chemistry professor Tomislav Friscic. He has been examining how these inputs could be removed through mechanochemistry, which would transform the solution-based process of building zeolitic frameworks into one that relies primarily on solid-state interaction of components during mixing. 

Such mixing is as straightforward as placing items in a sturdy closed vessel and shaking it vigorously. Ordinarily it is impossible to determine what is happening in this “black box” but with the aid of a heavy-duty transparent container and a synchrotron, Friscic and his colleagues were able to follow the interactions in real time. The results turned out to be fast and dramatic. “With the right catalytic supplements added to the reaction mixture you can get reactions to happen in under two minutes,” says Friscic. “It probably takes you less time to mix a couple of these solids in a milling jar than a conventional synthesis where you mix, filter or precipitate.”

Moreover, this approach yielded previously unknown versions of zeolitic imidazolate frameworks (ZIFs), which are topologically similar to zeolites. If the mechanical mixing process continues for just a few minutes more, however, these crystals break down into an amorphous powder similar to glass. When Friscic and his colleagues first made this observation, they were all set to consider the experiment concluded. Then, after a few minutes more mixing, the ZIFs formed all over again. “This was a complete surprise to us,” he says. “We made a crystalline material. We bashed it to death so it lost its crystalline structure. That sounds like the end of the story but no, if you keep on bashing the stuff it starts to crystallize again. You have an order-disorder-order transition under continuous milling.”

This material was dubbed katsenite, after Friscic’s post-doctoral student Athanassios Katsenis who subsequently became the lead author of the Nature Communications paper that resulted from this work. Although its structure is isomorphic to one of the most commonly employed commercial ZIFs, it cannot withstand the same rigours as an industrial reagent and will break down again. Nevertheless, Friscic regards this academic discovery as a valuable new way of examining mechanochemical processes that have become well established in industry. “Milling is being used all over the place on a very large scale,” he says, referring to the grinding of solids for mineral extraction, metal refining or pharmaceutical production. It is entirely possible, Friscic adds, that any of these processes could be yielding short-lived but chemically interesting and potentially valuable materials that have never been seen before. “If you are doing this in a black box, you can completely miss that there is an meta-stable intermediate crystal structure that temporarily exists,” he says. “It’s not just that the product might be interesting or new but there may be other things forming in the milling process that have only a transient existence. You can get structures that would otherwise be completely inaccessible.”