Stare at the past to catch a glimpse of the future

In our Perspective article "History and Utility of Zeolite Framework-Type Discovery from a Data-Science Perspective" published in Crystal Growth & Design, Maciej Haranczyk and I used a data-driven approach to find past trends in zeolite discoveries with a focus on utility of the materials. The insights were then used to assess current developments in the field.

Zeolites are nanoporous, crystalline solids, which occur in nature as minerals. But they can be also synthesized in the lab. Zeolites are so important because they are widely used in industry—mainly as acid-based catalyst for cracking purposes. The most striking structural feature of all zeolites is that they consist of tetrahedra where typically a silicon or an aluminum atom sits in the center. That atom is then tetrahedrally coordinated by four oxygen atoms. We call these tetrahedra primary building blocks.

In order to study trends in the discovery of zeolites we computed three metrics, two of which we are showing in the video. The first one is the annual discovery rate: how many new zeolite frameworks have been discovered per year. The second one is the average degree of tetrahedrality of the primary building blocks. That gives you an idea whether the average bonding pattern in the structure is almost perfect (1) or distorted and, in that case, to what extent (any number smaller than 1).

Based on the rolling 5-year average of the discovery rate, we can roughly divide the discovery trajectory into four phases. The first phase was dominated by World War II because new discoveries were entirely slashed until 1950.

The second phase took place between 1950 and the beginning of 1980s, at the end of which roughly 2 frameworks were discovered per year. During this time, the industrially most important zeolites were discovered. For example, FAU in 1958. One of its related materials, zeolite Y, is the most consumed zeolite catalyst. Another important example is MFI, which is the underlying framework of ZSM-5. ZSM-5 is outstanding because it is (probably) the zeolite catalyst implemented in the largest number of different processes. Both frameworks participate thus in mature, long-existing technologies, and their tetrahedrality is close to the average from all frameworks (green dashed line).

The third discovery phase, which we place between 1984 and 1994, is a good example of what Payra and Dutta noticed in 2003: "The evolution of materials development in the zeolite field over the last 50 years has followed a path of steady progress, along with steady leaps that introduce new paradigms of synthesis." The big leap in this phase was that more and more unconventional compositions were achievable; for example, silicoaluminophosphates. The discovery rate plateaued at the end of phase 3 at a little above 4 frameworks per year. Also at the end of phase 3, CZP was discovered, a chiral zeolite that was hoped to be exploitable for enantiomer separation.

We identified the fourth phase with an average annual discovery rate of 6 frameworks to last to date. An exceptionally intriguing discovery in this phase was RWY because the exotic sulfide and selenide composition led to semiconducting properties. This was hoped to be exploitable in photocatalysis for example. However, the structure is very distorted, which typically greatly reduces thermal stability under process-like conditions. In fact, this is a major problem with this framework type that has prevented deployability of the technology. The effect that such promising technologies are not deployable because of specific, insurmountable technology transfer barriers is known as the so-called valley of death.

Finally, the current state of zeolite discoveries is unprecedented because annual discovery rates and the five-year rolling average hardly differ for eight years in a row now. Also, we are currently seeing a second wave of exotic zeolites. But those are again very distored structures. Because such distorted structures have already in the past not lead to materials truly deployable for commercial or industrial purposes we caution high hopes in the currently on-going zeolite revolution.

The movie was mainly created with GULP for the relaxation of zeolite frameworks that were retrieved from IZA's database of zeolite structures and from the 2015 Nature Chemistry article by Mazur et al., gnuplot was used for the plots, VTK for snapshots/3D object rendering, ImageMagick's convert script for command line-based picture processing, QuickTime for audio recording, and ffmpeg for picture concatenation and final audio-video merging; some introductory slides were generated with PowerPoint 2011. Our research was supported by U. S. Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences, Geosciences and Biosciences, under Award DE-FG02-12ER16362 during this project; Lawrence Berkeley National Laboratory is funded by the U. S. Department of Energy under Award DE-AC02-05CH11231.