“Decline in Discovery of New Reaction Types Despite Exponential Growth in Reported Reactions, Study Finds”

Despite the exponential growth in the number of reactions reported by chemists over the past years, the discovery of new reaction types is dwindling, according to scientists. An analysis of millions of organic transformations deposited in databases and patent repositories, conducted using cheminformatics, has concluded that most synthetic chemistry relies on reusing well-known methods.

Bartosz Grzybowski, the study leader from the Polish Academy of Sciences and Ulsan National Institute of Science and Technology, South Korea, stated, “A couple of million reactions are reported every year, but many of them are applications of the same known methodologies. We’ve actually been discovering reactions – at most – at a linear rate, maybe a couple of thousand a year.”

To classify reaction types, the researchers defined rules, considering mechanistically relevant atoms located near the reaction centre, and scripted them into an algorithm. Grzybowski explained, “We then went through millions of reactions and categorised them. We did this at different points of time [starting with data from 1900] to see how the number of new reactions changed.”

The research team also found that while the discovery of reaction types has slowed down, processes are becoming more complex, with fewer steps being required to produce challenging scaffolds. Grzybowski noted that cascade transformations or multicomponent reactions are gaining popularity.

Guillaume Berionni, a physical organic chemist from the University of Namur, Belgium, commented, “This tells us that all simple organic and organometallic reactions between two reagents have already been discovered. Organic synthetic chemists now have to combine their research efforts. We have to team up with computational chemists and spectroscopists, and use new approaches such as automated chemistry and organic syntheses driven by artificial intelligence.”

Grzybowski pointed out that there was a crossover point about two decades ago, where the complex reactions overtook the simple ones. Simple transformations like SN2 reactions are as good as exhausted. “In over a hundred years, I think people have pretty much tried every possible combination,” he said.

However, Jean-Louis Reymond, who develops computational methods for organic synthesis at the University of Bern, Switzerland, believes that the field is not reaching saturation. He said, “The reactions might stay the same, but nowadays a lot of effort is put into making the reagents and reaction conditions more sustainable. Let’s consider the recent Nobel prize for organocatalysis – it’s not only about new reactions but also about simpler and better ways to carry them out.”

Maria Fernández-Herrera, a synthetic chemist from Cinvestav Mérida in Mexico, agrees that organic synthesis is moving in a direction where discovering new materials or greener processes for applications in different fields is gaining in importance. She cited the synthesis of metal–organic frameworks and other related structures as a clear example.

Grzybowski’s team found that only a few hundred to a thousand of the reaction types ever became popular. He said, “This means that you can only make predictive machine-learning models for a fraction of chemistry. There is simply too little data available for most reaction classes to train an algorithm on.”

David Nagib, an organic chemist at Ohio State University, US, noted that although simplification is great for teaching purposes or to train algorithms, scientists can’t ignore reactions’ subtleties. He said, “For example, the Diels–Alder [reaction] can couple an electron-rich diene with a deficient alkene or the exact opposite pairing – and can occur by concerted, stepwise, polar or radical mechanisms – catalysed by heat, light, acid, enzymes, metals and even organocatalysts. Merging these all together as a single reaction type risks ignoring the value and utility of each of these chemistries.”

Nagib added that many reaction classes have made huge leaps since they were first discovered. Carbon–hydrogen bond functionalisation has grown from a mechanistic oddity to revolutionising drug discovery, and click reactions can now be performed in living animals, he said.

As the field of organic synthesis continues to evolve, so do the reactions that remain on chemists’ wish list. Grzybowski said, “I, personally, am a big believer in multicomponent complex rearrangements. One step and you can build complexity like crazy. That’s my favourite!”