Computational chemistry can solve some global climate challenges

Environment and sustainability 28. dec 2023 3 min Professor Jan Halborg Jensen, Postdoc Maria Harris Rasmussen Written by Kristian Sjøgren

Researchers have developed an algorithm that can identify good candidates for catalysing many chemical reactions. For example, the researchers see an opportunity to find catalysts that can make manufacturing fertiliser more climate-friendly or even finding catalysts to more easily capture CO2 from the atmosphere.

For centuries, chemists have hunched over flasks and test tubes in search of catalysts to drive chemical processes.

The future will probably not look like this now that researchers have developed an algorithm that can manage the whole process on a computer.

The researchers can use the algorithm to suggest molecules that could drive a specific chemical process most optimally.

Then the researchers can test these suggestions in the laboratory, saving years of work with flasks and reagents.

According to the researchers behind the development of the algorithm, this has the potential to be enormously important not just for chemists but for the whole world.

“For example, much of the world’s ammonium fertiliser is produced by using the Haber–Bosch process, which requires substantial energy. If our algorithm can discover improved catalysts to drive the Haber–Bosch process, this can potentially reduce total global energy consumption by 2% just from manufacturing ammonium more efficiently,” explains Jan Halborg Jensen, Professor, Department of Chemistry, University of Copenhagen.

Jan Halborg Jensen and Maria Harris Rasmussen, Postdoctoral Fellow from the same department have been developing the new algorithm for several years and have published the results of their experiments in Angewandte Chemie here and here.

Seven years in the making

A catalyst is a molecule that can help drive a chemical process more favourably.

For example, a molecule could cause a chemical reaction at lower pressure or at a lower temperature – thus more rapidly and requiring less energy.

Discovering new catalysts normally requires many years of unsuccessful laboratory experiments.

In 2016, Jan Halborg Jensen therefore thought a computer algorithm might be able to determine the optimal molecules to drive a specific chemical process.

After seven years of developing the algorithm, the researchers published the first of two articles showing that it could identify a molecule for driving the chemical processes called Morita–Baylis–Hillman reactions.

In the second article, the researchers showed that the algorithm can be used widely to discover good catalyst candidates in general.

“We can find good organic and chemical catalyst candidates but cannot yet identify good catalysts with metal compounds, but we are working on this,” says Maria Harris Rasmussen.

Evolutionary process

The researchers ask the algorithm to discover a good catalyst to drive a specific chemical process.

The researchers then provide possible chemical building blocks from which to build the catalyst.

Discovering the optimal catalyst involves the algorithm developing in an evolutionary way.

It tests random molecules and then finds the molecules that will theoretically drive the chemical process optimally. It then pairs these molecules, and based on the pairing, it again selects the best catalyst candidates.

The algorithm continues in this evolutionary way until it has found the exact molecule that will be the optimal candidate.

Then the candidate can be tested in the laboratory to determine whether the algorithm has arrived at the correct calculation.

“In our first experiment, we wanted to find a molecule to drive Morita–Baylis–Hillman reactions, and the algorithm found a catalyst that did not look at all like the molecule that is considered the best catalyst today. We then examined our catalyst and found that it works a little bit better, showing that the algorithm worked,” explains Jan Halborg Jensen.

May reduce carbon emissions

Jan Halborg Jensen and Maria Harris Rasmussen see great potential in the algorithm.

This applies within industry, where better catalysts can produce commercially useful compounds less expensively. This applies especially to chemical processes for which there are currently no good catalysts.

The algorithm can also help to discover improved catalysts that can solve some of the great global challenges.

As previously mentioned, manufacturing fertiliser using the Haber–Bosch process requires substantial energy, but better catalysts can reduce this energy consumption by reducing the need to run the chemical process under pressure and at high temperatures.

Another example is CO2 capture and storage.

The researchers see an opportunity to identify catalysts that can drive a chemical process to capture CO2 from the atmosphere and turn it into solid matter.

An important parameter for whether this can be achieved, however, is not related to chemistry but rather to money.

The algorithm may well recommend a specific molecule for driving a chemical process, but if the molecule itself is extremely expensive to produce, it does not make much sense either in industry or as a solution to global climate challenges.

In addition, the algorithm needs to be disseminated and used widely.

“It needs to be adopted globally, and we will therefore aim to show that it works so that others can also have confidence in it. The algorithm needs to be tested on many chemical processes, and then researchers must investigate whether the catalysts that the algorithm has discovered actually perform well,” concludes Maria Harris Rasmussen.

The main research interest of the group is the development of new computational methods and applying them to important chemical problems. The group...

The main research interest of the group is the development of new computational methods and applying them to important chemical problems. The group...

© All rights reserved, Sciencenews 2020