KI meets Chemistry – Taking The Air Out Of
Greenhouse Gases
The world is searching for solutions to climate change – and one is currently being developed in an inconspicuous laboratory at LANXESS in Leverkusen. Here, a small team has developed a new product with promising potential: a new Lewatit® resin that can extract CO2 directly from the outside air. How did this come about? By bringing together the research team led by chemist Julian Krischel, BU LPT, with data science colleague Hanna Kahlfeld, GF IT.
Visually, when viewed individually, the small red, orange, black or white beads of the Lewatit® brand may not look particularly spectacular. But together, they are true powerhouses and impress with their versatility. The beads from the LPT business unit remove pollutants from water, help recycle batteries, and can even extract CO₂ from the air.
Until now, however, there has been one drawback: air filtration was limited to enclosed spaces such as greenhouses, submarines and the ISS space station. Impressive as that already is, the developers at LPT saw far greater potential in their already well-established Lewatit® VP OC 1065. “We wanted to filter CO₂ from outdoor air as well,” says Julian Krischel, Head of Product Innovation and project lead for Direct Air Capture at BU LPT. Only if that succeeded, he explains, could LANXESS’s beads make a significant contribution to combating climate change. The ambition was — and remains — high, and so is the pressure to succeed.
There are currently around 50 direct air capture systems in operation worldwide that permanently remove the greenhouse gas from the air. Last year, there were 20 fewer. One of the largest facilities is located in Iceland, where it extracts 4,000 tons of CO₂ from the air each year. A significantly larger plant is now being built there, designed to neutralize 36,000 tons of CO₂ annually. By 2030, the company plans to remove several million tons of CO₂ from the air each year — and by 2050, even several billion.
“The search for different technical solutions for CO₂ adsorption is intense,” says Krischel. It is a growing market. But for our task, Lewatit® had to become even more powerful. To make that happen, the LPT innovation team brought AI expert Hanna Kahlfeld on board.
Hanna Kahlfeld explains the role of AI in tackling challenges like this: “AI can help make product development faster and even more innovative.” The AI is trained with data in such a way that it can specifically generate formulation suggestions with the highest expected performance improvement. These formulations are then tested in the laboratory. Based on the test results, the AI can be “fed” with even better data — and continue to learn.
In the formulation currently used for the ion exchangers, there were around 30 variables and seven target parameters. Testing all of these variables in their many possible combinations in the laboratory to develop a new formulation would have taken decades and incurred immense costs. In this specific case, another challenge was added: only a small amount of experimental data was available. “This is exactly where our approach shows its strengths,” says Kahlfeld. The AI model is not only able to make predictions, but also to assess their uncertainty. Especially with small data sets, this is crucial in order to still arrive at better formulations. In this way, we find a clear path through the complex parameter space and significantly reduce the number of experiments required,” Kahlfeld summarize
Chemist Julian Krischel knew exactly what the new formulation needed to achieve. The beads were required to absorb large amounts of CO₂, but at the same time must not become too porous. Adsorption capacity, stability and swelling behavior therefore had to be carefully balanced in the right proportion.
The AI delivered formulation proposals, which Krischel and his team tested in the laboratory. Kahlfeld then fed the lab results back into the AI so that it could subsequently generate even better formulation suggestions. As early as the second iteration, the AI proposed two novel formulations which, according to the model’s predictions, were expected to perform very well in various areas. Krischel recognized even greater potential here and combined the two formulations by hand. His new formulation proved to be a complete success.
“After just two experimental iterations, we had found a formulation that met all of our expectations,” says Krischel. This is a prime example of how the key to success lies in the combination of AI-driven thinking and domain expertise.
The results speak for themselves: the new ion exchangers can absorb twice as much CO₂ as the materials used previously. Their kinetic properties — the “speed of uptake” — have also improved. This has a direct impact on their overall performance, placing them among the best in their class.
The team needed only ten months for this entire process. The first major step has thus been completed. Now the project moves on to production. “It is always an exciting moment when a development makes the transition from the laboratory to the plant. We are glad that we can rely on the expertise and experience of our colleagues in operations during this phase,” says Krischel.
The new ion exchanger is scheduled to be available from 2026. By then, the small beads may well live up to the high expectations and, in direct air capture systems, play their part in tackling the major challenge of global warming (see also the interview on the right).
Background information
The greenhouse gas CO₂ is distributed relatively evenly in our atmosphere. To operate direct air capture systems as efficiently as possible, they are therefore installed in locations where energy can be generated particularly cost-effectively. For this reason, the world’s largest facility is located in Iceland, near an active volcano. This allows the energy from the local geothermal power plant to be used to optimum effect. The plant draws in air and removes the CO₂ from it using chemical filters. Regenerable chemical filters such as those based on Lewatit® VP OC 1065 are ideally suited for this purpose. The use of this class of chemical adsorbers enables a continuous and sustainable process, while keeping environmental impact and the costs of regular material replacement to a minimum. The collected greenhouse gas is then stored underground, where it is naturally converted and thus permanently sequestered.
Interview
“We want to make a difference!”
In an Xpress interview, Julian Krischel, head of research for direct air capture, talks about the final step: bringing the new formula from the laboratory into operation. Only if the new Lewatit® for CO2 filtration is also scalable will a new, large future market open up for LANXESS.
Darting into the lab between two appointments with the plant managers, discussing the latest customer feedback or checking current measurements – for Julian Krischel, Head of Innovative Product Development in the LPT business unit, this has been part of his daily routine for around a year. He and his small team are working flat out to help the newly developed Lewatit® achieve a breakthrough: the goal is for it to extract CO2 from the outside air – permanently, efficiently and scalably. His office in the historic Q18 building at Chempark Leverkusen and the laboratory space available to his team are only a few meters apart. “That way, I can stay close to the action—even when my calendar is full,” he says.
He considers it a great privilege that the three researchers can focus entirely on a single topic—and a great responsibility to use the resources provided wisely. “Of course, we feel an obligation to deliver results as quickly as possible,” he says. “Fortunately, we have been very successful so far – thanks to the great support we receive from marketing, operations, our AI experts, and also from BU management.” New samples of Lewatit® Aeropure have just arrived from the factory – always a moment of excitement and anticipation for the team.
Mr. Krischel, how is the scaling of the new Lewatit® going in production?
Julian Krischel: The first operational trials are currently taking place. Of course, we always had the conditions in production in mind during development, but the so-called “upscaling” brings with it its own unique challenges. Instead of the three-liter kettles we have been using here in the laboratory, the process now has to be transferred to production facilities with a capacity of more than ten cubic meters. The initial results are promising, so we are confident that everything is going according to plan.
The new technology is still in its infancy. How big is its market potential?
Enormous. Political conditions, such as CO2 certificate trading, are driving demand for new climate technologies. We see a future market—especially in direct air capture plants—that is just emerging. And we want to be involved in shaping it from the very beginning. Our more than 80 years of experience with ion exchangers is, of course, a real advantage—we already have a wealth of expertise from more than 170 products that we can draw on. Our products are considered the benchmark in many markets – that opens doors.
How important is cooperation with our customers?
It is central – and the starting point for every new development. This is because every plant works a little differently from a technical standpoint or has its own specific requirements due to its location and the associated external conditions such as temperature or humidity. Capacity, durability, speed—we work with our customers to define all parameters down to the smallest detail and then have our product extensively tested on the customer's side. Our goal is to develop a flexible base material that can be adapted to different application scenarios with minimal effort.
When you delve so deeply into a topic for so long, what drives you personally?
The opportunity to make a real difference with research. Our technology makes it possible to permanently remove CO2 from the air – a major lever for a better climate. The product itself is also sustainable: although it has to be regenerated in appropriate cycles, our ion exchangers remain effective over a very long period of time. Clean air and clean water cannot be taken for granted – they are essential to our quality of life. Contributing to this is simply a great feeling.
If you are interested in AI support for your R&D project, please feel free to contact Hanna Kahlfeld or the “IT Analytics & AI – Procurement & R&D Analytics” team!