You’re working on two things that don’t seem quite compatible to most people: advanced quantum simulations and industrial processes like catalysts. We thought quantum physics was something abstract and played no role in our lives, and now it appears everywhere.
“In fact, quantum mechanics is already all around us. People don’t think about it, but it’s true. Every chemical bond is a proper interference of the wave functions of different electrons. This interaction can only be described by quantum mechanics.”
“In basic chemistry courses, many of us saw orbitals. Those were beautiful pictures, but this shape actually follows the solution of the Schrödinger equation to the many-body problem in quantum mechanics. I do this in my courses with students in relation to the hydrogen atom. This is the only atom we can solve Strictly, because it only has one electron. I think this is a very beautiful experiment. These students are in their second or third year of bachelor’s and when we solve that Schrödinger equation and suddenly they see those orbitals appearing mathematically, they find it very amazing.
“The fact that we personally experience quantum mechanics every day is also what interests me. A physicist by training, I was very interested in theoretical topics. But in the course ‘Quantum Theory of Chemical Bonding’ I saw that I could do something with it.”
“In our research we are trying to understand why chemical bonds are formed. Everything there has to be described in a quantum mechanical way, and that of course is the big challenge as well. After all, quantum mechanics is a beautiful and fairly abstract theory, but applying it to a real system has many Of atoms is different. It took a while before we got to where we are now. When I started in the late 1990s, we could study small molecules. Now we can design large systems, with a thousand atoms that we describe quantum mechanically.
Does this quantum modeling allow you to predict the process of chemical processes?
“Traditionally, many industrial chemical processes and the design of catalysts are often based on trial and error. Small changes happen and then the catalyst behaves differently. But how difficult this is to discover through experiments. Even with the most powerful microscopes, it is difficult to get down to the atomic scale. And therein lies the Our contribution: understanding what happens on the smallest scale – the nanoscale.
“Once you understand that, you can also design, of course. We try to make good predictions about which catalysts will work best. The idea is to design them in a more efficient way, with targeted design. The building blocks of the raw materials, ethylene and propylene, classically come from crude oil or Fossil raw materials Of course we have to get rid of that The world is now completely committed to new raw materials, e.g. The process of converting methanol to olefins. This starts with methanol, which can come from waste, or biomass, but can also be made from carbon dioxide2. So we need innovative catalysts. Also nanoporous materials to remove carbon dioxide2 It is becoming increasingly important.
“Today we still work in very strong synergy with experimentalists. We usually go back and forth. But in the future we hope to be able to fully simulate the process conditions as with a real catalyst – with realistic dimensions and defects – on a computer screen.”
How far are we from that, and how do you want to do it?
“The videos on our website are models with about five hundred atoms, but to simulate a real chemical process on a real catalyst, we have to simulate millions of atoms. Making such a movie with hundreds of atoms and calculating it all quantum mechanically would take two months with a supercomputer. That’s why We are now also working with artificial intelligence.
Molecular modeling of realistic conditions in methanol conversion (CMM)
“Based on quantum mechanical data, which we extract from smaller systems and quantify very precisely, we try to determine the energy of complex systems using machine learning models. Quantum mechanics remains extremely important. Because the training data you create has to be good. A machine learning model is not intelligent; It’s only as good as the data you put into it. We started with machine learning about three years ago, and only now have we published good articles about it on very complex real-world materials.
“I think this is really going to be a game-changer. Simulating a real-life catalytic experiment in which a realistic raw material is sent over a material that is fifty nanometers in diameter or a little more, while the process is going on… it’s really a dream. I hope we will make significant progress in this regard over the next five years.” .
This is of course very ambitious stuff. This is a clear. These are also things you can’t always do with a regular project. Because if you want to do these very innovative and challenging things, there’s sometimes some doubt about it: Is this even possible? It’s not always easy to fund the most innovative research. The European Commission, with grants from the European Research Council, funds such research. But we face very strong competition there.”
‘So, if we had money that I could choose what I would do with it, I would do very innovative and amazing things. The way forward.’
Did that stronger modeling or AI reveal anything you didn’t expect?
“Yes, we have witnessed amazing phenomena. Some materials exhibit elastic behavior. If you increase temperature or overpressure, they change shape. Experimentally it has been observed that elastic behavior is greatly affected by the size of the crystal. We have incorporated that into our model. When we work with small systems We see that all the pores open or close simultaneously. With a million atoms we see a completely different behavior. The material undergoes a phase transition in a completely different way: it starts at a certain nuclear point, and then domains are formed.
Does the demand for modeling come mainly from industrial process designers? Or does it also work in reverse?
‘It can also be the opposite. For a company2We can also do pickup High productivity offers We do where we start from hundreds of thousands of virtual materials. We then apply the selection criteria to this. This way we get access to a set of potentially interesting materials that our experimental partners can then try to synthesize. Of course: not everything can be manufactured. We can include almost anything in our model. This is like playing with Lego, assembling blocks on the nano scale. But it should be possible to make these basic elements experimentally.
“But it often happens that experimentalists have already decided that the material actually works well. Then we can try to understand why that works. Understanding that will allow us to make further changes to improve the design at the nanoscale. For example, we model materials that Recently developed for carbon dioxide2Storage, which has also been tested on an industrial scale.
Isn’t this more important because many of the chemical processes for making things have been around for 100 or 200 years? We only know that it works but not why. But the world is changing now: some raw materials are running out, some we no longer want to use, …
‘Lately there was, about the refinery of the future. FifthToday, most refineries still run almost entirely on crude oil. In the future, we will have to start with other raw materials, such as carbon dioxide2Biomass, as well as plastic waste. We will have to develop very good catalysts and processes for this, so molecular design will be important. Especially since these new raw materials are much more complex.
“Then we have to be able to work with models that represent real-world processes. We have left behind the days of modeling on small molecules. Now we are really trying to model the process as it is.”
What do you think is particularly important for the future of research?
‘Encouraging young people in science and technology. We face great challenges on our way, but we must look at them with optimism and do something about them. I strongly believe in the power of youth. When I see new people in the room and see how excited they are… I think that’s very important to me. I hope that this award will also be an incentive to show the importance of science and technology to younger generations. It’s not about me.’
I also think that cooperation between people is very important. I truly believe that great science can only be achieved through collaboration between people with different scientific backgrounds. I have always encouraged this in my research, including of course here within the Center for Molecular Modeling.
“It remains of course very important that adequate investment is made in independent basic research. Because we must continue to create that fertile ground! If you want to stay at the top of your discipline, you have to rethink everything every time. It is very important to ask Always do yourself again and then do new things. People who get a PhD here are specialists in our field, but later work in all kinds of sectors. The knowledge they gain has a domino effect.
“So I think that as a society we have to support research that we don’t see immediate demand for tomorrow, as well as with funding. This unrestricted research can really lead to major innovations. We also have to do fundamentally new things, like blue-sky research. Otherwise we simply We will go back, as scholars, but also as a society.
