Hydrogen, oxygen, carbon monoxide, carbon dioxide, methane – the steel industry releases a veritable cocktail of gases every hour. But how can these metallurgical gases be purified? This is where the research of Professor Peter Awakowicz from the Chair of Electrical Engineering and Plasma Technology and Professor Martin Muhler from the Lab of Industrial Chemistry comes in. The interdisciplinary research team at RUB studies how non-thermal plasma can be used for targeted cleaning and processing the metallurgical gas mixture. In the Carbon2Chem joint project funded by the German Federal Ministry of Education and Research (BMBF), in which both researchers have been involved since 2016, they are testing their innovative plasma technology on real gases. “The combination of a basic research project in Collaborative Research Centre 1316 and an application-oriented BMBF project has been a long-held dream of both of us that we can now fulfil,” says Awakowicz.

In sub-project L3 of Carbon2Chem, in which the RUB researchers are involved, the specific issue is pre-cleaning, the removal of oxygen from the coke oven gas. “This sounds simple, but it is tricky in detail,” explains research chemist Muhler. According to him, it is an intricate art to remove the oxygen from the predominantly hydrogenous coke oven gases. Traditional methods of exhaust gas purification, such as pressure swing adsorption, would not work if there was too much oxygen. The high chemical reactivity of oxygen would trigger dangerous gas reactions under normal pressure, such as an oxyhydrogen explosion. This is why Awakowicz and Muhler rely on pre-cleaning using plasma technology with cold plasma. How does it work? What makes non-thermal plasma so special? And how is it generated?

Innovative Technology for gas purification with cold plasma

Cold plasmas, or non-thermal plasmas, are plasmas in which the temperatures of ions, electrons and neutral particles vary. “The temperature of the electrons is high in these plasmas, while the temperature of the other gas particles is comparatively low,” explains Awakowicz. Since the plasmas are in thermal non-equilibrium, they are also often called non-equilibrium plasmas. They have an advantage with regard to gas purification processes: The ignited, cold plasma can be used for gas treatment without causing a significant increase in the temperature of the gas.

However, producing cold plasma is not easy. “The difficulty lies in supplying the gas with just enough energy so that the light electrons are accelerated and thus become hot, but the temperature of the large, heavy neutral particles and ions hardly changes,” explains Awakowicz. The research team from the Chair of Electrical Engineering and Plasma Technology has succeeded in producing precisely this state of non-thermal plasma in the purpose-built plasma reactor: The electrons become several tens of thousands of degrees Celsius hot, while the gas temperature of the entire plasma increases to barely more than room temperature.

“To achieve and understand this state, complex plasma diagnostics were necessary. We had to repeatedly readjust the individual parameters, such as the geometry and materials of the electrodes, the voltage amplitude and frequency, and associated with this the input power. Then the fundamental plasma parameters such as the electron density, the distribution function of the free electrons, but also the gas temperature had to be determined in order to optimise everything,” as Awakowicz describes the challenges.

While the team led by electrical engineer Awakowicz was fine-tuning the parameters for producing the cold plasma, the chemical researchers led by Muhler were analysing the chemical reactions triggered by the plasma discharge. It turned out that the cold plasma is so reactive that it animates the oxygen contained in the coke oven gas to react with hydrogen, so that water is formed. The gas mixture is freed from oxygen and is thus ready for further purification processes.

Waste gas purification on an industrial scale

What Awakowicz and Muhler have fundamentally researched in the RUB laboratory is being applied to specific gas mixtures in the steel industry in the BMBF Carbon2Chem project. In the first project phase from 2016 to 2020, the researchers already provided proof of feasibility: Their plasma technology can be applied to these specific metallurgical gases. In the second funding phase from 2020 to 2024, the technical processes will now be further validated and scaled up for industrial application from 2025.

Scale-up in the Carbon2Chem pilot plant

The relevant experiments take place on an area of 3,700 square metres in the pilot plant in Duisburg. The pilot plant was built in 2018 adjacent to the thyssenkrupp Steel Europe site and means that Carbon2Chem’s experiments can be conducted under industrial conditions. “The real exhaust gases are routed to the pilot plant site, where they are available to us,” explains Muhler. “We now have to show that our plasma system can operate with the real gases – on a much larger scale, of course. The reactor should be able to purify more than fifty times the amount of gas,” as he outlines the challenge. At RUB, the researchers have so far worked with small gas flows of ten litres per minute in the lab; at the pilot plant, they are dealing with flows with a much larger volume of 500 litres per minute and more. “A spectacular project, since the dimensions are so huge,” point out Awakowicz and Muhler.

Industrial implementation planned in 2025

The commercial implementation of the gas purification plant is scheduled for four years from now. “The final step, scaling up from tenfold to one hundredfold, will be an effort,” Awakowicz suspects, adding: “As researchers, we will have to hand over the baton to industry at some point.”

adapted from Lisa Bischoff (RUB)

The synthesis of many chemicals results not only in the desired product but also in its mirror image: the physico-chemical properties of the so-called enantiomers are very similar and they are therefore difficult to separate. This, however is necessary, because they often have different biological properties. This is important, for instance, when it comes to drugs. The (S)-ibuprofen isomer is effective against pain, but its twin (R)-ibuprofen is not. “Occasionally, one of the two forms is even toxic,” Professor Julia Bandow, Chair of Applied Microbiology at the RUB Faculty of Biology and Biotechnology points out.

Her research group employs enzymes, i.e. biological catalysts derived for example from bacteria or fungi to produce such chemicals. Some enzymes produce only one of the two enantiomers.

Enzymes are sensitive

However, enzymes are in general rather sensitive catalysts. Some are susceptible even to inactivation by the substrate they convert. “This is the case for our example enzyme. The unspecific peroxigenase, or UPO for short, extracted from the edible fungus Agrocybe aegerita or chestnut mushroom and synthesised by the research group of Professor Frank Hollmann from Technical University Delft can produce the fragrance (R)-1-phenylethanol. It requires hydrogen peroxide as a substrate for this purpose. If hydrogen peroxide is simply added to the enzyme-containing solution in the form of a concentrate, the enzyme is quickly inactivated,” explains Julia Bandow.

This dilemma the team resolved using several tricks. One of them was to use a plasma to produce hydrogen peroxide on demand. “That was a somewhat crazy idea,” Bandow admits in retrospect. “Because in fact, plasmas are the preferred means of destroying things.” In plasmas, which are created by adding energy to a gas, numerous reactive substances are formed, for example atomic oxygen, hydroxyl radicals, free electrons, and various excited species. Consequently, plasmas can be used to inactivate cancer cells, biofilms, viruses, or prions. In this instance, however, the plasma was supposed to help protect the biocatalysts by providing the enzyme with exactly the right dose of hydrogen peroxide needed to catalyse the fragrance at the push of a button.

“The generation of hydrogen peroxide using plasma has advantages over alternative production methods by means of enzymes or electrodes. Such enzymes are expensive to produce, they are attacked by hydrogen peroxide and hydrogen peroxide production is difficult to dose. Biocatalysts like UPO can precipitate at electrodes and clog them,” explains Julia Bandow.

Trick number two

The group therefore experimented with plasmas based on air or noble gases ignited directly above the enzyme-containing solution to produce the fragrance (R)-1-phenylethanol. However, the enzymes at the surface still were quickly destroyed by the reactive species. Here, trick number two came into play: the researchers attached the enzymes to beads, small spheres with a porous surface that lie at the bottom of the solution and hold the enzymes in place. They tested the optimal composition of the beads beforehand, because not every enzyme can dock equally well on every surface and still do its job, as this sometimes requires enzymes to move.

As a result, the beads sitting at the bottom of the container are covered by the solution, which acts as a buffer zone separating the enzyme from the plasma phase. The hydrogen peroxide produced by the plasma diffuses to the enzymes, where it is used in the reaction. Thanks to the buffer zone, the enzymes don’t come into contact with toxic doses of the substrate or other reactive species. Thus they remain intact and functional.

Tiny reactors

The reactors used so far are tiny. They can only hold up to five millilitres of liquid. “The protective layer that covers the beads is only about one millimetre thick,” says Julia Bandow. “This gap is sufficient to protect the enzymes from unstable and short-lived reactive substances that are created in the plasma. Some conveniently react to form hydrogen peroxide that is used as a substrate.” Hydrogen peroxide itself is comparatively long-lived. Its dose can be adjusted, for example, by pulsing the plasma, i.e. switching it on and off.

In the experiment, the enzymes were active for eight cycles of ten minutes without damage. Between cycles, the product (R)-1-phenylethanol was harvested and the second substrate, namely ethyl benzene, was replenished. The extraction of the product (R)-1-phenylethanol from the solution was carried out with ethyl acetate in just one step. “This is the great advantage of the biocatalytic over catalytic processes which produce both enantiomers,” stresses Julia Bandow.

Up to this point, the whole experiment was a proof-of-concept project, and it proved one thing: the approach works. Now, the research group is optimising the process, mainly with the aim of scaling up production volumes and optimising reactors. One reactor in which the enzyme-loaded beads rotate in the solution so that the enzymes have a steady substrate supply has already been tested successfully.

Further experiments have shown that the process works even better if the noble gas helium is used instead of air as the basis for plasma generation and if water vapour is added. “You find a lot more hydrogen peroxide in the solution, perhaps generated from plasma-generated OH radicals,” as Julia Bandow speculates. This increased the yield from about ten nanomole hydrogen peroxide per minute to 200 nanomole per minute. A further increase in hydrogen peroxide formation by 50 per cent was achieved by changing the voltage.

“Our greatest success is that we have come so far in a relatively short time,” Julia Bandow points out. “The plasma reactor might open up a new class of enzymes for commercial use.”

In the search for other enzymes that can produce valuable chemicals, the researchers are pursuing several strategies. Streptomyces, a group of soil-dwelling bacteria, also have hydrogen peroxide converting enzymes. However, the characterisation of the first three candidates hasn’t yet turned up any promising manufacturers of attractive products.

Another route to new enzymes could be through compost. Tests with model substrates, whose turnover is for example indicated by a colour change, revealed promising results. “We already know at what pH and temperature such reactions take place,” says Bandow. “But we haven’t yet identified which enzymes are responsible. That is the great unknown.”

adapted from Meike Drießen (RUB)

Even though plasmas at atmospheric pressure are often only a few cubic millimetres in size, they pack quite a punch. This is because special non-equilibrium states can be set up in them, which facilitate physical and chemical processes that are not possible in any other environment. The plasma thus becomes a special kind of laboratory, where atoms and molecules can be excited without their surroundings heating up. “Such excitations could theoretically also be generated in a gas, but to do so we would have to heat it to several thousand degrees Kelvin. As a result, the molecules would decompose,” explains Professor Uwe Czarnetzki, Head of the Chair of Plasma and Atomic Physics at the Faculty of Physics and Astronomy. For many years, he and his team have been developing methods to explore the processes inside plasmas and to characterise the plasmas.

Plasmas boast a unique feature: electric fields can be used to supply energy to the electrons in the plasma; the electrons in turn interact with molecules such as nitrogen or carbon dioxide while transferring the energy to them. The molecules are excited, and this happens without the environment heating up in the process, as would be the case in a gas. The molecules that are excited to vibrate have a much higher reactivity than those in the ground state. Plasma can therefore change chemistry or even enable certain chemical processes in the first place. Consequently, plasma provides basic researchers with a unique opportunity to study the excitation of molecules and the associated chemistry beyond thermodynamic equilibrium. Uwe Czarnetzki is therefore primarily interested in the vibrational states of molecules in plasmas.

Analysing vibrations in molecules

The individual atoms of molecules – for example, the carbon and oxygen atoms in the carbon dioxide molecule – are not rigidly bonded to each other. The bonds between the atoms periodically deform in different ways. These oscillations can take place on several energy levels, i.e. with different frequencies that can be excited by light energy. This requires the frequency of the light to correspond to the difference in frequencies between two neighbouring energy levels. It is possible to deduce the number of absorbing molecules in the light beam from the decrease in the intensity of the light at this particular frequency. The Bochum-based physicists led by Uwe Czarnetzki and Dr. Dirk Luggenhölscher in Collaborative Research Centre 1316 “Transient Atmospheric Pressure Plasmas” are using this fact to analyse the vibrational states of molecules in plasmas, specifically in collaboration with Humboldt Fellow Dr. Yanjun Du. She wants to find out how many different states occur and how often.

“Unfortunately, this method doesn’t work for all molecules or vibrations,” says Czarnetzki. That’s why he and his team, foremost among them Jan Kuhfeld, have refined the CARS method, short for Coherent anti-Stokes Raman Scattering. The complex laser method detects the otherwise forbidden transitions with high sensitivity, temporal and spatial resolution. In particular, the researchers revised the evaluation procedure for the theoretical calculation of the spectra so that they can also determine energy distributions that do not correspond to a thermodynamic equilibrium. “This is precisely the crucial point when using plasmas,” says Czarnetzki. The team also modified the laser system to detect all oscillation states simultaneously.

Understanding the mechanism behind excitation

The electric field is crucial to excite the oscillatory states. The crux of the matter is that a very high electric field is needed to generate the plasma, while a comparatively small field is needed for efficient vibration excitation.

At first, it was not clear why efficient oscillation excitation works despite this supposed contradiction. “Plasmas tend to do things that we don’t understand at first, but which are useful,” points out Czarnetzki. “Then, once we understand the mechanism, it turns out to be very mundane in retrospect.” Thanks to a new laser technique developed in Bochum by Dr. Nikita Lepikhin and his colleagues to measure the electric fields, the researchers finally understood what exactly happens when the plasma is created and the oscillations are excited.

First of all, one has to bear in mind how a plasma is created: To begin with, there is a gas to which energy is added in the form of electric current, until finally a high-energy state emerges in which a certain proportion of the gas is ionised. However, not all charges are generated at once. Initially, only some particles of the gas are ionised and accelerated, which in turn produces new charges. “In a few nanoseconds, an avalanche of charges forms and a high-density plasma is created,” outlines Uwe Czarnetzki. It is not possible to maintain such a high-energy state permanently, so the plasmas are operated in pulsed mode, i.e. switched on and off again and again, so to speak, typically a few thousand times per second.

Pulsed plasma solves supposed contradiction

Due to the extremely fast pulsing of the discharge, the plasma solves the supposed problem of the different optimal electric field for plasma generation and oscillation excitation all by itself, so to speak. In the first nanoseconds after applying the voltage, there is a low plasma density, resulting in a high electrical resistance and thus a high field. Following strong ionisation, an exceptionally high plasma density is eventually achieved. This reduces the resistance and thus also the field. In the subsequent phase, the field is small and ideal for the excitation of oscillations. In addition, the charges discharged by the current flow are suitably replenished in a very thin transition region between the plasma and the electrode. By far the largest part of the high-density plasma, however, doesn’t generate any new charges, but only acts as a current conductor. Here, the electrons are gently accelerated in a weak field and give up almost all of the energy thus absorbed to the vibrational excitation of the molecules.

This already works well enough, at least in nitrogen molecules. In future, the physicists also intend to use this method to study CO₂molecules. Initial measurements and theoretical considerations give them cause for optimism.

Combining different measuring methods

In the meantime, Czarnetzki’s team has also succeeded in combining the method for measuring the vibrational states with the method for measuring the electric fields in plasmas – a challenging task. “The smaller a plasma, the bigger the experimental setup required to study it,” points out Uwe Czarnetzki. “Our plasmas are so small that we don’t have room to insert a thick filament into them.”

Laser technology solves this problem, but it requires complex experimental setups. Therefore, the experimental setups for measuring the electric fields and the vibrational states couldn’t simply be coupled. Instead, the physicists created two identical plasmas with great effort and attention to detail, on which they can run the measurements simultaneously and then pool the data. Each little detail made a difference. “It can be enough if a cable is laid incorrectly to get interference in the system,” says Uwe Czarnetzki. The fact that it was possible to obtain a coherent picture of the processes in the plasma for the first time by combining the two sensitive measurement methods is therefore one of the highlights of his research.

adapted from Julia Weiler (RUB)