Mind meets electricity: The arrangement of the cooled reaction tubes in the printed microreactor is optimized with the help of artificial intelligence

Mind meets electricity: The arrangement of the cooled reaction tubes in the printed microreactor is optimized with the help of artificial intelligence

Microreactors from the 3D printer

Readingtime 8 minutes

How can chemical reactions be carried out as sustainably and efficiently as possible? A team from Evonik is researching this question in cooperation with external partners and taking advantage of the possibilities of 3D printing


TEXTKarl Hübner

Karl Hübner is a journalist who has a PhD in chemistry. He works part-time as a freelance author and often writes about research topics

If Dr. Senada Schaack has her way, the chemical reactor of the future will be small, custom-tailored, and homemade. “It’s disruptively different,” says Schaack, a process engineer who is researching a new generation of minireactors for Evonik in Hanau. The metal contraption is only about as big as a shoebox, but it promises great advances. Its abilities aren’t apparent from the outside. However, its inner workings are designed so that a chemical synthesis requires as little energy as possible and causes less CO2 emissions. Some reactors contain extremely thin tubes through which the reaction mixture or a cooling medium can be pumped during operation. Others contain intricately branched systems, which help to distribute flows from one tube to several.

“If we do this well, we can achieve a high degree of purity in the desired product and save energy that previously had to be used for cooling or complex material separations, such as distillation,” explains Schaack, who runs the Simulation and Additive Manufacturing competence center, or SAM 3D for short. Together with her employees, she develops concepts for the design of reactors—and then manufactures them from stainless steel powder using 3D printing.

These reactors could be used in many applications, for example in the production of methanol for solar or wind energy storage. The team is currently working on a new type of reactor for a synthesis step that often plays a role in the manufacture of active pharmaceutical ingredients: ortho-lithiation. “We designed the apparatus in such a way that it not only delivers the product with higher purity but also requires less cooling for the reaction than the conventional process,” says Schaack.

This development is part of the 3D-PROCESS project funded by the Federal Ministry for Economic Affairs and Climate Action (BMWK). The project is devoted to disruptive reactor concepts. These devices are digitally planned, manufactured in 3D printers, and optimized with the help of empirical process data and artificial intelligence. “We are completely rethinking the processes in chemical engineering,” says Schaack. Together with the project partners, she has set herself the goal of making chemical production more sustainable.

The potential for this seems enormous. Prior to the pandemic, in 2019, the chemical industry consumed 200 terawatt hours of energy in Germany alone in the form of natural gas, petroleum products, coal, and electricity—eight percent of the total demand.


Ortho-lithiation, for which the 3D-PROCESS team in Hanau is developing a reactor, is a case in point where a lot of energy can be saved with a better design. In this synthesis step, a lithium atom is temporarily introduced into an organic molecule. This releases so much heat that the reaction vessel has to be cooled down to at least minus 50 degrees Celsius. This technique uses liquid nitrogen, which requires a lot of energy to produce. A cooling temperature of -20 degrees Celsius could be sufficient for the new approach with the small metal block. The cooling could be carried out with an electrically temperature-controlled liquid, which would reduce energy consumption by 80 percent compared to the nitrogen variant.

The enormous savings are made possible by the innovative design of the reactor. It differs significantly from the conventional tank reactors commonly used in the chemical industry. In the latter reactor, the reaction heat is dissipated by cooling coils in the interior or in the double-walled outer shell. Because the cooling surface is relatively small in relation to the reactor volume, the cooling medium has to be very cold. The new reactor is a much smaller system known as a flow reactor. In the simplest case, the reaction mixture is passed through a linear reaction tube, which is wrapped in a cover containing the cooling medium. The ratio of the cooling surface to the reactor volume is significantly larger, which means that the medium does not have to be that cold. The heat dissipation becomes even more effective if the reaction mixture is not just passed through one tube but distributed over several channels.

Senada Schaack inspects the ortho-lithiation apparatus. Plans call for the lines to soon lead to the first self-printed test reactor

Senada Schaack inspects the ortho-lithiation apparatus. Plans call for the lines to soon lead to the first self-printed test reactor


This is exactly how the experts at Evonik designed their microreactor. “In the first prototype, we distribute the flow of the reaction mixture over a total of 256 tubes,” explains Dr. Hendrik Rehage, a process engineer on Schaack’s team. The advantages are obvious: More tubes increase the flow rate while making cooling more efficient at the same time, because each tube is surrounded by its own cooling channels. The experiments are intended to show whether the cooling temperature of -20 degrees Celsius calculated on the computer is actually sufficient.

The 256 reaction channels are each 0.8 millimeters in diameter. With conventional methods of metal processing, this would be an immense effort or not possible at all. But the situation is different when you use a 3D printer. “It doesn’t matter at all whether we manufacture a reactor with one or 256 channels,” says Rehage, explaining why additive manufacturing also makes very complex reactor geometries accessible. The decisive advantage of 3D printing compared to the classic production of microreactors is that the printed items can have any shape. Luckily, it’s long been possible to carry out 3D printing with metals such as aluminum, titanium, and stainless steel. The metal is fed in in finely powdered form, melted by a laser beam, and then deposited layer by layer on the workpiece to be manufactured.

“The microreactor will deliver products of higher purity and require less cooling”


“The dimensions and also the structural design of the test reactor are the result of the simulations that we performed beforehand,” says Rehage. The experts simulated how different geometries and process parameters affect where in the reactor it gets hot and to what extent, as well as how quickly the heat is dissipated. However, many of the mathematical descriptions are only approximations of reality. Therefore, it’s still important that an experiment be conducted afterward.


This was the case with the first prototype for ortho-lithiation. The team wants to use the possibilities of 3D printing to provide the components with integrated connections for the installation of sensors. These record temperatures, pressures, flow rates, and substance concentrations during operation. “This data lets the system know whether it is getting too hot somewhere in the reactor or whether unwanted byproducts are being produced,” explains Rehage. The engineers absolutely want to avoid both, because they reduce yield and product purity. If the product has to be laboriously separated from other molecules, this also increases energy consumption.

The team carries out test runs with the 3D printer for plastic components before the components are printed in stainless steel.

The team carries out test runs with the 3D printer for plastic components before the components are printed in stainless steel.

The 3D-PROCESS team uses methods of artificial intelligence to determine the parameters for which the yield and purity are as high as possible and the energy required for cooling is as low as possible. “On the basis of the sensor data, the system is able to adjust the process control and thus gradually get closer to optimal conditions in a self-learning loop,” explains Schaack. If this optimum cannot be achieved with the existing reactor, the data provide information on how the design should be changed, Schaack adds.

“We can produce the reactor for every new idea within two days”


Dr. Stefan Randl is also enthusiastic about such visions. As the head of research in the Health Care business line, Randl is following the experiments with the test reactor with great interest. He considers ortho-lithiation to be “a model example for many other syntheses in the pharmaceutical environment that previously had to be cooled or heated with a great deal of energy.”

Evonik produces active pharmaceutical ingredients and their precursors for external customers. It often happens that Randl’s team still has to develop the associated synthesis process. Wherever possible, attempts are made to implement a continuous process in a small flow reactor. “There is currently a trend in the industry away from classic batch production in large stirred tanks,” reports Randl. Continuous processes in microreactors enable temperature, pressure, and safety to be controlled much better. The miniaturized processes also require less solvent because more concentrated liquids are fed into the reactor. In addition, the catalyst required for the reaction can be used in the form of a much finer powder, which in turn leads to better product yields.

“For us, it’s always important to develop new processes as quickly and reliably as possible, and in such a way that they are economical and offer customers in the pharmaceutical industry products of the highest and consistent quality,” says Randl. It’s a great help that reactor concepts can be simulated in advance and that the reactor that is best suited for an application is printed in the end.

Process engineer Hendrik Rehage is a project manager in the 3D-Process project at Evonik

Process engineer Hendrik Rehage is a project manager in the 3D-Process project at Evonik


However, the printed microreactors should also bring their disruptive properties to bear in other areas. In another subproject of 3D-PROCESS, Evonik is working with researchers from the Karlsruhe Institute of Technology (KIT), Siemens Technology, and the KIT spin-off INERATEC. The partners are working on a reactor for the conversion of hydrogen and carbon dioxide (CO2) into methanol or dimethyl ether. Both substances are important basic chemicals and can also be used as fuels. While the CO2 comes from biogas or exhaust gases, the hydrogen can be obtained through water electrolysis—preferably when excess renewable electricity is available.

“Our goal is to create decentralized systems made up of standardized modules that can be set up next to a wind farm or a large-scale solar power system, for example,” says Professor Roland Dittmeyer from the KIT Institute for Micro Process Engineering. A corresponding reactor must be simple and robust and also meet a number of requirements, one of which is to separate the resulting methanol directly from the reaction mixture. In fact, 3D printing allows defined porous structures in the reactor wall. Thanks to these structures, this can now be achieved in connection with a suitable temperature control.


The project “Disruptive reactor concepts by means of additive manufacturing: From the digital design to industrial implementation—3D-PROCESS” kicked off in June 2021 and will conclude in May 2024. It has a budget of €9.8 million. It is funded as a joint project by the Federal Ministry for Economic Affairs and Climate Action under the funding code 03EN2065A-E. The consortium leader of the project is Evonik. Other partners include Siemens, the Karlsruhe Institute of Technology (KIT) with the Institute of Catalysis Research and Technology, the Institute for Chemical Technology and Polymer Chemistry, and the Institute for Micro Process Engineering, as well as INERATEC. The aim is to develop process engineering components for more energy-efficient chemical processes that cause fewer emissions and are more sustainable.


Initial trials in a test reactor have been promising. When all the small-scale studies are complete, the consortium will start to develop design rules to scale up the reactor to an industrial scale. Professor Christoph Klahn, who works on additive manufacturing in process engineering at KIT, praises the acceleration of the development process. “Thanks to the well-prepared design and manufacturing workflows, we can produce the corresponding experimental reactor for every new idea within two days,” he says.

The software modules for these end-to-end workflows come from Siemens and include simulation, design, and digital control of the 3D printers for reliable production of the fine structures. Dr. Christoph Kiener, Principal Key Expert Functional Design at Siemens, says than the 3D-PROCESS research project is an important contribution to the energy transition. He thinks that the new possibilities will transform the concept of chemical reactors. “Until now, this has been the place for a chemical reaction,” he says. “In the future, we will be able to bundle all the important process steps, i.e. mixing, reacting, separating, and so on, in a tailor-made way through simulation in one apparatus so that they then take place highly efficiently in defined zones.” The test reactor, in which the synthesis and separation of methanol are now combined in a very small space, gives an idea of what is possible.

The image of chemical parks could change in the future if small flow reactors were increasingly used instead of large tanks. But do these minisystems actually supply sufficient quantities? “That’s definitely the case for active pharmaceutical ingredients,” says Stefan Randl from Evonik. With a throughput of one liter per minute, the small reactors can produce tens to hundreds of tons of active ingredients per year. Moreover, the concept can easily be scaled. Depending on the desired product quantity, the necessary number of reactors can be connected in parallel.

The concept is a viable alternative even for large-scale chemical production operations, says Schaack. She sees the potential for radically smaller systems and processes that—thanks to precise data, better simulation, and precise control, require less steel, less solvent, and ultimately less energy. That, in turn, would also lead to less CO2 emissions. This really would be disruptive and enable chemistry to be more sustainable.

Credits: Andreas Doria / Upfront with motifs from Getty Images and Evonik Industries, Ramon Haindl (3), private

Illustration: Oriana Fenwick / Kombinatrotweiß with photo template from private

Efficiency enhancement

Stirring tank reactor vs. microreactor

During chemical reactions, through-air tanks from the 3D printer dissipate heat better than their predecessors.

Climate Change

More efficiency for climate protection

Despite many setbacks, an increase in energy efficiency is still the most viable way to limit the effects of global warming.

Monolith ROMEOVideo

Chemical Production

A reactor with a good energy balance

Researchers in the ROMEO project have developed a process that makes an industrial reaction more energy-efficient.

Senada Schaack

Additive manufacturing

The printed plant

Good for the environment and for cost-efficiency: Evonik is using plant components created by 3D printers to construct its own facilities.