Even when AI designs molecules with potential as medicines, drug development can be delayed if the process of actually producing them is slow. KAIST researchers have developed an “unmanned synthesis platform” that uses synthesis robots and light to automatically carry out the entire process from mixing raw materials to synthesis and analysis. The technology is expected to help secure a wider range of drug candidate compounds by reducing the burden of repetitive experiments.
KAIST (President Choongsik Bae) announced on October 18 that a research team led by Professor Yoonsu Park of the Department of Chemistry developed an unmanned synthesis platform combining visible light with a continuous-flow process. Using the platform, the researchers synthesized the “N-methyl pyrrole” scaffold, a fundamental molecular framework used in pharmaceutical development, and demonstrated the potential to scale up production through continuous manufacturing.
In drug development, molecules that have been designed must actually be synthesized so that their efficacy and safety can be tested. However, even slight differences in molecular structure can require entirely new reaction conditions, and even when small-scale synthesis is successful, scaling up production can alter the way light or heat is transferred, requiring the conditions to be adjusted again.
To reduce this burden, the research team integrated a raw-material dispensing robot, automated pumps, a light-irradiation reaction system, and analytical equipment into a single system. Once a researcher prepares the equipment and sets the conditions, the robot mixes the raw materials, and pumps continuously feed the raw-material solution through a narrow tube at a constant rate. Visible light is then irradiated onto the tube to induce the chemical reaction.
This method, in which raw materials are continuously fed through the system while undergoing a reaction, is called a “continuous-flow process.” It is similar to continuously supplying ingredients into a small pot rather than putting everything into a large pot at once. Using a narrow tube makes it easier to distribute light evenly throughout the solution, while production can be increased simply by operating the same process for a longer period.
After the reaction is completed, the solution is automatically transferred to analytical equipment. The equipment determines whether the desired compound has been successfully produced and how much has been made. From mixing the raw materials to checking the final results, the entire process proceeds automatically without requiring researchers to manually transfer and measure the solution each time.
To evaluate the performance of the automated system, the research team benchmarked a single-atom editing reaction. This is a new technology that selectively replaces a specific atom while keeping the overall molecular framework of a starting material called furan* intact. The compounds produced in this way can change how a drug binds to proteins in the body, and the technology is attracting attention in drug development for this reason.
* Furan: A five-membered ring compound composed of oxygen and carbon. It is a basic starting material that serves as a molecular scaffold for the synthesis of pharmaceuticals and other chemical substances.
The research team applied the synthetic method to more than 20 furan-based compounds. Even in structurally complex natural products, the desired portion could be modified while leaving the rest of the structure intact. In the continuous-flow process, the team also succeeded in continuously producing the N-methyl pyrrole scaffold.
Drug development involves producing compounds with slightly different structures and comparing their efficacy and safety. For example, part of an existing compound may be modified and then tested to determine whether it binds more effectively to a therapeutic target protein or has lower toxicity than the original compound. The new technology is expected to make it more efficient to obtain the compounds needed for such comparisons, helping researchers identify promising drug candidates.
Professor Yoonsu Park said, “How quickly and reliably designed molecules can be synthesized will determine the speed of future drug development,” adding, “We expect the platform developed in this study to provide a foundation for moving toward an era of automated organic synthesis by enabling reaction conditions to be quantified and reproduced.”
Hyeonbhin Byeon and Hojin Hong of the KAIST Department of Chemistry participated in the study as co-first authors, while Professor Yoonsu Park and InnoCORE Fellow Ha Eun Kim of the InnoCORE AI-CRED Innovative Drug Research Group participated as corresponding authors. The research findings were published online on August 7 in JACS Au , an international journal published by the American Chemical Society.
※ Paper title: Direct Access to the Simplest N-Substituted Pyrroles from Furans Enabled by Tailored Photocatalysts, DOI: 10.1021/jacsau.6c01035
This research was supported by the Korea Basic Science Institute’s Young Researcher Infrastructure Support Program (Project No. RS-2025-00642970) and the InnoCORE AI-CRED Innovative Drug Research Group (Project No. N10260004).
JACS Au
Direct Access to the Simplest N-Substituted Pyrroles from Furans Enabled by Tailored Photocatalysts
7-Aug-2026