Chemists at the University of Amsterdam (UvA; the Netherlands; ) have developed an autonomous chemical-synthesis robot with an integrated artificial-intelligence (AI)-driven machine learning unit. Dubbed “RoboChem,” the benchtop device can outperform a human chemist in terms of speed and accuracy, while also displaying a high level of ingenuity. Said to be the first of its kind, RoboChem could significantly accelerate chemical discovery of molecules for pharmaceutical and many other applications. RoboChem’s first results were published recently in Science.

RoboChem was developed by the group of professor Timothy Noël at UvA’s Van ‘t Hoff Institute for Molecular Sciences (). Their Science article shows that RoboChem can perform a variety of reactions while producing minimal amounts of waste. Working autonomously around the clock, the system delivers results quickly and tirelessly. “In a week, we can optimize the synthesis of about ten to twenty molecules,” says Noël. “This would take a PhD student several months.” The robot not only yields the best reaction conditions, but also provides the settings for scaleup. “This means we can produce quantities that are directly relevant for suppliers to the pharmaceutical industry, for example.”

In RoboChem, a robotic needle collects starting materials and mixes these together in volumes of just over 0.5 mL. These then flow through the tubing system to the reactor (photo), where the light from powerful light emitting diodes (LEDs) triggers the molecular conversion by activating a photocatalyst included in the reaction mixture. The flow then continues towards an automated nuclear magnetic resonance (NMR) spectrometer that identifies the transformed molecules. These data are fed back in real time to the computer that controls RoboChem, which processes the information using AI. “We use a machine learning algorithm that autonomously determines which reactions to perform. It always aims for the optimal outcome and constantly refines its understanding of the chemistry.” explains Noël.

The researchers also used RoboChem to replicate previous research published in four randomly selected papers. They then determined whether Robochem produced the same — or better — results. “In about 80% of the cases, the system produced better yields. For the other 20%, the results were similar,” Noël says.