Scientists at a leading Chinese university advance oxygen extraction for Mars using AI-guided chemistry
Researchers from the University of Science and Technology of China in Hefei have identified a molecule with the potential to generate oxygen on Mars. The discovery was achieved with the aid of artificial intelligence and was published in Nature Synthesis, a prominent scientific journal. The finding marks a significant step in planning future crewed missions to the red planet, where sustainable oxygen production will be essential for both life support and fuel, according to the NatSynth report.
For long-duration Mars missions, oxygen is not only vital for breathing but also a key propellant component. The most economical approach involves exploiting in-situ resources available on Mars rather than transporting all supplies from Earth. Water ice, known to be abundant in polar regions of Mars, presents a viable source if a catalyst system can confidently split frozen water into oxygen and hydrogen for practical use.
The research team concentrated on Martian meteorites—rocks that were ejected from Mars and later landed on Earth after space impacts. These samples provided a close analogue to Martian geology and offered a practical substrate for testing catalytic chemistry under realistic conditions. The AI-powered robotic system was deployed to collect samples and perform laser-based scans, enabling rapid, high-resolution analysis of the alien ore. The system’s algorithm evaluated the composition and suggested that millions of molecular candidates could emerge from a combination of six metal elements present in the rocks, including iron, nickel, manganese, magnesium, aluminum, and calcium as catalysts or catalyst components. The study estimates that virtually 3.7 million molecular configurations could be explored from these elements alone.
Over a six-day period, the robotic chemist efficiently progressed through the design, synthesis, and testing of 243 different molecules. Among these, one catalyst demonstrated the best performance by enabling water splitting at minus 37 degrees Celsius, a temperature that closely resembles Martian environmental conditions. This breakthrough highlights the potential for a robust catalytic system to operate under the cold, dry climate of the Martian surface and to advance oxygen production using local resources.
Experts note that using traditional trial-and-error methods would be prohibitively slow for identifying such an optimal catalyst. Estimates suggest that a human researcher applying conventional approaches could take around two millennia to arrive at a comparable solution. The AI-driven approach substantially accelerates discovery, enabling rapid screening and optimization of catalysts tailored to the Martian context. This development aligns with broader efforts to harness autonomous technologies for space exploration, reducing the need for constant Earth-based intervention and enabling more resilient mission architectures.
Earlier work in the field included explorations into extracting oxygen from lunar dust, a related challenge that informs strategies for in-situ resource utilization on airless bodies. The ongoing pace of discovery in this area continues to push the boundaries of how future explorers might live off the land, turning local resources into critical life support and propulsion assets. While the current focus is on Mars, the methodologies and AI-assisted workflows developed here may influence other planetary missions and, more broadly, the design of autonomous chemical systems for space environments. These efforts are widely reported in scientific outlets and summarized in synthetic chemistry reviews that emphasize practical, field-ready catalysts and scalable processes. (Attribution: NatSynth coverage and related scientific communications)