Chemists Build Autonomous Molecular Machines That Walk Using Chemical Fuel

A team at the University of Ulm has demonstrated artificial molecular machines that move autonomously along predetermined paths—powered by chemical fuel rather than external manipulation of each step. The work, published in Nature Chemistry on September 28, 2026, closes a long-standing gap in molecular machine research.

What they built

The system uses molecular "runners" that travel along backbones such as inositol structures. A phosphate group binds to the path and can relocate when fuel is present—moving to positions that would be thermodynamically unfavorable without energy input.

Professor Oliver von Delius's group showed three critical properties together for the first time:

1. Processivity—runners stay connected to their track
2. Autonomy—movement continues while fuel is available without external steering
3. Kinetic asymmetry—fuel drives the system away from its natural resting distribution

The numbers

On a shorter track, 64% of runners reached a less stable middle position after about two hours with fuel—versus negligible movement without fuel. On a longer inositol backbone, the phosphate group did not move at all without fuel even after eight weeks. With fuel, it traversed all five positions along the path.

Why it matters

Molecular machines have won Nobel recognition, but combining autonomy with directional processivity proved elusive. This demonstration provides a reaction cycle that could eventually position functional groups on complex molecules in ways conventional synthesis struggles to access—"moving uphill" thermodynamically to reach hard-to-get configurations.

Potential long-term applications include:

• Synthesizing challenging pharmaceutical intermediates
• Transporting molecular cargo along defined routes
• Building programmable nanoscale factories

Funding and credibility

The research was funded by the European Research Council (ERC) and Germany's DFG—signals of peer-reviewed confidence in fundamental science with multi-year horizons.

Connection to broader engineering trends

The same week, Karlsruhe Institute of Technology (KIT) published work on enzyme hydrogel beads for sustainable industrial biocatalysis—another step toward programmable chemistry at scale.

Meanwhile, SpaceX prepared orbital compute launches and AI labs debated rogue agents. Molecular machines remind us that not all frontier engineering lives in silicon.

Bottom line

We are still years from molecular assembly lines in factories. But September 28, 2026, added a credible mechanism for autonomous molecular motion—another brick in the bridge between chemistry and machine design.

Science continues quietly while AI dominates headlines. This result deserves a place in both conversations.