Scientists Tried to Kill Soil for Years, but It Kept Breathing
French researchers found that soil sterilized with radiation kept consuming oxygen and releasing carbon dioxide for six years. They suggest that metabolism-like reactions can occur outside living cells.

Sébastien Fontaine, a biochemist at the French National Institute for Agriculture, Food, and Environment, has spent 15 years trying to find out whether completely lifeless soil still releases carbon. His team sealed soil in jars and exposed it to sterilizing gamma radiation, then waited for carbon dioxide emissions, a sign of microbial respiration, to stop. They did not. The soil kept emitting the gas for weeks, then months.
The latest results, published in 2025 in Science Advances, show the samples kept consuming oxygen and releasing carbon dioxide for six years. The authors conclude that metabolism-like processes, which in living organisms take place inside cells, may also be possible without living cells. Fontaine says such reactions could even predate life on Earth.
How sterility was tested
The first results were published in 2013 in Biogeosciences, but some reviewers doubted the soil was truly sterile. The team therefore tried to kill it more thoroughly, using more radiation, pressure and heat, yet emissions persisted. Cells seen under an electron microscope showed no RNA or DNA, meaning they were dead. When researchers deliberately added microbes, these quickly recolonized the soil and released far more carbon dioxide.
Researchers Clémentin Bouquet and Benoit Kéraval followed two sets of sterilized samples, one of ordinary soil and one with added glucose. Emissions declined but did not vanish on days 1,606 and 2,442, and were higher in the glucose samples. In a specially built fuel cell, the sterile soil carried an electron current several times larger than a saltwater control. A preprint reported four of the eight intermediate molecules of the Krebs cycle.
Doubts and alternative explanations
Joshua Schimel, a soil ecologist at the University of California, Santa Barbara, says the findings are not very surprising, since iron and aluminum oxides in soil can catalyze such reactions. Markus Ralser of Charité University Hospital in Berlin and Sudha Rajamani of the Indian Institute of Science Education and Research, Pune, suggest enzymes may have persisted in the soil. Fontaine and Bouquet consider this very unlikely, as no enzyme is known to last six years.


