Earth’s Underground Worlds May Run on Radioactive Decay

No sunlight, no problem.

An illustration of the radioactive splitting of water particles
Samantha Mash / Quanta Magazine

Scientists poke and prod at the fringes of habitability in pursuit of life’s limits. To that end, they have tunneled kilometers below Earth’s surface, drilling outward from the bottom of mine shafts and sinking boreholes deep into ocean sediments. To their surprise, “life was everywhere that we looked,” says Tori Hoehler, a chemist and astrobiologist at NASA’s Ames Research Center. And it was present in staggering quantities: By various estimates, the inhabited subsurface realm has twice the volume of the oceans and holds on the order of 10^30 cells, making it one of the biggest habitats on the planet, as well as one of the oldest and most diverse.

Researchers are still trying to understand how most of the life down there survives. Sunlight for photosynthesis cannot reach such depths, and the meager amount of organic carbon food that does is often quickly exhausted. Unlike communities of organisms that dwell near hydrothermal vents on the seafloor or within continental regions warmed by volcanic activity, ecosystems here generally can’t rely on the high-temperature processes that support some subsurface life independent of photosynthesis; these microbes must hang on in deep cold and darkness.

Two papers published in February by different research groups seem to have solved some of this mystery for cells beneath the continents and in deep marine sediments. They show evidence that, much as the sun’s nuclear-fusion reactions provide energy to the surface world, a different kind of nuclear process—radioactive decay—can sustain life deep below the surface. Radiation from unstable atoms in rocks can split water molecules into hydrogen and chemically reactive peroxides and radicals; some cells can use the hydrogen as fuel directly, while the remaining products turn minerals and other surrounding compounds into additional energy sources.

Although these radiolytic reactions yield energy far more slowly than the sun and underground thermal processes, the researchers have shown that they are fast enough to be key drivers of microbial activity in a broad range of settings—and that they are responsible for a diverse pool of organic molecules and other chemicals important to life. According to Jack Mustard, a planetary geologist at Brown University who was not involved in the new work, the radiolysis explanation has “opened up whole new vistas” into what life could look like, how it might have emerged on an early Earth, and where else in the universe it might one day be found.


Barbara Sherwood Lollar set off for higher education in 1981, four years after the discovery of life at the hydrothermal vents. As the child of two teachers who “fed me on a steady diet of Jules Verne,” she says, “all of this really spoke to the kid in me.” Not only was studying the deep subsurface a way to “understand a part of the planet that had never been seen before, a kind of life that we didn’t understand yet,” but it “clearly was going to trample [the] boundaries” between chemistry, biology, physics, and geology, allowing scientists to combine those fields in new and intriguing ways.

Throughout Sherwood Lollar’s training in the 1980s and her early career as a geologist at the University of Toronto in the ’90s, more and more subterranean microbial communities were uncovered. The enigma of what supported this life prompted some researchers to propose that there might be “a deep hydrogen-triggered biosphere” full of cells using hydrogen gas as an energy source. (Numerous microbes found in deep subsurface samples were enriched with genes for enzymes that could derive energy from hydrogen.) Many geological processes could plausibly produce that hydrogen, but the best-studied ones occurred only at high temperatures and pressures. These included interactions between volcanic gases, the breakdown of particular minerals in the presence of water, and serpentinization—the chemical alteration of certain kinds of crustal rock through reactions with water.

By the early 2000s, Sherwood Lollar, Li-Hung Lin (now at National Taiwan University), Tullis Onstott of Princeton University, and their colleagues were finding high concentrations of hydrogen—“in some cases, stunningly high,” Sherwood Lollar says—in water isolated from deep beneath the South African and Canadian crust. But serpentinization couldn’t explain it: The kinds of minerals needed often weren’t present. Nor did the other processes seem likely, because of the absence of recent volcanic activity and magma flows.

“So we began to look and expand our understanding of hydrogen-producing reactions and their relationship to the chemistry and mineralogy of the rocks in these places,” Sherwood Lollar says.

A clue came from their discovery that the water trapped in those rocky places held not just large amounts of hydrogen but also helium—an indicator that particles from the radioactive decay of elements such as uranium and thorium were splitting water molecules. That process, water radiolysis, was first observed in Marie Curie’s laboratory at the beginning of the 20th century, when researchers realized that solutions of radium salts generated bubbles of hydrogen and oxygen. Curie called it “an electrolysis without electrodes.” (It took a few more years for scientists to realize that the oxygen came from hydrogen peroxide created during the process.)

Sherwood Lollar, Lin, Onstott, and their collaborators proposed in 2006 that the microbial communities under South Africa and Canada derived the energy for their survival from hydrogen produced through radiolysis. So began their long quest to unpack how important radiolysis might be to life in natural settings.


For much of the next decade, the researchers obtained samples from deep aquifers at various mining sites and related the complex chemistries of the fluids to their geological surroundings. Some of the water trapped beneath the Canadian crust had been isolated from the surface for more than 1 billion years—perhaps even for 2 billion. Within that water were bacteria, still very much alive.

“That had to be a completely self-sustained system,” Mustard says. By the process of elimination, radiolysis looked like a possible energy source, but could there be enough of it to support life?

In 2014, when Sherwood Lollar and her colleagues combined the results of nuclear chemists’ lab work with models of the crust’s mineral composition, they discovered that radiolysis and other processes were likely to be producing a huge amount of hydrogen in the continental subsurface—on par with the amount of hydrogen thought to arise from hydrothermal and other deep-sea environments. “We doubled the estimate of hydrogen production from water-rock reactions on the planet,” Sherwood Lollar says.

Microbes could directly utilize the hydrogen produced by radiolysis, but that was only half the story: To make full use of it, they needed not just hydrogen as an electron donor, but another substance as an electron acceptor. The scientists suspected the microbes were finding that in compounds made when the hydrogen peroxide and other oxygen-containing radicals from radiolysis reacted with surrounding minerals. In work published in 2016, they showed that radiolytic hydrogen peroxide was likely interacting with sulfides in the walls of a Canadian mine to produce sulfate, an electron acceptor. But Sherwood Lollar and her colleagues still needed proof that cells were relying on that sulfate for energy.

In 2019, they finally got it. By culturing bacteria from the groundwater in mines, they were able to show that the microbes made use of both the hydrogen and the sulfate. Water, some radioactive decay, a bit of sulfide—“and then you get a sustained system of energy production that can last for billions of years … like an ambient pulse of habitability,” says Jesse Tarnas, a planetary scientist and NASA postdoctoral fellow.

In their February paper, Sherwood Lollar and her colleagues showed that radiolysis is instrumental not just in the hydrogen and sulfur cycles on Earth, but in the cycle most closely associated with life: that of carbon. Analyses of water samples from the same Canadian mine showed very high concentrations of acetate and formate, organic compounds that can support bacterial life. Moreover, measurements of isotopic signatures indicated that the compounds were being generated abiotically. The researchers hypothesized that radiolytic products were reacting with dissolved carbonate minerals from the rock to produce the large quantities of carbon-based molecules that they were observing.

To cement their hypothesis, Sherwood Lollar’s team needed additional evidence. It arrived just one month later. Nuclear chemists led by Laurent Truche, a geochemist at Grenoble Alpes University in France, and Johan Vandenborre of the University of Nantes had been independently studying radiolysis in laboratory settings. In work published in March, they pinned down the precise mechanisms and yields of radiolysis in the presence of dissolved carbonate. They measured exact concentrations of various byproducts, including formate and acetate—and the quantities and rates they recorded aligned with what Sherwood Lollar was seeing in the deep fractures within natural rock.


While Sherwood Lollar was conducting her field research within the continental subsurface, a handful of scientists was trying to suss out the effects of radiolysis beneath the seafloor. Chief among them was Steve D’Hondt, a geomicrobiologist at the University of Rhode Island, who in February with his graduate student Justine Sauvage and their colleagues published the results of nearly two decades’ worth of detailed evidence that radiolysis is important for sustaining marine subsurface life.

In 2010, D’Hondt and Fumio Inagaki, a geomicrobiologist at the Japan Agency for Marine-Earth Science and Technology, led a drilling expedition to collect samples of sub-seafloor sediments from around the globe. Subsequently, D’Hondt and Sauvage suspended dozens of sediment types in water and exposed them to different types of radiation—and every time, they found that the amount of hydrogen produced was much greater than when pure water was irradiated. The sediments were amplifying the products of radiolysis. And “the yields were ridiculous,” D’Hondt says. In some cases, the presence of sediment in the water increased the production of hydrogen by a factor of nearly 30.

“Some minerals are just hotbeds of radiolytic-hydrogen production,” D’Hondt says. “They very efficiently convert the energy of radiation into chemical energy that microbes can eat.”

Yet D’Hondt and his colleagues found barely any hydrogen in the sediment cores they’d drilled. “Whatever hydrogen is being produced is disappearing,” D’Hondt says. The researchers think it’s being consumed by the microbes living in the sediments.

According to their models, in deep sediments more than a few million years old, radiolytic hydrogen is being produced and consumed more quickly than organic matter is—making radiolysis of water the dominant source of energy in those older sediments. While it accounts for only 1 percent to 2 percent of the total energy available in the global marine-sediment environment—the other 98 percent comes from organic carbon, which is mostly consumed when the sediment is young—its effects are still quite sizable. “It might be slow,” says Doug LaRowe, a planetary scientist at the University of Southern California, “but from a geologic perspective, and over geologic time … it starts to add up.”

This means that radiolysis “is a fundamental source of bioavailable energy for a significant microbiome on Earth,” Sauvage says—not just on the continents but beneath the oceans too. “It’s quite striking.”


The newfound scientific importance of radiolysis may not relate just to how it sustains life in extreme environments. It could also illuminate how abiotic organic synthesis may have set the stage for the origin of life—on Earth and elsewhere.

Sherwood Lollar has been invigorated by her team’s recent observations that, in the closed environmental system around the Canadian mines, most of the carbon-containing compounds seem to have been produced abiotically. “It’s one of the few places on the planet where the smear of life hasn’t contaminated everything,” she says. “And those are pretty rare and precious places on our planet.”

Part of their unique value is that they can be “an analogue for what might have been the prebiotic soup that our Earth might have had before life arose,” she says. Even if life didn’t arise in this kind of subsurface environment—higher-energy regions of the planet, such as hydrothermal vents, are still more probable venues for an origin story—it provided a safe place where life could be sustained for long stretches of time, far away from the dangers found at the surface (such as the meteor impacts and high levels of radiation that plagued the early Earth).

Modeling and experimental work have shown that even simple systems (consisting solely of hydrogen, carbon dioxide, and sulfate, for example) can lead to extremely intricate microbial food webs; adding compounds such as formate and acetate from radiolysis to the mix could significantly broaden the potential ecological landscape. And because acetate and formate can form more complex organics, they can give rise to even more diverse systems. “It’s important to see life operating with this amount of complexity,” says Cara Magnabosco, a geobiologist at the Swiss Federal Institute of Technology, in Zurich, “even in something that maybe you would view as very simple and very energy-poor.”

“Let’s say [radiolysis] can only make basic organic carbons, like formate and acetate,” LaRowe says. “If you move those compounds into a different environmental setting, perhaps they can react there to form something else. They become starter or feeder material for more complex reactions in a different setting.” That might even help bring scientists closer to understanding how amino acids and other important building blocks of life arose.

Sherwood Lollar is now collaborating with other scientists, including colleagues at the CIFAR Earth 4D project, to study how the organic molecules present in the ancient Canadian water might “complexify” the chemistry at hand. In work they’re hoping to publish later this year, “we show how the coevolution of organics and minerals is key for the diversification of these organic compounds,” says Bénédicte Menez, a geobiologist at the Paris Institute of Earth Physics and one of the leaders of the research. Her aim is to determine how more complicated organic structures could form and subsequently play a role in some of the earliest microbial metabolisms.

Astrobiologists are also realizing how crucial it might be to consider radiolysis when constraining the habitability of planets and moons throughout the solar system and the rest of the galaxy. Sunlight, high temperatures, and other conditions might not be strictly needed to sustain extraterrestrial life. Radiolysis should be practically ubiquitous on any rocky planet that has water in its subsurface.

Take Mars. In a pair of studies, one published a couple of years ago and the other last month, Tarnas, Mustard, Sherwood Lollar, and other researchers translated quantitative work being done on radiolysis on Earth to the Martian subsurface. They found that based on the planet’s mineral composition and other parameters, Mars today might be able to sustain microbial ecosystems akin to those on Earth—with radiolysis alone. The scientists identified regions of the planet where the microbial concentration would likely be greatest, which could guide where future missions should be targeted.

“It’s really fascinating to me,” Inagaki says, “as we are now in an era where particle physics is necessary to study microbial life in Earth’s planetary interior and other worlds in the universe.”