Between April 2021 and August 2023, a toaster-sized machine sitting inside the Perseverance rover produced 122 grams of oxygen on the surface of Mars.

That is roughly enough oxygen to keep a small dog breathing for ten hours. It is not, by any standard, a large amount.

But it is the first oxygen ever produced on another planet. And the specific reason it matters is not what most people assume. It is not primarily about breathing. It is about getting back home.

What MOXIE actually did

The experiment was called MOXIE — the Mars Oxygen In-Situ Resource Utilization Experiment. It was designed by a team at MIT’s Haystack Observatory led by Michael Hecht, and built into the Perseverance rover before its 2020 launch to Mars.

The device weighed about 15 kilograms and measured roughly 24 by 24 by 31 centimetres. It drew about 300 watts of power while running, roughly the same as a small kitchen appliance.

Its job was simple in principle. Mars’s atmosphere is about 95 percent carbon dioxide. Each CO₂ molecule contains two oxygen atoms bonded to one carbon atom. MOXIE was built to break that bond and collect the oxygen.

The process required heat. Inside the device, a component called a solid oxide electrolyzer — built by a Utah company called OxEon Energy — heated Martian air to approximately 800 degrees Celsius. At that temperature, an electric current could strip one oxygen atom off each CO₂ molecule. The result was molecular oxygen on one side and carbon monoxide on the other.

The oxygen was measured for volume and purity, then vented back into the Martian atmosphere. Nothing was kept.

Across 16 runs between April 2021 and August 2023, MOXIE produced 122 grams of oxygen at purities of 98 percent or better. At peak efficiency, it reached 12 grams per hour — twice NASA’s original target.

Why oxygen was the specific bottleneck

Most of the popular framing of MOXIE treats it as an experiment about breathing. That framing understates why the technology matters.

Astronauts on Mars will not primarily need oxygen for their lungs. A properly designed habitat can recycle breathing air using systems similar to those already flying on the International Space Station.

What astronauts on Mars will absolutely need — in vastly larger quantities than they need for breathing — is oxygen for rocket propulsion. Every gram of rocket fuel burned during a launch from Mars back to Earth requires several grams of oxygen to actually combust. Without oxygen, the fuel does nothing.

Water on Mars is largely a solved problem. Ice at the polar caps and in subsurface deposits can be melted and purified. Fuel is also tractable — methane can be produced from Martian CO₂ and hydrogen through a well-established chemical process called the Sabatier reaction.

The specific missing piece has always been oxygen. Nobody had ever demonstrated that oxygen could be reliably produced on Mars from Martian resources. MOXIE closed that gap.

The scaling problem

Here is where the story becomes interesting for anyone thinking about actual Mars missions.

Michael Hecht has estimated that lifting four astronauts off the Martian surface and returning them to Earth orbit would require somewhere between 25 and 30 tons of oxygen. Not grams. Tons.

MOXIE produced 122 grams total across 16 runs. To meet the requirements of a real crewed return mission, the technology would need to be scaled up by a factor of roughly 250,000 and run continuously for months rather than in short experimental bursts.

That is a substantial engineering challenge. A production-scale MOXIE would need to be perhaps a hundred times larger and physically deployed to the Martian surface well before a crewed mission arrived. It would need to operate autonomously through Martian dust storms, temperature swings between day and night, and seasonal variations in atmospheric density.

None of this is impossible. The 2022 Science Advances paper describing MOXIE’s early performance notes that the underlying chemistry works. What remains is the specific engineering of building a system a quarter of a million times larger than the one that has just been proven to work.

What OxEon Energy is doing next

The company that built the SOXE electrolyzer at the heart of MOXIE has continued developing the technology.

OxEon Energy, based in Utah, has since won additional NASA contracts to develop larger versions of the electrolyzer stack. Their current work targets ground-based demonstration of systems substantially larger than MOXIE, with the specific goal of validating the technology at scales relevant to actual Mars missions.

The company has also begun exploring terrestrial applications. The same technology that splits CO₂ on Mars can, with modifications, be used for carbon capture on Earth or for producing high-purity oxygen for industrial applications.

NASA has not yet announced a formal follow-up mission to MOXIE. The specific next step for Mars oxygen production is currently unclear. What is clear is that the technology has been proven, the scaling requirements are understood, and the specific engineering challenge that remains is now well-defined.

What still needs to be solved

Several substantial problems remain between MOXIE’s success and a working Mars return mission.

Dust is one. Martian dust storms can last for months and can reduce atmospheric density and clarity substantially. A production-scale oxygen system would need to keep running through these events, or store enough oxygen in advance to bridge the gaps.

Reliability is another. MOXIE ran for approximately one hour per session across 16 sessions. A real system would need to run for months or years continuously with minimal maintenance, at temperatures around 800°C, in an environment that no maintenance crew can currently reach.

Integration is a third. Any Mars return vehicle would need to combine oxygen production, fuel production, storage, and rocket loading — all of which currently exist only as separate concepts.

None of this is a reason to doubt that Mars return missions will eventually happen. It is a reminder that MOXIE, however successful, was a specific first step in a much longer engineering programme.

The 122 grams of oxygen that MOXIE produced between 2021 and 2023 is real. The specific scientific milestone — that oxygen has now been made on another planet — is genuinely historic. And the specific work required to turn that milestone into a mission that actually brings astronauts home is now, for the first time, something engineers can plan for.

The chemistry works. The next question is whether the scale-up does.