For the first time in history, oxygen was produced on the surface of another planet — not by astronauts, but by a toaster-sized NASA device aboard the Perseverance rover, running quietly between 2021 and 2023 as a test for something much bigger: getting people home from Mars alive

Somewhere in Jezero Crater, on an April evening in 2021, a gold box roughly the size of a toaster exhaled.

Not much came out. About 5.4 grams of oxygen, which NASA’s Jet Propulsion Laboratory reckoned was enough to keep an astronaut going for roughly ten minutes of normal activity. No plume, no sound anyone could hear, nothing a camera could have caught. Just a machine breathing for the first hour of its life, a couple of hundred million kilometres from anybody who cared.

That box was MOXIE, the Mars Oxygen In-Situ Resource Utilization Experiment, and it had become the first device to manufacture a usable resource on the surface of another planet.

How you make oxygen out of Martian air

Mars has an atmosphere that is thin, freezing and about 95 per cent carbon dioxide. That sounds thoroughly hostile until you remember what carbon dioxide is made of. One carbon atom, two oxygen atoms. Break the bond and the oxygen is right there waiting.

MOXIE did the breaking with solid oxide electrolysis: heat the gas to around 800 degrees Celsius, then use electricity to strip a single oxygen atom off each CO2 molecule. Oxygen out one side, carbon monoxide out the other. A team at MIT’s Haystack Observatory, led by Michael Hecht, the mission’s principal investigator, spent years shrinking that idea into a box that could survive a launch, seven months in deep space and a landing.

Nothing was kept. Every run was measured for volume, purity and efficiency, then vented straight back into the sky it came from.

The heavy part of coming home is the oxygen

Rockets need fuel, and something for it to burn in, because combustion without oxygen goes nowhere. On the climb up from Mars there is no useful air to grab on the way.

Hecht laid the numbers out for MIT News: lifting four astronauts off the Martian surface would take something like seven tonnes of rocket fuel and twenty-five tonnes of oxygen. Breathing barely rates a mention beside that. A crew spending a year on the surface might get through a single tonne between them.

The oxygen problem, then, is really a launch problem. Twenty-five tonnes hauled from Earth means an enormous, expensive, precarious delivery mission arriving ahead of any crew. Twenty-five tonnes made on site means a smaller ship and a shorter list of things that can go catastrophically wrong.

Sixteen runs, every season, every hour of the day

“The atmosphere of Mars is far more variable than Earth.”

That was deputy principal investigator Jeffrey Hoffman, explaining why the team kept sabotaging the conditions on their own instrument. Air density on Mars can halve across the year. Temperatures swing by a hundred degrees. A device that hums along at noon in summer and quits at midnight in winter is a device nobody sane bets a crew’s return trip on.

So MOXIE ran in daylight and in darkness, across seasons, in the wake of dust storms. NASA’s own two-year progress notes log a peak rate of 10.56 grams an hour in November 2022, achieved by controlling voltage rather than current across the electrolysis cells, a switch made partly for safety at high output.

By the time it shut down on 7 August 2023, MOXIE had run sixteen times and produced 122 grams of oxygen all up. Per JPL’s closing summary, that is about what a small dog breathes in ten hours. Peak output reached 12 grams an hour at a purity of 98 per cent or better, twice what the agency had originally asked for.

One paper, seven runs, one careful claim

The peer-reviewed account of MOXIE’s early performance, published in Science Advances in 2022 by Hoffman, Hecht, Donald Rapp and a long list of colleagues, covers only the first seven runs through the end of 2021, and describes the instrument as the first demonstration of in-situ resource use on another planet.

One paper on one instrument. What it establishes is that the chemistry holds up against real Martian air rather than a simulation chamber in Massachusetts. That is a narrower and far more useful finding than anything involving the word colony.

A scaled-up successor, as sketched in the accompanying MIT write-up, would need to run non-stop at the oxygen output of several hundred trees, land on Mars years before any astronaut does, and keep working with the nearest technician a planet away. Hecht has told My News 13 that a practical descendant would be washing-machine-sized, not toaster-sized.

Nobody breathed a molecule of it

Every gram MOXIE made went back into the Martian atmosphere within hours. Unbreathed, unburned, unbanked. As a demonstration it succeeded beyond its own targets. As a supply of anything at all it was useless by design. That was always the point.

MOXIE completed every technical requirement NASA had set for it, then stopped. Speaking to My News 13, Hecht put a full-scale successor at five to ten years of engineering away, and said the real bottleneck was never the chemistry. The harder question, in his view, is when NASA or a company such as SpaceX will actually be ready to fly humans to Mars. “Now NASA is focusing on getting to the moon,” he said.

So the instrument sits inert in the belly of a rover, question answered and power off, while the machine it was a rough draft of stays unbuilt.