Scientists in Cambridge, UK, have developed a reactor that can make oxygen from Moon rock -- a vital technology if plans to create a lunar base are to take off.

Whether tapping the Moon's resources or using the satellite as a jump-off point to explore the deeper reaches of space, occupants of any future lunar base will need oxygen to survive. Ferrying huge amounts of it to the Moon would be extremely expensive -- perhaps costing as much as US$100 million per tonne according to some estimates -- so researchers are examining potentially cheaper ways to produce oxygen on the Moon itself.

NASA has been looking for ways to get oxygen from Moon rock for several years. In 2005, as part of its Centennial Challenges programme, the agency offered a US$250,000 prize to the first team to come up with a piece of kit that could extract five kilograms of oxygen in eight hours from some simulated Moon rock. Despite raising the value of the prize pot to $1 million in 2008 with the help of the California Space Authority, the prize remains unclaimed. In addition, the agency's ongoing In Situ Resource Utilization programme is currently looking at several different technologies for extracting oxygen from Moon rock.

Now, Derek Fray, a materials chemist from the University of Cambridge, UK, and his colleagues have come up with a potential solution by modifying an electrochemical process they invented in 2000 to get metals and alloys from metal oxides1. The process uses the oxides -- also found in Moon rocks -- as a cathode, together with an anode made of carbon. To get the current flowing through the system, the electrodes sit in an electrolyte solution of molten calcium chloride (CaCl2), a common salt with a melting point of almost 800 °C.

Eroding anode

The current strips the metal oxide pellets of oxygen atoms, which are ionized and dissolve in the molten salt. The negatively charged oxygen ions move through the molten salt to the anode where they give up their extra electrons and react with the carbon to produce carbon dioxide -- a process that erodes the anode. Meanwhile, pure metal is formed over at the cathode.

To make the system produce oxygen and not carbon dioxide, Fray had to make an unreactive anode. This was crucial: "without those anodes, it doesn't work", says Fray. He discovered that calcium titanate, which is a poor electrical conductor on its own, became a much better conductor when he added some calcium ruthenate to it. This mixture produced an anode that barely erodes at all -- after running the reactor for 150 hours, Fray calculated that the anode would wear away by roughly three centimetres a year.

In their tests, Fray and his colleagues used a simulated lunar rock called JSC-1, developed by NASA. Fray anticipates that three reactors, each a metre high, would be enough to generate a tonne of oxygen per year on the Moon. Three tonnes of rock are needed to produce each tonne of oxygen, and in tests the team saw almost 100% recovery of oxygen, he says. Fray presented the results last week at the Congress of the International Union of Pure and Applied Chemistry in Glasgow, UK.

To heat the reactor on the Moon would need just a small amount of power, Fray notes, and the reactor itself can be thermally insulated to lock heat in. "It won't be a problem," he says. The three reactors would need about 4.5 kilowatts of power -- not much more than that used to heat an immersion heater in a domestic boiler -- which could be supplied by solar panels or even a small nuclear reactor placed on the Moon.

With an extra £10 million (US$16.5 million), Fray says he would be able to develop "a robust prototype" of a bigger reactor that could be operated remotely. He is currently working with the European Space Agency towards this goal.


A similar technique for oxygen extraction is being developed by Donald Sadoway at the Massachusetts Institute of Technology in Cambridge, Massachusetts, but his process works at a much higher temperature of up to 1600 °C -- which means that the Moon rock is molten and can act as the electrolyte itself. It produces molten metal, including iron, which sinks to the bottom. Fray says that his process is more efficient because it works at a lower temperature, but Sadoway insists that molten salt electrolysis, as his technique is called, makes up elsewhere for the extra heat it needs. "In Derek's process, the molten salt allows him to operate at a much lower temperature," says Sadoway, "but he still has to consolidate the Moon rock into a solid form." This is often difficult because of the fine sandy nature of Moon rock, he says.

Sadoway's reactor could even build itself. The interior would be Moon regolith -- the powdery rubble that forms the Moon's surface -- heated electrically to become molten, and the exterior would be solid regolith that has cooled. "We form the wall of the reactor by allowing the molten regolith to freeze," he says, but admits that starting the process is "tricky".

Sadoway says that with sufficient funding, he could have his system scaled up within two years. His process has been shortlisted by NASA and is receiving some funding from the agency. "Once we solve the materials problems at the lab scale we should be able to move quickly," he says.