Thursday, 20 November 2014 13:56

Making Oxygen on the Moon

    MIT Professor Donald R. Sadoway shows molten oxide electrolysis can produce life-supporting oxygen.

    Mat Day Sadoway Web 1181
    Molten oxide electrolysis is a platform for using electricity to make carbon-free metals, including iron and titanium, in a single-step, says Donald R. Sadoway, the John F. Elliott Professor of Materials Chemistry in the Department of Materials Science and Engineering at MIT.

    Techniques of molten oxide electrolysis (MOE) can stretch from refining structural metals on earth to producing oxygen to support life on the moon, MIT Professor Donald R. Sadoway says. That's because the same process that extracts iron and other metals from their metal oxides releases oxygen as a byproduct.

    There is sufficient metallic oxide content in the lunar surface to use molten oxide electrolysis to make oxygen from iron and sodium oxides, as well as from more plentiful aluminosilicates, oxygen-rich solid compounds of aluminum and silicon. "It doesn't matter where you go on the moon, you'll be able to use MOE and make oxygen," says Sadoway, who is John F. Elliott Professor of Materials Chemistry in the Department of Materials Science and Engineering at MIT. Sadoway spoke Oct. 21, 2014, at the Materials Day Symposium, "New Frontiers in Metals Processing," hosted by the Materials Processing Center at MIT.

    Sadoway and MIT colleagues have been studying molten oxide electrolysis techniques for separating a variety of metals, from iron to titanium for carbon-free production in a single step. His team has made carbon-free nickel, iron and ferrochromium, among others. "What I hope I've shown is this idea is really a platform," Sadoway says.

    Lunar Oxygen

    Since people need to breathe in several pounds of oxygen per day to sustain life, "if you're going to put people on the moon, you need oxygen; you need it for human life," Sadoway says.

    Using a meteor sample from Arizona as a proxy, Sadoway determined the approximate percentages of metal oxides in lunar surface materials and modeled their efficiency at yielding oxygen. "You get the cheapest oxygen from the sodium (oxide, Na2O) and the iron (oxides, Fe2O3 or FeO), then there is a 1.5-volt silicon reduction peak, and you get pretty good utilization depending on what your operating temperature is," he says.

    Oxygen is also needed as a rocket propellant, so being able to make oxygen on the moon could provide oxygen for the return trip and lead to its use as a base for future missions to Mars. "If you could generate oxygen on the moon, you only need a one way shipment, and you can come back," Sadoway says.

    Sadoway's research shows that a yield of about 300 grams of oxygen per kilogram of lunar surface material, which is called regolith, is possible from silicon dioxide reduction at 1,600 degrees Celsius. Despite its high cost, an iridium anode would be the best solution for molten oxide electrolysis on the moon, he suggests. "If you're a quarter million miles from home and you need six pounds of oxygen, you'll pay anything for it," he says.

    Mat Day Sadoway MOE Cell Web
    Diagram of molten oxide electrolysis cell. Molten oxide electrolysis is a platform for carbon-free production of metals such as iron and titanium fueled by electricity, says Donald R. Sadoway, the John F. Elliott Professor of Materials Chemistry in the Department of Materials Science and Engineering at MIT. Image courtesy, Don Sadoway.

    Making Cleaner Metals

    Sadoway is motivated to develop cleaner methods of producing steel as regulators focus on carbon dioxide and its contribution to climate change and global warming. Sadoway, who joined the MIT faculty in 1978, showed a New Yorker cartoon from that era showing two men in suits standing at a window and the caption, "Where there's smoke, there's money." The window overlooks a factory building and storage tanks with multiple chimneys and vent pipes belching dark fumes. "It was funny in its time," Sadoway says.

    Steelmaking releases carbon dioxide (CO2) as a byproduct, and production is rising at about 8 percent per year, overtaking concrete as the top man-made source of carbon dioxide, Sadoway says. Manufacturers worldwide produce more than 1.8 billion tons of steel per year, excluding recycling, and it takes about a half-ton of carbon to make a ton of steel, he explains. Aluminum, at production of about 70 million tons per year, also takes a half-ton of carbon to make a ton of aluminum.

    Charles Martin Hall in the United States and Paul Héroult in France independently discovered the process of refining aluminum by electrolysis in 1886. Inspired by the electrolysis process for aluminum, Sadoway and MIT colleagues, including Thomas B. King Assistant Professor of Metallurgy Antoine Allanore, have been working to develop molten oxide electrolysis techniques for refining iron, steel and other metals with the dual aim of reducing environmental impact and lowering capital and operating costs.

    A molten oxide electrolysis cell consists of a crucible with an anode at the top, a cathode at the bottom, which has a liquid pool, and a molten oxide electrolyte in between. Electric current activates the process. "You feed metal oxide into the cell and by the action of electric current, you break...the metal oxide into liquid metal which pools at the bottom. In the most optimistic case, when you have a carbon-free anode, you make oxygen gas which bubbles off the top," Sadoway explains.

    Unlike traditional production methods, which consume fossil fuels and produce carbon as a byproduct, molten oxide electrolysis can produce carbon-free virgin stainless steel. Since stainless steel requires a certain amount of carbon to be integrated into it, "you'd actually dope this just up to the level of carbon that you needed, so you'd have low oxygen, low nitrogen and so on," Sadoway explains.

    Although laboratory scale experiments require heating the crucible from the outside, larger scale versions of molten oxide electrolysis will be self-heated by the Joule effect once the process is running, Sadoway says.

    Sadoway and Allanore reported a new chromium-iron anode material in Nature in May 2013. "What you've got is metal oxide forming on metal, and
    it's a ferro-chromium alloy and it forms an iron oxide, chromium oxide on it, that's thin enough to allow current to flow without resistance penalty, but thick enough to prevent chemical attack," Sadoway explains. Experiments demonstrated the molten electrolysis cell with the new anode material produced oxygen as well as nearly pure iron metal for two and a half hours running at 1,560 degrees Celsius.

    Sadoway, Allanore and Jim Yurko (Ph.D., 2001) started Boston ElectroMetallurgical Corp., which is based in Woburn, Mass., to commercialize the technology. "There are about five people, a 3,000-square-foot facility, and we're working on something at pilot scale, hopefully in about a year's time," he says.

    Economical Method

    Molten oxide electrolysis offers an economical alternative to traditional steelmaking processes, Sadoway argues. "The capital cost of this is much, much lower than a greenfield steelmaking operation. And that means smaller minimum capacity, so instead of building a greenfield steelworks at several million tons, you could imagine doing so at the order of several hundred thousand tons, and finally, direct processing of oxide ore, which means that you can use the existing supply chain, but you don't have to have all the other unit operations," he says. MOE can take inputs from fine metal powders to large chunks of metal and make a marketable product.

    "As the price of electricity falls, you discover that in fact this is a more efficient use of energy," Sadoway says. "Ultimately you make two-thirds of a ton of oxygen every time you make a ton of iron. Instead of being a headache, it becomes another product."

    Future Challenges

    Although Sadoway demonstrated molten oxide electrolysis of iron and nickel, some other metals, such as rare earths are harder to produce purely, and electronic conductivity also is an issue with transition metals. "The molten oxides of transition metals exhibit electronic conduction, so if they're electronically conductive, you'll have shorting between the electrodes and your Faradaic efficiency goes way down," Sadoway says.

    "Then there is this materials problem of the inert anode that can pass current at about 1 amp per square centimeter, be immersed in a molten oxide as an electrolyte, and stand up to evolving oxygen. I can give you materials that can handle any two of those three, but to get all three, that's a special challenge, and 1,700 degrees C is, that's a high-temperature," Sadoway says.

    Still, Sadoway and colleagues pushed their MOE cell to 1,725 degrees Celsius to make titanium from its oxide. "We had a carbon anode, and the carbon converted to carbon monoxide, so you're making fuel at this electrode, and the electrolyte which contained TiO2, and optimistically, we're making titanium liquid in one step," he says. That work used a molybdenum crucible.

    Sadoway is working on MOE production of rare earths from their oxides. "It's easy to go from dirt to metal in an all oxide electrolysis, but I'm not going to misrepresent the situation to you and suggest that you can separate different rare earths by this process," he says. However, MOE will easily produce mixed rare earth mischmetal in one step, without fluorides and their accompanying pollution. For example, a mix of the rare earth elements cerium and lanthanum can be produced for applications such as lighter igniters where complete separation isn't required.

    back to newsletterWritten by Denis Paiste, Materials Processing Center