The Svartsengi power station sits on the banks of the Blue Lagoon, an artificial geothermal spring and one of Iceland’s most popular tourist attractions. For decades, it has supplied Icelanders with geothermal electricity and heat. The rub is that extracting this renewable energy from the ground requires fossil fuels to run the pumps. So in 2011, an Icelandic energy startup called Carbon Recycling International built the George Olah plant, which captures Svartsengi’s CO2 emissions and turns them into a carbon-neutral fuel.

The idea for CO2 recycling was around long before the George Olah plant became the first to put it into practice. The idea is to take carbon dioxide emitted by power plants and use some chemical wizardry to turn it into useful fuels like propane or methane. Aside from CO2, the main ingredients in this process are hydrogen and a metallic catalyst. Cook it all together at high temperatures and voila: You’ve got yourself a tank of liquid hydrocarbon fuel. Although emissions from hydrocarbon fuels are exactly the problem this process is trying to solve, in principle capturing the emissions from the newly made fuels can create a closed loop. The world pumps out nearly 40 billion tons of CO2 each year, so converting even a small fraction of that into carbon-neutral fuel would be a win.

Yet Iceland’s George Olah plant remains the only facility converting emissions into fuel on an industrial scale. The problem is that the most efficient techniques require nanoparticle catalysts that are expensive to produce, which stalled the technology on the road from the lab to the real world. But a new process for cheaply minting CO2-loving nanoparticles developed by chemists at the University of Southern California and the National Renewable Energy Laboratory may nudge carbon recycling toward mainstream adoption. “The sustainable production of catalysts has been a major bottleneck,” says Noah Malmstadt, a chemical engineer at the University of Southern California. “Nanoparticle catalysts are very promising, and the ability to produce them sustainably at scale is something we’ve really pioneered.”

At the heart of the USC system are carbide nanoparticles, a generic term for compounds of carbon and another element–in this case a silvery metal called molybdenum. The nanoparticles are like a magnet to CO2 and kick-start the chemical reaction that turns emissions into fuel. “Molybdenum carbide is particularly interesting to us because it is relatively low cost and is uniquely suited to perform the multiple functions that are required to convert CO2 to fuel, like breaking the carbon-oxygen bonds,” says Frederick Baddour, a nanomaterials scientist at the National Renewable Energy Laboratory.

Malmstadt and his colleagues aren’t the first to use metal carbide nanoparticles to recycle CO2. But in the past, producing these nanoparticles meant baking them in reactors at around 1,100 degrees Fahrenheit. Reaching these temperatures was super energy-intensive. Even then, the size of the resulting particles was all over the place–which kills efficiency, because the chemical reaction initiated by the particles happens only on their surface. A good catalyst is one in which the surface area of all the particles is maximized, which is one of the main benefits of using nanoparticles.

The new system uses a millifluidic reactor, which operates at only 650 degrees Fahrenheit and forces the metal carbide feedstock through channels less than a millimeter wide. The result is nearly uniform metal carbide particles–literal carbon copies–that can be produced cheaply at scale. Malmstadt says the team has a paper under peer review that demonstrates their control of 16 of these reactors working in tandem. It’s not exactly industrial scale, but it demonstrates that the process can easily be scaled up without needing to build a larger device.