Toxic mud from aluminium production can be used to make greener steel

The huge reservoirs of hazardous red mud that are produced as part of refining aluminium could be used to make greener steel.

There are roughly 4 billion tonnes of red mud stored around the world, which is an environmental hazard that can lead to deadly accidents. And producing a tonne of steel generates nearly 2 tonnes of CO2 – that is because most steel production involves burning fossil fuels to react carbon with the oxygen in iron ore, yielding iron but also carbon dioxide.

Isnaldi Souza Filho at the Max Planck Institute for Iron Research in Germany and his colleagues have come up with a way to tackle both problems. They have devised a method to extract iron from red mud by exposing it to a plasma of hydrogen and argon, and then using this iron to produce steel.

Key to the process is that red mud contains between 30 and 60 per cent iron oxide by weight, alongside hazardous elements such as arsenic and lead.

The researchers heated up red mud in a device called an electric arc furnace to a temperature of roughly 1850°C (3362°F), with a blend of argon and hydrogen to react with the oxygen. The resulting melt was then cooled, crushed and separated into iron pellets ready to be turned into steel.

Co-author Matic Jovičević-Klug, also at the Max Planck Institute for Iron Research, says that given the amount of red mud there is, the process could produce between 748 million and 942 million tonnes of steel, which would result in over a billion tonnes less of CO2 compared with conventional methods. However, this scale would still only be a fraction of the steel produced globally each year.

Using green hydrogen as part of the refining process is not a new idea, says Mark Jacobson at Stanford University in California. In 2021, a Swedish consortium called HYBRIT demonstrated a trial run that lowered the carbon footprint of steelmaking by up to 98 per cent. What is new is using the hazardous red mud as a feedstock, he says. “Whether it is less expensive than the first process is difficult to tell,” says Jacobson. “The authors claim it is inexpensive, but more information is needed to determine this.”