# Can molten aluminium be used to reduce iron oxide (rust) into iron?

Will surface rust on an iron crucible be converted if used to melt aluminium?

Can one convert bog/lake iron ore into iron using molten aluminium similar to the Thermite reaction?

Or perhaps the real question is what temperature would this occur at?

EDIT: I'm hoping one could use the molten aluminium instead of aluminium dust to achieve the same reaction as occurs in the Thermite process. I know about Thermite steel rail and copper earth connection welding. My goal is to reduce rust to iron by sprinkling it (with attendant sparks) onto a molten aluminium surface. Will the temperature be too low in practice or will the iron oxidise or blow off before consolidating? Trials today were inconclusive due to low temperature and small sample size.

• Short answer: you can Mar 6 '16 at 7:53
• Jun 4 '20 at 13:35

Take a look at the Ellingham Diagram.

The line for aluminium oxide is lower than the line for the all the iron oxides. This indicates greater stability of aluminium oxide, or in other words indicates aluminium's greater affinity for oxygen. Thus, based on thermodynamics alone one would expect aluminium to be able to reduce iron oxides.

Based on this figure, one would also expect this reaction to be spontaneous at all temperatures.

Important caveat, the ignition temperature remains critical. Just as a piece of paper will not combust unprovoked, even though the reaction is spontaneous, the thermite reaction needs to reach its ignition temperature.

Additionally, of interest to you would be this paper. Given below is a DSC curve from it

The endothermic peak ca. 660 corresponds to the melting of aluminium, while the exothermic peaks correspond to the thermite reaction.

The activation of the process is reported as 145 kJ/mol, and according to he authors in the study the product is controlled by the diffusion of $\ce{Al}$ in $\ce{FeAl2O4}$

• To the extent that this theory suggests the reaction to be spontaneous, it is woefully inadequate. The reaction is notorious for its high ignition temperature. Oct 12 '16 at 17:46
• @feetwet I agree with you completely. In fact, I plan to update my answer with some additional references :) Oct 12 '16 at 22:29
• If I am reading this correctly it implies that at a temperature somewhere around 850 to 950 degC we can expect the reaction to self start. If there are issues with diffusion it would also suggest to me that this would work best if the components were mixed before they were expected to react. Floating the iron oxide on aluminium might be expected to work at the interface though. I suppose there is hope but it is going to have very hot molten aluminium to get it going. Oct 17 '16 at 15:13
• Yes, I believe so. Although I would expect the ignition temperature to be > 1000 C...the information I could find showed some variance. However, I have no practical experience whatsoever so it may very well be lower. If you are carrying out the reaction, experiment with the temperature and see what results you get. I'd love to hear them myself :) Oct 17 '16 at 15:17

The trick to the thermite reaction of iron oxides with aluminum is the ignition temperature: somewhere over 1600 °C. Since aluminum melts at 660 °C you have to add (and sustain on contact) a lot of heat before it can begin to steal oxygen from rust.