Probably the best products to remove rust from steel parts and panels contain phosphoric acid. These are products like Naval Jelly (a phosphoric acid based gel) or Metalprep (a solution of phosphoric acid).
There are two main reasons for using phosphoric acid for rust removal:
1) It dissolves rust at a much faster rate than it dissolves the surrounding iron.
2) It leaves a nice iron phosphate coating on the clean metal surface, which affords some protection from further rusting.
The reactions are:
fast
Fe2O3 + 2 H3PO4 ———> 2 FePO4 + 3 H2O (1)
slow
2 Fe + 2 H3PO4 ———> 2 FePO4 + 3 H2 (gas) (2)
In reaction 1, rust (Fe2O3) gets turned into iron phosphate and water, which mostly gets washed away when you rinse the part or panel. The phosphate part of phosphoric acid is responsible for this reaction.
Reaction 2 is the reason you may see some bubbles. The iron itself is actually dissolving, but this is a relatively slow reaction. The H+ ions from the phosphoric acid are responsible for this reaction. You won’t lose any worthwhile amount of metal from this reaction, but this is the reason why you don’t leave the acid on the part for more than about 15 minutes.
The layer of FePO4 that is left on the surface adheres strongly enough due to surface effects that much of it does not wash away with the rinse, hence the good protection from further rusting.
Hydrochloric (“muriatic”) acid is also sometimes used for derusting and gives the two reactions shown below. In this case both reactions are relatively fast, so you lose significant metal when derusting with hydrochloric acid. Also, the FeCl3 produced does not stick to the metal surface, so there is no protection of the clean metal surface. As if that weren’t enough reason not to use hydrochloric acid, the Cl- ions present actually promote rust formation!
fast
Fe2O3 + 6 HCl ———> 2 FeCl3 + 3 H2O (3)
fast
2 Fe + 6 HCl ———> 2 FeCl3 + 3 H2 (gas) (4)
The reason that the iron oxide (rust) does not provide surface protection like the phosphate does is that the oxide flakes off, continually exposing new metal to the oxidation process. On aluminum parts, the oxide does stick and is the reason why “bare” aluminum is so stable (pure aluminum metal is actually more reactive that pure iron.) This is also the reason that antifreeze solutions that contain phosphates should not be used in engines with aluminum heads, i.e., the phosphates break down the protective aluminum oxide coating on the parts. Surface oxide on iron is the bad stuff, but surface oxide on aluminum is the good stuff, see?