Chromatography Forum

I've been looking for details on this proceedure and can't find any. I understand that you'd run 50% nitric in water through your HPLC, but other than cleaning it of residues I don't know of other purposes. What does nitric pacification do?

Thanks

Because nitric acid is a strong oxidising acid, it will destroy residual organic material, dissolve some deposited anions and cations, and then passivate the stainless steel.

Stainless steel is essentially "stainless" because it forms a well-bonded, non-porous, oxide skin on exposure to air. The skin is self-healing in ambient air, whereas ordinary steel forms oxide films ( rust ) that flake away and are porous, thus allowing continual corrosion of the substrate metal.

Howver, when stainless is subjected to an environment that is oxygen-depleted, eg purified water, HPLC mobile phases, etc., the oxide film can break down, and further corrosion results in the formation of a dark "smut" along with a fresh stainless steel surface that is highly chemically reactive, and the smut can also detact and block filters.

Every time stainless steel surfaces are physically ( grinding, welding elctropolished etc.,) or chemically ( eg halides, haloacids etc ) attacked, the protective oxide film needs to be restored by oxidation.

Nitric acid passivation accelerates the oxide film formation, creating a stronger film that is much more resistant to aqueous and oxygen-deficient environments. I should mentain that other passivation treatments, eg citric acid, are also used, and some stainless steels, eg 400 series, are not easy to passivate.

It's really important to only use the nitric acid concentration, time and temperature specified by your instrument manufacturer, otherwise you could destroy seals etc.. 50% concentration is very common, but it depends on the stainless steels used ( probably 316, 316L, in HPLCs ). Nitric acid cleans, deionises, and oxidises the digestive system of your valuable toy.

Hope this helps,

Bruce Hamilton

A search of the forum may yield important additional information. This has been discussed several times, it should be over a year ago that some references (LC.GC etc.?) were presented on studies about this.

A couple of years ago, there was an article in LC/GC by John Dolan in his troubleshooting column.

There's also an old, but good, procedure for doing this on the Waters web site, in pdf form, I think.

Bruce, thank you for the clear explanation.
YM