Passivation: No Dimensional Change, Nitric vs Citric, A967

Passivation is a chemical treatment that removes free iron from the surface of a stainless steel part so the alloy’s own chromium can rebuild its protective oxide layer. Nothing is added to the part and no measurable base metal comes off: dimensions change by less than 1 µm, so fits, threads, and bores come out of the bath exactly as machined. The treatment is for stainless steel only, and it exists because CNC machining, forming, and handling smear foreign iron onto a surface whose corrosion resistance depends on staying chromium rich. Within surface finishing, it is the lowest footprint option: where a coating adds a layer and a polishing process removes one, passivation only cleans and re oxidizes.

The mechanism explains the rest of this page, including what the process cannot fix. Stainless steel resists rust because it holds at least about 10.5% chromium. Exposed to oxygen, a clean stainless surface grows a chromium oxide film a few nanometers thick, thin enough to be invisible. That film, the passive layer, is what stands between the iron in the alloy and the environment. Shop work damages it in a specific way: steel tooling, grinding swarf, wire brushes, and ordinary shop dust deposit loose particles of plain iron on the surface. Those particles rust, and the rust spreads into the surrounding stainless. Acid passivation dissolves that embedded iron while leaving the alloying chromium in place, and the film regrows on the cleaned surface. The glossary entry for the passive layer covers the term in more depth.

How passivation works

A passivation line is a cleaning sequence first and an acid bath second. The part is degreased and cleaned, because oil, coolant, and chips block the acid from reaching the iron it is meant to dissolve. It then soaks in a dilute acid bath, nitric or citric, where the acid selectively dissolves free iron and any iron based contamination on the surface. The part is rinsed and dried. On exposure to air, the passive film reforms over a surface that is now genuinely stainless rather than stainless spotted with plain iron.

Two properties of this mechanism drive the decisions around it.

First, the process is selective. The bath chemistry attacks exposed free iron far faster than it attacks the alloy, which is why a correctly run treatment strips contamination without etching the part. When that selectivity breaks down, through chloride contaminated baths, excessive time, or excessive temperature, the surface can etch and darken instead. Shops call that failure flash attack, and it is the main quality risk of the process.

Second, the passive layer is self healing. Scratch it, and the exposed chromium re oxidizes on its own in air or aerated water. Passivation is not a coating that wears out; it removes what blocks a reaction the alloy performs by itself. That is also why the treatment is defined by what it removes rather than what it applies.

What passivation will not fix

The most common specification error is ordering passivation to repair a surface defect it cannot touch. Passivation removes microscopic iron, not visible damage. It will not remove rust stains and pitting, oxide scale from heat treatment, or the rainbow colored heat tint that welding leaves along a bead. Heat tint is chromium depleted oxide sitting on top of the metal, and mild passivation acid does not dissolve it. Those conditions need cleaning or pickling first. Pickling uses a stronger acid blend, typically nitric plus hydrofluoric, to strip the damaged surface layer, and ASTM A380 governs that cleaning and descaling work before passivation is attempted. A part that is descaled and then passivated ends up with both problems solved. A part that is only passivated keeps its weld tint and the corrosion cell that comes with it.

The same limit applies to cosmetic damage. Scratches, tool marks, and a rough machined finish remain after passivation, because the process neither adds material nor removes enough to level anything. For parts that need corrosion resistance and a smoother surface, electropolishing does both in one step and is accepted as a passivation route in its own right under ASTM B912; that alternative is linked in full below.

The passivation process, step by step

A production passivation cycle has five stages, and the first two decide whether the last one passes.

1. Degrease and clean. Oils, cutting fluid, and chips come off in alkaline cleaner or solvent. Acid cannot work through oil, and skipping this step is the most common cause of failed treatment.
2. Remove scale and heat tint where present. Welded or heat treated parts are pickled or mechanically cleaned per ASTM A380 before any passivation bath.
3. Acid immersion. The part soaks in a nitric or citric bath at a concentration, temperature, and time matched to the grade and the treatment family, typically minutes at temperature rather than hours.
4. Rinse and dry. Thorough rinsing stops the reaction and removes residual acid; the rinse water should be low in chlorides.
5. Test. The lot is checked against an acceptance test to confirm free iron is gone.

The immersion step is where the two treatment chemistries diverge. ASTM A967 recognizes both. It is the specification that defines chemical passivation treatments for stainless parts, along with the tests that accept or reject them.

Nitric and citric treatments

Nitric acid is the traditional bath and still the default in aerospace supply chains, where AMS 2700 governs and some customers accept nothing else. Common nitric treatments run roughly 20 to 25% acid at moderate temperature for around 30 minutes on the 300 series. Stronger and dichromate added variants serve precipitation hardening and free machining grades, and some treatments run up to about 50% acid. Nitric is aggressive: it emits corrosive fumes, needs dedicated ventilation and handling controls, and its dichromate variants carry hexavalent chromium, which is regulated and hazardous.

Citric acid passivation is the newer family, adopted for safety and disposal reasons and now standard in medical and much general work. Treatments run about 4 to 10% citric by weight: hotter baths finish in a few minutes, room temperature baths need tens of minutes. Citric is biodegradable, does not emit the fumes nitric does, and passes most 300 series parts without complications. Modern formulations have reduced its old weaknesses, bath growth and variability. Still, nitric holds the longer acceptance history with the most conservative customers.

For the buyer, the choice is usually settled by three inputs: what the customer specification allows, what the grade needs, and what the processor is equipped to run well. Both routes, run correctly and verified by test, produce a passive surface.

Specifying it on a drawing

A passivation note is only as good as what it names. The complete callout states the governing standard, the treatment family, the acceptance test, and the sequence. One full note reads: passivate per ASTM A967, citric treatment, verify by water immersion test, after heat treatment and final machining. The sequence clause matters on any heat treated or hardened part, because passivating before final machining simply recontaminates the surface the treatment cleaned. Standards context and certification practice are covered under material certifications and standards, and the inspection side of acceptance testing under quality inspection and metrology.

Stainless grades: 304, 316, and 17-4

All stainless grades passivate by the same mechanism, but they do not respond identically, and the grade on the drawing should steer the treatment. The comparison table above summarizes the three grades that dominate machined stainless work, and stainless steel as a material covers the compositions in full.

304 is the baseline. It carries about 18% chromium and 8% nickel, passivates readily in standard nitric or citric cycles, and its main risk is the ordinary one: embedded iron from tooling and handling. 316 is 304 plus 2 to 3% molybdenum, which is what buys its resistance to chlorides in marine, pharmaceutical, and food washdown service. Its passivation behavior matches 304, but the stakes of skipping it are higher, because parts are specified in 316 precisely for the environments where free iron rust becomes a failure.

17-4 PH is the different case. It is a precipitation hardening grade, heat treated to high strength, and it passivates less easily than the austenitic grades. The traditional treatment is a dichromate added nitric bath, citric cycles are now common, and under treatment is a known failure mode. Sequence matters as much as chemistry: machine the part in the annealed condition, age it, then finish machine and passivate last, so the treatment follows the final cut. At high hardness, acid exposure raises a hydrogen embrittlement concern, and hardened PH parts commonly get a relief bake after chemical processing per the governing material spec. In chloride service, even a well passivated 17-4 part pits before a 316 part would, a limit no surface treatment removes. Grade specific machining practice is covered under stainless steel CNC machining.

Two other families need a caution. Free machining grades such as 303 and 416 add sulfur or selenium inclusions that improve machinability but complicate passivation: the acid attacks the inclusions and can leave a degraded surface, so these grades need treatments specifically rated for them. Some high carbon martensitic grades, such as 440C, are not considered suitable for acid passivation at all.

Dimensional effects: the zero change advantage

Every other surface process on a stainless part changes its size. This one does not, and that single fact settles a lot of finishing decisions. Passivation deposits nothing and removes no measurable metal: total dimensional movement stays below 1 µm, which is beneath the resolution of most shop measurement and far beneath any tolerance a machined part carries.

Compare that with the coatings it competes with for corrosion protection. Powder coating adds 60 to 120 µm to every coated surface, which closes clearances, changes thread fits, and can push a machined-to-size feature out of tolerance. Plating builds a deposited layer measured in micrometers to tens of micrometers, calls for allowance on every critical dimension, and brings hydrogen considerations of its own. Anodizing consumes and rebuilds each aluminum surface by 10 to 15 µm, which is why anodized parts mask or re machine tight fits. Passivation needs none of that management. There is no coating thickness to budget, no masking of threads and bores, no growth toward an interference fit, and no re machining after the bath. A part that measured in spec before passivation measures in spec after it.

The consequences run in two directions. Passivation can be specified late, on a finished part, without redesigning anything around it. It is also the finishing step that leaves the tolerance stack exactly as machined, which matters most on precision features: sealing bores, bearing seats, gasket faces, and threaded connections hold their machined dimensions through the bath. The trade is that passivation buys corrosion resistance only. It adds no wear layer, no color, and no smoothing. The surface finish the part had going in, rough or smooth, is the finish it has coming out, so parts that need a specific roughness get there by machining, bead blasting, or polishing first, and the roughness side of that decision is covered under surface finish and roughness.

Take a machined 316 manifold block with a sealed bore tolerance of a few micrometers. Sent to powder coat, the bore would need masking or recutting to hold its fit. Sent through passivation, the whole part including the bore goes in the bath and comes out at machined size, with the free iron from cutting removed and the molybdenum bearing alloy doing its job in a washdown environment.

Worked examples

The two examples below show the decision as it lands on real parts.

Example: 316 food equipment bracket

A formed and machined 316 bracket for a washdown line needs corrosion protection after fabrication, and the customer specification names ASTM A967. The part carries two tapped mounting holes and a dowel pin hole on a tight center distance. Because passivation changes no dimensions, the holes need no masking and no post treatment recutting. The bracket is degreased, soaked in a 4 to 10% citric bath at temperature for the cycle the processor runs for 316, rinsed, dried, and lot tested by water immersion. The result is a bracket at machined size with free iron removed, fit for repeated chloride bearing washdowns. Had the same part needed color or wear resistance instead, the decision would move to powder coat, with masking on both holes.

Example: 17-4 aerospace shaft at high hardness

A 17-4 PH actuator shaft is machined in the solution treated condition, aged to the H900 condition for strength, then finish ground. The drawing note reads passivate per AMS 2700 after final machining and hardness verification. The processor runs a PH rated treatment, dichromate added nitric or an accepted citric cycle, followed by an embrittlement relief bake, because the part sits above the hardness threshold where hydrogen from acid exposure becomes a risk. Verification uses humidity or water immersion testing rather than copper sulfate, which can false fail on this grade. The shaft keeps every ground dimension through treatment, and the sequence note prevents the classic error: passivating before the last cut re embeds iron.

When passivation is the wrong call

Passivation earns its place on clean, mechanically sound stainless parts that need corrosion resistance restored. It is the wrong tool in five specific situations.

• The surface carries heat tint, scale, or existing rust. The bath removes none of it. Clean or pickle under ASTM A380 first, or take a mechanical route such as abrasive cleanup, then passivate.
• The part is not stainless. Passivation as specified here is a stainless treatment: aluminum parts go to anodizing, steel parts to plating or organic coatings.
• The job is cosmetic. Passivation removes no visible defects and changes no finish.
• The part needs wear resistance or a barrier layer. Those are coating or hard facing problems.
• The part is pitting through its own alloy limits. No surface treatment upgrades 304 or 17-4 to 316 performance in chloride service.

The adjacent decision is passivation versus electropolishing, covered above and on the electropolishing page. Choose passivation when only corrosion resistance is wanted. Choose electropolishing when the part also needs smoothness, cleanability, or deburring, at the cost of removing 5 to 25 µm of surface metal. Buyers choosing between organic coatings for the same part can start from the anodizing versus powder coat comparison, which lays out the buildup and appearance trade-offs those two processes carry.

What to verify with a processor

A few questions expose most passivation quality risks before they reach production:

• Which treatments does the processor actually run for your grade: nitric, citric, or dichromate added nitric?
• Which acceptance test does the lot get, and is that test right for the grade, given that copper sulfate false fails on some grades?
• Are parts of different grades, and plain carbon steel parts, kept out of the same load, since mixed loads invite galvanic damage and flash attack?
• Is the rinse water low in chlorides?
• For hardened PH parts, is a post treatment bake included?

A processor with straight answers to those questions is running the process under control rather than by feel.

The short decision summary: if the part is stainless, was machined, welded, or handled in a shop, and needs its corrosion resistance back without touching a single dimension, passivation is the correct specification. The drawing note should name the standard, the treatment, the test, and the sequence, so the result is verifiable rather than assumed.