The Ferric Chloride Test Tells the Truth: Pitting Is an Autocatalytic Runaway

A passive film is not a permanent shield. The moment a pit nucleates, the local chemistry inside it diverges violently from the bulk solution, and the corrosion accelerates itself. The only laboratory test that reproduces that reality honestly is the ferric chloride immersion, codified as ASTM G48. We ran it on P550 and on 316L, and the results explain both why the test is harsh and why it is the right filter for any chloride-service grade. Pitting is an autocatalytic runaway, and the test lets you watch it happen.

What the ferric chloride bath does differently

In 6 percent ferric chloride the polarisation curve of P550 shows no real passive plateau (Table 1). The corrosion current density jumps to 1.23 × 10ₘ A/cm² at 25°C and 1.97 × 10ₘ at 50°C, orders of magnitude above anything we measured in sodium chloride. Ferric ions hydrolyse to release acid, so the steel is bathed in an oxidising, acidic chloride mix that keeps the surface from re-passivating. The point is not that FeCl₃ is “unfair”; it is that it mimics the occluded cell inside a real pit far better than a quiet NaCl bath does.

The occluded-cell autocatalysis

Inside a pit, metal dissolves to form cations, chloride migrates in to preserve charge balance, and those metal chlorides hydrolyse to release acid. The pH inside the pit can drop far below the bulk, the chloride concentration can reach three to ten times the bulk value, and the acidified pocket stays in active dissolution while the outside surface remains passive. That is an occluded cell: a self-sustaining galvanic couple where the pit feeds itself. ASTM G48 works precisely because ferric hydrolysis supplies the same acidification externally, so the specimen experiences the worst-case local chemistry from the start. No stable passive film survives that, which is exactly the point.

Charge-transfer resistance collapses

Electrochemical impedance confirms the damage. In the FeCl₃ bath the charge-transfer resistance of P550 falls to 247.8 Ω·cm² at 25°C and 23.48 Ω·cm² at 50°C (Table 2). Compare those to the 10⁰ to 10⁲ Ω·cm² range we measured in NaCl, and the contrast is stark: the occluded-cell chemistry strips away orders of magnitude of protection. The impedance spectrum also develops an inductive loop, a fingerprint of the adsorption and dissolution processes driving the autocatalysis.

Immersion mass loss, the brutal comparison

After 72 hours at 25°C in 6 percent FeCl₃, P550’s average corrosion rate was 6.13 g/m²·h while 316L reached 14.04, more than double (Table 3). The deepest pit on P550 measured about 1839 µm; on 316L it reached 2328 µm, deep enough to threaten perforation of a thin section. At 50°C the gap persists: P550 averaged 14.44 g/m²·h against 316L’s 20.88, and several P550 pits grew deep enough to link up or punch through the plate (Table 4). The ferric chloride test is harsh, but it ranks the grades in the same order the field does, which is why we trust it.

Corrosion products confirm the chemistry

Energy-dispersive spectroscopy of the corrosion products (Table 5) found oxygen at 29.28 weight percent and chlorine at 42.33 weight percent, with chromium, manganese and iron making up the rest. The dominance of chlorides and oxides is the chemical signature of the autocatalytic hydrolysis: the pit products are metal chlorides and oxides, exactly what the occluded-cell model predicts. Manganese sulfide inclusions, common initiation sites, release sulfide under acid and feed the attack locally, which is one reason inclusion control matters as much as bulk chemistry.

Pitting follows a growth law

The most useful output is not a single number but a rate. Tracking pit depth against immersion time in FeCl₃ (Table 6) and fitting the data gives the growth law c = 382.17 × t⁺¹³₦³₈, where c is pit depth in micrometres and t is time in hours. That exponent, 0.3458, sits right next to the theoretical one-third predicted by the semi-ellipsoidal pit model, in which depth scales with the cube root of time. In plain terms, pit depth grows with the cube root of exposure, so doubling the service life does not double the depth, but it still grows without mercy once initiation has occurred.

The rate, not just the ranking, is the deliverable

A corrosion rate in g/m²·h is useful for comparing grades, but for a structural component the depth law is what protects you. Because pit depth scales near the cube root of time, a component that survives 72 hours in the test does not survive forever; it simply has not been tested long enough. The right move is to extrapolate the fitted law to your planned inspection interval, then set the wall thickness or the inspection frequency so the predicted depth never reaches the critical value. P550’s lower rate and shallower pits buy you a longer interval than 316L, but both still demand a scheduled look, not a set-and-forget assumption that the alloy will look after itself.

Where P550 still needs care

The ferric chloride test also exposes P550’s limits. Above roughly 50°C the pits begin to link and perforate, and manganese sulfide inclusions remain preferred initiation sites no matter how good the bulk chemistry. That means surface preparation, inclusion control, and avoiding stagnant chloride pockets matter as much as alloy choice. A high-nitrogen grade is not a licence to ignore geometry and cleanliness; it is a larger margin that good practice must protect, and the test is the fastest way to learn whether your practice is good enough.

Why this should change how you qualify steel

Do not trust a passivation certificate alone, and do not trust a quiet NaCl polarisation curve either, because neither reproduces the occluded-cell chemistry that kills components in the field. Run ASTM G48, read the corrosion rate and the maximum pit depth, and plan inspection around a cube-root growth law rather than a fixed lifetime. P550 still grows pits in ferric chloride, but it grows them slower and shallower than 316L, and that margin is what keeps a drill collar in the hole instead of on the failure report. The honest test is the one that tries to break your steel, and then tells you how long it took.