A slimy or whitish haze on a cured epoxy is not slow drying. It is hardener that has left the film, and it will take your intercoat adhesion with it.
A tank is blasted, primed with an epoxy zinc-rich, and left overnight. Next morning the surface has a faint white haze and feels slightly soapy to the hand. Most crews wipe it with a rag, decide it looks fine, and apply the intermediate coat. Two years later the intermediate layer can be peeled off in sheets, and the primer underneath is intact and clean.
What is actually on the surface
The film is not wet. It has cured, and while curing it produced a by-product that migrated to the surface.
- Amine hardeners react with carbon dioxide and water. The amine groups at the film surface combine with atmospheric CO2 and moisture to form carbamate and carbonate salts. These are water-soluble and they sit on top of the coating, not inside it.
- The film underneath is sound. This is the part that makes amine bloom so deceptive. Crosslink density, hardness and adhesion to steel are usually all acceptable. The problem exists purely at the interface that the next coat has to bond to.
- It is a surface phenomenon, not a bulk one. That is also why it can be removed and the coating then performs as designed, provided removal happens before the next coat and not after.
Conditions that produce it
- Low temperature with high humidity. The cure reaction is slow, so the surface stays chemically active for longer while moisture is available. Cold damp mornings are the classic window.
- Poor air movement. Still air lets CO2 accumulate at the surface. Enclosed tanks, the inside of a cofferdam and sheeted scaffolds all concentrate the effect.
- Slow hardener systems. Polyamide and phenalkamine hardeners are chosen for tolerance and low-temperature workability, and they are the ones most prone to bloom. Fast cycloaliphatic systems bloom less but are less forgiving to apply.
- High film build. A thick coat holds reaction heat briefly, then releases it slowly, keeping the surface active for hours.
- Zinc-rich and micaceous iron oxide coats. Porous, high-surface-area films expose more amine to the air and are over-represented in bloom complaints.
Why it is easy to miss
On a light grey or white coating the haze reads as uneven gloss and gets recorded as an appearance defect. On a dark colour it may be invisible under site lighting. The reliable indicator is tactile: pass a gloved hand across the surface. A sound epoxy feels dry and slightly slick; a bloomed one feels waxy or soapy.
The distinction matters because the usual reaction to a cosmetic defect is to accept it, and the usual reaction to an adhesion risk is to act. Bloom is the second, disguised as the first.
What it costs if it is ignored
- Intercoat adhesion. The salt layer is weak in shear and it is water-soluble. It forms a plane of weakness between two coats that were designed to bond chemically.
- Topcoat defects. Residual salt draws moisture and causes crawling, cratering and fish eyes in the finish coat, which then has to be stripped and redone.
- Blistering in service. Salt trapped under a topcoat creates osmotic cells. Water passes through the film, dissolves the salt, and the resulting pressure lifts the paint. This is the failure mode that appears at inspection and cannot be argued away.
Telling bloom from under-cure
These two look similar and have completely different remedies, so test before acting.
- Wipe test. Damp cloth, moderate pressure. Bloom wipes off and leaves a clean, glossy surface underneath. An under-cured film smears, feels soft, or shows the cloth picking up pigment.
- Solvent rub. A standard solvent rub on a bloomed but cured film gives a normal result. On an under-cured film it dissolves the surface quickly.
- pH indicator. A damp strip of indicator paper pressed on the surface will read alkaline where amine salts are present.
- Do not confuse with efflorescence. Efflorescence is a concrete and masonry phenomenon, salts migrating out of the substrate. On steel it does not occur; the white material comes from the coating itself.
Removal
- Warm water and a soft brush, or fresh water at moderate pressure. The salts are water-soluble, which is the whole point. Work in sections and rinse so that dissolved salt does not dry back onto the surface.
- Do not use solvent to wipe it off. Solvent wets and redistributes the salt across a larger area without removing it, and it can drive some of it into the film surface. It looks like it works, and it makes the adhesion problem worse.
- Abrade when the bloom is heavy or has been overcoated. Light sanding with the appropriate grade, then remove dust, then verify the surface is dry before recoating.
- Verify before recoating. The test is cheap and takes a minute. Recoating an unverified surface is the expensive alternative.
Prevention is a schedule decision
- Recoat before bloom forms. Bloom develops over hours. If the specification allows a 24-hour interval, consider overcoating earlier on a cold damp site rather than later.
- Move air without chilling the steel. Ventilation removes CO2 from the boundary layer. Blowing cold outside air onto a warm surface creates condensation, which is a different and worse problem. Recirculate or temper the air.
- Raise the surface temperature. Every few degrees shortens the window in which the surface remains amine-rich.
- Specify for the conditions. On winter work in humid climates, a hardener chosen for low bloom behaviour is worth the small sacrifice in application tolerance.
A bloomed epoxy is not a ruined coating. It is a coating with a contaminated interface, and the interface is exactly where the next coat has to do its work. The whole cost of getting this wrong is concentrated in a step that takes two minutes with a bucket of water, which is why it gets skipped and why it shows up in failure investigations years later.