How to Test Glossy Trim and Prime Latex Over Oil

Why this matters

Latex enamel brushed straight onto old oil trim looks perfect for a week. Then a vacuum bumps the baseboard, or a fingernail catches the door edge, and the new coat peels off in a strip that keeps going, because it was only ever resting on the oil, not stuck to it. The fix is to strip the failed coat and start again, at the painter's cost. The test that prevents it takes two minutes and a cotton ball, and the prep after it is a known sequence. This card owns that test and that procedure. Which primer to use belongs to 'Bonding Primer Vs Stain Blocking Primer Vs No Primer Decision'; this card assumes you have read it and picked one.

Why latex fails on oil

Old oil (solvent alkyd) enamel cures by oxidation and keeps hardening for years. It ends up hard, slick and nonporous, and often glossier than it started. Latex (waterborne acrylic) dries by its water evaporating and its acrylic particles fusing into a film. That film grips by mechanical tooth and by wetting a surface it can bond to; on glossy oil it gets neither. The bond is weak from the first day, and it shows the first time the trim is knocked or flexed. Sanding gives tooth; a bonding primer is formulated to grip a slick surface and give the latex something it can bond to.

Before the test

• Pre-1978 home. Presume lead until an EPA-recognized test kit or a certified renovator's sample says otherwise. The alcohol test disturbs a coin-sized spot and is not itself a problem, but the scuff sanding that follows an oil result is RRP work under 40 CFR 745 Subpart E in target housing done for compensation, unless it stays within the rule's interior minor repair and maintenance limit (6 square feet per room) and uses no prohibited practice; see 'Lead Paint RRP Protocol Reference'.
• Denatured alcohol is flammable and its denaturant is toxic. Nitrile gloves, a few cotton balls dampened from a small container rather than an open can, no flame or pilot light nearby, and the used cotton goes in a closed metal can, not the trash bag.

The test, done so it means something

1. Clean a spot first. Wipe a 2 inch area with a degreasing cleaner, rinse with water, and let it dry. Trim carries hand oil and dust, and grime on a cotton ball looks like paint transfer.
2. Rub with alcohol. Dampen a cotton ball with denatured alcohol and rub the cleaned spot firmly, back and forth, about twenty strokes.
3. Read the cotton and the surface.
   • Color on the cotton and a soft, slightly tacky spot: the top coat is latex. Alcohol softens acrylic.
   • Clean cotton and an unchanged surface: treat it as oil. Alkyd is not softened by alcohol.
4. Test more than one piece. Doors, casing, baseboard and window trim in the same house were often painted at different times. Test each type of trim in each room you are bidding, and every room with a different history (bathrooms and kitchens get repainted more often).

Limits of the test. Some waterborne alkyds and urethane-modified enamels resist alcohol and read as oil. That error is harmless, since you will then prep as if it were oil. The test also only reads the top coat. A latex result says nothing about what is under it, which is the reason for the adhesion check below.

The adhesion check on any trim you will not prime

A latex result does not mean the existing coat is well stuck. If a previous painter put latex over oil without priming, the alcohol test reads latex and the coat is already failing. Before deciding to paint directly over existing latex, check it: on an inconspicuous spot (inside a closet casing, behind a door), cut a small X through the coating with a sharp knife, press a strip of strong tape firmly over it, and pull it off sharply. This is a field version of the tape method in ASTM D3359, in substance; check the current edition for the formal method. Clean edges with nothing on the tape means a sound coat. Coating lifting off in pieces, especially in sheets that come away from a glossy layer underneath, means the stack is failing and painting over it buries a failure.

The procedure for oil trim

1. Clean. Degreasing cleaner, rinse, dry. Sanding over grease drives it into the scratches, where it blocks the primer.
2. Scuff sand until there is no shine left. 180 to 220 grit on a sanding sponge, until the whole surface is uniformly dull under a raking light, with no glossy spots in the profiles or corners. You are cutting tooth, not removing paint. In a pre-1978 house this is the regulated step: wet sanding or a HEPA-shrouded sander, under the RRP setup.
3. Remove the dust. Vacuum, then wipe with a damp cloth and let it dry. Primer laid over dust sticks to the dust.
4. Prime with a bonding primer chosen per the decision card. Apply at the film the product data sheet calls for and let it cure for the time the data sheet gives for topcoating. Bonding primers often develop full adhesion more slowly than they dry to the touch; topcoating early is a common way to lose the bond you just built.
5. Check the primer before topcoating. Repeat the X-and-tape check on a primed spot after the cure time. Clean edges mean the primer has gripped. Primer lifting on the tape means stop: the surface was not clean or dull enough, or the cure time was too short.
6. Topcoat with latex enamel using the lay-on and tip-off method in 'How to Brush Enamel on Baseboard and Casing'.

Liquid deglosser is a solvent wipe that dulls gloss chemically. It is useful on detailed profiles where a sponge cannot reach, but it is flammable, needs ventilation, and many primer makers want it used with, not instead of, a scuff. Where you use it, follow the deglosser label and the primer's data sheet together.

Worked example: three results in one 1988 house

The job. A two-bedroom house built in 1988, so the lead presumption does not apply. The customer wants all interior trim and doors repainted in a semi-gloss latex enamel. The lead tested before bidding, in four places.

Result one: doors and baseboard read oil. Cotton clean after twenty strokes on the hall doors and the bedroom baseboard, surfaces unchanged and noticeably glossy. Procedure: clean, scuff to no shine, dust off, bonding primer, latex enamel. This added one primer coat and the scuff to every door and all the baseboard, and the bid carried it.

Result two: bedroom casing read latex, tape check clean. Color on the cotton, surface tacky. The X-and-tape check inside a closet casing came away clean. Procedure: clean, light scuff for tooth, latex enamel directly, no primer. The previous painter had evidently primed when they went from oil to latex, or the casing had been latex from the start.

Result three: bathroom casing read latex, tape check failed. Color on the cotton. But the tape check behind the bathroom door pulled a piece of white coating the size of a thumbnail cleanly off a glossy cream layer underneath. The latex top coat had been put directly over oil, years earlier. Adding new latex on top would have added weight and another layer to a bond that was already failing.

The call on result three. Everything that would come off with a putty knife and tape was removed down to the sound oil layer, the edges between removed and sound areas were feathered with 180 grit so they would not show through, and the whole bathroom casing then went through the oil procedure: clean, scuff, bonding primer, latex. The cost driver was the removal time, which the lead put on the bid as its own line after showing the customer the piece on the tape. See related: The Bathroom Trim That Peeled in Sheets, which follows the same failure when nobody tested.

The tally. Four tests on three kinds of trim produced three different procedures. Testing only the doors, then assuming the rest matched, would have primed the casing for nothing and painted latex over a failing bathroom coat.

What would change the answer. In a 1965 house, results one and three both become RRP jobs at the sanding and removal steps, and result three's scraping is the heaviest disturbance in the house. If the doors had read latex too, the X-and-tape check on a door would decide whether they went straight to enamel.

What the alcohol test does not sort

Clear and stained finishes. Alcohol dissolves shellac and softens some lacquers, so a stained door or a cabinet face that turns gummy under the cotton is not latex; it is a clear coat with its own priming rules, usually a shellac-based or bonding primer chosen per the decision card. Factory finishes on cabinets and prefinished doors (conversion varnish, catalyzed lacquer) resist alcohol like oil does and need the cabinet prep in 'Cabinet Refinishing Job Standard'.

Exterior trim. The same test works, but an old exterior oil coat is usually chalky as well as hard. Wash the chalk off and let it dry before the test and before priming, or both the reading and the primer bond are to the chalk.

How to tell the job is right

• Primer test passed on a cured primed spot before any enamel went on, on each kind of trim that was primed.
• No glossy spots visible in raking light after scuffing, especially in profile coves and inside corners.
• A week after completion, a fingernail pressed and dragged firmly along an inconspicuous edge does not lift the new coat. Peeling that starts at an edge almost always traces back to a spot missed by the scuff or primed over dust.

References

• ASTM D3359, Standard Test Methods for Rating Adhesion by Tape Test (current edition)
• 40 CFR 745 Subpart E (EPA Renovation, Repair and Painting rule)
• See related: Bonding Primer Vs Stain Blocking Primer Vs No Primer Decision
• See related: The Bathroom Trim That Peeled in Sheets
• See related: How to Brush Enamel on Baseboard and Casing