Good Coatings Fail When Chemistry Is an Afterthought: Lessons From 12 Years of Field Failures

I keep an embarrassing spreadsheet. After 12 years in coatings, it lists seven formulation and specification mistakes that I either made or approved. Combined, those mistakes cost around $140,000 in wasted material, rework, and lost trust. I don't keep the list to punish myself. I keep it because every new coating project starts with the same temptation: treat a coating as a commodity and hope the chemistry works out.

It usually doesn't.

I'm now on the technical service side at Evonik, which means I work with people who manufacture coatings rather than people who just apply them. The phone calls I get rarely start with "we made a bad product." They start with photos: a roof that's peeling, a car coating that's spotting, a lens coating that's failing after a few months. The first instinct is always to blame the applicator or the weather. Often, that's wrong.

The surface problem that isn't the real problem

In September 2022, a customer in Los Angeles called about an elastomeric roof coating that had started bubbling after the first serious rain. The job was a 12,000-square-foot commercial roof. The contractor had followed the application instructions. The weather was fine. The coating still failed.

If you search for elastomeric roof coatings Los Angeles CA, you'll see dozens of products claiming to handle heat, UV, and rain. Most of them are acrylic-based, because acrylic is the default workhorse. But the word "acrylic" hides a lot of chemistry. Two products can both be called elastomeric roof coatings and still have completely different polymer structures, plasticizer levels, and UV stabilization packages.

That's what happened on that roof. The formulator had used a low-cost plasticizer to reach the flexibility spec on paper. It worked in the lab. It didn't work in Los Angeles. After a few months of intense UV, the plasticizer migrated to the surface. Dirt stuck to it. Mildew grew on it. The coating lost adhesion in patches and started peeling. The contractor blamed the product. The product manufacturer blamed the contractor. The roof got re-coated, and nobody was happy.

Here's the uncomfortable part: I should have caught it. The manufacturer bought one of our acrylic binders and asked for a starting formulation. We gave them one. But I didn't ask enough questions about what else they planned to add. That failure is on my spreadsheet. It's the kind of mistake that looks obvious in retrospect.

Category names hide the real chemistry

It's tempting to think that a coating is a coating. You compare the price per gallon, check the coverage rate, and make a decision. But the chemistry underneath the label determines everything: adhesion, flexibility, UV resistance, dirt pickup, and service life.

This isn't just a roofing problem. The same pattern shows up across the coatings world.

Take what is AR coating on glasses, for example. People ask this question because they're trying to decide whether the upgrade is worth it. An anti-reflective coating is not a single layer of "something" that makes glare disappear. It's a thin-film interference stack, usually made of alternating oxide layers with carefully controlled thicknesses. Each interface is designed to cancel reflected light. The coating's durability depends on the materials chosen, the deposition process, and how well the layers bond to each other and to the lens. That's why a cheap AR coating can look fine when it's new and then scratch or delaminate after a few weeks of cleaning.

I see an even better example in the automotive aftermarket. Products like Adam's graphene coating are marketed as if graphene alone does the work. Enthusiasts watch a video, apply the coating in an afternoon, and expect paint that never scratches and never fades. But the graphene is a component inside a polymer matrix. That matrix is what actually bonds to the paint, protects against UV, and holds up to washing. If the matrix chemistry is wrong for the application, or the surface isn't prepared properly, the graphene doesn't save you.

What most people don't realize is that the same principle applies at every scale. A roof coating, a car coating, and an eyeglass AR coating are all examples of thin films doing chemistry on an unpredictable surface. The moment you reduce them to a brand name or a price tag, you lose the information that actually determines success.

Documents are not paperwork

There's a second reason coatings fail that nobody wants to talk about: people ignore the data sheets. I'm not just talking about the SDS, although that's where the problem often starts.

Searching for Evonik Degussa MSDS is still common, because many coating formulations were written years ago when the company was still called Degussa. People find an old SDS, file it, and assume they've done their due diligence. But an SDS is a safety document, not a performance document. It tells you how to handle the material. It doesn't tell you whether your binder is compatible with the plasticizer you bought from a different supplier at a lower price.

I learned this the hard way. In 2017, I approved a raw material substitution to save money on a high-gloss floor coating. The safety data sheet looked fine. The technical data sheet looked fine. What I didn't check was how the new material behaved with the existing crosslinker in the formula. The result was a batch that cured unevenly and had to be stripped and reapplied. That mistake cost about $18,000 and took three weeks off the project timeline.

We didn't have a formal change-review process back then. After that failure, we created one. The third time a similar issue came up, we caught it before production. We've caught 14 potential failures using that checklist in the last five years. But the process should have existed from day one.

What a coating failure really costs

A coating failure is never just the cost of the coating. It's the cost of the labor, the delay, the equipment rental, the rework, and the damage to your reputation. It's also the cost of the end customer's perception.

When a roof coating fails on a commercial building, the building owner doesn't say "the chemistry was wrong." They say "the contractor did a bad job." When an inexpensive AR coating on eyeglasses scratches after three weeks, the customer doesn't blame the lens manufacturer. They blame the optical shop that sold them the glasses. When a graphene coating on a car streaks and fails to cure, the owner doesn't analyze the formulation. They say the brand is overhyped.

That's what quality perception means in this industry: the observable result becomes the brand.

I'm not saying every product needs to be the most expensive option. There are legitimate cases where a budget coating is fine. But the budget decision should be made with your eyes open. If you're choosing a coating for a surface that's expensive to reprepare, hard to access, or highly visible, the chemistry deserves more attention than the price per gallon.

The checklist I now use before approving any coating

After enough expensive lessons, I stopped trusting product categories and started trusting data. Here's what I do now, and it takes less than an hour:

• Read the SDS before the price sheet. It reveals the solvent system, solids content, and polymer family you're actually buying.
• Ask for the technical data sheet and look for performance numbers with test methods attached. If a supplier claims high elongation, I want to know which standard was used and under what conditions.
• Check compatibility of the whole system, not just the main binder. Additives, plasticizers, and crosslinkers can quietly make each other fail.
• Test on the actual substrate in the actual environment whenever possible. Lab panels are not roofs in Los Angeles.
• Create a formal review process before changing any material. Even a "minor" substitution needs sign-off.

This isn't bureaucratic caution. It's the difference between a coating that performs and a coating that just happens to fail after the warranty expires.

I still keep that embarrassing spreadsheet. New failures still happen, because coatings are chemistry and chemistry is never perfectly predictable. But I don't make the same mistake twice, and that's the real point. Whether you're coating a roof, a car, or a lens, the quality of the final result is the quality of your brand. The chemistry behind it decides which one people remember.