If you've ever approved a coating or chemical batch that passed the supplier's QC but failed on your line, you know the feeling. The paperwork is neat. The certificate of analysis is attached. And still the powder coating delaminates from glass, the acid bath leaves rusty streaks, or the color comes out visibly wrong.
Let me rephrase what usually happened: the supplier didn't necessarily deceive you. The product met the specification. The problem is that the specification was never tied to your actual process.
I'm a quality and compliance manager at IFF, and I review roughly 200 unique chemical and coating items every year before they are released to customers — or accepted from vendors. Take it from someone who spends part of every week rejecting things: in B2B chemistry, a product name is not a specification.
Powder coating glass isn't a "find me a powder" problem
Every few weeks, a sample request comes in with little more than the phrase "powder coating glass." From the buyer's side, that sounds like a material request. The deeper problem is that ordinary powder coating is designed around metal.
In a standard electrostatic spray line, powder particles are charged by the gun and attracted to a grounded workpiece. Glass is an insulator. It doesn't conduct charge the way steel does. If you run a soda-lime glass part through a conventional metal powder coating line, you get uneven film thickness, pinholes, and poor edge coverage. Some operations solve this by preheating the glass, using a different charging method, or applying an adhesion promoter. Some don't. The powder coating itself may be excellent — but it was never the only variable.
There's also the thermal expansion mismatch. Powder coatings cure at elevated temperatures. Glass and polymer expand and shrink at different rates. If the coating is too rigid or the cooling cycle isn't controlled, parts will fail after cure even when they look perfect coming out of the oven.
Color is another hidden gap. A RAL number or a Pantone name is not a complete color tolerance. I've rejected parts that matched the code but not the visual sample. Specifying a powder coating for glass should include a measured color tolerance, not just a code.
According to Pantone color matching guidance, a Delta E below 2 is generally considered acceptable for brand-critical colors. Values of 2 to 4 are visible to trained observers, and anything above 4 is visible to most people.
So when someone asks about powder coating glass, the first question is not "which powder?" It's "what glass, what process, what cure schedule, and what test tells us the coating actually stayed bonded?"
What products contain nitric acid — and why the answer isn't enough
I see the same trap in a different container when buyers search for "what products contain nitric acid." On one level, the answer is straightforward. Nitric acid is used in fertilizers, metal treatment and passivation products, pH adjustment, cleaning formulations, electronic etchants, and laboratory reagents. But knowing that a product contains nitric acid tells you very little about whether it will do the job.
Nitric acid is a strong oxidizing acid, but the commercial product comes in different forms. A 68% aqueous solution is different from red fuming nitric acid at 90% plus. Technical grade is not the same as reagent grade. And two drums of "70% nitric acid" can have completely different impurity profiles.
Take stainless steel passivation. The acid is supposed to remove free iron from the surface and help the chromium oxide layer reform. If the acid contains chlorides, it can attack the metal instead of protecting it. Iron impurities can leave deposits that look like rust after a supposedly successful passivation bath. The label says "contains nitric acid" and the concentration is right — but the trace chemistry is wrong for the application.
Let me be direct: the last 30% of a nitric acid certificate of analysis matters more than the 70% assay. If you're buying by product name alone, you're buying a gamble.
A $22,000 education on hidden spec gaps
I only learned this the expensive way. Early in my career, a supplier quoted "nitric acid, 70%" for a passivation process. The price was good. The COA showed the concentration. I approved it without digging into the impurity block. The first production batch left orange streaks on stainless parts, and we had to stop the line.
The cleanup, re-testing, and redo cost about $22,000. The original price difference we saved was maybe $800. It was the worst trade I've made in quality work, and I've checked the impurity section of every COA since.
Similar math applies on the coating side. Suppose a cheaper powder saving $0.10 per part is used on 50,000 parts. That's $5,000 in apparent savings. If that powder causes even a 5% adhesion issue — 2,500 parts to quarantine, re-test, and recoat — the invoice savings disappear fast. And if the failure happens in the field, the cost is much higher.
I'm not saying expensive always means better. I'm saying low price without verified performance is a red flag, not a bargain.
What a transparent specification actually looks like
In our Q1 2024 supplier audit, we found that first submissions from new material sources were the weakest point. The chemistry was usually fine. The missing pieces were test methods, actual measured values, storage conditions, and clear limits. That's not a supplier problem — it's a specification problem.
I now give every new product request the same test. Ask for the things that are not included on the datasheet. A supplier that lists all fees, limits, and test methods up front — even when the total looks higher — usually costs less in the end.
- What is actually measured, and what method is used? A COA should show real numbers, not "conforms." The method matters as much as the limit.
- What is not being disclosed? If iron, chlorides, particle size, adhesion, or cure response are missing, ask why. Those are the attributes that cause field failures.
- How is it validated for your process? For powder coating glass, run a test on the same glass with the same pre-treatment and cure cycle. For nitric acid, test it in the same bath chemistry and alloy condition.
- What happens if it doesn't work? Transparent suppliers give you a path for rework, rejection, and corrective action — before the PO is signed, not after a problem appears.
For color, put the tolerance in writing. For chemistry, put the impurity limits in writing. For coatings, put the adhesion test and cure schedule in writing. If a supplier hesitates to document those, that's a deal-breaker for me.
If you're already an IFF customer and need current product documentation, the fastest route is the IFF contact page. Some customers still find their way to the older customer portal — the one labelled Catalyst ABA login on certain accounts. That portal is fine if your account team gave you access. Otherwise, a direct request through IFF contact gets routed faster because it includes your application details.
This discipline matters even during ownership changes. The announced transition of IFF Pharma Solutions to Roquette is a good example: a product transfer only works if the quality system goes along with it. Buyers should ask about specification ownership, regulatory documentation, and change control — not just the new logo on the invoices. The integrity of the product is not guaranteed by a handshake.
To be clear, this was accurate as of early 2025. Chemical supply chains, ownership arrangements, and product grades evolve quickly. Verify the current COA, portal access, and technical contact before you finalize a purchase.
Bottom line: you are not buying "powder coating glass" or "nitric acid." You're buying a product that has to survive your process, your quality checks, and your customer's expectations. The product name gets the conversation started. The specification — and the supplier's honesty about what isn't in it — is what keeps you out of trouble.