Three Months In, the White Humidifiers Had Turned Yellow
A Guide to Plastic Yellowing and UV Lightfastness Testing - Why Plastics Turn Yellow, How ABS, PC, PP, PS and TPU Compare, Xenon Arc and UV Lamp Accelerated Tests with Color Difference (ΔE), Yellowness Index (YI) and Grey Scale Criteria, Resin and Additive Spec Items, Checks at Sample, First Article and Production, Request Wording, Real Cases and a Checklist

Hello, this is StartMade.

A small-appliance brand imported 3,000 humidifiers with white ABS housings from a factory in China. The samples were bright white with a good gloss, so mass production was approved right away, and the pre-shipment inspection found no visible color variation. Three months into sales, though, reviews started coming in: "I kept it by the window and the side facing the sun turned yellow," and "It's brand new but looks like something old." Nothing was wrong with how it worked, but one photo was enough to sink the star rating.

Tracing the cause led back to the raw material. The factory had switched to a cheaper ABS than the one used for samples, dropped the UV absorber, and was mixing in nearly 30% regrind (回料) made from ground-up runners. The purchase order said only "白色ABS" (white ABS). Nowhere did it say how yellowing would be tested or how much change was acceptable.

Yellowing doesn't show on a freshly made product. It appears only after weeks or months of exposure to sunlight, heat and air, which is why routine inspection rarely catches it. This article covers why plastics turn yellow, what lightfastness and weathering tests factories and labs run, how to set the criteria, and what to put in your spec and inspection plan.

Yellowing happens after shipment, on your customers' windowsills and desks. Lock the resin grade and additives down in writing, and set the color difference (ΔE) or yellowness index (YI) after accelerated testing as numbers.

1. Why Plastics Turn Yellow

Yellowing happens when polymer molecules slowly break down under light, heat and oxygen. The same material can change in a few weeks or hold up for years depending on which resin was used and how it was molded. These are the causes we see most often.

• Ultraviolet light (光老化): UV from sunlight or fluorescent lighting breaks polymer chains and drives oxidation. Products kept by windows, outdoor goods and clear parts are especially vulnerable.
• Heat (热老化): Parts that sit near heaters, LEDs or motors yellow slowly even without sunlight. If the barrel temperature is too high during molding, or the resin stays in the barrel too long, parts come out slightly yellow from the start.
• Missing additives: UV absorbers (紫外线吸收剂), light stabilizers (HALS, 受阻胺光稳定剂) and antioxidants (抗氧剂) all add cost, so they tend to disappear when the spec doesn't mention them.
• Regrind: Ground-up runners and rejects (回料) have already been heated once, so they yellow faster. On white or clear cosmetic parts, even a small amount makes a visible difference.
• Gas yellowing (酚黄): Nitrogen oxides from forklift exhaust or heaters in a warehouse react with phenolic antioxidants in poly bags and stain white products yellow. If stock that has never seen sunlight turns yellow, this is worth suspecting.

For a factory, resin grade and additives are the easiest places to cut cost. Making samples with good resin and quietly switching in production is not unusual, and you can't tell from the outside, so the resin and additives need to be tied down with documents and tests.

2. How Different Plastics Yellow

Each material handles light and heat differently. You need to know what your product is made of before you can set test conditions and criteria.

We covered how to check whether a resin grade has been swapped in Episode 125, and coating adhesion in Episode 130. For coated products, figure out whether the yellowing comes from the base material or the coating.

3. Lightfastness and Weathering Tests at Factories and Labs

Nobody can wait months in the sun, so labs run accelerated tests that push light, heat and humidity hard to speed up aging. Knowing the names alone lets you ask a factory "what conditions did you run, for how many hours, and how much did the color change?"

How to Read Accelerated Test Results

A question we hear a lot is "How many months of real use is one xenon hour?" Unfortunately there's no exact conversion. It depends on local sunlight, indoor or outdoor use, and whether the light passes through window glass. Rather than focusing on the hours, it is far more useful to compare side by side with the approved sample under the same conditions. Put specimens made from the resin approved at sample stage into the same chamber as production parts; if their ΔE or ΔYI differ a lot, the resin or additives have probably changed.

A ΔE of around 1 is hard to tell apart even side by side, while above 2–3 most people notice. On white products, b* (yellowness) moves well before L* (lightness), so check the b* change when you get the results.

4. Spec Items That Prevent Yellowing

"白色ABS" on its own doesn't tell a factory which resin to use or which additives to add, or how much. Write it down in numbers, from the resin to the test conditions.

See Episode 45 for managing color standards and Episode 59 for writing specs in general. Agreeing in advance whether tests run in the factory's own lab or at an outside lab such as SGS or CTI also cuts down on arguments over results.

5. Checks at Each Stage

Sample Stage

When samples arrive, pick two. Put one in a drawer and the other on a south-facing windowsill for two to four weeks. Cover half of the windowsill one with aluminum foil, and you can compare the exposed and covered areas on the same part. It costs nothing, yet tells you a surprising amount. At this stage, also collect the resin COA, additive information and any accelerated test results from the factory or a lab.

First Article Stage

Set the first parts off the production line next to the approved sample and measure ΔE with a spectrophotometer. Ask for photos of the resin bag labels and lot numbers to confirm the resin matches the sample, and run the accelerated test once more. Aligning resin and test conditions at the pre-production meeting makes this much easier; we covered that in Episode 123.

Production and Shipment Stage

Accelerated tests take days, so running one for every shipment isn't practical. Instead, have the inspector bring a portable colorimeter to pre-shipment inspection, measure ΔE against the approved sample, and check resin bag labels and regrind use on site. Keeping a retained sample from every lot lets you trace back which lot the problem started with if a yellowing claim comes in later. See Episode 100 for pre-shipment inspection.

6. Sample Request Wording for the Factory

Putting the wording below in your purchase order or spec makes it harder for the factory to swap resin or additives, and leaves you a basis for assigning responsibility if yellowing occurs.

"本产品为白色外观件，须满足以下耐黄变要求：须使用我方确认的原料品牌和牌号，并添加紫外线吸收剂和受阻胺光稳定剂，量产前提供原料COA和采购凭证。外观件不得添加回料。量产品与封样色差ΔE≤1.5（D65光源，分光测色仪测量）。氙灯老化按ISO 4892-2测试300小时后，ΔE≤3.0，黄变指数变化ΔYI≤5，不得出现开裂或粉化。样品和首件须完成全部测试，量产每批次须测量色差并留样，留样避光保存。如需更换原料品牌、牌号、色母或添加剂，或调整回料比例，须事先取得我方书面确认。"
(This product is a white cosmetic part and must meet the following anti-yellowing requirements. Use the resin brand and grade we have approved, add a UV absorber and a hindered amine light stabilizer (HALS), and provide the resin COA and proof of purchase before mass production. No regrind may be added to cosmetic parts. Color difference between production parts and the approved sample must be ΔE 1.5 or less (D65 light source, measured with a spectrophotometer). After 300 hours of xenon arc testing per ISO 4892-2, ΔE must be 3.0 or less and the yellowness index change ΔYI 5 or less, with no cracking or chalking. Samples and first articles must complete all tests; in production, measure color difference for every lot and keep retained samples stored away from light. Any change of resin brand, grade, masterbatch or additives, or any adjustment to the regrind ratio, requires our prior written approval.)

7. Three Real Cases

These are anonymized versions of patterns we often see in consultations. Amounts and figures are examples.

Case 1 — White ABS Humidifier

This is the case from the opening. Leftover stock and the original sample went into the same xenon test; after 300 hours the sample measured ΔE 1.8, while the production units exceeded ΔE 6. From the next run, the resin grade and UV absorber were written into the spec and regrind was banned on cosmetic parts. Unit cost rose by about RMB 1.5, which was small next to what the exchanges and partial refunds on units already sold had cost.

Case 2 — PP Outdoor Folding Chair

The color barely changed, but after one season claims came in saying "white powder rubs off the seat" and "the backrest cracked." PP degrades by chalking and weakening rather than by turning yellow. The factory was only checking color difference, and there was no light stabilizer in the resin. Only after the next version specified HALS and added cracking and chalking checks plus a bend test after 500 hours of UV lamp exposure did those claims drop off.

Case 3 — Clear TPU Phone Case

A month after launch, reviews piled up saying "the clear case has already gone yellow." It turned out the factory was using cheaper aromatic TPU. Switching to an anti-yellowing grade raised material cost, but the brand could now put "ΔYI ○ or less after 300 h UV testing" on its product page, and that one line actually became a selling point. We covered phone accessory imports in general in Episode 80.

8. Common Mistakes

• Specifying the resin only by material name, like "白色ABS"
• Assuming production will match because the sample came out white
• Converting accelerated test hours directly into real-world months
• Judging materials like PP, which degrade by chalking and cracking, on color difference alone
• Keeping the approved sample on an office windowsill, so the reference itself changed
• Mistaking gas yellowing in the warehouse for a resin defect

9. Yellowing and Lightfastness Checklist

Spec and Contract Stage

• The spec names the resin manufacturer and grade
• Whether UV absorber, HALS and antioxidant are added is stated
• A regrind ratio is set for cosmetic parts
• Accelerated test conditions (type and hours) and ΔE/ΔYI criteria are set
• Changes to resin, masterbatch and additives require approval

Sample and First Article Stage

• A sample sat on a windowsill for 2–4 weeks with half covered for comparison
• Resin COA and accelerated test results were received
• First articles were compared to the approved sample with a colorimeter
• Both sides keep approved samples stored away from light

Production and Shipment Stage

• Color measurements and retained samples are kept for every lot
• Resin labels and regrind use were checked during pre-shipment inspection

Wrapping Up - Start by Putting One Sample on a Windowsill

Yellowing may seem unrelated to how a product works, but customers see a yellowed shell and decide it's cheap. Count the returns, exchanges and falling star ratings, and holding the resin grade and adding the right additives is not a big cost.

There's something you can do right now. If you have stock or samples, take one, wrap half of it in foil, leave it on a windowsill, and unwrap it two weeks later. If you can see a clear line, rework your resin and additive spec before the next order.

StartMade supports you on the ground in China: writing resin and additive specs, checking resin COAs and lab results, measuring color on first articles, and verifying resin on site before shipment. If you already have yellowing claims, we start with root cause analysis; if you're preparing a new product, we start with the resin standard.