InsightsA known limit
Can NIR sort black plastic?
Direct answer
No. Black and carbon-black plastics absorb near-infrared light, so no usable spectral fingerprint returns and NIR cannot identify their polymer type. Black pieces are removed instead as visible rejects by color sorting, before the material separation runs on what is left. Removing them by color does not identify what polymer they were.
Why the signal does not come back
NIR sorting works because each polymer reflects near-infrared light in a distinctive way. The sensor reads that returned response and matches it against a validated library. The whole method depends on light coming back.
Carbon black does the opposite: it absorbs across the near-infrared range. A black flake returns too little signal to match against anything, so the machine has no fingerprint to compare. It is not a question of sensitivity or of a better spectral library — there is nothing to read.
What color sorting does instead
Black is trivially visible. A color sorter classifies color, transparency and appearance, so a black fragment is one of the easiest things it removes — the same way it removes an off-color flake, a wood chip or a metal fragment.
| Stage | The decision it makes | Platform |
|---|---|---|
| Color stage, first | Black and off-color pieces out, by appearance | MAS-C on flakes; MAS-B on whole bottles; the color half of MAS-PC and MAS-PC Pro |
| Material stage, second | Polymer identity, on what remains | MAS-P, MAS-P Pro; MAS-B Pro on whole bottles |
Worth stating precisely, because it is the sentence most often lost: removing black material by color does not mean identifying that black material's polymer type. It means taking it out of the way.
The fix is sequence, not a better sensor
A stream carrying a meaningful black fraction will not sort by material until that fraction is gone. The answer is to order the line so the color decision happens first and the polymer decision runs on the remainder.
This is also why "black is not an interference" is the wrong way to describe it. Black material is an interference for NIR material identification, and a routine reject for color sorting. Both halves of that sentence are true at the same time, and dropping either one misleads.
When the black fraction decides the project
If black pieces are a small contaminant, the color stage removes them and the project is unaffected. If black is a large share of the feed and the goal is to identify what those black pieces are made of, that goal is not available from NIR — and no configuration changes that.
Which case you are in is a question about your actual stream, not about the machine, so it is settled by sending a representative sample rather than by specification.
Where this stops applying
- NIR cannot reliably identify black or carbon-black plastics. This is a property of the material, not a configuration setting.
- Color sorting removes black pieces but says nothing about which polymer they are.
- The reverse framing is also wrong: black is not simply harmless. It is an interference for material identification and a routine reject for color sorting.
- Film, bags and pellets remain unsuitable for these platforms regardless of color, because air-jet ejection cannot control their trajectory.
- The achievable separation on a stream with a large black fraction is confirmed on a representative sample, not assumed.
Common follow-ups
Why can NIR not read black plastic?
Carbon black absorbs near-infrared light across the range NIR uses, so too little signal returns to match against a spectral library. Without a returned fingerprint there is nothing for the machine to classify.
Can black plastic be removed at all?
Yes, easily — by color. Optical color sorting removes black items as visible rejects, and platforms such as MAS-C, MAS-B, MAS-PC and MAS-PC Pro all do this.
Does removing black tell you what polymer it was?
No. Removing black material by color takes it out of the stream without identifying it. If you need to know what the black fraction is made of, optical sorting does not answer that and neither does NIR.
Should black be removed before or after material sorting?
Before. The standard sequence is to remove black items first through optical color sorting, then run the polymer separation on what remains, so the NIR stage is not being asked to read material it cannot read.
Is black material an interference or not?
Both statements are needed. It is an interference for NIR material identification, and it is a routine reject for optical color sorting. Saying only one of the two gives a misleading picture.