How Much Sample Do You Actually Need to Qualify a Pigment Batch?

TL;DR

100 grams is not a qualification sample. It’s a first-look drawdown. If you’re running dispersion, viscosity, colorimetry, and QUV aging, you need 200-300 g minimum, and that assumes nothing fails. Run tests in the right order—dispersion first, then hue, then viscosity, then aging—so you stop on the first failure and don’t burn material you can’t replace. The protocol below is what a real mid-size ink plant uses. Skip steps at your own risk.

The 100 g Lie

Most pigment suppliers ship 100-gram bags and call them “samples.” Here’s what 100 g actually buys you:

• One full-cone drawdown on a #3 K-bar
• A single grind in a 250 mL bead mill—if you’re lucky
• Exactly zero room for a replicate

That’s not qualification. That’s a handshake. If you’re signing off a new pigment for a production ink, you need data, not impressions. And data costs material.

Breakdown: Grams Per Test

Here’s what a standard qualification run actually consumes, test by test. Numbers based on a 50% pigment loading in a solvent-based ink system, dispersed on a Netzsch MiniCer or equivalent.

Full dispersion / grind check

50 g pigment

The mill needs a minimum fill volume. For a 250 mL chamber at 50% bead loading, that’s ~125 mL of paste. At 50% pigment concentration, you’re looking at 62.5 g pigment, but you can push it with 50 g if your binder-to-pigment ratio is tight. Anything less and the mill cavitates or you get false positives on grind gauge readings.

You also want enough paste to pull a 25 μm grind-gauge drawdown without scraping the bottom of the cup. Been there. Looks clean at 5 μm, but you had no material left to confirm.

Dilution + viscosity curve

30 g pigment

Once the grind checks out, you let down with binder and solvent to target viscosity. That means:

• Base letdown: uses the full dispersion, so this overlaps—count the binder, not the pigment twice
• But you need fresh material for viscosity-sweep on a cone-and-plate rheometer: 3-4 aliquots at different shear rates, each 8-10 g of paste → ~30 g pigment

If the flow curve looks wrong—thixotropic loop too wide, yield stress too high—you’re done. That pigment will print like glue. No point testing color on something that won’t transfer.

QUV accelerated weathering

20 g pigment

At 50% loading, 20 g pigment yields enough ink to draw down three coated panels (150 × 70 mm, 100 μm wet film). That’s two panels for QUV-B 313 exposure plus one unexposed control. You need the control because the human eye is terrible at remembering what color something was three weeks ago.

QUV cycle: 4 h UV at 60°C, 4 h condensation at 50°C. Run 500 hours minimum. Check ΔE at 250 h and 500 h. If gloss retention drops below 50% at either checkpoint, that pigment is indoor-only. Period.

Colorimetry — CIELAB with replicates

10 g pigment

One drawdown at full strength, one at 1:10 reduction with TiO₂. Three readings each, average the L*a*b* values. 10 g is tight but doable—it covers the drawdowns and the reduction paste. If you’re comparing against a retained standard from the last batch, those 10 g also cover the standard redraw, which shrinks as standards age and thicken.

The real minimum

Add it up:

• Dispersion: 50 g
• Viscosity: 30 g
• QUV: 20 g
• Colorimetry: 10 g

Total: 110 g

Looks like 100 g almost works, right? Wrong. That calculation assumes every test passes on the first try. It doesn’t include the scale-up trial at the end. And it assumes zero handling loss—which anyone who’s ever weighed pigment on a balance knows is fiction.

Real minimum with one scale-up trial and handling loss: 200-300 g. Ask your supplier for that number. If they say 100 g is plenty, they either don’t understand qualification or they don’t want you to find something.

The Test Sequence That Saves You Material

There’s a right order and a wasteful one. Follow this. Stop at the first failure. Don’t go back to the supplier asking for more sample unless you saw something worth chasing.

1. Dispersion / grind gauge (50 g) — If it doesn’t get below 5 μm within the standard mill residence time, reject. Reason: a pigment that won’t disperse on lab-scale equipment will definitely clog a production screen pack. We’ve pulled mesh filters out of a 300 L bead mill that looked like someone poured concrete through them.
2. Hue + colorimetry (10 g additional) — ΔE relative to standard. Threshold: ΔE > 1.5 means you won’t match batch-to-batch in production. If ΔE > 0.8 and you’re buying for a spot-color ink, you can maybe compensate with formulation—but ask yourself if you want that headache every delivery.
3. Viscosity / rheology (30 g additional) — Yield stress, thixotropic recovery, high-shear viscosity. A pigment with the right color but the wrong flow curve is the wrong pigment. Printers don’t care how nice the shade is if the ink won’t transfer at 200 m/min.
4. QUV aging (20 g additional) — If gloss retention drops below 50% or ΔE shifts more than 3.0 after 500 h, that pigment cannot go into an exterior-grade product. Don’t negotiate with the data.
5. Scale-up trial (50-100 g) — Same formula, scaled 5×. If dispersion time shifts, if color drifts, if viscosity doesn’t track linearly, your lab data is not predictive. Figure out why before you commit to a ton.

What Each Test Failure Actually Means for Production

Dispersion failure

If the pigment doesn’t break below 5 μm on your lab mill, it won’t on production either—but production will find out the hard way. The filter housing pressure gauge will climb, the operator will call you at 3 AM, and somewhere upstream a primary particle agglomerate you should have caught is now embedded in 800 kg of ink base. Good luck.

Color shift — ΔE > 1.5

A ΔE of 1.5 is visible to a trained eye under D65 lighting. Your customer’s QC department will see it. They will send photos. They will escalate. And formulation adjustments can only pull you so far—if the spectral reflectance curve has a different shape from your standard, no amount of shading will make it identical across all light sources. Metamerism is not fixable.

QUV gloss loss

Different application, same story. For outdoor inks and coatings, gloss loss in QUV is the canary. If you’re past 50% loss at 500 hours, real-world Florida exposure will destroy that surface in 12-18 months. Warranty claims. Replacement costs. A pigment that’s $2/kg cheaper suddenly isn’t.

Viscosity out of spec

If the letdown viscosity is wildly different from your standard pigment at the same loading, your formulator has to adjust solvent and resin levels. That changes the dry-film properties. That changes gloss, adhesion, and—you guessed it—color. Viscosity isn’t a standalone check; it’s coupled to everything downstream.

Reference: Mid-Size Ink Plant Qualification Protocol

This is a real protocol used at a ~2,000-ton/year solvent-ink plant. It’s not theoretical. It works.

1. Request 300 g from supplier. Specify the exact lot number. Ask for the supplier’s retained sample from the same lot—this matters when results don’t match.
2. Dispersion check: 50 g pigment + binder + solvent, grind on bead mill (0.6-0.8 mm YTZ beads, 250 mL chamber, 3200 rpm). Pull at 15, 30, 45, 60 min. Grind gauge reading must be ≤ 5 μm at 45 min or reject.
3. Letdown to target viscosity: Use the remaining dispersion paste. Target 25 ± 3 s (DIN 4 cup at 23°C). Log letdown ratio (binder:solvent). If ratio deviates more than 15% from the standard pigment, flag it.
4. Viscosity sweep: Fresh dispersion, letdown to 25 s. Cone-and-plate at 25°C, shear ramp 0.1-1000 s⁻¹ up and down. Acceptable thixotropic loop area: within ±20% of standard.
5. Color check: Drawdown full strength and 1:10 TiO₂ reduction. CIELAB, D65/10°. ΔE must be ≤ 1.5 full strength, ≤ 1.0 reduction. If borderline, run metamerism index under A/F11/D65 triad.
6. QUV aging: 500 h, QUV-B 313, cycle as above. ΔE ≤ 3.0, gloss retention ≥ 50% at 500 h.
7. Scale-up trial: 100 g pigment, same formula × 5 volume. Repeat grind gauge and color check. Results must track within 10% of lab-scale values.
8. Decision gate: All passes = approve lot for production. One failure = document and either reject or negotiate with supplier. Two failures in any category = no-go.

FAQ

Can I qualify a pigment with less than 100 g?

No, not if “qualify” means you’d stake a production order on it. You can do a color check and a grind gauge with 50-80 g. But you won’t have viscosity data, you won’t have aging data, and you won’t have a scale-up tieback. You’ll be making a $20,000 purchase decision on partial data. That’s a gamble, not a qualification.

What if the supplier only ships 100 g?

Ask for more. Specifically: “I need 300 g from the production lot for dispersion, rheology, color, and QUV qualification.” A serious supplier will send it. A supplier who refuses is either protecting a weak batch or has never been asked and assumes 100 g is the industry norm. Either way, that’s useful information about who you’re dealing with.

How long does a full qualification take?

Lab work: 3-4 days running one sample, plus 21 days for QUV. Scale-up adds a day. Realistic total: 4 weeks from sample receipt to go/no-go decision. Plan around that. Don’t order pigment today and schedule production for next week.

Is it worth qualifying if I’m only buying 200 kg?

If 200 kg of bad pigment wrecks a production run worth 10× that in ink, ink that ships to a customer who then rejects it—yes, it’s worth qualifying. The qualification costs maybe $500 in lab time and material. The alternative can cost you a customer. Run the math.