How to measure D90 in protein powder?

D90 means the particle diameter value where 90% of all powder particles are smaller than this size by volume. For dry-milled plant protein flour and protein concentrate fines, D90 is a critical index to judge cell rupture degree, starch integrity and air classification separation effect. Two mainstream industrial testing methods are widely applied: laser diffraction particle size analysis (laboratory standard) and online dry particle size monitoring (production line real-time measurement).

1. Core definition of D90 for protein-starch mixed powder

For milled pea feedstock before air classification:

• Target D90: 25–40 μm (balanced full cell wall rupture + intact starch granules)
• Protein fines after classification (70% protein concentrate): D90 = 4–12 μm
• Coarse starch by-product: D90 = 30–70 μm

If D90 is too small (over-grinding <15 μm): massive ultrafine starch pollutes protein fines, lowering purity. If D90 is too large (under-milling >50 μm): incomplete cell rupture traps protein inside composite particles, reducing separation efficiency and yield.

2. Laboratory offline measurement: Dry laser diffraction particle size analyzer (standard reference method)

This is the official testing method for protein-mill process quality control, high accuracy for batch sampling.

Required equipment

• Laser particle size analyzer with dry dispersion module (vibratory feeder + compressed air dispersion nozzle)
• Dry nitrogen / clean compressed air (oil-free, dew point < -40°C)
• Sample collection container, stainless steel sampling spoon, desiccator
• Sieve (100 mesh) to break soft agglomerates before testing

Step-by-step testing procedure

1. Representative sampling
   Collect 3–5 parallel samples from different positions of the mill/classifier discharge outlet, mix evenly to avoid segregation. Protein powder easily separates by density during pneumatic conveying; single-point sampling causes D90 deviation.
2. Sample pre-treatment
   Place powder in a low-humidity desiccator for 30 min to eliminate moisture-induced agglomeration. Gently sieve through 100 mesh to break loose clumps without crushing individual particles.
3. Dry dispersion parameter setup (critical for protein powder)
   Protein bodies carry strong static charge and easily agglomerate; set dispersion air pressure reasonably:
   • For milled mixed flour (protein + starch): 0.20–0.30 bar
   • For high-purity protein fines: 0.15–0.22 bar (lower pressure to avoid fracturing tiny protein particles)
     Feeding vibration speed: low continuous vibration to prevent over-feeding and particle overlap in laser detection window.
4. Background calibration & sample testing
   Run blank air background calibration first to eliminate dust interference. Inject sample into laser beam, record volume-based particle size distribution curve, collect 3 parallel test readings per sample.
5. Read D90 value from distribution report
   The instrument generates cumulative volume curve: the X-axis particle size corresponding to cumulative volume fraction = 90% is D90. Average 3 parallel results as final D90, discard abnormal outliers.

Key notes to avoid test error

• Do NOT use liquid wet dispersion for dry-fractionated protein powder: water hydration dissolves protein bodies and destroys original particle morphology, D90 data completely invalid.
• Control lab humidity below 45% RH; high moisture forms permanent agglomerates leading to falsely high D90 readings.
• Avoid excessive dispersion air pressure: high airflow impact shatters intact starch granules, making measured D90 smaller than actual production particle size.

3. Online real-time D90 monitoring (production line continuous measurement)

For 24h dry fractionation lines, install inline dry laser particle sensors at mill discharge and classifier fine/coarse outlets to track D90 without manual sampling.

Working principle

Pneumatic powder split sampling extracts a tiny stable material stream from the pipeline, disperses particles via low-pressure air, and passes through laser detection window. The control system automatically calculates D10/D50/D90 and uploads data to PLC.

Process linkage function

When D90 deviates from target window:

1. Milled feedstock D90 >40 μm: PLC automatically raises grinding rotor speed of split-axis ACM to enhance cell rupture.
2. Milled feedstock D90 <20 μm: reduce rotor speed to stop starch over-grinding.
3. Protein fines D90 >12 μm: increase secondary classifier wheel speed to cut large starch fragments out of fine fraction.

4. Alternative screening comparison method (rough rapid field test, low precision)

If laser equipment is unavailable on-site, use stacked vibrating sieves for approximate D90 estimation (only for quick trend judgment, cannot replace laser testing for formal quality records):

1. Weigh equal powder sample, pass through serial standard test sieves (10 μm, 20 μm, 30 μm, 45 μm, 63 μm).
2. Collect weight retained on each sieve, calculate volume cumulative distribution by mass conversion.
3. Interpolate particle size at 90% cumulative passing as approximate D90.

Limitation: Irregular fibrous cell wall fragments cannot be accurately sized by sieving, so D90 error reaches ±8–12 μm; only used for emergency on-site adjustment reference.

5. Data application of D90 measurement in protein processing

1. Judge milling performance: Stabilize D90 within 25–40 μm to balance cell dissociation and starch integrity.
2. Diagnose air classification defects:
   • Protein fines D90 jumps sharply = large starch granules mixed into fine stream (classifier wheel speed too low)
   • Coarse starch D90 drops drastically = severe starch over-grinding in mill
3. Standardize raw material batches: Different pea varieties show different baseline D90 after identical milling parameters; adjust split-axis rotor speed per D90 test results.
4. Verify recirculation effect: Monitor D90 of recirculated coarse stream to limit circulation cycles and prevent accumulated ultrafine starch.

Laser diffraction dry particle size analysis is the standard accurate method to measure D90 of dry protein powder, following strict low-humidity, low-pressure dry dispersion rules to eliminate static and moisture agglomeration. Online inline laser sensors realize real-time D90 tracking for automatic production parameter adjustment, while sieve analysis is only a rough auxiliary test. Stable control of D90 within target ranges is essential to guarantee high separation efficiency and 70%+ protein purity in dry fractionation lines.