How to Troubleshoot Wide Particle Size Distribution in NCM

Wide particle‑size distribution (high span value, broad PSD) shows as: D50 within specification, but D10 too low and D90/Dmax too high. Finished powder contains both excessive ultrafine fines and residual coarse agglomerates. It deteriorates tap density, electrode coating quality, calendar performance and battery safety. This article systematically analyses root causes across raw material, jet‑mill equipment, classifier parameters, feeding system and environment, provides step‑by‑step troubleshooting workflow and corrective actions for mass‑production NCM.

1. Typical symptoms of wide PSD

• Span＞1.2 (target span for NCM:0.8‑1.2)
• D50 acceptable, D90 is high (coarse tail), meanwhile D10 is low (many ultrafine fines)
• Laser PSD: broad distribution curve, obvious bimodal hump
• SEM observation: mix of complete spheres, broken fine fragments and bonded coarse agglomerates
• Consequential defects: unstable tap density, coating scratches, higher BET, unstable Dmax, batch‑to‑batch quality fluctuation.

Important note: Distinguish two different situations:

1. True wide PSD: existing both coarse hard agglomerates and over‑ground fine fragments.
2. Apparent false wide PSD: only soft electrostatic agglomerates, real primary spheres are intact.

2. Step‑wise troubleshooting workflow

Step 1: Distinguish true wide PSD vs false wide PSD

1. Check PSD test: perform ultrasonic dispersion during laser test.
   • After full dispersion, PSD becomes narrow → false wide PSD, caused by soft electrostatic agglomerates. Root cause: insufficient secondary air, static accumulation, high dew‑point / moisture.
   • After full dispersion, PSD still broad → true wide PSD from process and equipment problem, continue troubleshooting.
2. Use SEM inspection:
   • False wide PSD: individual spherical particles are intact, only loosely clustered together.
   • True wide PSD: mix broken particle fragments and real hard sintered agglomerates.

Step 2: Check upstream calcined clinker quality

Raw‑material defects will transfer to finished powder and cannot be fully fixed only by classifier tuning.

1. Too many hard fused sintered agglomerates from calcination. Even after jet‑milling, partial large agglomerates cannot be fully broken; meanwhile high grinding energy generates large amount of fine fragments.
   • Inspection: test feed clinker PSD before jet mill. If feed material already has broad distribution, optimize precursor and calcination curve. Reduce local over‑sintering.
2. Large fluctuation of clinker grindability: mixture of under‑sintered fragile particles and over‑sintered hard lumps. Fragile particles break into fines easily, hard lumps remain coarse.
   • Action: stabilize calcination temperature, holding‑time and oxygen atmosphere. Homogenize calcined clinker before feeding into jet mill.

Step 3: Jet‑mill grinding‑related causes (simultaneously produce coarse and fine fractions)

1. Excessive grinding pressure
   High collision energy breaks part of secondary spheres into ultrafine fines; yet some large hard agglomerates cannot be fully disintegrated → coexistence of fines and coarse particles.
   • Correction: reduce grinding pressure to minimum effective pressure which only breaks sintered necks, avoid over‑energy input. Monitor BET rise as warning signal for particle fracture.
2. Unstable feeding rate

• Too low feed rate: low particle concentration → severe particle‑to‑particle impact, generate many fine broken fragments.
• Periodic material surging / bridging: momentary over‑feed leads incomplete de‑agglomeration, residual coarse agglomerates escape to finished product.
  • Correction: check loss‑in‑weight gravimetric feeder; repair hopper anti‑bridging device; stabilize mass flow, keep proper gas‑solid ratio.

Step 4: Classifier system problems (most frequent cause for true wide PSD)

1. Poor matching of classifier speed, main airflow and secondary air
   • Too high main airflow: strong drag force pulls unseparated coarse agglomerates through rotor, D90 rises.
   • Insufficient secondary air: soft agglomerates are not dispersed; clusters are treated as large particles, deteriorate separation effect.
   • Only adjust rotor speed without matching airflow → classification sharpness drops significantly.
   • Correction: re‑optimize speed‑airflow recipe. Increase secondary‑air for soft‑agglomerate shear. Do not run rotor far away from optimal operating point.
2. Classifier mechanical degradation
   • Worn / deformed turbine rotor blades: flow‑field distortion, cut‑point blurs, classification sharpness drops. Even if frequency set‑value unchanged, separation performance deteriorates.
   • Powder buildup inside volute, guide vanes: dead‑zone powder periodically peel off, cause intermittent coarse particle leakage.
   • Seal leakage: bypass flow lets unclassified raw powder directly enter finished product.
   • Correction: inspect rotor wear; clean classifier housing and volute completely; check shaft‑seal and housing sealing condition. Replace worn ceramic rotor when necessary.
3. Over‑high circulation ratio
   Excessive classifier speed leads heavy circulation. Part of particles suffer repeated collision and break into ultrafine fines; meanwhile a small portion of hard agglomerates still escape classification.
   • Correction: adopt minimum sufficient classifier speed, reduce unnecessary circulation load. Sample return coarse stream: if large number of qualified fine particles in return flow, circulation ratio is too high.
4. Single‑stage classification limitation for high‑nickel NCM
   When raw clinker contains large amount hard agglomerates, single‑stage turbine classifier cannot realize sharp separation.
   • Correction: evaluate two‑stage series classification solution; first stage remove oversized agglomerates, second stage fine tune D50.

Step 5: Environment and gas‑source issues

1. Nitrogen pressure / airflow fluctuation: compressor instability leads variable collision energy and variable classification drag force. PSD drifts and broadens.
   • Correction: install nitrogen pressure‑stabilizing station; fix flow control loop.
2. High system dew‑point: moisture triggers soft agglomeration, creates apparent wide PSD.
   • Correction: control circulating nitrogen dew‑point below‑40℃. Improve system sealing to prevent humid ambient air ingress.
3. Static accumulation: poor grounding of classifier and pipelines, particles stick together to form fake coarse clusters.
   • Correction: inspect all grounding connections.

3. Fault diagnosis table

4. Key do‑and‑don’t in troubleshooting

1. ❌ Do not simply increase classifier speed to reduce coarse particles: higher speed raises circulation ratio, generates more fine broken fragments, span may stay high or become worse.
2. ❌ Do not blindly raise grinding pressure: it creates more ultrafine fines without fully eliminating hard agglomerates.
3. ✅ Always combine three means: laser PSD(D50/D10/D90/span), BET value and SEM morphology. PSD data alone cannot distinguish true wide PSD vs soft agglomeration artefact.
4. ✅ When span keeps high, first confirm whether problem comes from feed clinker, before modifying milling‑classification parameters.
5. ✅ For high‑nickel NCM, secondary‑air adjustment should be always included in inspection checklist.

Wide particle‑size distribution in NCM can be either apparent false broadening caused by soft electrostatic agglomerates, or true broad PSD from raw‑material quality, improper grinding energy, poor speed‑airflow matching or classifier mechanical wear. Troubleshooting sequence: distinguish true / false wide PSD → inspect calcined clinker → check feeding stability → optimize grinding pressure → tune classifier speed/main‑air/secondary‑air → inspect equipment wear and sealing. Do not only adjust rotor speed; improve classification sharpness and control circulation ratio, to get PSD with reasonable span while retaining spherical secondary‑particle integrity.