Dry fractionation (ultra-fine milling + air classification) is a clean, water-free route to produce pulse protein concentrates, yet it carries inherent technical, raw-material and application constraints, as summarised from pulse dry-fractionation practice on protein-mill.com.
1. Limited maximum protein purity
Dry fractionation only achieves physical separation based on particle size, density and aerodynamic behaviour. It cannot fully isolate protein bodies from residual starch and fibre. Typical final protein content falls in the range of 40–55% for chickpea, and up to ~60% for selected pea or fava bean materials. It can only manufacture protein concentrates, not high-purity protein isolates (>75% protein). For applications requiring isolate-grade material, downstream wet polishing is mandatory.
2. Protein purity vs yield trade-off
Raising protein concentration in the fine fraction always reduces the mass yield of the protein-rich stream and the overall protein recovery rate. Recirculating coarse fractions back to the mill can improve protein recovery, yet over-regrinding damages starch granules, blurs particle size differences and weakens classification efficiency. Operators must continuously balance these two targets during production.
3. Raw-material sensitivity
The performance strongly depends on pulse cultivar, seed hardness, starch granule morphology, moisture, growing origin and storage history. Chickpea is notably less forgiving than yellow pea. Small variations in incoming raw material will shift particle liberation after milling and change air classification cut-point behaviour, leading to batch-to-batch inconsistency. Strict pre-screening and frequent parameter recalibration are required. High-fat feedstocks are also difficult to process; surface fat causes particle agglomeration and interferes with airflow separation.
4. Antinutritional factors remain in the protein concentrate
Dry fractionation is purely physical and has no washing or solubilisation step. Phytic acid, trypsin inhibitors and oligosaccharides are not removed; instead, these anti-nutrients tend to concentrate together in the fine protein fraction. This imposes formulation constraints for sensitive food applications and may require separate pre-treatment or post-processing to reduce anti-nutrient levels.
5. Retention of native off-flavours
No aqueous extraction or washing step to remove volatile beany, grassy or astringent compounds. Dry-fractionated protein powder retains the inherent pulse flavour, which can negatively impact taste in neutral beverage or delicate food formulations. Additional flavour-masking or mild thermal treatment may be needed.
6. Functional limitations for certain end-use cases
While native protein structure is preserved, the co-existing starch and fibre modify functional performance. Dry-fractionated concentrates are less ideal for applications demanding extremely high solubility, clear liquid systems, or very strong foaming performance, compared with wet-extracted isolates. They are better suited for baked goods, extruded snacks and blended flour systems rather than clear plant protein drinks.
7. Co-product commercialisation pressure
The process generates large volumes of starch-rich coarse fraction as the main co-product. The overall process economics rely heavily on stable market outlets for this starch/fibre stream. Without viable sales channels for the coarse fraction, the unit production cost of protein concentrate rises sharply, reducing commercial viability.
8. Powder handling and dust risks
Fine milled pulse powders are highly dusty. Industrial air classification lines require robust dust containment, explosion protection and inerting controls. Powder flowability can also fluctuate with moisture, causing feeding instability and blocking in classifier rotors.
Dry fractionation is excellent for sustainable production of pulse protein concentrates, but it cannot replace wet extraction for high-purity isolate production. Its biggest limitations are capped protein purity, raw-material sensitivity, retained antinutrients and flavour compounds, plus the economic dependency on co-product utilisation. Material-specific milling and air-classifier tuning can mitigate these issues, yet the fundamental physical separation boundaries remain.
