Dry Phase Fortification Methods for Wheat Flour and Extruded Cereal Matrices

In wheat flour milling lines operating at 25–40 t/h per process stream, dry-phase fortification is executed by metering a prepared nutrient premix into the pneumatic transfer line between the final reduction roll and the centrifugal screen sifter, using a twin-screw loss-in-weight feeder with a hopper capacity of 30 L, screw diameter of 25 mm, and stepper motor resolution of 0.1 kg/h. The feeder discharge is introduced through a venturi eductor supplied with compressed air at 0.8–1.2 bar, and the injection point is positioned downstream of the final airflow split but upstream of the sifter so that agglomerates are dispersed before the flour enters the packer bin. Premix addition rates for folic acid are set to deliver 0.7 mg/lb in United States enriched flour under 21 CFR 137.165, which corresponds to a final folic acid concentration of approximately 1.54 mg/kg; for a 1:2000 dilution premix on a 30 t/h line, the feeder set point is commonly 900–1200 g/h. Uniformity is monitored by taking 10 consecutive packer samples in accordance with ISO 24333:2009 and analyzing folic acid by high-performance liquid chromatography; the acceptance limit for coefficient of variation is ≤ 5%, while iron spot checks are used to detect gross segregation of dense iron particles. To reduce segregation, the premix median particle size is matched to the flour bulk fraction within 30–60 µm by air-jet sieving, because hard wheat flour typically has a D50 of 50–70 µm and premix particles above 250 µm tend to migrate to the top of vibrated silos. Where ambient relative humidity exceeds 60%, flour is pre-dried to ≤ 13.5 g/100 g moisture by ISO 712:2009 before dry blending, because surface moisture above this value causes electrostatic adhesion of iron particles to the feeder hopper and increases mass-flow variability. Published data for this specific configuration is limited, but production audits frequently show a packaging variance of 2–4% coefficient of variation across 25 kg bags when the premix particle size span is held below 2.0 and the flour conveying velocity is maintained below 28 m/s.

What Controls Premix Uniformity When Folic Acid Is Diluted to 0.7 mg/lb in Wheat Flour?

Uniformity of folic acid in dry wheat flour is governed by the number of geometric dilution steps, the carrier particle size distribution, and the shear imparted during ribbon blending. In a standard two-stage dilution, folic acid is first ground in a pin mill with maize starch or triturated sucrose at a 100:1 weight ratio, then diluted again to 1:2000 in a horizontal ribbon blender with working capacity of 500 kg, trough length-to-diameter ratio of 2.5:1, and shaft speed of 20–25 rpm for 8–12 min. Extended mixing beyond 15 min does not improve homogeneity and can fracture the carrier particles, generating fines that promote dust loss and downstream segregation; therefore, the mixing interval is verified by sampling the blender discharge at 10 fixed points and measuring folic acid levels. The coefficient of variation for folic acid in the finished premix must remain below 5%, and when the value exceeds this threshold, the premix is re-ribboned or the particle size span from D10 to D90 is reduced below 2.0 by changing the carrier grade. Carrier particles with a D50 above 250 µm in a flour base with a D50 of 60 µm produce vertical segregation in silos and in packer bins, so laser diffraction data obtained according to ISO 13320:2020 are used to reject premix shipments outside the specified span. In the United States, the folic acid addition level of 0.7 mg/lb is fixed by 21 CFR 137.165, and mill enrichment records must demonstrate that the average of 12 consecutive packer samples falls within the regulatory range for the enrichment standard. The premix must not be dry-blended with amine-based leavening residues or free reducing sugars because folic acid degradation accelerates above 50°C in such reactive media, and the resulting breakdown products are not resolved by the standard chromatographic methods used for compliance testing. Moisture ingress during premix storage is also critical; premixes stored in low-density polyethylene liners with thickness below 75 µm can gain 0.5–1.0 g/100 g moisture within 4 weeks at 70% relative humidity, causing folic acid particles to adhere to carrier surfaces and reducing the efficiency of subsequent dry blending.

Where older volumetric screw feeders are replaced by continuous in-line microdosing systems, the mass flow of the dry premix is regulated by a loss-in-weight platform with load-cell resolution of 1 g and a feed screw turning at 60–120 rpm under closed-loop control. The control system monitors hopper weight loss over rolling 30 s windows and updates screw speed every 2 s; feed-rate deviations during a 10-minute audit must remain within ± 3% of the target set point to avoid folic acid overage or underage. Because the premix is injected into a pneumatic line with air velocity of 18–25 m/s, the eductor is positioned at a 30° angle to the horizontal to reduce compacting at the injection port and to direct the powder toward the centre of the flour stream. Batch-to-batch variance in premix density requires automatic recalibration when the bulk density measured by a 1 L graduated cylinder changes by more than 0.10 g/cm³ from the reference value, because such a shift alters screw fill ratio and can change the delivered mass by 5–8%. This microdosing arrangement is used on a twin-screw extruder feed line where the dry mix consists of wheat semolina, rice flour, and sugar at 8.0 g/100 g moisture; the premix addition of vitamins and minerals is kept below 0.5% of total dry feed to avoid affecting the expansion geometry at the die. The capability of the system is assessed with a moving average of 10 consecutive 30 s feeder outputs; when the standard deviation exceeds 1.5%, the mechanical agitation inside the hopper is adjusted or the screw is replaced with a larger pitch to reduce pulsing. Published data for this specific configuration is limited; however, comparative audits on production-scale cereal lines have shown that loss-in-weight microdosing maintains a coefficient of variation of 2–4% for folic acid in 25 kg bags, whereas volumetric screw feeders without feedback commonly produce coefficients of variation between 6% and 12% when the premix bulk density varies by more than 0.15 g/cm³. The dry premix must not be combined with hygroscopic carriers such as glucose syrup solids at relative humidities above 50%, because cohesive flow in the eductor results in partial blockage and feed-rate oscillation.

Ferrous Sulfate Versus Elemental Iron in Long-Stored Wheat Flour: Oxidation Thresholds and Carrier Selection

In dry wheat flour stored at 25°C and 65% relative humidity, ferrous sulfate heptahydrate particles with a D90 below 150 µm accelerate lipid oxidation because the exposed hydrated iron surface reacts with unsaturated fatty acids in wheat germ and endosperm lipid. The peroxide value of such flour, determined by AOAC 965.33, may increase from 0.5 to 3.5 meq O₂/kg within 16 weeks when iron is added as ferrous sulfate at 40 mg/kg; sensory rejection can occur above 5 meq O₂/kg in products formulated with whole wheat flour or wheat germ. Elemental iron, especially electrolytic iron with a particle size distribution of D90 45 µm and bulk density of 2.0–2.5 g/cm³, is less chemically reactive but has lower gastric dissolution and lower relative bioavailability in high-phytate wheat matrices; for this reason, millers serving populations with high phytate intakes may select sodium iron EDTA where regulation permits. Carrier selection for iron premixes must account for the density difference between the iron source and flour; triturated sucrose at bulk density 0.65 g/cm³ is closer to flour bulk density 0.55–0.60 g/cm³ than maize starch at 0.45 g/cm³, and therefore sucrose-based iron premixes remain suspended in the flour mass for a longer distance along vibratory conveyors. The dry blend must not be combined with ascorbic acid in the same premix without encapsulation because ascorbic acid reduces ferric iron to ferrous iron under moisture pockets above pH 3.0, accelerating oxidation and forming coloured complexes. Packaging in paper bags with low-density polyethylene liners of thickness 75 µm slows moisture ingress but is inadequate at warehouse relative humidity above 70%; under those conditions, flour moisture exceeds 14.0 g/100 g within 4 weeks and blending uniformity deteriorates because iron particles become coated with water films and adhere to conveying surfaces. Operations in tropical climates therefore require immediate packing into moisture-barrier formats and the use of desiccant inserts when the storage period exceeds 8 weeks. The operational boundary for ferrous sulfate in dry wheat flour is a moisture content below 13.0 g/100 g; above this value, oxidation rates increase sharply and iron recovery on sieving decreases due to agglomeration.

Following extrusion and hot-air drying to a product moisture of 2.0–3.0 g/100 g, expanded cereal pieces are transferred to a rotary enrobing drum where dry micronutrient powders are applied at 0.1–0.4% by mass of the base piece. The drum operates at 8–14 rpm with internal lifting flights; the product bed temperature at the dusting station is maintained below 45°C because vitamin C and vitamin B12 on the surface degrade rapidly above 60°C when the oil or gum Arabic adhesion layer remains in the liquid state. The adhesion promoter is a hot vegetable oil spray applied at 2.0–4.0 g/100 g before dusting; the oil droplet size is controlled with air-atomizing nozzles at 0.3–0.8 mm Sauter mean diameter to avoid wet pockets that cause uneven powder adhesion and soft texture. The dry powder is introduced through a brush feeder or twin-screw auger at 0.1–0.2 kg/h for a 100 kg/h line; adhesion is measured by shake-out tests of 10 pieces in a 2 L sealed bottle after 10 min, with failure defined as more than 15% of the vitamin label claim lost from the surface. Particle size of the dry fortificant powder should be D90 below 125 µm for dusting to adhere evenly; larger particles detach and create visible specks on the expanded cereal surface. The oil temperature at the spray bar must not exceed 90°C, and the time between oil spray and dusting must be less than 5 s to achieve adhesive wetting without allowing the oil to migrate into the porous internal voids. Vitamin losses on the surface become measurable when the oil spray temperature exceeds 90°C; cooling to 70–80°C within 10 s after dusting is required for heat-labile micronutrients. This post-extrusion dusting method is used for dry vitamin and mineral application because it avoids the thermal and shear losses that occur when the same micronutrients are added to the raw mix before twin-screw extrusion. The residual moisture of the finished pieces after dusting must remain below 5.0 g/100 g; higher values cause glass transition elevation and rubbery texture in starch-based puffs, and the product no longer meets the crispness specification associated with expanded cereal matrices.

Thermal Degradation Thresholds for Dry Blends Added Before Twin-Screw Extrusion

Dry-phase addition of vitamin and mineral premixes to the feed mix of a corotating twin-screw extruder with screw diameter 32 mm, L/D ratio 40:1, and specific mechanical energy input of 180–350 kJ/kg subjects heat-labile vitamins to barrel temperatures and shear fields that produce variable retention. When thiamine mononitrate is added at 1.0 mg/100 g to a wheat-maize mix with moisture 16–18 g/100 g, retention in a barrel zone profile of 90–130°C with residence time 45–60 s is commonly reported in the range 55–85%; raising the final zone temperature from 130°C to 150°C reduces thiamine retention by 15–25 percentage points. Ascorbic acid and its sodium salt are more sensitive, with retention below 50% at melt temperatures above 150°C and die pressures above 80 bar; therefore, dry-phase vitamin C addition before extrusion is generally avoided in favour of post-extrusion dusting unless encapsulated forms with lipid or starch matrices are used. Riboflavin is thermostable in dry matrices below 180°C but sensitive to light; extrusion barrels and downstream cooling conveyors must be shielded from UV below 420 nm to retain more than 90% of added riboflavin. The operational window for riboflavin added before extrusion is therefore focused on light shielding rather than barrel temperature, but the barrel profile is still limited to ≤ 180°C to avoid decomposition of the carrier starch and the formation of brown reaction products. Trace minerals added before extrusion can promote Maillard reactions and increase extrudate brownness; iron sulfate at 40 mg/kg in a sugar-fortified cereal recipe reduces expansion ratio from 3.2 to 2.6 at a die temperature of 150°C because the iron catalyses reducing sugar breakdown and modifies the viscoelastic properties of the melt. This property cliff-edge requires that sugar-rich formulations with added iron be extruded at melt temperatures of 140°C or lower to avoid collapse, but the lower temperature increases melt viscosity and demands higher torque from the twin-screw drive; the extruder must be equipped with a torque control loop that prevents overload above 90% of the gearbox rating. Published data for this specific configuration is limited; however, industrial retention audits on 32 mm corotating extruders generally show that dry premixes containing vitamin B12 and folic acid retain 70–90% of label claim when the barrel temperature never exceeds 130°C and the residence time is below 60 s.

For dry premix feeders on extruded cereal lines with a production rate of 500 kg/h, calibration is performed with a 25 L stainless steel hopper and a helical screw of 20 mm diameter; the feed factor is adjusted until a 10-minute accumulation sample differs from the target by no more than 3 g. The angle of internal friction of premixes measured by ring shear according to ASTM D6773 must remain below 45°; free-flowing premixes with flow function coefficient above 10 can be metered with volumetric screw feeders, while cohesive premixes below 4 require loss-in-weight dosing with mechanical agitation inside the hopper. Bulk density variations greater than 0.10 g/cm³ alter screw fill ratio and cause underdosing by 5–8%; therefore, the premix is re-tested by a 1 L bulk density cylinder before every shift, and the feeder is recalibrated if the measured value is outside the control limit. The feeder discharge is connected to the extruder raw mix inlet through a transparent flexible polyurethane tube with internal diameter 25 mm; operators inspect the tube for powder build-up at 30-minute intervals because static charge can reduce the effective internal diameter and produce pulsating feed. Premixes containing vitamin A as retinol acetate require dry atmosphere control below 40% relative humidity at the feeder hopper, because the ester hydrolyses and isomerises in the presence of free moisture and oxygen; nitrogen blanketing of the hopper is used where the line is located in a high-humidity environment. The calibration record must include the feeder set point in grams per minute, the hopper weight over time, the ambient temperature, and the relative humidity at the conclusion of the run; these records support ISO 22000:2018 process control requirements and provide traceability for the dry fortification step. Published data for this specific configuration is limited, but production-scale audits generally show that feeder recalibration at the start of each shift reduces the coefficient of variation of riboflavin in extruded cereal pieces from 8–10% to 3–5% when compared with volumetric feeding without gravimetric feedback.

When Free-Flowing Premix Is Injected at Air Velocities Above 25 m/s into Flour Conveying Lines

At air velocities above 25 m/s in a dilute-phase flour conveying line, dry fortificant particles impact against pipe bends, causing attrition of carrier agglomerates and separation of small vitamin particles from heavier iron particles. The consequence is a shift in particle size distribution measured by ISO 13320:2020 from D90 180 µm to D90 120 µm after passing only 3 elbows, which increases the proportion of fines below 20 µm and produces electrostatic cling to polyurethane pipe walls. The fines also migrate to the top of the packer bin, so the first bags discharged after a conveying run may contain an excess of low-mass vitamin carrier particles while the last bags may contain an excess of heavier iron particles. To limit this, dense-phase conveying at air velocity 10–15 m/s is used, but backpressure can exceed 1.5 bar and requires rotary airlock seals capable of withstanding the pressure differential. The injection point is placed 10 m before the centrifugal sifter so that any agglomeration is broken and re-mixed; however, excessive distance beyond 20 m increases segregation in horizontal runs because the saltation velocity of iron particles is lower than that of folic acid carrier particles. Conveying lines are designed with bend radii greater than 10 pipe diameters and with polished stainless steel surfaces of roughness Ra below 0.8 µm to reduce particle adhesion. When the conveying air velocity is reduced below 25 m/s, the mill must confirm that the flour remains suspended and does not deposit in the bottom of the line; the minimum pickup velocity for hard wheat flour at 12.5 g/100 g moisture is commonly in the range of 12–15 m/s, and operation below this value causes line plugging. Published data for this specific configuration is limited; however, industrial audits have shown that a reduction in conveying velocity from 30 m/s to 20 m/s decreases folic acid coefficient of variation in packed flour from 7% to 4% when the premix carrier is triturated sucrose.

Because sodium iron EDTA has a solubility product and density that differ from ferrous sulfate, its use in high-phytate wheat fractions changes the dry-phase blending requirement from a simple particle-size match to an electrostatic discharge control. The compound has a median particle size of 60–100 µm and bulk density of 0.80–0.90 g/cm³; when blended into flour with bulk density of 0.55 g/cm³, the density difference is 0.25–0.35 g/cm³, and severe segregation occurs in silos unless a triturated carrier is used to reduce the premix bulk density to 0.60–0.70 g/cm³. In high-phytate wheat flour with phytate-to-iron molar ratio above 6:1, sodium iron EDTA shows relative iron absorption two to three times that of ferrous sulfate, based on stable isotope studies in human volunteers; however, its use is not permitted in all markets and must be checked against Codex CAC/GL 09-1987 and national fortification regulations. The dry blend should not be held in unlined steel bins, because the iron complex can react with zinc and copper surfaces and form coloured deposits that require acid cleaning. Static charge is controlled by maintaining air humidity between 50% and 60% in the blending room; at lower humidity, the powder acquires a negative charge and clings to plastic feeder hoppers, while at higher humidity the flour moisture exceeds the oxidation threshold. The premix must be added after the flour has passed through the final sifter in some operations, but this arrangement increases the risk of uneven distribution in the packer bin; therefore, the final sifter is often moved downstream of the addition point and the flour is re-sifted as a blended mixture. Published data for sodium iron EDTA in wheat flour at industrial scale is limited, but pilot trials generally show that a coefficient of variation below 5% can be maintained when the premix is prepared with a starch-sucrose carrier blend at 1:1 weight ratio.

Post-Extrusion Dusting of Vitamin B12 and Niacinamide onto Expanded Cereal Shapes: Surface Adhesion and Particle Size Matching

Expanded cereal pieces with a bulk density of 0.10–0.15 g/cm³ and surface porosity of 0.2–0.4 mL/g accept dry vitamin powders only when the adhesion promoter is applied at 2.5–3.5 g/100 g oil or gum solution and the powder D90 is below 100 µm. Vitamin B12 is added at 0.5–1.0 µg/100 g, so it is first diluted in a lactose or maltodextrin carrier at 1:1000; niacinamide at 15–20 mg/100 g is less difficult to meter but can produce crystalline surface roughness if its D90 exceeds 250 µm and the carrier is not changed to a finer grade. The dry powder is introduced into the enrobing drum at 0.1–0.2 kg/h for a 100 kg/h line; adhesion is measured by shake-out tests of 10 pieces in a 2 L sealed bottle after 10 min, with failure defined as more than 15% of the vitamin label claim lost from the surface. Because vitamin B12 is light-sensitive, conveying and packaging areas must be fitted with UV filters below 420 nm; storage at 30°C and 65% relative humidity in metallized film limits loss to 10% over 12 months. The surface adhesion promoter must not be applied at temperatures above 90°C, because the oil becomes too thin and drains into the pores of the expanded cereal, leaving insufficient liquid at the surface to capture the dry vitamin particles. The drum is operated with a fill ratio between 15% and 25% of the total volume; higher fill ratios reduce the number of collisions per minute and cause incomplete dusting, while lower fill ratios increase breakage of the expanded pieces. After dusting, the pieces are cooled in a forced-air tunnel at 25°C and 50% relative humidity for 10–15 min before packaging; this step is required to return the surface oil phase to a solid or semi-solid state and prevent transfer of vitamins to the packaging film. Published data for this specific configuration is limited, but industrial audits on expanded maize-wheat shapes show that B12 retention exceeds 90% when the dusting drum is equipped with UV-filtered lighting and the product is packaged within 30 min of dusting.

Batch ribbon blending of high-dose vitamin C in extruded cereal fines requires a two-stage dry dilution because the pure active has a fine particle size D50 below 30 µm and a bulk density of 0.30–0.40 g/cm³, while the extruded fines have a bulk density of 0.25–0.35 g/cm³ and a D50 above 200 µm. The first dilution uses maltrin or silicon dioxide at 0.5% to reduce electrostatic clumping; the second dilution uses a ribbon blender with capacity of 200 kg and mixing time of 10 min at 25 rpm. The final premix must be used within 4 h when exposed to ambient humidity above 50% because ascorbic acid discolours rapidly under moist air and forms brown degradation products that stain the extruded cereal surface. The addition level of high-dose vitamin C in these fines is restricted to 100–250 mg/100 g of finished cereal; above this range, the surface pH decreases below 3.5 and the acid reacts with the starch matrix, producing a hard translucent layer on the expanded piece. The particle size of the vitamin C premix must not exceed D90 75 µm to avoid a sandy mouthfeel; this requirement is verified by air-jet sieving every 2 h during the production run. Because the fines are then packaged into composite cans with internal liners, the residual moisture of the blend must be below 5.0 g/100 g and the headspace oxygen must be reduced below 2% by nitrogen flushing; otherwise, vitamin C retention falls below 80% after 6 months at 25°C. Published data for this specific configuration is limited, but industrial stability audits on dry fortified cereal fines generally show that ascorbic acid retention under these conditions remains above 85% for 12 months when the moisture limit and oxygen exclusion are maintained.