Why Dry Plant Powders Defy Direct Compression in High-Dose Veterinary Tablets

Direct compression of the dry combined Huangqin Dijincao powder is rarely successful for cattle boluses with target weight 1–5 g. The crude herb material contains lignified fibre and residual moisture, which together produce high elastic recovery after compaction. Production-scale intermittent rotary presses show capping and edge chipping when main compression force exceeds 18 kN. Tablet hardness measured on a Schleuniger-type tester declines as machine speed rises, because plastic deformation time is insufficient. The fibrous powder also segregates in the feed frame when particle morphology differs between Huangqin root fragments and Dijincao herb particles.

Wet granulation is therefore required. A typical tablet formula uses combined Huangqin Dijincao powder at 25–45 wt%, microcrystalline cellulose at 20–30 wt%, lactose monohydrate at 10–20 wt%, crospovidone at 2–5 wt%, colloidal silicon dioxide at 0.5–1.0 wt%, and magnesium stearate at 0.5–1.0 wt%. The binder solution is povidone K30 in 50% v/v ethanol at 5–10% w/w of dry granulate. Excess free moisture is controlled before compression by loss on drying at 105°C per USP <731>; terminal granulate moisture above 5.0% w/w increases sticking on punch faces.

Granulation is performed in a high-shear mixer with chopper speed 1,500 rpm, followed by fluid-bed drying at inlet air temperature 55–65°C. Dried granules are dry-sieved through a 0.8 mm screen. Compression on a 27-station rotary press with 1 g target tablet weight is adjusted to produce friability below 1.0% by USP <1216> and disintegration below 30 min in water at 37°C by USP <701>. Stability batches are stored at 40°C/75% RH under VICH GL3 conditions to detect moisture migration from the film-coating layer into the herbal core. Residual ethanol from the granulation solvent is monitored under VICH GL18.

Injectable-grade purification is not a filtration-only problem

Injectable presentations of Huangqin Dijincao powder are not produced by direct dissolution of the crude powder. Plant cell debris, lipophilic waxes, and high-molecular-weight polyphenols produce visible precipitates and filter blockage. The manufacturing route begins with aqueous decoction at 95–100°C for 60–120 min, followed by disc-stack clarification and polish filtration. A 0.2 µm membrane filter removes microbial contamination but does not remove endotoxin. Endotoxin control therefore requires ultrafiltration or anion-exchange chromatography before final concentration.

The purified extract is formulated at 5–20 mg/mL of dried extract equivalent, adjusted with sodium chloride to 0.9% w/v, and buffered to pH 6.0–7.5. The pH range is critical because baicalin, a major flavonoid marker in Huangqin, shows pH-dependent solubility and may crystallize in acidic conditions. Nitrogen sparging during filling reduces oxidative discoloration of polyphenols. Vials are filled under laminar air flow and terminally sterilized at 121°C for 15 min only where short-cycle stability data confirm acceptable baicalin recovery. When terminal sterilization is not feasible, aseptic filtration through 0.22 µm PVDF membrane is used.

Quality tests follow parenteral chapters. Bacterial endotoxin is measured by USP <85>, visible particulate matter by USP <790>, and subvisible particulate matter by USP <788>. Published data for this specific combined herbal extract are limited, so endotoxin limits are derived from the animal weight-based dose rather than a fixed monograph value. Production experience indicates that pyrogen removal is more reproducible when the extract is concentrated to 20–30% total solids before ultrafiltration; higher solids content accelerates membrane fouling.

Hard gelatin capsule filling of the dry powder is constrained by moisture uptake in the crude herb matrix. The blend is prepared by delumping through a 0.5 mm sieve, then mixed in a low-shear twin-shell blender for 15 min at 50 rpm. A representative formula for companion animal capsules contains combined powder at 40–60 wt%, lactose monohydrate at 25–40 wt%, microcrystalline cellulose at 10–20 wt%, sodium starch glycolate at 2–4 wt%, and magnesium stearate at 0.5–1.0 wt%. Powder moisture is held below 8.0% w/w by pre-drying in a tray dryer at 40°C; higher moisture leads to punch adhesion on dosator-type capsule fillers and to brittleness in gelatin shells after storage.

Polyphenols in Dijincao can bind to gelatin under accelerated storage at 40°C/75% RH, causing delayed dissolution and cross-linked shell pellicle formation. HPMC capsules are therefore preferred for long-term stability batches where dissolution release is critical. Dissolution is monitored by USP <711> Apparatus 2 at 50 rpm in 0.1 M hydrochloric acid at 37°C for QC release. Uniformity of mass follows Ph. Eur. 2.9.5, and content uniformity follows USP <905> when the dose unit contains less than 50 mg of marker compound.

Capsule filling on a 30,000 capsule/h dosator machine produces weight variation below 3.0% RSD only when powder bed height is stable. Herbal powders with irregular particle shape show erratic flow through the feed gate. Addition of 0.5–1.0 wt% colloidal silicon dioxide improves flow but can slow dissolution if over-mixed; mix times beyond 30 min produce hydrophobic coating of the API particles. Production batches therefore use a two-stage mixing procedure. First, silicon dioxide is blended with a small portion of lactose; second, the remaining powder is added and mixed for 15–20 min.

Premix Carrier Selection and Segregation Control in Medicated Feed

The combined powder is incorporated into medicated premix at 5–20 wt% active powder, with the remainder carrier. Carrier selection is not inert. Ground limestone increases mix density and reduces dust, but its alkaline surface can accelerate degradation of flavonoid glycosides in the presence of moisture. Rice hull pellets and corn cob granules provide low-density carriers that improve sticking of fine herbal particles to the carrier surface. The carrier particle size is typically between 250 µm and 1.0 mm, with bulk density in the range 0.50–0.75 g/cm³.

Mixing is performed in a 1,000 L ribbon mixer for 10–15 min after a stepwise geometric dilution of the API preblend. Uniformity is assessed on 10 sampled points per batch, with an acceptance coefficient of variation below 5.0% for the marker compound. Segregation occurs during silo discharge and truck transport when coarse carrier particles move differently from the fine herbal fraction. Dust suppression with food-grade vegetable oil at 0.5–1.5 wt% can reduce segregation but increases bridging risk in bin hoppers at higher oil loadings.

Compliance in feed applications follows VICH GL3 for stability, VICH GL18 for residual solvents, and local GMP requirements for veterinary medicated feed. In the EU, premix operations are within the scope of Directive 2001/82/EC and the GMP principles for veterinary medicinal products. For materials exported to feed mills, heavy metal and microbial limits are documented by USP <232> and <233> or equivalent compendial methods.

When the Same API Is Blended into Drinking Water: Oral Solution Stability Boundaries

Oral solutions and drinking-water concentrates prepared from the crude powder require a dispersed system rather than a true solution. The aqueous extraction route is preferred because insoluble lignocellulose from the raw herb settles within minutes in drinking-water tanks. A production batch uses dried extract at 2–10 g/L, dissolved in purified water at 25–35°C under recirculation. The batch is adjusted to pH 6.5–8.0 because precipitation of baicalin increases below pH 4.5. Temperature during mixing is held below 40°C to avoid hydrolytic degradation of flavonoid glycosides.

Preservative selection is constrained by the nonsterile aqueous environment. Sodium benzoate at 0.1–0.2% w/v or potassium sorbate at 0.1–0.2% w/v is used when the pH is below 7.0. Above pH 7, antimicrobial activity of benzoate declines, and the batch is instead filled as a single-use drench rather than a multi-dose drinking-water product. Recirculation through a 100 µm inline basket strainer removes coarse plant fragments; final polishing through a 45 µm filter is applied before filling into 1 L or 5 L high-density polyethylene containers.

Microbial limit testing follows USP <61> and <62>. Total aerobic microbial count is typically controlled below 100 CFU/mL for oral liquid products, but the exact limit is set by the registration file. Published data for this specific combination in drinking-water matrices are limited, so stability protocols should include re-suspension tests under field temperature cycling. Bottle inversion and stirring-settling studies in 500 L water tanks provide more relevant data than bench-top beaker tests.

Direct oral powders for piglet and calf drenching are manufactured by dry granulation when the API fraction exceeds 30 wt%. The granulate reduces dust formation and improves flow through veterinary drench pumps. A dry-granulation route uses a roller compactor with roll force 10–20 kN/cm, followed by oscillating granulator sieve 0.8–1.25 mm. The final granulate is blended with dextrose monohydrate at 40–60 wt%, citric acid at 0.5–1.0 wt%, and colloidal silicon dioxide at 0.25–0.5 wt%. Sodium chloride may be included at 0.5–1.0 wt% for oral rehydration formulations.

Wet granulation in a fluid-bed top-spray system is used when higher binder distribution is required. Povidone K30 solution at 5% w/w is sprayed at a rate 100–150 g/min for a 100 kg batch. Inlet air temperature is held at 55–65°C, product temperature at 35–45°C, and final loss on drying is below 3.0% w/w by USP <731>. The dried granulate is passed through a 1.0 mm screen and packed in foil-laminated sachets to protect against moisture ingress. Sachet fill weight is verified by in-line checkweigher with rejection of any unit outside ±2.0% of target.

For powder presentations, sieve analysis per USP <811> is used to document the particle size distribution. Retention on a 150 µm sieve should be below 10% w/w for water-dispersible powders intended for piglet drenching, to prevent nozzle clogging. However, published particle-size data for the combined Huangqin Dijincao powder are limited, so milling validation must be specific to the incoming raw material lot. Raw material batches with high stem content require pre-milling through a pin mill fitted with a 0.5 mm screen before blending.