When direct compression is selected for an 80 mesh extract with moisture below 5.0 %

The dry extract powder is milled through an 80 mesh (177 μm) screen using a stainless-steel hammer mill with rotor speed not exceeding 3000 rpm to limit heat exposure. Loss on drying is determined under USP <731> and must remain below 5.0 % w/w before mixing; above this limit, capping and sticking appear on a rotary tablet press running at 40–60 rpm. Pre-drying is performed in a fluid-bed dryer at 55–65 °C with inlet air dew point ≤8 °C; the product bowl is discharged only when outlet air relative humidity is below 20 %. Blend preparation uses a double-cone blender at 60 % working volume. Microcrystalline cellulose 20–30 wt% and croscarmellose sodium 2–5 wt% are added as diluent and disintegrant. Magnesium stearate is withheld until the final 3 min of blending at 0.5–1.0 wt%, because prolonged lubrication reduces tablet hardness and increases disintegration time.

Compression is performed on a 16-station rotary press with 9.5 mm round tooling. Tablet weight uniformity is assessed using USP <905>; relative standard deviation below 2.0 % is achieved only when the press is equipped with a force feeder operated at 25–35 rpm. Friability is measured on 10 tablets for 4 min at 25 rpm according to USP <1216>; the acceptance value is ≤1.0 % for all formulations, and the high-herbal-load tablet in Table 1 must not exceed 0.5 %. Disintegration time is determined in 900 mL water at 37 ± 2 °C using USP <701>; tablets above 40 wt% extract are required to disintegrate within 30 min. Final packaging is aluminium–aluminium blister film with ≥60 g/m² foil thickness; moisture ingress during storage above 3.0 % increases friability. Published data for this specific extract in direct compression is limited; the ratios in the table represent commissioning runs with similar berberine-containing dry extracts.

High-speed capsule filling of the milled powder is constrained less by theoretical fill volume than by powder flow under low relative humidity and static charge retention after sieving. The raw extract is preblended with colloidal silicon dioxide at 0.5–2.0 wt% in a high-shear mixer operating at 150–300 rpm for 1–2 min; this addition reduces Carr Index from an observed raw-powder range of 32–38 % to below 25 % when measured by USP <1174> powder flow methodology. Bulk density and tapped density are measured by USP <616>; a Hausner ratio below 1.25 is required before loading the capsule machine hopper. The filling room is maintained at 18–25 °C and ≤40 % relative humidity; above 60 % relative humidity the powder surface softens and dosator pins on the filling machine show visible adherence within 20 min of continuous running. Capsule size 0 or 1 is selected for 250–400 mg fill weights; size 0 capsules are not used with high-dose formulations when the plug height exceeds 80 % of the open capsule body. Weight variation is measured on 20 capsules according to USP <905>; the acceptance criterion is ±5 % for fill masses above 300 mg. The filled capsules are passed through a metal detector and deduster before packing in HDPE bottles with silica gel canisters; the package must keep internal headspace moisture below 30 % relative humidity for a 24-month storage period.

What limits terminal sterilization of berberine-containing aqueous injections at neutral pH?

Parenteral solutions of Huanglian Baiguan Powder are prepared from a clarified aqueous extract that has been concentrated to a specified berberine chloride content. Solubility screening in water for injection shows adequate dissolution at pH 3.5–4.5, but precipitation occurs above pH 5.5 when the solution is cooled to 2–8 °C. Terminal moist-heat sterilization at 121 °C for 15 min is feasible only when the pH is maintained below 4.5 and disodium edetate is included at 0.01–0.05 wt%; under these conditions berberine degradation is below 2.0 % as measured by HPLC peak area normalization at 345 nm on a C18 column. Filtration before filling uses a 0.45 μm polyethersulfone prefilter followed by a 0.22 μm PVDF membrane under 0.5–1.0 bar differential pressure; filter integrity is tested before and after filtration according to ISO 13408-2:2018. Endotoxin limits are set from the maximum dose by USP <85>; for a 10 mL dose in cattle, the limit is typically ≤2.5 EU/mL. Subvisible particulate matter is controlled by USP <787>; only 10 μm and 25 μm particle counts within the specified limits are accepted. The solution is filled into amber type I glass ampoules of 5 mL or 10 mL and stored below 25 °C; freeze–thaw cycles are not recommended because precipitated berberine may not redissolve completely. Published data for this specific extract in parenteral formulations is limited; pilot-scale stability data should be generated before commercial batch release.

Dry drinking-water powders require carrier-based dilution well below the cohesion threshold of the raw extract. Lactose monohydrate or glucose monohydrate is milled to 60–80 mesh and dried to ≤2.0 % moisture before contact with the API. The extract is geometrically diluted in three stages at ratios of 1:10, 1:10, and 1:10 in a ribbon blender with a working volume of 65 %; mixing time for each stage is 8–12 min at 20–30 rpm. The final powder is passed through a 35 mesh (500 μm) sieve; retained material must not exceed 5 % of the batch. Reconstitution in drinking water requires water below 30 °C and pH ≤6.5, because hard water with total hardness above 250 mg/L CaCO₃ can reduce solubility and create a bitter precipitate. The final powder is packed in laminated foil sachets of 100 g, 500 g, and 1000 g; sachet seals must withstand 250 kPa internal pressure without rupture. In poultry and swine drinking water systems, the prepared solution is administered through proportioner pumps calibrated at 0.5–2.0 % dosing rates; flushing with clean water for 10 min after the treatment period prevents residue accumulation in nipple drinkers.

Pelleted premix steps, mixer coefficient of variation, and finished feed incorporation ratios

Medicated premix production uses a carrier with high oil-holding capacity and low dusting: ground corn cob granules 20–40 mesh or rice hulls are conditioned to ≤10 % moisture. The herbal API is first blended with calcium carbonate at 1:5 and then diluted to 1:50 with carrier in a horizontal paddle mixer. Mixer coefficient of variation must be below 5 % after 10 min; sampling follows ISO 6497:2002 with 10 increments taken from the mixer discharge. Finished feed incorporation ratios are calculated from the target active concentration in feed, commonly 1–5 kg/ton of finished feed for medicated premixes. Pelleting at 70–80 °C with 15–30 s conditioning residence time is used; exposure above 85 °C is avoided unless a stability study demonstrates berberine retention above 90 %. The pelleted product must meet applicable registration requirements for medicated feed premises under FDA 21 CFR 558.3 or equivalent national regulation.

Carryover between batches is controlled by flushing the mixer with 25 kg of ground corn at ≥40 mesh after each product batch; the flush material is either discarded or used in the next batch of the same formula. Residual analytical verification uses a rinse sample from the mixer surfaces; the acceptance limit is ≤10 ppm of berberine in the next batch. Feed mill air handling must keep carrier moisture below 12 %, because higher moisture increases bridging in the mill bin and reduces finished premix flow. Published data for this specific premix composition in commercial feed mills is limited; the mixing times in the table reflect commissioning trials for berberine-containing botanical premixes.

Granulation of the API with povidone binder is performed when oral powders must be dosed into drinking water at low final concentrations but dust generation must be controlled. The API and filler are preblended for 5 min in a high-shear mixer at 150 rpm; a solution of povidone K30 at 5 wt% in purified water or 50 % ethanol is added at 8–12 % of dry mass. Wet massing at 300 rpm for 2–3 min produces granules with a target endpoint of 1.5–2.0 N cohesion when measured by a texture analyzer. Drying in a fluid-bed dryer at 50–60 °C proceeds until loss on drying is below 3.0 %. Dried granules are sized through a 16 mesh (1180 μm) sieve and a 40 mesh (425 μm) sieve; fines below 40 mesh are limited to 10 wt% to prevent segregation during sachet filling. Dissolution is tested by adding the granule to water at 25 °C and pH 6.0 with stirring at 50 rpm; complete dispersion must occur within 5 min. The granules are packed in airtight sachets with an oxygen transmission rate below 50 cm³/m²·24 h·atm to limit oxidative darkening. The final product is intended for oral solution preparation in swine and poultry drinking water; batch-to-batch variation in granule bulk density should remain within ±5 % of the qualification value.

Administering oral drench solutions to cattle through proportioner pumps at low dosing volumes

Oral drench solutions are formulated as high-concentration liquids for automatic proportioner pumps and drenching guns. The extracted API is dissolved in a vehicle containing propylene glycol 10–20 %, glycerin 5–10 %, and purified water to volume; pH is adjusted to 4.0–5.0 with citric acid or hydrochloric acid. The solution is filtered through 10 μm and then 1 μm polypropylene cartridges to remove particulate matter. Viscosity must remain below 10 mPa·s at 20 °C to ensure accurate draw through proportioner pumps calibrated at 0.1–2.0 % dosing rates. Stability is verified at 25 °C/60 % RH and 40 °C/75 % RH for 6 months; berberine retention above 95 % is typical when the solution is protected from light in amber packaging. Packaging is in 1 L, 5 L, and 20 L HDPE containers with induction-sealed caps; the 20 L container is not recommended for use in direct drenching guns without a flow limiter. If the solution is injected into drinking water through proportioner pumps, the final concentration in drinking water should not exceed 0.1 % due to bitter taste and reduced water intake in cattle.