For tablet dosage forms, the API is first passed through a 600 µm stainless steel sieve to remove transport aggregates. A direct compression baseline is established on a 16-station rotary tablet press with B tooling at 50,000 tablets per batch. The powder blend consists of 20.0% w/w API, 72.5% w/w microcrystalline cellulose PH102, 4.0% w/w croscarmellose sodium, and 1.5% w/w magnesium stearate. Mixing is performed in a 250 L bin blender at 12 rpm for 15 min. Blend uniformity samples are withdrawn from 10 locations and must show RSD ≤ 5.0% under FDA 21 CFR 211.110. Compression force is held at 8–14 kN to produce tablet hardness of 60–90 N. Disintegration is tested according to USP <701> in 0.1 M HCl at 37°C and must be complete within 15 min. Dose uniformity is evaluated by USP <905> with an acceptance value of ≤15.0. If more than 30% of the API remains below 50 µm, punch sticking and weight variation above 3% occur on extended runs. The compression suite is maintained below 45% RH because moisture uptake above this threshold increases tablet capping at turret speeds above 45 rpm. The terminal oral tablets are produced at 50 mg and 250 mg strengths and packed in 10-tablet aluminium/PVC blisters.

Terminal Moist Heat Sterilization Limits for Injectable Formulations

Injectable formulation work begins with solubility screening in Water for Injection at 25°C and 40°C. If aqueous solubility is below 10 mg/mL, a co-solvent platform of 20–30% propylene glycol and 10–15% ethanol is evaluated. The bulk solution is adjusted to pH 5.0–6.5 with 0.1 M HCl or 0.1 M NaOH. Sodium chloride is added to target osmolality of 285–310 mOsm/kg. The solution is compounded in a 316L stainless steel vessel under nitrogen overlay at 200 rpm for 30 min.

Terminal moist heat sterilization at 121°C for 15 min is acceptable only when the resulting total degradation products remain below 0.5% and the assay remains 95.0–105.0%. The cycle must deliver an F0 value of ≥8 min at the cold spot. If the API is heat-labile, aseptic filtration through a 0.22 µm PVDF membrane is used and filling is conducted under Grade A conditions. Post-filtration hold time is limited to 24 h at 2–8°C. Endotoxin limits are calculated from the proposed maximum species dose; a limit of ≤0.5 EU/mg applies for a 10 mg/kg dose. Sterility testing follows USP <71>. Particulate matter is controlled under USP <788>. Residual solvent testing is performed according to VICH GL18 when ethanol or propylene glycol is present in the formulation. The finished solution is not compounded with strong oxidizing preservatives because visible discoloration occurs above pH 7.0. The terminal product is a 100 mL multidose vial sealed with butyl rubber stoppers. Production-scale filling on a 12-head peristaltic line without sufficient nitrogen overlay produces final fill-weight variability above 2%, a failure mode observed when the API solution releases gas bubbles in the holding tank.

What Determines Capsule Fill Weight Uniformity at Low API Loadings?

Low-dose capsule filling is controlled less by API mass than by excipient particle-size distribution. A size 2 HPMC capsule with target fill weight 120 mg is filled on a dosator-type encapsulation machine. The API is geometrically diluted in 3 steps at 1:1, 1:3, and 1:5 with lactose monohydrate D50 75 µm. The final blend is passed through a 500 µm screen and mixed for 10 min at 18 rpm. Blend uniformity must meet RSD ≤ 5.0% per 21 CFR 211.110. Fill weight is monitored at 10-minute intervals with a target deviation of ≤3%. Dissolution is tested by USP <711> using 900 mL of 0.1 M HCl at 37°C; Q is ≥80% in 45 min for immediate-release capsules. Moisture in the capsule filling suite is held below 40% RH to prevent HPMC shell deformation and API agglomeration. The terminal product is a 30-capsule HDPE bottle with a 1 g silica gel desiccant. Published data for this specific configuration is limited; therefore, the geometric dilution sequence is revalidated when the API lot particle-size D90 shifts by more than 100 µm.

If Hard Water Exceeds 200 ppm CaCO₃, Oral Powder Reconstitution Requires Buffering

When hard water from field boreholes contains more than 200 ppm CaCO₃, oral powder reconstitution produces visible clouding and may reduce syringeability in automated drenching equipment. The water-soluble oral powder is therefore pre-blended with 2.0% citric acid and 0.5% sodium lauryl sulfate to improve wetting. Powder flow is measured by USP <1174>; a Carr index below 20 is required for high-speed sachet filling at 60 sachets/min. The finished oral powder is packed in 100 g foil-lined jars with a moisture vapour transmission rate below 0.1 g/m²/day. If the powder is mixed into drinking water, the final medicated concentration is prepared at 0.5–1.0 g/L and must be consumed within 24 h. Pre-drying at 40°C for 2 h is required when storage relative humidity exceeds 60% RH. The terminal product is marked as a non-sterile oral powder and is not used for parenteral administration.

Drying Granules to ≤2.0% LOD Without Generating Fines

Granulation of the API into free-flowing granules requires a high-shear mixer with a 65 L bowl, impeller speed 150 rpm, chopper speed 1800 rpm, and wet massing time 3–5 min. The binder is 3.0% povidone K30 in purified water, added at 8–10% w/w. Drying is performed in a fluid-bed dryer with inlet air at 60–65°C and product temperature held at 30–35°C. Final loss on drying is controlled to ≤2.0% by USP <731>. Dried granules are sieved through a 1.0 mm screen, and the target D50 is 250–425 µm. Granule porosity is not measured directly at routine release; moisture and sieve profile act as release surrogates. If the LOD falls below 0.8%, the granules become friable and generate fines above 15% during sachet filling. The finished granules are packed in 1 kg foil-lined sachets for oral administration after mixing in feed. Published data for this specific configuration is limited; pilot-scale runs are therefore used to verify the water addition range before commercial granulation.

Why Carryover in Micro-Pellet Premixes Must Stay Below 1.0%

Premix production at feed-mill level differs from pharmaceutical blending because the API is dispersed onto a carrier, not agglomerated. Ground corncob fractions of 500–850 µm are loaded into a 500 kg ribbon mixer. Mineral oil at 1.0% w/w is sprayed onto the carrier to control dust. Shichuan Powder API is added as a micro-ingredient and mixed for 15 min at 20 rpm. Blend uniformity is assessed at 10 sampling points; the target is RSD ≤ 5.0%. Carryover to a subsequent batch must remain below 1.0% because residual drug in non-medicated feed can exceed carryover limits and spread the active ingredient into finishing barns. Electrostatic charge on dry corncob carrier is a known production failure at relative humidity below 30% RH; dust rises and API drifts onto mixer shafts. Adding water is not recommended because it promotes local clumping. Instead, mineral oil is increased to 1.5% w/w when carryover exceeds 1.0%. The terminal product is a 25 kg bag label-compliant with EU Regulation (EU) 2019/4 on medicated feed and appropriate national feed additive registers. The premix is intended only for incorporation into complete feed at the mill, not for top-dressing on farm.

Dispensing stock solutions through proportional dosing pumps exposes the API to prolonged dissolved-state contact with water and light. A stock solution is prepared at 1:20 dilution in a 1000 L high-density polyethylene tank. The solution is adjusted to pH 5.5–6.5 with citrate buffer. Sodium metabisulfite at 0.1% w/w is added only when dissolved oxygen exceeds 4 mg/L and the solution is stored beyond 6 h. Mixing is performed with a 300 rpm propeller stirrer for 20 min. The solution is then pumped through a 50 µm in-line filter to remove undissolved particles before entering the dosing pump. Proportional dosing pumps are calibrated to deliver 0.5–5% stock solution into drinking water. The final concentration is verified by conductivity or UV absorbance at the drinker line. The solution is protected from light and used within 24 h at 25°C. Chemical stability in the 1000 L tank is monitored at 0, 8, and 24 h by HPLC assay. The terminal product is a 1 L opaque high-density polyethylene bottle for farm use. The solution must not be mixed with chlorine-based disinfectants at the drinker line because oxidation causes rapid assay loss.