For direct compression of the Sishen Powder Veterinary Grade API, the powder is first screened through a 0.8 mm aperture on a conical mill equipped with a round-hole screen. The mill tip speed is kept between 15 m/s and 22 m/s to reduce the fibrous botanical fraction without generating excessive fine particles. A d50 of 125 µm is targeted because coarser material causes weight variation, while finer material increases dust and compacts poorly. The milled API is pre-blended with 30–45% w/w spray-dried lactose monohydrate and 25–40% w/w microcrystalline cellulose in a 300 L bin blender for 15 min at 12 rpm. Crospovidone is added at 2–4% w/w before lubricant addition. Magnesium stearate is withheld until the final 3 min and is used at 0.5–1.0% w/w because over-lubrication has been observed on rotary presses as a hardness drop below 50 N and delayed disintegration under USP <711> conditions at 37 °C. Compression is performed on a 24-station rotary tablet press with 12 mm concave tooling. Pre-compression force is set at 1.5–2.5 kN; main compression force is held between 8 kN and 15 kN. Target tablet hardness is 60–90 N, and friability is controlled below 1.0% according to USP <1216>. In-process weight checks are taken every 15 min with a ±3% individual limit. Production-scale failure modes include variation in the botanical particle size distribution when mill screens are worn; screen replacement is required when mass median diameter shifts above 150 µm. Because the powder may contain hygroscopic constituents, the compression suite is maintained at 20–25 °C and 40–45% RH. End-of-run segregation tests are performed by sampling the hopper at start, middle, and end; a marker content deviation above 5% between any two points requires blending correction. Dedicated compression tools are recommended because the plant powder can embed in tooling crevices and cause repeated punch sticking. The terminal product is an oral tablet for piglets or calves, packaged in aluminum-PVC blister units with a desiccant sachet. Because published data for this specific botanical API in low-dose direct compression is limited, a feasibility run at 20 kg scale is recommended before full commercial batch release.

What Limits Blend Uniformity in Low-Dose Capsule Filling?

At blend scales above 50 kg, low-dose capsule filling with the Sishen Powder Veterinary Grade API is constrained by agglomeration of hygroscopic fine particles after storage above 45% RH. The API is first triturated 1:5 with lactose monohydrate and screened through a 0.5 mm hand sieve. The pre-blend is then charged into a 500 L V-blender with the remaining lactose, microcrystalline cellulose, and crospovidone. Colloidal silicon dioxide is used at 0.2–0.5% w/w to improve flow into the dosing disc; talc is limited to 1.0–2.0% w/w because higher levels can delay capsule disintegration. The blend is mixed for 20 min at 10 rpm, sampled from 10 positions, and tested for marker content. Table 1 lists the release checks for the powder blend before capsule filling.

Dosator machine settings depend on the powder column height. The powder bed depth is maintained at 60–70% of the dosing disc thickness. When content uniformity RSD exceeds 5%, the usual corrections are to reduce the powder bed depth, increase colloidal silicon dioxide to 0.5% w/w, or pre-mill the API through 0.4 mm. Capsule filling is carried out on a continuous-motion dosator machine operating at 20,000–40,000 capsules/h. Hard gelatin capsule sizes range from 2 to 0 depending on dose; fill weight is held between 200 mg and 500 mg. Segregation checks are performed after the first 5 min and again after 30 min of filling by collecting 20 consecutive capsules. Capsule lock rings are checked every 30 min for closure defects. The terminal product is a capsule for companion animals or calves under veterinary prescription. The finished capsules must meet dissolution testing by USP <711> or the registered veterinary pharmacopoeial method. The capsule shell moisture is regulated because the botanical matrix can transfer moisture into the gelatin shell and cause cross-linking if the packaging headspace exceeds 60% RH during long storage.

Before a 2% w/w Sishen Powder Veterinary Grade API premix is released to a feed mill, ribbon mixer validation is completed on a production-scale horizontal mixer of 500 kg capacity. The carrier is ground corncob with moisture below 10%. The API is first triturated with 5–10 kg of carrier and passed through a 0.6 mm sieve to break up botanical aggregates. This intermediate is transferred to the ribbon mixer and blended for 12 min at 25 rpm. Mineral oil is introduced through a spray bar at 0.3–0.5% w/w during the final 3 min to suppress dust without forming hydrophobic carrier films. Sampling uses a sampler thief at 10 points, including the end zone and discharge gate. The coefficient of variation for the marker compound must remain below 5.0% before the premix is released to a feed mill. Ribbon mixer paddle clearance is set at 2–4 mm from the bottom; worn paddles above 5 mm increase dead zones. A premix with an assay CV above 5% is not released; it is re-blended for 5 min increments and re-sampled. The terminal product is a Type A medicated article or intermediate premix for further dilution where regulatory status permits; in the United States, facilities operate under 21 CFR 225, while in the European Union the product falls under Regulation (EU) 2019/4 for feed business operators. A known incompatibility risk exists with high-divalent-cation mineral sources; segregation or adsorption should be checked if copper sulfate or zinc oxide is included at 100–200 mg/kg in the final feed. Analytical recovery of the marker from finished feed should be within 90–110% of the calculated value. The premix must be stored in sealed paper bags with an inner polyethylene liner at not more than 25 °C and 60% RH to prevent moisture migration from the corncob carrier. Fill-off into bags is performed through a vibratory chute to prevent particle size segregation; bag weight is controlled to 25 kg ±0.2 kg.

Injectable-Grade Preparation Requires Removal of Pyrogens and Insoluble Fiber Residues

Injectable conversion of the Sishen Powder Veterinary Grade API is not a direct dissolution step. The parenteral route requires extraction of water-soluble constituents, removal of plant fiber and endotoxin, and terminal sterilization or aseptic filtration. Extraction is performed in a jacketed stainless steel reactor with 70–80 °C purified water for 2–3 h under constant agitation. The cooled extract at 25 °C is clarified through a plate-and-frame filter loaded with 0.45 µm cellulose acetate membranes. A second filtration through 0.22 µm hydrophilic polyethersulfone capsule filters follows. The filtrate is adjusted to pH 5.5–6.5 with sodium hydroxide or hydrochloric acid. Tonicity is brought to physiological range with sodium chloride 0.9% w/v or mannitol 5.0% w/v. Terminal sterilization at 121 °C for 15 min is used only when the marker assay remains within 95–105% of the pre-sterilization value and no visible precipitate forms. If the product is heat-sensitive, aseptic filtration is substituted, and the filling zone must meet ISO 14644-1:2015 Class 5 with Grade A surrounding. The terminal product is an injectable solution for cattle or swine, filled in amber Type II glass vials under nitrogen headspace if oxidative marker degradation is observed. The release panel in Table 2 applies.

Residual solvents are controlled under VICH GL18 if extraction solvents are introduced. The container closure must pass the elastomeric closure test defined by USP <381>. For multi-dose vials, antimicrobial effectiveness is tested by USP <51>, and the preservative content must remain within 90–110% through expiry. Polysaccharide or polyphenol constituents may reduce membrane flux at the 0.22 µm stage; therefore, the filtration train is sized by a filter compatibility study before scale-up. Published data for this specific botanical parenteral configuration is limited.

When Moisture Content Exceeds 1.5% in Wet Granulation, Granule Flow and Hardness Collapse

When direct compression fails due to poor flow or high dust generation, fluid-bed top-spray granulation is selected for the Sishen Powder Veterinary Grade API. The dry powder is charged into a 250 L fluid-bed granulator with lactose monohydrate 40–55% w/w and sodium starch glycolate 2–4% w/w. The binder solution is povidone K30 at 4–6% w/w in purified water. Spray rate is controlled at 80–150 g/min, inlet air temperature at 60–70 °C, and product temperature at 30–38 °C. After binder addition, drying continues until the loss on drying is below 1.5%. If the drying endpoint exceeds 2.5%, granule flow measured through a 10 mm orifice under USP <1174> falls below 4 g/s and tablet hardness becomes non-uniform on the subsequent compression line. Dried granules are screened through 1.0 mm and blended with crospovidone 2–4% w/w and magnesium stearate 0.5–1.0% w/w for 5 min. Granule moisture is monitored by near-infrared probe if the granulator is equipped; the inline endpoint is correlated with loss on drying every 2 min during the drying ramp. Sieve analysis uses 0.250 mm, 0.500 mm, and 1.000 mm screens; the target granule fraction is 70–80% between 0.250 mm and 1.000 mm. Fines below 0.250 mm above 25% indicate over-drying, while lumps above 1.000 mm indicate insufficient dry-milling. The granule bulk density target is 0.50–0.65 g/cm³. Tablet compression from granules usually runs at lower main compression force than direct compression; 6–10 kN is sufficient for a 12 mm concave tablet. The terminal product can be a unit-dose granule sachet for horses, a top-dressing granule for cattle, or granules for reconstitution into an oral suspension. Container closure is a foil-lined sachet or aluminum drum with desiccant; the barrier should have a water vapor transmission rate below 0.5 g/m²/day at 38 °C/90% RH. The finished granules must meet loss on drying under USP <731> and content uniformity under USP <905> or Ph. Eur. 2.9.40 if presented as a single-dose product.

To produce a drinking-water soluble powder, the Sishen Powder Veterinary Grade API is dry-blended with a low-moisture dextrose carrier under room relative humidity below 30%. The API fraction is 30–60% w/w; dextrose is 35–60% w/w; sodium citrate is 1–2% w/w; citric acid is 0.5–1.0% w/w; colloidal silicon dioxide is 0.3–1.0% w/w. Sodium citrate and citric acid act as pH regulators and chelate hard-water calcium and magnesium ions above 200 mg/L CaCO₃ hardness. The dry blend is mixed in a 300 L ribbon blender for 20 min and passed through a 0.5 mm screen before auger filling into heat-sealed aluminum-laminated sachets of 100 g to 1 kg. Dissolution is evaluated by adding 10 g of powder to 1 L of water at 20 °C and stirring at 200 rpm for 5 min; the dispersion must pass through a 0.25 mm sieve without retained particles. The terminal product is a soluble powder for pigs, poultry, or calves, administered through a proportional medicator. Metering pump compatibility with EPDM diaphragms must be confirmed because botanical residues can adhere to elastomer surfaces after 24 h exposure. The manufacturing line must comply with 21 CFR 210/211 or EudraLex Volume 4 when the product is registered as a veterinary medicinal product. Water content is controlled by USP <921> Method Ia, with a release limit not exceeding 5.0%. Published data for this specific botanical matrix in automatic drinking water delivery systems is limited; therefore, a field-scale cleanout study under 0.5–1.0% stock solution is recommended before routine use.

Thermomechanical Stability in Post-Pelleting Liquid Application Systems

A suspension is prepared by dispersing the Sishen Powder Veterinary Grade API in a food-grade vegetable oil or in a 2% w/w hydroxypropylcellulose solution at 25–35 °C. The carrier system is selected to avoid plugging of low-volume spray nozzles. The suspension is metered onto cooled pellets at 35–40 °C in a drum coater or continuous belt spray system. Heating above 40 °C for more than 2 h has been observed to form a sticky film on spray nozzles, requiring solvent-assisted cleaning. Uniform distribution is checked by collecting 10 pellet samples from the discharge stream after 5 min, 15 min, and 30 min of spraying. The target coefficient of variation for marker concentration is below 10%. Terminal product is a medicated feed pellet for swine, cattle, or sheep. The product must comply with Regulation (EU) 2019/4 for EU feed business operators and with 21 CFR 558 for U.S. Type B or Type C medicated feed applications. Analytical recovery of the marker from pelleted feed should be within 90–110% of the calculated value. The process is suitable when the feed formulation cannot tolerate the API before pelleting because of die temperatures between 70–90 °C. A post-pelleting liquid application line requires a mean residence time distribution check on the spray drum to ensure that no untreated pellets by-pass the spray zone.