Because the liquid extract of Angelica sinensis supplied as a veterinary-grade API contains water-extractable polysaccharides, low-molecular-weight sugars, ferulic acid, and phthalide derivatives such as Z-ligustilide, wet granulation tablet development cannot proceed by treating the extract as a simple active solution. The extract acts simultaneously as a binder, a hygroscopic agent, and a plasticizer in a lactose–microcrystalline cellulose granulation matrix. In production-scale high-shear granulation using a 10 L bowl and a chopper speed of 1,500–3,000 rpm, the aqueous dilution ratio is adjusted so that total granulation fluid remains between 8.0% w/w and 15.0% w/w of dry powder mass. Impeller tip speed is held within 3–5 m/s; impeller power draw, not elapsed time, determines the endpoint. Wet mass density of 0.62–0.70 g/mL and loss-on-drying of 8.0–12.0% w/w are more reproducible than fixed mixing times. Drying in a fluid-bed dryer at an inlet air temperature of 60–70 °C is preferred because tray drying at higher temperature causes surface crusting from the polysaccharide fraction. Granules dried below 6.0% w/w LOD show lamination during compression on a rotary tablet press at dwell times below 30 ms; granules above 12.0% w/w LOD cause picking and sticking because extract sugars hydrate at the punch face. Compression force is then adjusted to achieve tablet hardness of 60–90 N, while friability is monitored by Ph. Eur. 2.9.7 and disintegration by Ph. Eur. 2.9.1. Content uniformity testing follows Ph. Eur. 2.9.40 for tablets containing low-dose extract markers. In production suites below 40% RH, dried granules remain compression-ready for up to 24 h; above 60% RH, moisture uptake can exceed 1.5% w/w within 2 h and requires re-drying. These boundaries are operational controls, not theoretical estimates.

What Limits the Sterile Filtration Path for Polysaccharide-Rich Angelica sinensis Extract?

Injectable manufacturing is constrained less by chemical potency than by endotoxin, bioburden, and subvisible particle load. The liquid extract is not sterile and contains water-extractable polysaccharides that can foul membrane filters in a non-linear manner. For a 2–5% w/v solution in water for injection, apparent viscosity at 25 °C normally remains below 15 mPa·s, but aggregates can form above 10% w/v and after pH adjustment into the neutral range. The production sequence therefore begins with dilution and pH adjustment to 5.0–7.5, followed by cooling to 2–8 °C for 12–24 h. A 0.45 µm PVDF prefilter protects the downstream 0.22 µm sterilizing-grade membrane. Terminal sterilization by autoclaving is generally avoided because the phthalide fraction undergoes thermal degradation; aseptic filtration is the standard terminal processing step. Bacterial endotoxins are measured by Ph. Eur. 2.6.14 in the diluted bulk solution before filtration. Where no official veterinary monograph exists for the extract, endotoxin limits should be derived from the intended route and species body mass. For intravenous administration in large animals, a conservative working limit of less than 0.5 EU/mg of extract solids is often applied; intramuscular and intramammary products may be justified at higher limits based on route-specific risk assessment. Particulate matter is assessed by Ph. Eur. 2.9.19 for subvisible particles in small-volume parenterals, and sterility by Ph. Eur. 2.6.1 after membrane filtration. The extract should not be combined with divalent cation-rich diluents such as Ringer's solution containing calcium or magnesium without compatibility testing, because polysaccharide–ion interactions can form flocculent precipitates. Filter flux tests on production-scale disc filter capsules show that pressure differentials should be recorded rather than total filtration time; when differential pressure exceeds 0.8 bar at 20 °C, the batch should be re-clarified rather than forced through the filter, because fouling is generally irreversible and risks bacterial challenge. Published data for this specific extract configuration are limited, so each filtration campaign must include pre-validation of membrane lot variability and extract lot viscosity.

In feed premix production, the liquid extract is not directly incorporated at high inclusion rates because free moisture above 14% w/w in a finished premix creates conditions for mold growth and clumping. The preferred route is adsorption onto carriers with an oil absorption capacity above 100 g/100 g, such as precipitated silica, defatted rice bran, or calcium carbonate. The liquid extract is diluted to a sprayable viscosity and atomized at 0.2–0.4 MPa through a binary nozzle in a horizontal paddle mixer or ribbon mixer. Carrier addition is staged so that the free liquid is taken up before the next portion is sprayed; batch size, spray rate, and carrier absorption capacity are interlinked. Mixing is continued for 10–15 min after the final extract addition, and the coefficient of variation of 10 assay samples is maintained below 5.0% when sampling follows ISO 6497:2002. The adsorbed premix is then dried to a moisture content not exceeding 10% w/w measured by ISO 6496:1999. Alkaline carriers such as calcium oxide should be avoided because ferulic acid may form salts at elevated pH, shifting marker assay recovery. Precipitated silica has high absorption capacity but can reduce bulk density below 0.30 g/mL, which may require densification before sachet filling. Defatted rice bran provides nutritional compatibility but introduces natural oil variability; the residual oil fraction can exceed 5% w/w and must be controlled to prevent rancidity. Published loading-capacity data for Angelica sinensis liquid extract on these carriers are limited; adsorption ratios should be confirmed by batch loading trials at the production site rather than transferred directly from chemically similar botanical extracts.

When the Extract Is Diluted into Drench Solutions for Ruminant Administration

Oral drench formulations require aqueous stability and a viscosity low enough to pass through a standard drench gun nozzle without pulsation. The liquid extract is diluted to a final concentration of 5.0–15.0% v/v in a buffered vehicle containing sodium benzoate at 0.1% w/w and citric acid monohydrate at 0.05% w/w. The pH is adjusted to 4.0–5.5; above pH 6.0, precipitation of high-molecular-weight polysaccharide fractions can occur within 72 h at 40 °C. Propylene glycol may be incorporated up to 20% v/v to keep water-soluble and lipid-soluble marker compounds in a single phase at 5 °C. The final viscosity measured by a Brookfield viscometer at 20 rpm and 25 °C is maintained below 50 mPa·s; higher values can lead to incomplete discharge from a 16-gauge drench nozzle. Preservative efficacy is evaluated by Ph. Eur. 5.1.3 under the intended container closure. Because the extract contains reducing sugars, browning can occur if the vehicle is autoclaved; pasteurization at 80 °C for 30 min or aseptic filtration is preferred. Storage in amber glass or opaque HDPE at 15–25 °C reduces light-induced marker degradation. A precipitation stress test at 5 ± 3 °C for 7 days is a predictive tool for transport stability. Published data for this exact formulation are limited, so the stress-test operating range must be re-confirmed for each extract lot and final packaging configuration.

Hard capsule filling is not performed by direct addition of the liquid extract because free moisture migrates into the shell and produces zones of excess plasticization. The extract is first adsorbed onto microcrystalline cellulose or calcium silicate at a ratio of 1 part extract solids to 2–3 parts carrier. The resulting powder is dried to a water activity below 0.60 aw and milled through a 500 µm screen. The powder blend is then lubricated with 0.5% w/w colloidal silicon dioxide and 0.25% w/w magnesium stearate. Flowability is assessed by Ph. Eur. 2.9.36 powder flow, with an angle of repose below 40° considered acceptable for automatic capsule filling. Capsule fill weight is not controlled by volume alone because extract-carrier powder density varies with moisture; automatic tamping pin machines are set by fill weight rather than pin depth. Gelatin shells require storage below 60% RH at 15–25 °C; HPMC shells with lower equilibrium moisture may be used when fill hygroscopicity remains above 1.5% w/w moisture gain per hour at 50% RH. Dissolution testing in compendial media may show high variability because extract polysaccharides swell and form a viscous diffusion layer; the dissolution method should therefore include a surfactant such as 0.3% w/w sodium dodecyl sulfate in the aqueous phase. The absence of an official veterinary monograph means that dissolution acceptance criteria must be justified from pilot pharmacokinetic data, not from human botanical dietary supplement standards. This adsorption route is preferred for dry premix capsules and oral powders because it avoids the thermal stress of spray drying.

Residual Solvent Control and Drying Pathways Are Not Independent Variables

When the liquid extract is converted into powders or granules for tablets, premixes, or capsules, the drying method determines both residual solvent profile and marker retention. If ethanol or aqueous ethanol was used as the extraction solvent, residual ethanol in the final powder is measured by headspace gas chromatography according to Ph. Eur. 2.2.28 and classified under VICH GL18. Ethanol belongs to Class 3 and carries a limit of 5000 ppm in line with ICH Q3C Option 2, but veterinary products must still justify the limit in the target species. Vacuum drying at 50–60 °C and 0.08–0.10 MPa reduces residual ethanol below 5000 ppm within 8–12 h for a 1 cm bed depth. Spray drying on maltodextrin DE 15 at an inlet air temperature of 160–180 °C and outlet temperature of 70–85 °C produces a powder with bulk density of 0.35–0.50 g/mL and moisture below 5.0% w/w. The outlet temperature is critical because Z-ligustilide recovery declines sharply above 90 °C; published data for this extract under specific spray-drying settings are limited. To avoid unverified thermal degradation claims, each lot is tested by HPLC using a marker compound such as ferulic acid with quantification per Ph. Eur. 2.2.29. Vacuum belt drying is an alternative when the extract contains high sugar levels; belt speed and heating-zone temperatures are set to maintain wet mass temperature below 60 °C. Granule size distribution after milling is controlled through a 710 µm sieve for premix use and a 250 µm sieve for capsule powders. Sifting and blending should not be performed in areas with RH above 60% because dried powder regains moisture rapidly and may form hard agglomerates. The final moisture specification is not the only release criterion; residual ethanol, marker assay, bulk density, and sieve fraction are required to ensure reproducible downstream dosing.

Granulated oral powders for feed top-dress or water medication are produced by roller compaction rather than wet granulation when the extract has already been adsorbed onto a carrier. The primary process conflict is between adequate granule hardness and rapid disintegration in water at 37 °C. A dry blend containing 20–35% w/w adsorbate, lactose monohydrate, and 0.5% w/w magnesium stearate is compacted at a roll pressure of 3–5 MPa and milled to a 500–1,000 µm granule fraction. Granule hardness is not measured by tablet hardness testers; instead, the fraction remaining after tumbling in a Roche friabilitor for 10 min should exceed 95% w/w. Disintegration time in water at 37 °C should be below 3 min when measured by the method of Ph. Eur. 2.9.1 for dispersible granules. If disintegration time exceeds 5 min, surface polysaccharides have formed a gel plug; the formulation must be reworked with a disintegrant such as croscarmellose sodium at 2–5% w/w. Water medication cups and in-feed top-dress applications require that granules do not form a floating mat; the inclusion of 0.2% w/w food-grade simethicone emulsion may be necessary in high-moisture feed systems. The final product is packaged in aluminum-lined foil pouches because extract granules are hygroscopic and lose free-flow properties above 60% RH.