Tablet Compression of a Multi-Marker Herbal Powder: What Breaks First?

In a tableting line equipped with a 24-station rotary press with B-tooling, Xiao Chaihu powder containing Bupleuri Radix, Scutellariae Radix, Pinelliae Rhizoma, Glycyrrhizae Radix et Rhizoma, and Zingiberis Rhizoma Recens behaves as a poorly flowing, moderately hygroscopic botanical matrix. Bulk powder exposed to relative humidity above 60% RH is pre-dried at 45–50°C for 4–6 h before blending; otherwise the initial moisture load pushes the wet-massing stage into the picking range. The first failure observed in production is not insufficient hardness but picking and capping when residual moisture exceeds 5.0% w/w or when wet-massing time during high-shear granulation extends beyond 6–8 min, because the saponin fraction migrates to the granule surface under mechanical shear. Direct compression is confined to formulations with at least 20–30% w/w microcrystalline cellulose and 1.0–2.0% w/w colloidal silicon dioxide; even then, the Carr Index of the pre-blend typically remains above 25, requiring forced feeders and paddle speed adjustments. A 10-point blend uniformity study with an acceptance limit of relative standard deviation ≤ 5.0% for saikosaponin A and baicalin by HPLC is run before compression, because the multi-plant powder exhibits batch-to-batch marker variability. Tablets cut for swine dosing at a total weight of 1,000 mg are usually prepared with 500 mg of native powder, 25–35 mg of crospovidone, 12–20 mg of copovidone, and 5–10 mg of magnesium stearate; the compression speed is reduced below 35 rpm to minimise lamination caused by elastic recovery of the fibrous Pinellia fraction. Hardness values of 8–12 kp and friability below 1.0% under USP <1216> tablet friability testing are achieved when precompression is set at 2–4 kN and main compression is ramped to 18–22 kN. Coating with HPMC-based film is performed in a perforated pan at an inlet air temperature of 60–70°C, a product temperature of 38–42°C, and a spray rate of 8–12 g/min per kg of tablet load; the coating protects the saponin-rich core from oxygen and moisture ingress but does not remediate already fractured cores. The main incompatibility in tablet development is prolonged contact with highly alkaline lubricants or wet binder systems containing strong alkali, which promote deacylation of saikosaponin glycosides; neutral or slightly acidic granulating fluids at pH 5.5–6.8 are preferred. Dissolution release of baicalin in 0.1 M hydrochloric acid is typically limited, so an enteric or buffered subcoat may be required if early intestinal release claims are part of the veterinary target product profile.

Sterile ampoules of Xiao Chaihu require an extraction sequence that determines whether the finished product remains clear at pH 5.5–6.8 and whether the bacterial endotoxin load is reduced below the parenteral threshold. The powder is first dispersed in water for injection at a liquid-to-solid ratio of 8:1 to 10:1 and extracted at 90–95°C for 2 h; the decoction is then subjected to ethanol precipitation at 60–70% v/v and held at 2–8°C for 12–24 h to remove polysaccharides, glycoproteins, and heat-coagulable plant debris. Ethanol is recovered under vacuum at a jacket temperature not exceeding 65°C, because higher sustained temperature accelerates hydrolysis of saikosaponin A and baicalin glucuronides. Depyrogenation is performed with activated carbon at 0.1–0.3% w/v and 60°C for 30 min, followed by a 0.45 µm prefilter and a 0.22 µm PVDF membrane; the filtrate is filled into amber glass ampoules under nitrogen with headspace oxygen below 2.0% v/v. Terminal sterilization at 121°C for 15 min is generally not compatible with the aqueous multi-component decoction, because heat exposure can produce flocculent precipitation and an unacceptable insoluble particle count; aseptic filtration is therefore the default route when sterility under USP <71> is claimed. Bacterial endotoxin limits are calculated from the maximum intended veterinary dose volume and are commonly aligned with USP <85> injectable monographs, with an example threshold of ≤ 0.5 EU/mg when the injected dose is normalised by powder mass. The main operational boundary is the pH window: below pH 4.0, saikosaponin glycosides lose their terminal sugar moieties and the solution gradually darkens, while above pH 8.0, baicalin glucuronides become susceptible to oxidation. Published data for the specific configuration of Xiao Chaihu injection in companion animals is limited; therefore, the extraction, filtration, and packaging parameters must be revalidated per species and per batch because marker content varies with botanical seasonality.

Capsule Filling of High-Saponin Powder and the Moisture Window

When filling hard gelatin capsules with Xiao Chaihu powder, the primary constraint is the moisture window required to keep the shell from becoming brittle or tacky on a continuous dosator machine. The powder is pre-sieved through a 40-mesh screen to remove fibrous agglomerates and adjusted to a loss-on-drying value below 5.0% w/w under USP <731>; at higher moisture, the saponin-rich particles adhere to polished dosator pins and cause fill weight drift. A typical capsule blend for companion-animal dosing contains 100–200 mg of native Xiao Chaihu powder per #1 capsule, with 20–30% w/w microcrystalline cellulose as a diluent, 1.0–2.0% w/w colloidal silicon dioxide as a glidant, and 0.25–0.5% w/w magnesium stearate as a lubricant; the pre-compacted blend is adjusted to a bulk density of 0.45–0.65 g/cm³ so that the dosator chamber discharges a consistent slug. Fill weight uniformity studies are run at 25 ± 2°C and 45–55% RH, with a target relative standard deviation of less than 3.0% across 20 consecutive capsules. Hard gelatin capsules filled with hygroscopic botanical powder are packed with desiccant canisters and stored at 25°C/60% RH because moisture transfer from the shell into the powder can embrittle the shell over a 3-month stability window. Disintegration is affected by the hydrophobic surface film created by excessive magnesium stearate; if the dosage form must pass veterinary oral disintegration limits, the addition of 2–5% w/w croscarmellose sodium or 5% w/w sodium starch glycolate is preferred over extended mixing time. The main incompatibility is direct filling of powder that has not been de-lumped, because the fibrous Pinellia and jujube fractions segregate during transfer and produce capsules with a marker content variation exceeding the acceptance criterion of USP <905>. Capsules are typically administered as an adjunct to antimicrobial or antiparasitic therapy in dogs and cats; published data for the specific configuration of Xiao Chaihu powder in feline capsules is limited, so marker-normalised dosing rather than total powder weight is required.

Unlike granulated premixes, direct-use Xiao Chaihu powder relies on particle-size uniformity to prevent separation in creep feeders and feed top-dress applications. The native powder is blended with a water-soluble carrier such as lactose monohydrate or dextrose at a dilution of 1:10 to 1:50 before point-of-use administration, because the raw botanical powder is too cohesive to flow uniformly through a rotary valve. The carrier is selected to have a median particle size within 80–120 µm to match the saponin-rich powder and reduce percolation during mixer discharge. Batch mixing in a double-cone or V-blender is conducted at 60–70% of vessel capacity for 15–20 min, with a lump-breaking pass through a 40-mesh screen after the first 5 min to disperse jujube and Pinellia fragments. Direct powders are susceptible to dust generation; dust is controlled by adding 0.5–1.0% w/w of a food-grade vegetable oil or by keeping the finished blend moisture at 3.0–5.0% w/w. However, oil addition above 1.0% w/w can coat the saponin-rich particles, reduce water dispersibility, and create a sticky film on stainless-steel feed augers. The primary field failure observed in direct powder administration is marker content variation along the feed line; therefore, a sampling plan of at least 10 points per tonne is used for saikosaponin A and baicalin analysis by LC-MS/MS, with an acceptance limit of ± 10% of labelled marker content. Direct powder administration is used in nursery pigs as a short-term adjunct for reduced feed intake associated with respiratory stress; the powder is given as a top-dress rather than as a complete-feed incorporation to avoid segregation in long auger systems.

Granulation Yield Drops When Binder Level Exceeds 4.0% w/w

A top-spray fluid bed processing Xiao Chaihu powder shows a sharp yield drop when dry-binder level exceeds 4.0% w/w or falls below 1.5% w/w, because overwetting and twin formation dominate at high binder levels while low binder produces unacceptable fines below 80 µm. Aqueous polyvinylpyrrolidone K30 solution is sprayed to achieve a dry-binder level of 2.0–4.0% w/w, with an inlet air temperature of 55–65°C, a product temperature of 30–35°C, and an inlet air dew point below 10°C to prevent surface crusting. The saponin-rich Xiao Chaihu matrix exhibits natural surfactant behaviour, so foaming during binder addition is controlled by reducing atomisation air pressure to 1.5–2.0 bar and by intermittent spraying. Agglomeration is continued until the granule fraction between 16-mesh and 40-mesh exceeds 80% w/w; undersized and oversized granules are dry-screened and recycled, but recycle fractions above 30% w/w of the batch introduce marker heterogeneity because the recycled fines are enriched in baicalin and depleted in the more fibrous Pinellia fraction. Drying is stopped at a loss-on-drying of 2.0–4.0% w/w; overdrying below 2.0% w/w increases granule friability and generates dust during sachet filling, while residual moisture above 4.5% w/w leads to caking in aluminium-foil sachets under accelerated stability at 40°C/75% RH. The resulting granules are intended for reconstitution in drinking water or for direct oral syringe administration after dispersion in water at a concentration of 50 mg/mL; the dissolution of baicalin from granules is faster than from compressed tablets because the wet granulation step partially pre-dissolves the glucuronide fraction and deposits it on the lactose surface. High-shear granulation is less suited to this powder because the Pinellia fibre absorbs water unevenly and produces dense, slow-dissolving granules with a bimodal size distribution, unless a pre-wetting step of 60–90 s is applied before binder addition. Tablets and dry powders are generally preferred when the dosing route must avoid any reconstitution step that could expose the saponin fraction to microbial growth in stagnant water lines.

Because feed premixes are diluted at 1:200 to 1:2,000 in final rations, Xiao Chaihu powder must be pre-blended onto a carrier with matching particle size and low residual moisture before it enters the main feed mixer. Corn cob, calcium carbonate, or wheat middlings with a D50 of 150–250 µm are preferred carriers; fine limestone below 100 µm increases dust and marker segregation, while coarse carriers above 400 µm create a fast-flowing fraction that passes through auger flights and leaves saponin-rich fines behind. The powder is first mixed with 0.5–1.0% w/w soybean oil or propylene glycol in a ribbon mixer for 3–5 min to bind fine particles to the carrier surface. Sampling at 10 points per batch with an acceptance coefficient of variation ≤ 5.0% for baicalin is used to verify homogenous distribution; higher coefficients of variation have been traced to worn ribbon clearances exceeding 5 mm and to overfilled mixers above 70% nominal capacity. After mixing, the premix is passed over a 20-mesh safety screener and a plate magnet, and the transfer line is flushed with 1 kg of ground corn per 100 kg batch to control carryover into subsequent non-medicated batches. A major processing boundary is pelleted feed: exposure to pelleting temperatures of 75–85°C for 20–30 s can reduce marker recovery of saikosaponin A by thermal desiccation; therefore, post-pelleting liquid application or a lower-temperature pelleting profile is preferred when Xiao Chaihu premix is included. The premix should not be blended with a high-choline or high-copper mineral pack at ambient humidity above 70% RH, because hygroscopic choline chloride and copper sulfate can initiate local moisture uptake and cause the botanical fraction to cake within 48 h. Published data for pelleted feed stability of the complete Xiao Chaihu marker set is limited; batch-specific recovery studies are recommended before final ration formulation.

When Xiao Chaihu Powder Replaces Fluid Extract in Oral Solutions, Clarification and Preservative Partitioning Change

The replacement of a fluid extract with dry Xiao Chaihu powder alters oral solution behaviour in two immediate ways: the powder contributes water-insoluble plant residues that raise filtration load, and the saponin fraction can form colloidal aggregates that bind preservative molecules and reduce their free concentration. The powder is dispersed in purified water at 20–50 mg/mL, adjusted to pH 5.5–6.8 with citrate or phosphate buffer, and hydrated under low-shear agitation for 45–60 min. The mixture is clarified through a 0.45–1.0 µm polyethersulfone filter at 20–25°C; tighter membranes below 0.45 µm remove polysaccharide stabilisers and may reduce baicalin recovery by 2–5%, while looser filters leave visible sediment after 24 h at 2–8°C. If a suspension rather than a solution is desired, 0.2–0.4% w/v xanthan gum or microcrystalline cellulose/carboxymethylcellulose sodium is added after filtration, but the hydration must be complete before benzoate addition because the gum network can trap preservative pockets. Sodium benzoate at 0.1–0.2% w/w and potassium sorbate at 0.1% w/w are used as preservatives; antimicrobial effectiveness is confirmed under USP <51> because botanical colloids and polysaccharides can reduce the free preservative fraction below the killing threshold. The solution or suspension is filled into amber polyethylene terephthalate bottles with a nitrogen-purging step to keep headspace oxygen below 2.0% v/v. The main incompatibility is quaternary ammonium preservatives such as benzalkonium chloride, which interact with anionic botanical colloids and produce stringy precipitates; this formulation route should be avoided unless phase-behaviour screening demonstrates compatibility at the final concentration. The shelf life of a reconstituted Xiao Chaihu oral liquid is typically shorter than that of the dry powder, with a use period of 14–28 d after first opening when stored at 15–25°C; beyond this interval, preservative depletion and oxidative discolouration of baicalin are the limiting failure modes.