Direct compression of chlorotris(triphenylphosphine)rhodium(I) at a 0.1% w/w drug load is executed only after the API has been pre-sieved through a 150 µm stainless-steel mesh inside a nitrogen-flushed isolator set to ≤ 40% RH and 15–25 °C. The geometric dilution is performed with spray-dried lactose monohydrate at a 1:1 active-to-excipient fold ratio until a 10% w/w triturate is obtained; the triturate is then charged into a 600 L bin blender with microcrystalline cellulose (Ph. Eur. 4.4.2) and crospovidone (USP/NF) at 12 rpm for 20 min. Subsequent lubrication with magnesium stearate is limited to 0.25% w/w and 3 min to avoid shearing the triphenylphosphine ligand shell and generating catalytically active degradation fragments. Compression is run on a rotary press with 10 station tooling at 20–40 rpm, 8–12 kN main compression force, and 70 N target hardness. Friability is monitored per USP <1216> with a single run of 6.5 g for 100 revolutions; weight uniformity is verified per USP <905>. Aqueous film coating is not applied unless a non-aqueous seal coat has been deposited from an ethanol-acetone system and residual solvent has been reduced below USP <467> limits. Terminal tablets are white to pale yellow, packaged in cold-form foil with 1.0 g molecular sieve sachet, and labeled for controlled room temperature storage.

The principal processing barrier is low-dose segregation in the feed frame. Without a forced feeder, blend potency variability exceeds 6% RSD after 15 min of press operation; a rotating paddle feeder at 20 rpm maintains content uniformity below 3% RSD. The API is not screened through a cone mill because the high shear raises localized temperature above 60 °C, which accelerates triphenylphosphine oxidation to triphenylphosphine oxide. That oxidation product is quantified by a validated HPLC method with UV detection at 210 nm and a quantitation limit of 0.05 area%.

What Limits Dry Granulation of the Rh(I) Complex into Hard Capsule Fill?

Roller compaction is selected when direct compression fails flow specifications. A 0.15% w/w API blend containing mannitol (60.0% w/w), pregelatinized starch (25.0% w/w), croscarmellose sodium (2.0% w/w), and silicon dioxide (0.5% w/w) is compacted at 18 kN/cm roll force, 2.0 mm roll gap, and 8 rpm roll speed on an instrumented roller compactor with a side seal. The compacted ribbon is milled through a 1.25 mm rasp screen; granules are collected with a target particle size distribution of 0.2–0.8 mm and not more than 20% w/w fines below 100 µm. Fines above that threshold cause capsule weight variation to exceed 4% RSD on a dosator-type capsule filler. The granules are filled into size 3 hydroxypropyl methylcellulose capsules at 25,000 capsules/hour with target fill weight 120 mg and acceptance limits of ±3 mg. In-process controls include bulk density (0.45–0.65 g/mL) and angle of repose (35–40°) measured per USP <1174>. Dissolution testing follows USP <711> apparatus 2 at 50 rpm in 0.1 N HCl; capsules are stored in HDPE bottles with induction-sealed liners and 1 g desiccant.

Residual moisture after compaction must remain below 1.5% w/w by Karl Fischer (USP <921>). Ambient humidity above 50% RH causes the fines to adhere to the compaction rolls, producing ribbon splitting and a 10–15% reduction in yield. If magnesium stearate is omitted, the ribbon adheres to the rolls; if it is added above 0.5% w/w, capsule dissolution drops below 80% at 45 min due to hydrophobic film formation.

When the API Is Dispersed by Non-Aqueous Wet Granulation for Oral Granules

Oral granules in sachets require a coarser, non-friable particle that cannot be obtained by roller compaction alone. A non-aqueous wet granulation process is used with an 8.0% w/w binder solution of povidone (K30) in anhydrous ethanol. The API is dissolved in the binder phase at 0.05% w/w of the final granule mass, thereby avoiding discrete API particles and improving content uniformity. The granulation is performed in a high-shear mixer with an impeller speed of 150 rpm and a chopper at 1,500 rpm for 180 s. Wet mass is passed through a 1.5 mm screen and dried in a fluid bed at 40 °C for 20–30 min until loss on drying is ≤ 1.0% w/w. Ethanol residual is confirmed below 5,000 ppm by USP <467>. The dried granules are sized through a 1.0 mm sieve and blended with 0.5% w/w colloidal silicon dioxide before sachet filling. Fill weight is 500 mg with ±2.5% limits. Disintegration of the granules is tested per Ph. Eur. 2.9.1 in 37 °C water; complete dispersion occurs within 5 min when the granules contain 2.0% w/w crospovidone. This route avoids aqueous granulation because the API undergoes rapid ligand exchange in the presence of water and chloride ions, yielding a hydroxy-bridged degradation product.

Non-Aqueous Injectable Solution Production in Type I Glass Vials

Parenteral administration requires complete dissolution of the Rh(I) complex in a non-aqueous vehicle that can be sterile-filtered. A development vehicle containing 60% v/v PEG 400, 30% v/v dehydrated ethanol, and 10% v/v propylene glycol is pre-filtered through a 0.45 µm PVDF membrane and then chilled to 2–8 °C before the API is added under nitrogen sparging at 0.1 mg/mL. The bulk solution is mixed with a magnetic stirrer at 150 rpm for 30 min, then sterilized by passage through a 0.22 µm PVDF capsule filter at 1.5 bar differential pressure. Terminal steam sterilization at 121 °C is not used because the triphenylphosphine ligands oxidize to triphenylphosphine oxide within 15 min under autoclave conditions, causing visible yellow-to-brown discoloration. Filling is performed under Grade A laminar flow within an ISO 14644-1:2015 Class 5 cleanroom, with filling volume 2.0 mL ± 0.05 mL into 2 mL amber Type I glass vials. Headspace oxygen is displaced with sterile-filtered nitrogen to ≤ 2% O₂. Each lot is tested for sterility (USP <71>), bacterial endotoxins (USP <85>), particulate matter (USP <788>), and elemental impurities by ICP-MS per ICH Q3D. Rhodium is controlled as a Class 2B elemental impurity with an injectable permitted daily exposure consistent with the relevant route-specific PDE.

Solubility screening in aqueous media indicates precipitation at pH 4.0 and at chloride concentrations above 0.9% w/v. Therefore the product is not diluted with saline before administration unless a compatibility study at the intended final concentration has been completed. In-line filter integrity is tested before and after filling by the bubble point method; PVDF membrane failure at ≤ 2.5 bar leads to batch rejection. Process simulations use sterile culture media to demonstrate a contamination rate below 0.1%.

Lyophilization of chlorotris(triphenylphosphine)rhodium(I) from a tert-butanol-water cosolvent produces a stable, amorphous cake for reconstitution. The API is dissolved in warm tert-butanol at 35 °C and then mixed with 9% w/v mannitol in WFI at a final tert-butanol fraction of 20% v/v; at 0.1 mg/mL no precipitation is observed before the freezing ramp. The solution is filled into 5 mL tubing vials with 0.5 mL volume and lyophilized using a cycle that freezes the shelf to -40 °C at 1.0 °C/min, holds for 120 min, then applies primary drying at -25 °C and 0.10 mbar for 36 h. Secondary drying is ramped to 25 °C at 0.5 °C/min and held for 6 h at 0.05 mbar. Residual moisture is measured by Karl Fischer (USP <921>) and controlled to ≤ 1.0% w/w. The vial headspace is backfilled with nitrogen to 0.8 atm before stoppering. Reconstitution is performed with 5.0 mL of 5% dextrose injection; full dissolution occurs within 120 s with gentle swirling. The reconstituted solution is not stored for more than 4 h at 2–8 °C because the complex undergoes slow solvolysis. Visual inspection rejects cakes with shrinkage, meltback, or color darker than pale yellow. Container closure integrity is verified by dye ingress per USP <1207>. This dosage form is used where an aqueous solution is required but the API has insufficient long-term solution stability.

Control of Rhodium Migration from Lipid-Based Oral Capsule Filling

Self-emulsifying lipid filling of HPMC capsules is assessed when the API is released in a fed-state simulation and requires protection from gastric hydrolysis. A lipid preconcentrate composed of 40% w/w medium-chain triglycerides, 35% w/w polyoxyl 35 castor oil, and 25% w/w propylene glycol monocaprylate is prepared under nitrogen and dried with 0.5% w/w anhydrous sodium sulfate before the API is dissolved at 0.2 mg/g. The fill mass is 250 mg per size 1 HPMC capsule; filling is performed at 30 °C with a semi-automatic liquid capsule filler at 2,000 capsules/hour. Bubble-free filling requires degassing at -0.07 MPa for 20 min. Capsule shells are band-sealed with 25% w/w gelatin solution and stored below 25 °C. The main process conflict is oxidative degradation of the triphenylphosphine ligand by peroxides present in polyoxyl 35 castor oil. Peroxide value of the lipid phase is therefore controlled to ≤ 2 meq/kg by Ph. Eur. 2.5.5; above that threshold, the fill color shifts to brown within 7 days at 40 °C/75% RH. Compatibility is confirmed by monitoring triphenylphosphine oxide via HPLC after 14 days at 5 °C and 25 °C. Dissolution testing uses USP <711> apparatus 2 at 75 rpm with 900 mL of pH 6.8 phosphate buffer containing 0.5% w/w sodium lauryl sulfate. The dispersed oil droplet size after aqueous dilution is 100–300 nm as measured by dynamic light scattering. This lipid fill route is not used with gelatin capsules containing free water above 13% w/w.