Direct compression platforms require the API to be pre-sieved through a 425 µm screen before dry blending. The free carboxylic acid moiety on (R)-(−)-3-carbamoylmethyl-5-methylhexanoic acid creates a mildly acidic particle surface that can promote adhesion to uncoated tool steel when main compression force exceeds 22 kN on high-speed rotary presses. In production campaigns exceeding 8 h, punches fabricated from D2 steel with chromium nitride physical vapour deposition are preferred. A representative tablet formula for a 100 mg strength is: API 45.0% w/w, microcrystalline cellulose PH-102 38.5% w/w, croscarmellose sodium 5.0% w/w, colloidal silicon dioxide 1.0% w/w, sodium stearyl fumarate 1.5% w/w. Blending is performed in a 600 L V-blender at 12 rpm for 15 min after a geometric dilution step. The blend is sampled from 10 locations for blend uniformity per FDA 21 CFR 211.110 with acceptance 90.0–110.0% label claim and RSD ≤5.0%. Compression on a 45-station rotary press uses pre-compression of 4–5 kN and main compression of 12–18 kN to maintain tablet hardness 70–100 N. Friability testing is performed according to USP <1216>; the acceptance limit is NMT 0.8%. Disintegration in 900 mL of purified water at 37±2°C must be complete within 5 min for uncoated cores per USP <701> and Ph Eur 2.9.1. Content uniformity follows USP <905> with acceptance value ≤15.0. Dissolution testing is carried out in 900 mL of pH 6.8 phosphate buffer at 37±0.5°C, paddle speed 75 rpm, with sampling points at 10, 15, 20, 30, 45 min; the specification is Q=80% at 30 min. Film coating with an aqueous dispersion at 3.0% w/w weight gain is performed in a perforated pan coater with inlet air 60–70°C, product bed temperature 38–45°C, and pan speed 2–6 rpm. The final tablet is stored in HDPE bottles with desiccant; water activity is maintained below 0.60.

Chiral purity and related substance control on the direct compression blend use a validated HPLC method with a chiral stationary phase consisting of amylose tris(3,5-dimethylphenylcarbamate) coated on 5 µm silica, 250 mm × 4.6 mm column dimensions. The mobile phase is n-hexane:ethanol:trifluoroacetic acid 80:20:0.1 v/v/v, flow rate 1.0 mL/min, injection volume 10 µL, column temperature 30°C, and UV detection at 210 nm. The undesired enantiomer is quantitated by area normalization with a limit of NMT 0.15%; linearity is demonstrated from 0.05% to 1.0%, and sensitivity, carryover, and precision follow ICH Q2(R1).

Why does capsule filling require moisture below 2.0% w/w when the API controls blend flow?

Low-shear encapsulation is restricted by the powder’s Carr index and the hygroscopic response of the hard gelatin shell. After dry blending, the API is combined with lactose monohydrate 200 mesh, pregelatinized starch, and sodium starch glycolate in a 200 L bin blender rotating at 10 rpm for 18 min. The blend is sampled from the top, middle, and bottom; blend uniformity must be 95.0–105.0% label claim with RSD ≤5.0%. Hard gelatin capsule shells are conditioned at 20–25°C and 45–50% RH for at least 48 h before filling. If the blend moisture exceeds 2.0% w/w by Karl Fischer titration (USP <921>), powder adhesion to tamping pins increases and fill weight drift exceeds ±3.0% on a 150,000 capsule/h tamping-pin machine. A desiccant such as silica gel 0.5–1.0% w/w is added when the incoming API lot shows water content above 1.0% w/w. The fill weight for a size 1 capsule is typically 250 mg with an acceptance range of ±5.0%; machine settings use tamper pins set to 12–14 mm stroke and dosing chamber vacuum of 0.6–0.8 bar. The closed capsule is subjected to metal detection and weight sorting. Dissolution testing in 900 mL of 0.1 M hydrochloric acid at 37±0.5°C with basket speed 100 rpm uses Q=75% at 45 min for immediate-release capsules per USP <711>. Gelatin cross-linking is evaluated by the two-tier dissolution test per USP <711>; if the capsule fails the first medium and passes after enzyme addition of pepsin, cross-linking is documented but the batch is not released without stability verification. The secondary package includes a 1 g canister of molecular sieve desiccant per 60 cc HDPE bottle to keep closure water vapour transmission below 0.2 mg/day at 40°C/75% RH. For the acidic API, lubricant selection uses sodium stearyl fumarate 0.5% w/w instead of magnesium stearate, since the free carboxylic acid can hydrolyze magnesium stearate during prolonged storage at 40°C/75% RH and generate stearic acid crystals that reduce dissolution rate.

In a high-shear wet granulation line, the API is premixed with lactose monohydrate and maize starch at a 1.0:2.5:0.5 ratio in a 600 L high-shear granulator. The binder solution contains PVP K30 4.0% w/w dissolved in purified water adjusted to pH 5.0±0.2 with 0.1 M citric acid. The solution is sprayed through a 2.0 mm two-fluid nozzle at a rate of 200–250 g/min. Impeller speed is 150 rpm during binder addition and 250 rpm during wet massing for 90 s; chopper speed is maintained at 800 rpm. Wet granules are discharged through a 4.0 mm aperture and dried in a fluid bed dryer with inlet air 65±5°C and product temperature 35–40°C until loss-on-drying is 1.5–2.5% w/w. Dried granules are milled through a 1.0 mm screen using a conical mill at 2000 rpm with blunt-edge impeller. The resulting granule fraction has bulk density 0.48–0.56 g/mL, tapped density 0.62–0.70 g/mL, and Hausner ratio 1.25–1.35. Granules are filled into stick-pack sachets on a vertical form-fill-seal machine using laminated aluminium foil; fill weight 500 mg ± 3.0%. Seal integrity is tested per ASTM F2054 with burst pressure ≥180 kPa; vacuum leak testing uses methylene blue dye with no color wicking. The finished granular dosage form is reconstituted in 50 mL of water to form a suspension; the pH of the reconstituted mixture is 5.0–6.0. Dissolution of granules is assessed by USP <711> using a 900 mL pH 6.8 phosphate buffer, paddle 75 rpm, with sampling at 10, 20, 30, 45 min; Q=80% at 30 min. Microbial limits for non-sterile granules follow USP <61> and USP <62>; total aerobic microbial count ≤10³ CFU/g and total combined moulds and yeasts ≤10² CFU/g.

When terminal sterilisation is replaced by aseptic filtration for the free acid parenteral

Parenteral processing changes when the free carboxylic acid exhibits solution-state instability at autoclave temperatures above 110°C. A solution formulation is prepared by adding the API to 80% of the final volume of Water for Injection at 20–25°C, then adjusting to pH 6.8±0.3 with 1 M sodium hydroxide. The carboxylate salt increases solubility to ≥50 mg/mL at 25°C; published data for this specific compound’s temperature-dependent solubility in phosphate-buffered saline is limited, but the pH adjustment point is confirmed by potentiometric titration at 25°C. Tonicity is adjusted with sodium chloride to 280–320 mOsm/kg per USP <785>. The solution is sparged with pharmaceutical-grade nitrogen for 20 min before filtration, and dissolved oxygen is maintained below 2.0 ppm. Pre-filtration is performed through a 0.45 µm PVDF cartridge at 50–100 L/m²/h, followed by sterilizing filtration through a 0.2 µm double-layer PVDF filter, with a bubble point test before and after filtration per ASTM F838-20. The filtration train is validated for adsorptive losses; filtrate recovery at the first 5 L is tested for assay by HPLC and must be 98.0–102.0% of target. Filling is executed on an aseptic line inside an ISO Class 5 unidirectional airflow zone under EU GMP Annex 1 Grade A conditions, with a Grade B background. A maximum aseptic holding time of 4 h at 15–25°C is applied before filling. The filled vials are sealed with 20 mm bromobutyl rubber stoppers and subjected to visual inspection per USP <790>. Subvisible particulates are measured by light obscuration per USP <788>; limits for containers ≤ 100 mL are ≥10 µm ≤ 6000 per container and ≥25 µm ≤ 600 per container. Sterility testing follows USP <71> and Ph Eur 2.6.1 with direct inoculation or membrane filtration. Bacterial endotoxins are controlled according to USP <85> and Ph Eur 2.6.14; for an injectable product with a maximum dose of 100 mg, the limit is calculated as 0.25 EU/mg. The terminal product is stored upright at 2–8°C when photo-stability studies indicate light sensitivity; any light-protective overwrap uses amber film validated to reduce UV-A transmittance by ≥90%.

When annealing is incorporated into the freeze-drying cycle, the lyophilised dosage form reduces amorphous phase heterogeneity when 4.0% w/w mannitol is combined with 2.0% w/w sucrose as bulking agents. The filling solution is prepared at 50 mg/mL API and filtered as described for the parenteral solution. Vials of 20 mL Type I borosilicate glass are filled to 5.0 mL fill volume and partially stoppered with bromobutyl closures that allow water vapour transfer during freeze-drying. The freeze-dry cycle begins with shelf cooling to −40°C at 0.5°C/min and a 2 h hold. Annealing is performed at −20°C for 4 h to permit crystallisation of mannitol, which reduces primary drying resistance. Primary drying is run at shelf temperature −25°C and chamber pressure 80–120 mTorr for 48 h; the product temperature is held below the collapse temperature by ≥2°C. Secondary drying is executed at shelf temperature 35°C for 8 h until residual moisture is ≤1.0% w/w by Karl Fischer (USP <921>). Cake appearance is evaluated for height, color uniformity, and absence of melt-back. Reconstitution time in 5.0 mL of Water for Injection must be ≤120 s with gentle swirling; foaming is controlled by pre-wetting the cake with 1.0 mL of WFI and holding for 30 s before final dilution. The reconstituted solution is tested for pH 6.5–7.5, clarity per USP <790>, and subvisible particles per USP <788>. Container closure integrity is verified by dye ingress or helium leak per USP <1207>; the stopper is seated under 500 mbar vacuum. The lyophilised product is stored at 2–8°C and protected from light.

Roller-compacted granulate parameters and loss-on-drying limits

Dry granulation by roller compaction is used when the API is sensitive to aqueous granulation and direct compression yields poor flow. The API is first blended with microcrystalline cellulose PH-200, lactose monohydrate, and crospovidone in a 300 L bin blender at 10 rpm for 15 min. The pre-blend is fed into a roller compactor with smooth rolls of 200 mm diameter and 80 mm width. Roll pressure is maintained at 80–120 kN/m of roll width, roll speed 3–5 rpm, and screw feeder speed 20–40 rpm. Ribbons are milled through a 1.0 mm screen using a conical mill at 1800 rpm to produce granules with D50 180–250 µm. The granulate is lubricated with 0.5% w/w sodium stearyl fumarate for 3 min and compressed into tablets at main compression 10–15 kN. Loss-on-drying of the granulate is controlled at 0.5–1.5% w/w; below 0.5% w/w, static charge increases and tablet hardness decreases; above 1.5% w/w, sticking on upper punches occurs after 45 min of compression. Tablet hardness is 60–90 N, friability NMT 0.8% per USP <1216>, and disintegration time NMT 10 min per USP <701>. The roller-compacted tablet is coated with a moisture barrier film at 2.5% w/w weight gain and packaged with desiccant. Dissolution is tested in 900 mL of pH 4.5 acetate buffer at 37±0.5°C, paddle 50 rpm, sampling at 15, 30, 45, 60 min; Q=70% at 60 min for a non-immediate-release control strategy. Chiral purity and related substances are tested on the granulate, tablets, and stability samples by HPLC; the chiral impurity limit is NMT 0.15% and total impurities NMT 0.5% per ICH Q3A.