In direct compression of (S)-oxetan-2-ylmethanamine as a low-dose oral tablet, the API is first pre-blended with milled mannitol at a ratio of 1:10 before transfer to a 600 L bin blender. This initial pre-blend reduces segregation risk because the API particle-size distribution is typically finer than the filler. The main blend is prepared with mannitol or dibasic calcium phosphate dihydrate as the filler, crospovidone at 2–5% w/w as the disintegrant, colloidal silicon dioxide at 0.25–0.5% w/w as the glidant, and magnesium stearate at 0.5–1.0% w/w as the lubricant. Lactose monohydrate is not used because the primary amine in the API can undergo Maillard-type condensation with reducing lactose, producing brown discoloration and related substances. Published stability data for this exact amine-lactose interaction is limited, but the known amine-carbonyl reaction is sufficient to exclude lactose from initial formulation screening. Blending is performed in a bin blender at 10 rpm for 15 min, with blend uniformity sampling per FDA 21 CFR 211.110; if the relative standard deviation of API content exceeds 4.0%, the pre-blend ratio is increased to 1:20 or a milling step is introduced.

Compression is carried out on a rotary tablet press equipped with B-tooling and a force feeder. A representative 10 mm round tablet core with a total weight of 200 mg is compressed at 6–12 kN to achieve tablet breaking force of 40–80 N. Friability is controlled below 0.8% according to USP <1216>, and disintegration is specified at not more than 15 min in water at 37 °C according to USP <701>. Content uniformity follows USP <905> with an acceptance value of not more than 15. Because the API may be hygroscopic, incoming moisture is monitored; if moisture exceeds 1.0% w/w, pre-drying at 40–50 °C is applied before blending. Compression rooms are maintained below 40% RH. The terminal product is an immediate-release film-coated tablet sealed in aluminum/aluminum blister packs with desiccant.

Wet Granulation Intermediates for Controlled-Release Matrix Tablets

A high-shear wet granulation route is selected when the API is to be embedded in a hypromellose-based controlled-release matrix. The API is first dispersed in purified water or dissolved as the hydrochloride salt, then added to a high-shear granulator with a 65 L bowl. The binder solution contains hypromellose at 5% w/w; the liquid-to-solid ratio is maintained at 0.08–0.12 kg/kg. Granulation is performed with an impeller tip speed of 4–8 m/s and a chopper speed of 1500 rpm for 2–5 min. This energy range is selected to produce dense granules without excessive fines, but it must be qualified against API degradation because the oxetane ring may be sensitive to local temperature rises during wet massing. After granulation, the wet mass is dried in a fluid-bed dryer with inlet air at 60–70 °C and product temperature at 35–40 °C until loss on drying reaches 1.5–2.5% w/w. The dried granules are milled through a conical mill fitted with a 813 µm screen at 800 rpm.

The milled granules are blended with hypromellose K15M at 20–40% w/w, microcrystalline cellulose at 20–30% w/w, and magnesium stearate at 0.5–1.0% w/w. The API load in the final granule is typically 5–20% w/w. Dissolution testing is performed according to USP <711> Apparatus 2 at 50 rpm in pH 6.8 phosphate buffer; published dissolution data for this specific API matrix is limited, so the acceptance criterion is established during development rather than copied from a standard monograph. The terminal dosage form is a controlled-release film-coated tablet. During scale-up, batch-to-batch variance in granule density must be monitored because changes in granule porosity alter the diffusion path length and release rate from the hypromellose matrix.

For low-dose hard-gelatin capsule filling, particle-size control is the primary determinant of content uniformity. The API is pre-blended with microcrystalline cellulose at 1:5 and then geometrically diluted with a mannitol/starch mixture to a final drug load of 0.5–5.0% w/w. The final powder blend contains microcrystalline cellulose 50–70% w/w, mannitol 20–40% w/w, crospovidone 2–5% w/w, colloidal silicon dioxide 0.25–0.5% w/w, and magnesium stearate 0.5–1.0% w/w. Powder flow is controlled to a Carr index of 15–25 and a Hausner ratio of 1.2–1.4. If flow falls outside this range, a tamping-pin capsule filler is preferred over a dosator machine because dosator systems are more sensitive to blend compressibility and can produce fill weight drift during long runs. Capsule filling is performed on a tamping-pin machine at 50,000–100,000 capsules per hour with fill weight control at not more than ±3% relative standard deviation.

A representative fill is 150 mg of powder blend into a size 3 hard-gelatin capsule. The capsule shell is equilibrated at 35–45% RH before filling to prevent shell brittleness and crosslinking. If the API is hygroscopic, hypromellose capsule shells are substituted for gelatin. Dissolution is tested according to USP <711> Apparatus 1 at 100 rpm in 0.1 M hydrochloric acid. Container closure integrity follows USP <671>. The terminal product is a hard-shell immediate-release capsule in a high-density polyethylene bottle with a desiccant canister.

What Limits Sterile Filtration Throughput for Aqueous Amine API Solutions?

Aqueous ready-to-use injectable solutions of (S)-oxetan-2-ylmethanamine require careful filter selection because a primary amine can adsorb to nylon and, to a lesser extent, to some charged polyvinylidene fluoride membranes. The formulation is prepared with the API at 10 mg/mL in water for injection, adjusted to pH 4.5–5.5 with an acetate buffer at 5–20 mM. Tonicity is adjusted with sodium chloride at 0.9% w/v or mannitol at 5% w/v. Nitrogen blanketing is applied throughout compounding because the free-base form can react with atmospheric carbon dioxide, causing pH drift. Sterile filtration is performed through a 0.22 µm low-protein-binding polyethersulfone or polyvinylidene fluoride membrane. Filter adsorption studies are conducted per PDA Technical Report 26; if API adsorption exceeds 2%, the membrane area is increased or the first filtrate volume is discarded. Terminal sterilization, where selected, is performed at 121 °C for 15 min with an F0 of not less than 12. If heat-labile degradation products exceed 0.5% during terminal sterilization, aseptic filtration and filling in an EU GMP Annex 1 Grade A environment is used instead.

Primary packaging is Type I borosilicate glass vials according to USP <660> with bromobutyl rubber stoppers according to USP <381>. The terminal product is a single-dose injectable solution. The following control matrix applies:

Quality attribute	Test method	Control limit
Assay	HPLC	90.0–110.0% of label claim
Related substances	HPLC	Individual ≤0.2%, total ≤1.0%
pH	USP <791>	4.5–5.5
Osmolality	USP <785>	280–320 mOsm/kg
Particulate matter, subvisible	USP <788>	≥10 µm: NMT 6000 per container; ≥25 µm: NMT 600 per container
Visible particulates	USP <790>	Practically free
Sterility	USP <71>	No growth
Bacterial endotoxins	USP <85>	NMT 0.20 EU/mg
Elemental impurities	ICH Q3D	Parenteral PDE limits

Lyophilisation cycle development for (S)-oxetan-2-ylmethanamine hydrochloride in 10R borosilicate vials requires a mannitol-based crystalline cake to prevent collapse of the low-molecular-weight API. The pre-lyophilization solution contains the API at 20 mg/mL and mannitol at 4% w/v as the bulking agent. The fill volume is 5 mL per vial, and the solution is filtered through a 0.22 µm polyethersulfone membrane before filling. Freezing is performed by ramping the shelf to -40 °C at 0.5–1.0 °C/min and holding for 2–4 h. Primary drying is conducted at a shelf temperature of -20 °C and a chamber pressure of 0.2 mbar for 24–48 h; the product temperature is kept below the collapse temperature of the formulation. Secondary drying is run at 25–35 °C for 6–12 h. Final moisture is controlled to not more than 1.0% w/w according to USP <921>. Container closure integrity is verified by USP <1207> after stoppering.

Reconstitution time is controlled to not more than 2 min with 5 mL of water for injection. If reconstitution exceeds this limit, the mannitol-to-API ratio is adjusted or the fill depth is reduced to increase cake surface area. The lyophilized cake is a sterile powder for injection with no added preservative. Because the API is a primary amine, excipient nitrite levels must be controlled before lyophilization to limit potential nitrosamine formation under acidic reconstitution; residual nitrite in mannitol and water for injection is assessed under ICH M7(R2) guidance. The terminal product is a single-dose lyophilized powder for injection.

When the API Is Co-processed as Single-Dose Oral Granules in Sachet Blisters

Single-dose oral granules containing (S)-oxetan-2-ylmethanamine are produced by wet granulation using a sugar-free matrix to avoid lactose-amine incompatibility. The matrix contains milled mannitol at 60–80% w/w, xylitol at 10–20% w/w, and low-substituted hydroxypropyl cellulose at 2–5% w/w as the disintegrant. The API is loaded at 0.5–5.0% w/w. The binder solution is prepared with hypromellose or povidone K30 at 3% w/w in purified water. Granulation is performed in a high-shear granulator with an impeller tip speed of 3–6 m/s, and the wet mass is dried to a loss-on-drying value of 1.0–2.0% w/w. The dried granules are screened through a 1000 µm sieve and filled into paper/aluminum/polyethylene laminate sachets at 500 mg per sachet. Seal integrity is tested according to USP <671>; seal strength is set at not less than 6 N/15 mm to prevent moisture ingress in stability zones III and IV.

Taste masking may be required because the primary amine can produce a bitter or ammoniacal taste. If unacceptable, the granules are coated with ethylcellulose or Eudragit RL/RS in a fluid-bed coater at a weight gain of 5–10% w/w; published data on the taste intensity of this specific API is limited, so sensory evaluation is conducted on the granule batch before scale-up. The terminal product is an oral granule in single-dose sachet form, intended for dispersion in water or direct administration. Residual solvent is controlled according to ICH Q3C; ethanol from coating is limited to 5000 ppm as a class 3 solvent.

Sterile Dry Powder Filling, Reconstitution Time, and Particulate Control

Aseptic dry powder filling of (S)-oxetan-2-ylmethanamine in a mannitol-based carrier is performed when the API is not sufficiently stable in aqueous solution for ready-to-use injection. The sterile powder is prepared by spray drying or by lyophilization followed by milling, with an API-to-mannitol ratio of 1:4 or 1:9. The powder is filled into 20 mL Type I borosilicate vials using an automatic auger or vacuum drum filler inside an EU GMP Annex 1 Grade A isolator. Fill weight is controlled to 100–500 mg with a relative standard deviation of not more than 2.0%. The powder blend is passed through a 500 µm screen before filling to remove agglomerates and to ensure rapid reconstitution. Particle size is measured by laser diffraction; the D90 is controlled at not more than 100 µm for reconstitution time below 3 min with 10 mL of water for injection.

Following reconstitution, the solution concentration is 20 mg/mL. The reconstituted solution is controlled for subvisible particulates according to USP <788> and for visible particulates according to USP <790>. The sterility of the powder is confirmed according to USP <71>, and bacterial endotoxins are tested according to USP <85> with a limit of not more than 0.20 EU/mg. Because the API is a hygroscopic primary amine, vials are stoppered under dry nitrogen with bromobutyl closures and sealed with aluminum flip-off caps. Container closure integrity is verified by USP <1207>. The terminal product is a sterile powder for injection requiring reconstitution before administration.