Direct compression of celecoxib at production scale is treated as a high-risk unit operation because the API is assigned to BCS Class II, with reported aqueous solubility below 10 µg/mL at 25°C and high permeability. The downstream consequence is that dissolution rather than gastric emptying or permeability controls the rate-limiting step for oral absorption from tablets and capsules. Direct compression is therefore limited to lower dose strengths where the final tablet mass can be kept within a manufacturable range; at strengths of 200 mg and above, the proportion of API in the tablet core rises to a point where segregation risk and poor flow often force a change to granulation. The process begins with a dry blend prepared in a 1000 L tumble blender fitted with an intensifier bar. Celecoxib is pre-screened through a 0.5 mm sieve and blended with lactose monohydrate, microcrystalline cellulose, croscarmellose sodium, colloidal silicon dioxide, and magnesium stearate. Lactose monohydrate is included as a brittle filler that reduces the effect of API particle size change; croscarmellose sodium functions as a superdisintegrant; colloidal silicon dioxide is added at 0.2–0.5% w/w to improve flow without retarding dissolution. Final lubrication with magnesium stearate is held at 0.5–1.0% w/w and 3–5 min of mixing. Prolonged lubrication coats the hydrophobic API surfaces, reduces wetting, and produces a measurable delay in dissolution under USP <711> conditions. The blend is compressed on a rotary tablet press equipped with a force feeder and 10 mm round concave tooling; main compression force is typically held at 8–18 kN with a precompression force of 2–4 kN. Tablet hardness is monitored by diametral compression at 80–120 N, and friability is controlled at ≤1.0% w/w loss per USP <1216>. Content uniformity is assessed at beginning, middle, and end of the compression run with an acceptance value of ≤15.0 per USP <905>. The direct compression route should not be attempted without a defined particle size specification for the API; laser diffraction per USP <429> is used to track the cumulative distribution, and a D90 below 30 µm is a common development target for BCS Class II weak acids, although the exact upper limit is product-specific and published data for this specific configuration is limited. If ambient relative humidity exceeds 60%, the low-density API tends to agglomerate, so material conditioning and humidity-controlled dispensing are required.

Does High-Shear Wet Granulation Improve Content Uniformity Without Suppressing Dissolution?

In manufacturing lines where direct compression cannot maintain content uniformity at the required dose, high-shear wet granulation is selected, particularly when tablet mass exceeds approximately 400 mg. The operation is carried out in a top-drive or bottom-drive granulator with bowl capacities from 25 L to 600 L, a main impeller speed of 100–200 rpm, and a side chopper speed of 1500–3000 rpm. Dry mixing of celecoxib with lactose monohydrate, croscarmellose sodium, and povidone is followed by addition of a binder solution containing povidone K30 at 5–10% w/w in purified water. The binder is added at a controlled rate of 0.5–1.0 L/min per 100 kg of dry mix while impeller power consumption or torque is recorded. The wet massing end-point is defined by the inflection in torque or power rather than by fixed time, because batch-to-batch variability in API particle size and excipient moisture shifts the liquid requirement. Overgranulation beyond the end-point produces dense granules that resist disintegration and delay release; undergranulation leaves fine particles that segregate during transfer and cause content uniformity failure. The wet granules are dried in a fluid-bed dryer with inlet air temperature controlled at 55–65°C and product temperature not exceeding 40°C. Drying continues until loss on drying is ≤2.0% w/w by USP <731>. Higher inlet air temperatures are not automatically acceptable, because polymorphic form changes and chemical degradation must be excluded; monitoring by X-ray powder diffraction per USP <941> is required after drying if the crystalline form is part of the drug product specification. The dried granules are milled through a 1.0 mm screen and blended with extragranular croscarmellose sodium and magnesium stearate. Sieve analysis per USP <786> is used to control the granule size distribution; fines below 75 µm are generally limited to <20% w/w to reduce capping and sticking on the tablet press. Dissolution testing is performed by USP <711> in a surfactant-containing aqueous medium, because plain water cannot provide sink conditions for a BCS Class II weak acid with pKa near 11.1. The current compendial monograph for the specific marketed tablet or capsule must be consulted for the exact dissolution medium composition, apparatus speed, and Q acceptance value. Residual water above 2.0% w/w before lubrication can promote localized adhesive wear on tooling; magnesium stearate above 1.0% w/w should be avoided for the same reason as in direct compression.

Capsule filling with celecoxib is moved to roller-compacted dry granulation when direct powder filling shows low bulk density, high Hausner ratio, or fill weight variability outside the target. The preblend is compacted on a roller compactor with 200 mm diameter rolls and cheek-plate side seals; hydraulic pressure is maintained at 30–80 bar, roll speed at 3–8 rpm, and gap at 2–4 mm. Ribbon solid fraction is measured by envelope density and controlled between 0.55 and 0.75. Ribbons below 0.55 generate excessive fines during milling; ribbons above 0.75 can produce hard granules that remain intact in the capsule and delay drug release. Milling is performed through a conical mill fitted with a 1.0 mm rasp screen at 1200–1800 rpm. Oversize is recycled, but the recycle stream is limited to ≤30% w/w of the total granulate because repeated compaction changes particle hardness and dissolution behavior. The granulate is filled into hard gelatin or hypromellose capsules on a dosator or tamping-pin capsule filler. Fill weight is verified gravimetrically at start-up and at fixed intervals; high-speed lines operating above 100,000 capsules/hour often require a 100% checkweigher to reject underfilled or overfilled units. Capsule shell moisture is held at 15–20% w/w, and the filling suite is maintained at 40–45% RH to prevent shell deformation during storage. Content uniformity of the filled capsules is assessed by USP <905>; dissolution is tested by USP <711> in a surfactant-containing medium. The water content of the granulate is controlled by USP <731> or USP <921>, depending on whether the specification is expressed as loss on drying or Karl Fischer water. Bulk density and tapped density are measured by USP <616>; the Hausner ratio derived from these measurements is used to set fill weight and to predict flow through the dosator or tamping pin. Published data for specific roller compaction parameters for celecoxib is limited; the above ranges are development starting points that must be confirmed against ribbon solid fraction and dissolution response for each formulation. An operational boundary is the ambient relative humidity: at >60% RH, the powder may pick up moisture and the resulting granulate may stick to the mill screen and capsule filler contact parts.

The route comparison in the table summarizes critical process windows observed across development and scale-up campaigns; the ranges are not compendial release limits but serve as starting points for pharmaceutical development.

Oral Granule Sachet Lines and Dose Titration Process Boundaries

For dose titration or swallowing difficulty, a granule presentation for oral dispersion is selected when intact capsules or tablets are unsuitable. The granulate is produced by the same high-shear or dry granulation routes used for tablets and capsules, but the material is not compressed; instead it is filled into stick-pack or sachet form-fill-seal lines. Bulk density of the finished granulate is held within 0.45–0.65 g/mL, and tapped density is measured by USP <616> to calculate the Hausner ratio. A Hausner ratio above 1.35 indicates poor flow through the volumetric auger and is corrected by reducing fines below 75 µm or by adding colloidal silicon dioxide at 0.2–0.5% w/w. Fill weight is verified at start-up and every 30 min using an analytical balance with 0.1 mg resolution. The contents of each sachet are dispersed in water or a soft food vehicle before administration; because celecoxib has aqueous solubility below 10 µg/mL, the resulting dispersion is generally a suspension rather than a solution unless a co-solvent or surfactant is included in the granulate. Dissolution testing uses USP <711> with a surfactant-containing medium; content uniformity is assessed by USP <905> using the entire sachet contents as one dosage unit. Moisture control by USP <731> is critical; loss on drying above 2.0% w/w increases granule agglomeration and causes fill weight drift on long runs. Packaging must provide a moisture barrier; if aluminum laminate is used, a moisture vapor transmission rate below 0.1 g/m²/day at 38°C/90% RH is a typical target for hygroscopic granulate, though the specific value must be derived from stability data generated under ICH Q1A(R2). The granule sachet route is not interchangeable with a simple powder blend; the particle size distribution must be designed so that the dose delivered after dispersion matches the dissolution profile of the reference solid oral product. Data from a single development batch are not sufficient to set the recirculation limits on the form-fill-seal line because residual granulate in the hopper can segregate and change the fill weight over extended runs.

When Liquid Oral Delivery Is Required, Cosolvent Ratio Rather Than pH Governs Solubilized Fraction

When a liquid oral vehicle is specified, the aqueous solubility of celecoxib below 10 µg/mL prevents a simple water-based solution at therapeutic concentration. pH adjustment does not provide sufficient solubilization within the acceptable oral pH range because the API is a weak acid with pKa near 11.1; only at pH values far above the physiological oral range would ionization become significant. The practical formulation uses a co-solvent system containing ethanol, propylene glycol, PEG 400, or glycerin, optionally with a non-ionic surfactant. The ratio of co-solvent to water is selected by phase-solubility studies and is controlled during bulk compounding because small changes in the co-solvent fraction can shift the solubilized fraction and alter dose uniformity. The bulk solution is compounded in a 316L stainless steel vessel under nitrogen; mixing is performed with a top-mounted agitator, and the addition sequence is fixed so that celecoxib is first dissolved in the co-solvent-rich phase before aqueous dilution. Ethanol, if present, is assayed by gas chromatography per USP <611> and held within the approved labeling range. Fill volume is checked for multi-dose bottles by USP <698>; microbial enumeration follows USP <61>, and testing for specified organisms follows USP <62>. The filled product is protected from light because liquid-state degradation may proceed faster than in solid form; photostability data generated under ICH Q1B determine whether amber glass or an opaque plastic container is required. For an oral dispersion rather than a solution, viscosity is measured at 25°C with a rotational viscometer; the target range is validated against dose delivery because settlement of celecoxib particles can lead to underdosing if the container is not shaken. Commercial oral solution formulations have demonstrated feasibility, but their exact inactive ingredient concentrations are regulatory-confidential; an extemporaneous suspension prepared without the approved co-solvent system should not be assumed to match the dissolution characteristics of the commercial solid form. Residual solvent levels in the liquid vehicle must comply with USP <467> and ICH Q3C permitted daily exposure limits.

Injectable Presentation Requires Solubilizer Selection Before Terminal Sterilization Is Fixed

Injectable presentation of celecoxib is evaluated as an acute-care formulation-development target, not as a standard commercial parenteral monograph product. The API's aqueous solubility below 10 µg/mL and pKa near 11.1 exclude simple isotonic saline or dextrose vehicles; pH adjustment to above 10 is not acceptable for parenteral tissue tolerance. Solubilization candidates include co-solvents such as ethanol, propylene glycol, and PEG 400; non-ionic surfactants such as polysorbate 80; and cyclodextrin derivatives such as sulfobutyl ether β-cyclodextrin. Published data for specific celecoxib parenteral formulations is limited, so each candidate must be screened for precipitation upon dilution with simulated IV fluids, for hemolytic potential, and for filter compatibility. The drug solution is filtered through a 0.22 µm PVDF or PES membrane under aseptic conditions. If micellar or cyclodextrin complexes are used, filter validation must demonstrate that the membrane does not selectively retain the active or alter particle size distribution; filter integrity is tested by bubble point or diffusion according to the filter manufacturer's validated limits. Aseptic processing is performed in an ISO 14644-1 Class 5 environment, equivalent to Grade A, under FDA 21 CFR 211.42. Particulate matter is controlled by USP <788>; small-volume parenteral limits are ≤6000 particles per container ≥10 µm and ≤600 particles per container ≥25 µm. Bacterial endotoxin limit is calculated from the maximum adult dose per USP <85> using K=5 EU/kg; the resulting limit is expressed in EU/mg of celecoxib. Sterility testing follows USP <71>; container closure integrity testing follows USP <1207>. Terminal moist-heat sterilization at 121°C for 15 min is not automatically acceptable; thermal degradation products and co-solvent volatility must be controlled within ICH Q3B thresholds. If terminal sterilization is not feasible, aseptic filtration and/or lyophilization are used. A lyophilized cake, if developed, is controlled for residual moisture by USP <921> Method Ic at ≤2.0% w/w and reconstitution time at 25°C. The absence of a pharmacopeial monograph for celecoxib injection means that release specifications are established under ICH Q6A and the relevant general chapters rather than a compendial drug product monograph. The formulation must also demonstrate physical stability during short-term clinical use; precipitation on standing or during infusion with 0.9% sodium chloride injection is a known failure mode for poorly water-soluble injectable weak acids and must be evaluated with actual infusion sets, not glass beakers alone.