Empagliflozin Pharma Grade API received as crystalline Form I with a differential scanning calorimetry endotherm onset at approximately 152°C and a particle size distribution D90 in the range of 10–40 μm depending on the milling campaign is characterized by an acicular crystal habit that confers poor flow properties and a Hausner ratio typically exceeding 1.35 when measured per the methods described in USP <1174>. The bulk material cannot be charged directly to a high-speed rotary tablet press without an intermediate granulation step; roller compaction is therefore the primary robust manufacturing pathway for oral solid dosage forms. The API is pre-blended with lactose monohydrate (PhEur 0184), microcrystalline cellulose (PhEur 0316, grade Avicel PH-102 or equivalent), crospovidone (PhEur 1147, Type A), and colloidal anhydrous silica (PhEur 0434) in a bin blender or V-blender. Representative quantitative ranges for a 400 mg core tablet are: empagliflozin 10.0–25.0 mg (equivalent to 2.5–6.25 % w/w), crospovidone 2–5 % w/w, colloidal anhydrous silica 0.5–1.5 % w/w, microcrystalline cellulose 20–35 % w/w, and lactose monohydrate quantum satis to final core mass. Roller compaction is performed on a Gerteis Mini-Pactor or Alexanderwerk WP 120 at roll pressure 20–60 kN, roll speed 1–10 rpm, and gap width 1–3 mm, with ribbon density targeted at 0.9–1.3 g/cm³. The ribbon is subsequently milled through a screen aperture of 0.8–1.25 mm and the resulting granulate is classified by sieving. The thermal sensitivity of the active ingredient dictates that processing temperatures must not exceed 130°C at any downstream unit operation; thermogravimetric analysis under nitrogen shows mass loss onset near 160°C. Magnesium stearate (PhEur 0297) is added extragranularly at 0.75–1.5 % w/w and blended for a controlled 3–5 minutes at 15–25 rpm; over-lubrication exceeding 8 minutes compacts the magnesium stearate lamellae on the granulate surface and produces a measurable reduction in tablet hardness of 15–25 N at equivalent main compression force, a failure mode documented on production-scale rotary presses. Core tablets of 6 mm round concave tooling are compressed on a Korsch XL 400 or Fette 1200i with precompression force 3–6 kN and main compression force 8–18 kN, yielding hardness in the range 80–160 N measured with a Schleuniger 8M hardness tester. Disintegration time must comply with USP <701>, not more than 15 minutes in water at 37 ± 2°C. Content uniformity testing per USP <905> on the finished core typically achieves %RSD below 3.5 for 10 mg cores and below 2.8 for 25 mg cores across 30 units, provided the API D90 is held below 40 μm and the blending batch size exceeds 50 kg. At pilot scale below 20 kg, stratification of the active has been observed when the blender loading sequence places magnesium stearate onto the discharge port before the main blend; the corrective loading order introduces the stearate through a 500 μm hand screen only in the final 3-minute blend.

When Direct Compression Replaces Roller Compaction for Empagliflozin 10 mg Tablet Trains

Direct compression is the preferred economic route only when the API supplier provides a densified grade with D10 above 5 μm, D50 between 15 and 30 μm, and D90 below 50 μm; without this particle-size envelope, the 10 mg dose exhibits unacceptable content uniformity drift on continuous production runs. A forced feeding system on a Sejong or Killian rotary press induces a feed-frame dwell time that segregates coarse granules from fines, an effect amplified in blends containing more than 25 % w/w microcrystalline cellulose due to electrostatic charging behavior below 30 % RH. The blend bulk density must remain within 0.45–0.60 g/mL, tapped density 0.55–0.75 g/mL (USP <616> Method I), Carr index not more than 20, and static angle of repose below 35°. In process development on a 49-station rotary press operating at 60 rpm, the %RSD of 10 mg cores approached 6.0 after 4 hours of continuous operation when the API D90 exceeded 50 μm; the mechanism was identified as percolation of fine active particles through void spaces in the powder bed under sustained mechanical vibration. The compression force required to achieve 100 N hardness in a direct-compression formulation is typically 12–20 kN, compared to 8–12 kN for roller-compacted granulate, because the unprocessed blend exhibits higher elastic recovery. Tablets discharged from a direct compression train must pass friability testing per USP <1216> with weight loss below 0.8 % after 100 revolutions at 25 rpm. When these limits are not met, the manufacturing route is reverted to roller compaction rather than increasing crospovidone content, because crospovidone loadings above 5 % w/w prolong disintegration time beyond 20 minutes and shift dissolution profiles in pH 1.2 medium outside the acceptance window.

The capsule filling unit operation for Empagliflozin is not initiated with a standalone API charge; the active is first pre-blended with microcrystalline cellulose and lactose monohydrate at a 1:1 mass ratio to create a pre-mix, which is subsequently milled through a 0.5 mm screen to break up agglomerates. The pre-mix is quantified by high-performance liquid chromatography before being charged to the main blender. Hard gelatin capsules (size #3, 13.3 mm locked length) or hydroxypropyl methylcellulose capsules of the same size are filled on a Bosch GKF 1500 or MG2 Planeta 100 dosing-disc/tamping pin capsule filler. Tamping pin compression settings are adjusted to produce a plug density of 0.60–0.80 g/cm³, tamp depth of 12–16 mm, and tamp force not exceeding 150 N to avoid capsule shell deformation. The fill weight for a 25 mg capsule is generally 350–450 mg, yielding a drug load of 5.6–7.1 % w/w; for the 10 mg dose, the fill weight may be reduced to 200–300 mg to maintain acceptable powder flow. Shell closure integrity is verified in-process by weight sorting at ±3 % of target and visual inspection for split or dented caps. Capsule content uniformity per USP <905> should show %RSD below 4.5 for a 10 mg fill and below 3.5 for a 25 mg fill. Dissolution testing is performed per USP <711> using Apparatus 2 (paddle) at 50 rpm in 900 mL of pH 6.8 phosphate buffer at 37.0 ± 0.5°C, with acceptance criterion Q = 80 % dissolved at 30 minutes. Formulations containing crospovidone as disintegrant and lactose as water-soluble filler achieve greater than 85 % release at 15 minutes; HPMC capsule shells form a gel layer at the dissolution interface and delay release by 5–10 minutes relative to gelatin shells. When the capsule filler is operated above 80,000 capsules per hour, powder residence time in the dosing station drops below 200 milliseconds, and the granulate must exhibit critical-state flow metrics consistent with Jenike shear cell testing (ASTM D6128). In production environments where ambient temperature exceeds 30°C and relative humidity is above 60 %, HPMC shells absorb moisture up to 5 % w/w, reducing brittleness but increasing the risk of powder caking in the hopper; the filling suite is therefore maintained at 20–25°C and 35–50 % RH.

What Limits Powder Segregation During Low-Dose Capsule Filling of Empagliflozin 10 mg Formulations?

The dosing-disc/tamping pin principle creates a centrifugal acceleration that stratifies particles by size and density; the active, present as acicular Form I crystals with a particle density of approximately 1.32 g/cm³, segregates preferentially toward the outer wall when blended with denser lactose monohydrate (true density 1.54 g/cm³). The resulting localized drug concentration in the dosing plug can deviate by more than 15 % from target even when the bulk blend shows adequate uniformity. Mitigation is achieved by matching the aerodynamic diameter of the active to that of the diluent system: the API is specified to maintain D50 between 15 and 30 μm and D90 below 50 μm, while the lactose monohydrate grade is specified as 200 mesh (approximate D50 of 45 μm), thereby minimizing differential acceleration under centrifugal force. Electrostatic charging of the active becomes significant below 30 % RH; the charged particles adhere to metal dosing pins and accumulate on the powder bowl surface, and the observed loss in dose weight is in the range of 5–10 % after 30 minutes of uninterrupted operation. The control strategy includes antistatic grounding of the capsule filler and conditioning the blend at 45–55 % RH for a minimum of 12 hours before filling. The tamping-pin dwell time is a third determinant of content uniformity: at a pin cycle rate above 90 strokes per minute, the powder consolidation time is insufficient to form a stable plug in a size #4 capsule, and dose weight %RSD rises above 6.0. Published data for this specific configuration is limited; however, process development batches on a Bosch GKF 1500 with 16 tamping stations demonstrate that a pin penetration depth of 12–14 mm and a pin diameter of 7.8 mm produce acceptable plug integrity at 70–85 strokes per minute. The angle of the powder bowl fill level must also be controlled: a fill level above 80 % of bowl volume generates back pressure that compacts the powder prematurely and causes dose weight to drift upward over the course of a filling campaign.

Lyophilized Injectable Presentation, Sterile Filtration Limits, and Buffer Capacity for Empagliflozin Intravenous Formulations

No commercial injectable product containing Empagliflozin is currently approved by the FDA, EMA, PMDA, or NMPA; parenteral preparations have been developed only for clinical pharmacology and toxicology studies, and published data for this specific configuration is limited. The fundamental formulation challenge arises from the pH-dependent aqueous solubility of the active: the unionized phenolic form is practically insoluble below pH 5.0 (generally below 0.1 mg/mL), while deprotonation at pH values above 8.0 raises solubility above 1.0 mg/mL. An intravenous study formulation at a target concentration of 1 mg/mL therefore requires pH adjustment to between 7.5 and 8.5 using sodium hydroxide 0.1 M or tromethamine at 50–100 mM. The buffer system must maintain pH stability over 24 hours at 25°C; a 10 mM phosphate buffer at pH 7.8 shows measured pH drift of up to 0.3 units over 8 hours in non-buffered saline. Sterilization by aseptic filtration through a 0.22 μm PVDF membrane is mandatory; terminal autoclaving at 121°C for 15 minutes is contraindicated because the molecule degrades under alkaline conditions at elevated temperature, producing measurable degradants corresponding to the loss of the oxolan-3-yl ether side chain and modification of the glucitol moiety. A lyophilized presentation may be prepared with mannitol 5 % w/v and trehalose 2 % w/v as bulking and cryoprotectant excipients; the lyophilization cycle employs a freezing plate temperature of -45°C, primary drying at -25°C under 50 mTorr for 36 hours, and secondary drying at 25°C for 8 hours. The reconstituted solution must meet the osmolality range specified in USP <785> (isotonic target 280–320 mOsm/kg) and particulate matter limits per USP <788> (not more than 6,000 particles ≥10 μm and not more than 600 particles ≥25 μm per container). Type I borosilicate glass vials (USP <660>) are preferable to soda-lime glass, because leached alkali from soda-lime containers shifts the pH above 9.0 over 12 weeks at 25°C, accelerating degradation. Research batches containing 1 mg/mL Empagliflozin in phosphate-buffered saline at pH 7.8 have been reported as chemically stable for only 48 hours at 2–8°C; beyond this window, assay loss of 3–5 % is observed along with the appearance of the desmethyl impurity at levels exceeding 0.2 % w/w by HPLC. Published long-term stability data for lyophilized presentations beyond 6 months at 2–8°C is not available in the public domain.

Film coating of Empagliflozin core tablets functions as a moisture barrier and taste-masking layer rather than a purely cosmetic finish. The 10 mg and 25 mg film-coated tablets use a hypromellose 2910 (6 mPa·s, PhEur 0048) based coating system applied in a perforated pan coater (Glatt GC 750 or O'Hara LabCoat) with inlet air temperature 60–75°C, outlet air temperature 35–45°C, pan speed 8–15 rpm, and spray rate 20–80 g/min depending on batch size. The coating suspension has a solids content of 10–15 % w/w in purified water, with a final coated tablet weight gain of 2.5–4.0 % w/w. The applied coating must satisfy dissolution testing equivalence: profiles in 0.1 M hydrochloric acid (pH 1.2) using USP Apparatus 2 at 50 rpm show no more than 5 % difference in percent dissolved at 15 minutes between coated and uncoated cores. If the coating suspension contains titanium dioxide (E171, PhEur 0263) or iron oxide yellow (E172, PhEur 0303), the operator must verify that opacifier particles do not form aggregates larger than 50 μm that would bridge the spray nozzle (typical nozzle diameter 0.5–1.0 mm). The coating operation must be interrupted if the bed relative humidity exceeds 60 % RH, because the API core tablets absorb moisture at a rate of 0.03–0.05 % w/w per hour at 25°C/60 % RH, and elevated moisture content above 3.0 % w/w in the core promotes hydrolytic degradation of the ether side chain over shelf life. In-process control includes atomizing air pressure of 1.0–2.0 bar, nozzle-to-bed distance of 150–200 mm, and exhaust air humidity monitoring. Post-coating drying at 40°C for 15–30 minutes reduces loss on drying to below 1.0 % w/w (USP <731>). Aqueous wet granulation of Empagliflozin with high-shear mixers is not recommended as a primary route because the API exhibits moisture sensitivity above 3 % w/w water content and exposure to granulation temperatures above 60°C accelerates hydrolytic degradation of the oxolan-3-yl ether side chain; the hydro-alcoholic granulation approach, if unavoidable for equipment compatibility, must use ethanol or isopropanol at no more than 20 % w/w of granulation fluid and drying temperatures below 50°C in a fluid-bed dryer.

Blend Uniformity Nested Sampling Across Empagliflozin 10 mg and 25 mg Core Batches Requires Process-Development Documentation

Following the withdrawal of the FDA draft guidance on blend uniformity for ANDAs, blend uniformity testing is not a compendial release test; the statutory release requirement remains the Uniformity of Dosage Units test in USP <905> with acceptance value (AV) not exceeding 15. During process development, stratified nested sampling with a sampling thief at 10 pre-determined locations in a V-blender or bin blender is performed for batches from 50 to 800 kg. The documented performance of roller-compacted granulate containing Empagliflozin at 2.5–6.25 % w/w shows blend uniformity mean values from 95 to 105 % with %RSD below 5.0 after 15 minutes of blending at 25 rpm in a 500 L bin. In contrast, a direct-compression blend with the same drug load but with API D90 above 60 μm showed %RSD of 7.2 at the first sampling point due to poor deagglomeration. The parameter that best correlates with finished-product content uniformity in the 10 mg strength is the ratio of API D90 to excipient D50: values above 1.8 produce AV that approaches or exceeds 15 at the 10 mg dose. Process analytical technology using near-infrared spectroscopy has been implemented at some production sites for real-time blend uniformity monitoring; the spectrometer must be calibrated against a validated HPLC method per ICH Q2(R2) with a root mean square error of prediction below 1.5 % w/w for the active. The equipment validation record includes the sampling thief bias study per ASTM E2709, and the analytical HPLC method must be validated for linearity over 50–150 % of nominal concentration per ICH Q2(R1). The granulation, compression, and filling processes described above are not interchangeable without re-validation: changing the lactose monohydrate supplier from a crystalline to a spray-dried grade alters ribbon density by 8–12 % under identical roller compaction settings, and this shift propagates into a hardness deviation of up to 20 N in the finished core unless the compaction force is re-optimized.