3D-printing oral dosage forms by depositing shear-thinning gels layer-by-layer at near-room temperature — where infill geometry, not excipient chemistry, becomes the release-rate control variable
Direct compression and hot-melt extrusion (HME) remain the workhorses of solid oral dosage form manufacturing, but both bind release kinetics tightly to formulation chemistry. To shift a tablet from immediate-release to 12-hour modified-release, a formulator must re-select polymers, adjust coating thickness, or change the compression force — each change requiring fresh stability studies. Worse, HME and fused-deposition-modeling (FDM) 3D printing both require melting the polymer matrix at 160–220°C, a temperature window that degrades a large fraction of thermolabile actives (peptides, some antibiotics, ascorbic acid derivatives, several biologics-adjacent small molecules).
Conventional modified-release strategies and their structural limitations:
Direct compression + functional coating: • Release rate set by coat composition (Eudragit RS/RL ratio), coat thickness (typically 20–120 µm), and curing conditions • Changing target release profile requires re-optimizing spray-coating process — new pan-coating runs, new dissolution validation, weeks of development • Coating uniformity across a batch varies ±5–15% tablet-to-tablet, propagating into release-profile variability
Hot-melt extrusion (HME) monolithic matrices: • Drug dispersed molecularly or as solid dispersion in thermoplastic polymer (HPMCAS, Kollidon VA64, Eudragit E) • Barrel temperatures 130–200°C across multiple heating zones; residence time 1–5 min at elevated shear • Release profile fixed by polymer:drug ratio and matrix porosity established during extrusion — cannot be tuned post-manufacture • Thermolabile actives: extensive literature reports degradation onset for peptide-like or oxidation-prone APIs above 100–130°C, well below typical HME processing windows
FDM 3D printing (the "hot" alternative 3D-printing route): • Requires a two-step process: HME first produces a drug-loaded filament (1.75 mm or 2.85 mm diameter), which is then re-melted at the FDM printhead (180–220°C) and deposited • Double thermal exposure compounds degradation risk • Filament brittleness, poor drug loading (typically <30% w/w before filament becomes unprintable), and nozzle clogging are persistent engineering problems
Why geometry-based control matters: • If release rate could be set by the internal architecture of the dosage form — its porosity, surface-to-volume ratio, tortuosity — rather than by remelting or recoating, formulators could decouple "what drug, what polymer" from "how fast does it release" • This is precisely the proposition of semi-solid extrusion (SSE) printing: a cold, low-shear deposition process where the printed lattice itself becomes the release-rate control variable • Additional driver: WHO and FDA interest in point-of-care/pharmacy-based manufacturing of patient-specific doses (fractional or non-standard strengths) — batch compression tooling cannot economically produce single-unit customized doses; digital SSE printing can.
Semi-solid extrusion (SSE, also called pressure-assisted microsyringe or paste extrusion) prints a pre-formed gel or paste ink — not a thermoplastic filament — through a fine nozzle using pneumatic or mechanical (lead-screw) pressure. No heating element is required at the printhead: the ink is formulated to be printable at 20–40°C from the outset, which is what makes SSE compatible with thermolabile drugs, proteins, and moisture- or heat-sensitive excipients that FDM and HME cannot process.
What makes a paste "printable" — the rheological design window:
Shear-thinning (pseudoplastic) behavior: • Apparent viscosity must drop sharply under the shear applied during extrusion (typically 10–500 s⁻¹ in the nozzle) so the ink flows through a sub-millimeter orifice under achievable pressure • Power-law index n <1 (often 0.15–0.45 for HPMC/Carbopol gels) — viscosity falls by 1–2 orders of magnitude from rest to print shear rate • After deposition, shear rate drops to ~0 and viscosity must recover fast enough (thixotropic recovery, seconds timescale) to prevent the bead from slumping before the next layer is deposited
Yield stress requirement: • Below the yield stress (rest state), the gel must behave as a soft solid — this is what lets a printed filament hold a bridge, an overhang, or a defined cylindrical shape instead of flowing flat • Target range: 50–300 Pa, measured by oscillatory amplitude sweep (crossover of G′/G″) or vane rheometry • Too low (<20 Pa): printed layers slump, shape fidelity lost, layer height cannot be maintained • Too high (>500 Pa): requires excessive extrusion pressure, risk of nozzle clogging and irregular bead deposition
Common SSE ink systems and typical composition: • HPMC E15/K100M gels: 10–20% w/w polymer in water/ethanol cosolvent, drug dissolved or suspended, yield stress ~80–150 Pa • Gelatin-based hydrogels: 15–25% w/w gelatin, gelled at 4–10°C before printing, printed near gel point for controlled setting • Pluronic F127: 20–30% w/w in water exploits reverse thermal gelation — liquid when cold, gels near body temperature, useful for in-situ forming or low-temperature printing • Methocel (HPMC) + glycerin plasticizer: improves layer-to-layer fusion, reduces brittleness after drying • Carbopol 974P: 0.5–2% w/w cross-linked polyacrylic acid, neutralized to pH 6–7, very high yield stress at low solids content — used for high shape-fidelity small-volume printlets
Extrusion mechanics: • Pneumatic syringe pump: regulated air pressure (50–400 kPa) pushes plunger; simple, low-cost, widely used in FabRx/Nottingham academic printers • Mechanical lead-screw extruder: stepper-motor-driven plunger gives finer volumetric control, better for viscous high-yield-stress inks • Print head moves in X-Y per layer (G-code from sliced STL), Z-stage increments by layer height (typically 0.2–0.8 mm) between layers • Post-print: printlets typically air-dried or oven-dried at 40–50°C for 12–24 h to remove residual water/solvent and achieve final mechanical strength
Every SSE print is defined by a small set of process parameters that jointly determine bead geometry, layer bonding, and — critically for drug release — internal porosity. Because these are digital slicer settings rather than formulation changes, the same drug-loaded ink can be printed into dosage forms spanning immediate-release to extended-release simply by changing the G-code.
Nozzle diameter: • Directly sets minimum extruded bead (filament) width — typically bead width ≈ 1.0–1.3 × nozzle diameter due to slight die swell • Common gauges: 25G (0.41 mm), 23G (0.61 mm), 21G (0.84 mm), 18G (1.20 mm), 15G (1.55 mm) • Finer nozzles (0.4–0.6 mm) give higher print resolution and smoother surfaces but require higher pressure for a given ink viscosity and are more prone to clogging with suspended drug particles >20–50 µm • Coarser nozzles (1.0–1.5 mm) print faster with lower pressure but coarser surface texture and reduced feature resolution — usually paired with lower infill densities
Extrusion pressure and print speed: • Pressure must overcome the ink yield stress plus nozzle wall shear resistance (Hagen–Poiseuille-type scaling modified for power-law fluids): P ∝ (Q·L/r³)^n approximately, where Q is flow rate, L nozzle length, r radius • Typical working pressures: 50–150 kPa for low-viscosity HPMC gels through 0.84 mm nozzles; up to 300–400 kPa for high-yield-stress Carbopol pastes through 0.41 mm nozzles • Print speed (X-Y travel, 5–20 mm/s) must be matched to volumetric flow rate to maintain constant bead cross-section — mismatch causes over-extrusion (bulging, closed pores) or under-extrusion (gaps, poor layer bonding)
Infill density and pattern — the direct porosity control: • Infill density (%) is the fraction of the interior cross-section filled with material versus void, set in slicer software (e.g., Repetier-Host, Cura-derived slicers adapted for pharma printers) • Grid/rectilinear infill: parallel bead lines, alternating 0°/90° per layer — simple, predictable, moderate tortuosity • Gyroid infill: triply periodic minimal surface, uniform mechanical properties in all directions, smoothly interconnected pore network — favored when isotropic dissolution behavior is desired • Honeycomb: hexagonal cells, high mechanical strength per unit material, but more tortuous diffusion path • At fixed shell thickness (typically 2–4 perimeter walls, ~0.8–1.6 mm total shell), infill density from 20% to 80% changes total void volume, internal surface area exposed once the shell is breached, and the tortuosity factor the dissolution medium must traverse to reach embedded drug
Why this matters for release: infill density and pattern change the effective surface area available to dissolution medium and the diffusion path length through the porous matrix — the two dominant terms in Higuchi-type and Peppas–Korsmeyer release models — without altering a single excipient.
The defining experiment of SSE printlet research holds the drug, polymer, and total printlet mass constant while varying only the infill density (and sometimes pattern), then runs USP Apparatus II dissolution testing (paddle, 900 mL medium, 37°C, 50 rpm) to show that release half-life can shift several-fold from geometry alone — matching or exceeding the tunability previously achievable only by reformulating an HME matrix.
Standard in-vitro release testing setup:
• USP Apparatus II (paddle), 900 mL dissolution medium (typically 0.1 N HCl for gastric-stage or pH 6.8 phosphate buffer for intestinal-stage, or biorelevant FaSSGF/FaSSIF media for physiologically realistic profiles) • 37 ± 0.5°C, paddle rotation 50 rpm (sometimes 75 rpm for harder-to-dissolve infill patterns) • Sampling at 5, 10, 15, 30, 45, 60, 90, 120 min (and extended to 6–8 h for high-infill/extended-release printlets), UV-Vis or HPLC quantification of drug concentration
Representative comparative results (same formulation, varying infill only): • 20% infill: high open porosity, low tortuosity, large initial wetted surface area → >90% released within 45–60 min, profile approximates zero-order (constant rate) over the first 30–40 min because the dissolving front advances through a nearly unobstructed lattice • 50% infill: intermediate — t50 (time to 50% release) around 35–45 min, transitional kinetics between zero-order and Higuchi diffusion-controlled release • 80% infill: low porosity, high tortuosity, matrix behaves closer to a monolithic HME tablet — t50 extends to 60–70 min and full release requires >6 h; profile fits the Higuchi square-root-of-time model (Q ∝ √t) or Korsmeyer–Peppas with diffusional exponent n indicating anomalous (non-Fickian) transport as the gel matrix itself also erodes
Model fitting used in the literature: • Zero-order model: Q = k₀·t — best fit for low-infill, high-porosity printlets where release rate is roughly constant (surface-area-limited, not diffusion-limited) • First-order model: ln(1−Q) = −k₁·t — often better for denser, HME-comparator tablets where the concentration gradient across an intact matrix drives release • Higuchi model: Q = kH·√t — frequently the best-fitting model for intermediate-to-high infill SSE printlets, consistent with diffusion through a porous matrix • Korsmeyer–Peppas: Mt/M∞ = k·tⁿ — used to diagnose transport mechanism; n≈0.45 indicates Fickian diffusion, 0.45<n<0.89 indicates anomalous (coupled diffusion + matrix relaxation/erosion) transport, common in swellable HPMC/gelatin printlet matrices
Comparison to conventional HME tablets of identical formulation: • HME monolithic matrix tablets of the same drug:polymer ratio typically show a single fixed release profile (t50 ~50–60 min) that can only be shifted by reformulating polymer grade or drug loading • SSE printlets of the same ink, printed at different infill densities, reproduce a comparable RANGE of t50 values (19–70 min demonstrated above) — meaning one qualified ink formulation can serve immediate-release through moderate extended-release indications via slicer settings alone, with no new excipient qualification required.
Academic groups — most prominently FabRx (spun out of UCL School of Pharmacy) and the University of Nottingham 3D printing group — have used SSE to print paracetamol and captopril printlets with patient-tunable dose and infill-tuned release, benchmarked directly against hot-melt-extruded comparator tablets of the same formulation. These studies establish infill geometry as a legitimate, regulatorily discussable design variable within a Quality Target Product Profile (QTPP), not merely a laboratory curiosity.
Paracetamol SSE printlets (FabRx / UCL School of Pharmacy): • Printed from an HPMC- or gelatin-based hydrogel ink loaded with paracetamol, using a pressure-assisted micro-extrusion (PAM) printer • Printlet dose set by adjusting print volume/geometry (height, diameter) at fixed ink drug concentration — enabling non-standard, patient-specific doses (e.g., pediatric fractional dosing) without new tooling or a new formulation batch • Infill density varied to demonstrate both fast-release (near-immediate, geometry approximating 20–30% infill) and slower-release (60–80% infill) printlets from one qualified ink
Captopril SSE printlets (University of Nottingham): • Captopril, an ACE inhibitor with a relatively short half-life, is a candidate for modified-release dosing to smooth plasma concentration fluctuations • HPMC-based printable gel carried the captopril payload; SSE printlets compared against directly-compressed and HME-processed captopril tablets of matched drug content • Study demonstrated that varying internal infill pattern (grid vs. gyroid) and density shifted the dissolution profile across a therapeutically meaningful range while avoiding any thermal exposure of the API — relevant because captopril contains a free thiol group susceptible to oxidative/thermal degradation, making the near-ambient SSE process a meaningful stability advantage over HME/FDM routes
Why these case studies matter beyond the bench: • Both studies used real pharmacopeial dissolution testing (USP II) and standard analytical quantification (HPLC/UV), not just qualitative print demonstrations — giving regulator-legible in-vitro release data • They directly benchmark against a conventional HME comparator of the SAME formulation, isolating geometry (not chemistry) as the variable responsible for the release-profile shift • They support the core translational claim of SSE printlet technology: a single qualified printable ink can serve a family of doses and release profiles via digital slicing parameters, aligning with point-of-care and pharmacy-based personalized manufacturing models that batch compression cannot economically support
Remaining barriers to clinical/industrial adoption: • Batch-to-batch and printer-to-printer reproducibility of bead geometry at pharmaceutical GMP tolerances • Long-term physical/chemical stability of printed hydrogel matrices during drying and storage (moisture migration, polymorphic transitions) • Regulatory pathway for a "digital" manufacturing process where the CQA-determining variable is G-code/infill rather than a fixed formulation — actively being discussed with FDA and EMA under emerging frameworks for point-of-care and continuous manufacturing.
The central translational insight from FabRx and Nottingham work is that infill geometry functions as an independent Critical Process Parameter for release rate — orthogonal to drug loading, polymer grade, and coating. A single validated SSE ink can be sliced into immediate-release (20% infill, t50 ≈19 min), intermediate (50% infill, t50 ≈38 min), or extended-release (80% infill, t50 ≈70 min) printlets of the identical patient-specific dose, without requalifying a single excipient — a manufacturing flexibility no compression- or HME-based route can match at the point of care.