Phase-separated polymer blends for controlled drug delivery by tuning morphology

The study addresses how to control drug release from oral amorphous solid dispersions (ASDs) by tailoring the morphology of phase-separated polymer blends. Many new drug candidates have poor aqueous solubility and bioavailability, and ASDs are used to stabilize the amorphous drug and improve dissolution. Dual-polymer ASDs offer the possibility to combine a hydrophobic polymer for physical stability with a hydrophilic polymer to accelerate release and enable extended, uniform therapeutic effects. Hot-melt extrusion (HME) is a solvent-free, scalable process attractive for ASD manufacture and as a pre-step for additive manufacturing and personalized medicine. Phase separation has been observed in dual-polymer ASDs, but its deliberate use to tailor release has not been explored. This work tests the hypothesis that phase-separated morphologies formed from hydrophobic polylactic acid (PLA) and hydrophilic hydroxypropyl methylcellulose (HPMC) can be tuned via composition and drug loading so that the hydrophilic phase acts as a channelling network controlling the drug release rate.

Prior literature highlights the importance of ASDs for enhancing dissolution and stabilizing amorphous drugs and discusses strategies for controlled release using polymer matrices. Dual-polymer matrices can combine complementary properties (e.g., hydrophobic stability and hydrophilic release), while thermal processing via hot-melt extrusion avoids solvent-related drawbacks and supports continuous manufacturing and 3D printing. Although phase separation in dual-polymer ASDs has been reported post-extrusion, its purposeful exploitation to achieve tailored oral drug release profiles had not been systematically examined. This study builds on these insights by linking phase-separated morphology to dissolution behavior using advanced X-ray imaging.

Formulation and extrusion: Nicotinamide (drug), HPMC (AFFINISOL HME 4M), and PLA were weighed and physically mixed (6 g batches) at 0 or 10 wt% nicotinamide with PLA/HPMC weight ratios of 30/70, 50/50, and 70/30. Mixtures were processed in a 5 mL Xplore micro-compounder (two conical screws, circular die Ø 1.5 mm) at 180 °C, 50 rpm, for 6 min to produce filaments. The chosen HPMC grade has lower Tg and melt viscosity to facilitate HME. Thermal analysis (DSC): ~5 mg samples in hermetically sealed Al pans were analyzed on a TA Q1000 with He/N2 purge gases (50 mL/min). Protocol: cool to -50 °C, 1 min isotherm, heat to 180 °C, two cycles, heating/cooling rates 10 °C/min. Empty pan reference. X-ray diffraction (XRD): Performed on extruded rods (transmission) using a SAXSLAB Nordic instrument with Rigaku 003 micro-focus Cu Kα (λ=1.5406 Å), Pilatus 300K detector, 300 s exposure, sample-to-detector 134 mm (LaB6 calibrated). Dissolution studies: USP paddle apparatus, 900 mL phosphate buffer pH 6.8 at 37 °C, 25 rpm. Duplicate baths per sample; each bath contained ~100 mg filament (cylindrical, ~1.5 mm diameter, 5 cm length). Sampling every 15 min for 3 h, then every 30 min; filaments remained in buffer for 5 days before SEM. UV–Vis (Cary 60) at 262 nm quantified nicotinamide via calibration curve. Scanning electron microscopy (SEM): Post-dissolution residues were quenched in liquid nitrogen and freeze-dried for 24 h to minimize drying artifacts, fractured, sputter-coated with ~4 nm Au, and imaged at 5 kV (JEOL JSM-7800F Prime). Scanning transmission X-ray microscopy (STXM) and NEXAFS: Ultramicrotomed 150 nm sections were placed on 100 nm Si3N4 membranes. Measurements at the PolLux beamline (SLS) with 25 nm zone plate. Energy stacks 280–330 eV (C K-edge), step size 0.1 eV near edge, 1 eV post-edge. Images at selected resonance energies; beam/step size 30–100 nm depending on field of view. Spectral processing with MANTIS: conversion to optical density, normalization to post-edge, compositional maps by singular value decomposition using reference spectra (PLA, HPMC, nicotinamide). Ptychographic X-ray computed nanotomography (PXCT): ~40 µm pillars prepared from filaments using a cryogenic lathe (Preppy) and mounted on OMNY pins. Measurements at cSAXS (SLS) under cryogenic conditions (OMNY setup) at 6.2 keV with a coherently illuminated Fresnel zone plate (optimized illumination). Beam size ~9 µm; scan step 1.5 µm. Diffraction recorded with an in-vacuum Eiger 1.5M at 7.2 m; exposure 0.05 s per position. 1600 projections from 0–180°, acquired as two tomograms with doubled angular steps; flux ~3.5×10^6 ph/s; total time ~10 h per sample; estimated surface dose ~2×10^7 Gy. Absence of radiation damage confirmed by reconstructing sub-tomograms from each half-dataset. Reconstructions: ptychography via difference map then maximum-likelihood; tomography via filtered back projection; half-pitch resolution ~100 nm by Fourier shell correlation. Image analysis: PXCT volumes processed in Avizo and ImageJ: background subtraction (Gaussian blur for low-frequency removal), cropping of artifacts, median filtering, two-phase watershed segmentation. Segmented HPMC-rich domains visualized; local thickness computed in ImageJ as the diameter of the largest inscribed sphere per voxel to quantify hydrophilic domain size distributions and connectivity.

- Thermal/solid-state: DSC and XRD confirm amorphous ASDs at 10 wt% nicotinamide for all PLA/HPMC ratios studied. No nicotinamide melting endotherm; XRD shows broad amorphous halos. The PLA-rich phase Tg ~50 °C (vs ~45 °C expected if drug equally partitioned), implying ~5 wt% nicotinamide in PLA and preferential partitioning to HPMC. At 30 wt% drug loading, nicotinamide crystallized post-extrusion and ASDs could not be formed. - Dissolution behavior (pH 6.8, 37 °C): Strong dependence on PLA/HPMC ratio. 30/70 PLA/HPMC shows burst release with most drug released within 1–2 h. 50/50 shows faster initial release then transitions to slower, near-linear release after ~1 h. 70/30 shows slow, near-linear release with only ~20% released after 6 h. - Post-leaching morphology (SEM, after 5 days): HPMC leached, leaving a porous PLA-rich skeleton. Higher HPMC content (50/50 and 30/70) yields more brittle, highly porous remnants with thin PLA walls; 70/30 shows denser morphology with more spherical pores, indicative of lower HPMC connectivity. PLA phase remains macroscopically cohesive in all cases; filaments swell during dissolution due to HPMC hydration. - Drug distribution and pristine morphology (STXM/NEXAFS): Phase separation into PLA-rich and HPMC-rich domains is confirmed. Nicotinamide signatures present in both phases; NEXAFS shows doublet C 1s→π* resonance in PLA-rich regions and a single peak in HPMC-rich regions, consistent with different intermolecular environments, precluding direct quantitative partitioning by NEXAFS. Drug appears homogeneously distributed within each polymer domain at ~100 nm resolution. Rare bright spots correspond to tiny nicotinamide crystals; overall crystalline amount negligible at 10 wt%. - Morphology vs composition (STXM imaging at 288 eV): 50/50 has micron-scale HPMC-rich domains that are relatively uniform and well-connected. 70/30 exhibits HPMC-rich domains spanning <1 µm to tens of µm with reduced connectivity through the PLA-rich matrix. - 3D connectivity (PXCT): Electron density contrast distinguishes PLA (0.412 e/Å^3) from HPMC (0.407 e/Å^3). Both ASDs (70/30 and 50/50 at 10 wt% drug) possess globally well-connected HPMC networks (with some peripheral exceptions likely connected outside the field of view). Adding drug to a 70/30 polymer blend increases HPMC domain size and spacing between connected regions, attributed to drug plasticization during extrusion. - Local thickness analysis (PXCT): 70/30 shows a mixture of narrow domains and some very large HPMC regions (>7 µm), with narrower average local thickness for small domains (mean ~0.85 µm). 50/50 exhibits a more homogeneous thickness distribution with larger average local thickness for narrow domains (mean ~1.2 µm). Larger, more connected hydrophilic pathways in 50/50 facilitate penetration of dissolution medium and faster release; narrower, more tortuous HPMC networks in 70/30 correlate with extended, slower release. - Mechanistic insight: Swelling of HPMC during dissolution fractures the initially rigid PLA matrix into thin walls, increasing PLA surface exposure and shortening diffusion paths, enabling partial drug release from PLA within days, despite PLA’s typical long-term release behavior. Morphology thus acts as a rate-controlling factor superimposed on intrinsic polymer and drug dissolution characteristics.

The results directly link phase-separated morphology to controllable drug release in dual-polymer ASDs. By adjusting PLA/HPMC ratio, the hydrophilic HPMC network’s connectivity and thickness can be tuned, which governs the ingress of dissolution medium and the drug’s diffusion pathways. STXM and PXCT reveal that at higher HPMC content (50/50), larger, better-connected hydrophilic channels promote rapid wetting and diffusion, producing faster initial and sustained release. At higher PLA content (70/30), the hydrophilic pathways are narrower and less connected, slowing medium penetration and yielding extended, near-linear release. Drug is present in both phases, predominantly favoring HPMC, which ensures a significant fraction is available for release on oral timescales; however, morphological evolution upon HPMC swelling increases PLA exposure and enables appreciable drug release from PLA as well, improving overall utilization. The dissolution mechanism is context-dependent: for highly soluble drugs like nicotinamide, diffusion through the wetted HPMC network dominates; for poorly soluble drugs, erosion of the hydrophilic matrix may control release while drug concentration remains below congruent release limits. Overall, morphology engineering via composition and drug-induced plasticization provides a practical lever to tailor release profiles without changing drug or polymer chemistry.

Phase-separated polymer blends produced by hot-melt extrusion provide a robust strategy to tailor drug release in amorphous solid dispersions. In nicotinamide/PLA/HPMC systems, increasing hydrophobic PLA content narrows the hydrophilic HPMC network and slows release, whereas higher HPMC content creates larger, connected domains that accelerate dissolution. Advanced X-ray imaging (STXM, PXCT) establishes how polymer fraction and drug presence modulate domain size, connectivity, and drug distribution, and dissolution studies confirm the strong morphology–release correlation. Tailoring composition during extrusion thus enables manufacturing of oral dosage forms with targeted release profiles. Future work could explore broader drug chemistries and loadings, alternative hydrophilic/hydrophobic polymer pairs, quantitative phase-specific drug partitioning, and in vivo–in vitro correlations to generalize morphology–release design rules.

- Quantitative drug partitioning between phases by NEXAFS was not possible due to phase-dependent resonance shape differences and, in some cases (e.g., thicker 50/50 sections), higher uncertainty from sample thickness. - High drug loading (30 wt%) led to immediate nicotinamide crystallization post-extrusion, preventing ASD formation, limiting exploration of higher loadings. - Dissolution experiments were conducted with duplicate baths (n=2), and observations are limited to in vitro conditions over hours to days; generalizability to other drugs/polymers and in vivo settings may vary. - PXCT/STXM probe limited fields of view; while connectivity was high within imaged volumes, edge domains may connect outside the volumes, and nanoscale features below ~100 nm resolution were not resolved. - Morphology evolves during dissolution due to HPMC swelling, leading to PLA wall thinning and fracture; comparisons between pristine and leached morphologies should consider this dynamic restructuring.