The Evolving Science of Amorphous Solid Dispersions

Key Takeaways

• Polymer selection for ASDs requires evaluating crystallization resistance, dissolution, and membrane transport as distinct endpoints.

• Drug-loading limits depend on the compound and formulation; ASD loading also differs from drug content in the finished tablet.

• Phase separation and crystallization are different changes, and drug-rich species can accompany either effective release or dissolution failure.

• Spray drying and hot-melt extrusion require formulation-specific assessment of solvent behavior, thermal exposure, and product performance.

• Scale-up should assess product attributes alongside processing conditions, secondary drying, analytical transfer, and finished-dosage-form manufacture.

Before an orally administered small-molecule drug can be absorbed, it must dissolve in gastrointestinal fluids. For compounds that dissolve poorly in water, this step presents a formulation challenge: how can the medicine make the drug available for absorption? Amorphous solid dispersions (ASDs) address that challenge by changing the drug’s physical form and combining it with a polymer carrier.

In a crystalline drug, molecules occupy a repeating, ordered structure. An amorphous drug lacks that long-range order. Its higher-energy state can improve solubility and dissolution, but it also creates a tendency to crystallize. The polymer helps stabilize the amorphous material through interactions with drug molecules and restrictions on their movement.

Maintaining that balance becomes more complicated when the formulation encounters liquid. A polymer that protects the drug during storage may perform less favorably during dissolution, and increasing the amount of drug in a dispersion can compromise release. Manufacturing introduces further variables: solvent composition, drying conditions, and thermal exposure can influence the material ultimately produced.

Recent research is clarifying why these tradeoffs occur. Studies that follow drug-rich droplets during dissolution, compare polymer performance across different endpoints, and examine processing conditions across equipment scales are helping connect molecular behavior with practical development choices.^1—4 For sponsors and manufacturers, that understanding can guide the selection of a dispersion that releases effectively and can be made reproducibly as a finished dosage form.

Polymer selection can produce different answers depending on what is measured. A 2026 indomethacin study compared hydroxypropyl methylcellulose acetate succinate (HPMCAS), polyvinylpyrrolidone–vinyl acetate (PVPVA), and the amino methacrylate copolymer Eudragit E PO.²

Each performed best on a different endpoint. Eudragit E PO provided the greatest resistance to crystallization. HPMCAS generated the highest supersaturation during dissolution, while PVPVA produced the highest drug flux across the artificial membrane used in the study. Supersaturation means that a solution contains a higher drug concentration than would exist at equilibrium; flux measures the rate of transport per unit membrane area. The authors identified PVPVA as a favorable compromise between physical robustness and permeation efficiency in the investigated system.

The distinction between concentration and transport is particularly relevant. HPMCAS led on supersaturation, but PVPVA led on membrane flux. Selecting a polymer using only the concentration achieved during dissolution would therefore have produced a different ranking from the transport experiment. Neither measurement established differences in human exposure.

Composition further complicates selection. In felodipine dispersions, researchers examined drug solubility and miscibility in three polymeric carriers, then measured crystal growth at predefined drug loadings.⁵ Crystal-growth rates depended on viscosity and showed an exponential relationship with loading within the systems studied.

The polymer and its concentration consequently need to be evaluated together. The felodipine work provides a specific relationship to investigate within a formulation family, while the indomethacin comparison shows why resistance to crystallization should remain one selection criterion among several.^2,5

Loading percentages require a clear denominator. A posaconazole study selected an ASD containing 70% drug, but the finished tablet contained 50% after incorporation of other formulation components.⁴ Dispersion loading and tablet loading described different materials.

Dissolution testing helped select the composition. Dispersions containing 60% and 70% posaconazole showed appreciable dissolution, whereas the 80% formulation showed negligible dissolution under the tested conditions. The authors proposed rapid crystallization on contact with the medium as the likely explanation for the poor result at 80%.⁴

A different compound, GDC-6893, performed favorably at higher loading.6 Spray-dried dispersions containing either HPMCAS or PVPVA showed good drug–polymer miscibility and physical stability even at 80% drug loading. All tested formulations reached the maximum achievable diffusive flux in the study’s experimental system.

The researchers also assessed bioaccessibility using a gastrointestinal model. The ASDs improved bioaccessibility compared with crystalline GDC-6893, while polymer type and loading did not significantly affect the profiles within the tested formulation set.

The posaconazole and GDC-6893 studies involved different compounds and experimental conditions rather than a controlled comparison between formulations. Their value together is to demonstrate the variability of the loading problem. One system’s successful percentage cannot establish another’s target. The useful range must accommodate the particular formulation’s stability, release, and dosage-form requirements.

Crystallization and phase separation describe different changes. Crystallization involves formation of an ordered molecular structure. Phase separation concerns composition: a mixture can separate into drug-rich and polymer-rich regions. Compositional separation does not, by itself, establish that crystals have formed, although phase separation can affect release and lead to crystallization. This distinction helps frame investigations of poor dissolution. A formulation may change during manufacture, or release-limiting behavior may emerge only after contact with liquid. Identifying the timing and nature of the change gives a development team a more specific problem to address.

Solvent composition can influence phase behavior during spray drying. In a study exploring ritonavir–PVPVA dispersions, adding water to methanol induced phase separation, with more extensive heterogeneity at the higher water content investigated.⁷

At 25% drug loading, material prepared from a spray solvent containing 10% water and 90% methanol showed compromised dissolution compared with material prepared from water-free solvent. The solid-state characterization used did not detect phase heterogeneity in that formulation. The authors suggested that slow-releasing drug-rich phases could explain the difference.

Changes at the dissolution interface provide another explanation for poor release. In clotrimazole dispersions prepared with different molecular-weight grades of polyvinylpyrrolidone (PVP), researchers observed stable drug-rich nanodroplets in formulations with good release.¹ Rapid droplet agglomeration accompanied one of the poor-release mechanisms identified in the study.

The researchers examined these behaviors in relation to the limit of congruency, the drug loading above which a drug-rich barrier can hinder release. Their findings show why identifying drug-rich species is insufficient on its own: stable nanodroplets accompanied good release, while their rapid agglomeration accompanied a failure mechanism. Polymer molecular weight affected the observed behavior.

Mechanistic observations can therefore make formulation experiments more selective. An investigation aimed at preventing agglomeration addresses a clearer question than another broad search for an improved dissolution profile.

A release-enhancing adjustment can change other performance measures. In a study of HPMCAS-based dispersions of ivacaftor and ARV-825, release was very slow or undetectable at higher drug loadings.⁸ Adding 10% by weight glyceryl tributyrate substantially improved release.

The additive also affected the material generated during dissolution. Ivacaftor dispersions produced nanosized drug-rich amorphous droplets that enhanced permeation across Caco-2 cell membranes, with smaller droplets associated with faster permeation. Glyceryl tributyrate produced larger droplets and decreased permeation despite improving release.

The authors explicitly identified the implications for in vivo performance as unknown. The demonstrated tradeoff was between drug release and cell-based transport: the additive addressed slow liberation from the ASD, but the resulting formulation performed less favorably in the permeation experiment.

This example supports following a promising formulation adjustment beyond its immediate objective. A better dissolution curve identifies an improvement in that measurement; further experiments establish what else changed and whether the overall result supports advancement.

Spray drying and hot-melt extrusion expose formulations to different processing environments. In spray drying, the drug and other ingredients are dissolved in a solvent, the solution is atomized into droplets, and heated gas removes solvent to form a powder. Residual solvent can remain and require secondary drying.

Hot-melt extrusion uses heat and screw-driven mixing to combine the drug and polymer. In twin-screw equipment, rotating, intermeshing screws mix the components to produce a drug–polymer dispersion or solid solution. The solvent-free process requires attention to the material’s thermal experience, including the relationship between temperature, exposure time, and degradation.

These differences suggest distinct priorities for route-selection experiments. Spray-drying development should address solvent-dependent phase behavior and solvent removal. Extrusion development should establish whether the required thermal and mixing conditions preserve product quality. Both routes need evaluation through the properties of the resulting dispersion.

An indomethacin comparison demonstrates why product measurements matter.⁹ Hot-melt-extruded samples showed greater resistance to recrystallization, while spray-dried samples achieved higher intrinsic dissolution rates. PVP K30 outperformed hydroxypropyl methylcellulose E5 in both stability and intrinsic dissolution within the investigated system. That result is a formulation-specific tradeoff rather than a general ranking of the technologies. The preferred route should follow the attributes required for the drug–polymer combination under development.

Predictive methods can help connect formulation and process choices. Researchers combined drug–polymer–solvent phase behavior and glass transitions with mass and energy balances to identify spray-drying design spaces for ritonavir and naproxen with PVP or PVPVA.¹⁰ Glass transition describes the change between a rigid, glassy amorphous state and a more mobile, rubbery state. Including this behavior helped connect formulation properties with solvent-removal conditions.¹¹

Such work can guide experiments toward relevant processing conditions, provided the model’s scope remains clear. A design space established for selected compounds, polymers, and solvents supplies evidence for those systems, rather than a universal manufacturing recipe.

Matching a process characteristic across equipment does not necessarily reproduce the product. In an acetaminophen–PVP K30 spray-drying study, matching atomized droplet-size distributions alone did not reproduce powder characteristics between pilot- and production-scale dryers, although similar morphology and crystallinity could be achieved.¹²

The criterion did not account for differences in droplet–drying-gas mixing and residence-time distributions. The authors identified the need for production-scale experiments to obtain characteristics similar to those achieved at pilot scale.

Extrusion presents a related challenge through changes in heat generation. A dipyridamole–copovidone study examined extruders of different sizes and evaluated total energy imparted to the material, conducted heat, and differences between barrel and melt temperatures.³ Experiments and simulation addressed the changing contributions of heat conduction and viscous dissipation across scales.

A pilot-scale extrusion investigation focused on drug degradation illustrates the consequences of incomplete transfer.¹³ None of six tested scale-up variants reproduced the laboratory process’s product quality. Modeling enabled correlations between degradation, local melt temperature, and exposure time in particular screw zones; those relationships also applied to laboratory-scale results. Local thermal conditions helped explain degradation that the tested transfer strategies did not adequately control. The authors described prediction before initial extrusion trials as a possible future application, rather than an established capability.

These studies support checking whether scale-up preserves the material’s relevant processing conditions and product attributes. Comparable settings are a starting point for that investigation, while product measurements determine whether the transfer achieved its purpose.

For spray-dried materials, powder collection can be followed by substantial additional processing. Secondary drying removes residual solvent and, for some materials, requires balancing drying time against stability risks.

Controlled humidity offers one experimental approach. In an indomethacin–PVPVA ASD containing residual ethanol, water-assisted drying at 30 °C accelerated ethanol removal compared with conventional drying.¹⁴ Thermodynamic modeling guided selection of starting compositions that avoided water-induced phase separation. The experiments used dynamic vapor sorption with integrated Raman spectroscopy, rather than commercial-scale drying equipment. The result connects a process objective with formulation-specific constraints: improving solvent removal also required controlling the risk of phase separation.

Monitoring methods need attention during transfer as well. In another study, researchers used online near-infrared spectroscopy to determine secondary-drying endpoints and quantify residual acetone, water, and drug content in a spray-dried intermediate.¹⁵ Transfer to commercial scale on another analyzer at a different site required calibration-model updates to address instrumentation and scale differences.

Secondary drying and analytical transfer consequently deserve a defined place in process development. The evidence needed to support a larger-scale process extends beyond the initial step that forms the dispersion.

The posaconazole work followed formulation selection into tablet manufacture.⁴ The selected 70%-drug-loaded ASD had low bulk density and poor flowability; granulation improved both properties before preparation of the final tablet.

The dissolution evidence retained an important boundary. The authors did not extend testing to the point where precipitation could occur or add extra polymer to prevent it. They also cautioned that differing test conditions prevented a definitive dissolution comparison with another formulation.

The study thus documents progress through powder handling and tablet manufacture while leaving aspects of solution performance unresolved. Keeping those findings distinct gives the next development stage a concrete purpose: test the remaining performance question without overlooking the manufacturing progress already demonstrated.

A candidate ASD should advance with evidence that connects its composition to release performance, its processing conditions to product quality, and its powder properties to the finished dosage form. The remaining uncertainties should be equally explicit. This gives sponsors and manufacturing partners a practical basis for selecting the formulation to manufacture, defining what scale-up must preserve, and deciding which experiment comes next.

1. Deac, Alexandru, et al. “Dissolution mechanisms of amorphous solid dispersions: Role of polymer molecular weight and identification of a new failure mode.” Journal of Pharmaceutical Sciences. 114: 486–496 (2025).

2. Kumar, N S Krishna, et al. “Effect of Drug–Polymer Interaction on Solid-State Stability and In Vitro Drug Flux of Amorphous Solid Dispersions.” Molecular Pharmaceutics. 23: 3195–3205 (2026). https://doi.org/10.1021/acs.molpharmaceut.6c00196

3. Zecevic, Damir E, et al. “From benchtop to pilot scale-experimental study and computational assessment of a hot-melt extrusion scale-up of a solid dispersion of dipyridamole and copovidone.” International Journal of Pharmaceutics. 537: 132–139 (2018).

4. Xiang, Tianyi, et al. “Efficient development of high drug loaded posaconazole tablets enabled by amorphous solid dispersion.” RSC Pharmaceutics. 2: 178–185 (2025).

5. Tian, Yiwei, David S Jones, and Gavin P Andrews. “An investigation into the role of polymeric carriers on crystal growth within amorphous solid dispersion systems.” Molecular Pharmaceutics. 12: 1180–1192 (2015).

6. Shetty, Nivedita, et al. “Assessing the impacts of drug loading and polymer type on dissolution behavior and diffusive flux of GDC-6893 amorphous solid dispersions.” Journal of Pharmaceutical Sciences. 114: 103686 (2025).

7. Li, Na, et al. “Water-Induced Phase Separation of Spray-Dried Amorphous Solid Dispersions.” Molecular Pharmaceutics. 17: 4004–4017 (2020).

8. Shah, Dishan D, et al. “Addressing the Release and Permeation Challenges of High-Tg Drugs in Amorphous Solid Dispersions.” Molecular Pharmaceutics. 23: 2385–2397 (2026).

9. Martynek, Dominik, et al. “Stability and recrystallization of amorphous solid dispersions prepared by hot-melt extrusion and spray drying.” International Journal of Pharmaceutics. 672: 125331 (2025).

10. Dohrn, Stefanie, et al. “Predicting process design spaces for spray drying amorphous solid dispersions.” International Journal of Pharmaceutics: X. 3: 100072 (2021).

11. Solomos, Marina A, et al. “Evaluating Spray Drying and Co-Precipitation as Manufacturing Processes for Amorphous Solid Dispersions of a Low Tg API.” Journal of Pharmaceutical Sciences. 112: 2087–2096 (2023).

12. Thybo, Pia, et al. “Scaling up the spray drying process from pilot to production scale using an atomized droplet size criterion.” Pharmaceutical Research. 25: 1610–1620 (2008).

13. Matić, Josip, et al. “Towards predicting the product quality in hot-melt extrusion: Pilot plant scale extrusion.” International Journal of Pharmaceutics: X. 3: 100084 (2021).

14. Kerkhoff, Jana, et al. “Water-Assisted Drying of PVPVA-Based Amorphous Solid Dispersions.” Molecular Pharmaceutics. 22: 3771–3781 (2025).

15. Ikeda, Craig, et al. “Application of Online NIR Spectroscopy to Enhance Process Understanding and Enable In-process Control Testing of Secondary Drying Process for a Spray-dried Solid Dispersion Intermediate.” Journal of Pharmaceutical Sciences. 111: 2540–2551 (2022).