3D printing of spray-dried extracts: An additive manufacturing approach to plant-based solid dosage forms

Abstract

Plant-derived materials frequently exhibit poor flowability and compactibility, complicating their conversion into solid dosage forms. This preliminary study investigated whether chemically complex spray-dried extracts of Eugenia uniflora and Punica granatum could be processed using semi-solid extrusion (SSE) 3D printing and how extract identity and manufacturing route influence the resulting product attributes. For this purpose, printed units were compared with nominally composition-matched tablets produced by wet granulation followed by compression to characterize the mechanical, disintegration, and marker-dissolution behavior associated with each process. Following paste screening, a 2³ factorial design identified sodium starch glycolate, filled top/bottom layers, and drying at 24 °C as the combination providing the most favorable balance between mechanical strength and disintegration for both extracts. Under the selected geometries, printed units weighed 701–751 mg, compared with 501–502 mg for compressed tablets with conventional punches design, while all formulations showed friability below 1%. Printed units required higher breaking forces than compressed tablets for P. granatum (44.2 ± 2.95 vs 31.3 ± 4.27 N) and E. uniflora (57.4 ± 4.05 vs 30.8 ± 2.80 N), accompanied by longer disintegration and slower initial marker dissolution. Printed P. granatum units approached complete marker dissolution within 30 min, whereas printed E. uniflora units reached approximately 75–80%. These findings provide preliminary evidence that SSE can process complex plant extracts into mechanically resistant solid forms. However, extract-dependent differences demonstrate that fully exploiting the design and dose flexibility of SSE requires formulation and process optimization according to the composition of the incorporated botanical matrix and the intended product performance.

1. Introduction

Continuous innovation in pharmaceutical dosage forms is essential to meet the growing demand for products with variable shapes, doses, sizes, and applications [1]. In this context, additive manufacturing (AM), particularly three-dimensional (3D) printing, has emerged as a flexible strategy for pharmaceutical development. Several printing technologies have been investigated for drug-delivery applications, but their suitability depends strongly on the properties of the formulation and the intended product [2,3].

For example, fused deposition modeling requires the previous production of drug-loaded filaments and commonly involves elevated processing temperatures, which may limit its application to thermolabile materials [4]. Powder-based techniques depend on adequate powder flow, spreading, and binding, whereas vat photopolymerization requires photosensitive materials and raises additional concerns regarding drug compatibility and residual formulation components [3]. Semi-solid extrusion (SSE), in contrast, processes pastes or gels under relatively mild conditions, avoids filament preparation, accommodates high active-material loads, and enables geometric customization and on-demand production. Nevertheless, SSE also presents specific challenges related to paste rheology, extrusion continuity, shape retention, dimensional fidelity, and forms attributes required. [1,2,5,6]. These challenges are particularly relevant to enable manufacturing herbal dosages forms applying 3D printing technique [7].

Despite advances in drying and powder-processing technologies, plant extracts, including spray-dried powders, frequently remain unsuitable for direct compression [8]. Their poor flowability, low compactibility, and high hygroscopicity compromise manufacturability in conventional tableting [[9], [10], [11], [12]]. Wet, dry, or melt granulation may mitigate some of these limitations; however, producing mechanically robust tablets with satisfactory disintegration and dissolution behavior remains challenging, particularly when high proportions of plant-derived materials are required [[12], [13], [14]]. Alternative manufacturing approaches capable of accommodating these complex materials are therefore needed.

SSE-based 3D printing may expand the technological options available for converting poorly compressible plant extracts into solid dosage forms. However, despite the increasing application of SSE to pharmaceutical products, its use with chemically complex plant extracts remains relatively underexplored [7,15]. Most studies have focused on formulations containing isolated active pharmaceutical ingredients, and the transfer of these findings to multicomponent plant extracts is not straightforward [16]. Extract constituents may necessity functional polymers, which able the paste preparation and printability. Although, have an infinity of polymers, play an important role, because extract may interact with polymers and other excipients, altering liquid requirements, rheological behavior, extrudability, layer adhesion, drying shrinkage, and final product performance [7]. Moreover, direct comparisons between printing formulation for SSE-printed dosage forms and composition-matched tablets manufactured by conventional methods remain limited. Such comparisons are important for distinguishing the effects of formulation composition from those specifically associated with the manufacturing route.

In this context, the present study selected spray-dried extracts of Eugenia uniflora and Punica granatum as representative plant-derived active materials based on the maturity of the scientific evidence supporting their progression toward pharmaceutical development. The development of a plant-derived material into a viable pharmaceutical active is a progressive and multidisciplinary process that integrates chemical characterization and standardization, pharmacological and safety evidence, analytical development, and technological studies aimed at obtaining a reproducible pharmaceutical product [17]. In this context, both species have been investigated regarding their phytochemical composition, pharmacological activities, safety, and chemically characterized spray-dried extracts, with established analytical markers available for both materials: myricitrin for E. uniflora and granatins A and B for P. granatum [[18], [19], [20], [21], [22], [23]]. This accumulated evidence provides a rational basis for advancing these extracts to technological studies focused on their incorporation into pharmaceutical dosage forms. Accordingly, the present study addresses an additional stage in this development pathway by investigating the feasibility of incorporating these chemically complex extracts into SSE-printed solid dosage forms.

Therefore, we hypothesized that the distinct consolidation mechanisms involved in SSE printing and compression would result in different product attributes, although neither the direction nor the magnitude of these differences was predefined, and that the effects of the manufacturing route could vary according to the physicochemical complexity of the incorporated plant extract. Accordingly, this study aimed to develop an SSE-printable formulation platform containing spray-dried extracts of Eugenia uniflora or Punica granatum and to determine how extract identity and manufacturing route influence the properties of the resulting solid dosage forms. This comparative assessment was intended to provide a preliminary understanding of how complex botanical matrices respond to SSE processing, thereby supporting the future development of 3D-printed drug-delivery systems tailored to specific dose, geometry, and performance requirements.

3. Results and discussion

3.1. Definition of paste formulation

Among the preliminary formulations evaluated (Table 1), PF2 and PF3, containing high proportions of microcrystalline cellulose (MCC) and polyvinyl alcohol (PVA), respectively, did not form homogeneous and continuously extrudable pastes. These formulations exhibited visible grittiness, inadequate particle dispersion, and phase separation, which impaired manual extrusion and shape retention after deposition.

These results revealed inadequate dispersion ability between the ingredients selected for PF2 and PF3. This behavior may be related to the solubility properties of the excipients and how they disperse into binder. In the case of PVA, despite the well-established use for 3D printing in hot-melt extrusion processes for filament fabrication in fused deposition modeling (FDM) or in powder bed fusion techniques, its application in semisolid extrusion (SSE) presents significant challenges [4,33]. Effective incorporation of PVA in SSE systems typically requires relatively high concentrations (approximately 20–25%) and the presence of plasticizers, lubricants, and binders, in addition to elevated temperatures, to achieve sufficient dispersion and film formation to improve flowability through the syringe [34]. In the context of this study, such processing conditions were deliberately avoided, given that the plant-based materials employed may be thermosensitive and susceptible to degradation or undesired interactions with multiple excipients. Consequently, PVA was incorporated as a dry powder without prior solubilization or gelation. This hindered its integration into the paste matrix and prevented it from fulfilling its intended role as a film-forming polymer. Even at low concentrations, undissolved PVA compromised the homogeneity of the mixture, leading to insufficient cohesion and adversely affecting both flowability and printability of the formulation.

Similarly, the high MCC content of PF2 also impaired paste formation. Although MCC is widely used as a diluent in conventional solid dosage forms, its insolubility and strong water-binding capacity can be detrimental in extrusion-based 3D printing. In PF2, the solid content exceeded 45%, primarily cellulose, which likely surpassed the critical threshold for stable flow [35]. The reduced availability of free liquid impaired particle wetting and dispersion, resulting in a dry, granular, and poorly cohesive paste. These limitations are consistent with reports on cellulosic materials in 3D printing, where their hydrophilicity and poor matrix compatibility hinder dispersion and interfacial interaction. At high concentrations, cellulose fibers may reach the “gel crowding factor”, leading to entanglement and loss of flowability [35]. Similar effects are observed in hydrogel systems, where high solid content reduces shrinkage but compromises extrusion and structural fidelity [36].

Conversely, formulations containing higher proportions of lactose as filler and either low amounts of MCC (PF1) or Kollicoat® IR as the film-forming polymer (PF4) demonstrated markedly improved paste homogeneity and handling. These pastes exhibited more uniform flow during extrusion, with enhanced strand cohesion and ease of deposition. These results are closely related to the physicochemical properties and proportions of the ingredients used.

Lactose, being partially water-soluble, likely facilitated better dispersion of the solid components and reduced internal friction within the paste, contributing to smoother extrusion dynamics. When used at appropriate levels, MCC also contributed positively by reinforcing the structural strength of the paste [34,36]. In the same way, Kollicoat® IR, a graft copolymer composed of 75% polyvinyl alcohol (PVA) and 25% polyethylene glycol (PEG), further enhanced printability. Its physicochemical profile, characterized by low intrinsic viscosity and high flexibility, reduces the need for additional plasticizers. The PEG moiety acts as an internal plasticizer, increasing elasticity and supporting filament integrity during extrusion and deposition [24].

Supporting these findings, Zidan, Alayoubi, Asfari, Coburn, Ghammraoui, Aqueel, Cruz and Ashraf [37] demonstrated that the proportion of soluble excipients is among the most influential factors affecting the viscoelastic behavior of extrudable pastes, particularly their creep response. In the gel-based matrices evaluated, increasing the proportion of soluble materials improved paste flow, whereas high concentrations of swellable excipients increased crossover stress and reduced extrudability. Although these findings cannot be directly extrapolated to the present system, they may help explain the improved performance of PF1 and PF4, in which lactose accounted for more than 90% of the solid phase, compared with PF2, which contained a high proportion of MCC and exhibited poor extrudability.

Despite the improved performance of PF1 and PF4, the use of water or 90% ethanol as wetting liquids resulted in post-deposition defects, including syneresis, shrinkage, and interlayer delamination during drying [38]. These defects are commonly observed in SSE systems, in which rapid solvent evaporation may promote micropore formation. Without sufficient matrix elasticity or structural cohesion, the printed constructs may undergo dimensional collapse due to internal stresses and volume loss during drying [38,39].

To overcome these limitations, the wetting liquid was replaced by a 5% (w/v) Soluplus® solution, yielding formulations PF1.2 (lactose, glycerol, and MCC 102) and PF4.2 (lactose, glycerol, and Kollicoat® IR). These pastes exhibited improved processing behavior, including stable filament flow, improved layer stacking, and enhanced structural integrity. Soluplus® contributed to these properties by increasing the viscosity of the continuous phase. Previous studies have shown that Soluplus® can impart measurable yield stress, support elastic recovery, and provide film-forming capacity at strand interfaces. Together, these effects may mitigate syneresis, shrinkage, and interlayer slippage during drying [1,40].

Following this optimization, both formulations were evaluated after incorporation of the spray-dried Eugenia uniflora extract. The glycerol concentration was adjusted to improve lubricity and maintain consistent flow while remaining within the concentration range previously reported for pharmaceutical applications [37]. Among the formulations evaluated, PF4.2.1 exhibited the most favorable performance (Fig. 2). It maintained filament cohesion during extrusion, enabled smooth and continuous strand deposition with minimal die swell, and exhibited well-defined interlayer adhesion. Drying-related deformation, including warping and edge rounding, was visibly reduced, resulting in better definition of both the perimeter and infill regions of the printed structures. Thus, even after incorporation of the phenolic-rich E. uniflora extract, the optimized formulation retained the processing and structural characteristics required for printing. These findings suggest that the lactose–polymer–binder system provided sufficient stabilization to accommodate the chemically diverse constituents of the extract [23,41].

Regarding the dimensional fidelity of PF4.2.1 containing spray-dried E. uniflora extract, the units measured immediately after printing showed mean length, width, and thickness values of 17.29 ± 0.29, 10.46 ± 0.40, and 4.91 ± 0.40 mm, respectively. These measurements corresponded to deviations of +1.69%, +16.22%, and +40.38% from the nominal CAD dimensions. Length, width, and thickness differed significantly from their respective nominal values (p = 0.002, p < 0.001, and p < 0.001, respectively). After drying, the mean length, width, and thickness decreased to 16.67 ± 0.98, 9.80 ± 0.86, and 4.13 ± 0.20 mm, respectively, corresponding to deviations of −1.96%, +8.89%, and +18.10% from the CAD model. The length of the dried units did not differ significantly from the nominal value (p = 0.207), whereas width and thickness remained significantly greater than their respective CAD dimensions (p = 0.003 and p < 0.001, respectively). The relative standard deviations ranged from 1.66% to 8.17% immediately after printing and from 4.81% to 8.79% after drying, indicating limited dimensional variability among the units obtained from the three independent printing batches.

Paired analysis revealed significant mean reductions of 3.56 ± 5.92% in length (p = 0.034), 6.07 ± 10.21% in width (p = 0.034), and 15.25 ± 9.05% in thickness (p < 0.001) during drying. The more pronounced reduction along the build direction suggests anisotropic matrix consolidation during solvent removal. The initial oversizing likely resulted from filament spreading and flattening, as well as overlap between adjacent deposition paths. Panraksa, Zhang, Rachtanapun, Jantanasakulwong, Qi and Jantrawut [42] reported a material-spreading rate of 15.65 ± 5.58% for an optimized SSE formulation, which was comparable to the 16.22% initial width deviation observed in the present study. Pronounced dimensional changes along the build direction have also been reported by Teoh, Zhang, Belton, Chan and Qi [43] and Zhang, Belton, Teoh, Gleadall, Bibb and Qi [44], who associated thickness loss with sagging, filament merging, and vertical collapse during drying. Although the dried PF4.2.1 units retained 98.04% of the intended length and preserved their overall oblong geometry, systematic deviations remained in width and thickness.

The magnitude of dimensional change during drying is strongly dependent on formulation composition. Johannesson, Khan, Hubert, Teleki and Bergström [45], for example, reported reductions of approximately 15–21% in diameter and 17–33% in thickness for SSE-printed gel-emulsion tablets, showing a trend similar to that observed in the present study. In addition to formulation composition, excess liquid and high concentrations of hydrated polymers can intensify structural collapse and surface shrinkage during drying [46]. The relatively limited lateral contraction and preservation of visible layers in PF4.2.1 suggest that the formulation provided sufficient cohesion to maintain the overall structure during solvent removal, supporting its selection for further optimization. The remaining systematic deviations from the CAD dimensions may be addressed through adjustment of the extrusion parameters or compensation of the model dimensions without necessarily requiring reformulation. Overall, RSD values below 9% indicate reasonably consistent dimensional behavior across the three independent printing batches.

On this basis, the working base for printing was defined as lactose (filler), glycerin (humectant/plasticizer), Kollicoat® IR (film-forming polymer), and a 5% (w/v) Soluplus® solution as the liquid binder, into which the spray-dried extracts were incorporated at the target loads.

Download the full article as PDF here 3D printing of spray-dried extracts

or continue reading here

João Victor Rocha Bitú, Milena Fernandes da Silva, James Correia de Melo, Janaina Carla Barbosa Machado, Luiz Alberto Lira Soares, Magda Rhayanny Assunção Ferreira, 3D printing of spray-dried extracts: An additive manufacturing approach to plant-based solid dosage forms, Journal of Drug Delivery Science and Technology, Volume 127, Part 1, 2027, 108883, ISSN 1773-2247, https://doi.org/10.1016/j.jddst.2026.108883.

Read also our introduction article on 3D Printing here: