Emerging 3D printing technologies for solid oral dosage forms: Processes, materials and analytical tools for real-time assessment

Abstract

Three-dimensional (3D) printing is an emerging technology with the potential to increase manufacturing flexibility and enable personalized drug delivery. 3D printing may form tablets using digitally controlled layer-by-layer material deposition, permitting the tailoring of solid oral dosage geometry and facile modifications of drug release profiles without requiring extensive alterations to the pharmaceutical formulation and process. The challenge to assure the quality of drugs still lies in monitoring and controlling critical steps in the 3D printing process. Optimizing an 3D printing process requires a comprehensive understanding of the critical process parameters, material attributes and their impact on the performance of 3D-printed tablets. This review focuses on recent advances in 3D printing technologies for solid oral dosage forms, emphasizing critical process parameters and material attributes that may be considered for optimizing printing processes and enhancing the quality of printed tablets. Additionally, this review explores real-time analytical tools and the crucial considerations for ensuring the performance of building materials, printing processes, and manufactured solid drug products. This review contributes to the ongoing discourse on harnessing the potential of 3D printing in the pharmaceutical field while emphasizing the imperative need for quality assurance throughout additive manufacturing processes.

1. Introduction

Additive manufacturing, often recognized as three-dimensional (3D) printing, emerged during the 1980s as a technology for manufacturing complex designs and customized parts. Over the years, this technology has transcended its origins in manufacturing and engineering, impacting diverse industries. The versatility of 3D printing technology allows the manufacturing of customized parts from an array of materials such as metals, ceramics, and polymers (Chia and Wu, 2015; Ligon et al., 2017; Ngo et al., 2018). 3D printing has been recognized for manufacturing medical devices and a variety of dosage forms with tailored material formulations (Di Prima et al., 2016; Goole and Amighi, 2016; Hsiao et al., 2018). 3D printing holds the potential to increase manufacturing flexibility by providing on-demand production of personalized solid dosage forms and to enable the distributed and point-of-care manufacturing of complex and advanced medical products in remote or austere conditions (FDA, 2022, 2023a, c). The appeal lies in the capability of 3D printing to tailor tablet geometry and easily modify drug release profiles without requiring extensive alterations to the pharmaceutical formulation (El Aita et al., 2018). The flexibility extends to creating single dosage forms with multiple active pharmaceutical ingredients (APIs), catering to patients with multiple conditions and individualized needs based on their age, demography, medical conditions, and genetic factors (Alayoubi et al., 2022).

Pharmaceutical manufacturing traditionally has centered on large-scale production and may not designed to support the flexibility and ability to tailor multistep drug production for small batches. Herein lies the promise of 3D printing, offering an alternative to conventional methods that rely on a one-size-fits-all approach (Lepowsky and Tasoglu, 2018; Trenfield et al., 2018a). Despite these benefits, the full implementation of 3D printing in pharmaceuticals is still limited by quality and safety challenges. No new 3D-printed drug products have been approved since the approval of Spritam^™ by the U.S. Food and Drug Administration (U.S. FDA) in 2015. Contrastingly, additive manufactured medical devices have been approved at a nearly exponential rate by the U.S FDA since 2010 (Fogarasi et al., 2023). The lack of new drug products utilizing these technologies highlight the challenge of bringing 3D-printed drug products to market and the critical necessity for assessing reproducibility and quality while improving in-process analytical tools and control strategies. Understanding material formulation requirements is crucial for a successful printing and assuring the quality of the drugs fabricated using 3D printing.

Although several approaches have been proposed in literature for 3D printing of tablets and other dosage forms using various geometric designs and technologies (Chia and Wu, 2015; Goole and Amighi, 2016; Muhindo et al., 2023; Prasad and Smyth, 2016; Ursan et al., 2013), relatively few have delved into material properties and their interaction with printing parameters (Jamróz et al., 2018; Rahman et al., 2018). The complex interplay between material quality attributes, processing parameters, geometric design parameters and tablet quality remains underexplored. The diverse material types utilized in 3D printing necessitate novel characterization methods to evaluate their quality attributes and predict their printability. Additionally, assessing the quality of raw materials and the associated impact on material flow and printability is necessary to adequately control the manufacturing processes and determine the final quality of the printed dosage forms (Lamichhane et al., 2019; Ligon et al., 2017; Melocchi et al., 2020).

This review article discusses quality considerations that may be considered for adopting 3D printing technologies within the pharmaceutical field. These considerations include criticality of material attributes and printing parameters in evaluating the quality of manufactured tablets by various 3D printing processes. In this regard, recent studies that focus on assessing material quality attributes and process parameters across seven distinct 3D printing technologies, namely pressure assisted micro-syringe (PAM), fused filament fabrication (FFF), droplet deposition modeling (DDM), binder jetting (BJT), material jetting (MJT), stereolithography (SLA) and selective laser sintering (SLS) are compiled. This review also highlights various real-time analytical tools used in the quality assessment of material and printing process in the manufacturing of solid oral dosage forms using these 3D printing technologies.

2. Classification of 3D printing technologies

Additive manufacturing is an overarching term that includes a wide range of printing technologies, each offering unique attributes and applications. The American Society for Testing and Material (ASTM) classifies 3D printing technologies into seven main categories: material extrusion-based, binder jetting, material jetting, Vat polymerization, powder bed fusion, sheet lamination, and direct energy deposition (ASTM-F2792–12a, 2012). Table 1 narrows down this ASTM classification specifically for pharmaceutical applications based on the physical attributes of starting building materials and the targeted dosage forms for manufacturing. Table 1 provides also examples of various drug products manufactured using these technologies, including immediate-release (IR) and extended-release (ER) tablets, implants, oral films, intravaginal rings, pulsatory devices, API encapsulation, microporous controlled-release drug delivery devices and orodispersible films.

Table 1.

Common 3D printing technologies in pharmaceutical applications.

Group for Pharmaceutical Applications	3D Printing Technique	Starting Building Materials	Dosage Form	Solidifying Method
Material Extrusion-Based	FFF	Solid filaments plastics	IR and ER solid dosage forms Implants Oral films Intravaginal Rings	Temperature
	DDM	Pellets/Granules	IR and ER solid dosage forms Implants Intravaginal rings	Temperature
	PAM	Liquid Semisolid paste Polymer suspensions	IR and ER solid dosage forms Complex drug delivery device Implantable patches	Solvent evaporation
Ink Jet Printing	BJT	Powder/Solution	Implantable/subdermal drug IR and ER solid dosage forms Pulsatory devices Resorbable devices	Binder solution
	MJT	Solution/Hydrogel	IR solid dosage forms Mucosal films Orodispersible films	Temperature UV
Vat Polymerization	SLA	Liquid Resin/Hydrogel	IR solid dosage forms API encapsulation Microneedles	UV Digital light
Powder bed fusion	SLS	Powder	Microporous controlled-release drug delivery devices IR and ER solid dosage forms Reservoir-type drug delivery devices	Laser beam

Regardless of the printing technology, a 3D printing process starts with creating a digital model representing the desired geometry, crafted using computer-aided design (CAD) software. The CAD file is transformed into standard tessellation language file format (*.STL), which is then converted onto cross-sectional slices or layers, forming a blueprint for the intended model. A slicer program then translates these slices into a series of precise g-code commands. The 3D printer builds up the object layer-by-layer until completion. In the slicer program, numerous parameters can be adjusted, including layer thickness, infill density, increment angle and printing speed. All these parameters aid in precise control and facilitate the creation of complex structures; but also introduce challenges in terms of quality control, owing to a limited understanding of the impact of the process parameters on the performance of the final drug product. In pharmaceutical applications, these parameters may influence tablet characteristics such as kinetic release, friability, hardness and disintegration time (Ngo et al., 2018; Zidan et al., 2019a). Therefore, a thorough understanding of these parameters becomes imperative to implement a 3D printing manufacturing process.

3. Material extrusion-based 3D printing

Material extrusion-based 3D printing techniques use solid filaments, pellets, granules, semi-solids, or hydrogel as starting building materials. There are three leading extrusion-based printing technologies with applications in pharmaceuticals: fused filament fabrication (FFF), also known as fused deposition modeling, droplet deposition modeling (DDM), and pressure assisted micro-syringe (PAM). FFF was developed in 1998 (Crump, 1989), while PAM and DDM were developed in 2002 (Li et al., 2002; Vozzi et al., 2002). Researchers have shown a growing interest in these material extrusion-based techniques owing to their ability to tune drug release by adjusting the tablet geometry and polymer (Azad et al., 2020; Goyanes et al., 2014).

3.1. Operational principle for material extrusion-based 3D printing

Fig. 1 shows a schematic diagram of PAM technology. This technology requires a semisolid or hydrogel formulation as starting building materials. Semisolid formulations are usually prepared by dissolving suitable ratios of polymer with one or more organic or inorganic solvents (Alayoubi et al., 2022; El Aita et al., 2019; Khaled et al., 2014; Khaled et al., 2015b). PAM involves a syringe with a mobile piston connected to a nozzle. The formulation is then filled into an extrusion barrel with a plunger. To facilitate the extrusion process, pressurized air is applied to the piston, usually between 3 and 5 bar (El Aita et al., 2019; Khaled et al., 2014; Khaled et al., 2015b). This pressurization drives the material through the nozzle onto the building platform, which usually moves in the z-direction to enable the vertical addition of layers of material. A drying step is required after the printing process to ensure the removal of residual solvents and solidification of the 3D-printed tablet. One advantage of PAM is the ability to carry out the printing process at room temperature because the material need not be molten, making PAM particularly advantageous for printing thermolabile APIs.

Fig. 1.

Schematic representation of PAM.

Fig. 2 provides a schematic representation of the FFF technique. The starting building material for FFF is a continuous thermoplastic filament usually produced by hot melt extrusion (HME) (Kempin et al., 2018; Korte and Quodbach, 2018). These filaments possess a suitable range of mechanical properties, flexibilities, and melt viscosities to guarantee a smooth and successful printing process. The FFF process commences by loading the filament into a heated nozzle with a gear system for the melting or softening of the material. The melted filament is extruded through the nozzle and deposited into layers on the building platform, which usually moves in the z direction to allow the stacking of printed layers. After cooling and solidification, the layers bond and fuse with each other, resulting in a 3D-printed tablet (Prasad and Smyth, 2016). FFF is a solvent-free printing technique, eliminating the need for post-fabrication steps. Two approaches are employed in FFF to print objects with different materials and or colors: multi-nozzle and mixer single-nozzle systems (Fenollosa et al., 2019). The multi-nozzle method utilizes dual heads, each supplied with different materials having distinct melting points, which allow the fabrication of multi-compartment capsular devices loaded with different APIs (Jamróz et al., 2018; Kempin et al., 2018; Park et al., 2018). On the other hand, the mixer single-nozzle approach involves the passage of different filaments, mixed or not, through a single nozzle (Fenollosa et al., 2019).

Fig. 2.

Schematic representation of FFF.

A limitation of FFF is the inherent mechanical constraints of most pharmaceutical-grade polymers, which necessitates the addition of plasticizers to the formulation (Ligon et al., 2017). While this plasticization step may be essential for achieving suitable polymer properties, it may introduce challenges by potentially impacting the physicochemical properties of the resulting tablets. Several approaches have been reported to alleviate the need to prepare filaments for feeding into an FFF process (Fanous et al., 2020; Goyanes et al., 2019; Pietrzak et al., 2015; Zhang et al., 2020). One approach DDM, which employs pellets or granules as the starting building material (Hentschel et al., 2020; McDonagh et al., 2022a; Ramezani Dana et al., 2019; Sierra-Vega et al., 2024; Welsh et al., 2019; Zhang et al., 2021).

The schematic representation of DDM technique is shown in Fig. 3. In the DDM process, granules or pellets are fed into a plasticating cylinder, which may be heated by two separate heating zones. Initially, the granules or pellets are melted and processed by a screw mechanism, which doses the molten plastic material against a stagnation pressure. Before the plastic material leaves plasticating cylinder to enter the discharge unit, the system performs a decompression process to close the non-return valve. The discharge unit incorporates a high-frequency piezo-controlled nozzle, which extrudes the material as individual droplets onto a moveable building platform (Hentschel et al., 2020; McDonagh et al., 2022a). This platform may be moved along three axes. DDM may be applied to personalize the drug release profile of poorly water-soluble compounds (Welsh et al., 2019). Compared with FFF, the DDM process does not require the preparation of drug-loaded filaments as an intermediate product.

Fig. 3.

Schematic representation of DDM.

3.2. Critical process parameters for material extrusion-based 3D printing

The quality of 3D-printed tablets manufactured by material extrusion-based technologies may be affected by several process parameters, which need to be optimized for processing certain starting building materials and controlled for ensuring critical characteristics of printed tablets. These process parameters include extrusion temperature, temperature of the building platform, nozzle orifice diameter, printing speed and applied pressure.

Extrusion temperature is a main process parameter, exerting a significant influence on the quality of 3D-printed tablets, impacting API stability, amorphization and polymer viscosity (Tagami et al., 2017; Wang et al., 2020). PAM offers the advantage of operating at room temperature, while FFF requires higher temperatures to melt the polymeric matrix into a flowable state. This high-temperature requirement limits FFF for thermolabile materials and APIs. Precise control over extrusion temperature is needed to ensure that the viscosity of printing materials falls within the desired range of melt flow properties (Azad et al., 2020). The extrusion temperature profile is developed and optimized to process a printing material to have adequate physical and rheological characteristics during the extrusion process and further effective layer deposition on the building platform (Azad et al., 2020; Skowyra et al., 2015). An increase in extrusion temperature commonly decreases the viscosity of the printing material, causing a loss of the solid character and making it more liquid in the printing process. Conversely, operating at a temperature lower than the optimal extrusion range results in heightened viscosity of the material, leading to issues such as poor flow, nozzle blockage by solidifying the polymer, and low fusion bonding strength among deposited layers in FFF (Pietrzak et al., 2015; Yang et al., 2018). The high-viscosity printing material also requires applying elevated pressures in a PAM to achieve the desired flow rate (Diaz-Torres et al., 2022; El Aita et al., 2018; Shafiei et al., 2018). The applied pressure in PAM is directly proportional to the viscosity of the material (Mohammed et al., 2021). Additionally, increasing the extrusion temperature may improve the tensile strength and the elastic modulus of the 3D-printed object by reducing empty spaces in the structure (Pang et al., 2022) and influencing the drug content in the 3D-printed tablets (Tagami et al., 2017). For scenarios where FFF follows an HME continuous process (Pietrzak et al., 2015; Zhang et al., 2020), the printing temperature needs to surpass the HME processing temperature due to differences in heating time between the two processes (Pietrzak et al., 2015). The temperature of the building platform and temperature difference between the extrusion nozzle and the platform constitute other critical factors impacting the quality of the 3D-printed tablets. The temperature of the building platform is usually lower than the extrusion temperature, but the difference between both temperatures should be small enough to guarantee the adherence of the first layer deposited (El Aita et al., 2018; Pietrzak et al., 2015).

Nozzle orifice diameter may also impact the quality of the 3D-printed tablets produced by extrusion-based technologies. Nozzle orifice diameter may affect the extrusion rate of the printing material, resolution, density and tensile strength of deposited filament or droplets (Triyono et al., 2020; Zidan et al., 2019b). Nozzle diameter dictates the width and height of the printed path; therefore, a narrow diameter orifice leads to higher resolution and smoother topography of printed tablets, while a wider orifice leads to lower resolution and more visible layer lines at the tablet surface. Triyono and coworkers used FFF to demonstrate a nonlinear relationship between nozzle diameter and tensile strength and density of the printed objects (Triyono et al., 2020). It is worth noting that the selection of nozzle diameter is also a function of the viscosity of the printing material. Materials of high viscosity require a broader diameter to avoid nozzle clogging and buildup of back pressure in the extrusion vessel.

Printing speed plays a crucial role in defining the resolution and fine details of the printed tablets. Printing speed refers to the movement speed of the printing head. Optimizing the printing speed depends on the desired resolution and the viscosity of the printing material. High printing speed may not be used for high-viscosity materials because maintaining a continuous extruded filament to form a stable bed is challenging. This may lead to defects and irregularities during printing, resulting in significant dose variation and inconsistent performance of printed products. Low printing speed could enhance resolution, sacrificing efficiency or production rate. However, printing a low-viscosity material at a low printing speed could lead to over-extrusion of the material in the 3D-printed tablet (Mohammed et al., 2021). Mohammed and coworkers demonstrated that as the printing speed increases, the line width of a PAM extrusion decreases significantly (Mohammed et al., 2021). The interaction between printing speed and flow properties of printing materials may be considered in the optimization processes of PAM and FFF (Cotabarren and Gallo, 2020; Mohammed et al., 2021). This interaction may lead to several printing defects, such as over- and under-extrusion, when the printing speed is less or greater than the appropriate material flow rate, respectively.

Infill density and pattern may also influence the quality of printed tablets manufactured by material extrusion-based technologies. Infill density is the percent of fullness attained by deposited material in the inner area of a layer, while the infill pattern describes the orientation and arrangement of this material inside the layer (Tanveer et al., 2022). An increase in the infill density may lead to an increase in the weight, volume and overall densification of printed tablets owing to changes in the internal structure (Fina et al., 2020; Goyanes et al., 2014; McDonagh et al., 2022a; Tagami et al., 2017). Infill density decreases the porosity and drug release of the tablets manufactured using DDM (Ebrahimi et al., 2024; McDonagh et al., 2022b; Zhang et al., 2021) and may impact the thermal conductivity of the polymer (Ravoori et al., 2018). Fig. 4(a) and Fig. 4(b) show the impact of infill density on the 3D-printed tablet structure and drug released, respectively, highlighting that increased infill density reduces tablet porosity and decreases the percentage of Tylenol released (Zhang et al., 2021).

Fig. 4.

(a) SEM images of 3D-printed tablets at different infill densities (b) In vitro dissolution data of the 3D-printed tablets with infills from 30, 60 and 100%. Adapted from reference (Zhang et al., 2021) with permission from Elsevier.

The geometry of a printed tablet may also affect its quality characteristics. Goyanes and coworkers used FFF to manufacture five geometric-shaped tablets (cubic, pyramid, cylinder, sphere or torus shapes) that would be challenging to manufacture by traditional compaction methods (Goyanes et al., 2015b). They demonstrated that the surface area/volume ratio was the most significant factor impacting drug release properties based on the geometry employed. Increasing surface area exposes more drugs to the dissolution media, leading to faster disintegration and release. FFF has also been used to print hard capsules and suppositories (Tagami et al., 2020; Yang et al., 2020). The geometry of the capsules and coat thickness dictate their drug release kinetics. Yang and coworkers reported a pulsatile release with a lag time of 4 h when a coating thickness of 0.8 mm was used; however, increasing the thickness from 0.8 mm to 1.66 mm sustained the drug release and increased the lag time to 6 h (Yang et al., 2020). Additionally, Tagami and coworkers printed suppository shells of various thicknesses using polyvinyl alcohol filaments (Tagami et al., 2020). They reported that thickness and internal geometry dictated the hardness of these suppository shells. The dissolution profile also depended on the shell type and its internal structure (Tagami et al., 2020).

Droplet aspect ratio (DAR) and discharge rate are two specific parameters to describe the droplet deposition and their fusion into droplet chains in DDM. The DAR may be estimated as the ratio of width to height of a droplet as it is discharged from an extrusion nozzle and deposited on a printing platform. A lower DAR leads to the deposition of droplets too close to each other, resulting in an overfilled 3D-printed object with a wavy appearance (McDonagh et al., 2022a). However, a higher DAR results in a spaced deposition of droplets into an underfilled part with small gaps between droplet chains (McDonagh et al., 2022a). The DAR has also been shown to affect the dimensional accuracy of the 3D-printed object (Hentschel et al., 2020). The DAR can affect the mechanical properties of printed tablets. For example, it was revealed that the mechanical properties are higher when the DAR is reduced because of the increased degree of filling in the tablet (Eisele et al., 2023; Hentschel et al., 2023).

The droplet discharge rate describes the volume of material deposited in each drop through the nozzle (Sierra-Vega et al., 2024). The droplet discharge rate is modified to achieve the volume necessary for layer thickness. Higher droplet discharge rates may affect the dimensional accuracy of the 3D-printed object and may cause errors in the printing process (Hentschel et al., 2020). Charlon and Soulestin demonstrated that increasing the discharge rate improved the mechanical properties of the printed objects (Charlon and Soulestin, 2020). The authors justify this by a higher degree of filling. A reduction in the droplet discharge rate results in a longer residence time of the material in the discharge unit, which increases the building time. A balance between DAR, droplet discharge, and infill density is necessary for high-quality droplet deposition and accurate printing patterns.

3.3. Critical material attributes for material extrusion-based 3D printing

3.3.1. Critical material attributes for FFF and DDM

The starting building material commonly used in FFF is a thermoplastic filament composed of a single polymer or a polymer blend. This filament needs to possess adequate rheological and mechanical characteristics to ensure suitable processability. The API may be loaded to the filament matrix by soaking the polymer in an organic aqueous solution or integrated with the polymer during the HME process (Goyanes et al., 2014; Park, 2015; Pietrzak et al., 2015). The soaking polymer method may be employed for low API loading (<2%) due to the limited drug loading capacity of commercial filaments (El Aita et al., 2018; Goyanes et al., 2015a; Skowyra et al., 2015). It has been shown that the soaking process does not significantly impact the mechanical properties and physical appearance of commercial polymeric filaments (Goyanes et al., 2014; Goyanes et al., 2016). The starting building materials for DDM are pellets or granules loaded with API. The size of pellets for DDM usually ranges from 2 to 3.2 mm in diameter (McDonagh et al., 2022a). Pellets and granules may be manufactured using various well-established methods in the pharmaceutical field, including hot melt extrusion, high-shear wet granulation, and twin-screw wet granulation (McDonagh et al., 2022a). DDM may be applicable not only to FFF processable material but also to other materials that are brittle of high melt viscosities and tend to break up or clog FFF extrusion heads (McDonagh et al., 2022a, b).

The most widespread method for API loading to building materials of FFF is HME, which mixes the polymer with API and other excipients using a single and twin-screw extrusion, allowing dispersion of poorly soluble APIs in polymer matrices and continuous manufacturing of filaments (Tan et al., 2018; Verstraete et al., 2018). HME is also suitable for the loading of high drug concentrations, ensuring proper mixing and enhancing drug solubility within the polymer matrix (Rahman et al., 2018). Nevertheless, a plasticizer may be added to the formulation for high API concentration to soften the filament and provide suitable printing flexibility if the API cannot serve this role in the formulation (Aho et al., 2015; Goyanes et al., 2015b). The HME process may be unsuitable for thermolabile material due to aggressive heating and mixing that may impact the stability and crystallinity of the API (Melocchi et al., 2021; Pietrzak et al., 2015). Consequently, continuous evaluation of filament quality and API stability using various analytical tools, such as stability indicating chromatographic methods, Fourier Transform Infrared, differential scanning calorimetry, and X-ray diffraction (XRD), may be necessary to assess throughout the HME process.

Common standard polymeric materials for FFF processing include polylactic acid, polyvinyl alcohol, polyvinyl-pyrrolidine (Arafat et al., 2018a; Arafat et al., 2018b; Sadia et al., 2016; Varghese et al., 2022). These polymers are commercially available as filaments of diameters ranging from 1.75 to 3 mm (Andronov et al., 2023; Bandari et al., 2021). For pharmaceutical applications, polylactic acid and polyvinyl alcohol are frequently employed for sustained and immediate drug release, respectively. Other extrudable polymers have also been reported in the literature, including Eudragit RS (methacrylic acid and ethyl acrylate copolymer), Eudragit RL, Eudrgait RL 100, Eudrgait E, ethyl-cellulose, polycaprolactone, hydroxypropyl cellulose, hydroxypropyl methylcellulose, poly lactide-co-glycolide and Soluplus (Rahman et al., 2018). These polymers differ in their hydrophilicity, mechanical strength, biocompatibility and crystallinity, affecting their processability and release characteristics. Plasticizers are usually used with ethyl-cellulose to be extruded, forming sustained-release matrices (Borujeni et al., 2020; Goyanes et al., 2015a; Goyanes et al., 2015b; Zhang et al., 2017a; Zhang et al., 2017b). Eudragit are nonbiodegradable and amorphous polymethacrylate-based polymers that exhibit high thermal stability and melt miscibility with many APIs and achieve customizable controlled drug release profiles by changing their functional group (Senarat et al., 2022). Eudragit E PO is a well-known polymer that is not FFF printable without additives, but it may be processed in DDM (McDonagh et al., 2023). Soluplus has an amphiphilic chemical structure with low hygroscopicity and low glass transition temperature, allowing extrusion in FFF and DDM (Attia et al., 2023; McDonagh et al., 2022b, 2023).

To select a polymeric candidate printing material for FFF or DDM processing, a detailed characterization of its mechanical and rheological properties is required to meet the targeted quality and printing requirements (Arrigo and Frache, 2022; Azad et al., 2020). For instance, filaments that are too brittle or soft cannot be extruded as they may break or be squeezed by the extruder head gears (Jamróz et al., 2018). The rheological properties of the polymer may be assessed to determine its extrudability and extrusion temperature range (Mishra et al., 2022). At lower Reynold (Re) number flow through the nozzle, the polymer tends to have a Newtonian flow, but an increase in the Re number leads to non-Newtonian flow behavior (Mendes et al., 2019). Likewise, rheological properties may provide useful information that may be correlated with material performance attributes during manufacturing and provide insight into the properties of the final product. Therefore, the viscosity of the material is considered the most critical rheological parameter for defining the optimal printing conditions in FFF and DDM (Arrigo and Frache, 2022; Azad et al., 2020).

The viscosity represents the resistance of the material to flow and is given as a stress and strain rate ratio. Most filaments are non-Newtonian fluids; their viscosity depends on the strain rate, causing the long molecules to orient along the flow direction when the applied strain is sufficient. Different models have been proposed to estimate the viscosity of filaments (Cox and Merz, 1958; Roland, 2013). The Cross-WLF and Cox-Merz models have been used for the FFF process (Mendes et al., 2019; Mishra et al., 2022). The cross-WLF model applies to a broader range of shear rates and considers the dependency of the temperature of the shear rate and viscosity. The model determines the melt viscosity (η) using Equation (1), where η_o, represents the zero-shear viscosity, $\dot{\gamma }$ is the strain rate, τ* is the critical stress level at the transition to shear thinning, and n is the power law index in the high shear rate regime (Mishra et al., 2022).

$$\eta =\frac{{\eta }_o}{1+{\left(\frac{{\eta }_0\dot{\gamma }}{{\tau }^{\text{*}}}\right)}^{1-n}}$$ (1)

The empirical Cox-Merz model established that the magnitude of the complex dynamic viscosity (η*) at frequency (ω) is comparable to the magnitude of the shear viscosity (η*) at strain rate $\left(\dot{\gamma }\right)$ (Cox and Merz, 1958). This model is described by Equation (2).

$$\eta (\dot{\gamma })={\left|{\eta }^{*}\left(\omega \right)\right|}_{\omega =\dot{\gamma }}$$ (2)

The strain rate applied in the nozzle may be estimated by the following equation using the nozzle diameter (D_c) and the volumetric flow rate (Q):

$$\dot{\gamma }=\frac{4Q}{\pi D_c^3}$$ (3)

Determining the optimal viscosity range is essential for predicting the extrudability of a melt formulation. Capillary rheometers are effective tools to estimate complex viscosity within a high shear rate regime, typically between 1 and 10000 s⁻¹, while rotational viscometers are suitable for a range between 0.01 and 100 s⁻¹. Fig. 5(a) shows a plot of the typical conditions for shear viscosity or complex viscosity during flow through the printed nozzle and bonding during deposition (Acierno and Patti, 2023). The plot in Fig. 5(a) illustrates the specific viscosity values required to ensure successful extrusion through the nozzle (shear rate in the range of 30 – 500 s⁻¹) and the deposition process (shear rate in the range of 0.01 – 0.1 s⁻¹).

Fig. 5.

(a) Typical shear or complex viscosity and (b) Storage modulus and loss modulus conditions for extrusion through the printer nozzle and bond consolidation during deposition.

Adapted from reference (Acierno and Patti, 2023).

Storage modulus (G’) and loss modulus (G”) are also used to describe the viscoelastic properties of the polymeric filaments processed in FFF. G’ represents the elastic component, indicating the solid-state behavior of the polymer, while G” characterizes the viscous component, representing energy dissipation. Crossover point where G’=G” provides information regarding the solid or viscous behavior of the material at different temperatures and its ability to resist oscillation stress. Fig. 5(b) shows a plot of the typical conditions of the G’ and G” for bond consolidation and flow in the printer nozzle (Acierno and Patti, 2023). Conditions where G” exceeds G’ are needed to ensure proper extrusion through the nozzle (angular frequency between 50 – 150 rad/s), and conditions where G’ surpasses G” are required to establish interlayer bonding and maintain the geometric shape in the 3D-printed objects (angular frequency < 0.1 rad/s). However, G’>G” conditions during extrusion may lead to nozzle clogging (Aho et al., 2019).

3.3.2. Critical material attributes for PAM

The starting building materials for PAM are mostly a semi-solid paste or hydrogel composed of a combination of API, polymers, excipients and solvents. Semi-solid formulations are prepared by mixing suitable ratios of polymer with water or organic/ inorganic solvents (El Aita et al., 2019; Khaled et al., 2014; Khaled et al., 2015a). The API may be loaded onto the gel via entrapment or diffusion. Diffusion depends on the porosity of the hydrogel and hence is more suitable for small molecule drugs. Factors such as API size-hydrogel pore size ratio, spatial API distribution in the hydrogel, and drug loading method may impact the release kinetics attained (Lepowsky and Tasoglu, 2018).

For processing in PAM, materials with a minimum viscosity are usually used to ensure smooth flow through the micro-syringe, enabling precise deposition and layering on the printing platform (Yang et al., 2020). High-viscosity materials may be processed with high printing pressure, increasing the risk of nozzle clogging, while low-viscosity materials may compromise the structure of the 3D-printed tablet (El Aita et al., 2019). The specific viscosity requirement may vary based on different factors, including the composition of the paste, nozzle size, pressure exerted, and desired printing resolution. Generally, pastes with a viscosity range of 10² – 10⁴ cP may be processable in PAM to facilitate layering during printing.

Processable polymers in PAM applications for solid dosage forms include, but are not limited to carbopol, hydroxypropyl cellulose, hydroxypropyl methylcellulose, polyvinyl pyrrolidone and polyethylene glycol (Alayoubi et al., 2022; Azad et al., 2020; Conceição et al., 2019; El Aita et al., 2020; Khaled et al., 2018; Khaled et al., 2014; Khaled et al., 2015a, b; Zidan et al., 2019a; Zidan et al., 2019b). The use of natural and synthetic hydrogels is also common for drug delivery system using PAM. Commonly used hydrogel matrices of alginate, gelatin, agarose, fibrin, and chitosan have been successfully extruded in PAM and investigated for 3D printing of various drug products (Lepowsky and Tasoglu, 2018). Physicochemical properties of the polymers, including viscosity, molecular weight, crosslinking density and glass temperature, are critical for a successful printing process. For instance, the molecular weight and polydispersity of the polymer may impact printability due to unanticipated changes in flow behavior. High polydispersity increases the range of glass transition and melting endotherm peaks and transition areas from the Newtonian plateau to shear thinning (Azad et al., 2020). Lactose and microcrystalline cellulose have also been used as excipients in various studies to control dissolution, grittiness, and the pressure required for extrusion (Zidan et al., 2019a). Microcrystalline cellulose has also been used as a paste modifier to increase viscosity and amplify shear thinning for smooth extrusion and rapid deposition (Yang et al., 2020).

Physicochemical characteristics of excipients and API may affect the extrudability behavior of the semi-solid in PAM. A recent study examined the formulation attributes impacting the release profiles of atorvastatin and metoprolol from polypills (Alayoubi et al., 2022). The results highlight the critical role of the hydroxypropyl methylcellulose matrix in achieving the immediate release of atorvastatin, while the core–shell design is crucial for the extended release of metoprolol (Alayoubi et al., 2022). The nature of the solvent and its evaporation rate are the critical attributes that affect dimensional accuracy (Algahtani et al., 2020; Mohammed et al., 2021). The presence of solvents exerts some challenges due to the post-drying step that may lead to shrinkage, deformation and associated inaccuracy in the dimension of the printed tablets. Common solvents such as acetone and ethanol may improve the fluidity and deposition of the paste; however, the solvent evaporation rate during printing may impact consistency. A solvent of high vapor pressure may cause fast drying of the paste and nozzle blockage, while a low vapor pressure may slow down the solidification of the materials at the nozzle tip, as well as the collapse of the solid during the printing process (Algahtani et al., 2020; Mohammed et al., 2021). Water has been used as a continuous phase of pastes for PAM processing; however, optimizing the water content of the paste is crucial due to the hygroscopicity of the polymers and excipients used, which may lead to swelling and clogging.

Computational fluid dynamic simulations have been used to elucidate correlations between the viscosity of the starting material and nozzle diameter, as well as to predict the required extrusion pressure and movement speed of the printing head for material deposition (Yang et al., 2020). This simulation demonstrated that increasing the viscosity of paste greatly increased the maximum total pressure at the inlet of the extrusion barrel while slightly decreasing the maximum velocity at the outlet of the printing nozzle. Mechanistic rheology-based extrudability models have also been generated to predict flow behavior and provide insight into the impact of excipients on the rheological behavior of the semi-solid (Zidan et al., 2019b). This study found that creep recovery, cross-over modulus and extrudability models were valuable to understanding the effects of excipients on the rheological behavior of semi-solid during 3D printing. However, extrudability models generated by texture analyzers were more accurate than oscillating rheometer models in predicting the printing pressures of semi-solid materials, as the extrudability model accounts for the nozzle diameter factor (Zidan et al., 2019b).

4. Inkjet printing: binder and material jetting

Inkjet printing is a 3D technology that includes two distinct technologies, namely binder jetting (BJT) and material jetting (MJT). The fundamental process in 3D inkjet printing involves a precise formation of material droplets to be deposited onto a substrate under digital control. The droplets are subsequently solidified and culminated in the final product. BJT and MJT were developed in the early 1990 s (Piłczyáska, 2022; Sachs et al., 1993). 3D printing processes based on inkjet technology have promising potential to manufacture complex, miniaturized multifunctional systems with high-resolution geometric patterns (Kyobula et al., 2017). Spritam^™ (levetiracetam) is the first FDA-approved product manufactured using BJT 3D (Aprecia, 2015).

4.1. Operational principle for inkjet printing

Fig. 6 presents a schematic illustration of the BJT process. In BJT, a spreader (roller) places a thin powder layer on the bed while the printhead nozzle sprays a binder solution along the x- and y-direction in selected areas based on the 3D CAD model requirements. The precise discharge of binder droplets initiates the fusion of powder particles to form the first layer of an object being printed. The build platform incrementally descends to build more layers while the powder supply ascends in the z-direction, and the powder spread, binder jetting, deposition and fusion processes are repeated. This layer-by-layer deposition approach continues until the desired object is fully formed. The unbound powder particles that provide structural support during the printing process are then removed, often followed by post-processing steps such as drying and de-dusting (El Aita et al., 2018; Lamichhane et al., 2019). There are two common approaches for API loading based on the strength required (Curti et al., 2020). For example, API is generally added to the binder solution for 3D printing of tablets of potent drugs. For tablets of high API loading, the API is typically blended with the excipients in the powder blends.

Fig. 6.

Schematic representation of BJT.

The atomization mechanism of binder droplets varies based on the type of inkjet printing head, which is classified as continuous jet (CJ) or drop-on-demand (DOD). CJ printing utilizes a pressurized flow to produce a continuous stream of droplets discharged from the nozzle, even when no printing is required (Mostafaei et al., 2021a; Prasad and Smyth, 2016). CJ systems operate with drop generation rates in the 20 – 60 kHz region, and drop velocity at the nozzle is typically > 10 ms⁻¹ (Derby, 2010). DOD uses thermal or piezoelectric print heads to produce droplet volumes as low as 1–100 pL only when a drop is required, which may be more efficient and, thus, more frequently used (Prasad and Smyth, 2016; Varghese et al., 2022).

The schematic representation of MJT technique is presented in Fig. 7. In the MJT process, build and support materials droplets are selectively sprayed onto the build platform. An ultraviolet (UV) light of specific wavelengths of 190 and 400 nm is directed onto the molten material on the build platform for curing purposes (Gulcan et al., 2021). The print head moves along the x- and y-direction during the printing process, and the build platform is lowered along the z-direction by the height of each deposited layer of the material. MJT may require the inclusion of a support structure for the 3D-printed tablets since liquid or molten material is used as the starting building material. The build and support structure are printed simultaneously during the same printing task. A support structure of dissolvable material is usually employed, which can be removed afterward using bath sonication in sodium hydroxide solution, heating, or a high-pressure water jet (O’Neill et al., 2017). MJT process may create both matte and glossy surface finish options. A matte surface is obtained when the support material covers the whole printed tablet. In contrast, a glossy surface is generated when only structurally needed areas are supported and the model is exposed to air during the curing stage. MJT can be operated using either continuous jetting or droplet-on-demand jetting.

Fig. 7.

Schematic illustration of MJT.

4.2. Critical process parameters for inkjet 3D printing technology

The printing parameters, such as layer thickness, roller/spreader speed, printing speed, binder saturation, drying time and print orientation, may be optimized based on the mechanical design of printing heads, mechanism of droplet ejection, and jetting characteristics to produce consistent 3D-printed tablets (Elkasabgy et al., 2020; Kotta et al., 2018; Mostafaei et al., 2021a). These process parameters can influence the physical, chemical, and mechanical properties of the resulting tablets (Jamróz et al., 2018).

For a BJT process, the layer thickness is determined by the height of the powder bed along the z-axis during the printing process. Generally, layer thickness should be greater than the maximum particle size and may vary from 15 to 300 μm (Mostafaei et al., 2021a). For thin powder layers, binder droplets may quickly reach the previous wet layer and affect the stability of the printed tablets. Decreased layer thickness increases powder density and the overall mechanical strength of the tablets when binder saturation is fixed (Enneti and Prough, 2019; Mostafaei et al., 2021a). Layer thickness is also used to estimate the binder saturation, which affects the dimensional accuracy of the printed objects (Vaezi and Chua, 2011). For MJT, layer thickness represents droplet height and is usually between 16–32 μm (Sireesha et al., 2018). In MJT, the layer thickness is primarily determined by the physiochemical characteristics of starting materials due to the complex physics of droplet formation. Some studies have demonstrated that layer thickness may significantly impact the dimensional accuracy and build time of printed objects (Kechagias et al., 2014; Pugalendhi et al., 2020b). However, research indicates that layer thickness has a minimal effect on the surface roughness of printed objects (Aslani et al., 2019).

Roller/spreader speed is another critical process parameter to be controlled in BJT. This parameter represents the speed of spreading the powder on top of the pre-printing layer. Usually, this speed may vary between 0.1 to 16 mm/s. Slower roller speeds may enhance the resolution and dimensional accuracy of printed tablets while increasing the printing time (Shrestha and Manogharan, 2017). Conversely, fast roller speeds reduce the contact time between the roller and powder bed, thereby minimizing compression force on the powder bed. A recent study demonstrated that an increase in roller speed from 2 to 6 mm/s slightly decreased accuracy in the z-direction but improved it in the x-direction (Miyanaji et al., 2016b).

Printing speed is a critical process parameter for inkjet 3D printing technologies. Printing speed refers to the rate at which printheads navigate across the print bed. Fast printing speed may lead to insufficient evaporation of the liquid material from the pre-printed layer before the subsequent layer is printed, resulting in lower dimensional accuracy, higher surface roughness, and, therefore, an alteration of equilibrium saturation (Miyanaji et al., 2018a; Miyanaji and Yang, 2016; Myers et al., 2021; Sen et al., 2021a). Miyanaji and coworkers elucidated that increasing the BJT printing speed from 100 to 700 mm/s decreased the saturation level from ~ 85 to ~65 % due to a reduction of the x-y dimensional accuracy (Miyanaji et al., 2018a). The printing speed may also affect the mechanical properties of the printed objects and the reproducibility of the printing process (Miyanaji et al., 2018a; Myers et al., 2021). An increase in BJT printing speed may cause a decrease in the mechanical strength of the tablets due to the enhanced inertial forces (Miyanaji et al., 2018a; Miyanaji et al., 2016b). In contrast, MJT enables printing at high-speed or high-quality mode (Gulcan et al., 2021). The high-speed printing mode may produce tablets with better mechanical properties in a shorter time than those produced under the high-quality printing mode (Gulcan et al., 2021; Pugalendhi et al., 2020a).

Binder saturation impact also dimensional accuracy and properties of printed tablets in BJT. Binder saturation measures the volume of binder used for printing, and saturation is the ratio of the volume occupied by the binder to the volume available in the powder (Enneti and Prough, 2019). At low binder saturation, powder particles may not aggregate, leading to layer delamination and the formation of voids and pores in printed tablets (Chen and Zhao, 2016; Mostafaei et al., 2021a; Schmutzler et al., 2019). Higher binder saturation may lead to the bonding of excessive powder particles in a printed bed, resulting in higher surface roughness and dimensional inaccuracies (Chen and Zhao, 2016; Shrestha and Manogharan, 2017; Vaezi and Chua, 2011). Furthermore, higher binder saturation may lead to over-wetting, causing the particles to stick to the roller and subsequently affect powder bed homogeneity with cracks, roughness, and even shifts within the powder bed (Chen and Zhao, 2016). Kafara and coworkers have shown that higher binder saturations leads to inaccuracies in the dimensions of the printed objects compared to the CAD design (Kafara et al., 2018). Binder saturation may also affect the mechanical properties of the 3D-printed object (Miyanaji et al., 2016a; Miyanaji et al., 2016b; Mostafaei et al., 2021b; Shrestha and Manogharan, 2017). It has been reported that an increase in binder saturation from 50 to 75 % increased the mechanical strength of the object in BJT by approximately 50 % (Miyanaji et al., 2016a). Mostafaei and coworkers also reported that the increase in binder saturation with an adequate drying time reduces porosity and increases the density of the 3D-printed objects (Mostafaei et al., 2021b).

Optimization of the binder saturation parameter may be done experimentally or using mechanistic modeling. Miyanaji and coworkers proposed a mechanistic model to predict the binder saturation for the BJT (Miyanaji et al., 2018b). This study concluded that the degree of binder saturation attained depends on the physical and chemical interactions between binder droplet and powder bed, including spreading and penetration. The following equation provides an estimate of binder saturation considering the volume of binder per drop [pL], V_binder, packing rate percentage (PR), spacing between droplets (μm), X and Y, and layer thickness (μm), Z.

$$S=\frac{V_{\mathit{\text{binder}}}}{(1-PR{)}^{*}{X}^{*}{Y}^{*}Z}$$ (4)

Drying time in a BJT process represents the time needed for curing (removing the excess of binder) the surface of the powder bed after spraying and saturation with binder solution. Drying time depends mainly on the extent of binder saturation, layer thickness, and physical and chemical properties of the powder and binder. Some binders need a brief drying time, while others need over 15 s per layer (Mostafaei et al., 2021a). Optimization of binder saturation, drying time, and their interaction is critical to ensure the quality of printed tablets. Mostafaei and coworkers studied the interaction effects of binder saturation and drying time on microstructure and properties of 3D-printed coupons from tungsten carbide-cobalt (Mostafaei et al., 2021b). Fig. 8 summarizes this interaction, changing the drying time from top to bottom and the binder saturation from left to right (Mostafaei et al., 2021b). Lower binder saturation needs a shorter drying time for efficient binding between layers and achieving high density. Long drying time using a lower binder saturation leads to an increased porosity in the previous layer at the interface. Higher binder saturation needs a longer drying time to print tablets of high density and low porosity. On the other hand, short drying time keeps the binder solution between layers and results in poor binding of particles of a new layer because of high powder porosity.

Fig. 8.

Schematics show the effect of drying time and binder saturation on void and porosity formation. The blue areas indicate newly, undried binder in layer 2, while the yellow areas indicate dried binder in layer 1. Adapted from reference (Mostafaei et al., 2021b) with permission from Elsevier.

Print orientation of tablets in a powder bed is another critical parameter that may impact the outcomes of a BJT process. Print orientation of tablets refers to their alignment with respect to the x-, y-, and z-axes of the build platform and the powder stacking direction by the roller (Mostafaei et al., 2021a). This orientation may affect the porosity, mechanical properties, and surface roughness of printed objects (Myers et al., 2021; Shanjani et al., 2011; Zhang et al., 2009). The influence of the orientation pattern on the surface roughness and green density of the BJT printed object was investigated (Myers et al., 2021). This investigation concluded an increase in surface roughness and a decrease in green density by increasing the print angle from 0° to 45° in relationship to the z-direction. Other studies have also shown that y-orientated printed objects presented lower porosity and better mechanical properties than z-orientated printed objects due to a higher number of layers to be stacked up in comparison to either x- or y-orientated samples (Zhang et al., 2009). Similarly, the alignment may affect the mechanical properties and surface roughness of the printed tablets in the MJT process due to the simultaneous over-curing of some parts compared to others. Several studies have demonstrated that dimensional accuracy in the z-axis is lower than in the x- and y-axes (Pilipovic et al., 2020), and that x-orientated printed objects presented better results in terms of stiffness (Cazon et al., 2014). The effect of build orientation on the mechanical properties of objects printed in MJT was also investigated. Kesy and Kotlinski reported that objects printed in the yz-orientation have the highest mechanical properties than those printed in the xz-orientation (Kesy and Kotlinski, 2010). The authors attributed these variations in mechanical properties to the heterogeneous absorption of light energy by the photopolymer material during the jetting process. Build orientation may also affect the yield strength and breaking stress of printed objects in MJT (Tomar et al., 2019). Higher values of these properties were obtained when the printed objects were manufactured parallel to the xy plane.

4.3. Critical materials attributes for inkjet 3D printing technology

The effect of powder blend properties such as particle size distribution (PSD), flowability, and packing density on the BJT process performance has been reported (Averardi et al., 2020; Bai et al., 2017; Lepowsky and Tasoglu, 2018; Lu et al., 2009; Mostafaei et al., 2019; Sen et al., 2021a; Ziegelmeier et al., 2015). PSD of the powder bed may affect the properties of the printed tablets. Several reports demonstrated that changing PSD from Gaussian to bimodal improves the packing density and flowability of powders, and the mechanical properties of printed objects (Bai et al., 2017; Lu et al., 2009; Sen et al., 2021a). The typical mean particle size range for a BJT process ranges between 0.1 and 150 μm (Mostafaei et al., 2019). The impact of the particle size on the mechanical properties of tablets printed in BJT is not yet fully understood. Mostafaei and coworkers concluded that fine particles (16 – 25 μm) produce printed objects with lower mechanical strength than those produced from large particles (Mostafaei et al., 2019), while Lu and coworkers concluded that fine particles (< 20 μm) produced objects with higher mechanical strength (Lu et al., 2009). The packing density of the powder bed is affected by its PSD, morphology, and flowability characteristics (Averardi et al., 2020). Powder blends with a bimodal PSD usually exhibit higher packing density, resulting in lower porosity and higher mechanical strength for printed tablets (Ziegelmeier et al., 2015). Powder flowability is another attribute affecting the quality of printed tablets in BJT. Powder flowability describes the ability of the powder to spread quickly and precisely to form consistent layers of uniform thickness that can correlated to PSD parameters (Lepowsky and Tasoglu, 2018). Powder with poor flowability reduces resolution and dimensional accuracy and results in high variation in content uniformity due to particle agglomeration (Mostafaei et al., 2021a; Sen et al., 2021a).

Material attributes of binder solutions may also significantly affect the quality attributes of printed tablets in BJT and MJT processes. The selection of the appropriate solvent and solute for a binder solution depends on the physicochemical characteristics of the binder solution and pharmaceutical formulation requirements. Various studies reported the use of ethanol, chloroform, acetone, water, and some organic buffers as solvents and PVP, ethyl cellulose (EC), and poly-L-lactide (PLLA) as solutes (El Aita et al., 2018; Kozakiewicz-Latala et al., 2022). Chang and coworkers have used Kollidon as a binder to produce tablets with high breaking strength (Chang et al., 2020). Tian and coworkers have also demonstrated that filling excipients with high water solubility, binding agents with high viscosity, and moistening agents with high water content may increase bonding strength and hardness (Tian et al., 2019).

Viscosity, density, and surface tension of the binder solution affect the efficiency of the jetting process, the shape of the droplet formed and the spreading pattern in BJT and MJT. Drop formation occurs via jet breakup behavior of binder solution. Drop formation is then governed by dimensionless parameters such as the Reynolds number (Re) and Weber number (We). The Re number denotes the ratio between inertial and viscous forces, while the We number denotes the ratio between inertial and surface forces. However, to predict the binder solution behavior, the Ohnesorge (Oh) number and Z parameter may be determined (Elkaseer et al., 2022; Kyobula et al., 2017; Wickstrom et al., 2017). The Oh number is a dimensionless parameter correlating the viscous forces to the inertial and surface tension forces between the binder solution and particle surface. The following equations may be used to calculate these dimensionless numbers, where ν, ρ, α, μandγ are the velocity, density, droplet diameter, dynamic viscosity, and surface tension of the ink, respectively.

$$Re=\frac{\rho \nu \alpha }{\mu }$$ (5)

$$We=\frac{{\nu }^2\rho \alpha }{\gamma }$$ (6)

$$Oh=\frac{\sqrt{We}}{Re}$$ (7)

$$Z=\frac{1}{Oh}$$ (8)

Studies with Newtonian fluids revealed a recommended range for Z values between 1 and 10, and viscosities between 10 and 100 mPa s, based on the results from numerical simulations of drop formation (Reis and Derby, 2000). Materials with low viscosity result in Z ≥ 10, leading to satellite drops and low resolution. Conversely, high viscosity may lead to nozzle clogging and reduce the flowability of the material. A map was created with coordinates Re and We numbers, which helps define fluid properties suitable for inkjet systems (Derby, 2010; Elkaseer et al., 2022). Additionally, Surface tension also dictates droplet formation and may be optimized within a range of ~ 30 – 70 Mn/m to prevent dripping without affecting the spreading of the binder solution. For pharmaceutical applications, binder solutions containing non-Newtonian polymers, APIs, and other excipients complicate flow behavior; as a result, viscosity may deviate from these recommended ranges.

Incorporation of the API into the binder solution may affect its jetting behavior and quality attributes of printed tablets, such as hardness and disintegration. In a recent study, the incorporation of amitriptyline hydrochloride, as API, at high concentrations increased viscosity and decreased surface tension of the binder solution (Sen et al., 2020). Ideally, drug loading concentrations may be fine-tuned to modify critical attributes of the binder solution and to further optimize the outcomes of the printing process. Recent Studies have shown that Oh values within the range of 0.1 < Oh < 0.14 may be used for binder solutions in BJT and MJT (Sen et al., 2020).

The physical and chemical interactions between the powder bed and binder may also affect dimensional accuracy, mechanical properties, and the final surface roughness of the printed tablets in BJT (Bai et al., 2018; Mostafaei et al., 2021a). Polymorphic transformations and transitions from amorphous to crystalline states have been observed in some cases of these interactions, depending on the critical characteristics of the binder solution and that of powder particle size and morphology (Jacob et al., 2020). Sen and coworkers developed a drop test as a screening method to understand powder-binder interaction, representing this interaction by two parameters, namely binding capacity and binding index (Sen et al., 2021b). These screening tests implemented in different studies may aid in selecting suitable binder solutions for a specific powder formulation and in improving quality controls over the printing process. Similarly, critical material attributes of powder blends (hydrophilic vs. hydrophobic or smooth surface vs porous) may impact the spreading of the binder solution and the overall resolution attained in the MJT process. Additionally, properties such as gravitational forces, capillary forces, contact angle and surface energy should be considered to understand these interactions (Kim et al., 2018).

5. Powder bed fusion: selective laser sintering (SLS)

SLS falls under the category of powder bed fusion techniques, where a laser is employed to sinter or melt a powder bed, metal, or polymer. This process results in the fusion of powder particles and further solidification of build starting material according to a predesigned geometric pattern (Wang et al., 2017). SLS was first developed by Joseph J. Beaman and Carl R. Deckard in 1984 at the University of Texas and patented in 1990 (Beaman and Deckard, 1990). SLS was primarily based on a neodymium-doped yttrium aluminum garnet (Nd: YAG) laser with a power output of 100 W, paired with acrylonitrile butadiene styrene powder as starting building material (Tabriz et al., 2023). SLS is a fast technique that eliminates the need to use filaments or binder solutions. Likewise, SLS produces printed tablets directly from the printing process, eliminating the need for post-processing steps (Charoo et al., 2020; Varghese et al., 2022). SLS processes can also be adapted to produce large batch sizes of tablets, with a capacity to produce about 100 tablets per printing run (Tikhomirov et al., 2023).

5.1. Operational principle for SLS

SLS is an advanced rapid prototyping process that utilizes an external thermal energy source to fuse specific regions within a powder bed. Fig. 9 provides a schematic representation of the SLS process. An SLS system is primarily composed of a printing chamber with a build platform, a powder reservoir chamber, a powder roller, and a laser source. The SLS process starts with elevating the temperature of the powder bed to temperatures just below the melting points or the glass transition temperatures of the polymers employed. Subsequently, a high-power x-y axis laser beam scans above the powder to increase its temperature rapidly above the melting point. This thermal energy application induces the fusion of the particles to solidify according to the scanning pattern performed per the geometric design loaded into the printer. After sintering the first layer, the SLS process continues by lowering the building platform and introducing a fresh layer of the powder. This sequential layering continues until the tablet is formed in the powder bed. The un-sintered powder is removed from the final product for testing and recycling steps, as needed (Fina et al., 2018b).

Fig. 9.

Schematic representation of SLS.

5.2. Critical process parameters for SLS

The main critical parameters for SLS include layer thickness, geometry, powder bed temperature, and laser energy (Awad et al., 2020; Barakh Ali et al., 2019; Trenfield et al., 2018b). These parameters are optimized based on the physical and chemical properties of the materials used for manufacturing tablets.

Layer thickness and geometrical design of tablets affect their quality attributes. Layer thickness in SLS usually ranges between 100 and 500 μm (Charoo et al., 2020). Thinner layers result in higher printing resolution, while thicker layers lead to rougher surfaces with low resolution (Awad et al., 2020; Gibson and Shi, 1997). Layer thickness is set to be less than sintered thickness and higher than the average particle size of the powder to ensure layer integrity and dimensional accuracy, respectively (Awad et al., 2020; Yang et al., 2021). Higher layer thickness requires higher laser energy to sinter the particles along the longer z-axis of each layer and to avoid failures due to delamination (Charoo et al., 2020; Tabriz et al., 2023). Vande Ryse and coworkers demonstrated that layer thickness may impact the morphological, physical, and mechanical properties of printed objects in SLS (Vande Ryse et al., 2022). Increasing layer thickness may decrease the tensile and flexural moduli of printed objects (Caulfield et al., 2007; Vande Ryse et al., 2022). On the other hand, Kulinowski and coworkers concluded that the geometry design parameters of tablets are critical in achieving the dissolution characteristics desired for certain applications (Kulinowski et al., 2021). They found that decreasing the surface area to volume (S/V) ratio from 2.1 to 1.3 delayed the dissolution of acetaminophen from 0.75 to 2 h (Kulinowski et al., 2021).

Controlling powder bed temperature in the building platform is essential for optimizing an SLS process. Powder bed temperature is usually set slightly below the melting temperature of crystalline polymers or slightly above the glass transition temperature of amorphous polymers (Awad et al., 2020). In the case of a polymer blend, the bed temperature to be set may be calculated using the following equation, where x₁ and x₂ represent the weight fraction of each polymer and T_1g and T_2g correspond to the glass transition temperature of each polymer (Awad et al., 2020; Gibson and Shi, 1997).

$$\frac{1}{T_{bed}}=\frac{x_1}{T_{1g}}+\frac{x_2}{T_{2g}}$$ (9)

A pre-heated powder bed decreases the incident energy required for sintering (Awad et al., 2020). Applying high temperature to a powder bed improves the sintering action and produces tablets of high density and hardness (Singh et al., 2017; Tontowi and Childs, 2001). However, thermal degradation of incorporated polymer(s) may occur under these conditions, which may lead to the phase transformation of API from an amorphous to a crystalline state and a decrease in the density and hardness of printed tablets (Fina et al., 2017; Ho et al., 1999; Singh et al., 2017). Barakh Ali and coworkers used SLS to print diclofenac tablets at different powder bed temperatures (Barakh Ali et al., 2019). An increase in the powder bed temperature increased the tablet hardness and disintegration time and subsequently decreased drug release from tablets. The authors attributed this behavior to an increase in the extent of sintering at high temperatures of powder bed compared to that attained at lower temperatures.

The laser source used in the SLS process may vary from laser diodes to CO₂ lasers (Awad et al., 2020; Gueche et al., 2021a; Gueche et al., 2021b). The selection of a suitable laser source depends on the optical characteristics of the powder particles. The wavelength required for sintering depends on the laser beam diameter and the formulation attributes (Charoo et al., 2020). The energy density was selected as the main laser-related process parameter in some reported studies (Awad et al., 2020; Charoo et al., 2020; Gueche et al., 2021a; Gueche et al., 2021b). Energy density describes the amount of energy transmitted per unit area (J/mm²) and relates to the laser power, scanning speed, and hatch spacing, as shown in Equation (10) (Charoo et al., 2020).

$$\mathit{\text{EnergyDensity}}=\frac{\mathit{\text{Laserpower}}}{\mathit{\text{ScanningSpeed}}*\mathit{\text{Hatchspacing}}}$$ (10)

According to this equation, an increase in laser power increases the energy density, while an increase in the scanning speed decreases this parameter. Increasing the laser power improves the mechanical properties and the density of printed tablets in SLS (Singh et al., 2017; Tikhomirov et al., 2023; Vande Ryse et al., 2022). Tikhomirov and coworkers used SLS to print naproxen-loaded tablets to study the effect of laser energy input on tablet characteristics (Tikhomirov et al., 2023). This study demonstrated that the mass, hardness, and friability of the naproxen tablets may be tunable by changing the laser power applied. Increasing the laser power led to an increase in naproxen tablets hardness and weight while decreasing their friability. On the other hand, Bai and coworkers found that low laser power improves dimensional accuracy; however, it forms weak and highly porous printed objects (Bai et al., 2016).

Some SLS printers use a fixed laser power, in these cases, the energy density may be controlled by varying scanning speed and hatch spacing. Scanning speed in an SLS system typically ranges from 10 to 200 mm/s. An increase in scanning speed results in a reduced energy density, indicating that a lower amount of energy is transmitted to the powder, leading to less sintering and poor mechanical properties of tablets (Barakh Ali et al., 2019; Trenfield et al., 2018b). Barakh Ali and coworkers studied the effect of scanning speed (270, 300, and 330 mm/s) on the hardness, disintegration time, and drug release of diclofenac tablets (Barakh Ali et al., 2019). This study demonstrated that increasing scanning speed decreased hardness and disintegration time while increasing drug release. Trenfield and coworkers showed similar results for printing theophylline tablets at varying scanning speeds between 100 and 180 mm/s (Trenfield et al., 2018b). This change in scanning speed led to a decrease in tablet hardness from 485 N to 16 N and a significant reduction in their porosity.

The hatch spacing may also affect the quality attributes of printed tablets in SLS. Equation (10) shows that decrease hatch spacing increases the energy density. A longer hatch spacing may lead to un-sintered areas in printed layers and the formation of tablets with low mechanical properties and short printing time (Gibson and Shi, 1997; Singh et al., 2017). Nonetheless, hatch spacing that is too short may cause thermal deformations (Awad et al., 2020; Vande Ryse et al., 2022). Kulinowski and coworkers printed paracetamol tablets in SLS by varying hatch spacing as 150 μm (HS₁₅₀), 100 μm (HS₁₀₀), or 50 μm (HS₅₀). They reported changes in internal structure and dissolution properties from these tablets (Kulinowski et al., 2021). Fig. 10 shows that decreasing the hatch spacing from HS₁₅₀ to HS₅₀ resulted in thermal deformation of smaller particles with aggregation into clusters or melting followed by internalization into the larger grains. This behavior produced tablets with lower porosity and uniform surface structure (Kulinowski et al., 2021).

Fig. 10.

SEM microstructure of upper surface of (a) HS₁₅₀; (b) HS₁₀₀; (c) HS₅₀. Adapted from reference (Kulinowski et al., 2021) with permission from Elsevier.

5.3. Critical materials attributes for SLS

Polymeric powder blends are the starting building material used in the SLS process. The powder characteristics impact the properties of printed tablets, such as accuracy, internal stress, and mechanical properties. Theoretically, the thermoplastic polymers used in the HME and FFF may be used in the SLS process (Sarode et al., 2013). These blends must contain at least one component that can absorb laser energy at a certain wavelength for the sintering process (El Aita et al., 2018). These powder blends typically contain API and excipients; however, API may be added after printing in some SLS applications to avoid its degradation by the high temperatures used (Fina et al., 2018b). Selection of polymer candidates for SLS is made based on various factors such as molecular weight and structure, physiochemical characteristics, thermal stability, impurities profile, desired tablet design and drug release characteristics (Griehl and Ruestem, 1970; Ligon et al., 2017; Schmid et al., 2014). Likewise, these polymers used in SLS should be recognized as safe for human consumption and compatible with API and other excipients in the formulation (Charoo et al., 2020; El Aita et al., 2018; Ligon et al., 2017).

Various SLS polymers have been successfully used and reported in the literature for pharmaceutical applications including polycaprolactone, Kollicoat IR (polyvinyl-alcohol and polyethylene glycol co-polymer), hydroxypropyl methylcellulose, ethyl cellulose, Kollidon VA64 (vinylpyrrolidone-vinyl acetate co-polymer), Polyethylene oxide, Lactose monohydrate, Eudragit L100–55 (methacrylic acid and ethyl acrylate copolymer), and Eudragit RL (Allahham et al., 2020; Awad et al., 2019; Barakh Ali et al., 2019; Fina et al., 2017; Fina et al., 2018a; Fina et al., 2018b). The molecular structure and weight of the polymer affect the sintering process. For instance, previous studies have shown that polycaprolactone with a molecular weight of 40,000 g/mol exhibited poor sintering behavior, while polycaprolactone with a molecular weight of 50,000 g/mol showed better sintering properties (Charoo et al., 2020; Ligon et al., 2017).

In SLS, the critical material attributes of the powder blend include mean particle size, shape, PSD, flowability, thermal stability and absorptivity of laser energy (Ligon et al., 2017; Schmid et al., 2014). Particle size, shape and PSD play significant roles in the sintering process. A particle size ranging from ~ 60 to ~180 μm has been previously reported for pharmaceutical applications in SLS (Awad et al., 2020). Goodridge and coworkers recommended avoiding particles smaller than 45 μm, but no experimental data were provided to support the conjecture (Goodridge et al., 2012). Powder particle size may impact the process energy requirement, mechanical properties, physical and chemical stability, and release behavior of the printed tablets. Large particle size results in a reduced packing density with large spaces between particles and reduced mechanical properties of the tablets; therefore, more energy may be required to achieve efficient sintering. Smaller particles generally have poor flow properties that may cause agglomerations of the material, making the powder layering and processing in SLS more challenging (Schmid et al., 2014; Tabriz et al., 2023; Ziegelmeier et al., 2015). Additionally, powders of narrow PSD may be used to ensure uniform absorption of laser energy (Schmid et al., 2014). The shape of particles may also affect the outcomes of the SLS printing process. Spherical particles are most required to ensure acceptable flow properties and uniform energy absorption (Awad et al., 2020). Different studies have demonstrated that irregular particles may lead to non-uniform sintering and inconsistent powder layering due to poor flowability (Ziegelmeier et al., 2015). In this case, flow properties may be improved by the addition of flow modifiers such as talc, silicon dioxide, or magnesium silicates to the pharmaceutical formulation; however, this addition may significantly influence tablet characteristics (Charoo et al., 2020; Ligon et al., 2017; Tabriz et al., 2023).

Thermal characteristics of the polymers are essential in developing of SLS processes and products. Semi-crystalline and amorphous thermoplastic polymers of different thermal properties are commonly used in SLS (Charoo et al., 2020; Ligon et al., 2017). Semi-crystalline and amorphous thermoplastic polymers have very different thermal properties, which mainly affect their behavior when processed by SLS. Semi-crystalline polymers undergo a large change in viscosity and density within a narrow temperature range upon melting and crystallization, while amorphous polymers undergo less well-defined phase transitions over a broader temperature range (Tontowi and Childs, 2001).

The polymers used in SLS applications should have a high laser beam absorptivity to enable effective sintering and fusion of layers. When the powder has poor light absorptivity, incorporating a photo-absorber (absorbance enhancer) becomes necessary to promote sintering. Physicochemical characteristics and concentration of a photo-absorber may also impact the outcomes of a printing process. Two main types of photo-absorber have been reported in the literature for printing objects in SLS: Candurin^® Gold Sheen (Fina et al., 2017; Fina et al., 2018a; Fina et al., 2018b; Gueche et al., 2021b) and Candurin^® NXT Ruby Red (Barakh Ali et al., 2019). Candurin^® is a dye commonly used in the pharmaceutical industry as a tablet-coating agent (Gueche et al., 2021b). Low doses of its ingredients, namely titanium dioxide, potassium aluminum silicate and iron oxide, are generally reported as safe by the U.S. FDA (FDA, 2023d). However, further investigations on the safety of Candurin for 3D printing of drug products may still be needed (Gueche et al., 2021b; Winkler et al., 2018). Optimizing the loading concentration of a photo-absorber is essential to achieve the desired properties of printed tablets. Excessive amounts of the absorber may result in thermal degradations, while insufficient amounts lead to incomplete sintering (Yang et al., 2021). Various studies have reported that a Candurin^® NXT Ruby Red concentration of about 3 % w/w was sufficient for optimal sintering (Barakh Ali et al., 2019). Additionally, other studies have used tartrazine lake as a photo-absorber and showed that increasing its concentration beyond 3 % led to physical deformation at the edges and poor printing accuracy of tablets despite observed enhancements in the printing efficiency (Yang et al., 2021).

6. Vat photopolymerization: stereolithography

Stereolithography (SLA) is a Vat photopolymerization technique that utilizes photosensitive materials exposed to controlled radiation/light patterns to form polymerized layers of materials. SLA was invented and patented by Charles W. Hull in the late 1980 s (Hull, 1986). SLA has been used in the production of both IR and ER oral dosage forms with high printing resolution (Healy et al., 2019; Varghese et al., 2022; Wang et al., 2016). SLA does not require direct feeding of powder into the printer and eliminates concerns related to flowability and segregation. SLA is also used in printing matrixes with thermally-sensitivity APIs due to the minimal localized heating required in the process (Curti et al., 2020; Wang et al., 2016).

6.1. Operational principle for SLA

A diagram of the SLA printing process is illustrated in Fig. 11. The liquid photosensitive resin is placed inside the Vat and integrated with a movable building platform, an automated refilling pump, and a level adjustment device. A computer-controlled laser beam guided by scanner mirrors scans the pattern at the surface of a resin bath, initiating radical polymerization at the selected coordinates. After solidification, the first layer starts to adhere to the platform. The building platform is then lowered by the same depth as the first layer, and the laser scanning and curing process is repeated to complete tablet printing. At the end of the SLA process, excess resin is washed away using solvent, and the printed tablets are cured using UV light to complete the photopolymerization reaction at the surface and enhance mechanical properties. SLA has two orientation approaches depending on the direction of light projection used in the printer: top-down and bottom-up. In the top-down approach, a laser source is positioned above the Vat, and tablets are printed facing upward (Fig. 11). On the other hand, in the top-down approach, photopolymerization is performed by an irradiation source beneath the Vat and the solid is fabricated to face downward.

Fig. 11.

Schematic representation of SLA.

6.2. Critical process parameters for SLA

The quality of printed tablets in SLA is influenced by various process parameters, which should be controlled to avoid inconsistencies in dimensional accuracy, surface roughness, and mechanical properties of tablets. These parameters include layer thickness, laser power, scanning speed, print orientation, geometry, hatch spacing, hatch style, and hatch overcure.

Layer thickness is one of the most critical and studied parameters in SLA (Cekic et al., 2019; Chockalingam et al., 2008; Chockalingam et al., 2006a; Chockalingam et al., 2006b; Onuh and Hon, 1998b; Rahmati and Ghadami, 2014; Raju et al., 2010; Schaub et al., 1997; Zhou et al., 2000). A decrease in layer thickness often increases tensile strength, improves surface roughness, and decreases dimensional error in SLA printed tablets due to the shorter path of the laser through thin layers of liquid resin (Cekic et al., 2019; Chockalingam et al., 2006a; Onuh and Hon, 1998b; Raju et al., 2010; Schaub et al., 1997; Zhou et al., 2000). Nonetheless, short layer thickness leads to more slices to be printed and longer processing and build time (Onuh and Hon, 1998b).

The oligomer and monomer are exposed to the laser power of the UV laser beam. The laser power can improve the mechanical properties of the printed tablets (Lu et al., 1995; Zeng et al., 2023). A recent study by Zeng and coworkers showed an increase in tensile strength, Young’s modulus, and compressive strength of printed objects by increasing the laser power from 60 to 75 mW (Zeng et al., 2023). The use of laser beams of low intensity may lead to insufficient curing, material shrinkage and or deformity (Zeng et al., 2023). The laser power dictates the width of the cured layer and the overall resolution attained (Jacob et al., 2020). The correlation between the energy density, laser power, scanning speed, and hatch spacing is also described in Equation (10). Similarly, the interaction between laser power and scanning speed may be considered to optimize an SLA process. This interaction affects the kinetics of the curing reaction during polymerization and determines the exposure time to light (Chia and Wu, 2015; Zakeri et al., 2020). Using a low-intensity laser at a high scanning speed results in incomplete curing, while using a high-power laser at a low scanning speed may cause over-polymerization (Ni et al., 2018). Fine-tuning of laser power and scanning speed may then control the curing depth and thickness for a given resin. Since light penetration depth impacts printing time, it is critical to fine-tune all process parameters and their interactions for an efficient SLA printing process and uniform tablet characteristics (Melchels et al., 2010).

Geometrical designs of tablets and printing orientation parameters may also affect the properties of printed tablets in SLA. Robles-Martinez and coworkers varied geometry designs (cube, disc, pyramid, sphere, torus and multiple tori shapes) of tablets produced in SLA to evaluate the effect of tablet shape on drug release (Robles-Martinez et al., 2018). The results from this work showed that the ratio of surface area to volume (SA/V) of tablets significantly affected drug release kinetics from polyethylene glycol diacrylate tablets. A faster drug release was observed by increasing the SA/V ratio of tablet design. These results indicated that drug dissolution from these tablets may be modified by changing their geometry rather than altering the formulation composition (Robles-Martinez et al., 2018). Several reports showed variations in object characteristics by changing the printing orientation in SLA (Chockalingam et al., 2008; Lu et al., 2023). Lu and coworkers have printed objects of the same dimensions but with different printing orientations using SLA. The first orientation was parallel to the y-axis, while the second orientation was perpendicular to the y-axis (Lu et al., 2023). Objects printed at a parallel orientation exhibited higher strength, apparent elastic modulus, maximum principal stress peaks, and lower failure risk compared to those printed in the perpendicular orientation.

Hatch spacing, hatch overcure and hatch style are additional process parameters that may impact the quality of tablets printed using SLA processes (Khorasani and Baseri, 2012; Lee et al., 2001; Onuh and Hon, 1998b; Rahmati and Ghadami, 2014; Raju et al., 2010; Zhou et al., 2000). Hatch refers to a specific configuration of individually polymerized lines (duration and location) that describes the respective cross-sectional area on the resin surface (Onuh and Hon, 1998a). Hatch spacing describes the distance between parallel vectors used to hatch the interior of the part printed. Hatch overcure is the depth at which one cured vector string pierces into the adjacent lower layer to keep individual layers connected in printed tablets. Hatch overcure is typically set between 10 – 35 % of layer thickness (Zakeri et al., 2020). Lee and coworkers developed an artificial neural network model to understand the effects of layer thickness, hatch spacing, and hatch overcure on the dimensional accuracy of tablets (Lee et al., 2001). The outcome of this model demonstrated that dimensional accuracy decreased as hatch spacing decreased and/or hatch overcure increased for a given layer thickness. These results agreed with those reported by other research groups (Cho et al., 2000; Khorasani and Baseri, 2012; Onuh and Hon, 1998b). A smaller hatch spacing leads to overlap between strands, and higher hatch overcure indicates larger penetration of strands toward the previous layer, causing inadequate laser exposure of the strands and resulting in larger dimensional errors. Similarly, higher hatch spacing may lead to trapping liquid resin within the object formed; hence, this residual liquid must be solidified in a separate post-curing step. The hatch style affects the sequence and amount of polymer solidification in printed tablets and may alter their physical properties and internal stress distribution (Onuh and Hon, 1998a, b).

6.3. Critical materials attributes for SLA

Photopolymerizable resins used in SLA are usually blends of monomers and oligomers that cure and crosslink in the presence of a photoinitiator upon exposure to UV light at a specific wavelength. A photoinitiator is required to convert photolytic energy into reactive species (radicals or cations), which may drive the chain growth via a radical or cationic mechanism (Bagheri and Jin, 2019). These resin blends may also contain additives such as pigments, dyes, or light absorbers (Ligon et al., 2017). Methacrylate and acrylate-based resins are the most common monomers and oligomers used in SLA due to their fast reaction rates to radical polymerization, good stability and tunable mechanical properties (Ding et al., 2019). However, these common resins have a limitation of tendency to shrink during the chain growth-free radical polymerization. To overcome shrinking, synthesized alkyne-carbonate and flexible oligomers have been investigated. More recently, new hybrid resin formulations that include the addition of (methacryloxypropyl)-methylsiloxane to a siloxane-methacrylate composite and curable elastomers of epoxy aliphatic acrylates and aliphatic urethane diacrylates have been used (Xu et al., 2021). Both materials have many biomedical applications due to their enhanced mechanical properties, such as toughness and modulus. Other resins have been reported in the literature for solid oral dosage forms using SLA, including poly (ethylene glycol) diacrylate, poly(ethylene glycol), Polyethylene glycol dimethacrylate, Kollidon, hydroxypropyl methylcellulose, diphenyl (2,4, 6-trimethyl-benzoyl) phosphine oxide, (Kadry et al., 2019; Robles-Martinez et al., 2018; Robles-Martinez et al., 2019; Wang et al., 2006; Wang et al., 2016; Xu et al., 2020).

Vat photopolymerization has been used to print tablets via two different approaches. In the first approach, API is dispersed homogeneously in a crosslinked resin matrix containing a photoinitiator and a photopolymer. The release of API occurs then by diffusion through the swollen polymer matrix. In the second approach, API is loaded onto a printed tablet using after printing process through traditional drug loading techniques based on absorption, including adsorption by dipping and spray coating, where the polymer matrix is swelled in a concentrated solution of the API (Xu et al., 2021).

Photopolymerization reactions start with an interaction of the photoinitiator with UV to produce free radicals that propagate through the polymer until a termination reaction occurs. The rate of photopolymerization may be described by Equation (11) (Yu et al., 2020), where ν_PP, is the rate of photopolymerization, K_PP is the photo propagation rate constant, ϕ is quantum yield, ε is extinction coefficient, I_o is the incident light intensity, κ_t is termination rate constant, and M is monomer concentration.

$${\nu }_{PP}=K_{PP}{\left(\frac{\varphi \epsilon I_o}{{\kappa }_t}\right)}^{1/2}[M{]}^{3/2}$$ (11)

The photopolymerization rate is then dependent on the initial monomer concentration and efficiency of the photoinitiator. An increase in the initial monomer concentration leads to a nonlinear increase in polymerization rate, while the polymerization efficiency impacts printing time and resolution (Yu et al., 2020). The wavelength of the employed light source in photopolymerization is a critical parameter in selecting a suitable photoinitiator establishing optimum printing conditions.

Critical energy (E_c) and penetration depth (D_p) are resin-dependent process parameters that affect the photopolymerization rate. Optimization of these two parameters in SLA is essential to optimize the printing process through fine-tuning the laser power and scanning speed. D_p is related to the absorbance characteristic of the resin, while E_c represents the energy required to initiate polymerization. Equation (12) shows the relationship of E_c, D_p and the depth/thickness of cured resin (C_d) (Yu et al., 2020). E₀ represents the energy of incident light at the surface. Equation (12) may be used to develop a working curve by log-plotting C_d versus different applied irradiation doses, E₀ (Chiulan et al., 2021; Melchels et al., 2010).

$$C_d=D_P\text{ln}\left(\frac{E_0}{E_C}\right)$$ (12)

The energy of incident light at the surface must surpass the E_c required to reach the gel point of the resin to form a solidified layer. The level of E_c to be used may be determined by changing the initiator concentration and controlling dissolved oxygen and inhibitors. The fine-tuning of these parameters slightly increases macromonomer conversion at the interface to be more than the gel point. This tuning of the conversion rate achieves adequate bonding between layers and minimizes over-curing that can disrupt the geometric design of printed tablets, especially for those of porous internal structure (Chiulan et al., 2021; Melchels et al., 2010). Therefore, resins with high extinction coefficient and low penetration depth are preferred in SLA as they allow better control of the polymerization reaction.

Another important material attribute of the resin is its rheological behavior which affects the recoating procedure on the curing surfaces. Resins of high viscosity require a longer recoating time, decreasing the printing efficiency. On the other hand, low-viscosity resins may deteriorate the surface resolution and finish of the printed tablets (Lakkala et al., 2023). It has been reported that a viscosity range of 0.85–4.5 Pas is recommended for SLA resin in pharmaceutical applications (Sutton et al., 2018). Additionally, the degree of polymer network and molecular architecture may impact the strength and modulus of printed objects. Voet and coworkers have reported that increasing unsaturation sites in the resin backbone achieves a highly crosslinked polymer network of enhanced strength and stiffness (Voet et al., 2018).

The applicability of printing polypills of various APIs in SLA was tested recently. Robles-Martinez and coworkers used SLA to print multi-layer polypills containing five different APIs, namely paracetamol, acetylsalicylic acid, naproxen, chloramphenicol, caffeine, and prednisolone (Robles-Martinez et al., 2019). The concentration of API in each layer was determined using Raman mapping technology. Comparative analysis of Raman spectra collected from the mapped areas showed diffusion of naproxen, acetylsalicylic acid, and paracetamol between the layers printed due to their existence as amorphous solid state in the printed polypills (Robles-Martinez et al., 2019). The diffusion of drugs between layers is a quality attribute of polypills that affects their performance. Other studies evaluated the impacts of formulation attributes on weight variation and dimension accuracy of the printed tablets. Healy and coworkers manufactured tablets of paracetamol and acetylsalicylic acid with concentrations of 2.5 % and 5 %w/w, highlighting that the incorporation of the drug can impact the overall dimensions of the printed tablets even though the geometric design set by the CAD model was the same (Healy et al., 2019).

Various safety concerns have been raised for resin-based tablets printed in SLA processes (El Aita et al., 2018; Gueche et al., 2021b; Lakkala et al., 2023). These concerns are related to potential exposure to uncured resin and volatile organic compounds emitted during printing (Shirazi et al., 2015). The potential release of unreacted monomer residues may pose a safety risk, necessitating post-washing and post-curing steps to prevent traces of these residues (Lakkala et al., 2023). Other concerns related to unexpected reactions between drugs and photopolymers and temperature of photopolymerization that may adversely impact drug stability (Deshmane et al., 2021). To assure quality and address these concerns, analytical tools, such as Fourier Transform Infrared, NMR, differential scanning calorimetry, dissolution, etc., are utilized in various studies to assess the conversion rate and drug stability (Healy et al., 2019; Karakurt et al., 2020). Other studies optimized the resin matrix formulation by studying the effect of resin components on drug release. For instance, increasing the hydrophobic PCL triol ratio in the resin mixture slowed down drug release (Healy et al., 2019), while adding PEG 300 (PEGDA) as a plasticizer to the resin promoted release due to increased molecular mobility with decreasing crosslinking density (Wang et al., 2016).

7. Analytical tools for real-time assessment

Integrating PAT tools into 3D printing processes ensures robustness and repeatability while optimizing process parameters for certain formulations of specific material attributes. Table 2 provides a summary of main process parameters, material attributes and printing materials with some APIs referenced in the literature for these 3D printing techniques explored in this review. As demonstrated in previous sections, many process parameters that impact the CQAs of printed tablets should be considered when developing control strategies for 3D printing processes. Although each 3D printing technology defines its control, testing, and characterization techniques, the control strategies emphasize control over mass and energy transport parameters that affect the quality attributes of printed tablets (Norman et al., 2017). This review provides an in-depth assessment of the process parameters, critical material attributes, and the risks associated with the quality of 3D printed tablets that are essential to control 3D printing processes.

Table 2.

Summary of the main process parameters and material attributes for printing materials and APIs referenced in the literature.

3D Printing Technique	Process Parameters	Material Attributes	Excipients (Polymers, Plasticizers, Other) and APIs*	References
FFF	Extrusion temperature Building platform temperature Nozzle orifice diameter Printing speed Applied pressure Rate of cooling.	Melting and/or glass transition temperature of polymer/excipients Polymer viscosity Type and concentration of Plasticizer Filament thickness	Excipients: Polyvinyl alcohol, Hydroxypropyl cellulose, Eudragit RS PO, Kolliphor TPGS, Eudragit EPO, Triacetin, Sodium starch glycolate, Croscarmellose sodium, Crospovidone, Sorbitol, Parteck MXP, Ethyl cellulose, Hydroxypropyl methylcellulose, Soluplus, Polyethylene glycol, Microcrystalline cellulose, Lactose monohydrate, Polyvinyl pyrrolidone, Kollicoat IR, Hydroxypropyl methylcellulose acetate succinate. APIs: Paracetamol, 5 – aminosalicylic acid, 4-aminosalicylic acid, Carvedilol, Aripiprazole, Warfarin (sodium salt), Theophylline, Prednisolone, Lisinopril dihydrate, Indapamide, Amlodipine besylate, Rosuvastatin calcium, Carbamazepine, Triethyl citrate, Verapamil hydrochloride, Dipyridamole, Ibuprofen.	(Arafat et al., 2018a; Arafat et al., 2018b; Borujeni et al., 2020; Briatico-Vangosa et al., 2019; Goyanes et al., 2015a; Goyanes et al., 2015b; Ilyes et al., 2019; Jamroz et al., 2017; Li et al., 2018; Li et al., 2022; Pereira et al., 2019; Sadia et al., 2016; Shi et al., 2021; Skowyra et al., 2015; Zhang et al., 2017b)
DDM	Extrusion temperature Building platform temperature Nozzle orifice diameter Printing speed Droplet aspect ratio Discharge rate	Polymer Viscosity Melting and/or glass transition temperature of polymer Moisture content	Excipients:Eudragit E PO, Hypromellose acetate succinate, Polyethylene oxide N-10 grade, Polycaprolactone, Polyvinyl-pyrrolidone, Surplus, Kollidon VA64 (vinylpyrrolidone-vinyl acetate co-polymer). APIs: Paracetamol, Felodipine, Caffeine.	(Ebrahimi et al., 2024; McDonagh et al., 2022a, b, 2023; Zhang et al., 2021)
PAM	Applied pressure Building platform temperature Printhead temperature Nozzle orifice diameter Printing speed	Rheological properties API distribution Molecular weight Crosslinking density Glass temperature, Solent composition Solvent evaporation rate.	Excipients: Trisodium phosphate dodecahydrate,Hydroxypropyl methylcellulose, Poly(acrylic acid), Microcrystalline cellulose, Sodium starch glycolate, polyvinylpyrrolidone, Sodium phosphate monobasic and dibasic, Polyethylene glycol, Mannitol, Lactose, Polyplasdone, Polyvinyl alcohol, Starch 1500, Calcium phosphate, Eudragit NE30D. APIs: Guaifenesin, Paracetamol, Nifedipine, Captopril, Ramipril, Pravastatin sodium, Atenolol, Acetylsalicylic acid, Hydrochlorothiazide, Diclofenac sodium, Metoprolol succinate, Atorvastatin calcium dihydrate, Dipyridamole, Levetiracetam,	(Alayoubi et al., 2022; Conceição et al., 2019; El Aita et al., 2019; El Aita et al., 2020; Khaled et al., 2018; Khaled et al., 2014; Khaled et al., 2015a, b; Li et al., 2018; Zidan et al., 2019a; Zidan et al., 2019b)
BJT	Layer thickness Roller/spreader speed Printing speed Printing temperature Hatch spacing Drying time Powder bed temperature.	For powder bed: For binder liquid: Interaction binder-powder Particle size distribution Flowability Packing density Moisture content Surface energy Composition Solvent type Viscosity Density Surface tension Capillary forces Binder saturation Surface energy Contact angle Agglomeration	Excipients: Lactose monohydrate, Microcrystalline cellulose, Polyvinylpyrrolidone, Ethyl cellulose, Hydroxypropyl methylcellulose, Maltitol, Maltodextrin, Eudragit 100, Eudragit RLPO, Kollidon SR, D-sucrose, Pregelatinized starch, Micro-silica gel, Mannitol, Colloidal silica, Methylene blue, Sodium lauryl sulfate, Stearic acid, Sodium croscarmellose, calcium sulfate hemihydrate, APIs: Quinapril hydrochloride, Clotrimazole, Paracetamol, Ibuprofen, Naproxen, and Famotidine, Flufenamic acid, Amitriptyline hydrochloride, Captopril, Pseudoephedrine hydrochloride, Levetiracetam, Warfarin sodium, Indomethacin, Caffeine	(Chang et al., 2020; Infanger et al., 2019; Katstra et al., 2000; Kozakiewicz-Latala et al., 2022; Lee et al., 2003; Sen et al., 2020; Tian et al., 2018; Wang et al., 2021; Yu et al., 2009; Yu et al., 2007),
MJT	Applied pressure Nozzle orifice diameter Printing speed Substrate temperature UV wavelength Jetting frequency.	Viscosity Density Surface tension Moisture content Composition.	Excipients:Polyvinyl alcohol, Irgacure 2959, Poly (ethylene glycol) diacrylate, Nvinyl-2-pyrrolidone, 2-hydroxyethyl acrylate, sodium phosphate dibasic, potassium phosphate monobasic, Sodium lauryl sulphate, Ropinirole HCl, APIs: Minoxidil sulfate, Carvedilol, Ibuprofen, Beeswax,	(Clark et al., 2017; Clark et al., 2020; He et al., 2020; Junqueira et al., 2022; Kyobula et al., 2017; Rivers et al., 2024)
SLS	Layer thickness Laser energy density Laser beam diameter Scanning speed Powder-bed temperature Building platform temperature	Particle size distribution Particle shape Moisture content Surface energy Flowability Melting/glass transition temperature Thermal conductivity Absorptivity of laser energy Absorbance enhancer concentration.	Excipients: Kollicoat IR, Eudragit L100-55, Candurin Gold Sheen, Eudragit RL, Ethyl cellulose, Polyethylene oxide, Hydroxypropyl methylcellulose, Kollidon VA64, lactose monohydrate, Ethyl cellulose, APIs: Paracetamol, Ondansetron Hydrochloride, Cyclodextrin, Diclofenac sodium, Ibuprofen.	(Allahham et al., 2020; Awad et al., 2019; Barakh Ali et al., 2019; Fina et al., 2017; Fina et al., 2018a; Fina et al., 2018b)
SLA	Layer thickness Laser power Energy density UV light wavelength Scanning speed Hatch spacing Hatch overcure Hatch style.	Resin viscosity Concentration of photopolymer and photoinitiator Drug loading in photopolymer	Excipients:Poly(ethylene glycol) diacrylate, Polyethylene glycol dimethacrylate, Kollidon, Hydroxypropyl methylcellulose,Poly (ethylene glycol), Diphenyl (2,4, 6 – trimethyl – benzoyl) phosphine oxide, APIs: Theophylline, Paracetamol, Acetylsalicylic acid, Naproxen, Chloramphenicol, Caffeine, Prednisolone, Irbesartan, Atenolol, Hydrochlorothiazide, Amlodipine.	(Kadry et al., 2019; Robles-Martinez et al., 2018; Robles-Martinez et al., 2019; Wang et al., 2006; Wang et al., 2016; Xu et al., 2020).

^*It is expected that GRAS (Generally Recognized as Safe) classified materials will be utilized in the manufacturing of drug products intended for human consumption.

In traditional pharmaceutical manufacturing, PAT tools have been implemented using spectroscopic techniques such as Raman and NIR due to their non-destructive nature and quick analysis (De Beer et al., 2011; Rangel-Gil et al., 2023; Sierra-Vega et al., 2022; Sierra-Vega et al., 2019); however, only a few studies reported implementing real-time strategies in 3D printing in pharmaceutical applications. Different PAT tools have been used for real-time monitoring of the 3D printing process in other fields. Tools such as NIR and Raman spectroscopy, thermography, X-ray imaging, acoustic emissions, and visual imaging have been very helpful in the real-time monitoring of powder consolidation, deposition process, chemical composition, temperature distribution, geometry profile, droplet speeds, and defect detection (AbouelNour and Gupta, 2022; Everton et al., 2016; Fu et al., 2021; Kong et al., 2020; Oleff et al., 2021). Nevertheless, the integration of PAT tools in pharmaceutical 3D printing has been reported mainly for offline applications, which involves removing the sample from the process flow before analysis (Edinger et al., 2019; Jorgensen et al., 2023; Khorasani et al., 2016; Melendez et al., 2007; Trenfield et al., 2018a; Trenfield et al., 2018b; Trenfield et al., 2022; Trenfield et al., 2023).

The experience of PAT applications in other fields may be adopted in the pharmaceutical field for real-time monitoring of the 3D printing processes and ensuring the quality of printed tablets. For instance, NIR and Raman spectroscopic may be used for online quantification of API and its distribution in various printed layers and, consequently, for correlating printing parameters to drug release properties (Yang et al., 2023). Thermography may provide early warnings regarding potential hot spots during printing that may lead to API degradation, as well as provide insight into heat transfer between polymeric interfaces and solidification (Aho et al., 2019). NIR imaging visualizes the spatial distribution of APIs and excipients and detects any solid-state transformation (Khorasani et al., 2016). For BJT and MJT, flash illumination may be used to capture digital images of jets and drops and provide quantitative information regarding tail-width fluctuations, lateral deflections, and satellite velocities. Flash illumination may also be used to assess the accuracy of the process and the ability of the print head to precisely emit an accurate amount of binder at specific areas of bed powder. Similarly, high-speed X-ray imaging was used for real-time observation of the BJT process in terms of droplet shape, satellite drop formation, and drifts from the main drop (Parab et al., 2019). The images in Fig. 12 illustrate real-time monitoring of the BJT printing process using high-speed X-ray imaging. Fig. 12(a) shows the sequence of two consecutive binder droplets, while Fig. 12(b) shows the evolution sequence in droplet geometry during flight.

Fig. 12.

Illustration of real-time monitoring of binder jetting printing process using high-speed X-ray imaging (a) Image sequence showing the behavior of two consecutive binder droplets. Nominal velocity of the droplet head was around 8 m/s. (b) Image sequence showing the evolution in the geometry of the droplet during flight. Reprinted from reference (Parab et al., 2019).

Yang and coworkers reported the integration of NIR spectroscopy to monitor the caffeine content during/after an FFF process (Yang et al., 2023). Initially, the authors developed a NIR calibration model using full-completion caffeine tablets. However, the caffeine predictions of this NIR model were influenced by the surface morphology and stage of completion of the scanned tablets; for which the development of an NIR calibration model was necessary considering the tablet completion percentage. Seoane-Viano and coworkers also reported the use of a MicroNIR^™ spectrometer to quantify the efavirenz load in 3D printed tablets manufactured using an FFF process based on powder extrusion (Seoane-Viano et al., 2023). The MicroNIR^™ probe was attached to a moving part of the printer, parallel to the printer printhead and perpendicular to the building plate, as shown in Fig. 13. In-line NIR spectra were measured during the manufacturing process right after each tablet was printed. NIR calibration models were able to predict the API concentration with root mean square error (RMSE) < 1.1 %. NIR hyperspectral imaging (HSI) was also used to quantify metformin hydrochloride concentration in dosage forms printed using BJT (Stranzinger et al., 2021). The spectra were acquired just after the printing process. The concentration distribution maps provided by the developed NIR-HSI models were capable of clustering and predicting API loading in the formulations with a linear correlation with HPLC data.

Fig. 13.

Direct powder extrusion printhead with the attached NIR probe. Reprinted from reference (Seoane-Viano et al., 2023).

Drug concentration is not the only quality attribute subject to the application of PAT for process monitoring. A pressure sensor was integrated into a pharmaceutical PAM process to monitor the printing pressure and characterize the rheological properties of the semi-solid materials under the exact printing conditions (Diaz-Torres et al., 2022). The rheological properties of printing materials may change with the progression of extrusion cycles of 3D printing; hence, the implementation of this pressure sensor identified the most suitable printing conditions for the different semi-solid materials. In general, integrating PAT tools into printing processes enhances understanding, maintains control, and optimizes critical printing parameters for each printing material while developing a robust analysis of the 3D printed tablets and verifying the finished product meets the established quality specifications.

8. Quality considerations

The current regulatory framework encourages innovation in pharmaceutical manufacturing by implementing risk-based, systematic, and science-based approaches complemented by robust pharmaceutical quality systems (FDA, 2004; ICH-Q8(R2), 2009; ICH-Q9, 2006; ICH-Q10, 2009). The U.S. FDA recognizes the potential of additive manufacturing (e.g., 3D printing technology) to customize medical devices and pharmaceuticals for individual patients, as well as to enable the distributed manufacturing of complex and advanced medical products in remote or austere conditions (FDA, 2023a). In December 2017, the U.S. FDA issued guidance to provide technical considerations for the additive manufacturing of medical devices (FDA, 2017b), presenting comprehensive technical considerations on the design, materials, printing process parameters and validation methodologies. 3D-printed drug products are expected to adhere to the Current Good Manufacturing Practice (CGMP) and regulatory standards in terms of quality, aligning with the quality expectations of other pharmaceutical products manufactured by conventional methods.

To support the adoption of emerging technologies, including 3D printing, into pharmaceutical manufacturing, the U.S. FDA’s Center for Drug Evaluation and Research established the Emerging Technology Program. This initiative serves as a collaborative platform, enabling manufacturers to engage directly with the U.S. FDA to identify and address potential technical and regulatory challenges related to the novel technology before filing or submitting a regulatory application (FDA, 2017a). Concurrently, the U.S. FDA conducts research endeavors to understand the implications of emerging technology on product quality. This knowledge not only informs the Emerging Technology Program but also ensures that U.S. FDA regulatory frameworks remain harmonized with the latest advancements in manufacturing science (FDA, 2023c). Furthermore, the U.S. FDA actively engages in collaborative partnerships within the public–private domain, encouraging alliances across a diverse spectrum of interdisciplinary stakeholders.

Recognizing the diverse variables inherent to each 3D printing technology mentioned in previous sections, a one-size-fits-all control strategy may not be applicable. This review shows that the quality assurance of 3D-printed drug products depends on many variables, including 3D printing technique, process parameters and materials. As regulatory agencies and industry stakeholders gain more experience and expertise in 3D printing processes and 3D-printed drug product development, several quality considerations may help ensure their successful implementation. Below are some key quality considerations to support the integration of 3D printing technology in pharmaceutical manufacturing:

1. Compliance with CGMP Standards and Quality Assurance: CGMP standards are also applicable for the integration of 3D printing technologies into the manufacturing of drug products (U. S Department of Health and Human Services, 2014). CGMP helps to uphold the quality and integrity of pharmaceutical products by addressing key aspects such as incorporating 3D printing into existing pharmaceutical quality systems, establishing validation schemes, continuous verification plans, and robust protocols for materials handling, cleaning validations, in-process control, printer calibration, and post-processing quality checks (Awad et al., 2022).

2. Material selection and safety: 3D printing introduces a wide array of materials and processes, necessitating rigorous characterization of material properties, degradation profiles, and interactions with the API to guarantee safety, efficacy, and suitability of the 3D printing technology. In addition, the verification of batch-to-batch consistency reveals product changes resulted from variations in printing process. For example, material recycling in BJT and SLS may induce potential alterations in material properties from its original state (Long Ng et al., 2024).

3. Process monitoring and control: Continuous monitoring of the critical printing parameters may be useful to ensure that printed product meets the quality expectations. Real-time monitoring of process parameters, including temperature, pressure, and layer adhesion, may be considered to identify deviations and maintain control over product quality. Monitoring a printing process can provide insights into the relationship between critical starting and in-process material attributes and product CQAs. Additionally, an automated sampling process may be considered to collect representative samples throughout the printing process for quality analysis and to determine the possible root causes of failure modes.

4. Scale-up: Scaling up 3D printing processes from lab-scale prototypes to full-scale production may present challenges in optimization, validation, and supply chain management. Approaches such as increasing the print head number and using parallel units are common strategies for scaling 3D printing processes (Awad et al., 2022). Feasibility studies and risk assessments can be performed to identify potential barriers to scale-up and develop mitigation strategies.

5. Information technology and software architecture: A robust IT infrastructure and software architecture can be implemented to support data security, integrity, traceability, process control, and regulatory compliance (FDA, 2002, 2023b). Many 3D printing processes are engineered as automated systems with precise local controls over the critical parameters of the printing process.

6. Artificial Intelligence (AI): Ongoing advancements in AI have the potential to accelerate the development of safe and effective drugs, and ultimately improving patient care (FDA, 2025). AI models may be trained to analyze material properties, such as their compatibility with specific APIs, mechanical strength, solubility, and biocompatibility. By processing large dataset from experiments and simulations, historical data, material characteristics, and patient-specific information, AI models can identify patterns and recommend optimized drug formulations and 3D printing processes tailored to individual needs (Serrano et al., 2024). Additionally, a risk-based credibility assessment framework needs to be established to ensure the reliability and safety of AI-driven recommendations (FDA, 2025).

9. Conclusion

3D printing may provide a potential opportunity to personalize solid oral dosage forms to address diverse patient needs and to offer significant advantages over conventional manufacturing methods in terms of formulation flexibility, drug product design and programmed drug release profiles. 3D printing processes are intrinsically scalable, with the capacity to be integrated as a semi-continuous or continuous process. However, the full implementation of 3D printing in pharmaceutical manufacturing is contingent on addressing specific challenges highlighted in this review, including an understanding of thermal, mechanical, rheological, and physicochemical properties of raw materials, as well as the interactions between materials attributes and process parameters. Moreover, ensuring the quality attributes of 3D-printed products during development and over the product lifecycle is critical. These attributes include drug content and distribution, geometrical design, dissolution properties, and possibility of drug degradation and polymorphism associated with high temperature and other processing conditions. Integrating in-line analytical tools becomes beneficial in this regard, facilitating real-time insights that contribute to product quality and improvement in 3D printing processes in pharmaceutical manufacturing. As the industry evolves, these considerations will be important in harnessing the full potential of 3D printing for personalized and efficient healthcare solutions.

Data availability

No data was used for the research described in the article.