Exploring the Potential of Robocasting for High-Density Electrolytes in Solid Oxide Fuel Cells

Abstract

This study investigates the application of robocasting technology for fabricating high-density yttria-stabilized zirconia (8YSZ) electrolytes used in solid oxide fuel cells (SOFCs). The primary goal is to overcome the limitations of traditional manufacturing techniques, such as low density and poor microstructural control. Using a combination of hydrothermal synthesis, rheological testing, and robocasting, we achieved dense 8YSZ structures (over 95% density) with minimal porosity. The fabricated electrolytes underwent sintering and debinding processes, with thermal treatment profiles optimized for structural integrity. A microstructural analysis through SEM and XRD confirmed the formation of stable crystalline phase. This research opens new avenues for the use of additive manufacturing in electrochemical applications, particularly for producing complex ceramic components with superior characteristics.

1. Introduction

Solid oxide fuel cells (SOFCs) represent an environmentally friendly technology [1] that is becoming increasingly important in the field of electrochemical conversion, often outperforming other types of fuel cells. They have gained considerable interest due to their stable operational performance, environmental benefits, and enhanced fuel utilization efficiency [2]. A key challenge in enhancing the performance of SOFCs lies in the development of high-density electrolytes, which are critical for minimizing energy losses associated with ionic conduction.

There are many available materials used as electrolytes in SOFCs such as yttria-stabilized zirconia (YSZ), scandia-stabilized zirconia (ScSZ), magnesia-stabilized zirconia (MgSZ), calcia-stabilized zirconia (CaSZ) [3], and gadolinium-doped ceria [4]. Yttria-stabilized zirconia (YSZ) is recognized as a top-tier electrolyte for high-temperature applications, featuring excellent mechanical and chemical stability, as well as remarkable ionic conductivity [3].

Recent advancements in the fabrication techniques for yttria-stabilized zirconia (YSZ) electrolytes have significantly enhanced their performance and applicability in solid oxide fuel cells (SOFCs). Traditional ceramic processing methods, such as tape casting, screen printing, dip coating, and powder pressing, have been widely utilized to create dense YSZ structures; however, these techniques often limit design flexibility and exhibit some drawbacks including large ohmic resistance and pores [5,6]. Elias Shahsavari et al. [7] deposited YSZ on anode substrates using two different slurry spin coating techniques: one without thermal treatment and another with thermal assistance. The application of heat during the spin coating process enhanced the affected area and facilitated optimal particle densification prior to sintering. Scanning electron microscopy studies revealed that the porosity of the YSZ layer applied via the thermally assisted method was lower than that of the untreated layer.

Jayappa Manjanna et al. [5] synthesized cerium oxide co-doped with calcium and lanthanum electrolytes using an auto-combustion method, which demonstrated effective densification alongside an improved microstructure and modified transport properties.

In contrast, innovative approaches like 3D printing have emerged as promising alternatives, enabling the fabrication of complex geometries that can optimize electrolyte performance. Zhang et al. [8] provide a comprehensive review of additive manufacturing techniques applied to YSZ, highlighting the advantages of 3D printing in achieving tailored architectures that improve ionic conductivity and reduce overall weight.

One of the notable methods is digital light processing (DLP), as demonstrated by Zhang J. et al. [9], who showed that this technique facilitates the rapid fabrication of YSZ ceramics with enhanced mechanical strength and conductivity. Their work illustrates how DLP not only accelerates the manufacturing process but also achieves finer resolutions, crucial for developing intricate designs that can enhance the electrochemical properties of the electrolytes. Similarly, Mohammadi et al. [10] discussed robocasting, which offers precise control over the deposition of YSZ, facilitating the creation of scaffolds with optimized porosity and connectivity. This technique not only contributes to improved ionic conductivity but also enables the integration of multifunctional materials into the design.

S. Anelli et al. [11] used inkjet printing to obtain YSZ ultrathin electrolyte layer, with a thickness of approximately 3 μm. Inkjet printing has shown promise in the creation of dense and thin electrolytes.

Isabel María Peláez-Tirado et al. [12] fabricated 3D-printed electrolytes using filaments composed of 3-YSZ and 8-YSZ through the fused filament fabrication (FFF) 3D printing technique. The 3D-printed electrolytes demonstrated chemical stability during the debinding and sintering processes, showing minimal microstructural alterations when compared to traditional press and sintering methods. Additionally, they achieve very high relative densities (greater than 95%), making them suitable for operation in solid oxide fuel cells (SOFCs).

Wei et al. [13] made significant contributions to the development of 3D-printed YSZ electrolytes for SOFCs, but their method faces several challenges, including difficulties in achieving constant density, geometric accuracy, and post-processing efficiency.

In addition to these methods, hydrothermal synthesis has been identified as a traditional viable technique for producing high-purity YSZ powders with uniform particle size, as discussed by Sato et al. [14]. The resulting powders exhibit improved sintering behavior, which is essential for achieving dense structures necessary for effective SOFC performance. Moreover, Vinchhi et al. [3] emphasized the importance of refining the microstructure of YSZ through controlled fabrication techniques to enhance its mechanical and thermal stability, critical for operational efficiency in fuel cells.

Overall, the combination of traditional and advanced fabrication methods represents a significant step forward in the development of YSZ electrolytes. By integrating these methodologies, our study emphasizes the importance of refining fabrication techniques to enhance the functionality of YSZ electrolytes in energy applications, ultimately contributing to the advancement of solid oxide fuel cell technology.

Achieving dense ceramic structures is strongly influenced by the preparation of the starting powders, as emphasized in the previous literature, including Mistler and Twiname’s work [15] on tape casting techniques. Their findings underscore that uniform particle size distribution is essential for optimal packing density and successful densification, particularly at lower sintering temperatures. In our study, we employed hydrothermal synthesis to prepare the 8YSZ powders, which facilitated the formation of uniform particles and enhanced microstructural stability, thereby minimizing pore formation during sintering. This meticulous preparation not only improves the sinterability of the powders, resulting in denser structures with fewer voids, but also significantly enhances the printability of the material in the 3D printing process. The consistent particle characteristics achieved through hydrothermal synthesis ensure reliable flow properties, facilitating precise layer deposition and minimizing issues like clogging. Furthermore, the effective packing and sintering behavior of these powders contribute to improved layer density in the printed scaffolds, directly influencing the overall integrity and mechanical properties of the final structures. By connecting optimized powder preparation with the 3D printing process, we can fine-tune printing parameters to achieve high-quality ceramic structures that are suitable for applications in solid oxide fuel cell (SOFC) electrolytes, thereby enhancing the performance and efficiency of these energy systems.

This study focuses on overcoming the issues encountered by other researchers by using robocasting, an additive manufacturing method to create dense (over 95% density) and complex 3D structures—particularly disks and cylinders—from zirconia (ZrO₂) doped with 8% Y₂O₃. Robocasting facilitates the precise layer-by-layer deposition of ceramic materials, allowing the production of dense, non-porous components with highly controlled geometries. The resulting 8YSZ structures demonstrate enhanced mechanical properties and elevated ionic conductivity, effectively overcoming limitations associated with conventional fabrication methods.

2. Results and Discussion

2.1. Rheological Behavior of the Hydrogel and 8YSZ Ceramic Paste

Before the validation of the ceramic paste, rheological measurements were performed on the Pluronic hydrogel binding agent. In rheology, two of the most important parameters are represented by elastic modulus (G′) and viscous modulus (G″). G′ represents the elastic component of the viscoelastic behavior of the binder, indicating the material’s ability to store energy when deformed and return to its original shape, while G″ represents the viscous component reflecting the material’s ability to dissipate energy as heat when deformed.

In the case of Pluronic-based hydrogel, as shown in the graph in Figure 1, both G′ and G″ remain constant over time, indicating that the mechanical properties of the material do not change during the test. This suggests that the material has reached a stable state in its response to the applied stress. Moreover, when both G′ and G″ remain constant, the material maintains a consistent balance between its elastic and viscous responses throughout the test. This indicates that the material’s microstructure or molecular network remains intact without undergoing significant changes, such as degradation or phase transitions.

In practical terms, a material with constant G′ and G″ over time is desirable for applications requiring consistent mechanical properties, such as 3D printing, where uniform material behavior is crucial for producing high-quality parts with dimensional accuracy [16].

Moreover, according to the literature, depending on the objectives of the printing process, different intervals of the complex viscosity presented in

Table 1. Range of complex viscosity based on printing objective.

Printing Objective	Recommended Complex Viscosity Range (Pa·s)
High-resolution printing	Low complex viscosity 100–500
Higher printing speeds	Medium complex viscosity 500–1000
Structural integrity	High complex viscosity >1000

are recommended [17]. In this work, considering the intended application (high-density electrolytes for SOFCs), the main objective of the 3D printing process is structural integrity. In this study, given the target application—high-density electrolytes for SOFCs—maintaining structural integrity is paramount.

The rheological data presented in Figure 2 show a complex viscosity exceeding 1000 Pa·s, which strongly supports the selection of Pluronic as a binder for 8YSZ-based pastes. This high viscosity is crucial for ensuring the material’s stability and performance in the intended application. Therefore, the selection of Pluronic F-127 as a binder for preparing 8YSZ-based pastes is rheologically justified.

To evaluate the rheological behavior of the paste based on 80% 8YSZ and 20% Pluronic hydrogel, an amplitude sweep test was performed. The amplitude sweep test measures the elastic modulus (G′) and the viscous modulus (G″) over a range of shear deformations, providing insights into the viscoelastic properties of the material. For the 8YSZ-based ceramic paste, the elastic modulus (G′) and the viscous modulus (G″) are graphically represented as a function of shear deformation (γ) in Figure 3.

The data suggest that the 8YSZ-based paste is highly elastic at low deformations but transitions to a viscoelastic behavior at higher deformations. This information is crucial for applications where the paste needs to be processed or used under varying stress conditions. From this representation, the intersection between G′ and G″, known as the yield point, can be identified. This is the moment when the applied mechanical force exceeds the molecular forces, and the material begins to flow. A high yield stress (τ = 1156 Pa) indicates that the 8YSZ paste requires significant stress to start flowing, which is typical for highly viscous and structured materials.

According to the literature, to assess the printability of a ceramic paste, it must exhibit a shear-thinning rheological behavior as well as a yield point (G′ = G″) > 1 × 10³ Pa to obtain three-dimensional ceramic structures while maintaining shape integrity and self-supporting capability [18].

Thus, the graph in Figure 4 demonstrates how the apparent viscosity of the 8YSZ paste decreases with increasing shear stress, indicating shear-thinning behavior. Additionally, the yield point of the paste, based on the data presented above, is 1156 Pa, which is greater than 1 × 10³ Pa. Therefore, it can be concluded that the paste composed of 80% 8YSZ and 20% Pluronic hydrogel is suitable for use in the robocasting process.

2.2. X-Ray Diffraction Results

The X-ray powder diffraction (XRD) analysis (Figure 5) carried upon the sintered 8YSZ sample revealed the presence of two crystalline structures—one cubic (fluorite-like), reference PDF 04-023-7233, space group Fm-3m (225) and one tetragonal (ZrO₂-like), reference PDF 04-010-3269, space group P42/nmc (137), that can be attributed to the crystalline phases presented in

Table 2. Crystalline phases identified through X-ray diffraction (XRD) analysis.

Sample	Compound	PDF Reference	Chemical Formula	Crystallization System
Sintered 3D structure	Cubic ZrO2	04-023-7233	Y(x)Zr(1 − x) O2	Cubic
	Tetragonal ZrO2	04-010-3269	Y(x)Zr(1 − x) O2	Tetragonal

. The predominant structure can be attributed to a cubic solid solution yttrium zirconium oxide (Figure 5), which accounted for approximately 80% of the total crystalline phases. The second phase identified can be attributed to a tetragonal yttrium zirconium oxide, which accounts for the remaining approximately 20% of the total crystalline percentage.

The percentage of the crystalline phases was obtained by performing a Rietveld refinement upon the collected data. The refinement was performed using the Topas v5 software program (Figure 6).

The fitted profile was obtained by using a five-degree polynomial function for the background and a split Pseudo-Voight function for the peak’s shapes. The goodness of fit presented a value of 1.27, and the RWP of the refinement was 11.64. The obtained values show that the two structures can explain the collected experimental data very well.

For a better understanding of the contribution of each structure to the peaks profile in Figure 7 and Figure 8, the shape and height of the diffraction peak located at 30.135 2θ can be explained through the contribution of the (111) plane diffraction peak of the cubic structure and of the (101) plane diffraction peak of the tetragonal structure.

The semi-quantitative Rietveld analysis (Figure 9) used in order to determine the phase percentages of the studied sample assumes that all of the crystalline phases are identified and included within the analysis. The weight percentage calculation is defined according to the following equation (Hill and Howard) [19].

$$W_{\alpha }={\displaystyle \frac{S_{\alpha }{\left(ZMV\right)}_{\alpha }}{{\sum }_{j=1}^n{S}_j{\left(ZMV\right)}_j}}$$

(1)

where “S_α” represents the scaling factor of α phase, “ZM” the elementary cell mass, “V” the elementary cell volume, and “n” the number of phases present within the analysis. Thus, in Figure 9, the results of the semi-quantitative analysis are presented.

2.3. Microstructural Analysis Results

The representative SEM images for the initial and thermally treated samples are shown in Figure 10.

The 3D-printed initial samples consist of angular granules, represented by the ceramic material, embedded in an amorphous matrix attributed to the binder used. After thermal treatment, the samples exhibit a granular structure with polyhedral grains, showing no grain boundary defects, unlike findings in other research studies [20]. This indicates that both the debinding and sintering processes were successfully performed, leading to the densification of the 3D structures.

Energy-dispersive X-ray spectroscopy (EDS) was employed to determine the chemical composition before and after thermal treatment. The EDS results are presented in Figure 11. The compositional EDS results revealed an elemental configuration consistent with the design, confirming the presence of Zr and Y. In the initial samples, the elemental configuration was lower due to the presence of the binder.

The cross-sectional SEM images on the sintered cylindrical samples (Figure 12) confirm a high degree of densification in the sintered samples, with minimal porosity observed. This indicates the effectiveness of our thermal treatment process in achieving dense, mechanically robust structures, which is crucial for solid oxide fuel cell (SOFC) applications where porosity can compromise ionic conductivity and mechanical integrity.

As observed in the SEM cross-section, the presence of distinct grain boundaries without secondary phase segregation indicates that the sintering process effectively promoted phase homogeneity. This is advantageous for maintaining consistent electrochemical performance across the sample [21]. Also, the SEM cross-sectional view highlights the absence of significant defects, such as cracks or delamination, which underscores the compatibility of the robocasting technique with the thermal profile we employed. This observation supports the suitability of robocasting for creating dense ceramic components with controlled microstructures for high-performance applications.

2.4. Particle Size Distribution

The particle size distribution for the sintered samples, determined through SEM image analysis using the Java-based software ImageJ (Version 1.54k), is shown in Figure 13. The disk-shaped samples are labeled Z12.2C and Z12.3C, while the cylindrical sample is labeled Z10.4C.

The particle size distribution of the sintered samples, analyzed using SEM images and the Java-based software program ImageJ, provided important insights into the microstructure. The disk-shaped samples (Z12.2C and Z12.3C) had an average particle size of approximately 1.5 μm, while the cylindrical sample (Z10.4C) showed a slightly larger average particle size of 2.1 μm. The distribution was narrow, with over 85% of particles falling within the 1.0 to 2.5 μm range, ensuring consistent sintering and densification. The granular structure, featuring well-defined polyhedral grains, exhibited minimal porosity (less than 5%) and no visible defects, which is essential for achieving high-density electrolytes. These findings confirm that the debinding and sintering processes were effectively executed, enhancing both mechanical strength and ionic conductivity in the 3D-printed structures.

2.5. Density Measurement

The density of the structures was determined for a representative sample from each type—disk and cylinder—after thermal treatment using the Archimedes principle-based method. The results are presented in Figure 14. In both cases (5.6 g/cm³ for the disk and 5.69 g/cm³ for the cylinder), the density exceeded 95% of the theoretical density considered to be 5.879 g/cm³ obtained by theoretical density calculation through Rietveld refinement of the XRD data.

The theoretical density was calculated using the following parameters:

• The unit cell mass, volume, and weight percent of the phases within the mixture.
• The weight fraction for each phase.

This was calculated as follows:

$$w_p={\displaystyle \frac{Q_p}{{\sum }_{p=1}^{N_p}{Q}_p}}$$

(2)

where the following are defined: N_p—number pf phases; Q_p − S_pM_pV_p/B_p; S_p—Rietveld scale factor for phase p; M_p—unit cell mass for phase p; V_p—unit cell volume for phase p; B_p—Brindley correction for phase p [22].

Brindley correction is a function which considers the linear absorption coefficient of phase p, with a packing density of 1 and the linear absorption coefficient of the mixture, also for a packing density of 1 [22]. The density results for each phase obtained through Rietveld refinement offered the results presented in

Table 3. Theoretical density results obtained through Rietveld refinement of the XRD data.

Sample	Phases	Chemical Formula	Theoretical Density, g/cm3
Sintered 3D structure	Cubic ZrO2	Y(x)Zr(1 − x) O2	5.853
	Tetragonal ZrO2	Y(x)Zr(1 − x) O2	5.982

.

2.6. Electrical Behavior Evaluation by Impedance Spectroscopy

The evaluation of the electrical behavior of the 3D-printed 8YSZ-disk shaped samples was performed by electrical impedance spectroscopy (EIS) measurements. The Nyquist diagrams corresponding to sample at 650 °C, 700 °C, and 750 °C are presented in Figure 15. The data are corrected for sample geometry. It can be easily observed that the arcs of the Nyquist plots are spread on progressively narrower ranges of the real part of the impedance as the temperature of the measurement increases. This observation indicates an increase in conductivity with the temperature. The temperature-dependent increase in conductivity is typical for ionic conductors. YSZ is a well-known ionic conductor, and therefore, it could be inferred that the main transport mechanism observed for our sample is the ionic type of conductivity.

The Nyquist diagrams present two semicircles. The smaller radius semicircle is observed at a high frequency and is ascribed to the bulk conductivity. The larger radius semicircle is observed at low frequency and is ascribed to the conductivity at the grain boundaries. The grain conductivity was estimated using the following equation:

$${\sigma }_G={\displaystyle \frac{l}{R_{G}S}}$$

(3)

where l represents the thickness of the sample, R_G represents the bulk resistance of the sample, and S represents the cross-sectional area of the sample, perpendicular to the axis of the electrodes. The value of the bulk resistance R_G is determined by fitting the experimental data.

The insert of Figure 15 shows the equivalent circuit used for modeling the sample. The equivalent circuit consists of two parallel RQ circuits connected in series to a single resistor. The parallel RQ circuits model the grain and the grain boundary conductivity, respectively, where the R values represent the resistances corresponding to the conduction processes and the Q values represent constant phase elements. The series resistor models the ohmic resistance of the contacts. Using the resistance values returned by the fit, it was estimated that the grain conductivity of the sample reaches 7.5 × 10⁻⁵ S/cm. This evaluation was performed at 650 °C. The observed value for the grain conductivity is consistent with the literature data [23].

3. Materials and Methods

3.1. Hydrothermal Synthesis

The ZrO₂ powder doped with 8% Y₂O₃, referred to as 8YSZ, was synthesized using a hydrothermal method. Zirconium tetrachloride (ZrCl₄, 99%, Merck, Darmstadt, Germany) was used as the precursor for the zirconium source. To prepare the doped material, yttrium oxide (Y₂O₃) was dissolved in the ZrCl₄ solution under vigorous mechanical stirring until a clear and homogeneous solution was obtained.

Ammonia solution (NH₃, 25% p.a.) was added as a mineralizer to adjust the pH of the solution to approximately 9, creating an alkaline suspension. The pH was continuously monitored using a precise digital pH meter. The formation of the doped zirconia powder was achieved through hydrothermal treatment at 200 °C for 2 h in a Teflon-lined autoclave (Berghof, Berchtesgaden, Germany) with a 5 L capacity equipped with a water-cooling system.

Following the hydrothermal process, the resulting precipitate was thoroughly washed and filtered to remove any soluble impurities, and then dried in an oven at 110 °C until a stable weight was achieved. The final 8YSZ powder was then subjected to calcination at 1200 °C for 90 min, with a controlled heating and cooling rate of 3 °C/min.

3.2. Pluronic Hydrogel Preparation

An amount of 30% of Pluronic F-127 powder (Sigma Aldrich, St. Louis, MO, USA) was dissolved in distilled water (70%) at 90 °C to create a hydrogel with appropriate rheological properties for 3D printing applications.

3.3. Ceramic Paste Preparation

To prepare the paste for 3D printing of 8YSZ-based high-density electrolytes for solid oxide fuel cells (SOFC), a blend of 80% 8YSZ powder and 20% Pluronic hydrogel was created. This mixture was homogenized using a Thinky-ARE 250 planetary centrifugal mixer (Thinky Corporation, Tokyo, Japan) for 2 min at 2000 rpm. After homogenization, the paste was loaded into the syringe of the 3D printer and subjected to a defoaming step in the same mixer, in order to remove any air bubbles that could compromise the quality of the printing. The resulting homogeneous paste was then utilized for the additive manufacturing of two types of high-density electrolytes, disks and cylinders, using the 3D-Bioplotter EnvisionTEC Starter System (EnvisionTEC GmbH, Gladbeck, Germany).

3.4. Manufacturing of the YSZ-Based Electrolytes

The 3D structures were fabricated using the robocasting technique based on extrusion using the 3D BioPlotter EnvisionTEC Starter Series system (4th generation) equipped with the necessary software for the import of STL files and the operation of the printer. The 3D printing process began with the creation of the final models (a disk and a cylinder with a diameter of 20 mm) using SolidWorks software (Version 2019 SP0.0) and exporting them in STL format (Figure 16). RP Perfactory software (Version 3.2) was used to slice the model into layers and convert the STL file into a *.BPL file, which is compatible with the Visual Machines software (Version 2.10.130r5) responsible for operating the printer.

The extruder syringe was loaded with 8YSZ-based paste and mounted onto the X-Y-Z motion stage of the printer. Extrusion was performed by applying direct pressure from compressed air. Conical nozzles with a diameter of 800 μm were used for ceramic deposition. Printing was conducted at room temperature, with various extrusion parameters (presented in

Table 4. Printing parameters for 8YSZ-based structures.

Model	Diameter (mm)	Distance Between Strands (mm)	Applied Pressure (bar)	Extrusion Speed (mm/s)
Disk	20	0.8 mm	1.25–3.1	12–25
Cylinder	20	No infill

) adjusted to ensure continuous and uniform deposition. Examples of the obtained structures are presented in Figure 17.

3.5. Post-Printing Treatment of 8YSZ 3D-Printed Structures

The thermal treatment profile required for the debinding process was determined using thermal analysis (DSC). The analysis was performed with a NETZSCH DSC 200 F3 Maia differential scanning calorimeter in argon atmosphere. Experimental data were processed using the PROTEUS ANALYSIS software program (Version 4.8.5). Figure 18 presents the DSC curve used to identify the relevant temperatures for the debinding treatment of the obtained structures. The information gathered enables the precise adjustment of the debinding temperature profile to avoid cracks and defects in the final structure.

The organic binder Pluronic F-127 forms micelles in water at around 20–25 °C. During this process, the polymer chains self-assemble into spherical structures, with hydrophobic groups oriented inward and hydrophilic ends directed outward. This interaction with water is energetically favorable, releasing heat into the surroundings. The exothermic peak observed at approximately 60 °C indicates a dehydration process of the Pluronic solution. At this elevated temperature, water molecules bound to the hydrophilic ends of F127 may be released, emitting heat as the interaction weakens.

A distinct endothermic peak at ≈388 °C corresponds to the decomposition of the Pluronic F-127 binder, as reported in various research studies [24] on the thermal analysis of Pluronic F-127.

The thermal treatment process, shown in Figure 19, follows a multi-step approach that ensures efficient removal of organic components while promoting the desired properties in the ceramic material. After debinding at 420 °C, sintering was carried out at 1450 °C for 2 h to fully densify the structures. A slow heating rate of 1 °C/min helped preserve the integrity of the 3D structures and facilitated the decomposition of the binder.

3.6. Characterization Methods

The rheological properties of the Pluronic F-127-based hydrogel and ceramic paste were characterized at a constant temperature (T = 25 °C) using the Anton Paar MCR 302e rheometer, with a plate-plate PP25 measuring system and a constant angular frequency of ω = 1 rad/s by performing an amplitude sweep test.

Phase analysis was performed using the BRUKER D8 ADVANCE diffractometer, operated through the DIFFRACplus XRD Commander software package (Bruker AXS, Karlsruhe, Germany), utilizing the Bragg–Brentano diffraction method in a Θ-Θ coupled vertical configuration. Spectra were recorded in the range of 20–80°, with a step size of 0.020° and an acquisition time of 3.3 s per step. Data processing was carried out using DIFFRAC.EVA VER.5 2019, part of the DIFFRAC.SUITE.EVA software package (Version 5/2019), along with the ICDD PDF-5+ 2024 database.

Microstructural analysis was conducted using a QUANTA INSPECT F50 scanning electron microscope (SEM) produced by FEI, Eindhoven, The Netherlands, equipped with a field emission gun (FEG) with a resolution of 1.2 nm, and an energy-dispersive X-ray spectroscopy (EDS) system with a MnK resolution of 133 eV.

The grain size of the sintered microstructures was determined by analyzing SEM images using the ImageJ software program (Java-based).

The density of the 3D-printed structures subjected to thermal treatment was measured using the method based on Archimedes’ principle.

Electrical impedance spectroscopy (EIS) measurements were performed using an Autolab PGSTAT 128N impedance analyzer (Metrohm Autolab B.V., Utrecht, The Netherlands). Blocking electrodes of gold were deposited using magnetron sputtering on the parallel surfaces of the sample without other sample preparation. The measurements were carried out using the two contacts method in AC mode with a peak-to-peak perturbation voltage of 100 mV. The shape of the applied signal was sinusoidal and the frequency range of the measurement span from 1 MHz to 0.1 Hz. The temperature range of the experiment was established between 650 °C and 750 °C at a 50 °C interval. In order to control the temperature, the sample was mounted on a Probostat equipment (NORECS, Oslo, Norway), itself placed into a vertical tubular furnace (ELITE, Leicestershire, UK). For analysis and impedance data fitting, the NOVA 2.1 software from Methrom Autolab (Metrohm Autolab B.V., Utrecht, The Netherlands) was used.

4. Conclusions

This study demonstrates that robocasting is a viable technique for producing high-density 8YSZ electrolytes for solid oxide fuel cells, offering advantages over conventional fabrication methods. The 3D-printed structures exhibited a high density and very low porosity making them suitable for SOFC applications. The combination of precise robocasting technology and optimized thermal treatments resulted in improved microstructural control and mechanical integrity of the electrolytes. Future research could focus on scaling this approach for industrial applications and exploring the performance of the obtained electrolytes in SOFCs.