On the manufacturing of potassium sodium niobate piezoceramics with low viscosity slurry via digital light processing using high refractive index monomers

The doped KNN piezoceramic powder was purchased from (Entekno Industrial, Technological and Nano Materials Corp., Turkey) with \(d_{50} = 1~\upmu \text {m}\). Piezoceramic contains some smaller and larger particles as shown in the scanning electron microscopy (SEM) image in Fig. 1. The particles are highly non-sperical, with a few straight edges and rounded corners, which is typical of KNN piezoceramics. KNN piezoceramic is chosen for its relatively high piezoelectric coefficients and lead-free composition, which ensures the production of eco-friendly, high-performance components. The UV resin used in the study was prepared by combining 1,6-hexanediol diacrylate (HDDA), technical grade purity 80%, and 4-acryloylmorpholine (ACMO), purity 97%, both purchased from Merck KgaA, Germany. The photoinitiator (PI) under the name Fine Tuner FT1 (most likely Ethyl(2,4,6-trimethylbenzoyl)phenylphosphinate oxide (liquid form of TPO, also known as TPO-L)) was purchased from Resyner Technologies S.L., Spain. A PI is essential in every UV curable system because it generates free radicals when exposed to UV light, which initiate the polymerization of ACMO and HDDA. The PI enables the layer-by-layer curing of the ceramic slurry using UV light. This specific PI was selected because it is highly efficient in the UV range used by common vat photopolymerization systems (405 nm), ensures fast curing times and high-resolution prints. Its liquid form is easier to disperse and increases slurry viscosity less than its powder form (known as TPO). The solvent-free, polymeric nature wetting and dispersing additive DISPERBYK-111, containing phosphoric acid ester, was purchased from BYK-Chemie GmbH, Germany. Section 3 describes the justifications for selecting the UV resin and dispersing additive and their functions. In short, KNN provides functional piezoelectric properties, ACMO and HDDA form a photopolymerizable matrix with suitable viscosity and mechanical strength, PI ensures efficient curing, and DISPERBYK-111 maintains slurry homogeneity.

The 3D printable slurry is prepared by adding 76.7 wt.% of the as-received, undried, unprocessed KNN powder to UV resin consisting of 75 wt.% HDDA and 25 wt.% ACMO. Both the PI and the additive DISPERBYK-111 were calculated based on the weight of the slurry (ceramic particles + UV resin). The best printable composition contains 0.5 wt.% PI and 0.25 wt.% DISPERBYK-111. To prepare the 3D printable ceramic slurry, both monomers, the PI, the additive DISPERBYK-111, and the ceramic particles were added to the cup and mixed under vacuum for 5 minutes using a centrifugal mixer (800-1800 rpm).

A DLP type 3D printer (ELEGOO Inc., China) with a UV light wavelength of 405 nm and a light intensity of 4 mW/cm² was used to fabricate ceramic specimens. The 3D printer was modified with a tape casting system to ensure high quality prints by casting a thin layer of ceramic slurry after each layer. After 3D printing, the parts were washed with isopropyl alcohol and then slowly debinded by heating in an air atmosphere at \(1^{\circ }\text {C}\)/min up to \(600^{\circ }\text {C}\), where it was held for another hour. The debinded specimens were sintered in a conventional one-step sintering process in a closed alumina crucible at \(3^{\circ }\text {C}\)/min up to the sintering temperature (\(1120^{\circ }\text {C}\)), where the specimens were held for one hour, followed by relatively rapid cooling (first 5 minutes at \(60^{\circ }\text {C}\)/min, further at \(25^{\circ }\text {C}\)/min). Electrodes were formed on the sintered specimens by manually applying a thin layer of conductive silver ink (CW2205, Chemtronics, USA) with a brush. After drying, the specimens were heated in an oven at \(120^{\circ }\text {C}\) for 10 minutes to achieve maximum conductivity of the silver ink. Finally, the sintered piezoceramic specimens were polarized in silicone oil heated to \(120^{\circ }\text {C}\) for a total of 6 minutes (1 minute ramp time, 5 minutes hold time, 10 seconds ramp-down time) with an electric field of 3 kV/mm.

For the SEM images, the KNN particles were simply dropped onto a sticky conductive tape on the specimen holder without applying any pressure. The cross-sectional SEM images of the specimens were obtained by breaking the ceramic specimens at room temperature and mounting them on a special holder to expose their cross-sections. All specimens used for SEM were sputtered with a 4 nm thick platinum layer for higher resolution. A Helios G4 CX DualBeam^TM SEM imaging system (Thermo Fisher Scientific, Waltham, MA, USA) was used.

The 3D printed specimens were analyzed using thermogravimetric analysis (TGA) to understand the degradation of UV light-cured resin with temperature. Measurements were performed with a TG 209 F1 Libra Netzsch instrument (NETZSCH-Gerätebau GmbH, Germany) from \(30^{\circ }\text {C}\) to \(800^{\circ }\text {C}\) at a rate of \(10^{\circ }\text {C}\)/min using a synthetic air atmosphere (20/80% oxygen and nitrogen).

The viscosities of the prepared ceramic slurries were measured using a rheometer (Anton Paar MCR 702, Anton Paar Germany GmbH, Ostfildern-Scharnhausen, Germany) at varying shear rates from 1 to \(50~\text {s}^{-1}\) at \(25^{\circ }\text {C}\). The different shear rates were chosen based on our previous studies where a clear shear thinning effect was observed [51, 52]. The metal plates used for viscosity measurements had a diameter of 25 mm and the distance between the plates was set to 0.5 mm due to the very low viscosity of the slurries.

The specimens used to measure cure depth were manufactured on the same 3D printer used to manufacture the specimens, exposing a single layer for a given time. The specimens were measured using a digital caliper, with a minimum of three specimens per exposure time and material configuration to obtain an average.

The relative permittivity was calculated usingafter measuring the capacitance of the sintered and polarized piezoceramics at room temperature at 1 kHz. In Eq. 1, C defines the capacitance of the sensor at a frequency of interest, d is the average thickness of the sensor, \(\varepsilon _0\) is the vacuum permittivity, and A describes the overlapping electrode area of the sensor. Due to the hollow cylinder geometry, the approximate area of the internal electrode was used. The capacitance was measured with an LCR meter (Voltcraft LCR-300, Conrad Electronic SE, Hirschau, Germany). The dissipation factor and the mechanical quality factor \(Q_m\) were measured directly with the same LCR meter. Seven specimens were used to obtain an average. The capacitance and dissipation factor versus temperature were measured with the same LCR meter, while slowly heating the specimens in an oven. Two specimens were measured.

The electrical impedance of 3D printed piezoceramics was measured over frequency (10 kHz to 5 MHz) at room temperature of a free-standing sample on a table using a vector network analyzer (Bode 100, OMICRON electronics GmbH, Attn. OMICRON Lab, Austria).

The high refractive index and low viscosity of the UV resin are crucial for achieving a printable slurry with a high ceramic concentration. Monomers with low viscosity and high refractive indices, such as polyethylene glycol diacrylate (PEGDA) and HDDA, are often used to prepare ceramic slurries [2, 32, 35, 39, 49, 50, 53‐64], but the maximum printable ceramic concentration is limited. Recently, some studies have reported very low viscosities of slurries prepared with the acrylate monomer ACMO [2, 65]. ACMO is a monofunctional monomer with a high refractive index that is commonly used as a reactive diluent. It reduces the viscosity of the slurry and improves printability while participating in the photopolymerization reaction to form the rigid plastic, which binds piezoceramic particles. ACMO’s high refractive index makes it ideal for achieving high piezoceramic concentrations with high cure depths. The extremely low viscosity of ACMO is important for maximizing ceramic concentration in printable slurries. Furthermore, the polar nature of ACMO enhances its compatibility with ceramic particles, thereby improving slurry stability. Therefore, in this study, a high refractive index and low viscosity monomer ACMO was selected based on the results of the study [2], which reported high curing properties along with low viscosity. However, contrary to what was reported in [2], we found that ACMO alone cannot be used as a ceramic matrix due to its low cross-linking density, resulting in extremely soft and sticky parts after printing, which are practically unprocessable. Other studies have combined ACMO with other monomers such as polyethylene glycol diacrylate (PEG(508)DA) [65], which solves the low crosslink density problem. PEGDA can have different molecular weights, with lower values (250-400 g/mol) usually preferred due to lower viscosity [2, 60, 62‐64]. However, PEG(400)DA does not have the lowest viscosity of the most commonly used monomers [2]. Therefore, the most commonly used monomer in the scientific community for the preparation of ceramic slurry with very low viscosity is HDDA [2, 39, 50, 54‐61, 63]. Compared to a low molecular weight PEGDA, HDDA has a lower viscosity but also a slightly lower refractive index [2]. HDDA is a bifunctional crosslinking monomer with a relatively high refractive index and low viscosity. The dual reactive groups of HDDA enhance the mechanical strength and structural integrity of the printed green ceramic body and promote rapid curing and robust interlayer bonding. Its low viscosity helps maintain the viscosity of the ceramic slurry low. To increase the low cross-linking density of pure ACMO, we combined it with HDDA at a 75:25 (by weight) ratio, forming a UV light curable resin to which ceramic particles were added. The HDDA-to-ACMO ratio has not been precisely optimized and will be addressed in a subsequent study. A 50:50 (HDDA:ACMO) ratio results in similar slurry viscosity and slightly higher cure depths, but it produces flexible as-printed ceramic parts that deform easily during washing and subsequent processing steps. A higher HDDA ratio would most likely help achieve stiffer green ceramic objects. However, we believe it is unnecessary because parts printed with a 75:25 (HDDA:ACMO) weight ratio are sufficiently stiff. A higher HDDA concentration would result in reduced cure depth due to HDDA’s lower refractive index compared to ACMO. Furthermore, HDDA shrinks more than ACMO, due to its dual reactive groups. This higher HDDA shrinkage could create significant internal stresses within the printed part as the polymer network contracts during curing. These stresses could result in defects between layers or cracking during the subsequent washing step.

The low viscosity of the UV resin used is critical to the success of ceramic printing. By combining only ceramic particles and UV resin, ceramic concentrations close to 70 wt.% (approximately 40 vol.%) typically result in very high viscosity slurries with poor ceramic particle dispersion. This effect worsens as ceramic particle size decreases or ceramic particle size distribution increases. In addition, most ceramic particles are generally hydrophilic due to the many hydroxyl groups on their surface, resulting in a tendency to agglomerate [66]. On the other hand, most UV resins that are used are non-aqueous and contain hydrophobic chains, resulting in poor particle wetting when UV light curable ceramic slurries are formed. Addition of dispersing additives to ceramic slurries can increase particle wetting, thus reduce slurry viscosity. By modifying the surface of the ceramic particles with additives such as polymeric surfactants, hydrophobic properties can be imparted to the ceramic particles [66]. This results in better dispersion and stability of the ceramic slurry and enhanced curing efficiency, leading to higher-quality, fewer-defect ceramic parts. The exact process of introducing additives into the ceramic slurry to reduce its viscosity can be divided into two groups: Particle functionalization [3, 32, 34, 36, 37, 49, 56, 57, 59‐61, 65, 67, 68] or simply adding the additive to the suspensions prior to slurry mixing [2, 4, 35, 39, 40, 50, 53, 54, 58, 62‐64, 69‐77]. The latter approach, in which the dispersing additive is added to the slurry prior to mixing, is a much simpler, more sustainable, and time-saving approach. Therefore, in this study, the dispersing additive is also added directly to the slurry. Among the various dispersing additives reported in the literature, very low viscosities of ceramic slurries have been achieved with DISPERBYK-111 [3, 63, 65]. DISPERBYK-111 is a polymeric dispersant that stabilizes ceramic powders in UV resins. Its phosphate-based chemistry enables strong adsorption onto ceramic surfaces, improving particle dispersion by forming chemical bonds with the hydroxyl (-OH) groups on hydrophilic ceramic particle, such as KNN. The polymeric chain extends into the non-aqueous UV resin and creates a steric barrier around the particle. This steric hindrance prevents particle-particle interactions and reduces agglomeration. Additionally, the phosphate groups may impart a slight surface charge, which contributes to electrostatic repulsion between particles and further enhances dispersion stability. Furthermore, its polymeric nature ensures clean decomposition during debinding, leaving no residual particles in the final ceramic components. For these reasons, DISPERBYK-111 was selected for ceramic slurry preparation in this study.

The optimal composition of a printable piezoelectric ceramic slurry is 76.7 wt.% KNN and 23.3 wt.% UV resin made of 75% HDDA and 25% ACMO monomers (by weight). To the prepared suspensions, 0.5 wt.% PI and 0.25 wt.% DISPERBYK-111 are added on top of its full weight.

For the UV-based 3D printing, the viscosity of ceramic slurries must be low to ensure smooth layer coating, uniform spreading, and defect-free printing. According to the literature, viscosities grater than 3 Pa\(\cdot\)s at printing-relevant shear rates of \(10-100~\text {s}^{-1}\) can impede slurry leveling [50]. This can lead to incomplete layer formation, particularly in the initial layers, and increased printing failure rates [50]. High ceramic concentrations usually result in high slurry viscosities, which limits the maximum ceramic concentration to less than 70 wt.% [78]. This study developed a ceramic slurry with 76.6 wt.% solids content that exhibits a viscosity of 2.34 Pa\(\cdot\)s at \(1~\text {s}^{-1}\) and decreases to 0.41 Pa\(\cdot\)s at \(50~\text {s}^{-1}\), see Fig. 2, which is indicative of pronounced shear-thinning behavior. The low viscosity of 3 Pa\(\cdot\)s, well below the threshold over the wide shear rate range, facilitates smooth flow under the recoating blade. This enables the formation of a uniform layer (e.g., \(20~\upmu \text {m}\) thickness) without bubbles or voids. This is critical for the first layers, as maximum surface area contact with the build platform enhances adhesion. This viscosity profile reduces the mechanical force required for recoating, thereby improving process efficiency and minimizing wear on printer components. Shear-thinning behavior, see Fig. 2, results from particle alignment and agglomerate breakdown at higher shear rates, which reduces flow resistance. Compared to typical slurries with a similar ceramic concentrations (>5 Pa\(\cdot\)s at \(50~\text {s}^{-1}\) [3, 4, 78]), the low viscosity at 76.6 wt.% solids reflects effective dispersant optimization. This enables the manufacturing of high-density green ceramic objects.

Bild vergrößern

The dispersant DISPERBYK-111, used at an optimal concentration of 0.25 wt.% relative to the weight of the ceramic slurry, enables low viscosity of the ceramic slurry. However, the low viscosity increases the sedimentation rate of ceramic particles; a homogeneous suspension is maintained for approximately 30 minutes under static conditions. Using a layer recoater helps gently stir the slurry and prevent particle settling for at least several hours of printing. If left undisturbed, most ceramic particles will settle within 12 hours. However, manually stirring the slurry for two to three minutes will homogenize it, offering a practical solution.

Increasing the concentration of DISPERBYK-111 beyond the optimum concentration of 0.25 wt.% relative to the weight of the ceramic slurry does not alter the viscosity or sedimentation stability. Therefore, to be on the safe side, we recommend using concentrations of \(\ge\)0.25 wt.%. No negative influences were observed with higher concentrations. Below the minimum concentration of the dispersant (0.25 wt.% relative to the weight of the whole ceramic slurry), the ceramic slurry (with 76.6 wt.% ceramic) is unmixable due to poor wetting of the ceramic particles by the UV resin, resulting in extremely high viscosity. The minimum dispersing agent concentration required should vary depending on the type of ceramic, particle size, and distribution.

Overall, low ceramic slurry viscosity at high ceramic concentration enhances printability, reduces porosity in printed parts, and supports high-resolution ceramic printing. This makes it suitable for applications requiring complex geometries and high part density.

The piezoceramic slurry was optimized by varying the PI concentration between 0.25 wt.% and 1 wt.% relative to the weight of the ceramic slurry to achieve optimal printing results. Figure 3 shows the cure depth of piezoceramic slurries as a function of the PI contents. As shown in Fig. 3(a), PI concentration has a minimal impact on cure depth, with variations of only a few micrometers across the tested range. Figure 3(b) shows that cure depth follows the expected trend of decreasing at PI concentrations below and above the optimal range. This is consistent with prior studies [51]. The optimal PI concentration for the slurry in this study is 0.5 wt.% (relative to the weight of the ceramic slurry). However, concentrations between 0.4 wt.% and 0.6 wt.% yield comparable curing parameters, reducing the need for precise material mixing. Figure 3(c) presents the linear Beer-Lambert fit for the 0.5 wt.% PI slurry, which confirms a proportional relationship between cure depth and the logarithm of exposure energy, indicating high quality of the results presented.

Bild vergrößern

Notably, the piezoceramic slurry in this study exhibits superior curing characteristics compared to other slurries with similar particle sizes, concentrations, and refractive indices [73, 79]. This enhanced performance highlights the potential of the optimized slurry for advanced ceramic additive manufacturing applications.

Bild vergrößern

TGA was performed to characterize the thermal decomposition behavior and quantify the weight loss of the organic components in the UV-curable piezoceramic slurry. The results are shown in Fig. 4. The UV curable polymer mixture developed in this study and used to prepare the ceramic slurry fully decomposes in an air atmosphere below \(500^{\circ }\text {C}\). Peak gas formation occurs between \(355^{\circ }\text {C}\) and \(455^{\circ }\text {C}\), necessitating pauses in the heating curve during debinding to prevent cracking, blistering, or deformation of the specimen. After complete polymer evaporation, the residual mass is 79 wt.%, which is slightly higher than the initial amount of 76.7 wt.% KNN piezoceramic used for material preparation. At first glance, the higher residual weight of the ceramic added could indicate residue from polymers or additives. However, both PI and DISPERBYK-111 should fully decompose below \(500^{\circ }\text {C}\) in an air atmosphere due to thermo-oxidative degradation. In this process, oxygen facilitates the rapid cleavage of C-P, C-C, and C-H bonds, forming volatile species such as CO₂, H₂O, and phosphorus oxides. The organic components and phosphate groups of both PI and DISPERBYK-111 undergo complete oxidative volatilization. PI decomposes primarily between 250 and \(400^{\circ }\text {C}\), and DISPERBYK-111 decomposes primarily between 300 and \(450^{\circ }\text {C}\). There is no significant residue remaining at \(500^{\circ }\text {C}\) because the oxidative environment ensures nearly complete conversion to gaseous products. This leaves behind less than 1-5 wt.% of nonvolatile material, which is typically volatile phosphorus compounds or trace impurities and constitutes less than 0.75% of the entire ceramic slurry. The same applies to both monomers due to their low molecular weight and polar functional groups, which undergo complete volatilization in the presence of oxygen. Therefore, the remaining mass of 79 wt.% must indicate KNN piezoceramic, suggesting small error in the weight measurements during slurry preparation.

Compared to conventional debinding protocols, which typically heat green ceramic bodies to \(600^{\circ }\text {C}\) at \(1^{\circ }\text {C}\)/min [2, 4, 39, 77], the slurry in this study requires shorter debinding time of at least one hour. This efficiency stems from the optimized UV-curable polymer mixture, which enables faster debinding, reduces manufacturing time, eliminates the need for initial debinding in a nitrogen atmosphere, and lowers energy and material (if inert atmosphere is used) consumption. These findings underscore the superior performance of the developed slurry for additive manufacturing applications.

Bild vergrößern

The quality of the 3D printed piezoceramics was visually inspected under an optical microscope, see Fig. 5. The printing direction is indicated by the arrows on the left. The parts are white after printing and slightly yellowish after sintering. While the quality of the printed surfaces of the green ceramic parts immediately after printing appears visually smooth with the naked eye, the microscopic analysis reveals distinct layer lines in both as-printed, see (a), and the sintered parts, see (b). Figure 5(a) shows a side view of as-printed rectangular specimen, with dashed lines indicating the roughness of the sidewalls, which is typical of UV-curable ceramic printing. This roughness is approximately \(20~\upmu \text {m}\), matching the printing layer height. It results from overexposure near the transparent perfluoroalkoxy (PFA/nFEP) film and slight underexposure deeper in the layer due to UV light scattering by ceramic particles. In some areas, the roughness exceeds \(20~\upmu \text {m}\), as indicated by the shorter dashed line in Fig. 5(a). From our experience, most commercially 3D printed parts exhibit similar, if not higher, sidewall roughness. To achieve optimal printing outcomes with flat, smooth walls, it is essential to eliminates visible layer lines. Incorporating UV blockers or absorbers into the material composition could mitigate these issues and enhance the print quality of the piezoceramic slurry developed in this study. However, extensive experimental testing is required, which is beyond the scope of this study and be addressed in future research.

Figure 5(b) shows the cross-section of the fractured sintered part, revealing persistent layer lines and thin cavities. Note that the two images are shown at different magnifications. These are most likely caused by cleaning the 3D printed parts with isopropyl alcohol. This process creates slight, nearly invisible cracks between layers that transfer to the sintered part, reducing its quality and density. The deeper cracks (\(> 20~\upmu \text {m}\)) visible in (a) are likely caused by the same cleaning process. These findings suggest that alternative cleaning methods or materials could improve the structural integrity of as-printed and sintered piezoceramic parts. For example, water could be used since the ACMO monomer is water soluble. Further investigation into alternative cleaning methods is required.

We evaluated the quality and processability of the developed piezoelectric ceramic material by fabricating and analyzing hollow cylinders with a wall thickness of 0.3 mm. This geometry allows for effective debinding and provides valuable data on the quality of the sintered part and the material properties. It also demonstrates the ability to print thin walls using the developed UV-curable piezoceramic slurry. The specimens at different stages of manufacturing are shown in Fig. 6. In (a), the specimen is shown immediately after 3D printing (green ceramic body). In (b) sintered specimen is shown and (c) a specimen covered with conductive silver ink and attached cables prior to polarization. Note that the thin edges of the specimen in (c), on both sides, were cleaned of silver ink in order to separate the internal and external electrodes before polarization.

Bild vergrößern

Due to the lack of programmable pauses at critical temperatures (\(355-455^{\circ }\text {C}\), as shown in Fig. 4, in the current debinding equiment), thicker samples (>0.5 mm) exhibit cracks, bubbles, and other defects during uninterrupted debinding. This renders the sintered piezoceramics unusable. In contrast, printing thicker geometries poses no significant challenges.

The dimensional analysis of the hollow cylinders is summarized in Table 1. The green parts have slightly increased wall thickness (0.33 mm) due to minor overexposure during printing, as observed in Fig. 5. Thickness measurements were obtained by fracturing a printed part and analyzing it under a microscope. After sintering, the specimens shrinks, and the final dimensions are reported in Table 1. To achieve the desired sintered dimensions, we calculated a shrinkage compensation factor (printed dimension/sintered dimension \(\times\) 100), which indicates that the parts should be scaled by approximately 20.5-22.2% in the XY plane and 26.6% in the Z-axis before printing. These scaling factors align with trends reported in the literature [3, 35, 36], confirming the consistency of the material’s processing behavior.

The microstructure of the sintered piezoceramic, which was sintered at a temperature of \(1120^{\circ }\text {C}\), a heating rate of \(3^{\circ }\text {C}\)/min, and a holding time of 1 hour, was analyzed using SEM and is shown in Fig. 7. These sintering parameters were selected based on available equipment and literature data. The SEM images reveal a densely packed microstructure, suggesting that high-density piezoceramics can be achieved by mitigating the interlayer cracks observed in Fig. 5(b). However, the microstructure contains a mixture of larger grains (up to \(4~\upmu \text {m}\)) surrounded by smaller, uniformly sized grains (\(< 1~\upmu \text {m}\)). This variation in grain size indicates that further optimization of sintering parameters for KNN piezoceramics is necessary to achieve uniform grain sizes, which are critical for optimal piezoelectric performance. Although optimizing these parameters is beyond the scope of this study, the observed density demonstrates the potential of the developed UV-curable piezoceramic slurry for high-quality additive manufacturing.

Bild vergrößern

To validate the suitability of the 3D printable ceramic slurry for piezoceramic production, the dielectric properties and impedance curves of polarized piezoceramic hollow cylinders manufactured using the developed slurry, were measured. Seven hollow cylinder specimens were fabricated and polarized successfully without failure. The average dielectric properties at room temperature are summarized in Table 2. The series and parallel capacitances are almost exactly the same, so only one is given. Unfortunately, the manufacturer of piezoceramics does not provide any material properties for comparison. The measured relative permittivity of 613 is consistent with that of typical lead-free KNN piezoceramics (500-800) [80], making it competitive within this category. The dielectric loss of 0.026 is low and comparable to that of lead-based PZT. It is also superior to that of many lead-free piezoceramics, which often exhibit higher losses. The series resistance of 2.253 \(\mathrm k\Omega\) is relatively high, which is typical of KNN due to its intrinsic material properties. However, it is suitable for low-frequency applications. The parallel resistance, in the \(\mathrm M\Omega\) range, indicates excellent electrical insulation and minimized leakage currents. The mechanical quality factor \(Q_m\) is only 38.35, which is generally very low and indicates “soft” piezoceramic material suitable for sensors and low-power applications. However, we believe this value could be increased by optimizing sintering parameters.

Figure 8 shows the capacitance and dielectric loss of piezoceramics manufactured at temperatures up to \(200^{\circ }\text {C}\). Temperature-dependent measurements reveal that capacitance gradually increases with temperature due to enhanced dipole mobility in the crystal lattice. Meanwhile, dielectric loss rises due to increased domain wall motion. This leads to higher frictional losses as the domains reorient under the applied electric field. Furthermore, a rise in temperature increases ionic and electronic conductivities, resulting in leakage currents within the material. The increase in dielectric loss with temperature is relatively high, which is typical for soft piezoceramics with more mobile domains. Above \(150^{\circ }\text {C}\), both capacitance and dielectric loss increase sharply, but neither peaks until above \(200^{\circ }\text {C}\). This suggests a phase transition slightly above this temperature, which is consistent with KNN piezoceramics. Testing above \(200^{\circ }\text {C}\) was limited by cable durability. However, the observed trends align with standard KNN piezoceramic behavior, confirming the ceramic slurry’s suitability for manufacturing functional piezoceramic devices.

Figure 9 shows the impedance analysis, which further validates the piezoceramic performance. (a) displays resonance and anti-resonance peaks in the impedance magnitude curve accompanied by decreases in phase angle at frequencies corresponding to vibration modes (e.g., thickness, radial, and shear). These features confirm the successful polarization and functionality of the piezoceramics. Figure 9(b) highlights the first resonance frequency and rounded impedance peaks, which are indicative of soft piezoceramics with a low mechanical quality factor \(Q_m\) and average coupling coefficients. Most of the KNN piezoceramic compositions are considered as soft PZT substitutes.

Bild vergrößern

This study demonstrates the efficiency of the developed 3D printable slurry for additive manufacturing. Optimized for high cure depth and low viscosity, the slurry produces functional piezoceramics, suitable for sensors and low-power applications. Although sintering optimization requires further investigation, the current findings confirm the potential of the piezoceramic slurry for producing high-quality piezoceramic production.