Ferroelectric materials are widely used as a piezoelectric ceramics, serving as sensors, motors, energy harvesters or parts of advance measurement devices among many others applications [1–4]. Their prominent piezo- and ferroelectric properties derived from the perovskite crystal structure are still utilized in new emerging fields of engineering, leading to new solutions and possibilities. However, those materials are also a current challenge for many researchers as their most commonly used element, lead, is proven to be dangerous to natural environment . Therefore Pb-based piezoelectrics, such as PbZrxTi(1−x)O3 (PZT), Pb1 − xLax(Zr1 − yTiy)1−x/4O3 (PLZT) or Pb(Mg1/3Nb2/3)O3 although dominating at this moment, are aimed to be replaced with lead-free solutions.
There are many possible replacements considered by various scientific associations, specialized in functional ceramics:
Barium-titanate ceramic (BT) systems – those materials are among the very first developed lead-free piezoelectrics, that could potentially replace lead-based systems. They are advantageous mainly due to their easier appliance in industry, as they are very stable at room temperatures and are said to possess low dielectric loss. However, because of their limitations, new, optimized solutions utilizing barium are required, to compensate for low Curie temperature (Tc) and relatively small deformation coefficients (d33), including: BaZrxTi(1−x)O3 - BaxCa(1−x)TiO3 (BZT-BCT) and BaTiO3 – BiFeO3 (BF-BT) materials.
Bismuth layer-structured (BLSF) systems – those materials are generally based on the layered structure in which Bi-based layer occurs interchangeably with other elements oxides, such as K, Na, Ca, Ti, Nb, Bi. Their target applications include aerospace and nuclear industries, because of relatively high Tc. However their d33 coefficients are magnitude smaller than any other popular lead-free piezoelectrics, regardless of method or dopants and therefore are decisively limiting their usage. Recent materials include Bi4Ti3O12 (BLSF), Bi4Ti3O12 (BLaT) and Bi3.1Nd0.9Ti3O12 (BNdT) layers, obtained by solid-state reactions .
Bismuth-sodium-titanate (BNT) systems – known for its relatively high piezoelectric polarization (Pr), those materials are based on following lead-free structure: (BiNa0.88K0.08Li0.04)0.5Ti0.995 Mn0.015O3, which can be doped towards better sinterability by addition of other Bi-based ferroelectric such as Bi0.5K0.5TiO3 (BKT), as shown by Taghaddos et al.. It also occurs in ((Na0.5Bi0.5)TiO3)1−x–(BaTiO3)x (NBBT) system and could be very promising in variants with relatively high Tc and decreased dielectric loss, that often occurs.
Piezoelectric nanomaterials are considered due to their improved sinterability, high density of polycrystals and homogeneous microstructure. For example, properly sintered and processed n-ZnO is a prominent material within this category, with promising piezoelectric and functional features in general. However, nanomaterial piezoelectrics applications are limited due to difficulties in spontaneous polarization alignment by applied electric field, which is crucial for piezoelectric properties.
Organic lead-free piezoelectrics – most prominent organic lead-free piezoelectric is Polyvinylidene fluoride (PVDF), which is developed mainly for energy harvesters as it possesses high conversion efficiency due to high coupling factor and easy manufacturing for such applications.
Sodium-potassium-niobate (KNN) systems – widely recognized, due to their versatile properties and generally prominent piezoelectric performance, derived from phase composition in case of 1:1:2 stoichiometric ratio (K0.5Na0.5NbO3). Both Pr and Tc parameters are relatively high when compared to other piezoelectrics with large coupling factor to follow but main issue is proper synthesis and densification of polycrystals.
Ferroelectric materials, such as described above (except for PVDF), used as piezoelectrics have the crystalline structure of perovskite, which is commonly known as a structure similar to BaTiO3. It’s specific positioning combined with proper compositions stands for many functional properties, including ferromagnetism or thermoelectricity. KNN materials consist of two phases: ferroelectric potassium niobate (K1 − xNbO3) and antiferroelectric sodium niobate (NaxNbO3) which differ in transition but crystallize in the almost the same perovskite structure. The basic transition dependency of K/Na ratio results in possibility of obtaining functional perovskite structure with outstanding piezoelectric properties, given that mentioned stoichiometric proportions are preserved, because of morphotropic phase boundary (MPB) that it creates in the tetragonal - rhombohedral transition.
The main challenge of implementing KNN lead-free piezoelectric materials is the proper synthesis. While material can be relatively simply obtained by high temperature processing of uniaxially pressed samples, it’s expected properties could be decreased due to the small density of samples. Amongst possible reasons of this situation, most possible are: evaporation of sodium, alkali vacancies created during sintering and lack of liquid phase in the microstructure of pure KNN[13–15].
In face of those problems, a proper solutions were pursued by many researching teams experimenting with both environment and composition. Shimizu et al. showed that pure alkali niobate ferroelectric and dielectric properties can be significantly improved using only low oxygen partial pressure processes.Those studies were inspired by articles on KNN ceramics sintered under low oxygen partial pressure,which showed suppressing of vacancies formation in used process, both proving that reducing oxygen interaction with those materials is desired on every step of technological process. Studies concerning sintering in H2 and Ar atmospheres respectively showed, that similar improvements of properties can be obtained by changing the environment. Moreover, similar studies shown that different atmospheres can affect dopants behavior.
Yang et al. while obtaining transparent lead-free ferroelectric ceramics, went one step further, employing sealed sintering process and comparing it to sintering in air. Although used KNN materials were differing due to the doping of Ca(Sc0.5Nb0.5)O3 (CSN), it was shown, that sealed ceramic samples were denser that their unsealed counterparts. The overall improvement of electrical properties was also noted. Additionally, more advanced, alternative processes were involved including, for example, spark plasma sintering (SPS), and while reducing volatility of obtained samples and offering overall better performance, they are still more technologically demanding.
In addition to process changes, various dopants were reported to improve KNN synthesis processes and were considered as a possible solution to this problem. The simplest idea of doping with alkaline earth metals was proposed by Malič et al. but was generally lacking significant, if any, positive effect on densification of samples. Previously mentioned articles on low oxygen partial pressure in KNN processes involved addition of LiF, to furthermore improve the densification by usage of Lithium with similar solution proposed by Saito et al. but additionally involving alkaline earth metals, resulting in advanced systems, such as: (K0.44Na0.52Li0.04)(Nb0.86Ta0.10Sb0.04)O3 in which lithium subtracts either potassium or sodium atoms, while alkaline earth metals are in niobium positions and resulting in additional improvements of piezoelectric charge coefficients (d33) with relatively high conversion ratios. Due to those findings, later iterations of solution were to follow[23, 24], showing similar increase of the same coefficients. Important process distinction was provided in studies, in which KNN was firstly annealed, and then milled with different doping oxides. Such process improved densities of sample with relatively small additions of dopants .
Another chance of improvement came with involvement of rare earth metals oxides. Those materials were expected to improve samples, based on previous research, performed to improve PZT material systems. Those studies showed that Yb3+ substitution in perovskite structure could reduce coupling factor of materials, but greatly improve its other piezoelectric properties . In case of KNN materials, coupling factor is considered high, and therefore such dopant was considered in our previous studies prior to this paper  alongside Er2O3 which was investigated by Zhao et al. and shown that it could result in creation of highly desired liquid phase during sintering .
It is worth mentioning that reversed approach, in which KNN serves as dopant was also considered, to obtain ternary systems with MPB compositions based on BNT materials. However, as of this day, this solution did not had a big impact on the lead-free piezoelectrics, as it is merely opening opportunities on further improving of BNT-based systems, which are already considered inferior due to the complicity of processes, when compared to KNN materials [10, 29, 30].
The aim of this study was to obtain a dense KNN-based material, using industrially applicable process, using the listed findings and preliminary studies performed prior to this paper, which would be tested in stress sensor. To achieve this task, dopants of rare-earth elements, namely Er2O3, were employed. Additionally, the process was improved by numerous factors. Usage of hot pressing in h-BN form would expectedly improve the density of material in comparison to commonly used techniques, additionally texturing the microstructure of samples, which is desirable in case of piezoelectrics . Moreover, the decrease of sintering temperature may reduce the overall alkali evaporation, allowing to manufacture stoichiometric 0.5 KNN material when combined with additional sodium in initial powder mixture.