Acceptor and Donor Dopants in Potassium Sodium Niobate Based Ceramics

B-site doping in potassium sodium niobate (KNN) with Mn2+ (M⁢nNb″′) and Ti4+ (T⁢iNb′) dopants were soluble but prevented KNN from achieving a high relative density, while Sn4+ (S⁢nNb′) was not soluble in the structure as evidenced by second phase peaks in X-ray diffraction (XRD) traces. However, SnO2 was an effective sintering aid in KNN-50/50. A-site doping with Sr2+ (S⁢r(Na,K)⋅) up to 1 mol% initially improved dielectric properties but higher sintering temperatures were required for compositions with >1 mol% Sr. Samples with 5% and 7% of Sr-doping completely shifted the transition of TO–T to below RT and broadened the TC peaks as the relaxor. All Ti-doped and Sr-doped compositions showed an increase in conductivity, manifested as high values of dielectric loss (tanδ). More than 1% of acceptor and donor dopants showed the ionic-type conduction mechanism, while 1% displayed the electronic mechanism as attributed from the strongly frequency-dependent tanδ. In conclusion, these samples have the potential to open up new applications in the field of electroceramics.


2012;
The aim of this research was to introduce acceptor-dopants (Mn 2+ , Ti 4+ , and Sn 4+ ) and donor-dopants (Sr 2+ ) into KNN to see the individual effect of their solubility and structural changes on the electrical properties of KNN-50/50-based ceramics.

EXPERIMENTAL PROCEDURE
Nb 2 O 5 was obtained from Stanford Materials Corporation with 99.999% purity, whereas K 2 CO 3 and Na 2 CO 3 were obtained from Fisher Scientific with 99.9% anhydrous used. In addition to this, TiO 2 was obtained from Aldrich Chemistry with 99.99%, and MnO 2 , SnO 2 , and SrCO 3 were obtained from Aldrich with 99.9% purity each. Carbonates were dried at 300 • C, whereas all other oxides were heated at 900 • C for 24 h. Batches of 50 g of doped-KNN formulations (K 0 . 5 Na 0 . 5 Ti and (K 0 . 5 Na 0 . 5 ) 1−x Sr x NbO 3 where 0.0 ≤ x ≤ 0.07) were prepared from dried raw powders in a hot condition (∼200 • C) to avoid the non-stoichiometric conditions caused by moisture, especially in carbonates. All compositions were attrition milled for 1 h in a 500 ml jar at 300 rpm in isopropanol using 3 mm Dia. Y 2 O 3 stabilized zirconia milling media prior to calcination. After milling, the slurry was washed with further isopropanol, separated from milling media through a sieve, and volatiles were removed at 80 • C for 24 h in a drying oven. The dried material was sieved through a 150-micron mesh and calcined for 6 h at 850 • C at 3 • C/min and 5 • C/min, heating and cooling rates, respectively. Reacted powders were re-milled before pressing a pellet. 10 mm diameter disks of all compounds were pressed uniaxially with 2ton force and fired at in a temperature range of 1120-1165 • C Densities of pellets were measured as per the Archimedes method, which were around 90%-bulk in average. XRD traces of sintered pellets were obtained using a Siemens D500 diffractometer at the 2θ range of 10 o -80 o , using CuKα radiation. The dielectric properties were characterised using an LCR meter (Model 4284A, Hewlett Packard).

XRD
The XRD traces from KNN-50/50 doped on the B-site with Mn 2+ , Ti 4+ , and Sn 4+ are shown in Figures 1, 2. The XRD traces from KNN-50/50 doped with 1% Mn and 1% Ti both appear to be similar, but the shape of 2θ ∼ 45 • {220}, and {002} peaks shows a slight difference (Figure 1). In both cases, the intensity of the 2θ = 32 • {111} is lower than that of the undoped KNN . Increasing the concentration of Ti to 5% Ti emphasized the structural change observed at 1% Ti. There are a few studies on single Mn acceptor dopant in KNN. Peaks shifting toward higher 2θ is evidence of shrinkage in the lattice volume, which took place for Mn 2+ (Rafiq et al., 2015) and Ti 4+ acceptor dopants as compared to the undoped KNN peak (shown in Figure 4; 2θ = 45-46 • ). Conversely, Lin et al. (2010) reported that XRD peaks shifted toward lower 2θ (i.e., expansion in the lattice volume) in the case of Mn 4+ doped KNN.
For KNN-50/50 doped with Sn 4+ , secondary peaks of SnO 2 are visible in Figure 2. Moreover, within the resolution limits of in-house XRD, there was no discernible change in the trace of the major KNN peaks, confirming that Sn does not enter into solid solution with KNN. Su et al. (2010) investigated SnO 2 and CuO co-doping in KNN and concluded that Sn 4+ was not soluble after 1 mol%. Their XRD traces also depicted secondary peak positions similar to those in this study. Akça and Yılmaz (2015) also reported insolubility issues for Sn 4+ in KNN with secondary peaks visible in their XRD data. The reasons behind the insolubility of Sn 4+ issues in KNN is unclear, since Sn 4+ (0.69 Å) has a similar ionic radius to Nb 5+ (0.68 Å; Shannon, 1976). However, Sn 4+ is more covalently bonded to O than Nb 5+ , which may influence its solubility (Barret, 1962;IBchem, 2016). For KNN-50/50 doped with Sr 2+ on the A-site, the XRD traces revealed a broadening of the {111} peaks as a function of x with respect to undoped KNN-50/50 when fired in both air and N 2 . Figures 3, 4 should be evaluated together, indicating that Sr 2+ was incorporated within the KNN lattice.

Dielectric Properties
Acceptor Dopants: Mn 2+ , Ti 4+ , and Sn 4+ The temperature dependence of relative permittivity (ε r ) and loss (tanδ) of 1 mol% Mn doped KNN-50/50 are shown in Figures 5A,B as a function of temperature and frequency. Two phase transition temperatures, i.e., orthorhombic-tetragonal (T O−T ), and tetragonal-cubic (T T−C ) are clearly presented, though both transitions are shifted toward lower temperatures compared to undoped KNN-50/50 sintered in air. The ε r and tanδ at room temperature of 1% Mn doped KNN-50/50 was 380 and 0.05 at 100 kHz, respectively. However, at lower frequencies, the tanδ increased dramatically when increasing the temperature. The most likely cause of tanδ relates to oxygen vacancy (V ·· O ) formed according to the defect equation: At high frequencies (1 MHz), tanδ was suppressed with acceptor Mn 2+ dopant. At these frequencies, the loss mechanism relating to V ·· O may clamp out, resulting in a decrease in the overall tanδ. There are a number of potential loss mechanisms relating to V ·· O , such as rotation of defect dipoles and movement of space charge. It is not known which mechanism dominates in this study.
The dielectric properties of Ti-doped KNN-50/50 are shown in Figures 5C,D. For low concentrations, (1%) the T T−C temperature decreased with little change in T O−T . However, for high concentrations, T T−C phase transition temperatures increased, and T O−T decreased compared to the undoped and 1% Ti doped compositions and the phase transitions become broader. The peak in permittivity at around room temperature in 3 and 5% Ti-doped KNN may relate to the observation of minor changes to the shape of some XRD peaks (at {220} and {002} planes; Figure 1). Dielectric loss in Ti-doped suppressed from x = 0.01 to x = 0.05 with respect to the temperature and frequency, which suggested that the conduction mechanism shifted from electronic to ionic as the Ti dopant increased. Nevertheless, electronic species are lighter than ionic ones; that's why ionic species need more activation energy and consequently higher conductions at higher temperatures (see Supplementary Data).
The dielectric properties of Sn 4+ doped KNN are shown in Figures 6A,B. Unsurprisingly, there was no effect on the dielectric properties with respect to undoped compositions since there was no evidence that Sn 4+ enters the KNN lattice.

Donor Dopant: Sr 2+
Dielectric properties of Sr donor dopant (A-site) in KNN-50/50 are presented in Figures 6C,D. 1% Sr-doped KNN-50/50 compositions were sintered in both air and N 2 to compare the behavior with undoped in a previous study (Hussain et al., 2018) and acceptor doped compositions in the previous sections. However, for undoped KNN sintered in N 2 , the transition temperature decreased in agreement with 1% Sr-doped KNN (N 2 ). N 2 sintered Sr 2+ KNN becomes more conductive as compared to undoped KNN, presumably because Sr 2+ donates extra electrons (h · ) and low pO 2 creates V ·· O . It has been proposed that the increase in conductivity contributes to the larger permittivity for N 2 sintered KNN-1 Sr. T T−C with acceptor (Mn 2+ ) and donor (Sr 2+ ) in air are 390 • C and 384 • C, respectively, which suggests that disruption of the ferroelectric order to reduce T T−C is greater for A-site Sr 2+ doping in comparison to B-site Mn 2+ . At higher Sr concentrations, x = 0.05 and 0.07, the phase transitions broaden in temperature (inset of Figure 6C), consistent with the broadening of peaks in the XRD traces (Figure 3). In addition, the dielectric loss increased when samples became more conductive at higher than 1% Sr (see Supplementary Data). Nonetheless, this broadening and higher dielectric loss phenomenon with donor doping looks similar to the results reported in the case of W 6+ B-site donor doping in KNN, investigated by Wu S. et al. (2014).

CONCLUSION
Acceptor dopants such as Mn 2+ (Mn Nb ) and Ti 4+ (Ti Nb ) at the B-site of KNN were incorporated to modify properties, but they both inhibited densification of KNN. Nonetheless, both acceptor species were soluble in the lattice, as revealed by XRD. Dielectric losses increased dramatically at lower frequencies with increasing temperatures by using acceptors but were moderately lower at higher frequencies. A further B-site dopant (Sn 4+ ) was also attempted but was insoluble, as evidenced by the appearance of secondary phase peaks in XRD data at low concentrations. Nevertheless, Sn 4+ was an effective sintering aid in KNN-50/50 and improved its relative density. Ceramic density improved with 1 mol% Sr 2+ but at a higher sintering temperature. Sr 2+ -doped formulations showed higher conductivity, which manifested itself in higher values of tanδ. T C of KNN-1 Sr in N 2 decreased with respect to air-sintered samples.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AUTHOR CONTRIBUTIONS
FH: Investigation, software, data analysis, original draft writing, and critical discussion of results, etc. AK: Formal discussion, review and editing, and language. ZL: Conceptualization, review, and editing. NA, GW, and DZ: Review, editing, and technical additions. AM: Review, editing, and technical discussion. DW: Review, editing, formal analysis, and critical discussion. All authors contributed to the article and approved the submitted version.

FUNDING
FH acknowledges NED University of Engineering and Technology, Pakistan, for funding support. All authors acknowledge financial help from the Sustainability and Substitution of Functional Materials and Devices EPSRC grant (EP/L017563/1) through Prof. Ian M. Reaney. The data of this work is from Ph.D. thesis of FH (Corresponding Author; Hussain, 2016). ZL acknowledges the Henry Royce Institute for Advanced Materials, funded through EPSRC grants EP/R00661X/1, EP/S019367/1, EP/P02470X/1, and EP/P025285/1.