Transparent conductive oxides (TCO) as transparent electrodes have been widely used in thin-film solar cells and flat panel display devices 1 2 . The commonly applied TCO materials are In₂O₃:Sn (ITO), SnO₂:F (FTO), and ZnO:Al (AZO) 1 2 . AZO has attracted much interest as a potential substitute for ITO due to the abundance of its constituent elements in nature, relatively low deposition temperature, and stability in hydrogen plasma 2 3 .

The magnetron-sputtering ceramic target is one of the most widely used methods for AZO film deposition 3 . In the sputtering system, the target plays a major role in achieving high-quality films 4 5 6 7 . Generally, the target for sputtering TCO films should have a high density, finer grain size, and better conductance 7 8 9 10 11 , which will be helpful for avoiding the formation of nodules to prolong the target lifetime 7 8 , increasing the deposition rate and film uniformity 9 and meeting the requirement of direct current (DC) sputtering. The attempts to enhance the density of the AZO ceramic target become a crucial issue for both researchers and target manufacturers 7 8 12 13 14 . Sun et al. 12 fabricated an ultrahigh-density AZO sintered body (>99.7% theoretical density) after pressureless sintering at 1,400°C by adjusting the mass fraction of polyacrylic acid when slip casting a mixture slurry of commercial ZnO and 2 wt.% Al₂O₃ powders. Hwang et al. 13 found that the preliminary heat treatment under external pressure increased the density and uniformity after a final sintering. The maximum density value of 2 wt.% Al-doped ZnO sintered at 1,350°C was about 5.52 g/cm³ (approximately 98.9% of the theoretical density). Recently, Zhang et al. 14 used a two-step sintering process to obtain the AZO ceramic with a relative density of more than 99% by sintering the 30-nm sol–gel-synthesized AZO nanoparticles at the second-step sintering temperature of 1,000°C for 12 h. To our best knowledge, the nearly full-dense (namely, exceeding 99.9% of the theoretical density) AZO ceramic has rarely been reported and still kept a challenge as before, especially by a simple and low-cost pressureless sintering at a relatively low temperature.

Few studies have also shown that a small amount of a rare earth element such as yttrium introduced into a ZnO matrix can obviously improve the properties of both ZnO films and the corresponding ceramic sputtering targets 15 16 17 18 . For example, Han et al. 16 have utilized an electrochemical deposition method to obtain a 3.7 at% yttrium-doping ZnO film with a resistivity of as low as 6.3 × 10⁻⁵ Ω cm after post-deposition annealing in nitrogen at 300°C. GfE Co. (Nuremberg, Germany) has produced a novel aluminum-doped zinc oxide ceramic with yttria doping (AZO:Y) target containing a small amount of Y₂O₃ besides Al₂O₃, which can be stably sputtered by pulsed DC sputtering technology due to the higher conductivity 17 18 . Using this kind of target, Tsai et al. 19 found that the thin AZO:Y film deposited at 300°C had the lowest resistivity of 3.6 × 10⁻⁴ Ω cm, the highest mobility of 30.7 cm² V⁻¹·s⁻¹, and the highest carrier concentration of 5.6 × 10²⁰ cm⁻³.

However, the above mentioned research results mainly focused on the properties of AZO:Y films; the detailed investigation on the influence of Y doping on the mircrostructure and densification of the AZO ceramic target itself is lacking. In this work, we attempted to fabricate highly dense AZO:Y ceramics by pressureless sintering by a coprecipitation process using Y and Al co-doped ZnO nanoparticles as raw materials, and the microstructure and densification of AZO:Y ceramic were investigated.

The SEM images of the unmodified (Figure 1a) and different amounts of yttria-modified AZO (Figure 1b,c,d) nanoparticles after calcination at 600°C for 2 h as well as the plot (Figure 1e) of particle sizes estimated from the SEM images as a function of yttria content are shown. All nanoparticle samples exhibit a nearly spherical morphology, and the average particle sizes are 45.5, 21.6, 19.1, and 15.1 nm for the AZO:Y₀, AZO:Y_0.1, AZO:Y_0.15, and AZO:Y_0.2 samples, respectively, which show a trend of size decrease with increasing Y₂O₃ contents. The result of the SEM images suggests that a small amount of Y₂O₃ addition can remarkably retard the growth of AZO nanoparticles during calcination. In order to explore the phase of AZO:Y nanoparticles after calcination, a 0.2 wt.% Y₂O₃-doping nanoparticle sample was used to carry out the XRD analysis, as shown in Figure 2. It can be seen that all diffraction peaks are only labeled to the wurtzite ZnO structure (JCPDS card no. 036–1451) without any secondary phase diffraction peaks from Al₂O₃ and Y₂O₃ or their compound. According to the energy-dispersive X-ray spectroscopy (EDS) analyses of AZO:Y_0.2 nanoparticles shown in Figure 3, Al species are found, while the trace of Y species cannot be detected due to its trivial content to the EDS detection. It can be summarized that the role of yttria doping in the calcined AZO nanoparticles is to decrease the particle size.

The AZO:Y ceramic was fabricated by pressureless sintering the calcined AZO:Y nanoparticles at 1,300°C in air. Figure 4 shows the XRD patterns of the sintered specimens at 1,300°C. For all sintered specimens, except for the main diffraction peaks corresponding to the hexagonal wurtzite ZnO structure, other small peaks are assigned to the ZnAl₂O₄ phase. Similarly, the phase related to the yttrium dopant cannot be observed. Because the ionic radius of yttrium (approximately 0.90 Å) is larger than that of zinc (approximately 0.74 Å), yttrium is hardly doped into the ZnO lattice. Almost all yttrium will react with Al₂O₃ to form an Al₂Y₄O₉ phase, as described in the patent 18 and the literature 20 . In addition, we also use EDS to detect a tiny area with Y aggregation around the grain boundary and further verify that the atom ratio of Y to Al is 2.2, which is close to the stoichiometric ratio of the Al₂Y₄O₉ phase (not shown in this paper). Considering the fact that the solubility of the Al element in ZnO is about 0.9 at% as determined by Zhang 21 and that excess Al₂O₃ totally reacts with both ZnO and Y₂O₃ to transform into ZnAl₂O₄ and Al₂Y₄O₉, respectively, the final composition of the sintered ceramic can be approximatively expressed as 97.4 wt.% ZnO:Al + (2.6 to 2.5) wt.% ZnAl₂O₄ + (0 to 0.2) wt.% Al₂Y₄O₉ with varied Y₂O₃ doping . Taking the theoretical density (TD) value of ZnO:Al, ZnAl₂O₄, and Al₂Y₄O₉ as 5.610 g/cm³, 4.640 g/cm³, and 4.520 g/cm³, the reasonable TDs of the final sintered AZO ceramics can be deduced to be 5.585 g/cm³ for the yttria-undoped ceramic and approximately 5.583 g/cm³ for the yttria-modified ceramics using a weighted average calculation. The measured densities and the TDs as a function of Y₂O₃ concentrations are plotted in Figure 5. Increasing the Y₂O₃ content, the density of the sintered specimen increased from 5.524 g/cm³ (98.90% of TD) without any Y₂O₃ doping to the highest one of 5.583 g/cm³ (99.98% of TD) with a Y₂O₃ content of 0.2 wt.%. So, the small amount of Y₂O₃ addition can obviously enhance the density to be close to the theoretical density. Furthermore, our research also reveals that the Y₂O₃ content beyond 0.2 wt.% will severely deteriorate the density of the AZO ceramic (not shown here). A doping content of 0.2 wt.% Y₂O₃ should be the most optimal one.

In order to understand the effect of Y₂O₃ modification on the microstructure of the sintered AZO ceramics, SEM was employed to observe the polished sample surface, as shown in Figure 6a,b,c,d. For the AZO:Y₀ sample, one can observe that the grain is not uniform with an average size of about 5.7 μm, and some small white particles with a size of approximately 0.5 μm uniformly disperse in an inner grain and/or grain boundary. These smaller white particles are deemed to be the secondary-phase ZnAl₂O₄, as described by Han et al. 22 and also determined by EDS analysis as shown in Figure 6e,f, where the abundant Al element in the white particles is much higher than that in the AZO:Y grain. In addition, there are some small cavities and microstructural defects in the sintered body. However, with increasing Y₂O₃ content (shown in Figure 6b,c,d), the grain size decreases to 2.0, 2.4, and 2.2 μm for AZO:Y_0.1, AZO:Y_0.15, and AZO:Y_0.2, respectively, where the grains become more uniform than those in the AZO:Y₀ sample. Meanwhile, the small cavities and microstructural defects almost disappear with the increase of the Y₂O₃ content, implying an increase in density. This result is also in accordance with the measured density shown in Figure 5. So, the observation on the microstructure demonstrates that the addition of a small amount of Y₂O₃ may also play a key role in increasing the density, refining the grain, and improving the microstructure uniformity. The reason for achieving a highly dense AZO ceramic with a fine grain can be ascribed to the following two factors: The first may be due to the use of finer AZO:Y raw nanoparticles, which possess a much higher sintering activity. Due to the smallest particle size, AZO:Y_0.2 nanoparticles can be easily sintered to bemore dense at a relatively low temperature. The second may be due to the existence of the secondary-phase particles in the grain boundary. Besides ZnAl₂O₄ particles, the introduction of Al₂Y₄O₉ into the grain boundary by Y₂O₃ doping can further refine the grain size by inhibiting the grain growth by pinning or dragging the migration of grain boundaries 22 .

Except for a high density and fine grain, a low resistivity of the AZO ceramic target will meet the requirement of DC sputtering with a high deposition rate 5 . To study the effect of yttria addition on the bulk resistivity of the AZO ceramic, PPMS was used to measure the resistivity; the results are listed in Table 1. It can be found that the Y₂O₃ addition in the AZO matrix does not change the resistivity to a great extent. It just varies from 2.1 × 10⁻³ Ω·cm without yttria addition to 4.6 × 10⁻³ Ω·cm for the AZO:Y_0.2 sample, which is yet to meet the requirement of DC sputtering. The increase in bulk resistivity can be interpreted by the fact that more AZO grains are refined by yttria doping, resulting in more serious grain boundary scattering 23 .

Bulk resistivity of AZO:Y ceramic with varied Y₂O₃ contents.

The main focus of this study was to improve the density and microstructure of the AZO ceramic target by introducing a yttria dopant with a low-cost pressureless sintering process. SEM, EDS, XRD, PPMS, and a densimeter were employed to characterize the precursor nanoparticles and the sintered AZO ceramics. Increasing the yttria doping to 0.2 wt.%, the AZO:Y nanoparticle synthesized by a coprecipitation process has a nearly sphere-shaped morphology and a mean particle diameter of 15.1 nm. With the same amount of yttria, a fully dense AZO ceramic (99.98% of TD) with a grain size of 2.2 μm and a bulk resistivity of 4.6 × 10⁻³ Ω·cm can be achieved. This kind of AZO:Y ceramic has a potential to be used as a high-quality sputtering target to deposit ZnO-based transparent conductive films with higher optical and electrical properties.

PPMS: physical property measurement system; TD: theoretical density.

The authors declare that they have no competing interests.

YY conceived the study and participated in its design and coordination. PL, MW, TW, and RT carried out the experiments on the fabrication and characterization of the nanoparticles and corresponding sintered ceramic. WS supervised the result analysis and conducted a paper modification as a corresponding author. All authors read and approved the final manuscript.