In Fig. 1a, BET SSAs of the ITO-NPs heat-treated at 400, 600, 800 and 980 °C were 102.32, 39.54, 18.49 and 8.05 m2/g respectively. The BET SSA decreased along with increasing heat-treatment temperature. Also, their average particle size was calculated from the BET SSA as follows;
$$ {\text{D }} = 6/\uprho \cdot {\text{d }} $$
(1)
where D, ρ and d are particle size, SSA and density, respectively. In Fig. 1b, the calculated average particle sizes of ITO-NPs heat-treated at each temperature were 8.2, 21.1, 45.2 and 103.8 nm, respectively. The particle size increased along with increasing heat-treatment temperature. The increase of particle size is owing to particle surface migration [7]. According to the transformation kinetics;
$$ {\text{D}} = {\text{D}}_{ 0} \cdot {\exp }\left( - {\text{E/kT}} \right) $$
(2)
where D, D0, E, k and T are diffusivity of particle surface, its initial diffusivity, activation energy for particle surface migration, Boltzmann constant and temperature, respectively. When the temperature becomes high at constant activation energy, the diffusivity of ITO-NPs surface becomes high as a function of their temperature. Accordingly, the ITO-NPs are grown when the heat-treatment temperature is raised, and their physical properties such as particle size can be simply controlled by changing the heat-treatment temperature.
We controlled BET SSA and particle size of ITO-NPs by changing heat-treatment temperature, and ITO-NPs heat-treated at 980 °C, of which BET SSA and particle size were 8.05 m2/g and 103.8 nm, was chosen to improve target density. The target density is an important factor that affects electrical and optical properties of ITO coating layer by sputtering method [8]. Lower resistivity and higher transparency can be obtained from the sputtering target with higher density and, in general, the higher density can be obtained with larger sized nanoparticles. The reason was mentioned at the later part of this section.
Thus, we synthesized ITO-NPs and heat-treated at 980 °C. HRTEM observation of ITO-NPs heat-treated at 980 °C is shown in Fig. 2a. Primary particle sizes are c.a. 30–40 nm and they were agglomerated to exhibit larger size. In Fig. 2b morphology of the ITO-NPs are round mound shape and lattice structure were obviously observed. In fact, in Fig. 2c highly crystallized diffraction pattern was observed from SAED pattern. It is supposed that crystal structure of ITO-NPs was highly ordered to (222) preferred orientation while heat-treating at the high temperature, 980 °C. In fact, XRD analysis shows that they are crystalline ITO-NPs. In Fig. 3, all the detected peaks of the nanoparticle samples was corresponded with that of crystallized ITO. From the patterns, very intense peaks were observed at the three most important peaks of In2O3 namely 〈222〉, 〈400〉, 〈440〉 reflections. The peaks do not deviate from the PDF intensities, implying random non-oriented arrangement of the ITO-NPs. The major peaks due to SnO2 at 26.5° and SnO at 33.2° 2θ were absent in the observed pattern, indicating complete miscibility of In and Sn in the proposed composition [9]. It is known that each Sn4+ replaces In3+ in ITO lattice and, thereby, donating a free electron for the conductivity. Therefore, the ITO materials retain the cubic In2O3 structure up to the solid solubility limit of the SnO2 in In2O3 [10]. ICP analysis revealed that composition ratio of Into Sn of the reused ITO-NPs was 89.95–9.98 in weight. Also, full width half maximum (FWHM) of the peak was wider than that of the commercialized ITO particles. The result indicates that the size of the ITO-NPs is nanocrystal according to the Scherrer’s equation [11]. From the X-ray diffraction peak, particle size can be calculated by using Scherrer’s equation as;
$$ {\text{t}} = 0.9\uplambda / {\text{B cos}}\,\uptheta_{\text{B}} $$
(3)
where t, λ, B, and θB are particle size, wavelength (0.1542 nm for CuKα radiation), FWHM of a peak in radians, and diffracted angle, respectively. In Eq. (3), the intensity peak increases along with a reduction in the peak half width indicating the growth of ITO-NPs. Accordingly, particle size becomes smaller as the FWHM is widened. Average particle size calculated from FWHM was c.a. 34.5 nm, which is almost in accordance with HRTEM observation in Fig. 2a. The average particle size, 103.8 nm, calculated from BET SSA, is attributed to agglomeration between particles by diffusion and particle growth under high temperature.
Nevertheless, smaller particle size shown in the HRTEM and FWHM is concerned with Derjaguin, Landau, Verwey and Overbeek (DVLO) theory [12]. In DLVO theory, the potential energy of van der Waals attraction and the potential energy of the electrical double layer interaction [13] are summed to provide a total interaction potential energy between colloidal particles. Due to differences in the surface chemistries of dispersing agent and ITO-NPs, obtaining a stable dispersion of polymeric dispersing agent and ITO-NPs can be challenging. The importance of the colloidal stability of starting dispersions on the final properties of ink pastes has been demonstrated. Zhao et al. [14] found that the presence of aggregated titanium dioxide particles in dispersions deleteriously affects the optical and mechanical properties. Researchers have found that the colloidal stability of dispersions can be disrupted by changes in surface potential of dispersing agent [15], ionic strength [15], concentration of dispersing agent [13] and particle size [13]. One way to address the questions about clustering and stability in ink paste is through predictions using DLVO theory [15]. Under drying above the latex glass transition temperature, particles consolidated and compacted, forcing the ITO-NPs to segregate into the boundary regions between dispersing agent, PVP.
Using the ITO-NPs, we fabricated ITO target. In Fig. 4a, 2-in. sized ITO target was well fabricated, and its microstructure was highly dense as shown in Fig. 4b. In fact, sintered density of the ITO target was measured to be 7.126 g/cm3. Considering that theoretical density of ITO is 7.155 g/cm3, we calculated as follows; (7.127/7.155) × 100 = 99.61%. That is, we got an ITO target with density of 99.6%. The value is very high enough to commercialization. It is attributed to using the ITO-NPs of which size was 103.8 nm, as mentioned earlier. In fact, we experienced that relatively lower value of density (87.9%) was obtained when smaller sized (21 nm) ITO-NPs was used to make a target. In contrary, target density was improved to 96% when we used larger sized (103.8 nm) ITO-NPs. Then, using the 103.8 nm sized ITO-NPs, we achieved ITO target of which density is 99.6% by optimizing the sintering conditions. It is reported that [16] the driving force of sintering leads to the reduction of total surface energy in system. Solid sintering can be divided into three stages initial-stage sintering, mid-stage sintering and final-stage sintering. The interfaces, namely “necks”, are formed among raw powders in the initial-stage sintering. Grain growth and pore connection occur simultaneously in the middle-stage sintering. In final-stage sintering, the pores become isolated while grain boundaries are linked each other and grains grow rapidly, so that the densification rate decreases evidently.” From the behavior, it is supposed that larger sized ITO particles reduce pore size by decreasing ITO grain growth rate leading to improvement of generation and isolation of pores. Thus, it is supposed that target density was enhanced by using larger-sized ITO-NPs.
The microstructures of ITO layers coated by using sputtering method are shown in Fig. 5a–d, respectively. Thickness of the ITO layers were 130, 180, 250, and 350 nm, respectively. It was observed that grain was grown when the thickness was raised. In Fig. 6a, their sheet resistances were 266.5, 114.1, 47.1, and 29.5 Ω/sq. respectively. The sheet resistance decreased as the film thickness increased. It is owing to thickness dependence of metallic layer [17]. The sheet resistance is determined using the simple equation;
$$ {\text{R}}_{\text{s}} = \uprho / {\text{t}} $$
(4)
where Rs, ρ, and t are sheet resistance, resistivity, and thickness, respectively. From the Eq. (4), assuming that the resistivity is constant, the sheet resistance is reversely proportional to film thickness. Whereas, in Fig. 6b, their optical transmittances at 550 nm (T550) were 82.9, 83.9, 81.3, and 82.3%, respectively. It is attributed to clearly coated layer from highly densified ITO target. Those T550 values are lower than that of commercial TCEs. We suggest that it is attributed to amorphous structured TCE layer just after deposition without heat-treatment. Thus, further work is to improve optical and electrical properties by enhancing physical properties of ITO film by, for example, optimizing heat-treatment conditions (temperature, environments), etc. Although further works are being done to improve ITO layers, ITO-NPs reused from ITO target scraps is feasible to apply to make sputtering target for TCEs.