- Email: [email protected]
Preparation and sintering of indium–gallium–zinc oxide ceramics with different zinc oxide contents
Journal of the European Ceramic Society 35 (2015) 3893–3902
Contents lists available at www.sciencedirect.com
Journal of the European Ceramic Society journal homepage: www.elsevier.com/locate/jeurceramsoc
Preparation and sintering of indium–gallium–zinc oxide ceramics with different zinc oxide contents Ming-Wei Wu a,∗ , Shih-Hsien Chang a , Wei-Ming Chaung b , Hung-Shang Huang c a b c
Department of Materials and Mineral Resources Engineering, National Taipei University of Technology, Taipei 10608, Taiwan, ROC Department of Materials Science and Engineering, National Formosa University, Huwei, Yunlin 63201, Taiwan, ROC New Materials Research & Development Department, China Steel Corporation, Kaohsiung 81233, Taiwan, ROC
a r t i c l e
i n f o
Article history: Received 16 February 2015 Received in revised form 26 June 2015 Accepted 27 June 2015 Available online 15 July 2015 Keywords: IGZO Sputtering target Sintering Microstructure Electrical properties
a b s t r a c t Amorphous IGZO film has been extensively used as the channel layer of thin-film transistors. To investigate the IGZO sputtering targets, the effects of sintering temperatures on the sintering, microstructure, and electrical properties of IGZO ceramics with In2 O3 :Ga2 O3 :ZnO mole percentages of 1:1:1 (IGZO-111) and 1:1:2 (IGZO-112) were studied. In IGZO-111 ceramics, the In2 O3 and ZnGa2 O4 phases are completely replaced by In2 Ga2 ZnO7 and InGaZnO4 phases when the sintering temperature is increased from 1300 ◦ C to 1400 ◦ C. Moreover, the crystal structure of IGZO-112 ceramic is a single phase of InGaZnO4 , and no phase transformation occurs between 1200 ◦ C and 1500 ◦ C. The optimum relative densities of IGZO-111 and IGZO-112 ceramics are 99.8% and 99.0%, respectively. After 1500 ◦ C sintering, the resistivities of IGZO111 and IGZO-112 ceramics are 1.5 × 10−3 cm and 2.5 × 10−3 cm, respectively. The properties of IGZO ceramics are comparable to those of AZO and GZO ceramics reported in the literature. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Amorphous indium–gallium–zinc-oxide (IGZO) film exhibits outstanding electrical performance and visible transparency, so it is widely used as the channel material of thin-film transistors [1–9]. Among the deposition techniques of thin films, magnetron sputtering is a dominant preparation method due mainly to its low coating temperature, high coating rate, good quality of the film, and high feasibility for large-area deposition [10–12]. The effects of various sputtering parameters, particularly the oxygen partial pressure, on the properties of IGZO films have been the focus of previous research [1–9]. Recently, several studies [13–21] have demonstrated that the characteristics of the sputtering targets for various oxide thin films play important roles on the preparation and the performances of sputtered films. Thus, in addition to the sputtering parameters, the sputtering target is an important factor in determining the properties of oxide films. To optimize the performances of sputtered films, the correlations between target performance and film properties must be thoroughly clarified. The characteristics of the oxide targets, including the sintered density [13,14], microstructural uniformity [15], stoichiometry [17–19], and electrical properties [20,21], have been
∗ Corresponding author. Fax: +886 2 27317185. E-mail addresses: [email protected], [email protected] (M.-W. Wu). http://dx.doi.org/10.1016/j.jeurceramsoc.2015.06.029 0955-2219/© 2015 Elsevier Ltd. All rights reserved.
found to obviously affect both the sputtering process and the film properties. A tin-doped indium oxide (ITO) target with higher sintered density and more homogeneous SnO2 distribution can suppress the arcing phenomenon and nodule generation during sputtering [14,15]. Furthermore, Minami et al. [20] found that using aluminum-doped zinc oxide (AZO) targets with lower resistivity can increase the deposition rate and lower the arcing counts. In addition to the sputtering process, the properties of sputtering targets also affect the film performances significantly. Using an ITO target with a higher sintered density can result in a film with lower resistivity [13]. Neves et al. [17] found that a ZnO target produced with non-stoichiometric nanometer powder can generate a ZnO film with lower resistivity. Furthermore, Minami et al. [20] and Huang et al. [21] investigated the effects of the electrical properties of AZO targets on the performances of AZO films. These two studies clearly demonstrated that an AZO target with lower resistivity can result in a more uniform film having lower resistivity. In addition, Yamada et al. [18] demonstrated that using Ti2 O3 :Nb and TiO2 :Nb ceramic targets instead of a Ti:Nb metallic target further facilitates the production of a niobium-doped titanium oxide (TNO) film with excellent electrical properties under a wide process window of oxygen pressure. On the other hand, radio-frequency (RF) power, whose coating rate is lower than that of direct-current (DC) power, is usually used for the sputtering of various oxide films due to the high intrinsic resistivities of oxide targets [6]. However, DC power has been
3894
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
widely applied to deposit various transparent conductive oxide (TCO) films because of the low resistivities of their targets [20–23]. Fang et al. [22] showed that an AZO target with a resistivity of 0.2 cm can be used to manufacture an AZO film by DC sputtering. TNO and TiO2 films can also be prepared by DC sputtering using TiO2−x targets with resistivities of about 0.3 cm [23,24]. Therefore, it is clear that the electrical properties of TCO targets affect the performances of the films, the stability of the sputtering process, and the feasibility of DC sputtering. Given the importance of sputtering targets, the processes and properties of sputtering targets for TCO films have been extensively investigated, particularly those of AZO and gallium-doped zinc oxide (GZO) [20,21,25–37]. Unfortunately, the process and characteristics of IGZO targets have been studied relatively little [3,4]. These unresolved questions prevent IGZO films and their devices from being optimized. Lo and Hsieh [3] studied the fabrication of an IGZO (In2 O3 :Ga2 O3 :ZnO = 1:1:2 mol%) target and reported that the crystal structure and relative density of the target are a single InGaZnO4 phase and 93%, respectively. Lee et al. [4] investigated the preparations of IGZO targets with atomic ratios (In:Ga:Zn) of 2:2:1, 1:1:1, and 1:1:2 and the correlations between the compositions and crystal structures of IGZO powders. For IGZO powders with an atomic ratio (In:Ga:Zn) of 2:2:1, the main In2 O3 phase decreased and the ZnGa2 O4 phase increased as the calcination temperature was increased from 730 ◦ C to 1030 ◦ C. When the atomic ratio of Zn is gradually increased, the amounts of InGaZnO4 and InGaZnO6 phases increase and those of In2 O3 and ZnGa2 O4 phases decrease. The above findings indicate the complexity of the microstructural changes during sintering of the IGZO powders. The objective of this study was thus to clarify the influences of sintering temperature (1200–1500 ◦ C) on the sintering, microstructures, and electrical properties of IGZO targets with different ZnO contents.
2. Experimental procedure In2 O3 , Ga2 O3 , and ZnO powders with median sizes of 0.1 m, 0.1 m, and 0.4 m, respectively, were used to produce
the IGZO ceramic targets. IGZO ceramics with molar fractions (In2 O3 :Ga2 O3 :ZnO) of 1:1:1 and 1:1:2 were investigated because these two fractions are the predominant compositions of IGZO films [1–9]. In this study, these two IGZO ceramics were designated as IGZO-111 and IGZO-112. To prepare the IGZO ceramic slurry with a solid content of 25 vol%, a 0.5 wt% dispersant of ammonium polyacrylate was first added into distilled water. The In2 O3 , Ga2 O3 , and ZnO powders were added into the aqueous solution, and then the powder slurry was ball milled with ZrO2 grinding balls for three hours. The ball to powder ratio (BPR) of ball milling was 3:1. Afterwards, a 0.5 wt% binder of polyacrylic emulsion was added into the slurry and the slurry was ball milled for one additional hour. The previous slurry was subsequently spray-dried in 140 ◦ C air using a spray dryer (L-8, Ohkawara Kakohki Co., Yokohama, Japan). The IGZO spray-dried granules were spherical and ranged in size from 10 to 40 m. To prepare the green compact, the IGZO spraydried granules were uniaxially compacted at a pressure of 150 MPa into disks of 12.5 mm diameter and 4.5 mm thickness. For binder removal, the green compacts were heated at 5 ◦ C/min to 600 ◦ C and held for 30 min in air. Afterwards, the specimens were directly heated at 10 ◦ C/min to the sintering temperatures (1200 ◦ C, 1300 ◦ C, 1400 ◦ C, and 1500 ◦ C) and were sintered for three hours in air, followed by furnace cooling. Archimedes’ method was used to evaluate the sintered densities as a function of sintering temperature. A dilatometer (DIL 402C, NETZSCH, Selb, Germany) was used to identify the amount and rate of the dimensional change during heating and sintering. To understand the evaporation behavior during high temperature sintering, thermogravimetric analysis (TGA, STA 449 F3, NETZSCH, Selb, Germany) and weight loss after sintering were measured. The sintered specimens were then fractured and thermally etched at 1100 ◦ C for one hour. The etched fracture surfaces were observed by a field-emission SEM (LEO-1530, Zeiss, Oberhochem, Germany). The fracture surfaces were observed to collect valuable information on the three-dimensional distribution of pores and grains in the material. To understand the distributions of different phases, the cross-section of sintered specimens were sampled, ground, polished, thermally etched at 1100 ◦ C for one hour, and then examined
Fig. 1. X-ray diffraction patterns of IGZO-111 ceramics sintered at (a) 1200 ◦ C and (b) 1400 ◦ C.
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
3895
by SEM in SE and BSE modes. Quantitative metallography was used to estimate the average grain sizes [38]. The crystal structures of the IGZO ceramics as a function of sintering temperature were analyzed using an X-ray diffractometer (D8, Bruker, Karlsruhe, Germany) with Cu K␣ radiation. Furthermore, the four-point probe method was applied to investigate the influences of chemical stoichiometry and sintering temperature on the resistivities of IGZO ceramics using a source meter (2400, Keithely Instruments Inc., OH, USA).
and 1500 ◦ C, as shown in Fig. 3. Nevertheless, the porosity of IGZO-112 ceramic sintered at 1300 ◦ C was still high. When the sintering temperature was further increased to 1400 ◦ C or 1500 ◦ C, the densification of IGZO-112 ceramic was highly activated, and low porosity was obtained. It should be noted that the porosities of the two ceramics were increased slightly by raising the sintering temperature from 1400 ◦ C to 1500 ◦ C. This phenomenon could be attributed to the volatilization of ZnO-based ceramics at high temperatures [29,31,32,36,39,40]. The weight losses of two IGZO ceramics sintered at various temperatures are discussed in Section 3.4. To further clarify the distributions of In2 Ga2 ZnO7 and InGaZnO4 phases in the IGZO-111 ceramics, the cross-section of the IGZO111 ceramics sintered at 1400 ◦ C and 1500 ◦ C were examined and are shown in Figs. 5 and 6, respectively. Observation of the crosssection microstructure using the BSE mode clearly indicated that a dispersion phase, shown by the dark contrast, was distributed in a matrix phase, shown by the bright contrast. In2 Ga2 ZnO7 and InGaZnO4 phases should appear as bright and dark contrasts, respectively, in the BES images because the average atomic number of In2 Ga2 ZnO7 is higher than that of InGaZnO4 . Moreover, as shown in Fig. 1(b), the diffraction intensity of In2 Ga2 ZnO7 phase is obviously higher than that of InGaZnO4 . These findings demonstrate that the dispersion phase of InGaZnO4 was distributed in the matrix phase of In2 Ga2 ZnO7 in the IGZO-111 ceramics sintered at 1400 ◦ C and 1500 ◦ C. Furthermore, Figs. 5 and 6 indicate that raising the sintering temperature from 1400 ◦ C to 1500 ◦ C increased the volume percentage of the InGaZnO4 phase in the IGZO-111 ceramic. The stability of the InGaZnO4 phase may be higher than that of In2 Ga2 ZnO7 , and the amount of the InGaZnO4 phase was thus increased at the expense of In2 Ga2 ZnO7 . The cross-section microstructures of IGZO112 ceramics sintered at 1400 ◦ C and 1500 ◦ C were also observed using the BSE mode, as shown in Fig. 7. The XRD and BSE results demonstrate that the single phase of InGaZnO4 had been generated at 1200 ◦ C and remained unchanged even after sintering at 1500 ◦ C.
3. Results and discussion
3.3. Densification
3.1. X-Ray diffraction patterns
Fig. 8 shows the sintered densities of two ceramics as a function of sintering temperature. The results indicate that the sintered densities of IGZO-111 ceramics peaked at 6.55 g/cm3 at 1300 ◦ C and then gradually declined to 6.40 g/cm3 at 1500 ◦ C. The sintered densities of IGZO-112 ceramics were improved from 3.86 g/cm3 to 6.32 g/cm3 by increasing the sintering temperature from 1200 ◦ C to 1400 ◦ C. However, the sintered density was slightly decreased to 6.30 g/cm3 as the sintering temperature was further increased to 1500 ◦ C. The trend of sintered densities of IGZO-112 ceramics was identical to that of porosity (Fig. 4). In contrast, in the IGZO111 ceramics, the trend of sintered density did not match that of porosity (Fig. 3). This inconsistency could be attributed to the phase transformation of IGZO-111 ceramic during sintering, as shown in Fig. 1. When the sintering temperature was increased from 1300 ◦ C to 1400 ◦ C, the In2 O3 and ZnGa2 O4 phases were completely replaced by the In2 Ga2 ZnO7 and InGaZnO4 phases. The theoretical densities of In2 O3 and ZnGa2 O4 are 7.119 g/cm3 (JCPDS 89-4595) and 6.168 g/cm3 (JCPDS 86-0410), respectively. Thus, the theoretical densities of IGZO-111 ceramics sintered at 1200 ◦ C and 1300 ◦ C should be 6.618 g/cm3 . However, the theoretical densities of IGZO111 ceramics sintered at 1400 ◦ C and 1500 ◦ C were correlated with the volumetric percentages of In2 Ga2 ZnO7 and InGaZnO4 phases. The theoretical densities of In2 Ga2 ZnO7 and InGaZnO4 are 6.494 g/cm3 (JCPDS 38-1097) and 6.380 g/cm3 (JCPDS 38-1104), respectively. Based on quantitative metallography of the crosssection microstructure, the percentage of InGaZnO4 phase in the
Fig. 2. X-ray diffraction patterns of IGZO-112 ceramics sintered at 1200 ◦ C and 1500 ◦ C.
The crystal structures of IGZO-111 and IGZO-112 ceramics are shown in Figs. 1 and 2, respectively. Several phases, including In2 O3 , ZnGa2 O4 , InGaZnO4 , and InGaZnO6 , had previously been found in IGZO ceramics [3,4]. Fig. 1(a) shows that In2 O3 and ZnGa2 O4 were present in the IGZO-111 ceramic sintered at 1200 ◦ C, indicating that Ga2 O3 and ZnO reacted to ZnGa2 O4 entirely. These two crystal structures did not change after 1300 ◦ C sintering. However, after sintering at 1400 ◦ C, In2 Ga2 ZnO7 and InGaZnO4 phases completely replaced the previous In2 O3 and ZnGa2 O4 ones, as shown in Fig. 1(b). When the sintering temperature was further increased to 1500 ◦ C, In2 Ga2 ZnO7 and InGaZnO4 phases still existed in the IGZO111 ceramic. On the other hand, the crystal structure of IGZO-112 ceramics sintered at 1200 ◦ C and 1500 ◦ C was a single InGaZnO4 phase, as demonstrated in Fig. 2. No phase transformation occurred in the IGZO-112 ceramics between 1200 ◦ C and 1500 ◦ C. 3.2. Microstructure The thermally etched fracture surfaces of IGZO-111 and IGZO112 ceramics sintered at various temperatures are shown in Figs. 3 and 4, respectively. After sintering at 1200 ◦ C, the two ceramics were still highly porous, indicating inferior sintering densification. However, the densification of IGZO-111 was much more obvious than that of IGZO-112 after 1300 ◦ C sintering. Only minor pores existed in the IGZO-111 ceramics sintered at 1300 ◦ C, 1400 ◦ C,
3896
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
Fig. 3. The thermally etched fracture surfaces of IGZO-111 ceramics sintered at (a) 1200 ◦ C, (b) 1300 ◦ C, (c) 1400 ◦ C, and (d) 1500 ◦ C.
Fig. 4. The thermally etched fracture surfaces of IGZO-112 ceramics sintered at (a) 1200 ◦ C, (b) 1300 ◦ C, (c) 1400 ◦ C, and (d) 1500 ◦ C.
Fig. 5. The cross-section microstructure of the IGZO-111 ceramic sintered at 1400 ◦ C. (a) SE image (b) BSE image.
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
3897
Fig. 6. The cross-section microstructure of the IGZO-111 ceramic sintered at 1500 ◦ C. (a) SE image (b) BSE image.
Fig. 7. The cross-section microstructures of the IGZO-112 ceramics sintered at (a) 1400 ◦ C and (b) 1500 ◦ C.
IGZO-111 ceramics sintered at 1400 ◦ C and 1500 ◦ C were 8 vol% and 30 vol%, respectively. Therefore, the theoretical densities of IGZO111 sintered at 1400 ◦ C and 1500 ◦ C should be 6.485 g/cm3 and 6.460 g/cm3 , respectively. On the other hand, the crystal structure
of IGZO-112 ceramic was a single InGaZnO4 phase, and no phase transformation occurred between 1200 ◦ C and 1500 ◦ C. The theoretical density of IGZO-112 sintered at various temperatures was thus regarded as 6.380 g/cm3 .
Fig. 8. The sintered densities of IGZO-111 and IGZO-112 ceramics as a function of sintering temperature.
3898
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
Fig. 9. The relative densities of IGZO-111 and IGZO-112 ceramics as a function of sintering temperature.
To investigate the actual porosities of the two ceramics as a function of sintering temperature, the sintered densities in Fig. 8 were divided by the theoretical densities to calculate the relative densities, as plotted in Fig. 9. The results indicate that the relative densities of IGZO-111 and IGZO-112 ceramics sintered at 1200 ◦ C were only 83.6% and 63.5%, respectively. The densification was apparently activated by raising the sintering temperature from 1200 ◦ C to 1400 ◦ C. The relative densities of these two ceramics peaked at 1400 ◦ C and then slightly decreased at 1500 ◦ C. The results of the relative densities corresponded well to those of the fracture surfaces, as can be seen from comparing Figs. 3, 4, and 9. The highest relative densities of IGZO-111 and IGZO-112 ceramics were 99.8% and 99.0%, respectively, in this study. 3.4. Thermal analysis To further understand the sintering behaviors of these two ceramics, a dilatometer and TGA were also used to investigate the dimensional change and weight loss, respectively, during sintering. For the dilatometer test, the two ceramics were heated at 10 ◦ C/min to 1400 ◦ C without holding in air. The amount and rate of the dimensional change are presented in Fig. 10. The results indicated normal thermal expansion at temperatures of 25–800 ◦ C. However, an apparent swelling phenomenon was observed in the two ceramics, particularly IGZO-112. The maximum linear expansions of the IGZO-111 and IGZO-112 ceramics were 1.4% at 1020 ◦ C and 7.9% at 1150 ◦ C, respectively. Moreover, the maximum expansion rates of the IGZO-111 and IGZO-112 ceramics were 0.03%/min at 880 ◦ C and 0.38%/min at 960 ◦ C, respectively. The extraordinary expansion of the IGZO-112 ceramic could be attributed to the difference in volume between the original reactants (In2 O3 , Ga2 O3 , and ZnO) and the final products (InGaZnO4 ) during the heating period. After heating to 1400 ◦ C without holding, the linear shrinkages of the IGZO-111 and IGZO-112 ceramics were 10.3% and 0.8%, respectively. These above findings demonstrate that the IGZO-111
ceramic densified faster than the IGZO-112 ceramic, and the relative density of the IGZO-112 ceramic was thus sufficiently lower than that of the IGZO-111 ceramic, particularly at temperatures below 1300 ◦ C, as shown in Figs. 9 and 10. However, the shrinkage rate of the IGZO-112 ceramic increased greatly above 1150 ◦ C, and the maximum shrinkage rate of the IGZO-112 ceramic was even higher than that of the IGZO-111 ceramic, as shown in Fig. 10(b). When these two ceramics were sintered at 1400 ◦ C for three hours, the difference in the relative densities of the two ceramics was minor. Due to the obvious evaporation phenomenon during hightemperature sintering of ZnO-based ceramics, the weight losses of the two IGZO ceramics sintered at various temperatures were studied and are displayed in Fig. 11 [29,31,32,36,39,40]. The results showed no clear difference between the two ceramics when the sintering temperature was lower than 1500 ◦ C. However, after 1500 ◦ C sintering, the weight losses of IGZO-111 and IGZO-112 ceramics were 4.7% and 5.1%, respectively. To differentiate the evaporation behaviors of the original In2 O3 , Ga2 O3 , and ZnO powders used in this study, these three powders were loosely packed in a crucible and then examined by TGA. The TGA specimens were heated at 10 ◦ C/min to 1500 ◦ C and held for one hour in air. Fig. 12 indicates that the weight loss of ZnO powder was much higher than that of Ga2 O3 , which was followed by In2 O3 . The ZnO powder exhibited a weight loss as high as 42% after holding at 1500 ◦ C for one hour. When the sintering temperature was increased from 1400 ◦ C to 1500 ◦ C, the relative densities of the two IGZO ceramics were decreased due mainly to the evaporation of ZnO, as shown in Fig. 9. 3.5. Grain size Fig. 13 reveals the average grain sizes of two IGZO ceramics as a function of the sintering temperature. The main phases of IGZO-111 (In2 Ga2 ZnO7 ) and IGZO-112 (InGaZnO4 ) were used to calculate the average grain sizes. When the sintering temperature
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
3899
Fig. 10. The dilatometric curves of IGZO ceramics as a function of sintering temperature. (a) Dimensional change (b) dimensional change rate.
was increased from 1200 ◦ C to 1500 ◦ C, the grain sizes of IGZO-111 and IGZO-112 ceramics increased from 0.4 m to 9.5 m and from 0.3 m to 12.7 m, respectively. The grain sizes of IGZO-111 and IGZO-112 ceramics sintered at 1400 ◦ C were 6.0 m and 10.9 m, respectively. The results also show that the grain sizes of IGZO112 ceramics sintered at 1200 ◦ C and 1300 ◦ C were smaller than
those of IGZO-111. This phenomenon could be attributed to the increased number of pores, which dragged the grain boundary and thus inhibited the grain growth, in the IGZO-112 ceramics. However, the trend reversed after 1400 ◦ C sintering, and the grains of IGZO-111 ceramics were larger than those of IGZO-112. After sintering at 1400 ◦ C, these two ceramics contained only about 1 vol%
3900
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
Fig. 11. The weight losses of IGZO ceramics as a function of sintering temperature.
pores, and the dragging effect contributed by the pores could be excluded. When these two ceramics were sintered at 1400 ◦ C or 1500 ◦ C, the crystal structures of IGZO-111 and IGZO-112 ceramics were two phases, In2 Ga2 ZnO7 and InGaZnO4 , and a single
phase of InGaZnO4 , respectively. According to the Zener effect, the presence of InGaZnO4 dispersion in the In2 Ga2 ZnO7 matrix could inhibit the grain growth and reduce the grain size in the IGZO-111 ceramics.
Fig. 12. The TGA curves of the original In2 O3 , Ga2 O3 , and ZnO powders used in this study.
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
3901
Fig. 13. The average grain sizes of IGZO ceramics as a function of sintering temperature.
3.6. Electrical properties To assess the electrical properties of the IGZO ceramics prepared in this study, the resistivities were tested using the four-point probe method. The resistivities of the two IGZO ceramics sintered at 1200 ◦ C and 1300 ◦ C could not be measured by the sourcemeter used in this study because they exceeded the range of the instrument. However, after sintering at 1400 ◦ C or 1500 ◦ C, the conductance of IGZO ceramics was much improved, and their resistivities could be obtained. The resistivities of IGZO-111 and IGZO-112 ceramics sintered at 1400 ◦ C were 4.9 × 10−3 cm and 4.5 × 10−3 cm, respectively. When the sintering temperature was increased to 1500 ◦ C, the resistivities of IGZO-111 and IGZO-112 ceramics were further improved to 1.5 × 10−3 cm and 2.5 × 10−3 cm, respectively, indicating that both In2 Ga2 ZnO7 and InGaZnO4 phases were highly conductive. In addition, the resistivities of amorphous IGZO-111 and IGZO-112 sputtered films reported in the literature range from 5 × 10−3 cm to 104 cm and 4 × 10−3 cm to 105 cm, respectively, depending on the sputtering parameters [2,4,5,7,8]. The amorphous structure of IGZO-111 film can be retained even after annealing at 600 ◦ C in argon [9]. These results show that the resistivities of In2 Ga2 ZnO7 and InGaZnO4 crystal phases were generally lower than those of amorphous IGZO structure. In the literature, the processes and properties of AZO and GZO ceramic targets sintered in air have been extensively investigated [21,25,28–36]. The relative densities, grain sizes, and resistivities of AZO targets sintered in air range from 98.6% to 99.8%, from 2.7 m to 5.0 m, and from 8.9 × 10−4 cm to 5 × 10−1 cm, respectively [21,25,28–31]. Moreover, the relative densities, grain sizes, and resistivities of GZO targets sintered in air range from 96.0% to 99.3%, from 2.0 m to 4.5 m, and from 1.4 × 10−3 cm to 1 cm, respectively [31–36]. This study demonstrates that the various properties, including the relative densities, grain sizes, and resistivities, of IGZO
ceramics sintered in air are comparable to those of AZO and GZO ceramics. The resistivities of IGZO ceramics could approach the lowest resistivities of AZO and GZO ceramics sintered in air. 4. Conclusions The influences of chemical stoichiometry and sintering temperature on the sintering, microstructure, and electrical properties of IGZO ceramics for sputtering targets were studied. The findings can be summarized as follows. 1. The crystal structures of IGZO-111 ceramics sintered at 1200 ◦ C or 1300 ◦ C were a mixture of In2 O3 and ZnGa2 O4 phases. When the sintering temperature was increased to 1400 ◦ C, the previous In2 O3 and ZnGa2 O4 phases were completely replaced by In2 Ga2 ZnO7 and InGaZnO4 phases. The InGaZnO4 dispersion phase was distributed in the In2 Ga2 ZnO7 matrix phase of IGZO111 ceramics sintered at 1400 ◦ C or 1500 ◦ C. However, the crystal structure of IGZO-112 ceramic was a single InGaZnO4 phase, and no phase transformation occurred between 1200 ◦ C and 1500 ◦ C. 2. The highest relative densities of IGZO-111 and IGZO-112 ceramics were 99.8% and 99.0%, respectively, in this study. The relative densities of the two IGZO ceramics peaked at 1400 ◦ C and then slightly decreased at 1500 ◦ C due mainly to the evaporation of ZnO. 3. Both ceramics exhibited an obvious swelling phenomenon during heating, particularly IGZO-112. The IGZO-111 ceramic densified faster than the IGZO-112 ceramic when the sintering temperature was lower than 1400 ◦ C. However, the shrinkage rate of the IGZO-112 ceramic increased greatly above 1150 ◦ C, and the difference in the relative densities of the two ceramics sintered at 1400 ◦ C for three hours was minor. 4. After sintering at 1500 ◦ C, the resistivities of IGZO111 and IGZO-112 ceramics could achieve 1.5 × 10−3 cm and 2.5 × 10−3 cm, respectively, showing that both In2 Ga2 ZnO7 and InGaZnO4 phases were highly conductive. In addition, the
3902
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
resistivities of In2 Ga2 ZnO7 and InGaZnO4 crystal phases were generally better than those of the amorphous structure in the IGZO film reported in the literature. Acknowledgements The authors thank the Ministry of Science and Technology of the Republic of China for support and funding under contract number NSC 99-2218-E-150-046. Gratitude is also extended to Dr. FangCheng Chou and Dr. Guo-Jiun Shu in the Center for Condensed Matter and Science of National Taiwan University for the TGA experiments. References [1] H. Kumomi, K. Nomura, T. Kamiya, H. Hosono, Amorphous oxide channel TFTs, Thin Solid Films 516 (2008) 1516–1522. [2] T. Aoi, N. Oka, Y. Sato, R. Hayashi, H. Kumomi, Y. Shigesato, DC sputter deposition of amorphous indium–gallium–zinc-oxide (a-IGZO) films with H2 O introduction, Thin Solid Films 518 (2010) 3004–3007. [3] C.C. Lo, T.E. Hsieh, Preparation of IGZO sputtering target and its applications to thin-film transistor devices, Ceram. Int. 38 (2012) 3977–3983. [4] Y.S. Lee, Z.M. Dai, C.I. Lin, H.C. Lin, Relationships between the crystalline phase of an IGZO target and electrical properties of a-IGZO channel film, Ceram. Int. 38 (2012) S595–99. [5] M.J. Gadre, T.L. Alford, Highest transmittance and high-mobility amorphous indium gallium zinc oxide films on flexible substrate by room-temperature deposition and post-deposition anneals, Appl. Phys. Lett. 99 (2011) 051901. [6] J. Yao, L. Gong, L. Xie, S. Zhang, Comparison of the electrical and optical properties of direct current and radio frequency sputtered amorphous indium gallium zinc oxide films, Thin Solid Films 527 (2013) 21–25. [7] B.D. Ahn, H.S. Shin, G.H. Kim, J.S. Park, H.J. Kim, A novel amorphous InGaZnO thin film transistor structure without source/drain layer deposition, Jpn. J. Appl. Phys. 48 (2009) 03B019. [8] C.H. Jung, H.I. Kang, D.H. Yoon, The electrical, optical, and structural properties of amorphous indium gallium zinc oxide films and channel thin-film transistors, Solid-State Electron. 79 (2013) 125–129. [9] H. Kim, J.Y. Moon, Y.W. Heo, H.S. Lee, Effect of thermal annealing on Ni/Au contact to a-InGaZnO films deposited by dc sputtering, Thin Solid Films (2010) 6348–6351. [10] X.B. Zhang, Z.L. Pei, J. Gong, C. Sun, Investigation on the electrical properties and inhomogeneous distribution of ZnO:Al thin films prepared by DC magnetron sputtering at low deposition temperature, J. Appl. Phys. 101 (2007) 014910. [11] W. Yang, Z. Liu, D.L. Peng, F. Zhang, H. Huang, Y. Xie, Z. Wu, Room-temperature deposition of transparent conducting Al-doped ZnO, Appl. Surf. Sci. 255 (2009) 5669–5673. [12] R. Cebulla, R. Wendt, K. Ellmer, Al-doped zinc oxide films deposited by stimultaneous RF and DC excitation of a magnetron plasma: relationships between plasma parameters and structural and electrical film properties, J. Appl. Phys. 83 (1998) 1087–1095. [13] K. Utsumi, O. Matsunaga, T. Takahata, Low resistivity ITO film prepared using the ultra high density ITO target, Thin Solid Films 334 (1998) 30–34. [14] T. Omata, M. Kita, H. Okada, S. Otsuka-Yao-Matsuo, N. Ono, H. Ikawa, Characterization of indium–tin oxide sputtering targets showing various densities of nodule formation, Thin Solid Films 503 (2006) 22–28. [15] K. Nakashima, Y. Kumahara, Effect of tin oxide dispersion on nodule formation in ITO sputtering, Vacuum (2002) 221–226. [16] K.L. Ying, T.E. Hsieh, Y.F. Hsieh, Colloidal dispersion of nano-scale ZnO powders using amphibious and anionic polyelectrolytes, Ceram. Int. 35 (2009) 1165–1171. [17] N. Neves, R. Barros, E. Antunes, I. Ferreira, J. Calado, E. Fortunato, R. Martins, Sintering behavior of nano- and micro-sized ZnO powder targets for rf magnetron sputtering applications, J. Am. Ceram. Soc. 95 (2012) 204–210.
[18] N. Yamada, T. Hitosugi, J. Kasai, N.L.H. Hoang, D. Nakao, Y. Hirose, T. Shimada, T. Hasegawa, Transparent conducting Nb-doped anatase TiO2 (TNO) thin films sputtered from various oxide targets, Thin Solid Films 518 (2010) 3101–3104. [19] T. Kubart, J. Jensen, T. Nyberg, L. Liljeholm, D. Depla, S. Berg, Influence of the target composition on reactively sputtered titanium oxide films, Vacuum 83 (2009) 1295–1298. [20] T. Minami, J.I. Oda, J.I. Nomoto, T. Miyata, Effect of target properties on transparent conducting impurity-doped ZnO thin films deposited by DC magnetron sputtering, Thin Solid Films 519 (2010) 385–390. [21] H.S. Huang, H.C. Tung, C.H. Chiu, I.T. Hong, R.Z. Chen, J.T. Chang, H.K. Lin, Highly conductive alumina-added ZnO ceramic target prepared by reduction sintering and its effects on the properties of deposited thin films by direct current magnetron sputtering, Thin Solid Films 518 (2010) 6071–6075. [22] G. Fang, D. Li, B.L. Yao, Fabrication and vacuum annealing of transparent conductive AZO thin films prepared by DC magnetron sputtering, Vacuum 68 (2003) 363–372. [23] Y. Sato, H. Akizuki, T. Kamiyama, Y. Shigesato, Transparent conductive Nb-doped TiO2 films deposited by direct-current magnetron sputtering using a TiO2−x target, Thin Solid Films 516 (2008) 5758–5762. [24] H. Ohsaki, Y. Tachibana, A. Mitsui, T. Kamiyama, Y. Hayashi, High rate deposition of TiO2 by DC sputtering of the TiO2−x target, Thin Solid Films 392 (2001) 169–173. [25] M.W. Wu, D.S. Liu, Y.H. Su, The densification, microstructure, and electrical properties of aluminum-doped zinc oxide sputtering target for transparent conductive oxide film, J. Eur. Ceram. Soc. 32 (2012) 3265–3275. [26] N. Neves, A. Lagoa, J. Calado, A.M.B. Rego, E. Fortunato, R. Martins, I. Ferreira, Al-doped ZnO nanostructured powders by emulsion detonation synthesis—improving materials for high quality sputtering targets manufacturing, J. Eur. Ceram. Soc. 34 (2014) 2325–2338. [27] J.R. Lee, T.J. Chung, S.H. Yang, G.S. Hong, K.S. Oh, Prevention of tapering in the tube-shaped sputtering target via initial heat treatment under external pressure, Ceram. Int. 41 (2015) 3677–3683. [28] C.P. Liu, G.R. Jeng, Properties of aluminum doped zinc oxide materials and sputtering thin films, J. Alloys Compd. 468 (2009) 343–349. [29] J. Zhang, W. Zhang, E. Zhao, H.J. Jacques, Study of high-density AZO ceramic target, Mater. Sci. Semicond. Process. 14 (2011) 189–192. [30] T. Wei, Y. Zhang, Y. Yang, R. Tan, P. Cui, W. Song, Effects of ZnAl2 O4 segregation in high temperature sintered Al-doped ZnO sputtering target on optical and electrical properties of deposited thin films, Surf. Coat. Technol. 221 (2013) 201–206. [31] J. Liu, W. Zhang, D. Song, Q. Ma, L. Zhang, H. Zhang, X. Ma, H. Song, Comparative study of the sintering process and thin film sputtering of AZO, GZO and AGZO ceramics targets, Ceram. Int. 40 (2014) 12905–12915. [32] J. Liu, W. Zhang, D. Song, Q. Ma, L. Zhang, H. Zhang, X. Ma, H. Song, Gallium-doped zinc oxide targets fabricated by sintering: impact of target quality on sputtered thin film properties, Mater. Sci. Semicond. Process. 27 (2014) 1–11. [33] C. Liu, X. Cai, Z. Zhang, L. Gao, C. Xiao, F. Shang, H. Yang, Effect of gallium doping on the properties of GZO target, Adv. Mater. Res. 779–780 (2013) 363–367. [34] J.P. Wiff, Y. Kinemuchi, K. Watari, Hall mobilities of Al- and Ga-doped ZnO polycrystals, Mater. Lett. 63 (2009) 2470–2472. [35] M.S. Jang, M.K. Ryu, M.H. Yoon, S.H. Lee, H.K. Kim, A. Onodera, S. Kojima, A study on the Raman spectra of Al-doped and Ga-doped ZnO ceramics, Curr. Appl. Phys. 9 (2009) 651–657. [36] M.W. Wu, P.H. Lai, C.H. Hong, F.C. Chou, The sintering behavior, microstructure, and electrical properties of gallium-doped zinc oxide ceramic targets, J. Eur. Ceram. Soc. 34 (2014) 3715–3722. [37] M.W. Wu, Y.C. Chen, The sintering trajectory and electrical properties of niobium-doped titania sputtering targets, J. Am. Ceram. Soc. 98 (2015) 163–170. [38] M.I. Mendelson, Average grain size in polycrystalline ceramics, J. Am. Ceram. Soc. 52 (1969) 443–446. [39] J. Han, P.Q. Mantas, A.M.R. Senos, Densification and grain growth of Al-doped ZnO, J. Mater. Res. 16 (2001) 459–468. [40] N. Neves, R. Barros, E. Antunes, J. Calado, E. Fortunato, R. Martins, I. Ferreira, Aluminum doped zinc oxide sputtering targets obtained from nanostructured powders: processing and application, J. Eur. Ceram. Soc. 32 (2012) 4381–4391.
Contents lists available at www.sciencedirect.com
Journal of the European Ceramic Society journal homepage: www.elsevier.com/locate/jeurceramsoc
Preparation and sintering of indium–gallium–zinc oxide ceramics with different zinc oxide contents Ming-Wei Wu a,∗ , Shih-Hsien Chang a , Wei-Ming Chaung b , Hung-Shang Huang c a b c
Department of Materials and Mineral Resources Engineering, National Taipei University of Technology, Taipei 10608, Taiwan, ROC Department of Materials Science and Engineering, National Formosa University, Huwei, Yunlin 63201, Taiwan, ROC New Materials Research & Development Department, China Steel Corporation, Kaohsiung 81233, Taiwan, ROC
a r t i c l e
i n f o
Article history: Received 16 February 2015 Received in revised form 26 June 2015 Accepted 27 June 2015 Available online 15 July 2015 Keywords: IGZO Sputtering target Sintering Microstructure Electrical properties
a b s t r a c t Amorphous IGZO film has been extensively used as the channel layer of thin-film transistors. To investigate the IGZO sputtering targets, the effects of sintering temperatures on the sintering, microstructure, and electrical properties of IGZO ceramics with In2 O3 :Ga2 O3 :ZnO mole percentages of 1:1:1 (IGZO-111) and 1:1:2 (IGZO-112) were studied. In IGZO-111 ceramics, the In2 O3 and ZnGa2 O4 phases are completely replaced by In2 Ga2 ZnO7 and InGaZnO4 phases when the sintering temperature is increased from 1300 ◦ C to 1400 ◦ C. Moreover, the crystal structure of IGZO-112 ceramic is a single phase of InGaZnO4 , and no phase transformation occurs between 1200 ◦ C and 1500 ◦ C. The optimum relative densities of IGZO-111 and IGZO-112 ceramics are 99.8% and 99.0%, respectively. After 1500 ◦ C sintering, the resistivities of IGZO111 and IGZO-112 ceramics are 1.5 × 10−3 cm and 2.5 × 10−3 cm, respectively. The properties of IGZO ceramics are comparable to those of AZO and GZO ceramics reported in the literature. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Amorphous indium–gallium–zinc-oxide (IGZO) film exhibits outstanding electrical performance and visible transparency, so it is widely used as the channel material of thin-film transistors [1–9]. Among the deposition techniques of thin films, magnetron sputtering is a dominant preparation method due mainly to its low coating temperature, high coating rate, good quality of the film, and high feasibility for large-area deposition [10–12]. The effects of various sputtering parameters, particularly the oxygen partial pressure, on the properties of IGZO films have been the focus of previous research [1–9]. Recently, several studies [13–21] have demonstrated that the characteristics of the sputtering targets for various oxide thin films play important roles on the preparation and the performances of sputtered films. Thus, in addition to the sputtering parameters, the sputtering target is an important factor in determining the properties of oxide films. To optimize the performances of sputtered films, the correlations between target performance and film properties must be thoroughly clarified. The characteristics of the oxide targets, including the sintered density [13,14], microstructural uniformity [15], stoichiometry [17–19], and electrical properties [20,21], have been
∗ Corresponding author. Fax: +886 2 27317185. E-mail addresses: [email protected], [email protected] (M.-W. Wu). http://dx.doi.org/10.1016/j.jeurceramsoc.2015.06.029 0955-2219/© 2015 Elsevier Ltd. All rights reserved.
found to obviously affect both the sputtering process and the film properties. A tin-doped indium oxide (ITO) target with higher sintered density and more homogeneous SnO2 distribution can suppress the arcing phenomenon and nodule generation during sputtering [14,15]. Furthermore, Minami et al. [20] found that using aluminum-doped zinc oxide (AZO) targets with lower resistivity can increase the deposition rate and lower the arcing counts. In addition to the sputtering process, the properties of sputtering targets also affect the film performances significantly. Using an ITO target with a higher sintered density can result in a film with lower resistivity [13]. Neves et al. [17] found that a ZnO target produced with non-stoichiometric nanometer powder can generate a ZnO film with lower resistivity. Furthermore, Minami et al. [20] and Huang et al. [21] investigated the effects of the electrical properties of AZO targets on the performances of AZO films. These two studies clearly demonstrated that an AZO target with lower resistivity can result in a more uniform film having lower resistivity. In addition, Yamada et al. [18] demonstrated that using Ti2 O3 :Nb and TiO2 :Nb ceramic targets instead of a Ti:Nb metallic target further facilitates the production of a niobium-doped titanium oxide (TNO) film with excellent electrical properties under a wide process window of oxygen pressure. On the other hand, radio-frequency (RF) power, whose coating rate is lower than that of direct-current (DC) power, is usually used for the sputtering of various oxide films due to the high intrinsic resistivities of oxide targets [6]. However, DC power has been
3894
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
widely applied to deposit various transparent conductive oxide (TCO) films because of the low resistivities of their targets [20–23]. Fang et al. [22] showed that an AZO target with a resistivity of 0.2 cm can be used to manufacture an AZO film by DC sputtering. TNO and TiO2 films can also be prepared by DC sputtering using TiO2−x targets with resistivities of about 0.3 cm [23,24]. Therefore, it is clear that the electrical properties of TCO targets affect the performances of the films, the stability of the sputtering process, and the feasibility of DC sputtering. Given the importance of sputtering targets, the processes and properties of sputtering targets for TCO films have been extensively investigated, particularly those of AZO and gallium-doped zinc oxide (GZO) [20,21,25–37]. Unfortunately, the process and characteristics of IGZO targets have been studied relatively little [3,4]. These unresolved questions prevent IGZO films and their devices from being optimized. Lo and Hsieh [3] studied the fabrication of an IGZO (In2 O3 :Ga2 O3 :ZnO = 1:1:2 mol%) target and reported that the crystal structure and relative density of the target are a single InGaZnO4 phase and 93%, respectively. Lee et al. [4] investigated the preparations of IGZO targets with atomic ratios (In:Ga:Zn) of 2:2:1, 1:1:1, and 1:1:2 and the correlations between the compositions and crystal structures of IGZO powders. For IGZO powders with an atomic ratio (In:Ga:Zn) of 2:2:1, the main In2 O3 phase decreased and the ZnGa2 O4 phase increased as the calcination temperature was increased from 730 ◦ C to 1030 ◦ C. When the atomic ratio of Zn is gradually increased, the amounts of InGaZnO4 and InGaZnO6 phases increase and those of In2 O3 and ZnGa2 O4 phases decrease. The above findings indicate the complexity of the microstructural changes during sintering of the IGZO powders. The objective of this study was thus to clarify the influences of sintering temperature (1200–1500 ◦ C) on the sintering, microstructures, and electrical properties of IGZO targets with different ZnO contents.
2. Experimental procedure In2 O3 , Ga2 O3 , and ZnO powders with median sizes of 0.1 m, 0.1 m, and 0.4 m, respectively, were used to produce
the IGZO ceramic targets. IGZO ceramics with molar fractions (In2 O3 :Ga2 O3 :ZnO) of 1:1:1 and 1:1:2 were investigated because these two fractions are the predominant compositions of IGZO films [1–9]. In this study, these two IGZO ceramics were designated as IGZO-111 and IGZO-112. To prepare the IGZO ceramic slurry with a solid content of 25 vol%, a 0.5 wt% dispersant of ammonium polyacrylate was first added into distilled water. The In2 O3 , Ga2 O3 , and ZnO powders were added into the aqueous solution, and then the powder slurry was ball milled with ZrO2 grinding balls for three hours. The ball to powder ratio (BPR) of ball milling was 3:1. Afterwards, a 0.5 wt% binder of polyacrylic emulsion was added into the slurry and the slurry was ball milled for one additional hour. The previous slurry was subsequently spray-dried in 140 ◦ C air using a spray dryer (L-8, Ohkawara Kakohki Co., Yokohama, Japan). The IGZO spray-dried granules were spherical and ranged in size from 10 to 40 m. To prepare the green compact, the IGZO spraydried granules were uniaxially compacted at a pressure of 150 MPa into disks of 12.5 mm diameter and 4.5 mm thickness. For binder removal, the green compacts were heated at 5 ◦ C/min to 600 ◦ C and held for 30 min in air. Afterwards, the specimens were directly heated at 10 ◦ C/min to the sintering temperatures (1200 ◦ C, 1300 ◦ C, 1400 ◦ C, and 1500 ◦ C) and were sintered for three hours in air, followed by furnace cooling. Archimedes’ method was used to evaluate the sintered densities as a function of sintering temperature. A dilatometer (DIL 402C, NETZSCH, Selb, Germany) was used to identify the amount and rate of the dimensional change during heating and sintering. To understand the evaporation behavior during high temperature sintering, thermogravimetric analysis (TGA, STA 449 F3, NETZSCH, Selb, Germany) and weight loss after sintering were measured. The sintered specimens were then fractured and thermally etched at 1100 ◦ C for one hour. The etched fracture surfaces were observed by a field-emission SEM (LEO-1530, Zeiss, Oberhochem, Germany). The fracture surfaces were observed to collect valuable information on the three-dimensional distribution of pores and grains in the material. To understand the distributions of different phases, the cross-section of sintered specimens were sampled, ground, polished, thermally etched at 1100 ◦ C for one hour, and then examined
Fig. 1. X-ray diffraction patterns of IGZO-111 ceramics sintered at (a) 1200 ◦ C and (b) 1400 ◦ C.
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
3895
by SEM in SE and BSE modes. Quantitative metallography was used to estimate the average grain sizes [38]. The crystal structures of the IGZO ceramics as a function of sintering temperature were analyzed using an X-ray diffractometer (D8, Bruker, Karlsruhe, Germany) with Cu K␣ radiation. Furthermore, the four-point probe method was applied to investigate the influences of chemical stoichiometry and sintering temperature on the resistivities of IGZO ceramics using a source meter (2400, Keithely Instruments Inc., OH, USA).
and 1500 ◦ C, as shown in Fig. 3. Nevertheless, the porosity of IGZO-112 ceramic sintered at 1300 ◦ C was still high. When the sintering temperature was further increased to 1400 ◦ C or 1500 ◦ C, the densification of IGZO-112 ceramic was highly activated, and low porosity was obtained. It should be noted that the porosities of the two ceramics were increased slightly by raising the sintering temperature from 1400 ◦ C to 1500 ◦ C. This phenomenon could be attributed to the volatilization of ZnO-based ceramics at high temperatures [29,31,32,36,39,40]. The weight losses of two IGZO ceramics sintered at various temperatures are discussed in Section 3.4. To further clarify the distributions of In2 Ga2 ZnO7 and InGaZnO4 phases in the IGZO-111 ceramics, the cross-section of the IGZO111 ceramics sintered at 1400 ◦ C and 1500 ◦ C were examined and are shown in Figs. 5 and 6, respectively. Observation of the crosssection microstructure using the BSE mode clearly indicated that a dispersion phase, shown by the dark contrast, was distributed in a matrix phase, shown by the bright contrast. In2 Ga2 ZnO7 and InGaZnO4 phases should appear as bright and dark contrasts, respectively, in the BES images because the average atomic number of In2 Ga2 ZnO7 is higher than that of InGaZnO4 . Moreover, as shown in Fig. 1(b), the diffraction intensity of In2 Ga2 ZnO7 phase is obviously higher than that of InGaZnO4 . These findings demonstrate that the dispersion phase of InGaZnO4 was distributed in the matrix phase of In2 Ga2 ZnO7 in the IGZO-111 ceramics sintered at 1400 ◦ C and 1500 ◦ C. Furthermore, Figs. 5 and 6 indicate that raising the sintering temperature from 1400 ◦ C to 1500 ◦ C increased the volume percentage of the InGaZnO4 phase in the IGZO-111 ceramic. The stability of the InGaZnO4 phase may be higher than that of In2 Ga2 ZnO7 , and the amount of the InGaZnO4 phase was thus increased at the expense of In2 Ga2 ZnO7 . The cross-section microstructures of IGZO112 ceramics sintered at 1400 ◦ C and 1500 ◦ C were also observed using the BSE mode, as shown in Fig. 7. The XRD and BSE results demonstrate that the single phase of InGaZnO4 had been generated at 1200 ◦ C and remained unchanged even after sintering at 1500 ◦ C.
3. Results and discussion
3.3. Densification
3.1. X-Ray diffraction patterns
Fig. 8 shows the sintered densities of two ceramics as a function of sintering temperature. The results indicate that the sintered densities of IGZO-111 ceramics peaked at 6.55 g/cm3 at 1300 ◦ C and then gradually declined to 6.40 g/cm3 at 1500 ◦ C. The sintered densities of IGZO-112 ceramics were improved from 3.86 g/cm3 to 6.32 g/cm3 by increasing the sintering temperature from 1200 ◦ C to 1400 ◦ C. However, the sintered density was slightly decreased to 6.30 g/cm3 as the sintering temperature was further increased to 1500 ◦ C. The trend of sintered densities of IGZO-112 ceramics was identical to that of porosity (Fig. 4). In contrast, in the IGZO111 ceramics, the trend of sintered density did not match that of porosity (Fig. 3). This inconsistency could be attributed to the phase transformation of IGZO-111 ceramic during sintering, as shown in Fig. 1. When the sintering temperature was increased from 1300 ◦ C to 1400 ◦ C, the In2 O3 and ZnGa2 O4 phases were completely replaced by the In2 Ga2 ZnO7 and InGaZnO4 phases. The theoretical densities of In2 O3 and ZnGa2 O4 are 7.119 g/cm3 (JCPDS 89-4595) and 6.168 g/cm3 (JCPDS 86-0410), respectively. Thus, the theoretical densities of IGZO-111 ceramics sintered at 1200 ◦ C and 1300 ◦ C should be 6.618 g/cm3 . However, the theoretical densities of IGZO111 ceramics sintered at 1400 ◦ C and 1500 ◦ C were correlated with the volumetric percentages of In2 Ga2 ZnO7 and InGaZnO4 phases. The theoretical densities of In2 Ga2 ZnO7 and InGaZnO4 are 6.494 g/cm3 (JCPDS 38-1097) and 6.380 g/cm3 (JCPDS 38-1104), respectively. Based on quantitative metallography of the crosssection microstructure, the percentage of InGaZnO4 phase in the
Fig. 2. X-ray diffraction patterns of IGZO-112 ceramics sintered at 1200 ◦ C and 1500 ◦ C.
The crystal structures of IGZO-111 and IGZO-112 ceramics are shown in Figs. 1 and 2, respectively. Several phases, including In2 O3 , ZnGa2 O4 , InGaZnO4 , and InGaZnO6 , had previously been found in IGZO ceramics [3,4]. Fig. 1(a) shows that In2 O3 and ZnGa2 O4 were present in the IGZO-111 ceramic sintered at 1200 ◦ C, indicating that Ga2 O3 and ZnO reacted to ZnGa2 O4 entirely. These two crystal structures did not change after 1300 ◦ C sintering. However, after sintering at 1400 ◦ C, In2 Ga2 ZnO7 and InGaZnO4 phases completely replaced the previous In2 O3 and ZnGa2 O4 ones, as shown in Fig. 1(b). When the sintering temperature was further increased to 1500 ◦ C, In2 Ga2 ZnO7 and InGaZnO4 phases still existed in the IGZO111 ceramic. On the other hand, the crystal structure of IGZO-112 ceramics sintered at 1200 ◦ C and 1500 ◦ C was a single InGaZnO4 phase, as demonstrated in Fig. 2. No phase transformation occurred in the IGZO-112 ceramics between 1200 ◦ C and 1500 ◦ C. 3.2. Microstructure The thermally etched fracture surfaces of IGZO-111 and IGZO112 ceramics sintered at various temperatures are shown in Figs. 3 and 4, respectively. After sintering at 1200 ◦ C, the two ceramics were still highly porous, indicating inferior sintering densification. However, the densification of IGZO-111 was much more obvious than that of IGZO-112 after 1300 ◦ C sintering. Only minor pores existed in the IGZO-111 ceramics sintered at 1300 ◦ C, 1400 ◦ C,
3896
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
Fig. 3. The thermally etched fracture surfaces of IGZO-111 ceramics sintered at (a) 1200 ◦ C, (b) 1300 ◦ C, (c) 1400 ◦ C, and (d) 1500 ◦ C.
Fig. 4. The thermally etched fracture surfaces of IGZO-112 ceramics sintered at (a) 1200 ◦ C, (b) 1300 ◦ C, (c) 1400 ◦ C, and (d) 1500 ◦ C.
Fig. 5. The cross-section microstructure of the IGZO-111 ceramic sintered at 1400 ◦ C. (a) SE image (b) BSE image.
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
3897
Fig. 6. The cross-section microstructure of the IGZO-111 ceramic sintered at 1500 ◦ C. (a) SE image (b) BSE image.
Fig. 7. The cross-section microstructures of the IGZO-112 ceramics sintered at (a) 1400 ◦ C and (b) 1500 ◦ C.
IGZO-111 ceramics sintered at 1400 ◦ C and 1500 ◦ C were 8 vol% and 30 vol%, respectively. Therefore, the theoretical densities of IGZO111 sintered at 1400 ◦ C and 1500 ◦ C should be 6.485 g/cm3 and 6.460 g/cm3 , respectively. On the other hand, the crystal structure
of IGZO-112 ceramic was a single InGaZnO4 phase, and no phase transformation occurred between 1200 ◦ C and 1500 ◦ C. The theoretical density of IGZO-112 sintered at various temperatures was thus regarded as 6.380 g/cm3 .
Fig. 8. The sintered densities of IGZO-111 and IGZO-112 ceramics as a function of sintering temperature.
3898
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
Fig. 9. The relative densities of IGZO-111 and IGZO-112 ceramics as a function of sintering temperature.
To investigate the actual porosities of the two ceramics as a function of sintering temperature, the sintered densities in Fig. 8 were divided by the theoretical densities to calculate the relative densities, as plotted in Fig. 9. The results indicate that the relative densities of IGZO-111 and IGZO-112 ceramics sintered at 1200 ◦ C were only 83.6% and 63.5%, respectively. The densification was apparently activated by raising the sintering temperature from 1200 ◦ C to 1400 ◦ C. The relative densities of these two ceramics peaked at 1400 ◦ C and then slightly decreased at 1500 ◦ C. The results of the relative densities corresponded well to those of the fracture surfaces, as can be seen from comparing Figs. 3, 4, and 9. The highest relative densities of IGZO-111 and IGZO-112 ceramics were 99.8% and 99.0%, respectively, in this study. 3.4. Thermal analysis To further understand the sintering behaviors of these two ceramics, a dilatometer and TGA were also used to investigate the dimensional change and weight loss, respectively, during sintering. For the dilatometer test, the two ceramics were heated at 10 ◦ C/min to 1400 ◦ C without holding in air. The amount and rate of the dimensional change are presented in Fig. 10. The results indicated normal thermal expansion at temperatures of 25–800 ◦ C. However, an apparent swelling phenomenon was observed in the two ceramics, particularly IGZO-112. The maximum linear expansions of the IGZO-111 and IGZO-112 ceramics were 1.4% at 1020 ◦ C and 7.9% at 1150 ◦ C, respectively. Moreover, the maximum expansion rates of the IGZO-111 and IGZO-112 ceramics were 0.03%/min at 880 ◦ C and 0.38%/min at 960 ◦ C, respectively. The extraordinary expansion of the IGZO-112 ceramic could be attributed to the difference in volume between the original reactants (In2 O3 , Ga2 O3 , and ZnO) and the final products (InGaZnO4 ) during the heating period. After heating to 1400 ◦ C without holding, the linear shrinkages of the IGZO-111 and IGZO-112 ceramics were 10.3% and 0.8%, respectively. These above findings demonstrate that the IGZO-111
ceramic densified faster than the IGZO-112 ceramic, and the relative density of the IGZO-112 ceramic was thus sufficiently lower than that of the IGZO-111 ceramic, particularly at temperatures below 1300 ◦ C, as shown in Figs. 9 and 10. However, the shrinkage rate of the IGZO-112 ceramic increased greatly above 1150 ◦ C, and the maximum shrinkage rate of the IGZO-112 ceramic was even higher than that of the IGZO-111 ceramic, as shown in Fig. 10(b). When these two ceramics were sintered at 1400 ◦ C for three hours, the difference in the relative densities of the two ceramics was minor. Due to the obvious evaporation phenomenon during hightemperature sintering of ZnO-based ceramics, the weight losses of the two IGZO ceramics sintered at various temperatures were studied and are displayed in Fig. 11 [29,31,32,36,39,40]. The results showed no clear difference between the two ceramics when the sintering temperature was lower than 1500 ◦ C. However, after 1500 ◦ C sintering, the weight losses of IGZO-111 and IGZO-112 ceramics were 4.7% and 5.1%, respectively. To differentiate the evaporation behaviors of the original In2 O3 , Ga2 O3 , and ZnO powders used in this study, these three powders were loosely packed in a crucible and then examined by TGA. The TGA specimens were heated at 10 ◦ C/min to 1500 ◦ C and held for one hour in air. Fig. 12 indicates that the weight loss of ZnO powder was much higher than that of Ga2 O3 , which was followed by In2 O3 . The ZnO powder exhibited a weight loss as high as 42% after holding at 1500 ◦ C for one hour. When the sintering temperature was increased from 1400 ◦ C to 1500 ◦ C, the relative densities of the two IGZO ceramics were decreased due mainly to the evaporation of ZnO, as shown in Fig. 9. 3.5. Grain size Fig. 13 reveals the average grain sizes of two IGZO ceramics as a function of the sintering temperature. The main phases of IGZO-111 (In2 Ga2 ZnO7 ) and IGZO-112 (InGaZnO4 ) were used to calculate the average grain sizes. When the sintering temperature
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
3899
Fig. 10. The dilatometric curves of IGZO ceramics as a function of sintering temperature. (a) Dimensional change (b) dimensional change rate.
was increased from 1200 ◦ C to 1500 ◦ C, the grain sizes of IGZO-111 and IGZO-112 ceramics increased from 0.4 m to 9.5 m and from 0.3 m to 12.7 m, respectively. The grain sizes of IGZO-111 and IGZO-112 ceramics sintered at 1400 ◦ C were 6.0 m and 10.9 m, respectively. The results also show that the grain sizes of IGZO112 ceramics sintered at 1200 ◦ C and 1300 ◦ C were smaller than
those of IGZO-111. This phenomenon could be attributed to the increased number of pores, which dragged the grain boundary and thus inhibited the grain growth, in the IGZO-112 ceramics. However, the trend reversed after 1400 ◦ C sintering, and the grains of IGZO-111 ceramics were larger than those of IGZO-112. After sintering at 1400 ◦ C, these two ceramics contained only about 1 vol%
3900
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
Fig. 11. The weight losses of IGZO ceramics as a function of sintering temperature.
pores, and the dragging effect contributed by the pores could be excluded. When these two ceramics were sintered at 1400 ◦ C or 1500 ◦ C, the crystal structures of IGZO-111 and IGZO-112 ceramics were two phases, In2 Ga2 ZnO7 and InGaZnO4 , and a single
phase of InGaZnO4 , respectively. According to the Zener effect, the presence of InGaZnO4 dispersion in the In2 Ga2 ZnO7 matrix could inhibit the grain growth and reduce the grain size in the IGZO-111 ceramics.
Fig. 12. The TGA curves of the original In2 O3 , Ga2 O3 , and ZnO powders used in this study.
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
3901
Fig. 13. The average grain sizes of IGZO ceramics as a function of sintering temperature.
3.6. Electrical properties To assess the electrical properties of the IGZO ceramics prepared in this study, the resistivities were tested using the four-point probe method. The resistivities of the two IGZO ceramics sintered at 1200 ◦ C and 1300 ◦ C could not be measured by the sourcemeter used in this study because they exceeded the range of the instrument. However, after sintering at 1400 ◦ C or 1500 ◦ C, the conductance of IGZO ceramics was much improved, and their resistivities could be obtained. The resistivities of IGZO-111 and IGZO-112 ceramics sintered at 1400 ◦ C were 4.9 × 10−3 cm and 4.5 × 10−3 cm, respectively. When the sintering temperature was increased to 1500 ◦ C, the resistivities of IGZO-111 and IGZO-112 ceramics were further improved to 1.5 × 10−3 cm and 2.5 × 10−3 cm, respectively, indicating that both In2 Ga2 ZnO7 and InGaZnO4 phases were highly conductive. In addition, the resistivities of amorphous IGZO-111 and IGZO-112 sputtered films reported in the literature range from 5 × 10−3 cm to 104 cm and 4 × 10−3 cm to 105 cm, respectively, depending on the sputtering parameters [2,4,5,7,8]. The amorphous structure of IGZO-111 film can be retained even after annealing at 600 ◦ C in argon [9]. These results show that the resistivities of In2 Ga2 ZnO7 and InGaZnO4 crystal phases were generally lower than those of amorphous IGZO structure. In the literature, the processes and properties of AZO and GZO ceramic targets sintered in air have been extensively investigated [21,25,28–36]. The relative densities, grain sizes, and resistivities of AZO targets sintered in air range from 98.6% to 99.8%, from 2.7 m to 5.0 m, and from 8.9 × 10−4 cm to 5 × 10−1 cm, respectively [21,25,28–31]. Moreover, the relative densities, grain sizes, and resistivities of GZO targets sintered in air range from 96.0% to 99.3%, from 2.0 m to 4.5 m, and from 1.4 × 10−3 cm to 1 cm, respectively [31–36]. This study demonstrates that the various properties, including the relative densities, grain sizes, and resistivities, of IGZO
ceramics sintered in air are comparable to those of AZO and GZO ceramics. The resistivities of IGZO ceramics could approach the lowest resistivities of AZO and GZO ceramics sintered in air. 4. Conclusions The influences of chemical stoichiometry and sintering temperature on the sintering, microstructure, and electrical properties of IGZO ceramics for sputtering targets were studied. The findings can be summarized as follows. 1. The crystal structures of IGZO-111 ceramics sintered at 1200 ◦ C or 1300 ◦ C were a mixture of In2 O3 and ZnGa2 O4 phases. When the sintering temperature was increased to 1400 ◦ C, the previous In2 O3 and ZnGa2 O4 phases were completely replaced by In2 Ga2 ZnO7 and InGaZnO4 phases. The InGaZnO4 dispersion phase was distributed in the In2 Ga2 ZnO7 matrix phase of IGZO111 ceramics sintered at 1400 ◦ C or 1500 ◦ C. However, the crystal structure of IGZO-112 ceramic was a single InGaZnO4 phase, and no phase transformation occurred between 1200 ◦ C and 1500 ◦ C. 2. The highest relative densities of IGZO-111 and IGZO-112 ceramics were 99.8% and 99.0%, respectively, in this study. The relative densities of the two IGZO ceramics peaked at 1400 ◦ C and then slightly decreased at 1500 ◦ C due mainly to the evaporation of ZnO. 3. Both ceramics exhibited an obvious swelling phenomenon during heating, particularly IGZO-112. The IGZO-111 ceramic densified faster than the IGZO-112 ceramic when the sintering temperature was lower than 1400 ◦ C. However, the shrinkage rate of the IGZO-112 ceramic increased greatly above 1150 ◦ C, and the difference in the relative densities of the two ceramics sintered at 1400 ◦ C for three hours was minor. 4. After sintering at 1500 ◦ C, the resistivities of IGZO111 and IGZO-112 ceramics could achieve 1.5 × 10−3 cm and 2.5 × 10−3 cm, respectively, showing that both In2 Ga2 ZnO7 and InGaZnO4 phases were highly conductive. In addition, the
3902
M.-W. Wu et al. / Journal of the European Ceramic Society 35 (2015) 3893–3902
resistivities of In2 Ga2 ZnO7 and InGaZnO4 crystal phases were generally better than those of the amorphous structure in the IGZO film reported in the literature. Acknowledgements The authors thank the Ministry of Science and Technology of the Republic of China for support and funding under contract number NSC 99-2218-E-150-046. Gratitude is also extended to Dr. FangCheng Chou and Dr. Guo-Jiun Shu in the Center for Condensed Matter and Science of National Taiwan University for the TGA experiments. References [1] H. Kumomi, K. Nomura, T. Kamiya, H. Hosono, Amorphous oxide channel TFTs, Thin Solid Films 516 (2008) 1516–1522. [2] T. Aoi, N. Oka, Y. Sato, R. Hayashi, H. Kumomi, Y. Shigesato, DC sputter deposition of amorphous indium–gallium–zinc-oxide (a-IGZO) films with H2 O introduction, Thin Solid Films 518 (2010) 3004–3007. [3] C.C. Lo, T.E. Hsieh, Preparation of IGZO sputtering target and its applications to thin-film transistor devices, Ceram. Int. 38 (2012) 3977–3983. [4] Y.S. Lee, Z.M. Dai, C.I. Lin, H.C. Lin, Relationships between the crystalline phase of an IGZO target and electrical properties of a-IGZO channel film, Ceram. Int. 38 (2012) S595–99. [5] M.J. Gadre, T.L. Alford, Highest transmittance and high-mobility amorphous indium gallium zinc oxide films on flexible substrate by room-temperature deposition and post-deposition anneals, Appl. Phys. Lett. 99 (2011) 051901. [6] J. Yao, L. Gong, L. Xie, S. Zhang, Comparison of the electrical and optical properties of direct current and radio frequency sputtered amorphous indium gallium zinc oxide films, Thin Solid Films 527 (2013) 21–25. [7] B.D. Ahn, H.S. Shin, G.H. Kim, J.S. Park, H.J. Kim, A novel amorphous InGaZnO thin film transistor structure without source/drain layer deposition, Jpn. J. Appl. Phys. 48 (2009) 03B019. [8] C.H. Jung, H.I. Kang, D.H. Yoon, The electrical, optical, and structural properties of amorphous indium gallium zinc oxide films and channel thin-film transistors, Solid-State Electron. 79 (2013) 125–129. [9] H. Kim, J.Y. Moon, Y.W. Heo, H.S. Lee, Effect of thermal annealing on Ni/Au contact to a-InGaZnO films deposited by dc sputtering, Thin Solid Films (2010) 6348–6351. [10] X.B. Zhang, Z.L. Pei, J. Gong, C. Sun, Investigation on the electrical properties and inhomogeneous distribution of ZnO:Al thin films prepared by DC magnetron sputtering at low deposition temperature, J. Appl. Phys. 101 (2007) 014910. [11] W. Yang, Z. Liu, D.L. Peng, F. Zhang, H. Huang, Y. Xie, Z. Wu, Room-temperature deposition of transparent conducting Al-doped ZnO, Appl. Surf. Sci. 255 (2009) 5669–5673. [12] R. Cebulla, R. Wendt, K. Ellmer, Al-doped zinc oxide films deposited by stimultaneous RF and DC excitation of a magnetron plasma: relationships between plasma parameters and structural and electrical film properties, J. Appl. Phys. 83 (1998) 1087–1095. [13] K. Utsumi, O. Matsunaga, T. Takahata, Low resistivity ITO film prepared using the ultra high density ITO target, Thin Solid Films 334 (1998) 30–34. [14] T. Omata, M. Kita, H. Okada, S. Otsuka-Yao-Matsuo, N. Ono, H. Ikawa, Characterization of indium–tin oxide sputtering targets showing various densities of nodule formation, Thin Solid Films 503 (2006) 22–28. [15] K. Nakashima, Y. Kumahara, Effect of tin oxide dispersion on nodule formation in ITO sputtering, Vacuum (2002) 221–226. [16] K.L. Ying, T.E. Hsieh, Y.F. Hsieh, Colloidal dispersion of nano-scale ZnO powders using amphibious and anionic polyelectrolytes, Ceram. Int. 35 (2009) 1165–1171. [17] N. Neves, R. Barros, E. Antunes, I. Ferreira, J. Calado, E. Fortunato, R. Martins, Sintering behavior of nano- and micro-sized ZnO powder targets for rf magnetron sputtering applications, J. Am. Ceram. Soc. 95 (2012) 204–210.
[18] N. Yamada, T. Hitosugi, J. Kasai, N.L.H. Hoang, D. Nakao, Y. Hirose, T. Shimada, T. Hasegawa, Transparent conducting Nb-doped anatase TiO2 (TNO) thin films sputtered from various oxide targets, Thin Solid Films 518 (2010) 3101–3104. [19] T. Kubart, J. Jensen, T. Nyberg, L. Liljeholm, D. Depla, S. Berg, Influence of the target composition on reactively sputtered titanium oxide films, Vacuum 83 (2009) 1295–1298. [20] T. Minami, J.I. Oda, J.I. Nomoto, T. Miyata, Effect of target properties on transparent conducting impurity-doped ZnO thin films deposited by DC magnetron sputtering, Thin Solid Films 519 (2010) 385–390. [21] H.S. Huang, H.C. Tung, C.H. Chiu, I.T. Hong, R.Z. Chen, J.T. Chang, H.K. Lin, Highly conductive alumina-added ZnO ceramic target prepared by reduction sintering and its effects on the properties of deposited thin films by direct current magnetron sputtering, Thin Solid Films 518 (2010) 6071–6075. [22] G. Fang, D. Li, B.L. Yao, Fabrication and vacuum annealing of transparent conductive AZO thin films prepared by DC magnetron sputtering, Vacuum 68 (2003) 363–372. [23] Y. Sato, H. Akizuki, T. Kamiyama, Y. Shigesato, Transparent conductive Nb-doped TiO2 films deposited by direct-current magnetron sputtering using a TiO2−x target, Thin Solid Films 516 (2008) 5758–5762. [24] H. Ohsaki, Y. Tachibana, A. Mitsui, T. Kamiyama, Y. Hayashi, High rate deposition of TiO2 by DC sputtering of the TiO2−x target, Thin Solid Films 392 (2001) 169–173. [25] M.W. Wu, D.S. Liu, Y.H. Su, The densification, microstructure, and electrical properties of aluminum-doped zinc oxide sputtering target for transparent conductive oxide film, J. Eur. Ceram. Soc. 32 (2012) 3265–3275. [26] N. Neves, A. Lagoa, J. Calado, A.M.B. Rego, E. Fortunato, R. Martins, I. Ferreira, Al-doped ZnO nanostructured powders by emulsion detonation synthesis—improving materials for high quality sputtering targets manufacturing, J. Eur. Ceram. Soc. 34 (2014) 2325–2338. [27] J.R. Lee, T.J. Chung, S.H. Yang, G.S. Hong, K.S. Oh, Prevention of tapering in the tube-shaped sputtering target via initial heat treatment under external pressure, Ceram. Int. 41 (2015) 3677–3683. [28] C.P. Liu, G.R. Jeng, Properties of aluminum doped zinc oxide materials and sputtering thin films, J. Alloys Compd. 468 (2009) 343–349. [29] J. Zhang, W. Zhang, E. Zhao, H.J. Jacques, Study of high-density AZO ceramic target, Mater. Sci. Semicond. Process. 14 (2011) 189–192. [30] T. Wei, Y. Zhang, Y. Yang, R. Tan, P. Cui, W. Song, Effects of ZnAl2 O4 segregation in high temperature sintered Al-doped ZnO sputtering target on optical and electrical properties of deposited thin films, Surf. Coat. Technol. 221 (2013) 201–206. [31] J. Liu, W. Zhang, D. Song, Q. Ma, L. Zhang, H. Zhang, X. Ma, H. Song, Comparative study of the sintering process and thin film sputtering of AZO, GZO and AGZO ceramics targets, Ceram. Int. 40 (2014) 12905–12915. [32] J. Liu, W. Zhang, D. Song, Q. Ma, L. Zhang, H. Zhang, X. Ma, H. Song, Gallium-doped zinc oxide targets fabricated by sintering: impact of target quality on sputtered thin film properties, Mater. Sci. Semicond. Process. 27 (2014) 1–11. [33] C. Liu, X. Cai, Z. Zhang, L. Gao, C. Xiao, F. Shang, H. Yang, Effect of gallium doping on the properties of GZO target, Adv. Mater. Res. 779–780 (2013) 363–367. [34] J.P. Wiff, Y. Kinemuchi, K. Watari, Hall mobilities of Al- and Ga-doped ZnO polycrystals, Mater. Lett. 63 (2009) 2470–2472. [35] M.S. Jang, M.K. Ryu, M.H. Yoon, S.H. Lee, H.K. Kim, A. Onodera, S. Kojima, A study on the Raman spectra of Al-doped and Ga-doped ZnO ceramics, Curr. Appl. Phys. 9 (2009) 651–657. [36] M.W. Wu, P.H. Lai, C.H. Hong, F.C. Chou, The sintering behavior, microstructure, and electrical properties of gallium-doped zinc oxide ceramic targets, J. Eur. Ceram. Soc. 34 (2014) 3715–3722. [37] M.W. Wu, Y.C. Chen, The sintering trajectory and electrical properties of niobium-doped titania sputtering targets, J. Am. Ceram. Soc. 98 (2015) 163–170. [38] M.I. Mendelson, Average grain size in polycrystalline ceramics, J. Am. Ceram. Soc. 52 (1969) 443–446. [39] J. Han, P.Q. Mantas, A.M.R. Senos, Densification and grain growth of Al-doped ZnO, J. Mater. Res. 16 (2001) 459–468. [40] N. Neves, R. Barros, E. Antunes, J. Calado, E. Fortunato, R. Martins, I. Ferreira, Aluminum doped zinc oxide sputtering targets obtained from nanostructured powders: processing and application, J. Eur. Ceram. Soc. 32 (2012) 4381–4391.