Abstract: IGZO/PET specimens were respectively prepared with 0°, 15°, 30°, 45°, and 60° as the inclination angle of the thin film deposition. The n-type conductivity was identified in these specimens. The following parameters, including the film thickness and morphologies of the top and lateral surfaces, the mechanical properties, chemical compounds and their primary lattices in X-ray diffraction (XRD) patterns, and the X-ray photoelectron spectrometries (XPS) for Ga2p, Ga3d, In3d, Zn2p, and O1s are presented. IR is defined as the intensity ratio of the InGaO3(ZnO)3 peak value to the sum of the peak values of InGaO3(ZnO)3 and InGaZnO4. Decompositions of the O1s and Ga3d spectra provide nearly Gaussian profiles of Ga-Ga, In4d, Ga-O, O1, O2, and O3. The O2 intensity ratio IRO2, which is defined as the ratio of O2 peak intensity to the peak intensity sum of O1 and O2, was evaluated for these five specimens. Depth profiles of the distributions of O, Zn, Ga, and In ions were obtained by secondary ion mass spectrometry (SIMS), and the slope (Go) of the O-ion profile in the decaying region was obtained as a function of inclination angle. Increases in the inclination angle can effectively reduce surface roughness. The peak intensities of Ga2p, Ga3d, In3d, Zn2p, and O1s formed in the specimen preparations with a nonzero inclination angle were always lower than those of the IGZO_0° specimen. Although an increase in inclination angle can raise the IRO2 value, a nonzero inclination angle yielded an IR value lower than that of the IGZO_0° specimen. Moreover, the IRO2 value increased with (negative) Go decrease as the response. The combined result of increasing IRO2 and the In-O and Ga-O bonds, and decreasing the IGZO film thickness can increase light transmission. Specimens with a larger surface roughness result in a higher reflection. The IR parameter for the specimens with a nonzero inclination angle has a value smaller than that of the IGZO_0° specimen; as such, an increase in IR is advantageous for increasing the n-type conductivity. The decreasing rate of extinction coefficient k with respect to wavelength in the visible light region increases significantly with increasing IRO2. IR and IRO2 are therefore the governing factors of the peak intensities for the three decomposed profiles of micro-photoluminescence (PL). Increasing IR and IRO2 or decreasing film thickness is advantageous for increasing the PL peak intensities. Increases in the product values of IR and IRO2 are favorable for increasing the peak intensities of violet and ultraviolet.
© 2016 Optical Society of America
Oxide semiconductors, such as zinc oxide (ZnO), indium-tin oxide (ITO), indium-zinc oxide (IZO) and indium-gallium-zinc oxide (IGZO) are good candidates for channel layers of thin film transistors (TFTs) [1–4]. Compared with amorphous silicon and organic semiconductor TFTs, amorphous IGZOs have attracted much attention as a promising candidate for an active matrix flat-panel display due to their larger field effect mobility . Polyethylene terephthalate (PET) is a semicrystalline polymer whose composites were widely used in various industries. Light-weight, flexible display technologies have currently gained considerable interest in the display and electronic industries.
In Kim et al. , the fabricated hybrid solutions of InGaO3(ZnO)m thin films with superlattice structures were prepared for the crystallization of multicomponent oxides. The solid-phase epitaxy of a single phase decreased the electrical resistivity, improved the power factor and lowered the thermal conductivity. Crystallizations of InGaZnO4 (IGZO) by excimer laser on SiO2 substrates  were obtained at low temperature without a seed substrate. The IGZO film was fabricated by the conventional solid state reaction method in air using a pulsed laser. Amorphous IGZO films with low-resistance ohmic contacts were obtained from the exposure with hydrogen plasma after deposition by radio frequency magnetron sputtering . The a-IGZO thin films were air annealed at different temperatures, the data of which show that air annealing decreased the densities of the oxygen-vacancy-representing states and generated the contaminant component at the O1s peak . IGZO films were also prepared by pulsed laser deposition at room temperature . The carrier concentration was varied by several orders of magnitude depending on the oxygen partial pressure during deposition, after which the deposited films were all highly transparent throughout the visible region and into the near-infrared.
The a-IGZO films were deposited at room temperature, and then annealed with different hydrogen and oxygen gas flow ratios . For the a-IGZO thin film with a high O content and low density of electrons, the electrical and photosensitive characteristics of the TFT degrade. The carrier concentration and resistivity of the a-IGZO films were greatly affected by the addition of hydrogen and heat treatment. The effects of substrate temperature on the surface roughness, electrical resistivity, mobility, charge carrier concentrations, transmission and optical bandgap of a-IGZO thin film deposited by radio-frequency magnetron sputtering were observed . Investigation of the O1s core level and the Ga3d, In4d, and Zn3d shallow-core levels spectra revealed that as the substrate temperature increased, an O1s component representing the oxygen vacancies increased and the IR of In/Ga also increased; however, the IR of Zn/Ga decreased. The fabrication process and electrical characteristics of bottom-gate IGZO TFTs regarding the influence of oxygen or nitrogen post-annealing ambient was reported in . Results showed that the IGZO layers after annealing in N2 have a higher concentration of oxygen vacancies. The evolution of the electrical properties and thin-film transistor characteristics of a-IGZO films synthesized by RF sputtering with O2 plasma immersion has been examined . The effects of O2 plasma immersion on the electrical properties and the transistor performance can be attributed to the reduction of the oxygen-related defects in the IGZO thin films. Further, the effects of high energy electron beam irradiation on the properties of IGZO thin films were explored . The strong reduction of donor-like oxygen-vacancy defects occurred due to the formation of either oxygen interstitials or zinc-vacancy acceptor defects. Upon thermal annealing for the a-IGZO at temperatures up to 500°C, although the amorphous characteristics were maintained, the electronic properties were considerably enhanced . This could be ascribed to the increased optical band gap and increased oxygen vacancies. Moreover, the RF sputtered a-IGZO films produced better stoichiometry, lower electrical conductivity, higher refractive index, larger band gap, and increased the concentration of electrons in the conduction band with a reduced concentration of oxygen vacancies .
The optical properties of amorphous IGZO and polycrystalline ZnO thin films were investigated by spectroscopic ellipsometry . The optical gap of the IGZO film increased with increasing Ga content and by annealing. This behavior is attributed to the formation of Ga2O3 with a large band-gap energy and the structural relaxation after annealing, respectively. In Nomura et al. , the electronic states in a-IGZO were investigated through optical spectroscopy and hard x-ray photoelectron spectroscopy (HX-PES). All films had tail-like optical absorptions due to the extra subgap densities of states (DOSs) near valence band maximas (VBMs). In order to reduce defects and increase the carrier concentrations of amorphous IGZO thin films, the effects of annealing with different gases and ion implantation were examined . The amorphous phase underwent transformation to a nanocrystalline phase due to annealing. As the annealing temperature increased, the optical gap energy increased due to crystallization.
Glancing angle deposition (GLAD) is based on deposition at oblique angles, where the trajectory of the incident vapor flux is not parallel to the substrate normal. The review  examines the GLAD process and column growth, the properties observed in the fabricated films, and the applications of this technology. The ability to reliably fabricate oriented structures with varying degrees of structural anisotropy allowed GLAD to produce film for a variety of application areas including the surface area enhancement , the mechanical response of the columnar microstructure to a nanoindentation load , the dominance in the optical coating fabrication techniques , and its influence on birefringent properties [25, 26].
Applications of non-zero inclination angle to the deposition plate in the preparations of IGZO film have produced noticeable reduction in the mean surface roughness and great improvement in the undulation behavior of coating film and substrate demonstrated in the IGZO_0° specimen. These two advantages inspire us to investigate the optical, electrical, and mechanical properties and microstructures of specimens induced by the inclination angle. In the present study, five inclination angles, 0°, 15°, 30°, 45°, and 60°, were applied to the deposition plate during the coatings of the IGZO film onto a PET substrate. Chemical compounds were altered and the microstructure, including the degree of crystallization and oxygen vacancy defects, were affected by the difference in the inclination angle. Specimen thickness and morphology, and optical properties are expected to have significant changes. X-ray diffraction (XRD) patterns identify the IGZO chemical compounds and their primary lattices. The intensities of these compounds were used to define the IR, which is of importance to the optical parameters. XPS analyses provided the spectra of Ga2p, Ga3d, In3d, Zn2p, and O1s. The O1s spectra for these five specimens were further deconvoluted into the O1, O2, and O3 Gaussian profiles, and the intensity ratio IRO2, relating to the oxygen vacancy defects, was also evaluated for these specimens. SIMS was applied to determine the depth profiles of O, Zn, Ga, and In ions. The profile slope of O ions was employed to establish their interaction with IRO2 and evaluate the influence of inclination angle on the depth of oxygen vacancy defects. Then, the effects of IRO2 and IR on the film morphology and thickness, microstructures, optical properties, and PL peak intensities were evaluated. Finally, the connections of inclination angle with these parameters were clarified.
2. Experimental details
The IGZO thin films were deposited on a Poly(ethylene terephthalate) (PET) substrate using an RF magnetron sputter system (Helix, HLLS-87, Taiwan). A single IGZO film with In: Ga: Zn = 1: 1: 1 as the mole ratio was used as the target. The deposition details are shown in Table 1. The PET substrate has a thickness of 188 μm, and the physical and mechanical properties are shown in Table 2 .
It is found that the IGZO_15° sample has the largest bulge, but its mean surface roughness is the second largest of these five specimens. This behavior can be explained in that the IGZO_15° sample has a smoother topography than that of the IGZO_0° sample because the bulges in the former are created more uniformly in size compared to those of the latter. A wider distribution in the bulge size in general has a larger difference in the heights between large and small bulges, thus resulting in a relatively large surface roughness.
The inclination angle is defined as the angle between the ion-beam direction and the direction normal to the PET substrate. The rig was designed for the PET substrate tilting at a specified inclination angle before coating the film via the rectangular deposition plate. A horizontal circular disc was prepared with four drilling holes for locking a block in a form of rectangular plate by the screws on its bottom surface. This block has been prepared with a tilt surface on the left-upper corner in order to support the deposition plate with a smooth contact. The three intervals between any two adjacent drilling holes have been properly designed such that this block can move flexibly on the horizontal disc and the inclination angle of the deposition plate can be achieved as we need. In the present study, 0°, 15°, 30°, 45° and 60° are chosen as the inclination angles with an error of ± 1.0°. This rig was installed in the deposition chamber during the sputtering process. The central line of the specimen was thus adopted as the baseline to measure the distance between the target and the substrate, for which 4 cm was given as the distance in this study.
A surface roughness measurement instrument (Solver-P47H, NT-MDT, Russia) was applied to measure the topographies of the IGZO films. The top and lateral profiles of the IGZO films were prepared by a dual focused ion beam (FIB, FEI Nova-200, USA). Images of the specimens’ lateral profile were taken to evaluate the voids formed in the interface. The TEM diffraction patterns are provided to identify the crystallization behavior in these specimens. X-ray diffraction (XRD) analyses were carried out on a diffractometer (Rigaku, ATX-E, USA) in the 2θ range of 30°-80° using a CuKα X-ray as the radiation source. All of the XRD measurements were carried out at room temperature. The 2θ scans show the angular positions of the oxide chemical compounds and their primary lattices. X-ray photoelectron spectroscopy (XPS, PHI-5000 Versaprobe, ULVAC-PHI, Osaka, Japan) was utilized to observe the variation tendency of the chemical state for each element. Depth profiles of phosphorous were carried out on a secondary ion mass spectrometry instrument (SIMS, Cameca 6F, France), for which Cs + ions were used as the primary ion beam. The optical properties, including the absorption and transmission spectra of the specimens, were recorded on a spectrophotometer (Hitachi U-4000, Japan). Photoluminescence (PL) spectra (Jobin Yvon/Labram HR, France) were obtained at room temperature with a 325 nm He–Cd laser with a spot size of 0.8 μm and a power of 30 mW incident at the sample. And finally, a nanoidentation tester (MTS, Nano Identer XP, USA) with a scratch function was used to measure the material hardness (H) and reduced modulus (Er).
3. Results and discussion
Preparations of the IGZO specimens by varying the inclination angle resulted in significant effects on several of the inherent properties, including the surface topography and microstructure of the film at the interface. Figures 1(a)–1(e) show the SEM morphologies of the specimens prepared at 0°, 15°, 30°, 45°, and 60°, respectively. For the specimens prepared at 0° and 15°, large bulges in the micron to sub-micron size range were found. The mean size of the bulge for the IGZO_15° sample was relatively larger than that of IGZO_0° specimen. Moreover, as the inclination angle increased, the mean bulge sizes greatly reduced, and the topography became very smooth when the inclination angle rose to 45° and 60°. These five specimens show the sequence of mean bulge size (BL) as (BL)15° > (BL)0° > (BL)30° > (BL)45° (BL)60°. The mean surface roughness of the 0°, 15°, 30°, 45°, and 60° specimens were 14.49 ± 0.70 nm, 13.61 ± 0.52 nm, 4.03 ± 0.11 nm, 2.20 ± 0.31 nm, 2.47 ± 0.24 nm, respectively. Generally, thin films deposited on a flexible substrate can bend the substrate if the externally-applied stress, which is governed by the incident/inclination angle, and the residual stress in the film, which is frequently induced by an elastic modulus difference of more than 5 GPa between a flexible substrate and the metal oxide film, are in total sufficiently high to produce the bending behavior . In this study, the former stress term has the highest value if an inclination angle of 0° is used. It is significantly reduced with increasing inclination angle. Only for the total stress created in the IGZO_0° specimen is it possible to exceed the threshold value needed for substrate bending. This can explain why only the IGZO_0° specimen exhibits a wavy substrate in these five specimens. Lateral profiles of these five specimens were prepared to investigate the microstructure of the IGZO films and their thicknesses. The ones for 0° and 60° as the inclination angle are shown in Figs. 2(a) and 2(b), respectively. As can be seen, for the IGZO_0° specimen, there exists a wavy pattern in the substrate surface after the thin film was deposited, which caused numerous bulges in the coating film. As the inclination angle increased, the interface between the coating film and its substrate were found to be quite tight and smooth. Results of the surface topography suggest that the bulges were formed in relation to the wavy interface pattern formed beneath them. Figure 2(a) is adopted as an example to investigate the effect of film thickness on the topography and surface roughness of coating film. The film thickness measured at A, B, and D positions are 1492.0, 1466.2 and 1549.6 nm, respectively. The bulge thickness of the wavy PET substrate at point C is measured to have 79.8 nm above the connection line of two points in the interfaces beneath A and D. The film thickness corresponding to position B is obviously smaller than those of positions, A and D. The SEM lateral profile in Fig. 2(a) shows the topography of the IGZO film to be a wavy pattern too. The undulations in the pattern are obviously affected by the wavy PET substrate. This waviness is also included in the measure of the mean surface roughness (Ra) more or less. It is found that the sum of the thickness at position, B and C, is larger than that of position A but close to that of position D, and the difference in thickness between the positions, A and B + C, is much larger than the mean surface roughness (Ra) at position B. In Fig. 2(b), the waviness of the PET substrate is seldomly present in the interface, the coating film is presented to have a quite uniform thickness and a flat topography. The mean surface roughness (Ra) is little affected by the surface morphology, thus resulting in a very small value ( = 2.47 ± 0.24 nm). It is thus concluded that the local film thickness is indeed affected by the waviness of the PET substrate formed under the deposition conditions; the thin film topography undulations have more or less influence on the measures of mean surface roughness. The larger the substrate waviness is, the greater the influence on the mean surface roughness is. Increases in the inclination angle are useful to eliminate both the wavy patterns in the substrate and the residual stress in the thin film, and thus reduce the surface roughness significantly. The IGZO film thicknesses were strongly dependent upon the inclination angle. The film thickness results of these five specimens are shown in Fig. 2(c). The film thicknesses varying with the inclination angle are prepared quite randomly. The film thickness of the IGZO_15° specimen is the largest of these five specimens, which is believed to be due to the mean grain size of this specimen, as discussed later, much greater than those of the other four specimens.
The TEM images of the lateral surfaces of the IGZO_15° and IGZO_45° specimens are shown in Figs. 3(a) and 3(b) as representative. In the small island-shaped dark areas, the IGZO film was formed via the deposition of particles with various grain sizes. The grain size in the thin film was determined by importing the TEM images into “Solidworks” software via the “Drawing” code. Then, the images were magnified to have their maximum size with a resolution of nanometers. The brightness and contrast were then adjusted to make the particles encircled by dash curves clearly visible. Then, the “Spline” code was applied to evaluate the grain sizes. The grains in the IGZO_15° specimen have their mean size much larger than that of the IGZO_45° specimen. The mean grain sizes (GS) of these five specimens in Fig. 3(c) shows the sequence that (GS = 6.6 ± 0.71 nm)15° > (GS = 4.5 ± 0.8 nm)0° > (GS = 3.5 ± 0.18 nm)30° ≥ (GS = 3.5 ± 0.12 nm)45° ≥ (GS = 3.5 ± 0.07 nm)60°. It is found that the sequence of mean grain size (GS) is quite consistent with that of bulge size (BL). That is, the bulges created in the specimens have their mean size proportional to the mean grain size in the IGZO film.
Nano-indentation tests were carried out to evaluate the hardness (H) and reduced modulus (Er) of the five specimens, the results of which are given in Table 3. The H values monotonically lowered with increasing inclination angle, with significant reductions in hardness occurring at the inclination angles ranging from 0° to 45°. The causes of the reduced hardness are discussed below. The behavior where a nonzero inclination angle yielded a lower Er value compared to that of IGZO_0° is also manifested. However, the minimum Er was found in the IGZO_30° specimen, in which the Er value lowered to about only 1/4 that of the IGZO_0° specimen. Apart from the microstructure variation of the coating film, the effect of the film thickness and thus the substrate properties on H and Er should be evaluated too. The coating film thickness is significantly decreased as the inclination angle ≥ 30° was applied. The reduction of film thickness allowed the indentation measurements of these two mechanical properties to be substantially affected by the soft PET substrate. The big reductions of these two parameters in the specimens of IGZO_30°, 45°, and 60° can provide evidence of the substrate’s effect on them.
Figure 4 shows the XRD patterns of the five specimens. The highest peak at 2θ 25.0° was identified to be in association with the PET substrate. The peaks at 2θ 12.8° and 2θ 46.5° display a pattern indicating the relatively noticeable polycrystalline phases. The former was identified as InGaO3(ZnO)3 (006), and the latter as InGaO(ZnO)1 (1010). The relatively smaller peak at 2θ 18.0° was identified as InGaO3(ZnO)3 (009) [7, 29]. The peak intensities of these three polycrystalline materials are shown in Table 4. It can be found that the peak intensity of IGZO_30° for all three polycrystalline materials is always the lowest of the five specimens.
The ratios of the peak intensity of InGaO3(ZnO)3 (including (006) and (009)) to the sum of the peak intensities of InGaO3(ZnO)3 and InGaZnO4 are defined as IR. IR is one of the governing factors affecting the mechanical properties. Moreover, the atomic percentages of the In, Ga, Zn and O ions are certainly affected by this parameter. The IR values in Table 5 for the specimens with nonzero inclination angles are smaller than that of the IGZO_0° specimen; that is, the content of InGaO3(ZnO)3 (006) + InGaO3(ZnO)3 (009) is slightly reduced when a nonzero inclination angle was used.
The TEM diffraction patterns have been obtained for these five specimens. Figures 5(a) and 5(b) show those of the IGZO films deposited at 15° and 45° respectively as inclination angle. InGaO3(ZnO)3 (006), InGaO3(ZnO)3 (009) and InGaZnO4 (1010) appear in the patterns of the IGZO_0°, _45°, and _60° specimens, respectively. The clarity of the pattern associated with InGaO3(ZnO)3 (006) is much higher than those of the other crystalline phases, which is consistent with the peak intensities of the XRD patterns. The ED patterns created with the IGZO_15° and IGZO_30° specimens were found to be amorphous, despite these polycrystalline phases being locally detectable in the XRD analyses. The TEM morphology of IGZO_45° is shown in Fig. 5(c). As indicated by the circled area, the polycrystalline phase is identified to be InGaO3(ZnO)3 (006) + InGaO3(ZnO)3 (009).
XPS analyses were conducted for the In, Ga, Zn, and O elements in the five specimens, for which Figs. 6(a)–6(e) respectively show the intensity variations of the In, Ga, Zn and O elements with binding energy. Sharp patterns are present at In (In3d5/2 and In3d3/2), Ga (Ga2p3/2 and Ga2p1/2), and Zn (Zn2p3/2 and Zn2p1/2), while broad patterns with a blunt peak are shown in Ga3d and O1s only. Chemical compositions can be evaluated from the relative integrated areas corresponding to the In, Ga, Zn and O peaks; thus, the atomic ratio of these chemical elements could be estimated. The atomic percentage ratios of In3d: Ga2p: Zn2p: O1s obtained from the XPS analyses for the five specimens are shown in Table 6. The data indicate that the application of a nonzero inclination angle can reduce the atomic percentages of Ga2p and Zn2p but increase the atomic percentage of O1s. Increase in the atomic percentage of O1s is expected to produce oxygen interstitial (Oi) formation due to in-diffusion of O from the ambient [15, 29]. Decrease in the atomic percentage of Zn2p is related to zinc vacancy (VZn) formation due to out-diffusion of Zn from films . The reductions of Zn in the atomic percentage ratio implies that the InGaO3(ZnO)3 content in the IGZO film was lowered by the nonzero inclination angle, which is consistent with the behavior shown in Table 5 for IR. Table 7 shows the peak intensities of (Ga2p3/2, Ga2p1/2, Ga3d), (In3d5/2, In3d3/2), O1s, and (Zn2p3/2, Zn2p1/2) for the specimens prepared at the five inclination angles. The peak intensities of the IGZO_0° and 30° specimens, for all four atoms, have the highest and lowest values, respectively, of the five specimens. The peak intensities (PIs) for Ga2p3/2, Ga2p1/2, Ga3d, In3d5/2, In3d3/2, and O1s share the sequence (PI)0° > (PI)15° > (PI)60° > (PI)45° > (PI)30°; while the sequences for Zn2p3/2 and Zn2p1/2 indicate (PI)0° > (PI)60° > (PI)15° > (PI)45° > (PI)30°. Moreover, application of nonzero inclination angles lowered the peak intensities of all these ions.
In Fig. 6(a), the center peaks of Ga2p3/2 and Ga2p1/2 are slightly shifted to a higher binding energy with an increasing inclination angle. All changes are less than 1 eV. This small change can be attributed to the following factor: the decrease in the density of valence electrons, an increase in the oxidation state of the Ga-O bond, and/or a Fermi-level shift in the band gap due to irradiation [30, 31]. The Ga2p3/2 and Ga2p1/2 spin-orbital-splitting photoelectrons were located at binding energies of 1117-1118 eV and 1144-1145 eV, respectively, for the five specimens, and were assigned to Ga2O3 . The differences in binding energies between Ga2p3/2 and Ga2p1/2 were about 27 eV for the five specimens. Ga2p peaks are created corresponding to the Ga-Ga bond. The deconvolutions (decompositions) were made for the Ga3d spectrum of these five kinds of specimen. Figure 7 is a representative one to show the peaks formed at about 17.55 eV, 18.95 eV, and 19.75 eV, respectively. Those at 17.55 eV and 19.75 eV indicate Ga-Ga and Ga-O bonds, respectively . The peak intensity of the Ga-Ga bond at these inclination angles is found to be higher than that of the Ga-O bond. This behavior is possibly due to the excess Ga on the surface. The peak intensities (PIs) of these two bonds show the same sequence of (PI)0° > (PI)15° > (PI)60° > (PI)45° > (PI)30°. According to the data shown in Table 7, the peak values of Ga3d due to the difference in inclination angle have a magnitude sequence the same as those of the Ga-Ga and Ga-O bonds because these two bonds are the dominant components in the Ga3d spectrum. The peak at 18.95 eV, which is identified as In4d , is concealed in the Ga3d spectrum as a small bulge with 19 eV as its binding energy. The peak intensities of the In4d profile for applications of nonzero inclination angles are somewhat lower than those of the IGZO_0° specimen. The sequence of the In4d peak intensity due to the change in inclination angle is also close to those of the Ga-Ga and Ga-O bonds, except for the interchange in the magnitude order of 45° and 60°.
In Fig. 6(b), the peaks at about 444.4 eV and about 452 eV are identified as In3d5/2 and In3d3/2, respectively. They are also shifted with slightly higher binding energy compared to that of the IGZO_0° specimen if a nonzero inclination angle was applied. These shifts are also attributable to the increase in the number of In-O bonds . The In peaks are located at 444-445 eV and 451-452 eV, corresponding to the electronic states of In3d5/2 and In3d3/2, respectively. The energy difference (about 7 eV) between the two In peaks indicates that In is in the + 3 oxidation state (In2O3) and substituted into the lattice . The In3d5/2 peak at about 444.4 eV corresponds to the In-In bond [29, 34]. Content changes in In and Ga are relevant to light transmittance, and their connection with optical behavior will be discussed later. In Fig. 6(c), the highest and second highest peaks are identified as Zn2p3/2 (about 1021.5 eV) and Zn2p1/2 (about 1045 eV), respectively, and are produced corresponding to the Zn-Zn bond [29, 34]. The shifts to higher binding energies due to the use of nonzero inclination angles are also attributable to the Zn-O bond . The differences in binding energies between Zn2p3/2 and Zn2p1/2 were in a range of 22 eV to 23 eV for all catalysts in the five specimens, which is quite close to the standard reference value of ZnO .
The broad spectra shown in Fig. 6(d) are identified as O1s. Deconvolutions were made for these five specimens prepared at the different inclination angles. Only the spectra of the IGZO_0° are shown in Fig. 8 as the representative pattern. Every O1s spectrum in these specimens can be decomposed into three nearly fitting Gaussian curves with about 530 eV, 531 eV and 532 eV as the peaks for O1, O2, O3 respectively. The peak at a binding energy of ~530 eV (O1) is ascribed to the O2- ions in the a-IGZO structure, and represents oxygen in the oxide lattice without oxygen vacancies. The O2 peak ( 531 eV) reflects the oxygen vacancies or OH groups, and is typically associated with the presence of oxygen-deficient regions. The O3 peak (about 533 eV) corresponds to the existence of weakly bonded oxygen species on the film surface, and is attributable to H2O and absorbed species integrated into the materials [35–37]. The ratio of the O2-peak intensity to the sum of the O1 and O2 intensities (IRO2) represents the indication of the relative quantity of the oxygen-related defects (vacancies). A lower ratio of oxygen vacancies and surface absorbed oxygen is related to charge trapping and increases the electrical stability of TFT devices . Table 8 shows the O1, O2, and O3 peak intensities and intensity ratio values of IRO2 for the five specimens. The applications of nonzero inclination angle have the intensities of O1 and O2 lower than those of IGZO_0°. However, the IRO2 value, and thus the oxygen vacancies increased as the inclination angle increased. A molecular dynamics study  of nickel vapor deposition shows the behavior that vacancy concentrations at various incident energies increase with increasing incident angle. The application of a nonzero incident angle is in a sense equivalent to the application of an inclination angle at the same value in this study. The content of InGaO3(ZnO)3, as shown in the IR results in Table 5, is thus reduced slightly by increasing the oxygen vacancies. The influence of IRO2 on the specimen’s hardness shows that a small increase in IRO2, and thus oxygen vacancies, directly produced a significant hardness reduction. However, the behavior demonstrated in the reduced modulus shows that the sequence of the Er results associated with the inclination angle is exactly the same as those of the peak intensities of Ga2p3/2, 1/2, Ga3d, In3d5/2, 3/2 and O1s, and is slightly different from that of Zn2p3/2, 1/2 in the order interchange between 15° and 60°. It is therefore concluded that the reduced modulus is governed by the peak intensities of Ga2p, Ga3d, In3d, Zn2p and O1s; it is increased by increasing the intensities of these ions.
SIMS was used to measure the ion counts (C/s) of O, Zn, Ga, and In varying with the depth beneath the specimen’s top surface. The distributions of these four ions in the IGZO_0° specimen are presented in Fig. 9 as the representative results. The ion counts of O, Zn, Ga, and In are almost constant over a depth ((D)upper) as their upper bound; each chemical element lowers to a smaller value in response to the ions’ content drop starting at different depths. In Fig. 9, a depth (D) of 1493 nm is marked. Before it a constant O ion count (C/s) (about 1.05 × 106) existed as the upper bound; and at a depth of 1945 nm behind it an approximately constant O ion count (about 1.0 × 105) is obtained as the lower bound. Then the gradient (Go) of the secondary O ions defined as [(C/s)upper - (C/s)lower] / [(D)upper - (D)lower] is used to evaluate the effect of inclination angle on the degenerating rate of O content between the upper and lower bounds. The absolute Go values in Table 9 show the sequence |Go|0° > |Go|15° > |Go|30° > |Go|45° > |Go|60°. This sequence, due to the change in inclination angle, is exactly opposite to that of IRO2. Accordingly, the IRO2 of the specimen, which reflects the concentration of oxygen vacancies, increases in response to a decrease in Go.
Figures 10(a)–10(c) respectively show the transmittance, reflection, and absorption results of the specimens operating in the wavelength region of 300 nm to 800 nm and their local magnification in a partial region of 300 nm to 400 nm. In Fig. 10(a), light transmission in the range of 300 nm to about 400nm is shown to increase significantly with increasing wavelength, irrespective of the inclination angle being applied. In the wavelength range > 400 nm, the specimen transmission rise turned out to be much mild. Due to the combined effects of increasing the oxygen-related vacancies in terms of IRO2 (shown in Table 8) and decreasing the IGZO film thickness (in Fig. 4), the transmission was increased by increasing the inclination angle, especially for an inclination angle ≥ 30°. The effects of oxygen deficiency or metallic interstitials on the transmission have been discussed in the literature [40, 41]. According to , the increasing numbers of in In-O and Ga-O bonds and thus the changes in the chemical compositions of InGaO3(ZnO)3 and InGaZnO4 due to the increase in inclination angle are also combined as the cause of the transmission increase. Figure 10(b) shows the reflection results in the same wavelength range. In the visible light region, the reflection is strongly affected by the thin film topography, including the surface roughness and the size and pattern of surface asperities. In this wavelength range, the results in Fig. 10(b) indicate that surface roughness (Ra) becomes the dominant factor of reflection; the specimen with a larger Ra value has a higher reflection as the result. Figure 10(c) shows the absorptions of the five specimens varying with wavelength. The results indicate that significant reductions in absorption occurred only in the wavelength region < 400 nm; and the absorption lowered to nearly zero at wavelengths of roughly > 400 nm. Accordingly, increases in the inclination angle are advantageous for reductions of absorption only in the wavelength region of < 400 nm. This behavior can be anticipated due to the chemical composition transition between InGaO3(ZnO)3 and InGaZnO4 , and the change in oxygen deficiency or metallic interstitials [40, 41].
Figures 11(a)–11(e) show the results of the refractive index (n) and extinction coefficient (k) obtained from the IGZO_0°, 15°, 30°, 45°, and 60° specimens, respectively. In the visible light region (380 nm – 780 nm), the n values of the IGZO_0° specimen were found to be monotonically lowered by increasing the light wavelength slightly. When a nonzero inclination angle was applied, the n values at various wavelengths in the visible light were always smaller than that of the IGZO_0° specimen. The n values for the IGZO specimens with nonzero inclination angles had quite random patterns such that the monotonically decreasing behavior was found in the IGZO_45° specimen only. Here, IN is defined as the integration of the n values over the wavelengths in the visible light region. The results in Table 10 show the following sequence: (IN)0° > (IN)30° > (IN)45° > (IN)60° > (IN)15°. The refractive index (n) of the material is related to its density. If the density can be lowered by introducing some kinds of grain size and voids (or vacancies), it becomes possible to decrease the refractive index. In the geometric treatment of Tait et al. , the film density normalized to the bulk density (ρ) is related to the deposition angle α which is here equal to the inclination angle of deposition plate, by the expression :
This formula was developed without taking the voids in the film into account. In this study, the film density ρ is lowered by increasing the inclination angle if Eq. (1) is adopted. The voids in the film can reduce the density. In general, a film deposited with a larger grain size (GS) had a larger quantity of voids in response to it. Define Vo as the volume proportion of nano voids (or vacancies) of the IGZO film. In general, a larger Vo value is produced in the specimen with a larger GS value. Therefore, the Vo results are expected to have a sequence of (Vo)15° > (Vo)60° > (Vo)45° > (Vo)30° > (Vo)0°. Define (Wv) as the weighting factor of ρ due to the voids formed in the coating film, it has a value of ≤ 1. The sequences of GS and Vo for these five specimens imply the Wv results satisfying the sequence of (Wv)0° > (Wv)30° ≥ (Wv)45° ≥ (Wv)60° >> (Wv)15°. The combined effect of the inclination angle in deposition and the quantity of voids in the film on the film density can be evaluated on the basis of the product of Wv and ρ ( = Wvρ). Due to the fact that (Wv)15° is much smaller than those of the other four specimens, the Wvρ results have the film density sequence of (Wvρ)0° > (Wvρ)30° > (Wvρ)45° > (Wvρ)60° > (Wvρ)15°. This Wvρ sequence for these five specimens is exactly consistent with that of the IN parameter. From the definition for IR, a higher value of IR means a film with a greater content of InGaO3(ZnO)3. Since the grain size of InGaO3(ZnO)3 is larger than that of InGaZnO4, a great content of InGaO3(ZnO)3 implies a film with a larger grain size and a lower material density. By the relation of IN and density mentioned previously, the sequence for the IR values of these five specimens is almost the same as that for the IN parameter except the order exchange of either of these two parameters at 45° and 60°. This characteristic implies that an increase in IR is advantage for the rise of n.
The k curves in the five specimens were found to be monotonically decreasing in the full or a part of the wavelengths of visible light. The ones for the IGZO_0° and IGZO_45° specimens can be categorized to the class of the full wavelength, while the remaining three specimens had k values that dropped to zero at a wavelength beyond which the k value remained unchanged. k = 0 prevailed for the entire wavelengths of visible light in the IGZO_60° specimen. Herein, Δk kmax – kmin, where kmax and kmin denote the maximum and minimum k values, respectively. The Δk values of the specimens prepared by a nonzero inclination angle are always lower than that of the IGZO_0° specimen. The k gradient (kg) is defined as , where λmax and λmin are the wavelengths corresponding to the kmax and kmin values in the visible light region, respectively. The kg results in Table 10 show that the negative kg value increases with increasing inclination angle. The contrast between the IRO2 in Table 8 and the kg in Table 10 indicates that the negative kg, which is an indication of the decreasing rate of k with respect to light wavelength, rises significantly by increasing the O2 intensity ratio (IRO2), especially for those at high inclination angles. The optical properties of transmission (T) and reflection (R), and the refractive index (n) and extinction coefficient (k) of the IGZO_0° specimen are shown to compare with those reported in the literature. The rough ranges of T and R in the visible light wavelengths are 64% ~90% and 36% ~10%, respectively. They are fairly close to the results of these two parameters shown in the studies of Aoi et al.  and Liu et al. , in which the specimens were prepared by coating conditions distinct from those of this study. The n and k values in the present study vary between 2.08 ~2.25 and 0.055 ~0.52, respectively. They are found to be also close to those results in the study of Galca et al. . Very few literature studies had the reports as to the mechanical properties of hardness and reduced modulus. Since these two mechanical properties are strongly dependent upon the film thickness, the substrate material, and the indentation depth, comparisons for the measurements of these two properties make sense only if the above three controlling factors are available.
The PL intensities evaluated for the IGZO_0°, 15°, 30°, 45°, and 60° have been obtained and shown in Figs. 12(a)–12(e). These PL spectra, in the wavelength region of 300 to 800 nm, show undulations with their peaks present in the neighborhoods of three fixed wavelengths. These peaks were identified as being associated with some colors including ultraviolet. Therefore, considering these wavelengths, nearly Gaussian deconvolutions were made for every PL spectrum. The peaks, with their wavelengths and intensities, are shown in Table 11 with the possible color indicated. The wavelengths in the range of 359.57 nm to 365.24 nm in the peak-1 class were identified as ultraviolet; they are close to the band edge emission which is ascribed to recombination of free excitons through exciton-exciton collision process . The wavelengths in the range of 395.57 nm to 404.37 nm in the peak-2 class were identified as violet; they were created due to free exciton recombination or donor-acceptor transition . The wavelengths in the broad visible range from 400 nm to 800 nm were attributed to various defects such as Zn vacancies, Zn interstitials or O vacancies [47, 49, 50]. Wavelengths in the peak-3 class were distributed in a relatively wide range. The color for these peaks varies from violet to green. The peak with green emission should be related to O vacancies . The blue emission is attributed to an electron transition from a shallow donor level of an O vacancy to the valence band . For the IGZO_0° and 60° specimens, their peak-3 was identified as being green; for the IGZO_15° specimen’s peak-3 wavelength (450.16 nm) was identified as being in the mixing region between violet and blue; and, the peak-3 values for the IGZO_30° and IGZO_45° were classified as blue. The three peak intensities of the IGZO_15° specimen have the minimum values of the five IGZO specimens, respectively; while those of the IGZO_60° specimen have the respective maximum values. As the inclination angle increases to ≥ 30°, the peak intensities of these three classes rise significantly such that they are much higher than the corresponding peak of IGZO_0°, particularly for the peak-3 intensity in IGZO_60° associated with green. Film thickness, IR (in Table 5) and IRO2 (in Table 8) appear to be the controlling factors to the intensities of peak-1 to peak-3 shown in Table 11; however, it seems hard to establish the simple relationships among these three peak intensities and these three controlling factors. Nevertheless, increases in specimen IR and IRO2 values are advantageous for increases in the PL intensities of these three peaks; while conversely, increases in film thickness are favorable for reductions in the PL intensities of these peaks. The combined effect of these three controlling factors has caused the values of peak-1 to peak-3 for IGZO_60° to be the highest and those of the IGZO_15° specimen to be the lowest amongst all five specimens. The above IRO2 parameter characteristic has also been reported in Pu et al. . Increases in the inclination angle caused oxygen vacancy (Vo) formation, and such vacancies were reported as being responsible for the green emission (GE) . However, this is only partly true in the present study because the peak-3 wavelengths of the five specimens were distributed over a relatively wider region, including violet and mixed violet-blue emissions. Violet emission (VE) peaks at wavelengths ranging from 395.57 nm to 404.37 nm are related to a transition between the shallow donor levels of the zinc interstitial (Zni) [15, 40–42] or double ionized oxygen vacancy (Vo2+) . In the PL analyses, the peak-1 intensities (Iultraviolet) show the following sequence (Iultraviolet)60° > (Iultraviolet)45° > (Iultraviolet)30° > (Iultraviolet)0° > (Iultraviolet)15°; while the peak-2 intensities show the sequence (Iviolet)60° (Iviolet)45° > (Iviolet)30° > (Iviolet)0° > (Iviolet)15°. The peak-1 and 2 intensities were affected by both IR and IRO2 rather than the individual one. Let the value of (IRO2) × (IR) be defined as CI, and in the sequence of (CI = 0.448)60° > (CI = 0.428)30° > (CI = 0.420)45° > (CI = 0.414)0° > (CI = 0.396)15°. The sequences shown in the parameters of Iviolet and Iultraviolet are also quite close to that of CI except for the order exchange of either of these two parameters evaluated at 30° and 45°. We can thus conclude that Iviolet and Iultraviolet are strongly dependent upon the combined effect of IRO2 and IR. Increases in the product of IRO2 and IR are advantageous for increases in Iviolet and Iultraviolet.
Application of nonzero inclination angles in the specimens can lower the intensities of O1 and O2 below those of the IGZO_0° specimen. Increases in the inclination angle can increase the IRO2 value related to oxygen vacancies; however, nonzero inclination angles can cause the IR values to always be lower than that of the IGZO_0° specimen. The combined results of increasing the oxygen-related vacancies in terms of IRO2, and decreasing the IGZO film thickness while increasing the number of In-O and Ga-O bonds due to the increase in the inclination angle can increase the specimen’s transmission. Reflection is strongly affected by the thin film topography. The specimens with a larger value of mean surface roughness result in a higher reflection. Due to the changes in IR and IRO2, increasing the inclination angle is advantageous for lowering the absorption in wavelengths shorter than 400 nm only. Increases in IR or film density are advantageous for the increases of refractive index, n; the IR parameter for specimens prepared with a nonzero inclination angle always has its value smaller than that of the IGZO_0° specimen. The Δk ( = kmax - kmin) values for the specimens prepared by nonzero inclination angles are lower than that of the IGZO_0° specimen. The decreasing rate of k with respect to light wavelength increases significantly with increasing IRO2, especially for those created at high inclination angles. The deconvolutions of micro-photoluminescences can obtain: a peak-1 associated with ultraviolet emission; a peak-2 associated with violet emission; and, a peak-3 associated with emissions varying in a wide range of light wavelengths, dependent upon the inclination angle. IR and IRO2 appear to be the controlling factors for the intensities of these three peaks. Increasing IR and IRO2 is advantageous for raising the PL intensities of these three peaks. Increasing the film thickness is disadvantageous for the PL intensities. The peak intensities, Iviolet and Iultraviolet, created at various inclination angles, are strongly dependent upon both IRO2 and IR. Moreover, increases in the product of IRO2 and IR promote increases of Iviolet and Iultraviolet.
References and links
1. D. C. Paine, B. Yaglioglu, Z. Beiley, and S. Lee, “Amorphous IZO-based transparent thin film transistors,” Thin Solid Films 516(17), 5894–5898 (2008). [CrossRef]
2. K. Song, D. Kim, X. S. Li, T. Jun, Y. Jeong, and J. Moon, “Solution processed invisible all-oxide thin film transistors,” J. Mater. Chem. 19(46), 8881–8886 (2009). [CrossRef]
3. K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. Hirano, and H. Hosono, “Room-temperature fabrication of transparent flexible thin-film transistors using amorphous oxide semiconductors,” Nature 432(7016), 488–492 (2004). [CrossRef] [PubMed]
4. A. Sato, K. Abe, R. Hayashi, H. Kumomi, K. Nomura, T. Kamiya, M. Hirano, and H. Hosono, “Amorphous In–Ga–Zn–O coplanar homojunction thin-film transistor,” Appl. Phys. Lett. 94(13), 133502 (2009). [CrossRef]
5. J. S. Park, T. W. Kim, D. Stryakhilev, J. S. Lee, S. G. An, Y. S. Pyo, D. B. Lee, Y. G. Mo, D. U. Jin, and H. K. Chung, “Flexible full color organic light-emitting diode display on polyimide plastic substrate driven by amorphous indium gallium zinc oxide thin-film transistors,” Appl. Phys. Lett. 95(1), 013503 (2009). [CrossRef]
6. J. H. Kim, D. K. Seo, C. H. Ahn, S. W. Shin, H. H. Cho, and H. K. Cho, “Hybrid solution processed InGaO3(ZnO)m thin films with periodic layered structures and thermoelectric properties,” J. Mater. Chem. 22(32), 16312–16317 (2012). [CrossRef]
7. T. Chen, M. Y. Wu, R. Ishihara, K. Nomura, T. Kamiya, H. Hosono, and C. I. M. Beenakker, “Excimer laser crystallization of InGaZnO4 on SiO4 substrate,” J. Mater. Sci. Mater. Electron. 22(11), 1694–1696 (2011). [CrossRef]
8. S. H. Yang, J. Y. Kim, M. J. Park, K. H. Choi, J. S. Kwak, H. K. Kim, and J. M. Lee, “Low resistance ohmic contacts to amorphous IGZO thin films by hydrogen plasma treatment,” Surf. Coat. Tech. 206(24), 5067–5071 (2012). [CrossRef]
9. S. J. Kang, J. Baik, H. J. Shin, J. G. Chung, K. H. Kim, and J. Lee, “X-ray photoelectron spectroscopic investigation of air-annealed amorphous In-Ga-Zn-O thin-film surface electronic and photonic devices, and systems,” ECS J. Solid State Sci. Technol. 2(9), 192–194 (2013). [CrossRef]
10. A. Suresh, P. Gollakota, P. Wellenius, A. Dhawan, and J. F. Muth, “Transparent, high mobility InGaZnO thin films deposited by PLD,” Thin Solid Films 516(7), 1326–1329 (2008). [CrossRef]
11. C. H. Jung, H. I. Kang, and 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, 125–129 (2013). [CrossRef]
12. A. Thakur, H. Yoo, S. J. Kang, J. Y. Baik, I. J. Lee, H. K. Lee, K. Kim, B. Kim, S. Jung, J. Park, and H. J. Shin, “Effects of substrate temperature on structural, electrical and optical properties of amorphous In-Ga-Zn-O thin films,” ECS J. Solid State Sci. Technol. 1(1), 11–15 (2012). [CrossRef]
13. T. T. T. Nguyen, O. Renault, B. Aventurier, G. Rodriguez, J. P. Barnes, and F. Templier, “Analysis of IGZO thin-film transistors by XPS and relation with electrical characteristics,” J. Disp. Technol. 9(9), 770–774 (2013). [CrossRef]
14. P. Liu, T. P. Chen, Z. Liu, C. S. Tan, and K. C. Leong, “Effect of O2 plasma immersion on electrical properties and transistor performance of indium gallium zinc oxide thin films,” Thin Solid Films 545, 533–536 (2013). [CrossRef]
15. S. H. Jung, H. J. Moon, M. K. Ryu, K. I. Cho, B. S. Bae, and E. J. Yun, “The effects of high-energy electron beam irradiation on the properties of IGZO thin films prepared by rf magnetron sputtering,” J. Ceram. Process. Res. 13, s246–s250 (2012).
16. L. Li, L. Fan, Y. Li, Z. Song, F. Ma, and C. Liu, “Effect of thermal annealing on the properties of transparent conductive In–Ga–Zn oxide thin films,” J. Vac. Sci. Technol. 32(2), 021506 (2014). [CrossRef]
17. J. Yao, L. Gong, L. Xie, and 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, 21–25 (2013). [CrossRef]
18. D. Kang, I. Song, C. Kim, Y. Park, T. D. Kang, H. S. Lee, J. W. Park, S. H. Baek, S. H. Choi, and H. Lee, “Effect of Ga/In ratio on the optical and electrical properties of GaInZnO thin films grown on SiO2/Si substrates,” Appl. Phys. Lett. 91(9), 091910 (2007). [CrossRef]
19. K. Nomura, T. Kamiya, H. Yanagi, E. Ikenaga, K. Yang, K. Kobayashi, M. Hirano, and H. Hosono, “Subgap states in transparent amorphous oxide semiconductor, In–Ga–Zn–O, observed by bulk sensitive x-ray photoelectron spectroscopy,” Appl. Phys. Lett. 92(20), 202117 (2008). [CrossRef]
20. J. W. Park, P. S. Jeong, S. H. Choi, H. Lee, B. H. Kong, and H. K. Cho, “Optical and structural properties of ion-implanted InGaZnO thin films studied with spectroscopic ellipsometry and transmission electron microscopy,” Jpn. J. Appl. Phys. 48(11), 111603 (2009). [CrossRef]
21. M. M. Hawkeye and M. J. Brett, “Glancing angle deposition: fabrication, properties, and applications of micro- and nanostructured thin films,” J. Vac. Sci. Technol. A 25(5), 1317–1335 (2007). [CrossRef]
22. M. Suzuki and Y. Taga, “Numerical study of the effective surface area of obliquely deposited thin films,” J. Appl. Phys. 90(11), 5599–5605 (2001). [CrossRef]
23. M. W. Seto, B. Dick, and M. J. Brett, “Microsprings and microcantilevers: studies of mechanical response,” J. Micromech. Microeng. 11(5), 582–588 (2001). [CrossRef]
24. M. Harris, M. Bowden, and H. A. Macleod, “Refractive index variations in dielectric films having columnar microstructure,” Opt. Commun. 51(1), 29–32 (1984). [CrossRef]
26. S. Wang, G. Xia, H. He, K. Yi, J. Shao, and Z. Fan, “Structural and optical properties of nanostructured TiO2 thin films fabricated by glancing angle deposition,” J. Alloys Compd. 431(1-2), 287–291 (2007). [CrossRef]
27. S. L. Zhang and J. C. M. Li, “Anisotropic elastic moduli and Poisson’s ratios of a poly(ethylene terephthalate) film,” J. Polym. Sci. Pt. B-Polym, Phys. 42, 260–266 (2004).
28. K. T. Huang and H. C. Chen, “Automatic measurement and stress analysis of ITO/PET flexible substrate by shadow moiré interferometer with phase-shifting interferometry,” J. Disp. Technol. 10(7), 609–614 (2014). [CrossRef]
29. G. H. Kim, H. S. Kim, H. S. Shin, B. D. Ahn, K. H. Kim, and H. J. Kim, “Inkjet-printed InGaZnO thin film transistor,” Thin Solid Films 517(14), 4007–4010 (2009). [CrossRef]
30. Y. J. Lin, C. L. Tsai, Y. M. Lu, and C. J. Liu, “Optical and electrical properties of undoped ZnO films,” J. Appl. Phys. 99(9), 093501 (2006). [CrossRef]
31. D. H. Fan, Z. Y. Ning, and M. F. Jiang, “Characteristics and luminescence of Ge doped ZnO films prepared by alternate radio frequency magnetron sputtering,” Appl. Surf. Sci. 245(1-4), 414–419 (2005). [CrossRef]
32. J. F. Moulder, W. F. Stickle, P. E. Sobol, and K. D. Bomben, Handbook of X-ray Photoelectron Spectroscopy (Perkin-Elmer Corporation: Eden Prairie, 1992).
33. S. Y. Bae, C. W. Na, J. H. Kang, and J. Park, “Comparative structure and optical properties of Ga-, In-, and Sn-doped ZnO nanowires synthesized via thermal evaporation,” J. Phys. Chem. B 109(7), 2526–2531 (2005). [CrossRef] [PubMed]
34. B. Kumar, H. Gong, and R. Akkipeddi, “A study of conduction in the transition zone between homologous and ZnO-rich regions in the In2O3–ZnO system,” J. Appl. Phys. 97(6), 063706 (2005). [CrossRef]
35. Y. S. Rim, D. L. Kim, W. H. Jeong, and H. J. Kim, “Effect of Zr addition on ZnSnO thin-film transistors using a solution process,” Appl. Phys. Lett. 97(23), 233502 (2010). [CrossRef]
36. T. Kamiya and H. Hosono, “Material characteristics and applications of transparent amorphous oxide semiconductors,” NPG Asia Mater. 2(1), 15–22 (2010). [CrossRef]
37. S. Jeong, Y. G. Ha, J. Moon, A. Facchetti, and T. J. Marks, “Role of gallium doping in dramatically lowering amorphous-oxide processing temperatures for solution-derived indium zinc oxide thin-film transistors,” Adv. Mater. 22(12), 1346–1350 (2010). [CrossRef] [PubMed]
38. B. Y. Su, S. Y. Chu, Y. D. Juang, and S. Y. Liu, “Effects of Mg doping on the gate bias and thermal stability of solution-processed InGaZnO thin-film transistors,” J. Alloys Compd. 580, 10–14 (2013). [CrossRef]
39. X. W. Zhou, R. A. Johnson, and H. N. G. Wadley, “A molecular dynamics study of nickel vapor deposition: temperature, incident angle, and adatom energy effects,” Acta Mater. 45(4), 1513–1524 (1997). [CrossRef]
40. K. Matsubara, P. Fons, K. Iwata, A. Yamada, and S. Niki, “Room-temperature deposition of Al-doped ZnO films by oxygen radical-assisted pulsed laser deposition,” Thin Solid Films 422(1-2), 176–179 (2002). [CrossRef]
41. D. Beena, K. J. Lethy, R. Vinodkumar, V. P. Mahadevan Pillai, V. Ganesan, D. M. Phase, and S. K. Sudheer, “Effect of substrate temperature on structural, optical and electrical properties of pulsed laser ablated nanostructured indium oxide films,” Appl. Surf. Sci. 255(20), 8334–8342 (2009). [CrossRef]
42. C. Yang, X. M. Li, X. D. Gao, X. Cao, R. Yang, and Y. Z. Li, “Effects of the oxygen pressure on the structural and optical properties of ZnBeMgO films prepared by pulsed laser deposition,” J. Cryst. Growth 312(7), 978–981 (2010). [CrossRef]
43. R. N. Tait, T. Smy, and M. J. Brett, “Modelling and characterization of columnar growth in evaporated films,” Thin Solid Films 226(2), 196–201 (1993). [CrossRef]
44. T. Aoi, N. Oka, Y. Sato, R. Hayashi, H. Kumomi, and Y. Shigesato, “DC sputter deposition of amorphous indium–gallium–zinc–oxide (a-IGZO) films with H2O introduction,” Thin Solid Films 518(11), 3004–3007 (2010). [CrossRef]
45. S. J. Liu, S. H. Su, H. W. Fang, J. H. Hsieh, and J. Y. Juang, “Effects of Cr doping on physical properties of amorphous In–Ga–Zn–O films,” Appl. Surf. Sci. 257(23), 10018–10021 (2011). [CrossRef]
46. A. C. Galca, G. Socol, and V. Craciun, “Optical properties of amorphous-like indium zinc oxide and indium gallium zinc oxide thin films,” Thin Solid Films 520(14), 4722–4725 (2012). [CrossRef]
47. N. Tiwari, H. P. D. Shieh, and P.-T. Liu, “Structural, optical, and photoluminescence study of ZnO/IGZO thin film for thin film transistor application,” Mater. Lett. 151, 53–56 (2015). [CrossRef]
48. K. U. Sim, S. W. Shin, A. V. Moholkar, J. H. Yun, J. H. Moon, and J. H. Kim, “Effects of dopant (Al, Ga, and In) on the characteristics of ZnO thin films prepared by RF magnetron sputtering system,” Curr. Appl. Phys. 10(3), S463–S467 (2010). [CrossRef]
49. J. L. Zhao, X. W. Sun, H. Ryu, and Y. B. Moon, “Thermally stable transparent conducting and highly infrared reflective Ga-doped ZnO thin films by metal organic chemical vapor deposition,” Opt. Mater. 33(6), 768–772 (2011). [CrossRef]
50. F. Wu, L. Fang, Y. J. Pan, K. Zhou, H. B. Ruan, G. B. Liu, and C. Y. Kong, “Effect of annealing treatment on structural, electrical, and optical properties of Ga-doped ZnO thin films deposited by RF magnetron sputtering,” Thin Solid Films 520(2), 703–707 (2011). [CrossRef]
51. Q. Li, Z. X. Song, F. Ma, Y. H. Li, and K. W. Xu, “Effects of high-temperature thermal annealing on the electronic properties of In-Ga-Zn oxide thin films,” J. Vac. Sci. Technol. A 33(2), 021520 (2015). [CrossRef]
52. D. H. Zhang, Z. Y. Xue, and Q. P. Wang, “The mechanisms of blue emission from ZnO films deposited on glass substrate by r.f. magnetron sputtering,” J. Phys. D Appl. Phys. 35(21), 2837–2840 (2002). [CrossRef]
53. H. Pu, Q. Zhou, L. Yue, and Q. Zhang, “Investigation of oxygen plasma treatment on the device performance of solution-processed a-IGZO thin film transistors,” Appl. Surf. Sci. 283, 722–726 (2013). [CrossRef]
54. T. Prasada Rao, M. C. Santhosh Kumar, A. Safarulla, V. Ganesan, S. R. Barman, and C. Sanjeeviraja, “Physical properties of ZnO thin films deposited at various substrate temperatures using spray pyrolysis,” Physica B 405(9), 2226–2231 (2010). [CrossRef]