Investigation of the Structural, Electrical, and Optical Properties of the ...

nanomaterials Article

Investigation of the Structural, Electrical, and Optical Properties of the Nano-Scale GZO Thin Films on Glass and Flexible Polyimide Substrates Fang-Hsing Wang 1 , Kun-Neng Chen 2 , Chao-Ming Hsu 3 , Min-Chu Liu 1 and Cheng-Fu Yang 4, * 1

2 3 4

*

Department of Electrical Engineering and Graduate Institute of Optoelectronic Engineering, National Chung Hsing University, Taichung 402, Taiwan; [email protected] (F.-H.W.); [email protected] (M.-C.L.) Department of Electrical Engineering, Kun-Shan University, Tainan 710, Taiwan; [email protected] Department of Mechanical Engineering, National Kaohsiung University of Applied Science, Kaohsiung 807, Taiwan; [email protected] Department of Chemical and Materials Engineering, National University of Kaohsiung, Kaohsiung 81141, Taiwan Correspondence: [email protected]; Tel.: +886-7-591-9283; Fax: +886-7-591-9277

Academic Editor: Chih-hung Chang Received: 7 March 2016; Accepted: 29 April 2016; Published: 10 May 2016

Abstract: In this study, Ga2 O3 -doped ZnO (GZO) thin films were deposited on glass and flexible polyimide (PI) substrates at room temperature (300 K), 373 K, and 473 K by the radio frequency (RF) magnetron sputtering method. After finding the deposition rate, all the GZO thin films with a nano-scale thickness of about 150 ˘ 10 nm were controlled by the deposition time. X-ray diffraction patterns indicated that the GZO thin films were not amorphous and all exhibited the (002) peak, and field emission scanning electron microscopy showed that only nano-scale particles were observed. The dependences of the structural, electrical, and optical properties of the GZO thin films on different deposition temperatures and substrates were investigated. X-ray photoemission spectroscopy (XPS) was used to measure the elemental composition at the chemical and electronic states of the GZO thin films deposited on different substrates, which could be used to clarify the mechanism of difference in electrical properties of the GZO thin films. In this study, the XPS binding energy spectra of Ga2p3/2 and Ga2p1/2 peaks, Zn2p3/2 and Zn2p1/2 peaks, the Ga3d peak, and O1s peaks for GZO thin films on glass and PI substrates were well compared. Keywords: Ga2 O3 -doped ZnO (GZO) thin film; glass; polyimide (PI); X-ray photoelectron spectroscopy (XPS)

1. Introduction Recent advances in nanotechnology have contributed to the development of photon upconversion materials as promising new-generation candidates of fluorescent bioprobes and spectral converters for biomedical and optoelectronic applications [1]. Also, dimension-reduced process to make them approaching two-dimensional nanostructures into precisely controlled lamellar or called layer nanomaterials are currently achievable. In particular, layer-by-layer (LbL) assembly is known as a highly versatile method for fabrication of controlled layered structures from various kinds of component materials [2]. Zhuang et al. published a review focusing on networks‘ growth processes of metal-organic frameworks and crystalline coordination from the view of surface chemistry [3]. They mainly focused on the methods involving liquid phases, and those methods can be summarized under the term of “liquid phase epitaxy” (LPE). The LPE method can deposit thin films having crystalline, high quality coatings and with predefined orientations as the thin films are deposited on the different Nanomaterials 2016, 6, 88; doi:10.3390/nano6050088

www.mdpi.com/journal/nanomaterials

Nanomaterials 2016, 6, 88

2 of 13

substrates. Indium-tin oxide (ITO)-based transparent conduction oxide (TCO) thin films have been widely used in the applications of solar cells, flat panel displays [4], and other optoelectronic products due to their excellent conductivity and transparency. Much more interest has been given to ZnO-based thin films such as undoped ZnO thin films [5], Al-doped ZnO (AZO) thin films [6], and F-doped ZnO (FZO) thin films [7] because they have the stability properties even under hydrogen plasma. For that, they are potential candidates for the applications of solar cell technology based on thin-film silicon. Al-doped ZnO (AZO) are the major materials for ZnO-based thin films, because they present favorable electrical properties and can be investigated as indium-free TCO thin films. In spite of studies on the preparation, characterization, and effect of doping on the properties of ZnO-based thin films, certain effects of either some dopants or preparation procedures still remain unclear. Because Ga is less receptive to oxidation, Ga2 O3 -doped ZnO TCO materials have been reported to have a better stable property [8,9]. When Ga2 O3 is served as a dopant, the resistivity of Ga2 O3 -doped ZnO thin films is lower than that using other group-III elements [10,11], because Ga has a more compatible covalent bond length with Zn than other group-III elements [12]. TCO thin films deposited on different substrates [13], for example glass and polyimide (PI), are widely used throughout the semiconductor and electronics industries. The polymer substrates are cheaper, lighter, and more flexible as compared with the conventional glass substrates; they could be effectively used in applications such as flexible display and flexible solar cells. For that, the necessity of studying the deposition process of TCO thin films on polymer substrates has been increased, as the polymer substrates are suitable for flat-panel displays (FPDs) and optoelectronics [14–16]. Many researchers have reported on the Ga2 O3 -doped ZnO thin films in regard to the different doping concentrations of Ga2 O3 . In the past, the highly conductive and transparent Ga2 O3 -doped ZnO thin films had been deposited at high growth rates by radio frequency magnetron sputtering. In the present paper, a ceramic target of ZnO with 3 wt % Ga2 O3 (ZnO:Ga2 O3 = 97:3 in wt %, abbreviated as GZO) was used to deposit the GZO thin films on the Eagle 173 glass and flexible polyimide (PI) substrates by using the radio frequency (RF) magnetron sputtering process. Then the detail analysis of defects for nano-scale thin films on different substrtaes could be achieved by X-ray photoelectron spectroscopy (XPS) [3]. In this study, the dependences of the structural, electrical, and optical properties of the GZO thin films on different deposition temperatures and substrates were investigated. Also, XPS is a surface-sensitive and a quantitative spectroscopic technique that can be used to measure the variation of electron or x-ray excited from one lower energy layer to one higher energy layer. In order to clarify the mechanism of difference in electrical properties, we would investigate the change of the chemical structures of the GZO thin films deposited on different substrates by XPS. 2. Experimental Procedures In this work, the radio frequency (RF) (13.56 MHz) magnetron sputtering process was used to deposit the GZO thin films. ZnO (97 wt %, 5 N, Admat Inc., Norristown, PA, USA) and Ga2 O3 (3 wt %, 5 N, Admat Inc., Norristown, PA, USA) were mixed, ground, calcined at 1273 K for 2 h, and sintered at 1673 K for 2 h to form the ceramic target with a 2 in diameter. The area of Eagle 173 glass (Corning Inc., New York, NY, USA) and polyimide (abbreviated as PI, Taimide Tech. Inc., Hsinchu County, Taiwan) were used as substrates and their areas were about 33 mm ˆ 33 mm. Before the deposition process was started, the base chamber pressure of the sputtering system was pumped to less than 1 ˆ 10´6 Torr, then the deposition parameters were controlled at different pressures and powers. The optimal deposition parameters were a RF power of 50 W and a working pressure of 5 mTorr because the deposited GZO thin films had the flattest surface and the acceptable deposition rate. During the deposition process, only pure argon (99.999%) was introduced in the chamber, the flow rate of argon was 20 sccm. Deposition temperature was also used as a parameter, where GZO thin films were deposited at room temperature (RT, 300 K), 373 K, and 473 K, respectively. The RT was characterized as that the chamber was unheated and the desired temperatures of 373 K and 473 K were achieved by heating of the sample stage using a neon lamp and monitoring the temperature

Nanomaterials 2016, 6, 88

3 of 13

using a thermocouple. Thickness of the GZO thin films was one of the most important parameters to influence the characteristics of the GZO thin films. For this reason, thicknesses of the GZO thin films were measured using field emission scanning electron microscopy (FESEM, JSM-6700, JEOL, Tokyo, Japan). Deposition rate and thickness of the GZO thin films were determined by averaging five data sets obtained by FESEM. After calculating the deposition rate, the thicknesses of all GZO thin films were about 150 ˘ 10 nm by controlling the deposition time. The roughness (or flatness) was measured using atomic force microscopy (AFM), and the crystalline structure was measured using X-ray diffraction (XRD) patterns with Cu Kα radiation. The electrical properties of the GZO thin films were determined by Hall effect measurement (Ecopia, HMS-3000, Bridge Technol., Chandler Heights, AZ, USA) at room temperature using the Van der Pauw method with four pressed indium balls onto the corners of the samples under a 0.55 T magnetic field. While the optical transmittances of the GZO thin films on glass and PI substrates were measured by using an ultraviolet-visible spectroscopy (UV-Vis) spectrophotometer (Hitachi U3300, Kenichi Sato, Osaka, Japan). The surface chemical composition and bonding of the GZO thin films were analyzed using the X-ray photoemission spectroscopy (XPS, ULVAC-PHI, 5000 Versaprobe, Physical (PHI) Electronics, Chanhassen, MN, USA) after the thin films’ surfaces were pre-cleaning for 3 min by sputtering. The parameters for XPS were monochromatized Al Kα 187.85 eV, presputter setting: 2 kV, sputter area: 2 mm ˆ 2 mm, sputter rate: 12.5 nm/min, sputter time: 0.3 min, and the Ar ions with energy of ~2 keV, respectively. 3. Results and Discussion Because the GZO thin films were deposited using a sputtering method in a pure Ar atmosphere, we believe that substrate would be the important factor that would influence the properties of the deposited thin films. We used FESEM to observe the surface morphologies of the GZO thin films deposited on different substrates under different deposition temperatures. The morphologies of the GZO thin films as a function of the deposition substrate and temperature are shown in Figure 1, which indicates that as those deposition parameters are changed, the surface morphologies apparently changed as well. As Figure 1a–e show, only nano-scale particles were observed, and the surface morphologies had no apparent change, regardless of the variations of substrates and deposition temperatures. As RT (Figure 1a,b) and 373 K (as Figure 1c,d show) were used as the deposition temperatures, the GZO thin films deposited on glass substrates had less pores than those deposited on PI substrates. As different deposition temperatures were used, the morphology of the GZO thin films exhibited a roughness surface. However, the variations of average crystallization sizes are dependent on the substrate and deposition temperature and they have not easily been calculated from the surface observation. We will illustrate the variations of grain sizes from the XRD patterns and the following equation [17], and the results are also shown in Table 1: D “ p0.9λq{pβcosθq

(1)

where D is the size of nano-scale particles; λ is 1.54 Å; β is the full width at half-maximum (FWHM); and θ is the diffraction angle, respectively. As Table 1 shows, the crystallization size increased with the increase of the deposition temperature. This is caused by the fact that as the deposition temperature increases, the surface energy of the GZO thin films during the deposition process increases, and the chance for particles’ growth of the GZO thin films increases, and then their crystallization sizes increase. The surface roughness was also calculated as a function of substrate and deposition temperature; the results of the RT-deposited GZO thin films are shown in Figure 2 and the results of all the GZO thin films are also compared in Table 1. The root mean square (RMS) surface roughness of the GZO thin films on glass and PI substrates was measured to be 0.30 nm by AFM. As the deposition temperature was RT, 373 K, and 473 K, the measured values were 1.46 nm, 1.25 nm, and 1.12 nm when the substrate was glass and the measured values were 1.44 nm, 1.22 nm, and 1.10 nm when the substrate was PI, respectively. Apparently, the flatness was improved as the deposition temperature was raised.

Nanomaterials 2016, 6, 88

4 of 13

Nanomaterials 2016, 6, 88

4 of 13

Nanomaterials 2016, 6, 88

4 of 13

Figure 1. Surface morphologies of the Ga2O3-doped ZnO (GZO) thin films as a function of substrate

Figure 1. Surface morphologies of the Ga2 O3 -doped ZnO (GZO) thin films as a function of substrate and deposition temperature. (a) Glass; (b) polyimide (PI); (c) 373 K-Glass; (d) 373 K-PI; (e) 473 K-Glass; and deposition temperature. (a) Glass; (b) polyimide (PI); (c) 373 K-Glass; (d) 373 K-PI; (e) 473 K-Glass; and (f) 473 K-PI, respectively. and (f) 473 K-PI, respectively. Figure 1. Surface morphologies of the Ga2O3-doped ZnO (GZO) thin films as a function of substrate Table 1. The 2θtemperature. value of the (002) peak, size (D), constant full(e) width at halfand deposition (a) Glass; (b)crystallite polyimide (PI); (c)lattice 373 K-Glass; (d)(c), 373the K-PI; 473 K-Glass;

Tablemaximum 1. (f) The value of the and (002) peak, crystallite size (D), ZnO lattice constant (c), the full width (FWHM) value, roughness of the Ga2O 3-doped (GZO) thin films deposited on at and 4732θ K-PI, respectively. different substrates and at different temperatures and ZnO deposited on different substrates. RT: half-maximum (FWHM) value, and roughness of the Ga2 O3 -doped ZnO (GZO) thin films deposited on Table 1. The 2θand value the (002)temperatures peak, crystallite size (D),deposited lattice constant (c), the full width at RT: half-room room temperature; PI: polyimide. different substrates at of different and ZnO on different substrates. maximum (FWHM) value, and roughness of the Ga 2O3-doped ZnO (GZO) thin films deposited on temperature; PI: polyimide. Substrates 2θ (°) Crystallite Size (nm) c (nm) FWHM (°) Roughness (nm) different substrates and at different temperatures and ZnO deposited on different substrates. RT: RT glass 34.20 31.34 0.5240 0.407 1.46 room temperature;˝ PI: polyimide. ˝

Substrates 2θ ( 34.14 ) RT PI Substrates 2θ (°) 373 K glass 34.20 34.24 RT glass RT glass 34.20 RT PI373 K PI 34.14 34.18 PI 34.14 373 K glass 34.24 473RT K glass 34.36 373 K373 PI 34.24 473KKglass PI 34.18 34.36 473 KRT glass 373 K glass PI 34.36 34.18 ZnO 34.12 473 K473 PI K glass 34.36 34.36 RT ZnO PI 34.10 RT ZnO glass 473 K PI 34.12 34.36 RT ZnO PI 34.10 RT ZnO glass 34.12 RT ZnO PI 34.10

(a)

Crystallite Size (nm) 28.12 Crystallite Size (nm) 24.71 31.34 31.34 28.12 23.08 28.12 24.71 20.42 23.08 24.71 20.18 20.42 23.08 23.42 20.18 20.42 24.45 23.42 20.18 24.45 23.42 24.45

c (nm) 0.412 FWHM ( ) Roughness (nm) 0.5249 1.44 c (nm) (°) Roughness 0.5234 0.336 0.407 1.25 (nm) 0.5240 FWHM 1.46 0.5240 0.407 0.412 1.46 1.44 0.5249 0.360 0.5243 1.22 0.5249 0.412 0.336 1.44 1.25 0.5234 0.265 0.5216 1.12 0.5243 0.296 0.5234 0.336 0.360 1.25 1.22 0.5216 1.10 0.5216 0.382 0.5243 0.360 0.265 1.22 1.12 0.5252 1.24 0.5216 0.296 0.5216 0.265 1.12 1.10 0.5255 0.406 1.26 0.5252 0.382 0.5216 0.296 1.10 1.24 0.5255 0.406 1.26 0.5252 0.382 1.24 0.5255 0.406 1.26

(b)

Figure 2. Atomic force microscopy (AFM) analysis of the GZO thin films as a function of substrate at room temperature. (a) Glass and (b) PI, respectively.

(a)

(b)

XRD themicroscopy GZO thin(AFM) films analysis as a function of deposition and substrate Figurepatterns 2. Atomicof force of the GZO thin films as temperature a function of substrate at Figure 2. Atomic force microscopy (AFM) analysis of the GZO thin films as a function of substrate at temperature. andthe (b) GZO PI, respectively. wereroom shown in Figure(a) 3, Glass and all thin films exhibited the (002) peak. XRD patterns of the room temperature. (a) Glass and (b) PI, respectively. RT-deposited ZnO thin films are also added in Figure 3 as a reference for the GZO thin films. ofasthe thin films as peaks a function deposition temperature anddeposition substrate WhenXRD glasspatterns was used theGZO substrate, the (002) of theofGZO thin films prepared with were shown in and all thefilms thin the (002) XRD patterns the XRD patterns of GZO thin a films function deposition temperature andPIof substrate temperatures of Figure RT,the 3733, K, and 473 KGZO wereas situated atexhibited 2θ of = 34.20°, 34.24°,peak. and 34.36°; when was ZnO 3, thin films are GZO also added in Figure 3 as a the reference for theXRD GZO patterns thin films. were RT-deposited shown in Figure and all the thin films exhibited (002) peak. of the When glass was used as the substrate, the (002) peaks of the GZO thin films prepared with deposition RT-deposited ZnO thin films are also added in Figure 3 as a reference for the GZO thin films. When temperatures of RT, 373 K, and 473 K were situated at 2θ = 34.20°, 34.24°, and 34.36°; when PI was

glass was used as the substrate, the (002) peaks of the GZO thin films prepared with deposition

Nanomaterials 2016, 6, 88

5 of 13

temperatures of RT, 373 K, and 473 K were situated at 2θ = 34.20˝ , 34.24˝ , and 34.36˝ ; when PI was used as the substrate, the (002) peaks of the GZO thin films prepared with deposition temperatures 2016, 6, 88 5 of 13 of RT,Nanomaterials 373 K, and 473 K were situated at 2θ = 34.14˝ , 34.18˝ , and 34.36˝ , respectively. The (002) peaks of the GZO thin films prepared with a deposition temperature of RT on glass and PI substrates were used as the substrate, the (002) peaks of the GZO thin films prepared with deposition temperatures situated at 2θ = 34.12˝ and 34.10˝ . The lattice constant (c) was calculated by using the 2θ values shown of RT, 373 K, and 473 K were situated at 2θ = 34.14°, 34.18°, and 34.36°, respectively. The (002) peaks in Figure 3 and the results were also compared in Table 1. When glass was used as the substrate, of the GZO thin films prepared with a deposition temperature of RT on glass and PI substrates were the calculated constant was 0.5240 nm,constant 0.5234 nm, and 0.5215 nm for deposition temperatures situated atlattice 2θ = 34.12° and 34.10°. The lattice (c) was calculated by using the 2θ values shown of RT,in373 K, 3and K; when was used asinthe substrate, the calculated constant Figure and 473 the results werePI also compared Table 1. When glass was used as lattice the substrate, the was 0.5249calculated nm, 0.5243 nm,constant and 0.5215 fornm, deposition temperatures of RT, 373 K, andtemperatures 473 K, respectively. lattice was nm 0.5240 0.5234 nm, and 0.5215 nm for deposition of RT, 373 K,lattice and 473 K; when of PI the wasZnO usedthin as the substrate, the calculated wasglass 0.5249 The calculated constants films were 0.5252 nm andlattice 0.5255constant nm when and PI nm, 0.5243 nm, and 0.5215 nm for deposition temperatures of RT, 373 K, and 473 K, respectively. The were used as substrates. All the calculated lattice constants c of the GZO thin films in Table 1 being 3+ ions calculated lattice constants of the ZnOisthin films were 0.5252the nmradius and 0.5255 when(62 glass PI smaller than those of the ZnO thin films significant, because of Ganm pm)and is smaller were used as substrates. All the calculated lattice constants c of the GZO thin films in Table 1 being than that of Zn2+ ions (72 pm). Table 1 also shows an important result: the lattice constant of the GZO smaller than those of the ZnO thin films is significant, because the radius of Ga3+ ions (62 pm) is thin films decreases with 2+the increase of the deposition temperature, independent of the substrate. smaller than that of Zn ions (72 pm). Table 1 also shows an important result: the lattice constant of The lattice parameters of TCO-based thin films usually depend on the concentration of foreign atoms, the GZO thin films decreases with the increase of the deposition temperature, independent of the defects, external the difference of their ionic radii withdepend respectontothe theconcentration substituted matrix substrate. Thestrain, latticeand parameters of TCO-based thin films usually of 3+ ions will substitute ions. foreign Those results suggest that as the deposition temperature increases, more Ga atoms, defects, external strain, and the difference of their ionic radii with respect to the 3+ the sites of Zn2+ matrix ions and the lattice constant decreases. study oftemperature Li et al. suggested theGaoxygen substituted ions. Those results suggest that as theThe deposition increases,that more 2+ ions will substitute of parameters, Zn ions andatthe lattice decreases. The study of Li et al. vacancies might reduce the the sites lattice least the constant c value [18]. suggested the oxygen vacancies might the lattice parameters, c value As Figure that 3 shows, the full width at halfreduce maximum (FWHM) valuesatofleast the the (002) peak[18]. of the GZO As Figure 3 shows, the full width at half maximum (FWHM) values of the (002) peak of the GZO thin films deposited on different substrates were calculated, and the results are compared in Table 1. thin films deposited on different substrates were calculated, and the results are compared in Table 1. The FWHM values of the (002) peak for GZO thin films deposited on glass substrates were 0.407, 0.336, The FWHM values of the (002) peak for GZO thin films deposited on glass substrates were 0.407, and 0.265 for substrate temperatures of RT, 373 K, and 473 K; the FWHM values of the (002) peak for 0.336, and 0.265 for substrate temperatures of RT, 373 K, and 473 K; the FWHM values of the (002) GZO peak thin films deposited on PI substrates were 0.412, 0.360, and 0.292 for substrate temperatures of for GZO thin films deposited on PI substrates were 0.412, 0.360, and 0.292 for substrate RT, 373 K, and 473 respectively. These results suggest that the GZO thinthat films at higher temperatures ofK,RT, 373 K, and 473 K, respectively. These results suggest the deposited GZO thin films temperatures have a better crystalline structure. deposited at higher temperatures have a better crystalline structure.

30

40

50

60

2(Degree)

70

(002)

RT-ZnO RT-GZO 373K-GZO 473K-GZO

80

20

(004)

(004)

Intensity (a.u.) 20

(b)

Intensity (a.u.)

RT-ZnO RT-GZO 373K-GZO 473K-GZO

(002)

(a)

30

40

50

60

2(Degree)

70

80

Figure 3. X-ray diffraction (XRD) patterns of the undoped ZnO thin films and GZO thin films as a

Figure 3. X-ray diffraction (XRD) patterns of the undoped ZnO thin films and GZO thin films as a function of deposition temperature and on different substrates: (a) glass and (b) PI. a.u.: arbitrary unit. function of deposition temperature and on different substrates: (a) glass and (b) PI. a.u.: arbitrary unit.

Figure 4 shows the transmission spectra of the GZO thin films plotted against the different deposition temperature and on different substrates; measured wavelength was in the region of Figure 4 shows the transmission spectra of thethe GZO thin films plotted against the different 300–1000 nm. The transmission spectrum of the RT-deposited ZnO thin films is also added in Figure deposition temperature and on different substrates; the measured wavelength was in the region of 4 as anm. reference for the GZO thin films. For thin films, the absorption edges were blue-shifted 300–1000 The transmission spectrum ofthe theGZO RT-deposited ZnO thin films is also added in Figure 4 and the transmittance ratios were higher as compared with that of the RT-deposited ZnO thin films. as a reference for the GZO thin films. For the GZO thin films, the absorption edges were blue-shifted As Figure 4 shows, the maximum transmittance ratios of all the GZO thin films in the range of visible and the transmittance ratios were higher as compared with that of the RT-deposited ZnO thin films. light (400–700 nm) were more than 95%, regardless of deposition temperature and substrate. We As Figure 4 shows, the maximum transmittance ratios of all the GZO thin films in the range of visible know that many factors will affect the transmission spectrum of the GZO thin films. The increase of light (400–700 nm) were 95%, regardless and because substrate. We know the transmission ratiomore in thethan optical band is caused of bydeposition the increasetemperature in carrier density, a higher that many factors will affect transmission spectrum of thefewer GZOdefects. thin films. The increase of the deposition temperature canthe cause the GZO thin films to have The higher deposition transmission ratio in the optical is caused by the increase carrier density, higher temperature could even causeband the GZO thin films to have betterincrystallization; thebecause average a and deposition temperature can cause the GZO thin films to have fewer defects. The higher deposition temperature could even cause the GZO thin films to have better crystallization; the average and

Nanomaterials 2016, 6, 88

6 of 13

Nanomaterials 2016, 6, 88

6 of 13

maximum transmittance ratios of all the GZO thin films in the region of visible light had no apparent variation as thetransmittance deposition temperature Theinreason is that the optical band gap (Eg) of maximum ratios of all theincreased. GZO thin films the region of visible light had no apparent variation as theisdeposition temperature increased. Thethus reason that theenergy optical in band (Eg) of the the GZO thin films in the region of ultraviolet light, theisoptical thegap region visible GZO thin filmsthe is inelectrons, the region and of ultraviolet thefilms optical in the of light visiblewill lighthave light cannot excite then thelight, GZOthus thin in energy the range ofregion visible cannot excite the electrons, and then the GZOthe thinaverage films in transmittance the range of visible light have highthin high transmittance ratios. As Figure 4 shows, ratios of will all the GZO transmittance ratios. As Figure 4 shows, the average transmittance ratios of all the GZO thin films films in the range of visible light were about 90.9%–92.8% and in the range of near-infraredinlight the range of visible light were about 90.9%–92.8% and in the range of near-infrared light (700–1500 (700–1500 nm) they were about 84.8%–88.9%, respectively. The detailed optical properties of the GZO nm) they were about 84.8%–88.9%, respectively. The detailed optical properties of the GZO thin films thin films deposited on different substrates and at different temperatures and ZnO thin films deposited deposited on different substrates and at different temperatures and ZnO thin films deposited on on different substrates in Table Table2.2. different substratesare arecompared compared in 100

(a)

80

60

40 RT-GZO 373K-GZO 473K-GZO RT-ZnO

20

0

400

600

800

Wavelength (nm)

1000

Transmittance (%)

Transmittance (%)

100

(b)

80

60

40 RT-GZO 373K-GZO 473K-GZO RT-ZnO

20

0

400

600

800

Wavelength (nm)

1000

Figure 4. Transmittance spectra of the undoped ZnO thin films and GZO thin films deposited as a

Figure 4. Transmittance spectra of the undoped ZnO thin films and GZO thin films deposited as a function of deposition temperature and different substrates: (a) glass and (b) PI. function of deposition temperature and different substrates: (a) glass and (b) PI.

In the transmission spectra of all the GZO thin films, a greater sharpness was noticeable in the curves of the absorption edges, the GZO opticalthin band edgesa had no apparent shift as the deposition In the transmission spectra of and all the films, greater sharpness was noticeable in the increased. In theand past,the determination of the optical band gap (Eg) was often to curvestemperature of the absorption edges, optical band edges had no apparent shift as necessary the deposition developincreased. the electronic bandpast, structure of a thin-film However, using extrapolation temperature In the determination of material. the optical band gap (Eg) was often methods, necessary to the Eg values of the GZO thin films can be determined from the absorption edge for direct interband develop the electronic band structure of a thin-film material. However, using extrapolation methods, transition, which can be calculated using the relation in Equation (3) as follows:

the Eg values of the GZO thin films can be determined from the absorption edge for direct interband αhνrelation = C × (hν Eg)1/2 (2) transition, which can be calculated using the in−Equation (3) as follows: where C is the constant for the direct transition, h is Planck’s constant, and ν is the frequency of the “ C ˆcoefficient, phν ´ Egq1/2 incident photon [19]; α is the opticalαhν absorption which is calculated using Lambert’s law (2) as follows:

where C is the constant for the direct transition, h is Planck’s constant, and ν is the frequency of the α = ln((1/T)/h) (3) incident photon [19]; α is the optical absorption coefficient, which is calculated using Lambert’s law where 𝑇 and h are the thin film’s transmittance ratio and thickness. as follows: Figure 5 plots (αhν) against (hν) (energy) in accordance with Equation (3), and the Eg values can (3) α “ lnpp1{Tq{hq be found by extrapolating a straight line at (αhν) = 0. The calculated Eg values of the GZO thin films

function deposition temperature and onand different substrates are also shown in Figure 5. whereas T aand h are of thethe thin film’s transmittance ratio thickness. The linear dependence of (αhν) on (hν) indicates that the GZO films are direct Figure 5 plots (αhν) against (hν) (energy) in accordance withthin Equation (3), and transition-type the Eg values can semiconductors. As the deposition temperature increased from RT to 473 K and was used be found by extrapolating a straight line at (αhν) = 0. The calculated Eg values ofglass the GZO thinasfilms the substrate, the Eg values increased from 3.612 eV to 3.677 eV and the wavelength of the absorption as a function of the deposition temperature and on different substrates are also shown in Figure 5. edge decreased from about 360 nm (transmittance ratio over 60%) to about 352 nm, where the The linear dependence of (αhν) on (hν) indicates that the GZO thin films are direct transition-type wavelengths of the absorption edge were blue-shifted as compared with those investigated by Roth semiconductors. As the deposition temperature increased from RT to 473 K and glass was used as the et al. [20]. As the deposition temperature increased from RT to 373 K and PI was used as the substrate, substrate, thevalues Eg values increased eV3.622 to 3.677 eV and wavelength of the absorption the Eg increased fromfrom 3.5513.612 eV to eV and the the wavelength of the absorption edgeedge decreased from about 360 nm (transmittance ratio over 60%) to about 352 nm, where the wavelengths decreased from 356 nm to 350 nm. As compared with the results deposited at room temperature by of the Gong absorption edge were blue-shifted as compared with those investigated by3.50 Roth al. and [20].had As the et al., the deposited GZO thin films had Eg values changed from 3.42 eV to eVet[21] an absorption edge increased at the wavelength nm. the GZO the thinEg films deposition temperature from RToftoabout 373 K400 and PI Apparently, was used asfor theallsubstrate, values investigated in this study the values were larger than those investigated Gong et al. The356 Eg nm increased from 3.551 eV to 3.622 eVEgand the wavelength of the absorption edgebydecreased from to 350 nm. As compared with the results deposited at room temperature by Gong et al., the deposited GZO thin films had Eg values changed from 3.42 eV to 3.50 eV [21] and had an absorption edge at

Nanomaterials 2016, 6, 88

7 of 13

the wavelength of about 400 nm. Apparently, for all the GZO thin films investigated in this study the Nanomaterials 6, 88than those investigated by Gong et al. The Eg values are 3.320 eV (glass) 7 of 13 and Eg values were2016, larger 3.317 eV (PI) in this study (Eg is about 3.24 eV in [22]). As compared with those of the RT-deposited values are 3.320 eV (glass) and 3.317 eV (PI) in this study (Eg is about 3.24 eV in [22]). As compared ZnO thin films, the reason for the blue-shift in the absorption edge of the GZO thin films (Figure 3) is with those of the RT-deposited ZnO thin films, the reason for the blue-shift in the absorption edge of caused by the increase of the Eg values. The detailed Eg values of the GZO thin films deposited on the GZO thin films (Figure 3) is caused by the increase of the Eg values. The detailed Eg values of the GZO different andon at different differentsubstrates temperatures ZnOtemperatures thin films deposited on different substrates thin substrates films deposited and atand different and ZnO thin films deposited are also compared in Table 2. on different substrates are also compared in Table 2. 11

3.0x10

11

2.0x10

(a)

(b)

2

(h) (cm x eV)

11

2.0x10

11

1.5x10

-1

-1

(h) (cm x eV)

2

11

2.5x10

11

11

1.0x10

2

2

1.5x10

11

1.0x10

RT 373K 473K

10

5.0x10

0.0 3.2

3.4

3.6

3.8

4.0

10

5.0x10

0.0 3.2

RT 373K 473K

3.4

3.6

3.8

Photonenergy h (eV)

Photonenergy h (eV)

Figure 5. The (αhν)2 vs. hν-Eg plots of the GZO thin films deposited as a function of deposition

Figure 5. The (αhν)2 vs. hν-Eg plots of the GZO thin films deposited as a function of deposition temperature and on different substrates: (a) glass and (b) PI. temperature and on different substrates: (a) glass and (b) PI. Table 2. Optical properties and the optical band gap (Eg) values of the GZO thin films deposited on

properties the optical band gap (Eg) of the GZO thin films Tabledifferent 2. Optical substrates andand at different temperatures and values ZnO thin films deposited on deposited different on different substrates and at different temperatures and ZnO thin films deposited on different substrates. substrates. Transmittance Substrates Transmittance Ratio-Visible Ratio-Visible RT glass 90.9% 373 K glass 91.6% RT glass 90.9% K glass 90.5% 373 K473 glass 91.6% RT PI 92.8% 473 K glass 90.5% 373 K PI 92.4% RT PI 92.8% 92.4% 373 K473 PI K PI 92.4% 87.1% 473RT K ZnO PI glass 92.4% RTglass ZnO PI 86.1% RT ZnO 87.1%

Substrates

Transmittance

Maximum

Absorption

Eg Value

Transmittance Maximum Ratio-Infrared Transmittance Ratio EdgeAbsorption (nm) (eV)Eg Value Ratio-Infrared Transmittance Ratio Edge (nm) (eV) 85.0% 95.9% 360 3.612 84.8% 85.0% 87.2% 84.8% 86.8% 87.2% 87.0% 86.8% 88.9% 87.0% 82.6% 88.9% 82.1% 82.6%

95.8% 95.9% 95.6% 95.8% 95.8% 95.6% 96.1% 95.8% 95.5% 96.1% 90.2% 95.5% 89.9% 90.2%

354 352 356 354 350 370 368

3.674 3.612 360 3.677 3.674 354 3.551 3.677 352 3.580 3.551 356 3.622 3.580 354 3.320 3.622 350 3.317 3.320 370 RT ZnO PI 86.1% 82.1% 89.9% 368 3.317 Two competing phenomena are generally agreed to be dominant in affecting the absorption edge in heavily doped semiconductors. The first is the well-known Burstein-Moss band-filling effect Two competing are generally agreed towith be dominant in of affecting theconcentration; absorption edge which negatively phenomena shifts the measured band-edge energy the decrease the carrier in heavily doped semiconductors. The first is the well-known Burstein-Moss band-filling effect which the second phenomenon that affects the optical absorption edge is the increase of donor density, negatively shifts the measured band-edge energy with the decrease of the carrier concentration; which will cause a change in the nature and strength of the interaction potentials between donors the second phenomenon thatThe affects theconcentration, optical absorption the increase of donor density, which and the host crystal. carrier which edge will beisshown in Figure 5a, is higher than that will Roth et al. [20], the wavelength of the absorption edge can be shifted a lower causeinvestigated a change inbythe nature and and strength of the interaction potentials between donorstoand the host value about 350 nm. The Eg value Ga2be O3 shown is aboutin 4.9Figure eV [23]5a, and the Eg value of ZnO is about by crystal. Theofcarrier concentration, whichofwill is higher than that investigated Weand believe the Eg values of absorption all the deposited films higher than those ofof theabout Roth 3.24 et al.eV. [20], the that wavelength of the edgeGZO can thin be shifted to a lower value ZnO thin films are caused by the addition of Ga2O3 into the ZnO. 350 nm. The Eg value of Ga2 O3 is about 4.9 eV [23] and the Eg value of ZnO is about 3.24 eV. We believe The variations of the carrier concentration, carrier mobility, and resistivity of the GZO thin films that the Eg values of all the deposited GZO thin films higher than those of the ZnO thin films are at different deposition temperatures on glass and PI substrates are compared Figure 6. Figure 6a caused by the of Ga 2 O3 into the shows thataddition the carrier concentration ofZnO. the GZO thin films increased with the raising deposition The variations of the carrier concentration, carrier mobility, GZOthe thin temperature. When the GZO thin films are deposited on glass and or PIresistivity substratesofbythe using RFfilms at different deposition on glass substrates are the compared Figure 6. Figure 6a shows sputtering process,temperatures many defects and poresand willPI exist and inhibit electrons’ movement. Using a higher deposition temperature during deposition process can lead an enhancement of temperature. the thin that the carrier concentration of the GZOthe thin films increased with thetoraising deposition densification andare crystallization; theglass increase in substrates the diffraction intensity thesputtering (002) peak,process, as Whenfilms’ the GZO thin films deposited on or PI by using theofRF shown in Figure 3, has proven those results. That can decrease the numbers of defects and pores in many defects and pores will exist and inhibit the electrons’ movement. Using a higher deposition

temperature during the deposition process can lead to an enhancement of the thin films’ densification and crystallization; the increase in the diffraction intensity of the (002) peak, as shown in Figure 3, has proven those results. That can decrease the numbers of defects and pores in the GZO thin films

Nanomaterials 2016, 6, 88

8 of 13

and can decrease the inhibition of the barrier electron transportation [24]. Also, as the deposition temperature was raised from RT to 473 K, the carrier (electron) concentration increased from 8.92 ˆ 1020 to 11.1 ˆ 1020 cm´3 for glass substrates and from 8.74 ˆ 1020 to 11.6 ˆ 1020 cm´3 for PI substrates. Nanomaterials 2016, 6, 88 8 of 13 Figure 6b shows that the carrier mobility of the GZO thin films increased with the raising deposition temperature of the used substrates. Figure 6b also shows as the the GZO and thin was filmsalmost and canindependent decrease the inhibition of the barrier electron transportation [24].that Also, sameas deposition temperature was was used,raised the carrier mobility thin films on glass substrates the deposition temperature from RT to 473 of K, the theGZO carrier (electron) concentration 20 cmon −3 for 20 to 11.6 × 1020 was higher than that8.92 of ×the thin films PIglass substrates. Asand the from deposition temperature was raised increased from 1020GZO to 11.1 × 10 substrates 8.74 × 10 cm−3 2 for PI substrates. Figure 6b shows that the carrier mobility of the GZO thin films increased with the from RT to 473 K, as Figure 6b shows, the carrier mobility increased from 3.58 to 8.25 cm /V-s for glass raisingand deposition temperature and independent of the usedThe substrates. Figure 6b substrates increased from 2.92 to was 8.28almost cm2 /V-s for PI substrates. improvement in also electron shows that as the same deposition temperature was used, the carrier mobility of the GZO thin films the mobility could be attributed to the re-crystallization of the GZO thin films that takes place during on glass substrates was higher than that of the GZO thin films on PI substrates. As the deposition deposition process with the increase of the deposition temperature. temperature was raised from RT to 473 K, as Figure 6b shows, the carrier mobility increased from Figure 6c shows the dependence of the resistivity of the GZO thin films on deposition temperature 3.58 to 8.25 cm2/V-s for glass substrates and increased from 2.92 to 8.28 cm2/V-s for PI substrates. The and the used substrate. Resistivity ofcould the GZO thin films is proportional to the reciprocal value improvement in electron mobility be attributed to the re-crystallization of the GZO thin filmsof the product the place carrier concentration N and the mobility µ: thatof takes during the deposition process with the increase of the deposition temperature. Figure 6c shows the dependence of the resistivity of the GZO thin films on deposition ρ “ 1{pN ˆ GZO e ˆ µqthin films is proportional to the reciprocal temperature and the used substrate. Resistivity of the value of the product of the carrier concentration N and the mobility μ:

(4)

As Equation (4) shows, both the carrier concentration and the carrier mobility contribute to the ρ = 1/(N × e × μ) (4) resistivity. The results in Figure 6c indicate that the resistivity decreased with the increase of the Astemperature. Equation (4) shows, the carrierinconcentration and the carrier the be deposition Thus, both the increases carrier concentration andmobility electroncontribute mobility to should Thewhen resultswe in analyze Figure 6cthe indicate thatinthe resistivityofdecreased increase the in takenresistivity. into account decrease resistivity the GZOwith thin the films. Thoseofresults deposition temperature. Thus, the increases in carrier concentration and electron mobility should be Figure 6a,b suggest that reduction in resistivity also comes from an increase in electron mobility in taken into account when we analyze the decrease in resistivity of the GZO thin films. Those results addition to the increase of the electron concentration. In this study, the minimum resistivity of the GZO in Figure 6a,b suggest that reduction in resistivity also comes from an increase in electron mobility in thin films at ato deposition temperature of 473 K on different substrates is mainly influenced addition the increase of the electron concentration. In this study, the minimum resistivityby of both the the carrier concentration and the carrier mobility being at their maximum. As the deposition temperature GZO thin films at a deposition temperature of 473 K on different substrates is mainly influenced by ´3 Ω¨ cm for glass was raised from RT to 473 K, the resistivity decreased from 1.95atˆtheir 10´3maximum. to 0.680 ˆAs 10the both the carrier concentration and the carrier mobility being deposition 3 Ω¨ cm for PI substrates, respectively. substrates and from ˆfrom 10´3RT toto0.651 ˆthe 10´resistivity temperature was 2.44 raised 473 K, decreased from 1.95 × 10−3 to 0.680 × 10−3 Ω·cm

12

2

Carrier Concentration (x10

10

(a)

Mobility (cm /V-s)

14

22

-3

cm )

for glass substrates and from 2.44 × 10−3 to 0.651 × 10−3 Ω·cm for PI substrates, respectively.

10

8 glass PI

6

300

350

400

450

500

(b)

8

6

4

2 Glass PI

0

50

Resistivity (*cm)

150

200

Substrate temperature ( C)

-2

10

100

o

Deposition temperature (K) (c)

Glass PI

-3

10

-4

10

50

100

150

o

200

Substrate temperature ( C)

Figure 6. Variations of the (a) Hall mobility; (b) carrier concentration; and (c) resistivity of the GZO

Figure 6. Variations of the (a) Hall mobility; (b) carrier concentration; and (c) resistivity of the GZO thin films as a function of used substrates and deposition temperature. thin films as a function of used substrates and deposition temperature.

Nanomaterials 2016, 6, 88

9 of 13

Nanomaterials 2016, 6, 88

9 of 13

0

0

200

400

600

Zn2p3

800

1000

800

1000

Binding Energy (eV)

1200

Ga2s

Zn LMM1

Ga LMM1 Ga LMM

Zn3s Ga3s

Zn3d Ga3d

40000

Ga3p Zn3p

80000

Ga2p3 Ga2p1 Zn2s

c/s

120000

Zn2p1

160000

O KLL

(a)

O1s

200000

Zn LMM-Ga LMM2-Ga LMM3 Ga LMM4 Zn LMM2 Zn LMM3

Figure 7 shows a typical widescan spectrum of the GZO thin films as a function of the used Figure shows a typicalpeaks widescan spectrum of the GZOZn, thinO,films a function of theZn used substrates. The7photoelectron of the main elements, andasGa, and Auger and Ga substrates. The photoelectron peaks of the main elements, Zn, O, and Ga, and Auger Zn and Ga LMM LMM (electron excited from L layer to M layer) and O KLL (electron excited from K layer to L layer) (electron excited from L layer to M layer) and O KLL (electron excited from K layer to L layer) peaks peaks were obtained. As Figure 7 shows, some small peaks with intensities lower than each main were obtained. As Figure 7 shows, some small peaks with intensities lower than each main photoelectron peak were also observed. photoelectron peak were also observed.

1400

(a) 200000

0

0

200

400

600

Zn2p3

Binding Energy (eV)

1200

Ga2s

Ga2p3 Ga2p1 Zn2s

40000

Zn3d Ga3d Zn3p Ga3p Zn3s Ga3s

80000

Zn2p1

c/s

120000

O KLL

160000

Ga LMM1 Ga LMM Zn LMM1 Zn LMM-Ga LMM2-Ga LMM3 C1s Ga LMM4 O1s Zn LMM2 Zn LMM3

PI

1400

(b) Figure 7. Typical widescan spectrum of the GZO thin films as a function of used substrates. (a): glass;

Figure 7. Typical widescan spectrum of the GZO thin films as a function of used substrates. (a): glass; (b): PI. (b): PI.

To clarify the mechanism of improvement in electrical properties, the chemical structures of the GZO thin films were investigated by XPS. The XPS spectra for properties, the Ga2p1/2, Ga2p 3/2, Zn2p1/2,structures Zn2p3/2, To clarify the mechanism of improvement in electrical the chemical of and Ga3d peaks of the GZO thin films are shown in Figure 8 as a function of the used substrates, and the GZO thin films were investigated by XPS. The XPS spectra for the Ga2p1/2 , Ga2p3/2 , Zn2p 1/2 , the values are compared in Table 3; those thin films were deposited at RT. The XPS spectra for the Zn2p3/2 , and Ga3d peaks of the GZO thin films are shown in Figure 8 as a function of the used Ga2p1/2, Ga2p3/2, Zn2p1/2, Zn2p3/2, and Ga3d peaks of the ZnO thin films are also shown in Figure 8 for substrates, and the values are compared in Table 3; those thin films were deposited at RT. The XPS comparison. When glass was used as the substrate, the binding energy of each constituent element spectra the Ga2p , Ga2p , Zn2p , and Ga3d peaks of the ZnO thin films are also 1/2 , Zn2p wasfor positioned at 1/2 1145.22 eV3/2 (Ga2p 1/2), 1118.36 eV3/2 (Ga2p 3/2), 1045.23 eV (Zn2p1/2), 1022.10 eV (Zn2p3/2), shown Figure 8 for comparison. glass was used as the substrate, theasbinding energy of andin20.72 eV (Ga3d), respectively, asWhen calibrated to 285.43 eV (C1s); When PI was used the substrate, each the constituent element was positioned at 1145.22 (Ga2p1/2 1118.36 (Ga2p ), 1045.23 binding energy of each constituent element waseV positioned at),1143.35 eVeV (Ga2p 1/2),3/2 1116.52 eV eV (Ga2p 3/2 ), 1043.64 eV (Zn2p 1/2 ), 1020.47 eV (Zn2p 3/2 ), and 19.06 eV (Ga3d), respectively. We know the (Zn2p ), 1022.10 eV (Zn2p ), and 20.72 eV (Ga3d), respectively, as calibrated to 285.43 eV (C1s); 1/2 3/2 binding energies of Ga 2 O 3 and Ga metal are 1117.4 ± 0.5 eV and 1116.5 ± 0.2 eV. As Figure 8a shows, When PI was used as the substrate, the binding energy of each constituent element was positioned at no eV apparent be eV observed XPS spectra of the Ga3d peak ofeV the(Zn2p ZnO thin),films. For eV 1143.35 (Ga2ppeak), could 1116.52 (Ga2pin the ), 1043.64 eV (Zn2p ), 1020.47 and 19.06 1/2

3/2

1/2

3/2

(Ga3d), respectively. We know the binding energies of Ga2 O3 and Ga metal are 1117.4 ˘ 0.5 eV and 1116.5 ˘ 0.2 eV. As Figure 8a shows, no apparent peak could be observed in the XPS spectra of the

Nanomaterials 2016, 6, 88

10 of 13

Ga3d peak of the ZnO thin films. For that, the results in Figure 8a suggest that the binding state of Ga2 O3 will dominate the GZO thin films on glass and the binding state of Ga will dominate the GZO thin films on PI.2016, 6, 88 Nanomaterials 10 of 13 Table 3. Comparison of binding energies of Ga2p , Zn2p Zn2p1/2 , and Ga3dthin peaks for 3/2 , Ga2p 1/2Ga 3/2 , dominate that, the results in Figure 8a suggest that the binding state of 2O3 will the GZO films GZO thin films on different substrates. on glass and the binding state of Ga will dominate the GZO thin films on PI. Title energies of Ga2p3/2, Ga2p1/2 Glass Table 3. Comparison of binding , Zn2p3/2, Zn2p1/2, and PI Ga3d peaks for GZO thin films on Ga2P different substrates. Binding energy (eV) 1145.22 1143.35 1/2

Ga2P3/2 Title Binding energy (eV) 1118.36 1116.52 PI Glass Zn2p1/2 Binding energy (eV) 1045.23 1043.64 Ga2P1/2 Binding energy (eV) 1145.22 1143.35 Zn2p3/2 Binding energy (eV) 1022.10 1020.47 Ga2P3/2 Binding energy (eV) 1118.36 1116.52 Ga3d Binding energy (eV) 20.72 19.06 Zn2p1/2 Binding energy (eV) 1045.23 1043.64 Zn2p3/2 Binding energy (eV) 1022.10 1020.47 Ga3d Binding energypeaks (eV) were observed in 20.72 19.06 As Figure 8b shows, two similar the XPS spectra of the ZnO and GZO thin

films. The binding energy peak at 1022.40 ˘ 0.10 eV could be due to Zn2p3/2 in the ZnO1-x structure As Figure 8b shows, similar˘peaks were observed the XPS spectra theresults ZnO and and the binding energy peak two at 1021.1 0.40 eV was due to in metallic zinc [25].ofThe in GZO Figure 8b thin films. The binding energy peak at 1022.40 ± 0.10 eV could be due to Zn2p3/2 in the ZnO1-x structure suggest that the binding state of ZnO1-x will dominate the GZO thin films on glass and the binding and the binding energy peak at 1021.1 ± 0.40 eV was due to metallic zinc [25]. The results in Figure state of Zn will dominate the GZO thin films on PI. The binding energy of the Ga3d peak for Ga2 O3 8b suggest that the binding state of ZnO1-x will dominate the GZO thin films on glass and the2+binding (Ga3+state ) is 20.7 eV; the binding energies of the Ga3d peak for Ga2 O (Gaof1+the ) orGa3d GaOpeak (Gafor)Ga are about of Zn will dominate the GZO thin films on PI. The binding energy 2O3 20.7 eV; energy energies of the Ga3d Gafor metal is (Ga 18.61+)˘or0.3 eV,(Ga respectively. 3+) is the 2+) are aboutFigure (Gaand 20.7binding eV; the binding of thepeak Ga3dfor peak Ga2O GaO 20.7 8c shows the binding energy of the Ga3d peak of the GZO thin films on different substrates. When eV; and the binding energy of the Ga3d peak for Ga metal is 18.6 ± 0.3 eV, respectively. Figure 8cglass was used the substrate, the of the peakonwas aboutsubstrates. 20.7 eV, which suggests showsasthe binding energy of binding the Ga3d energy peak of the GZOGa3d thin films different When glass was used as theof substrate, thepeak binding energy of 3+ the); Ga3d wasused aboutas20.7 which suggests the binding energy the Ga3d is Ga whenpeak PI was theeV, substrate, the binding 2 O3 (Ga 3+); when PI was used as the substrate, the binding theofbinding energy ofwas the Ga3d 2OeV, 3 (Ga energy the Ga3d peak aboutpeak 18.6is˘Ga 0.3 which suggests the binding energy of the Ga3d peak energy of the Ga3d peak was about 18.6 ± 0.3 eV, which suggests the bindingofenergy of the Ga3d is the Ga metal. No apparent peaks could be observed in the XPS spectra the Ga3d peak ofpeak the ZnO is the Ga metal. No apparent peaks could be observed in the XPS spectra of the Ga3d peak of the ZnO thin films. For that, the binding energies in Figure 8c reveal the same results as those of Figure 8a. thin films. For that, the binding energies in Figure 8c reveal the same results as those of Figure 8a. The results in Figure 8 showed significant variations in the core binding energies (~1.55 eV) as we The results in Figure 8 showed significant variations in the core binding energies (~1.55 eV) as we varied the substrates. The change can be related to the difference between the intrinsic properties of varied the substrates. The change can be related to the difference between the intrinsic properties of the deposition materials as well asas totothe the interaction interactionwith with substrates the deposition materials as well thedifferences differences in in the thethe substrates [26].[26].

1110

1120

1140

1150

1015

1045

GZO-glass GZO-PI ZnO-glass ZnO-PI

Intensity (a.u.)

20

1035

Binding energy (eV)

(c) Ga3d

18

(Zn2p1/2)

1025

Binding energy (eV)

16

GZO-glass GZO-PI ZnO-glass ZnO-PI

(Zn2p3/2)

(Ga2p1/2) 1130

(b) Zn2P

Intensity (a.u.)

GZO-glass GZO-PI ZnO-glass ZnO-PI

(Ga2p3/2)

Intensity (a.u.)

(a) Ga2P

22

24

26

28

30

32

Binding energy (eV)

Figure 8. Binding energy spectraofof(a) (a)Ga2p Ga2p3/2 and Ga2p1/2 peaks; (b) Zn2p3/2 and Zn2p1/2 peaks; and Figure 8. Binding energy spectra 3/2 and Ga2p1/2 peaks; (b) Zn2p3/2 and Zn2p1/2 peaks; (c) Ga3d peak for the undoped ZnO thin films and GZO thin films on different substrates. and (c) Ga3d peak for the undoped ZnO thin films and GZO thin films on different substrates.

Nanomaterials 2016, 6, 88

11 of 13

Nanomaterials 2016, 6, 88the XPS spectra for the Gaussian-resolved components of O of the11 of 13 thin Figure 9a shows GZO 1s films on glass substrates; those thin films were deposited at RT. The bonding energy components Figure 9a shows the XPS spectra for the Gaussian-resolved components of O1s of the GZO thin were films centered at 531.17 eV and eV were for glass and PI substrates, respectively. As shown in on glass substrates; those529.35 thin films deposited at RT. The bonding energy components Figure 9a, two shoulders were visible on the low-energy side and high-energy side of the GZO were centered at 531.17 eV and 529.35 eV for glass and PI substrates, respectively. As shown in thin films,Figure indicating that at least three different bonds existed. For GZO thin films deposited on 9a, two shoulders were visible on the low-energy side and high-energy side of the GZO thinglass substrates, the bonding ofthree the O resolved intothin three components centered at films, indicating that state at least different bondswas existed. For GZO films deposited on glass 1s spectrum stateeV of the O1s spectrum wasOresolved into three components centered at 530.21substrates, eV, 531.18the eV,bonding and 532.39 for the OI , OII , and peaks, respectively. As shown in Figure 9b, III eV, 531.18 eV,visible and 532.39 eVhigh-energy for the OI, OIIside , andof OIII peaks, As shown in Figure only 530.21 one shoulder was on the the GZOrespectively. thin films, and we also found that 9b, only one shoulder was visible on the high-energy side of the GZO thin films, and we also foundof the three different bonds existed. For GZO thin films deposited on PI substrates, the bonding state that three different bonds existed. For GZO thin films deposited on PI substrates, the bonding state O1s spectrum was resolved into three components centered at 529.03 eV, 529.83 eV, and 531.02 eV for of the O1s spectrum was resolved into three components centered at 529.03 eV, 529.83 eV, and 531.02 the OI , OII , and OIII peaks, respectively. Table 4 shows the area ratios of the OI , OII , and OIII peaks; the eV for the OI, OII, and OIII peaks, respectively. Table 4 shows the area ratios of the OI, OII, and OIII area beneath peaks is clearly different fordifferent these two peaks; thethe area beneath the peaks is clearly for substrates. these two substrates.

(b)

Intensity (a.u.)

Intensity (a.u.)

(a) O

O

528

530

O O O

O

532

534

Binding energy (eV)

536

526

528

530

532

534

Binding energy (eV)

Figure 9. Binding energy spectra of O1s peaks for GZO thin films on different substrates: (a) glass

Figure 9. Binding energy spectra of O1s peaks for GZO thin films on different substrates: (a) glass and and (b) PI. (b) PI. Table 4. Comparison of area ratios of OI, OII, and OIII peaks for GZO thin films on different substrates.

Table 4. Comparison of area ratios of OI , OII , and OIII peaks for GZO thin films on different substrates. Title Glass-RT PI-RT O I peak 14.1% 20.3% Title Glass-RT PI-RT OII peak 81.1% 52.6% OOI IIIpeak 14.1% 20.3% peak 4.8% 27.0% OII peak 81.1% 52.6% OIII peak 4.8% 27.0% For GZO thin films deposited on glass (PI) substrates, the low bonding energy component of the OI peak located at 530.21 ± 0.15 eV (529.03 ± 0.15 eV) is attributed to O2− ions on the wurtzite structure of the hexagonal Zn2+ deposited ion array, surrounded bysubstrates, Zn atoms with full complement nearest- of For GZO thin films on glass (PI) the their low bonding energyof component 2− 2 ´located O ions [25,27]. The second binding energy component of the O II peak at 531.18 ± the Oneighbor peak located at 530.21 ˘ 0.15 eV (529.03 ˘ 0.15 eV) is attributed to O ions on the wurtzite I 2− ions in the oxygen vacancies within the matrix of 2+ 0.15 eV (529.83 ± 0.15 eV) is associated with O structure of the hexagonal Zn ion array, surrounded by Zn atoms with their full complement of ZnO1−x [28], which also known as VO-like bonding. The highest bonding energy component of the nearest-neighbor O2 ´isions [25,27]. The second binding energy component of the OII peak located OIII peak located at 532.39 ± 0.15 eV (531.02 ± 0.15 eV) implies 2the presence of hydrated oxide species at 531.18 ˘ 0.15 eV (529.83 ˘ 0.15 eV) is associated with O ´ ions in the oxygen vacancies within on the film surface [29]. As Figure 9a shows, when glass was used as the substrate and the deposition the matrix of ZnO1 ´ratio which is also known as V -like bonding. The highest bonding energy x [28], was RT, the area of the OII peak was larger than Othose of the OI and OIII peaks. These results component of the O peak 532.39 ˘ 0.15 eV (531.02 0.15toeV) implies the presence suggest that thereIIIare morelocated oxygenat vacancies existing in GZO thin˘films cause the increase in the of hydrated oxide species on the film surface [29]. As Figure 9a shows, when glass was used carrier concentration. When PI was used as the substrate, because the PI substrates had less surfaceas the substrate and the deposition was RT, thefilms areatoratio theadhesive OII peak wasand larger thanthin those ofon the OI energy, that would cause the GZO thin have of poor force, the GZO films would have suggest poor crystallization, asmore Figureoxygen 3 shows. The areaexisting ratio ofinthe OII thin peakfilms and OPIIIIsubstrates peaks. These results that there are vacancies GZO apparently decreased andcarrier the area ratios of the OIWhen and OIII increased. These results are causedthe PI to cause the increase in the concentration. PIpeaks was used as the substrate, because by the decrease of oxygen vacancies and the increase of oxygen absorption, which will degenerate substrates had less surface energy, that would cause the GZO thin films to have poor adhesive force, the electrical properties of the GZO thin films, as Figure 6 shows. These results suggest that even with

and the GZO thin films on PI substrates would have poor crystallization, as Figure 3 shows. The area ratio of the OII peak apparently decreased and the area ratios of the OI and OIII peaks increased. These results are caused by the decrease of oxygen vacancies and the increase of oxygen absorption, which

Nanomaterials 2016, 6, 88

12 of 13

will degenerate the electrical properties of the GZO thin films, as Figure 6 shows. These results suggest that even with the same deposition parameters, the GZO thin films deposited on different substrates will have different results. 4. Conclusions In this study, the high-transmittance nano-scale GZO thin films were successfully deposited on glass and PI substrates by using the RF sputtering process. For all GZO thin films, the transmittance ratios of the GZO thin films at 400–700 nm were more than 94% regardless of deposition temperature and substrate. As the deposition temperature increased from RT to 473 K and glass (PI) was used as the substrate, the FWHM values of the (002) peak for GZO thin films deposited on glass (PI) substrates were 0.407 (0.412), 0.336 (0.360), and 0.265 (0.292) for substrate temperatures of RT, 373 K, and 473 K; the Eg values increased from 3.612 (3.551) eV to 3.677 (3.622) eV and the wavelength of the absorption edge decreased from about 360 (356) nm to about 352 (350) nm, respectively. From the binding energy spectra of Ga2p3/2 and Ga2p1/2 peaks, Zn2p3/2 and Zn2p1/2 peaks, the Ga3d peak, and O1s peaks, we had proved that GZO thin films on different substrates had different chemical and electronic states, and thus different electrical properties. Acknowledgments: The authors acknowledge financial supports of MOST 104-2221-E-390-013-MY2, MOST 104-2622-E-390-004-CC3, and MOST 103-2221-E-005-040-MY2. Author Contributions: Kun-Neng Chen, Chao-Ming Hsu, and Min-Chu Liu helped carrying out the deposition and measurement processes of GZO thin films and data analysis; Fang-Hsing Wang and Cheng-Fu Yang organized the paper, encouraged in paper writing, and helped proceeding the experimental processes and measurements. Conflicts of Interest: The authors declare no conflict of interest.

References 1. 2. 3. 4. 5.

6.

7. 8.

9. 10. 11. 12.

Tsang, M.K.; Baia, G.; Hao, J. Stimuli responsive upconversion luminescence nanomaterials and films for various applications. Chem. Soc. Rev. 2015, 44, 1585–1607. [CrossRef] [PubMed] Ariga, K.; Yamauchi, Y.; Rydzek, G.; Ji, Q.; Yonamine, Y.; Wu, K.C.-W.; Hill, J.P. Layer-by-layer Nanoarchitectonics: Invention, Innovation, and Evolution. Chem. Lett. 2014, 43, 36–68. [CrossRef] Zhuang, J.L.; Terfort, A.; Wöll, C. Formation of oriented and patterned films of metal–organic frameworks by liquid phase epitaxy: A review. Coord. Chem. Rev. 2016, 307, 391–424. [CrossRef] Stadler, A. Transparent conducting oxides—An up-to-date overview. Materials 2012, 5, 661–683. [CrossRef] Fay, S.; Kroll, U.; Bucher, C.; Vallat-Sauvain, E.; Shah, A. Low pressure chemical vapour deposition of ZnO layers for thin-film solar cells: Temperature induced morphological changes. Sol. Energy Mater. Sol. Cell 2005, 86, 385–397. [CrossRef] Huang, C.C.; Wang, F.H.; Yang, C.F. Effects of deposition temperature and hydrogen flow rate on the properties of the Al-doped ZnO thin films and amorphous silicon thin-film solar cells. Appl. Phys. A 2013, 112, 877–883. [CrossRef] Wang, F.H.; Yang, C.F.; Lee, Y.H. Deposition of F-doped ZnO transparent thin films using ZnF2 -doped ZnO target under different sputtering substrate temperatures. Nanoscale Res. Lett. 2014, 9. [CrossRef] [PubMed] Assuncao, V.; Fortunato, E.; Marques, A.; Aguas, H.; Ferreira, I.; Costa, M.E.V.; Martins, R. Influence of the deposition pressure on the properties of transparent and conductive ZnO:Ga thin-film produced by r.f. sputtering at room temperature. Thin Solid Films 2003, 427, 401–405. [CrossRef] Kim, B.G.; Kim, J.Y.; Lee, S.J.; Park, J.H.; Lima, D.G.; Park, M.G. Structural, electrical and optical properties of Ga-doped ZnO films on PET substrate. Appl. Surf. Sci. 2010, 257, 1063–1067. [CrossRef] Ohkubo, I.; Ohtomo, A.; Ohnishi, T.; Mastumoto, Y.; Koinuma, H. In-plane and polar orientations of ZnO thin films grown on atomically flat sapphire. Surf. Sci. 1999, 433, L1043–L1048. [CrossRef] Hartnagel, H.L.; Dawar, A.L.; Jain, A.K.; Jagadish, C. Semiconducting Transparent Thin Films; Institute of Physics Publishing: Philadelphia, PA, USA, 1995. Park, H.C.; Byun, D.; Angadi, B.; Park, D.H.; Choi, W.K.; Choi, J.W.; Jung, Y.S. Photoluminescence of Ga-doped ZnO film grown on c-Al2 O3 (0001) by plasma-assisted molecular beam epitaxy. J. Appl. Phys. 2007, 102. [CrossRef]

Nanomaterials 2016, 6, 88

13. 14.

15.

16.

17. 18. 19.

20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

13 of 13

Lu, S.K.; Huang, J.T.; Lee, T.H.; Wang, J.J.; Liu, D.S. Flexibility of the Indium Tin Oxide Transparent Conductive Film Deposited onto the Plastic Substrate. Smart Sci. 2014, 2, 7–12. Fortunato, E.; Goncalves, A.; Assuncao, V.; Marques, A.; Aguas, H.; Pereira, L.; Ferreira, I.; Martins, R. Growth of ZnO:Ga thin films at room temperature on polymeric substrates: Thickness dependence. Thin Solid Films 2003, 442, 121–126. [CrossRef] Yu, Z.N.; Zhao, J.J.; Xia, F.; Lin, Z.J.; Zhang, D.P.; Leng, J.; Xue, W. Enhanced electrical stability of flexible indium tin oxide films prepared on stripe SiO2 buffer layer-coated polymer substrates by magnetron sputtering. Appl. Surf. Sci. 2011, 257, 4807–4810. [CrossRef] Wang, F.H.; Kuo, H.H.; Yang, C.F.; Liu, M.C. Role of SiNx Barrier Layer on the Performances of Polyimide Ga2 O3 -doped ZnO p-i-n Hydrogenated Amorphous Silicon Thin Film Solar Cells. Materials 2014, 7, 948–962. [CrossRef] Azaroff, L.V. Element of X-ray Crystallography; McGraw-Hill: New York, NY, USA, 1968. Li, X.; Wang, Y.; Liu, W.; Jiang, G.; Zhu, C. Study of oxygen vacancies’ influence on the lattice parameter in ZnO thin film. Mater. Lett. 2012, 85, 25–28. [CrossRef] Wang, F.H.; Huang, C.C.; Yang, C.F.; Tzeng, H.Z. Optical and Electrical Properties of the Different Magnetron Sputter Power 300 ˝ C-Deposited Ga2 O3 -ZnO Thin Films and Applications in p-i-n α-Si:H Thin-Film Solar Cells. Int. J. Photoenergy 2013, 2013. [CrossRef] [PubMed] Roth, A.P.; Webb, J.B.; Williams, D.F. Band-gap narrowing in heavily defect-doped ZnO. Phys. Rev. B 1982, 25, 7836–7839. [CrossRef] Gong, L.; Lu, J.; Ye, Z. Transparent and conductive Ga-doped ZnO films grown by RF magnetron sputtering on polycarbonate substrates. Sol. Energy Mater. Sol. Cell 2010, 94, 937–941. [CrossRef] Khan, Z.R.; Khan, M.S.; Zulfequar, M.; Khan, M.S. Optical and Structural Properties of ZnO Thin Films Fabricated by Sol-Gel Method. Mater. Sci. Appl. 2011, 2, 340–345. [CrossRef] Orita, M.; Ohta, H.; Hirano, M.; Hosono, H. Deep-ultraviolet transparent conductive β-Ga2 O3 thin films. Appl. Phys. Lett. 2000, 77, 4166–4168. [CrossRef] Igasaki, Y.; Saito, H. Substrate temperature dependence of electrical properties of ZnO:Al epitaxial films on sapphire (1210). J. Appl. Phys. 1991, 69, 2190–2195. [CrossRef] Choi, B.H.; Im, H.B.; Song, J.S.; Yoon, K.H. Optical and electrical properties of Ga2 O3 -doped ZnO films prepared by r.f. sputtering. Thin Solid Films 1990, 193–194, 712–720. Burkstrand, J.M. Substrate effects on the electronic structure of metal overlayers—An XPS study of polymer-metal interfaces. Phys. Rev. 1979, 20, 2853–4858. [CrossRef] Yu, C.F.; Chen, S.H.; Sun, S.J.; Chou, H. Influence of the grain boundary barrier height on the electrical properties of Gallium doped ZnO thin films. Appl. Surf. Sci. 2011, 257, 6498–6502. [CrossRef] Rao, L.K.; Vinni, V. Novel mechanism for high speed growth of transparent and conducting tin oxide thin films by spray pyrolysis. Appl. Phys. Lett. 1993, 63, 608–610. Fan, J.C.C.; Goodenough, J.B. X-ray photoemission spectroscopy studies of Sn-doped indium-oxide films. J. Appl. Phys. 1997, 48, 3524–3531. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).