Evaluation of the optoelectronic properties and corrosion behavior of Al2O3-doped ZnO films prepared by dc pulsed magnetron sputtering

1 © 2014 IOP Publishing Ltd Printed in the UK 1. Introduction Transparent conductive oxides (TCOs) have attracted great attention in a wide variety of applications where high transmittance in the visible range and low resistivity are necessary. The actual and potential applications of TCO thin films include flat panel displays, photovoltaic cells, low emissivity windows, electrochromic windows, transparent thin film transistors, light-emitting diodes and semiconductor lasers. When the applications entail long-term environmental exposure, the performance of the devices deteriorates, diminishing their reliability. TCO acting as an electrode is one of Journal of Physics D: Applied Physics Evaluation of the optoelectronic properties and corrosion behavior of Al2O3-doped ZnO films prepared by dc pulsed magnetron sputtering CZubizarreta1, EGBerasategui1, RBayón1, REscobar Galindo2,4, RBarros3, DGaspar3, DNunes3, TCalmeiro3, RMartins3, EFortunato3 and JBarriga1 1 IK4-TEKNIKER, Polo Tecnológico de Eibar, c/ Iñaki Goenaga, 5, 20600 Eibar, Spain 2 Instituto de Ciencia de Materiales de Madrid (ICMM -CSIC), Cantoblanco, 28049 Madrid, Spain 3 CENIMAT/I3N, Departamento de Ciências dos Materiais, Faculdade de Ciências e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal E-mail: [email protected] Received 20 June 2014, revised 23 September 2014 Accepted for publication 7 October 2014 Published 13 November 2014 Abstract The main requirements for transparent conducting oxide (TCO) films acting as electrodes are a high transmission rate in the visible spectral region and low resistivity. However, in many cases, tolerance to temperature and humidity exposure is also an important requirement to be fulfilled by the TCOs to assure proper operation and durability. Besides improving current encapsulation methods, the corrosion resistance of the developed TCOs must also be enhanced to warrant the performance of optoelectronic devices. In this paper the performance of aluminum-doped zinc oxide (AZO) films deposited by pulsed dc magnetron sputtering has been studied. Structure, optical transmittance/reflectance, electrical properties (resistivity, carrier concentration and mobility) and corrosion resistance of the developed coatings have been analyzed as a function of the doping of the target and the coating thickness. Films grown from a 2.0 wt% Al2O3 target with a thickness of approximately 1 µm showed a very low resistivity of 6.54 × 10–4 Ωcm and a high optical transmittance in the visible range of 84%. Corrosion studies of the developed samples have shown very low corrosion currents (nanoamperes), very high corrosion resistances (in the order of 107 Ω) and very high electrochemical stability, indicating no tendency for electrochemical corrosion degradation. Keywords: AZO, PVD, electrochemical impedance, polarization, optoelectronics (Some figures may appear in colour only in the online journal) C Zubizarreta et al Printed in the UK 485501 JPD © 2014 IOP Publishing Ltd 2014 47 J. Phys. D: Appl. Phys. JPD 0022-3727 10.1088/0022-3727/47/48/485501 Papers 48 Journal of Physics D: Applied Physics IOP 4 Author to whom any correspondence should be addressed. Abengoa Research, Energia Solar 1, Palmas Altas, Seville 41014, Spain 0022-3727/14/485501+13$33.00 doi:10.1088/0022-3727/47/48/485501 J. Phys. D: Appl. Phys. 47 (2014) 485501 (13pp) C Zubizarreta et al 2 the most exposed materials in the different stacks. To warranty the performance of these layers, not only are the optoelectronic properties important; durability must also be ensured. The degradation rate of this layer is influenced by factors such as film thickness, deposition conditions and exposure history [1]. To our knowledge, there is little literature on this topic [2–4]. Thus, more research is needed in order to improve its corrosion resistance. Among the different TCOs, aluminum-doped zinc oxide (AZO) has been recognized in recent years as a potential replacement for commercial indium-tin-oxide and fluorinedoped tin-oxide. It has good optical and electrical properties [5, 6] and other advantages such as its non-toxicity and high thermal and chemical stability, associated with low cost and resource availability [7]. AZO coatings can be prepared by different deposition techniques, including dc magnetron sputtering (MS [8]), RF MS [9–11], pulsed dc MS [12, 13], pulsed laser deposition [14], reactive MS [15, 16], vacuum arc plasma evaporation [17] and sol-gel [18], among others. Pulsed dc MS presents high sputtering yields with high plasma density, long-term process stability, enhanced dynamic deposition rates [12, 19], and improved deposition speed and film quality [20]. These excellent properties make it the most attractive technology for commercial needs, being suitable for large-scale deposition of high-quality films at low temperature and being suitable for use on polymeric substrates [21, 22]. Nevertheless, in order to optimize the deposition process and the films obtained, indepth research is required so as to understand the influence of growth conditions on the structure and optoelectronic properties of the coatings. In the present paper, a thorough study of the influence of two different process parameters (Al2O3 doping and film thickness through deposition time) on the optoelectronic properties and the corrosion behavior of the developed layers has been carried out. 2. Experimental The AZO coatings were deposited using the pulsed dc MS method, a type of physical vapor deposition (PVD) technique, using the industrial MIDAS 450 equipment developed by IK4-TEKNIKER. This system has three rectangular unbalanced magnetron evaporators of 550  × 125 mm target size that are magnetically linked with the help of a dummy magnet to form a closed field configuration. The distance between the evaporator and substrate was 120 mm, the substrates were rotating planetarily in a double rotation in a 400 mm diameter substrate holder at 8 rpm. Two different AZO ceramic targets (ZnO:Al2O3, 99.5:0.5 wt% and ZnO:Al2O3, 98:2.0 wt%) were used for the sputtering depositions. An argon pressure of 3  × 10–3 mbar was applied for all the experiments. The power level used to investigate the mechanism of AZO thin film growth was 2.9 W cm−2 corresponding to 2,000 W. The substrates were heated by means of industrial heaters up to a temperature of 350°C. The power source was dual-output pulsed dc (5–350 kHz) from AEI Pinnacle Plus + 5/5 kW. The duty cycle (4%) and the frequency (100 kHz) were fixed in all the processes. Films with different thicknesses (150, 310, 710 and 1,025 nm) were deposited. The samples were deposited on microscope slides (Menzel-Gläser) and on silicon wafers. The structural analysis of the samples was performed by x-ray diffraction (XRD; PANalytical, X’Pert Pro model) in grazing incidence geometry with Cu Kα line radiation (λ = 1.5406 Ǻ) at room temperature. In order to analyze the chemical composition of the coatings, Rutherford backscattering spectroscopy (RBS) was carried out. RBS experiments were performed with a 5 MV HVEE Tandetron [23], and the spectra were collected using 3.035 MeV He+ ions at an ion dose of 10 μC. The data were acquired with a silicon surface barrier detector located at a scattering angle of 170°, and with an energy resolution of 20 keV. The experimental spectra were fitted with the software programs RBX [24] and SIMNRA [25]. Hall effect measurements at room temperature were used to determine the resistivity, the carrier concentration and the mobility of the majority carriers with a Bio-Rad HL5500 instrument. The total (direct and diffuse) optical transmittance and the specular reflectance were measured in the wavelength range from 200 to 2,500 nm on microscope slides (Menzel- Gläser) 76 × 26 mm2 in size using a Perkin-Elmer Lambda 950 spectrophotometer equipped with a 150 mm diameter integrating sphere. Atomic force microscopy (AFM; Asylum Research MFP-3D Stand Alone) together with scanning electron microscopy (SEM/FIB, Zeiss Auriga Crossbeam microscope) revealed the roughness (Sq, in nm) and microstructure of the films. Corrosion behavior of AZO-coated silicon wafers was studied by means of electrochemical impedance spectroscopy and polarization techniques. Coatings from targets with different Al2O3 dopant concentrations and different resultant thicknesses were evaluated. The measurements were performed in a three-electrode electrochemical cell using an Ag/AgCl (KCl 3M) reference electrode (SSC) and a platinum wire counter electrode. Tests were carried out at room temperature and under aerated conditions on an exposed area of 1 cm2. For simulating the corrosive media a solution of NaCl 0.06 M was used. Despite the fact that in many applications TCOs should work under acid or alkaline environments, a neutral and standard solution of NaCl was selected for this preliminary electrochemical characterization. Further studies in aqueous media under different pH values will be performed in the future when suitable deposition process parameters are selected according to preliminary characterization studies. Table 1. Chemical composition obtained by RBS measurements of the coatings grown from different dopant concentrations (0.5 wt% and 2.0 wt% Al2O3) in the target. Sample Zn (at%) O (at%) Al (at%) AZO_0.5% Al2O347.5 50.1 2.4 AZO_2.0% Al2O345.7 50.6 3.7 J. Phys. D: Appl. Phys. 47 (2014) 485501 C Zubizarreta et al 3 The procedure for evaluating the electrochemical response of the coatings doped with different aluminum content and grown with different thicknesses consisted of the measurement of the open circuit potential for 1,800 s of immersion in NaCl 0.06 M solution. After this time, electrochemical impedance spectroscopy was performed with a sinusoidal ac perturbation of±10 mV amplitude at a frequency range from 10 kHz to 10 mHz, 10 freq/decade. The impedance spectra were registered at open circuit potential. After the impedance measurements, a potentiodynamic polarization curve was registered on each surface from ‒0.4 to 1.2 V (versus open circuit potential) at a scan rate of 0.5 mV s−1. All potentials are referred to the Ag/AgCl electrode (0.207 V versus the standard hydrogen electrode). 3. Results and discussion 3.1. Influence of target doping The effect of target doping on the electrical and optical properties of the AZO layers has been investigated considering two distinct Al2O3 concentrations, where ZnO targets with 0.5 wt% Al2O3 and 2.0 wt% Al2O3 concentrations have been evaluated, leading to thickness samples of 1,025 nm. From the previous studies, the argon pressure of 3 × 10–3 mbar has been established as the optimum parameter and has been employed for all the experiments [26]. As the first step, the composition and structure of the deposited coatings were analyzed. Table1 shows the chemical composition obtained from RBS measurements of the coatings grown from targets with different dopant concentrations. Both samples are practically stoichiometric (oxygen content ~50 at%). Also, in both cases, the Al content in the AZO films is higher than the nominal one in the target, which was already reported in previous studies such as [27]. The sample prepared from a target with 0.5 wt% Al2O3 has 2.4 at% Al in the coating compared to 3.7 at% Al in the sample prepared from a target doped with 2.0 wt% Al2O3. Figure 1 shows the x-ray diffraction patterns for both studied conditions. When the ZnO films are doped with Al2O3, Al3+ ions become part of the ZnO lattice in the substitutional sites of Zn2+ ions, and there may also be Al interstitial atoms. Both analyzed samples show the strong ZnO (0 0 2) peak of hexagonal wurtzite-type structure and also the (0 0 4) peak with very weak intensity. This indicates that the films grew along the c-axis orientation, perpendicular to the substrates. Thus, Al3+ ions substituting Zn2+ did not change the ZnO hexagonal wurtzite structure. No other phases, such as Al or Al2O3, were detected in the films. For the 0.5 wt% Al2O3 concentration (1 0 1), (1 1 2) and (1 0 0) peaks are also shown, indicating that for lower dopant content there is a poorer c-axis orientation. In addition to the c-axis orientation, a small shift of the peak position of the (0 0 2) plane is observed to a higher position when increasing the Al content in the film (from 34.28° for 0.5 wt% Al2O3 to 34.36° for 2.0 wt% Al2O3), which indicates Figure 1. XRD diffractograms of the AZO thin films as a function of the Al2O3 concentration. Table 2. Full width at half maximum (FWHM) of the (0 0 2) peak and grain size for the two Al2O3 dopant concentrations in the target, 0.5 wt% and 2.0 wt%. [Al2O3] (wt%) FWHM (rad.) Grain size XRD (nm) 0.5 0.01074 14.1 2.0 0.00350 43.3 Table 3. Electrical resistivity (ρ), Hall mobility (µ) and carrier concentration (n) values for the ZnO targets with 2.0 wt% Al2O3 and0.5 wt% Al2O3 concentrations. Target doping Resistivity (Ωcm) Hall mobility (cm2/Vs) Carrier concentration (cm−3) 0.5 wt% Al2O32.07 × 10–3 13.8 2.18 × 1020 2.0 wt% Al2O36.54 × 10–4 20.7 4.61 × 1020 J. Phys. D: Appl. Phys. 47 (2014) 485501 C Zubizarreta et al 4 an increased influx of Al atoms at the Zn sites with increasing Al2O3 concentration [28]. For 0.5 wt% Al2O3 concentration, the (0 0 2) peak is slightly broader than for 2.0 wt%, indicating that the grain size of this film is smaller. The grain or crystallite size was obtained from the width of XRD peaks using Scherrer’s formula [29]. As seen in table2, the grain size of the films for 0.5 wt% Al2O3 concentration is 14.1 nm, and for the 2.0 wt% Al2O3 concentration it is 43.3 nm. Table 3 shows the results from Hall measurements for the AZO films grown from targets with different Al2O3 concentrations. As the Al2O3 concentration of the target increases from 0.5 to 2.0 wt%, the resistivity decreases, reaching values of 6.54  × 10–4 Ωcm. This decrease is due to an increase in carrier concentration and in Hall mobility, which can also be related to the higher Al content in the sample (table 1). The conduction characteristics of ZnO are basically dominated by electrons generated from oxygen vacancies and Zn interstitial atoms [30]. However, when the ZnO films are doped with Al2O3, the conductivity increases due to the contribution from Al3+ ions in the substitutional sites of Zn2+ ions and Al interstitial atoms as well as from oxygen vacancies and Zn interstitial atoms [27]. The bibliography indicates that there is a limit in doping concentration as the excess of Al atoms above a certain critical concentration in the ZnO films segregates into grain boundaries, decreasing the conductivity of the layer [27]. Correlating the electrical properties with the structural ones, the increase in resistivity for samples grown from the target with the lower dopant concentration could also be explained in terms of the segregation of Al content into the grain boundaries, since the grain boundaries increase with the decrease in grain size (table 2), which makes Al electrically inactive, causing this increase in resistivity [31]. The transmittance, reflectance and optical band gap for both samples have been measured and calculated (figure 2). Both films exhibit high optical transmittance in the visible wavelength range (84–86%). It is observed that the transmittance near the infrared region strongly decreases by increasing the amount of Al2O3 (2.0 wt%). The decrease in transmittance in the near-infrared region corresponds to the increasing carrier density. This behavior is caused by the increase in plasma oscillation frequency with the increasing carrier density, which can be explained by the Drude theory and is well known in degenerated TCO films [32, 33]. The optical band gap, Eg, of a direct gap semiconductor can be determined by extrapolating the linear portion of (αhν)2 against the photon energy (hν [34]), where α is the absorption coefficient. The band gap increases with the increase in the Al2O3 concentration, obtaining values of 3.41 eV for 0.5 wt% Al2O3 and values of 3.67 eV for 2.0 wt% Al2O3. This increase in the band gap is due to an increase in free electron concentration in the films as a result of Al doping. This shift of the band gap can be explained by the Burstein–Moss effect [35, 36], which suggests that a band gap can be widened with an increase in the carrier concentration. As the Fermi level moves up into the conduction band in an n-type semiconductor, the lower states in the conduction band get filled up, which means that more energy is required for the electrons to be excited from the valence band to the conduction band, leading to an extension of the optical band gap. Thus, the absorption edge shifts towards higher energy. SEM and AFM analyses have been performed in order to determine the morphology of the deposited films regarding the variation in Al2O3 concentration, and the results are presented in figure 3. A columnar/polygon grain structure is clearly observed for the 0.5 wt% Al condition. Moreover, with the increase in Al2O3 concentration, a crater-like structure is evident, confirming that the dopant has a vital effect on the final film morphology [37]. The AFM measurements revealed maximum roughness values of approximately 23 and 21 nm for the 0.5 and 2.0 wt% Al2O3 doping conditions, respectively. In summary, regarding the influence of the target dopant, it can be seen that although this parameter slightly affects the Figure 2. Optical properties of the AZO films obtained from targets with different dopant concentrations: (a) transmittance and reflectance spectra, (b) plot of (αhν)2 against photon energy (hν). J. Phys. D: Appl. Phys. 47 (2014) 485501 C Zubizarreta et al 5 optical properties in the visible region, the electrical properties are clearly improved using the target with a higher concentration of Al2O3. 3.2. Influence of coating thickness To investigate the influence of coating thickness on the electrical and optical properties of the films, and taking into account the results obtained regarding the target dopant, AZO films of 150, 310, 710 and 1,025 nm thickness were deposited from a target of ZnO doped with 2.0 wt% Al2O3 under a pressure process of 3 × 10–3 mbar, a power of 2 kW and a pulse frequency of 100 kHz. The XRD patterns of the AZO films with different thicknesses are shown in figure4. It shows that the films obtained were polycrystalline with a hexagonal structure and highly oriented with the c-axis perpendicular to the substrate, having a Bragg peak along the (0 0 2) plane. All the samples also show the (0 0 4) peak with much less intensity, imperceptible in the figurefor most of the samples. For all the films studied, apparent peak shifts towards the low-angle side were observed compared to that of bulk ZnO (2θ = 34.45°), which could indicate that stress exists in all the AZO films [38]. The surface and cross-section morphologies are presented in figure5 for the various AZO film thicknesses studied. As Figure 3. AFM and SEM images of the AZO films for the ZnO target with (a) 0.5 wt% Al2O3 and (b) 2.0 wt% Al2O3 concentrations. Figure 4. (a) XRD diffractograms and (b) grain size of the AZO thin films as a function of thickness. J. Phys. D: Appl. Phys. 47 (2014) 485501 C Zubizarreta et al 6 expected [39], when the film thickness increases, the crystallinity is also increased, with evident enlargement in the crystallite size (figure 5). This tendency was confirmed by SEM images, with a clear increase in grain size related to the evolution of the film thickness. Moreover, AFM measurements also showed this behavior, reaching a maximum roughness value of approximately 21 nm (figure 6). A nanocrystalline structure is clearly seen for the AZO films with lower thicknesses [40]. Moreover, the cross-sec- tion morphologies of these lower thickness conditions exhibited a columnar structure perpendicular to the substrates. The fine-scaled grains grow with deposition time, forming V-shaped structured grains. During the initial stages of deposition, random orientation nuclei were formed, and later the nanocrystalline structures engaged into a competitive growth stage, where crystals with higher vertical growth rate form large and well-oriented V-shaped grains [40]. This specific grain shape has been largely reported for ZnO:Al films subjected to etching processes [40–42], which may indicate that Figure 5. AFM and SEM images of the AZO films with (a) 150 nm, (b) 310 nm, (c) 710 nm and (d) 1,025 nm thickness. Figure 6. Roughness for AZO films as a function of thickness. (roughness mean square, RMS) J. Phys. D: Appl. Phys. 47 (2014) 485501 C Zubizarreta et al 7 a similar process may be occurring during the film deposition for longer deposition times. As previously mentioned, from figures5(a)–(d), it is clear that the increase in film thickness resulted in a coarsening and roughening of the films; however, the columnar bulk structure observed at the lower thickness conditions evolves into a dense and heterogeneous structure. Moreover, a uniform distribution of grains can be observed for all the conditions studied. Figure 7 shows the resistivity, Hall mobility and carrier concentration of the AZO as a function of the thickness. The lowest resistivity, obtained for the sample with the highest thickness (1,025 nm) is 6.54 × 10–4 Ωcm. For thinner samples (150 and 310 nm) the electrical properties are very similar (almost identical) but for higher thicknesses the conductivity increases significantly (58% when increasing the thickness to 710 nm and 42% for the thickest layer of 1,025 nm). This decrease in resistivity with the thickness is related to an increase in carrier concentration and in carrier mobility. Usually, as the thickness decreases the crystallinity of the layers worsens, having many defects due to large stress, and many free electrons can be caught by the large number of traps formed by these defects. Longer sputtering times (i.e. thicker coatings) entail higher temperatures on the substrate due to the ion bombardments. This effect increases the diffusion mobility of the sputtered particles, improving crystallinity and crystal size. Thus, the number of grain boundaries (scattering centers for free carriers) decrease, improving electrical properties [39]. Figure 8 shows the optical transmittance and the specular reflectance spectra of the AZO films deposited with different thicknesses. The average optical transmittance in the visible range (400–800 nm) was above 84% for all the films. When increasing film thickness the mean transmittance decreases slightly, which may be caused by optical scattering arising from longer optical paths and also by changes in the carrier concentration [43]. The conduction electrons are important because they dominate the optical properties of the materials in the near-infrared wavelengths. As the thickness of the films increases, the transmittance in the near-infrared wavelength region decreased. As mentioned in the previous section, this is a consequence of the higher carrier absorption caused by the increased carrier concentration in thicker films [32]. The reflectance results from the interference effect in the air/TCO. The visible range of wavelengths (400–800 nm) basically consists of interference fringes due to the thin film thicknesses, and a mean reflectance between 10 and 15%. Above 1,500 nm the films reveal high reflectance, overlapping a low transmittance region. This phenomenon, i.e. the transition from high transmittance/low reflectance to high reflectance/low transmittance, can be attributed to the increase in carrier concentration observed. The optical band gap of AZO films decreases gradually when increasing the thickness (from 3.76 eV for 150 nm to 3.67 eV for 1,025 nm). This behavior could indicate that apart from carrier concentration and mean crystallite size [44, 45], there are other parameters that affect the band gap energy shift, such as the residual stress in the films [46] and the merging of an impurity band into the conduction band which compensates the Burstein–Moss effect [44]. From figure 8(c), we can observe that the haze factor increases with the evolution in morphology and increase in thickness due to light deflection at the film/air interface as a consequence of surface roughness [47]. 3.3. Corrosion behavior 3.3.1. Effect of target doping. The evolution of the open circuit potential with immersion time in NaCl 0.06 M electrolyte was studied for two coatings with different aluminum content. For both coated surfaces, the potential increases slightly and reaches a stable value after few minutes of immersion. After 1,800 s of immersion, the AZO coating with 0.5% of Al2O3 doping shows an open circuit potential of ‒0.024 V (versus Ag/ AgCl); the AZO coating with higher aluminum content shows a nobler potential (0.135 V). The aluminum concentration in AZO structures influences the electrochemical potential of the samples in saline media. It is known that the structure and composition of PVD coatings strongly influence the electrochemical response of the total coating due to differences in the coating density, micro-defects, grain sizes, etc [48]. Figure 7. Variation of resistivity (◼), Hall mobility (● ) and carrier concentration (▲) as a function of thickness. J. Phys. D: Appl. Phys. 47 (2014) 485501 C Zubizarreta et al 8 After these 1,800 s of immersion in NaCl the open circuit potential was stable in both coatings. Then, electrochemical impedance spectroscopy was performed for each surface. It is well known that the impedance technique can be very useful for coating characterization because it provides information about the possible presence of pores and defects on their structure, and also about the system degradation with time. Figure9 shows the Nyquist diagrams registered. Both coatings show similar curves reflecting an incomplete semi-circle. From the Bode diagrams of figure9 it is easier to detect two time constants represented by two maximums in the -Phase-f curve and a shift in the slope of the |Z|-f curve in the medium frequency region (figure 10). Both AZO samples developed with different aluminum content presented very similar impedance diagrams and no significant differences in electrochemical response were found. Experimental data obtained by electrochemical impedance spectroscopy measurements were analyzed by means of an equivalent circuit composed of two time constants (figure11). This circuit, composed of two pairs of constant phase element/resistance (CPR/R) elements combined in parallel is frequently employed to simulate the behavior of PVD coatings in aqueous media. At high frequencies, R1 is related to the resistance across the coating defects or pores Figure 8. Optical behavior of AZO films of different thicknesses: (a) transmittance and reflectance spectra, (b) plot of (αhν)2 against photon energy (hν). The inset shows the optical band gap obtained by extrapolating the linear part, (c) haze factor. Figure 9. Nyquist diagram obtained from impedance measurements for AZO samples with different dopant concentrations. J. Phys. D: Appl. Phys. 47 (2014) 485501 C Zubizarreta et al 9 and CPE1 represents the capacitance of the coating (interface coating/electrolyte). In the second sub-circuit, R2 represents the corrosion resistance at the coating pores/defects bottom and CPE 2 the capacitance in the interface of the electrolyte/ silicon substrate. Rs represents the resistance of the electrolyte. The impedance of the constant phase elements (CPE) is represented by the equation ω=ZYi1/ () CPE n 0 where Y0 is the admittance of an ideal capacitance and n is an empirical constant, ranging from 0 to 1. When n = 1, the CPE behaves as a pure capacitor and when n = 0, the constant phase element behaves as a pure resistor. Due to the high corrosion resistance of substrate by itself, some authors proposed an equivalent circuit with only one time constant to simulate the system [53]. The passive state of the silicon substrate at the bottom of the coating pores sometimes makes their contribution from the TCO coatings barely distinguishable. At high frequencies (104 Hz), the angle phase registered in both coatings is close to 90°. Additionally, the low values of the Yo-CPE1 element (1.1  × 10‒6 F cm−2 for 0.5 wt% Al2O3 and 1.6  × 10‒6 F cm−2 for 2.0 wt% Al2O3) as well as the n-values close to 1 obtained in both coatings, suggest a capacitive behavior of the surface of both TCOs. Thecorrosion resistance R1 was relatively higher in the case of the coating doped with 2.0 wt% Al2O3. The R1 values determined after fitting the experimental data were 44 kΩ for the 0.5 wt% Al2O3-doped coating and 74 kΩ for the 2.0 wt% Al2O3-doped coating. The second time constant detected in Bode plots appears at medium-low frequencies and is related to the electrochemical process taking place at the substrate/ electrolyte interface across the coating defects or pores. In both cases the value of R2 is high enough, implying a low density of defects in both AZO coatings. Total corrosion resistance of coated systems taking into account R1 and R2 resistances is very high, in the order of 107 Ω. Comparing both coatings, the R2 value measured in the case of the 0.5 wt% Al2O3 dopant is approximately 8 × 107 Ω, whereas the R2 value in the case of the sample with 2.0 wt% Al2O3 dopant is around 2 × 107 Ω. After impedance measurements, a potentiodynamic polarization test was additionally performed on coatings to evaluate their response in the anodic region and their respective corrosion currents. Figure12 shows the curves obtained. Both Figure 10. Bode plots of both coatings with different aluminum content. Figure 11. Equivalent circuit model used in the fitting process of experimental data. Figure 12. Polarization curves of AZO coatings in NaCl solution for AZO samples with different dopant concentrations (Al2O3 concentration in the target). J. Phys. D: Appl. Phys. 47 (2014) 485501