1 Introduction

ZnO based thin films have gained great interest due to its outstanding properties such as non-toxicity, being abundant in nature, high thermal stability, high transparency and conductivity etc. ZnO has a direct band gap (3.30 eV) [1] with a high exciton binding energy (60 meV) [2] at room temperature. High exciton binding energy makes ZnO an important material for many applications such as UV-lasers [3], solar cells [4], photo detectors [5], LEDs [6] etc. Adjusting the some properties of a thin film is an important prerequisite for many applications. Thus, ZnO thin films should be doped. After the doped ZnO, some properties can change drastically. Al doping element changes to optical band gap of the thin film. Al element is a cheap material. It is non-toxic material. It is know Al doped ZnO is an alternative material for the transparent conductive oxide layer.

ZnO thin films have been fabricated by several techniques such as chemical vapor deposition [7], pulsed laser deposition [8], chemical bath deposition [9], sol–gel [10] and thermionic vacuum arc (TVA) [11] deposition technique.

In this study TVA deposition has been used for fabrication of the Al doped ZnO thin films. TVA is a unique deposition technique among the others. Thin films produced by TVA deposition technique are extremely flat, having low roughness, nano structured, transparent, as reported earlier [12,13,14]. In this study, Al doped ZnO thin films has been produced onto glass, polyethylene terephthalate (PET) and silicon (Si) substrate. Morphological, structural and optical properties and substrate effect of the Al doped ZnO thin films have been investigated. ZnO and ZnO semiconductors show magnetic and photocatalytic properties [15], optical properties [16] electronic properties [17], transparent conductive oxide [18], etc. have been widely studied around the world. However, it is a still an attractive material to be studied.

2 Experimental setup

TVA deposition technique is an anodic plasma generator and it works under the high vacuum condition. It is a good technique to produce high purity, nano structured, thin films with less roughness. In principle, TVA experimental system consists of anode, which represents tungsten crucible with spoon-like shape, and cathode, which represents electron gun (tungsten filament). The anode material is located inside the crucible and both electrodes are located inside the vacuum chamber. The cathode is a coil with four-turn, made of tungsten wire with a thickness of 0.4 mm and 1 mm inner radius. Experimental setup of the TVA system is shown in Fig. 1. As seen in figure AC voltage applied to the tungsten filament. As a result, cathode emits thermal electrons. Emitted electrons are collimated via Wehnelt cylinder and are accelerated towards the anode under the high DC voltage. When accelerated electrons reach the anode from the heated cathode, a continuous evaporation of the anode material is established. With a further increase of the applied voltage, a bright discharge is established between the electrodes. Finally, deposition will take place when the energetic ions in the plasma of the coating material reach the substrate.

TVA deposition is a very promising and relatively a new method for fabrication of the thin films. The advantages of this method, ion energies can be controlled even during the experiment. In addition to it does not require any buffer or mixture gas thus it provides to produce thin films by keeping impurity level at the minimum. Furthermore, it can provide fabrication of the multiple thin films even with the different coating materials simultaneously [18, 19].

Fig. 1
figure 1

Illustration of experimental setup of the thermionic vacuum arc system

Full size image

All doping Al and ZnO material were in powder form with purity of 99.99%. Materials prepared in disk shape by using high-pressure pellet machine before the deposition process. The material to be evaporated consists of ZnO:Al disk with 3 mm diameter and 1 mm thickness. The used substrates were prepared by using ultrasonic bath apparatus for 20 min for several times and then cleaned with ethyl alcohol before adjusting on the substrate holder. Deposition was performed under the high vacuum (from 7 × 10−5 to 2 × 10−4 torr) conditions. The applied voltage between the electrodes decreased from 1500 to 400 V and the deposition performed last 1 min. The ignition of the plasma causes a decrease of the chamber pressure and the applied voltage whereas the discharge current occurs (from 0 to 0.4 A). The experimental parameters also summarized in Table 1.

Table 1 Experimental parameters of the AZO thin films deposition
Full size table

3 Results and discussion

In this study Al doped ZnO thin films were fabricated make use of the TVA technique onto amorphous glass, polycrystalline PET and single crystal Si (111) substrate. The morphological, structural and optical properties of the samples have been investigated with various analysis devices.

Considering the XRD results, all fabricated films are in polycrystalline form and are show different crystalline orientation in terms of substrate effect. The measurement was taken place in the range of 20°–80° and results are shown in Fig. 2. ZnAl2O4 bi-metal oxide has been found for all substrate in cubic form and matched with crystallographic database number 96-900-7033 for Match software and literature [20]. ZnAl2O4 is a spinel type oxide and is known to have high chemical and thermal stability, high mechanical resistance and low surface acidity all of which make it a suitable material for a wide range of applications [21]. Aluminum oxide (Al2O3) peak that correspond to highest peak in XRD pattern was found in orthorhombic space group and obtained results show a good associated with Match software crystallographic database number 96-810-3789 and literature [22]. The other constructive interference peak of the Al, ZnO, AlO2 and Al2O3 was found in the XRD pattern at various degrees. The XRD peaks are also associated with the literature [23,24,25,26,27].

Fig. 2
figure 2

XRD pattern of the deposited AZO thin films onto glass, PET and Si

Full size image

The surface morphology, were determined by using an Ambios Q-Scope atomic force microscope (AFM) in non-contact mode. The measurement was taken place in the range of 4 µm × 4 µm as seen in Fig. 3. The AFM images of the samples showed granular film structure with spherical grains of varying sizes.

Fig. 3
figure 3

AFM images and height distribution of the AZO thin films deposited onto a glass, b PET, and c Si substrate

Full size image

The crystallite sizes were calculated by Scherrer equation [28, 29];

$${\text{D}}=\frac{{k\lambda }}{{\beta {\text{cos}}\theta }}$$
(1)

where D is the crystallite size, K is a shape factor which is selected a 0.94, λ is used wavelength in XRD measurement, β is a full width at half maximum (FWHM) value of the respective diffraction peak, θ represents the peak position. Crystallite size values were obtained between ~ 20 nm and ~ 100 nm for three different substrates. Dislocation density (δ) and internal strain (ε) were calculated by the following equations:

$$\delta =\frac{1}{{{D^2}}}$$
(2)
$$\varepsilon =\frac{{\beta {\text{cos}}\theta }}{4}$$
(3)

The obtained results are summarized in Table 2. Obtained results show that the crystallite size values relatively higher for Si substrate compared to PET and glass substrate. It can be concluded that nano-crystallites can be grown with lower size onto amorphous substrates rather than crystal substrates.

Table 2 XRD patterns, peak positions and calculated crystallite sizes, dislocation density, micro strain values
Full size table

The FESEM measurements were taken in order to investigate surface topography of the samples using Carl Zeiss Supra VP40 FESEM device. The measurement results are shown in Fig. 4 with 100k times magnification. As seen in figure, the deposited film onto the glass substrate has lower grain size compared to grain size of the PET and Si substrate. The FESEM results are in good agreement with the AFM measurements.

Fig. 4
figure 4

FESEM images of the AZO thin films deposited onto a glass, b PET, and c Si substrate

Full size image

The optical analysis of the Al doped ZnO thin films were conducted by using UV–Visible spectrophotometry and Filmetrics F20 thin film thickness measurement system. Transmittance values of the thin films were obtained as ~ 50% and ~ 30% in visible spectrum for glass and PET substrate respectively. Reflectance, refractive index and extinction coefficient data are also shown in Fig. 5. Cauchy dispersion model, was used in order to obtain the refractive index n(λ) as a function of the wavelength. Cauchy model is given by:

Fig. 5
figure 5

AFM images of the AZO thin films deposited onto (a) glass, (b) PET, and (c) Si substrate

Full size image
$$n(\lambda )={A_n}+\frac{{{B_n}}}{{{\lambda ^2}}}+\frac{{{C_n}}}{{{\lambda ^4}}}+ \cdots$$
(4)

where \({A_n}\), \({B_n}\), and \({C_n}\) are the Cauchy model parameters. Mean thickness values of the deposited films were found as, 50 nm for glass wafer, 20 nm for PET wafer and 30 nm for Si wafer, respectively.

The differences in the thickness have been formed due to different distances between samples and anode crucible. As can be seen in figure refractive indices of the samples slightly decrease with the increasing wavelength for PET and Si unlike the glass substrate. Tauc plot was used to obtain optical band gap of the AZO thin film for glass and PET substrate. Tauc plot is a simple technique to determine the band gap of the thin films [30]. The method is simply based on a linear extrapolation of the energy axis that gives band gap energy. The absorption coefficient is related to photon energy by the known equation as:

$$\alpha hv=\beta {(hv - {E_g})^m}$$
(5)

where β is constant, alpha represents the absorption coefficient, h is the Planck constant, ν represents the incident photon frequency, and \({E_g}\) represents the optical band gap of the semiconductor material. Here m, selected as 2, which corresponds to direct allowed transitions. In order to prove calculated optical band gap value photoluminescence measurement was taken using PerkinElmer LS 45 device. The obtained results were indicated in Fig. 5. The optical band gaps of the thin films were determined as 3.48 eV and 3.5 eV for thin films onto glass and PET substrates, respectively. The band gap of the un-doped ZnO thin film is 3.30 eV. When Al doped the ZnO material, the band gap value shift towards to the higher energy value. The obtained band gap values are good harmony with the value determined from the optical method and literature.

Raman spectroscopy measurement was taken place in order to support XRD results. The obtained results are shown in Fig. 6. ZnO, Al2O3 and substrate peaks, which are PET and Si, were found in Raman spectra [31,32,33,34,35]. The results are also compatible with XRD measurement.

Fig. 6
figure 6

Raman spectrum of the AZO thin films deposited onto glass, PET, and Si substrate

Full size image

4 Conclusion

In this paper, the Al doped ZnO thin films successfully grown on glass, PET and Si substrates via thermionic vacuum arc deposition technique. According to XRD pattern, all films are in polycrystalline form. The one should be noted that bigger crystallite sizes (~ 100 nm) were calculated on Si substrate and lower sizes (~ 20 nm) were calculated on glass substrate. Results are also in accordance with FESEM images. The chosen substrate plays a crucial role during the formation of the crystallites. According to the optical results, average transmittance values of the films defined as 50% and 30% on glass and PET substrate in the visible region, respectively. The optical band gaps of the samples were determined as 3.48 eV and 3.5 eV for thin films onto glass and PET substrates, respectively. In Raman analysis, ZnO and Al2O3 peaks were determined. However, ZnAl2O4 oxide was not found in Raman spectra. In addition, any impurity such as NH3, water etc. was not detected.