Figure 1 displays an overview of the thin film structure and a schematic diagram of the four types of (In, Ga) co-doped ZnO multi-deposition thin films. As shown in Fig. 1 (a), lattice mismatch occurs owing to the difference in lattice parameters between the substrate and the bottom layer. The four different types of thin films have in common the top six layers of (In, Ga) co-doped ZnO multi-layer thin films in varying bottom layers: with In-doped ZnO (IZO) as the bottom layer for type 1 (Fig. 1 (b)), Ga doped ZnO (GZO) as the bottom layer for type 2 (Fig. 1 (c)), (In, Ga) co-doped ZnO (IGZO) as the bottom layer for type 3 (Fig. 1 (d)), and Ti doped ZnO (TZO) as the bottom layer for type 4 (Fig. 1(e)). Each type represents a structure with a different lattice mismatch between the substrate and the bottom layer. And all the films had a thickness of about 450 nm.
Figure 2 illustrates the X-ray diffraction (XRD) patterns of the ZnO-based thin films, which were deposited as the bottom layers of the four types of (In, Ga) co-doped ZnO samples. All four patterns showed that the (002) peak positions were slightly shifted to lower angles after the carbon-dioxide laser annealing process.
$$d=\frac{n\lambda }{2sin\theta }$$
1
$$\frac{1}{{d}_{hkl}^{2}}=\frac{4}{3}\left(\frac{{h}^{2}+hk+{k}^{2}}{{a}^{2}}\right)+\frac{{l}^{2}}{{c}^{2}}=\frac{{l}^{2}}{{c}^{2}}$$
2
$$\frac{c}{a}=1.6028-13.4\times {10}^{-7}T-2.7\times {10}^{-9}{T}^{2}$$
3
The interplanar spacing was calculated using Bragg’s law (Eq. (1)), where n is an integer (1, 2, 3, 4 ⋯), λ (0.154 nm) is the X-ray wavelength of the CuKα source, and θ is the Bragg diffraction angle [30]. The lattice parameter c is calculated using Eq. (2) where \({d}_{hkl}\) is the interplanar spacing of the (hkl) index and a and c are the lattice parameters [31]. Eq. (3) represents the ratio of the lattice parameters a to c according to the measurement temperature, where T is the absolute temperature [32]. All the measurements were performed at room temperature (T = 298 K). By using Eq. (2) and Eq. (3), we can derive each film’s lattice parameter a. Table 1 shows the (002) peak angles, lattice parameter c, and lattice parameter a for the ZnO based thin films subjected to different annealing processes; electrical furnace annealing (F) was performed, followed by sequential carbon-dioxide laser annealing (FL). As presented in Table 1, the Ga doped ZnO thin film showed the highest value of lattice parameter a, followed by (In, Ga) co-doped ZnO thin films, In doped ZnO thin films, and Ti doped ZnO thin films for both two different annealing processes.
Table 1
(002) peak angle, lattice parameter c, and lattice parameter a for the ZnO-based thin films with 2 different annealing processes.
Materials | (002) peak angle (2θ) (F / FL) | Lattice parameter c (Å) (F / FL) | Lattice parameter a (Å) (F / FL) |
IZO | 34.5927 / 34.5723 | 5.1797 / 5.1827 | 3.2329 / 3.2348 |
GZO | 34.4744 / 34.4458 | 5.1969 / 5.2011 | 3.2437 / 3.2463 |
IGZO | 34.5112 / 34.4703 | 5.1915 / 5.1975 | 3.2403 / 3.2441 |
TZO | 34.6213 / 34.5927 | 5.1755 / 5.1797 | 3.2303 / 3.2329 |
Figure 3 depicts the analysis results of X-ray diffraction (XRD) in the four types of (In, Ga) co-doped multi-deposition thin films on sapphire substrates. The (002) peaks, which are the preferred orientation of the hexagonal wurtzite structure of ZnO, clearly appeared without secondary phases. After carbon-dioxide laser annealing, the intensity of the (In, Ga) co-doped ZnO multi-deposition thin films increased. This implies that the carbon-dioxide laser post annealing process improved the crystallinity of the thin films. Hence, we established that all types of (In, Ga) co-doped ZnO thin films were well deposited and crystallized on the sapphire substrates.
Figure 4 depicts the lattice mismatch rates of the four different types of (In, Ga) co-doped multi-deposition thin films against the sapphire substrate. The mismatch rates were calculated by following equation [33]:
$$Mismatch rate= \frac{{a}_{sapphire}-{a}_{film }}{{a}_{sapphire}}\times 100 \left(\%\right)$$
4
where \({a}_{sapphire}\) and \({a}_{film}\) are lattice parameter a of sapphire substrate (a = 4.754 Å) [34] and bottom layer of (In, Ga) co-doped multi-deposition thin film, respectively. After the electrical furnace process, type 2 exhibits the lowest mismatch rate between the substrate and the film. It can be observed that types 2, 3, 1, and 4 have the lowest resistance values in that order. After the carbon-dioxide laser annealing process was applied, the lattice mismatch rates of all the (In, Ga) co-doped multi-deposition thin films diminished. This was caused by a rise in the lattice parameter a of the thin film, which was calculated from the XRD peaks in Fig. 2 and presented in Table 1. Even when carbon-dioxide laser annealing was applied, there was no change in the order of the lattice mismatch rates of the thin films. The calculated mismatch rates were 31.95%, 31.71%, 31.76%, and 31.99% for type 1, type 2, type 3, and type 4 in the (In, Ga) co-doped ZnO multi-deposition thin films, respectively. The misfit strain can be expressed as follows [35]:
$$\epsilon =\frac{{a}_{substrate}-{a}_{film}}{{a}_{substrate}}$$
5
and is caused by the lattice mismatch rate. According to Vlassak [18], for a thin film prepared on a sapphire substrate with a grain size L0 and the grain size increased to L after the annealing in the electrical furnace, the volumetric strain compared to the initial state is:
$${\varDelta V}^{XS}=3\varDelta a(\frac{1}{L}-\frac{1}{{L}_{0}})$$
6
where Δa is the excess volume per unit of the grain boundary. The misfit strain and volumetric strain have the following correlation:
$$\epsilon =-\frac{1}{3}\varDelta {V}^{XS}=\varDelta a(\frac{1}{L}-\frac{1}{{L}_{0}})$$
7
and stress:
$$\sigma =M\varDelta a\left(\frac{1}{{L}_{0}}-\frac{1}{L}\right)=M\epsilon$$
8
where M is the biaxial modulus of the thin film. Thus, through the above-mentioned equations, it can be seen that lattice mismatch, misfit strain, and stress are in a directly proportional relationship with each other. Therefore, the stress was reduced owing to a decrease in the lattice mismatch due to carbon-dioxide laser annealing.
Figure 5 shows the FE-SEM surface morphology of (In, Ga) co-doped multi-deposition thin films with electrical furnace, electrical furnace plus carbon-dioxide laser annealing analyzed by field emission scanning electron microscopy (FE-SEM). As shown in the Figureure, type 2 exhibited the largest crystallite size for both the electrical furnace and furnace & carbon-dioxide laser annealing processes. The crystallite size decreased in the order of type 3, type 1, and type 4. By comparing the images of two different annealing processes, we can see that the crystallite size was slightly increased after carbon-dioxide laser annealing was applied. This result is caused by the high-energy irradiation of the carbon-dioxide laser annealing process. The crystallites of the (In, Ga) co-doped ZnO multi-deposition thin films absorbed the energy of the laser irradiation process and the crystallite size became larger than that of the films for which only the electrical furnace was processed. The calculated crystallite sizes of the (In, Ga) co-doped multi-deposition thin films are sorted in Table 2. Table 2 indicates the crystallite sizes of the four types of (In, Ga) co-doped multi-deposition thin films deposited on sapphire substrates. The crystallite sizes were calculated by the Scherrer equation [36–37]:
Table 2
Grain size and grain size difference of (In, Ga) co-doped multilayered thin films with 2 different annealing processes.
Type | Grain size (nm) (F) | Grain size (nm) (FL) | Grain size difference (nm) |
Type 1 | 21.16 | 21.82 | 0.66 |
Type 2 | 21.44 | 22.81 | 1.37 |
Type 3 | 21.32 | 22.09 | 0.77 |
Type 4 | 19.95 | 20.56 | 0.61 |
$$D= \frac{K\times \lambda }{B\times cos{\theta }_{B}}$$
9
where λ is the X-ray wavelength (0.154 nm) of the CuKα source, K is a appointed constant of 0.89, B is the constant of Bragg’s angle, and B is the full width at half maximum (FWHM) values of the (002) diffraction peaks in the XRD data presented in Fig. 3. The calculated crystallite size values coincided with the FE-SEM images. Moreover, the grain growth values after carbon-dioxide laser annealing were 0.66, 1.37, 0.77, and 0.61 nm for type 1, type 2, type 3, and type 4, respectively. Here, the role of the energy of carbon-dioxide laser is the driving force for grain growth in thin films [38]. When carbon-dioxide laser is irradiated on a thin film, energy is used to relieve the stress of the thin film. The remaining energy assists in grain growth. Grain growth tends to decrease, where high stress or strain remains in the structure. Figure 4. showed that the lattice mismatch was reduced through carbon-dioxide laser annealing, indicating that carbon-dioxide laser annealing relieved the stress in the thin film. Thus, we can derive the relieved stress data from the reduced lattice mismatch rate. Because the carbon-dioxide laser energy is used to relieve the stress, the remaining energy of the carbon-dioxide laser can be estimated from the relieved stress data. Additionally, grain growth tends to decrease in total energy stored in thin films, such as strain energy and dislocation energy. These energies are produced by the existence of defects or dislocations [19]. Considering the residual stress in the thin film and the remaining carbon-dioxide laser energy, type 2 is the most suitable state for grain growth. Therefore, the calculated lattice mismatch in the film can be the main factor in increasing the grain size of the structure. The optical annealing energy from the carbon-dioxide laser was first applied to the lattice-mismatched area to cure the mismatch instead of the grain growth process. As a result, we believe that the lower lattice-mismatched type 2 has a lower resistance with higher grain growth values, while the higher lattice-mismatched type 4 has higher resistance with lower grain growth values.
Figure 6 illustrates the sheet resistance of four different types of (In, Ga) co-doped ZnO multi-deposition thin films measured following electrical furnace and after furnace and carbon-dioxide laser annealing processes sequentially. Measured after the electrical furnace process, the type 2 specimen depicted the lowest sheet resistance value of 1.24 MΩ/sq., while the other types exhibited higher sheet resistances than type 2, and all types of (In, Ga) co-doped multi-deposition thin films had MΩ/sq. level of sheet resistance. When carbon-dioxide laser annealing was applied following the electrical furnace process on the thin films, the sheet resistance drastically decreased. In order to exist as a substitutional form of (In, Ga) at the Zn site, more than the formation energy of substitutional (In, Ga) should be applied [39]. (In, Ga) were well substituted with Zn through the energy of the carbon-dioxide laser, which generated free electrons and increased the conductivity of the thin films. Moreover, the carbon-dioxide laser annealing process can remove defects in the thin film and induce grain growth. The electrons, which are charge carriers of (In, Ga) co-doped multi-deposition thin films, are trapped in defects or scattered in grain boundaries [40]. These phenomena degrade the conductance of the thin films. However, by introducing carbon-dioxide laser irradiation, high optical energy can be transmitted to the thin film, and the stress can be relieved. As a result, grain growth could be induced. Therefore, carbon-dioxide laser annealing leads to a reduction in defects and grain boundary scattering. As depicted in the Figureure, the sheet resistance decreased from MΩ/sq. to kΩ/sq. level after the carbon-dioxide laser annealing process. Type 2 depicted the lowest sheet resistance of 34.5 kΩ/sq.
Figure 7 displays the optical transmittance measurement for four different types of (In, Ga) co-doped multi-deposition thin films following the electrical furnace (a) and after the furnace and carbon-dioxide laser sequential irradiation process (b). As depicted in Fig. 7 (a) and Fig. 7 (b), the average optical transmittance of all types at wavelength from 400 to 800 nm, which is the wavelength range of the visible region, were higher than 85%. The average optical transmittance data indicate that all four different types of (In, Ga) co-doped multi-deposition ZnO thin films have high optical transmittance properties in the visible region and are suitable for optical device applications such as transparent conducting oxides.
Figure 8 shows the energy band gaps of the four different types of (In, Ga) co-doped ZnO multi-deposition thin films. The calculated energy band gap was derived by employing Tauc’s plot. Figure 8 (a) and Fig. 8 (b) showed the energy band gaps of the thin films after the electrical furnace and after the furnace and carbon-dioxide laser annealing, respectively. By comparing Fig. 8 (a) and Fig. 8 (b), we can Figureure out that the energy band gap slightly rised after carbon-dioxide laser annealing was performed. This increased energy band gap is originated from the Burstein-Moss effect [41–42]. carbon-dioxide laser annealing induced an increase in carrier concentration. This results in a shift of the Fermi level into the conduction band, which leads to the broadening of the optical bandgap energy.
Figure 9 illustrates the photoluminescence (PL) spectra of the four different types of (In, Ga) co-doped multi-deposition thin films after electrical furnace (a) and after electrical furnace and carbon-dioxide laser post annealing processes. All types of (In, Ga) co-doped multi-deposition thin films depicted major luminescent peaks at approximately 380 nm. When the excited electron generated by the He-Ag laser returns to the ground state, light is emitted. The wavelength of the light became shorter as the energy band gap increased. The energy band gap can be enlarged by the increase of carrier concentration derived from the Burstein-Moss effect. Table 3. explains each type’s energy band gap calculated by following equation [43]:
Table 3
Peak wavelength and energy band gap of (In, Ga) co-doped multilayered thin films with 2 different annealing processes.
Type | Peak wavelength (nm) (F) | Peak wavelength (nm) (FL) | Energy band gap (eV) (F) | Energy band gap (eV) (FL) |
Type 1 | 378.8 | 378.6 | 3.271 | 3.274 |
Type 2 | 378.4 | 377.9 | 3.277 | 3.280 |
Type 3 | 378.7 | 378.4 | 3.274 | 3.277 |
Type 4 | 379.7 | 379.3 | 3.265 | 3.268 |
$$E=\frac{hc}{\lambda }$$
10
where h is the Planck constant, c is the velocity of light, and \(\lambda\) is the wavelength of the excited photons detected by the luminescent peaks in the PL spectrum. The peak position slightly blue-shifted toward a shorter wavelength, and an increase in the optical energy band gap derived after the carbon-dioxide laser irradiation process. Figure 10 displays the energy band gap of the four types of (In, Ga) co-doped multi-deposition thin films extracted and derived by Tauc’s plot and PL spectra, respectively. We can see that the energy band gap data of Fig. 9 and Table 3 have a similar tendency to Tauc’s plot presented in Fig. 8. Thus, it can be confirmed that the electrical conductivities of the (In, Ga) co-doped ZnO multi-deposition samples were enhanced by carbon-dioxide laser annealing.
Figure 11 illustrates the Figure of merit for four different types of (In, Ga) co-doped multi-deposition thin films defined by Haacke [44]:
$${\varphi }_{TC}=\frac{{T}^{10}}{{R}_{s}}$$
11
where \(T\) is the optical transmittance value of the visible region in the wavelength range of 400 to 800 nm, and \({R}_{s}\) is the sheet resistance of the thin films. For optical device applications, low sheet resistance and high optical transmittance are required. Therefore, Haacke’s Figure of merit can be a good indicator of its ability for optical device applications. As shown in the Fig. 11, type 2 showed the highest Figure of merit value of 9.75 × 10− 6 among the four different types of (In, Ga) co-doped multi-deposition thin films. This result indicates that type 2 is the most suitable structure and process for optical applications.