Introduction

Silver oxide thin films gained the attention of many research groups in recent years [1,2,3,4,5,6,7,8,9,10,11,12,13], mainly because they exhibit significant applications in gas sensors, passivation, organic field-effect transistors, perovskite solar cells, organic light-emitting diode encapsulation, optical data storage, and surface plasmon optics, etc.

There are numerous deposition techniques for obtaining Al2O3 thin films as thermal evaporation [2],atomic layer deposition (ALD) [13,14,15], magnetron sputtering [16, 17], pulsed-laser deposition (PLD) [18], spray pyrolysis [19, 20], and sol-gel coating [21].

Valence electrons in any conductor act like free electrons, where the collision among the electrons is similar among gas molecules that are described in the kinetic theory of gases [22]. Optical properties of metals are produced due to the interaction between the fallen photons on the surface and the electronic cloud [22]. Optical reflectivity is considered the most important physical property of the metal layers, which generally relates to the interaction between the light and the free electrons which could be expressed by the equation of dispersion as follows [22, 23]:

$$ {\omega}_{\mathrm{p}}=\sqrt{\frac{4\pi N{e}^2}{\varepsilon_{\infty }{m}^{\ast }}} $$
(1)

where N is the concentration of conducting electrons, e the electron charge, ∞ the high frequency (visible wavelength), ɛ the dielectric constant, and m* the electron effective mass.

Both electron concentration and electron effective mass affect this plasma edge [22] and could be directly determined from the reflecting spectra where dramatic change of reflectivity happens at the plasma edge as a result of photon reflection from the conduction band electron plasma oscillations [23]. Furthermore, equation indicates that, when carrier concentration and hence typically the conductivity are increased, the plasma edge moves to high frequencies (shorter wavelengths) [23]. In this paper, we are interested in identifying the factors affecting the plasma edge formed in the reflectance spectra of aluminum oxide thin films prepared by the thermal oxidation of the aluminum thin films deposited by DC magnetron sputtering of aluminum targets.

Experimental

Sample Preparation

Pure Al thin films (99.99%) were deposited at room temperature onto thoroughly cleaned n-type Si (100) and glass substrates from a high-purity Al (99.99%) target using DC magnetron sputtering process. The substrate placed below the source in the direction of ablated material flux. The deposition conditions are given in Table 1.

Table 1 The thermal treatment conditions for each sample
Full size table

Each deposited film was annealed in air at specific temperature (100, 200, 300, 400, and 500 °C) in order to oxidize the aluminum atoms. Table 2 contains the thermal treatment conditions for each sample.

Table 2 The thermal treatment conditions for each sample
Full size table

Sample Characterization

Crystallite structure of the films was measured by X-ray diffraction (XRD) using (Stoe StadiP) transmission X-ray diffractometer employing a Cu-Kα1 (λ = 1.54060 Å) source. The optical transmittance and reflection spectra were recorded with a UV–Vis spectrophotometer (Cary 5000).

Results and Discussion

X-ray diffraction measurements were carried out on the aluminum oxide thin films. The XRD profiles of these samples are shown in Fig. 1. As seen in this figure, the A sample spectrum contains a large number of peaks and broad bands, some of which are related to aluminum metal and the rest are related to some phases of aluminum oxide (alpha, beta, kappa, gamma). This great diversity of phases makes it difficult to make an accurate description of the structural composition of this sample. On the other hand, when the annealing temperature increases, the number of peaks becomes limited and the spectrum becomes clearer. Therefore, we determined the main peaks in the spectra of samples from B to F. Table 3 contains the main features of all the peaks observed in XRD spectra. We have taken into account the determination of the samples in which each of the main peak clearly shows. It can be seen that the alpha phase peak appears in the spectra of the C, E, and F samples. In addition, the samples that have a high percentage of aluminum metal are A, D, and E. The rhombohedra phase of aluminum oxide has a high percentage in three samples (A, B, and D). It is interesting to note that the peak of the silicon does not show in the spectra of the samples except for the sample (F) spectrum, which indicates the low density of this sample and that it has a porous structure.

Fig. 1
figure 1

The XRD spectra of the prepared samples

Full size image
Table 3 The main features of all the peaks observed in XRD spectra
Full size table

The structures we obtained are more diversified than that recorded in previous work [2], which adopted the method of thermal evaporation to obtain aluminum thin films.

The ratio of non-oxidizing aluminum atoms in each sample can be used to investigate the stability of the bonds (Al–O) during thermal treatment. This ratio is proportional to the sum of the relative areas of the aluminum peaks in the XRD spectrum. In order to calculate the relative areas, a fitting process performed for each peak adoption of the Gaussian and Lorentzian shape. Table 4 contains the position and the area of each peak, in addition to the total area (SAl%) of the prepared samples.

Table 4 The position, the area of each peak, and the total area (SAl%) of the prepared samples
Full size table

Figure 2 illustrates the total area (SAl%) as a function of annealing temperature.

Fig. 2
figure 2

SAl% as a function of annealing temperature

Full size image

It can be observed that in general, the ratio of aluminum atoms decreases with increasing annealing temperature except for the sample D (400 °C), where the ratio of non-oxidizing aluminum atoms increases. This indicates that oxidation processes are limited at the temperature 400 °C where a low-concentration mixture of the rhombohedra and cubic Al2O3 phases is formed. On the other hand, the annealing at 300 °C (sample C) appears to be suitable for the formation of the phase alpha and kappa p with high oxidation rates. Almost complete oxidation process also occurs in the case of the sample F (500 °C/2 h) where the alpha stable phase predominates.

Figure 3 shows the reflectance spectra of the prepared samples.

Fig. 3
figure 3

The reflectance spectra of the prepared samples

Full size image

By comparing this figure with Fig. 2, we notice that the effect of annealing temperature on reflectance is not the same as its effect on the proportion of aluminum in the film, and this makes us believe that the reflectivity is not related to the ratio of aluminum atoms. On the other hand, we note that in all spectra, there is a wide band to the left of the plasma edge and that this band slopes sharply toward the plasma edge, and we note that the length of the slope decreases with both the increase in annealing temperature and in the reflectance. This can be attributed to the different optical behavior on both sides of the plasma edge. In the case where λ < λp, light interacts with individual aluminum nanoparticles. In the case where λ > λp, light interacts with surface plasmons of larger Al nanoparticles, where, the concentration of surface plasmons depends on the annealing temperature, and accordingly, we can conclude from Fig. 3 that samples C, E, and D have largest Al nanoparticles compared to the rest of the samples.

The presence of two types of aluminum nanostructures leads to an important conclusion, which is the presence of a plasma edge for individual aluminum atoms (λ′p). Depending on Fig. 3, it can be expected that this edge is in the ultraviolet range (λ′p < 300 nm).

As mentioned in the introduction, samples with plasma edges with short wavelengths have high concentration of charge carriers. By applying this rule to the Fig. 4 that illustrates plasma edge as a function of annealing temperature, we conclude that sample D has the highest concentration of charge carriers followed by sample E. The charge carriers here are intended to be aluminum surface plasmons.

Fig. 4
figure 4

Plasma edge as a function of annealing temperature

Full size image

Figure 5 shows plasma edge as a function of the total area (SAl%). We notice that C and F samples are on the left side of the figure and have high plasma edges because of their low concentrations of surface plasmons.

Fig. 5
figure 5

Plasma edge as a function of the total area (SAl%)

Full size image

The A and B samples, which are located on the right side of the shape, also have high plasma edges despite having high concentrations of aluminum metal. This is due to the fact that the high concentration of metal particles and their interaction with each other prevents the occurrence of response vibrations of surface plasmons. We conclude from the above that the samples D and E are best suited as Al/Al2O3 composites for surface plasmon applications because they have high concentrations of surface plasmons.

Because the prepared films were opaque, we calculated the optical energy band gap from the reflectance spectra using Kubelka-Munk method [24, 25]. Figure 6 shows the method of calculating the energy band gap in the case of sample E.

Fig. 6
figure 6

Variation of (αhʋ)2 against hʋ for sample E

Full size image

Figure 7 shows the optical band gap as a function of annealing temperature. We notice that the band gap increases with increasing annealing temperature, but at the temperature of 400 °C (sample D), the band gap decreases. However, it increases again with doubling the treatment time using the same temperature.

Fig. 7
figure 7

The optical band gap as a function of annealing temperature

Full size image

This behavior can be understood by investigating the relationship between the optical energy band gap and the plasma edge (Fig. 8).

Fig. 8
figure 8

The optical band gap as a function of plasma edge

Full size image

We notice that, for λp < 830 nm, the band gap decreases when the plasma edge increases. For higher plasma edges, the band gap becomes independent of the plasma edge, in which case the band gap related only to the structural composition of the thin film. The great effect of the ratio of the non-oxidizing aluminum atoms appears here, as the A and B samples have low values for the band gap because they have a high content of the non-oxidized aluminum atoms. Figure 8 is very important because it helps to precisely determine the state at which the plasma effect begins to appear.

We found that, in the samples A, C, E, and F, the band gap related only to the structural composition of the thin film. This also appears applicable to the plasma edge. Despite the many phases formed in these samples, it appears that the plasma edge is proportional to the concentration of the alpha phase. The formation of this phase is associated with high oxidation rates; this is the reason for obtaining a small concentration of aluminum in the sample F, which leads to a low concentration of surface plasmons, and it will naturally increase the film’s dielectric constant. In addition, the significant effect of the grain size on the surface plasmon resonance [26] must be taken into account.

Finally, the optical measurement of the plasma frequency is seen as a non-contact probe to measure the electrical properties [23]. In our case, this seems to be incorrect for all cases, especially for λp < 300 nm, as the band gap (which can be expressed as a measure of conductivity) decreases with increasing the plasma edge. The diversity of structural phases, along with the diversity of preparation conditions (annealing temperature), is the main factor determining the overall optical and electrical characteristics of the prepared thin films. However, investigating the relationship between the optical band gap and the plasma edge appears necessary for the engineering design of Al/Al2O3 composites, which make them suitable for specific applications.

Conclusions

In this work, aluminum oxide films are deposited by DC magnetron sputtering of aluminum metal on silicon and glass substrates with subsequent oxidation of deposition products at different annealing temperatures. Annealing processes were performed in the air using an oven. The effect of temperature on XRD and reflectance spectra of the prepared films was studied. The results obtained included the following:

  • The structural composition of the prepared films is highly dependent on the annealing temperature.

  • The annealing at 500 °C for 2 h provides almost complete oxidation with a high alpha stable phase ratio.

  • In addition to the plasma threshold (λp) caused by aluminum nanoparticles, we monitored the presence of a plasma edge for the individual aluminum atoms (λ′p).

  • The annealing temperatures in the range 400–500 °C are best suited as Al/Al2O3 composites for surface plasmon applications because they have high concentrations of surface plasmons.

  • The curve Eg vs. λp is very important for the engineering design of Al/Al2O3 composites for surface plasmon applications.