Open Access
Add to BibSonomy
Abstract
Aluminum-doped zinc oxide (AZO)/Ag/aluminum-doped zinc oxide (AAA) specimens are prepared by varying O2 and N2 flow rates during the AZO deposition to investigate the gas effects on surface morphology, microstructure, conductivity type and optical and electrical properties. The AZO specimens are found to have decreases in grain size and RMS surface roughness (SRq), and an increase in compressive residual stress when the N2 flow rate increases higher than 2.5 sccm. The addition of O2 can contribute a similar tendency to grain size, SRq and residual stress; the grain size and SRq are always comparatively smaller, while the residual stress is higher than that prepared with N2. The grain size and SRq are obtained to be proportional to the AZO thickness, whereas they are inversely proportional to the compressive residual stress of AZO. Specimen D prepared with (N2: 0, O2: 2.5) sccm and having 60-nm AZO thickness possesses the strongest anti-reflection effect of the AAA structure, and therefore impedes the reflection from the Ag interlayer significantly. P-type conductivity is achieved by introducing the N2 into the AZO layers and (N2: 15, O2: 0) sccm is used to achieved the highest carrier concentration (CC) and mobility (Mb), and thus the lowest resistivity (R) of all specimens in this study. Increasing the O2 flow rate leads to decreases in Mb and CC, but can obtain the highest average transmittance ($\overline {\textrm {T}} $) as (N2: 0, O2: 2.5) sccm is applied. The O2 flow rate becomes the dominant factor for the electrical and optical properties and microstructure as AZO films are deposited in the N2 and O2 mixed atmosphere. (N2: 15, O2: 0) sccm is used to achieve the highest CC and the lowest R of the oxide/metal/oxide specimens reported in the literatures.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Transparent conducting oxides (TCOs), such as fluorine-doped tin oxide (FTO), indium tin oxide (ITO), and indium gallium zinc oxide (IGZO) [1–3], have been widely used in the designs of electronic devices and solar cells due to its high transparency in visible light wavelengths (380-750 nm) and superior electrical conductivity. Among all transparent conducting oxides materials, Zinc oxide (ZnO) has drawn considerable attentions because it possesses the advantages including non-toxicity, low fabrication temperature, and wide band gap features [4,5]. Previous studies [6–8] have reported the doping of Al, Ga, or In into ZnO to be an effective method to lower its electrical resistivity.
The study of Nasr et al. [9] indicates that as the doping level of Al into ZnO increases, the grain size of aluminum-doped zinc oxide (AZO) becomes smaller, and it is owing to the dopant segregations at grain boundary. Furthermore, the electrical resistivity initially decreased when the Al doping level was lower than 2% because Al carriers played a role of electron donor. However, as the doping level was higher than 2%, excessive Al carriers led to the impurity scattering, and hence increased the electrical resistivity.
ZnO thin film originally exhibits a n-type conductivity, and it is reported to be very difficult to produce a high-quality p-type ZnO due to the abundant existence of natural defects, such as Zn interstitials, Zn antisite defects, and oxygen vacancies in the thin films [10,11]. Previous study of Xiu et al. [12] indicated that a strong improvement of the n-type property of ZnO was achievable via the doping of group V elements, such as N, P, As, and Sb. Among these group V elements, N is especially the most promising p-type dopant for ZnO due to its electronic structure similar to oxygen as well as a low ionization energy [13,14]. Dress and Wuttig [15] investigated the oxygen effect on the growth of ZnO thin films by sputtering pure zinc target in the oxygen-rich environment. As the oxygen contents reached the critical amount, the particles transformed from metallic mode to oxidic mode, which influence the evolution of residual stress as well as surface morphology significantly. In the study of Horwat and Billard [16], it shows that the ZnO:Al thin films deposited in the presence of argon–oxygen reactive gas mixtures would change the stoichiometry and enhance the transparency of thin films. Bhuvana et.al. [17] proposed a method to synthesize the AlN-doped ZnO (ANZO) thin films by sputtering the ZnO and AlN targets at 500°C simultaneously. The p-type conductivity of ANZO thin films was achieved as the dopant (Al, N) concentrations reach 1 mol. %. Meanwhile, the carrier concentration and carrier mobility were affected significantly by the Al and N doping effects since the Al and N atoms acted as the electron donors and acceptors, respectively.
Several studies have shown that high quality AZO thin films deposition has been achieved by different deposition methods. Knoops et al. [18] reported that AZO films could be deposited using the deposition precursors (Zn(C2H5)2, Al(CH3)3, and O2) by a plasma-enhanced chemical vapor deposition (CVD) process. The study of Shkondin et al. [19] indicates that AZO nanopillars and nanotubes were fabricated using the advanced reactive ion etching and atomic layer deposition (ALD) techniques. In the study of Liu et al. [20], a pulsed laser deposition (PLD) method was applied to fabricate AZO thin film with c-axis direction.
Several studies [21–23] have shown that the TCO/metal/TCO (TMT) structure is a promising structure to achieve a high conductivity and low energy consumption function for transparent conducting thin films. The study of Alford et.al. [24] indicates that the embedded metal layer is the crucial component in the TMT thin films since its thickness can significantly influence the optical and electrical properties, and morphological behaviors of thin films. Among all metals, Ag is the most potential material used as an interlayer on account of its lowest electrical resistivity.
In present study, AZO/Ag/AZO specimens are deposited with the two AZO layers exposed to N2, O2, and N2 + O2 to investigate the gas flow rate effects on the morphological, electrical and optical properties. The correlations among the oxygen vacancy ratio, the carrier concentration, and the optical bandgap are established in sequence for the n-type and p-type specimens, respectively. A figure of merit (FOM) function proposed by Haacke [25] is then applied to evaluate the efficiency of the opto-electrical conversion in the AZO/Ag/AZO triple layer transparent conductive thin films. The three electrical properties of the ANZO/Ag/ANZO specimen prepared with (N2: 15, O2: 0) sccm are compared with those properties reported in the literatures in other to highlight the present results superior to the other TMT specimens.
2. Experimental details
2.1 Specimen preparations
In present study, the triple layers of the AZO/Ag/AZO (AAA) films were deposited using a co-supttering system (ACS-4000-C3, ULVAC, Japan) for the AZO and Ag layers by a direct current (DC) power sources of 90 and 50W, respectively. The depositions for both AZO and Ag were implemented at room temperature (25°C) as well as a rotational speed of 15 rpm to ensure the uniformity of thin films. The depositions of the AZO layers in AAA films were performed at the argon flow rate of 15 sccm, while O2 and N2 were introduced during the depositions for specimens, A – G, by following the design of experiments as shown in Table 1, and the Ag interlayer was deposited at a DC power of 50 W and an argon flow rate of 15 sccm. Due to the restriction of co-sputtering, the total gas flow rate should be controlled no more than 30 sccm. The deposition processes are performed in the nitrogen and oxygen deficient and rich environments by setting the gas flow rate as 2.5 and 15 sccm, respectively. The AZO (ZnO: 98%, Al2O3: 2%) and pure Ag (99.99%) targets were deposited as the oxide layers and the metal interlayer, respectively. In order to remove the contaminants on target surface, AZO and Ag targets were pre-sputtering for 5 minutes using the DC sputtering with Ar flow rate set as 15 sccm. Prior to deposition, bare quartz glass substrates (2.5mm x 2.5mm) were cleaned ultrasonically in ethanol, acetone, and deionized water sequentially. Each cleaning step was performed for 6 minutes by an ultrasonic cleaner and then dried in flowing nitrogen gas.
Table 1. The gas flow rates for the depositions of AZO layers for the AZO/Ag/AZO thin films.
2.2 Characterization
The transmittance (T), absorptance (Ab), and reflectance (Re) of specimens in the wavelengths from 200 to 1000 nm were measured using an UV/Visible/NIR spectrophotometer (HITACHI U4100, Japan). Surface morphology and RMS surface roughness were characterized using an Atomic Force Microscope (AFM) (Bruker Dimension ICON, Germany) in the tapping mode. Analysis of crystallinity structure was conducted by the multipurpose X-ray Diffractometer (XRD) (Bruker AXS, model D8 DISCOVER with GADDS, Germany) with its gracing incident diffraction (GID) identification function and the Cu Kα (λ = 1.54184 Å) radiation at 1° glancing angle. Microstructures and defects of AZO were identified using a High-Resolution Transmission Electron Microscopy (HRTEM; JEM-2100F, JEOL, Japan). TEM samples placed on the standard TEM lacey grid were prepared by using a Dual-Beam Focused Ion Beam (DB-FIB) (FEI Nova-200 NanoLab Compatible, America). Hall effect measurements were conducted at room temperature (25 °C) to measure the electrical properties of AAA specimens, such as carrier mobility, carrier concentration, and resistivity. The chemical state of thin films surface was characterized by an X-ray photoelectron spectroscopy (XPS, ULVAC-PHI PHI 5000 VersaProbe, Japan).
3. Results and discussion
XRD measurement were carried out to investigate the crystal and microstructural properties of AZO/Ag/AZO (AAA) films affected by the variations of N2 and O2 flow rates. Figure 1 shows the diffraction patterns for the AAA specimens coded by A-G, and the polycrystalline phases for all these 7 specimens are identified. The XRD patterns present the preferential AZO (002) orientation with the c-axis perpendicular to the glass substrate, while a minor peak with the AZO (103) orientation is also observed. They show that all these 7 specimens still maintain the hexagonal wurtzite crystalline structure of ZnO in spite of the addition of gas and their flow rate in the deposition. It can also be noticed that specimens B and C possess the comparatively higher diffraction intensity for the (002) orientation compared to the other 5 specimens. The XRD pattern of specimen B, which was prepared using a N2 flow rate of 2.5 sccm, presents the highest diffraction intensity of all specimens. As the N2 gas flow rate increased to 15 sccm, the diffraction intensity started to decline by increasing N2 flow rate. Specimens, D-G, which were prepared with a non-zero oxygen flow rate have noticeable decreases in the diffraction intensity. It shows that the crystallinity of AZO (002) is significantly reduced by the O2 addition during the deposition process, irrespective of the N2 flow rate.
To further investigate the microstructural property of thin films affected by the gas flow rates, the grain size of the AZO(002) orientation is calculated using the Debye-Scherrer’s equation [26]:
where D is the average grain size along the (002) orientation, θ is the Bragg diffraction angle, λ is the X-ray wavelength (λ = 1.54184 Å), and β denotes the full width at half maximum (FWHM) of the AZO (002) peak.The XRD result shows that the AZO layers of the AAA thin films will maintain its original hexagonal crystal structure, irrespective of the addition of N2 or O2. For the hexagonal structure with preferable c-axis orientation, the in-plane stress (${\sigma _{{\mathop{\textrm {film}}\nolimits} }}$) is evaluated using the biaxial stress model [27]:
According to the Eqs. (1 4), the values of grain size, c-axis lattice constant and residual stress have been obtained and shown in Table 2. The largest AZO (002) grain size occurs as a N2 flow rate of 2.5 sccm is used. The grain size of AZO thin films increases from 17.72 to 21.21 nm when the N2 flow rate increases from 0 to 2.5 sccm, and then it decreases when the N2 flow rate is supplied to be 15 sccm. It can be explained that the segregation of ions in the grain boundary deteriorates the crystallinity of films as the N doping concentration is beyond the critical concentration [29]. The mismatches of ionic radii among Zn, Al, and N (rzn2+ = 0.074 nm, rAl3+ = 0.054 nm, rN3- = 0.146 nm) can induce the compressive stresses [30]. Specimens A, B, and C have the N atomic ratios of 0, 2.0, and 3.6%, respectively, and results indicate that the N atomic ratio of 3.6% is beyond the critical concentration, thus reducing the grain size. The variations of RMS surface roughness (SRq) with the nitrogen flow rate present a trend similar to that of grain size. The highest SRq and the largest grain size occur when the N2 flow rate is set as 2.5 sccm. When the N2 flow rate increases from 0 to 2.5 sccm, a reduction in compressive residual stress from -2.12 to -0.96 GPa is observed, and it can be ascribed to the decrease in lattice constant from 5.266 to 5.239 Å. Since the bond length of Zn-N bond (2.04 Å) is longer than that of the Zn-O bond (1.93 Å) [28], the lattice expansion occurs when the N atoms are incorporated into the films, and the substitution of N atoms for the O sites took place in the ZnO lattices. Besides, an increase in lattice constant from 5.239 to 5.246 Å and an enhancement in compressive residual stress from -0.96 to -1.27 GPa are observed when the N2 flow rate increases from 2.5 to 15 sccm. Since the nitrogen atomic ratio has been a value larger than the critical value, the dopanFigt segregation at grain boundary occurs and reduces the lattice constant, thus resulting in an increase in compressive residual stress.
Table 2. The microstructural parameters of the AZO layers for AZO/Ag/AZO triple layer specimens.
Increasing the O2 flow rate will bring in a decrease in grain size. The grain size is 12.73 nm when the oxygen flow rate is 2.5 sccm, and it drops to 11.02 nm as the oxygen flow rate is risen to 15 sccm. The grain sizes for the specimens prepared with the O2 addition are noticeably smaller than that prepared with the addition of N2. Besides, both the surface roughness and grain size of single layer AZO films are reduced, whereas the compressive residual stress is conversely elevated by increasing the O2 flow rate. The rise in compressive residual stress from -2.55 to -2.88 GPa is obtained when the O2 flow rate increases from 2.5 to 15 sccm as a result of the expansion of lattice constant from 5.275 to 5.283 Å. The behaviors demonstrated in these results could be attributed to: (1) the decreasing deposition rate of the films because the sputtering energy is consumed by the unionized neutral O atoms, leading to the crystal defects and a decline in grain size [31]; (2) the interstitial sites of ZnO lattices tend to be occupied by O atoms, and therefore it shows the non-stoichiometric oxygen inside the thin film in the oxygen-rich environment [32]. An increase in grain size from 15.14 to 16.83 nm is observed as the (N2, O2) flow rates vary from (2.5, 12.5) to (12.5, 2.5) sccm. It can be noticed that the effect of the O2 addition on the grain size is more significant than that caused by the N2 addition. Besides, even a small amount of oxygen can reduce the crystalline size significantly. The grain size and SRq are elevated, but the compressive residual stress is conversely lowered by reducing the O2 flow rate even though the N2 flow rate in the mixing gas increases. Since the chemical activity of O2 is higher than that of N2, O2 will be ionized by the plasma prior to N2 during the deposition process; therefore, the effects of O2 on the grain size and surface roughness are more noticeable than N2.
The thin film thickness and deposition rate of the AZO film in the AZO/glass specimens affected by the variations of gas flow rates are characterized by the TEM images, and then the thickness data are applied as the references in the preparations of the triple layer AAA specimens. Figure 2 shows the correlations of AZO thickness with the SRq, grain size, and residual stress along the (002) orientation of the AAA specimens. It shows that the grain size and SRq are varied almost linearly proportional to the thin film thickness, but the residual stress is inversely proportional to it. The results indicate that the variations of the (N2, O2) flow rates can result in a significant change of thin film thickness, and therefore influence the grain size, SRq, and compressive residual stress. Similar trend between film thickness and grain size was also presented in the study of Inguva et al. [33]. The study of Shkondin et al. [19] indicated that the grain size values of AZO (002) varied between 10-20 nm when the AZO thickness was prepared around 100 nm with deposition temperature between 150-250 °C via the use of atomic layer deposition (ALD). Besides, the grain sizes along the AZO (100) orientation were observed between 30-50 nm under the same deposition conditions. In present study, there is no AZO (100) peak observed, and the grain size values of AZO (002) is ∼20 nm when the AZO thickness is ∼100 nm and the deposition was implemented at 25 °C when using the co-sputtering technique. These two results reveal that the grain size of AZO (002) is around 20 nm when the thickness of AZO is around 100 nm, irrespective of the deposition techniques and temperatures. The relatively larger grain size in Ref. [19] was created due to the formation of AZO (100) at 150-250 °C. In the initial stage of growing the film, the crystalline structure of films is significantly affected by the interface between the amorphous quartz substrate and the thin film nanocrystals due to the lattice mismatch. In the next stage of growing the film, it starts to create a higher crystallinity along the c-axis since the lattice construction has become stronger and ordered. Therefore, relatively larger grain size and rougher thin film surface are created by increasing the film thickness when the more ordered microstructure is formed in the thicker film. Generally, the origin of residual stress can be ascribed to the factors including lattice mismatch, thermal mismatch, and interface dislocations [34]. The evolution of residual stress is mainly dependent on the thin film growth, such as its pattern and uniformity. In the present study, an increase in grain size can bring in fewer grain boundaries as the thin film thickness increases; then the reductions of grain boundary induce the drop of compressive residual stress. The effect of interface dislocation between amorphous quartz substrate and crystalline AZO thin film is gradually dwindled by increasing film thickness.
Fig. 2. The grain size, RMS surface roughness, and compressive residual stress for the AZO/Ag/AZO thin films expressed as a function of the AZO thickness.
HR-TEM characterization has been conducted to investigate the influences of N2 and O2 on the defects and crystallines in the AZO thin films. The TEM images of the AZO film in the specimens prepared with the (N2, O2) flow rates of (15, 0), (0, 15), and (12.5, 2.5) sccm are shown in Figs. 3(a)–(c) respectively. The thickness of the AZO film growing along the (002) orientation is determined by the image method of TEM. It can be seen that the AZO film thickness is slightly affected by the variations of N2 flow rate, whereas it is reduced by increasing the O2 flow rate significantly. Defects such as cracks and voids are not detected along the interface between the thin films and the quartz substrate. The crystalline areas are identified via the selected area electron diffraction (SAED) patterns. These patterns reveal the coexistence of AZO (201), (110), (101), (100), (103), (004), and (203) orientations; besides, the (002) is the dominant orientation. The local crystalline area is magnified and shown in Figs. 3((a-3)–(c-3)). The orientations of poly-crystalline structures are identified by measuring the d-spacing values and referring the International Centre for Diffraction Data (ICDD) files. In the study of Schmidt-Mende et. al [35], the results indicate that the defects of ZnO films is dependent on its defect chemistry, and a nearly defect-free ZnO nanostructure can be achieved by chemical vapor deposition (CVD) with the ratio of (ZnO:graphite) set to be (1:1). However, defects in the AZO and ZnO are observed either with or without the addition of gases to Ar by the sputtering method [36–38]. In this study, the line defect of dislocation is present in Fig. 3(a-3), and the lattice distortions in Figs. 3(a-3)–(c-3) are also detected. The dislocation line is one kind of defect usually formed during the thin film depositions owing to the irregularity in the crystal structure along the line. Lattice distortion is a structural disorder as a result of the misalignment of unit cells in the crystals, and it is created mainly due to the atomic scale mismatch of crystals. The interatomic charge transition can also lead to the formation of lattice distortion even though the atomic radii of elements are close each other. The effects of local lattice distortion on lattice constant and bulk modulus are negligible [39]. Several defects are observed in these locally magnified areas, and the formation of defects can be ascribed to the behavior that either the added N atoms had occupied the O sites of AZO lattices or the deterioration of the AZO crystalline had occurred because the energy of ion bombardment on the target surface was insufficient to form the AZO thin film with high crystallinity in the oxygen-rich environment.
Fig. 3. TEM images and SAED patterns for the AZO layer in the AZO/glass specimens. (N2, O2) flow rates are: (a) (15, 0) sccm; (b) (0, 15) sccm; (c) (12.5, 2.5) sccm.
Figures 4(a)–(c) show that all specimens except for specimen A can be classified as the three groups (B and C, D and G, E and F) and the two specimens in the same group have presented the transmittance (T), reflectance (Re), and absorptance (Ab) profiles in a wavelength range of 200 to 1000 nm with the similar trend. The behavior demonstrated in these curves indicate that change in the N2 flow rate did not bring in a significant effect on these three optical properties, irrespective of the O2 content in the deposition process. Define $\overline {\textrm T} $ as the average value of T for the wavelengths between 200 and 1000 nm. Figure 4(a) shows that the sequence of ($\overline {\textrm T} $)D${\geqq}$ ($\overline {\textrm T} $)G${\geqq}$ ($\overline {\textrm T} $)F${\geqq}$ ($\overline {\textrm T} $)E${\geqq}$ ($\overline {\textrm T} $)A${\geqq}$ ($\overline {\textrm T} $)B${\geqq}$ ($\overline {\textrm T} $)C is valid in the visible light region (380-760 nm). This sequence reveals the characteristics that the use of pure N2 and an increasing N2 flow rate are disadvantageous for the rise of $\overline {\textrm T} $; $\overline {\textrm T} $ is elevated by the O2 content when it is below the critical value, bit it is conversely lowered by increasing the O2 content when it is beyond the critical value. Figure 4(b) shows that specimens B and C have the highest Re in the green-yellow region; the D and G specimens have the highest Re in the violet-blue region, and the E and F specimens have the highest Re arising in the orange-red region. All these 7 specimens have the Ab values higher than 70% in the ultra violet region but lower than 20% in the orange-red and near infrared regions, as shown in Fig. 4(c).
Fig. 4. (a) Transmittance, (b) reflectance, and (c) absorptance spectrum for triple layer AZO/Ag/AZO specimens with code A-G.
According to the previous studies of the TCO/Metal/TCO (TMT) structure thin films [40–42], the top and bottom TCO layers have been reported to possess the anti-reflection effect and therefore suppress the reflection from the Ag layer, and the anti-reflection effect of the TCO layers is mainly dependent on the thickness of TCO. Define $\overline {\textrm T} $ and $\overline {{\mathop{\rm Re}\nolimits} } $ as the average value of transmittance and reflectance arising at the wavelengths of the visible light, respectively. On the basis of the microstructures and optical properties of this study, it can be concluded that the variations of gas flow rates affect the AZO thickness and thus the $\overline {{\mathop{\rm Re}\nolimits} } $ significantly, and the relationship between reflection and AZO thickness is shown in Fig. 5. The highest $\overline {\textrm T} $ and the lowest $\overline {{\mathop{\rm Re}\nolimits} } $ are presented in specimen D with the AZO thickness around 60 nm. It can also be noticed that as the AZO thickness is prepared below 65 nm (for D, E, F, G specimens), the $\overline {\textrm T} $ is always higher than 60% and the $\overline {{\mathop{\rm Re}\nolimits} } $ is lower than 26%. $\overline {\textrm T} $ is lower than 55% and $\overline {{\mathop{\rm Re}\nolimits} } $ is higher than 26% as the AZO thickness is larger than 80 nm (for A, B, C specimens). That is, specimens can possess the highest anti-reflection effect when the AZO layer in AAA thin films is prepared with a thickness between 40 and 65 nm. This behavior has a great consistency with the previous study of Mohamed [43], and it can also explain the results in Fig. 5 why specimens D, E, F, and G deposited with the O2 addition always present a relatively lower $\overline {{\mathop{\rm Re}\nolimits} } $ as well as higher $\overline {\textrm T} $ compared to the specimens A, B, C with the N2 addition only.
The absorption coefficient (α) is calculated from the measured spectral transmittance via the formula [44]:
where t is the total thickness of the composite film, and $\overline {\textrm T} $ denotes the mean transmittance. The value of optical bandgap for the thin films is determined using Tauc plot by extrapolating the linear portion of α2 to intersect the hν axis. The relation of absorption coefficient (α) and optical bandgap energy (Eg) can be described as [44]: where C is a constant and hν is the incident phonon energy (h: Planck’s constant 6.63×10−34 J ; ν: the wave frequency of incident light). Figure 6 shows that the α2-Eg curves for specimens, A-G. The Eg values for these seven specimens are shown in Table 3, which are varied in a range of 3.09 to 3.27 eV. Specimens, B and C, prepared with the N2 addition only have been detected with the p-type conductivity; while the other five specimens prepared with a non-zero O2 flow rate, regardless of the N2 addition, are presented to have the n-type conductivity.Table 3. Electrical properties, FOM, and optical bandgap of specimens A-G.
XPS measurement was conducted to verify the chemical state of element on specimen surface. The XPS O 1s core spectrum for specimen A is shown in Fig. 7 as an example. Each O 1s spectrum can be decomposed into three Gaussian-like spectra with their binding energy about 530.3, 531.5, and 532.5 eV, respectively. The peak concentrated at ∼530.3 eV (O1) is attributed to the oxygen in the ZnO crystal lattice, representing the Zn-O bonding without the oxygen vacancy. The peak located at ∼ 531.5 eV (O2) is presented to be the oxygen vacancies, which are produced in the oxygen-deficient regions. The O3 peak at a binding energy of ∼532.5 eV is assigned to the O-H group absorbed on the specimen surface as a result of the atmospheric contamination during the sample handling.
In order to evaluate the relative quantity of the chemical bonding state on specimen surface, the peak areas, AO1, AO2, and AO3, are calculated from the XPS O1s core level spectra, and applied to evaluate the amounts of O1, O2, and O3 in the form of relative peak area ratios (AO1/Atot, AO2/Atot, and AO3/Atot) as shown in Table 4, where Atot is defined as Atot≡AO1+ AO2+ AO3. The influence of oxygen vacancy ratio on electrical properties will be discussed in the later section.
Table 4. Relative area ratios of O1s.
Hall effect measurements were conducted to obtain the carrier mobility (Mb), carrier concentration (CC), and resistivity (R) for specimens, A-G, working at the room temperature (25 °C). Figure 8(a) shows the dependences of these three electrical properties on the N2 flow rate for specimens A-C. For specimen A, its top and bottom AZO layers were prepared in the pure argon atmosphere, it presents a n-type conductivity in accordance with the natural characteristic of ZnO. Specimens B and C, in which AZO layers are deposited in the pure N2 atmosphere, show the p-type conductivity. During the transition of conductivity from n- to p- type, it can be noticed that both the carrier mobility and carrier concentration are reduced from 21.59 (specimen A) to 15.67 cm2/Vs (specimen B) and 7.46 ×1021 (specimen A) to 5.58 ×1021 cm−3 (specimen B), respectively, by introducing a nitrogen flow rate of 2.5 sccm. This decline in carrier concentration can be interpreted as the result of the major carrier transfer from electrons to holes since the N atoms act as a shallow acceptor [45]. Besides, excessive N atoms become the source of the impurity defects, leading to a decrease in carrier mobility, and thus an increase in resistivity due to the ion-impurity effect. When the N2 flow rate is risen from 2.5 to 15 sccm, the carrier mobility and carrier concentration are elevated from 15.67 (specimen B) to 24.78 cm2/Vs (specimen C) and 5.58 ×1021 (specimen B) to 7.96 ×1021 cm−3 (for specimen C), respectively. When more N acceptors are trapped in the AZO layers owing to the doping level enhanced by increasing N2 flow rate, it produces more and more carrier holes, and therefore finally leads to the increases in the p-type carrier concentration and carrier mobility.
Fig. 8. Carrier mobility, carrier concentration, and resistivity for triple layer specimens as a function of (a) N2 flow rate; (b) O2 flow rate; and (C) N2 and O2 mixed gas flow rate.
Figure 8(b) shows that the carrier concentration and carrier mobility are reduced while the resistivity is elevated by increasing the O2 flow rate. The thin films deposited without O2 and N2 (specimen A) tend to result in a higher carrier concentration and mobility as well as a lower resistivity compared to those in specimens, D and E. The AZO thin films deposited with the higher O2 flow rate possess a higher electrical resistivity and hence a lower electrical conductivity. All these three specimens present the n-type conductivity, and this behavior is in agreement with the previous study [46]. The reduction of the n-type conductivity is induced by the addition of O2 because the amount of oxygen vacancies is reduced by the O2 supplement. The oxygen vacancies usually formed under the oxygen-deficient condition acting as electron donors, thus resulting in the enhancement of electrical conductivity. The mechanism can be described as:
Figure 9 shows that the carrier concentration is varied proportional to the oxygen vacancy ratio for the n-type specimens (A, D, E, F, and G). When the oxygen vacancy ratio increases, it can lead to the rise of carrier concentration. However, the opposite tendency is performed for the p-type specimens (B and C). The p-type carrier concentration is reduced by increasing oxygen vacancy ratio.
The variation of gases flow rates can lead to the changes in carrier mobility, carrier concentration, and resistivity. Besides, the variation of thin film thickness is also reported as an important factor for the variation of electrical properties [48]. As shown in Table 3, the AZO thickness is positively proportional to carrier mobility and carrier concentration for specimens A, C, D, E, F, and G. However, specimen B presents a completely different tendency. For example, the relationship between specimens A and B reveals that when the AZO thickness increased from 87.5 (specimen A) to 94.3 (specimen B) nm, carrier mobility decreased from 21.2 (specimen A) to 15.7 (specimen B) cm2/Vs and carrier concentration also drop from 7.46 ×1021 (specimen A) to 5.58 ×1021 cm−3 (specimen B). In this study, it is believed that the variation of oxygen vacancy ratio has its contribution on the change in electrical properties greater than the variation of AZO thickness.
Figure 10 shows the dependence of optical bandgap on carrier concentration, and it indicates that specimens A, D, E, F, and G possess a positive correlation between the optical bandgap (Eg) and carrier concentration (CC). The bandgap widening effect, namely the blue-shift of absorption edge, is presented along with an increase in carrier concentration as a result of lifting the fermi level into the conduction band of the degenerate semiconductor, and this effect is the well-known Burstein-Moss effect (BM effect). The absorption edge in the n-type AZO films is given as [49]:
In order to determine the optimal parameters, which is able to express the best opto-electrical conversion efficiency with the lowest power consumption and the highest efficiency in the transparent conductive electrodes, a Haacke’s figure of merit (FOM) is applied to evaluate [25]:
where ${\textrm{T}^{\ast }}$ is the transmittance obtained at a typical wavelength of 550 nm, and Rsh denotes the sheet resistance of thin films. The values of FOM for these 7 specimens are shown in Table 3, they show that specimen D has the highest FOM (2.15×10−2 Ω−1) while specimen B has the lowest one (4.55×10−6 Ω−1), even though the Rsh of specimen B (27.2 Ω/sq) is comparatively lower than that of specimen D (90.5 Ω/sq). The electrical properties and FOM for the specimen C and D in the present study and the reported works are summarized in Table 5. The results indicate that FOM is determined to be the combined behavior affected by the thicknesses and materials of triple layer films.Table 5. Comparisons of the electrical properties for specimen C in this study and for the reported literatures.
4. Conclusions
The N2, O2, and N2+O2 gases with different flow rates are added in the deposition process of AZO layers for AZO/Ag/AZO thin films in order to investigate their effects on the electrical and optical properties. Following conclusions are made:
- (1) When only N2 is supplied at a flow rate ${\geqq}$2.5 sccm, the segregation of ions in the grain boundary of AZO layer will bring in the reductions of grain size and SRq and the increases in compressive residual stress. The p-type conductivity is available but the transmittance ($\overline {\textrm T} $) is lowered by adding the N atoms into the AZO layers. (15, 0) sccm for the (N2, O2) flow rates can be applied to achieve the highest carrier concentration and mobility, and thus the lowest resistivity of all seven specimens.
- (2) An increase in O2 flow rate can result in a comparatively smaller grain size and SRq, and higher compressive residual stress. The n-type conductivity of the AAA specimen is prevalent. The carrier mobility and concentration are lowered by introducing O2 into the AZO layers. The highest $\overline {\textrm T} $ and the lowest $\overline {{\mathop{\rm Re}\nolimits} } $ are obtained when the (N2, O2) flow rates are (0, 2.5) sccm. The O2 flow rate becomes the dominant factor for the electrical and optical properties, and microstructure of specimens prepared in the N2 and O2 mixing atmosphere.
- (3) Grain size and SRq are elevated, and the compressive residual stress is reduced by increasing the AZO thickness of the AAA specimen. As the AZO layers of specimen D are prepared with a thickness around 60 nm, they can bring in the strongest anti-reflection effect and suppress the reflection from the Ag interlayer, thus presenting the lowest $\overline {{\mathop{\rm Re}\nolimits} } $ and the highest $\overline {\textrm T} $ of all specimens in this study.
- (4) Carrier concentration of the n-type AAA specimens (A, D, E, F, and G) is positively correlated to oxygen vacancy ratio and optical bandgap, whereas the p-type specimens (B and C) show the exactly opposite trend. Both the highest carrier concentration and the lowest resistivity are produced by specimen C. Specimen D possesses the highest FOM (84.39% at the 550-nm wavelength) value.
Funding
Ministry of Science and Technology, Taiwan (106-2221-E-006-099-MY2).
Acknowledgements
This work was financially supported by the Ministry of Science and Technology, Taiwan, R.O.C., under grant 106-2221-E-006-099-MY2.
Disclosures
The authors declare no conflicts of interest.
References
1. D. Zhang, K. Ryu, X. Liu, E. Polikarpov, J. Ly, M. E. Tompson, and C. Zhou, “Transparent, conductive, and flexible carbon nanotube films and their application in organic light-emitting diodes,” Nano Lett. 6(9), 1880–1886 (2006). [CrossRef]
2. H. Yao, G. Zheng, P.-C. Hsu, D. Kong, J. J. Cha, W. Li, Z. W. Seh, M. T. McDowell, K. Yan, and Z. Liang, “Improving lithium–sulphur batteries through spatial control of sulphur species deposition on a hybrid electrode surface,” Nat. Commun. 5(1), 3943 (2014). [CrossRef]
3. K. Nomura, H. Ohta, A. Takagi, T. Kamiya, M. Hirano, and H. Hosono, “Room-temperature fabrication of transparent flexible thin-film transistors using amorphous oxide semiconductors,” Nature 432(7016), 488–492 (2004). [CrossRef]
4. V. Srikant and D. R. Clarke, “On the optical band gap of zinc oxide,” J. Appl. Phys. 83(10), 5447–5451 (1998). [CrossRef]
5. C. Yang, X. Li, Y. Gu, W. Yu, X. Gao, and Y. Zhang, “ZnO based oxide system with continuous bandgap modulation from 3.7 to 4.9 eV,” Appl. Phys. Lett. 93(11), 112114 (2008). [CrossRef]
6. G. Valle, P. Hammer, S. H. Pulcinelli, and C. V. Santilli, “Transparent and conductive ZnO: Al thin films prepared by sol-gel dip-coating,” J. Eur. Ceram. Soc. 24(6), 1009–1013 (2004). [CrossRef]
7. C. Moditswe, C. M. Muiva, and A. Juma, “Highly conductive and transparent Ga-doped ZnO thin films deposited by chemical spray pyrolysis,” Optik 127(20), 8317–8325 (2016). [CrossRef]
8. D. Sun, C. Tan, X. Tian, and Y. Huang, “Comparative Study on ZnO Monolayer Doped with Al, Ga and In Atoms as Transparent Electrodes,” Materials 10(7), 703 (2017). [CrossRef]
9. B. Nasr, S. Dasgupta, D. Wang, N. Mechau, R. Kruk, and H. Hahn, “Electrical resistivity of nanocrystalline Al-doped zinc oxide films as a function of Al content and the degree of its segregation at the grain boundaries,” J. Appl. Phys. 108(10), 103721 (2010). [CrossRef]
10. F. Selim, M. Weber, D. Solodovnikov, and K. Lynn, “Nature of native defects in ZnO,” Phys. Rev. Lett. 99(8), 085502 (2007). [CrossRef]
11. M. D. McCluskey and S. Jokela, “Defects in zno,” J. Appl. Phys. 106(7), 071101 (2009). [CrossRef]
12. F. Xiu, Z. Yang, L. Mandalapu, D. Zhao, and J. Liu, “Photoluminescence study of Sb-doped p-type ZnO films by molecular-beam epitaxy,” Appl. Phys. Lett. 87(25), 252102 (2005). [CrossRef]
13. B. Chavillon, L. Cario, A. Renaud, F. Tessier, F. Cheviré, M. Boujtita, Y. Pellegrin, E. Blart, A. Smeigh, and L. Hammarstrom, “P-type nitrogen-doped ZnO nanoparticles stable under ambient conditions,” J. Am. Chem. Soc. 134(1), 464–470 (2012). [CrossRef]
14. H. Lu, P. Zhou, H. Liu, L. Zhang, Y. Yu, Y. Li, and Z. Wang, “Effects of nitrogen and oxygen partial pressure on the structural and optical properties of ZnO: N thin films prepared by magnetron sputtering,” Mater. Lett. 165, 123–126 (2016). [CrossRef]
15. R. J. Drese and M. Wuttig, “Stress evolution during growth in direct-current-sputtered zinc oxide films at various oxygen flows,” J. Appl. Phys. 98(7), 073514 (2005). [CrossRef]
16. D. Horwat and A. Billard, “Effects of substrate position and oxygen gas flow rate on the properties of ZnO: Al films prepared by reactive co-sputtering,” Thin Solid Films 515(13), 5444–5448 (2007). [CrossRef]
17. K. Bhuvana, J. Elanchezhiyan, N. Gopalakrishnan, and T. Balasubramanian, “Codoped (AlN) and monodoped (Al) ZnO thin films grown by RF sputtering: A comparative study,” Appl. Surf. Sci. 255(5), 2026–2029 (2008). [CrossRef]
18. H. C. Knoops, B. W. van de Loo, S. Smit, M. V. Ponomarev, J.-W. Weber, K. Sharma, W. M. Kessels, and M. Creatore, “Optical modeling of plasma-deposited ZnO films: Electron scattering at different length scales,” J. Vac. Sci. Technol., A 33(2), 021509 (2015). [CrossRef]
19. E. Shkondin, O. Takayama, M. A. Panah, P. Liu, P. V. Larsen, M. D. Mar, F. Jensen, and A. Lavrinenko, “Large-scale high aspect ratio Al-doped ZnO nanopillars arrays as anisotropic metamaterials,” Opt. Mater. Express 7(5), 1606–1627 (2017). [CrossRef]
20. Y. Liu, Q. Li, and H. Shao, “Optical and photoluminescent properties of Al-doped zinc oxide thin films by pulsed laser deposition,” J. Alloys Compd. 485(1-2), 529–531 (2009). [CrossRef]
21. J. Leng, Z. Yu, W. Xue, T. Zhang, Y. Jiang, J. Zhang, and D. Zhang, “Influence of Ag thickness on structural, optical, and electrical properties of ZnS/Ag/ZnS multilayers prepared by ion beam assisted deposition,” J. Appl. Phys. 108(7), 073109 (2010). [CrossRef]
22. K. Ravichandran, K. Subha, A. Manivasaham, M. Sridharan, T. Arun, and C. Ravidhas, “Fabrication of a novel low-cost triple layer system (TaZO/Ag/TaZO) with an enhanced quality factor for transparent electrode applications,” RSC Adv. 6(68), 63314–63324 (2016). [CrossRef]
23. N. Ren, J. Zhu, and S. Ban, “Highly transparent conductive ITO/Ag/ITO trilayer films deposited by RF sputtering at room temperature,” AIP Adv. 7(5), 055009 (2017). [CrossRef]
24. A. Dhar and T. Alford, “High quality transparent TiO2/Ag/TiO2 composite electrode films deposited on flexible substrate at room temperature by sputtering,” APL Mater. 1(1), 012102 (2013). [CrossRef]
25. G. Haacke, “New figure of merit for transparent conductors,” J. Appl. Phys. 47(9), 4086–4089 (1976). [CrossRef]
26. V. Ţucureanu and D. Munteanu, “Enhanced optical properties of YAG: Ce yellow phosphor by modification with gold nanoparticles,” Ceram. Int. 45(6), 7641–7648 (2019). [CrossRef]
27. Q. Shi, K. Zhou, M. Dai, H. Hou, S. Lin, C. Wei, and F. Hu, “Room temperature preparation of high performance AZO films by MF sputtering,” Ceram. Int. 39(2), 1135–1141 (2013). [CrossRef]
28. S. Park, J. Chang, H. Ko, T. Minegishi, J. Park, I. Im, M. Ito, D. Oh, M. Cho, and T. Yao, “Lattice deformation of ZnO films with high nitrogen concentration,” Appl. Surf. Sci. 254(23), 7972–7975 (2008). [CrossRef]
29. R. Amiruddin, S. Devasia, D. Mohammedali, and M. S. Kumar, “Investigation on PN dual acceptor doped p-type ZnO thin films and subsequent growth of pencil-like nanowires,” Semicond. Sci. Technol. 30(3), 035009 (2015). [CrossRef]
30. J.-H. Lee and B.-O. Park, “Transparent conducting ZnO: Al, In and Sn thin films deposited by the sol–gel method,” Thin Solid Films 426(1-2), 94–99 (2003). [CrossRef]
31. Y. Tao, S. Ma, H. Chen, J. Meng, L. Hou, Y. Jia, and X. Shang, “Effect of the oxygen partial pressure on the microstructure and optical properties of ZnO: Cu films,” Vacuum 85(7), 744–748 (2011). [CrossRef]
32. C. R. Gobbiner, M. A. A. Veedu, and D. Kekuda, “Influence of oxygen flow rate on the structural, optical and electrical properties of ZnO films grown by DC magnetron sputtering,” Appl. Phys. A 122(4), 272 (2016). [CrossRef]
33. S. Inguva, R. K. Vijayaraghavan, E. McGlynn, and J.-P. Mosnier, “Highly transparent and reproducible nanocrystalline ZnO and AZO thin films grown by room temperature pulsed-laser deposition on flexible Zeonor plastic substrates,” Mater. Res. Express 2(9), 096401 (2015). [CrossRef]
34. A. Moridi, H. Ruan, L. Zhang, and M. Liu, “Residual stresses in thin film systems: Effects of lattice mismatch, thermal mismatch and interface dislocations,” Int. J. Solids Struct. 50(22-23), 3562–3569 (2013). [CrossRef]
35. L. Schmidt-Mende and J. L. MacManus-Driscoll, “ZnO–nanostructures, defects, and devices,” Mater. Today 10(5), 40–48 (2007). [CrossRef]
36. Y. Hu, Y. Chen, Y. Wu, M. Wang, G. Fang, C. He, and S. Wang, “Structural, defect and optical properties of ZnO films grown under various O2/Ar gas ratios,” Appl. Surf. Sci. 255(22), 9279–9284 (2009). [CrossRef]
37. I. Sieber, N. Wanderka, I. Urban, I. Dörfel, E. Schierhorn, F. Fenske, and W. Fuhs, “Electron microscopic characterization of reactively sputtered ZnO films with different Al-doping levels,” Thin Solid Films 330(2), 108–113 (1998). [CrossRef]
38. S. Yin, M. M. Shirolkar, J. Li, M. Li, X. Song, X. Dong, and H. Wang, “Influences of defects evolvement on the properties of sputtering deposited ZnO: Al films upon hydrogen annealing,” AIP Adv. 6(6), 065020 (2016). [CrossRef]
39. H. Song, F. Tian, Q.-M. Hu, L. Vitos, Y. Wang, J. Shen, and N. Chen, “Local lattice distortion in high-entropy alloys,” Phys. Rev. Mater. 1(2), 023404 (2017). [CrossRef]
40. G. Wang, Z. Li, S. Lv, M. Li, C. Shi, J. Liao, and C. Chen, “Optical absorption and photoluminescence of Ag interlayer modulated ZnO film in view of their application in Si solar cells,” Ceram. Int. 42(2), 2813–2820 (2016). [CrossRef]
41. T. Yang, Z. Zhang, S. Song, Y. Li, M. Lv, Z. Wu, and S. Han, “Structural, optical and electrical properties of AZO/Cu/AZO tri-layer films prepared by radio frequency magnetron sputtering and ion-beam sputtering,” Vacuum 83(2), 257–260 (2008). [CrossRef]
42. W.-S. Liu, Y.-H. Liu, W.-K. Chen, and K.-P. Hsueh, “Transparent conductive Ga-doped MgZnO/Ag/Ga-doped MgZnO sandwich structure with improved conductivity and transmittance,” J. Alloys Compd. 564, 105–113 (2013). [CrossRef]
43. S. Mohamed, “Effects of Ag layer and ZnO top layer thicknesses on the physical properties of ZnO/Ag/Zno multilayer system,” J. Phys. Chem. Solids 69(10), 2378–2384 (2008). [CrossRef]
44. D. Mardare, M. Tasca, M. Delibas, and G. Rusu, “On the structural properties and optical transmittance of TiO2 rf sputtered thin films,” Appl. Surf. Sci. 156(1-4), 200–206 (2000). [CrossRef]
45. K. Iwata, P. Fons, A. Yamada, K. Matsubara, and S. Niki, “Nitrogen-induced defects in ZnO: N grown on sapphire substrate by gas source MBE,” J. Cryst. Growth 209(2-3), 526–531 (2000). [CrossRef]
46. M. Lalanne, J. Soon, A. Barnabé, L. Presmanes, I. Pasquet, and P. Tailhades, “Preparation and characterization of the defect–conductivity relationship of Ga-doped ZnO thin films deposited by nonreactive radio-frequency–magnetron sputtering,” J. Mater. Res. 25(12), 2407–2414 (2010). [CrossRef]
47. H.-W. Ra, R. Khan, J. Kim, B. Kang, K. Bai, and Y. Im, “Effects of surface modification of the individual ZnO nanowire with oxygen plasma treatment,” Mater. Lett. 63(28), 2516–2519 (2009). [CrossRef]
48. Y. S. Jung, Y. S. Park, K. H. Kim, and W.-J. Lee, “Properties of AZO/Ag/AZO multilayer thin film deposited on polyethersulfone substrate,” Trans. on Electr. and Electron. Mater. 14(1), 9–11 (2013). [CrossRef]
49. H. Lee, S. Lau, Y. Wang, K. Tse, H. Hng, and B. Tay, “Structural, electrical and optical properties of Al-doped ZnO thin films prepared by filtered cathodic vacuum arc technique,” J. Cryst. Growth 268(3-4), 596–601 (2004). [CrossRef]
50. D. B. Potter, M. J. Powell, I. P. Parkin, and C. J. Carmalt, “Aluminium/gallium, indium/gallium, and aluminium/indium co-doped ZnO thin films deposited via aerosol assisted CVD,” J. Mater. Chem. C 6(3), 588–597 (2018). [CrossRef]
51. A. Yamamoto, T. Kido, T. Goto, Y. Chen, and T. Yao, “Bandgap renormalization of ZnO epitaxial thin films,” Solid State Commun. 122(1-2), 29–32 (2002). [CrossRef]
52. B. Zhang, B. Yao, Y. Li, Z. Zhang, B. Li, C. Shan, D. Zhao, and D. Shen, “Investigation on the formation mechanism of p-type Li–N dual-doped ZnO,” Appl. Phys. Lett. 97(22), 222101 (2010). [CrossRef]
53. J. Lu, Z. Ye, L. Wang, J. Huang, and B. Zhao, “Structural, electrical and optical properties of N-doped ZnO films synthesized by SS-CVD,” Mater. Sci. Semicond. Process. 5(6), 491–496 (2002). [CrossRef]
54. L. Pauling, The Nature of the Chemical Bond (Cornell University Press, 1960), Vol. 260.
55. M. D. Kumar, Y. C. Park, and J. Kim, “Impact of thin metal layer on the optical and electrical properties of indium-doped-tin oxide and aluminum-doped-zinc oxide layers,” Superlattices Microstruct. 82, 499–506 (2015). [CrossRef]
56. H.-W. Wu and C.-H. Chu, “Structural and optoelectronic properties of AZO/Mo/AZO thin films prepared by rf magnetron sputtering,” Mater. Lett. 105, 65–67 (2013). [CrossRef]
57. J. H. Kim, Y.-J. Moon, S.-K. Kim, Y.-Z. Yoo, and T.-Y. Seong, “Al-doped ZnO/Ag/Al-doped ZnO multilayer films with a high figure of merit,” Ceram. Int. 41(10), 14805–14810 (2015). [CrossRef]
58. M. G. Varnamkhasti, H. R. Fallah, M. Mostajaboddavati, and A. Hassanzadeh, “Influence of Ag thickness on electrical, optical and structural properties of nanocrystalline MoO3/Ag/ITO multilayer for optoelectronic applications,” Vacuum 86(9), 1318–1322 (2012). [CrossRef]
59. J. H. Kim, J. H. Lee, S.-W. Kim, Y.-Z. Yoo, and T.-Y. Seong, “Highly flexible ZnO/Ag/ZnO conducting electrode for organic photonic devices,” Ceram. Int. 41(5), 7146–7150 (2015). [CrossRef]
60. J. H. Kim, T.-W. Kang, S.-I. Na, Y.-Z. Yoo, and T.-Y. Seong, “ITO-free inverted organic solar cells fabricated with transparent and low resistance ZnO/Ag/ZnO multilayer electrode,” Curr. Appl. Phys. 15(7), 829–832 (2015). [CrossRef]
61. Y.-T. Li, C.-F. Han, and J.-F. Lin, “Characterization of the electrical and optical properties for a-IGZO/Ag/a-IGZO triple-layer thin films with different thickness depositions on a curved glass substrate,” Opt. Mater. Express 9(8), 3414–3431 (2019). [CrossRef]
