Synthesis and Characteristics of Zn-Doped CuCrO₂ Transparent Conductive Thin Films

Abstract

The effects of doping a p-type CuCrO₂ film with zinc on its structural and optoelectronic properties were investigated by experiments using CuCr_1−xZn_xO₂ thin films (x = 0, 0.025, 0.065, 0.085). An increase in the amount of zinc dopant in the thin films affected the lattice constant and increased its Gibbs free energy of phase transformation. Cross-sectional images of the CuCrO₂ thin film samples exhibited a dense polygonal microstructure and a surface morphology with protruding nanoscale granules. With the increase in the amount of Zn dopant, the surface roughness decreased, thereby increasing the amount of incident photons as well as the visible-light transmittance and ultraviolet-light absorption of the thin films. With the zinc doping in the CuCrO₂ thin films, the band gap increased from 3.09 to 3.11 eV. The substitution of Cr³⁺ with Zn²⁺ forms hole carriers in the crystals, which was demonstrated by X-ray photoelectron spectroscopy and Hall effect measurements. The conductivities and carrier concentrations of the Zn-doped CuCrO₂ thin films were greater than those of undoped CuCrO₂. The CuCr_1−xZn_xO₂ film (x = 0.065) exhibited the best optoelectronic properties; its carrier concentration and resistivity were 1.88 × 10¹⁷ cm⁻³ and 3.82 Ωcm, respectively.

1. Introduction

Transparent conductive oxides (TCOs), with a band gap of approximately 3.0 eV, exhibit visible-light transmittance, UV-light absorbance, and good conductivity. Because of these interesting characteristics, TCOs can meet the requirements for transparent, translucent p–n components [1,2,3]. Recently, studies of the application of p-type materials in p–n junctions have reportedly converted n-type TCOs into p-type materials. For instance, novel p–n diodes could be developed by converting n-type ZnO to its p-type form [4,5].

P-type TCO delafossite CuCrO₂ exhibits optoelectronic, catalytic, and magnetic properties, as well as potential for development and applications [6,7]. Because thin-film materials are crucial components of various products, CuCrO₂ thin films were focused on in this study. It is anticipated that the CuCrO₂ thin film possesses optimal physical properties, which is conducive to the development of practical devices. The conductive properties of CuCrO₂ can be improved by the addition of doping elements to increase its hole concentration and reduce its resistance. A majority of studies have reported the doping of CuCrO₂ with Mg²⁺ [8,9,10,11]. On the other hand, in this study, zinc (Zn²⁺) was doped in CuCrO₂. Zinc was used as the doping element because it is a non-toxic, non-precious metallic element. The substitution of Cr³⁺ at lattice positions with Zn²⁺ forms point defects, which generate hole carriers; hole carriers in turn increase conductivity. Previous studies have applied Zn-doping of CuCrO₂ to nanoparticles and thin films [12,13,14]. In contrast to these relevant studies, different processing conditions were employed, that is, a lower phase-transformation temperature, lower partial oxygen pressure, higher passive annealing atmosphere of argon, and more varied quantities of the Zn dopant. This experiment aimed to investigate the effects of zinc doping on the structural, optical, and electrical properties of CuCrO₂. We have confirmed that the zinc dopant in CuCrO₂ thin film can increase the transmittance and carrier concentration and reduce the resistivity. This research demonstrates effective doping and establishes a theory of zinc doping on the structural and optoelectronic characteristics of CuCrO₂.

2. Materials and Methods

The sol–gel method was employed to prepare p-type semiconductor thin films. The substrate was quartz glass. Before modulation, the precursor solutions used were copper acetate (Showa, Saitama, Japan, 98%), chromium acetate (Showa, Saitama, Japan, 98%), zinc acetate (Showa, 99%), triethanolamine (Tedia, Fairfield, OH, USA, 99%), and anhydrous ethanol (Shimakyu, Niigata, Japan, 99.5%). The CuCr_1−xZn_xO₂ films with different Cr/Zn ratios (x = 0, 0.025, 0.065, 0.085) were prepared, and various experimental parameters were used to modulate the proportions for the experiment. The amount of copper was 0.02 mol; the amounts of chromium were 0.02 mol (x = 0), 0.0195 mol (x = 0.025), 0.0187 mol (x = 0.065), and 0.0183 mol (x = 0.085); and the amounts of zinc were 0 mol (x = 0), 0.0005 mol (x = 0.025), 0.0013 mol (x = 0.065), and 0.0017 mol (x = 0.085). Anhydrous ethanol (99.5%) was added, and then the mixture was stirred at a rotational speed of 200 rpm for 2 h at room temperature. Triethanolamine was added, and then the mixture was stirred for 24 h to complete the modification of the precursors. The deposition of thin films was carried out by spin coating at a spinning speed of 1500 rpm for 15 s. After completion, the thin films were subjected to a pre-annealing treatment: the films were dried at 300 °C for 10 min, and the temperature was then increased to 500 °C for 2 h. This step was performed three times to complete the spin coating and pre-annealing treatment. Finally, the film specimens were placed in a tubular atmosphere furnace filled with argon gas (99.99% purity) at a flow rate of 200 sccm. The partial pressure of oxygen in the tube was very low. The temperature was 600 °C for 2 h to carry out controlled-atmosphere argon annealing.

An X-ray diffraction (XRD, D8 Discover, Bruker, Karlsruhe, Germany) measurement was employed to analyze the crystal structures of the CuCr_1−xZn_xO₂ thin films (x = 0, 0.025, 0.065, 0.085). The XRD was operated at 40 KV, 30 mA with Cu Kα radiation (wavelength = 1.54 Å). The cross-sectional microstructure and thickness of films were analyzed by field emission scanning electron microscopy (FESEM, JSM-7800F, JEOL, Tokyo, Japan). Atomic force microscopy (AFM, Dimension Icon, Bruker, Karlsruhe, Germany) was performed to measure the surface morphology and roughness of the film samples. The electron binding energy of the undoped and Zn-doped CuCrO₂ was analyzed using X-ray photoelectron spectroscopy (XPS, K-Alpha, Thermo Scientific, Loughborough, UK). The XPS instrument was operated at a wavelength of 0.83 nm with Al target radiation. Thin-film transmittance measurements were conducted on an ultraviolet–visible spectroscopy system. The Hall effect measurement was employed to analyze the electrical properties of the CuCr_1−xZn_xO₂ films. Hall effect measurement can identify the electrical features of resistivity, mobility, concentration, and Hall coefficient. The measured parameter of the magnetic field was 0.6 T at room temperature.

3. Results and Discussion

Figure 1 shows the XRD results for each film specimen. All of the undoped and Zn-doped CuCrO₂ thin films had delafossite structures with (006), (101), (012), (104), (018), (110), and (0012) diffraction peaks. Zn and Zn-related oxide structures were absent because the solid solution of Zn was in the CuCrO₂ lattice, and the doping concentration did not exceed the solubility. However, traces of CuO and CuCr₂O₄ remained in the CuCr_1−xZn_xO₂ (x = 0.085) thin film, related to the fact that CuO and CuCr₂O₄ are at low-temperature stable phases [15]. The phase transformation of CuCrO₂ is as follows:

CuO + CuCr₂O₄ → 2CuCrO₂ + ½O₂

(1)

The thermodynamics for this phase transformation may be described as follows [16]:

ΔG = ΔG° + RT ln(pO₂)

(2)

ΔG = −RT ln(pO₂)_equil. + RT ln(pO₂)

(3)

ΔG = RT ln(pO₂/(pO₂)_equil.)

(4)

Here, ΔG is the transfer of Gibbs free energy (J/mol), ΔG is the change in the standard Gibbs free energy (J/mol), R is the gas constant, T is the absolute temperature, pO₂ is the partial pressure of oxygen in the flowing gas, and (pO₂)_equil. represents the specific partial pressure of oxygen, with CuO, CuCr₂O₄, and CuCrO₂ in the material at thermodynamic equilibrium. According to the thermodynamic data, the changes in the Gibbs free energy may be explained by the temperature and partial pressure of oxygen [17]. According to theories in materials science, Cr³⁺ can be substituted with Zn²⁺ as the ionic radii of Cr³⁺ and Zn²⁺ are 0.63 and 0.74 Å, respectively (discussed in a later section with the band gap, XPS and Hall effect measurements). Using the total pattern analysis solution to determine the lattice constant of CuCr_1−xZn_xO₂, the a- and c-axis constants increased with the Zn dopant content (

Table 1. Lattice constants of delafossite CuCr_1−xZn_xO₂ thin films (x = 0, 0.025, 0.065 and 0.085 Å).

CuCr1−xZnxO2 Thin Films	a-Axis	c-Axis
x = 0	2.972	17.069
x = 0.025	2.974	17.070
x = 0.065	2.974	17.074
x = 0.085	2.975	17.071

). XRD JCPDS data file No. 89-6744 indicates that the delafossite structure lattice constants of a and c are 2.973 and 17.100 Å, respectively. This difference in dimensions of the ions distorts the lattice and forms a stress field [18]. The greater stress field in CuCr_1−xZn_xO₂ (x = 0.085) caused by the amount of zinc in the solid solution impeded atomic diffusion during CuCrO₂ phase transformation, increasing its required Gibbs free energy. Under the same annealing conditions (600 °C), residual low-temperature structures of CuO and CuCr₂O₄ were observed in the thin film. According to the thermodynamic theory, the CuO and CuCr₂O₄ phases may be removed by increasing the annealing temperature or by decreasing the partial pressure of oxygen (both were not performed in our study). Lowering the Gibbs free energy permits the reaction between CuO and CuCr₂O₄, affording CuCrO₂. A previous related study has used Equations (2)–(4) for its thermodynamic calculations and for the construction of CuCrO₂, CuO, and CuCr₂O₄ phase diagrams describing the relationship between the partial pressure of oxygen and annealing temperature [17]. Several arguments can be formulated by this study and the previous study: 1. The appropriate chemical composition, low partial pressure of oxygen, and high temperature are crucial conditions for the formation of single-phase CuCrO₂; 2. CuO and CuCr₂O₄ react to form CuCrO₂ at a high process temperature [15]; 3. An oxygen partial pressure of less than 10⁻³ atm and a temperature of ≥600 °C enable the formation of a single-phase CuCrO₂ structure.

FESEM cross-sectional images of the thin films revealed irregularly shaped, densely packed microstructures. The formation of these microstructures was dependent on the annealing process. Zn doping did not significantly affect this structural characteristic, as evidenced by the results for undoped and heavily doped CuCr_1−xZn_xO₂ (x = 0 and 0.085; Figure 2a,b). The thicknesses of the CuCr_1−xZn_xO₂ films (x = 0, 0.025, 0.065, 0.085) were in the range of 163 to 169 nm. The minor variation of the film thickness was in agreement with the processing capabilities of sol–gel spin coating and annealing treatment, and the cross-sectional microstructures were similar to those reported in the literature [19,20].

As can be observed by the two- and three-dimensional AFM images in Figure 3a and Figure 4a, the surface of the undoped CuCrO₂ thin film consisted of uniformly dispersed granules of different sizes. The root-mean-square roughness of the thin film surface was 20.6 nm. Figure 3b–d and Figure 4b–d show the two- and three-dimensional AFM surface morphologies of Zn-doped CuCrO₂, which were similar to the surface structures of undoped CuCrO₂. The surface root-mean-square roughness values of the CuCr_1−xZn_xO₂ thin films with x values of 0.025, 0.065, and 0.085 were 19.7, 15.4, and 15.8 nm, respectively. Therefore, the increase in the Zn dopant content resulted in a decrease in the surface root-mean-square roughness. This decrease is related to the stress field resulting from the insertion of zinc into the delafossite lattice. During annealing, the stress field hinders the migration of atoms, dislocations, and grain boundaries during growth, thereby limiting the microstructural growth of surface granules and decreasing the root-mean-square roughness of the surface.

Transmittance measurements (Figure 5) indicated that all of the thin films exhibit translucency at wavelengths longer than around 380 nm. The transmittance spectra of the films were similar, and the transmittance of the undoped CuCrO₂ thin film was slightly lower. The fundamental transmittance edge of the Zn-doped thin films exhibited a blue-shift phenomenon. With increasing zinc content, the transmittance of the CuCr_1−xZn_xO₂ thin films increased. At wavelengths between 500 and 800 nm, the transmittance of the Zn-doped CuCrO₂ thin film increased by approximately 2% to 7%. This increase is related to the lowering of surface roughness, reducing the diffuse reflection of photons; the normal directions of the formations were different while the incident parallel light ray impinged a rough surface. Therefore, a lower surface roughness of the film causes a low diffuse reflection and an increase in the incident light ray (photons), resulting in increased transmittance. At a wavelength of 800 nm, the transmittances of the undoped and Zn-doped CuCr_1−xZn_xO₂ thin films were approximately 68% and 70–75%, respectively.

Initially, the absorption coefficients of all of the CuCrO₂ thin films between wavelengths of 400 and 800 nm (Figure 6) were relatively small, considerably increasing at wavelengths shorter than 400 nm. The thin films mainly absorbed ultraviolet light with short wavelengths; thus, doping the CuCr_1−xZn_xO₂ films (x = 0.065, 0.085) with higher amounts of Zn can increase the absorption of photons with wavelengths of 300 to 350 nm. This trend is related to the low surface roughness and high carrier concentration. Factors affecting the absorption coefficients of CuCrO₂ thin films include (1) the surface roughness, which affects the amount of incident rays (photons) into the thin film; (2) typically, the increase in the dopant content results in different lattice arrangement symmetries, thus changing the wave and path of ray (photons) propagation and resulting in increased scattering (the degree of lattice distortion caused by the dopant content of the thin films affects the optical path of rays inside the thin films); (3) the level of carrier concentration in the material can affect the amount of photons absorbed. The unit volume of film has a high density of carriers; this also results in a high probability and number of absorbed photons. In summary, thin films with high absorption coefficients exhibit characteristics of low roughness, a low degree of diffuse reflection, high incident photons, and a high probability of photons being absorbed.

Tauc’s relation of (αhν)² versus hν was plotted to measure the direct band gap of the material [21]. The band gap of undoped CuCrO₂ film was 3.09 eV (Figure 7a). As shown in Figure 7b–d, the band gaps of the CuCr_1−xZn_xO₂ films with x values of 0.025, 0.065, and 0.085 were 3.09, 3.10, and 3.11 eV, respectively. Therefore, doping the CuCrO₂ thin film with zinc leads to an increased band gap. The trend is similar to the blue-shift phenomenon as observed in the transmittance measurements. As the energy level of Cu 3d¹⁰ in the delafossite oxide structure is close to the O 2p levels, the hybridized orbitals can cause the delocalization of the hole states, conferring mobility to the holes [22]. With the disruption of the interatomic bonds adjacent to the holes by external energy, valence electrons fill the positions of the holes, resulting in hole displacement. In the zinc-doped CuCrO₂ film, Zn²⁺ replaces Cr³⁺; concurrently, Zn 4s forms an acceptor energy level slightly above the high valence band in the band gap structure. Energy at room temperature can excite electronic transitions from the valence band to these acceptor energy levels, causing an increase in hole carrier concentration in the valence band and an increase in the band gap of the CuCrO₂ thin films. The band gap values obtained in this study (3.09–3.11 eV) are similar to those reported for CuCrO₂ subjected to extrinsic doping [13,23,24].

Using the mathematical theory of extrapolation of measurement data, the calculation can be performed in line with the actual measurement results. If the energy gap for CuCr_1−xZn_xO₂ (x = 0.085) is assumed to be unknown, then the known data for CuCr_1−xZn_xO₂ (x = 0, 0.025, 0.065) can be used to calculate the band gap by extrapolation. The calculated band gap of CuCr_1−xZn_xO₂ (x = 0.085) was 3.105 eV. This value is close to the measured energy gap (3.11 eV).

Figure 8a–d shows the results of XPS measurements of the CuCr_1−xZn_xO₂ thin films (x = 0, 0.025, 0.065, 0.085). The photoelectron binding energies of Cu 2p_3/2, Cu 2p_1/2, Cr 2p_3/2, Cr 2p_1/2, and O1s were 932.7, 952.5, 576.8, 586.4 and 530.5 eV, respectively; their corresponding atomic valences were monovalent cation (Cu⁺), trivalent cation (Cr³⁺), and divalent anion (O²⁻), respectively. These values of electron binding energy are close to those reported in the literature [25]. The binding energies of Zn 2p_3/2 were 1021.7 eV, indicating that zinc was a divalent cation (Zn²⁺). Based on the theories of ceramic science, since the CuCr_1−xZn_xO₂ thin films are crystalline delafossite, the anions and cations bond to one another. The increase of carrier concentration by zinc doping observed in the Hall effect measurement suggests that zinc replaces chromium site (see below).

XPS is sensitive to surface contamination [26]. Before conducting the XPS analysis, the surfaces of the CuCr_1−xZn_xO₂ were bombarded with argon ions to remove absorbed oxygen elements and contaminants, in order to acquire the correct compositions. The stoichiometric compositions of CuCr_1−xZn_xO₂ are Cu:Cr_1−xZn_x:O = 25:25:50. For each set of quantitative values of Cu, Cr, Zn, and O in these films, the anion O contents were 57.57–58.56%, whereas the cation Cu and Cr_1−xZn_x elements were less than the stoichiometric ratio inside the films. The Cu contents were 17.66–19.20%. The Cr contents were 22.95–23.78%. The Zn contents were 0.18–0.44%. The contents of the O were twice as high as that of the Cu and Cr_1−xZn_x elements. This is favorable for the generation point defects of hole carriers, such as Cu and Cr vacancies or interstitial oxygen. In this study, the Zn-doping technique was also implemented to enhance hole carrier concentrations. The CuCr_1−xZn_xO₂ (x = 0, 0.025, 0.065) films with different Zn-doping contents displayed the same single-phase structure, as demonstrated by XRD. No impurity phase contributions to the sub-peak were observed in the CuCr_1−xZn_xO₂ film (x = 0, 0.025, 0.065). The CuCr_1−xZn_xO₂ (x = 0.085) film had traces of CuO and CuCr₂O₄, and the impurity phases may contribute to the sub-peaks in the spectra. The Zn element is present in the film, which suggests the effect of the O–Zn bond in a minor peak shift of oxygen. No Cu satellite transition was observed in the diagram.

Table 2. CuCr_1−xZn_xO₂ thin films’ Zn content versus semiconductor characteristics.

Specimen	Zn Content (at %)	Band Gap (eV)	Resistivity (Ωcm)	Carrier Concentration (cm−3)
CuCrO2	–	3.09	34.35	1.09 × 1015
CuCr0.975Zn0.025O2	0.18	3.09	20.16	8.80 × 1015
CuCr0.935Zn0.065O2	0.26	3.10	3.82	1.88 × 1017
CuCr0.915Zn0.085O2	0.44	3.11	3.98	1.67 × 1017

shows the relationship between the Zn contents and optoelectronic properties of the CuCr_1−xZn_xO₂ films. The Hall effect measurements for the electrical properties of CuCr_1−xZn_xO₂ thin films (x = 0, 0.025, 0.065, 0.085) are as follows: for the undoped CuCr_1−xZn_xO₂ (x= 0) thin film, the resistivity was 34.35 Ωcm, and the carrier concentration was 1.09 × 10¹⁵ cm⁻³. The substitution of Cr³⁺ with Zn²⁺ in the solid solution permitted the lattice point defects to generate hole carriers. The effect of the Cr³⁺–Zn²⁺ substitution on the carrier formed in delafossite-CuCrO₂ may be described as follows:

(Cu_Cu)^× + (Cr_Cr)^× + 2(O_O)^× → (Cu_Cu)^× + (Cr_Cr)^× + (Zn_Cr)^’ + 2(O_O)^× + h˙

(5)

Here, (Cu_Cu)^×, (Cr_Cr)^×, and (O_O)^× represent copper, chromium and oxygen in the original lattice sites, respectively; (Zn_Cr)^’ represents that Zn²⁺ replaces Cr³⁺, which then creates a hole (h˙) carrier. Hence, the carrier concentrations in CuCr_1−xZn_xO₂ with x = 0.025, 0.065 and 0.085 after doping of the CuCrO₂ thin films with zinc were 8.80 × 10¹⁵, 1.88 × 10¹⁷ and 1.67 × 10¹⁷ cm⁻³, respectively. Their corresponding resistivity values were 20.16, 3.82 and 3.98 Ωcm. These resistivities are in a lower range of 1.16 × 10² to 8.57 × 10⁴ Ωcm, as reported in our previous study on intrinsic CuCrO₂ films prepared by the sol–gel method [19]. In addition, the electrical properties are similar to those reported for Zn-doped CuCrO₂ (2.12−3.84 Ωcm) [13]. The CuCr_1−xZn_xO₂ (x = 0.065, 0.085) films with lower resistivities were 3.82 and 3.98 Ωcm, respectively, and the carrier mobilities were 8.65 and 9.38 cm²/Vs, respectively. The values are close to conventional p-type semiconductors. General speaking, when compared to n-type semiconductors, p-type materials have lower mobility due to the fact that a hole current can only be evolved when electron holes are filled with valence electrons in the neighborhood, not free electrons. Positive Hall coefficients indicate that the material is a p-type semiconductor. The results thus confirm that the zinc dopant in the CuCrO₂ thin film increases its carrier concentration and reduces its resistivity. On the contrary, if Zn-dopant atoms are at interstitial sites (Zn_i²⁺) or substitute copper sites (Cu⁺), then they contribute to the outer orbital electrons. The combination of an electron with a hole results in a lower hole carrier concentration, thereby increasing the resistivity. This study therefore demonstrated the effective doping of p-type CuCrO₂ with Zn²⁺, which substitutes Cr³⁺.

Previous literature has reported the importance of developing p-type transparent/oxide semiconductors and its applicability to device electronics [27,28]. The work functions of conventional n/p-type TCOs and their problems and solutions have also been discussed comprehensively. We share the same purposes, which is conducive to the practical optoelectronic applications of these semiconductors. Fabrication of such a p–n heterojunction device makes it possible to develop the p-type CuCr_1−xZn_xO₂ applications, such as selective contacts in photovoltaics, and p-channel transparent film transistors. To achieve the available temperature of the mass-production below 400 °C, the high energy generated by the excimer laser is applied to the surface of the CuCrO₂ film, and a thermal energy effect is generated only at a depth of 100 nm on the film surface. Therefore, not too much thermal energy is transferred to the substrate or material underneath. This annealing method is beneficial for future CuCrO₂ development.

4. Conclusions

In this study, the sol–gel method, spin-coating method, and annealing treatment were employed to prepare Zn-doped CuCrO₂ thin films. Zn in the solid solution can exchange with Cr at the lattice sites, which changes the structure and optoelectronic characteristics of the thin film. All of the CuCr_1−xZn_xO₂ thin films (x = 0, 0.025, 0.065) exhibited a single-phase structure; at higher amounts of the Zn dopant in CuCr_1−xZn_xO₂ (x = 0.085), low-temperature phase structures of CuO and CuCr₂O₄ were retained. The film thicknesses ranged from 163 to 169 nm. Doping zinc into the CuCrO₂ thin film gradually reduced the surface roughness from 20.6 to 15.4 nm, thereby increasing the visible-light transmittance and capability of UV-light absorption. The Zn-doped CuCrO₂ thin film exhibited an increased hole concentration from 10¹⁵ to 10¹⁷ cm⁻³ as well as an increase in its band gap from 3.09 to 3.11 eV. For intrinsic and optimal doping of CuCr_1−xZn_xO₂ (x = 0, 0.065), resistivities of the thin films were respectively 34.35 and 3.82 Ωcm. The substitution of Cr³⁺ with Zn²⁺ in the CuCr_1−xZn_xO₂ thin film was identified by the XPS and Hall effect measurements. Doping CuCrO₂ with Zn is therefore suitable for improving its electrical and optical properties. The CuCr_1−xZn_xO₂ film could be a candidate for the p-type semiconductor layer in practical devices.