The crystal structure and optical bandgap of the films were characterized using X-ray diffraction XRD and UV–Vis spectroscopy.. Despite these promising features, CuO still requires furth
OVERVIEW
Limitations and challenges of p-type transition metal oxides
Global energy demand is rising due to population growth, improved living standards, and the widespread use of modern technologies To ensure universal access by 2030, energy must be affordable, reliable, and environmentally friendly, in line with the United Nations Sustainable Development Goal 7 Achieving this requires not only optimizing existing materials but also developing new materials with superior properties and performance to enable cleaner, more efficient energy systems.
Metal oxides play a pivotal role in modern electronic devices, especially as thin films or as matrices that integrate nanostructures such as nanoparticles, nanowires, and nanofibers They offer advantages including high sensitivity, long lifespan, low cost, ease of maintenance, and compact size, making them highly suitable for a wide range of sensors and electronic components These materials can be synthesized using simple techniques such as solution processing and physical deposition methods like sputtering, enabling fabrication and customization of their electrical, physical, and chemical properties for diverse applications.
Transition metal oxides (TMOs) stand out among metal oxides for their distinctive electronic structures The partially filled 3d valence orbitals and unpaired, readily delocalized electrons confer a broad range of tunable electronic, magnetic, and optical properties As a result, TMOs have attracted extensive research interest for energy conversion and storage applications TMOs are typically categorized by their dominant charge carriers into p-type and n-type materials, with typical p-type examples including NiO and Mn3O4.
Co3O4 and CuO are examples of p-type transition metal oxides, whereas commonly studied n-type TMOs include SnO2, ZnO, Fe2O3, and TiO2 Although n-type TMOs have been extensively studied and successfully applied in a wide range of electronic and optoelectronic devices—primarily due to their higher charge-carrier mobility and lower effective electron mass—the development of complementary p-type TMOs remains a key area of ongoing research in materials science and device engineering.
TMOs remains limited in both availability and performance In the field of gas sensor applications, a statistical analysis based on a Web of Knowledge search conducted on July
In 2013, it was revealed that only about 10% of research focused on p-type semiconductors (Figure 1.1) [7] The shortage of p-type semiconductor materials poses a significant obstacle to advancing future electronic and optoelectronic technologies The persistent imbalance between n-type and p-type materials hinders material design for devices such as p–n junction diodes, transistors, and solar cells Consequently, expanding research on p-type transition metal oxides (TMOs) is crucial for the development of metal oxide–based electronic and optoelectronic technologies.
Figure 1.1 Statistical analysis of n-type and p-type oxide semiconductors in gas sensor applications (Web of Knowledge search, July 15, 2013) [7]
CuO stands out among p-type transition metal oxides due to its intrinsic p-type conductivity, narrow bandgap of approximately 1.2–1.9 eV, and favorable physicochemical properties Recent studies also highlight CuO’s potential for high-temperature superconductivity, closely tied to the specific Cu–O atomic arrangement Structurally, CuO adopts a monoclinic lattice rather than the cubic rock-salt structure seen in many oxides In addition, its low cost, abundant natural availability, and environmentally friendly profile further boost its appeal for sustainable and scalable device fabrication Collectively, these attributes position CuO as one of the most promising materials for future technological development.
CuO thin films doped with transition metal Ni
Due to its advantageous properties, CuO has attracted significant attention for applications in electronic and optoelectronic devices, including solar cells [16], photodiodes [17], and field-effect transistors [18]
To realize these applications, various fabrication methods have been developed for producing CuO thin film, such as molecular beam epitaxy [19], chemical vapor deposition
[20], sputtering [21], thermal evaporation [22], spray pyrolysis [23], pulsed laser deposition
[24], successive ionic layer adsorption and reaction (SILAR) [25], and spin-coating [26]
Building on advances in fabrication methods, metal-doping strategies have been employed to further enhance CuO thin films Basith et al show that metal doping can significantly tailor the electronic structure of CuO, leading to substantial changes in its optical, electrical, morphological, and magnetic properties Transition metals are often selected as dopants because of their potential applications in spintronics, ferromagnetic materials, and spin-phonon interaction phenomena Recent studies have investigated dopants such as Mg, Al, Mn, and Co in CuO thin-film systems, revealing dopant‑specific improvements in performance.
Nickel is chosen as a dopant for CuO because of its distinctive attributes The ionic radius of Ni²⁺ (0.69 Å) is very close to that of Cu²⁺ (0.73 Å), enabling Ni to readily substitute Cu in the CuO lattice with minimal distortion Moreover, NiO has a wide bandgap ranging from 3.65 to 4.0 eV, together with low electrical resistivity and high thermal stability, properties that are expected to enhance the electrical and optical characteristics of CuO, particularly by increasing its conductivity.
Ni-doped CuO has been widely studied for enhancing the structural and optical properties of pure CuO Baturay et al prepared Ni-doped CuO thin films using the spin-coating method and confirmed a polycrystalline tenorite structure in all samples by X-ray diffraction The optical bandgap of the films showed a non-monotonic trend, decreasing from 2.03 eV to 1.96 eV at 6% Ni doping, then increasing to 2.22 eV at 10% Ni, with this variation attributed to Ni substituting Cu in the lattice and altering the electronic structure.
6 structure Moreover, Ni doping significantly reduced the carrier concentration, affecting the p-type conductivity of CuO [33]
Ha et al synthesized Ni-doped CuO nanorods on CuO films grown on fluorine-doped tin oxide (FTO) substrates via a modified chemical bath deposition approach Varying Ni concentrations from 0.5% to 20% widened the CuO bandgap from 2.33 eV to 3.46 eV, thereby altering the material’s electronic structure The doped films achieved a maximum photocurrent density of 1.75 mA/cm² at 0.55 V, indicating enhanced light absorption and photoelectrochemical conversion, and demonstrating that Ni doping improves CuO photoelectrochemical performance.
Dolai et al prepared CuO thin films on fused silica substrates by the sol–gel method and doped them with Ni at 0.1, 1.9, and 4.7 at.% Analyses confirmed that the Ni ions are predominantly in the Ni2+ oxidation state within the films The Ni-doped CuO thin films exhibit ferromagnetic behavior at room temperature, with magnetization gradually decreasing as the temperature increases [38].
Ni-doped CuO thin films were deposited on glass and silicon substrates by radio-frequency sputtering, as reported by El Aakib et al The Ni doping modified the structural properties, with all films exhibiting a polycrystalline structure Increasing the Ni content from 0 to 4.5 at.% widened the optical bandgap from 1.62 to 1.76 eV, while all samples retained p-type conductivity regardless of the doping level [39].
Nickel doping markedly influences the photoelectric and magnetic properties of copper oxide (CuO) Yet, the effect of nickel on CuO’s conduction mechanism remains underexplored, with only a limited number of studies addressing this facet Alongside experimental investigations, simulation methods have become powerful tools for analyzing and predicting material behavior from theoretical principles Among these, density functional theory (DFT)–based simulations have attracted substantial attention due to their ability to reveal detailed information about the electronic structure and the fundamental mechanisms that govern material properties.
DFT-based study on pure and doped CuO
Density Functional Theory (DFT), developed by Kohn and Sham [40], is a modern computational framework for studying the electronic structure of many-particle systems in solid-state materials It enables calculation of key quantities such as band structure, density of states (DOS), band gap values, and the optimized geometric structures of materials In semiconductor materials like CuO, DFT is an effective tool for investigating how dopant atoms influence electronic properties and the material’s structure By modeling doped CuO with DFT, researchers can predict the physical and electronic characteristics and guide experimental efforts to synthesize materials with desired properties.
Ekuma et al employed the DFT+U method to investigate the electronic and optical properties of monoclinic CuO, revealing strong hybridization between Cu-d and O-p states near the Fermi level, a band gap of 1.25 eV, and magnetic moments of 0.68 μB on Cu and 0.18 μB on O Comparison with experimental X-ray photoelectron spectroscopy and optical spectra shows good agreement, validating the reliability of their computational model [41].
Recent DFT simulations have investigated CuO's band gap characteristics In a comparative study, Cao et al analyzed the structural and electronic properties of CuO in both monoclinic and cubic phases Their results indicate that the monoclinic phase is the more stable polymorph, with lattice parameters that closely align with experimental data and a bandgap near 1.2 eV, whereas the cubic phase displays a larger bandgap of about 1.7 eV.
CuO's electronic structure is governed by O 2p orbitals that dominate both the valence and conduction bands In terms of optical properties, the monoclinic phase shows strong absorption in the visible region (~2.0 eV), whereas the cubic phase exhibits a blue-shifted absorption spectrum These results underscore the significant influence of crystal structure on CuO's properties.
Additionally, the effect of vacancies in the CuO lattice has been investigated Wang et al conducted experimental work and DFT simulations on copper vacancies in CuO photocathodes for water splitting The study found that copper vacancies increase charge carrier density and improve charge separation and transport, effectively doubling the relevant performance metric.
8 photoelectrochemical efficiency when fabricated under pure oxygen DFT simulations confirmed that these vacancies create intermediate energy levels that aid charge transfer, highlighting their key role in enhancing photocathode performance [43]
Jamal and colleagues investigated the effect of doping CuO with transition metals—Fe, Co, and Ni—on its structural, electronic, and optical properties using the DFT+U approach Their DFT+U calculations reveal that CuO films exhibit p-type conductivity with an indirect bandgap of 1.58 eV, in agreement with experimental observations Doping with Fe reduces the indirect bandgap to 1.51 eV, Co increases it slightly to 1.60 eV, and Ni shows minimal impact on the bandgap.
Density Functional Theory (DFT) has proven to be a powerful tool for investigating the structural, electronic, magnetic, and optical properties of both pure and doped CuO, while integrated theoretical–experimental studies remain limited, especially in revealing how metal dopants alter CuO's electrical conductivity To address this gap, we combine experimental measurements with DFT simulations to systematically assess the impact of nickel (Ni) doping on CuO, aiming to advance understanding and provide theoretical support for future experimental investigations.
Research objectives
This study aims to reveal the formation mechanism of p-type conductivity in CuO, examining both undoped CuO and CuO doped with the transition metal nickel (Ni) It also investigates how different dopant species influence CuO’s conductivity mechanism, a topic that remains limited in both simulation and experimental studies To achieve the overall aim, the study formulates specific objectives.
This study's primary objective is to experimentally investigate how Ni doping influences the crystal structure and optical band gap of CuO thin films prepared by the sol-gel method By varying the Ni content during sol-gel synthesis, we examine structural changes using X-ray diffraction and monitor shifts in the optical band gap with UV–visible spectroscopy, with the aim of understanding dopant-induced modifications that can enhance the performance of CuO in optoelectronic applications such as sensors and photovoltaics.
The second objective is to employ DFT-based simulations to predict the electronic structure and key properties of both undoped and Ni-doped CuO Hubbard U parameter is
9 examined to determine an appropriate U value that aligns the computational model to closely match the experimental data
The third objective is to build CuO structures featuring copper and/or oxygen vacancies across different charge states to identify which vacancy type predominantly drives hole formation and p-type conductivity Additionally, the study investigates how varying nickel doping levels alter CuO’s conduction mechanism, using undoped CuO as the baseline.
To understand how dopant behavior affects CuO, this study extends the analysis to third-row elements with varying numbers of valence electrons It examines how these valence variations influence CuO conductivity and provides predictions to guide future experimental investigations.
METHODOLOGY
Experimental procedure
This study investigates nickel (Ni) doping of copper oxide (CuO) to produce Ni-doped CuO, highlighting the role of transition-metal dopants in modifying CuO properties The precursor synthesis relies on the sol-gel method and uses a defined set of chemical compounds to prepare the precursor solutions.
- Copper (II) acetate monohydrate: Cu(CH3COO)2.H2O
- Nickel(II) acetate tetrahydrate: Ni(CH3COO)2.4H2O
Copper(II) acetate and nickel(II) acetate salts are the primary precursors for synthesizing CuO and NiO, respectively Ethanol serves as the solvent to dissolve the mixed acetates, while MEA, or monoethanolamine, acts as a stabilizer in the reaction.
Figure 2.1 Chemicals used to prepare the precursor solution
Following the procedure outlined in Figure 2.2, copper(II) acetate and nickel acetate were dissolved in 19.395 mL of absolute ethanol, with nickel acetate present at concentrations of 4%, 6%, 8%, and 10% To prevent agglomeration, 0.605 mL of MEA was added, and the beaker was covered with plastic wrap After stirring for 15 minutes, the temperature was raised to 75°C and the solution was gently stirred for 1 hour to ensure complete dissolution Finally, the solution was cooled to room temperature and could be stored in the refrigerator for 24 hours.
Figure 2.2 The precursor synthesis process
The resulting solution was completely homogeneous, with no sign of precipitation, as shown in Figure 2.3
Figure 2.3 Resulting precursor sample post-synthesis
After the post-synthesis process, the precursor was used for thin-film deposition The deposition was carried out using the sol-gel method, as illustrated in Figure 2.4
Figure 2.4 Preparation of the glass substrate and the spin-coating process for thin-film deposition
The 22 mm × 22 mm glass substrate is used for the precursor deposition process First, to ensure the substrate is clean, it is ultrasonically cleaned sequentially in acetone, ethanol, and distilled water to remove contaminants
Next, to enhance adhesion, it is then treated with a 2% hydrofluoric acid solution, generating an H⁺ ion layer on the surface
The precursor solution is applied to the substrate by spin-coating, with 50 μL dispensed onto the surface and spun at 1500 rpm for 40 seconds to form a uniform film The coated substrate is then dried on a hot plate at 100°C for 7 minutes This coating step is repeated three times to achieve the desired film thickness The spin-coating apparatus used in this study is shown in Figure 2.5.
Finally, the coated substrate undergoes a two-step thermal treatment: annealing at 250°C for 30 minutes, followed by a heat treatment at 500°C for 30 minutes to form a crystalline thin film This annealing sequence, conducted with the annealing system used in this study, is shown in Figure 2.6.
The annealed thin films will be characterized using the following techniques to evaluate their properties
X-ray diffraction (XRD) is an analytical technique used to characterize material properties It probes a material's structure with X-ray beams whose wavelengths are comparable to the spacing between atoms As the X-rays interact with the sample, part of the radiation is scattered by the crystal's atomic planes, producing a diffraction pattern By analyzing the resulting XRD pattern, one can determine crystalline phases, estimate crystallite size, and assess the degree of structural order within the material.
The correlation between the X-ray wavelength and the interplanar spacing of the sample's atomic planes is governed by Bragg's Law (illustrated in Figure 2.7): n𝜆 = 2𝑑 ℎ𝑘𝑙 sin𝜃 (2.1) where:
• n is the order of reflection (n =1, 2, 3, …);
• λ is the wavelength of the incident X-ray;
• 𝑑 ℎ𝑘𝑙 is the interplanar spacing of atomic planes with Miller indices (ℎ𝑘𝑙);
• θ is the angle between the incident X-ray beam and the crystallographic plane of the sample
In this study, the crystal structure of thin films was investigated using an X-ray diffractometer (XRD, Bruker, D2 Phaser) with Cu-Kα radiation (wavelength 𝜆 = 1.5406 Å) in the scanning range of 20°– 80°
UV-Vis spectroscopy is a widely used spectroscopic analytical technique for determining the optical properties of materials by measuring their absorption or transmission of light in the ultraviolet (UV) and visible (Vis) regions of the spectrum This method provides essential information on how materials interact with UV and visible light, making it a fundamental tool in chemistry and materials science.
The absorption coefficient (α) in the spectral region of light absorption was determined using the Beer-Lambert law It was calculated from the measured absorbance (A), film thickness (d), and transmittance (T) using the standard relation α = −ln(T)/d, which can also be written as α = (2.303 A)/d when absorbance is expressed on a base‑10 scale This approach links the optical density of the film to the exponential attenuation of transmitted light across its thickness, providing a quantitative measure of its light-absorbing properties across the spectrum.
The optical band gap energies were evaluated through the Tauc plot method, utilizing the following relation:
Where: 𝐸 𝑔 is optical bandgap energy; 𝛼 is the absorption coefficient and h𝜈 is the incident photon energy
In this study, the Shimadzu UV-2450PC dual-beam spectrophotometer was used to investigate the optical absorption of the thin films in the wavelength range of 300 to 800 nm (Figure 2.8)
Figure 2.8 UV-Vis Spectroscopy (Shimadzu 2450)
Simulation method
Computational simulations were performed using the plane-wave pseudopotential method within density functional theory (DFT), a widely used quantum framework for investigating atomic structures, electronic properties, and magnetism in semiconductors This section first provides an overview of DFT and then describes its application to material simulations, with a specific focus on CuO and Ni-doped CuO analyzed in this study.
The electron’s behavior in a quantum system is described by the wave function Ψ, governed by the Schrửdinger equation Solving this equation reveals key electron properties
- such as quantum states and orbital shapes - based on energy conservation For a single, non-relativistic particle in a stationary state, this is expressed by the time-independent Schrửdinger equation (2.1) [45]:
2m∇ 2 + V(r)] Ψ = EΨ (2.1) Where, ℏ is the reduced Planck constant; m = 9.109 × 10 −31 kg, is the mass of the electron;
∇ 2 is is the Laplacian (differential operator);
For simple atomic systems such as hydrogen and hydrogen-like ions (for example, He⁺), the Schrödinger equation can be solved exactly, yielding both the wavefunction and the energy spectrum with relative ease In contrast, multi-electron systems introduce intricate electron-electron interactions and additional degrees of freedom that make the full equation far more complex To manage this challenge, physicists and chemists rely on approximations, with the Born–Oppenheimer approximation being one of the most fundamental This approach leverages the large mass difference between nuclei and electrons to effectively decouple nuclear motion from electronic motion in a molecule, enabling a practical separation of electronic and nuclear dynamics By solving the electronic structure problem at fixed nuclear positions and then treating nuclear motion on the resulting potential energy surface, the Born–Oppenheimer approximation dramatically simplifies calculations and provides a robust framework for understanding molecular spectra, reactivity, and other essential properties.
Under the Born-Oppenheimer approximation, the molecular wave function is expressed as a product of separate nuclear and electronic components This separation allows the electronic structure to be studied independently, with the nuclei treated as relatively stationary, thereby simplifying quantum chemistry calculations by decoupling fast electronic motion from slower nuclear motion.
The time-independent, non-relativistic Schrửdinger equation that describes a system involving multiple electrons interacting with multiple nuclei is given by the following expression equation (2.2) [46]:
𝑈(𝑟 𝑖 , 𝑟 𝑗 ) represents the interaction between an electron located at position 𝑟 𝑖 and another electron at position 𝑟 𝑗 ;
E is the ground state energy of the electrons; Ψ is the electronic wave function for all N electrons
Density Functional Theory (DFT) replaces the complex many-electron wavefunction with the electron density as its fundamental variable, a function of three spatial coordinates regardless of how many electrons are present This effective dimensionality reduction makes DFT a practical and computationally efficient framework for approximating solutions to the Schrödinger equation in complex, many-electron systems Consequently, DFT is widely regarded as one of the most robust and reliable approaches for studying the quantum behavior of matter At its core, the Hohenberg–Kohn theorems establish that a system's ground-state properties are uniquely determined by its electron density, even in the presence of external potentials.
Theorem 1: The external potential 𝑉(𝑟⃗) acting on an electron system is uniquely determined by the electron density 𝜌(𝑟⃗) This implies that no two different external potentials, 𝑉(𝑟⃗) and 𝑉′(𝑟⃗), can yield the same 𝜌(𝑟⃗)
Theorem 2: The electron density corresponding to the system’s exact ground state minimizes the total energy functional In other words, the true ground-state electron density is the one that produces the lowest total energy for the system
The Hohenberg–Kohn theorems state that the ground-state electron density uniquely determines all properties of a quantum system, but they do not provide a practical way to compute that density To make this concept usable, Kohn and Sham introduced a practical method within density functional theory by formulating the problem in terms of non-interacting reference electrons and solving the Kohn–Sham equations to obtain the ground-state density [40].
E[𝜌(𝑟)] = 𝑇[𝜌(𝑟)] + ∫ 𝜌(𝑟)𝑣(𝑟)𝑑𝑟 + 𝐸 𝑒𝑒 (2.3) Where: 𝑇[𝜌(𝑟)] is the kinetic energy;
∫ 𝜌(𝑟)𝑣(𝑟)𝑑𝑟 is the interaction with external potential and electron-nuclei interaction
𝐸 𝑒𝑒 represents the electron-electron interaction energy This interaction can be expressed as:
In equation (2.4), the electron-electron interaction energy is the sum of the classical electrostatic interaction and the exchange-correlation energy Kohn and Sham reintroduced the single-particle wave functions Ψ_i to formulate a set of independent-particle equations (equation (2.5)) for n electrons, where the electron density is given by ρ(r) = ∑_{i=1}^n Ψ_i*(r) Ψ_i(r) The kinetic energy in equation (2.3) is expressed in terms of these Kohn–Sham orbitals as the non-interacting kinetic energy T_s[Ψ], which contributes to the total energy in density functional theory.
2𝑚∑ 〈Ψ 𝑛 𝑖 𝑗 |∇ 2 |Ψ 𝑖 〉 (2.6) Since the wave functions Ψ 𝑖 follow the orthonormality condition, they must satisfy:
The functional of a wave function can be determined as shown in equation (2.8): Ω[Ψ 𝑖 ] = E[𝜌(𝑟)] − ∑ ∑ 𝜀 𝑖 𝑗 𝑖𝑗 ∫ Ψ 𝑖 ∗ (𝑟)Ψ 𝑗 (𝑟)𝑑𝑟 (2.8) Lagrange multipliers 𝜀 𝑖𝑗 are introduced to enforce the orthonormality of the wave functions
By minimizing Ω[Ψ 𝑖 ] with respect to Ψ 𝑖 ∗ (𝑟), the Kohn-Sham equations are derived
In equation (2.9), the effective potential 𝑣 𝑒𝑓𝑓 (𝑟) can be expressed in more detail as:
In equation (2.10), 𝑣 𝑋𝐶 (𝑟) represents the exchange-correlation potential energy, which is expressed as:
To obtain the solution, the complete system must be solved iteratively using the self- consistent field method [49, 50] The steps involved in this method are summarized below:
First, an initial trial 𝜌(𝑟) must be defined
Second, based on 𝜌(𝑟), the Kohn-Sham equations are solved to determine the orbitals Ψ 𝑖 (𝑟)
Third, 𝜌(𝑟) is recalculated from the Kohn-Sham orbitals obtained in step two, using the expression:
Within Kohn–Sham density functional theory, the KS electron density is constructed from the occupied orbitals as ρ_KS(r) = 2 ∑_i |Ψ_i(r)|^2, accounting for spin degeneracy This density is updated iteratively through self-consistent field cycles until convergence is reached, defined by the condition that |ρ_KS(r) − ρ(r)| < cc for all points r, where cc is the convergence criterion Once convergence is achieved, the final total energy of the system is computed using the converged density and orbitals.
VASP (Vienna Ab initio Simulation Package) is among the most widely used tools for performing density functional theory (DFT) simulations As a first-principles computational package, it enables atomistic modeling of materials and detailed analyses of electronic structure and quantum-mechanical dynamics It solves the Kohn–Sham equations within DFT and also supports Roothaan equations under the Hartree–Fock framework, including hybrid functionals that blend DFT and HF components Furthermore, VASP includes Green’s function techniques, such as quasiparticle GW, and the adiabatic-connection fluctuation-dissipation theorem within the random phase approximation, as well as many-body perturbation methods like MP2 In this thesis, all simulations were carried out using VASP [51, 52].
The simulation process employed in this study is illustrated in Figure 2.9, and the steps are described in the following sections
Figure 2.9 Overview of the simulation process and sequential steps
A Cu48O48 supercell in a 2×3×2 cell is used in this study to model copper oxide The spin configuration of the Cu atoms in CuO is set up based on the suggestions of B.X Yang et al [53] and Avishek Maity et al [54], as shown in Figure 2.10 Structural models are visualized using the VESTA software [55].
Figure 2.10 depicts a Cu48O48 monoclinic CuO supercell, highlighting spin polarization of copper atoms with spin-up shown in blue and spin-down in green, while oxygen atoms are represented in red.
Figure 2.11 Ni-doped CuO bulk supercell structure (Cu₄₈O₄₈) Cu, O, and Ni atoms are shown as blue, red, and yellow spheres, respectively
Systems containing copper (Cu) or oxygen (O) vacancies were created by removing the corresponding Cu or O atom from the CuO crystal lattice In the main Ni-doped system, a single Ni atom was introduced by substituting a Cu atom in the Cu48O48 supercell, resulting in a Ni doping concentration of 2.08% The Ni-doped Cu48O48 supercell with one substituted Ni atom is shown in Figure 2.11.
2.2.2.2 Setup of parameters and functionals for DFT simulations
This section outlines the computational setup for the simulations, with first-principles calculations based on density functional theory (DFT) performed using the projector-augmented wave (PAW) method [56], and exchange-correlation effects treated by the generalized gradient approximation (GGA) in the Perdew-Burke-Ernzerhof (PBE) functional [57].
We use a plane-wave basis set with a 450 eV cutoff energy and model the electronic structure with pseudopotentials Brillouin-zone integration is carried out on a 4×4×4 gamma-centered k-point mesh During geometry optimization, ionic positions, the cell volume, and the cell shape are all allowed to relax The electronic self-consistency loop is converged when the total-energy change between successive steps is less than 10^-5 eV Typically, 40–60 ionic relaxation steps are required.
DFT accuracy hinges on the choice of exchange–correlation functional, a key factor in determining how well calculated results align with experimental data To enhance precision, researchers commonly employ a range of functionals—from local and semi-local approaches to hybrid and meta-GGA formulations and newer range-separated methods—each offering different trade-offs between accuracy and computational cost for diverse systems and properties.
RESULTS AND DISCUSSION
Experiment results
The results of the XRD analysis for the synthesized thin films are presented in Figure 3.1 The XRD pattern reveals that pure CuO exhibits prominent diffraction peaks at 35.52° and 38.70°, corresponding planes (002) and (111) of CuO, respectively These peaks are characteristic of the monoclinic phase of CuO and are consistent with previous studies on CuO thin films synthesized via the sol-gel method [65]
Figure 3.1 XRD pattern of CuO and Ni-doped CuO thin films
Figure 3.1 shows that as the Ni doping concentration increases, the intensity of the characteristic diffraction peaks of CuO gradually decreases and significantly drops at the 10% doping level Moreover, no diffraction peaks related to other oxide phases such as Cu₂O or NiO are detected, indicating that the sample remains in a single-phase CuO structure In addition, the characteristic peaks of CuO in the doped samples exhibit a slight shift toward higher 2θ angles, reflecting the substitution of Cu²⁺ ions by Ni²⁺ ions within the CuO crystal lattice These results were also observed in Ni-doped CuO thin films in our previous study [66] The present study confirms that Ni²⁺ ions can effectively replace Cu²⁺
At higher doping concentrations, 28 ions are incorporated into the CuO lattice, a result supported by XRD data This finding provides a solid basis for modeling the substitution of Cu atoms by Ni in a CuO supercell, enabling accurate theoretical calculations of Ni-doped CuO systems.
Figure 3.2 presents the UV-vis absorption spectra of Ni-doped CuO thin films
Figure 3.2 UV-vis measurement of Ni-doped CuO thin films
UV–visible (UV–vis) absorption measurements reveal that the optical bandgap of Ni-doped CuO thin films lies between 2.12 and 2.28 eV, indicating that Ni incorporation modulates the electronic structure of the CuO host The corresponding bandgap values are listed in Table 3.1.
Table 3.1 Band gap value of CuO undoped and Ni-doped CuO thin films
% Doping level Band gap energy (eV)
Results indicate that increasing the concentration of Ni doping leads to a slight increase in the band gap of the thin film Compared to the UV–visible results at lower Ni doping, this trend shows that higher Ni incorporation marginally widens the band gap.
Consistent with our previous work [66], the current study shows that the band gap increases as Ni concentration rises This upward trend is also reported in other studies involving higher Ni doping levels [33], which further supports the reliability and consistency of these findings.
Simulation results
Effective simulation modeling begins with constructing a model that faithfully mirrors experimental data, making accuracy the top priority In this context, selecting the appropriate U value is crucial, as it directly influences the calculated properties of the system By calibrating the U parameter to align with experimental results, researchers achieve more reliable predictions of the system's behavior Ultimately, careful U-value selection and validation against experiments ensure simulations yield meaningful insights.
Geometry optimization was used to determine the lattice parameters of the Cu48O48 supercell in simulations performed both without the Hubbard U term and with several values of U The resulting lattice parameters were then compared with experimental measurements to evaluate the accuracy of the simulations The discrepancy between the computed and experimental lattice parameters was quantified using Equation (3.1).
Error=calculated value−experimental value experimental value × 100% (3.1)
The variation of each lattice parameter (a, b, and c) of Cu 48 O 48 as a function of the
U parameter is illustrated in Figure 3.3 In this study, a wide range of U values (U = no U,
This study investigates the effect of the U parameter in the DFT+U method on lattice parameters a, b, and c As the U value increases, lattice parameter a grows while b and c contract, showing that the U parameter strongly influences lattice geometry All three lattice parameters intersect the experimental values (red line) near U = 6 eV, and the calculation error decreases to nearly zero for U in the 5–7 eV range A U value of 6 eV provides the closest agreement between calculated and experimental lattice parameters [13], although further investigation of U values on other properties remains necessary.
Figure 3.3 CuO lattice parameters: (a) a, (b) b, (c) c, and (d) error, obtained from DFT (noU) and DFT+U simulations with varying U values, compared with experimental data
Continuing our investigation of Cu–O bonding in CuO, we compared the bond lengths obtained from simulations with the experimental data cited in [13] In this analysis, the Cu–O distances were extracted from the simulation results and measured using the VESTA software, enabling a direct assessment of how well the computed structures reproduce the experimental bonding environment.
[55] Experimentally, CuO exhibits two shorter Cu–O bonds (1.95 Å) and two longer ones (1.96 Å) [13] The measured bond lengths from the simulations are summarized in Table 3.2
Table 3.2 Variation of Cu−O bond lengths with different U values, compared to experimental data [13]
According to Table 3.2, omitting the U parameter causes Cu–O bond lengths to deviate substantially from the experimental data, ranging from a minimum of 1.94 Å to a maximum of 1.978 Å Introducing and increasing the U value gradually draws the calculated Cu–O distances closer to the experimental measurements of about 1.95–1.96 Å The results identify a U range of 6–9 eV as yielding the most accurate Cu–O bond lengths in CuO, emphasizing that selecting an appropriate U parameter is essential in DFT+U calculations to realistically reproduce the material's structural properties.
3.2.2 DOS analysis of undoped and Ni-doped CuO
Pristine and doped CuO electronic structures were investigated through density of states (DOS) calculations The DOS analysis was supported by the VASPKIT tool [61], with the Fermi level referenced to 0 eV From the resulting DOS profiles, the band gap values for each CuO system were estimated.
Figure 3.4 presents the density of states (DOS) for the Cu48O48 supercell calculated at several U values Before applying the U parameter, the DOS shows a continuous distribution from the valence-band maximum (VBM) to the conduction-band minimum (CBM), indicating no band gap and metallic behavior in CuO (Figure 3.4a) When U is applied, a band gap appears in the DOS, characterized by a clear separation between the VBM and CBM As U increases, the band gap widens The calculated band gap values corresponding to the different U values are shown in Table 3.3, illustrating how the U parameter tunes the electronic structure of CuO.
7 eV yields results that are consistent with the experimental band gap of CuO (1.2–1.9 eV)
[8, 9] It is therefore essential to select the U value that best matches the experimental data for further analysis a) b) c) d) e) f)
Figure 3.4 DOS of Cu 48 O 48 : (a) without U (U = 0), and with different U values (b) 3 eV, (c) 5 eV, (d) 6 eV, (e) 7 eV, and (f) 9 eV The Fermi level is indicated by black dashed lines
Table 3.3 Bandgap value of Cu 48 O 48 cell with different U values
U value (eV) Bandgap value (eV)
Experimental measurements by Marabelli et al show that the CuO band gap decreases from 1.56 eV at 0 K to 1.35 eV at 300 K To ensure consistency between simulation and experimental conditions and avoid misinterpretation of computational accuracy, it is essential to compare theoretical results with the 0 K value The absorption edge of CuO thin films is highly sensitive to film morphology and defects, so a direct comparison with DFT results (which assume a perfect, defect-free crystal) may not capture experimental characteristics In this context, the experimental data reported by Marabelli et al from bulk samples synthesized via sintering provide a more reliable reference than thin-film data Consequently, when comparing the present computational results with Marabelli et al., a good agreement between theory and experiment can be observed.
Using a Hubbard U of 6 eV, the simulations produce a band gap of 1.65 eV that aligns with experimental measurements The corresponding magnetic moment per Cu atom is 0.63 μB, which is in close agreement with the experimental range of 0.65–0.68 μB reported in the literature.
First-principles calculations of CuO were performed to determine its lattice parameters, band gap, and magnetic moment, with results benchmarked against experimental values In a Cu48O48 supercell using a Hubbard U of 6 eV, the calculated lattice constants, band gap, and magnetic moment show good agreement with experimental data, demonstrating the reliability and accuracy of the simulations This validation provides a solid basis for subsequent computational investigations and further analyses of CuO properties.
Ni, as a transition metal, requires establishing the U value for its 3d orbitals before doping We first evaluated the U value for Ni using pure NiO, which crystallizes in a rocksalt structure with a lattice parameter of 4.19 Å A full exploration of NiO would be computationally demanding, so we adopted a U value of 6.4 eV reported in reference [44], which yields a calculated band gap of 3.58 eV in our calculations, closely matching the experimental value of 3.65 eV [34] The calculated density of states (DOS) of NiO is shown in Figure 3.5.
Figure 3.5 DOS of NiO calculated with U (Ni) = 6.4 eV
After determining the U values for Cu and Ni, we analyzed the Ni-doped CuO system The density of states (DOS) for a single Ni atom, both without and with the U parameter applied, is shown in Figure 3.6, panels a and b.
Figure 3.6 DOS structure of CuO with Ni doped a)without applying U parameter;
(b) with applied U values: U(Cu) = 6 eV and U(Ni) = 6.4 eV
The doped CuO system exhibits a band gap of 1.62 eV, indicating a slight reduction relative to pure CuO, though the change is not significant This result aligns with experimental observations at low doping concentrations reported in previous studies [66].
CuO is known to be a p-type semiconductor Intrinsic defects in CuO, such as V Cu ,
V_O and interstitial oxygen atoms (O_i) are the primary contributors to the material’s p-type conductivity Among defects, copper vacancies (V_Cu) and O_i promote the formation of free holes, thereby enhancing p-type conduction In contrast, V_O, an anion-related defect, acts as a compensating center by introducing electrons that neutralize the holes, reducing p-type conductivity.
First, we examine the density of states (DOS) for copper vacancies (V_Cu) and oxygen vacancies (V_O) in CuO, as shown in Figure 3.7 Compared with the DOS of pristine CuO (Figure 3.4), the introduction of these vacancies creates impurity states within the band gap Specifically, the Cu vacancy (V_Cu) shifts the Fermi level toward the valence band, indicating hole formation and p-type conductivity in CuO.
In contrast, for 𝑉 𝑂 , the Fermi level is located closer to the conduction band, indicating electron accumulation and n-type conductivity a) b)
Figure 3.7 DOS calculation for vacancy cases a) 𝑉 𝐶𝑢 and b) 𝑉 𝑂 in CuO
Subsequently, the formation energies of 𝑉 𝐶𝑢 and 𝑉 𝑂 were calculated according to reactions (3.3) and (3.4):
Here, N denotes the total number of Cu and O atoms in the supercell, which is set to