Electronic and Magnetic Contribution for CuO and CuO Nanofibers Doped with Mn at 3.0%

Abstract

The copper (II) oxide nanofibers (NFs) synthesized with the electrospinning method showed a necklace-like morphology and nanometric size. The use of the XPS (X-ray Photoelectronic Spectroscopy) technique allowed the analysis of the Cu 2p and O 1s orbitals showing a CuxO type stoichiometry (x = 1, 2, 3), in turn, the UPS (Ultraviolet Photoelectronic Spectroscopy) region determined the conduction state associated to the dielectric function. These data are compared with the EELS technique. The NFs have presented a behavior with double magnetic phase associated to the non-stoichiometry and oxygen vacancies, and the non-presence of the AFM phase due to the increase of the vacancies. In addition, their electronic and magnetic structure reveal spin-orbit related changes shown in the Cu 2p spectra. The results showed in the conduction band holes and the Cu 2p and O 1s orbitals.

1. Introduction

The electrospinning technique has been widely used in recent years for the synthesis of nanomaterials, especially micro and nanometer scale fibers for a wide range of applications in areas such as biotechnology, spintronics and electronics. The technique is characterized by being versatile and easy to assemble, allowing the processing of a wide variety of polymers, integrating in recent years ceramics, semiconductors and dielectrics [1]. The technique consists of using polar solutions (PVP or PVA) in which acetates are dissolved, once mixed, the solutions are placed in a syringe connected to a hose and needle, the latter placed at a certain distance to be deposited on an aluminum plate. To this process an electric field is applied between (5 to 20 kV), the process is diversified in terms of uses, time, synthesis process.

The CuO study is a p-type semiconductor with a monoclinic crystal structure, with a $C_{2h}^6$ (C2/c space group. The copper atom is coordinated to four coplanar oxygen atoms located at the corner of a rectangular parallelogram and the oxygen atom is coordinated by four copper atoms located at the corner of a distorted tetrahedron. These chains traverse the [110] and $\left[1\overline{1}0\right]$ directions, respectively [2]. The crystallographic structure is octahedral. The molecule has 4 Cu atoms in the Wickoff position 4c (1/4,1/4,0) and 4 O atoms in the Wickoff position 4e (0, y, 1/4) where y = 0.4184 [3]. The lattice parameters in the unit cell are a = 4.6837 Å, b = 3.4226 Å, c = 5.1288 Å, the angle β = 99.45° between a and c.

In current reports Cu shows three oxidation states Cu¹⁺, Cu²⁺ y Cu³⁺ [4, 5, 6] and excess holes in its valence band structure (VC) which are associated with the presence of Cu³⁺ microtraces [7]. The copper oxide faces V_O due to the synthesis process the crystalline structure undergoes cation exchanges. This originates changes in its chemical coordination octahedral (O_h) or tetrahedral (D_4h) generating a superexchange thus giving the magnetic origin [8, 9, 10, 11].

In recent studies the oxide has been highlighted by various nanostructures (nanosheets, nanoflowers, nanoparticles, nanorods, etc.) [7, 8, 12, 13, 14, 15]. The study of the surface of these nanostructures has shown relevance through X-ray Photoelectron Spectroscopy (XPS) showing a stoichiometric quantification approach and identification of cations/anions. The analysis of CuO has focused on the Cu 2p and O 1s orbitals, due to its physical characteristics (hybridization and high energy edges). In addition, it is a transition metal because it has incomplete d level, therefore, its study is focused on showing its configuration (1s² 2s² 2p² 3s² 3p⁶ 3d⁹ 4s²) and the acceptance of electrons from the O²⁻ ion, this phenomenon allows charge transfer. However, this is not orange blossom, because a chemical coordination O_h and D_4h is governed. These coordination’s show distortions caused by the packing of ions of different sizes generating the Jahn-Teller effect. Other authors have shown that the 3d orbital shows perturbations due to the possibility of displacements due to bond distortion, clarifying that it is not due to the generation of V_O [12, 13, 14, 15, 16]. Some authors have focused on studying the Cu 3d orbital because of its proximity to the VB [17]. Some authors have shown modifications in their forbidden gap (E_g) because of quantum effects forbidden in the continuum. These changes are due to the quantum confinement of the electronic states which are modified by states not yet clarified in the literature because on the one hand the crystal structures are modified due to the nanometric dimensions and their resonance from the VB to the CB [9, 10, 11, 18, 19, 20, 21, 22, 23, 24].

XPS is used to understand the oxidation behavior of metals. The study of CuO NFs has led to the elucidation of the initial excitation processes [5, 6, 24, 25, 26].

The analysis of the Cu 2p_3/2 and Cu 2p_1/2 doublets shows changes in the electrical behavior. In addition, the association of the O 1s orbital, which is quantified to obtain the stoichiometry of the oxide, has been added to this study. The spectroscopists suggest to calculate these by the block method which provides higher accuracy ±4.0% error [27]. On the other hand, the background of the spectra is important to associate chemical species, this quantification is relevant due to the resonance of electrons in the valence band [28].

The 2p photoemission spectra of the transition metals in the first row are distinguished by being highly complex, showing a large and complex background, peak asymmetries and a broad multiplet structure. However, the Cu 2p spectrum is particularly peculiar, even among them. We will discuss these peculiarities and how the associated difficulties can be overcome by using the Shirley - Vegh - Salvi - Castle (SVSC) based fitting methods under the active approach, as well as compare with the Tougaard contribution, in order to know the electronic contribution associated with the chemical species in the CuO [5, 29, 30].

The study for the identification of the chemical state and its quantification by XPS, using the NIST database [31] and Electron Effective Attenuation Length (EAL), generally provide the necessary information for the identification of metal oxides.

Investigations of multiplet structure calculations suggest that transition metals exhibit charge transfer which can complicate the intensity of the spectral multiplet peaks. Analysis of 2p spectra can be complicated to distinguish the ion structure and charge effect of neighboring bonded ions as both the oxidation state and the multiplet splitting can be affected by this perturbation. These phenomena have been observed principally in copper oxide [5, 32, 33].

The multiplet structure in recent years has been analyzed using CTM4XAS (Charge Trasfer Multiplet) software. The software solves quantum data through density functional theory (DFT) and Hartree-Fock function to calculate the spin-orbit entanglement.

DFT (Density Functional Theory) and the Hartree-Fock function to calculate the spin-orbit entanglement. The calculation is used to determine the final energy of the orbital to be treated using three primitives: (a) atomic multiplet, (b) crystal field and (c) charge transfer. Employing the Schrodinger equation, the minimum energy, and e-e interactions and spin-orbit interaction are determined by providing information of the electron location employing the atom symmetry and charge fluctuation in the energy states $\left(3d^n\to 3d^{n+1}L\right)$, considering L as a hole, respectively. The multiplet structure is important in the study of XPS due to the contribution of the VB electrons and the core-hole (generated from the photoelectric effect). Moreover, in transition metals the 3d (L_3,2) edge is the most important due to the unpaired electron, which provides electronic information [30, 31]. Stavitski and de Groot [32], point out that one of the first features of XPS is the absorption of electrons in the BV and the core-hole effect, because of this the L_3,2 edge is important due to the $2p^{6}3d^n\to 2p^{5}3d^{n+1}$ transition, due to this absorption the multiplet effect is affected in the deeper 2p, 3 s or 3p levels [30, 32].

The calculation process using CTM4XAS software employs empirical data: a reduction parameter of the Slater integral, which refers to the final configuration of the calculated electronic structure. The crystal field (10 Dq) would indicate the separation of e_g and t_2g. Δ (charge transfer energy) refers to the average energy for the 3d orbital, Udd (the core hole potential) is the energy transferred to a hole (unoccupied state) and Upd (Hubbard) is a potential related to the charge fluctuation in the 3d orbital. In addition, the Lorentzian and Gaussian values taken from some database are fixed [32].

Furthermore, the study of the multiplet structure of CuO has not been generated in recent years unlike the undersigned of this manuscript [26, 33], therefore, authors Okada and Kotani to extend and determine the data obtained [25]. These authors studied three Cu oxides by approximation calculations using a Hamiltonian proposing the energy of the free ion (Cu) and the gap (electronic repulsion due to its neighboring O 2p atom) in the BV, binding the energy level Cu 2p⁶ 3d⁹ 4s² and its final state 3d⁸ L (where L denotes the gap and the relationship with O 2p). This spectrum has been discussed by several authors due to the importance of the Columbian interaction, spin-orbit entanglement and the screening shielding effect [25].

The EELS technique evaluates the energy loss of inelastic scattered electrons passing through the sample in the transmission electron microscope (TEM). The valence electron energy loss study (VEELS) focuses on the lower energy loss spectrum, typically <50 eV, this region shows the valence and conduction band characteristics due to electron beam interaction. These are appreciated using scanning transmission electron Microscopy (STEM) Park and Yang [34]. Egerton mentions that the use of STEM provides nanometer resolution spectra, thus allowing surface and crystal structure information to be obtained [35]. The author Herrera-Pérez et al. presents the VEELS technique as a approach to obtain Eg characteristics, electronic and electrical properties through the cole-cole diagram (the graph is obtained from the Im vs. Re function), from the Single Scattering Distribution (SSD) [36]. This analysis is important to determine the resolution of the Zero Loss Peak (ZLP) from the method proposed by Park and Yang [34]. In addition to the above, obtaining the modern spectra and analysis proposed by Eljarrat et al. from the Kramers-Kronig analysis (KKA) and the elimination of the relativistic effects (Cerenkov effect) [37]. Herrera-Pérez et al. shows in their optoelectronic studies the elimination of this effect, as well as convenient methods to calculate the complex dielectric function (CDF) of the materials [36]. In recent years, the electronic study has been complemented by joint density of states (JDOS) convolution showing the electrons present in the conduction band [38].

On the other hand, the theoretical calculations obtained by means of density functional theory (DFT) show similarities to the experimental ones by EELS and XPS, techniques that are exposed in this research, but how to approximate it, for years it has been tried to solve the Schrodinger equation, Born-Oppenheimer in the early 30’s proposed an equation to model multielectronic systems (a many-body wave solution) with N variables. The first-principles or ab initio method is based solely on the laws of physics, without reference to empirical parameters. Kohn-Sham to give solution to this uncertainty treats the set of N particles as electron density, broadens the potential energy with a potential that has been neglected to obtain the non-interacting parts (VXC Exchange-correlation potential) in the Hamiltonian (the Hamiltonian supplements ψ in the Schrödinger equation; the latter has no physical meaning); relative to the term shows that the motions of the electrons are correlated with each other. Software (CASTEP -Cambridge Serial Total Energy Package-, Wien2K, VASP) have implemented for such theory two approximations: Local Density Approximation (LDA) and Generalized Gradient Approximation (GGA), which perform estimates considering variables for Eg’s energy solution, density of states (DOS). However, the calculations have shown estimates that do not agree with those measured experimentally, especially for the d and f orbitals. In the last two decades Hubbard - U effects implemented in DFT have been shown, described as DFT + U in strongly correlated systems that approximate the d and f systems to the expected ones. What does the Hubbar value in the approximation mean, it is a value that serves to describe the electron-electron interaction without shielding (e.g., in the CTM4XAS software it uses the Slater integral to consider such interaction to estimate the type of chemical bond), this interaction is important to determine the magnetic moment and E_g.

The aim of this work is to determine by modern fits to XPS and EELS spectroscopies the Cu 2p, O 1s electronic states and the interband transition states. Furthermore, to calculate the stoichiometry of the A-NFs and the reference (Sigma Aldrich). The EELS section will show a comparative study of CuO and CuO doped with 3.0% Mn to know the changes associated with Mn compared to that calculated with CASTEP 7.0 software. Finally, the magnetic study will show the change of magnetization with respect to temperature and the association of the magnetic field with respect to the shape of the NFs.

2. Experimental procedure

For the synthesis of polymeric fibers by the electrospinning method, the following reagents were used:

1. PVA (High purity, 130,000 Da, Sigma Aldrich)

2. Tri-distilled water (1.1 μohms/cm, J. T. Baker)

3. Copper Acetate II (99.6% of purity, Sigma Aldrich)

The 8.0% PVA polymer solution was made by weighing 8.0 g of PVA in tri-distilled water at 100°C with constant agitation at 100 rpm for 24 hours.

For the polymeric solution of copper acetate II, 1.0 g of it was weighed and added over 2 ml of tri-distilled water at 50°C and left stirring for 2 hours. Finally, 10 g of PVA was added and the whole solution was stirred for 24 hours.

Finally, the polymer solution was placed in a 10 ml syringe connected to a hose for dosing by means of a number 22 needle. The parameters used in the pump were 0.3 ml/hr. and the distance from the needle to the collection plate was 20 cm.

2.1 Characterizations and computational methods

2.1.1 Scanning electron microscopy (SEM)

The sample used was a piece of aluminum with an area of 1 cm², taken from the collector sheet, and used as an electrode for electrospinning. The collected images were taken from the SEM Hitachi SU3500 equipment used to verify that the material was obtained in the form of fibers.

2.1.2 TGA-DSC analysis

The TGA-DSC analysis aimed to show the chemical reactions with their weight distribution in relation to the heat of the reaction (entropy).

A simultaneous thermal thermogravimetric difference test (TGA-DSC) was performed using the SDT-TGA Q600 equipment of TA instruments, through the air and a temperature ramp of 10°C/min, with a range of 25–1000°C, placing 0.25 mg of the synthesized material in a sample holder of the equipment, to determine the calcination temperature of the study compound.

2.1.3 Structure characterization and XRD analysis

Structure characterization was performed by X-ray diffraction (XRD) using an X’Pert Pro diffractometer equipped with an X’Celerator detector. Diffraction patterns were taken in an interval 2θ = 30°–65° using a step size of 0.05 s⁻¹ with a Cu – Kα radiation source λ = 1.5418 Å.

From the lattice parameters obtained from the XRD characterization, these spectra were used for Rietveld refinement using a Pseudo-Voight peak shape (least squares method) in Fullprof suite software version 2022 [39]. A CIF structure was obtained from the refinement, which was used to determine the 2-D and 3-D electron density using VESTA software (free version provide by VESTA) [40]. Especially this analysis took into account the thermal factors of the crystal structure of Arbrik et al. to optimize the calculated versus experimental spectrum using the above-mentioned software [3].

2.1.4 High resolution-transmission electron microscopy (HR-TEM)

The synthesized samples were dispersed at 10 mgL⁻¹ isopropanol for 20 min, then deposited on a 3 mm diameter nickel grid with carbon membrane.

The BF-HR-TEM micrographs were obtained in the JEOL-JEM-2200FS equipment operated at 200 kV. In addition, JImage software was used to determine the average particle size distribution by entering the micrographs and calibrating each one. In this way a histogram was constructed using a Log Normal function [26].

2.1.5 X-Ray photoelectronic spectroscopy (XPS)

The XPS study was obtained using the following parameters: scan energy 10 eV, resolution 0.1 eV, dwell time 200 msec, 40 scanning scans per spectrum and a 90° angle. The monochromatic source used was an Al Kα ($\mathit{h\nu }=1486.6\mathrm{eV}$).

The nanofibers were deposited on a graphite tape, adhered to the sample holder of the XPS equipment, introduced into the pre-chamber at a vacuum pressure of 10⁻⁶ Torr, and finally into the analysis chamber at a pressure of 10⁻¹⁰ Torr.

The peak analysis of the spectra was calculated using AAnalyzer® software (DEMO version) by means of Gaussian and Lorentzian functions (block method, which consists of assigning individual values to each peak, that is, it is not cumulative [27]), these areas obtained were used to calculate the stoichiometry of the oxide. In addition, the convolution considered the Shirley model type SVSC (Shirley-Vegh-Salvi-Castle) suggested for high resolution spectra (in this case Cu 2p), also, for the case of O 1s a Slope line was applied. The data were plotted in Origin Pro 8.6 software [26].

To calculate the chemical composition of NFs, a spherical model was used (model suggested by Shard et al. [41] for 1D materials) that considers the electron attenuation effect (λ), in relation to this Bravo-Sánchez et al. [42] and German-Cabrera et al. [29], relate the λ factor for 2D materials (the difference of this method to the one mentioned above is that it does not consider the shape) as the effect of electron interactions with the atoms of the material, thus affecting the binding energy (E_b) [43, 44].

2.1.6 CTM4XAS

The free version of CTM4XAS software was used to elucidate the multiplet structure (possible energy states calculated by the Schrodinger equation) for the Cu 2p spectrum. The calculation procedure consisted of choosing the Cu element with valence 2+, then the XPS 2p option. By choosing XPS, the charge transfer flag was automatically activated. In addition, the stacks options (to plot the multiplet structure) were selected. To accurately determine the theoretical spectrum, a tetrahedral symmetry was chosen as suggested by Okada et al. [25], the spectrum obtained was calibrated at 933.60 eV to compare the experimental (A-NFs) and reference spectra (Sigma Aldrich) [43].

2.1.7 Electron energy loss spectroscopy (EELS)

The study of NFs by EELS spectroscopy was carried out on a Gatan spectrometer (Tridiem 866 ERS), coupled to a Titan G2 60–300 transmission electron microscope. A Wien monochromator was used to correct the electron beam, the operating conditions were carried out at 200 kV, the collection angle was 17 mrad (corresponding to a parallel beam α = 0 mrad), in diffraction mode at 0.05 eV/Channel, a pitch of 5.0 nm and 2.0 s per pixel.

Through the valence electron energy loss spectroscopy (VEELS) method, and using a scanning transmission electron microscope (STEM), the spectra of A-NFs and B-NFs (CuO, doped with 3.0% Mn) were fitted to the center of the zero loss peak (energy calibration) using the asymmetric Pearson VII function to calculate the experimental resolution, this calculation will be carried out through Origin Pro 8.6 software [44].

On the other hand, the data obtained were processed in the DigitalMicrograph software where the measurement noise in the spectrum was subtracted. Also, the dielectric function and the loss function were obtained mainly by means of the Kramers-Kronig analysis. To remove the relativistic effect, they were treated through a difference method working with angles of: β1 = 17.3 mrad and β2 = 9.1 mrad, that is, two spectra taken with different angles were subtracted. This method is used to determine the bandgap [36].

On the other hand, the thickness value of the samples was calculated using the Log-ratio method (absolute, with relative thickness), taken from the author Egerton [35], from the parameters of: electron beam energy E₀ (kV), the effective collection angle (β), and the effective atomic number Z_eff. In addition, take into account t = λ/0.75, where λ is an attenuation factor of the electron on the sample (statistical value, this value is taken from a graph proposed by the author Egerton [35]. To obtain the single scattering distribution (SSD), it was analyzed by means of the Fourier function using the following equation:

$$S\left(E\right)=\frac{I_{0}t}{\pi a_0{m}_0{v}^2}\mathit{Im}\left(\frac{-1}{\epsilon \left(E\right)}\right)ln\left(1+{\left(\frac{\beta }{{\theta }_E}\right)}^2\right)$$E1

where $I_0$ is the intensity of the zero-loss beam, t is the thickness of the sample, $v$ is the incident electron velocity, $\beta$ is semi-angle of collection, $a_0$ is the Bohr radius, $m_0$ is the rest mass of the electron and ${\theta }_E$ is the characteristic scattering angle for energy loss [34, 45].

To obtain the electron loss function (ELF) was from the SSD employing Kramers-Kroning [35] given by:

$$1-\frac{1}{n^2}=\frac{2}{\pi }\int \mathit{Im}\left[\frac{-1}{\epsilon \left(E\right)}\right]\frac{\mathit{dE}}{E}=\frac{2}{\pi }\int \left[\mathit{\text{scale}}\_\mathit{\text{Factor}}\times \mathit{SSD}\right]\frac{\mathit{dE}}{E}$$E2

The scale factor was determined using the refractive index of CuO (2.3), the same as that used for the doped material. Then the real function of the dielectric function, $Re\left[1/\mathrm{\epsilon }\left(\mathrm{E}\right)\right]$, experimentally obtained from the loss function, $Im\left[-1/\mathrm{\epsilon }\left(\mathrm{E}\right)\right]$, using the Kramers-Kronig transformation.

$$\mathit{Re}\left[\frac{1}{\epsilon \left(E\right)}\right]=1-\frac{2}{\pi }P{\int }_0^{\propto }\mathit{Im}\left[\frac{-1}{\epsilon \left(E\right)}\right]\frac{E^{´}{\mathit{dE}}^{´}}{E{´}^2-E^2}$$E3

where P denotes the Cauchy integral part (it is a weighting to correct for any imbalance in the sample data). Finally, the complex dielectric function $\left(\epsilon ={\epsilon }_1+i{\epsilon }_2\right)$, ${\mathrm{\epsilon }}_1$ and ${\mathrm{\epsilon }}_2$, was obtained from the relation $Re\left[1/\mathrm{\epsilon }\left(\mathrm{E}\right)\right]$ and $Im\left[-1/\mathrm{\epsilon }\left(\mathrm{E}\right)\right]$ [34].

To distinguish the inter-band transitions from the imaginary function, according to the equation [35, 36]:

$$J_{\mathit{CV}}=J_{\mathit{cv}1}+i{J}_{\mathit{cv}2}=\frac{m_o^2}{e^2{\mathrm{\hslash }}^2}\frac{E^2}{8{\pi }^2}i\left({\epsilon }_2\left(E\right)+i{\epsilon }_1\left(E\right)\right)$$E4

where $m_0$ is the electron mass, $e$ is the Charge and $E$ is the energy, $J_{\mathit{CV}}=Re\left[J_{\mathit{CV}}\right]$ is the joint density function of states, these provide the inter-band transitions obtained from the imaginary part of the dielectric function.

As a comparative method to the complex dielectric function and the joint state function, these were calculated using Materials Studio 7.0 software, through the CASTEP (Cambridge Serial Total Energy Package) program using 10 cores on the Xeon10 server of the CIMAV cluster, Chihuahua, Mexico. This was possible using a P1 crystal structure, and atomic positions of Cu (0.25, 0.25, 0) and O (0, 0.4185, 0.25). The structure was optimized and energized using the GGA (Generalized Gradient Approximation) method and the PBE (Perdew Burke Ernzerhof) functional. The convergence was place at $2\bullet {10}^{-5}\mathit{eV}/\mathrm{a}\mathit{\text{tomo}}$, the force on the atom 0.01 eV/Å, the stress on the atom less than 0.02 GPa, and the maximum atomic displacement no more than $5\bullet {10}^{-4}\AA$. The electron exchange of the correlation energy was treated in the GGA framework using PBE, as well as a U potential (7.5 eV) as a correction method in the forbidden band, the energy cutoff of the plane wave basis set was chosen at 500 eV. Directions over the Brillouin zone (BZ) were fixed by the Monkhorst-Pack method with a grid (k-points) of 8 × 8 × 8. In addition, the convergence criterion for the total energy was set with a self-consistent field (SCF) tolerance at $2\bullet {10}^{-6}\mathit{eV}/\mathrm{a}\mathit{\text{tomo}}$. For the calculation as an antiferromagnetic structure without fixed spin orientation. The DOS (density of states) values were taken from the calculations suggested by the author Absike et al. [46].

3. Results and discussion

3.1 SEM analysis

The polymeric fibers obtained by the electrospinning technique were analyzed through SEM, in the secondary electron mode, at magnifications of 5000×. The images are shown in Figure 1a, by means of JImage software calculated the average sizes of the fibers (panel b), resulting to be 47.24 nm.

Figure 1.

Polymeric fibers obtained by electrospinning method, CuO (panel a), and average diameter quantified by JImage software shown in panel b.

3.2 TGA-DSC analysis

The graphs shown in Figure 2 present the result of the simultaneous TGA-DSC test. The same figure illustrates the synthesized and calcined material (panels a and b).

Figure 2.

Panel (a) illustrates an example of the finished polymer solution for use in electrospinning. Panel (b) shows the electrospun polymeric fibers and panel (c) shows the calcined polymeric fibers. Finally, panel (d) illustrates the TGA-DSC analysis showing the decomposition of the precursors of the polymer solution.

This technique was used in the thermal analysis to find the calcination temperature, which after analyzing panel d illustrating the different chemical reactions of decomposition in the polymeric fibers until CuO was obtained, was found after 650°C. Temperatures from 40.5–650°C show the decomposition of the organic and inorganic groups that compose the polymeric fibers such as: alcohol, water, chemically bound water, acetate, and PVA, mainly (see panel d). The analysis was complemented with a DSC that exhibits two types of reactions for the research materials: positive enthalpy (exothermic), related to the release of thermal energy, manifested by the peak at 230 J/g (180°C), and a negative enthalpy (endothermic), related to energy adsorption −25 J/g, at 650°C. The final concentration shows thermal stability for pure CuO, corroborating that, from 650°C, there are no energy and mass changes, respectively.

On the other hand, the authors Kroger et al. [47], Jeong et al. [18], and Piñon-Espitia et al. [26] indicate that the negative enthalpy energy is related to the formation of V_O (oxygen vacancies) from an uncontrolled atmosphere.

Based on the results shown in Figure 2d, in an uncontrolled atmosphere, the defects associated with V_O are corroborated by TGA due to a negative enthalpy.

3.3 DRX analysis

Figure 3 shows the XRD obtained from the NFs, indicating that the predominant phase is tenorite (panel a), which was corroborated with PDF 80–1268 (panel b). Relation to the fits by the Rietveld method show agreement with the experimental result (Y_obs-Y_calc), which is shown in the same figure.

Figure 3.

Experimental XRD data are shown (panel a), which are compared with the PDF (power diffraction file) data shown in panel b.

In Table 1, the variations related to the Rietveld refinement settings for the XRD spectra are presented. Their variants allow elucidation of their crystal structure with respect to the average crystallite size, lattice parameters, and changes in the O⁻² anion in the crystal structure.

Structure	a (Å)	b (Å)	c (Å)	V (Å^3)	χ^2	Y (O^2−) (Å)
Sigma Aldrich	4.6837	3.4226	5.1288	81.199	—
Asbrik [3]	4.68370	3.42260	5.12880	81.079	—	0.41840
A-NFs	4.6858	3.4230	5.1298	82.282	1.44	0.41600

Table 1.

The data obtained from the Rietveld refinement for the NFs under study compared with two references (sigma Aldrich and S. Asbrik) are shown.

The parameters a, b, and c of the structure are compared with those of the reference sample of Asbrik [3], showing consistency with the calculated ones. In addition, the Y parameter shows changes due to the presence of oxygen vacancies [40, 48].

3.3.1 Electron density in the tenorite structure

According to the XRD results explained above, the structure obtained from the refinement is shown in Figure 4. The 3-D structures exhibit the (110) plane. In panel a shows the comparison of the electron density for the A-NFs and the references (Asbrik [3], Cao et al. [49]), indicating charge transfer from oxygen to Cu²⁺. In relation to this point, this phenomenon is characteristic of MOs as a consequence of thermal treatment, nanometric dimensions, and thermodynamic changes (the oxide formation energy change generally presents negative peaks associated with endothermic absorption) [16, 32, 40, 48, 50, 51]. In addition, the energies exhibited in panel a suggest a covalent bond, increasing due to energy transfer this suggests an increase in the forbidden E_g and a greater number of defects [40, 48].

Figure 4.

In panel (a) the comparison of the electron density of S. Asbrik [3] and A-NFs is observed, in panel b and c the 3-D crystal structures of CuO in the (110) plane are shown, respectively. Finally, in panels (d) and (e) the 2-D electron density is plotted for the reference and experimental structure.

3.4 TEM analysis

According to the XRD analysis, morphological changes were studied compared to the Sigma Aldrich reference sample.

The bright field (BF) analysis by HR-TEM shows in Figure 4b the formation of fibers in the form of a collar, which shows the characteristic planes of the tenorite phase (diffraction pattern attached to the image) compared to the Sigma Aldrich reference (panel b). In addition, an analysis was performed concerning the defects shown in panel c (it was amplified in panel d) and utilizing the edge marked with the yellow arrow in panel c. An EELS analysis (high energy between 400 and 700 eV) was performed to identify oxygen, this data corroborates the TGA-DSC analysis explained in Section 3.2.

The distribution and elemental composition of the NFs, using the EDS of the Hitashi 7700 TEM. Panel a show the pure CuO NFs showing the Cu (yellow) and O (red) components, furthermore, this can be corroborated with the signals obtained from the EDS analysis (see panel b). The distribution of the elements in the mapping NFs showed homogeneity, in a 1:1 ratio (see reference [33]).

3.5 XPS and CTM4XAS analysis

The XPS results are displayed in Figure 5, these plots show the spectral differences and spin-orbit coupling energy for the A-NFs and the Reference (Sigma Aldrich). The spectra in panels a and b show the contributions of Cu¹⁺, Cu²⁺, and Cu³⁺ cations in the main Cu 2p_3/2 and Cu 2p_1/2 peaks. The presence of Cu1+ and Cu2+ cations are bound to oxygen (panel c and d), this can be corroborated by the NBD diffraction patterns in Refs [48, 52]. These spectra show a spin-orbit splitting $\overrightarrow{\mathit{J}}=\overrightarrow{\mathit{L}}+\overrightarrow{\mathit{S}}$ at the peaks located at: 19.97 and 19.91 eV, respectively. Furthermore, it can be corroborated that in the left panel of Figure 3, the thin films reported by Pauly et al., up on oxidation of the orbitals also present a splitting attributed to the thermal treatment of the material [53].

Figure 5.

An analysis of the A-NFs is shown regarding the corroboration of the tenorite phase shown in panel b, compared with the sigma Aldrich reference (panel a), these show the electron diffraction pattern which shows the characteristic planes (−111) and (111). In panel c and d, a surface analysis of the A-NFs due to the defects shown in the edges is exhibited, panel e shows by EELS (high energy edges) the presence of oxygen.

Three settings indicating three valence states for Cu (1+, 2+ and 3+) have been proposed in the main branches, reported by the authors: Piñon-Espitia, et al., Ochoa-Torres, et al., Pauly, et al., Sarkar, et al., [5, 26, 50, 53]. These authors have also discussed the presence of mixed oxides by the chemical reaction Cu₂O.

On the other hand, these peaks are related to phenomena such as: paramagnetism and electrical conductivity, due to the presence of the Cu²⁺ cation associated to the CuO [54, 55]. In Figure 5 The contributions of the chemical species present in the peaks (Cu¹⁺, Cu²⁺, Cu³⁺) given by the areas (see Table 2) indicate that the greatest contribution to the conductivity is due to the cations of the A-NFs, which coincides with their geometry that includes a greater area-volume ratio compared to the Reference (Sigma Aldrich).

Sample	Cu^1+ 2p (eV)	O 1 s Cu^1+ (eV)	Cu^2+ 2p (eV)	O 1 s Cu^2+ (eV)	Cu^3+ 2p (eV)	Vacancies (eV)
A-NFs	19104.6	4543.56	16063.7	5865.18	15,228	869.60
Reference	2357.15	7766.1	5944.72	10721.00	6214.70	3169.50

Table 2.

Analysis of areas for the adjustments made.

In the case of Cu-bonded oxygen, the evolution in Table 2 of sample A and the commercial sample is presented. There is a relationship between the areas between the cation and the anion (oxygen). In the particular case of oxygen, the reactions obtained for this work have been shown in Section 3.2, as well as their derivatives (e.g., enthalpy and entropy). Sample A showed a higher amount of Cu¹⁺ as well as a tendency to decrease the amount of the other cations, so according to the author Chusei et al. [56] there is a tendency to higher reactivity because the surface oxygen area increases according to the Cu¹⁺ peak, meaning that it increases. While the amount of oxygen vacancy is like the Reference.

On the other hand, as mentioned in the experimental part, the Shirley of the spectra was analyzed by modifying the block method and the SVSC function proposed by Herrera-Gómez et al. [28], adding the Tougaard function to the analysis. The calculated values for the A-NFs and the Reference were: 0.24 to 0.04 eV⁻¹ and 3864 to 2000 eV⁻¹, respectively, associated with recombination of the 2p band with the conduction band (see Figure 5), which is corroborative with Figure 3 of author Herrera-Gómez et al. [28].

3.5.1 Analysis of O 1s spectra and Vo contributions

Figure 5c and d show the results of the O 1s spectrum and its homogeneity in the samples by means of the Background Slope calculation. In addition, this analysis is supported by the TGA-DSC technique (see Section 3.2) that other authors have related to XPS, especially on the generation of vacancies in the tenorite structure [55, 57].

The vacancies have been located in the O 1s spectra between 531 and 533 eV and are related to the enthalpy due to the negative reaction suggested by the authors [48, 55] in the formation of the oxide, being the TGA-DSC technique an alternative for the detection of vacancies (they have been identified as S2 in panels c and d). The importance of the generation of V₀ is how they influence its physical properties such as magnetic, in addition to the electrical properties of the material indicated by the literature [41, 58, 59, 60, 61].

Table 3 shows the percentage variations of Cu¹⁺ and Cu²⁺ cations related to the O2- ion, which are compared with the Reference (Sigma Aldrich). The A-NFs have a Cu¹⁺/O²⁻ ratio of 0.92 and 0.72% for the Reference. The percentage corresponding to the Cu²⁺/O²⁻ ratio is 1.08% in the A-NFs, and 1.38% for the Reference. These non-stoichiometric ratios are associated with structural and surface defects and other phenomena related to the nanometer dimensions of the material (e.g., plasmonic effects) [62, 63].

Sample	Pico Energía (eV)	[Cu^1+]%	[Cu^2+]%	VO
A-NFs	529.36	0.92
	531.35		1.08
	532.83			0.096
	536.06
Reference	529.83	0.72
	530.95		1.38
	531.87			0.653
	532.65
	533.66
	535.81

Table 3.

Percentage of Cu¹⁺ and Cu²⁺ spices associated with oxygen (O₁(S₀)-O₂(S₁)), respectively.

In addition, the quantification of V_O, as well as the ratio of V_O with respect to the network oxygen.

These percentage ratios show that the A-NFs presented an increase of Cu¹⁺ in the crystalline structure, compared with the Reference which has a lower amount and a higher amount of Cu²⁺; in relation to this, the research samples have slightly an excess of Cu²⁺, this shows that the presence of both cations are important for their electrical, magnetic and optical properties, mainly [12, 54, 61, 64].

3.5.2 Multiplete structure analysis

In this section we show evidence of changes in chemical coordination, binding energy, hybridization, and charge transfer (∆), related to the nanometer dimensions of the synthesized material compared to the Reference (Sigma Aldrich). Other related changes are: scramming shielding effect, electronic changes in the main 2p_3/2 peak (Cu 2p – L_3,2 edge electrons) and satellites (holes).

Figure 6 shows the final states of the NFs compared to the commercial and calculated material (CTM4XAS). Image a show the set of study spectra compared to the calculated one (spectral shape and multiplet structure). In addition, the panels show the changes mainly in the satellites discussed in the previous section [32, 65, 66].

Figure 6.

Panels a and b show the Cu 2p spectra of the A-NFs and the reference (sigma Aldrich) and O 1 s (panels c and d), respectively.

In Figure 6a the ∆ according to the obtaining of the materials is exposed, which is seen to be higher for the A-NFs and much lower for the commercial one. According to Okada and Kotani [25] the ∆ effects are due to the electron donors (in this case the two oxygens located in the c plane), data that agrees with the percentages presented in Table 3, which would show that the A-NFs show higher charge transfer due to the proximity of the oxygen ion to Cu and generating a higher scramming shielding.

The authors showed that the satellites are due to the holes in the oxide, which is observed in panel a that physically exemplifies the effect of the exposed edge related to electrons and holes in the BV. Finally, the leading-edge shifts and hole variation are mainly due to the hybridization of the Cu 2p and O 1s edges, and their shift to higher binding energy, as is the case in the experimental sample exhibiting the screening shielding effect.

About the satellites, it can be appreciated that according to the ions and their coordination, variations were obtained with those shown in Figure 5 (image a and b), indicating for the experimental ones a D_4h type structure (tetrahedral) and the Reference type O_h (octahedral). On the other hand, the screaming shielding effect is directly related to electronic effects related to electron and hole resonance.

3.5.3 Chemical composition analysis based on a spherical model

The chemical composition of NFs has been calculated through the spherical model. This proposal is derived from recent reports by Shard et al., and Cardona, [61, 67]. In Table 4 shows that A-NFs do not retain stoichiometry.

Sample	Species chemicals		Pico; BE; (eV)	Pico; BE; (eV)	Compounds	XCuO	I1/I2	A%	SF
A-NFs	Cu^+1	O^−2	933.36	529.3	Cu^+1 - O^−2	0.33–0.54	0.35–0.64	0.98	Cu1.02O1.15
	Cu^+2	O^−2	934.91	531.33	Cu^+2 - O^−2	0.33–0.57	1.05–0.64	0.98	Cu2.06O2.03
Reference	Cu^+1	O^−2	933.33	529.76	Cu^+1 - O^−2	0.35–0.48	0.35–0.51	0.98	Cu0.97O1.02
	Cu^+2	O^−2	936.06	530.66	Cu^+2 - O^−2	0.81–0.54	1.66–1.36	0.98	Cu2.43O2.01

Table 4.

Primary XPS signal data for Cu 2p and O 1s spectra used to calculate the chemical composition in the nanopowders.

Note: the data obtained in SF, were obtained from calculating λ = 1.25 (Electron attenuation factor in the LEA sample), this factor was used to calculate the chemical composition by means of the MML model which was proposed as a bulk model, without layers. X_CuO is the Cation compositional factor, I₁/I₂ are the intensities of these, A% is the atomic percentage and SF is the final stoichiometric factor.

According to the geometries estimated with the spherical model of Shard [41, 61] the compositions were for A-NFs: Cu_1.02O_1.16, Cu_2.22O_2.12, and for the Reference (Sigma Aldrich): Cu_0.97O_1.02, Cu_2.43O_2.01, with a 4.0% error, respectively. The chemical composition was related to the E_k and to the photoelectric effect applied to the sample (cross section).

3.6 EELS analysis

For this analysis, two samples, A-NFs and B-NFs (CuO doped with 3.0% Mn) were taken for comparison. To corroborate the distribution of Mn and its percentage can be observed in the reference [33].

Figure 7 shows the inelastic contributions of the electrons in the A-NFs and B-NFs, respectively. The resolution of the peaks exhibited in panels a-b was 0.97 eV, as an asymmetric result of the deconvolution. In addition, the loss function (blue line) is shown in both panels, in panel a a a plasmon is exhibited at 22 eV, in panel b 2 plasmon peaks are exhibited, 10.0, 22 eV, while the ∼32.0 eV peak represents a M_2,3 Mn [35]. On the other hand, the analysis using the Pearson VII function allowed us to clean the spectra obtained, and thus obtain the single scattering distribution (SSD) (blue line).

Figure 7.

Multiplet structure for the A-NFs and the reference (sigma Aldrich), compared to the calculated one (panel a). In addition, in panels b and c it differentiates the O_h and D_4h structures.

In addition, the relative thickness calculated by the Log-ratio method yielded values of: 35.55 (A-NFs) and 38.72 nm (B-NFs) suitable for analysis.

3.6.1 Surface and bulk plasmon analysis

Egerton reports plasmons for all materials between 10–30 eV [35], while Charles, 1996 by means of an analysis of metals and EELS analysis determined an energy of 0 to 10 eV for these metals [35]. In the study materials, plasmons were found to be associated at 22.52 and 22.15 eV, for materials A and B, respectively. Material B showed a signal of 10 eV which is consistent with being doped.

Figure 7 (Panels a,b) show the collective excitations of electrons (plasmons) where they interact in low regions in the BC and BV, known as bulk plasmon and surface plasmon, respectively. In this regard, the authors Herrera-Pérez et al., Egerton, [35, 36] propose to evaluate the energy of the free electron plasmon as:

$${\mathrm{E}}_{\mathrm{p}}=28.82\mathrm{eV}\ast {\left(\frac{\mathrm{z\rho }}{\mathrm{A}}\right)}^{1/2}$$E5

Where z is the number of electrons per molecule, ρ is the density of A-NFs = 6.520 g/cm³ and B-NFs = 6.473 g/cm³ (these values were calculated by Fullprof suite software), respectively and A is the molecular weight 79.57 and 134.483 g/mol, respectively.

According to the calculations proposed by the authors Herrera-Pérez et al., Egerton [35, 36] the plasmon energy for A-NFs and C-NFs turn out to be: E_p = 16.49 eV and E_p = 17.88 eV, respectively. The change between these values is due to the interband transition below the plasmon peak [35].

These results suggest that the plasmons found are transiting in low regions between the valence and conduction bands. According to Egerton, this would indicate that they are semiconducting materials [35].

3.6.2 Comparative analysis of the dielectric function and the edges obtained from XPS-survey

Figure 8 shows the dielectric function, the peaks associated with the interband transitions obtained in EELS compared to the XPS survey spectra, for the A-NFs and C-NFs. This comparison indicates according to the author Meyer and co-workers [63], the data obtained showed a similarity.

Figure 8.

Valence energy loss spectra (VEELS) of A-NFs (a) and B-NFs (b), extracted from the zero peak and dispersion spectrum (SSD).

${\mathrm{\epsilon }}_1$ is the real part of the dielectric function which presents in this case a negative transition indicator showing a response below the value of the semiconductive behavior [37]. In addition, the authors Johann Toudert and Rosalía Serna [64], performed a study of the effects of collective oscillations (free charges) in Ag and Au oxides to know the optical and interband and intraband transition effects by means of the complex dielectric function ($\mathrm{\epsilon }={\mathrm{\epsilon }}_1+\mathrm{i}{\mathrm{\epsilon }}_2$) [64]. The discussion focuses on the plasmonic effect, because on electromagnetic radiation, especially the infrared and UV-Vis zones. The results of those authors show similarity to those presented for the real function (see panels a and b), ${\mathrm{\epsilon }}_1$ being negative suggesting that it has the metallic character [68, 69, 70, 71]. However, the relationship does not comply with what is exposed by these authors because ${\mathrm{\epsilon }}_1>1$ so such harmonic dispersion is in the UV-Vis region, but it complies with the imaginary relationship due to its growth [66]. In addition, the authors mention that this effect is for alternative plasmonic materials. In studies with doped materials, they improve the response to such an effect [72]. The author Manuel Cardona, 1986, finds that the negative value for ε₁ is related to excitons (phonon-electron) [67].

On the other hand, the peaks of the imaginary function (${\mathbf{\epsilon }}_{\mathbf{2}}$) were associated with the Cu 2p, 3d, and O 2p transition interbands in the conduction band, and the peaks A, D, E, H, and I are associated with the holes. In Figure 26 of reference Meyer et al. [63] present the imaginary part of the dielectric function pointing out five transition states, associated with the Cu 2p, 3d, O 2p, and holes bands.

Moreover, the bands obtained by XPS in the UPS (Ultraviolet Photoelectronic Spectroscopy) region (see panels e and f) show similarity to those reported in CuO [41, 46, 67]. Furthermore, Wang and co-workers, [73], performed the theoretical and experimental study of CuO by XPS- UPS, finding that the obtained peaks corresponded to the Cu 2p, 3d and O 2p orbitals [46, 67]. On the other hand, panel f shows the oxidative evolution of metallic Cu to the study oxides. In the study materials the leading edge presented a Cu-2p and Cu-3d splitting, while the O-2p shows an overlap with these orbitals (A-NFs and Reference), while for doping there is slightly a peak associated with it (hybridization), which agrees with the model of Wang et al. [73]. In addition, in the case of metallic Cu, the peak corresponding to oxygen was not present.

Table 5 shows the values of the electron transition shifts in the conduction band. When Mn is added to the NFs, it is observed that the main peak shows a smaller peak width (0.25 eV) and a smaller number of holes. It is also seen that the number of peaks for the A-NFs is like the theoretical one. The Cu 3d transitions for the study materials with respect to O 2p, mainly, are exposed. In Figure 8c and d a shift from 1.33 to 3.69 eV (transition Cu 2p → Cu 3d, see DOS in Table 5) in the doping with respect to pure is shown. For the case of Cu 2p with the O 2p of materials A and C there was a shift between these peaks due to doping [74, 75, 76, 77, 78, 79, 80, 81, 82].

Sample	Transition	Assignation	EELS (eV)	UPS (eV)	DOS (eV); CASTEP GGA-PBE	DOS (eV); Wang, et al. [73]	Energía de transición ${\mathbf{\epsilon }}_{\mathbf{2}}\left(\mathbf{\omega }\right)$ Meyer et al. [63]
A-NFs	B	O 2p → Cu 2p	7.75	6.60	1.94	1.98	2.71
	C	O 2p → Cu 3d	10.73	12.88	2.56	2.96	2.89
	J	O 2p	30	25.46	2.47		5.24
B-NFs	A	O 2p → Cu 2p	3.08	1.97
	B	O 2p → Cu 3d	3.43	4.60
	F	O 2p	30.14	21.49

Table 5.

Proposed inter-band transitions and compared to references ^A[73], ^B[63].

In addition, the DOS-CASTEP label shows the energy obtained from the model by this software and compares it with the authors mentioned in the Table.

Figure 9 shows the joint density of states calculated for the interband transitions. In the A-NFs (panel a), three states corresponding to those calculated by CASTEP (panel c) were observed, which coincide with those presented by the authors [41, 46, 67, 73], furthermore, the O 2p shift coincides with that of Meyer et al. [63]. In the case of the doped material, the Cu 3d transition increased the energy (see panel b) and decreased the number of holes (see Figure 9d). Panel c shows the holes obtained, similar to those reported by the author cited above [63]. On the other hand, the DOS model of CuO is calculated by Meyer et al. [63]. Absike, et al., [46], by calculating the TDOS and PDOS show agreement with the results (see panels c and d) [46]. Panel d shows the agreement with the one calculated by CASTEP through the GGA-PBE (Generalized Gradient Approximation - Perdew Burke Ernzerhof) functional and the one reported by Meyer and collaborators, [69].

Figure 9.

Panels a and b indicate the dielectric functions $\widehat{\mathit{\epsilon }}\left(\mathit{E}\right)={\mathit{\epsilon }}_{\mathrm{1}}\left(\mathit{E}\right)+\mathit{i}{\mathit{\epsilon }}_{\mathrm{2}}\left(\mathit{E}\right)$, for the A-NFs and C-NFs, respectively. In panels c and d, the interband transitions are exhibited. Panel e shows the survey spectra of: Cu metallic, A-NFs, B-NFs, and the reference. Panel f shows the Cu 3d and O 2p bands for the materials in the UPS region.

3.6.3 Bandgap determination

Figure 10 compares the normalized ELF (Energy Loss Function) (black dotted line) calculated using the KKA (Kramers-Kronig Analysis) logarithm in the 0–10 eV range with the VEELS spectra using the difference method (red line). This method allows us to calculate the onset energy associated with the (optical) bandgap in the ELF. The Eg energy was determined using the method of Rafferty and Brown [38] as the intercept of a polynomial fit $A+{\left(E-E_g\right)}^n$ (A is the background level), the plots suggest an indirect Eg of 2.03 and 2.85 eV, respectively. The Eg reported by Meyer et al. is 1.0 eV [69], which we attribute to working with bulk material. The calculated bandgap and that obtained in our materials suggest that the bandgap values may be associated with electronic resonance problems at the Cu 3d and Cu 2p levels due to its nanometer character, which generates changes in the orbitals, for this reason, the curve was not entirely possible to smooth it [26, 28].

Figure 10.

Panels a and b show the joint density of states for A-NFs and B-NFs, respectively. Panel c exhibits the DOS for CuO calculated by CASTEP. In addition, the BC region with the interband transitions is exposed. Panel d shows the interband region obtained by Mayer and the gaps obtained from this model [69].

3.7 Magnetic analysis

Figure 11 shows the paramagnetic behavior shown by the magnetic susceptibility of A-NFs and C-NFs (panel b). Mariammal et al. [83] performed a study with CuO and Mn-doped CuO in nanoflakes, which suggests the coupling of Mn on CuO creates spin decompensation at the surface (superparamagnetism) and spin-glass behavior that the antiferromagnetic phase disappears. According to Zhao et al., possible couplings can appear when CuO is doped with Mn, are: Cu-O-Cu-O-Mn-O-Cu-O-Cu-O-Cu, Mn-O-Mn, and Mn-O-Cu-O-Mn attributed to paramagnetic or spin-glass, antiferromagnetic, and ferromagnetic behaviors, respectively [70]. On the other hand, at d levels of Mn a strengthening in the magnetic aspect was expected, however, the antiferromagnetic effect did not occur, as in the case of NPs mentioned by Zhao et al. [70], above.

Figure 11.

In panels a-b, the comparisons of the VEELS spectra with deconvoluted ZLP and the fit of these with the difference method are shown. In addition, the blue dashed line indicates the polynomial fit for E_g. a) Corresponds to the A-NFs and b) are the B-NFs.

Figure 12a and b show the magnetization versus temperature (ZFC-FC) results of A-NFs and B-NFs pointing to temperatures of 58 and 60 K, corresponding to ferromagnetic behavior. This behavior is also observed in the results of NPs in CuO and Mn-doped CuO (nanowires) of the author Han et al. [12] shown in panel c. Similar to this work, it shows a blocking T greater than 80 K which is attributed to the different grain size and higher doping than this work.

Figure 12.

Magnetic susceptibility from 2 K to 300 K, a) for CuO NFs, pure and doped with Mn (2.5%).

Recalling that bulk CuO exhibits two antiferromagnetic phases at 213 and 230 K it is remarkable that the same in Figure 12a and b no such phases are observed in both cases due to spin decompensation at the surface of the nanostructures [84]. The absence of these phases also in the work of Narsinga Rao et al., [71] in nanoparticles shown in the image d.

Figure 13a and b show the magnetic behaviors using hysteresis loops for A-NFs and B-NFs at 300 K, which were found to be superparamagnetic and paramagnetic, respectively. The superparamagnetic behavior is comparable with that obtained for NPs by Narsinga Rao et al. [71].

Figure 13.

Magnetic analysis panel (a and (b for A-NFs and panel c) and d) for B-NFs at 300 K.

In Table 6, the data of magnetic saturation variations and coercivity for the synthesized materials and CuO NPs of authors Borzi et al. [85] and Narsinga Rao et al. [71] at room T are shown. It is observed that the coercivity is higher in A-NFs than those reported by those authors. The difference is attributed to the nanometer dimensions between the nanostructures handled and the difference in the geometries between the NFs and the NPs of these mentioned authors. On the other hand, C-NFs present higher remanence with respect to A-NFs, but also lower coercivity.

Sample	Hc (KOe)	Mr (emu/g)
A-NFs	0.25	2.80 × 10^−5
B-NFs	0.006	0.002
Borzi, et al. [85]	0.200	0.007
Narsinga Rao et al. [71]	0.200	0.306

Table 6.

Coercivity and Remanence for A and B materials, compared with Borzi et al. and Narsinga Rao et al. [71, 85].

4. Conclusions

The synthesis of A-NFs by electrospinning showed a monoclinic phase corresponding to tenorite. These NFs show multivalent states (Cu¹⁺, Cu²⁺ and Cu³⁺). Their lattice parameters calculated by the Rietveld method suggest lattice distortion. This was corroborated by the electron density in the crystal structure with charge transfer from O²⁻ to Cu²⁺ appreciated as peanut shape which is characteristic in semiconducting materials [1].

The XPS study showed the identification of the cations/anions in the Cu 2p and O 1s orbitals, corroborated with the authors [1, 2, 3, 4, 5, 6, 7], respectively. The distances of the main branches (Cu 2p_3/2 and Cu 2p_1/2) were: 19.97 and 19.91 eV, respectively, because of the Jahn-Teller effect [6, 86, 87]. The increase in the distance of the main branches is a consequence of paramagnetism in CuO. The spectra were used to calculate the stoichiometry using the geometrical topofactor (sphere) method showing non-stoichiometric oxides: Cu_1.02O_1.15 and Cu_2.06O_2.03, and the reference: Cu_0.97O_1.02 and Cu_2.43O_2.01. The CTM4XAS software corroborated the cationic states in the Cu 2p orbital through the multiplet structure. In addition, the calculation employed a higher charge transfer than previous reports. Moreover, the multiplet structure of A-NFs suggests a D_4h structure presents higher number of holes (this chemical coordination is suggested as the one with the lowest energy according to the experimental spectral shape). On the other hand, the O 1s spectra allowed us to calculate the oxygen vacancies generated in these, found in the same samples at 533.17, 532.22 eV. These energies agree with previous reports.

The difference method in VEELS spectra allowed to calculate the onset energy associated with Eg (optical) using the energy loss function (EFL). In addition, this method suggests the elimination of the relativistic effect of the electrons (10⁶) from the spectra [13, 14]. The polynomial fit for the spectra presented an indirect Eg, in agreement with that calculated by CASTEP agrees and differs with previous reports.

The comparative XPS-EELS study for A-NFs and B-NFs (3.0% Mn-doped CuO) in the UV-VIS region (0–50 eV) through the dielectric function showed similarity in spectral shape, deconvolution of the EELS spectra determined the interband states in the CB for Cu 3d, Cu 2s and O 2p orbitals. Furthermore, the comparative study of the A-NFs and B-NFs suggests changes in the interband transitions in agreement with those calculated, the E_g by CASTEP and VEELS obtained were: 1.27, 2.03, 2.85 eV. Meyer et al. [63, 69] shows.

According to the magnetization study, it was observed that sample A and B Ferromagnetism and Paramagnetic (mixed phase) behavior, which is attributed to the change obtained in the Cu 2p multiplet peaks of XPS (change in electric current arises due to changes in $\overrightarrow{\mathit{J}}=\overrightarrow{\mathit{L}}+\overrightarrow{\mathit{S}}$). These results are compared to those of reference [16], finding a mixed phase. This is due to the spin decompensation at the 3d⁹ orbital, which is attributed to its nanometer size and doping and B a paramagnetic behavior, which is attributed to its nanometer size and doping. In addition, the antiferromagnetic phase that normally appears in a CuO (II) bulk material at 232 K, in these samples is not present. This is attributable to the nanometer size, particle geometry, oxygen vacancies, charge decompensation (Cu³⁺) and occurs in other nanostructures [16, 70, 88].

Acknowledgments

The author M. Piñón-Espitia thanks the Ph.D. CONACYT scholarship grant No. 467043. We would like to make two important mentions for the completion of this work: to Dr. Francesca Péiro and Dr. Luis Gerardo Silva for the support received in the EELS and XPS studies. As well as to the national laboratory of nanotechnology CIMAV, S.C. Chihuahua.

Conflict of interest

The authors have no conflicts to disclose.

Abbreviations

NFs	nanofibers
A-NFs	CuO Nanofibers
B-NFs	CuO doped with Mn at 3.0% nanofibers
NPs	Nanoparticles
SEM	scanning electron microscopy
HR-TEM	high-resolution transmission electron microscopy
XRD	X-ray diffraction
TGA-DSC	thermogravimetric analysis-Differential scanning calorimetry
XPS	X-ray photoelectronic spectroscopy
UPS	ultraviolet photoelectronic spectroscopy
CTM4XAS	charge transfer multiplet
UV–Vis	ultraviolet visible
EELS	electron energy loss spectroscopy
VEELS	valence electron energy loss spectroscopy
GGA-PBE	generalized gradient approximation- Perdew Burke Ernzerhof
CASTEP	Cambridge serial total energy package
DFT	density functional theory
DOS	density of state
AFM	antiferromagnetic
FM	ferromagnetic