Structural Defect Impact on Changing Optical Response and Raising Unpredicted Ferromagnetic Behaviour in (111) Preferentially Oriented Nanocrystalline NiO Films

Abstract

NiO thin films deposed on a glass substrate, “NiO/glass”, are successfully prepared using a spray pyrolysis technique (SPT) at 460 °C and characterized via X-ray diffraction (XRD), scanning electron microscopy (SEM), energy dispersive X-ray, Atomic force microscopy (AFM), spectroscopic ellipsometry (SE), Photoluminescence (PL) and diverse electric and magnetic studies. The structural investigation shows that the synthesized films crystallized in a cubic structure with (111) preferential orientation. The NiO layers exhibit a uniform grain of regular sizes with aggregates randomly distributed across their surface. The optical properties of the NiO thin films evidenced a normal optical dispersion as well as good transparency of the NiO films. An unpredicted ferromagnetic aspect was raised due to the high oxygen presence in the synthetized material. A high thermal dependency of the conductivity, as well as a semiconductor behavior of the grown NiO material, is also demonstrated.

1. Introduction

Metal oxides (MOx) semiconductors as inexpensive materials owned fascinating physical properties and excellent chemical stability [1,2,3]. Such characteristics increase their demand in solar cells [4,5], gas sensors [6,7,8], optical windows [9], and acousto-optic devices applications [10,11,12]. They usually exist in n-type transparent conductive oxide (TCO) film. Nevertheless, few of them exist in p-type semiconductors despite their challenging performance [13,14,15,16]. In fact, p-type TCOs are known for their wideband energy bandgap (Eg) varying between 3.6–4.0 eV [17,18,19,20]. NiO is one of the p-type MOx materials crystallized in cubic structure ensuring a good anti-ferromagnetic property [21,22,23,24,25,26] and eminent optical properties [27,28]. Nevertheless, it has not been widely investigated in the industry compared to other MOx semiconductors, such as ZnO and CuO.

In the last decade, NiO thin film synthesis attracted research interest due to its large application in transparent conductive films [29], electrochromic devices [30], organic light-emitting diodes and anodes [30], and chemical sensors as functional layer material [31]. NiO films have been widely synthesized through various physical and chemical routes including plasma-enhanced chemical vapor deposition [32,33,34,35], reactive spraying [36,37,38], pulsed laser deposition [39,40], sol-gel [41], and SPT [42,43,44]. The latest remains one of the most chemical facile routes utilized to prepare NiO films. It is largely applied to prepare films of different components [45]. SPT allows the deposition of films of various thicknesses, multi-layered films, and porous films across a wide surface area [46,47]. NiO sprayed film synthesis starts from a fine droplet token from a solution comprising the chosen substance that is sprayed on the substrate surface after being preheated. This technique guarantees a good crystallization of the films without resorting to any heat treatment cycles using a low-cost raw material and excellent adhesion. Various defects and vacancies accompanied the synthesized films that play a major role in exhibiting new properties [48]. For instance, nickel vacancy affects the conduction and its type. On other hand, the magnetic properties of MOx are influenced by the growth defects [49,50,51].

Several NiO films of uniform thickness have been fabricated via SPT [52]. However, the optimization of the right growth conditions of the NiO films remains a challenge for this technique, where the proper flow rate, substrate temperature, and droplet size vary from one case to another in order to achieve the desired target [53]. Previous works have examined particular growth conditions on the prepared NiO thin films. For instance, Sajilal, et al. [54] succeed in preparing NiO thin layers at different thicknesses using nickel acetate as a precursor at T = 15 K. Nandhini J., et al. [55] also synthesized NiO via an aerosol technique through nickel nitrate, where they reveal its crystallization into a monoclinic structure. Since the properties depend on the structure, S. Viswestwaran, et al. [56] Studied nickel oxide using the electronic paramagnetic resonance technique and deduced the Lande factor. Recently, K.O. Ukoba, et al. [57] similarly reviewed generally the change in the growth conditions′ effect on the fabricated NiO films. Despite the existing work, the optimal synthesizing parameter is still required where no precise flow rate, substrate temperature, and droplet size could be precisely admitted for growing NiO films. Indeed, each synthesized film via SPT requires particular growth conditions to reach the desired research and technology goal. To the best of our knowledge, the optical study of NiO spray thin film from ellipsometry does not investigate so far.

In this context, we report on the current research work on NiO layer synthesis via the spray pyrolysis technique. We represent a broad study of structural, morphological, optical, electric, and magnetic properties of a selected NiO thin film. Our approach differs from those used in earlier works. We have determined the optimal growth conditions such as the concentration of the solution, the constant flow rate, and the deposition time (i.e., thickness), to grow films with high crystallinity. We reported, for the first time, the optical study of NiO spray thin film from ellipsometry. The NiO lattice dependency on the effect of the intrinsic defects has been investigated. Various elements raise defects in the NiO films like chlorine and oxygen. Indeed, we find that the nickel chloride used during the synthesis process leads generally to a high density of defects due to chlorine. This is in addition to the located stress and dislocations derived from the oxygen constructing the NiO material. These defects emerged as an important unpredicted ferromagnetic aspect. We also demonstrate a high thermal dependency of the conductivity as well as a semiconductor behavior of the grown NiO material. We establish a normal optical dispersion and excellent transparency of the NiO films. Such results constitute great finds since they have been archived via a low-cost technique.

2. Experimental Details

2.1. Samples Preparation

Nickel chloride hexahydrate (NiCl₂,6H₂O) purchased from Sigma Aldrich with purity ≥ 99% was dissolved in deionized water. A 0.01 M of the obtained mixture was agitated with a magnetic stirrer at room temperature until the complete disappearance of the nickel chloride salt. Meantime, chloride acid has been added to ensure a good dissolving of nickel hydroxide formed during NiCl₂ dissolution. The substrate, i.e., glass was cleaned ultrasonically via ethanol, Acetic acid, and deionized water and then naturally dried in air. We used a constant flow rate of 4 mL/min through a 0.5 mm diameter nozzle during the spray process at 460 °C. Nitrogen has been used as a carrier gas at a pressure equal to 0.35 bar. We fixed the plane distance in-between the nozzle-substrate to 27 cm [58]. Then, the samples were cooled naturally in the air. We may resume the growth process as below:

• − The formation mechanism of nickel oxide:

• − The residual nickel hydroxide obtained via spray process at 460 °C through an acid-base reaction:

• − Nickel hydroxide transformed into oxide NiO:

2.2. Samples Characterization

X-ray diffractometer (Analytical X Pert PROMPD) with monochromatized copper-source (λ = 1.54056 Å) scanned the samples from 20 to 90°. The morphological properties have been investigated initially using an SEM. We operate with a JEOL-JSM 5400 model SEM supplied by an EDX spectroscopy for elemental microanalysis. Further morphological investigations have been carried out via AFM through VEECO digital instrument 3A apparatus in tapping mode. The optical properties have been characterized with a GES5 SOPRA ellipsometer equipped with a rotating polarizer in the wavelength range between 250–900 nm (4.97–1.38 eV) for an incidence angle of 75° and a xenon arc lamp light source. The existing optical fiber cut the light beam from the light source output to the polarizer input. Then, the polarizer converts random polarization of the light state into a linearly polarized light that illuminates and reflects on the sample surface. The polarized state of the reflected light is determined by combining the rotating analyzer and the photonic detector. The optical calculations have been accomplished with Winelli_II software. A VSM has been addressed to investigate the magnetic properties. Such an instrument ensures rapid measurements and neutralization from parasitic drifts, where it exemplified signals eliminated by synchronous detection by measuring the induced flow in a winding by periodic displacement of the sample. This clarifies VSM characterization where the displacement of the sample at the center of a measuring coil with a constant amplitude vibrates vertically (z-axis) at a fixed frequency of 200 Hz. The electric conductivity of NiO films has been examined via a four-point probe in the temperature range [150–450 °C].

3. Results and Discussions

3.1. Structural Characterizations

3.1.1. Structural Parameters Determination

XRD measurements showed a typical pattern of NiO sprayed thin films (see, Figure 1). We located the well-defined NiO peaks (111), (002), (022), and (222). These latest orientations correspond to the cubic structure NiO according to the JCPDS card 040-835 [59,60]. Indeed, the NiO film grown here is highly oriented following the (111) direction. Therefore, the NiO films grown here have a cubic structure along with the (111) orientation [59,60].

The interplanar spacing “d_hkl” and lattice parameter “a” were determined using the correlation between the Bragg formula and the characteristic relation of the cubic phase [35,61,62]. We reported in

Table 1. Interplanar distance “d_hkl” and lattice parameter “a” of NiO sprayed thin films.

	d111(Å)	d200(Å)	d022(Å)	d222(Å)	a(Å)
NiO	2.4224	2.0925	1.4813	1.2171	4.1850

the calculated values of interplanar spacing and lattice parameters.

3.1.2. Defect Impact on the Change of the Lattice Parameter

The lattice parameter “a” shows a slight difference compared to that known for NiO powder [61]. This could be assigned to the local defect in the prepared NiO films. This confirms the changes in the X-ray diffraction peaks width obtained here. Indeed, the structural defects could be seen in terms of crystallite size as micro-strain and dislocation density. These are determined via the following relation [62,63,64]:

$$D=\frac{k.\lambda }{{\beta }_{\frac{1}{2}}\mathit{cos}(\theta )}$$

(4)

$$\epsilon =\frac{{\beta }_{\frac{1}{2}}}{4tan(\theta )}$$

(5)

$$\delta =\frac{1}{D^2}$$

(6)

where k = 0.90 is the Scherrer constant and β_1/2 the corrected half-width of the peak.

Table 2. Values of crystallite size D, dislocation density ε, micro-strain δ and texture coefficient TC of NiO films.

Orientation	(111)	(002)	(022)	(222)
D (nm)	81.68	24.35	21.58	223.52
ε (10−4)	6.99	20.72	18.00	1.57
δ (1014 lines/m2)	1.50	16.86	21.48	0.20
TC	3.37	0.23	0.13	0.27

represents the values of the crystallite size, micro-strain, and dislocation density values for the different orientations. The obtained results prove the origin of the lattice parameter “a” change in the NiO film, which is associated with the local defect. Indeed, the crystallite size is larger than the mesh size in all orientations. Thus, the crystallites are formed from a large number of meshes which make some properties of the synthesized thin NiO film close to the ideal case.

3.1.3. Crystallite Formation Proof

The probability of crystallite growth could be determined by the mean values of the structural parameters related to the texture coefficient (TC) obtained by XRD measurements. The texture coefficient (TC) of the maximum preferred orientation (hkl) plane “TC _{(h k l)}” is deducted from the XRD spectrum according to the following formula [63,64]:

$$T{C}_{\left(hkl\right)}=\frac{I\left(hkl\right)/I_0\left(hkl\right)}{N^{-1}\sum I\left(hkl\right)/I_0\left(hkl\right)}$$

(7)

where I (hkl) represents the measured relative intensity of (hkl) plane, I₀(hkl) is the standard intensity of the same plane taken from the JCPDS 040-835, and N indicates the reflection number.

The TC _(hkl) values of the NiO sprayed films confirm that the orientation of the crystallites is along (111) direction [63,64]. Admitting that TC _{(h kl)} may give an indication about the probability of crystallite formation. The mean values of the structural parameters, i.e., the crystallites size <D>, the residual stress <ε>, and the dislocation density <δ> could be found using Williamson and Hall relations [65,66]:

(8)

(9)

(10)

Table 3. NiO films average values of crystallite size, dislocation density, and micro-strain.

<D> (nm)	ε (10−4)	δ (1014 lines/m2)
86.00	7.77	2.94

resumes the average calculated values of crystallite size, dislocation density, and micro-strain of NiO films. It is possible then to estimate the number of meshes per crystallites as reported in [67,68] using the following relation:

$$N_{\mathrm{M}/\mathrm{C}}={\left(\frac{D}{a}\right)}^3$$

(11)

The probability of crystallite growth is assigned to the number of meshes per crystallite. Herein, we estimated the number of meshes per crystallite in NiO prepared films, we found it equal to ≈8.709 × 10⁶. The obtained finite number is coherent with the presence of the defects. Indeed, as mentioned by E. Blanc, et al. [69], the typical dimension of the crystallite in the ideal case assigned to a single crystal must be higher than 1 µm. Thus, for nickel oxide, such a dimension corresponds to at least 10¹² meshes per crystallite. This latest is greater than the estimated value in our NiO films. This confirms the defect influence on the growing NiO films where it reduces the meshes per crystallites.

3.2. Morphological Investigation

The growth mechanism, the quality, and the topography of the NiO films have been studied by correlating both SEM and AFM. The SEM images of NiO films across a large area of the sample (≈100 μm) illustrate a rough surface with pores randomly located (see, Figure 2a). At high magnifications, this image clearly confirms a typical microstructure of the NiO thin film (see, Figure 2b). The observed roughness comes from a nonhomogeneous visible grain with a regular shape through the NiO film (see, Figure 2b). The located grains showed a large size variety, where their sizes change from 200 nm to 400 nm (see, Figure 2b). Similarly, wide pores have been found to induce a porous aspect to the NiO film (see, Figure 2b). Other aggregates, i.e., surface defects have been also located. Such defects are mainly formed during a deposition process. Their origin could be attributed to the residual wet phases and some hydrochloric acid molecules penetrating pores.

We also studied a 3 × 3 μm² area across the sample surface using AFM microscopy to image the morphology of the synthesized NiO film. Figure 3a illustrates the 2D AFM images of the NiO sprayed films. A rough surface of the NiO film has been located similarly to that seen with SEM. Herein, a high-intensity region with grains like that revealed by SEM appeared clearly, with a good precision of their size and shape (see, Figure 3a). Additionally, weak intensity regions have been located where minor roughness could be reported assigned to the pores observed in SEM. Therefore, AFM measurements are consisting of SEM results. For better imaging, we represented this area using a 3D image (see, Figure 3b). This latest confirms the presence of grains and pores.

3.3. EDX Analysis

For a better analysis of the binary material ratio in the NiO prepared films. Nickel and oxygen stoichiometric ratios have been determined using EDX analysis (see, Figure 4). We located three main Ni peaks at around 1 keV, 7.5 keV, and 8.4 keV, respectively. Another peak has been observed at 0.5 keV which corresponds to oxygen. On the other hand, a fifth peak was also found near 2.8 keV. This latest could be originated from residual chloride descending from the protocol followed for the growth process. We also deduced the weight and atomic percentage of nickel and oxygen in the grown NiO thin films from EDX measurements.

Table 4. Chemical compositions in NiO films.

Element	Weigh %	Atomic %
Ni	76.4	48.22
O	21.65	50.12
Cl	1.95	1.66

summarizes the related mass and atomic percentages of the elements oxygen and nickel. We found an atomic ratio of [O]/[Ni] slightly higher than 1 (≈1.040) highlighting the p-type of the NiO material synthesized here.

The root means square roughness (R_ms) usually expressed as the average square deviation from the mean line within the assessment length L (see, Figure 4). Indeed, R_ms is given by the relation [70,71,72]:

$$R_{ms}=\sqrt{\frac{1}{L} {{\displaystyle \int }}_0^L{\left(z-z_m\right)}^{2}dx}$$

(12)

We found R_m in the order of 20 nm. It is less than $\frac{1}{10}$ of the sample thickness which demonstrates promising optical applications of the grown NiO films.

3.4. Dispersion Parameters

We studied the optical properties of NiO films with SE. A simplified diagram of the ellipsometry measuring device is given in Figure 5. This technique is a self-referencing technique, measuring the polarization states before and after reflection. It is based on the interaction of polarized light with a sample described by the Jones matrix formalism [73,74,75]:

$$\left[\begin{array}{c}{E}_{rp}\\ E_{rs}\end{array}\right]=\left[\begin{array}{cc}{\tilde{R}}_{pp}& {\tilde{R}}_{sp}\\ {\tilde{R}}_{sp}& {\tilde{R}}_{ss}\end{array}\right]\left[\begin{array}{c}{E}_{ip}\\ E_{is}\end{array}\right]$$

(13)

where E_ip and E_is correspond to the parallel (p) and perpendicular (s) components of the incident electric field beam light(i), while E_rs and E_rs are assigned to the reflected (r) ones. ${\tilde{R}}_{pp}$ and ${\tilde{R}}_{ss}$ correspond to the complex Fresnel reflection coefficients of the sample for both polarization p and s, correspondingly. The coefficients ${\check{R}}_{sp}$ and ${\check{R}}_{ps}$ are equal to zero since the cubic structure is isotropic.

The ${\tilde{R}}_{pp}$ and ${\tilde{R}}_{ss}$ coefficient expressed for the simple ambient/material case in the top panel of Figure 5 are given by the following relations [75]:

$${\tilde{R}}_{pp}=r_p=\frac{N\mathit{cos}{\Phi }_i-N_0\mathit{cos}\Phi }{N\mathit{cos}{\Phi }_i+N_0\mathit{cos}\Phi }=\left|r_p\right|e^{i{\delta }_p}$$

(14)

$${\tilde{R}}_{ss}=r_s=\frac{N_0\mathit{cos}{\Phi }_i-N\mathit{cos}\Phi }{N_0\mathit{cos}{\Phi }_i+N\mathit{cos}\Phi }=\left|r_s\right|e^{i{\delta }_s}$$

(15)

where δp and δs correspond to the phase changes for the parallel (p) and perpendicular (s) components upon reflection.

Since NiO crystallizes in a cubic structure. This imposes an isotropic optical property on the material. In this case, the off-diagonal terms R_ps and R_sp describe the cross-coupling of both p- and s-polarized light${\widetilde{ R}}_{pp}$ and ${\tilde{R}}_{sp}$become zero. On the other hand, ellipsometry normally describes both parameters: the amplitude ratio upon reflection tanΨ(λ) and the phase shift difference between p and s polarizations Δ(λ). These parameters determine several optical properties, i.e., the refractive index n(λ) and the extinction coefficient k(λ). They are described by using the following equation [76]:

$$\rho =\frac{r_p}{r_s}=tan\mathsf{\Psi } e^{i\mathsf{\Delta }}$$

(16)

where $\mathit{tan}\mathsf{\Psi }=\left|\frac{r_p}{r_s}\right|$ and Δ(λ) corresponds to Δ = δ_p − δ_s.

The experimental and simulated study of tan(Ψ) and cos(Δ) with their standard deviation and average values have been addressed to verify the obtained optical results (see, Figure 6a,b). The ratio (ρ) is related to the complex dielectric function, ε = ε₁ + iε₂, by the following relation [77,78]:

$$\epsilon =si{n}^2{\mathsf{\Phi }}_i \left[1+tan{\mathsf{\Phi }}_i {\left(\frac{1-\rho }{1+\rho }\right)}^2\right]$$

(17)

Both results, i.e., experimental and simulated, of tan(Ψ) and cos(Δ) demonstrate a good coherence. The standard deviation has been found also partially zero. This confirms our findings and gives the reliability to the calculated optical parameters via ellipsometry measurements and Winelli_II software.

We could determine the dispersion parameter, i.e., the refractive index n(λ) and the extinction coefficient k(λ) using the obtained dielectric parameter. This is using the three following relations [76,79]:

ε(λ) = (n(λ) + ik(λ))² = ε₁(λ) + iε₂(λ)

(18)

ε₁(λ) = n² − k²

(19)

ε₂(λ) = 2n*k

(20)

Figure 7 images the n(λ) and k(λ) change. We located a decrease in the value of the extinction coefficient k, where it reaches a low value in the visible range. This illustrates the high transparency of binary materials. However, we found that the extinction coefficient k increased with the wavelength in the infrared domain. This is assigned mainly to the free carrier, mainly the holes. This demonstrates that nickel deficiencies in NiO films induce a p-type character in the semiconductor.

3.5. Magnetic Properties

NiO materials display particular magnetic properties [80,81,82]. The saturation magnetization (Ms), the remanent magnetization (Mr), and the coercivity (Hc) of the NiO thin films have been studied using M-H hysteresis loops at room temperature (see, Figure 8). The associated values have been given in

Table 5. Values of Coercivity (Hc), Magnetization (Ms), and Retentivity (Mr) of NiO thin films.

Ms (10−4 emu)	Mr (10−5 emu)	Hc (Oe)
3.95	1.05	38.08

after subtracting the magnetic contribution of the glass substrate from the overall magnetic behavior. The NiO films display a typical ferromagnetic behavior, despite the anti-ferromagnetic nature of the NiO material [83,84,85]. This unique magnetic effect found here could be related to the: non-stoichiometric defect. Indeed, the prepared thin films contain some defects such as stress and dislocations. This explains the structural defect and Weiss domains in the grown NiO thin films. Here, a minor interaction between domains may cause results that limit the polarization of the material with the coming magnetic field. This induces a ferromagnetic behavior of these films. On the other hand, nickel oxide as binary material is a p-type semiconductor [86]. This p-type has a non-stoichiometric defect that emerged due to the dual effect of chlorine and the excess oxygen. Indeed, the nickel chloride used during the synthesis process leads to a high density of defects due to chlorine that become pronounced when added to the local defect originated by oxygen forming NiO material. This stoichiometric defect may result in such a ferromagnetic behavior. Our find is consistent with previous studies. Indeed, a close ferromagnetic behavior has been reported for NiO powder at high and room temperature [87,88]. This confirms our results obtained for NiO thin film.

3.6. Electrical Properties

The electrical conductivity of the NiO films has been investigated via four points probe at different temperatures. The measurements at each temperature were performed after reaching the thermodynamic equilibrium (i.e., the stabilization of the temperature). The four probes have been arranged linearly, i.e., in a straight line equally distanced from each other. Herein, the potential drop V is measured across the middle of the two probes after the passage of a constant current through them. Meanwhile, the high-temperature effect on the sample has been considered after exposing it to an oven. Figure 9 illustrates the variation of Ln(σT) as a function of (1000/T). This image the electrical conductivity variation with temperature. The thermal conductivity exhibited a semiconductor behavior. Indeed, we found a linear decrease of Ln(σT) versus 1000/T, which confirms that the electrical conductivity follows the Arrhenius law [89,90,91]:

$$\sigma =\frac{A}{T} e^{-\frac{E_a}{KT}}$$

(21)

where A is a constant and $E_a$ is the activation energy.

The calculated activation energy value is equal to 0.5 eV. The obtained energy is smaller than that of $\frac{E_g}{2}$. This proves that NiO cannot be considered an intrinsic semiconductor under our current measurement conditions.

4. Conclusions

In the present work, we reported NiO layers synthesis via the spray pyrolysis technique. A diverse structural, morphological, optical, electric, and magnetic properties of NiO thin film have been revealed. We represent a new approach by specifying the optimal: concentration of the solution, the constant flow rate, and the deposition time. A quasi-monocrystalline structure has been demonstrated for the prepared NiO film with a preferential (111) direction. For the first time, the optical study of NiO spray thin film from ellipsometry has been given here. We illustrated NiO lattice dependency to the effect of the intrinsic defect. Various elements induce defects in the NiO films like chlorine and oxygen. Indeed, nickel chloride used in the synthesis process generates a high density of defects due to chlorine. This is in addition to the located stress and dislocations derived from the oxygen constructing the NiO material. Such defects arose an important unpredicted ferromagnetic aspect. A high thermal dependency of the conductivity, as well as a semiconductor behavior of the grown NiO material, have been revealed. We establish a normal optical dispersion and excellent transparency of the NiO films. Our findings archived herein via a low-cost technique will facilitate the synthesis of NiO films and open the door for their application in industry. A further study is in progress to reach electro-optical properties and magneto-electrical properties in terms of the colossal magneto-resistance properties.