Investigation of Microstructure and Magnetic Properties of Zn1-xMnxO and Zn0.98-xMnxFe0.02O (x = 0, 0.05, and 0.09) prepared by Solid-state Reaction Method

In this study, we investigated the microstructure and magnetic properties of Zn1-xMnxO and Zn0.98-xMnxFe0.02O (x = 0, 0.05, and 0.09) powders prepared by the solid-state reaction method. The starting material, which consisted of ZnO, Mn, and Fe powders, were wet milled for 3 hours using high-energy milling. We then used an X-ray diffractometer (XRD), scanning electron microscope, and vibrating sample magnetometer to investigate the effects of doping and codoping on the microstructure, morphology, and magnetic properties, respectively. The XRD results suggest that Mn and Fe ions had substituted into the ZnO matrix, as illustrated by the resulting single-phase polycrystalline hexagonal wurtzite structures. The diffraction intensity was observed to decrease as the Mn composition increased. The analysis showed that the lattice parameters decreased due to Mn and Fe ion substitution in the ZnO matrix. The co-doping of Mn-Fe ions in the ZnO structure enhanced the magnetic properties, particularly due to the Zn0.89Mn0.09Fe0.02O composition. The increase in the Mn dopant and Mn-Fe co-dopant concentrations strongly contributed to the improved morphology and magnetic properties. Therefore, we can conclude that the presence of Mn and Fe co-dopants in the ZnO system contributed to its magnetic properties, as confirmed by high-saturation magnetization.


Introduction
Zinc oxide is a semiconductor with a hexagonal structure, wide direct band gap (3.3 eV), and an exciton binding energy of 60 MeV at room temperature. This type of semiconductor is an environmentally friendly and low-cost host material that can be processed with different metal ions, including transition metal (TM) elements such as Cr, Mn, Fe, Ni, Co, and others [1][2][3][4]. Among all the TM impurities, Mn is widely used for doping into ZnO, because its radius (0.066 nm) is equivalent to that of Zn atoms (0.060 nm) and it has high solubility [5].
However, Fe-doped ZnO has difficulty obtaining a high Curie temperature (T c ) due to the low solubility of Fe in ZnO and the occurrence of phase separation, which tends to form a cluster or a secondary-phase ZnFe 2 O 4 at doping concentrations >2% [6]. Several research methods have been used in the co-doping of Mn-Fe into ZnO thin films to improve the magnetic saturation, including the sol-gel technique [7], solid-state reaction [8], and solid-state sintering at low temperature [9]. These methods have showed that ZnO co-doped with Mn and Fe can increase the magnetic saturation, as compared to that achieved when using only Mn as a dopant. Co-doping two different TM impurities into the ZnO wurtzite structure is an effective approach for increasing its ferromagnetic properties and Curie temperature [10].
To the best of our knowledge, there has been no discussion of the physical aspects of the co-doping (Mn-Fe) effect with Mn variation into ZnO and the resulting magnetic properties by the solid-state reaction method. The advantages of the solid-state reaction are its relative simplicity, low cost, and easy synthesis as compared to other methods such as ion implantation and thin-film deposition. In this study, we employed the solid-state reaction method to fabricate Zn 1-x Mn x O, and Zn 0.98x Mn x Fe 0.02 O (x = 0, 0.05, and 0.09), which are commonly classified as diluted magnetic semiconductors. June 2020  Vol. 24  No. 2 Our findings suggest that Mn and Mn-Fe co-dopants directly influence the microstructure and morphology of ZnO as well as enhancing its ferromagnetic properties.

Experiments
The starting materials used in the synthesis of Zn 1x Mn x O and Zn 0.98-x Mn x Fe 0.02 O were ZnO (> 99.9%), Mn (> 99.9%), and Fe (> 99%) powders. Table 1 lists the various compositions of the dopant and co-dopants used in this study. The Mn dopant and Mn-Fe co-dopant were mixed and milled by high-energy milling, with a powder-to-ball ratio of 1:10. The wet milling process was performed with the addition of a 20-ml toluene solution for 3 h and drying in ambient air at 100°C for 1 h. Subsequently, the dried powder was put into a mold with a diameter of 12 mm, then pressed in a hot press at a pressure of 1.5 tonf/cm 2 and temperature of 150°C. The microstructures of the Mn dopant and Mn-Fe co-dopant were characterized using X-ray diffraction (XRD, Smartlab Rigaku Cu-Kα radiation wavelength λ = 1.5406 Å). The magnetic properties were measured using a vibrating sample magnetometer (VSM250, Dexing Magnet Ltd.), and the surface morphology was determined using a scanning electron microscope (SEM, Hitachi 3500 SU). We can see that no secondary phase is observed, as compared to a previous work that used the co-precipitation method and found evidence of a secondary phase of Mn 3 O 4 [11]. The authors of another study reported that the use of the solid-state reaction method resulted in the appearance of secondary phases of Mn 2 O 3 at x = 0.2 [12].

Results and Discussion
Interestingly, we found the (002) peak intensities of Zn 1x Mn x O and Zn 0.98-x Mn x Fe 0.02 O to be lower than that of pure ZnO and that these peaks were shifted to a higher 2θ angle, as shown in Figures 1(b) and 1(c), respectively. This trend is in contrast with the findings based on the sol-gel method with x = 0.01-0.1 [13,14] and those based on the solid-state reaction method with x = 0.02, 0.04, and 0.06 [15], which demonstrate peak shifts to a lower 2θ angle. However, our findings showed adequate agreement with previous work that had used the solid-state reaction method, with shifts to a   [12]. Accordingly, the diffraction peaks shifted to higher angles for all Mn 2+ doping concentrations as a result of the occupancy of Zn 2+ sites by Mn 2+ , which causes lattice distortion by tensile stress [16]. The XRD pattern of Zn 0.98-x Mn x Fe 0.02 O shifted to a 2θ angle of 35.35°when the Mn-Fe co-dopant was incorporated into ZnO.
Furthermore, an increase in the Mn-Fe co-dopant concentration resulted in a decrease in peak intensity, particularly at a concentration of x = 0.09. The 2θ angle peak shifted to a higher 2θ angle, as depicted in Figure  1b. It has been reported that this angle peak shifted to lower 2θ angles than those of undoped ZnO [7][8][9]. Other authors have reported a shift in the co-doping (Cr-Ni) angle peak to a higher 2θ angle [2].
Based on the above XRD patterns, we can determine lattice parameters such as the lattice constant, lattice parameter ratio, d-spacing, volume, crystal size, and strain. Lattice parameters a and c in the hexagonal structure can be calculated from the (100) and (002) planes using the following equations: To determine the d-spacing and volume (V) of the lattice, we used the following equations: where λ is the wavelength, θ is the Bragg angle of diffraction, and h, k, l are the Miller indices. Table 2 shows the calculated lattice parameters. We found the lattice parameters a and c of Zn  [17]. Alternatively, this tendency could be caused by a defect in which Mn was only partially doped and did not substitute but only adhered to the surface of ZnO [12]. The ratio of c/a tended to be approximately constant for both the pure and doped samples, which suggests that the dopant atoms were well incorporated into the ZnO crystal lattice without altering the overall crystal structure [18]. To determine the crystallite size (D) and strain (ε), the following equations were used, respectively: where β is the FWHM of a plane (002). As we can see in Table 2, the crystal size (D) and strain (ε) of the Zn  [20].  [19,21]. Moreover, ferromagnetism is reported to emerge in diluted magnetic semiconductors for one of the following three reasons: (1) the formation of secondary phases that do not match; in our case the XRD results in Figures 1(b) and1(c) confirm the absence of a secondary phase [10]. (2) Dopant peaks are present; but no peak corresponding to the dopants are observed [22]. (3) RKKY interaction occurs between the conductive electrons of the host and the spinpolarized electrons of the dopant [23].
It is well known that magnetic properties are possible in ZnO due to the interaction of the spin polarization of the conductive electrons in the ZnO host lattice [24]. Other researchers have reported that the magnetization of ZnO becomes possible with higher or lower ferromagnetic properties with the presence of magnetic impurities that may not be detected by XRD [6]. It has also been demonstrated that either the secondary-phase Mn 2-x Zn x   [25]. The ferromagnetic properties of Zn 0.98-x Mn x Fe 0.02 O are minimal when the addition of doping Mn is x = 0.05 and maximal when the concentration of Mn doping is x = 0.09. The sol-gel method on a thin film can be used to determine the magnetic saturation value, which was found to nearly double when x = 0.01 to x = 0.03 and was almost the same when x = 0.04 and 0.05 [7]. Table  3 summarizes the magnetic saturation values of Mn and the Mn-Fe co-dopant into ZnO, which are much higher than that of pure ZnO and those reported previously [8,9]. On the other hand, Ashraf et al. used the sol-gel method in thin-film samples and reported a similar ratio for the Mn concentration x = 0.03 [7].   Der Walls forces among the ions [26]. This agglomeration caused the particle size to become heterogeneous, ranging from 150 nm to 500 nm, which represents an extension of the nanoparticles size range (100 nm). Figure 3(b) shows histograms of the particle-size distributions from 100 random sampling points for each sample with the fitted Gaussian distribution indicated by the black dashed line. The particle-size distributions of the samples without the addition of Mn dopant (x = 0) tend to be narrow around the peak, which occurs at 245.0 nm and 262.5 nm for the Zn