Free Access

This article has an erratum: [https://doi.org/10.1051/epjap/210039s]


Issue
Eur. Phys. J. Appl. Phys.
Volume 94, Number 2, May 2021
Article Number 20101
Number of page(s) 9
Section Semiconductors and Devices
DOI https://doi.org/10.1051/epjap/2021210012
Published online 19 May 2021

© EDP Sciences, 2021

1 Introduction

II–VI semiconductor compound material, such as zinc oxide (ZnO), has a wide band gap (i.e., 3.370 eV) and high exciton binding energy (i.e., 60 meV) [1]. ZnO is rich in reserves and isnon-toxic and harmless. On the one hand, improving the magnetic moment of ZnO can help improve ZnO-based spintronic device. On the other hand, scholars focus on the magnetic source of ZnO-based semiconductor materials. Manganese (Mn) is opted to be used as a doping element to improve the magnetic properties of ZnO and avoid controversy over its magnetic source. The radius of Mn2+ (0.083 nm) [2] is similar than that of Zn2+ (0.074 nm) [2], and substituting Zn with Mn can reduce lattice distortion and improve crystal quality [3]. Experimental literature [4] reported that Mn and its oxides are non-magnetic [4], and their precipitating capabilities did not affect their magnetic properties, thereby making the use of Mn beneficial in the study of magnetic sources.

In recent years, researchers have conducted extensive experimental and theoretical studies on the magneto-optical properties of ZnO doped with Mn. Senol et al. [5] investigated the magnetic properties of ZnO doped with Mn by using hydrothermal method. The results showed that the doped system did not demonstrate any second magnetic phase at 600 °C, and the system was ferromagnetic (FM). Thamaraiselvan et al. [6] examined the effect of the magneto-optical properties of ZnO doped with Mn by using dip coating method. The results show that the doped system has a red-shift into the ultraviolet-visible region, FM properties are reduced, and oxygen vacancy (VO) is observed when Mn is annealed at 500 °C, and its doping amount is increased. Pazhanivelu et al. [7] investigated the effects of ZnO co-doped with Mn and cobalt (Co) on the magneto-optical properties of ZnO by using co-precipitation method. The results show that the FM Curie temperature of the doped system can reach above room temperature. Raman spectroscopy analysis and photoluminescence spectroscopy results show that VO exists in the doped system. Cao et al. [8] investigated the magnetic properties of ZnO co-doped with Mn and Co by using O plasma-assisted molecular beam epitaxy. The results show that the magnetic moment of the doped systemincreases with the amount of Mn doping, and VO/VZn are observed in the doped system when the amount of Co doping is constant. In theoretical calculations, Park et al. [9] used first-principles method to examine the effects of Mn and VO on the magnetic properties of ZnO. The results show that the doped system is FM when the Mn doping amount is 4%, and VO is found between two Mn atoms. Aimouch et al. [10] investigated the magnetic properties of ZnO co-doped with Mn and chromium (Cr) by using the first-principles method. The results show that ZnO systems co-doped with Mn and C rare half-metallic and have improved ferromagnetism. Liu et al. [11] examined the magnetic properties of ZnO co-doped with Mn and Co by using first-principles method. The results show that the Co and Mn co-doped system can achieve room temperature ferromagnetism. Moulai et al. [12] investigated the magnetic properties of Mn-doped ZnO by using first-principles method. The results show that the magnetic moment of the doped system increases with the amount of Mn doping. Yan et al. [13] used the first-principles method to investigate the room temperature ferromagnetism of ZnO, in which Mn doping co-exists with VZn. The results show that the doped systemis FM, and the Curie temperature can reach above room temperature. Ponnusamy et al. [14] used first-principles to study the magnetic effect mediated by Zn vacancies in Mn-doped ZnO. The results show that the ferromagnetism in undoped ZnO originates from VZn. VZn defects and Mn ions coexist in ZnO, and Mn ions also have obvious magnetization effects. Wang et al. [15] used first-principles to study the influence of point vacancies and Mn co-existing on the magnetoelectric properties of ZnO. The results show that the ZnO: Mn-VZn system has p-type conductivity, and the greater the distance between VZn and Mn, the better the electrical conductivity of the system. In contrast to VZn, the ZnO: Mn-VO system is n-type conductivity. The longer the distance between VO and Mn, the worse the conductivity of the system. In addition, the ZnO: Mn system containing neutral single VZn or single VO may be in the antiferromagnetic phase. Although there are certain studies on the magnetic moment produced by the coexistence of Mn doping and oxygen vacancy in ZnO at home and abroad, the source and mechanism of the magnetic moment of ZnO doped with Mn double doping and Zn vacancy is still controversial. The calculation in literature [16] pointed out that the origin of ferromagnetic order in the ZnO system in which Mn double doping and Zn vacancy coexist is caused by the d-d hybrid coupling of Mn-Mn. That is, the source of the magnetic moment of the ZnO system in which Mn double doping and Zn vacancy coexist still adopts the traditional magnetic theory of double exchange interaction with oxygen atoms as an intermediate medium [17] to explain unscientifically. This is contrary to the magnetic theory of carrier as media proposed by Sato and Yoshida [18] and the organic combination of dual exchange mechanism proposed by Dietl et al. [19]. So far, a reasonable theoretical explanation is unavailable. The control of the amount of VO/VZn is experimentally challenging, when Mn doping and oxygen vacancies or Zn vacancies coexist in ZnO. Although previous [9,13] theoretical calculations have determined the effect of the magnetic properties of ZnO with MnZn and VO/VZn, only the one with single MnZn and VO/VZn is discussed. Therefore, the effect of different MnZn to VO ratios on the magnetic properties of ZnO is not investigated. Moreover, the problem of single Mn doping to VO rations in ZnO is common in the study. Explaining the effect of different MnZn to VO ratios on the magnetic properties of the system is also challenging. Theoretical calculations [18] emphasized that the total magnetic moments of ZnO with MnZn and VO depended on the sum of the net magnetic moments of the atoms. This conclusion is inconsistent with the results in the literature [19] and in the present work. Thus, neglecting the net magnetic moment of Zn atoms in the calculation of total magnetic moment is improper. Therefore, this study investigates the effects of different MnZn to VO/VZn ratios on the magnetic properties of ZnO by using first-principles method to solve such problems. Studies have shown that the amount of VZn is large and the total magnetic moment of the doped system is small when the amount of Mn doping is constant. The magnetic moment of ZnO system with MnZn and VO is related to the relative proportion of VO and MnZn and the doped system with MnZn and VO/VZn can achieve room temperature ferromagnetism (RTF); this is consistent with the results reported in the literature [13]. Also, when the MnZn to VZn ratio (MnZn:VZn) is2:1 or 2:2, the doped system exhibits p-type half-metallic ferromagnetism, which suggests the use of the dilute magnetic semiconductor (DMS) as a hole injection source. It has certain theoretical value for the design of new DMS.

2 Theoretical models and calculation methods

2.1 Theoretical model

This work was selected 2 × 2 × 3 supercells to perform all the calculationsto study the magnetic properties of ZnO with MnZn and VO/VZn. First, pure 2 × 2 × 3 supercell of Zn24O24 and one/two Mn atom/s-doped supercell of Zn23Mn1O24/Zn22Mn2O24 were created. Second, literature [14,20] showed that the doped system was relatively stable when MnZn was close to VO/VZn. Therefore, the model with the closest MnZn to VZn was selected in subsequent studies. The ratio (MnZn:VO) of the four kinds of models were 2:2, 2:1, 1:2, and 1:1 with chemical formulas of Zn22Mn2O22, Zn22Mn2O23, Zn23Mn1O22, and Zn23Mn1O23, respectively. The ratio (MnZn:VZn) of the four kinds of models were 2:2, 2:1, 1:2 and 1:1 with chemical formulas of Zn20Mn2O24, Zn21Mn2O24, Zn21Mn1O24 and Zn22Mn1O24, respectively. Figure 1 shows 10 different ratios. Positions 1 and 2, 3 and 4, and 5 and 6 correspond to MnZn, VO and VZn, respectively. The results in literature [21] showed that the phase transition of the crystal structure did not occur when Mn was substituted into Zn and the Mn doping amount reached 10 mol%. The maximum doping amount of Mn herein is approximately 8.33 mol%. Thus, the phase transition will not occur.

thumbnail Fig. 1

Calculation model.

2.2 Calculation method

All calculations were performed using the CASTEP [22] (Material studio 8.0) software package, which used the generalized gradient approximation (GGA) + U plane wave pseudo potential method and Perdew–Burke–Ernzerh of [23] exchange correlation function to improve structure and calculate energy, respectively. Traditional density functional theory cannot accurately describe systems that contain d and f electrons, especially transition metal oxides and nitrides. Meanwhile, GGA+U can accurately describe electronic structures that contain doped transition metal oxides. For localized electrons, GGA+U introduce a Hubbard parameter U (i.e., repulsive energy) that describes the strong correlation among atoms in the model. Computing via test and referring to literature, Ud value of Zn-3d electrons, Up value of O-2p electrons, Ud value of Mn-3d, and plane wave cut-off energy were 10 [24], 7 [24], 6.9 [25], and 380 eV, respectively. K point was set to 4 × 4 × 2, and the ion nuclear is calculated through super soft pseudo potential [26]. The valence electrons of Zn, O, and Mn were 4s23d10, 2s22p4, and 3d54s2, respectively. The accuracy of the atomic plane wave energy was set to 1.0 × 10−5 eV, and the force that acts on each atom has a value of not more than 0.01 eV/Å. The force that acts on each atom and internal stress has a value of not more than 0.3 eV/nm and 0.05 Gpa, respectively; and the tolerance offset was 1.0 × 10−4 nm. The calculated band gap of the ZnO crystal is 3.40 eV, which is consistent with the experimental results [27], indicating that the parameter setting is reasonable, the U value is selected properly, and all structural optimizations and property calculations are performed on the basis of the aforementioned parameters. The valence of Mn in ZnO was calculated to be +2 by using the Bader charge method [28]. The finding is consistent with the experimental results in this work [29]. When judging whether the doped system is ferromagnetic (FM), all atoms spin upward; when judging whether the doped system is antiferromagnetic (AFM), half of the atoms spin up and the other half spin down.

3 Results and discussion

3.1 Formation energy and total magnetic moment

The formation energies of different models were calculatedto analyze the relative stability of doping and defect systemsas follows [30,31]:(1)where ETot(D) is the total energy of the defect system, ETot(ZnO) is the total energy of the pure ZnO system that is as big as the defect system, and q is the charge of the system. Ev represents the energy value of the valence top band; ni is the number of atoms i added or removed; ni is positive when removed, otherwise, it is negative; and μi is the chemical potential of the i atom. The formation energy affects the experimental conditions and is divided into O-enriched (O-rich) and Zn-enriched (Zn-richor O-poor) conditions. Considering the model with the closest Mn to V was selected in subsequent studies. The thermal stability of O and Zn atoms in ZnO, their chemical potentials must satisfy the following condition: μo + μZn + μZnO. In O-rich conditions, the chemical potential of the O atom is represented by half the energy of the O molecule, that is, . Moreover, the chemical potential of the Zn atom is obtained by μZn = μZnO − μo. In O-poor conditions, μZn and μMn are calculated by Zn and Mn bulk materials, which are and , respectively. The chemical potential of O is determined by using μo = μZnO − μZn. The calculated formation energies are listed in Table 1. The smaller the formation energy is, the easier the doping and defects are formed [32], and the stronger the stability is.

It is calculated that the total magnetic moment M, the total anti-ferromagnetic and ferromagnetic energy, the total antiferromagnetic and ferromagnetic settings, and the energy difference between the magnetic states of Zn23Mn1O24, Zn22Mn2O24, Zn22Mn2O22, Zn22Mn2O23, Zn23Mn1O22, Zn23Mn1O23, Zn20Mn2O24, Zn21Mn2O24, Zn21Mn1O24, Zn22Mn1O24 systems are listed in Table 1.

In order to visually observe the formation energy and total magnetic moment of all doped systems in Table 1, the graphs of statistical bars of all doped systems are drawn, as shown in Figure 2.

It can be seen from the organic combination of Table 1 and Figure 2, the formation energy of Zn22Mn2O22 and Zn23Mn1O23 are the same in O-rich and Zn-rich conditions, in which ZnO with different proportions of MnZn and VO co-exists. The formation energy of Zn22Mn2O23 and Zn23Mn1O2 can be smaller and more stable in Zn-rich conditions than in O-rich conditions. In O-rich or Zn-rich conditions, the formation energy sequence of the four doped systems with different ratios is Zn22Mn2O23 < Zn22Mn2O22 < Zn23Mn1O23 < Zn23Mn1O22. The formation energies of Zn20Mn2O24, Zn21Mn2O24, Zn21Mn1O24, and Zn22Mn1O24 systems in Zn-rich conditions, in which ZnO with different proportions of MnZn and VZn co-exists, are larger than those in O-rich conditions. In O-rich or Zn-rich condition, the formation energy sequence of the four doped systems with different ratios is Zn20Mn2O24 < Zn21Mn2O24 < Zn22Mn1O24 < Zn21Mn1O24. The calculation results show that no matter under oxygen-rich conditions (O-rich) or zinc-rich conditions (Zn-rich), the formation energies of all doped systems are negative and compounds are easily formed.

It can be seen from the organic combination of Table 1 and Figure 2, the total magnetic moments of the Zn23Mn1O24 and Zn22Mn2O24 systems are 5 μB and 0 μB, respectively; such findings are the same as the results in the previous study [33]. In contrast with the magnetic moment of ZnO doped with Mn, the total magnetic moment of ZnO, in which MnZn and VO co-exist, is 5μB when Vo: MnZn is 2:1/1:1. When Vo: MnZn is 1:2/2:2, the total magnetic moment of both doped systems is 10μB,which is larger than that of ZnO doped with Mn. The results show that compared with Zn23Mn1O24 system, the total magnetic moments of Zn22Mn2O22, Zn22Mn2O23, Zn23Mn1O22, Zn23Mn1O23, and Zn23Mn1O23 systems are related to the relative Mn doping to VO ratios. This phenomenon is consistent with the reported experimental and theoretical results in the previous study [17,34,35]. The calculation results show that the conclusions of the experimental literature [16] are not contradictory to that of the experimental literature [17]. The magnetic size of ZnO with MnZn and VO is related to the relative MnZn to VO ratio. Also, the total magnetic moments of doped systems with MnZn and VZnare 4.98, 2.75, 8.49, and 7.06μB when VZn: Mn ratios are 1:1, 2:1, 1:2,respectively. When the Mn ratio is constant, the total magnetic moment of the doped system with MnZn and VZndecreases with the increase in VZn ratio. Due to controlling the ratio of Zn in the experiment is difficult, no experiment supports the aforementioned results.

It can be seen from Table 1 that, except for the ferromagnetic properties of Zn22Mn2O22 and Zn20Mn2O24, all other doped systems exhibit antiferromagnetic properties.

Table 1

Formation energy of all systems, total energy of antiferromagnetic and ferromagnetic settings, total magnetic moment, and energy difference between antiferromagnetic state and ferromagnetism state.

thumbnail Fig. 2

Formation energy and total magnetic moment of all systems.

3.2 Analysis of magnetic and half-metallic properties of doped system

It is calculated that the total density of states distribution of the Mn-doped non-vacancy system, the Mn-doped O-containing vacancy system and the Mn-doped Zn-containing system are shown in Figures 3a–3i.

Figures 3a–3i show that the total state density distributions of the electron spin up and spin down are asymmetric, thereby exhibit a net magnetic moment. This result is consistent with the results of the total magnetic moment analysis in Section 3.1.

Spin polarizability (P) refers to the ratio of the density of states of major carriers near the Fermi level less the minor carriers to the sum of major and minor carriers. Ps obtained by using the following equation:(2)

Figures 3f and 3g show that Zn20Mn2O24 and Zn21Mn2O24 have a band tail effect, respectively. Both Fermi levels enter the upper spin valence band, and the Fermi energy levels that did not enter lower spin valence band formed a half-metallized p-type degenerate semiconductor. This phenomenon is due to the difference in the relative Mn doping to VZn ratios in the Zn20Mn2O24 and Zn21Mn2O24 systems, the concentration of free holes in the doped system, and the adjustment of the ferromagnetic interaction of DMS. According to formula (2), the Ps of the Zn20Mn2O24 and Zn21Mn2O24 systems are both 100%, that is, both of them have conduction hole Ps of up to 100%, and they can be used as hole injection source for the promising new type of half-metallized DMS.

thumbnail Fig. 3

Total state density of distribution: (a) Mn doped without vacancy; (b–e) Mn doped O–containing vacancy system; (f–i) Mn doped Zn–containing vacancy system.

3.3 Analysis of magnetic sources in doped system

According to the results in Section 3.1, the total magnetic moments of Zn23Mn1O22, Zn23Mn1O2, Zn21Mn1O24, and Zn22Mn1O24 systems do not increase, and their relative research values are insignificant compared with those of the Zn23Mn1O24 system. However, the total magnetic moments of Zn22Mn2O22, Zn22Mn2O23, Zn20Mn2O24, and Zn21Mn2O24 systems greatly increased. Therefore, the research of the next paragraph mainly focuses on the magnetic sources of Zn22Mn2O22, Zn22Mn2O23, Zn20Mn2O24, and Zn21Mn2O24 systems.

The calculated partial density of state (PDOS) distributions of Zn22Mn2O22, Zn22Mn2O23, Zn20Mn2O24, and Zn21Mn2O24 are shown in Figures 4a–4d.

Results show that the total magnetic moment of Zn22Mn2O22 and Zn22Mn2O23 is 10μB, and μB is Bohr magneton. For example, the magnetic moments of two Mn atoms in Zn22Mn2O22 system are 5.13μB and 5.09μB, and the magnetic moments of the main O atoms near VO are 0.02μB or 0.01μB. The magnetic moments of Mn and O atoms are arranged in the same direction. The magnetic moments of the main Zn atoms near the VO are −0.07, −0.06, −0.04, and −0.02μB, and they are arranged in the opposite direction. Therefore, the total magnetic moment of Zn22Mn2O22 system is 10μB. Since the magnetic moment source of the Zn22Mn2O23 system is similar to that of the Zn22Mn2O22 system, it will not be repeated.

From Figures 4a and 4b, it can be seen that the magnetic source of Zn22Mn2O22 and Zn22Mn2O23 system is O-2p orbital, there is a strong coupling electron exchange between Zn-4s orbital and Mn-3d orbital, sp-d caused by this strong interaction. That is, the O-2p partial density of state upper-spin orbit and the lower-spin orbit are asymmetric; the Zn-4s orbital partial density of state is asymmetric; the Mn-3d orbital partial density of state is asymmetric.

Results show that the total magnetic moments of Zn20Mn2O24 and Zn21Mn2O24 are 7.06μB and 8.49μB, respectively. For example, the magnetic moments of the two Mn atoms in the Zn20Mn2O24 system are 4.35μB and 4.33μB, and the magnetic moments of the main O atoms near the VZn are −0.40,−0.29,−0.27, and −0.26μB. Therefore, the total magnetic moment of Zn22Mn2O22 system is 7.06μB. In the same way, the magnetic moment source of Zn21Mn2O24 is similar to that of Zn22Mn2O22, it will not be repeated.

Figures 4c and 4d show that a strong coupled-electron exchange exist among the O-2p, Zn-4s, and Mn-3d orbits of Zn20Mn2O24 and Zn21Mn2O24 systems. The PDOS of O-2p, Zn-4s, and Mn-3d orbits are asymmetric in spin-up and spin-down orbits because of the strong sp-d interaction. This can also be explained from the theory of defect chemistry. For example, due to 2 Mn substitutions for Zn doping and 2 Zn vacancies in ZnO, the reaction formula is as follows.(3)

It can be seen from the formula (3) that the 2MnZn − 2VZn complex is formed in the doped system.

In the equilibrium state, the defect chemical reaction of ZnO is(4) (5) (6)

From equations (4)(6), we know that after Zn vacancies are formed, hole carriers appear in the doped system. This is because in ZnO, Zn ions are usually positive divalent. Once there is excess oxygen, in order to maintain the neutral condition, the divalent Zn ions trap a hole. According to the density of states Figures 4c and 4d, it can be seen that the Fermi level enters the valence band, the doped system is a p-type degenerate semiconductor, and the ionization energy is 0. Therefore, free hole carriers appear in the doped system to conduct electricity. It can be concluded that the magnetic sources of Zn20Mn2O24 and Zn21Mn2O24 systems are consistent with the magnetic theory of carrier as media proposed by Sato and Yoshida [18] and the organic combination of dual exchange mechanism proposed by Dietl et al. [19]. However, the literature [16] not only neglected the magnetic source of O-2p orbital and Zn-4s orbital electron spin polarization, but also neglected the mechanism of magnetic source mediated by carriers. It is worth further study.

Figures 5a–5d show the calculation of the net spin density distributions of Zn22Mn2O22, Zn22Mn2O23, Zn20Mn2O24, and Zn21Mn2O24 to intuitively analyze the magnetism sources of doped systems from the electron's structure.

It can be seen from Figures 5a and 5b that the total magnetic moment of the Zn22Mn2O22 and Zn22Mn2O23 systems is contributed by the difference between the up-spin density and the down-spin density of the spin-polarized O atom, contributed by the difference between the up-spin density and the down-spin density of spin-polarized Mn atoms, the opposite is true for Zn atoms, which is contributed by the difference between the down-spin density and the up-spin density. The results are in good agreement with the findings in PDOS. It can be seen from Figures 5c and 5d that the total magnetic moment of the Zn20Mn2O24 and Zn21Mn2O24 systems is contributed by the difference between the down-spin density and the up-spin density of the spin-polarized O atom, contributed by the difference between the up-spin density and the down-spin density of spin-polarized Mn atoms, contributed by the difference between the up-spin density and the down-spin density of the spin-polarized Zn atom. This is consistent with the analysis results of the PDOS distribution.

In order to visually observe the difference in charge transfer before and after doping, we calculated that the differential charge density distribution in the (0,0,1) plane of Zn24O24, Zn23Mn1O24, Zn22Mn2O22, Zn22Mn2O23, Zn20Mn2O24, Zn21Mn2O24 systems are shown in Figures 6a–6f. As can be seen from Figures 6a–6f, compared with the pure system Zn24O24, the charge density around O and Mn in the system Zn23Mn1O24, Zn22Mn2O22, Zn22Mn2O23, Zn20Mn2O24, Zn21Mn2O24 is relatively denser. The calculation results show that the covalent bonds of Zn23Mn1O24, Zn22Mn2O22, Zn22Mn2O23, Zn20Mn2O24, Zn21Mn2O24 are enhanced after doping.

thumbnail Fig. 4

PDOS distributions of (a) Zn22Mn2O22, (b) Zn22Mn2O23, (c) Zn20Mn2O24, and (d) Zn21Mn2O24.

thumbnail Fig. 5

Net spin density distributions of (a) Zn22Mn2O22, (b) Zn22Mn2O23, (c) Zn20Mn2O24, and (d) Zn21Mn2O24.

thumbnail Fig. 6

Differential charge density distribution (a) Zn24O24; (b) Zn23Mn1O24; (c) Zn22Mn2O22; (d) Zn22Mn2O23; (e) Zn20Mn2O24; (f) Zn21Mn2O24.

3.4 Curie temperature analysis of doped system

As can be seen from Table 1, because the energy difference between the antiferromagnetic state and the ferromagnetic state of the Zn22Mn2O22 and Zn20Mn2O24 systems is relatively large, the Zn22Mn2O22 and Zn20Mn2O24 systems are used as the object to study the Curie temperature. That is, the Curie temperatures of the Zn22Mn2O22 and Zn20Mn2O24 systems were studied separately. The Curie temperature of DMS can be estimated from the energy difference (ΔE) between the AFM and FM states according to the Heisenberg model in mean field approximation, which is illustrated by the following formula [36].(7)Where kB is the Boltzmann's constant, C is the magneticion concentration, and Tc is the estimated Curie temperature of DMS. Formula (7) shows that the increase in ΔE results in high Tc value. As can be seen from Table 1, the total energies of the AFM settings of Zn22Mn2O2 and Zn20Mn2O24 systems correspond to −48515.083 and −45951.560 eV, respectively. The total energies of the FM settings of Zn22Mn2O22 and Zn20Mn2O24 systems correspond to −48515.123 and −45951.598 eV, respectively. The ΔE between the AFM and FM states of the Zn22Mn2O22 system has a value of 48 meV. Moreover, the ΔE between the AFM and FM states of the Zn20Mn2O24 system has a value of 38 meV. The known data were substituted into formula (7), and the Curie temperatures of Zn22Mn2O22 and Zn20Mn2O24 were calculated to be 400 and 380 k, respectively. The calculated results are in agreement with the experimental results [13,37]. The results show that Zn22Mn2O22 and Zn20Mn2O24 systems can meet the requirements of Curie temperature, which is above room temperature.

4 Conclusion

The effect of different Mn doping to point vacancy ratios on the magnetic properties of ZnO has been investigated using the first principle method. In the formation energy of doped systems, ZnO with different ratios of MnZn and VO/VZn can be smaller and more stable in Zn-rich conditions than in O-rich conditions. The doped system exhibits p-type half-metallic ferromagnetism when the MnZn: VZn is 2:1 or 2:2 in ZnO. The increase in VZn results in small total magnetic moments when the doping amount of Mn is constant. For the ZnO system in which Mn doping and oxygen vacancies coexist, when the amount of oxygen vacancies is constant, with Mn doping increase, the magnetic moment becomes larger. Both Zn22Mn2O22 and Zn20Mn2O24 can achieve ferromagnetic characteristics above room temperature.

Conflict of interest: the authors declare that they have no conflict of interest.

Author contribution statement

Qingyu Hou: writing − original draft, project administration, funding acquisition.

Yuqin Guan: conceptualization, methodology, writing − review & editing, data curation.

Zhichao Wang: conceptualization, methodology, writing − review & editing, data curation.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (Grant Nos. 61366008, 61664007 and 61964013) and Science and Technology Major Project of Inner Mongolia Autonomous Region (2018-810).

References

  1. Z.L. Wang, ACS Nano 2, 1987 (2008) [CrossRef] [PubMed] [Google Scholar]
  2. J. Panda, I. Sasmal, T.K. Nath, AIP Adv. 6, 035118 (2016) [Google Scholar]
  3. T. Yu, P. Cheng, S.P. Huang, P. Wang, H.P. Tian, Comput. Theor. Chem. 1057, 15 (2015) [Google Scholar]
  4. S.A. Ahmed, Results Phys. 7, 604 (2017) [Google Scholar]
  5. S.D. Senol, A. Guler, C. Boyraz, L. Arda, J. Supercond. Nov. Magn. 32, 2781 (2019) [Google Scholar]
  6. P. Thamaraiselvan, S. Kannan, P. Gowthaman, M. Kutraleeswaran, Int. J. Sci. Res. Sci. Technol. 4, 679 (2018) [Google Scholar]
  7. V. Pazhanivelu, A.P.B. Selvadurai, R. Murugaraj, J. Mater. Sci.-Mater. Electron. 27, 8580 (2016) [Google Scholar]
  8. Q. Cao, M.X. Fu, G.L. Liu, H.J. Zhang, S.S. Yan, Y.X. Chen, L.M. Mei, J. Jiao, J. Appl. Phys. 115, 243906 (2014) [Google Scholar]
  9. S.Y. Park, H.W. Lee, J.Y. Rhee, J. Korean Phys. Soc. 51, 1497 (2006) [Google Scholar]
  10. D.E. Aimouch, S. Meskine, A. Boukortt, A. Zaoui, J. Magn. Magn. Mater. 451, 70 (2018) [Google Scholar]
  11. Z.X. Liu, X.M. Yuan, P. Yang, J. Magn. Magn. Mater. 461, 1 (2018) [Google Scholar]
  12. N. Moulai, N. Bouarissa, B. Lagoun, D. Kendil, J. Supercond. Nov. Magn. 32, 1077 (2019) [Google Scholar]
  13. W.S. Yan, Z.H. Sun, Q.H. Liu, Z.R. Li, Z.Y. Pan, J. Wang, S.Q. Wei, Appl. Phys. Lett. 91, 062113 (2007) [Google Scholar]
  14. R. Ponnusamy, S.C. Selvaraj, M. Ramachandran, P. Murugan, P.M.G. Nambissan, D. Sivasubramanian, Cryst. Growth Des. 16, 3656 (2016) [Google Scholar]
  15. C.A. Wang, J.X. Li, S.L. Fu, J.Y. Bao, T. Lei, J. Miao, Int. J. Mod Phys B 34, 2050210 (2020) [Google Scholar]
  16. E. Salmani, O. Mounkachi, H.E. Zahraouy, M. Hamedoun, A. Benyoussef, J. Supercond. Nov. Magn. 26, 229 (2013) [Google Scholar]
  17. C. Zener, Phys. Rev. 82, 440 (1951) [Google Scholar]
  18. K. Sato, H.K. Yoshida, Jpn. J. Appl. Phys. 39, L555 (2000) [Google Scholar]
  19. T. Dietl, H. Ohno, F. Matsukura, Science 287, 1019 (2000) [Google Scholar]
  20. Q.Q. Gao, Y.Q. Dai, C.B. Li, L.G. Yang, X.C. Li, C.J. Cui, J. Alloys Compd. 684, 669 (2016) [Google Scholar]
  21. K. Manoj, N. Kuldeep, C. Suvarcha, U. Ahmad, K. Ramesh, M. Yoshitake, C.M. Singh, J. Nanosci. Nanotechnol. 19, 8095 (2019) [Google Scholar]
  22. M.D. Segall, P.J.D. Lindan, M.J. Probert, C.J. Pickard, P.J. Hasnip, S.J. Clark, M.C. Payne, J. Phys.: Condens. Matter. 14, 2717 (2002) [NASA ADS] [CrossRef] [Google Scholar]
  23. J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77, 3865 (1996) [Google Scholar]
  24. H.C. Wu, Y.C. Peng, C.C. Chen, Opt. Mater. 35, 509 (2013) [Google Scholar]
  25. C.H. Chien, S.H. Chiou, G.Y. Guo, Y.D. Yao, J. Magn. Magn. Mater. 282, 275 (2004) [Google Scholar]
  26. D. Vanderbilt, Phys. Rev. B: Condens. Matter 41, 7892 (1990) [Google Scholar]
  27. J.V. Foreman, J.G. Simmons, W.E. Baughman, J. Liu, H.O. Everitt, J. Appl. Phys. 113, 133513 (2013) [Google Scholar]
  28. R.F.W. Bader, Atoms in Molecules-A Quantum Theory, International Series of Monographs on Chemistry 22 (Oxford University Press, Oxford, 1990) [Google Scholar]
  29. K. Samanta, S. Dussan, R.S. Katiyar, P. Bhattacharya, Appl. Phys. Lett. 90, 261903 (2007) [Google Scholar]
  30. A. Janotti, C.G.V.D. Walle, Rep. Prog. Phys. 72, 126501 (2009) [Google Scholar]
  31. T.T. Guo, G.B. Dong, Q. Chen, X.G. Diao, F.Y. Gao, J. Phys. Chem. Solids 75, 42 (2014) [Google Scholar]
  32. H.G. Sun, W.L. Fan, Y.L. Li, X.F. Cheng, P. Li, J.C. Hao, X. Zhao, Phys. Chem. Chem. Phys. 13, 1379 (2011) [Google Scholar]
  33. A.L. He, X.Q. Wang, R.Q. Wu, Y.H. Lu, Y.P. Feng, J. Phys.: Condens. Matter. 22, 175501 (2010) [Google Scholar]
  34. D. Iuşan, B. Sanyal, O. Eriksson, Phys. Rev. B 74, 235208 (2006) [Google Scholar]
  35. Q. Wang, Q. Sun, P. Jena, Y. Kawazoe, Phys. Rev. B 79, 115407 (2009) [Google Scholar]
  36. K. Sato, P.H. Dederichs, Y.H. Katayama, Europhys. Lett. 61, 403 (2003) [CrossRef] [Google Scholar]
  37. U. Philipose, S.V. Naira, S. Trudel, C.F. Souza, S. Aouba, R.H. Hill, H.E. Ruda, Appl. Phys. Lett. 88, 263101 (2006) [Google Scholar]

Cite this article as: Qingyu Hou, Yuqin Guan, Zhichao Wang, Effect of different Mn doping and point vacancy ratios on the magnetic properties of ZnO, Eur. Phys. J. Appl. Phys. 94, 20101 (2021)

All Tables

Table 1

Formation energy of all systems, total energy of antiferromagnetic and ferromagnetic settings, total magnetic moment, and energy difference between antiferromagnetic state and ferromagnetism state.

All Figures

thumbnail Fig. 1

Calculation model.

In the text
thumbnail Fig. 2

Formation energy and total magnetic moment of all systems.

In the text
thumbnail Fig. 3

Total state density of distribution: (a) Mn doped without vacancy; (b–e) Mn doped O–containing vacancy system; (f–i) Mn doped Zn–containing vacancy system.

In the text
thumbnail Fig. 4

PDOS distributions of (a) Zn22Mn2O22, (b) Zn22Mn2O23, (c) Zn20Mn2O24, and (d) Zn21Mn2O24.

In the text
thumbnail Fig. 5

Net spin density distributions of (a) Zn22Mn2O22, (b) Zn22Mn2O23, (c) Zn20Mn2O24, and (d) Zn21Mn2O24.

In the text
thumbnail Fig. 6

Differential charge density distribution (a) Zn24O24; (b) Zn23Mn1O24; (c) Zn22Mn2O22; (d) Zn22Mn2O23; (e) Zn20Mn2O24; (f) Zn21Mn2O24.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.