Investigation on the structural and magnetic properties of M$_x$Bi$_{2–x}$Te$_3$ (M = Gd, Fe, Cr) (x = 0, 1) using colloidal hot-injection method

Investigation on the structural and magnetic properties of M

Bi

Te

(M = Gd, Fe, Cr) (x = 0, 1) using

colloidal hot-injection method

N SYED KALEEMULLAH¹, M MALAIDURAI², R THANGAVEL²and J KUMAR^1,*

1Crystal Growth Centre, Anna University, Chennai 600025, India

2Department of Applied Physics, Indian Institute of Technology, Dhanbad 826004, India

*Author for correspondence (marsjk@gmail.com) MS received 29 May 2021; accepted 14 November 2021

Abstract. Gadolinium-doped bismuth telluride (GBT), ferric-doped bismuth telluride (FBT) and chromium-doped bismuth telluride (CBT) nanocrystals have been synthesized using colloidal hot-injection method with same doping ratio.

Field-oriented uniaxial anisotropic ferromagnetic properties for all the samples were analysed from the squareness ratio (M_r/M_s),K₁(magnetocrystalline anisotropic constant),K₁^v(shape anisotropic constant),K_eff(effective anisotropy con- stant) and magnetic energy using vibration sample magnetometer. The magneto-impedance (MI) spectral ratio (DZ/Z₀)%

has been influenced by rotational magnetization, as well as domain wall motion with respect to applied magnetic field.

The results have proven that the GBT thin film sample exhibits the maximum MI effect. Our results may shed light on the simple method of synthesis and development of the effective MI materials based on rare-earth/transition-doped bismuth telluride for the realization of magnetic sensor applications in future.

Keywords. Magneto-impedance; ferromagnetic particles; bismuth telluride; magnetic material.

1. Introduction

BSTS (bismuth-antimony-tellurium-selenium) materials (TIs) have been generating enormous interest among researchers in the past few years [1–5]. For realizing spin-based devices, magnetic doping of BSTS is crucial, which requires dopant atoms to be present in the host system and not just on its surfaces [6–8]. Doping magnetic elements like transition 3dor rare-earth 4fatoms substitutedin-situchalcogens atom breaks the time- reversal symmetry, which leads to occurrence of unique phe- nomena in BSTS, while the spin-polarization generated by them induce magnetism [9–11]. Bismuth telluride (BT) is the important member in the BSTS family (Bi2Te3, Bi2Se3, Sb2Te3, Sb2Se3). Doping BT with varying percentages of Cr can be used to dominate various magnetic phenomena, while varying con- centrations of Fe controls magnetic phases of the material [12,13]. Also, Fe-doped BT reveals unique structures that has temperature-dependent properties [14]. Fe-doped BT shows increased electrical and magnetic properties [15]. Influence of exchange correlation potential on structural and magnetic properties of Gd-doped BT is being studied immensely [10].

Recent literature survey has seen an increasing trend of various protocols and techniques for the synthesis of transi- tion/rare-earth metal-doped BT [3]. Some of these prominent techniques include chemical vapour deposition [16],

Bridgman [17–19], solvothermal synthesis [20], molecular beam epitaxy [21–30], standard solid-state reaction method [23,31,32], solution phase route [33] and wet chemical syn- thesis [34,35]. To our knowledge, we have not come across any reported literature about the synthesis of transition/rare- earth-doped BT using colloidal hot-injection method.

Therefore, in this article, we have synthesized undoped and Fe_xBi_2–xTe₃ (x= 1), Cr_xBi_2–xTe₃ (x= 1) and Gd_xBi_2–xTe₃ (x= 1) powders using colloidal hot-injection method and made thin films using spin coating method. Structural prop- erties of the various powders have been studied using X-ray diffractometer (XRD), while morphology and magnetic properties of the different thin films have been analysed using scanning electron microscopy (SEM), vibrating sample magnetometer and magneto-impedance (MI), respectively.

2. Experimental

Bismuth(III) acetate (99.99%), tellurium (99.8%), iron(III) nitrate (98%), chromium(III) nitrate (non-hydrated) (97%), gadolium(III) oxide (99.99%), tricytylphosphine (TOP) (97%) and oleic acid were purchased from Sigma-Aldrich.

The schematic representation of the synthesis process is shown in figure 1a–e.

https://doi.org/10.1007/s12034-021-02632-x

2.1 Synthesis of undoped Bi₂Te₃nanoparticles

Bi₂Te₃ nanopowders were prepared according to the method reported by Anet al[36]. In a typical procedure, 2 mmol of bismuth(III) acetate was dissolved in 20 ml oleic acid at 40°C for 30 min under vacuum to obtain a trans- parent solution in a beaker sealed with a spiral tube sur- rounded by continuous water flow. The apparatus is as shown in figure 1a. The solution was further heated to 150°C under a N2atmosphere. A Te-TOP solution (10 mol Te in 2 ml TOP) was swiftly injected into the flask while maintaining the final temperature. The transparent solution changes to black immediately after injection of the Te-TOP source. After reaction is complete (about 1 min), the solu- tion was cooled to room temperature by placing the flask in a water bath. The sample is centrifuged to 10,000 rpm for 30 min (figure 1b). The black precipitate so obtained is dried and the powder is mixed with N,N-dimethyl for- mamide and poly-vinylidene fluoride (PVDF)–co-hexaflu- oropropylene and heated in vacuum (figure 1c). The resultant solution is coated on an aluminium substrate using spin coating method (figure1d) and annealed for an hour at 500°C in vacuum (figure1e) [36].

2.2 Synthesis of Gd/Fe/Cr-doped Bi2Te3nanoparticles

For doping Gd/Fe/Cr, the procedure is same as above except additionally the dopants are mixed in the initial step of the reactions. Chromium(III) nitrate is dissolved in 2.5 ml of

distilled water and mixed along with bismuth(III) acetate and oleic acid for doping Cr in Bi₂Te₃ (Cr_xBi_2–xTe₃). In the case of Gd dopant (Gd_xBi_2–xTe₃), gadolinium(III) oxide is mixed with 2.5 ml of HNO₃. For Fe-doped Bi₂Te₃ (Fe_xBi_2–xTe₃), iron(III) nitrate can be directly used in the initial step. For all the dopant systems,x= 1.

2.3 Characterization of materials

The synthesized samples were investigated using scanning electron microcopy (Carl Zeiss MA15/EVO 18), energy dispersive microanalysis (Oxford Liquid Nitrogen free SDD X MAX 50 EDS), X-ray diffractometer (PANalytical X’Pert Powder XRD System) and vibrating sample mag- netometer (Lake shore model:7404). The surface morphol- ogy of all the samples was examined using scanning electron microscopy (Carl Zeiss MA15/EVO 18 Scanning Electron microscopy). Current–voltage (I–V) measurements are made using 6221 Kiethley power source.

Magneto-impedance is measured using the setup reported by Malaiduraiet al[37]. Magnetic field between the poles were measured using the ‘VIJAYANDA DGM-100’ gauss meter. A magnetic field in the range of 0.0–0.1 T was created by changing the current through the electromagnet from 0 to 2 A. Current of 1 mA is supplied to the sample through 6221 Kiethley power source. Analog discovery (II) is used to provide f= 100 kHz and record magnetic impe- dance data [37]. For all the samples, surface of the thin films is kept parallel to the applied magnetic field, i.e.,x-axis.

Figure 1. Schematic illustration of the colloidal hot-injection method synthesis procedure.

3. Results and discussion

Low magnification of SEM images of the synthesized doped (Gd, Fe and Cr) and undoped Bi₂Te₃ samples revealed nanoparticles to be in the range of 100 to 200 nm, as shown in figure 2. SEM analysis showed the evolution of flakes- like structure for BT (undoped Bi₂Te₃), as shown in figure 2a; feather-like structure for GBT (Gd-doped Bi_2- Te₃), as shown in figure2b; and granular structure for FBT (Fe-doped Bi₂Te₃) and CBT (Cr-doped Bi₂Te₃), as shown in figure 2c and d, respectively, with variation in particle density. Energy dispersive X-ray analysis (EDAX) plot for BT, CBT, FBT and GBT is shown in figure2e, f, g and h, respectively. The corresponding molecular weight and atomic weight percentage of elements of all the samples are tabulated in table1. Although equal stoichiometric quantity of dopants (Cr/Fe/Gd) were taken with respect to Bi, the EDAX results showed atomic weight of only 12.61% of Cr, 7.08% of Fe and 5.96% Gd. This is due to the fact that the Cr atomic radius is smallest, while Gd atomic radius is

largest among the dopants used. So a greater number of Cr atoms can be incorporated in Bi sites than compared to Gd atoms.

Figure 2. SEM image of (a) BT, (b) GBT, (c) FBT, (d) CBT; EDAX plot for (e) BT, (f) CBT, (g) FBT, (h) GBT; and (i) voltagevs.

current plot of thin film samples of BT, CBT, FBT and GBT.

Table 1. Molecular weight and atomic weight percent of values of various samples calculated from EDAX.

Sample Atom name Mol. weight % Atomic %

BT Bi 43.05 31.58

Te 56.95 68.42

GBT Gd 5.91 5.96

Bi 47.31 35.91

Te 46.77 58.13

FBT Fe 2.61 7.08

Bi 48.97 35.48

Te 48.41 57.44

CBT Cr 4.66 12.61

Bi 41.33 27.83

Te 54.01 59.56

Figure 2i shows the current vs. voltage plot for all the samples. For pure BT, the I–V plot shows non-ohmic behaviour. As the dopants are added, the plots move further towards non-ohmic behaviour. From the graph it can be seen that in all the samples, when voltage increases there is a non-linear increase in the flow of current because of the carriers generated in the sample.

Figure 3 shows the X-ray diffraction patterns of the undoped BT, GBT, FBT, CBT prepared by hot-injection method. In figure3, the major peaks indexed at (1 0 1) (0 1 5) (1 0 10) (1 1 0) (0 0 15) (2 0 5) (0 2 10) (1 1 15) (1 2 5) correspond to the reflections of rhombohedral phase R3m.

The lattice constantsa= 4.45 A˚ andc= 31.19 A˚ calculated for this experiment are in agreement with the reported values for BT a = 4.385 A˚ and c= 30.48 A˚ (JCPDS 15-0863). The significant figure of change in lattice parameter values ofa,bandcis observed in table2. Such a small variation is attributed to the differences in the atomic radii of doped atoms Cr (128 pm), Fe (126 pm) and Gd (238 pm) in comparison to Bi (230 pm) and Te (210 pm), respectively, and also due to the non-intercalation of dopant

metal atoms into the Van der Waals interlayer as an inter- stitial atom [38–44].

The interplanar spacing d_hkl was calculated using lattice geometry equation [45,46]. Knowing hkl positions, different parameters were calculated as crystalline size (Debye–Sherrer and William–Hall method), dislocation density (1/D²), and volume of unit cell. The average nanocrystalline size ‘D’ is calculated using Deby–Sherrer’s formula as:

D¼ Kk

bcosh; ð1Þ

whereK= shape factor, k= wavelength of Cukaand b is the instrumental broadening [45]. The cell parameters (a =b, c), particle size and other structural properties are shown in table 2. With the addition of dopants, there is a slight variation in the value of lattice parameters. The dopant ions (Gd^?3, Fe^?3, Cr^?3) replace the Bi^?3ions and forms strong bonding between a and b phases. Since ‘c’

parameter is greater than ‘a’ parameter in an a–c phase, therefore weak covalent bond betweena andcphase may be responsible for the variation in the lattice parameters [47,48]. It can be observed that in comparison to BT, cell volume (shown in table2) increases for all the systems.

Using Williamson–Hall method, strain was calculated as:

e¼ b

4tanh: ð2Þ

By rearranging equations (1) and (2),

bcosh¼Kk

D þ4esinh: ð3Þ

The above equation is W–H equation [45].

Figure4a and b shows the variation of lattice constants, coercivity, remanent magnetization with respect to particle size, while figure 4c shows the trend in the variation of lattice constant with respect to all the samples. Figure 4d shows their initial magnetization curves of H (Oe) vs. M (emu g^–1) plot for all the samples. Study of particle size in this experiment for different dopants systems shows that the particle size decreases with the addition of dopants when compared to BT all the cases except for GBT where it increases, as shown in figure4c.

Table3 shows all the magnetic parameters calculated in this study. The M–H loops for BT, GBT, FBT and CBT samples are as shown in figure5a and their magnified image is shown as inset in figure5b. The variation of dM/dHwith applied magnetic field is shown in figure 5c, d, e, f for various systems. From figure5a, it can be observed that BT exhibits paramagnetic nature only between the region

?750 and-750 Oe, while beyond that it shows diamag- netic-like behaviour [49]. The presence of free charge car- riers (holes) results in the paramagnetic contribution. Also, there is a possibility of the presence of group of electrons having a large diamagnetic effect, which is reduced when these electrons are trapped at acceptors or excited to the conduction band to produce holes by the influence of Cr or Figure 3. XRD spectrum of (a) GBT, (b) FBT, (c) CBT, (d) BT

and (e) standard data.

Fe or Gd dopants. Hence, introduction of Cr, Fe and Gd in BT gives them ferromagnetic behaviour, which can be seen in the curves of CBT, FBT and GBT, respectively, in figure 5a. The saturation magnetization (M_s) is calculated

from theM(1/H) plot at 1/H?0. We have calculated the experimental value of the zero-field saturation magnetiza- tion and magnetocrystalline anisotropy constant K₁=‘H_sM_s from the initialM–Hcurve of figure4d. The Figure 4. (a) Plot for particle sizevs.lattice constants ‘a’ (on left axis) and ‘c’ (on right axis). (b) Plot for particle sizevs.coercivity (on left axis) and remnant magnetization (on right axis). (c) Plot for various systemsvs.lattice constant ‘a’ (on left axis) and ‘c’ (on right axis). (d) Field dependence of magnetization for undoped BT (black), GBT (red), FBT (blue) and CBT (green) systems.

Table 2. Various parameters calculated from XRD.

Physical quantities BT GBT FBT CBT

Lattice constant ‘a’ (A˚ ) 4.45 4.42 4.44 4.46

Lattice constant ‘c’ (A˚ ) 31.19 30.69 31.13 31.23

a/cratio 0.136 0.144 0.142 0.142

Cell volume (A˚ ) 535.85 520.06 532.49 539.42

Lattice strain calculated 0.003 0.008 0.006 0.006

Strain from graph 0.004 0.04 0.01 0.01

Dislocation density (910¹²) (nm^–2) 16.27 92.75 46.77 46.76

Particle size (D–S method) (nm) 24.03 14.05 14.26 14.60

Particle size (W–H method) (nm) 24..96 16.97 17.39 17.36

definition of effective anisotropy constantK_eff =‘H_aM_sis related to the anisotropy field throughH_a=M_s/slope. It can be observed that the GBT system has the largest value of K_eff followed by FBT and CBT with BT having the lowest value. It is observed that the system GBT exhibits maxi- mum remnant magnetism followed by FBT, while CBT is the lowest. GBT and FBT systems have highest values for Ms,Msp,K1,K1v

,Keff and energy among the four systems, while having least value for squareness ratio. The square- ness ratio (M_r/M_s, i.e., 0.08 and 0.09 for FBT and GBT) is well below of the typical value*0.5 for randomly oriented uniaxial anisotropic ferromagnetic particles [50]. The FBT and GBT systems are soft ferromagnets, showing rapid increase in magnetization at the initial increase in magnetic field. This is followed by magnetic non-saturation, i.e., non- linear increase of magnetization with field above 3 kOe for the systems CBT, FBT and GBT, respectively. The mag- netic parameters, e.g., spontaneous magnetization (M_sp: calculated from the extrapolation of high fieldM(H) data to

zero field limit) and remnant magnetization (M_r: magneti- zation retained in the sample after reducing the field from 20 to 0 kOe) have enhanced remarkably. A comparative plot of the initial magnetization curves (figure 4d) shows that field-dependent increase of magnetization of the systems CBT, FBT and GBT are controlled mainly by magnetic domain rotation [51].

The decrease of coercivity in CBT, FBT and GBT in comparison to BT is also an indication of strong interpar- ticle interactions in the studied system. The increase in the value of anisotropic constant (K_eff) as we move from CBT?GBT?FBT suggests that a transformation of mag- netic grains from single-domain state to multi-domain state in this magnetic material [42]. It can be observed from figure5c to f that the peaks are almost symmetric about the H= 0 axis. The peaks are separated by magnetic field 2H_m. It may be noted (table3) that the peak position atH_m, i.e., the inflection point in M–H curve, is very close to the coercivity (H_c) of the samples. All the peaks exhibit Figure 5. (a) TheM–Hloops for the undoped BT, GBT, FBT and CBT systems. (b) Inset graph shows the magnified image of6a.

Field dependence of dM/dHfor (c) CBT, (d) FBT, (e) GBT and (f) BT systems.

Table 3. Various parameters calculated from vibrating sample magnetometer.

Magnetic properties BT GBT FBT CBT

Saturation magnetization,Ms(emu g^–1) (910^–4) -90 117.38 115.17 10.23

Coercive field,Hc(Oe) 126.6 76.2 83.65 102.29

Remanent magnetization,Mr(emu g^–1) (910^–4) 4.41 9.94 8.43 1.38

Hm(Oe) 156.22 80.86 133.15 75.87

Spontaneous magnetization,Msp(emu g^–1) (910^–4) 4 109.2 104.64 5.99

Susceptibility,v(emu g^–1Oe) (10^–7) -108.5 15.6 16.9 4.1

Magnetocrystalline anisotropic constant,K₁ 0.185 41.04 35.65 0.741

Shape anisotropic constant,K1V

(910^–12) 63.4 86.6 83.4 0.65

Energy,E(J) 0.38 7.47 5.75 0.34

Squareness ratio,Sr 1.03 0.09 0.08 0.23

Effective anisotropic constant,Keff 0.19 110.88 125.93 1

asymmetric behaviour. The effect of single-domain and multi-domain nature of the grains is also realized from the field dependence of dM/dH curves. For an ideal single- domain particle with square-shaped M–H loop, the dM/dHis very large at the coercive field (H_c) and zero at H?0. The finite value of dM/dHatH?0 (2.86910^–6, 1.1152910^–6, 9.16956910^–6, 1.0971910^–5inlBper f.u.

Oe for BT, CBT, FBT and GBT systems, respectively) further establishes the presence of the regimes of single- domain and multi-domain grains in the studied samples [50]. The dM/dHcurves (figure5c–f) showed peaks at about H_m. The larger value of H_mthan that of the coercive field (H_c) indicates switching field distribution due to disordered shell contributions in pseudo-single domain or multi- domain grains [50]. The peak height of dM/dH at H_m is (*3.67084910^–6,1.23092910^–6, 9.51724910^–6, 1.1513 910^–5, in emu g^–1 Oe unit for BT, CBT, FBT, GBT, respectively) noted to be higher than dM/dH at H?0. Figure 4b shows the plot for particle size vs. H_c (coercivity) and M_r(remnant magnetization).H_cdecreases with increase in the particle size reaching a minimum value for GBT and then increasing in the case of BT, while it is opposite in the case of Mr where it increases reaching a maximum in the case of GBT and then decreases for BT.

The longitudinal dc field dependence of the MI is expressed as:

DZ Z0

ð Þ ¼% ZZ₀ Z0

%: ð4Þ

Here, Z₀ is magneto-impedance at zero applied magnetic field value. The impedanceZincrease with an increase of dc magnetic field, showing positive magneto-impedance effect.

Figure 6a–d shows the longitudinal dc magnetic field dependence of MI DZ/Z₀, where the dc fields are applied parallel to the direction of thin film for constant frequency f = 100 kHz. Note that the spectra shown in figure6a–d is strongly dependent on the magnetic field (H Oe) and they present two peaks for all the systems. A typical example involving a hysteretic feature in MI profiles with respect to increasing and decreasing applied dc magnetic fields is displayed in figure 6a–d. A two-peak behaviour with a sensitivity (DZ/Z0)%/Oe near zero field was observed for all the systems. For undoped Bi₂Te₃, the sensitivity (DZ/Z0)%

vs. magnetic fields hysteresis in the magneto-impedance profiles first increases up to 200 Oe and decreases then onwards for both negative and positive field directions, as shown in figure6a. From the spectra, it can be understood that the systems exhibit crystalline anisotropic behaviour.

In figure6b, for GBT, the hysteresis increases from 0 to 400 and -400 Oe on both sides and then decreases.

However, in the case of FBT, the hysteresis decreases from 0 to 200 and -300 Oe and then it increases, as shown in figure 6c. The same behaviour is observed in CBT except the peak is attained at 200 and-200 Oe, respectively. For all the systems, the magnitude of the anisotropic field induced in the thin film is equal to the magnetic field for

which maximum value of impedance occurs. The peak value of the MI ratio is attained at higher magnetic fields for GBT among the four systems, as shown in figure6d. Two peaks in the profile indicates the rotational magnetization, as well as domain wall motion that contributes to the MI [52]. It can be observed that figure6a–d shows symmetrical behaviour. The MI ratio (DZ/Z)% for BT, GBT, FBT and CBT is 10, 21, 13 and 18%, respectively. MI ratio (DZ/Z)%

shows variations less than 30%, which reveals anisotropic MI behaviour in all the cases. When the anisotropic field is reached in the system, then the magnetization of domain wall becomes free to rotate under the influence of the magnetic field. Magnetization occurs because of dc mag- netic field and the magnetic vector begins to rotate increasing the impedance [37]. This phenomenon is explained in figure6e.

Figure6e illustrates schematically the mechanism of MI in the sample. In figure 6e, the red arrowheads are the representation (BT/CBT/FBT/GBT) of direction of the domains in the sample. The red arrow heads point in the direction of orientation of the majority of the domains in the sample. When B is increased in one direction, say negative side of thex-axis, the direction of the domains in the sample changes and they begin to point in same direction (arrow- heads pointing towards the downward direction). This results in the accumulation of the charges in the junction of Figure 6. The longitudinal dc field of magneto-impedanceDZ/Z₀ for the (a) BT, (b) GBT, (c) FBT, (d) CBT and (e) corresponding orientation of magnetic domains and accumulation of charges in the sample.

the top layer in the sample as shown in figure 6e. With further increase in the magnitude of B, the unidirectional orientation of domain is disturbed, consequently decreasing the number of charges in the top layer. The same phe- nomenon is repeated when the direction of B is reversed but the direction of orientation of the domains is mirror oppo- site (arrowheads pointing upward) compared to the previous case and so is the sign of the accumulation of charges in the top layer of the sample.

4. Conclusion

SEM and XRD results indicate the presence of feather- and granular-like morphology in the range of 100–200 nm with rhombohedral structure. Current–voltage characterization revealed the non-ohmic behaviour of all the samples. As the dopants are added, the samples further move towards non- ohmic behaviour. The particle size has increased for GBT in comparison to BT, while for FBT and CBT it has decreased.

GBT have highest values for M_r, M_s, M_sp, K₁, K₁^v and magnetic energy. FBT has highest value of H_c and K_eff among all the systems. The squareness ratio (M_r/M_s, i.e., 0.08 and 0.09 for FBT and GBT) is well below of the typical value *0.5 for randomly oriented uniaxial aniso- tropic ferromagnetic particles. The magnetic parameters (M_sp and M_r) have enhanced remarkably. All the systems exhibit symmetrical behaviour. The MI ratio (DZ/Z₀)% for BT, GBT, FBT and CBT is 10, 21, 13 and 18%, respec- tively. The results have proven that the GBT thin film sample exhibits the maximum MI effect. Among the materials studied in this study, gadolinium-doped BT has proven to be more effective candidate for the realization of future magnetic sensor applications.

Acknowledgements

We would like to thank the Solar Energy Laboratory, Department of Applied Physics, Indian Institute of Technology (IIT-ISM), Dhanbad, Jharkhand, India, for providing experimental facilities. We would also like to thank Pondicherry University for providing vibration sample magnetometry facility.

References

[1] Tokura Y, Yasuda K and Tsukazaki A 2019Nat. Rev. Phys.

1126

[2] Tian J, Chang L T, Kou X, Murata K, Choi E S, Lang Met al 2014Nano Lett.145423

[3] Ashalley E, Chen H and Tong X 2015Front. Mater. Sci.9 103

[4] Tian W, Yu W, Shi J and Wang Y 2017Materials10814 [5] Hasan M Z and Kane C L 2010Rev. Mod. Phys.823045

[6] Harrison S E, Collins-Mcintyre L J, Li S, Baker A A, Shelford L R, Huo Yet al2014J. Appl. Phys.115023904 [7] Zhang J M, Zhu W, Zhang Y, Xiao Y and Yao Y 2012Phys.

Rev. Lett.1091

[8] Shelford L R, Hesjedal T, Collins-Mcintyre L, Dhesi S S, Maccherozzi F and Van der Laan G 2012Phys. Rev. B861 [9] Ju L, Xu T, Hu D, Cui H, Zhang Y and Shi C 2015Chin.

J. Chem. Phys.28173

[10] Van Quang T and Kim M 2013J. Appl. Phys.11317A934 [11] Jo N H, Lee K J, Kim C M, Okamoto K, Kimura A,

Miyamoto Ket al2013Phys. Rev. B871

[12] Kou X, Lang M, Fan Y, Jiang Y, Nie T, Zhang Jet al2013 ACS Nano79205

[13] Kim M and Song J H 2012J. Appl. Phys.11107E307 [14] West D, Sun Y Y, Zhang S B, Zhang T, Ma X, Cheng Pet al

2012Phys. Rev. B85081305(R)

[15] Jo N H, Lee K, Kim J, Jang J, Kim J and Jung M H 2014 Appl. Phys. Lett.104252413

[16] Cha J J, Williams J R, Kong D, Meiste S, Peng H, Bestwick A Jet al2010Nano Lett.101076

[17] Checkelsky J G, Ye J, Onose Y, Iwasa Y and Tokura Y 2012 Nat. Phys.8729

[18] Hor Y S, Roushan P, Beidenkopf H, Seo J, Qu D, Check- elsky J Get al2010Phys. Rev. B81195203

[19] Li H, Song Y R, Yao M Y, Yang F, Miao L, Zhu Fet al2012 Appl. Phys. Lett.101072406

[20] Liu B, Xie W, Li H, Wang Y, Cai D, Wang Det al2015Sci.

Rep.44639

[21] Chang C Z, Zhang J, Feng X, Shen J, Zhang Z, Guo Met al 2013Science340167

[22] Chang C Z, Zhang J, Liu M, Zhang Z, Feng X, Li Ket al 2013Adv. Mater.251065

[23] Kou X F, Jiang W J, Lang M R, Xiu F X, He L, Wang Yet al 2012J. Appl. Phys.112063912

[24] Kou X, Fan Y, Lang M, Upadhyaya P and Wang K L 2015 Solid State Commun.21534

[25] Liu M, Zhang J, Chang C Z, Zhang Z, Feng X, Li Ket al 2012Phys. Rev. Lett.1081

[26] Tan X, Zhang D W, Liu Q, Xue Yu H F, Zhu Y Wet al2018 Phys. Rev. Lett.120130503

[27] Wang J, Chang C Z, Li H, He K, Zhang D, Singh Met al 2012Phys. Rev. B851

[28] Xia B, Ren P, Sulaev A, Liu P, Shen S Q and Wang L 2013 Phys. Rev. B871

[29] Liu W, He L, Xu Y, Murata K, Onbasli M C, Lang Met al 2015Nano Lett.15764

[30] Han M K, Ryu H and Kim S J 2013J. Electron. Mater.42 2758

[31] Li B, Fan Q, Ji F, Liu Z, Pan H and Qiao S 2013Phys. Lett. A 3771925

[32] Vaney J B, Aminorroaya S Y, Takaki H, Kobayashi K, Kobayashi N and Mori T 2019 Mater. Today Phys.

9100090

[33] Liang Y, Wang Y, Zhang G, Zeng D, Zhu M, Fu Jet al2020 Appl. Surf. Sci.501144268

[34] Garnero C F, Gyawali P, Lermechin A, Pegg I L and Philip J 2013J. Mater. Sci.268

[35] Zhang H, Liu C X, Qi X L, Dai X, Fang Z and Zhang S C 2009Nat. Phys.5438

[36] An J, Han M K and Kim S J 2019J. Solid State Chem.270 407

[37] Malaidurai M, Bulusu V, De S and Thangavel R 20162018 In Proc. AIP Conf. Proc.,00300501

[38] Taziwa R, Meyer E, Katwire D and Ntozakhe L 2017 J. Nanomater.20171

[39] Haldolaarachchige N, Karki A B, Phelan W A, Xiong Y M, Jin R, Chan J Y et al 2011 J. Appl. Phys. 109 103712

[40] Ali N, Ahmed R, Hussain A, Fu Y Q, Khan M, Haq B Uet al 2020 Optik8

[41] Augustine S and Mathai Ma E 2001 Res. Bull.

362251

[42] Zhang Z, Cao Y, Tao Q, Yan Y, Su X and Tang X 2020 J. Solid State Chem.292121722

[43] Knauer A, Krausslich J, Kittner R, Staske R and Arwolff A B 1990Cryst. Res. Technol.25441

[44] Hegde G S, Prabhu A N, Ashok R and Chattopadhyay M K 2021Mater. Sci. Semicond. Process.127105645

[45] Mote V, Purushotham Y and Dole B N 2012J. Theor. Appl.

Phys.61

[46] Sarma H and Sarma K C 2014IJSRP41

[47] Agarwal M B, Sharma A, Malaidurai M and Thangavel R 2018Superlattices Microstruct.117342

[48] Bindu P and Thomas S 2014 J. Theor. Appl. Phys. 8 123

[49] Mansfield R 1959IOP Sci.1959599

[50] Paneer Muthuselvan I and Bhowmik R N 2010 J. Magn.

Magn. Mater.322767

[51] Bowmik R N 2012Compos. B Eng.43503

[52] Kim C G, Jang K J, Kim H C and Yoon S S 1999J. Appl.

Phys.855447