Magnetic nanosized rare earth iron garnets

Journal of Magnetism and Magnetic Materials 422 (2017) 425–433

Contents lists available at ScienceDirect

Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm

crossmark

Magnetic nanosized rare earth iron garnets R3Fe5O12: Sol–gel fabrication, characterization and reinspection ⁎

Olga Opuchovica, , Aivaras Kareivaa, Kestutis Mazeikab, Dalis Baltrunasb a b

Department of Inorganic Chemistry, Vilnius University, Vilnius, LT-03225 Lithuania State Research Institute Center for Physical Sciences and Technology, Vilnius, LT-02300 Lithuania

A R T I C L E I N F O

A BS T RAC T

Keywords: Iron garnets Sol–gel processing Microstructure Mössbauer spectroscopy Magnetization measurements

The magnetic nanosized rare earth iron garnets (R3Fe5O12, where R=Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) were prepared by an aqueous sol–gel method. Herein we present, that all these garnets can be obtained by this effective synthesis method simply by changing the temperature of the final annealing. It was also demonstrated, that a different annealing temperature leads to a different particle size distribution of the final product. The SEM analysis results revealed that the smallest particles were formed in the range of 75–130 nm. The phase purity and structure of the rare earth iron garnets were estimated using XRD analysis and Mössbauer spectroscopy. Magnetic properties were determined by magnetization measurements. The relation between the particle size, composition and magnetic properties of the sol-gel derived garnets were also discussed in this study.

1. Introduction In recent years magnetic materials have been in the focus of interest, with much attention paid to their potential usage in modern technologies. Iron containing garnets and spinels with non-equivalent and antiferromagnetically coupled spin sublattices represent two of the most important classes of these materials [1]. The ferrimagnetic rare earth iron garnets, R3Fe5O12, characterize a unique group of materials which have long been studied for their novel magnetic and magnetooptical properties. The previously mentioned properties have made these materials attractive and useful for practical purposes. Ferrimagnets have found the application in passive microwave components such as isolators, circulators, phase shifters, and miniature antennas operating at a wide range of frequencies (1–100 GHz), as magnetic recording media [2–4]. It has been shown, that nanocrystalline gadolinium iron garnet samples prepared by the microwave hydrothermal method are suitable for circulator, optical isolators and in the fiber communication systems [4,5]. The dysprosium iron garnets are also used in the manufacture of TV screens and data storage due to large Faraday rotation [6]. R3Fe5O12 belong to the space group Ia3¯d, whose structure contains three crystallographic sites: a dodecahedral site 24c, which is occupied by R3+; an octahedral site 16a, occupied by two Fe3+ and a tetrahedral site 24d, taken by three Fe3+, presented in a formula unit [7]. The Fe3+ magnetic moments of the octahedral and tetrahedral sites are coupled antiferomagnetically, and the net magnetic moment is antiparallel to the rare earth ions on the c sites [2,8–12]. The magnetic coupling ⁎

between Fe3+ in a 16a and Fe3+ in a 24d sublattices is strong and dominates above 150 K temperature. When the temperature is below 70 K this magnetic coupling is complete and the magnetic coupling of R3+ in a 24c and Fe3+ in a 24d becomes substantial [13]. To date, numerous methods were used for the preparation of rare earth iron garnets, among which solid-state reaction [3,11,14], coprecipitation [15–17], hydrothermal synthesis [4,18], sol–gel synthesis [19–23], low temperature liquid phase epitaxy (LPE) [24,25] and pulsed laser deposition (PLD) [26] have been suggested. Synthesis technique is significant for the preparation of magnetic particles, since they can exhibit very different magnetic properties [27]. Magnetic properties, such as coercivity, Hc, and saturation magnetization, Ms, are dependent on the morphology and microstructure of the materials [27]. For example, an enhancement of coercivity was obtained as the particle size was reduced [28]. Saturation magnetization, the Néel temperature (TN) and coercivity are very important for practical applications [29,30]. A large remanent magnetization, moderate coercivity, and (ideally) a square hysteresis loop are important characteristics of the materials for the application of the magnetic recording [27]. The sol–gel synthesis route is attractive due to the lower synthesis temperature, better homogeneity and controlled properties of the final product [19,31,32]. Herein we present an aqueous sol–gel preparation of rare earth iron garnets R3Fe5O12 (R=Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu), where a variation of the annealing temperature lets us prepare desirable single phase compounds. To the best of our knowledge, it is the first time when series of such materials are prepared by

Corresponding author. E-mail address: [email protected] (O. Opuchovic).

http://dx.doi.org/10.1016/j.jmmm.2016.09.041 Received 6 June 2016; Received in revised form 26 August 2016; Accepted 7 September 2016 Available online 09 September 2016 0304-8853/ © 2016 Elsevier B.V. All rights reserved.


O. Opuchovic et al.

an aqueous sol–gel processing route and the correlation between microstructure, composition and magnetic properties is shown. 2. Experimental 2.1. Materials and synthesis Reagents and materials used for the experiments in this study were iron (III) nitrate nonahydrate, Fe(NO3)3·9H2O (98.0% Duro-GalvanitChemie), samarium oxide, Sm2O3 (99.99% Aldrich), europium oxide, Eu2O3 (99.9% Alfa Aesar), gadolinium oxide, Gd2O3 (99.9% Alfa Aesar), terbium oxide, Tb4O7 (99.99% Alfa Aesar), dysprosium oxide, Dy2O3 (99.9% Aldrich), holmium oxide, Ho2O3 (99.9% Aldrich), erbium oxide, Er2O3 (99.9% Aldrich), thulium oxide, Tm2O3 (99.99% Alfa Aesar), ytterbium oxide, Yb2O3 (99.9% Aldrich), lutetium oxide, Lu2O3 (99.99% Treibacher), 1,2-ethanediol C2H6O2 (99.5% Aldrich) and nitric acid HNO3 (65% Eurochemicals). Rare earth iron garnets (R3Fe5O12, where R=Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) were prepared by an aqueous sol–gel method. Firstly, stoichiometric amount of an appropriate rare earth oxide was dissolved in hot nitric acid under stirring. When the clear solution was obtained, the excess of acid was removed by evaporation till the dry residue. In the following step, 20 ml of deionized water was added and clear solution was obtained. Fe(NO3)3·9H2O was dissolved in 50 ml of deionized water and both solutions were mixed. After stirring it for 1 h at 55–65 °C, the complexing agent 1,2-ethanediol was added to this solution in the molar ratio of 1:1 to the total metal ions. After stirring for 1 h the obtained solutions were evaporated to form gels. The synthesized gels were dried in an oven (120 °C) for 24 h. The obtained dry xerogels were ground in an agate mortar and preheated at 800 °C for 2 h at a heating rate of 10 °C/min. After an intermediate grinding in an agate mortar the powders were additionally sintered at 800, 900 or 1000 °C for 10 h.

Fig. 1. XRD patterns of Sm–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Sm3Fe5O12.

2.2. Characterization methods X-ray diffraction (XRD) analysis of the samples was carried out using Rigaku MiniFlex II diffractometer (Tokyo, Japan) working in the Bragg–Brentano (θ/2θ) geometry. The data were collected within 2θ angle from 5° to 70° at a step size of 0.01°2θ and at speed of 5°2θ/min, using CuKα radiation. Thermogravimetric/differential scanning calorimetry (TG/DSC) measurements were performed in the flowing air at a heating rate of 10 °C/min using Simultaneous Thermal analyzer STA6000 from PerkinElmer (Waltham, USA). Scanning electron microscope (SEM) FE-SEM Hitachi SU-70 (Tokyo, Japan) was used for the characterization of surface morphology. Images were analyzed using ImageJ program (20 particles in the image) to calculate the particle size. The vibrating sample magnetometer was applied for the magnetization measurements. Magnetic data were taken on powders of prepared garnet samples, which were encapsulated into a plastic straw in order to place into the magnetometer. Therein the lock-in amplifier SR510 (Stanford Research Systems, Sunnyvale, USA) was applied to measure the signal from the sense coils generated by vibrating sample. The gauss/teslameter FH-54 (Magnet Physics) was used to measure the strength of magnetic field between the poles of the laboratory magnet supplied by the power source SM 330-AR-22 (Delta Elektronika, Zierikzee, Netherlands). Mössbauer spectra of solid samples were measured in transmission geometry using 57Co(Rh) source and a Mössbauer spectrometer (Wissenschaftliche Elektronik GmbH, Starnberg, Germany) at room temperature and at elevated temperature applying special Mössbauer furnace. The spectra were fitted to subspectra (sextets) using WinNormos software.

Fig. 2. XRD patterns of Eu–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Eu3Fe5O12.

Fig. 3. XRD patterns of Gd–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Gd3Fe5O12.

sol–gel method using 1,2-ethanediol as complexing agent and by changing the final annealing temperature. The XRD patterns of all synthesized samples at different temperatures are presented in Figs. 1– 10. Obviously, the garnet as dominant crystalline phase could be

3. Results and discussion Different rare earth iron garnets were synthesized by an aqueous 426


O. Opuchovic et al.

Fig. 4. XRD patterns of Tb–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Tb3Fe5O12.

Fig. 7. XRD patterns of Er–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Er3Fe5O12.

Fig. 5. XRD patterns of Dy–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Dy3Fe5O12.

Fig. 8. XRD patterns of Tm–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Tm3Fe5O12.

Fig. 6. XRD patterns of Ho–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Ho3Fe5O12.

Fig. 9. XRD patterns of Yb–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Yb3Fe5O12.

determined in all cases. However, the phase purity of different lanthanide iron garnets depends on the final annealing temperature. It was determined, that samarium, europium and gadolinium orthoferrites (RFeO3) were dominant after heat-treatment of precursor gels

at 800 °C (see Figs. 1, 2 and 3, respectively). By increasing the temperature to 900 °C, SmFeO3 still remained the main crystalline phase in the end product. However, in the europium and gadolinium cases almost monophasic garnets Eu3Fe5O12 and Gd3Fe5O12 have 427


O. Opuchovic et al.

Fig. 10. XRD patterns of Lu–Fe–O gels annealed at 800, 900 and 1000 °C. The vertical lines represent the standard XRD pattern of Lu3Fe5O12.

Fig. 13. Mössbauer spectra of a) Gd–Fe–O annealed at 800 °C, b) Gd–Fe–O, 1000 °C, c) Lu–Fe–O, 800 °C, d) Lu–Fe–O, 1000 °C, e) Sm–Fe–O, 800 °C, f) Sm–Fe–O, 1000 °C, g) Yb–Fe–O, 800 °C, h) Yb–Fe–O, 1000 °C.

synthesis temperature in the range of 800–1000 °C (Fig. 5). In the case of holmium and erbium, the opposite effect was observed. The most pure Ho3Fe5O12, Er3Fe5O12 and Tm3Fe5O12 garnet phases were obtained already at the lowest used synthesis temperature (800 °C). With the increasing synthesis temperature the amount of side perovskite HoFeO3, ErFeO3 and TmFeO3 phases also increased (Figs. 6, 7 and 8, respectively). Interestingly, the monophasic ytterbium and lutetium iron garnets (Yb3Fe5O12 and Lu3Fe5O12) were successfully synthesized only at 800 °C. Further increase of temperature facilitates formation of lanthanide metal orthoferrites (Figs. 9 and 10). It is evident, that crystallization of rare earth iron garnet phase in the series of Sm–Lu elements depends on the final annealing temperature significantly. As it is seen from XRD results, the formation of iron garnets with the marginal lanthanides in the series (Sm3Fe5O12 and Lu3Fe5O12) occurs at quite different temperatures. The thermal decomposition mechanism of the precursor gels used for the fabrication of these garnets was studied by TG/DSC measurements. The representative TG/DTG/DSC curves of Sm–Fe–O and Lu– Fe–O precursor gels are shown in Figs. 11 and 12, respectively. The DTG curves reveal, that thermal behaviour of both samples is associated with three main steps of mass loss. First two mass loss steps in both gels appear at very similar temperatures (221 °C and 263 °C for Sm–Fe–O gel and 226 °C and 272 °C for Lu–Fe–O gel). These results let us conclude that similar mechanism of initial and main decomposition of precursor gels occurs due to the similarity of chemical composition of complexes formed during the synthesis of the gels. The last mass loss step is seen at 739 °C for Sm–Fe–O and at 769 °C for Lu–Fe–O precursors. The DSC curve of Sm–Fe–O gel shows the existence of two exothermic peaks at 880 °C and 970 °C. The first one could be attributed to the crystallization of mixture of perovskite and garnet phases, which is in a good agreement with XRD results. The second exothermic peak could be related to the final re-crystallization

Fig. 11. TG/DTG/DSC curves of the Sm–Fe–O precursor gel synthesized with 1,2ethanediol as a complexing agent.

Fig. 12. TG/DTG/DSC curves of the Lu–Fe–O precursor gel synthesized with 1,2ethanediol as a complexing agent.

formed. Finally, the pure Sm3Fe5O12 phase was obtained only after annealing the precursor gel at 1000 °C. The terbium iron garnet Tb3Fe5O12 was already predominant crystalline phase in the synthesis product obtained at 800 °C (see Fig. 4). Only minor amount of TbFeO3 could be detected in the XRD pattern. The reflections, attributable to perovskite phase, almost disappeared with the increasing temperature up to 900–1000 °C. The phase purity of Dy3Fe5O12 garnet was almost independent on the 428


O. Opuchovic et al.

Table 1 Data from Mössbauer spectra measured at room temperature: S is relative area of subspectrum, Γ, δ, 2ε, B are line width, isomer shift, quadrupole splitting and hyperfine field, respectively. Sample

Tann, °C

S, % ± 1

Γ, mm/s ± 0.01

δ, mm/s ± 0.01

2ε, mm/s ± 0.02

B, T ± 0.07

Phase

Gd–Fe–O

800

34 66

0.32** 0.40

0.37** 0.37

−0.20** 0.02

51.5** 50.35

α-Fe2O3 GdFeO3

Gd–Fe–O

1000

39 61

0.37 0.46

0.38 0.16

0.04 0.03

49.65 40.55

Gd3Fe5O12 a sites Gd3Fe5O12 d sites

Lu–Fe–O

800

36 64

0.43 0.62

0.36 0.14

0.03 0.10

48.76 39.26

Lu3Fe5O12 a sites Lu3Fe5O12 d sites

Lu–Fe–O

1000

50 50

0.32** 0.26

0.37** 0.35

−0.20** −0.04

51.5** 49.78

α-Fe2O3 LuFeO3

Sm–Fe–O

800

31 31 14 24*

0.32** 0.27 0.32 0.71

0.37** 0.36 0.38 0.20

−0.20** −0.07 −0.05 0.06

51.5** 50.67 49.22 40.40

α-Fe2O3 SmFeO3 Sm3Fe5O12 a sites Sm3Fe5O12 d sites

Sm–Fe–O

1000

8 34 58*

0.32** 0.35 0.54

0.37** 0.39 0.16

−0.20** 0.02 0.07

51.5** 49.93 40.74

α-Fe2O3 Sm3Fe5O12 a sites Sm3Fe5O12 d sites

Yb–Fe–O

800

37 63

0.44 0.60

0.37 0.14

0.06 0.10

48.83 39.29

Yb3Fe5O12 a sites Yb3Fe5O12 d sites

Yb–Fe–O

1000

18 14 25 43*

0.32** 0.21 0.38 0.55

0.37** 0.35 0.37 0.14

−0.20** 0.02 0.01 0.07

51.5** 49.96 48.78 39.50

α-Fe2O3 YbFeO3 Yb3Fe5O12 a sites Yb3Fe5O12 d sites

* **

Fixed ratio (d sites)/(a sites)=1.7. Fixed value.

Table 2 Data from Mössbauer spectra of Sm–Fe–O samples measured at elevated temperature Tm: S is relative area of subspectrum, Γ, δ, 2ε(Δ), B line width, isomer shift, quadrupole shift (splitting) and hyperfine field, respectively. Tann, °C

Tm, K

S, %

Γ, mm/s ± 0.02

δ, mm/s ± 0.01

2ε (Δ), mm/s ± 0.02

B, T ± 0.1

Phase

800

475

27 ± 1 33 ± 1 14 ± 1 26 ± 1

0.26 0.35 0.39 0.74

0.24 0.24 0.27 0.08

−0.23 −0.04 0.04 0.09

47.5 40.9 35.2 27.0


800

580

28 ± 2 43 ± 5 11 ± 1 18 ± 2

0.31 0.38* 0.28 0.28

0.17 0.15* 0.17 −0.03

−0.25 −0.02* 0.33 0.90

44.0 30.2** – –


1000

660

11 ± 1 35 ± 1 55 ± 1

0.55 0.33 0.31

0.16 0.12 −0.06

−0.20 0.27 0.90

41.2 – –

α-Fe2O3 Sm3Fe5O12 a sites Sm3Fe5O12 d sites

* **

Fixed values for hyperfine field distribution. Average hyperfine field value for hyperfine field distribution.

The room temperature Mössbauer spectra of Gd–Fe–O, Lu–Fe–O, Sm–Fe–O and Yb–Fe–O gel samples annealed at 800 °C and 1000 °C are shown in Fig. 13. The Mössbauer spectra for the iron garnets were found to consist of two magnetic subspectra (sextets). One corresponds to the octahedral sublattice and the other to the tetrahedral sublattice [34,35]. It is obvious that overall shapes of all the single phase rare earth iron garnets spectra (Fig. 13b, c and g) are essentially independent on the nature of the rare earth ion. However, an overlapping of the spectrum peaks of garnet a site, orthoferrite and hematite can be seen in the other cases. For the better separation of the overlapping peaks wellknown fixed values of hyperfine parameters of hematite [36] were used. In the samples the garnet phase can be identified by the presence of sextet attributable to d site which peaks do not overlap with the peaks of side phases at room temperature. For studied garnets, the magnetic

of the garnet structure material from perovskite, as similar crystallization temperature was observed for Sm3Fe5O12 fabricated by coprecipitation method [16]. Contrary, these two peaks are not visible in the DSC curve of Lu–Fe–O gel, indicating that the final temperature required for the formation this garnet is lower. We can also predict, that a negligible exothermic peak located at around 887 °C can be attributed to the initial decomposition of lutetium iron garnet to perovskite. This temperature again is in a good agreement with XRD data, which show the formation of monophasic Lu3Fe5O12 at 800 °C and LuFeO3 at 900–1000 °C (see Fig. 10). Sol–gel derived R3Fe5O12 garnets were investigated using Mössbauer spectroscopy. For garnet structure materials Mössbauer spectroscopy is an important tool to estimate structural changes, hyperfine interaction, magnetic behaviour and to determine the distribution of Fe3+ ions among sublattices of garnet structure [33]. 429


O. Opuchovic et al.

Fig. 14. SEM micrographs of Sm–Fe–O gels annealed at 800 °C (a) and 1000 °C (b), Ho–Fe–O gels annealed at 900 °C (c) and 1000 °C (d), and Lu–Fe–O gels annealed at 800 °C (e) and 1000 °C (e).

hyperfine fields are found to be around 49 T for a sites and 39 T for d sites and are in a good agreement with reported elsewhere [33,37]. It was found that for the single phase garnet samples (Gd3Fe5O12 annealed at 1000 °C and Lu3Fe5O12 annealed at 800 °C) the area ratio of subspectra attributable to 24d tetrahedral and 16a octahedral sites slightly exceeds the 3:2 (Table 1). The average garnet subspectra area ratio of 1.7 was used when fitting to Mössbauer spectra in case of all three possible phases (Sm–Fe–O annealed at 800 °C and Yb–Fe–O at 1000 °C). The reliability of obtained results was tested comparing them with the results obtained at a higher temperature of measurement. At an elevated temperature the peaks of subspectra of hematite, orthoferrite and 16a site of garnet separate because the corresponding hyperfine field B (Table 2) decreases with the temperature differently – depending on magnetic ordering (Néel) temperatures of the phase. Table 1 summarizes the hyperfine parameter values obtained for the

Table 3 Average particle size calculated from SEM images. Gel composition

Sm–Fe–O Eu–Fe–O Gd–Fe–O Tb–Fe–O Dy–Fe–O Ho–Fe–O Er–Fe–O Tm–Fe–O Yb–Fe–O Lu–Fe–O

Particle size (nm) at different annealing temperatures 800 °C

900 °C

1000 °C

99.88 88.40 101.44 76.46 122.17 112.57 120.87 120.79 129.35 114.58

161.14 141.93 397.66 353.42 279.11 142.38 298.75 290.31 189.80 125.43

778.93 359.05 656.93 408.89 430.87 391.01 483.66 516.67 304.89 267.20

430


O. Opuchovic et al.

Table 4 Magnetic characterization data obtained M versus H curves at room temperature. Hc is coercivity, Mr is remanence, Ms is magnetization at H=4.4 kOe. Sample

Annealing temperature (°C)

Particle size (nm)

Hc (Oe)

Mr (emu/g)

Ms (emu/g)

Sm3Fe5O12 Eu3Fe5O12

1000 900 1000

778.93 88.40 359.05

43.25 97.13 67.39

3.51 4.75 3.98

20.4 12.1 13.8

Gd3Fe5O12

900 1000

397.66 656.93

401.38 350.08

0.32 0.47

0.7 1.0

Tb3Fe5O12

800 900 1000

76.46 353.42 408.89

372.71 363.37 162.16

1.11 1.19 1.77

2.3 3.5 3.6

Dy3Fe5O12

800 900 1000

122.17 279.11 430.87

401.37 274.65 235.62

2.79 2.33 2.59

4.8 4.4 5.0

Ho3Fe5O12

800 1000

112.57 391.01

72.23 58.35

3.92 2.85

9.8 9.6

Er3Fe5O12

800 900 1000

120.87 298.75 483.66

40 34 35

3.7 2.4 2.4

13.5 13.7 14.2

Tm3Fe5O12

800 1000

120.79 516.67

41.56 22.93

3.19 1.64

15.4 14.9

Yb3Fe5O12 Lu3Fe5O12

800 800

129.35 114.58

33.25 32.86

2.99 3.03

17.0 18.1

Fig. 15. M versus H curves for Eu3Fe5O12.

Gd3Fe5O12 formed at 1000 °C, while Yb3Fe5O12 and Lu3Fe5O12 formed at 800 °C and these are supporting results for the XRD analysis data. Moreover, the results obtained for the other measured samples show that the formed impurity phases are orthoferrite and hematite. These side phases are also identified by XRD measurements. Representative images of scanning electron microscopy (SEM) of some rare earth iron garnets revealed that surface morphology of prepared samples is dependent on the final annealing temperature (Fig. 14). It is evident from SEM images, that independently from the phase composition the larger particles have formed at higher temperatures. Table 3 presents the results of particle size calculated from SEM images using ImageJ program. The garnets prepared at 1000 °C are composed of particles 250–800 nm in size. The particle size does not exceed 400 nm for the synthesis products obtained at 900 °C. The smallest particles with the narrow particle size distribution (75– 130 nm) were obtained at 800 °C. Moreover, the particles are highly agglomerated independently on the annealing temperature. The formation of agglomerates covered by smaller particles is typical for different sol-gel derived compounds [38,39]. Compounds, sintered at lower temperatures show also open porous surface microstructure. In order to find correlation between the particle size, composition and magnetic properties the magnetization measurements of the prepared garnet samples were investigated. The hysteresis loops measured for several samples are shown in Figs. 15–17. The hysteresis of the samples varies with different rare earth ions in the composition. In most of the cases, the magnetization of the samples apparently do not saturate at applied magnetic field. Possibly, this is due to the contribution of magnetic moments of rare earth atoms which paramagnetic behaviour gives a linear dependence of magnetization on the strength of the magnetic field [9]. Such contribution is mostly apparent for Gd and Tb garnets (Figs. 16 and 17) for which the magnetization compensation temperature was near measurement (room) temperature. However, for nanomaterials saturation of magnetization can also be affected by surface spin canting and other size-related effects [40,41]. Moreover, in most cases the saturation magnetization has tendency to increase with particle size for the same rare earth iron

Fig. 16. M versus H curves for Gd3Fe5O12.

Fig. 17. M versus H curves for Tb3Fe5O12.

few studied garnet systems. Relative area of the subspectrum S shows relative amount of the phase in the sample considering the Mössbauer effect probabilities for the phases. It is clear, that Sm3Fe5O12 and 431


O. Opuchovic et al.

References

garnet. It is claimed, that the larger magnetic nanoparticles generally have the higher saturation magnetization value [42]. Moreover, the formed side phases orthoferrites and hematite are antiferromagnetics. Thus, it can be concluded that both, particle size and composition of the material is essential for the magnetic properties. Other magnetic characterization data are given in Table 4. Particle size of around 100 nm is small enough for single domain particles to exist. In our case, garnets obtained at 800 °C could be single domain particles. In the case of bigger particles, multi domain structure exists, where regions of uniform magnetization are separated by domain walls, contrary to single domain structure. Because there are no domain walls to move, large coercivity appears for the particles that are small enough [43]. It can be clearly seen, that coercivity decreases with the increasing particle size for the same garnet annealed at different temperature. These results are in a good agreement with the ones reported earlier [15]. The coercivity is dramatically increased in Gd3Fe5O12, Tb3Fe5O12 and Dy3Fe5O12, while remanence magnetization is lowest for these compounds. The same trend is observed for magnetic saturation. These results is consistent with other reports with similar compounds involving lanthanide ions [44]. Magnetic properties are claimed to be influenced essentially by magnetic moments [45], so obtained values might be partially related to the higher magnetic moments of appropriate lanthanide ions in the garnet composition. While some authors claimed, that in compounds with some lanthanide ions in the composition, different magnetic properties did not convincingly correlated with the total number of unpaired electrons of lanthanide ion or experimentally obtained magnetic moments. On the other hand, the strength of spin-orbit coupling alone could not explain different magnetic properties as well [44]. Such magnetic properties as coercivity is important for the practical application of the material. According to Coey, hard magnets have Hc > 400 kA/m (≈5 kOe), soft magnets – Hc < 10 kA/m (≈125 Oe) and magnetic recording media have intermediate values of coercivity [45]. It can be observed from Table 4, that Gd3Fe5O12, Tb3Fe5O12 and Dy3Fe5O12, can act as magnetic recording media while the rest of the presented garnets have properties of soft magnetic material. Depending on the field of application, various magnetic properties are required. Microwave switchers and other devices using the remanence or latching principle require high remanent magnetization and low coercive force which are achieved with large grain size. While in some cases, a small grain size is necessary [46]. Thus, in order to use the materials for devices, the microstructure and particle size must be controlled during the processing.

[1] M. Sugimoto, The past, present, and future of ferrites, J. Am. Ceram. Soc. 82 (1999) 269–280. [2] Ü. Özgür, Y. Alivov, H. Morkoç, Microwave ferrites, part 1: fundamental properties, J. Mater. Sci.: Mater. Electron. 20 (2009) 789–834. [3] F. Grasset, S. Mornet, J. Etourneau, H. Haneda, J.L. Bobet, Effects of ball milling on the grain morphology and the magnetic properties of Gd3Fe3Al2O12 garnet compound, J. Alloy. Compd. 359 (2003) 330–337. [4] T. Ramesh, R.S. Shinde, S.R. Murthy, Nanocrystalline gadolinium iron garnet for circulator applications, J. Magn. Magn. Mater. 324 (2012) 3668–3673. [5] G.-Y. Zhang, X.-W. Xu, T.-C. Chong, Faraday rotation spectra of bismuthsubstituted rare-earth iron garnet crystals in optical communication band, J. Appl. Phys. 95 (2004) 5267–5270. [6] R. Tholkappiyan, K. Vishista, Tuning the composition and magnetostructure of dysprosium iron garnets by Co-substitution: an XRD, FT-IR, XPS and VSM study, Appl. Surf. Sci. 351 (2015) 1016–1024. [7] M.A. Gilleo, S. Geller, Magnetic and crystallographic properties of substituted yttrium-iron garnet, 3Y2O3⋅xM2O3–(5−x)Fe2O3, Phys. Rev. 110 (1958) 73–78. [8] J. Rossat-Mignod, Neutron Scattering, Academic Press, London, 1987. [9] C.D.G. Soshin Chikazumi, Physics of Ferromagnetism, Oxford University Press, New York, 1997. [10] F. Sayetat, Huge magnetostriction in Tb3Fe5O12, Dy3Fe5O12, Ho3Fe5O12, Er3Fe5O12 garnets, J. Magn. Magn. Mater. 58 (1986) 334–346. [11] A. Seidel, L. Häggström, D. Rodić, Search for spin canting in gadolinium iron garnet using the Mössbauer technique, Hyperfine Interact. 73 (1992) 265–275. [12] P. Manimuthu, C. Venkateswaran, On the dodecahedral site substituted Lu3Fe5O12 ceramics for room temperature magneto-dielectric applications, Mater. Lett. 155 (2015) 8–10. [13] D. Petrov, Lattice enthalpies and magnetic loops in lanthanide iron garnets, Ln3Fe5O12, J. Chem. Thermodyn. 87 (2015) 136–140. [14] J.S. McCloy, B. Walsh, Sublattice magnetic relaxation in rare earth iron garnets, IEEE Trans. Magn. 49 (2013) 4253–4256. [15] V. d R. Caffarena, T. Ogasawara, Microstructure and hysteresis curves of samariumholmium-iron garnet synthesized by coprecipitation, Mater. Res. 6 (2003) 569–576. [16] V.R. Caffarena, T. Ogasawara, M.S. Pinho, J.L. Capitaneo, Samarium-iron garnet nanopowder obtained by co-precipitation, Lat. Am. Appl. Res. 36 (2006) 137–140. [17] D. Aoki, M. Shima, Magnetic and fluorescence properties of cerium-doped yttrium gadolinium aluminum iron garnet crystals, Jpn. J. Appl. Phys. 53 (2014) 113001-1113001-113004. [18] L. Guo, H.M. Yuan, K.K. Huang, L. Yuan, S.K. Liu, S.H. Feng, Hydrothermal synthesis, characterization and magnetic properties of rare earth iron garnets Sm3Fe5O12 and Eu3Fe5O12, Chem. Res. Chin. Univ. 27 (2011) 715–719. [19] P. Vaqueiro, M.A. López-Quintela, Influence of complexing agents and ph on yttrium-iron garnet synthesized by the sol−gel method, Chem. Mater. 9 (1997) 2836–2841. [20] R.D. Sánchez, J. Rivas, P. Vaqueiro, M.A. López-Quintela, D. Caeiro, Particle size effects on magnetic properties of yttrium iron garnets prepared by a sol–gel method, J. Magn. Magn. Mater. 247 (2002) 92–98. [21] Z. Cheng, H. Yang, Synthesis and magnetic properties of Sm–Y3Fe5O12 nanoparticles, Physica E Low-Dimens. Syst. Nanostruct. 39 (2007) 198–202. [22] D.T.T. Nguyet, N.P. Duong, T. Satoh, L.N. Anh, T.D. Hien, Magnetization and coercivity of nanocrystalline gadolinium iron garnet, J. Magn. Magn. Mater. 332 (2013) 180–185. [23] K. Praveena, S. Srinath, Effect of Gd3+ on dielectric and magnetic properties of Y3Fe5O12, J. Magn. Magn. Mater. 349 (2014) 45–50. [24] L.N. Singh, Ferromagnetic resonance and relaxation studies in LPE-grown bismuth substituted europium iron garnet films, J. Magn. Magn. Mater. 272 (2004) 2244–2246. [25] A.M. Kalashnikova, V.V. Pavlov, A.V. Kimel, A. Kirilyuk, T. Rasing, R.V. Pisarev, Magneto-optical study of holmium iron garnet Ho3Fe5O12, Low Temp. Phys. 38 (2012) 863–869. [26] H. Yamahara, M. Mikami, M. Seki, H. Tabata, Epitaxial strain-induced magnetic anisotropy in Sm3Fe5O12 thin films grown by pulsed laser deposition, J. Magn. Magn. Mater. 323 (2011) 3143–3146. [27] D.L. Leslie-Pelecky, R.D. Rieke, Magnetic properties of nanostructured materials, Chem. Mater. 8 (1996) 1770–1783. [28] B. Xavier, L. Amílcar, Finite-size effects in fine particles: magnetic and transport properties, J. Phys. D: Appl. Phys. 35 (2002) R15. [29] C.S. Kim, Y.R. Uhm, S.B. Kim, J.-G. Lee, Magnetic properties of Y3−xLaxFe5O12 thin films grown by a sol–gel method, J. Magn. Magn. Mater. 215–216 (2000) 551–553. [30] H. Xu, H. Yang, Magnetic properties of Ce,Dy-substituted yttrium iron garnet ferrite powders fabricated using a sol–gel method, Phys. Status Solidi A 204 (2007) 1203–1209. [31] C.J. Brinker, G.W. Scherer, Sol-gel Science: The Physics and Chemistry Of Sol-gel Processing, Academic Press, London, 1990. [32] A. Kareiva, Aqueous sol-gel synthesis methods for the preparation of garnet crystal structure compounds, Mater. Sci. 17 (2011) 428–437. [33] P.U. Sharma, K.B. Modi, V.K. Lakhani, K.B. Zankat, H.H. Joshi, Mössbauer spectroscopic studies of Fe3+-substituted YIG, Ceram. Int. 33 (2007) 1543–1546. [34] L. Sung, Ho, C. Kwang Pyo, H. Seok Won, L. Young Bae, Mössbauer study of Y3 −xFe5+xO12 and Y3−xInxFe5O12 (x=0.0, 0.18, 0.33) system, Solid State Commun. 83 (1992) 97–100. [35] M. Lataifeh, A.F. Lehlooh, S. Mahmood, Mössbauer spectroscopy of Al substituted

4. Conclusions Rare earth iron garnets with the composition of R3Fe5O12 (R=Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu) were fabricated by a simple and effective aqueous sol–gel method. For the first time to the best our knowledge, it was demonstrated that the final sintering temperature to obtain monophasic rare earth iron garnets is slightly different for different lanthanides, and could be selected with care. In the group of these compounds it was clearly seen, that the temperature of crystallization of R3Fe5O12 is changing from 1000 °C to 800 °C in the sequence of lanthanides from Sm3+ to Lu3+. The larger particles in the range of 250–800 nm of garnets have formed at higher temperatures (1000 °C), while at 800 °C the smaller particles (∼75 nm) were obtained. Mössbauer spectroscopy was used to evaluate relative amount of garnet phases in some prepared samples and confirmed the results obtained by XRD measurements. Correlations between composition, particle size and magnetic properties was observed by measuring magnetic hysteresis loops of the materials. Magnetization measurements revealed, that magnetic properties of synthesized rare earth iron garnet samples depend on the composition of the material and particle size. All these properties can be successfully controlled by changing the synthesis parameters. 432


O. Opuchovic et al.

[41] A.L. Tiano, G.C. Papaefthymiou, C.S. Lewis, J. Han, C. Zhang, Q. Li, C. Shi, A.M. M. Abeykoon, et al., Correlating size and composition-dependent effects with magnetic, mössbauer, and pair distribution function measurements in a family of catalytically active ferrite nanoparticles, Chem. Mater. 27 (2015) 3572–3592. [42] Y. Tan, Z. Zhuang, Q. Peng, Y. Li, Room-temperature soft magnetic iron oxide nanocrystals: synthesis, characterization, and size-dependent magnetic properties, Chem. Mater. 20 (2008) 5029–5034. [43] B. Issa, I.M. Obaidat, B.A. Albiss, Y. Haik, Magnetic nanoparticles: surface effects and properties related to biomedicine applications, Int. J. Mol. Sci. 14 (2013) 21266–21305. [44] M.L. Kahn, Z.J. Zhang, Synthesis and magnetic properties of CoFe2O4 spinel ferrite nanoparticles doped with lanthanide ions, Appl. Phys. Lett. 78 (2001) 3651–3653. [45] J.M.D. Coey, Magnetism and Magnetic Materials, Cambridge University Press, Cambridge, New York, 2009. [46] A.E. Paladino, E.A. Maguire, Microstructure development in yttrium iron garnet, J. Am. Ceram. Soc. 53 (1970) 98–102.

Fe in holmium iron garnet, Hyperfine Interact. 122 (1999) 253–258. [36] E. Kuzmann, S. Nagy, A. Vértes, Critical review of analytical applications of Mössbauer spectroscopy illustrated by mineralogical and geological examples (IUPAC Technical Report), Pure Appl. Chem. 75 (2003) 801–858. [37] D. Vandormael, F. Grandjean, H. Dimitri, J.L. Gary, Mössbauer spectral evidence for rhombohedral symmetry in R3Fe5O12 garnets with R=Y, Eu and Dy, J. Phys.: Condens. Matter 13 (2001) 1759. [38] R. Skaudzius, T. Juestel, A. Kareiva, Luminescence properties of Ln3+–doped (Ce3+, Eu3+, Tb3+ or Er3+) mixed–metals Y3(Al,In)5O12 and Y3Al4.75Cr0.25O12 garnets synthesized by sol–gel method, Mater. Chem. Phys. 170 (2016) 229–238. [39] A. Zarkov, A. Stanulis, T. Salkus, A. Kezionis, V. Jasulaitiene, R. Ramanauskas, S. Tautkus, A. Kareiva, Synthesis of nanocrystalline gadolinium doped ceria via solgel combustion and sol-gel synthesis routes, Ceram. Int. 42 (2016) 3972–3988. [40] C. Suchomski, C. Reitz, D. Pajic, Z. Jaglicic, I. Djerdj, T. Brezesinski, Large-pore mesoporous Ho3Fe5O12 thin films with a strong room-temperature perpendicular magnetic anisotropy by sol–gel processing, Chem. Mater. 26 (2014) 2337–2343.

433