Elsevier

Journal of Alloys and Compounds

Volume 689, 25 December 2016, Pages 894-898
Journal of Alloys and Compounds

Domain structure transformation and magnetic susceptibility of Ho2Fe17 single crystals

https://doi.org/10.1016/j.jallcom.2016.08.041Get rights and content

Highlights

  • Magnetic domain structure of Ho2Fe17 single crystals was investigated at 4–350 K.

  • A considerable transformation of domain structure at 110 K was found.

  • This transformation is caused by change of magnetic anisotropy in basal plane.

  • Non-SRT anomalies on AC susceptibilities correlates with the transformation of DS.

Abstract

For the first time, an investigation of the domain structure (DS) of Ho2Fe17 single crystals in the temperature range between 4 K and the Curie temperature of the crystals was carried out. Despite having ‘easy plane’ anisotropy in all temperature ranges and, therefore, no spin reorientation transitions and no expected changes in the DS, the Ho2Fe17 single crystals demonstrated a considerable transformation of magnetic domain structure at temperatures near 110 K caused by an increasing role of the magnetic anisotropy in the (001) basal plane at low temperatures. The domain walls of this low-temperature DS demonstrated coercivity to some degree, and this results in anomalies on temperature dependences of AC susceptibilities χ′(T) and χ″(T) measured in the excitation field μ0Hac = 1 mT (so-called non-SRT anomalies).

Introduction

The presence of anomalies in the temperature dependence of AC susceptibilities χ′(T) and χ″(T) measured in a low magnetic excitation field hac are usually considered to be an indicator of magnetic phase transitions [1]. Being relatively simple yet providing a high level of perceptibility, this method is widely used to determine Curie temperature TC and temperatures of spin-reorientation transition (SRT) [2], [3]. However, this method has a drawback if used to determine the SRT temperature: if a sample has a degree of coercivity and if, at the same time, the coercive field Hc exceeds the excitation field hac at low temperatures but Hc is negligible at high temperatures, then the χ′(T) and χ″(T) show additional anomalies at temperature Ts, that are unrelated to the magnetic phase transformation. Indeed, the Hc decreases when the temperature is increased. By the time that temperature Ts is reached, Hc becomes comparable with the magnitude of the excitation field hac. At temperature Ts, being driven by external AC magnetic field hac, the domain walls (DWs) begin to move. This immediately results in the increasing values of χ′ and χ″ (these are so-called non-SRT anomalies) [4], [5], [6], [7]. These anomalies can lead to misinterpretations of the AC data. One of the goals of this work is to show a correlation between local coercivity of low temperature DWs and non-SRT anomalies appearing on χ′(T) and χ″(T).

Despite the fact that these non-SRT anomalies depend directly on the mobility of domain walls and the DS plays an important role in this phenomenon, the previous work [4], [5] emphasized the role of activation energy, while the possibility of changes within the domain structure itself was not taken into consideration. On the other hand, different configurations of magnetic domains imply different types of domain walls and, therefore, presume specific activation energies for each type of domains. The domain structure of R2Fe17 compounds with Tb and Dy was previously investigated [8], [9], [10]. However, all available data in earlier literature was obtained on polycrystalline samples at room temperature, which is far above temperatures where the non-SRT anomalies can be observed. The aim of this work is to show how the transformation of the DS caused by temperature change correlates with non-SRT anomalies. For this, an investigation of the domain structure of a Ho2Fe17 intermetallic compound in the temperature range between 4 K and the Curie temperature was carried out.

All previous investigations of magnetic susceptibility of rare earth (RE) intermetallics in the non-SRT anomalies region were carried out in polycrystalline samples or powders made from polycrystalline samples. However, the grinding of a sample leads to the appearance of defects and additional pinning centers, that should generally distort the effect [11]. Furthermore, the inevitable redistribution of the particle size and their arbitrary orientation with respect to direction of the magnetic excitation field hac did not allow researchers to investigate this effect in detail. In order to exclude these undesired contributions, to observe the DS, and to measure χ′(T) and χ″(T) for the same specimen, a ferromagnetic Ho2Fe17 single crystal was chosen for this investigation.

The magnetic anisotropy of the given compound is of an “easy plane” type for all temperature regions (the Curie temperature of Ho2Fe17, TC = 350 K [12], [13], [14]). From this point of view, no magnetic phase transitions should occur in the sample below TC, and all anomalies on the temperature dependences of susceptibility in the low temperature region may be directly ascribed to non-SRT anomalies. In this article we will show that the magnetic domain structure is an even more sensitive indicator for transformation in the magnetic subsystem of the solid than the conventional magnetisation or magnetic susceptibility. Furthermore, the analysis of the domains configuration transformation together with susceptibility measurements allows us to obtain additional information on the magnetisation processes.

The Ho2Fe17 single crystal was obtained using the following technology: a 150 g alloy was produced by using induction melting from pure materials under an argon atmosphere. In order to increase the size of grains, the ingot was slowly cooled down from 1300°С to 700°С (2 K/min) with a subsequent rapid cooling from 700°С to room temperature (10–20 K/min). After this, the obtained ingot was sealed in a quartz tube under a vacuum and annealed at 1100°С for 8 days with a subsequent slow cooling (2 K/min) in order to obtain both a homogeneous structure and a relatively low concentration of defects in the crystal lattice. The single phase state of the ingot was checked by means of optical metallography, x-ray phase analysis, and energy-dispersive x-ray spectroscopy. Single crystalline grains up to 5 mm in diameter were extracted from the ingot and were orientated using the back scattering Laue method. The stressed surface was relaxed by means of electropolishing in a CrO3 solution. The single crystal selected for DS observation and the susceptibility measurements had a cylindrical shape with a height of 2.5 mm and a diameter of 3.5 mm.

The magnetic domain structure on the basal plane (100) of the Ho2Fe17 single crystal was investigated by means of the polar magneto-optical Kerr effect in a zero magnetic field and in a magnetic field up to 0.3 T in the temperature region from RT down to 4.2 K. The Kerr contrast was improved by depositing a quarter-wave film of ZnS on the surface of the sample [15]. References more thoroughly describe the details of the experimental technique used in this study [16], [17].

The field and temperature dependences of AC susceptibilities were measured in a Quantum Design PPMS commercial system. The real χ′ and imaginary χ″ components of the magnetic susceptibility were obtained in Am2kg−1 units which correspond to the amplitude of the maximal change of the sample magnetic moment in the excitation field hac when oscillating with frequency f.

The temperature dependence of Hc of the Ho2Fe17 single crystal is shown in Fig. 1 (a). The magnetic field was applied along [120] crystallographic directions (b’ axis), which is the direction of easiest magnetisation. The Hc value decreases relatively rapidly with an increase of temperature until 80 K, when it becomes practically undetectable. The inset in Fig. 1(a) shows the major hysteresis loop measured at 10 K where the sample has a coercivity of approximately Hc = 15 mT and small remanent magnetisation of 12 Am−2kg−1. (The arrows and numbers on the insert correspond to the fields in which the transformation of magnetic domain structure is presented in Fig. 3).

It is important to note that the given massive single crystal has easy plane anisotropy, and that the existence of magnetic hysteresis here is quite unusual. The dashed line in Fig. 1 (a) corresponds to the amplitude of the excitation AC field μ0hac = 1 mT which was subsequently used for all our χ′(T) and χ″(T) measurements. One can see that for temperatures below 50 K the Hc value exceeds the amplitude of excitation field hac and therefore the χ′(T) and χ″(T) below 50 K are expected to be of zeroth magnitude.

In Fig. 2 (a) and 2 (b) the temperature dependencies of real and imaginary parts of the magnetic susceptibility χ′(T) and χ″(T) are shown. The measurements were made at various frequencies f, and the AC magnetic field hac was applied along the easiest direction of magnetisation – the b’-axis of the single crystal. In fact, the reduction of Hc induces an increase of both components of the susceptibility, but only above 50 K. Above 110 К (Fig. 2 (a)), the χ′(T) has a plateau up to the Curie temperature. This indicates that the change of magnetisation is caused only by the movement of DWs influenced by hac and the χ′ value that is determined mainly by the demagnetisation factor of the sample.

Temperature dependences of χ″(T) are shown in Fig. 2 (b). The magnitude of this component of susceptibility reflects a temporary delay in the magnetisation change with respect to the changes in the excitation AC field. Obviously, the χ″ shows a frequency-dependent maximum. In low frequencies the maximum on χ″(T) occurs at the temperature of 70 K. With an increase in frequency, it moves to the high temperature region and at f = 10 kHz it is observed at the temperature of 86 K. It is necessary to note that the maximum of χ″ near this temperature is essentially more distinct than the peak which corresponds to the Curie temperature of Ho2Fe17 (the TC in Fig. 2 is designated by arrows). The position of the maximum on χ″(Т) closely corresponds to the inflection points on the χ(Т) curves.

Fig. 3 shows the domain structure at T = 10 K of the Ho2Fe17 single crystal on the prismatic plane (120), which contains both a and c axes ([100] and [001] crystallographic directions). The a-axis of the crystal lies in the figure plane top-down, the c-axis is directed from right to left (hard axis), and the b’-axis (easy axis) is perpendicularly directed to the figure plane but slightly tilted (approx. 1°). An external DC field up to 0.3 T was applied along the b’-axis (out of the plane). Despite the fact that in a sample with the “easy plane” anisotropy six directions of easy magnetisation are possible, the magnetisation in the single crystal (in absence of an applied magnetic field) is actually located just along one selected direction of easy magnetisation, forming a system of 180° stripe domains. This selected easy direction (b’-axis) makes an angle of 30° with the plane of the sample, making it possible to observe rather good Kerr contrasts in 180° stripe domains. Both the experimental scheme of the surface on which the domain structure was observed and the sketch of stripe magnetic domains separated by 180° domain walls are shown in Fig. 1 (c).

In our experiment, a magnetic field was applied in two opposite directions. The left column in Fig. 3 corresponds to positive fields, and the right column shows the DS in negative fields. In fields up to 40 mT, the domain structure does not undergo essential changes (Fig. 3a and e). For magnetic fields larger than 40 mT (Fig. 3b and f), new types of magnetic domains appear from 180° domain walls and grow with the increasing field. The magnetisation direction of these new domains lies along the b’-axis, which is parallel to the external magnetic field and perpendicular to the sample's surface. However, it also makes an angle of 60° with the magnetisation direction in the zeroth field. This type of magnetisation process is accompanied by rather significant magnetic hysteresis, even when a single crystal of a very good quality is obtained and its magnetic anisotropy is of an ‘easy plane’ type.

The field at which the new domains appear on the surface of the sample is different in different areas of the sample. It is worthwhile to note that for the selected area shown in Fig. 3, this field has the same value of 0.04 T up to 100 K. This is rather unusual because under heating the coercivity decreases rapidly (see Fig. 1 (a)) and at 80 K Hc becomes practically undetectable. At temperatures above 100 K, the field at which the new domains appear on the surface begins to gradually increase, and at 110 K it becomes higher than 0.3 T – the maximum available field in our experiment.

In the temperature interval from 110 K up to the Curie temperature of the crystals, no significant transformations in surface domains occur under the magnetic field change up to 0.3 T. Above 110 K, with increasing magnetic field, the Kerr contrast of 180° stripe domains gradually decreases but there is no new type of magnetic domains appearing from the 180° domain walls. Apparently, this is related to the fact that in this temperature interval, the magnetisation proceeds due to the growth of domains only in the volume of the single crystal, thus preserving the surface magnetic domain structure [18].

Section snippets

Discussion and conclusion

As it follows from the six-fold symmetry of the crystal (the Ho2Fe17 single crystal has the ‘easy plane’ type of anisotropy), we could expect 6 different types of magnetic domains where magnetisation would lie along all possible easy directions of the magnetisation. In fact, in the Ho2Fe17 single crystal the magnetisation lies only along one direction with (probably) the lowest demagnetisation field and only 2 types of magnetic 180° stripe domains exist in the sample's volume (this experimental

Acknowledgments

The work was supported by grant of gs1:Ministry of Education and Science of the Russian Federation No 1598. K.S. gratefully acknowledges the financial support of the gs2:Ministry of Education and Science of the Russian Federation in the framework of Increase Competitiveness Program of NUST ‘MISiS’ (K3-2015-029). This work was partially financed with projects of the Spanish MINECO MAT2014-53921-R and DGA IMANA E34 The authors would like to thank D. Goll for the cooperation, fruitful discussions

References (19)

  • J. Bartolomé et al.

    J. Magn. Magn. Mater

    (1995)
  • J. Bartolomé et al.

    J. Alloys Compd.

    (2003)
  • J.J.M. Franse et al.
    (1993)
  • Yu G. Pastushenkov et al.

    J. Magn. Magn. Mater

    (1997)
  • Yu G. Pastushenkov et al.

    J. Magn. Magn. Mater

    (1999)
  • Yu Pastushenkov et al.

    J. Alloys Compd.

    (2008)
  • R.A. Hein
  • V. Skumryev et al.

    J. Appl. Phys.

    (1998)
  • F. Canepa et al.

    Phys. Rev. B

    (2007)
There are more references available in the full text version of this article.

Cited by (0)

View full text