Phase transformation and exchange bias effects in mechanically alloyed Fe/magnetite powders

https://doi.org/10.1016/j.jallcom.2011.03.147Get rights and content

Abstract

Nanostructured powders processed by ball milling of a mixture of Fe and Fe3O4 at room temperature are shown to undergo an incomplete redox reaction with formation of FeO during the milling process. This reaction is favored by the high energy introduced during the mechano-alloying process. Concurrent effects of milling such as grain refinement down to the nanometre scale lead at the end of the milling processes to a mixed multiphase powder of nanograins, with Fe and Fe oxide grains inter-dispersed. We show that in the as-milled Fe/Fe3O4 powder, during milling process, wüstite (FeO) is formed as a consequence of the redox reaction. Moreover, with increasing temperature, the system undergoes an inverse phase transformation towards the initial Fe and Fe3O4 phases until about 450 °C. Above this temperature the reduction reaction Fe + Fe3O4 = 4FeO is reinitiated, resulting in sharp decrease of Fe and Fe3O4 content from about 550 °C and almost complete disappearance of these phases at about 900 °C. This transformation was investigated via an energy-dispersive in situ X-ray diffraction experiment using the synchrotron radiation. This study allows direct collection of X-ray patterns after few minutes exposure, at selected temperatures, ranging between 20 °C and 1000 °C. The structural and magnetic characterizations of the nanograin powders, as-milled and annealed at several temperatures, are studied using XRD, SEM and magnetic measurements. Such ferromagnetic–antiferromagnetic composites are extensively studied as they exhibit exchange bias effect, with a large impact in technological applications. The magnetic behaviour and intrinsic mechanisms leading to the occurrence of exchange bias effects are discussed and related to the samples microstructural features. A significant exchange bias effect, related to FeO content, is observed for as-milled sample, the effect being less pronounced upon annealing the nanograin powder.

Highlights

► Phase evolution in Fe/Fe3O4 nanopowder monitored by synchrotron X-ray diffraction. ► Fe/Fe3O4 nanopowder undergo an incomplete redox reaction with formation of FeO. ► FeO decomposes gradually into initial constituents at T up to 500 °C. ► At higher T the redox reaction is reversible, at 900 °C only FeO is observed. ► For the first time a strong exchange bias effect, related to FeO content, is observed.

Introduction

Granular systems of Fe nanoparticles embedded in iron oxide matrix have attracted a great deal of interest due to the observed magnetic exchange bias effect in such systems [1], [2], [3]. The exchange bias effect, defined as the shift of the hysteresis loop in ferromagnetic (FM)–antiferromagnetic (AFM) systems upon cooling down below the Néel temperature of the AFM phase [4], [5], has been shown to occur in granular ball-milled Fe/Fe oxide powders [1], [2]. It usually arises due to the coupling between FM and AFM grains at the interfaces, but has been revealed also in systems containing ferrimagnetic (FI) phases and disordered phases with spin-glass behaviour [3]. Therefore, the degree of structural disorder induced by the ball milling and the nanograin size distributions within the granular mixed FM/AFM powders are key factors for the exchange bias in granular systems. Mechanically alloyed mixtures of Fe and magnetite (Fe3O4) [6], and Fe and hematite (Fe2O3) [7] were shown to result in formation of wüstite (FeO) via a chemical reduction reaction occurring during the ball milling. Upon annealing FeO may decompose into the constituent powders [6]. The evolution of phase composition in such mixtures has a strong influence on the magnetic properties and the exchange bias effect, in particular and therefore needs to be strictly controlled during annealing, if the granular systems are to be used for technological applications.

We have processed a non-equimolar mixture of iron and magnetite by ball milling and we have studied the phase transformation that occurs during dynamic annealing in such systems via a unique tool of in situ characterization: temperature dependent X-ray diffraction of synchrotron radiation. Magnetic behaviour of as-milled and annealed samples are presented and interpreted in correlation with the phase composition in the samples.

Section snippets

Experimental

The Fe/Fe3O4 mixed powder (2:8 mass%) has been prepared by ball milling in a Retzsch PM 400 planetary ball mill with 4 vials. The total amount of the sample was 10 g, and the powders used were of high purity (99.99+%). The experimental conditions were: sun wheel frequency 200 rpm, sun wheel/vial frequency ratio 1.5, the loading constant 8; ball size 20 mm, ball-to-powder mass ratio 1/40 and number of tungsten carbide balls used was 8. The powder and balls were sealed together with hexane as

Structural phase transformation

In Fig. 1 we present the synchrotron XRD patterns taken at several, selected temperatures, from 45 °C to 450 °C. The pattern at 45 °C presents broad peaks, typical for nanostructured materials. We are able to identify the main Bragg reflections of the α-Fe, FeO (wüstite) and Fe3O4 (magnetite). As the initial mixed powders were α-Fe and Fe3O4 the observation of the Bragg reflections of the wüstite is the proof of an un-complete redox reaction that develops during the milling, in agreement with

Conclusions

The temperature evolution of the phase structure of a mixture of Fe and Fe3O4 (magnetite) powders, obtained by ball milling for 50 h at room temperature, has been studied by energy-dispersive in situ X-ray diffraction of synchrotron radiation. The nanocomposite powder is proven to undergo during preparation an incomplete redox reaction with formation of FeO. This reaction is favored due to the high energy developed during the milling and alloying. Two different stages of phase transformation are

Acknowledgements

The authors wish to acknowledge the help of Radu Nicula and Christian Lathe from MAX 80 beamline at HASYLAB during the XRD measurements. The financial support in the frame of the Marie Curie Host Development project HPMD-CT-2001-00089 as well as from Romanian Ministry of Education and Research via PN II project 12-129 is gratefully acknowledged.

References (20)

  • J. Nogués et al.

    J. Magn. Magn. Mater.

    (1999)
  • J. Nogués et al.

    Phys. Rep.

    (2005)
  • J. Ding et al.

    Scripta Mater.

    (1996)
  • J. Ding et al.

    J. Magn. Magn. Mater.

    (1998)
  • M. Seqqat et al.

    J. Magn. Magn. Mater.

    (1996)
  • D. Predoi et al.

    Thin Solid Films

    (2007)
  • O. Crisan et al.

    Sens. Actuators, A

    (2003)
  • E. Bonetti et al.

    J. Appl. Phys.

    (2001)
  • L. del Bianco et al.

    Phys. Rev. B

    (2002)
  • L. del Bianco et al.

    Phys. Rev. B

    (2004)
There are more references available in the full text version of this article.

Cited by (39)

  • Core-shell Fe@Fe<inf>x</inf>O<inf>y</inf> nanoring system: A versatile platform for biomedical applications

    2022, Materials and Design
    Citation Excerpt :

    The dimensions were chosen to ensure an aspect ratio (i.e., height/ external diameter) close to 1 to prevent the adverse effects of a high degree of asymmetry in the morphology [31,32]. The volume of the nanorings decreased as the unit cell volume decreased (black line), because the unit cell volume of the metallic Fe is smaller than that of wustite, which, in turn, is smaller than that of magnetite/maghemite [48]. Therefore, increasing the amount of metallic Fe on the nanorings while ensuring that their morphology remains unchanged decreases the nanoring volume.

  • Nanoarchitecture of advanced core-shell zero-valent iron particles with controlled reactivity for contaminant removal

    2018, Chemical Engineering Journal
    Citation Excerpt :

    Moreover, in conjunction with samples nZVI_4 (64.3 nm), nZVI_5 (64.8 nm), and nZVI_7 (56.7 nm), we observed the tendency of the MCL to decrease as the shell thickness increased (slight deviations from this trend are within experimental error; see Table 1). The cell parameter 0.2867 nm determined for all samples was in agreement with the reported values for α-Fe [55]. The XRD patterns of the samples with the shell contained broad diffraction lines belonging to 220 and 311 structural planes of magnetite/maghemite structure (see inset in Fig. 3), the intensity of which grew as the shell thickness increased.

View all citing articles on Scopus
View full text