Issue
Eur. Phys. J. Appl. Phys.
Volume 98, 2023
Special Issue on ‘Amorphous alloys and multiscale materials: Fundamental aspects and Energy applications’, edited by Zhao Zhankui, Wang Hongli and Tai Cheuk-Wai
Article Number 11
Number of page(s) 6
Section Surfaces and Interfaces
DOI https://doi.org/10.1051/epjap/2023220307
Published online 07 February 2023

© EDP Sciences, 2023

1 Introduction

Fe-based soft magnetic amorphous alloys, especially Fe-Si-B(C) system, are widely used in energy-economic transformer systems, as they show excellent soft magnetic properties [1,2]. For example, the no-load loss of the transformer made of Fe-Si-B amorphous alloy is about 70% lower than that of silicon steel (referring to the power loss measured at the primary when the secondary of the transformer is open circuit) [3]. Nevertheless, soft magnetic amorphous alloys are mostly in service in the ribbon shape, which are obtained by melt spinning method. This spin-quenching technology produces a temperature gradient between surface and core of the alloys, which consequentially generates internal stress in the process of their inherent rapid solidification (105–106 K/s). The internal stress leads to magnetic anisotropy that impairs good soft magnetic properties in Fe-based amorphous alloys [4].

Amorphous alloys produced by fast melt spinning inevitably typify a natural surface roughness (Δδr) due to the roller surface roughness, inhomogeneous solidification process and airflow environment [5]. This surface irregularities caused by the roughness act as pinning centres for domain walls, hindering the motion of magnetic domains through ribbon [6], and coercivity (Hc) is directly proportional to the relative roughness (Δδr/δ, δ average ribbon thickness) [7]. Furthermore, as the saturation flux density (Bs) of currently mass-produced Fe-based soft magnetic amorphous alloys is 1.56 and 1.64 T, such as Metglas® 2605 alloy series, considerably lower than those of oriented Fe-3%Si (2.03 T) and Fe-6.5%Si (1.79 T) steel [8]. Therefore, many studies have done to pursue high Bs amorphous alloy toward the value of silicon steel. Designing alloy components with high Fe content is a major approach to high Bs for Fe-based amorphous alloy, as Bs radically depends on the Fe content of amorphous alloy. However, in the absence of other large transition metal elements, there is a strict upper limit of Fe content for the formation of a single amorphous structure with high Bs [9]. The compositions of developed high Fe-content alloys are usually close to amorphous forming boundary, leading to partial crystallization in the as-quenched state, commonly at the alloy surfaces. The surface crystallization of amorphous ribbons occurs at a routine cooling rate (industrially acceptable spinning rate, 20–30 m/s) and even at the high spinning speed [10]. In Fe-based alloys, from the structural point of view, the surface crystallization with an oriented texture results in an in-plane magnetic anisotropy near the surface; and from the point of view of stress distribution, it results in an out-of-plane magnetic anisotropy in the amorphous bulk, originating from a compressive stress from the formation of higher-density crystalline surface layers [11]. Hence, these magnetic anisotropies significantly affect the soft magnetic properties [12].

The surface states of amorphous alloys have important effects on their soft magnetic properties. The influence of each of the major factors constituting the surface state on the soft magnetic properties has been thoroughly investigated. However, in fact, these factors often work together to affect the soft magnetic properties which is very complex and the underlying origin of which is important for improving the magnetic properties, but rarely reported. Since the existence of these factors will undesirably cause adverse effects on soft magnetic properties, it is curious whether the surface treatment to remove the surface layer can eliminate this adverse effect. In this paper, with this curiosity, we investigated the effects of surface state on magnetic properties and thermal stability for completely amorphous and amorphous with partial crystallization at their surface by layer polishing removal method. The purpose is to comprehensively evaluate the effects of polishing treatment and heat treatment on the soft magnetic properties of amorphous alloys with different surface states, and to explore the related mechanism.

2 Experimental

The alloys with nominal composition of Fe82Si1B12C5 and Fe83Si1B12C4 (at.%), designated as “Fe82” and “Fe83” alloy hereafter, were prepared induction melting with pure elementsof Fe (99.99 wt.%), Si (99.99 wt.%), B (99.9 wt.%), C (99.99  wt%) in a high vacuum Ar atmosphere. As shown in Figure 1a, the synthesis of Fe-based amorphous alloy includes two main processes, i.e., master alloys melting and rapid quenching. The wide and the thickness of the ribbons are about 1 mm and 23 µm with the spinning rate of 40 m/s, respectively. Figures 1b and 1c illustrate the stress model of amorphous alloy and amorphous alloy with surface crystallization, respectively. The alloys were mechanical polished by glass paper (2000#) from the free surface to center. The polishing subtly proceeds along the direction from free surface to center. Starting from the initial thickness, samples were taken for every surface thickness of 1 μm removed. The thickness was quantitatively measured by screw-thread micrometer and electronic balance.

The structure was detected by X-ray diffraction (XRD: Bruker, Advanced D8) with Cu- radiation. Thermal analysis of alloy was performed by differential scanning calorimetry (DSC: NETZSCH 404F3) with heating rate of 0.667 °C/s. The coercivity (Hc) was measured by B-H loop tracer (Linkjion, MATS-2010SD) under 800 A/m.

thumbnail Fig. 1

(a) Schematic diagram of the preparation process of Fe-based amorphous alloy, (b) and (c) schematic diagram of the stress models of completely amorphous ribbon and amorphous alloy with partial crystallization at surface.

3 Results and discussion

3.1 Influence of surface polishing on the thermal stability

The free surface is more prone to crystallization than roller surface, since the cooling rate at the free surface is lower than that at the roller surface attached to the cooling roller [1315]. Therefore, the free surfaces were selected for ascertainment of the microstructure. As shown in Figure 2, the free surface of as-quenched Fe82 alloy shows only a board peak around 45°, while a sharp peak appears in about 45° for as-quenched Fe83 alloy. This confirms the formation of a completely amorphous structure for as-quenched Fe82 alloys and a partial crystallization of precipitation of α-Fe (110) phase for Fe83 alloy at free surface. After the removal of surface layer with thick of 1 µm, Fe83 alloy exposes a complete amorphous structure, implying the thickness of the surface crystalline layer (Δδc) for this kind of alloys is usually small relative to their initial thickness of ribbon.

In order to reveal the influence of the surface state on the properties, the thickness range (0–4 μm) of the gradually removed surface of alloys is beyond that (1–2 μm) of the normal crystallization layer. The thermal stability and crystallization behavior during the heating process were analyzed by DSC. In Figure 3a, there are two exothermic peaks in all the DSC curves, indicating that the crystallization in these alloys occurs through two steps for both Fe82 and Fe83 alloys. The α-Fe phase begins to precipitate at the first onset crystallization temperature (Tx1), followed by Fe(B, C) compound in the secondary crystallization temperature (Tx2) [16,17]. In addition, Tx1 and Tx2 of Fe83 alloy shift to the low temperature compared to those of Fe82 alloy, due to easy precipitation of α-Fe and formation of Fe(B, C) compound in Fe83 alloy with higher Fe content. This indicates that the thermal stability of Fe83 alloys is lower than Fe82 alloy. However, Tx1 remains unchanged (419.3 and 403.2 °C) with thickness for each alloy. According to the area of the first crystallization peak, the onset crystallization enthalpy (ΔH1) of the two alloys is calculated. In Figure 3b, ΔH1 is shown as a function of the thickness removed by polishing (Δδp). The ΔH1 of Fe82 alloys gradually decrease with thickness decreasing, implying the reduction in their thermal stability. The main factor associated with this phenomenon is the alteration of internal stress state in as-quenched amorphous ribbons. In the very beginning of a quenching process, the surface cools more quickly and shrinks more severely than the core, and thus forming an internal stress distribution of tensile stresses at the surface and compressive stresses in the core. But soon the core with large volume shrinks faster than the surface with the temperature difference decreasing between the surface and the core, which causes the state of stress distribution to be reversed, i.e., the surface under compression and the core under tension [18]. If a uniform layer is removed from a surface of alloy, the constraint of compressive stresses on the core reduces. Meanwhile, the equilibrium of internal stresses is broken, and a new equilibrium will be established through an elastical deformation of the remainder of ribbon. Thus, the tensile stress distribution in the dominant core is strengthened across the remainder of the ribbon, resultantly increasing the total internal stress. Besides, mechanical polishing also induces plane stresses at a new surface of the remainder of ribbon. Furthermore, after the alloy is thrown out of the copper roller surface, the temperature of the alloy is about 200–300 °C, and the surface heat of alloys dissipates faster than the that at the core. These results in a more disordered and homogeneous structure and thus a larger ΔH1 at the surface than those at the core for the as-quenched Fe82 alloys without crystallization. This is consistent with the corresponding stress distribution described above [4]. Therefore, the enhance of internal stress caused by the polishing removal of surface stress constraint, the polishing-introduced stress and the removal of more homogeneous amorphous surfaces together are responsible for promoting crystallization of amorphous alloys. This also quantitatively explains the fact that ΔH1 decreases linearly with Δδp increasing. ΔH1 of as-quenched Fe83 alloy is generally higher than that of Fe82 alloy. ΔH1 of as-quenched Fe82 alloy is generally higher than that of Fe83 alloy. It is worth noting that an increase of only 1.2% in Fe content causes increase of about 8.4% in ΔH1 for the alloy system with an initial casting thickness (Δδp = 0). This indicates that Fe-based amorphous alloys with high Fe content close to boundary of amorphous forming ability are easier to crystallize and can precipitate more initial crystalline phases when heated. After the removal of surface crystallization layer (Δδc = 1 μm) of as-quenched Fe83 ribbons, ΔH1 of the amorphous bulk increases abnormally. This is because the removed crystalline part in the sample for DSC does not contribute ΔH1. Whereas, the percentage increase in ΔH1 (∼1.23%) caused by the small removal of surface crystallization layer (Δδc = 1 μm) is significantly smaller than the corresponding percentage of volume reduction (∼4.35%), which is attributed to the fact that the precipitated α-Fe phase in the components of the surface crystallized layer is very small relative to the amorphous matrix. Additionally, the precipitation α-Fe phase promotes B and Si to enrich in the amorphous bulk near the crystalline layer, which can improve the amorphous forming ability of this part and thus increase its amorphicity. This also favors increasing ΔH1 proceeding to polishing thinning after removing the crystalline layer, the change in ΔH1 with Δδp for as-quenched Fe83 alloy shows the same trend as that of as-quenched Fe82 ribbon, i.e., linearly decreasing, which can be reasonably explained by the same mechanism as that for quenched Fe82 ribbon.

thumbnail Fig. 2

XRD patterns of as-quenched Fe82 and Fe83 ribbons, and Fe83 ribbon after the removal of surface layer with a thick of 1 µm.

thumbnail Fig. 3

(a) DSC traces and (b) corresponding ΔH1 as a function of Δδp for as-quenched Fe82 and Fe83 ribbons.

3.2 Improvement of soft magnetic properties by annealing

Figures 4a and 4b are reciprocal thickness (1/δ) dependence of Hc. In the case of as-cast state, Hc of as-quenched alloys linear increases with the decreasing of the 1/δ for both Fe82 and Fe83 alloys. The results are consistent with previous studies in thinning Fe-based amorphous alloys and the relationship between Hc and δ could be express as [19]

(1)

where fs is the surface pinning force per unit length, fv is the volume pinning force per unit area of wall, Ms is the spontaneous magnetization, δ is the thickness of the alloy. The formula comprehensively reflects the influence of internal stress and thickness of ribbon on Hc. In addition to the factor of thickness, as the thickness removed by polishing increases, the tensile stress distribution dominated by the core is strengthened across the remainder of the ribbon and the polishing treatments also introduce stresses, resultantly increasing the total internal stress (fs) and thus increasing the effect of pinning magnetic domain. This intensifies the in-plane magnetic anisotropy near the surface and the out-of-plane magnetic anisotropy in the amorphous bulk, resulting in the increase in Hc.

The surface states of amorphous ribbons, including the internal stress and the partial crystallization at surface, cause magnetic anisotropy to impair soft magnetic properties [20,21]. It has been desirable to eliminate the adverse effects of these factors from the point of view of general knowledge, which is important to exert the intrinsic soft magnetic properties of Fe-based amorphous alloy. Annealing is considered as an effective technological means. Here, the effect of annealing on Δδp dependence of Hc is given experimentally. Comparing Figures 4a and 4b, Hc of both annealed Fe82 and Fe83 ribbons is roughly halved relative to as-quenched state, the reason of which has been proved by extensive studies to be the elimination or reduction of internal stress through annealing [22]. It is noted that Hc of annealed Fe83 ribbon is improved more greatly than annealed Fe82 alloy, even lower than that of Fe82 ribbon when Δδp = 0. In the as-quenched state (Fig. 2), only a small peak of α-Fe is observed, indicating the initial formation of a number of extremely small and heterogeneous α-Fe grains and clusters, embedded in an amorphous matrix at the surface of as-quenched Fe83 alloy. This heterogeneous structure generates an in-plane magnetic anisotropy at the surface. In the annealed state, as shown in Figure 4c, the peak of α-Fe is intensified, forming homogeneous α-Fe nanocrystalline grains (12–28 nm). The detailed clarification of crystallization behavior for the present Fe83 alloy is the work we will report soon, which is beyond the objective of this paper. This homogeneous nanocrystalline structure at the surface of as-quenched Fe83 ribbon, together with the effect of annealing to remove stress, is beneficial to improving the surface state and thus reducing the in-plane magnetic anisotropy at the surface. Therefore, annealing makes Hc of Fe83 ribbon decrease more significantly.

thumbnail Fig. 4

Hc as a function of 1/δ for (a) as-quenched and (b) annealed Fe82 and Fe83 amorphous ribbons, and (c) XRD patterns of annealed Fe82 and Fe83 amorphous ribbons.

4 Conclusion

Influences of surface state on thermal stability and magnetic properties for Fe82Si1B12C5 completely amorphous ribbons and Fe83Si1B12C4 amorphous ribbons with partial crystallization at each free surface have been investigated by polishing removal method. The conclusions are summarized as following:

  • the polishing removal of surface layer makes ΔH1 of the two kinds of Fe-based amorphous ribbons decrease obviously. This is due to the combined action of the intension of internal stress caused by the polishing removal of surface stress constraint, the polishing-introduced stress and the removal of more homogeneous amorphous surfaces together are responsible for promoting crystallization of amorphous alloys.

  • an increase of only 1.2% in Fe content causes an increase of 8.4 % in ΔH1 for the alloy system with an initial casting thickness (Δδp = 0). This indicates that Fe-based amorphous ribbons with high Fe content close to amorphous forming boundary are easier to crystallize and can precipitate more crystalline phases when heated.

  • the internal stress and the polishing-introduced stress magnetic anisotropy enhance the effect of pinning magnetic domain, thus increasing Hc and impairing soft magnetic properties.

  • annealing makes Hc of both annealed Fe82 and Fe83 ribbons with different Δδp decrease significantly, which is attributed to the elimination or reduction of internal stress.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

This work was supported by the National Natural Science Foundation of China (Grant No. 52101203 and U1704159).

Author contribution statement

Suo Zhang proposed the idea, designed the investigation procedures, and wrote the manuscript. Wenqiang Li and Chengfu Han conducted the experiments including XRD, DSC and tests for coercivity. Shaojie Wu, Chen Chen, Ran Wei and Tan Wang contributed to the data analysis of DSC and coercivity. Fushan Li supported theoretical instruction, data analysis and manuscript modification. All the authors contributed to the discussion of this work.

References

  1. R. Hasegawa, Applications of amorphous magnetic alloys, Mater. Sci. Eng., A 375–377, 90 (2004) [Google Scholar]
  2. N. DeCristofaro, Amorphous metals in electric-power distribution applications, MRS Bull. 23, 50 (1998) [Google Scholar]
  3. R. Hasegawa, D. Azuma, Impacts of amorphous metal-based transformers on energy efficiency and environment, J. Magn. Magn. Mater. 320, 2451 (2008) [Google Scholar]
  4. M. Tejedor, J.A. García, J. Carrizo, L. Elbaile, J.D. Santos, Stress relief and magnetic properties of magnetostrictive Fe79B16Si5 amorphous magnetic ribbons, J. Magn. Magn. Mater. 202, 485 (1999) [Google Scholar]
  5. S. Tsukahara, T. Satoh, T. Tsushima, Magnetic anisotropy distribution near the surface of amorphous ribbons, IEEE Trans. Magn. 14, 1022 (1978) [CrossRef] [Google Scholar]
  6. B. Groger, W. Beck, X.Z. Dong, N. Moser, H. Kronmuller, Analysis of the coercivity of amorphous ferromagnetic alloys, J. Magn. Magn. Mater. 26, 264 (1982) [Google Scholar]
  7. H. Kronmüller, Theory of the coercive field in amorphous ferromagnetic alloys, J. Magn. Magn. Mater. 24, 159 (1981) [Google Scholar]
  8. Y. Ogawa, M. Naoe, Y. Yoshizawa, R. Hasegawa, Magnetic properties of high Fe-based amorphous material, J. Magn. Magn. Mater. 304, e675 (2006) [Google Scholar]
  9. L.X. Shi, K.F. Yao, Composition design for Fe-based soft magnetic amorphous and nanocrystalline alloys with high Fe content, Mater. Des. 189, 108511 (2020) [Google Scholar]
  10. J. Pang, A.D. Wang, S.Q. Yue, F.Y. Kong, K.Q. Qiu, C.T. Chang, X.M. Wang, C.T. Liu, Fluxing purification and its effect on magnetic properties of high-Bs FeBPSiC amorphous alloy, J. Magn. Magn. Mater. 433, 35 (2017) [Google Scholar]
  11. G. Herzer, H.R. Hilzinger, Surface crystallisation and magnetic properties in amorphous iron rich alloys, J. Magn. Magn. Mater. 62, 143 (1986) [Google Scholar]
  12. E. Lopatina, I. Soldatov, V. Budinsky, M. Marsilius, L. Schultz, G. Herzer, R. Schäfe, Surface crystallization and magnetic properties of Fe84.3Cu0.7Si4B8P3 soft magnetic ribbons, Acta Mater. 96, 10 (2015) [CrossRef] [Google Scholar]
  13. F.P. Wan, T. Liu, F.Y. Kong, A.D. Wang, M.Q. Tian, J.C. Song, J.H. Zhang, C.T. Chang, X.M. Wang, Surface crystallization and magnetic properties of FeCuSiBNbMo melt-spun nanocrystalline alloys, Mater. Res. Bull. 96, 275 (2017) [Google Scholar]
  14. J.G. Wang, H. Zhao, C.X. Xie, C.T. Chang, S.M. Zhou, J.Q. Feng, J.T. Huo, W.H. Li, In-situ synthesis of nanocrystalline soft magnetic Fe-Ni-Si-B alloy, J. Alloys Compd. 790, 524 (2019) [CrossRef] [Google Scholar]
  15. S. Zhang, W.Z. Zhu, C.H. Han, W.Q. Li, T. Wang, C. Chen, R. Wei, S.J. Wu, F.S. Li, Effect of melt thermal history on magnetic properties and thermal stability in Fe-Si-B-C amorphous alloy, Intermetallics 151, 107741 (2022) [CrossRef] [Google Scholar]
  16. S. Zhang, Y.Z. Fan, Y.L. Li, W.Q. Li, H.J. Yi, H.L. Li, T. Wang, J.-Q. Wang, F.S. Li, Revealing the effects of melt temperature on structure and magnetic properties in Fe-Si-B-C alloy, Intermetallics 141, 107395 (2022) [CrossRef] [Google Scholar]
  17. A. Inoue, X.M. Wang, Bulk amorphous FC20 (Fe-C-Si) alloys with small amounts of B and their crystallized structure and mechanical properties, Acta Mater. 48, 1383 (2000) [CrossRef] [Google Scholar]
  18. M. Tejedor, J.A. Garcia, J. Carrizo, L. Elbaile, Mechanical determination of internal stresses in as-quenched magnetic amorphous metallic ribbons, J. Mater. Sci. 32, 2337 (1997) [Google Scholar]
  19. J.J. Becker, Effects of surface treatment and thinning on magnetic properties of rapidly quenched amorphous alloy ribbons, J. Appl. Phys. 52, 1905 (1981) [CrossRef] [Google Scholar]
  20. G. Herzer, Grain size dependence of coercivity and permeability in nanocrystalline ferromagnets, IEEE Trans. Magn. 26, 1397 (1990) [CrossRef] [Google Scholar]
  21. H.N. Ok', A.H. Morrish, Origin of the perpendicular anisotropy in amorphous Fe82B12Si6 ribbons, Phys. Rev. B 23, 2257 (1981) [Google Scholar]
  22. F.E. Luborsky, J.J. Becker, R.O. McCary, Magnetic annealing of amorphous alloys, IEEE Trans. Magn. 11, 1644 (1975) [CrossRef] [Google Scholar]

Cite this article as: Suo Zhang, Wenqiang Li, Chengfu Han, Shaojie Wu, Chen Chen, Ran Wei, Tan Wang, Fushan Li, Influences of surface state on thermal stability and magnetic properties of Fe-Si-B-C amorphous alloy, Eur. Phys. J. Appl. Phys. 98, 11 (2023)

All Figures

thumbnail Fig. 1

(a) Schematic diagram of the preparation process of Fe-based amorphous alloy, (b) and (c) schematic diagram of the stress models of completely amorphous ribbon and amorphous alloy with partial crystallization at surface.

In the text
thumbnail Fig. 2

XRD patterns of as-quenched Fe82 and Fe83 ribbons, and Fe83 ribbon after the removal of surface layer with a thick of 1 µm.

In the text
thumbnail Fig. 3

(a) DSC traces and (b) corresponding ΔH1 as a function of Δδp for as-quenched Fe82 and Fe83 ribbons.

In the text
thumbnail Fig. 4

Hc as a function of 1/δ for (a) as-quenched and (b) annealed Fe82 and Fe83 amorphous ribbons, and (c) XRD patterns of annealed Fe82 and Fe83 amorphous ribbons.

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.