NANOCRYSTALLINE SOFT MAGNETIC ALLOYS FOR APPLICATION IN ELECTRICAL AND ELECTRONIC DEVICES

V.R. Ramanan
ABB-Electric Systems Technology Institute
1021 Main Campus Drive
Raleigh, NC 27606

Introduction

Nanocrystalline soft magnetic materials are derived from crystallizing amorphous ribbons of specific families of (Fe,B)-based alloy chemistries. This new class of materials is characterized by 10-25 nm sized grains of a (bcc) (-(Fe,X) phase consuming 70-80% of the total volume, homogeneously dispersed in an amorphous matrix. The Japanese have pioneered and developed the work on these materials, especially in so far as the impact on major industrial applications is concerned.

As a result of research to date, two families of alloys show the best performance characteristics and have emerged as the leading candidates for reduction to application: Fe-Cu-Nb-B-Si (the "Finemet" family) and Fe-Zr-(Cu)-B-(Si) (the "Nanoperm" family). The Finemet family is characterized by an optimum grain size of about 15 nm, provides a saturation induction of about 1.2 T, and exhibits very good properties at high frequencies, comparable to some of the best Co-based amorphous materials. On the other hand, the grain sizes consistent with optimum performance are larger, around about 25 nm, in the Nanoperm family. The distinguishing feature of the Nanoperm family of alloys is the very low energy loss exhibited at low frequencies (60 Hz), offering the potential for application in electrical power distribution transformers.

The nanocrystalline materials are obtained by crystallizing precursors cast as amorphous alloy ribbons. The amorphous alloys typically crystallize in two stages: a magnetically desirable bcc-(Fe,X) phase appears first, followed by a boride phase, the presence of which is deleterious to good, soft magnetic behavior. In the optimized chemistries, the separation between the two crystallization events is large (~150 K), so that crystallizing heat treatments may be conducted above the temperature for the first event, while safely avoiding the onset of the other.

The excellent soft magnetic properties noted from these materials -- extremely low coercivities, high permeabilities, low energy losses, etc. -- has triggered major interest and research activity in both the academic/research community and the industrial community. Additionally, these Fe-based materials have potential as lower cost alternatives to the costly Co-based amorphous materials in many applications. Application of these materials in cores of transformers in switched-mode power supplies, chokes, ISDNs, etc., are already well on their way to commercialization. Other applications in high accuracy current transformers, ground fault interrupters, etc., are being investigated. The applications of these materials that have the greatest cost and performance impact, and yet are the most challenging, is in cores of transformers for electrical power distribution. This is the application that needs to be facilitated through additional work on nanocrystalline soft magnetic alloys.

Why Should These Materials be Studied and What are the Necessary Studies?

Alloy-Chemistry-Related Issues

The soft ferromagnetic behavior in these materials arises from a spatial averaging of the magnetic anisotropy of the aggregate of randomly oriented nanosized particles. Concurrently, the saturation magnetostriction of the materials is also reduced to near zero values.

Consequent to the low, averaged anisotropy, the ferromagnetic exchange length in these materials is estimated to be about 35 nm. In the Finemet family, the nanosized phase is (-(Fe,Si), with about 18 at % Si and possibly some Nb. In the Nanoperm family, an (-(Fe, Zr) phase is obtained, with the Zr content not well established and possibly with some B. It is now widely accepted that the ferromagnetic grain boundary (amorphous) phase facilitates the coupling between the nanosized bcc grains. Absent this phase, the considerably weaker magnetic bcc nanophase particles are decoupled, leading to rather poor magnetic characteristics in the material.

Clearly, the optimized grain size and the fractional volume coverage of the nanosized phase (which, in turn, is related to the width of the grain boundary region) intimately govern the efficiency of the spatial averaging of the anisotropy essential for the soft magnetic behavior in these materials.

There is evidence to indicate that the presence of Cu aids in profuse nucleation, and the heavier elements such as Nb and Zr act as growth inhibitors. The precise mechanism of this nanocrystallization behavior is not yet well established. In addition, the reason that Cu is essential only in Nb containing alloys is not explained.

Processing-Related Issues

Rapid solidification using melt spinning is the technique employed to prepare the precursor amorphous alloys. Without question, this should be the chosen route for manufacturing large quantities of nanocrystalline alloys. However, even though large quantities (about 3 metric tons per hour in continuous production) of amorphous Fe-B-Si ribbons are now produced commercially, the state of knowledge is still in its infancy about cost-efficient casting of the nanocrystalline alloys and, in particular, the Nanoperm family of alloys, in commercially interesting quantities.

The Finemet family of alloys, with their good performance at high frequencies, will be primarily targeted for applications requiring a few tens of grams per core. Available state of knowledge in batch casting is expected to be sufficient for the production of commercially viable quantities of these alloys.

In contrast, large scale, continuous production of the Nanoperm family of alloys has to be addressed. There are two reasons for this: (1) This family of alloys, with demonstrated low loss characteristics at line frequencies, has a great potential for application in cores of transformers for electrical power distribution; and, (2) saturation induction levels of about 1.7 T may be attained in Fe-Zr-B nanocrystalline alloy systems, as opposed to the 1.55 T typically attained in the commercially available Fe-B-Si amorphous alloys. Given that Fe-3%Si grain-oriented electrical steel saturates at about 2 T, the larger core sizes necessitated by a lower saturation induction level have been a deterrent to a much wider use of amorphous materials in transformers. The promise of the combination of low loss and high saturation induction available from the Nanoperm type of alloys makes them very attractive for application in large transformers. From available information, it seems safe to say that the Japanese, at best, are able to produce about 5 kg of 30 mm wide material at a time.

AlliedSignal, Inc., in the United States is the leading producer of amorphous metal products in the world and has clear technology leadership in melt spinning. There has been no interest on the part of AlliedSignal to undertake the large investments in time and resources necessary to define the manufacturing process for the Nanoperm family of alloys.

Application-Related Issues

It is well known that Fe-based amorphous materials tend to embrittle following the necessary heat treatments to optimize their performance. The material manufacturer has learned to optimize the alloys and casting processes to help alleviate this problem related to handling of the materials. The industrial end user, too, has established a large background of experience in cutting, shearing, and handling procedures to minimize breakage and loss of material during the course of processing it for the application.

The nanocrystalline alloy materials discussed here are generally more brittle and more difficult to handle than annealed amorphous alloy ribbons. In the case of small cores (such as those prepared from the Finemet family of alloys), the anneal to obtain the nanocrystalline phase, which embrittles the material, is conducted after the core has been manufactured. The processed core is then not handled. This is not a viable alternative in the case of manufacturing larger cores for distribution transformers.

As an unavoidable consequence of rapid solidification, the materials are confined to ribbon shapes with maximum thicknesses of about 25 µm. Therefore, it is expected that nanocrystalline soft magnetic materials will find their use only in wound core designs for distribution transformers covering the smaller ratings (up to 50 kVA). This, however, is still a sizable market, covering about 1 to 1.5 million transformers per year in the United States alone.

The Technical Community in the United States Must Address These Materials

Since most of the materials development has been carried out in Japan, and since related broad chemistry patents are held by Japanese firms, the U.S. companies have, in general, ignored these materials. There is no known significant activity, even in the U.S. academic community, perhaps due to the absence of supply of material from local sources. A considerable amount of research on these materials, albeit from a "pure science" perspective, is being conducted in Europe and Japan.

With the increase in the number of electronic devices in everyday lives, nanocrystalline soft materials are now finding their way into devices in the marketplace. Soon, with increased emphasis on energy conservation, these materials will have to be considered in the design of larger transformers. With a well conceived and directed effort, a U.S. source for these materials can be established, and such a source is necessary if the U.S. industry is to take full advantage of the benefits indicated in these materials.

The Finemet family of alloys is marketed by Hitachi Special Metals of Japan, Vacuumschmelze GmbH of Germany, and Imphy (a subsidiary of Usinor Sacilor) of France. They all make and sell small cores for many of the "specialty" applications at high frequencies, and others mentioned previously.

The Nanoperm alloys have been developed by Alps Electric Co. of Japan and are more recent than the Finemet family. Alps Electric has been working with Japanese end users to explore the use of nanocrystalline Fe-Zr-B alloys in various applications.

Key References

Makino et al. 1991. Materials Transactions, Japan Institute of Metals 32: 551.

_____. 1995a. Materials Transactions, Japan Institute of Metals 36: 924.

_____. 1995b. Nanostructured Materials 6: 985.

Suzuki et al. 1990. Materials Transactions, Japan Institute of Metals 31: 743.

Yoshizawa et al. 1988. J. Appl. Phys. 64: 6044.

_____. 1990. Materials Transactions, Japan Institute of Metals 31: 307.

_____. 1991. MRS Symposium Proc. 232: 183.

[Previous Section][Top of Report][Send Your Comments][WTEC Welcome Page][Next Section]

Published: January 1998; WTEC Hyper-Librarian