Nanocomposite Permanent Magnets

Jeff Shield

Permanent magnet materials continue to be one of the most technologically diverse classes of materials, with applications ranging from motors and generators to traveling wave tubes in industries just as diverse—automotive, consumer electronics, and the petrochemical industry, to name just a few. The figure of merit that describes permanent magnets is the maximum energy product. The energy product defines the amount of energy that a given permanent magnet can produce, and is the largest product between the induction and magnetic field in the second quadrant of the hysteresis loop. During the twentieth century, the development of new materials and processes led to dramatic improvements in the maximum energy product, doubling every 15 years (Figure 1).

energy product

Figure 1. The maximum energy product has increased exponentially over the past 100 years. This figure shows the increase for both isotropic and anisotropic materials.

The development of higher energy products is complicated by the necessity of both a high coercivity and a high magnetization. Obtaining a high coercivity usually requires dilution of the primary magnetic species (Fe or Co), which lowers the magnetization and thus the achievable maximum energy product. The theoretical maximum energy product, then, for the current best material—Nd2Fe14B—is 64 MGOe for anisotropic materials and 16 MGOe for isotropic materials.

To overcome this limitation, and to push the maximum energy products toward 100 MGOe, researchers began taking advantage of intergranular exchange interactions, ultimately forming two-phase structures combining high-magnetization, low coercivity phases with high coercivity, lower magnetization phases. In order to maximize the exchange interactions, this approach requires the phases to be assembled at the nanoscale. While this approach has led to materials with higher energy products compared to single-phase materials, the properties achieved thus far have been less than projected.

It turns out that the performance of these nanocomposite permanent magnets is limited largely by the demagnetization process. Normally, magnetization reversal in materials with small grains is controlled by nucleation of reverse domains, and ideally each grain or particle reverses its magnetization direction independent of other grains or particles. In highly interacting systems, however, magnetic domains involving tens, hundreds or thousands of individual grains or particles develop, with the domain boundaries following along grain boundaries (termed “interaction domains”). In these highly interacting materials, the magnetic reversal is controlled by expansion of the interaction domains and the movement of domain walls. Unfortunately, in the nanocomposite permanent magnets the motion of domain walls is much easier than the nucleation of reverse domains, and there is a dramatic loss of coercivity that limits the energy product.

A possible solution to this dilemma is to eliminate the development of interaction domains. Then, magnetic reversal could again be controlled by nucleation events, resulting in high coercivity. The approach that we took to eliminate interaction domains was to make Fe-Pt nanoclusters isolated from each other in a non-magnetic matrix. The nanoclusters were designed to have a composition in the two-phase region involving the soft magnetic Fe3Pt that provides the high magnetization and hard magnetic FePt that provides the high coercivity. The clusters were produced by inert gas condensation, a highly tunable process that allows monodispersed, sub-10 nm clusters (Figure 2).


Figure 2. Inert-gas condensation very effectively produces monodispersed, nanoscale clusters. Shown here are Fe clusters, imaged using atomic force microscopy. The ratio between the standard deviation to the average size is a good measure of uniformity; this ratio for clusters produced by inert gas condensation can be less than 0.1, which is considered monodispersed.

The Fe-Pt alloy clusters form in the solid-solution face-centered cubic phase, and therefore required heat treatment to force the formation of the appropriate phases. X-ray diffraction after heat treatment revealed the presence of both Fe3Pt and FePt, and that the percentage of soft magnetic phase was more than 50 volume percent. This is in comparison to the ~20 percent soft phase content common in previous nanocomposite permanent magnets. Transmission electron microscopy of samples after heat treatment (Figure 3) revealed some agglomeration and Ostwald ripening, but also that the scale of the Fe-Pt regions remained on the order of 10 nm. Internal features were also consistent with dissolution to the Fe3Pt and FePt phases. The scale of the system—less than 10 nm—ensures that the dimension of the phases, particularly Fe3Pt, enables excellent exchange coupling. Hysteresis loops of the two-phase structures revealed a dramatic increase in remanence compared to single-phase FePt, although a lower coercivity was also observed (Figure 4).

Fe Pt clusters

Figure 3. Transmission electron microscopy image of the heat-treated Fe-Pt clusters. Even after heat treatment, Fe-Pt regions remain on the order of 10 nm, and internal features show possible dissolution into the Fe3Pt and FePt phases.

hysteresis loops

Figure 4. Hysteresis loops of single-phase (Fe58Pt42) and two-phase (Fe67Pt33) Fe-Pt clusters after heat treatment.

However, the energy product is most sensitive to remanence in these hard magnetic materials, and the two-phase nanocomposite had a significantly higher energy product. The maximum energy product for two-phase clusters was 25.5 MGOe. For comparison, single-phase FePt clusters were found to have an energy product of 12 MGOe, very close to the theoretical value of 12.4 MGOe for single-phase FePt. The values are more than 25 percent higher than that previously achieved for isotropic Fe-Pt-based permanent magnets.

The dramatic improvement in the maximum energy product achieved here shows that the early projections of very high energy products for exchange-coupled permanent magnets can be realized. The key will be to construct materials and systems that have nanoscale features to ensure effective exchange coupling but that have particles or grains that reverse independently of one another.