This article has been accepted for publication in a future issue of this journal, but has not been
fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TMAG.2016.2535482, IEEE
Transactions on Magnetics
MAGCON-15-11-1796.R1
Fabrication, Processing Technologies and New
Advances for RE-Fe-B Magnets
David N. Brown, Member, IEEE
 powder and (2) consolidating rapidly quenched powder.
Abstract— Over the last 30 years rare earth permanent
magnets have become an integral part of our lives, driving
computers, home appliances and automobiles. They are set to
become an even more prominent component of our future with
the growth in clean technologies, like electric vehicles and wind
power generation. This market growth was briefly challenged in
2011 by the threat of supply restrictions and inflated prices from
the world’s principle source of rare earths. However, a rational
calm has now returned to the market and magnet users can
continue to benefit from the high efficiency and miniaturization
these phenomenal magnets provide to electromechanical devices.
The rare earth magnet industry and research community
continues to innovate and push these magnetic materials towards
their theoretical potential, specifically in the areas of higher
magnetic performance, lower costs and greater thermal stability.
These development areas for the two principle types of RE-Fe-B
magnet, sintered and rapidly quenched, are subtly different. The
coarser microstructure of sintered magnets has made them more
reliant on the less abundant heavy rare earth elements for
magnetic performance at elevated temperatures, so development
activity on sintered RE-Fe-B magnets has focused on minimizing
this Dy and Tb dependence. Magnets from rapidly quenched RE-
Fe-B materials have fine nanostructures and developments have
targeted higher remanence, lower material costs and protecting
the material from oxidizing environments.
Index Terms— Rare earth permanent magnets, Nd-Fe-B, rapid
solidification, magnet fabrication
I. INTRODUCTION
T HERE HAVE been so many excellent reviews on rare earth
iron boride (RE-Fe-B) permanent magnet developments
[1-6] since their first announcement in 1983 [7, 8]. This article
will attempt to provide a brief overview of the types of RE-Fe-
B magnet available, their fabrication techniques, their relative
magnetic characteristics and the current development activities
in the industry.
Rare earth magnets are known for their superior magnetic
properties of high induction and coercive force. These
properties arise from the extremely high magnetocrystalline
anisotropy made possible by the unique 3d-4f spin-orbit
interactions between transition metal and rare earth
components.
There are a number of ways to fabricate RE-Fe-B magnets.
The principle manufacturing routes are (1) sintering of aligned
David Brown is a senior research scientist at the Magnequench Technology
Centre, 01-19 The Galen, 61 Science Park Rd, Singapore 117525
(dbrown@magnequench.com).
0018-9464 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
1
Sumitomo Special Metals (SSM) and General Motors (GM)
simultaneously developed Nd2Fe14B into a commercial
product in the early 1980s. GM opted to produce rapidly
quenched Nd-Fe-B for nanocrystalline net-shaped magnets,
believing that rapid solidification technology was necessary to
maintain coercivity. SSM chose to produce anisotropic
microcrystalline sintered magnets, following on from its
considerable experience with sintering SmCo magnets [9].
The way in which the two businesses developed was also
different. SSM chose to license a number of magnet makers to
manufacture sintered RE-Fe-B magnets, while GM decided to
be the sole producer of rapidly quenched RE-Fe-B powder and
allowed its customers to fabricate polymer bonded and hot
formed magnets from this powder.
Both sintered and rapidly quenched magnets contain the
same RE2Fe14B crystals. The fundamental difference between
the two types of RE-Fe-B magnet is the microstructure. The
grain size and grain orientation ultimately controls the
magnetic performance. Smaller grain sizes enhance the
magnets ability to resist demagnetization, while orientating the
RE2Fe14B grains into a preferred direction can enhance the
magnet flux density, Br. Rapidly quenched magnet materials
have finer grain structures and a higher coercivity advantage,
while sintered magnets have the grain orientation advantage
and superior Br. There are alternative RE2Fe14B magnets that
attempt to have both fine grains and grain alignment, like hot
deformed rapidly quenched magnets and anisotropic polymer
bonded magnets.
II. APPLICATIONS OF RE-FE-B MAGNETS
The needs of a magnet application or device tend to drive
the developments in materials and fabrication processes. RE-
Fe-B magnets with high flux allow motors to be designed
smaller, lighter and more efficient. Over the last 30 years they
have become part of our everyday life, as seen in computer
hard drives, automotive accessories and office automation.
However, these magnets are set to become an even greater part
of our future, as they are crucial to the optimization of low-
carbon, clean technologies such as electric vehicles, wind
power generation systems and energy-saving home appliances,
as illustrated by the market growth predictions presented in
Fig. 1 [10].
Table 1 lists some typical applications for RE-Fe-B magnets
and Fig. 2 illustrate the relative amount of these magnets being
used in each application segment by mass [11, 12, 13].
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TMAG.2016.2535482, IEEE
Transactions on Magnetics
MAGCON-15-11-1796.R1 2

III. KEY PERFORMANCE INDICATORS FOR PERMANENT

MAGNETS
For a RE-Fe-B magnet to be chosen for a specific
application it needs to satisfy certain factors, which include
magnetic performance, thermal stability and cost.
A. Magnetic Performance
Magnetic performance is best illustrated by the second
quadrant demagnetization curve of a permanent magnet. Fig. 3
illustrates the key points of remanence (B r magnetic flux
density), intrinsic coercivity (Hci resistance to being
demagnetized) and energy product ((BH)max the amount of
Fig. 1. Predicted global demand for rare earth magnets [10]
energy stored within the magnet). From the general shape of
this curve it can be seen that the magnetization slowly
TABLE I decreases with an increasingly negative H field, and then it
Typical Magnet Applications abruptly falls away at about H = -12 kOe. This point on the
demagnetization curve is known as the ―knee‖ and it is
generally unwise to use a magnet where it operates at or
beyond this knee.
Br can be increased by aligning the RE2Fe14B grains in a
preferred crystallographic orientation. Sintered [14] and hot
deformed [6] magnets can achieve a high degree of grain
alignment, while isotropic magnets are limited in B r by their
very nature of having a randomly orientated crystal structure.
However, by combining the RE2Fe14B phase with a small
amount of -Fe or Fe3B within the nanostructure, can enhance
the Br of isotropic rapidly quenched magnets via the
exchange-coupling phenomenon [3, 15].

Fig. 3. Example of an anisotropic RE-Fe-B magnet second quadrant

demagnetization curve

Fig. 2. Estimated relative size of RE-Fe-B market segments
B. Thermal stability
One limitation of RE-Fe-B magnets is the operation at
elevated temperatures, >150oC. At these higher temperatures
the magnetic domains tend to relax and lose alignment,
reducing Br and Hci. Also, RE metals are susceptible to
oxidization and RE-Fe-B material can start to oxidize at
elevated temperatures in air if not suitably protected.
The Curie temperature (Tc) of Nd2Fe14B is 312oC, and the
decrease in magnetic performance with increasing temperature
can be predicted from the thermal coefficients of B r and Hci.
These parameters of  and are approximately -0.1 %/oC and

0018-9464 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TMAG.2016.2535482, IEEE
Transactions on Magnetics
MAGCON-15-11-1796.R1
-0.3 to -0.7 %/oC, respectively, meaning that Hci is most
significantly affected by working at elevated temperatures. For
example, a sintered Nd-Fe-B magnet with a room temperature
of Hci = 15 kOe and = -0.6 %/oC will lose half of its Hci by
being heated and operated at 100oC.
Oxidation resistance and Tc can be improved by substituting
cobalt for the iron component (+10oC to Tc per 1 wt% Co
substitution), but essentially RE-Fe-B magnets are limited to
operating below 200oC. Higher temperature applications tend
to use Sm-Co based magnets.
Temperature stability of magnets can be compared by
exposing samples of the same dimensions to a temperature for
a certain time, and measuring the room temperature flux
before and after the thermal exposure. The amount of flux lost
will be proportional to the degree of magnetic domain
relaxation and structural damage to the magnet sample.
For example, Fig. 4(a) relates the demagnetization curves of
three different types of polymer bonded magnet to their
relative thermal ageing behavior. The three magnets have
different Hci levels and J-H curve shapes, but were
manufactured to the same dimensions (OD=9.8 mm x H=6.4
mm) and had the same Br (magnet) / Br (powder) ratio of 78%. The
sample dimensions mean that the magnets operate with a
permanence coefficient of 2 in open circuit. This operating
point is indicated on the demagnetization curves. The relative
flux loss experienced by these magnets is illustrated in Fig.
4(b). The RE-lean, lower Hci magnet has a much higher flux
loss than the stoichiometric RE2Fe14B and RE-rich based
magnets. This poor thermal stability can be attributed to this
magnet operating at a high-gradient point on its
demagnetization curve. Ageing stability generally improves
with magnets of higher Hci, and when they operate on a low-
gradient part of the demagnetization curve.
(a)
(b)
Fig. 4. (a) Demagnetization curves and (b) ageing performance profiles of
three different types of PC=2 bonded magnets
0018-9464 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
3
C. Economics
RE-Fe-B magnets command a higher material price than the
lower performance ferrite magnet alternatives, but generally
lower than SmCo and FePt alternatives. Magnet users have
been concerned by recent fluctuations in the price and
availability of REs since a price spike in 2011. However, the
market has now returned to pre-2011 levels and the future
appears more stable [16], omitting the possibility of a
geopolitical catastrophe.
Nevertheless, RE-Fe-B magnet producers have addressed
the matter by offering more cost sensitive products (e.g.
replacing some Nd with lower cost Ce [17]) and attempting to
use less heavy rare earth elements (HREEs) by developing
technologies like grain boundary diffusion treatments. It is
these HREEs that raise most concern as they are far less
abundant than the lighter REs of Ce, La, Nd and Pr, and
therefore command a price of between 5 and 10 times that of
Nd metal.
Selectively diffusing the HREE into the grain boundaries of
fully dense magnets, via a post-sinter heat treatment, allows
the HREE to be concentrated in the grain boundary regions.
This will increase the magnet Hci without compromising the
Br, which happens if the HREE is allowed to enter the 2:14:1
phase and lower the Ms of this phase [1].
More excitingly, researchers from NIMS, Daido and Toyota
have recently used grain boundary diffusion treatments with
NdCu, NdAl and NdGa additives. They have modified the
grain boundary phase by concentrating these non-magnetic
additives in the grain boundary with an appropriate thermal
treatment, which in turn exchange-decouples the Nd2Fe14B
grains from one another and maximizes the magnet Hci
without any Dy or Tb [18, 19].
The cost of manufacturing is also a concern. There are a
number of novel consolidation techniques available to magnet
makers that offer certain technical advantages over traditional
methods, like higher density magnets (e.g. Spark Plasma
Sintering and Combustion Driven Compaction). However, the
cost of some of these more exotic processes (e.g, cycle time,
tool life…) can add too much of a premium to the final
magnet price, and it would be difficult for them to compete
with the well-established sintering and polymer bonding
fabrication routes.
IV. TYPES OF RE-FE-B MAGNET
The principle fabrication routes are illustrated in Fig. 5.
Generally, all start with an alloying step, where the constituent
elements are combined into the desired chemical composition.
This is followed by a casting stage to achieve the targeted
microstructure. Then there is a powder crushing stage,
followed by a consolidation step, a finishing stage and finally
a magnetizing step. However, there are significant differences
in the detail of each of these stages between rapidly quenched
magnets and sintered magnets.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TMAG.2016.2535482, IEEE
Transactions on Magnetics
MAGCON-15-11-1796.R1
Fig. 5. Illustration of RE-Fe-B magnet production routes
The resulting microstructures created from these four
processes are illustrated in Fig. 6. A sintered magnet has
regular shaped 5-10 m grains, magnetically orientated in a
preferred direction, providing it anisotropic magnetic
performance. The rapidly quenched ribbon microstructure has
regular shaped 20-30 nm grains with a random
crystallographic orientation, and will be magnetically
isotropic. These rapidly quenched ribbons can be hot pressed
into a fully dense compact and the grains will have an
isotropic orientation and be approximately 100 nm in
diameter. If this hot pressed magnet is then hot deformed or
extruded, its grains will grow in the orthogonal directions to
the deforming force, resulting in 500 x 50 nm platelet-like
grains which are magnetically aligned perpendicular to the
grain growth directions, and this type of magnet will have
anisotropic magnetic performance.
(a) (b)
(c) (d)
Fig. 6. RE2Fe14B grain structures in (a) a sintered magnet x2000, (b) Melt
spun powder x70000, (c) hot pressed magnet x50000 and (d) a hot deformed
magnet x30000
0018-9464 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
4
V. SINTERED RE-FE-B MAGNETS
Sintered magnets can offer extremely high anisotropic
magnetic performance as the RE2Fe14B grains are aligned in
one preferred crystallographic orientation. The theoretical B r
measured from a RE2Fe14B single crystal is ~16 kG [1].
However, if we assume there is 8 vol.% non-magnetic grain
boundary phase required, and the RE2Fe14B grains have a 95%
degree of alignment, then the actual magnet B r will be 14 kG.
This level of Br can be achieved from RE13.5(Fe,Co)balB6 type
compositions (N50 grade: Br~14 kG, Hci~13 kOe, (BH)max~50
MGOe). However, for practical applications where the
operating temperature exceeds 80oC, higher Hci will be
required. Conventionally, sintered magnets have achieved this
by substituting some Nd with HREEs. Combining Dy or Tb in
the alloy will increase the magnetocrystalline anisotropic field,
and raise the Hci of the magnet by approximately 2 kOe per 1
wt.% of Dy substituted for Nd.
Sintered magnets are marketed based on energy product and
coercivity level, which in turn can be related to their
traditional reliance on HREEs. Fig. 7 illustrates the
relationship between energy product, intrinsic coercivity,
required Dy level and the typical application for these
commercially available sintered magnets. Automotive
applications tend to operate at higher temperatures than
electronic goods, and these require higher Hci levels to
maintain magnetic performance stability at temperatures like
150oC.
Fig. 7. Sintered Magnet Grades Relative to Magnetic Performance, Dy-
content and Application [6, 12]
The reliance of sintered magnets on HREEs for Hci and high
temperature stability comes at a price. The B r is reduced as
(Nd,Dy,Tb)2Fe14B has lower magnetization saturation than
Nd2Fe14B. Also, the natural abundance of Dy and Tb is a lot
less than Nd (Dy/Nd<0.1), which results in the HREEs
commanding a higher price and having more uncertainty over
their future availability. These factors have driven sintered
magnet developments towards minimizing the amount of
HREE required. Such development activities include
producing finer grain sizes and exploring grain boundary
modification technologies.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TMAG.2016.2535482, IEEE
Transactions on Magnetics
MAGCON-15-11-1796.R1
A. Targeting finer grains in sintered magnets
RE-Fe-B magnet Hci is dependent on grain size, with Hci
being maximized when the grain size is equivalent to a single
magnetic domain [3]. This single domain size is equivalent to
about 100 nm for Nd2Fe14B. Traditional sintered RE-Fe-B
magnets are produced with grain sizes in the order of 5-10 m
(Fig. 6a). Recent developments have demonstrated that finer
grain sizes can yield higher Hci in sintered magnets, as
illustrated in Fig. 8 [20]. Several technologies have been
harnessed to increase sintered magnet Hci via finer
microstructures, minimized oxygen levels and grain boundary
phase modifications.
Fig. 8. Effect of finer grain sizes on Hci of a Nd-Fe-B sintered magnet [18]
Firstly, optimizing the strip casting stage [20, 21] can
produce alloy with a fine uniform columnar grain structure,
which can then be crushed into a fine precursor powder with
more structural and chemical uniformity than powders
prepared from traditionally cast ingots. The typical strip
casting apparatus is illustrated in Fig. 9 [22]. Strip cast alloy
appears as ~250 m thick platelets and has a fine columnar
microstructure, as shown in Fig. 10. After casting, these
platelets are crushed to ~3 m single crystal particles via jet
milling. The benefits of these casting developments can be
further enhanced by incorporating high-velocity helium gas jet
milling or HDDR technologies to further minimize particle
size [23].
Fig. 9. Strip casting process [22]
In parallel with these finer precursor powder developments,
researchers have also targeted greater crystallographic
0018-9464 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
5
alignment of the grains during the sintered magnet fabrication
process (higher Br). The Pressless-Processing technique is the
latest technology which offers a greater degree of grain
alignment than traditional axial and transverse alignment
techniques [14].
(a) (b)
Fig. 10. Appearance of (a) strip cast flakes and (b) the cross-sectional
microstructure of a 250 m thick flake
B. Grain boundary engineering for higher Hci
In addition to these finer grain-size and better
crystallographic developments, researchers have been
attempting to maximize the HREE-free Hci of sintered
magnets via grain boundary engineering. Post-sintering heat
treatments in themselves can improve magnet Hci via
smoothing of the grain boundary structure. Further Hci
enhancement can be seen with grain boundary diffusion
(GBD) treatments.
Generally, GBD involves a post-sintering thermal treatment
where a HREE-rich or RE-Cu surface coating is permeated
through the grain boundary network of the as-sintered
compact, concentrating the HREE component in the key grain
boundary areas and enhancing the barrier to magnetic domain
movement [6]. Published results have indicated that GBD
treatments can reduce the amount of HREE required, for a
certain Hci level, by at least half in sintered magnets [24].
More recently, researchers from TDK have demonstrated
that certain Ga-containing RE-Fe-B magnet compositions can
be thermally treated to develop the grain boundary phase to
achieve Hci levels of up to 25 kOe in Dy-free sintered magnet
samples [25].
VI. RAPIDLY QUENCHED POLYMER BONDED RE-FE-B
MAGNETS
The rapid quenching or melt spinning process involves
ejecting a jet of liquid alloy on to a high speed quench surface,
as illustrated in Fig. 11. The resulting product is a ~30 m
thick ribbon with an isotropic nanocrystalline RE2Fe14B
structure. This ribbon is then crushed to a desired particle size
and shape (from 5 m to 400 m flakes) and compacted into
magnets [7, 26].
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TMAG.2016.2535482, IEEE
Transactions on Magnetics
MAGCON-15-11-1796.R1 6

(a) (b)

Fig. 11. (a) Melt spinning process and (b) <440m powder product [28]

The alloy composition and grain size are the main factors
which determine the magnetic performance of a bonded
magnet. A wide range of compositions can be rapidly
quenched (RE5-15FebalB5-10), depending on the required
magnetic performance. For example, stoichiometric Nd 2Fe14B
will yield powder magnetic properties of B r = 9 kG, Hci = 10
kOe and (BH)max ~16 MGOe. If an excess of the RE
component is used (e.g. Nd2.2Fe13.8B), a RE-rich grain
boundary phase will be present and Hci will increase at the
expense of some Br. Similarly, if the RE component is reduced
in favor of the Fe component (e.g. Nd1.8Fe14.2B) then a
multiphase nanocomposite structure (Nd2Fe14B + -Fe) will
be formed and Br will increase at the expense of Hci.
Such Nd2Fe14B/-Fe nanocomposite magnet materials offer
great theoretical potential for higher B r via exchange-coupling
[3]. However, the full potential of exchange-coupled magnets
has not been achieved so far, as fabricating the ideal
microstructure for the two phase system with current
processing technologies remains a challenge. Current
commercial nanocomposite magnets do offer higher flux than
single phase magnets but their coercivity and thermal stable
remain limitations.
There is a great deal of flexibility with the type of magnet
that can be formed from melt spun powder. The powder can
be combined with a polymer and compression molded (~80%
powder Br) or injection molded (~65% powder Br) into
isotropic bonded magnets. Alternatively, the powder can be
hot pressed into a fully dense isotropic magnets or hot
deformed into anisotropic magnets [26, 27, 28]
A. Advantages of Polymer Bonded Magnets
This category of RE-Fe-B magnets tend to be isotropic in
magnetic performance (Br ~ 6-7 kG, (BH)max~ 8-12 MGOe),
and available with a range of Hci values (6-17 kOe), as
illustrated by the chart of commercial isotropic rapidly
quenched powders in Fig. 12. The powders to the right of this
chart have high Hci and are more suitable for high temperature
applications (150-180oC). Powders to the left of the chart have
high flux and are suitable easy magnetization applications that
operate up to 100oC. Powders in the lower left-hand corner
have increasing amounts of Ce substitutions and are designed
for cost sensitive applications.
Fig. 12. Range of rapidly quenched isotropic RE-Fe-B powders for bonded
magnet applications [28]
While isotropic bonded magnets do not have the high flux
of anisotropic sintered magnets they do offer a great deal of
freedom for engineering designs. These magnets can be
pressed into small and complex shapes, without fear of
reduced magnetic performance due to their unique
nanostructure. The isotropic structure also allows great
flexibility for intricate magnetization patterns. Polymer
bonded magnets are non-conductive so they do not suffer from
eddy current losses like fully dense, metallic magnets. This is
a great advantage for applications operating at high speed or
elevated temperatures.
Bonded Nd-Fe-B magnets have found a number of
applications in automotive parts, office automation, hard disc
drives, circulation pumps and home appliances where greater
flexibility in design and magnetization are required. The
general appearance of a compression molded magnet cross-
section is illustrated in Fig. 13.
(a) (b)
Fig. 13. (a) Typical -100mesh crushed melt spun powder and (b) the cross-
section of a compression molded magnet.
B. Development activity with bonded RE-Fe-B magnets
Market interest in lower cost bonded magnet materials has
led to the introduction of melt spun powders with Ce- and La-
components [17, 28]. Ce and La are more abundant than Nd,
and can replace Nd by up to 80% in Nd 2Fe14B for price-
sensitive applications that operate at low- to moderate-
temperatures. The effect of substituting Nd with Ce on the
magnetic performance of a compression bonded magnet is
illustrated in Fig. 14.

0018-9464 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TMAG.2016.2535482, IEEE
Transactions on Magnetics
MAGCON-15-11-1796.R1
Fig.14. Effect of Ce substitution on the remanence and coercivity of ((NdPr) 1-
xCex)2Fe14B bonded magnets [17].
Another request from bonded magnet users has been for
higher magnetic performance. The magnetic flux of bonded
magnets can be increased by either making the isotropic
magnet more dense or making the powder anisotropic and
aligning this powder while pressing the magnet. Such
anisotropic powders can be produced by using either hot
deformed rapidly quenched RE-Fe-B material [29] or HDDR
processed RE-Fe-B powder [30]. Typical anisotropic powder
magnetic properties are illustrated in Fig. 15 ((BH) max= 30~40
MGOe).
Fig. 15. Anisotropic RE-Fe-B powders available to be formed into bonded
magnets [28, 29]
There are practical challenges to achieving full alignment of
anisotropic powder in high-density compression molded
magnets. However, achieving full alignment is possible with
the injection molding technique, where there is more binder
material between the individual anisotropic powder particles.
Another key area of development has focused on improving
the thermal stability of bonded magnets. Generally, this
involves creating a protective barrier between the RE-Fe-B
particles and the potentially oxidizing operating environment.
The challenge to this goal is that the powder particles tend to
fracture when being compacted into a bonded magnet, as
illustrated in Fig. 16. The freshly fractured surfaces are more
susceptible to oxidation then the as-quenched surfaces.
Therefore, considerable efforts have gone into developing
protective coatings to either encapsulate the magnet or the
individual rapidly quenched powder particles [28, 30]. One
such technology is known as the anti-ageing powder coating
0018-9464 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
7
treatment, AA4. Fig. 17 illustrates the improved ageing
performance of magnets that receive the AA4 treatment [31].
Fig. 16. A melt spun RE-Fe-B flake that after being compression molded
into a polymer bonded magnet.
Fig.17. Improvement in flux ageing loss observed in samples with the AA4
treatment (180oC for 1000 hours)
VII. HOT FORMED RAPIDLY QUENCHED RE-FE-B MAGNETS
Rapidly quenched powders designed for hot forming require
compositions with an excess RE component to facilitate the
hot densification process (RE 13-15-Febal-B6). The process starts
with the powder being cold compacted into a low-density
perform. This compact is then coated with a lubricant and
pressed to full density at ~750°C within a few seconds under a
pressure of approximately 100 MPa. The product is an
isotropic magnet with extremely high Hci, considering it has a
HREE-free composition (Br = 8.3 kG, Hci = 18-22 kOe,
(BH)max ~15-18 MGOe).
If higher remenance is required then this isotropic part can
be placed in a larger die and hot die-upset, as illustrated in Fig.
18, or hot extruded into a variety of shapes including radially
aligned rings. The resulting fully dense parts have equivalent
anisotropic, high-energy magnetic performance to sintered
magnets (Br >12 kG, (BH)max >35 MGOe), but with higher Hci
as the grain size is close to the critical diameter of a single
magnetic domain for Nd2Fe14B [32].
Fig. 19 illustrates the flake microstructure of (a) hot-
pressed and (b) hot deformed magnets at x200 magnification.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TMAG.2016.2535482, IEEE
Transactions on Magnetics
MAGCON-15-11-1796.R1
The rapidly quenched flakes can be observed to align
perpendicular to the pressing-force direction in the densified
structure. The flakes in the MQ3 deformed magnet can be seen
to have become elongated, much like the RE2Fe14B grains in
Fig. 6(d).
Fig.18. Illustration of the MQ3 die-upsetting process
(a) (b)
Fig.19. An illustration of the hot pressed flakes after being (a) hot pressed,
MQ2 and (b) hot deformed, MQ3
Fig. 20 illustrates the relative magnetic performances of
isotropic hot-pressed (MQ2) and anisotropic hot die-upset
(MQ3) magnets. The degree to which the isotropic magnet is
deformed into the anisotropic magnet can be controlled, and
hence so can its magnetic properties. The die tooling
dimensions can be designed to provide a variety of degrees of
deformation, between 0 and 70%. The range of magnetic
performances possible for these various degrees of
deformation is illustrated in Fig. 21.
Fig. 20. Demagnetization curves from typical non-HREE hot pressed (MQ2)
and hot deformed (MQ3) magnets.
Fig. 21. Relationship between the degree of deformation and the magnetic
properties of melt spun magnets.
0018-9464 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
8
Recent work on hot deformed magnets has involved
targeting finer grain sizes, and exploring grain boundary
diffusion technologies [6, 18, 33, 34].
Also, novel rapid hot consolidation technologies have been
investigated (e.g. Spark Plasma Sintering). However, all new
technologies face the challenge of being economically viable.
For example, if the cost of consumables, tooling, and labor,
cycle time and alike, for a new fabrication process is higher
than for traditional techniques, and the relative improvement
in magnetic performance is nominal, then it is unlikely that the
new process will be adopted by magnet makers.
VIII. RECYCLING RE-FE-B MAGNETS
With higher RE metal prices, scarcity of HREEs and the
introduction of governmental legalization there has been
increasing interest in recycling RE-Fe-B magnets across the
industry [35]. New technologies to harvest waste magnets and
reprocess them into useable new alloy, powder or magnets are
being studied. This activity is most commercially attractive
when the RE metal prices are high and there is sufficient
financial interest in converting waste. The major technical
challenge to recycling RE-Fe-B magnets is removing the
levels of impurities within scrap magnets (e.g. O, C, Ni).
Sophisticated techniques are required to remove impurities
and make the scrap as comparable to the virgin raw material as
possible.
IX. ALTERNATIVE PERMANENT MAGNET TECHNOLOGIES
There are a number of long term permanent magnet
research activities that have the potential to deliver
alternatives to these RE-Fe-B magnets in the future. These
include developing SmFeN with higher Hci [6], Nd2Fe14B/-
Fe nanocomposites with improved microstructures [36],
anisotropic surfactant-assisted ball milled NdFeB powders
[37] and Fe16N2 with sufficient Hci and thermal stability [38].
REFERENCES
[1] K. H. J. Buschow, Handbook of Magnetic Materials, vol. 12,
Amsterdam, North Holland, 1999.
[2] J. F. Herbst and J. J. Croat, ―Neodymium-iron-boron permanent
magnets.‖ J. Magn. Mag. Mat. vol. 100, pp. 57-78, 1991.
[3] G. C. Hadjipanayis, ―Nanophase hard magnets.‖ J. Mag. Magn. Mat.
vol. 200, pp. 373-391, 1999.
[4] E. Burzo, ―Permanent magnets based on R–Fe–B and R–Fe–C alloys.‖
Rep. Prog. Phys. vol. 61, pp. 1099–1266, 1998.
[5] O. Gutfleisch, ―Controlling the properties of high energy density
permanent magnetic materials by different processing routes.” J. Phys.
D: Appl. Phys. vol. 33, pp. R157–R172, 2000.
[6] T. Iriyama, ―Recent Trends and Perspectives on Development of
Permanent Magnets in Japan.‖ Proc. 23rd Int. Workshop on Rare Earth
Permanent Magnet & Their Applications, (Annapolis, USA), pp. 180-
188, 2014.
[7] J. J. Croat, J. F. Herbst, R. W. Lee and F. E. Pinkerton, ―PrFe and NdFe
based Materials: A New Class of High Performance Permanent
Magnets.‖ J. Appl. Phys. vol. 55, pp. 2078, 1984.
[8] M. Sagawa, S. Fujimura, N. Togawa, H. Yamamoto and Y. Matsuura,
―Permanent Magnet Materials Based on the Rare Earth-Iron-Boron
Tetragonal Compounds.‖ J. Appl. Phys. vol. 55, pp. 2083, 1984.
[9] H. H. Stadelmaler, E. T. Henig, G. Petzow, ―A Chronicle of the
Development of Iron Based Rare Earth High Performance Magnets.‖ Z.
Metallkde. vol. 83, pp. 163-168, 1991.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TMAG.2016.2535482, IEEE
Transactions on Magnetics
MAGCON-15-11-1796.R1 9

[10] G. Rieger, ―The use of Permanent Magnets in Industrial and Energy processed by Nd–Cu eutectic diffusion under expansion constraint.‖ Scr.
Applications at Siemens.‖ Proc. 23rd Int. Workshop on Rare Earth Mat. vol. 81, pp. 48-51, 2014.
Permanent Magnet & Their Applications (Annapolis, USA) pp. 61-65, [34] H. Sepehri-Amin, J. Liu, T. Ohkubo, K. Hioki, A. Hattori and K. Hono,
2014. Enhancement of coercivity of hot-deformed Nd–Fe–B anisotropic
[11] S. Shaw and S. Constantinides, ―Permanent Magnets: the Demand for magnet by low-temperature grain boundary diffusion of Nd60Dy20Cu20
Rare Earths.‖ 8th International Rare Earths Conference, Hong Kong, eutectic alloy.‖ Scr. Mater. vol. 69, pp. 647-650, 2013.
November 2012. [35] A. Walton, H. Yi, N. A. Rowson, J. D. Speight, V. S. J. Mann, R. S.
[12] S. Kozawa, ―Trends and Problems in Research of Permanent Magnets Sheridan, A. Bradshaw, I. R. Harris and A. J. Williams, ‗The use of
for Motors.‖ Science and Technology Trends, Japan 2010. Available on hydrogen to separate and recycle neodymium-iron-boron-type magnets
line at: http://data.nistep.go.jp/dspace/bitstream/11035/2854/1/NISTEP- from electronic waste.‖ Journal of Cleaner Production, vol. 104, pp.
STT038E.-40.pdf 236, 2015.
[13] W. Benecki, T. Clagett, and S. Trout, ―Permanent magnets 2010–2020: [36] R. Skomski and J. M. D. Coey, ‗Giant energy product in nanostructured
A Comprehensive Overview of the Global Permanent Magnet Industry,‖ two-phase magnets.‖ Phys. Rev. B vol. 48, pp. 15812, 1993.
Report No. 9-21, vol. 55, 2010 [37] N. Poudyal, V. V. Nguyen, C. Rong and J. P. Liu ―Anisotropic bonded
[14] M. Sagawa, ―Development and Prospects of the Nd-Fe-B Sintered magnets fabricated via surfactant-assisted ball milling and magnetic-
Magnets.‖ REPM Proc. 21st Int. Workshop on RE Permanent Magnets field processing.‖ J. Phys. D: Appl. Phys. vol. 44, pp. 335002, 2011.
and their Apps. (Bled, Slovenia), pp.183, 2010. [38] K. Takagi, M. Akada, K. Ozaki, N. Kobayashi, T. Ogawa, Y. Ogata and
[15] D. Goll, M. Seeger, H. Kronmüller, ―Magnetic and microstructural M. Takahashi, ―High-pressure sintering behavior of α-Fe16N2 nano-
properties of nanocrystalline exchange coupled PrFeB permanent powder.‖ J. Appl. Phys. vol. 115, pp. 103905, 2014.
magnets.‖ J. Magn. Mag. Mat. vol. 185, pp. 49–60, 1998.
[16] WTO Dispute Settlement DS431: ―China — Measures Related to the
Exportation of Rare Earths, Tungsten and Molybdenum.‖ May, 20,
2015. Available online at : David Brown has worked as a senior
https://www.wto.org/english/tratop_e/dispu_e/cases_e/ds431_e.htm
research scientist at Magnequench
[17] Z. Chen, Y. K. Lim, and D. Brown ―Substitution of Ce for (Nd,Pr) in
Melt-Spun (Nd,Pr)–Fe–B Powders.‖ IEEE Trans. Magn. vol. 51, pp. International since 2000. He has gained an
210214, 2015. immense amount of experience in
[18] H. Sepehri-Amin, D. Prabhu, M. Hayashi, T. Ohkubo, K. Hioki, A. materials and process developments in an
Hattori and K. Hono ―Coercivity enhancement of rapidly solidified Nd-
industrial environment. He has been
Fe-B magnet powders.‖ Scr. Mat. vol. 68 pp. 167-170, 2013.
[19] T. T. Sasaki, T. Ohkubo, Y. Takada, T. Sato, Y. Kaneko and K. Hono involved with many new product
―Formation of non-ferromagnetic grain boundary phase in a Ga-doped developments, working at facilities in the
Nd-rich Nd-Fe-B sintered magnet.‖ Scr. Mat. vol. 113 pp. 218-221, USA and Asia.
2016.
David has extensive experience in powdered metal processing,
[20] W. F. Li, T. Ohkubo, K. Hono and M. Sagawa, ―The origin of coercivity
decrease in fine grained Nd–Fe–B sintered magnets.‖ J. Mag. Magn. alloying, rapid quenching, thermo-mechanical forming and
Mat. vol. 321, pp. 1100–1105, 2009. recycling of rare earth magnetic materials, as well as other
[21] J. Bernardi, J. Fidler, M. Sagawa and Y. Hirose, ―Microstructural functional materials that require similar fabrication techniques.
analysis of strip cast Nd–Fe–B alloys for high (BH)max magnets.‖ J.
David gained a B.Eng.(Hons) from the University of Bath and
Appl. Phys. vol. 83, pp. 6396-6398, 1998.
[22] S. Sasaki, ―Alloy flake for rare earth magnet production method thereof, a Ph.D. from the University of Birmingham in Materials
alloy powder for rare earth sintered magnet, rare earth sintered magnet..‖ Science and Engineering during the 1990s.
U.S. Patent 7 571 757, Jul, 12, 2007.
[23] M. Nakamura, M. Matsuura, N. Tezuka, S. Sugimoto, Y. Une, H. Kubo
and M. Sagawa, ―Preparation of ultrafine jet-milled powders for Nd-Fe-
B sintered magnets using HDDR and HD processes.‖ Appl. Phys. Lett.
vol. 103, pp. 022404, 2013.
[24] H. Sepehri-Amin, T. Ohkubo and K. Hono, ―The mechanism of
coercivity enhancement by the grain boundary diffusion process of Nd–
Fe–B sintered magnets.‖ Acta Mater. vol. 61, pp. 1982-1990, 2013.
[25] M. Miwa, T. Iwasa, T. Ishiyama, W. Ohkawa, Y. Tanaka and Y.
Enokido, ―Coercivity improvement of Dy-free NdFeB sintered
magnets.‖ The Ceramic Society of Japan. The 28th Fall Meeting 18th
September 2015, session 3D16
[26] J. J. Croat ―Current status and future outlook for bonded neodymium
permanent magnets.‖ J. Appl. Phys. vol. 81 pp. 4804-4808, 1997.
[27] J. Ormerod and S. Constantinides, ―Bonded permanent magnets: Current
status and future opportunities.‖ J. Appl. Phys. vol. 81, pp. 4816-4820,
1997.
[28] www.mqitechnology.com
[29] D.N. Brown, Y.K. Lim, W. Zhao, F. He, J.W. Herchenroeder, ―Recent
Developments with Melt-Spun NdFeB for Permanent Magnet
Applications.‖ Proc. 23rd Int. Workshop on Rare Earth Permanent
Magnet & Their Applications, (Annapolis, USA), pp. 45-50, 2014.
[30] H. Mitarai, K. Noguchi, C. Mishima, H. Matsuoka, M. Yamazaki, Y.
Kawasugi, ―Development of Compound for Anisotropic Bonded Nd
Magnets Using d-HDDR Magnet Powder.‖ IEEE Trans. Magn. vol. 50,
pp. 2105303, 2014.
[31] Z. Wei, H. Zhisan and D. Miller, ―Capsule‖ US Patent 9 028 705, May,
12, 2015.
[32] R. W. Lee, ―Neodymium-iron-boron permanent magnets prepared by
rapid solidification.‖ Appl. Phys. Lett. vol. 46, pp. 790, 1985.
[33] T Akiya, J Liu, H Sepehri-Amin, T Ohkubo, K Hioki, A Hattori, K
Hono, ―High-coercivity hot-deformed Nd–Fe–B permanent magnets

0018-9464 (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.