Experimental investigation of the Mg-Mn-Nd isothermal section at 450 °c

Article abstract of DOI:10.1016/j.jallcom.2014.04.113

The Mg-Mn-Nd isothermal section at 450°C was constructed experimentally using solid-solid diffusion couples and key alloys. Diffusion couples and key alloys were annealed at the same temperature for different annealing time periods to understand the diffusion behavior of the system components, and to achieve the phase equilibrium in the system. The phase relations and solubility limits were determined using scanning electron microscope coupled with wave/energy dispersive X-ray spectrometers, X-ray diffraction and metallography. No ternary compounds could be observed in this system at 450°C. The microstructures of the diffusion zones showed Mn either as a pure element or as a solute atom. However, microstructures of the annealed key alloys near the Mn-rich side showed the formation of Mn₁₇Nd₂ at 450°C. Other Mn binary compounds, Mn₂Nd and Mn₂₃Nd₆, were observed only in the as-cast alloys, because they are not stable at 450°C. The experiments showed that Mn remained stationary in some diffusion zones forming Mn+(MgNd), Mn+(Mg₃Nd) and Mn+Mg₄₁Nd ₅ two-phase regions. However, some of Mn atoms were diffusing in the system forming a continuous thin layer of pure Mn at the interface of pure Mg end-member, or to form (α-Nd)^Mg,Mn, (MgNd) and (Mg ₃Nd) ternary solid solutions. The solubility of Mn in these solutions was estimated as 3.0, 13.0 and 3.5 at.% Mn, respectively. Since Mn is in equilibrium with most of the phases in the system, the phase boundary lines are pointing toward Mn-rich corner except that of (α-Nd) ^Mg,Mn+(MgNd)+Mn₁₇Nd₂ three-phase field. This was confirmed based on the consistency of the results obtained from diffusion couples and key alloys.

Full text of DOI:10.1016/j.jallcom.2014.04.113

Accepted Manuscript
Experimental investigation of the Mg-Mn-Nd isothermal section at 450°C
Ahmad Mostafa, Mamoun Medraj
PII:
S0925-8388(14)00926-8
DOI:
http://dx.doi.org/10.1016/j.jallcom.2014.04.113
JALCOM 31078
Reference:
To appear in:
Received Date:
Revised Date:
Accepted Date:
20 March 2014
15 April 2014
17 April 2014
Please cite this article as: A. Mostafa, M. Medraj, Experimental investigation of the Mg-Mn-Nd isothermal section
at 450°C, (2014), doi: http://dx.doi.org/10.1016/j.jallcom.2014.04.113
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Experimental investigation of the Mg-Mn-Nd isothermal
section at 450°C
Ahmad Mostafa and Mamoun Medraj
Mechanical and Industrial Engineering Department, Concordia University,
1455 de Maisonneuve Blvd. West, Montreal, Quebec, Canada, H3G 1M8
Tel: (514) 848 2424 ext. 3146, Fax: (514) 848 - 3175
Email of the corresponding author: mmedraj@encs.concordia.ca
Abstract
The Mg-Mn-Nd isothermal section at 450°C was constructed experimentally using solid-solid
diffusion couples and key alloys. Diffusion couples and key alloys were annealed at the same
temperature for different annealing time periods to understand the diffusion behaviour of the
system components, and to achieve the phase equilibrium in the system. The phase relations and
solubility limits were determined using scanning electron microscope coupled with wave/energy
dispersive X-ray spectrometers, X-ray diffraction and metallography. No ternary compounds
could be observed in this system at 450°C. The microstructures of the diffusion zones showed
Mn either as a pure element or as a solute atom. However, microstructures of the annealed key
alloys near the Mn-rich side showed the formation of Mn17Nd2 at 450°C. Other Mn binary
compounds, Mn2Nd and Mn23Nd6, were observed only in the as-cast alloys, because they are not
stable at 450°C. The experiments showed that Mn remained stationary in some diffusion zones
forming Mn+(MgNd), Mn+(Mg3Nd) and Mn+Mg41Nd5 two-phase regions. However, some of
Mn atoms were diffusing in the system forming a continuous thin layer of pure Mn at the
interface of pure Mg end-member, or to form (α-Nd)Mg,Mn, (MgNd) and (Mg3Nd) ternary solid
solutions. The solubility of Mn in these solutions was estimated as 3.0, 13.0 and 3.5 at.% Mn,
respectively. Since Mn is in equilibrium with most of the phases in the system, the phase
boundary lines are pointing towards Mn-rich corner except that of (α-
Nd)Mg,Mn+(MgNd)+Mn17Nd2 three-phase field. This was confirmed based on the consistency of
the results obtained from diffusion couples and key alloys.
Keywords: metals and alloys; rare earth alloys and compounds; solid state reactions; diffusion;
phase diagrams; scanning electron microscopy (SEM)
1

1. Introduction
Magnesium and its alloys show an increasing commercial importance in automotive and
electronic consumer products because of many advantages. Their light-weight, high specific
toughness and rigidity, good machinability, castability and weldability using commonly known
techniques, made them attractive for the use in industry [1]. In terms of structural properties, pure
Mg is not suitable for direct commercial use. The liquid of pure Mg, however, is used to extract
neodymium from solid FeNdB magnet scrap materials [2]. It was found that addition of Nd to
Mg alloys improves their strength, ductility, toughness, weldability and machinability [3-5]. This
is attributed to grain refinement and the formation of secondary phase precipitates at the grain
boundary regions. On the other hand, addition of Mn improves the hot and cold strengths,
resistance to corrosion and formability compared with the pure Mg [6]. This is due to high
reactivity of Mn with other elements such as Fe [7], which leads to form a complex sequence of
intermediate compounds [8]. Thus, it is necessary to understand the phase equilibria in the Mg-
Mn-Nd system; which is essential for alloy design and development.
2. Literature review
2.1 The Mg-Mn binary system
The Mg-Mn system has been thoroughly studied [9-12]. According to these authors, no
intermediate compound exists between Mg and Mn terminal sides. The maximum solubility of
Mn in solid Mg was found to be 0.998 at.% Mn at the peritectic temperature, just below 650°C.
Another invariant reaction of monotectic type L1→L2+δ-Mn, takes place at 96 at.% Mn and
1198°C [11, 12]. Because of the difficulty of measuring the temperature of the liquid miscibility
gap experimentally, attempts to estimate that temperature were carried out using thermodynamic
calculations [12-14]. The calculated temperature of the liquid miscibility gap was reported as
2

3202°C [12], 1902°C [15] and 3415°C [14]. The temperature is significantly lower in the work of
[15]. They adjusted the Mg-Mn binary model parameters to obtain consistency for the
extrapolated Mg-Mn-Y ternary system.
2.2 The Mg-Nd binary system
According to Nayeb-Hashemi and Clark [16], who assessed the Mg-Nd phase diagram based on
the literature data, five intermetallic compounds, MgNd, Mg2Nd, Mg3Nd, Mg41Nd5 and Mg12Nd,
along with the terminal solid solutions of Nd in Mg, Mg in α-Nd and Mg in β-Nd, were reported.
The Mg12Nd compound was considered metastable, since it was observed in the samples
quenched directly from liquid and not after annealing [17]. Three eutectic reactions were
observed at 42.5 at.% Mg and 775°C, 64.5 at.% Mg and 750°C and 91 at.% Mg and 546°C. The
terminal solubility of Mg in solid α-Nd was estimated as 8.2 at.% Mg [18] and in β-Nd as 34 at.%
Mg [17] at 551°C and 775°C, respectively. Whereas, the terminal solubility of Nd in solid Mg
was determined as 0.1 at.% Nd [19, 20]. Moreover, the two congruently melting compounds,
MgNd and Mg3Nd have solid solubility ranges. The solid solubility range of Mg3Nd was
estimated as 6 at.% extending towards the Mg-rich side [17]. The solid solubility range of MgNd
was not determined quantitatively, its existence was proposed by Delfino et al. [17] based on the
change in the lattice parameters. Mg2Nd phase was suggested to appear as a Laves phase with
narrow stability range between ~660 and 755°C [21].
2.3 The Mn-Nd binary system
Many contradictory information on the Mn-Nd system were found in the literature [8, 22, 23].
Thus, three different versions of Mn-Nd phase diagram are available in the literature. To
overcome these uncertainties, a self-consistent Mn-Nd phase diagram was reconstructed based on
3

combination of experiments, CALPHAD methodology and first-principle calculations by our
group [24]. A eutectic was found to occur at 27 at.% Mn and 685°C. Three incongruent
intermetallic compounds, Mn2Nd, Mn23Nd6 and Mn17Nd2, were reported at 850, 940 and 1025°C,
respectively. The compounds Mn2Nd and Mn23Nd6 were found to decompose eutectoidally in the
temperature ranges of 550-650°C and 400-550°C, respectively. The maximum solid solubility of
Mn in α-Nd at the eutectic temperature was estimated as 2.3 at.% Mn based on EPMA results.
However, the solubility of Mn in β-Nd was determined as 5.0 at.% Mn by extrapolating the DSC
points to 728°C. No solid solubility of Nd in Mn was detected.
2.4 The Mg-Mn-Nd ternary system
The system was first assessed by Drits et al. [25] using thermal analysis, microscopic
investigation, and electrical resistivity measurements. The study was limited to the Mg-rich
corner up to 4 wt.% Mn and 10 wt.% Nd to determine the boundaries of the solid solution region
(by rapid quenching and annealing at 570°C, 600°C, and 630°C). Based on the resistivity
measurements, they [25] reported a peritectic reaction L+Mn ↔ α+Mg9Nd at 559.6°C. Also, they
found that Mn and Mg9Nd are in equilibrium with Mg solid solution. The compound Mg9Nd was
later given the formula Mg41Nd5 with tI92-Ce5Mg41 type structure [17]. The results obtained by
Drits et al. [25] were used later by Drits and Bochvar [26] to investigate the Mg-Mn-Nd-Ni
quaternary alloys.
Very limited experimental information on the Mg-Mn-Nd ternary phase diagram could be found
in the literature. The available study [25] covered the Mg-rich corner only and focused on finding
the solid solubility ranges of Mn and Nd in Mg. The main objective of the present work is to
construct the isothermal section of Mg-Mn-Nd phase diagram at 450°C experimentally for the
4

whole composition range by means of solid-solid diffusion couples and key alloys using XRD,
SEM/WDS and metallography.
3. Experimental procedure
The samples for diffusion couples and key alloys were prepared from pure metals. Mn-99.90
wt.% and Nd-99.95 wt.% purity were supplied by Alpha Aesar®, Johnson Matthey Company,
Inc. Mg-99.8 wt.% purity was supplied by CANMET Materials Technology Laboratory
(CANMET-MTL). Samples of 4-5g in weight were prepared in an arc-melting furnace with
water-cooled copper crucible and non-consumable tungsten electrode under argon. Excess Mg
(about 10%) was added to compensate for Mg losses due to evaporation, since Mg has to be
alloyed with relatively high melting temperature elements (Mn and Nd). Samples were melted
several times to ensure the composition homogeneity. The actual global composition was
determined using Ultima2 inductively coupled plasma with optical emission spectrometer (ICP-
OES). For each sample, three pieces from different locations were chosen. Average of the three
corresponding compositions was considered as the actual composition. The difference between
the three locations was found in the range 2 at.%. Key alloys were subjected to a step annealing
approach, which includes annealing for different time intervals. Then, the as-cast and annealed
microstructures were compared in order to understand the phase equilibrium.
Solid-solid diffusion couples were prepared from two blocks of alloys and/or pure metals. The
contacting surfaces were ground gradually up to 1200 grit SiC emery paper, then polished down
to 1µm using alcohol-based diamond suspension and 99% ethanol as a lubricant. The blocks were
attached together using stainless steel clamp rings. The samples and diffusion couples were
wrapped in tantalum foil, encapsulated inside an argon-flushed quartz tube with inside pressure
of (~8×10-1 torr) and annealed at 450°C for different time intervals. Annealing time intervals
5

were chosen systematically, based on the alloy composition, by performing annealing and
quenching after different time increments to address the microstructural changes with time. The
microstructures, of the diffusion couples and/or alloys, were then compared to understand the
phase equilibria and phase evolution.
Philips X’Pert pro X-ray diffractometer (XRD) was used. X-ray diffraction analysis was
performed on powder-form samples in the range from 20 to 120 degree 2θ with 0.02° step size to
identify the phases. Rietveld analysis was used to measure the relative amounts of these phases.
The crystal structure of the compounds was characterized using X’Pert Highscore plus software
[27]. The diffusion couples and equilibrated alloys were analysed using Hitachi S-3400 scanning
electron microscope (SEM) equipped with wavelength dispersive x-ray spectrometer (WDS) and
energy dispersive X-ray spectrometer (EDS). WDS spot analysis was used to identify the phase
composition of the diffusion zones and key alloys. WDS line-scans were performed across the
diffusion zones to determine the solubility ranges of the various phases in the system. The WDS
experimental error was estimated to be ~1 at.%, based on the actual difference between the
readings. EDS elemental mapping was used to determine phase composition and distribution of
elements within diffusion zones.
4. Results and discussion
In the following section, the binary and ternary solid solutions are described using parentheses
and superscripts. For example, (MgNd) indicates the ternary solid solubility of Mn in MgNd, the
binary solid solution of Mg in α-Nd is represented as (α-Nd)Mg, and the extended solid solubility
of Mg and Mn in α-Nd is represented as (α-Nd)Mg,Mn
.
Solid-solid diffusion couples were annealed at 450°C to accelerate diffusion without causing
melting, since the lowest liquidus temperature was estimated as 500°C based on the preliminary
6

thermodynamic calculations of the system. Diffusion couples were annealed for different
intervals to reveal the phase equilibrium between the formed layers and their constituents based
on the microstructural changes of diffusion couples versus time. Different annealing time
intervals were also essential to assure that all diffusion layers are present.
4.1 Solid-solid diffusion couples
The end-members of diffusion couple #1 were made from two binary alloys. The first end-
member was made from an alloy containing 78Mn-22Nd at.%. The second end-member was
made from an alloy containing 87Mg-13Nd at.%. Diffusion couple #1 was annealed at 450°C for
10 days. SEM micrographs of this diffusion couple are shown in Figure 1. WDS spot analysis
was carried out to identify the composition of the diffusion zones and their constituents. The
results of WDS spot analysis are listed in Table 1. WDS line-scan performed across diffusion
zones of diffusion couple #1 is shown in Figure 2. Initially, the as-cast composition of the first
end-member (zone #1) contained (α-Nd)Mn+Mn23Nd6. Mn23Nd6 is a peritectic compound forms at
930°C and decomposes eutectoidally at ~500°C [8, 24]. End-member (zone #1) was
homogenized at 450°C for 15 days, before coupling with the second end-member (zone #5), in
order to achieve the (α-Nd)Mn+Mn17Nd2 equilibrium structure. It is necessary to start with
homogenized end-member to avoid forming diffusion layers amongst non-stable compounds.
Although diffusion is a dynamic process and relies on the thermodynamic equilibrium between
the adjacent layers and their constituents, it will take very long time to reach equilibrium if meta-
stable compounds are presented.
Diffusion zone #2, shown in Figure 1 (c), represents the (MgNd) diffusion layer. The
composition profile across this layer is shown in Figure 2 (a). The profile shows that Mn is
7

substituting Mg with about 3.5 at.% at constant concentration of 50 at.% Nd. These points were
obtained from a very thin layer (1.7 µm width). Thus, this solubility limit needs to be confirmed
by key alloys. Nevertheless, the results obtained from the composition profile still important for
the preliminary prediction of the phase relationships. It is worth noting that the elemental scan
shown in Figure 2 (a) is done in a selective way by avoiding the secondary phases in the diffusion
layers (i.e. Mn and Mn17Nd2), whereas a blind line-scan is shown in Figure 2 (b).
(MgNd) layer seems to appear only when (α-Nd)Mn has a large contact with (Mg3Nd). Hence, it is
not a continuous diffusion layer. Figure 1 (c) is redrawn to show the phase relations between the
constituents of diffusion zones #1, 2 and 3 as illustrated in Figure 3. As can be seen from Figure
3, (MgNd) (zone #2) is in equilibrium with (α-Nd)Mn and Mn17Nd2 from zone #1 and with
(Mg3Nd) from zone #3. Thus, the three-phase equilibrium (α-Nd)Mn+Mn17Nd2+(MgNd) was
determined at the interface between zone #1 and zone #2. Furthermore, the equilibrium between
(MgNd) and (Mg3Nd) was determined at the interface between zone #2 and zone # 3. The contact
between zone #1 and zone #3 represents the phase equilibrium between Mn17Nd2 and pure Mn.
These phase relations are shown on the Mg-Mn-Nd triangle below at the end of diffusion couple
#1 discussion.
Diffusion zone #3 appeared with a total thickness of ~100µm. This zone resulted mainly from the
interatomic diffusion between Mg and Nd, when (α-Nd)Mn first reacted with diffusing Mg.
Because of the high affinity of Mg to Nd, diffusing Mg continues to react with the Nd stemming
from the decomposition of Mn17Nd2 that is originally present in the first end-member (zone #1).
The decomposition of Mn17Nd2 liberates Mn particles due to Nd consumption. More Mn was
liberated from the initial reaction between (α-Nd)Mn and the diffusing Mg. 3.7 at.% of the
8

liberated Mn was dissolved in Mg3Nd to form (Mg3Nd) solid solution in zone #3, which is
relatively wider than other zones. When the maximum solubility of Mn in (Mg3Nd) was reached,
no more Mn could be dissolved. Therefore, pure Mn regions remained stationary in zone #3 as
shown in Figure 1 (b) and (d). WDS line-scan in Figure 2 (b) shows scattered data in this zone
due to the random distribution of Mn in this layer. The phase morphology of Mn17Nd2 within
zone #1 and later becoming Mn in zone #3 is shown in Figure 1 (b).
In terms of phase equilibrium, diffusion zone #3 represents the Mn+(Mg3Nd) two-phase region in
the Mg-Mn-Nd isothermal section at 450°C. It is worth noting that Mn17Nd2 (zone #1) is not in
equilibrium with (Mg3Nd) (zone #3). This is clear from the microstructure shown in Figure 1 (b)
and Figure 3, where the extension of Mn17Nd2 morphology becomes pure Mn regions in zone #3.
Because zone #1 contains two phases, two possible paths can be depicted from diffusion couple
#1. The first possibility can be represented as follows: Mn17Nd2 → Mn → Mn+(Mg3Nd) →
Mg3Nd → Mg3Nd+Mg41Nd5. The second possibility can be as follows: (α-Nd)Mn+Mn17Nd2 →
(α-Nd)Mn+Mn17Nd2+(MgNd) → (MgNd) → (Mg3Nd) → (Mg3Nd)+Mn → Mg3Nd →
Mg3Nd+Mg41Nd5.
Zone #4 represents the Mg3Nd binary compound, which resulted from the diffusion reaction of
Nd and Mn41Nd5. The spot analysis in Table 1, line-scan in Figure 2 (b) and EDS elemental
mapping in Figure 4 revealed that there is no Mn in zone #4. Accordingly, the original interface
of diffusion couple #1 could be determined at the interface between zones #3 and #4 as
demonstrated in Figure 1 (a).
EDS elemental mapping, shown in Figure 4, was performed to detect the distribution of the three
elements among diffusion zones of diffusion couple #1. Figure 4 (a) represents the backscattered
9

electron image of diffusion couple #1. Figure 4 (b-d) represents the elemental maps of Mg, Mn
and Nd, respectively.
Mg map in Figure 4 (b) shows that Mg concentration decreases from 89.1 at.% (zone #5) to 46.5
at.% (zone #2). No Mg was detected in zone #1, which is the Mn-Nd binary alloy end-member.
Because of the low magnification of the elemental mapping image, Mn was only observed in
zones #1 and #3 as shown in Figure 4 (c). However, WDS line-scan (Figure 2 (b)) showed that
zone #2 contains 3.5 at.% Mn as discussed earlier. Mn trace could be evidence of the original
interface location, where the Mn liberated from Mn17Nd2 disappeared at the interface between
zone #3 and zone #4 marking the original interface. On the other hand, Nd map (Figure 4 (d))
shows that Nd concentration decreases from 97.4 at.% in zone #1 to 10.7 at.% in zone #5.
According to the phase equilibrium analysis of diffusion couple #1 using WDS and EDS
techniques, the phase relationships can be determined graphically on the Mg-Mn-Nd isothermal
section at 450°C as shown in Figure 5. The two end-members (A-B) are connected by a dashed
line. The width of the two-phase region Mn+(Mg3Nd) is taken as 1.7 at.% Mg based on the
binary solubility range of Mg3Nd at 450°C [17]. The drawing in Figure 3 is used to describe the
phase relationships on the Mg-Mn-Nd triangle at 450°C based on the microstructure of diffusion
couple #1.
Diffusion couple #2, prepared from two ternary alloys whose details are listed in Table 2, was
annealed at 450°C for 4 days. The annealing time here was shorter than that for couple #1,
because the quartz tube was observed to be dark indicating significant evaporation of some of the
elements. Longer annealing resulted in losing one of the end-members, which was experienced
during the several attempts to obtain a successful diffusion between these two end-members.
10

The first end-member (zone #1) contains Mn, Mn17Nd2 and (MgNd). The second end-member
(zone #4) contains Mn, Mg41Nd5 and (Mg3Nd). WDS spot analysis was used to determine the
composition of the diffusion zones and their constituents. The results of WDS spot analysis are
listed in Table 2. The SEM micrograph of diffusion couple #2 is shown in Figure 6. Figure 6 (a)
shows that diffusion couple #2 contains four diffusion zones. The original interface between the
two end-members is marked by a dashed line in the same figure. Figure 6 (b) is a magnified part
from diffusion zone # 2 showing the Mn particles resulting from the dissolution of Mn17Nd2 to
provide Nd which reacts with the diffusing Mg to form (Mg3Nd). The same figure shows that the
Mn particles are dispersed in the (Mg3Nd) solid solution matrix with 3.7 at.% Mn.
Besides the two three-phase equilibria Mn17Nd2+(MgNd)+Mn and Mn+(Mg3Nd)+Mg41Nd5
inferred from the end-members of couple #2, the Mn+(MgNd)+(Mg3Nd) three-phase equilibrium
was observed at the interface between zones #1 and #2. The WDS results in Table 2 show that
Mn concentration in zones #1 and 2 is relatively higher than that in zones #3 and 4. The relatively
high Mn concentration can be as a result of Mn17Nd2 decomposition, which produced appreciable
amount of free Mn atoms that have enough driving force to diffuse in a wide (Mg3Nd) matrix
(zone #2). However, most of the Mn in the second end-member alloy (zone #4) remained pure in
the microstructure as can be seen in zone #4 of Figure 6 (a), because the composition of the end-
member alloy lies in the Mn+Mg41Nd5+(Mg3Nd) triangulation. Thus, only a little amount of Mn
was dissolved by (Mg3Nd) as indicated in Table 2.
The solubility of Mn in (MgNd) was determined from the composition of the end-member alloy
(zone #1) in diffusion couple #2. WDS spot analysis revealed that (MgNd) solid solution contains
11

11.2 at.% Mn. Whereas, the solubility of Mn in (Mg3Nd) was measured as 3.7 at.% in both
diffusion couples #1 and 2.
Based on the obtained results from diffusion couple #2, the phase relationships are represented in
Figure 7. The two end-members (A-B) are connected by a dashed line. The shaded areas
represent the triangles of the end-members’ alloys used in diffusion couple #2.
The SEM micrograph of diffusion couple #3 is shown in Figure 8. The first end-member contains
three phases, Mn17Nd2, (MgNd) and (α-Nd)Mg,Mn. The second end-member is made from pure Mg
(99.98 wt.% purity). During annealing, various phases formed due to the interdiffusion between
Mg and Nd. The diffusion layers of this diffusion couple were similar to those of Mg-Nd binary
system, i.e.: MgNd, Mg3Nd and Mg41Nd4. The only difference is that the ternary diffusion couple
contains pure Mn within these layers, forming two-phase region within each. The Mn content in
all diffusion layers stemmed originally from Mn17Nd2 in the ternary alloy of the second end-
member (zone #6). WDS spot analysis was used to identify the composition of the phases in each
zone. The results of WDS spot analysis are summarized in Table 3.
Six diffusion zones were observed in diffusion couple #3 as shown in Figure 8 (a). Figure 8 (b)
represents a magnified region from diffusion couple #3, showing that (Mg3Nd) is in equilibrium,
selectively, with (MgNd) (the grey phase). Figure 8 (c) shows the presence of pure Mn, resulting
from the decomposition of Mn17Nd2, in diffusion zones #3 and 4.
Diffusion zone #1 represents the pure Mg end-member. Diffusion zone #2 represents a
continuous thin layer of pure Mn whose exact composition was difficult to acquire using WDS
spot analysis, because it is very thin. The composition profile in Figure 9, however, shows the
drop in both Mg and Nd compositions with increase in Mn content in zone #2. This supports the
12

presence of the observed Mn thin layer in zone #2. It is suggested that the pure Mn layer formed
due to the diffusion of Mn atoms that were liberated from the Mn-containing phases. This
suggestion was supported experimentally by diffusion couple #4, which will be discussed later.
The formation of this thin layer is necessary to fulfil the equilibrium condition near the Mg-rich
side, where Mg (zone #1) equilibrates with Mn (zone #2) and Mn equilibrates with Mg41Nd5 to
form the Mg41Nd5+Mn two-phase region in zone #3. Diffusion zones #4 to 5 represent other two-
phase layers, where Mn is in equilibrium with (Mg3Nd) and (MgNd). Diffusion zone #6
represents the three-phase alloy end-member. Besides the three two-phase equilibria
Mg41Nd5+Mn, (Mg3Nd)+Mn and (MgNd)+Mn, two three-phase equilibria; namely:
Mn+Mg41Nd5+(Mg3Nd) and Mn+(Mg3Nd) +(MgNd), could be obtained at the interface of zones
#3 and 4 and zones #4 and 5, respectively.
Figure 9 shows the WDS line-scan that was performed with a step size of ~1µm across the
diffusion zones of diffusion couple #3 as illustrated in Figure 8 (a). It is worth noting that zone #6
contains three phases. However, the WDS line-scan shows only one phase, because the line did
not cross the other two phases.
The line-scan showed that no remarkable solubility of Nd and Mn in Mg end-member (zone #1).
The composition profile through the Mg41Nd5+Mn diffusion layer (zone #3) showed constant
composition of the three elements. The profile across (Mg3Nd)+Mn (zone #4) shows an irregular
distribution of the three elements at certain points. This was attributed to the presence of Mn in
this diffusion zone. The composition profile across (MgNd)+Mn (zone #5) shows that up to 5.2
at.% Mg is substituted by Mn at constant concentration of Nd. This was also supported by the
WDS spot analysis summarized in Table 3. According to these profiles, the solubility of Mn in
13

both (Mg3Nd) and (MgNd) was determined as 2.6 and 5.2 at.% Mn, respectively. This will be
confirmed by key alloys in the next section.
The diffusion path was determined based on the local equilibria between the phases of diffusion
couple #3 as follows: (MgNd)+(α-Nd)Mg,Mn+Mn17Nd2 (end-member) → (MgNd)+Mn →
(MgNd)+Mn+(Mg3Nd) → (Mg3Nd)+Mn → Mg41Nd5+(Mg3Nd)+Mn → Mg41Nd5+Mn → Mn →
Mg (end-member). The phase equilibria determined by diffusion couple #3 are shown in Figure
10. The two end-members (A-B) are connected by a dashed line. The shaded area represents the
triangle of the three-phase end-member.
Diffusion couple #4 was prepared to provide better understanding of the formation of pure Mn
layer near the Mg end-member. The first end-member of this diffusion couple was made from
pure Mg (zone #1). The second end-member was from a ternary alloys containing
Mn+Mn17Nd2+(MgNd) (zone #5). Diffusion couple #4 was annealed only for 2 days. Because the
objective of this couple was to interrupt annealing at early stages of diffusion process to confirm
the observations about Mn as pure metal and as solute atoms as discussed in previous diffusion
couples #1-3. SEM micrograph of diffusion couple #4 is shown in Figure 11. Five diffusion
zones were observed in this diffusion couple. The WDS results of the constituents of every
diffusion zone are listed in Table 4.
It is clear that diffusion couple #4 did not reach equilibrium, because three phases were observed
in zone #3. This is not acceptable from thermodynamics point of view, since the maximum
number of phases per diffusion layer in a ternary diffusion couple must be two, except at the
interface. Nevertheless, this diffusion couple is discussed here to shed some light on the kinetics
of diffusion in this system. The three phases seen in zone #3 are, (Mg3Nd), Mg41Nd5 and Mn.
Figure 11 shows that Mn grain morphology did not change from one diffusion zone to another,
14

while Mg and Nd atoms are actively diffusing. Not only that, the pure Mn grain (in the rectangle
in Figure 11 (a)) is also blocking the movement of the other diffusing atoms as shown clearly in
Figure 11 (b), where a plate of non-diffusing Mg exists on the left-hand side of the Mn grain.
Other parts of the pure Mg end-member were diffusing across the interface. On the other hand,
Mn freed from Mn17Nd2 and (MgNd) in zone #5, is observed to be dispersed or dissolved by the
growing (Mg3Nd) ternary solid solution with time.
According to the phase equilibria determined from diffusion couple experiments and alloys of the
end-members, the phase relationships are shown in Figure 12. In the next section, key alloys are
used to confirm the phase relations obtained by diffusion couples and to determine the solubility
limits of the solid solutions extended in the ternary.
Different annealing time intervals were used to assure that no diffusion layer was missing. It is
worth emphasizing that Dybkov [28] reported that diffusion layer formation is a sequential
process, where the first formed layers are those adjacent to the end members. Other layers start to
form in between when the initial layers reach their maximum thickness. Hence, some thin layers
may be consumed by the reaction between the diffusing species and the constituents of the layer
when they are over annealed. Thus, it was necessary to observe the microstructural changes after
every annealing step.
As can be understood from the diffusion behaviour of Mn, it has relatively lower diffusivity than
that of Mg and Nd. Thus, pure Mn was observed stationary in some diffusion zones of the
diffusion couples. However, the mobility of Mn atoms improves as they liberate from the Mn-
rich compounds. These atoms can be dissolved by some of the stable Mg-Nd compounds to form
extended ternary solubility.
15

4.2 Key alloys experiments
Although diffusion couple studies are very useful to determine the phase relationships, they are
not always successful. The possibility of missing a phase is a common phenomenon [29].
Sometimes the formation of some phases is too slow to form a detectable diffusion layer. Hence,
key alloys experiments are designed to verify the experimental results obtained from the diffusion
couples. Also, they are important to determine the maximum solid solubility of the binary and
ternary components in the system.
The constituent phases and their compositions in the annealed alloys are listed in Table 5. XRD
analysis was carried out to confirm the WDS findings. The XRD and WDS results of the
annealed samples are consistent.
Because of slow kinetics of the solid-solid phase transformation, step annealing approach was
used and the key alloys were analyzed after each annealing step including the as-cast alloys.
Accordingly, the phase equilibrium can be determined based on the microstructural and
morphological changes that occur after each annealing step.
Some alloys showed significant difference between their as-cast and annealed microstructures.
For instance, the as-cast microstructure of the sample #1 is shown in Figure 13 (a). Four phases
(Nd, Mg41Nd5, Mg12Nd and Mn) were observed and labeled in the same figure indicating non-
equilibrium condition. However, the microstructure of the same alloy after annealing for 14 days
showed three phases (Mg41Nd5, (Mg)Mn,Nd and Mn) as shown in Figure 13 (b). Mg12Nd was
observed in the as-cast alloy only, because it is a peritectic compound [16, 30] formed during the
non-equilibrium cooling. Mg12Nd was reported as metastable phase [17, 21, 31] below 552°C.
Thus, during annealing at 450°C, Mg12Nd decomposed to form Mg41Nd5 and (Mg)Nd, because its
composition falls between these two phases. The formation of Mg41Nd5 from Mg12Nd explains
16

the significant increase in the amount of Mg41Nd5 in the annealed alloy as shown in Figure 13
(b).
The as-cast microstructure of sample #2, shown in Figure 14 (a), represents the three phases, Mn,
(MgNd) and Mg2Nd. Figure 14 (b) is for the microstructure of the same alloy after annealing at
450°C for 14 days. The annealed microstructure contains the three phases Mn, (MgNd) and
Mg3Nd. It is assumed that (Mg3Nd) resulted from the eutectoidal decomposition of Mg2Nd,
which is known as a Laves phase with a narrow thermal stability range of 657-750°C [16, 17].
Thus, Mg2Nd is not stable at 450°C. The phase morphology in Figure 14 (b) is evidence that
Mg2Nd decomposed eutectoidally as following Mg2Nd↔MgNd+Mg3Nd.
WDS spot analysis was not able to acquire the composition of the lamellae of the eutectoid phase
individually, because they are very thin. However, the WDS area analysis of the eutectoid
structure (Figure 14 (b)) revealed the following composition: 57.4Mg-9.2Mn-33.4Nd at.%, which
falls in the (MgNd)+(Mg3Nd) two-phase field. Therefore, X-ray powder diffraction was
performed on the sample before and after annealing to identify the composition of the phases.
The X-ray spectra of sample #8, before and after annealing, are shown in Figure 15. Silicon was
used in the powder samples as a calibration standard to correct the zero shift and specimen
displacement. XRD spectrum of sample #8 before annealing showed the presence of MgNd and
Mg2Nd, which is consistent with WDS results as can be seen in Table 5. The XRD spectrum of
sample #8 after annealing showed Mg3Nd peaks instead of Mg2Nd. This elucidates the
eutectoidal decomposition of Mg2Nd.
Sample #3 was prepared to investigate the phase relationships near the Nd-rich corner. This can
also provide valuable information on the maximum solid solubility of Mg and Mn in α-Nd at
17

450°C. The as-cast and annealed microstructures of sample #3 are shown in Figure 16 (a) and (b),
respectively.
The as-cast microstructure contains Mn, Mn2Nd, (MgNd) and (α-Nd)Mg,Mn. The annealed
microstructure contains Mn17Nd2, (MgNd) and (α-Nd)Mg,Mn. Mn2Nd is another peritectic Laves
phase in the system with a thermal stability range of ~600-830°C [8, 24]. It was observed with a
lamellar morphology in the Mn2Nd+(α-Nd)Mg,Mn eutectoid microstructure of the as-cast sample as
shown in Figure 16 (a). The target annealing temperature in this study is 450°C, which is lower
than the stability temperature of Mn2Nd. According to the Mn-Nd phase diagram [8, 24], Mn2Nd
decomposes eutectoidally at ~570°C to form Mn23Nd6 and (α-Nd)Mn, and Mn23Nd6 decomposes
further at ~485°C to form Mn17Nd2 and (α-Nd)Mn. From the microstructures of this alloy,
Mn17Nd2 was obtained due to Mn2Nd decomposition. Evidence is seen in Figure 16 (b) where
Mn17Nd2 exists and the lamellar eutectoid microstructure changed to equiaxed grains.
Mn2Nd was not the only source of Mn17Nd2 formation, pure Mn played additional role in the
formation of Mn17Nd2. Based on the morphology of Mn grains, it seems that Mn reacted with Nd
to form Mn17Nd2 as shown in Figure 16 (b). This can be an indication that the affinity of Mn to
Nd atoms is higher than that of Mn to Mg; since Nd forms intermetallic compounds with Mn
unlike Mg and Mn. The Mn17Nd2 formation reaction in this case is reversible to produce pure Mn
due to Mn17Nd2 decomposition in the presence of Mg atoms as concluded from diffusion couples
experiments.
It was also noticed that the available Mn was not completely transformed into Mn17Nd2. Part of
Mn atoms were dissolved by (MgNd) and (α-Nd)Mg,Mn to form ternary solid solutions. It is
evident from the WDS analysis that the solid solubility of Mn in (MgNd) varied from 9 at.%, in
18

the as-cast alloy, to 13 at.% Mn upon annealing. The solubility of Mn in (α-Nd)Mg,Mn was
determined as 3.0 at.% Mn.
4.3 The Mg-Mn-Nd isothermal section at 450°C
According to the diffusion couples and key alloys experiments, the Mg-Mn-Nd isothermal
section at 450°C was constructed in the full concentration range as shown in Figure 17. The
solubility of both Mn and Nd in Mg was relatively small; it was measured as <1.0 at.% Mn and
<1.0 at.% Nd. These values fall within the WDS error of measurements, therefore, the solubility
limits of both Mn and Nd in Mg at 450°C are taken from the work of Drits et al. [25] as 0.5 wt.%
Mn and 1.3 wt.% Nd.
Five three-phase regions were observed and confirmed by WDS and XRD analysis. The system
showed unique phase relationships due to the presence of Mn. For instance, because Mn does not
form intermetallic compounds with Mg, it was always seen as either solute in the Mg-Nd
compounds or pure element in all microstructures. However, Mn forms a stable compound with
Nd at 450°C, and this compound, Mn17Nd2, is in equilibrium with (α-Nd)Mg,Mn, (MgNd) and pure
Mn. Therefore, the phase boundary lines are pointing towards the Mn-rich corner, except for the
(α-Nd)Mg,Mn+(MgNd)+Mn17Nd2 three-phase field, as shown in Figure 17.
5. Conclusions
The Mg-Mn-Nd isothermal section at 450°C was constructed in the full composition range using
diffusion couples and key alloys, analyzed by XRD, SEM/EDS/WDS and metallography. No
ternary compounds were observed in the system at 450°C. Key alloys were used to confirm the
phase relations obtained from diffusion couples and to determine the solid solubility limits.
19

Diffusion couples revealed that Mn has relatively lower diffusivity than that of Mg and Nd. Thus,
pure Mn was observed stationary in some diffusion zones forming Mn+(MgNd), Mn+(Mg3Nd)
and Mn+Mg41Nd5 two-phase regions. However, relatively small amount of Mn diffused towards
the interface with Mg end-member forming a continuous thin layer of pure Mn. The remainder of
Mn was dissolved in (α-Nd)Mg,Mn, (MgNd) and (Mg3Nd).
(α-Nd)Mg,Mn solid solution is considerably wider than the other single-phase regions, because of
the wide binary solid solubility of Mg in α-Nd at 450°C, which is about 7.5 at.% Mg. The
solubility of Mn in (MgNd) was characterized as a substitution of Mg by Mn atoms at constant
Nd concentration with maximum solubility of 13 at.% Mn. Other extended solid solution,
(Mg3Nd), was found to have ~3.7 at.% Mn.
The microstructure of the studied alloys showed that Mn is in equilibrium with most of the
phases in the system, since it appeared as individual particles within matrices of Mg-Nd phases.
Accordingly, the phase boundary lines are pointing towards the pure-Mn corner, except for the
(α-Nd)Mg,Mn+Mn17Nd2+(MgNd) phase field.
6. Acknowledgment
The authors would like to acknowledge NSERC for the financial support of the project through
MagNET network for excellence. Dr. Dmytro Kevorkov from Concordia University is also
acknowledged for providing useful suggestions.
20

References
[1] J.A. Gann, A.W. Winston, Magnesium and its alloys, Industrial and Engineering Chemistry,
19 (1927) 1193-1201.
[2] Y. Xu, L.S. Chumbley, F.C. Laabs Liquid metal extraction of Nd from NdFeB magnet scrap,
Journal of Materials Research, 15 (2000) 2296-2304.
[3] B. Mordike, K. Kainer, Magnesium alloys and their applications, Willey-VCH, Weinheim,
2000.
[4] J. Cai, G. Ma, Z. Liu, H. Zhang, A. Wang, Z. Hu, Influence of rapid solidification on the
mechanical properties of Mg-Zn-Ce-Ag magnesium alloy, Materials Science and Engineering: A,
456 (2007) 364-367.
[5] T. Takenaka, Y. Narazaki, N. Uesaka, M. Kawakami, Improvement of corrosion resistance of
magnesium alloys by surface film with rare earth element, Materials Transactions, 49 (2008)
1071-1076.
[6] D. Wellbeloved, P. Craven, J. Waudby, Manganese and Manganese Alloys, in: Ullmann's
Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co., 2000.
[7] V.T. Witusiewicz, F. Sommer, E.J. Mittemeijer, Reevaluation of the Fe-Mn phase diagram,
Journal of Phase Equilibria and Diffusion, 25 (2004) 346-354.
[8] A. Saccone, S. Delfino, R. Ferro, Systematics of rare earth-manganese alloys: Nd-Mn phase
diagram, International Journal of Materials Research, 84 (1993) 563-568.
[9] N. Tiner, The Solubility of Manganese in Liquid Magnesium, Metals Technology, (1945) 1-7.
[10] D. Petrov, M. Mirgalovskaya, I. Strelnikova, E. Komova, The Constitution Diagram for the
Magnesium-Manganese System, Institute of Materials Science, Academy of Sciences of the
Ukrainian SSR, 1 (1958) 142-143.
[11] A. Nayeb-Hashemi, J. Clark, The Mg−Mn (Magnesium-Manganese) system, Journal of
Phase Equilibria, 6 (1985) 160-164.
[12] J. Gröbner, D. Mirkovic, M. Ohno, R. Schmid-Fetzer, Experimental investigation and
thermodynamic calculation of binary Mg-Mn phase equilibria, Journal of Phase Equilibria and
Diffusion, 26 (2005) 234-239.
21

[13] Y. Kang, A. Pelton, P. Chartrand, P. Spencer, C. Fuerst, Critical evaluation and
thermodynamic optimization of the binary systems in the Mg-Ce-Mn-Y system, Journal of Phase
Equilibria and Diffusion, 28 (2007) 342-354.
[14] M. Asgar-Khan, M. Medraj, Thermodynamic description of the Mg-Mn, Al-Mn and Mg-Al-
Mn systems using the modified quasichemical model for the liquid phases, Materials
Transactions, 50 (2009) 1113-1122.
[15] Y. Kang, A. Pelton, P. Chartrand, P. Spencer, C. Fuerst, Thermodynamic database
development of the Mg-Ce-Mn-Y system for Mg alloy design, Metallurgical and Materials
Transactions A, 38 (2007) 1231-1243.
[16] A. Nayeb-Hashemi, J. Clark, The Mg−Nd system (Magnesium-Neodymium), Bulletin of
Alloy Phase Diagrams, 9 (1988) 618-623.
[17] S. Delfino, A. Saccone, F. Ferro, Phase relationships in the neodymium-magnesium alloy
system, Metallurgical and Materials Transactions A, 21 (1990) 2109-2114.
[18] R. Joseph, K. Gschneidner, Solid solubility of magnesium in some lanthanide metals,
Transactions of the American Institute of Mining and Metallurgical Engineers, 233 (1965) 2063-
2069.
[19] J.J. Park, L.L. Wyman, Phase Relationships in Magnesium Alloys., in, WADC Technical
report: Astia Document No. AD142110, 1957, pp. 1-27.
[20] M. Drits, E. Padezhnova, N. Miklina, The combined solubility of noedymium and zinc in
solid magnesium, Russian Metallurgy (Metally), 3 (1974) 143-146.
[21] L.L. Rokhlin, Magnesium alloyscontaining rare earth metals: Structure and properties,
Taylor and Francis, New York, 2003.
[22] H. Kirchmayr, W. Lugscheider, Die Zustandsbilder Neodym-Samarium- und Terbium-
Mangan, Zeitschrift für Metallkunde, 61 (1970) 22-23.
[23] S. Makhalenko, Y. Kuz'ma, X-ray diffraction investigations of the Nd-Mn system, Izvestiya
Akademii Nauk SSSR, Neorganicheskie Materialy (Inorganic Materials (USSR)), 26 (1990)
2432-2433.
[24] A.O. Mostafa, A.E. Gheribi, D. Kevorkov, M. Mezbahul-Islam, M. Medraj, Experimental
investigation and first-principle calculations coupled with thermodynamic modeling of the Mn–
Nd phase diagram, Calphad, 42 (2013) 27-37.
22

[25] M. Drits, Z. Sviderskaya, L. Rokhlin, Alloys of the Mg-Mn-Nd system adjacent to the
magnesium corner, Zhurnal neorganjoheskoi Khimii (Journal of Inorganic Chemistry), 7 (1963)
2771-2777.
[26] M. Drits, N. Bochvar, Phase equilibria in the Mg-Nd-Mn-Ni system, Doklady Akademii
Nauk SSSR 178, No. 2 (1968) 403-405.
[27] PANalytical, Ver2.2b (2.2.2), in, Almelo, The Netherlands, 01-11-2006.
[28] V.I. Dybkov, Growth Kinetics of Chemical Compound Layers, Cambridge Int Science
Publishing, Cambridge, UK, 2004.
[29] A.A. Kodentsov, G.F. Bastin, F.J.J. Van Loo, The diffusion couple technique in phase
diagram determination, Journal of Alloys and Compounds, 320 (2001) 207-217.
[30] M. Mezbahul-Islam, A. Mostafa, M. Medraj, Essential magnesium alloys binary phase
diagrams and their thermochemical data, Journal of Materials, (2014) Article ID: 704283, (In
press).
[31] S. Gorsse, C. Hutchinson, B. Chevalier, J. Nie, A thermodynamic assessment of the Mg-Nd
binary system using random solution and associate models for the liquid phase, Journal of Alloys
and Compounds, 392 (2005) 253-262.
23

Figures captions
Figure 1: (a) SEM micrograph of the solid-solid diffusion couple #1 annealed at 450°C for 10
days; (b) magnified part showing diffusion zones #2; (c) microstructure of
Mn17Nd2+(α-Nd)Mn end-member; (d) magnified part showing the phase morphology of
diffusion zone #3. The numbers represent the diffusion zones and correspond to those
in Table 1 and Figure 2
Figure 2: (a) Selective line-scan across zone #2 avoiding the secondary phases (Mn and
Mn17Nd2); (b) blind line-scan across diffusion couple #1. The numbers represent the
diffusion zones and correspond to those in Table 1 and Figure 1
Figure 3: Schematic representation of the phase relations observed in Figure 1 (c) showing the
phase equilibrium between (MgNd) (zone #2) and both (α-Nd)Mn from zone #1 and
(Mg3Nd) from zone #3. Also, it demonstrates the three-phase equilibria (α-
Nd)Mn+Mn17Nd2+(MgNd) between zones #1 and #2
Figure 4: EDS elemental mapping of diffusion couple #1. Reader is encouraged to see the online
version for the colored images
Figure 5: Phase equilibria obtained from diffusion couple #1 annealed at 450°C for 10 days. The
numbers represent the corresponding phase field based on the obtained phase
equilibrium. A and B represent the actual composition of the two end-members
obtained by ICP
Figure 6: (a) SEM micrograph of the solid-solid diffusion couple #2 annealed at 450°C for 4
days; (b) magnified part of zone #2 showing the fine particles of Mn due to Nd
consumption from Mn17Nd2. The numbers represent the diffusion zones and
correspond to those in Table 2
Figure 7: The phase relations determined from diffusion couple #2 annealed at 450°C for 4 days.
The numbers represent the corresponding phase field based on the obtained phase
24

equilibrium. A and B represent the actual composition of the two end-members
obtained by ICP
Figure 8: (a) SEM micrographs of the solid-solid diffusion couple #3 annealed at 450°C for 5
days; (b) magnified portion showing zones #4, 5 and 6; (c) magnified portion showing
Mg41Nd5/Mg3Nd interface (zones #3 and 4). The numbers represent the diffusion zones
and correspond to those in Table 3 and Figure 9
Figure 9: Measured composition profile across diffusion couple #3 using WDS line-scan with a
spatial distance of ~1µm
Figure 10: The phase equilibria determined by diffusion couple #3 annealed at 450°C for 5 days.
The numbers represent the corresponding phase field based on the obtained phase
equilibrium. A and B represent the actual composition of the two end-members
obtained by ICP
Figure 11: (a) SEM micrograph of the solid-solid diffusion couple #4 annealed at 450°C for
2days; (b) magnified portion showing different forms pure of Mn. The numbers
represent the diffusion zones and correspond to those in Table 4
Figure 12: The phase equilibria inferred from the diffusion couple experiments
Figure 13: The microstructure of sample #1 (a) the as-cast condition; (b) annealed at 450°C for
14 days
Figure 14: The microstructure of sample #8 (a) the as-cast condition; (b) annealed at 450°C for 5
days; (c) (MgNd)+(Mg3Nd) eutectoid
Figure 15: X-ray spectrum of sample #8 (a) before annealing; (b) annealed at 450°C for 5 days
25

Figure 16: The microstructure of sample #3 (a) the as-cast condition; (b) annealed at 450°C for
14 days
Figure 17: The Mg-Mn-Nd isothermal section at 450°C based on the phase equilibria of the
diffusion couples and key alloys
26

List of tables
Table 1: WDS spot analysis of diffusion couple #1
Composition (at.%)
Zone #
Corresponding phase
Mg
Mn
2.6
Nd
1 (end-member)
-
-
97.4
10.9
(α-Nd)Mn
Mn17Nd2
89.1
2
3
46.5
3.5
50.0
(MgNd)
21.3
-
3.7
100
25.0
-
(Mg3Nd)
Mn
4
75.0
-
25.0
Mg3Nd
5 (end-member)
89.3
75.0
-
-
10.7
25.0
Mg41Nd5
Mg3Nd
Table 2: WDS spot analysis of diffusion couple #2
Composition (at.%)
Zone
Corresponding phase
Mg
Mn
100
89.3
11.2
Nd
-
10.7
49.1
1 (end-member)
-
-
Mn
Mn17Nd2
(MgNd)
39.7
2
74.0
-
3.7
100
22.3
-
(Mg3Nd)
Mn
3
75.8
<2.0
22.5
(Mg3Nd)
4 (end-member)
-
100
-
<2.0
-
Mn
Mg41Nd5
(Mg3Nd)
88.3
75.1
10.7
23.6
27

Table 3: WDS spot analysis of diffusion couple #3
Composition (at.%)
Zone
Corresponding phase
Mg
100
Mn
-
Nd
-
1 (end-member)
Mg
2
3
-
100
-
Mn
88.4
-
-
100
11.6
-
Mg41Nd5
Mn
4
72.6
-
2.6
100
24.8
-
(Mg3Nd)
Mn
5
44.8
-
5.2
100
50.0
-
(MgNd)
Mn
6 (end-member)
44.6
6.0
-
6.1
2.0
89.5
49.3
92.0
10.5
(MgNd)
(α-Nd)Mg,Mn
Mn17Nd2
Table 4: WDS spot analysis of diffusion couple #4
Composition (at.%)
Zone
Corresponding phase
Mg
100
Mn
-
Nd
-
1 (end-member)
Mg
2
3
-
100
-
Mn
88.4
-
<2.0
100
10.5
-
Mg41Nd5
Mn
(Mg3Nd)
73.2
<2.0
25.3
4
73.1
-
75.0
<2.0
100
-
24.1
-
25.0
(Mg3Nd)
Mn
Mg3Nd
5 (end-member)
44.6
6.1
89.1
100
49.3
10.9
-
(MgNd)
Mn17Nd2
Mn
-
-
28

Table 5: Selected key alloys and their experimental results
Phase composition
ICP
(Actual composition at.%)
Sample
1
(at.%)
WDS
Mn
XRD
Results
WDS
Results
Mg
Mn
2.3
Nd
9.2
Mg
0.0
Nd
0.0
88.5
Mn
Mg
Mn
(Mg)Mn,Nd
Mg41Nd5
100
99.0
88.0
<1.0
<1.0
<0.1
11.0
Mg41Nd5
2
3
4
5
6
52.2
24.7
10.9
91.1
71.6
16.5
19.1
70.0
6.5
31.3
56.2
19.1
2.4
Mn
Mg3Nd
(MgNd)
Mn
(Mg3Nd)
(MgNd)
-
Small to be detected
100
-
41.7
9.1
49.2
α-Nd
(MgNd)
Mn17Nd2
(α-Nd)Mg,Mn
(MgNd)
Mn17Nd2
6.0
37.3
<2.0
3.0
91.0
49.5
11.3
13.2
87.5
Mn
Mn17Nd2
(MgNd)
Mn
Mn17Nd2
(MgNd)
0.0
<1.0
38.3
100
88.9
12.0
0.0
10.6
49.7
Mn
Mg
Mn
(Mg)Mn,Nd
Mg41Nd5
0.0
99.0
88.8
100
<1.0
<1.0
0.0
<1.0
10.4
Mg41Nd5
12.4
16.0
Mn
Mg41Nd5
(Mg3Nd)
Mn
Mg41Nd5
(Mg3Nd)
-
100
<2.0
2.7
-
88.0
75.0
10.9
22.3
7
8
70.8
57.6
5.8
2.4
23.4
40.0
Mn
(Mg3Nd)
Mn
(Mg3Nd)
-
75.0
100
3.5
-
21.5
MgNd
Mg3Nd
(MgNd)
(Mg3Nd)
50.7
Small to be detected
2.4
46.9
29