Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In magnetic shape memory alloy with enhanced ductility

Abstract

Two of the current challenges facing producers of Ni–Mn–In alloys are the achievement of small hysteresis and good ductility. Here, we present a dual-phase (b-Ni 51.8 Mn 31.4 In 16.8 and c-Ni 62.4 Mn 32.5 In 5.1) Ni 52 Mn 32 In 16 alloy prepared by the zone melting liquid metal cooling directional solidification method, which simultaneously shows small hysteresis (DT < 10 K) and good ductility (6.6%). In addi-tion, and more importantly, an inter-martensitic transition with a large magnetization jump occurs in this alloy. This is expected to fur-ther broaden the working temperature range of actuators and sensors that use this magnetic shape memory alloy. The sequence of the martensitic transformation can be shown by in situ X-ray diffraction to be austenite ! 10M ! 14M. Additionally, the second (c) phase dramatically enhances the entropy change of these structural transformations and shifts them to higher temperatures. During the direc-tional solidification, a novel banded-like microstructure, consisting of two layers, one of the b single phase and the other of the two phases coupled, forms at the low growth rate. A qualitative model is presented to explain the experimental observation, taking into account both the competitive nucleation and the growth of the phases. Experimental and theoretical analysis in the present work shows a linear relationship between the maximum spacing of the b single phase layer and the growth rate.

Full text

Banded-like morphology and martensitic transformation of dual-phase
Ni–Mn–In magnetic shape memory alloy with enhanced ductility
Y.J. Huang
a
, Q.D. Hu
a,⇑
, J. Liu
b
, L. Zeng
a
, D.F. Zhang
a
, J.G. Li
a,⇑
a
School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240, People’s Republic of China
b
Ningbo Institute of Materials Technology and Engineering, Chinese Academy of Sciences, Ningbo, People’s Republic of China
Received 23 April 2013; received in revised form 5 June 2013; accepted 5 June 2013
Abstract
Two of the current challenges facing producers of Ni–Mn–In alloys are the achievement of small hysteresis and good ductility. Here,
we present a dual-phase (b-Ni
51.8
Mn
31.4
In
16.8
and c-Ni
62.4
Mn
32.5
In
5.1
)Ni
52
Mn
32
In
16
alloy prepared by the zone melting liquid metal
cooling directional solidification method, which simultaneously shows small hysteresis (DT< 10 K) and good ductility (6.6%). In addi-
tion, and more importantly, an inter-martensitic transition with a large magnetization jump occurs in this alloy. This is expected to fur-
ther broaden the working temperature range of actuators and sensors that use this magnetic shape memory alloy. The sequence of the
martensitic transformation can be shown by in situ X-ray diffraction to be austenite !10M !14M. Additionally, the second (c) phase
dramatically enhances the entropy change of these structural transformations and shifts them to higher temperatures. During the direc-
tional solidification, a novel banded-like microstructure, consisting of two layers, one of the bsingle phase and the other of the two
phases coupled, forms at the low growth rate. A qualitative model is presented to explain the experimental observation, taking into
account both the competitive nucleation and the growth of the phases. Experimental and theoretical analysis in the present work shows
a linear relationship between the maximum spacing of the bsingle phase layer and the growth rate.
Ó2013 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Keywords: Magnetic shape memory alloys (MSMAs); Ni–Mn–In; Banded-like morphology; Martensitic transformation; Directional solidification
1. Introduction
The Heusler-type Ni–Mn–Z (Z = In, Sn, Sb) metamag-
netic shape memory alloys originally reported by Sutou
et al. [1] have attracted considerable attention due to their
multifunctionalities, such as large magnetic-field-induced
strain (MFIS) [2,3], inverse magnetocaloric effect [2,4],
giant magnetoresistance effect [5,6] and high magnetother-
mal conductivity [7]. All of these properties originate from
the magnetic field-induced structural transformation from
ferromagnetic austenite to paramagnetic/antiferromagnetic
martensite, which offers the possibility of application in
high-performance actuators [8], environment-friendly mag-
netic refrigerators [9], etc. However, these practical applica-
tions of Ni–Mn-based alloys are seriously restricted by
their large hysteresis (DT) and inherent brittleness. Thus,
both reducing the hysteresis and improving the ductility
are the two critical challenges for these alloys.
Much effort has been put to reduce the hysteresis of alloys
by tuning the chemical composition or microalloying with
the addition of a fourth constituent, e.g. changing the Mn/
Z[6,10], Ni/Z [11] and Ni/Mn ratios [12], or substituting
Ni, Mn or Z by Co [13],Fe[14],Al[15] or Si [16].Itis
reported that the magneto-structural transformation tem-
peratures and hysteresis of Ni–Mn-based alloys are very sen-
sitive to such factors as the alloy composition [1], heat
treatment [17] and hydrostatic pressure [18]. Therefore, the
composition-spread technique is used in the search for opti-
mal Ni–Mn–In alloy compositions with further reduced hys-
teresis [19,20]. In our previous work [21], we showed that
1359-6454/$36.00 Ó2013 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.actamat.2013.06.012
⇑
Corresponding authors. Tel./fax: +86 21 54744119.
E-mail addresses: qdhu@sjtu.edu.cn (Q.D. Hu), lijg@sjtu.edu.cn (J.G.
Li).
www.elsevier.com/locate/actamat
Available online at www.sciencedirect.com
Acta Materialia xxx (2013) xxx–xxx
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

most previous investigations have focused on Ni–Mn–In
alloys containing 650 at.% Ni. Recently, a large magnetic
entropy change with a small thermal hysteresis was observed
in high-Ni alloys (Ni > 50 at.%) under a low magnetic field
[22,23]. Hence, we proposed to investigate a high-Ni alloy
with a nominal composition of Ni
52
Mn
32
In
16
in this work.
Furthermore, Liu et al. [18] suggested that the large hyster-
esis of Heusler alloys can be reduced significantly by heating
without bias stress but cooling under optimized hydrostatic
pressure. However, a good ductility of the Heusler alloys is
required for such a deformation case.
If Ni–Mn–In alloys had good ductility, there would be
more opportunities to exploit them in applications that
also required functional phenomena such as good work-
ability, mechanical training [24,25], fatigue response [26]
and elastocaloric [27,28]. In general, the ductility of a mate-
rial is determined by its microstructure, including the grain
morphology, grain size and characteristics of the second
phase. For structural materials, the ductility can be
improved significantly by the preparation process, includ-
ing the thermo-mechanical process [29] and directional
solidification [30]. Other potential ways to improve the
ductility of Ni–Mn-based alloys are through tailoring the
alloy composition to introduce a second phase [31] or to
control the solidification process to obtain a microstructure
of highly oriented polycrystals [32]. However, only a few
documents have reported the preparation of highly tex-
tured polycrystals by controlling the solidification process.
Additionally, a large reversible magnetic field-induced
strain could be achieved in such highly textured polycrys-
tals [3,33], as the anisotropy of such a structure is analo-
gous to that of single crystals.
In this work, Ni
52
Mn
32
In
16
alloy prepared by a zone
melting liquid metal cooling (ZMLMC) directional solidifi-
cation method simultaneously shows small hysteresis
(DT< 10 K) and good ductility (6.6%). Highly oriented
polycrystals with uniform axial composition are obtained
by controlling the temperature gradient (G) and the crystal
growth rate (R). An inter-martensitic transition with a
large magnetization jump occurs in this alloy, meaning that
the working temperature range of actuators and sensors
could be broadened further by using this magnetic shape
memory alloy [34]. Additionally, it is interesting that a
banded-like morphology along the growth direction forms
during the directional solidification. To the best of our
knowledge, this novel microstructural feature has never
been reported before in Heusler alloys, and we expect it
to offer new options for the preparation of functionally
gradient materials and in situ composites. A qualitative
model considering both competitive nucleation and the
growth of phases is proposed to explain the experimental
observations.
2. Experimental procedure
Homogeneous Ni
52
Mn
32
In
16
(in at.%) button ingots
were arc-melted four times using the starting materials
nickel, manganese and indium (with 99.99% purity). The
buttons were remelted and drop cast in chilled copper
molds to obtain master rods with a diameter of 7.0 mm
and a length of 50 mm. The rods were subsequently
remelted and directionally grown in alumina crucibles
under different conditions, as listed in Table 1. Directional
solidification was carried out in a homemade ZMLMC fur-
nace heated by high-frequency induction. The rods direc-
tionally solidified under different conditions were labeled
as in Table 1. Due to the use of Ga–In–Sn liquid alloy as
the cooling medium during directional solidification, the
temperature in the directionally solidified part should be
close to room temperature. Therefore, the temperature gra-
dient at the solid–liquid (S/L) interface front can be esti-
mated by
GSGL¼TMT0
L
where G
S
is the temperature gradient in the solid phase, G
L
is the temperature gradient in front of the solid–liquid
interface, T
M
is the melting point of Ni
52
Mn
32
In
16
alloy
(1251 K, as determined by differential scanning calorime-
try, DSC), T
0
is the room temperature (293 K) and Lis
the distance between the solidification interface and the
cooling liquid metal.
Longitudinal sections of the directionally solidified rods
were etched in a solution of 99 ml of methanol and 5 g of
FeCl
3
for optical observation (OM; Axio Imager A1m).
The dual-phase microstructural characteristics were
observed by compositional model in a scanning electron
microscope (JEOL 7600F). The compositional distribution
of the directionally solidified rods was tested by energy-dis-
persive spectroscopy (EDS). Identification of the crystal
structure and the preferred orientation was conducted by
X-ray diffraction (XRD; Rigaku, Ultima IV) using Cu
Karadiation at a scan speed of 5°min
–1
. An in situ
XRD measurement was conducted using a D8 ADVANCE
instrument at a heating/cooling rate of 10 K min
–1
.
After being annealed at 1173 K for 24 h, all character-
istic transformation temperatures were taken from the
peak values determined by DSC (NETZSCH 204F1) at
a rate of 5 K min
–1
for both heating and cooling. A Quan-
tum Design PPMS-9T (EC-II) magnetometer was used to
measure the temperature dependence of the DC magneti-
zation, both the heating and cooling rates of which were
the same as that for the DSC. Stress–strain curves were
obtained by testing with a compression machine (MTS
858) at room temperature, with an initial strain rate of
310
4
s
1
.
3. Results and discussion
3.1. Microstructural characteristics
Fig. 1a shows an overall image of the longitudinal sec-
tion in G1000R2 alloy, showing three parts along the
growth direction, i.e. the initial zone, the steady zone and
2Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

the quenching zone. Surprisingly, numerous periodic bands
(indicated by black arrows in Fig. 1a) are evident along the
growth direction in the steady zone. At higher magnifica-
tion, the periodic banded-like microstructure is constituted
of two layers, one of the bsingle phase and the other of two
phases (b+c) coupled (Fig. 1b). The bphase has a mar-
tensitic structure (Fig. 1c), indicating that the martensitic
transformation temperature is above the ambient tempera-
ture. Furthermore, the period bands are concave and
nearly parallel to the S/L interface, so the band spacing
should be related to the solidification condition (discussed
below in Section 4).
It should be noted that the microstructural observation
of the two phases coupled layer and the functional tests
were carried out in the steady zone, because the crystals
grow relatively steadily in this zone.
The composite microstructure of the two phases (b+c)
is presented in Fig. 1d, exhibiting a discrepancy from the
single phase predicted in the phase diagram [35] (Fig. 2a).
According to the Ni–Mn–In ternary phase diagram, the b
phase (B2/L2
1
-ordered phase) refers to the off-stoichiome-
tric Ni
2
MnIn Heusler alloy at 1123 K and the cphase is
the second phase, consisting of disordered face-centered
cubic (fcc) Ni existing in the Ni-rich corner. The EDS maps
Table 1
Nomenclature, growth morphology of Ni
52
Mn
32
In
16
alloy directionally solidified under different conditions.
Alloys Temperature gradient (G
L
,Kcm
–1
) Growth rate (R,lms
–1
) Two phases coupled growth morphology
G200R5 200 5 Dendritic structure
G200R15 15 Dendritic structure
G200R100 100 Well-developed dendritic structure
G500R5 500 5 Cellular structure with leaf-like side branches
G500R20 20 Dendritic structure + cellular structure
G500R100 100 Dendritic structure
G1000R2 1000 2 Cellular structure
G1000R5 5 Cellular structure
G1000R15 15 Cellular structure
Fig. 1. (a) Photograph of a longitudinal section of the G1000R2 alloy, showing the banded-like microstructure (indicated by black arrows) in the steady
zone during crystal growth. (b) Enlarged optical image of the banded-like microstructure, exhibiting two layers, one of the bsingle phase and the other of
the two phases (b+c) coupled. (c) Enlarged SEM image of the black rectangle marked in (b), showing the martensitic structure. (d) SEM (compositional
model) microstructure in the steady zone of the G1000R2 alloy, exhibiting a composite microstructure of two phases (b+c). The black phase refers to c;
the bright one is bphase. (e–g) EDS maps of the elements Ni, Mn and In corresponding to (d), respectively. The yellow dotted circles indicate the cphase.
Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx 3
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

indicate that the cphase is an In-poor phase (marked by
yellow dotted circles in Fig. 1e–g), in accordance with the
phase diagram. The cphase is in fact similar to the second
phase in other alloys, such as Ni–Mn–Ga [31], Ni–Mn–In–
Co [13,17] and Ni–Fe–Ga–Co [36], which have been
reported to contribute to the improvement of ductility.
To investigate the formation of this dual-phase micro-
structure, the solidus and liquidus lines of the cphase at
1273 K were extrapolated approximately by combining
the melting point of Ni–Mn–In alloys in the bsingle phase
region and the solidus/liquidus lines of cphase in the Ni–
Mn [37] and Ni–In [38] binary phase diagrams, as indicated
by the blue and red dotted lines in Fig. 2a, respectively.
Although an apparent compositional transition between
the two phases is observed (Fig. 2b), the compositional dis-
tribution shows that the Mn content in the three regions is
almost steady at 32 at.%. This implies that the solidification
path of Ni
52
Mn
32
In
16
alloy is nearly along the black dotted
line (the Mn content of 32 at.%) marked in Fig. 2a. The
formation of the composite microstructure in the present
ternary alloy could thus possibly be deduced during cool-
ing. Two possible formation reactions may be written as
follows:
(1) L !L0þcðfcc-NiÞ!bðperitectic reactionÞ.
(2) L !b(isomorphous reaction, i.e. bis the primary
phase and the residual liquid transforms directly into
cphase).
On the one hand, the cphase should be the high temper-
ature phase in the present work, implying that the primary
phase is the cphase during solidification and the peritectic
reaction most likely occurs. On the other hand, the micro-
structural observation in the quenching zone (Fig. 2c)
shows that the cphase with a well-developed dendritic
structure grows first, then reacts peritectically with liquid
to form the bphase. The solidification path could thus be
determined to be the peritectic reaction. Similar solidifica-
tion paths were in fact observed in Ni–Mn–Ga [39] and
Ni–Fe–Ga–Co alloys [40]. Due to the kinetics of the peri-
tectic reaction, this formation reaction is hardly completed
during the non-equilibrium process. Therefore, the residual
cphase has a dendritic/cellular morphology at room tem-
perature and is expected to grow along the cellular/den-
dritic interface (Fig. 2b).
During the directional solidification, the microstructural
characteristics and the growth morphology in the two
phases coupled layer are presented in Fig. 3 and summa-
rized in Table 1, respectively.
At G
L
= 200 K cm
–1
, the main morphology of the two
phases coupled growth is dendritic, as shown in Fig. 3a–
c. When the growth rate increases to 100 lms
–1
(Fig. 3c),
the primary dendrite seems to be shorter, implying a ten-
dency to change from dendritic to equiaxed growth.
At G
L
= 500 K cm
–1
and R=5lms
–1
, the dendritic
morphology turns into cellular structure with leaf-like side
branches (Fig. 3d). At the high temperature gradient of
G
L
= 1000 K cm
–1
, the cellular morphology of the two
phases coupled growth is more evident. Moreover, the
maximum spacing of the bsingle phase layer is reduced
by the increments in the growth rate, showing a linear rela-
tionship with reciprocal growth rate (Fig. 4a). When the
growth rate increases to 100 lms
–1
, the bsingle phase layer
disappears, indicating that the two phases coupled growth
is stable. Nevertheless, the temperature gradient has little
effect on the layer spacing of the bsingle phase (Fig. 4a).
Additionally, the fraction of the cphase in the two phases
coupled layer can be adjusted by controlling the tempera-
ture gradient and the growth rate, as shown in Fig. 4b.
For a given temperature gradient, the fraction of the c
phase decreases with decreasing growth rate. It is impor-
tant to note that the fraction of the cphase in directionally
solidified alloys is lower than that of suction casting and
drops down to 4% at the temperature gradient of
1000 K cm
–1
.
3.2. Preferred orientation
In order to investigative the structural anisotropy, the
preferred orientation and phase structure were identified
Fig. 2. (a) Ternary phase diagram of Ni–Mn–In system at 1123 K [35]. The green line refers to the composition of a constant 50 at.% Ni alloy. The solidus
and liquidus lines of fcc-Ni (the cphase) at 1273 K were extrapolated approximately, as indicated by the blue and red dotted lines, respectively. (b) SEM
morphology of dual-phase composite microstructure in the steady zone of the G1000R2 alloy, indicating a compositional transition from the bphase to c.
(c) SEM photographs of the S/L interface (the yellow lines), exhibiting a faceted/plane front interface. (For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of this article.)
4Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

by XRD, as shown in Fig. 5, which indicating that the
strongest peak is in the ð0010Þplane of 10M modulated
martensite (Fig. 5b) at G= 1000 K cm
–1
and R=2lms
–1
.
However, at G= 200 K cm
–1
and R= 100 lms
–1
, two
stronger peaks, ð0010Þand ð1210Þ, of 10M martensitic
phase are detected. More peaks are evident in arc-melted
Ni
52
Mn
32
In
16
alloys, indicating random orientations. Addi-
tionally, the diffraction of the cphase can be also be
observed in all conditions (Fig. 5a). This is consistent with
the microstructural observation (Fig. 3).
The preferred orientation of the high-temperature b
phase (austenitic structure) should be investigated further
because of the martensitic structure at room temperature.
Although the bphase with L2
1
atomic order is transformed
from the B2 structure, the B2 structure has conventionally
been treated as being L2
1
ordered. In view of the crystallo-
graphic information on 10M martensite, it is long period
stacking of 10 ð110ÞAclose-packed planes. The unit cell
can be described as a monoclinic lattice, the crystallo-
graphic axes being aligned along ½
110(the a-axis), ½001
(the b-axis) and ½110(the c-axis) of the cubic austenite
[41], as shown in Fig. 5c. These axes are referred to as
“monoclinic axes”. Thus, it is reasonable to conclude that
the basal planes ð0010ÞM(//(1 1 0)
A
) are stacked along
Fig. 3. (Compositional model in SEM) Microstructure of the two phase (b+c) coupled layer in the steady zone: (a) G200R5, dendritic structure; (b)
G200R15, dendritic structure; (c) G200R100, well-developed dendritic structure; (d) G500R5, cellular structure with leaf-like side branches; (e) G500R20,
dendritic structure + cellular structure; (f) G500R100, dendritic structure; (g) G1000R2, cellular structure, showing the broad bsingle phase layer; (h)
G1000R5, cellular structure; (i) G1000R15, cellular structure. The periodic bands disappear at the high growth rate of 100 lms
1
.
Fig. 4. (a) The maximum spacing for the bsingle phase layer as a function of reciprocal growth rate for Ni
52
Mn
32
In
16
alloys during directional
solidification; (b) fraction of the cphase in the two phases coupled layer vs. the temperature gradient and the growth rate, in reference to the fraction of the
cphase at suction casting.
Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx 5
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

the c-axis. Thus, the c-axis direct ion ([1 1 0]
A
) of the mar-
tensitic phase is developed as the preferred direction of
Ni
52
Mn
32
In
16
alloy during directional solidification. In
other words, the preferred direction of directionally solidi-
fied Ni
52
Mn
32
In
16
alloys is the ð0010Þplane of 10M mod-
ulated martensite, in line with that of Co–Ni–Al alloys [42].
3.3. Martensitic transformation and mechanical properties
It is well known that magnetic field-induced martensitic
transformation has led to many interesting phenomena
being reported in Ni–Mn–In alloys, and that the transfor-
mation temperature is highly sensitive to the alloy’s com-
position [1,2]. A dual-phase microstructure is presented in
this work. Hence, the effect of the cphase on the martens-
itic transformation is crucial. The samples for functional
tests were cut from the two phases coupled layer and from
the bsingle phase layer in the G1000R2 rod.
The DSC curves depicted in Fig. 6a clearly show that the
ordinary forward and reverse martensitic transformations
are manifested as large exo- and endothermal peaks,
respectively, in the vicinity of 350 K. However, another
two low-temperature peaks with the same signs, much less
pronounced than that of martensitic transformation, are
presented around 270 K with a certain of thermal hystere-
sis (DT), indicating a structural transition took place. Con-
sidering the results of scanning electron microscopy (SEM)
and XRD at room temperature (Figs. 1c and 5, respec-
tively), an inter-martensitic transition is assumed to occur.
Although the martensite at room temperature was identi-
fied by XRD (Fig. 5), it was nevertheless labeled M1 for
consistency with the low-temperature martensite (M2).
The corresponding DSC peak temperatures marked by
arrows in Fig. 6a are indicative of the forward (M
p
,T
If
)
and reverse (A
p
,T
Ir
) martensitic and inter-martensitic
M1–M2 transformations, respectively. The entropy changes
(DS) across these structural transformations were calculated
by DH/T
m
, where T
m
is an average of the forward and
reverse transformation temperatures. These characteristic
parameters are summarized in Table 2. The presence of
the cphase causes an obvious shift in the martensitic and
inter-martensitic transformation temperatures, and dramat-
ically enhances the DSfor these transformations. The DSof
Fig. 5. (a) XRD pattern of the transverse section (perpendicular to the growth direction) in directionally solidified rods and arc-melted Ni
52
Mn
32
In
16
alloys. The diffraction of the cphase (arc-melted Ni
62
Mn
33
In
5
alloy) is also presented as a reference. (b) Details in the black rectangle indicated in Fig. 5a,
showing 10M modulated martensite at room temperature. (c) The crystallographic relationship between austenite and martensite, the unit cells of which
are represented by black and orange lines, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web
version of this article.)
Fig. 6. (a) DSC curves and (b) M–T curves under a magnetic field of 2 T
(heating/cooling rate of ±5 K min
–1
) for the two phases (b+c) and the b
single phase in G1000R2 alloy. The inset shows M–T plotted in the
temperature range from 335 to 375 K.
6Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

the inter-martensitic transition is much smaller than that of
the martensitic transformation. Additionally, the DTof the
martensitic transformation is increased significantly by
the formation of the cphase. This could be attributed to
the pinning effect on the martensitic transformation
and the non-uniform compositional distribution of these
two phases.
It is worth noting that the single broad peak at 296 K
without thermal hysteresis should be the Curie temperature
of 10M martensite. However, the Curie temperature of
10M martensite is absent in the two phases coupled layer,
which may be devoured as the intermartensitic transforma-
tion is shifted to a higher temperature by the cphase.
Fig. 6b shows M–T curves under a magnetic field of 2 T.
The formation of the cphase enhances the change in mag-
netization with the martensitic transformation. A similar
effect of the cphase was also observed in Ni–Mn–Ga alloys
[43]. A higher value of the magnetization jump than that of
martensitic transformation occurs around 280 K with a
large thermal hysteresis. This can be ascribed to the inter-
martensitic transformation. However, this is different from
the inter-martensitic transition observed in Ni
52.6
Mn
23.6
-
Ga
23.8
alloy, where no jump in magnetization is seen [34].
This therefore suggests that the working range of use of
meta-magnetic shape memory actuators and sensors could
be broadened further using such alloys.
In order to identify further the low-temperature mar-
tensitic structure, an in situ XRD analysis was conducted
out, as shown in Fig. 7, in which a 14M martensitic struc-
ture is presented at 200 K. As a result, the transformation
sequence on cooling can be manifested as: A (paramag-
netic) !A (ferromagnetic) !10M (paramagnetic) !10M
(ferromagnetic) !14M (paramagnetic) !14M (ferromag-
netic). This is in accordance with that of other alloys
(Ni
52.6
Mn
23.6
Ga
23.8
[34],Ni
54
Fe
19
Ga
27
[44]). However, the
detailed features of this inter-martensitic transition could
be further elucidated by additional investigations, such as
of the MFIS [3], the magnetocaloric effect [4] and elastoca-
loric [27]. These are beyond the scope of the present work,
but could be subjects for study in the future.
Here, the compressive tests for the two layers (of the b
single phase and the two phases coupled) were applied
to evaluate the ductility of the directionally solidified
Ni
52
Mn
32
In
16
rod, as shown in Fig. 8. Good ductility with
a compressive strain of 7% for the two phases coupled layer
is achieved at room temperature, which is higher by a fac-
tor of 1.7 than that in the bsingle phase layer and is com-
parable to that of Ni
49
Fe
18
Ga
27
Co
6
(7% [45]). In
comparison to the bsingle phase layer, there is also a nota-
ble increase of 300 MPa in the compressive strength.
These enhancements of ductility and strength are attrib-
uted to the introduction of the cphase. Similarly, microal-
loying with Co introduced the cphase, which increase the
ductility of Ni
50
Mn
34
In
16
[13]. This good strength–ductility
combination is of great importance for broadening the
working range of Ni–Mn–In alloys.
The stress–strain curve of the two phases coupled layer
shown in Fig. 8 can be divided into two stages (I and II).
In stage I, a transition point and concave work hardening
are evident, which are attributed to inter-martensitic trans-
formation before dislocation slip. Moreover, the transition
stress for the two phases coupled layer (256 MPa) is higher
than that of the bsingle phase layer (202 MPa). Both tran-
sition stresses are much higher than that for Ni
54
Fe
19
Ga
27
single crystal (90 MPa [44]) because both the cphase and
the grain boundaries hinder the transition. In the stage
II, there is a distinct stress peak followed by work softening
down to fracture, which can be ascribed to the onset of
plastic instability.
4. The formation of the banded-like microstructure
The above experimental results reveal a banded-like
morphology, and a linear relationship between the spacing
of the bsingle phase layer and the growth rate. However,
the temperature gradient has a limited effect on the spacing
(Fig. 4). Indeed, a banded-like microstructure was also
observed in the directional solidification of various other
peritectic systems, including Nd–Fe–B [46], Fe–Ni [47]
and Cu–Sn [48]. A qualitative explanation for the forma-
tion of peritectic bands in binary systems and at medium/
high growth rates has been given by Trivedi and Kurz
[49,50], but several aspects of ternary systems and such a
low growth rate as features in this work are still unclear.
Furthermore, there is a serious shortage of thermodynam-
ics data for Ni–Mn–In alloys, such as general knowledge of
the solidification behavior, the shape of the liquidus surface
and more precise phase diagrams. As a result, we have tried
our best here to describe the formation of the banded-like
morphology qualitatively.
From the compositional analysis (Fig. 2b), it can be seen
that the content of Mn is almost steady at 32 at.% during
the solidification. Consequently, a rough schematic peritec-
tic diagram for Ni–Mn–In alloys at constant 32 at.% Mn is
given by Fig. 9. For the sake of simplicity, a single cycle
was taken as an example to investigate the formation of
the banded-like microstructure, and this is shown in
Fig. 10.
It is worth noting that no second phase (the cphase) has
previously been reported in Ni–Mn–In alloys with an Ni
component of 650 at:%. This could be ascribed to the
alloy composition being below the peritectic composition
(C
p
), as shown in Fig. 9. In other words, the previous alloys
would not go through the peritectic isotherm during
solidification.
To establish a qualitative explanation for the formation
of the banded-like morphology, the following assumptions
were made by considering the experimental results.
(1) The solute transport is governed by diffusion in the
liquid only.
(2) No convection effects are present in the liquid.
Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx 7
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

(3) The diffusion in the solid is negligible due to the tem-
perature of the Ga–In–Sn liquid alloy close to the
ambient temperature.
(4) The growth conditions are assumed to give a faceted/
plane front solidification (Fig. 2c).
(5) Local equilibrium is established very rapidly at the
solid/liquid interface in the mushy zone.
As soon as solidification starts, a solute boundary layer
is built up ahead of the S/L interface, where the liquid com-
position varies with time during the initial period of solid-
ification until the steady-state condition is reached. At the
steady state, neither the bphase nor the cphase is stable
(Fig. 3), indicating that the cphase is expected to form
along the cellular/dendritic interface. Since the banded-like
microstructure is present in the steady zone, the following
discussion is limited to the steady state.
For cellular growth (at the low growth rate of
R< 100 lms
–1
), the primary cphase forms first from the
liquid and quickly becomes enveloped by the peritectic b
phase. Growth of the two phases coupled layer subse-
quently starts (step 1 in Fig. 9) when the cooling passes
through the peritectic isotherm (T
p
in Fig. 9). At this time,
the composition field and the growth temperature ahead of
the cellular S/L interface at the steady state can be approx-
imated by [46]:
Ci
LðzÞ¼C0þGD
miRexp Rz
D
 ð1Þ
Ti
c¼Ti
LðCðzÞÞ  GD
Rði¼c;bÞð2Þ
where Ci
LðzÞis the solute concentration of the liquid phase
at the growing phase front, Ti
cðzÞis the cellular growth tem-
perature for the growing phase at the S/L interface, C
0
is
the alloy composition or far-field composition, m
i
is the liq-
uidus slope of the growing phase, Dis the diffusion coeffi-
cient in the liquid, zis the distance from the S/L interface
to the liquid and Ti
LðCðzÞÞ is the liquidus temperature of
the growing phase for the composition CðzÞ.
From Eq. (1), it can be seen that a compositional differ-
ence for the liquid phase is present between the c/liquid
interface and the b/liquid interface. As a consequence,
there are two simultaneous processes. First, solute Ni
atoms at the c/liquid interface are depleted by the growth
of the Ni-rich cphase and solute Ni atoms at the b/liquid
interface are enriched by the growth of the bphase. How-
ever, the enrichment rate at the b/liquid interface is much
slower than the depletion rate at the c/liquid interface,
because the b/liquid interface is much larger and the com-
position of the bphase is closer to that of the alloy (C
0
).
Secondly, the peritectic bcolonizes the primary cphase
by the solute inter-diffusion in the liquid (Fig. 3).
The depletion of solute Ni atoms thus decreases the tem-
perature at the c/liquid interface. As the temperature
decreases to a critical value, the bphase stabilizes and
begins to nucleate directly from the liquid phase (step 2
in Fig. 9), probably through heterogeneous nucleation on
the growing bphase. Meanwhile, some of the bphase
grows out rapidly in the form of a sphere, as shown in
Fig. 10b. In other words, the starting condition for growth
of the bsingle phase layer is that the local temperature at
the c/liquid interface should be lower than the local nucle-
ation temperature of the bphase at any z>0:
Tc
LðzÞ<Tb
LðCðzÞÞ  DTb
Nð3Þ
In this case, the bphase sphere engulfs the cand prop-
agates sideways at a rate much larger than that of the
solute-flux-controlled cellular front because of the colli-
sion-limited growth front of the bsphere (V
1
and V
2
in
Fig. 10b). During this period of explosive growth (step 3
in Fig. 9), the growth velocity of the bsingle phase layer
perpendicular to the macroscopic isotherms (V
1
) slows
down drastically as it reaches warmer liquid, while the lat-
eral growth rate stays high, as shown in Fig. 10c.
Subsequently, the growing interface falls behind the iso-
therm velocity, which develops an inverse recalescence [49]
and reduces the interface temperature. Meanwhile, the sol-
ute Ni atoms at the b/liquid interface are enriched by the
solute rejection with the explosive growth of the bphase,
which increases the driving force for the nucleation of c
at the interface. Hence, the cphase will nucleate again at
the b/liquid interface (step 4 in Fig. 9), and the condition
can be given by
Tb
LðzÞ<Tc
LðCðzÞÞ  DTc
Nð4Þ
After the cphase has nucleated from the growing b
interface, the interface of the bsingle phase layer becomes
morphologically unstable (Fig. 10d). With the growth of
the cphase, peritectic reaction occurs again between the
cphase and the liquidus as the local temperature at the
c/liquid interface is below T
p
(step 5 in Fig. 9), then
the bphase again nucleates from the cphase to form two
Table 2
Inter-martensitic transition temperature (K), martensite–austensite transformation temperature (K), Curie temperature of 10M martensite (K), thermal
hysteresis (DT, K) and entropy change (DS, J (kg K)
–1
) for different phases of G1000R2 alloy.
Phase Inter-martensitic
transition temperature
Martensite–austensite
transformation temperature
Curie temperature of 10M
martensite
Thermal hysteresis Entropy change
T
If
T
Ir
A
p
M
p
T10M
CDT
(14M–10M)
DT
(10M–A)
DS
(14M–10M)
DS
(10M–A)
b258 271 343 339 296 13 4 0.12 0.25
b+c277 286 356 347 – 9 9 0.66 1.64
8Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

phases coupled growth (Fig. 10e). As a result, a thin layer
of the cphase is evident at the interface between the two
layers (the bsingle phase and the two phases coupled), as
shown in Figs. 1b and 3g. Thus the cycle is completed
and starts again at Fig. 10a, which leads to the formation
of the banded-like microstructure.
Considering the temperature function of the steady-state
growth behaviors, the local temperature at the c/liquid
interface and the b/liquid interface can be written, respec-
tively, as follows:
Ti
LðzÞ¼Ti
LðCðzÞÞ  GD
RþGzð5Þ
where
Tc
LðCðzÞÞ ¼ TpþmcC0CpGD
mbRexp Rz
D

ð6aÞ
Tb
LðCðzÞÞ ¼ TpþmbC0CpGD
mcRexp Rz
D

ð6bÞ
The starting condition for the explosive growth of the b
single phase layer can then be rewritten by
Fig. 7. In situ XRD pattern of a longitudinal section of the G1000R2
alloy. The crystal structure at 200 K is 14M with a monoclinic unit cell.
Fig. 8. Compressive stress–strain curve of G1000R2 alloy, showing that
some inter-martensitic transitions occurred during the compression test
and also showing good ductility of 6.6%.
Fig. 9. A schematic peritectic diagram with linear solidus and liquidus
lines at constant 32 at.% Mn. The diagram shows the undercoolings
required for the nucleation of the cand bphases. DTc
Nand DTb
Nare the
undercoolings required for the heterogeneous nucleation of cand b
phases, respectively.
Fig. 10. Schematics for the formation of a cycle band during directional
solidification.
Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx 9
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

mcC0CpGD
mbRexp Rz
D

GD
R
þGz<mbC0CpGD
mcRexp Rz
D

DTb
N
ð7Þ
Considering that the bsingle phase layer should start at
the c/liquid interface, only nucleation at the interface is
considered. Consequently, the value of zshould be zero.
After simplification, one obtains:
G
R>1
DðmbmcÞðC0CpÞDTb
N
mc
mbmb
mcþ1
!
ð8Þ
This indicates that the formation of the bsingle phase
layer is easier at a lower growth rate for a given tempera-
ture, which agrees with the microstructural observation
(Fig. 5).
From Eq. (4), it can be seen that the growth interface of
the bsingle phase layer becomes unstable as the cphase
nucleates again at the b/liquid interface, which means a
maximum spacing (z
max
) for the bsingle phase layer. Sim-
ilarly, by substituting Eq. (6), the instability condition for
the bsingle phase layer can be given by:
mbC0CpGD
mcRexp Rz
D

GD
R
þGz<mcC0CpGD
mbRexp Rz
D

DTc
N
ð9Þ
The maximum undercooling for nucleation of the c
phase ahead of the bgrowing interface, which is at the
maximum spacing (z
max
) of the bsingle phase layer, can
be obtained by making the derivatives equal with respect
to zin both terms in Eq. (9):
G¼mb
mcmc
mb

GD
Rexp Rz
D

R
D
 ð10Þ
z
max
can then be obtained from Eq. (10):
zmax ¼D
Rln mc
mbmb
mc
 ð11Þ
This suggests that the maximum spacing of the bsingle
phase layer is inversely proportional to the growth rate for
a given composition. Moreover, the effect of the tempera-
ture gradient can be ignored. This is consistent with the
experimental results, as shown in Fig. 5, where the linear
relationship between the maximum spacing of the bsingle
phase layer and the reciprocal growth rate is obvious.
For dendritic growth at the high growth rate (100 lms
–1
),
the curvature undercooling and solute undercooling result-
ing from radial solute rejection at the dendritic tip cannot
be ignored, due to there being a dendritic front rather than
the faceted/plane front assumed for cellular growth. Thus,
the temperature field for the cand the bat the dendritic S/
L interface can be obtained as follows:
Ti
dðzÞ¼Ti
LðCðzÞÞ  GD
RDTi
SDTi
Cð12Þ
where DTi
Sand DTi
Care the solute undercooling and curva-
ture undercooling expressed by [46]:
DTi
Sffimið1kiÞC0CR
rD

1=2
ð13Þ
DTi
C¼2mið1kiÞC0CRr
D

1=2
ð14Þ
where r¼1=4p2,Kis the solute distribution coefficient
and Cis the Gibbs–Thomson coefficient.
Similar to cellular growth, by combing Eqs. (3), (5),(6)
and (12), and substituting Eqs. (13) and (14), the starting
condition for the bsingle phase layer at the high growth
rate can be given by:
G
R>1
DAR1=2þðmbmcÞðC0CpÞDTb
N
mc
mbmb
mcþ1
!
ð15Þ
where
A¼C0C
D

1=21
r

1=2
þ2ðrÞ1=2
!
mbð1kbÞ

1=2ðmcð1kcÞÞ1=2

In comparison with Eq. (8), the additional section,
AR1=2, on the right-hand side of Eq. (15) is the resisting
force for the formation of the bsingle phase layer at the
high growth rate, which comes mainly from the curvature
undercooling and solute undercooling of dendritic growth,
and is proportional to the square root of the growth rate.
Hence, a larger temperature gradient is needed to form
the bsingle phase layer at the high growth rate. Such a
solidification condition is hardly satisfied in this work,
though it can be reached by other rapid solidifications.
Similar structures have also been known to form during
rapid solidification of dendritic alloys [50]. Additionally,
the oscillation effect induced by the high growth rate is lar-
ger than that of the cellular growth. Consequently, the
starting condition for growth of the bsingle phase layer
given in Eq. (15) is hardly satisfied.
5. Conclusion
ZMLMC directional solidification was applied to pre-
pare highly oriented polycrystal Ni
52
Mn
32
In
16
alloy, which
simultaneously shows small hysteresis (DT<10K) and
good ductility (6.6%). During solidification at the high
growth rate of 100 lms
–1
, a kind of two phases coupled
growth microstructure is presented. However, at a low
growth rate, a novel microstructure of a banded-like mor-
phology is developed, which consists of two layers, one of
the bsingle phase and the other of the two phases coupled.
To a certain extent, this is a nonuniform microstructure,
but it opens up new avenues for preparing functionally gra-
dient materials and in situ composites, which can broaden
the working temperature range of Ni–Mn–In alloys [18].
According to this work, uniformly directional rods with
10 Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012

good ductility can be obtained at a high growth rate. Fur-
thermore, a qualitative model is presented to explain the
formation of the banded-like morphology. The theoretical
analysis is consistent with the experimental results, and a
linear relationship between the maximum spacing of the
bsingle phase layer and the growth rate is obtained.
The directionally solidified Ni
52
Mn
32
In
16
alloy has a
strong structural anisotropy analogous to that of single
crystals, the preferred direction of which is identified as
being the (0 0 10) plane of 10M modulated martensite, i.e.
the h110idirection of austenite.
More significant is that an inter-martensitic transforma-
tion with a large magnetization jump is observed in this
alloy. Therefore, the transformation sequence for the cur-
rent alloy on cooling can be manifested further by in situ
XRD as: A (paramagnetic) !A (ferromagnetic) !10M
(paramagnetic) !10M (ferromagnetic) !14M (paramag-
netic) !14M (ferromagnetic). This means that the work-
ing temperature range of actuators and sensors could be
broadened further by using this magnetic shape memory
alloy. The cphase causes an obvious shift of the martens-
itic and intermartensitic transformation temperatures, and
dramatically enhances the entropy change for these
transformations.
Acknowledgements
This work was financially supported by the National
Natural Science Foundation of China under Grant No.
51161120362 and Shanghai Science & Technology Com-
mittee under Contract No. 11JC1405900. The authors are
grateful to the Instrumental Analysis Center of Shanghai
Jiao Tong University for technical support.
References
[1] Sutou Y, Imano Y, Koeda N, Omori T, Kainuma R, Ishida K, et al.
Appl Phys Lett 2004;85:4358.
[2] Krenke T, Duman E, Acet M, Wassermann EF, Moya X, Man
˜osa L,
et al. Phys Rev B 2007;75:104414.
[3] Liu J, Aksoy S, Scheerbaum N, Acet M, Gutfleisch O. Appl Phys Lett
2009;95:232515.
[4] Moya X, Manosa L, Planes A, Aksoy S, Acet M, Wassermann EF,
et al. Phys Rev B 2007;75:184412.
[5] Sharma VK, Chattopadhyay MK, Shaeb KHB, Chouhan Anil, Roy
SB. Appl Phys Lett 2006;89:222509.
[6] Yu SY, Liu ZH, Liu GD, Chen JL, Cao ZX, Wu GH, et al. Appl
Phys Lett 2006;89:162503.
[7] Zhang B, Zhang XX, Yu SY, Chen JL, Cao ZX, Wu GH. Appl Phys
Lett 2007;91:012510.
[8] Karaca HE, Karaman I, Basaran B, Ren Y, Chumlyakov YI, Maier
HJ. Adv Funct Mater 2009;19:983–98.
[9] Liu ZH, Aksoy S, Acet M. J Appl Phys 2009;105:033913.
[10] Krenke T, Acet M, Wassermann EF, Moya X, Manosa L, Planes A.
Phys Rev B 2006;73:174413.
[11] Xuan HC, Ma SC, Cao QQ, Wang DH, Du YW. J Alloys Compd
2011;509:5761.
[12] Kumar DMR, Rao NVR, Raja MM, Rao DVS, Srinivas M, Muthu
SE, et al. J Magn Magn Mater 2012;324:26.
[13] Feng Y, Sui JH, Gao ZY, Dong GF, Cai W. J Alloys Compd
2009;476:935.
[14] Krenke T, Duman E, Acet M, Moya X, Manosa L, Planes A. J Appl
Phys 2007;102:033903.
[15] Pathak AK, Dubenkoa I, Mabon JC, Stadler S, Ali N. J Phys D:
Appl Phys 2009;42:045004.
[16] Pathak AK, Dubenkoa I, Stadlerb S, Ali N. J Alloys Compd
2011;509:1106.
[17] Liu J, Woodcock TG, Scheerbaum N, Gutfleisch O. Acta Mater
2009;57:4911.
[18] Liu J, Gottschall T, Skokov KP, Moore JD, Gutfleisch O. Nat Mater
2012;11:620.
[19] Takeuchi I, Famodu OO, Read JC, Aronova MA, Chang K-S,
Craciunescu C, et al. Nat Mater 2003;2:180.
[20] Cui J, Chu YS, Famodu OO, Furuya Y, Hattrick-Simpers JAE,
James RD, et al. Nat Mater 2006;5:286.
[21] Huang YJ, Hu QD, Li JG. Appl Phys Lett 2012;101:222403.
[22] Hu FX, Wang J, Shen J, Gao B, Sun JR, Shen BG. J Appl Phys
2009;105:07A940.
[23] Liu FS, Wang QB, Ao WQ, Yu YJ, Pan LC, Li JQ. J Magn Magn
Mater 2012;324:514.
[24] Gaitzsch U, Potschke M, Roth S, Rellinghaus B, Schultz L. Scr
Mater 2007;57:493.
[25] Dadda J, Maier HJ, Karaman I, Karaca HE, Chumlyakov YI. Scr
Mater 2006;55:663.
[26] Efstathiou C, Sehitoglu H, Kurath P, Foletti S, Davoli P. Scr Mater
2007;57:409.
[27] Cui J, Wu Yiming, Muehlbauer J, Hwang Yunho, Radermacher R,
Fackler S, et al. Appl Phys Lett 2012;101:073904.
[28] Man
˜osa L, Gonza
´lez-Alonso D, Planes A, Bonnot E, Barrio M,
Tamarit Josep-Lluı
´s, et al. Nat Mater 2010;9:478.
[29] Huang YJ, Chen ZG, Zheng ZQ. Scr Mater 2011;64:382.
[30] Xie J, Fu H, Zhang Z, Jiang Y. Intermetallics 2012;23:20.
[31] Cai W, Zhang J, Gao ZY, Sui JH, Dong GF. Acta Mater
2011;59:2358.
[32] Jiang C, Liu J, Wang J, Xu L, Xu H. Acta Mater 2005;53:1111.
[33] Gaitzsch U, Potschke M, Roth S, Rellinghaus B, Schultz L. Acta
Mater 2009;57:365.
[34] Seguı C, Chernenko VA, Pons J, Cesari E, Khovailo V, Takagi T.
Acta Mater 2005;53:111.
[35] Miyamoto T, Nagasako M, Kainuma R. J Alloys Compd
2013;549:57.
[36] Sui J, Gao Z, Yu H, Zhang Z, Cai W. Scr Mater 2008;59:874.
[37] Guo C, Du X. Intermetallics 2005;13:525.
[38] Massalski TB. In: Massalski TB, editor. Binary alloys phase
diagrams. Materials Park (OH): ASM International; 1990. p. 2267.
[39] Chen J, Gharghouri MA, Hyatt CV. Proceedings of SPIE, vol.
5387. Bellingham (WA): SPIE; 2004. p. 549.
[40] Liu J, Scheerbaum N, Hinz D, Gutfleisch O. Acta Mater
2008;56:3177.
[41] Pons J, Chernenko VA, Santamarta R, Cesari E. Acta Mater
2000;48:3027–38.
[42] Li JZ, Huang B, Li JG. J Cryst Growth 2011;317:110.
[43] Cai W, Zhang J, Gao ZY, Sui JH. Appl Phys Lett 2008;92:252502.
[44] Hamilton RF, Sehitoglu H, Efstathiou C, Maier HJ. Acta Mater
2007;55:4867.
[45] Panchenko E, Chumlyakov Y, Maier HJ, Timofeeva E, Karaman I.
Intermetallics 2010;18:2458.
[46] Zhong H, Li S, Lu H, Liu L, Zou G, Fu H. J Cryst Growth
2008;310:3366.
[47] Hunziker O, Vandyoussefi M, Kurz W. Acta Mater 1998;46:6325.
[48] Liu D, Li X, Su Y, Luo L, Zhang B, Guo J, et al. Mater Lett
2011;65:1628.
[49] Trivedi R. Metall Trans 1995;26A:1583.
[50] Kurz W, Trivedi R. Metall Trans 1996;27A:625.
Y.J. Huang et al. / Acta Materialia xxx (2013) xxx–xxx 11
Please cite this article in press as: Huang YJ et al. Banded-like morphology and martensitic transformation of dual-phase Ni–Mn–In
magnetic shape memory alloy with enhanced ductility. Acta Mater (2013), http://dx.doi.org/10.1016/j.actamat.2013.06.012