PHYSICAL REVIEW MATERIALS 3, 044406 (2019)
Editors’ Suggestion
Giant barocaloric effect in all-d-metal Heusler shape memory alloys
Araceli Aznar,1 Adrià Gràcia-Condal,2 Antoni Planes,2 Pol Lloveras,1,* Maria Barrio,1 Josep-Lluís Tamarit,1
Wenxin Xiong,3 Daoyong Cong,3,† Catalin Popescu,4 and Lluís Mañosa2,‡
1Departament de Física, EEBE and Barcelona Research Center in Multiscale Science and Engineering,
Universitat Politècnica de Catalunya, Eduard Maristany, 10-14, 08019 Barcelona, Catalonia
2Departament de Física de la Matèria Condensada, Facultat de Física, Martí i Franquès 1,
Universitat de Barcelona, 08028 Barcelona, Catalonia
3State Key Laboratory for Advanced Metals and Materials, University of Science and Technology Beijing,
No. 30 Xueyuan Road, Haidian District, Beijing 100083, People’s Republic of China
4CELLS-ALBA Synchrotron, 08290 Cerdanyola del Vallès, Catalonia, Spain
(Received 3 December 2018; revised manuscript received 26 February 2019; published 16 April 2019)
We have studied the barocaloric properties associated with the martensitic transition of a shape memory
Heusler alloy Ni50Mn31.5Ti18.5 which is composed of all-d-metal elements. The composition of the sample
has been tailored to avoid long-range ferromagnetic order in both austenite and martensite. The lack of
ferromagnetism results in a weak magnetic contribution to the total entropy change, thereby leading to a large
transition entropy change. The combination of such a large entropy change and a relatively large volume change
at the martensitic transition gives rise to giant barocaloric properties in this alloy. When compared to other shape
memory Heusler alloys, our material exhibits values for adiabatic temperature and isothermal entropy changes
significantly larger than values reported so far for this class of materials. Furthermore, our Ni50Mn31.5Ti18.5 also
compares favorably to the best state-of-the-art magnetic barocaloric materials.
DOI: 10.1103/PhysRevMaterials.3.044406
I. INTRODUCTION
Shape memory materials undergo a martensitic transition
from a high-temperature high-symmetry cubic phase (austen-
ite) to a low-temperature lower-symmetry close-packed phase
(martensite). The occurrence of this structural transition con-
fers these alloys a unique thermomechanical behavior which is
at the origin of a series of functional properties such as shape
memory, pseudoelasticity, and superelasticity [1]. More re-
cently, giant magnetocaloric [2,3] and mechanocaloric effects
[4,5] have also been reported to occur in shape memory alloys
which make them excellent candidates for environmentally
friendly solid-state refrigeration technologies [6,7].
For nonmagnetic alloys, the structural change at the
martensitic transition is mostly described by a shear of {110}
planes along the [1¯10] direction of the cubic phase, with
a negligible volume change [8]. Therefore, the martensitic
transition in these alloys is strongly sensitive to uniaxial stress
(either compressive or tensile) but almost insensitive to hy-
drostatic pressure, and these alloys exhibit giant elastocaloric
effects but negligible barocaloric effects [9]. However, in
magnetic shape memory alloys the strong coupling between
magnetism and structure results in noticeable volume changes
at the phase transition which may give rise to giant barocaloric
effects [5].
*pol.lloveras@upc.edu
†dycong@ustb.edu.cn
‡lluis.manosa@fmc.ub.edu
In their austenitic phase most magnetic shape memory
alloys exhibit a Heusler structure with d-group elements (Ni
and Mn) at the 8c (Ni) and 4a (Mn) positions (in Wyckoff no-
tation), and p-group atoms (Ga, Sn, In, etc.) at the 4b positions
[10]. Recently magnetic shape memory Heusler alloys have
been developed with all-d-metal elements [11,12]. In these
alloys Ti replaces p-group elements and it has been proved
that the d-metal Ti provides a stabilization of the Heusler
structure by d-d hybridization between Ti at the 4b position
and its nearest-neighbor element at the 8c or 4a position [12].
An interesting property of these all-d-group magnetic shape
memory alloys is the large relative volume change at their
martensitic transition which is significantly larger than for any
other magnetic shape memory alloy. This peculiarity points
to a marked sensitivity of the transition to the application of
mechanical stresses, and particularly to a hydrostatic pressure,
thereby making these materials excellent candidates to exhibit
giant barocaloric effects.
The giant magnetocaloric and mechanocaloric effects ex-
hibited by shape memory alloys originate from the first-order
martensitic transition which involves a significantly large en-
tropy change (St ). While in nonmagnetic alloys the entropy
is mainly due to lattice vibration (phonon contribution), in
magnetic shape memory alloys there is also a contribution
from magnetic degrees of freedom which usually competes
with the vibrational term [13], in such a way that the transition
entropy change decreases as the austenite becomes more and
more ferromagnetically ordered [14]. This competing sce-
nario provides a dilemma [15] because on the one hand large
entropy changes are desirable for caloric effects but on the
other hand large magnetization and volume changes are also
2475-9953/2019/3(4)/044406(7) 044406-1 ©2019 American Physical Society
ARACELI AZNAR et al. PHYSICAL REVIEW MATERIALS 3, 044406 (2019)
required for a good tunability of the transition temperature
with external field and pressure. While magnetic order seems
to be unavoidable to achieve giant magnetocaloric properties,
it would be desirable to find shape memory alloys with
large volume changes and weak magnetic order since they
would exhibit enhanced barocaloric properties. In this work
we report on a material meeting these requirements.
We have developed a Heusler alloy formed by all d ele-
ments (Ni, Ti, and Mn). The composition has been tailored
such that the alloy undergoes a martensitic transition slightly
below room temperature. The interplay between magnetism
and structure results in a relatively large volume change at the
martensitic transition. However, the absence of ferromagnetic
order results in a weak magnetic contribution to the entropy
giving rise to a large-transition entropy change. It is shown
that this combination of large entropy and volume changes at
the martensitic transition brings about outstanding barocaloric
performances to this alloy.
II. EXPERIMENTAL DETAILS
Polycrystalline button ingots were prepared by arc-melting
pure Ni, Mn, and Ti elements in a high-purity Ar atmosphere.
A small amount of B was also added to enhance the grain
boundary cohesion and to improve the mechanical properties
[16]. The ingots were sealed into evacuated quartz tubes
and annealed at 1173 K for 48 h to ensure homogenization,
followed by water quenching. The nominal composition of the
sample was (Ni50Mn31.5Ti18.5)99.8B0.2. For the sake of sim-
plicity in the following we label our sample Ni50Mn31.5Ti18.5.
A sample with dimensions 8.86 × 8.81 × 5.92 mm3
was cut from the ingot for calorimetric measurements un-
der hydrostatic pressure (barocaloric measurements). Smaller
samples were also cut for conventional differential scanning
calorimetry (DSC) and magnetic measurements. A number
of these small samples were ground to produce powder for
x-ray measurements. These powdered samples were further
annealed under vacuum at 773 K for 5 h in order to minimize
internal strains.
High-pressure powder diffraction experiments were per-
formed at the beamline BL04-MSPD at the CELLS-ALBA
synchrotron using a monochromatic beam of λ = 0.4246 Å
[17]. The beamline is equipped with Kirkpatrick-Baez mirrors
to focus the x-ray beam down to 20 × 20 μm2 (FWHM).
The sample was loaded into a membrane diamond anvil cell
(DAC) of 400-μm culet size. The pressure chamber was a
200-μm hole drilled on a 45-μm preindented stainless steel
gasket. NaCl powder was used as a pressure marker [18].
A cryostat was used to control the temperature of the DAC.
The sample-to-detector distance (200 mm) and the beam
center position were calibrated using FIT2D software [19] from
LaB6 diffraction data measured in the same conditions as the
sample.
Conventional DSC measurements were performed with
a Q2000 TA Instruments calorimeter, and magnetic mea-
surements were performed using a superconducting quan-
tum interference device (SQUID) magnetometer. For the
thermomagnetization measurements, the heating and cool-
ing rates were 1 K min−1. Calorimetry under hydro-
static pressure was carried out by means of a bespoke
FIG. 1. (a) Differential scanning calorimetry curves on cooling
(bottom curve) and heating (upper curve) runs. (b) Temperature
dependence of the magnetization measured under different applied
magnetic fields. (c) Inverse of the magnetic susceptibility as a
function of temperature (line is a linear fit to the data in the austenitic
phase). (d) Isothermal magnetization as a function of magnetic
field in austenite (green) and martensite (blue). Black solid line
corresponds to the predicted Curie-Weiss behavior computed using
Tc and μ determined from susceptibility data.
experimental device based on a Cu-Be pressure cell (op-
erating up to 6 kbar) adapted as a differential thermal an-
alyzer (DTA) by using Peltier modules as thermal sen-
sors [20]. DW-Therm M0.200.02 (Huber Kältemaschinenbau
GmbH) was used as pressure-transmitting liquid. The temper-
ature was controlled by an external circulating thermal bath
(Lauda Proline RP 1290, 183–473 K) and typical temperature
rates of ∼2 K min−1 were chosen.
III. RESULTS AND DISCUSSION
DSC curves taken at a temperature rate 5 K min−1 across
the martensitic transition of Ni50Mn31.5Ti18.5 are shown in
Fig. 1(a). The lower curve corresponds to the exothermal
transition on cooling (forward martensitic transition) and the
upper curve to the endothermal transition on heating (reverse
martensitic transition). By taking the characteristic transition
temperature (Tt ) as the peak temperature of each DSC curve
we obtain 243 and 255 K for the forward and reverse transi-
tions, respectively, with a 12 K thermal hysteresis. The latent
heat of the transition is obtained by integration of DSC curves,
with values of −20.8 ± 1.0 and 19.4 ± 1.0 kJ kg−1, which
correspond to transition entropy changes (St ) of −85 ± 6
and 76 ± 5 J kg−1 K−1 for the forward and reverse transitions,
respectively.
Figure 1(b) shows the temperature dependence of the
magnetization upon cooling and heating at selected values
of the magnetic field. The martensitic transition is clearly
observed as a sharp decrease (increase) in magnetization on
cooling (heating). For the studied range (up to 5 T) there
is no noticeable shift in the martensitic transition with mag-
netic field, and transition temperatures are in good agreement
with those determined from DSC. Figure 1(c) displays the
044406-2
GIANT BAROCALORIC EFFECT IN ALL-d-METAL … PHYSICAL REVIEW MATERIALS 3, 044406 (2019)
inverse of magnetic susceptibility, χ−1, as a function of
temperature computed from magnetization data at 5 mT. The
linear increase of χ−1 with increasing temperature observed
at high temperatures points to a paramagnetic behavior of
the austenite. A fit of a Curie-Weiss dependence to the data
renders a magnetic moment per formula unit μ = 5.6μB and
a paramagnetic Curie temperature Tc = 74 K. It is worth
noticing that Tc can be affected by a considerable error
because it is obtained from the extrapolation of a linear
fit to a set of data which are at temperatures far from Tc.
In the austenitic phase the magnetization linearly increases
with magnetic field, as illustrated in Fig. 1(d), which shows
measured isothermal magnetization at 270 K as a function of
magnetic field. Actually, the reliability of μ and Tc derived
from low field susceptibility data is confirmed by the excellent
agreement found between experimental M vs H data (solid
green symbols) and the predicted Curie-Weiss behavior using
μ = 5.6μB and Tc = 74 K, which is plotted as a black line in
Fig. 1(d). Furthermore, it has been reported that the austenitic
phase of Ni50Mn50−xTix is antiferromagnetic for x  20 [11].
For our sample x = 18.5 and the austenitic phase is paramag-
netic. Magnetization in martensite is significantly lower than
in austenite, and although the isothermal magnetization (at
230 K) also increases linearly with increasing magnetic field
(with a slope lower than in austenite) the lack of paramagnetic
behavior in χ−1 and the weaker magnetic field dependence of
the magnetization point to the existence of antiferromagnetic
correlations in martensite, as reported for the martensitic
phase in other magnetic shape memory alloys [21].
Diffraction patterns recorded at high temperature (299 K)
and low temperature (200 K) and atmospheric pressure are
shown in Fig. 2(a). The austenite can be indexed as a cubic
(Fm¯3m) structure and the martensite is well described by
an orthorhombic structure (Pmma). The lattice parameters
were obtained by identifying the peak positions and using
WIN_AFMAIL software (included in the WIN_CELL package)
[22]. We obtained a = 5.938(2) Å for the cubic austenite,
and a = 8.544(43) Å, b = 5.523(35) Å, and c = 4.376(18) Å
for the orthorhombic martensite. Diffraction patterns were
also recorded along one isobaric cooling ramp at atmospheric
pressure [Fig. 2(b)] and at 9 kbar [Fig. 2(c)]. At room
temperature and atmospheric pressure the sample is in the
austenitic state (with a small amount of retained marten-
site) and below ∼215 K diffraction peaks from the low-
temperature orthorhombic phase begin to develop. At the
lowest temperature the sample is in the orthorhombic phase
although a small peak from the cubic austenite is still present.
For an applied pressure of 9 kbar the sample is predominantly
in the austenitic phase at room temperature. There is some
presence of martensitic phase in a larger amount than at
atmospheric pressure. Upon cooling, diffraction peaks from
the martensitic structure start to grow at T ∼ 265 K, and at
the lowest measured temperature (200 K) the sample is in the
martensitic phase.
A simple and straightforward analysis of the evolution
of the martensitic transition can be achieved by monitoring
the intensity of the 220 cubic peak. In Fig. 2(d) we plot
the number of counts normalized by the value at ambient
temperature, for atmospheric pressure and 9 kbar. The shift
of the martensitic transition towards higher temperatures with
FIG. 2. (a) Diffraction patterns in the austenite (299 K, black
line) and in the martensite (200 K, red line), with indexation of the
peaks corresponding to the cubic (black labels) and orthorhombic
(red labels) phases for austenite and martensite, respectively. Peaks
from NaCl and MnO are also indicated. Isobaric evolution of selected
diffraction patterns obtained on cooling (b) at atmospheric pressure
and (c) under 9 kbar, respectively. (d) Normalized number of counts
of the 220 peak of the cubic phase as a function of temperature for
atmospheric pressure (black squares) and for 9 kbar (red circles).
Lines are guides to the eye. For clarity only a typical error bar is
shown for each run.
applied pressure is clear from the data. The transition at
atmospheric pressure is sharp, but under applied pressure it
spreads over a broader temperature range. The spreading of
the transition is not a particular effect of Ni-Mn-Ti but rather
it is related to nonhydrostatic stress due to the choice of NaCl
as a pressure-transmitting medium.
Illustrative examples of the baseline-corrected calorimetric
curves obtained under constant hydrostatic pressure are shown
in Figs. 3(a) (heating runs) and 3(b) (cooling runs). Appli-
cation of pressure shifts the martensitic transition to higher
temperatures, as shown in Fig. 4(a), where we plotted the peak
temperature of calorimetric curves as a function of hydrostatic
pressure, for both forward and reverse martensitic transitions.
The increase in transition temperatures is linear, with slopes
dT/d p = 3.2 ± 0.1 and dT/d p = 1.9 ± 0.1 K kbar−1 for
forward and reverse transitions, respectively.
The transition entropy change is computed as
St =
∫ T2
T1
1
T
dQ
dT
dT, (1)
044406-3
ARACELI AZNAR et al. PHYSICAL REVIEW MATERIALS 3, 044406 (2019)
FIG. 3. Calorimetric curves at selected values of applied hydro-
static pressure recorded during (a) heating and (b) cooling runs.
Entropy (with respect to the entropy at T0 = 220 K) as a function
of temperature for selected values of applied hydrostatic pressure for
(c) heating and (d) cooling runs.
where dQdT =
˙Q
| ˙T | , and T1 and T2 are freely chosen tempera-
tures below and above the martensitic transition, respectively.
The values obtained are plotted as a function of pressure in
Fig. 4(b), which shows a decrease in |St | with increasing
hydrostatic pressure, this decrease being more pronounced
for the forward transition. A linear fit to the data gives
d|St |/d p = −4.4 ± 0.8 J K−1 kg−1 kbar−1 for the forward
FIG. 4. (a) Transition temperature and (b) transition entropy
change as a function of pressure. Blue symbols stand for the forward
transition and red symbols stand for the reverse transition. Lines are
linear fits to the data.
transition and d|St |/d p = −2.6 ± 0.9 J K−1 kg−1 kbar−1
for the reverse transition.
Following the procedure described in Ref. [9] we have
computed the barocaloric effect using our calorimetric data
under hydrostatic pressure [Figs. 3(a) and 3(b)] and specific
heat data at atmospheric pressure [23] (which is assumed to
be pressure independent).
We first obtained the entropy vs temperature curves at
selected values of applied hydrostatic pressure from
S(T, p) =
⎧⎪⎪⎪⎪⎨
⎪⎪⎪⎪⎩
∫ T
T0
CMp
T
dT, T  T1
S(T1, p) +
∫ T
T1
1
T
(
Cp + dQdT
)
dT, T1 < T  T2
S(T2, p) +
∫ T
T2
CAp
T
dT, T2 < T,
(2)
where CMp and CAp are, respectively, the specific heats of
martensite and austenite, and Cp = xCMp + (1 − x)CAp , where
x is the fraction of martensite. The other quantities have the
same meaning as explained before. S(T, p) represents the
entropy referred to the value at a selected temperature of T0 =
220 K. The obtained entropy curves are shown in Figs. 3(c)
and 3(d) for reverse and forward transitions, respectively. The
scatter in transition entropy values [Fig. 4(b)] gives rise to
a relative error in S(T, p) at high temperatures in the range
8–10 %, and therefore the small differences in the entropy of
the austenitic phase for different applied pressures [Figs. 3(c)
and 3(d)] are not relevant.
From these entropy curves it is straightforward to compute
the isothermal entropy change induced by the application of a
pressure p as
S(T, 0 → p) = S(T, p) − S(T, 0), (3)
and the adiabatic temperature change as
T (S, 0 → p) = T (S, p) − T (S, 0). (4)
For a release of pressure (p → 0), equivalent expressions
hold.
Since pressure stabilizes the martensitic phase (as cooling
does), the entropy and temperature changes induced by ap-
plying a pressure p are computed from calorimetric curves
on cooling. Similarly, a release of a pressure p promotes the
transition from martensite to austenite and therefore entropy
and temperature changes in that case are computed from
heating curves. Results are shown in Fig. 5. As reported for
other magnetic shape memory alloys, the barocaloric effect in
Ni50Mn31.5Ti18.5 is conventional; i.e., isothermal application
(removal) of pressure reduces (increases) entropy and adia-
batic application (removal) of pressure increases (decreases)
temperature. It is found that the maximum values for the
isothermal entropy change (|S|max) and adiabatic temper-
ature change (|T |max) increase as pressure increases, as
illustrated in Fig. 6. For all pressures, values for |S|max
and |T |max are larger for the forward transition than for the
reverse one. The difference is ascribed to a larger transition
entropy change and a stronger sensitivity of the transition
temperature to pressure.
044406-4
GIANT BAROCALORIC EFFECT IN ALL-d-METAL … PHYSICAL REVIEW MATERIALS 3, 044406 (2019)
FIG. 5. Isothermal entropy changes corresponding to the (a) re-
moval and (b) application of selected values of hydrostatic pressure.
Vertical dashed lines indicate the reversibility region. Adiabatic tem-
perature changes corresponding to the (c) removal and (d) application
of selected values of hydrostatic pressure.
Application of barocaloric materials to refrigeration tech-
nologies requires a good reversibility of their barocaloric
effect under cyclic application and removal of pressure. Such
a reversibility near a first-order transition depends on the
competition between the width of thermal hysteresis and the
sensitivity of transition temperatures upon pressure. For a
conventional barocaloric effect reversible barocaloric effects
have been shown to occur within a temperature interval
bounded by the start of the reverse transition on heating
at atmospheric pressure and the start of the forward transi-
tion on cooling under an applied pressure [24,25]. Within
this temperature region application and removal of pressure
FIG. 6. (a) Maximum isothermal entropy change and (b) maxi-
mum adiabatic temperature change. Red symbols correspond to the
reverse transition upon application of pressure and blue symbols
correspond to the forward transition upon removal of pressure.
FIG. 7. Isothermal entropy and adiabatic temperature changes
for state-of-the-art magnetic barocaloric materials. For the sake of
clarity each material is designated by an indicative label and not
by its actual composition. All data are obtained from calorimetric
measurements (quasidirect method) except |T |max for GdSiGe
and LaFaSi which correspond to thermometric measurements. The
minimum pressure required to obtain |S|max is indicated for each
compound. Data correspond to the first application (or first removal)
of pressure, and are extracted from Refs. [24,25,27–34].
carries the state of the material through minor hysteresis
loops, and the reversibility in the barocaloric effect is directly
related to the reversibility in the fraction of material that
undergoes the forward and reverse transitions in each cycle.
For our Ni50Mn31.5Ti18.5 the reversibility region [indicated
by dashed vertical lines in Figs. 5(a) and 5(b)] spans from
245 to 260 K, and the maximum estimated reversible isother-
mal entropy change within this region amounts to |Srev| 
35 J kg−1 K−1.
The barocaloric effect has recently been studied in different
shape memory Heusler alloys. In Table I we compare present
results for Ni50Mn31.5Ti18.5 to those reported for other shape
memory Heusler alloys [5,24,26–29] for which we have listed
the largest reported entropy and temperature change values
(corresponding either to the first application or first removal
of pressure). It is clear that our Ni50Mn31.5Ti18.5 exhibits the
largest |S|max and |T |max values among all shape mem-
ory Heusler alloys. Furthermore, |Srev| is also larger than
reversible values reported for other shape memory Heusler
alloys [24].
In Fig. 7 we compare our values for isothermal entropy
change and adiabatic temperature change to those reported for
state-of-the-art magnetic barocaloric materials [24,25,27–34].
Because reversible values (upon repeatedly applying and re-
moving pressure) are not reported for many of the materials,
our comparison is based on entropy and temperature values
for the first application and first removal of pressure. It is
apparent that Ni50Mn31.5Ti18.5 exhibits the largest values,
only comparable to those for hexagonal Ni2In-type Mn-Co-
Ge-In (and related) compounds [33,34]. However, compared
to Ni2In-type compounds our alloy exhibits much better
mechanical properties. While Ni2In-type compounds rapidly
044406-5
ARACELI AZNAR et al. PHYSICAL REVIEW MATERIALS 3, 044406 (2019)
TABLE I. Sample composition, transition entropy change |St |, applied pressure p, isothermal entropy change |S|max, adiabatic
temperature change |T |max, and refrigerant capacity RC.
|St | dTd p p |S|max |T |max RC
Sample (J kg−1K−1) (K kbar−1) (kbar) (J kg−1K−1) (K) (J kg−1) Ref.
Ni49.26Mn36.08In14.66 27 1.8 2.6 24.4 – – [5]
Ni51.2Mn32.5In16.3 40.8 1.65 2.5 32 4 200 [24]
Ni42.47Co8.87Mn31.67Ga14.98In2.01 22.8 3.2 2.5 16 – 200 [26]
Ni58.3Mn17.1Ga24.6 16 0.4 10 13.5 2.8 56 [27]
Ni44.6Co5.5Mn35.5In14.4 16 4.4 6 15.6 6 399 [28]
Ni42.3Co7.9Mn38.8Sn11.0 28 4.7 6 23 10 786 [29]
(Ni50Mn31.5Ti18.5)99.8B0.2 85 3.3 4 74 12 1100 This work
degrade when cycled through the transition (owing to the
large volume change at the magnetostructural transition) and
bulk samples pulverize, our Ni50Mn31.5Ti18.5 has an excellent
stability in its thermodynamic and structural properties [23].
In addition to magnetic alloys, giant barocaloric effects
have also been reported in a broad variety of materials includ-
ing ferroelectric [35–37] and ferrielectric [38] materials, fluo-
rides [39,40], hybrid perovskites [41], and superionic conduc-
tors [42]. Among all these materials only (NH4)2NbOF5 [40]
exhibits an isothermal entropy change (S = 100 J kg−1 K−1)
larger than the value found for our Ni50Mn31.5Ti18.5 shape
memory Heusler alloy. However, the density of our material
(ρ = 7040 kg m−3) is almost five times larger than that of
(NH4)2NbOF5 (ρ = 1450 kg m−3) [9], which implies a much
larger isothermal entropy change in our compound when
normalized to volume. Furthermore, metallic materials have
a larger thermal conductivity which is desirable for practical
and efficient solid-state refrigeration.
IV. SUMMARY AND CONCLUSIONS
We have studied the barocaloric properties of
Ni50Mn31.5Ti18.5, a shape memory Heusler alloy composed
by all-d-metal elements. The composition has been tailored
so that the sample undergoes a martensitic transition but it
does not ferromagnetically order in any of the two structural
phases. The entropy change at the martensitic transition is
very large due to the weak magnetic contribution resulting
from the absence of ferromagnetic order. On the other
hand, the relatively large volume change at the martensitic
transition gives rise to a marked pressure dependence of the
transition temperatures. It has been shown that, due to the
combination of this pressure dependence and a large transition
entropy change, the alloy exhibits outstanding barocaloric
performances. The values found for pressure-induced
isothermal entropy and adiabatic temperature changes
outperform those reported for other shape memory Heusler
alloys and are among the largest values reported for the best
state-of-the-art barocaloric materials.
Our Ni50Mn31.5Ti18.5 alloy is composed of accessible mate-
rials. It exhibits an excellent behavior under cycling which to-
gether with the outstanding barocaloric performances reported
here places this alloy at the forefront of caloric materials to be
used in clean solid-state refrigeration technologies.
ACKNOWLEDGMENTS
We acknowledge financial support from CICyT (Spain)
Projects No. MAT2016-75823-R and No. FIS2017-82625-P,
and from AGAUR (Catalonia), Project No. 2017SGR-0042.
D.Y.C. acknowledges support from the National Natural Sci-
ence Foundation of China (Projects No. 51822102 and No.
51731005) A.G. acknowledges financial support from Univer-
sitat de Barcelona under the APIF scholarship. Experimental
support from E. Xuriguera is acknowledged.
[1] K. Otsuka and C. M. Wayman, Shape Memory Materials
(Cambridge University Press, Cambridge, UK, 1998).
[2] F. X. Hu, B. G. Shen, and J. R. Sun, Appl. Phys. Lett. 76, 3460
(2000).
[3] T. Krenke, E. Duman, M. Acet, E. F. Wassermann, X. Moya, L.
Mañosa, and A. Planes, Nat. Mater. 4, 450 (2005).
[4] E. Bonnot, R. Romero, L. Mañosa, E. Vives, and A. Planes,
Phys. Rev. Lett. 100, 125901 (2008).
[5] L. Mañosa, D. González-Alonso, A. Planes, E. Bonnot, M.
Barrio, J. L. Tamarit, S. Aksoy, and M. Acet, Nat. Mater. 9,
478 (2010).
[6] L. Mañosa, A. Planes, and M. Acet, J. Mater. Chem. A 1, 4925
(2013).
[7] X. Moya, S. Kar-Narayan, and N. D. Mathur, Nat. Mater. 13,
439 (2014).
[8] A. Planes and L. Mañosa, Solid State Phys. 55, 159
(2001).
[9] L. Mañosa and A. Planes, Adv. Mater. 29, 1603607 (2017).
[10] A. Planes, L. Mañosa, and M. Acet, J. Phys.: Condens. Matter
21, 233201 (2009).
[11] Z. Y. Wei, E. K. Liu, J. H. Chen, Y. Li, G. D. Liu, H. Z. Luo,
X. K. Xi, H. W. Zhang, W. H. Wang, and G. H. Wu, Appl. Phys.
Lett. 107, 022406 (2015).
[12] Z. Y. Wei, E. K. Liu, Y. Li, X. L. Han, Z. W. Du, H. Z. Luo,
G. D. Liu, X. K. Xi, H. W. Zhang, W. H. Wang, and G. H. Wu,
Appl. Phys. Lett. 109, 071904 (2016).
044406-6
GIANT BAROCALORIC EFFECT IN ALL-d-METAL … PHYSICAL REVIEW MATERIALS 3, 044406 (2019)
[13] T. Kihara, X. Xu, W. Ito, R. Kainuma, and M. Tokunaga, Phys.
Rev. B 90, 214409 (2014).
[14] V. Recarte, J. I. Pérez-Landazabal, V. Sánchez-Alarcos, V.
Zablotskii, E. Cesari, and S. Kustov, Acta Mater. 60, 3168
(2012).
[15] T. Gottschall, K. P. Skokov, D. Benke, M. E. Gruner, and O.
Gutfleisch, Phys. Rev. B 93, 184431 (2016).
[16] Z. Yang, D. Y. Cong, X. M. Sun, Z. H. Nie, and Y. D. Wang,
Acta Mater. 127, 33 (2017).
[17] F. Fauth, I. Peral, C. Popescu, and M. Knapp, Powder Diffr. 28,
S360 (2013).
[18] P. I. Dorogokupets and A. Dewaele, High Pressure Res. 27, 431
(2007).
[19] A. P. Hammersley, S. O. Svensson, and A. N. Fitch, High
Pressure Res. 14, 235 (1996).
[20] A detailed description of calibration protocols for DTA un-
der pressure can be found in E. Stern-Taulats, Ph.D. thesis,
Universitat de Barcelona, 2017.
[21] S. Aksoy, M. Acet, P. P. Deen, L. Mañosa, and A. Planes, Phys.
Rev. B 79, 212401 (2009).
[22] https://www.iucr.org/resources/other-directories/software/
wincell.
[23] W. Xiong (unpublished).
[24] E. Stern-Taulats, A. Planes, P. Lloveras, M. Barrio, J. L.
Tamarit, S. Pramanick, S. Majumdar, S. Yüce, B. Emre, C.
Frontera, and L. Mañosa, Acta Mater. 96, 324 (2015).
[25] E. Stern-Taulats, A. Planes, P. Lloveras, M. Barrio, J. L.
Tamarit, S. Pramanick, S. Majumdar, C. Frontera, and L.
Mañosa, Phys. Rev. B 89, 214105 (2014).
[26] L. Mañosa, E. Stern-Taulats, A. Planes, P. Lloveras, M. Barrio,
J. L. Tamarit, B. Emre, S. Yüce, S. Fabbrici, and F. Albertini,
Phys. Status Solidi B 251, 2114 (2014).
[27] X. J. He, K. Xu, S. X. Wei, Y. L. Zhang, Z. Li, and C. Jing,
J. Mater. Sci. 52, 2915 (2017).
[28] X. J. He, S. X. Wei, Y. Kang, Y. Zhang, Y. Cao, K. Xu, Z. Li,
and C. Jing, Scr. Mater. 145, 58 (2018).
[29] X. J. He, Y. R. Kang, S. X. Wei, Y. L. Zhang, Y. M. Cao, K.
Xu, Z. Li, C. Jing, and Z. B. Li, J. Alloys Compd. 741, 821
(2018).
[30] L. Mañosa, D. González-Alonso, A. Planes, M. Barrio, J. L.
Tamarit, I. S. Titov, M. Acet, A. Bhattacharyya, and S.
Majumdar, Nat. Commun. 2, 595 (2011).
[31] D. Matsunami, A. Fujita, K. Takenaka, and M. Kano, Nat.
Mater. 14, 73 (2015).
[32] S. Yüce, M. Barrio, B. Emre, E. Stern-Taulats, A. Planes, J. L.
Tamarit, Y. Mudryk, K. A. Gschneidner, V. K. Pecharsky, and
L. Mañosa, Appl. Phys. Lett. 101, 071906 (2012).
[33] R. R. Wu, L. F. Bao, F. X. Hu, H. H. Wu, Q. Z. Huand, J. Wang,
X. L. Dong, G. N. Li, J. R. Sun, F. R. Shen, T. Y. Zhao, X. Q.
Zheng, L. C. Wang, Y. Liu, W. L. Zuo, Y. Y. Zhao, M. Zhang,
X. C. Wang. C. Q. Jin, G. H. Rao, X. F. Han, and B. G. Shen,
Sci. Rep. 5, 18027 (2015).
[34] T. Samanta, P. Lloveras, A. U. Saleheen, D. L. Lepkowski, E.
Kramer, I. Dubenko, P. W. Adams, D. P. Young, M. Barrio, J. Ll.
Tamarit, N. Ali, and S. Stadler, Appl. Phys. Lett. 112, 021907
(2018).
[35] E. Stern-Taulats, P. Lloveras, M. Barrio, E. Defay, M. Egilmez,
A. Planes, J. L. Tamarit, L. Mañosa, N. D. Mathur, and X.
Moya, APL Mater. 4, 091102 (2016).
[36] E. A. Mikhaleva, I. N. Flerov, M. V. Gorev, M. S. Molokeev,
A. V. Cherepakhin, A. V. Kartashev, N. V. Mikhashenov, and
K. A. Sablina, Phys. Solid State 54, 1832 (2012).
[37] Y. Liu, J. Wei, P. E. Janolin, I. C. Infante, X. J. Lou, and B.
Dkhil, Appl. Phys. Lett. 104, 162904 (2014).
[38] P. Lloveras, E. Stern-Taulats, M. Barrio, J. L. Tamarit, S.
Crossley, W. Li, V. Pomjakushin, A. Planes, L. Mañosa, N. D.
Mathur, and X. Moya, Nat. Commun. 6, 8801 (2015).
[39] I. N. Flerov, A. V. Kartashev, M. V. Gorev, E. V. Bogdanov,
S. V. Mel’nikolva, M. S. Molokeev, E. I. Pogoreltsev, and N. M.
Laptash, J. Fluorine Chem. 183, 1 (2016).
[40] M. V. Gorev, E. V. Bogdanov, and I. N. Flerov, Scr. Mater. 139,
53 (2017).
[41] J. M. Bermúdez, M. Sánchez-Andújar, S. Castro-Garcia, J.
López-Beceiro, R. Artiaga, and M. A. Senaris-Rodríguez, Nat.
Commun. 8, 15715 (2017).
[42] A. Aznar, P. Lloveras, M. Romanini, M. Barrio, J. L. Tamarit,
C. Cazorla, D. Errandonea, N. D. Mathur, A. Planes, X. Moya,
and L. Mañosa, Nat. Commun. 8, 1851 (2017).
044406-7