Home Search Collections Journals About Contact us My IOPscience

Large magnetic entropy change near room temperature in antiperovskite SnCMn3

This content has been downloaded from IOPscience. Please scroll down to see the full text.

2009 EPL 85 47004

(http://iopscience.iop.org/0295-5075/85/4/47004)

View the table of contents for this issue, or go to the journal homepage for more

Download details:

IP Address: 155.246.15.35
This content was downloaded on 21/02/2015 at 06:45

Please note that terms and conditions apply.

 February 2009
EPL, 85 (2009) 47004 www.epljournal.org
doi: 10.1209/0295-5075/85/47004

Large magnetic entropy change near room temperature

in antiperovskite SnCMn3
B. S. Wang1 , P. Tong1,2(a) , Y. P. Sun1(b) , X. Luo1 , X. B. Zhu1 , G. Li1 , X. D. Zhu1 , S. B. Zhang1 ,
Z. R. Yang1 , W. H. Song1 and J. M. Dai1
1
Key Laboratory of Materials Physics, Institute of Solid State Physics, and High Magnetic Field Laboratory,
Chinese Academy of Sciences - Hefei 230031, PRC
2
Department of Physics, University of Virginia - Charlottesville, VA 22904, USA

received 14 November 2008; accepted in final form 28 January 2009

published online 3 March 2009

PACS 75.30.Sg – Magnetocaloric effect, magnetic cooling

PACS 75.50.Gg – Ferrimagnetics
PACS 75.30.Kz – Magnetic phase boundaries (including magnetic transitions, metamagnetism,
etc.)

Abstract – We report the observation of a large magnetocaloric effect near room temperature
in the antiperovskite SnCMn3 . The maximal magnetic entropy change at the first-order
ferrimagnetic-paramagnetic transition temperature (TC ∼ 279 K) is about 80.69 mJ/cm3 K and
133 mJ/cm3 K under a magnetic field of 20 kOe and of 48 kOe, respectively. These values are
close to those of typical magnetocaloric materials. The large magnetocaloric effect is associated
with the sharp change of lattice, resistivity and magnetization in the vicinity of TC . Through
the measurements of the Seebeck coefficient and normal Hall effect, the title system is found to
undergo a reconstruction of the electronic structure at TC . Considering its low-cost and innocuous
raw materials, Mn-based antiperovskite compounds are suggested to be appropriate for pursuing
new materials with larger magnetocaloric effect.

c EPLA, 2009
Copyright

Materials with large magnetocaloric effect (MCE) have entropy change −∆SM (15 J/kg K under 20 kOe) with a
attracted much attention in recent years due to their plateau-like temperature dependence [7,8]. In the carbon-
potential application in magnetic refrigeration, which deficient sample GaC0.78 Mn3 , the AFM state observed
provides a promising alternative to the conventional in the stoichiometric sample is collapsed. Instead, the
vapor-cycle one [1–3]. The research has been mainly ground state is FM with a transition temperature of
focused on the materials with expensive rare-earth 295 K, at which the value of −∆SM is 3.7 J/kg K under
elements, sometimes with poisonous elements (e.g., As). 50 kOe [11]. For GaCMn3−x Cox , the substitution of Co for
Therefore, practically, it is of great significance to explore Mn sites decreases the AFM-FM transition temperature
new systems with large MCE, but with innocuous and without significant loss of −∆SM , while it broadens the
low-cost raw materials. plateau-like temperature dependence [12].
The antiperovskite intermetallic compounds AXM3 In this paper, we report a large room temperature
(A: main group elements; X: carbon, boron or nitrogen; MCE in another antiperovskite compound, SnCMn3 ,
M: transition metal) have displayed lots of interest- corresponding to a ferrimagnetic (FIM)-paramagnetic
ing properties, such as superconductivity [4,5], giant (PM) transition (TC ∼ 279 K). The large MCE is related
magnetoresistance [6], large negative MCE [7,8], giant with a sharp change of magnetization at TC , which a
negative thermal expansion (NTE) [9,10], etc. Up to now, reconstruction of the electronic structure can account for.
GaCMn3 is the only antiperovskite compound that has Along with the previous reports on Mn-based antiper-
been reported to exhibit MCE. In GaCMn3 there exists an ovskite materials, we suggest that this type of compounds
abrupt first-order antiferromagnetic (AFM)-ferromagnetic could be an alternative for searching new MCE materials
(FM) transition at ∼ 165 K, which produces a large at various temperatures.
A polycrystalline sample of SnCMn3 was prepared
(a) E-mail: tongpeng@issp.ac.cn from powders of Sn (4N), Graphite (3N) and Mn (4N).
(b) E-mail: ypsun@issp.ac.cn The starting materials were mixed in the desired mole

47004-p1
B. S. Wang et al.

Fig. 1: X-ray diffraction pattern (solid curve) and Rietveld

Fig. 2: Temperature dependence of the magnetization, M -T ,
refinement result (crosses) for SnCMn3 at room temperature.
under ZFC and FC modes at H = 100 Oe and inverse suscepti-
The vertical lines show the Bragg peak positions as denoted
bility as a function of temperature. The FC M -T was measured
by the index (hkl). The difference between the data and the
on both warming (FCW) and cooling (FCC).
calculation is shown at the bottom. The goodness parameters
of refinement are indicated.

ratio, sealed in evacuated quartz tubes and then heated

at 1073 K for two days, slowly increased to 1150 K at
1.5 K/min, and then annealed at 1150 K for five days. After
the tube was quenched to room temperature, the prod-
ucts were pulverized, grounded, and pressed into pellets
which were annealed again under the same condition to
obtain homogeneous samples. X-ray powder diffraction
(XRD) pattern was collected using a Philips X
pert
PRO X-ray diffractometer with Cu Kα radiation at room
temperature. The XRD data was analyzed by using the
standard Rietveld technique. The magnetic measurements
were performed on a Quantum Design superconducting
quantum interference device (SQUID) magnetometer
(1.8 K
T
400 K, 0
H
50 kOe). The Seebeck coef- Fig. 3: Resistivity measured at zero field as a function of
ficient and Hall effect were measured in a commercial temperature. The inset shows the resistivity at zero field in
cooling and warming cycles in the vicinity of TC .
Quantum Design Physical Property Measurement System
(PPMS) (1.8 K
T
800 K, 0
H
90 kOe).
Figure 1 shows the typical refinement of the XRD first-order phase transition is associated with a structural
pattern for the SnCMn3 sample. All diffraction peaks can transformation. As for SnCMn3 , with increasing temper-
be indexed by the antiperovskite structure (space group, ature the unit cell volume undergoes a discontinuous
P m3m) and no detectable secondary phases can be found. contraction of ∼ 0.1% at TC , without changing the crystal
The refined lattice parameter a (3.981 Å) is slightly less structure [14]. As shown on the right hand of fig. 2, the
than the calculated one (3.989 Å) [13]. inverse magnetic susceptibility 1/χ(T ) above TC roughly
The measured magnetization as a function of temper- agrees with the Curie-Weiss law. However, the 1/χ(T ) at
ature, M -T , at a field of 100 Oe is shown in fig. 2. A clear elevated temperatures does not obey this law as reported
magnetic transition can be found at TC ∼ 279 K, which is by other authors with unclear reasons [16].
determined as the reflection point of the derivative of the Figure 3 displays the temperature-dependent resistivity
M -T curve. This value of TC is roughly in accordance with ρ(T ) measured at zero field. At low temperatures, ρ(T )
the previously reported one [14]. The neutron diffraction shows a metallic behavior. With increasing temperature,
measurement has revealed that the ground state of ρ(T ) presents a sudden reduction at TC . As the temper-
SnCMn3 is a noncollinear FIM one, which consists of ature increases further, ρ(T ) turns back to the metallic
both AFM and FM sublattices [15,16]. Thus the magnetic behavior with a smaller dρ(T )/dT ratio than at low
transition we observed at TC should be a FIM-PM one. temperatures. As shown in the inset of fig. 3, an obvious
As shown in fig. 2, a weak thermal hysteresis (∼ 2.5 K) thermal hysteresis between warming and cooling modes
between warming and cooling FC curves can be found can be seen, suggesting a first-order transition near TC ,
near TC , indicating a first-order transition. Generally, the which is consistent with the result of fig. 2.

47004-p2
 Large magnetic entropy change near room temperature in antiperovskite SnCMn3

Fig. 5: Magnetic entropy change of SnCMn3 for the field

Fig. 4: Magnetization isotherms of SnCMn3 at various temper- changing from 0–5 kOe to 0–48 kOe.
atures from 272 K to 288 K in the presence of external magnetic
fields up to 48 kOe.

Figure 4 presents the magnetization isotherms, M -H,

of SnCMn3 at selected temperatures near TC . These
M -H curves were measured after the sample was cooled
down from 300 K to each measurement temperature. A
metamagnetic transition induced by external fields can
be observed clearly in the M -H curves around TC . At
several selected temperatures near TC , the isotherms
in the increasing field/decreasing field cycles are also
measured and there are small hystereses in these cycles.
Such a small hystereses of M (H) is significant for the
practical application of magnetic refrigeration [17,18].
Based on the M -H curves, the magnetic entropy change
Fig. 6: The maximal −∆SM values of SnCMn3 and of several
can be evaluated using the Maxwell equation [2],
prototype magnetocaloric materials (derived from ref. [3]) as
a function of the external magnetic field. The temperatures to
∆SM (T, H) = SM (T, H) − SM (T, 0) =

H 
H  which the maximal −∆SM correspond are marked in brackets
∂S ∂M after the compounds.
dH = dH. (1)
0 ∂M T 0 ∂T H

In the case of magnetization measurements at small La0.7 Ca0.3 MnO3 and LaFe11.44 Si1.56 , has been displayed
discrete field and temperature intervals, ∆SM can be in fig. 6. It is evident that −∆SM of SnCMn3 is compa-
approximated as [2] rable to those of the candidates for magnetic refrigerant
      materials plotted in this figure. Compared with GaCMn3 ,
  (Mi − Mi+1 )Hi the maximal −∆SM value of SnCMn3 has a similar value,
∆SM Ti + Ti+1  = ∆Hi , (2)
 2  Ti+1 − Ti but it occurs near room temperature. Practically, the rela-
tive cooling power (RCP), which is a measure of how much
where Mi and Mi+1 are the experimental data of the heat can be transferred between the cold and hot sinks in
magnetization at Ti and Ti+1 , respectively, under the same an ideal refrigerant cycle, is also an important parameter
magnetic field. for selecting potential substances as magnetic refrigerants.
Figure 5 shows the thermal variation of the magnetic Generally, the RCP is defined as the product of the maxi-
entropy change −∆SM under different ranges of magnetic mum magnetic entropy change −∆S max and the full width
M
field up to ∆H = 48 kOe. For each ∆H, −∆SM reaches at half maximum δTFWHM [3,19]:
the maximal value at TC , i.e., 279 K. The value of −∆SM ,
which increases with increasing ∆H, is 80.69 mJ/cm3 K RCP = −∆SM max
δTFWHM . (3)
3
and 133 mJ/cm K for the characteristic ∆H = 20 kOe
and 48 kOe, respectively. A comparison between the peak According to eq. (3), the estimated RCP of SnCMn3
value of −∆SM in SnCMn3 and those in prototype MCE (∼ 212.8 mJ/cm3 , ∆H = 48 kOe) is less than those of
materials [3], such as Gd, Gd5 Si2 Ge2 , MnAs, DyCo2 , Gd (5300 mJ/cm3 , ∆H = 50 kOe) [3] and Gd5 Si2 Ge2

47004-p3
B. S. Wang et al.

Fig. 8: Normal Hall coefficient as a function of tempera-

ture, RH (T ). Inset: temperature-dependent Hall carrier density
Fig. 7: Temperature dependence of the Seebeck coefficient nH (T ) near TC . Solid lines are guidance for eyes.
α(T ).

for SnCMn3 (1.03 × 1022 cm−3 ) is a little less than that

(3360 mJ/cm3 , ∆H = 50 kOe) [3]. Despite this, SnCMn3 for a common metal like Cu (∼ 8.5 × 1022 cm−3 ), in line
could be a candidate for magnetic refrigerant materials with the metal-like character of electrical conductivity.
in the relative temperature region due to its inexpensive Consequently, the abrupt changes of α(T ) and RH (T ) in
and innoxious raw materials, as well as its considerable SnCMn3 may indicate evidently a sudden reconstruction
small thermal hysteresis. of the electronic structure, e.g., the DOS near EF . Such a
Intuitively, as indicated by eqs. (1) and (2), the large change of DOS could not arise from the volume collapse
−∆SM in SnCMn3 can be attributed to the sharp change at the phase transition which usually leads to a reduction
of magnetization in the vicinity of TC (fig. 2). For a of the DOS (namely of the carrier density) [22].
better understanding of the origin of the large MCE, the As discussed above, the large MCE in SnCMn3 ,
normal Hall effect and Seebeck coefficient of SnCMn3 accompanied by abrupt changes of magnetization, resis-
have been measured as a function of temperature. α(T ) tivity and lattice, originates from the reconstruction
(fig. 7) is inversely proportional to temperature at low of the electronic structure at the FIM-PM transition
temperatures, indicating that the main contribution to temperature. Such kind of behavior seems to be universal
α(T ) comes from the thermal diffusion of carriers. Thus, in Mn-based antiperovskite compounds, since sharp
the value of α(T ) can reflect the type of carriers [20], magnetic/structural phase transitions have been proved
namely, the negative α value means electron-type carriers to be characteristic of these compounds [15]. In addition,
in SnCMn3 . Another remarkable feature of α(T ) is the the types of phase transition and the corresponding
remarkable change occurring at TC . In a Boltzmann temperatures can be tuned by varying the chemical
picture, the Seebeck coefficient α can be described as [21], composition [9–12,15]. For example, the properties of
 
π 2 κ2B ∂ ln N (E) GaCMn3 can be reproduced by alloying 50% SnCMn3
α=− T , (4) with 50% ZnCMn3 [15]. Therefore, it is possible to get a
3e ∂E EF
higher value of RCP by increasing the MCE or broadening
where N (EF ) is the electronic density of state (DOS) at its temperature span via chemical doping. In this regard,
the Fermi energy EF and kB the Boltzmann constant. this type of materials can be considered as an alternative
Equation (4) indicates that any change in the Seebeck for pursuing new materials with large MCE at various
coefficient is a direct consequence of the modification temperatures. On the other hand, the mechanism of
in the DOS near EF . Figure 8 displays the tempera- magnetic orders in Mn-based antiperovskite compounds
ture dependence of the normal Hall coefficient, RH (T ), is far away from clear. Initial work suggests that the
measured at a magnetic field of 10 kOe. RH is negative in Ruderman-Kittel-Kasuya-Yosida (RKKY) interaction,
the entire temperature range investigated, indicating the which connects the itinerant electrons and localized
dominant carriers are of electron type. As temperature magnetic moments, may work in these materials [15].
increases, a sudden change of RH (T ) occurs at TC . The Alternatively, a simple mechanism of direct exchange
Hall carrier density nH can be reasonably estimated with interaction between Mn sites has been employed to
the formula nH = 1/|eRH | [20], where e is the elementary explain the experimental results [11,14]. Theoretically, a
electric charge. As shown in the inset of fig. 8, around well-dispersed d-orbital has been found [13], suggesting
TC , nH in the high-temperature PM phase increases three an itinerant mechanism for magnetic orders, consistent
times in comparison with that in the low-temperature FIM with the result of Mössbauer spectroscopy [23]. In this
phase. The calculated value of nH at room temperature scenario, however, it is a challenge to understand the

47004-p4
 Large magnetic entropy change near room temperature in antiperovskite SnCMn3

first-order magnetic transition with an abrupt change [6] Kamishima K., Goto T., Nakagawa H., Miura N.,
of lattice [23,24]. In this context, more researches on Ohashi M., Mori N., Sasaki T. and Kanomata T.,
Mn-based antiperovskite compounds will give a deeper Phys. Rev. B, 63 (2000) 024426.
insight into the nature of magnetic orders and competition [7] Tohei T., Wada H. and Kanomata T., J. Appl. Phys.,
between them, and vice versa. 94 (2003) 1800.
[8] Yu M. H., Lewis L. H. and Moodenbaugh A. R.,
In summary, we report the large magnetic entropy
J. Appl. Phys., 93 (2003) 10128.
near room temperature in the antiperovskite metallic
[9] Takenaka K. and Takagi H., Appl. Phys. Lett., 87
compound SnCMn3 . The large value of −∆SM is (2005) 261902.
comparable to those observed in contemporary magnetic [10] Takenaka K., Asano K., Misawa M. and Takagi H.,
refrigerant materials. The large magnetic entropy change Appl. Phys. Lett., 92 (2008) 011927.
is suggested by thermal transport measurements to be [11] Lewis L. H., Yoder D., Moodenbaugh A. R., Fischer
associated with the reconstruction of the Fermi surface D. A. and Yu M. H., J. Phys.: Condens. Matter, 18
in the vicinity of the FIM-PM transition. Along with the (2006) 1677.
advantages of the raw materials, the discovery of MCE in [12] Tohei T., Wada H. and Kanomata T., J. Magn. &
SnCMn3 may provide a new field for seeking good candi- Magn. Mater, 272-276 (2004) e585.
dates for magnetic refrigeration at various temperatures. [13] Motizuki K. and Nagai H., J. Phys. C: Solid State
Phys., 21 (1988) 5251.
∗∗∗ [14] Li Y. B., Li W. F., Feng W. J., Zhang Y. Q. and
Zhang Z. D., Phys. Rev. B, 72 (2005) 024411.
This work was supported by the National Key Basic [15] Fruchart D. and Bertaut E. F., J. Phys. Soc. Jpn.,
Research under contract No. 2007CB925002, and the 44 (1978) 781.
National Nature Science Foundation of China under [16] Kaneko T., Kanomata T. and Shirakawa K., J. Phys.
contract No. 50701042, No. 10774146 and Director’s Fund Soc. Jpn., 56 (1987) 4047.
of Hefei Institutes of Physical Science, Chinese Academy [17] Provenzano V., Shapiro A. J. and Shull R. D.,
Nature, 429 (2004) 853.
of Sciences.
[18] Mohapatra N., Iyer K. K. and Sampathkumaran
E. V., Eur. Phys. J. B, 63 (2008) 451.
REFERENCES [19] Gschneidner K. A. jr., Pecharsky V. K. and
Pecharsky A. O. et al., Mater. Sci. Forum, 315 (1999)
[1] Tishin A. M., J. Magn. & Magn. Mater, 316 (2007) 351. 69.
[2] Phan M. H. and Yu S. C., J. Magn. & Magn. Mater., [20] Siebold Th. and Ziemann P., Phys. Rev. B, 51 (1995)
308 (2007) 325. 6328.
[3] Gschneidner K. A. jr., Pecharsky V. K. and Tsokol [21] Behnia K., Jaccard D. and Flouquet J., J. Phys.:
A. O., Rep. Prog. Phys., 68 (2005) 1479. Condens. Matter, 16 (2004) 5187.
[4] He T., Huang Q., Ramirez A. P., Wang Y., Regan [22] Park M. S., Giim J. S., Park S. H., Lee Y. W., Lee
K. A., Rogado N., Hayward M. A., Haas M. K., S. I. and Choi E. J., Supercond. Sci. Technol., 17 (2004)
Slusky J. S., Inumara K., Zandbergen H. W., Ong 274.
N. P. and Cava R. J., Nature (London), 411 (2001) 54. [23] Grandjean F. and Gérard A., J. Phys. F: Met. Phys.,
[5] Uehara M., Yamazaki T., Kôri T., Kashida T., 6 (1976) 451.
Kimishima Y. and Hase I., J. Phys. Soc. Jpn., 76 (2007) [24] Kim I. G., Jin Y. J., Lee J. I. and Freeman A. J.,
034714. Phys. Rev. B, 67 (2003) 060407(R).

47004-p5