J Mater Sci (2018) 53:2405–2412

Ceramics
CERAMICS

Magnetocaloric effect in (La0.7Sr0.3MnO3)12x–(BaTiO3)x

solid solution spin-glass system
C. Nayek1,2 , M. K. Ray2 , A. Pal1 , I. M. Obaidat2 , and P. Murugavel1,*

1
Department of Physics, Indian Institute of Technology Madras, Chennai 600036, India
2
Department of Physics, United Arab Emirates University, Al-Ain 15551, United Arab Emirates

Received: 15 August 2017 ABSTRACT

Accepted: 13 October 2017 (La0.7Sr0.3MnO3)1-x–(BaTiO3)x solid solution samples (x = 0, 0.03, and 0.08)
Published online: were synthesized by sol–gel route. Pure La0.7Sr0.3MnO3 has orthorhombic
23 October 2017 structure with Pbnm space group whereas solid solution samples possess
rhombohedral structure with R-3c space group. Temperature-dependent ac
Ó Springer Science+Business susceptibility at different frequencies and memory effect confirmed the spin-
Media, LLC 2017 glass behavior in these samples. The samples exhibited magnetocaloric coeffi-
cient around room temperature. The maximum values of magnetic entropy
change observed in our samples are 0.45, 0.86, and 1.07 J Kg-1 K-1 for x = 0.03
and 0.62, 1.23 and 1.56 J Kg-1 K-1 for x = 0.08 compounds at 2, 4, and 5 T
applied magnetic fields, respectively. The relative cooling power values for
x = 0.03 and 0.08 are 55.9 and 116.5 J Kg-1, respectively, at 5 T. The observed
multifunctional properties of the studied system give a new direction to explore
the magnetocaloric effect in similar such systems.

Introduction magnetic cooling, but their ferromagnetic Curie

temperature (TC) is far below room temperature
The magnetocaloric effect (MCE), where the entropy (* 270 K) and they show high MCE (change in
changes occur in a magnetic material upon applica- magnetic entropy |DSM| * 19 J Kg-1 K-1 at TC in
tion of an external magnetic field, offers an alterna- the magnetic field range of 0–5 T). Large MCEs
tive eco-friendly magnetic refrigeration technique (|DSM| * 40 J Kg-1 K-1 at TC * 318 K and 32 J
with improved efficiency [1–4]. Though it has vast Kg-1 K-1 at TC * 310 K for x = 0 and 0.05 samples
application [5, 6], the materials showing the MCE at under the change of magnetic field from 0 to 5 T,
room temperature are still the area to be improved. respectively) are also observed in MnAs1-xSbx com-
Despite the reported large value of MCE in many pounds [10], but they are not environmental friendly.
compounds, few drawbacks restricted their use for Strong room temperature MCE is reported in
practical purposes. As an example, Gd5Ge2Si2 and Fe0.49Rh0.51 alloy, but high cost and attenuating
related Gd-based compounds [7–9] are used as a behavior of the MCE (|DSM| * 12.58 J Kg-1 K-1 at
prototypical refrigerant for room temperature TC * 315.6 K at zero magnetic field) due to an

Address correspondence to E-mail: muruga@iitm.ac.in

DOI 10.1007/s10853-017-1718-x
2406 J Mater Sci (2018) 53:2405–2412

irreversibility of the antiferromagnetic–ferromagnetic Materials and methods

(FM) transition make them not suitable for real
applications [11]. Therefore, the search for new The (La0.7Sr0.3MnO3)1-x–(BaTiO3)x (x = 0, 0.03, and
magnetic materials showing magnetocaloric effect is 0.08) compounds presented in this work are synthe-
imperative so as to find a suitable magnetic refrig- sized by sol–gel route. To prepare La0.7Sr0.3MnO3
erant which is cheaper, environmental friendly, and nanoparticles, 0.05 M precursor solutions are made
having larger MCE. by dissolving stoichiometric amount of lanthanum
In this context, the perovskite manganites having nitrate, manganese nitrate, and strontium nitrate in
complex magnetic phase diagram exhibiting inter- concentrated nitric acid (69%) which is used as sol-
esting functionalities such as magnetic, magnetore- vent. The precursor solutions are mixed at 90 °C and
sistance, and even magnetocaloric effect due to their stirred by a magnetic stirrer. 10 ml aqueous citric acid
intriguing magnetic orderings are gaining impor- solution (2 M) is added to the mixture solution fol-
tance [12, 13]. Among these perovskites, La1-xCax lowed by adding 12 ml of ethylene glycol after 3 h for
MnO3 series are reported to show large MCE (for the polymerization reaction. The total solution dried
x = 0, |DSM| * 5.5 J Kg-1 K-1 at TC = 230 K under at 90 °C under continuous stirring till the brown
the change in magnetic field from 0 to 1.5 T and for color resin is formed. The dried resin is pre-annealed
samples x = 0.33 and 0.45, |DSM| * 4.3 and 2 J at 450 °C for 2 h and finally annealed at 700 °C for
Kg-1 K-1 at TC = 257 and 234 K, respectively, under 6 h to get desired La0.7Sr0.3MnO3 nanoparticles. For
the change in magnetic field from 0 to 1.5 T), but preparing solid solution, appropriate amount of
their TC is quite low [14, 15]. Another perovskite aqueous barium carbonate solution (0.005 M) is
La1-xPbxMnO3 has high TC but low MCE (|DSM| of added with aqueous citric acid solution (0.015 M)
La0.8Pb0.2MnO3 is 1.22 J Kg-1 K-1 at TC = 294 K and 0.005 molar concentration of titanium tetra-iso-
for an applied field of 1.5 T) [16, 17]. However, propoxide in ethanol is added with 0.15 molar citric
La1-xSrxMnO3 manganite shows the possibility of acid solution. These two solutions are mixed and
tuning its TC well above room temperature with maintained at room temperature with continuous
change in composition [18–20]. Since, the maximum stirring condition for 30 min. Later, appropriate
MCE is envisaged near the transition temperatures, amount of La0.7Sr0.3MnO3 nano-powder is added to
La1-xSrxMnO3 is a potential material where large the gel, sonicated for 1 h, and dried at 60 °C under
MCE can be observed at room temperature [18, 20]. continuous stirring in a magnetic stirrer. The dried
In addition, other dopants are tried in La0.7Sr0.3MnO3 powder is then annealed at 780 °C for 5 h to get
(LSMO) to investigate the change in magnetic (La0.7Sr0.3MnO3)1-x–(BaTiO3)x nanoparticles. The
ordering and its influence on the MCE effect. The powders are characterized for phase formation by
substitution of Al and Ti at Mn-site in LSMO is PANalytical X’Pert Pro X-ray diffractometer. The
reported to lower the TC from * 360 K to below refinement is done using general structure analysis
300 K [21]. The substitution of Ti4? (3d0) ions at system (GSAS) code. The morphological study is
Mn4?-site changes the Mn4?/Mn3? ratio along with performed by JEOL high-resolution transmission
the structural change due to the increase in average electron microscope (HRTEM). Magnetic measure-
B-site ionic radius [22]. In this context, we are inter- ments are done using physical property measure-
ested in investigating the magnetic and magne- ment system (PPMS), and ac susceptibility (vac)
tocaloric effect of LSMO with doping on both A- and measurements are performed by commercial Cryo-
B-site ions. For this purpose, we have chosen the com- BIND system.
pounds (La0.7Sr0.3MnO3)1-x–(BaTiO3)x [(LSMO)1-x-
(BTO)x] (x = 0.03 and 0.08) and studied their struc-
ture, morphology, magnetic, spin-glass, and magne- Results and discussion
tocaloric properties. Interestingly, the doped
compounds exhibited an exciting spin-glass behavior Structural characterization
along with MCE effect. The detailed investigations are
Figure 1a shows the X-ray diffraction patterns of
presented in this work.
(La0.7Sr0.3MnO3)1-x–(BaTiO3)x nano-materials for
x = 0, 0.03, and 0.08. The patterns show clear
J Mater Sci (2018) 53:2405–2412 2407

Morphological characterization

The morphology of the samples is obtained by

transmission electron microscopy (TEM). Figure 2
shows TEM images for x = 0.03 and 0.08 samples
along with the selected area electron diffraction pat-
tern (SAED). The particles are arbitrary in shape, and
the average particle size as inferred from the TEM
images is around 40 nm for all samples. The ring
patterns in SAED picture as shown in Fig. 2b, d
confirm the polycrystalline nature of the samples.
The d values estimated from the SAED patterns are
well matching with the peaks in XRD pattern.

Magnetic properties

Temperature (T)-dependent magnetization (M) mea-

surement is performed from 4 to 350 K under an
external magnetic field of 100 Oe in zero field cooled
Figure 1 a X-ray diffraction patterns for (La0.7Sr0.3MnO3)1-x–
(ZFC) and field cooled (FC) mode for x = 0.03 and
(BaTiO3)x (x = 0, 0.03, and 0.08) solid solution samples (inset
shows the peak shift). b Rietveld refinement data for x = 0.03
0.08 samples and the results are shown in Fig. 3a, b,
sample. respectively. The figures show paramagnetic (PM) to
FM type magnetic transition around 330 and 294 K
diffraction peaks matching with pure La0.7Sr0.3MnO3 for x = 0.03 and 0.08 samples, respectively. It is
but with small shift. No additional peak representing observed that TC decreases with increase in Ti4?
secondary phase is seen which confirms the phase content. This may be due to the effect of dilution of
purity of all the samples. Shift in the diffraction peaks exchange interaction between Mn3? and Mn4? ions
(shown as an inset in Fig. 1a) is attributed to the via oxygen. Also, the disorderness induced by the
formation of the solid solution. substitution of Ba2? and Ti4? in A- and B-site sub-
The Rietveld refinements are carried out for all the lattices could be the reason of the decrease in TC. The
samples by GSAS software and as a representative ZFC–FC curve exhibits large deviation below TC. The
example Fig. 1b shows the refined data for x = 0.03 large difference in ZFC and FC values in tempera-
sample. The lattice parameters extracted from the ture-dependent magnetization measurements may
refinement data for bare La0.7Sr0.3MnO3 reveal that indicate spin-glass state in the samples. Figure 3c, d
the sample is crystallized in orthorhombic structure show the M versus applied magnetic field (H) curves
with Pbnm space group with lattice parameters for x = 0.03 and 0.08 samples recorded at 20 K. The
a = 3.864 Å, b = 3.852 Å, and c = 3.871 Å. However, figures reveal a typical ferromagnetic M–H curve
the refinement shows that for x = 0.03 and 0.08 with well-saturated moment at 20 K. The effective
compounds, the structure changes into rhombohe- moments (leff) retrieved from saturated magnetiza-
dral phase with R-3c as space group. The lattice tions for x = 0.03 and 0.08 are 2.64 and 2.54 lB/Mn,
parameters are a = b = 5.483 Å, c = 13.353 Å and respectively, at 1 T. These values are small compared
a = b = 5.502 Å, c = 13.458 Å for x = 0.03 and 0.08, to the calculated values using the spin-only moment
respectively. The c/a ratio is 1.002 for bare LSMO and of free Mn ions (3.87 lB for Mn4? and 4.90 lB for
it is increased to 2.435 and 2.446 for x = 0.03 and 0.08 Mn3?). This reduction in effective moment may be
samples, respectively, indicating structural distortion due to the dilution of Mn ions by Ti4? ion and the size
in the solid solution compared to the parent effect [23].
compound.
2408 J Mater Sci (2018) 53:2405–2412

Figure 2 a Transmission electron microscope picture and b SAED pattern for x = 0.03 sample and c transmission electron microscope
picture and d SAED pattern for x = 0.08 sample.

Figure 3 Temperature-
dependent magnetization
recorded under ZFC and FC
condition at an applied
magnetic field of 100 Oe for
a x = 0.03 and b x = 0.08
samples. The M–H curve
recorded at 20 K for
c x = 0.03 and d x = 0.08
samples.
J Mater Sci (2018) 53:2405–2412 2409

Spin-glass properties halting and reference mode susceptibilities, are

measured as a function of T with halting temperature
The bifurcation of M versus T curves done under ZFC Thalt at 25 and 30 K for x = 0.03 and 0.08 samples,
and FC conditions with large difference between respectively. The Dv0 [= (v0 ref - v0 mem)] versus T data
them at low temperature and the appearance of a is plotted for x = 0.03 and 0.08 samples and shown in
cusp in the ZFC signifies the presence of spin-glass the inset of Fig. 4a, b. The inset shows a peak exactly
magnetic phase. In order to confirm this, tempera- at Thalt ascertaining the memory effect in the system.
ture-dependent ac susceptibility v0 (T) is performed A detailed study on the spin-glass state of the system
from 95 Hz to 1 kHz frequency variation in a probing is reported elsewhere [24].
ac field (Hac) of 0.17 Oe. The normalized v0 (T) (in
phase) components are plotted with respect to tem- Magnetocaloric effect
perature and shown in Fig. 4a, b for x = 0.03 and 0.08
samples, respectively. The samples exhibit a fre- To further understand the nature of magnetic order
quency-dependent behavior below the broad peak fluctuations, we measured the field dependence of
temperature (assigned as spin-glass transition tem- isothermal M at various temperatures from 230 to
perature TP) and almost frequency-independent 335 K with an interval of DT = 5 K, for x = 0.03 and
behavior above TP, indicating spin-glass state nature 0.08 samples. Note that at room temperature,
of the samples. To ensure further, memory effect isothermal M versus H curve does not show any sign
measurement is carried out on x = 0.03 and 0.08 of saturation even at 5 T which may be attributed to
samples by recording the data in reference and halt- the presence of competing interactions at this tem-
ing mode. In halting mode, the sample is halted perature range giving rise to spin-glass behavior. The
below TP with a halting time thalt = 12 h whereas in broad transitions in M versus T curves shown in
reference mode, the measurement is done without Fig. 3 hint a second-order magnetic phase transition.
any halting. The v0 mem and v0 ref, corresponding to the To confirm it further, Arrott plots are drawn and
presented in Fig. 5c, d for x = 0.08 and 0.03 samples,
respectively. Arrott plot is defined as a plot between
square of magnetization (M2) of a substance against
the ratio of the applied magnetic field to magnetiza-
tion (H/M). According to Banerjee’s criterion, a
negative or positive sign of the slope of Arrott plots
attributes to a first-order or second-order magnetic
phase transition, respectively [25]. The positive slope
at all the points of M2 in the patterns confirms the
second-order spin-glass phase transition for the solid
solution samples.
To investigate the magnetocaloric effect (MCE),
DSM is extracted from the measured isothermal M (H)
curves, shown in Fig. 5a, b, using the expression
R H  ðH;TÞ
DSM ðT; DH Þ ¼  H01 dMdT dH. Figure 6a, b are
H
the temperature dependence of DSM plotted at 2, 4,
and 5 T applied magnetic field for x = 0.03 and 0.08,
respectively. It is observed that the maximum in
|DSM| is occurred at the FM to PM TC with distri-
bution of |DSM| over a relatively broad temperature
range. This broadening may be due to the glassy
nature of the materials (second-order phase
transition).
Figure 4 Temperature-dependent ac susceptibility for a x = 0.03
and b x = 0.08 samples. The insets show the memory effect The magnitude of |DSM| is strongly suppressed
performed for respective samples. upon lowering the field. The maximum values of
2410 J Mater Sci (2018) 53:2405–2412

Figure 5 External field-

dependent isothermal
magnetization for a x = 0.03,
b x = 0.08 and the Arrott
plots for c x = 0.08,
d x = 0.03 samples.

1.56 J Kg-1 K-1 for x = 0.08 compound at 2, 4, and 5

T applied magnetic fields, respectively. These values
are smaller than that of Gd, but comparable to other
manganite materials [18, 20]. We have calculated the
RCP which is defined as - DSM 9 DTFWHM, where
DTFWHM denotes the full width at half maximum of
- DSM versus T curve (48.3 and 75.5 for x = 0.03 and
0.08 samples, respectively). The RCP values for
x = 0.03 and 0.08 are 55.9 and 116.5 J Kg-1, respec-
tively, at 5 T which is greatly improved compared to
that of bare LSMO value reported in the literature
[18, 20]. Refrigerants with a wide working tempera-
ture span and high RCP are beneficial for the mag-
netic cooling application. The observed MCE in
(LSMO)1-x–(BTO)x over wide temperature range
gives an opportunity for further tuning of these
properties toward device applications.

Conclusion

In this report, nanoparticulate solid solution samples

of (LSMO)1-x–(BTO)x (x = 0.03 and 0.08) were syn-
thesized by sol–gel method and their phase purity
was confirmed by X-ray diffraction and Rietveld
Figure 6 |DSM| versus T plots for a x = 0.03 and b x = 0.08 refinement analysis. The difference in temperature-
samples. dependent field cooled and zero field cooled mag-
netization data, the behavior of temperature-depen-
|DSM| observed in our samples are 0.45, 0.86, 1.07 J dent ac susceptibility at different frequencies, and
Kg-1 K-1 for x = 0.03 compound and 0.62, 1.23, and memory effect confirmed the spin-glass behavior of
J Mater Sci (2018) 53:2405–2412 2411

the doped samples. The spin-glass was observed to [8] Pecharsky VK, Gschneidner KA Jr (1997) Tunable magnetic
be second order in nature with broad magnetic regenerator alloys with a giant magnetocaloric effect for
transition just below the ferromagnetic to paramag- magnetic refrigeration from * 20 to * 290 K. Appl Phys
netic transition. The magnetocaloric measurements Lett 70:3299–3301
revealed that the magnetocaloric coefficient was [9] Pecharsky VK, Gschneidner KA Jr (1997) Giant magne-
1.07 J Kg-1 K-1 for x = 0.03 and 1.56 J Kg-1 K-1 for tocaloric effect in Gd5(Si2Ge2). Phys Rev Lett 78:4494–
x = 0.08 sample. Interestingly, the high value of 4497
DTFWHM yielded larger relative cooling power (55.9 [10] Wada H, Tanabe Y (2001) Giant magnetocaloric effect of
and 116.5 J Kg-1 for x = 0.03 and 0.08, respectively). MnAs1-xSbx. Appl Phys Lett 79:3302–3304
Overall, the magnetocaloric effect observed in [11] Annaorazov MP, Nikitin SA, Tyurin AL, Asatryan KA,
(LSMO)1-x–(BTO)x solid solution samples with large Dovletov AKh (1996) Anomalously high entropy change in
relative cooling power and wide range of working FeRh alloy. J Appl Phys 79:1689–1695
temperature brings out the importance of the system [12] Martin C, Maignan A, Hervieu M, Raveau B (1999) Mag-
toward potential application side. netic phase diagrams of L1-xAxMnO3 manganites (L = Pr,
Sm: A = Ca, Sr). Phys Rev B 60:12191–12199
[13] Phan MH, Yu SC (2007) Review of the magnetocaloric
Acknowledgements effect in manganite materials. J Magn Magn Mater
308:325–340
The authors acknowledge the support given by Prof. [14] Guo ZB, Du YW, Zhu JS, Huang H, Ding WP, Feng D
P. S. Anil Kumar and Kaustuv Manna, Department of (1997) Large magnetic change in perovskite-type manganese
Physics, Indian Institute of Science, Bangalore for the oxides. Phys Rev Lett 78:1142–1145
ac susceptibility measurement. This work has been [15] Guo ZB, Zhang JR, Huang H, Ding WP, Du YW (1997)
partially supported by UAEU Program for Advanced Large magnetic entropy change in La0.75Ca0.25MnO3. Appl
Research (UPAR) under the Grant No. 31S241. Phys Lett 70:904–905
[16] Chau N, Nhat HN, Luong NH, Minh DL, Tho ND, Chau NN
(2003) Structure magnetic magnetocaloric and magnetore-
References sistance properties of La1-xPbxMnO3 perovskite. Physica B
327:270–278
[1] Bruck E (2005) Developments in magnetocaloric refrigera-
[17] Min SG, Kim KS, Yu SC, Suh HS, Lee SW (2005) Mag-
tion. J Phys D Appl Phys 38:R381–R391
netocaloric properties of La1-xPbxMnO3 (x = 0.1, 0.2, 0.3)
[2] Takeya H, Pecharsky VK, Gschneidner KA Jr, Moorman JO
compounds. IEEE Trans Magn 41:2760–2762
(1994) New type of magnetocaloric effect: implications on
[18] Kallel S, Kallel N, Hagazaa A, Pena O, Oumezzine M
low-temperature magnetic refrigeration using an Ericsson
(2010) Large magnetic entropy change above 300 K in
cycle. Appl Phys Lett 64:2739–2741
(La0.56Ce0.14)Sr0.3MnO3 perovskite. J Alloys Compd
[3] Gschneidner KA Jr, Pecharsky VK, Tsokol AO (2005)
492:241–244
Recent development in magnetocaloric materials. Rep Prog
[19] Phan MH, Tian SB, Hoang DQ, Yu SC, Nguyen C, Ulyanov
Phys 68:1479–1539
AN (2003) Large magnetic-entropy change above 300 K in
[4] Tishin AM, Spichkin YI (2014) Recent progress in magne-
CMR materials. J Magn Magn Mater 258:309–311
tocaloric effect: mechanisms and potential applications. Int J
[20] Morelli DT, Mance AM, Mantese JV, Micheli AL (1996)
Refrig 37:223–229
Magnetocaloric properties of doped lanthanum, manganites
[5] Yu BF, Gao Q, Zhang B, Meng XZ, Chen Z (2003) Review
films. J Appl Phys 79:373–375
on research of room temperature magnetic refrigeration. Int J
[21] Nam DNH, Bau LV, Khiem NV, Dai NV, Hong LV, Phuc
Refrig 26:622–636
NX, Newrock RS, Nordblad P (2006) Selective dilution and
[6] Kumar NP, Reddy PV (2014) Magnetocaloric effect in
magnetic properties of La0.7Sr0.3Mn1-xM0 xO3 (M’ = Al,
RMnO3 (R = Gd, Tb and Dy) multiferroics. Mater Lett
Ti). Phys Rev B 73:184430
122:292–295
[22] Jemaa FB, Mahmood SH, Ellouze M, Hlil EK, Halouani F
[7] Dan’kov SY, Tishin AM, Pecharsky VK, Gschneidner KA
(2015) Structural magnetic magnetocaloric and critical
Jr. (1998) Magnetic phase transitions and the magne-
behavior of selected Ti-doped manganites. Ceram Int
tothermal properties of gadolinium. Phys Rev B
41:8191–8202
57:3478–3490
2412 J Mater Sci (2018) 53:2405–2412

[23] Yadav PA, Deshmukh AV, Adhi KP, Kale BB, Basavaih N, solutions: experimental and density-functional studies. Phys
Patil SI (2013) Role of grain size on the magnetic properties Rev B 93:094401
of La0.7Sr0.3MnO3. J Magn Magn Mater 328:86 [25] Banerjee BK (1964) On a generalised approach to first and
[24] Nayek C, Samanta S, Manna K, Pokle A, Nanda BRK, second order magnetic transitions. Phys Lett 12:16
Kumar PSA, Murugavel P (2016) Spin-glass state in
nanoparticulate (La0.7Sr0.3MnO3)1-x(BaTiO3)x solid
Journal of Materials Science is a copyright of Springer, 2018. All Rights Reserved.