Influence of Ni content on structural, magnetocaloric and electrical properties in manganite La0.6Ba0.2Sr0.2Mn1−xNixO3 (0 ≤ x ≤ 0.1) type perovskites

We present a detailed study on the physical properties of La0.6Ba0.2Sr0.2Mn1−xNixO3 samples (x = 0.00, 0.05 and 0.1). The ceramics were fabricated using the sol–gel route. Structural refinement, employing the Rietveld method, disclosed a rhombohedral R3̄c phase. The magnetization vs. temperature plots show a paramagnetic–ferromagnetic (PM–FM) transition phase at the TC (Curie temperature), which decreases from 354 K to 301 K. From the Arrott diagrams M2vs. μ0H/M, we can conclude the phase transition is of the second order. Based on measurements of the isothermal magnetization around TC, the magnetocaloric effects (MCEs) have been calculated. The entropy maximum change (−ΔSM) values are 7.40 J kg−1 K−1, 5.6 J kg−1 K−1 and 4.48 J kg−1 K−1, whereas the relative cooling power (RCP) values are 232 J kg−1, 230 J kg−1 and 156 J kg−1 for x = 0.00, 0.05 and 0.10, respectively, under an external field (μ0H) of 5 T. Through these results, the La0.6Ba0.2Sr0.2Mn1−xNixO3 (0 ≤ x ≤ 0.1) samples can be suggested for use in magnetic refrigeration technology above room temperature. The electrical resistivity (ρ) vs. temperature plots exhibit a transition from metallic behavior to semiconductor behavior in the vicinity of TM–SC. The adiabatic small polaron hopping (ASPH) model is applied in the PM-semiconducting part (T > TMS). Throughout the temperature range, ρ is adjusted by the percolation model. This model is based on the phase segregation of FM-metal clusters and PM-insulating regions.


Introduction
Magnetic refrigeration (MR) technology based on the magnetocaloric effect (MCE) is advancing to become a suitable technology, compared to conventional gas refrigeration, 1-3 due to a number of advantages. 4 The MCE is generally characterized by two factors: change in entropy (DS M ) and relative cooling power (RCP).
Gadolinium (Gd) is a pure lanthanide element and is the rst material that has a high MCE with a Curie temperature (T C ) near room temperature (RT). 5 Interestingly, further MCE investigations were performed for binary Gd-M compounds, such as Gd 3 SiGe 2 , 6 which shows a MCE twice that of Gd. 7 The researchers focused on nding new cheaper materials with larger MCEs. In this context, manganites with perovskite structure have certain advantages over Gd: their elements are not expensive, they are chemically stable, they have high resistivity and they exhibit a good MCE under low magnetic elds. 8,9 Among these, ABO 3 compounds are known as perovskite manganites. These materials have a general formula Re 1Àx where Re 3+ is a rare earth element (Nd 3+ , La 3+ , Pr 3+ , Sm 3+ , .) and A 2+ is an alkaline earth ion (Sr 2+ , Ba 2+ , Ca 2+ ). 10,11 They present some interesting properties, which make them very attractive materials for industrial applications.
The pure stoichiometric lanthanum manganite LaMnO 3 is antiferromagnetic, insulating at 150 K, and the substitution of the rare earth element by a lower valence ion causes the oxidation of Mn 3+ into Mn 4+ to ensure electroneutrality in the material. It is followed by the appearance of macroscopic magnetization, i.e. a ferromagnetic coupling between the Mn 3+ ions (t 2 g 3 e 1 g ) and the Mn 4+ ion (t 2 g 3 e 0 g ). Substitution of La 3+ by a divalent or monovalent ion can result in a wide Curie temperature range which can vary from 150 K to 375 K. Experimentally, manganites, in particular manganese oxides La 1Àx -Sr x MnO 3 with x ¼ 0.3, are well studied systems. They present an FM-PM transition accompanied by a metal-semiconductor transition close to T C . Several studies have been performed on the magnetocaloric properties of this compound, which exhibits a large change in magnetic entropy, with a narrow range of working temperatures in the vicinity of T C . In addition, several investigations have been carried out to estimate the substitution effects of the Mn site and these have shown that, for the La 1Àx Sr x MnO 3 family, even a low rate of substitution at the Mn site would induce a signicant change in the properties of magnetotransport. [12][13][14] The values of T C and DS M are generally affected by the partial substitution of manganese ions by certain transition metals, for example the In ion, 15,16 the Al ion, 17 etc. On the other hand, the substitution of the rare earth La 3+ by certain metals lead to a signicant change in the magnetic and electrical properties. 18,19 Indeed, the substitution of the Re site with divalent ions 20 proves the oxidation of Mn 3+ to Mn 4+ , which is the origin of the ferromagnetic character. 21,22 The magnetic coupling between Mn 4+ and Mn 3+ is usually governed by the movement of the electron, for example between the two partly lled d layers with a strong Hund's coupling on site.
The issue of replacing magnetic and non-magnetic ions in the Mn site is very important. For example, the partial substitution of Mn 3+ ions with Ni 2+ ions modies the ratio of the Mn 4+ -O 2 -Mn 3+ network and leads to a decrease of the double exchange (DE) interactions. 23 Several studies have been carried out [24][25][26][27][28][29] to explain the relationship between the magnetotransport and magnetic properties of Re 1Àx A x MnO 3 substituted by different elements on the Mn site. Based on the information given in this article, we have carefully discussed the physical properties in La 0.6 Ba 0.2 Sr 0.2 Mn 1Àx Ni x O 3 (0 # x # 0.1) compounds.

Preparation
The La 0.6 Ba 0.2 Sr 0.2 Mn 1Àx Ni x O 3 ceramics were prepared using the sol-gel route. High purity precursors La(NO 3 2 and Ni(NO 3 ) 2 $6H 2 O were weighed in stoichiometric proportions and then dissolved in distilled water with continuous stirring, whilst on a hot plate. The mixtures were dispersed in solutions containing a complexation agent (citric acid) and a polymerizing agent (ethylene glycol). The citric acid was used as a chelating agent and the ethylene glycol was used as a gelication agent. In order to form a homogenous yellowish gel, the solutions were heated on a hotplate at about 100 C for 1 h under magnetic stirring. Then, to remove the excess solvent, the temperature was increased to 400 C and the combustion led to a very ne and very homogeneous powder (black powder). At the end of this process, the calcinated powder was ground and pressed into pellets. The pellets were subjected to sintering at 900 C for 24 hours in air.

Characterization
X-ray powder diffraction (XRD) was used to examine the structural behavior of the samples. Using a "Panalytical X'Pert Pro" diffractometer, the XRD was conducted through Cu-Ka radiation (l Cu ¼ 1.54056Å) with a 0.0167 step size and 19 # 2q # 90 angular range. The renement was analyzed by Rietveld's program using FULLPROF soware (version 0.2-March 1998-LLB-JRC). 30 A Philips XL30 scanning electron microscope (SEM) and an energy dispersive X-ray (EDX) spectrometer working at 15 kV were used to carry out a morphological study of the compounds. The magnetization measurements were recorded with a BS1 and BS2 magnetometer, which was developed in the Louis Néel laboratory in Grenoble.

X-ray analysis
The XRD patterns of the La 0.6 Ba 0.2 Sr 0.2 Mn 1Àx Ni x O 3 (LBSMNO) samples, recorded at room temperature (RT), are presented in Fig. 1. The analysis of these spectra indicates that all the compounds were successfully prepared with a good crystallinity and a single phase of La 0.6 Ba 0.2 Sr 0.2 Mn 1Àx Ni x O 3 ; we have not detected a second phase. In the inset of Fig. 1, we show the crystalline structure of these samples. A good t agreement between the simulation and the experimental pattern was observed.
The patterns of our samples were indexed in the rhombohedral (R 3c) symmetry (no. 167), with (La, Ba, Sr): 6a (0, 0, 0.25), (Mn, Ni): 6b (0, 0, 0) and O: 18 (x, 0, 0.25). The different structural parameters are tabulated in Table 1. When the Ni substitution increases, the volume and the lattice parameters decrease. A similar behavior has also been previously observed. 31 This decrease is explained by the fact that the average ionic radius of the manganese site (0.599 # r Mn+Ni # 0.592) decreases, which can be assigned to the formulation of a higher level of Mn 4+ , compared to Mn 3+ .  0.1), these results can be justied as follows: whenever the x ratio of Ni is increased, the proportion of Mn 3+ weakens by y ¼ 2x, while the ratio of Mn 4+ increases by x. In Table 2 The perovskite structure can be distorted from the ideal cubic structure, which greatly affects the properties. These distortions are principally given by the relationship between the ionic radius of the cations, dened by the tolerance factor t G : 32 where r A is the radius of the A site ions, r B is that of the B site ions and r O is the radius of the oxygen ions, which are found in the tables of Shannon. 33 Generally, a perovskite exhibits a cubic structure if t G is equal to 1 and it undergoes distortions if t G deviates from 1. 34 In our work, the tolerance factor t G decreases with the increase of Ni ( Table 2).
The rate of rhombohedric deformation D% can be calculated employing the expression: D% ¼ 1 3 From Table 2, the value of D decreases with the decrease in mean r Mn+Ni .
On the other hand, the average crystallite size D SC can be calculated using Scherrer's relationship: 37 Here, b is the full width at half maximum (FWHM) of the peak (104), l represents the wavelength of the Cu-Ka radiation (¼1.54056Å) and q corresponds to the angle of the most intense peak (104). The results are given in Table 2. It can be deduced that the grain size (D SC ) decreases from 55 to 49 nm when we introduce the Ni 2+ ions. As in Scherrer's method, the crystallite size values were determined from the Williamson-Hall equation q is the Bragg angle, 3 is the strain and D WH is the crystallite size. The slope of the plot of b cos q (y-axis) vs. 4 sin q (x-axis) gives the strain (3) and the crystallite size (D WH ) can be calculated from the intercept of this line on the y-axis (Fig. 2). The calculated values are grouped in Table 2. We can deduce from this result that the average crystallite size determined by Williamson-Hall is greater than that obtained by Scherrer's method, which is due to the broadening effect caused by the strain exhibited in this technique.    ImageJ soware was employed to determine a statistical count of the grain size, which was performed on the SEM images. This technique consists of measuring the diameters of all the particles in the SEM image. Then, we adjusted these data using the log-normal function. Where, s is the median diameter obtained from the data dispersions and D 0 is the median diameter obtained from the SEM images. Fig. 3(a) and (b) present the grain number (counts) versus the particle size. Using the t results, the mean diameter Table 2). It is remarkable that the average size of the particles obtained is greater than the average size of the crystallites determined by XRD. This can be explained by the fact that each particle observed by SEM is made up of several crystallites. The energy dispersive X-ray microanalysis (EDX) spectrum of La 0.6 -Ba 0.2 Sr 0.2 Mn 0.9 Ni 0.1 O 3 is presented in Fig. 3(c) as an example. This technique conrms the composition and purity of the samples. The spectra reveal the homogeneous distribution of La, Ba, Sr, Mn, Ni and O atoms over a wide surface area.

Morphological characterization
To evaluate the porosity p ð%Þ ¼ 1 À d d x of the compounds, we calculated the X-ray density, where d is the bulk density, a is the lattice constant, M is the molecular weight and N a is Avogadro's number 39 (see Table 2).

Magnetic properties
The evolution of M(T) measured at 0.05 T in cooled zero eld (ZFC) and cooled eld (FC) modes is presented in Fig. 4. It is observed that, in the low temperature region, the FC and ZFC curves diverge considerably for all samples, which proves the existence of typical spin glasses, which may also be ascribed to magnetic anisotropy. A spin glass-like state is generally prompted by the coexistence of competing AFM and FM interactions. 40 With a decreasing temperature, a PM-FM phase transition was observed at the Curie temperature. T C is given at the lowest point of the rst derivative of the curve M(T) (dM/dT) (inset Fig. 4). The T C values go from 354 K for x ¼ 0.00 to 301 K for x ¼ 0.10. This change has been attributed to the modication of the Mn-O-Mn bond angle. Doping with the slightly larger Ni 2+ (r Ni The most important cause of the decrease in T C is the reduction of the one-electron bandwidth W d . It is given as: 43    Table 2) which causes the decrease of T C . 44,45 In fact, this reduction in W leads to a reduction in the FM coupling between neighboring manganese atoms.
In addition, it is interesting to understand the behavior of M vs. temperature in the FM region. In this context, and according to Lonzarich and Taillefer, 46 magnetization obeys the theory of spin waves. At low temperatures, this theory is that the magnetization has multiplied in T 3/2 (Bloch's law) and over a wide range of temperatures in T 2 , yet in the vicinity of T C it varies as follows: (1 À T 4/3 /T C 4/3 ) 1/2 .
In the FM region, the M data has been adjusted by this relation: Here M 0 is the spontaneous magnetization. Fig. 5 presents the best t curves. It can be conrmed that the FM behavior of La 0.6 Ba 0.2 Sr 0.2 Mn 1Àx Ni x O 3 may be owing to spin waves. For T > T C , the PM region, the Curie-Weiss law was used to analyse the inverse of magnetic susceptibility (c À1 ¼ H/M): Here, C ¼ N A m B 2 3k B m eff 2 is the Curie constant and q CW is the paramagnetic-Curie temperature. These parameters were obtained using the t of curve c m À1 (T) (Fig. 5) and its values are also given in Table 3. The positive value of q CW suggests that the ferromagnetic interactions between the nearest neighbors are dominant in the system, which could be due to DE Mn 3+ -O 2À -Mn 4+ coupling. When x increased, this parameter decreased, which indicates the weakening of the ferromagnetic interactions. 47 The values of q CW are higher than those of T C , which indicates the presence of magnetic inhomogeneity above T C .
The experimental effective moment m exp eff can be calculated using the parameter C and the values are given in Table 3.
The theoretical effective paramagnetic moment for La 0. 6 Table 3, a difference between the experimental and theoretical values of m eff can be observed. This can be claried by the presence of FM clusters within the PM phase. 48 The evolution of magnetization vs. The calculated magnetic moment can be determined by: The magnetic moments of Ni 2+ , Mn 3+ and Mn 4+ have 2, 4 and 3 m B , respectively. The M sp values are 3.88 m B for x ¼ 0.00, 3.73 m B for x ¼ 0.05 and 3.58 m B for x ¼ 0.10. This reduction can be attributed to competition between the ferromagnetic and antiferromagnetic interactions. In addition, the Ni 2+ ion at the M site inuences the valence states of the manganese ions, i.e. it decreases the Mn 3+ /Mn 4+ ratio. This proves a reduction in the double Zener exchange (DE), which leads to a decrease in magnetization.
We analyzed the hysteresis loops at 10 K (m 0 H ¼ AE5 T) to better understand the magnetic properties at low temperatures (Fig. 6). The curves are similar, with hysteresis loops which are weak, with reasonably good coercive elds (m 0 H C ), which decrease with the increase of Ni content . This reduction can be assigned to the decrease in spin dependent electron hopping. In the weak m 0 H region, M grew signicantly and reached saturation as the eld increased. The value of M s (saturation magnetization) can be estimated at high m 0 H at about 5 T. From Table  3, it was found that M s decreased, which may be due to antiferromagnetic alignments. Likewise, the insertion of Ni in the Mn site modies the valence states of the manganese ions and decreases the level of Mn 3+ /Mn 4+ , which in turn weakens the DE interaction. The inset of Fig. 6 Table 3. The weak hysteresis loop with large saturation values conrms the characteristic so FM behavior of the compounds. In this context, it can be concluded that our compounds may be applicable to read and write processes in high density recording media or for information storage. 49 The remanence ratio (R) is utilized to comprise the isotropic nature of our investigated compounds. The values of remanence varied in the range of 5-13 emu g À1 . The ratio (R) is given by: R ¼ M r /M s . We have summarized the values of R in Table 3. The small obtained values conrm the isotropic natures. 50 In this context, for magnetic recording and memory devices, 51 it is advantageous to have higher remanence ratios. The values of R reveal an increasing trend with Ni 2+ substitution.

Magnetocaloric properties
We have shown previously that there is a phase transition around T C . Therefore, to calculate the change in magnetic entropy (ÀDS M ), it is necessary to know the order of this transition.
We have presented in Fig. 7(a-c), the external m 0 H variation of the isothermal magnetization at different temperatures around T C . M increases rapidly at low m 0 H and then achieves saturation, which shows ferromagnetic behavior. Above T C , thermally unsettled magnetic moments yield to increase magnetizations linearly at high temperatures, which means PM behavior. This phenomena demonstrates the magnetic phase transition.
In the inset of Fig. 7(a-c), we have presented the Arrott plots (M 2 vs. m 0 H/M) for the La 0.6 Ba 0.2 Sr 0.2 Mn 1Àx Ni x O 3 (0 # x # 0.1) compounds, from these curves we can conclude the nature of the magnetic phase transition. The slope of the curves is positive, so the transition is second order, according to Banerjee's criteria. 52 The MCE is an intrinsic characteristic of magnetic materials. [53][54][55] Its principle is based on cooling or heating compounds when subjected to a magnetic eld under adiabatic conditions, which is maximized when materials are close to their magnetic control temperature.
jDS M j can be determined, using Maxwell's equations, by the formula:   (ÀDS max M ) is not the only factor that determines the applicability of such a material, but also the temperature range over which it remains considerable is signicant.
The relative cooling power (RCP) is another very important parameter along with (Àdelta), which denes the amount of heat that can be released between cold and hot sinks in an ideal refrigeration cycle and can be given by the following formula: 57 Here DS max M represents the maximum of DS M and dT FWHM is the full width at half maximum. Fig. 9 present the relative cooling power as a function of m 0 H for our compounds. The values of RCP increase with increasing m 0 H and reache about 214 J kg À1 for x ¼ 0, 230 J kg À1 for x ¼ 0.05 and 285 J kg À1 for x ¼ 0.1 at m 0 H ¼ 5 T. To better understand the performance of the MCE of our compounds, the values of (ÀDS max M ) and RCP are compared to other manganites, as given in Table 4. [58][59][60][61][62] It can be noted that our samples, especially for x ¼ 0.1, have a suitable T C value, close to RT, and a relatively large magnetic entropy change to include other materials. This proves that our materials can be used in the eld of magnetic refrigeration.
On the other hand, to affirm the nature of phase transition, Franco et al. 63 proposed a phenomenological universal curve for the eld dependence of DS M . We can construct the universal curve by normalizing all the DS max M : DS M (T, m 0 H)/DS max M below and above T C , by imposing that the position of two additional reference points in the curve correspond to q ¼ AE1.
Here T r1 and T r2 are chosen as reference temperatures, such that DS M (T r1,2 ) ¼ 1/2DS max M . By referring to Banerjee's criteria for a 2 nd order PM-FM transition, all DS M curves at different m 0 H values should merge into one curve with temperature scaling. If not, the samples follow a 1 st order phase transition. Fig. 10 represents the evolution of DS M (T, m 0 H)/DS max M vs. temperature q at different m 0 H values for x ¼ 0.05, for example. In this gure, all the data collapses into a single master curve around T C , indicating the 2 nd order nature of this phase transition. These results are in good accordance with those obtained by the Banerjee criterion discussed before. In addition, this universal curve can be used for practical purposes, such as extrapolating results to elds or temperatures not available in the laboratory, improving data resolution and deconvoluting the response of overlapping magnetic transitions. 64 We can adjust this curve by the Lorentzian function: Here a, b and g are adjusted parameters. Given the asymmetry of the curve, two ensembles of different constants must be used: Fig. 9 Evolution of RCP vs. magnetic field for the LBSMNO compounds.  From eqn (7), the stance and magnitude of the peak, namely, (DS max M , T C ) and T r 1 and T r 2 are the only ones that are to describe DS, where T r 2 > T C and T r 1 < T C . In the end, to transpose DS(q) into the real DS M (T), we only use these values, which are xed by the properties of the compounds.
It is essential to study how the MCE evolves over the ranges of applied magnetic elds and the desired temperatures, taking into account that the 2 nd order transition has been proven for all samples.
The evolution of the DS M vs. the eld is given by the expression, according to Parker and Oesterreicher: 65 DS max M ¼ b(m 0 H) n , b is a constant and n is an exponent, which depends on the magnetic state of the sample. The exponent n can be expressed by the following expression: According to the mean eld approach for conventional ferromagnetic compounds, a minimum value of "n" is 2/3 at T C .
Below T C , n has been predicted to be 1 and the materials are in the FM state. However, above T C , it is equal to 2 in the PM zone, according to the Curie-Weiss law. Yet, recent experimental data indicates a deviation from n ¼ 0.66, in the case of a few so magnetic amorphous compounds. The temperature dependence of n is shown in Fig. 11. The values of n are found to be 0.67, 0.45 and 0.32 for x ¼ 0.00, 0.05 and 0.10, respectively. For x ¼ 0, the value is close to the values of the mean eld model. However, for x ¼ 0.05 and 0.1, these values do not coincide with the predicted value of the mean eld of 0.66. This difference is probably due to local inhomogeneities around T C . 66 3.5 Electrical properties Fig. 12 presents the variation of electrical resistivity (r) vs. temperature (T) for our samples. All the samples are magnetically ordered and the resistivity exhibits a metallic behavior for low temperatures, resulting from a strong ferromagnetic coupling. Semiconductor behavior (SC) is reported at higher temperatures. It can be concluded that these samples undergo a semiconductor-to-metal (SC-M) transition at T ¼ T M-SC . With increasing Ni concentration, this peak temperature T M-SC decreases ( Table 5). From this table, one can see that T M-SC for all the doped materials is much lower than T C . From this result, it can be said that the transport properties are governed by the presence of inter-grain boundaries.
To better understand the contribution of the different factors causing the conduction mechanism below the transition temperature (T < T M-SC ), the r(T) curve was tted using different theoretical models.
Conduction electrons meet different competitors, including scattering of the grain/domain boundary, electron-magnon scattering and electron-electron scattering. Using the following empirical relation, 67 the electrical resistivity data is analyzed: Fig. 11 Variation of the exponent n vs. temperature for the LBSMNO samples. Table 5 The best fit parameters gated with the experimental resistivity data utilizing eqn (17) x ¼ 0.00  Here r 0 is residual resistivity due to the domain or grain boundaries, r 2 T 2 is the contribution from the electron-electron scattering process to the electrical resistivity and r 5 T 5 is associated with the electron-phonon interaction. 68 The values of the electrical resistivity were adjusted using eqn (14) and in the inset of Fig. 13, we have given the best t. In Table 5 we have grouped together the estimated values of the adjusted parameters.
Above (T > T M-SC ), the resistivity is simulated by the SPH mechanism. 69 r is expressed as in the adiabatic SPH model: where A and E a are the coefficient of resistivity and activation energy associated to polaron binding energy, respectively. We have tted the variation of electrical resistivity (see the inset of Fig. 12) and the results are given in Table 5. We calculated the hopping energies E a and we have deduced that the values of E a are 180, 86 and 67 meV for x ¼ 0, 0.05 and 0.1, respectively. To understand the transport mechanism of the total resistivity over the whole temperature range, we used a phenomenological percolation model. 70 For this model, resistivity is dened based on the contributions of the FM clusters in the PM region. Thus, the resistivity is expressed by the following expression: Here f is the volume fraction of the ferromagnetic phase and (1 À f) is the volume fraction of the paramagnetic phase. The volume fraction follows the Boltzmann distribution and this is expressed via the following equation: where DU ¼ ÀU 0 (1 À T/T mod C ) is the energy gap between the FM and PM states. T mod C is the temperature of resistivity maxima and U 0 is taken as the energy gap for temperatures lower than T mod C .
With T ¼ T mod C , f ¼ f c ¼ 0.5 where f c is called a percolation threshold. 71 Where f < f c , the sample remains semiconducting and for f > f c it acquires a metallic phase. 72 Hence, in the entire temperature range eqn (16), can be given as: The data evaluated from eqn (17) are in agreement with the experimental results. It can be seen that the percolation model adequately describes the resistivity behavior over a wide range of temperatures, including the phase transition region. We have grouped the most suitable parameters in Table 5 and in Fig. 13, we have presented the t of the data.
The inset in Fig. 13 shows f(T) vs. temperature for all samples. Below T M-SC , the volume concentration of the ferromagnetic phase remains equal to 1. This proves the strong dominance of the FM part in this zone. Aerwards, f(T) starts to lower to zero, since the metallic state (FM) moves to a semiconductor state (PM). This result proves the validity of the percolation approach.

Conclusion
We have investigated the physical properties of polycrystalline La 0.6 Ba 0.2 Sr 0.2 Mn 1Àx Ni x O 3 samples. Their crystal structures correspond to a rhombohedral structure, with the R 3c space group without any secondary phase. When the substitution rate increases, the unit cell volume decreases. The magnetic and electrical measurement data indicate that our compounds show FMM behaviour at low temperature (T < T M-SC ) and PMS behaviour above T M-SC . This temperature decreases as the Ni substitution increases, due to the bandwidth reduction. The values of (ÀDS max M ) at m 0 H ¼ 5 T are 7.40 J kg À1 K À1 , 5.6 J kg À1 K À1 and 4.48 J kg À1 K À1 for x ¼ 0.00, 0.05 and 0.10, respectively. The magnetocaloric performance of these samples indicates that the polycrystalline La 0.6 Ca 0.1 Sr 0.3 Mn 1Àx Ni x O 3 compounds are good candidates for magnetic refrigeration at room temperature.

Conflicts of interest
There are no conicts to declare.